ERS Sleep Handbook

Visit www.ersnet.org/sleephandbook to download your electronic copy of the book and take the online CME test.

This book has been accredited by EBAP for 8 CME credits.

handbook

RespiratorySleep MedicineEditorsAnita K. SimondsWilfried de Backer

PUBLISHED BY EUROPEAN RESPIRATORY SOCIETY

CHIEF EDITORSAnita K. Simonds (London, UK)

Wilfried de Backer (Antwerp, Belgium)

AUTHORS

ERS STAFFMatt Broadhead, Jonathan Hansen, Sarah Hill, Sharon Mitchell,

Victoria Morton, Elin Reeves, Rachel White

© 2012 European Respiratory Society

Design by Claire Turner and Lee Dodd, ERSTypeset in China by Charlesworth GroupPrinted in the UK Page Bros (Norwich) LtdIndexed by Merrall-Ross International Ltd

All material is copyright to the European Respiratory Society. It may not be reproduced in any way including electronically without the express permission of the society.

CONTACT AND PERMISSIONS REQUESTS: European Respiratory Society, 442 Glossop Road, Sheffield, S10 2PX, UKTel: 44 114 2672860 Fax: 44 114 2665064 e-mail: [email protected]

ISBN 978-1-84984-023-1

Stefan AndreasChiara BaglioniFerrán BarbéMaria R. BonsignoreAn BoudewiynsGary CohenViliam DonicAthanasios KaditisMiriam Katz-SalomonBrian D. KentEric Konofal

Michel LecendreuxPatrick LevyMarie MarklundJuan Fernando MasaWalter T. McNicholasJosep M. MontserratMary J. MorrellGimbada B. MwengeArie OksenbergPaolo PalangeDirk Pevernagie

Winfried RanderathDieter RiemannRenata L. RihaDaniel RodensteinSilke RyanZoltan TomoriHa TrangJohan VerbraeckenStijn VerhulstMaria Pia Villa

Contributors viiPreface xiGet more from this Handbook xiiList of abbreviations xiii

Chapter 1 - Physiology and anatomy of sleep and breathing

Neuroanatomy and neurobiology of sleep 1Mary J. Morrell, Paolo Palange, Patrick Levy and Wilfried De Backer

Breathing during sleep and wakefulness 6 Mary J. Morrell, Paolo Palange, Patrick Levy and Wilfried De Backer

Homeostatic regulation during sleep 13Mary J. Morrell, Paolo Palange, Patrick Levy and Wilfried De Backer

Chapter 2 - Respiratory conditions

Definitions of sleep disordered breathing 21Johan Verbraecken

Obstructive sleep apnoea syndrome 25Johan Verbraecken

Obstructive sleep apnoea: epidemiology and risk factors 26Johan Verbraecken

Obstructive sleep apnoea: pathophysiology 29Johan Verbraecke and Wilfried de Backer

Obstructive sleep apnoea: clinical aspects 32Stefan Andreas

Obstructive sleep apnoea: consequences 35Stefan Andreas

Central sleep apnoea and Cheyne–Stokes respiration 39Winfried Randerath

Contents

Central sleep apnoea/Cheyne–Stokes respiration: epidemiology 40and risk factors Winfried Randerath

Central sleep apnoea/Cheyne–Stokes respiration: pathophysiology 42Winfried Randerath

Central sleep apnoea/Cheyne–Stokes respiration: clinical aspects 45Winfried Randerath

Central sleep apnoea/Cheyne–Stokes respiration: consequences 46Stefan Andreas

Hypoventilation syndromes/chronic respiratory insufficiency in sleep 48Viliam Donic and Zoltan Tomori

Chapter 3 - Nonrespiratory conditions

Nonrespiratory sleep disorders 52Renata L. Riha

Insomnia 53Dirk Pevernagie

Parasomnia and movement disorders during sleep 61Renata L. Riha

Hypersomnia 67Renata L. Riha

Circadian rhythm disorders 74Renata L. Riha

Psychiatric aspects of sleep 80Chiara Baglioni and Dieter Riemann

Chapter 4 - Clinical assessment

Sleep history 84Silke Ryan and Walter T. McNicholas

Differential diagnosis 87Walter T. McNicholas

Questionnaires on sleep 91Brian D. Kent and Walter T. McNicholas

Predisposing factors 95Maria R. Bonsignore

Diagnostic algorithms 100Josep M. Montserrat, Ferrán Barbé and Juan Fernando Masa

Quality of life 107Maria R. Bonsignore

Surgical and anaesthesia risk assessment 111Maria R. Bonsignore

Comorbidity assessment 115Maria R. Bonsignore

Chapter 5 - Diagnostic techniques

Polysomnography 120Renata L. Riha

Assessment of daytime sleepiness 131Renata L. Riha

Cardiorespiratory monitoring during sleep 136Maria R. Bonsignore and Juan Fernando Masa

Chapter 6 - Treatment

Treatment of obstructive sleep apnoea 147Johan Verbraecken, Arie Oksenberg, Marie Marklund and An Boudewyns

Positive airway pressure treatment 157Anita K. Simonds

Treatment of central sleep apnoea 164Winfried Randerath

Treatment of hypoventilation/chronic respiratory insufficiency 169Anita K. Simonds

Treatment of nonrespiratory sleep disorders 176Dirk Pevernagie

Chapter 7 - Medicolegal and economic aspects of sleep disorders

Medicolegal and economic aspects of sleep disorders 189Gimbada B. Mwenge and Daniel Rodenstein

Chapter 8 - Paediatric respiratory sleep medicine

Development of breathing and sleep and physiopathology of 200apnoea in the first years of lifeGary Cohen, Miriam Katz-Salamon and Ha Trang

Sleep disordered breathing in children 205Ha Trang and Anita K. Simonds

Comorbid respiratory disorders in children 210Anita K. Simonds

Nonrespiratory conditions in children 213Ha Trang, Michel Lecendreux and Eric Konofal

Clinical assessment in children 218Maria Pia Villa

Diagnostic techniques in children 221Stijn Verhulst and Wilfried De Backer

Management of sleep disordered breathing in children 228Athanasios Kaditis, Maria Pia Villa, Anita K. Simonds and Ha Trang

Index 237

Editors

Anita K. Simonds Wilfried de BackerNIHR Respiratory Disease Biomedical Dept of Pulmonary MedicineResearch Unit University and University Hospital ofRoyal Brompton and Harefield NHS AntwerpFoundation Trust Edegem, BelgiumLondon, UK [email protected]@rbht.nhs.uk

Authors

Stefan Andreas Chiara BaglioniPneumologische Lehrklinik Dept of Psychiatry and PsychotherapyUniversität Göttingen Freiburg University Medical CentreGöttingen, Germany Freiburg, [email protected] [email protected]

Ferrán Barbé Maria R. BonsignoreRespiratory Diseases Research Unit Biomedical Dept of Internal and Universitari Arnau de Vilanova Specialist Medicine (DiBiMIS) Lleida, Spain/ University of PalermoCIBER de Enfermedades Respiratorias Palermo, ItalyBunyola, Spain [email protected]@arnau.scs.es

An Boudewiyns Gary CohenDept of Otorhinolaryngology Dept of Woman and Child HealthUniversity Hospital Antwerp Karolinska InstitutetEdegem, Belgium Stockholm, [email protected] [email protected]

Viliam Donic Athanasios KaditisDept of Human Physiology and Sleep First Dept of PaediatricsLaboratory University of Athens School ofSchool of Medicine Medi cine and Aghia Sophia Children’sSafarik University HospitalKosice, Slovakia Athens, [email protected] [email protected]

Contributors

vii

Miriam Katz-Salomon Brian D. Kent Dept of Woman and Child Health, Pulmonary and Sleep Disorders UnitKarolinska Institutet St. Vincent’s University HospitalStockholm, Sweden Dublin, [email protected] [email protected] Eric Konofal Michel LecendreuxPaediatric Sleep Disorders Centre, Dept of Child and AdolescentAP-HP, Hôpital Robert Debré, PsychopathologyParis, France/ Robert Debré HospitalPediatric Clinical Investigation Centre Paris VII University(CIC 9202) INSERM Paris, FranceAP-HP, Hôpital Robert Debré [email protected], [email protected]

Patrick Levy Marie MarklundJoseph Fourier University and INSERM Dept of Dentistry/OrthodonticsGrenoble, France University of Umeå[email protected] Umeå, Sweden [email protected]

Juan Fernando Masa Walter T. McNicholasRespiratory Service Pulmonary and Sleep Disorders UnitHospital San Pedro de Alcantara St. Vincent’s University HospitalCáceres, Spain/ Dublin, IrelandCIBER de Enfermedades Respiratorias [email protected], Spain [email protected]

Josep M. Montserrat Mary J. MorrellServei de Pneumologia National Heart & Lung InstituteHospital Clinic-IDIBAPS Imperial College LondonBarcelona, Spain/ Academic Unit of Sleep and BreathingCIBER de Enfermedades Respiratorias Royal Brompton HospitalBunyola, Spain London, [email protected] [email protected]

vii

Gimbada B. Mwenge Arie OksenbergCentre for Sleep Medicine and Sleep Disorders UnitPneumology Department Loewenstein Hospital – Université Catholique de Louvain Rehabilitation CenterBrussels, Belgium Raanana, [email protected]` [email protected]

Paolo Palange Dirk PevernagieDept of Public Health and Infectious Sleep Medicine CentreDiseases Kempenhaeghe FoundationSapienza University of Rome Heeze, The Netherlands/Rome, Italy Dept of Internal [email protected] University of Gent Gent, Belgium [email protected]

Winfried Randerath Dieter RiemannCentre of Sleep Medicine and Centre of Sleep Medicine and Respiratory Care Respiratory CareBethanien Hospital Bethanien HospitalSolingen, Germany Solingen, [email protected] [email protected] Medical Centre Freiburg, Germany [email protected]

Renata L. Riha Daniel RodensteinDepartment of Sleep Medicine Centre for Sleep Medicine and Royal Infirmary Edinburgh Pneumology DepartmentEdinburgh, Scotland Université Catholique de [email protected] Brussels, Belgium [email protected]

Silke Ryan Zoltan TomoriRespiratory Sleep Disorders Unit Dept of Human Physiology and SleepSt. Vincent’s University Hospital LaboratoryDublin, Ireland School of Medicine Safarik [email protected] Kosice, Slovakia [email protected] ix

Ha Trang Johan VerbraeckenUniversity of Paris-Diderot Dept of Pulmonary Medicine andParis, France/ Multidisciplinary Sleep DisordersAP-HP, Robert Debré Hospital CentrePaediatric Sleep Centre, Antwerp University Hospital andCentre of reference for CCHS University of AntwerpParis, France Edegem, [email protected] [email protected]

Stijn Verhulst Maria Pia Villa Paediatric Sleep Lab and Paediatric Regional Sleep Disorders CentrePulmonology Sant’Andrea HospitalAntwerp University Hospital Rome, ItalyEdegem, Belgium [email protected]@ua.ac.be

x

Preface“A thing of beauty is a joy for ever: Its loveliness increases; it will never

pass into nothingness; but still will keep a bower quiet for us, and a sleep full of sweet dreams, and health, and quiet breathing...” John Keats

The ERS Handbook of Respiratory Sleep Medicine furthers the ERS HERMES project to Harmonise Education in Respiratory Medicine for European Specialists. Sleep medicine is truly multidisciplinary, and yet the huge expansion in sleep medicine facilities over the past two to three decades can be attributed to the recognition of and need to manage sleep disordered breathing, particularly obstructive sleep apnoea. The field is therefore of very great importance to respiratory physicians and respiratory team members, who are instrumental in running many European sleep laboratories. Moreover, treatment methods for obstructive sleep apnoea and nocturnal hypoventilation with continuous positive airway pressure and noninvasive ventilation have improved and diversified, and the links between sleep disordered breathing and co-morbidities are now much better understood, making it an area with an extensive and growing evidence base.

This handbook is part of the planned development of a training programme in sleep medicine, following the creation of a syllabus and curriculum in respiratory sleep medicine. Its aim is to help physicians and trainees meet the curriculum requirements. However, it covers the field extensively with detailed reference to nonrespiratory disorders and paediatric sleep medicine. As such, it is a valuable resource for any practitioner from a respiratory, neurology, cardiology, dental or ENT background who sees patients with sleep disorders, and needs an up-to date reference book that covers succinctly the causes, diagnosis and management of these conditions.

The Editors are very grateful to the ERS Task Force in Respiratory Sleep medicine, and other contributors who have written the chapters; and to the ERS staff who helped coordinate both the Task Force and the book.

Anita K. Simonds, Wilfried de BackerEditors

xi

By buying the ERS Handbook of Respiratory Sleep Medicine, you also gain access to the electronic version of the book, as well as an accredited online CME test.

To log in, simply visit www.ersnet.org/handbook and enter the unique code printed on the back cover of the book. Once logged in, you’ll be able to download the entire book in PDF or EPUB format, to read on your computer or mobile device.

You’ll also be able to take the online CME test. This handbook has been accredited by the European Board for Accreditation in Pneumology (EBAP) for 8 CME credits.

Get more from this Handbook

Also available from the ERS

European Respiratory Monograph 50: Sleep ApnoeaEdited by Walter T. McNicholas and Maria R. Bonsignore.

This Monograph takes an in-depth look at sleep apnoea, focusing on OSA, from pathophysiology and epidemiology to comorbidities, treatment and future directions. The 24 chapters cover all aspects of the disease, including the lat-est interventions and new diagnostic technologies. To see the full table of contents and read the editors’ introduction, visit erm.ersjournals.com/content/ermsa/1.toc

To buy a copy of this Monograph, please contact [email protected]

Your passcode to access online content is:

abc123

xii

(C)HF (congestive) heart failure(N)REM sleep (non-) rapid eye movement sleepAHI apnoea–hypopnoea indexASV adaptive servo venti lationBMI body mass indexCOPD chronic obstructive pulmonary diseaseCPAP continuous positive airway pressureCSA central sleep apnoeaCSR Cheyne–Stokes respirationECG electrocardiographyEEG electroencephalographyEMG electromyographyENT ear, nose and throatEOG electrooculographyESS Epworth Sleepiness ScaleICSD International Classification of Sleep DisordersMRI magnetic resonance imagingMSLT Multiple Sleep Latency TestMWT Maintenance of Wakefulness TestNIV noninvasive ventilationOHS obesity hypoventilation syndromeOSA(S) obstructive sleep apnoea (syndrome)OSLER test Oxford Sleep Resistance testPaCO2 arterial carbon dioxide tensionPaO2 arterial oxygen tensionPSG polysomnographyPtcCO2 transcutaneous carbon dioxide tensionSAHS sleep apnoea–hypopnoea syndromeSaO2 arterial oxygen saturationSDB sleep disordered breathingV’E minute ventilation

List of abbreviations

xiii

Neuroanatomy andneurobiology of sleep

Mary J. Morrell, Paolo Palange, Patrick Levy and Wilfried De Backer

The neurobiology of sleep

Sleep is a dynamic process involvingcomplex neural activation. In 1929 thepsychiatrist Hans Berger established thatbrain activity was different duringwakefulness and sleep by recording cortical

electrical potentials. He termed theserecordings electroencephalograms and bythe mid-1930s the cyclical patternsassociated with NREM sleep had beencategorised. The first observations of REMsleep occurred in the 1950s (Aserinsky et al.,1953). Understanding of the controlmechanisms of NREM and REM sleepcontinues to develop.

The neural regulation of wakefulness andarousal from sleep Neurons of the reticularactivating system are central to theregulation of wakefulness. Specifically, thereare two ascending pathways: one, a dorsalroute from the cholinergic laterodorsal andpedunculopontine tegmental nuclei,activates thalamic neurons to promote EEGactivity via glutamatergic thalamocorticalprojections (Saper et al., 2005). The ventralroute through the hypothalamus includesthe aminergic arousal system that originatesfrom the brainstem with serotonergic(dorsal raphe nuclei), noradrenergic (locuscoeruleus), histaminergic(tuberomammillary nucleus) anddopaminergic (ventral periaqueductal grey)neurons (Horner, 2008). Cortical activationduring wakefulness is also influenced byorexinergic (hypocretin) neurons originatingin the hypothalamus, and cholinergicneurons from the basal forebrain (fig. 1a).During wakefulness these pathways allowsensory information to be transmitted toareas of the association cortex via thethalamic gate.

The neural regulation of NREM and REMsleep The transition between wakefulnessand sleep occurs through a process ofreciprocal inhibition between arousal- andsleep-promoting neurons by way of a ‘‘flip-flop’’

Key points

N Wakefulness is maintained byactivation of the ascending reticularactivating system involving severalneurotransmitters includingglutamate, acetylcholine and themonoamines.

N NREM sleep onset is associated witha reduction in activation of theascending reticular activating systemand an increase in neural activitywithin the ventrolateral pre-optic area,anterior hypothalamus and basalforebrain.

N REM sleep is triggered by activation ofcholinergic neurons in thelaterodorsal and pedunculopontinetegmental nuclei. The suppression ofmotor activity in REM sleep isgenerated by glutamate-mediatedactivation of descending medullaryreticular formation.

N Cycles of NREM and REM sleepalternate throughout the night in apredictable manner.

N Ageing is associated with difficulty inmaintaining sleep and more frequentarousals.

ERS Handbook: Respiratory Sleep Medicine 1

a)

b)

Upper airway

1 Pineal2 Pons3 Medulla oblongata4 Spinal cord5 Pituitary6 Tuberomammillary nucleus7 Preoptic area8 Hypothalamus9 Corpus callosum10 Cerebral cortex11 Hippocampus12 Thalamus13 Midbrain14 Cerebellum

Upper airway

Upper airway

c)

Olfactorybulb

1

2 3

45

6

78

910

11

1213

14

• GABAergic

• Histaminergic• Serotonergic• Noradrenergic• Cholinergic• Orexinergic• Glutaminergic

Figure 1. a)Wakefulness-, b) NREM sleep- and c) REM sleep-generating neuronal systems in the rat brain.Descending projections to the respiratory and hypoglossal motor neurons in the medulla are also shown.Solid lines indicate active neuronal groups and projections, respectively. Dashed lines and decreasedsymbol size indicate suppressed activity. Lines terminating with an arrow indicate excitatory projections,lines terminating with an oval indicate inhibitory projections and lines terminating in a diamond indicatemixed excitatory and inhibitory projections for acetylcholine. The progressive suppression of hypoglossalmotor output to genioglossus muscle from wakefulness to NREM and REM sleep is illustrated by reducedline thickness. Figure adapted from Horner (2008), with permission from the publisher.

2 ERS Handbook: Respiratory Sleep Medicine

switch (McGinty et al., 2000). This is anall-or-nothing process that prevents theoccurrence of intermediate consciousstates. At sleep onset, neurons in theventrolateral pre-optic area (VLPO),anterior hypothalamus and basal forebrainare activated and inhibit the arousalsystems detailed previously. In particular,the VLPO neurons containing theinhibitory neurotransmitters c-aminobutyricacid (GABA) and galanin, project to (andinhibit) the wake-promoting regions of theascending reticular system (Sherin et al.,1998) and the descending brainstemarousal neurons (fig. 1b).

REM sleep occurs with activation ofcholinergic neurons in the laterodorsal andpedunculopontine tegmental nuclei. Thischolinergic activation occurs whenwithdrawal of the aminergic arousal systems(noradrenergic neurons in the locuscoeruleus and serotonergic neurons in thedorsal raphe nuclei) produces disinhibition.This causes the release of acetylcholine,which triggers the increased neural activitythat is a feature of REM sleep. Suppressionof motor activity, the other marker of REMsleep, is generated by glutamate-mediatedactivation of descending medullary reticularformation relay neurons (fig. 1c). Theactivity of these neurons is inhibitory tospinal motor neurons via the release ofglycine and to a lesser extent GABA(Reinoso-Suarez et al., 2001).

The cycle of wakefulness and sleep Thetransition from wake to sleep can be difficultto determine as there are typically briefperiods of drowsiness with transient burstsof wakefulness before sleep consolidation.

NREM sleep is conventionally divided intothree or four stages according to theguidelines laid out by the AmericanAcademy of Sleep Medicine in 2007. Thesestages approximately represent the depth ofsleep and are analysed using standardisedcriteria (see chapter 5).

In adults sleep is most often initiatedthrough NREM sleep and is marked bysynchronisation of EEG activity (for furtherdescription of the EEG that defines thestages of sleep see chapter 5). The overnightsleep patterns in a healthy young adult areshown in figure 2. NREM predominatesearly in the night with episodes of REMsleep occurring in approximately 90-minintervals. The 90-min NREM–REM cycle isrepeated approximately three to six timesduring the night, and the duration of REMsleep increases as the night progresses. Thepreferential occurrence of NREM sleep(particularly slow-wave sleep) early in thenight is coincidental with sleephomeostasis, while the predominance ofREM sleep later in the night is thought to beassociated with the circadian rhythm of corebody temperature.

Sleep cycles in ageing Sleep is essential forlife in humans. Total sleep deprivation over2–3 weeks impairs thermal regulation,energy balance and immune function,eventually causing death. The requirementsfor sleep vary with age. In infants, active(REM) sleep dominates in the first 1–2 months of life. After 3 months NREMsleep begins to dominate, and by 5 yrs ofage adult sleep stages are established. Thepercentage of REM sleep is reduced to adultlevels by 10 yrs of age.

AwakeREM

Stage 4Stage 3Stage 2Stage 1

Hours of recording1 2 3 4 5 60

Figure 2. A typical overnight sleep hypnogram illustrating the sleep cycles that occur overnight in a youngmale. REM sleep is seen approximately every 90 min and there are occasional brief arousals from sleep.


In adults, optimal sleep duration varies.Sleep restriction to ,5 h a night causes areduction in psychomotor vigilance (Dingeset al., 1997), a decline in mood andmotivation and a worse performance onmemory tests. Increasing sleep opportunityup to 10 h a night improves cognitivefunction. Most estimates suggest that 7.5–8.5 h of sleep are required for optimalperformance.

Ageing influences sleep cycles, with olderpeople reporting that they experiencedifficulty in maintaining sleep and increasedawakenings (fig. 3). Morning preference alsoincreases with age (Taillard et al., 2004);however, this may be due to changing workschedules or variation in social activities, aswell as changes in the physiologicalrequirements for sleep (Dijk et al., 2000).The increased number of arousals per nightmay be a consequence of the decline in theneural systems that regulate sleep or an age-related change in the arousal thresholds toexternal stimuli. Interestingly, althougholder adults experience more awakeningduring sleep, they do not seem to have anymore problems returning to sleep onceawake (Klerman et al., 2004).

By the age of 75 yrs there may be no deepsleep. The loss of the deep sleep with

increasing age may be due to a reduction inthe amplitude of delta waves detected onthe EEG recordings, meaning that stage 3–4sleep is not documented despite thepresence of delta waves. Alternatively,disrupted synchronisation of neuronalactivation may occur as a result of an age-related decline in the neural systems thatregulate sleep. An increase in the lightersleep partially compensates for the loss ofdeep sleep (Van Cauter et al., 2000) butthere is also a reduction in the number ofsleep spindles and K complexes. Theduration of REM sleep tends to remainconstant throughout adulthood (Landoltet al., 1996), although a reduction in theproportion of REM sleep has been reportedby some (Van Cauter et al., 2000).

Further reading

N Aserinsky E, et al. (1953). Regularlyoccurring periods of eye motility, andconcomitant phenomena, during sleep.Science; 118: 273–274.

N Boselli M, et al. (1998). Effect of age onEEG arousals in normal sleep. Sleep; 21:361–367.

N Dijk DJ, et al. (2000). Contribution ofcircadian physiology and sleep home-ostasis to age-related changes in humansleep. Chronobiol Int; 17: 285–311.

N Dinges DF, et al. (1997). Cumulativesleepiness, mood disturbance, and psy-chomotor vigilance performance decre-ments during a week of sleep restricted to4–5 hours per night. Sleep; 20: 267–277.

N Horner RL. (2008). Neuromodulation ofhypoglossal motoneurons during sleep.Respir Physiol Neurobiol; 164: 179–196.

N Iber C, et al. (2007). The AASM Manualfor the Scoring of Sleep and AssociatedEvents: Rules, Terminology and TechnicalSpecifications. 1st Edn. Westchester,American Academy of Sleep Medicine.

N Klerman EB, et al. (2004). Older peopleawaken more frequently but fall backasleep at the same rate as youngerpeople. Sleep; 27: 793–798.

N Landolt HP, et al. (1996). Effect of age onthe sleep EEG: slow-wave activity andspindle frequency activity in young andmiddle-aged men. Brain Res; 738: 205–212.

35

30

25

20

15

10

5

0

Aro

usal

/hou

r of

sle

ep

Age yrs100 20 30 40 50 60 70 80

r=0.852

p<0.00001

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●●

●●

●●

●●

●

●

●

●

●

●●

●

●●●

●●

●

Figure 3. The influence of age on the number ofarousals (awakenings o3 s) per hour of sleep.Reproduced with permission from Boselli et al.(1998), with permission from the publisher.


N McGinty D, et al. (2000). The sleep-wakeswitch: a neuronal alarm clock. Nat Med;6: 510–511.

N Reinoso-Suarez F, et al. (2001). Brainstructures and mechanisms involved inthe generation of REM sleep. Sleep MedRev; 5: 63–77.

N Saper CB, et al. (2005). Hypothalamicregulation of sleep and circadianrhythms. Nature; 437: 1257–1263.

N Sherin JE, et al. (1998). Innervation ofhistaminergic tuberomammillary neuronsby GABAergic and galaninergic neurons

in the ventrolateral preoptic nucleus ofthe rat. J Neurosci; 18: 4705–4721.

N Taillard J, et al. (2004). Validation ofHorne and Ostberg morningness-eveningness questionnaire in a middle-aged population of French workers. J BiolRhythms; 19: 76–86.

N Van Cauter E, et al. (2000). Age-relatedchanges in slow wave sleep and REMsleep and relationship with growth hor-mone and cortisol levels in healthy men.JAMA; 284: 861–868.


Breathing during sleep andwakefulness


Control of breathing during sleep andwakefulness

Neural control of breathing Neurons withrespiratory-related activity are located withinthe ventrolateral medulla in a columnextending from the facial nucleus to theupper spinal cord in animals (fig. 1). Whenisolated, these neurons are able to maintainrhythmic motor nerve output. A bilateralcluster of ventrolateral medullary neurons,known as the preBotzinger Complex(preBotC), is essential for thisrhythmogenesis (Smith et al., 1991). Acritical subpopulation of preBotC neurons

are anatomically distinguished from otherrespiratory neurons within the ventrolateralcolumn by expression of neurokinin-1receptor (NK1R) (Gray et al., 1999). Thegradual loss of the preBotC NK1R-expressingneurons may explain why SDB is prevalent insome people. Studies of human post mortemtissue have anatomically identified a clusterof NK1R-expressing neurons within thereticular formation of the caudal brainstemand this is presumed to be the site of thepreBotC in humans (Lavezzi et al., 2008).

A second cluster of respiratory neurons inthe brainstem has also been identified.Ventral to the facial nucleus and rostral tothe preBotC, these neurons are called theretrotrapezoid nucleus (RTN) in adults andthe parafacial respiratory group (pFRG) inneonates (fig. 1). The anatomicalboundaries and functional significance ofthese neurons are still under debate andappear to depend on the experimentalconditions under which the studies areperformed. The pFRG neurons may have aprimary role in rhythm generation bycontrolling the timing of inspiratory bursts.They are also opiate-insensitive, a propertythat distinguishes them from the opiate-sensitive neurons within the preBotC. Theimpact of opiates on respiratory rhythm-generating neurons is of interest because oftheir wide spread use in pain control.

The RTN is comprised of glutamatergic,chemosensory interneurons that can beanatomically identified by expression of theNK1R and the transcription factor Phox2b(Mulkey et al., 2004). The RTN neurons arenot rhythmically active, but they arechemosensitive and stimulation in vivoincreases ventilation during wakefulness.

Key points

N Neurons with respiratory-relatedactivity are situated in theventrolateral medulla.

N Sleep onset is associated with areduction in the wakefulness drive tobreathe, chemosensitivity andmetabolism; such that ventilation fallsand PaCO2 increases above theapnoeic threshold.

N Patency of the upper airway isdetermined by airway size, negativeintraluminal pressure, extraluminalpressure (such as adipose tissue)and compliance.

N Susceptibility of the upper airway tocollapse during sleep is increased bythe sleep-related reduction in lungvolume and the reduction in reflexactivation of the upper airwaydilator muscles.


In contrast, during sleep there is noventilatory response to stimulation. Thisdiffers from central chemosensitive neuronswithin the medullary raphe, which have agreater response to focal CO2 stimulationduring sleep compared with wakefulness(Nattie et al., 2001). Overall the neurobiologyof central chemosensitivity remains unclearbecause there are many types of neuronswhich contribute to the chemoresponse.

It has been suggested that developmentaldisruption of the RTN neurons (expressingPhox2b) results in a loss in CO2/pHsensitivity and lethal respiratory abnormalitiesat birth (Dubreuil et al., 2008). Clinically,mutation of the PHOX2B gene (whichencodes the transcription factor Phox2b)causes congenital central hypoventilationsyndrome (CCHS). Future histologicalstudies will determine whether Phox2b-positive neurons in the facial motor nucleus,which is presumed to be the site of the humanRTN, are deficient in CCHS patients.

Control of breathing at the wake/sleeptransition Sleep onset is associated with aloss of the wakefulness influences onbreathing, changes in chemosensitivity, anda sleep-related reduction in metabolism thatcombine to produce the blood gas changesdescribed later in this section. The term‘wakefulness drive to breathe’ describes theinfluence of cerebral activity on the regulationof breathing. Recent brain imaging studieshave shown that the primary and sensorycortices, basal ganglia and thalamic nuclei areimplicated in volitional and behaviouralcontrol and activation of these centres arepresumably lost during sleep. The ability tomaintain stable breathing during sleep onsetis a function of how the respiratory controlsystem responds to ventilatory perturbationsthat occur at this time.

Anatomy and function of the upper airway

Anatomy The upper airway can be dividedinto the nasopharynx (from cranial base toposterior border of the hard palate), theoropharynx (from posterior border of thehard palate to the tip of the soft palate) andthe hypopharynx (from the tip of the softpalate to the tip of the epiglottis).

The shape of the pharyngeal lumen has beenstudied by looking at the anteroposteriorand tranverse diameters at these levels. Thefactors that influence the prevalence of OSA,such as obesity, age and sex may change theupper airway anatomy.

The shape of the oropharynx andhypopharynx becomes more spherical(compared with the normal oval shape) withincreasing BMI, mainly due to a decrease inthe transverse axis. Moreover, in lean OSApatients the anatomy of the upper airway iscorrelated with the AHI, indicating that theanatomy of the upper airway in obesepatients is only one of several factorsinfluencing the upper airway occlusion(Mayer et al., 1996).

Caudal

Ventral

VII

CB

SO

LCNTS

RTN/pFRG preBötC

LRN

VRC

rVG

NA

BC cVG

Figure 1. A parasagittal section through themedulla and caudal pons highlighting theanatomical arrangement of respiratory neuronswithin the ventrolateral medulla. Theventrolateral respiratory column (VRC), includingthe Botzinger complex (BC), preBotzingercomplex (preBotC), rostral ventral respiratorygroup (rVG) and caudal ventral respiratory group(cVG) are shown in pink. The RTN/pFRG andpreBotC, the regions discussed in this chapter, arehighlighted in red. CB: cerebellum; LC: locusceruleus; NTS: nucleus of the solitary tract; NA:nucleus ambiguus; LRN: lateral reticular nucleus;VII: facial motor nucleus; SO: superior olive.Taken from McKay et al. (2010), with permissionfrom the publisher.


Older OSA patients have larger airways at allpharyngeal levels, compared with youngerpatients, although this does not appear toprevent sleep-related airway collapse, sincethe prevalence of OSA increases with age.Interestingly, ageing correlates closely andpositively with pharyngeal resistance in men,but not in women (White et al., 1985).

The anatomy of the upper airway iscorrelated with the AHI in lean OSA patients(BMI ,27 kg?m-2), indicating that theanatomy of the upper airway in obesepatients is only one of several factorsinfluencing the upper airway occlusion(Mayer et al., 1996).

The size of the tongue can be measured bycalculating the interhypoglossal length andthe genioglossal width. This approach canfurther be optimised by looking at a three-dimensional reconstruction of the upperairway lumen from computed tomography(CT) or magnetic resonance images andcalculating the airway volume at the threelevels (fig. 2) (De Backer et al., 2008).Beyond looking at the upper airway volumein three dimensions, one can also constructa computational grid starting from thesegmented upper airway and calculateresistance using computational fluid dynamics.

Alternative methods of describing the upperairway anatomy, such as cephalometry, giveless direct information regarding the airwaydimensions, but it is striking to see thatthere are still significant differences for anumber of cephalometric measures betweenOSA and snorers for instance (Mayer et al.,1996). Using cephalometry, the distancefrom the mandibular plane to the hyoid(MP-H) can be measured; the posteriorairway space can also be measured from thedorsum of the tongue to the posteriorpharyngeal wall, as well as the length of thesoft palate. In male OSA patients the posteriorairway space is narrow, and the tongue andsoft palate are enlarged with the hyoid boneinferiorly positioned (measured with theMP-H) (Kapsimalis et al., 2002; Kapsimaliset al., 2002a). When women with OSA werecompared with men, they had shorter softpalates, but the MP-H distance was similarlyabnormal in both men and women.

Fat deposition around the airway can also bemeasured using MRI. The volume of the softtissue structures surrounding the upperairway is larger in patients with OSAcompared with healthy people (Schwab et al.,2003) with the volume of the lateralpharyngeal walls and total soft tissue beingsignificantly larger in OSA patients,compared with healthy people. Thisphenomenon influences the upper airwayvolume per se, but also influences thecharacteristics, mainly collapsibility, of theupper airway.

The upper airway volume (in absolute terms)is larger in men than in women, although menhave a greater tendency to collapse theirairways, which may explain the differences inOSA prevalence between men and women.Overall there is a relationship between theupper airway volume and collapsibility, andthis relationship is sex specific.

Reflex activation of upper airway dilatormuscles The patency of the upper airway isdependent on the upper airway size, negativeintrapharyngeal pressures and compliance.These factors are influenced by hypoxia andhypercapnia, sleep/wake transitions, bloodpressure and sex-specific hormones.Moreover, the reflex activation of the upperairway dilator muscles also determines airwaypatency. Specifically, the reflex activationresponses are important to explain thedifferences in upper airway resistanceobserved between healthy people and OSApatients, and between men and women.

Influence of lung volumes on upper airwaygeometry Lung volumes may influence theactivation of the pharyngeal dilator muscles.An association between upper airway calibreand changes in upper airway volumes hasbeen described for both healthy people andOSA patients during wakefulness. Reducedlung volumes lead to increased pharyngealresistance. During sleep, reduced lung volumeleads to increased inspiratory airflowresistance and increased genioglossus muscleactivation. Specifically, the pharynx is morecollapsible during sleep at low lung volumes.

Critical closing pressure Overall, both upperairway geometry and neuromuscular


activation will contribute to the upper airwaycollapsibility, which can be measured usingthe critical closing pressure. Thismeasurement is based on the principle thatthe upper airway can be modelled as acollapsible segment or Starling resistor. Thecollapsible segment is subject to thesurrounding or critical pressure (Pcrit) thatdetermines the collapse. The airwaycollapses when the intraluminal pressuredrops below the Pcrit. Elevations in Pcrit inOSA patients can be related to anatomicfactors and/or disturbances inneuromuscular control. Passive Pcrit

(measured shortly after reducing upperpressure) and active Pcrit (after a longer timeperiod of pressure drop whenneuromuscular activation occurs inresponse to the airflow obstruction) aregenerally considered to be differentmeasurements. Active Pcrit reflectsneuromuscular activation, while passive Pcrit

reflects collapsibility related only toanatomic characteristics. Active responsesthat can compensate for high passive Pcrit

are blunted in OSA patients. Thus, OSApatients have disturbed passive and activePcrit. Obesity may enhance passive Pcrit

directly, or by lowering lung volumes.Weight loss can reduce active Pcrit byrecovery of active neuromuscular control.

Visceral adiposity is related to a decrease inneuromuscular control by an increase in theamount of circulating inflammatorycytokines such as tumour necrosis factor(TNF)-a, TNF-a receptor I, interleukin (IL)-6and IL-1b. In addition, inflammation of theupper airway structures themselves maylead to cell infiltration and remodelling ofthe extracellular matrix in the upper airwayproducing neurosensory deficits. The lattermay impair protective reflex responses tonegative intraluminal pressures and, thus,also compromise the neuromuscularresponses (Schwartz et al., 2010).

Anatomy as a predictor of therapeuticinterventions All treatments for OSA thatincrease upper airway volume will help toprotect the airway from collapse. Measuringupper airway volume, however, is notsufficient to predict the response totreatment since some interventions mayincrease overall upper airway volume, butstill induce an increase in upper airwayresistance, by narrowing the upper airway at

Nasopharynx (NP)

Trachea (T)HP

OLOPNP

CHN

Caudal viewMandible

MandibleTonguePharyngeal airwayEpiglottis

Lateral view

NostrilsChoanae

(CH) OverlapRegion(OL)

Oropharynx (OP)

Hypopharynx(HP)

O PL

Figure 2. Three-dimensional reconstruction of the upper airway from CT and MRI scans. L: larynx; O:oropharynx; P: pharynx. Reproduced from De Backer et al. (2008), with permission from the publisher.


a specific point. Therefore, predicting theeffect of upper airway interventions is notpossible by simply measuring the overallincrease in upper airway volume. There is aneed to study the effect on the upper airwayresistance. Changes in upper airwayresistance following interventions are moresensitive and specific for the improvementin the upper airway collapsibility thanchanges in upper airway volume alone(De Backer et al., 2007).

Overall, the upper airway anatomy and sizeis a crucial factor for upper airway function.A larger airway may protect against collapse,but this phenomenon should be consideredas sex specific since women have a reducedtendency to upper airway collapse eventhough they have anatomically smallerupper airways. Neuromuscular control ofthe upper airway, reflected as the responsesto applied negative pressure, also has animportant effect on the maintenance ofupper airway patency.

Respiratory mechanics and ventilation

Interest in sleep-related respiratory changeshas increased progressively from the early1980s. This was mainly due to therecognition of the clinical relevance of sleep-related respiratory diseases. In addition, theavailability of technologies has increased. Inearly studies, small cohorts of healthy youngvolunteers were studied, and mouthpiecesor face masks were utilised to track changesin breathing pattern. The problem with thesestudies was that the measurementtechniques influenced the results (e.g. a facemask increasing levels of inspired CO2).Later on, less invasive approaches were

utilised. These studies helped to identify thetechniques required to diagnose patientswith respiratory disorders during sleep.

In healthy young individuals, the transitionfrom wakefulness into sleep is associatedwith significant changes in lung mechanicsand ventilation (see table 1 for summary).A reduction in tidal volume (VT) occurs withdeepening levels of NREM sleep. A furtherreduction in VT, up to 25% compared withthe awake state, has been described duringREM sleep (Douglas et al., 1982a). Thereduction in VT produces a comparablereduction in V9E during the different sleepphases. The results of Douglas et al. 1982a,were obtained at an altitude of 1600 m andby the use of a face mask; however, theywere partially confirmed by Stradling et al.1985, who found a 20% reduction in alveolarventilation (VA) and 13% reduction in VT

during REM sleep. The reduction in VT wasattributed to a reduction in ventilatory driveas reflected by the decrease in VT/inspiratorytime (TI) ratio, again during REM, whereasrespiratory rate was substantially unchanged.

In early studies the reduction in V9E duringsleep was attributed, at least in part, to areduction in metabolic rate, although theobservation of an increase in PaCO2 and theproportional decrease in PaO2 favoured areduced ventilatory drive. Further studiesclarified that the ventilatory response tohypercapnia and hypoxia are reduced duringsleep, particularly during phasic REM, whenmotor output is reduced and variable inresponse to a rising chemoreceptor sensoryinput. The sleep-related reduction inchemosensitivity is shown in figure 3.

During sleep upper and lower airway calibreis reduced, as discussed previously. Centraldrive to upper airway dilator muscles, e.g. thegenioglossus and the alae nasi, and torespiratory pump muscles e.g. thediaphragm, is also reduced during thetransition from wakefulness to sleep. As aconsequence, a small increase in totalpulmonary resistance of ,3 cmH2O?L-1?s-1

can be observed in young non-obeseindividuals. The drive to upper airwaymuscle is further reduced during thetransition from NREM to REM sleep. Also,

Table 1. Changes in ventilatory parameters in thetransition from wakefulness to sleep in normalindividuals.

NREM REM

V9E Q QQ

VT Q QQ

RR « «VT/TI « Q

RR: respiratory rate.


phasic REM can lead to a reduction andfractionation of the diaphragm’s activityduring inspiration (Pack, 1995). As aconsequence of reduced muscle activity,upper airway patency is at risk particularlyduring REM sleep. In addition topredisposing anatomical factors, two mainphysiological abnormalities seem to play asignificant role in the development of upperairway obstruction during sleep: 1) thereduced dilator activity of the genioglossusand 2) a reduction in lung volumes(discussed previously). The role of thereduced activity of the genioglossus is welldocumented in the literature; the effect ofthe reduction in lung volumes on upperairway patency is still under investigation.

All the above-mentioned abnormalities inthe respiratory responses during sleep aremore pronounced in elderly individuals, withassociated more frequent episodes ofperiodic breathing and apnoeas duringsleep, which may result in ‘‘normal’’transient hypoxaemia (Shore et al., 1985).

Further reading

N De Backer JW, et al. (2007). Functionalimaging using computational fluiddynamics to predict treatment successof mandibular advancement devices insleep-disordered breathing. J Biomech;40: 3708–3714.

N De Backer JW, et al. (2008). Novelimaging techniques using computermethods for the evaluation of the upperairway in patients with sleep-disorderedbreathing: a comprehensive review. SleepMed Rev; 12: 437–447.

N Douglas NJ, et al. (1982a). Respirationduring sleep in normal man. Thorax; 37:840–844.

N Douglas NJ, et al. (1982b). Hypercapnicventilatory response in sleeping adults.Am Rev Respir Dis; 126: 758–762.

N Dubreuil V, et al. (2008). A humanmutation in Phox2b causes lack of CO2

chemosensitivity, fatal central apnea, andspecific loss of parafacial neurons. ProcNatl Acad Sci USA; 105: 1067–1072.

N Gray PA, et al. (1999). Modulation ofrespiratory frequency by peptidergic inputto rhythmogenic neurons in the pre-Botzinger complex. Science; 286: 1566–1568.

N Kapsimalis F, et al. (2002). Gender andobstructive sleep apnea syndrome, part 2:mechanisms. Sleep; 25: 499–506.

N Kapsimalis F, et al. (2002a). Gender andobstructive sleep apnea syndrome, part 1:clinical features. Sleep; 25: 412–419.

N Lavezzi AM, et al. (2008). Functionalneuroanatomy of the human pre-Botzinger complex with particular refer-ence to sudden unexplained perinatal andinfant death. Neuropathology; 28: 10–16.

N Mayer P, et al. (1996). Relationshipbetween body mass index, age and upperairway measurements in snorers andsleep apnoea patients. Eur Respir J; 9:1801–1809.

N McKay LC, et al. (2010). Physiology andneural control of breathing during sleep.Eur Respir Monogr; 50: 1–16.

N Mulkey DK, et al. (2004). Respiratorycontrol by ventral surface chemoreceptorneurons in rats. Nat Neurosci; 7: 1360–1369.

N Nattie EE, et al. (2001). CO2 dialysis inthe medullary raphe of the rat increasesventilation in sleep. J Appl Physiol; 90:1247–1257.

20

15

10

5

0

Ven

tilat

ion

L·m

in-1

End-tidal PCO2 mmHg3530 40 45 50

Awake

Stage 3/4

Stage 2

REM sleep

Figure 3. The relationship between ventilation andend-tidal CO2 tension measured duringwakefulness, NREM (light stage 2, and deep stage3/4) and REM sleep in 12 healthy people.Reproduced from Douglas et al. (1982b), withpermission from the publisher.


N Pack AI. (1995). Changes in respiratorymotor activity during REM sleep. In:Dempsey J, Pack AI, eds. Regulation ofbreathing. New York, Marcel Dekker; pp.983–1002.

N Schwab RJ, et al. (2003). Identificationof upper airway anatomic risk factorsfor obstructive sleep apnea with volu-metric magnetic resonance imaging.Am J Respir Crit Care Med; 168:522–530.

N Schwartz AR, et al. (2010). Obesity andupper airway control during sleep. J ApplPhysiol; 108: 430–435.

N Shore ET, et al. (1985). Ventilatory andarousal patterns during sleep in normalyoung and elderly subjects. J Appl Physiol;59: 1607–1615.

N Smith JC, et al. (1991). Pre-Botzingercomplex: a brainstem region that maygenerate respiratory rhythm in mammals.Science; 254: 726–729.

N Stradling JR, et al. (1985). Changes inventilation and its components in normalsubjects during sleep. Thorax; 40: 364–370.

N White DP, et al. (1985). Pharyngeal resistancein normal humans: influence of gender, age,and obesity. J Appl Physiol; 58: 365–371.


Homeostatic regulationduring sleep


Autonomic and cardiovascular regulation

The autonomic nervous system (ANS) acts asa motor system undertaking a large numberof specialised tasks, stimulatory andinhibitory, in a wide range of target organs,tissues, blood vessels, glands and even singleunits such as mast cells. Closely associatedwith the efferent nerves is a complex networkof afferent fibres that relay sensorychemoreceptor, mechanoreceptor,nociceptor, thermoreceptor, baroreceptor andosmoreceptor signals back to central nervoussystem (CNS) centres where the complexhomeostatic reflex arcs are regulated.

The ANS can be divided into three based onanatomical and physiologicalconsiderations. The sympathetic divisionforms the major part of the ANS, innervatingmore structures than the parasympathetic

system. Pre-ganglionic cell bodies originatein the thoracic and lumbar regions of thespinal cord, and synapse with post-ganglionic fibres that are conveyed byperipheral nerves to their targets. Theparasympathetic division arises in the brainstem and supplies the seventh, ninth andtenth cranial nerves.

Sleep onset is associated with markedcardiorespiratory changes. Depending onthe stage of sleep, different patterns ofhaemodynamic and autonomic responsesare observed. During NREM sleep there aredecreases in heart rate, systolic bloodpressure and cardiac output of up to 15%.These changes, which are most marked inslow-wave sleep, are thought to occur as aresult of changes in autonomic activity. Dataon autonomic function during sleep inhumans are still limited owing tomethodological problems (Smith et al., 1998).

Methodological issues In view of theanatomical position of the ANS, directlymeasuring autonomic activity, particularlyduring sleep, is not easy. Advances intechnology and pharmacology have allowedthe knowledge base of autonomic activity inhumans to expand significantly. Techniquesare available to directly measuresympathetic activity to muscle relativelynoninvasively (Smith et al., 1998) and anumber of easily measured surrogatemarkers of ANS activity, such as pulsetransit time (PTT) (Smith et al., 1999) andR–R variability (Ferini-Strambi et al., 1992;Roche et al., 1999; Khoo et al., 2001), arefeasible. However, for many ANSparameters, significant variations may existeven in healthy populations and betweenage groups. Intra-individual variability may

Key points

N Sleep onset is associated withchanges in the ANS that reduce heartrate, systolic blood pressure andcardiac output.

N OSA causes hypertension with elevatedsympathetic outflow to the vasculature.

N Circadian rhythms regulate the releaseof many hormones during sleep, suchas cortisol, growth hormone, andleptin and ghrelin.

N Sleep deprivation influences metabolicregulation and is associated withglucose intolerance and dysregulationof appetite hormones.


also be present. In view of this variation, careneeds to be taken when interpreting datafrom studies with small samples or those withnormal volunteer subjects who are notrepresentative of the general population.

Heart rate variability Accurately measuringparasympathetic activity is difficult and, todate, vagal–cardiac nerve traffic has not beendirectly measured in humans. Quantificationof vagal activity therefore has to be measuredindirectly using variations in cardiovascularparameters. Increased vagal tone mayproduce a varying pulse rate, atrioventriculardissociation or even transient sinus arrest.The sinoatrial node is also under the influenceof the sympathetic system. Whereas theresponse of heart rate to vagal stimulation isprompt (,0.25 s), the response tosympathetic activity takes ,5 s, with an evenlonger delay in the decay of the effect.

Heart rate variability (HRV) analysis, whichcan be performed easily and noninvasively,can therefore give an insight into bothsympathetic and parasympathetic activityduring sleep (Stein et al., 2012). Using R–Rinterval analysis by fast Fouriertransformation, a number of differentpatterns can be identified. A cyclical patternrelating to respiratory rate (sinusarrhythmia; ,0.3 Hz) is correlated withvagal tone. There are also lower frequencypatterns associated with baroreflex controlof heart rate mediated via both sympatheticand parasympathetic efferents (0.15 Hz),and an even slower component relating tohumoral and thermodynamic control of thecardiovascular system (0.05 Hz). Acutechanges in the R–R interval can also bedetected without the need for fast Fouriertransformation technology and, in patientswith OSA, have been used as indirectmarkers of sympathetic and vagal activity(Stein et al., 2012). The ratio of the longestR–R interval before a body movementduring sleep to the shortest in the periodimmediately after movement has been usedas an index of sympathetic activity. Likewise,the sleep/wake R–R ratio has been used toestimate parasympathetic tone.

The effects of the different sleep stages onR–R variability have been illustrated by

several studies establishing a very closerelationship between parasympathetic anddelta activities (fig. 1) (Jurysta et al., 2003)although this is modulated by age (Jurystaet al., 2006).

Autonomic ‘stress tests’ There is a number oftests of autonomic dysfunction that aresimple, do not require sophisticatedtechnology and can be performed at thebedside. However, a major limitation is thelack of standardisation. The earliest testswere devised by Ewing et al. (1980) forclinical evaluation of diabetics and consistedof the heart rate responses to the Valsalvamanoeuvre, standing up and breathingdeeply, and the blood pressure responses tostanding up and sustained hand gripping.The outcomes are usually expressed asdichotomous (positive or negative). Theycan suffer from poor reproducibility and lackof standardisation, which may limit theirusefulness. Nevertheless, they have beenpowerful enough to detect abnormalities ina population of OSA patients that weredistinct from normal controls (Veale et al.,1992) and that are corrected by CPAPtherapy (Veale et al., 1996). The acetylcholinesweat-spot test has been proposed as analternative test that is a more reproducibleand sensitive marker of parasympatheticdysfunction (Stewart et al., 1994). Whenperforming any of the above tests, it isimportant to take into account concurrentcardiopulmonary disease and drug therapy,which may interfere with interpretation.

Blood pressure monitoring Blood pressurechange is a surrogate marker of sympatheticactivity and considerable data on this topichave been accumulated from both animaland human studies. Continuous beat-to-beat monitoring is necessary to follow thefluctuating haemodynamic changes that canoccur over a number of seconds and thiscan be performed noninvasively using aninfrared plethysmographic volume clamp(Portapres1; Finapres Medical Systems BV,Amsterdam, the Netherlands) (Imholz et al.,1993) method, which produces a tracesimilar to an intra-arterial line. Thistechnique, which takes measurements fromthe digital artery, gives values that are not


quite the same as the true systemic pressureand are affected by hand position. It istherefore not accurate at determining absoluteblood pressure recordings, but it is very goodat tracking changes in blood pressure.

PTT and peripheral arterial tonometry Pulsetransit time (PTT) is the time taken for thearterial pressure wave to travel from theaortic valve (measured by the ECG R-wave)to the periphery (measured by fingerphotoplethysmography). With rises in bloodpressure, the arterial wall tension increasesand becomes stiffer and, hence, transmitsthis pressure wave more swiftly. PTT istherefore inversely correlated to changes inblood pressure (Smith et al., 1999). Usingthe ECG R-wave as the starting point for PTTinevitably creates a small error as the pre-ejection systolic period (PEP) will also beincluded in the interval measured. The PEPmay also be affected by factors thatinfluence blood pressure and not necessarilyin the same direction as the true PTT. As aconsequence, PTT is not particularly good atpredicting absolute beat-to-beat bloodpressure values as there is a tendency of therelationship between PTT and bloodpressure to drift over time (Smith et al.,1999). This does not affect its ability todetect short-term changes in blood pressureand PTT has been shown to be capable of

detecting acute changes in blood pressureassociated with sympathetic activationduring arousal and with the pulsus paradoxusgenerated by the high pleural pressureswings (Pitson et al., 1998) encountered inincreased upper airway resistance duringsleep (Argod et al., 2000). Many of thechanges in measured PTT in response to alarge inspiratory effort appear, in fact, to bedue to lengthening of the PEP rather than ofthe true PTT. PTT is a marker of inspiratoryeffort (amplitude of the oscillations in timewith respiration) (Pitson et al., 1998) and ofautonomic activation at point of arousal (dipin baseline level corresponding to surge inblood pressure at apnoea termination)(Pepin et al., 2005).

Peripheral arterial tonometry (PAT) wasintroduced into the field of sleep medicineby Pillar et al. (2002). Upper airwayobstruction (O’Donnell et al., 2002) andarousals (Pillar et al., 2003) have beenshown to be adequately identified by thistechnique. This led to the concept of ascreening technique (Penzel et al., 2004;Zou et al., 2006). In addition, the PATtechnique enables REM sleep recognition(Lavie et al., 2000; Dvir et al., 2002;Herscovici et al., 2007). Thisplethysmographic technique measuresperipheral arterial tone. The device envelops

REMAwake

III

IIIIV

1.5

10.5

00.80.60.40.2

0

Del

ta P

(nu

)H

Fnu

Time min

300 60 90 120 150 180 210 240 270 300 330

Figure 1. Sleep recording evidencing a close correlation between ECG high frequencies (as measured byHRV) and delta activity in the EEG. Reproduced from Jurysta et al. (2003), with permission from thepublisher. Delta P: delta EEG power; nu: normalised units; HFnu: normalised high frequency.


the finger up to and beyond its tip with auniform pressure field. Pressure within theprobe originates from a pressurised balloonlocated over its outside wall. Pulsatilevolume signals are recorded as opticaldensity changes from the finger’s palmarsurface within the applied pressure field.

REM sleep has been shown to be associatedwith considerable attenuation of the PATsignal. The decrease in amplitude beganduring NREM sleep and reached a nadirduring REM sleep (Lavie et al., 2000). Thisdemonstrated that REM sleep in humanswas associated with considerable peripheralvasoconstriction (Lavie et al., 2000). REM-related vasoconstriction may, for instance,predispose patients with compromisedcoronary arteries to ischaemic events duringsleep (Hanly et al., 1993).

Microneurography The development ofmicroneurographic techniques has enabledmonitoring of sympathetic autonomicactivity to be performed relativelynoninvasively in humans. A microelectrodeis passed into a superficial nerve or muscleand a recording taken that is thought torepresent the sympathetic vasoconstrictorsignal. This technique is not withoutdrawbacks in that any muscle activity in thevicinity of the probe may render readingsmeaningless and caution is required whenextrapolating recordings of sympathetictraffic taken from superficial sites to that ofdeeper tissues. However, microneurographyhas proven to be a highly effective researchtool in tracking the acute and chronicchanges in sympathetic activity occurring asa consequence of OSA (Somers et al., 1995).The technique has evidenced the changes insympathetic activity occurring during thevarious sleep stages as well as sympatheticactivation associated with sleepfragmentation and sleep apnoea (Somerset al., 1993; Leuenberger et al., 1995;Waradekar et al., 1996; Grassi et al., 2005;Carlson et al., 1993; Hedner et al., 1995).

Physiological data: insights into OSApathophysiology

Physiological changes Sympathetic trafficmeasured by microneurography has been

correlated to the blood pressure changesobserved in humans during sleep (Somerset al., 1993) and surgical sympathectomy hasbeen observed to attenuate the decrease inblood pressure associated with NREM sleep(Baccelli et al., 1969). Parasympatheticactivity tends to increase during NREMsleep (Jurysta et al., 2003; Furlan et al.,1990) and is largely responsible for thedecrease in heart rate and accentuation ofany sinus arrhythmia. Not uncommonly,this increase in vagal tone may inducefirst-degree heart block or evenatrioventricular dissociation.

The haemodynamic changes that occurduring REM sleep include erratic increasesin pulse rate and blood pressure. Somerset al. (1993) recorded instability in heart rateand blood pressure, which was associatedwith a level of sympathetic traffic significantlyhigher than that observed during wakefulnessin normal healthy people. It is possible thatthis higher level of sympathetic activity andcardiovascular load associated with REMsleep could trigger platelet aggregation,atheromatous plaque rupture or coronaryvasospasm, and explain why the incidence ofmyocardial infarction and ischaemic stroke ishigher in the early morning hours when REMsleep is more prevalent.

The increases in blood pressure and musclesympathetic activity tend to coincide withthe phasic eye movements of REM sleep andbecome less pronounced as the duration ofREM sleep increases. Overall, the increasesobserved in the haemodynamicmeasurements underestimate the increasein sympathetic activity during REM sleep.Whether this is due to the sympatheticoutflow to the cardiovascular system andskeletal muscle being dissociated or to thebaroreceptor reflexes buffering this responsewith increased parasympathetic activity, isnot clear. Nevertheless, REM sleep is aperiod of labile sympathetic andhaemodynamic activity.

Arousals from NREM sleep are associatedwith transient increases in heart rate andblood pressure, which are thought to bebrought about by a rise in sympatheticactivity and a withdrawal of parasympathetic


tone (Somers et al., 1993). Indeed, arousalstimuli that produce K-complexes on EEGhave been associated with the samehaemodynamic changes and increases inmuscle sympathetic traffic (Pillar et al.,2002; O’Donnell et al., 2002; Somers et al.,1993). The recurrent activation of thissympathetic and haemodynamic responsein OSA patients may be an important factorexplaining the enhanced cardiovascularmorbidity associated with this condition.Most of the human studies on thesympathetic system during sleep have beenperformed using recordings of skeletalmuscle nerve traffic and it is not clearwhether these findings are representative ofactivity in other vascular beds such as in theheart or brain. The neural effectormechanisms controlling the short-termhaemodynamic changes during sleep appearto be highly complex. For instance, there is amarked modulation of baroreceptorsensitivity during the different phases ofsleep (Conway et al., 1983).

ANS assessment and OSA pathophysiologyAlthough it remains uncertain whatfacilitates cardiovascular disease in apnoeicpatients, OSA is recognised as a cause ofhypertension (Baguet et al., 2005; Peppardet al., 2000), and the development ofhypertension might be the initiating factorthat results in the comorbidities ofischaemic stroke, cardiac ischaemia andchronic heart failure (McNicholas et al.,2007). Elevated tonic sympathetic outflowto the vasculature elicited by sleep apnoeapersists into the waking daytime hours(Somers et al., 1995; Grassi et al., 2005)and this, it is thought, may mediate theelevated pressure seen in those with OSAS(Fletcher et al., 1992a, b). Recently, wedescribed a new model exposing normalvolunteers to intermittent hypoxia (Tamisieret al., 2009). Using this model, we haveshown that intermittent hypoxia isassociated with an increase in sympatheticactivity (fig. 2) and a persistent increase inblood pressure during daytime (Tamisieret al., 2011). This supports the role ofsympathetic activation in hypertensiongenesis in OSA patients.

Metabolic regulation

Circadian rhythmicity and sleep/wakehomeostasis both have profound effects onmetabolic regulation. The release of growthhormone (GH) and cortisol, as well as otherhormones that control appetite regulation,such as leptin and ghrelin, are influenced bycircadian and sleep/wake processes(Leproult et al., 2011; Morselli et al., 2012).Circadian rhythmicity is generated in thesuprachiasmatic nuclei of thehypothalamus, and transmitted to otherareas of the brain and to the periphery viadirect neuronal connections and viasympathetic and hormonal signalling. Bycontrast, the molecular mechanisms bywhich sleep/wake homeostasis influencesthe metabolic and hormonal responses havenot been fully elucidated. Experimentalstudies, involving sleep deprivation duringthe night and sleep recovery during the day,

35.532.0

25.0

18.0

11.0

c)

a)

b)

28.5

14.5

21.5

7.5Pre-exposure Post-exposure

MSN

A b

urst

s·m

in-1

●

●■

■

●●

●

●

●

●●

●

●

●

Figure 2. Exposure to intermittent hypoxia after 13nights led to an increase in sympathetic activitymeasured by muscle sympathetic nerve activity(MSNA). Representative MSNA neurograms fromhealthy subjects a) before and b) after exposure tointermittent hypoxia. c) Sympathetic activityexpressed as MSNA bursts before and afterexposure. Reproduced from Tamisier et al. (2011),with permission from the publisher.


have been used to study the effect of sleepon metabolism and endocrine systems.Evidence for a modulatory effect of sleep onmetabolic pathways is well documented. GHand cortisol production, glucose toleranceand insulin release, and production of leptinand ghrelin are influenced by sleep timing,duration and quality.

Sleep and glucose metabolism For energygeneration, the brain depends entirely onblood glucose delivered by the circulation.Glucose homeostasis is controlled by thelevel of insulin, which lowers blood glucoselevels, and by hormones such as GH,glucagon, cortisol and catecholamines, whichincrease glucose levels in the blood. Inhealthy individuals, GH levels increase duringsleep, particularly during NREM sleep,whereas they are markedly decreased withsleep deprivation. Sleep/wake homeostasishas little influence on cortisol release,although sleep onset is associated with ashort-term inhibition of cortisol release andawakenings induce small increases.

By contrast, glucose metabolism issignificantly influenced not only by circadianrhythmicity, but also by sleep. Despiteprolonged fasting, glucose levels in theblood remain relatively constant duringsleep, thus preventing a reduction inglucose-mediated energy production in thebrain. Both nocturnal and daytime sleep areassociated with increased glucose levels.Moreover, studies using constant-rateintravenous glucose infusion have showndeterioration in glucose tolerance duringsleep. Because insulin also increases duringsleep, the most likely explanation for thereduction in glucose tolerance is a reductionin tissue glucose utilisation, including thatof the brain (Boyle et al., 1994). Thereduction in glucose utilisation seems tooccur mainly during NREM sleep. Bycontrast, during REM sleep, blood glucoseand insulin decrease toward normal levels.

Both total and partial sleep deprivationprotocols have been used in experimentalstudies. Fasting glucose levels increase after

prolonged total sleep deprivation (i.e. .120 h),but not with shorter total sleep deprivationduration. Glucose tolerance is also alteredwith total sleep deprivation, irrespective ofthe duration of sleep deprivation. Finally,reduced insulin sensitivity, after a glucoseload, has been reported as theconsequence of total sleep deprivation.Insulin sensitivity was found to be reducedafter sleep loss, but also after recoverysleep. The intravenous glucose tolerancetest has mainly been used to demonstratethe effect of partial sleep deprivation onglucose metabolism. Studies using partialsleep deprivation, a model that betterrepresents real lifestyles, demonstrated areduction in glucose tolerance and insulinsensitivity associated with sleep debt(Spiegel et al., 1999; Buxton et al., 2010).

Sleep and regulation of appetite During thepast 10 yrs, evidence has accumulated onthe modulatory effect of sleep on appetiteregulation mediated through the release ofhormones, such as the adipocyte-derivedleptin (that inhibits the ‘hunger’ signal) andstomach-derived ghrelin (that inhibits the‘satiety’ signal) (Spiegel et al., 2004). Leptinlevels are influenced by both circadianrhythmicity and sleep/wake homeostasis.Leptin levels increase at night and duringsleep, facilitating overnight fasting. Therelationship between sleep and leptin seemsto be bidirectional, as demonstrated byanimal studies: leptin administrationincreases NREM duration and decreasesREM sleep, whereas leptin deficiencydisrupts sleep architecture.

In contrast to leptin, ghrelin stimulateshunger and food intake. Very few studieshave examined the effect of sleep deprivationon nocturnal ghrelin release in normalindividuals. The results of these studies werenot univocal; however, they suggested areduction in ghrelin levels induced by sleepdeprivation (Dzaja et al., 2004). These resultsare apparently in contrast with thoseobtained in animals, where sleep deprivationpromotes food intake.


Further reading

N Argod J, et al. (2000). Comparison ofesophageal pressure with pulse transittime as a measure of respiratory effort forscoring obstructive nonapneic respiratoryevents. Am J Respir Crit Care Med; 162:87–93.

N Baccelli G, et al. (1969). Neural and non-neural mechanisms influencing circula-tion during sleep. Nature; 223: 184–185.

N Baguet JP, et al. (2005). Night-time anddiastolic hypertension are common andunderestimated conditions in newly diag-nosed apnoeic patients. J Hypertens; 23:521–527.

N Boyle PJ, et al. (1994). Diminished brainglucose metabolism is a significantdeterminant for falling rates of systemicglucose utilization during sleep in normalhumans. J Clin Invest; 93: 529–535.

N Buxton OM, et al. (2010). Sleep restric-tion for 1 week reduces insulin sensitivityin healthy men. Diabetes; 59: 2126–2133.

N Carlson JT, et al. (1993). Augmentedresting sympathetic activity in awakepatients with obstructive sleep apnea.Chest; 103: 1763–1768.

N Conway J, et al. (1983). Involvement of thebaroreceptor reflexes in the changes inblood pressure with sleep and mentalarousal. Hypertension; 5: 746–748.

N Dvir I, et al. (2002). Evidence for fractalcorrelation properties in variations ofperipheral arterial tone during REM sleep.Am J Physiol Heart Circ Physiol; 283:H434–H439.

N Dzaja A, et al. (2004). Sleep enhancesnocturnal plasma ghrelin levels in healthysubjects. Am J Physiol Endocrinol Metab;286: E963–E967.

N Ewing DJ, et al. (1980). Assessment ofcardiovascular effects in diabetic auto-nomic neuropathy and prognostic impli-cations. Ann Intern Med; 92: 308–311.

N Ferini-Strambi L, et al. (1992). Heart ratevariability during sleep in snorers withand without obstructive sleep apnea.Chest; 102: 1023–1027.

N Fletcher EC, et al. (1992a). Sympatheticdenervation blocks blood pressure eleva-tion in episodic hypoxia. Hypertension; 20:612–619.

N Fletcher EC, et al. (1992b). Repetitive,episodic hypoxia causes diurnal elevationof blood pressure in rats. Hypertension;19: 555–561.

N Furlan R, et al. (1990). Continuous 24-hour assessment of the neural regulationof systemic arterial pressure and RRvariabilities in ambulant subjects.Circulation; 81: 537–547.

N Grassi G, et al. (2005). Obstructive sleepapnea-dependent and -independent adre-nergic activation in obesity. Hypertension;46: 321–325.

N Hanly P, et al. (1993). ST-segment depres-sion during sleep in obstructive sleepapnea. Am J Cardiol; 71: 1341–1345.

N Hedner J, et al. (1995). Reduction insympathetic activity after long-term CPAPtreatment in sleep apnoea: cardiovascularimplications. Eur Respir J; 8: 222–229.

N Herscovici S, et al. (2007). DetectingREM sleep from the finger: an automaticREM sleep algorithm based on peripheralarterial tone (PAT) and actigraphy. PhysiolMeas; 28: 129–140.

N Imholz BP, et al. (1993). Feasibility ofambulatory, continuous 24-hour fingerarterial pressure recording. Hypertension;21: 65–73.

N Jurysta F, et al. (2003). A study of thedynamic interactions between sleep EEGand heart rate variability in healthy youngmen. Clin Neurophysiol; 114: 2146–2155.

N Jurysta F, et al. (2006). The link betweencardiac autonomic activity and sleep deltapower is altered in men with sleep apnea-hypopnea syndrome. Am J Physiol RegulIntegr Comp Physiol; 291: R1165–R1171.

N Khoo MC, et al. (2001). Cardiac auto-nomic control in obstructive sleep apnea:effects of long-term CPAP therapy. Am JRespir Crit Care Med; 164: 807–812.

N Lavie P, et al. (2000). Peripheral vaso-constriction during REM sleep detectedby a new plethysmographic method. NatMed; 6: 606.

N Leproult R, et al. (2011). Effect of 1 week ofsleep restriction on testosterone levels inyoung healthy men. JAMA; 305: 2173–2174.

N Leuenberger U, et al. (1995). Surges ofmuscle sympathetic nerve activity duringobstructive apnea are linked to hypoxe-mia. J Appl Physiol; 79: 581–588.


N McNicholas WT, et al. (2007). Sleepapnoea as an independent risk factor forcardiovascular disease: current evidence,basic mechanisms and research priori-ties. Eur Respir J; 29: 156–178.

N Morselli LL, et al. (2012). Sleep and meta-bolic function. Pflugers Arch; 463: 139–160.

N O’Donnell CP, et al. (2002). The effect ofupper airway obstruction and arousal onperipheral arterial tonometry in obstruc-tive sleep apnea. Am J Respir Crit CareMed; 166: 965–971.

N Penzel T, et al. (2004). Ambulatoryrecording of sleep apnea using peripheralarterial tonometry. Conf Proc IEEE EngMed Biol Soc; 5: 3856–3859.

N Pepin JL, et al. (2005). Pulse transit timeimproves detection of sleep respiratoryevents and microarousals in children.Chest; 127: 722–730.

N Peppard PE, et al. (2000). Prospectivestudy of the association between sleep-disordered breathing and hypertension.N Engl J Med; 342: 1378–1384.

N Pillar G, et al. (2003). An automaticambulatory device for detection ofAASM defined arousals from sleep: theWP100. Sleep Med; 4: 207–212.

N Pillar G, et al. (2002). Autonomic arousalindex: an automated detection based onperipheral arterial tonometry. Sleep; 25:543–549.

N Pitson DJ, et al. (1998). Value of beat-to-beat blood pressure changes, detected bypulse transit time, in the management ofthe obstructive sleep apnoea/hypopnoeasyndrome. Eur Respir J; 12: 685–692.

N Roche F, et al. (1999). Reduced cardiacsympathetic autonomic tone after long-term nasal continuous positive airwaypressure in obstructive sleep apnoeasyndrome. Clin Physiol; 19: 127–134.

N Smith RP, et al. (1999). Pulse transit time:an appraisal of potential clinical applica-tions. Thorax; 54: 452–457.

N Smith RP, et al. (1998). Obstructive sleepapnoea and the autonomic nervoussystem. Sleep Med Rev; 2: 69–92.

N Somers VK, et al. (1995). Sympatheticneural mechanisms in obstructive sleepapnea. J Clin Invest; 96: 1897–1904.

N Somers VK, et al. (1993). Sympathetic-nerve activity during sleep in normalsubjects. N Engl J Med; 328: 303–307.

N Spiegel K, et al. (1999). Impact of sleepdebt on metabolic and endocrine func-tion. Lancet; 354: 1435–1439.

N Spiegel K, et al. (2004). Sleep curtail-ment in healthy young men is associ-ated with decreased leptin levels,elevated ghrelin levels, and increasedhunger and appetite. Ann Intern Med;141: 846–850.

N Stein PK, et al. (2012). Heart ratevariability, sleep and sleep disorders.Sleep Med Rev; 16: 47–66.

N Stewart AG, et al. (1994). Autonomicnerve dysfunction in COPD as assessedby the acetylcholine sweat-spot test. EurRespir J; 7: 1090–1095.

N Tamisier R, et al. (2009). A new model ofchronic intermittent hypoxia in humans:effect on ventilation, sleep, and bloodpressure. J Appl Physiol; 107: 17–24.

N Tamisier R, et al. (2011). 14 nights ofintermittent hypoxia elevate daytimeblood pressure and sympathetic acti-vity in healthy humans. Eur Respir J; 37:119–128.

N Veale D, et al. (1992). Autonomic stresstests in obstructive sleep apnea syn-drome and snoring. Sleep; 15: 505–513.

N Veale D, et al. (1996). Abnormal auto-nomic stress responses in obstructivesleep apnoea are reversed by nasalcontinuous positive airway pressure. EurRespir J; 9: 2122–2126.

N Waradekar NV, et al. (1996). Influence oftreatment on muscle sympathetic nerveactivity in sleep apnea. Am J Respir CritCare Med; 153: 1333–1338.

N Zou D, et al. (2006). Validation a portablemonitoring device for sleep apnea diag-nosis in a population based cohort usingsynchronized home polysomnography.Sleep; 29: 367–374.


Definitions of sleep disorderedbreathing

Johan Verbraecken

Sleep disordered breathing (SDB) is a termused to describe a spectrum of respiratorydisturbances that occur during sleep. TheInternational Classification of SleepDisorders (ICSD-2) has defined the threemajor categories of SDB: obstructive sleepapnoea syndrome (OSAS), central sleepapnoea syndrome (CSAS) and sleep-relatedalveolar hypoventilation. The fundamentaldifference between the first two majorcategories is the pathophysiologicalmechanism that causes the respiratorydisturbance. In OSAS, the upper airwayobstruction is most often caused byabnormal anatomy and/or abnormal controlof the muscles that maintain the patency ofthe upper airway. In CSAS, dysfunction ofventilatory control in the central neurons isinvolved, resulting in loss of ventilatoryeffort (fig. 1). An obstructive apnoea/

hypopnoea can be defined as an event thatlasts for o10 s and is characterised by a

transient reduction in (hypopnoea), orcomplete cessation of (apnoea) breathing.According to the American Academy ofSleep Medicine (AASM) criteria from 1999, ahypopnoea can be defined as a decreasefrom baseline in the amplitude of a validmeasure of breathing during sleep thateither reaches .50% with an oxygendesaturation of 3% or an arousal, oralternatively a 30% reduction with 4%oxygen desaturation. Central sleep apnoea(CSA) refers to the cessation of ventilationlasting for o10 s (in adults) due to transientloss of neural output to the respiratorymuscles. The degree of severity is defined onthe basis of the number of apnoeas andhypopnoeas occurring during 1 h of sleep(this is the apnoea–hypopnoea index (AHI))and the severity of daytime symptoms.According to ICSD-2, the criteria for thediagnosis of a clinically significantobstructive sleep apnoea–hypopnoeasyndrome are: presence of criteria A, B andD, or C and D (see table 1). Based on thesecriteria, sleep apnoea occurs in 4% of menand 2% of women aged 30–60 yrs. Thedefinition of OSAS using two components,daytime symptoms and breathing patterndisturbances during sleep, indicates thatthere are also subjects who present withsleep apnoea without symptoms. Thesecases are referred to as obstructive sleepapnoea (OSA) and have an even higherprevalence, recently estimated to be 20% inthe male population.

OSAS can be subdivided into adult andpaediatric types, since the diagnostic criteriaand clinical presentation for abnormalbreathing during sleep are different foradults and children. Obstructive breathing

Key points

N OSAS is characterised by recurrentepisodes of partial or complete upperairway collapse during sleep.

N UARS is no longer considered as anindependent disease, but as part ofOSAS.

N Minimal diagnostic criteria exist forOSAS and CSAS.

N OHS is also part of the spectrum ofsleep disordered breathing.

N Hypoventilation syndromes can occurindependent of obesity.


events may include apnoeas, hypopnoeas,or respiratory effort-related arousals(RERAs). A RERA can be defined as asequence of breaths characterised byincreasing respiratory effort leading toarousal from sleep, but not meeting thecriteria for apnoea or hypopnoea. Moreover,these events present with a pattern ofprogressively more negative oesophagealpressure, terminated by a sudden change inpressure to a less negative level and anarousal. Oesophageal pressure isrecommended as the measurement methodof choice (AASM 2007), but the flattening ofthe flow curve obtained by nasal pressure isexplicitly mentioned, together with inductionplethysmography, as a feasible alternative.In reality, nasal pressure is the method ofchoice for more than 90% of sleeplaboratories. These events last 10 s or more.Upper airway resistance syndrome (UARS)

is characterised by increased upper airwayresistance, followed by frequent arousals,resulting in daytime sleepiness. It has thefollowing essential PSG features: absence ofobstructive sleep apnoeas, an AHI,5 events?h-1, and a lack of significantoxygen desaturation, which differ from thelaboratory findings of OSAS. At present, theterm UARS is no longer used as anindependent disease, but is subsumedunder the diagnosis of OSAS because itspathophysiology does not significantly differfrom that of OSAS.

The diagnosis of CSAS is made by criteriarecommended by the ICSD-2 manual.Patients with primary CSA present with: a)excessive daytime sleepiness; or b) frequentnocturnal arousals and awakenings duringsleep or insomnia complaints; or c)awakening short of breath, combined with 5

a)

Effort Oesophageal pressure (cmH2O)

Effort Abdomen

Effort Ribcage

Airflow

EEG Arousal

10 s

Sa,O2

Central apnoea

10075

-60-40-20

0

50

b)


Effort Abdomen

Effort Ribcage

Airflow

EEGArousal

10 s

Sa,O2

Obstructive hypopnoea

10075

-60-40-20

0

50

d)


Effort Abdomen

Effort Ribcage

Airflow

EEGArousal

10 s

Sa,O2

Upper airway resistance

10075

-60-40-20

0

50

c)


Effort Abdomen

Effort Ribcage

Airflow

EEGArousal

10 s

Sa,O2

Obstructive apnoea

10075

-60-40-20

0

50

Figure 1. Different presentations of apnoeas and hypopnoea. a) Central event; b–d) obstructive events.


or more central apnoeas per hour of sleep.The disorder should not be better explainedby another current sleep disorder, medicalor neurological disorder, medication use, orsubstance use disorder. Many patients withCSAS have mild hypocapnia ornormocapnia, but rarely hypercapnia andhypoventilation are also observed. Aperiodic pattern of waxing and waning ofventilation with periods of hyperventilationalternating with central apnoea/hypopnoeais termed Cheyne–Stokes respiration (CSR).According to the ICSD-2 manual CSR can beconsidered if: a) polysomnography shows atleast 10 central apnoeas and hypopnoeasper hour of sleep in which the hypopnoeahas a crescendo–decrescendo pattern oftidal volume accompanied by frequentarousals from sleep and derangement ofsleep structure; b) the breathing disorderoccurs in association with a serious medicalillness, such as heart failure, stroke, or renalfailure; c) the disorder is not betterexplained by another current sleep disorder,medication use or substance use disorder.

Although symptoms are not mandatory tomake this diagnosis, patients often reportexcessive daytime sleepiness, frequentarousals and awakenings during sleep,insomnia complaints, or awakening shortof breath.

A last group in the spectrum of SDB is calledsleep-related hypoventilation/hypoxaemicsyndrome. Sleep-induced hypoventilation ischaracterised by elevated PaCO2 of.45 mmHg while asleep, ordisproportionately increased relative tolevels during wakefulness. This groupcomprises idiopathic sleep-relatednonobstructive alveolar hypoventilation,congenital central alveolar hypoventilationsyndrome, and sleep-related hypoventilationdue to a medical condition (pulmonaryparenchymal or vascular pathology, lowerairway obstruction, neuromuscular andchest wall disorders). Obesityhypoventilation syndrome (OHS) isprobably the most common clinicalpresentation of this syndrome. This disorderis defined as the association of obesity(body mass index .30 kg?m-2) and SDBwith daytime hypersomnolence andhypercapnia (PaCO2 .45 mmHg) in theabsence of any other respiratory disease.There is however no commonly accepteddefinition for OHS.

Different definitions have been published bythe AASM for the different entities of SDB(ICSD-2 versus AASM Scoring Manual),which may complicate the understanding ofthe problem. In this article, the ICSD-2criteria have been used. The Task Force of

Table 1. Criteria for the diagnosis of a clinically significant obstructive sleep apnoea–hypopnoea syndrome (ICSD-2criteria).

A. At least one of the following applies:i. The patient complains of unintentional sleep episodes during wakefulness, daytime

sleepiness, unrefreshing sleep, fatigue or insomnia.ii. The patient wakes with breath holding, gasping or chokingiii. The bed partner reports loud snoring, breathing interruptions or both during the patient’s

sleep

B. PSG recording shows the following:i. o5 scoreable respiratory events (i.e. apnoeas, hypopnoeas, or RERAs) per hour of sleepii. Evidence of respiratory effort during all or a portion of each respiratory event

OR

C. PSG recording shows the following:i. o15 scoreable respiratory events (i.e. apnoeas, hypopnoeas, or RERAs) per hour of sleepii. Evidence of respiratory effort during all or a portion of each respiratory event

D. The disorder is not better explained by another current sleep disorder, medical or neurologicaldisorder, medication use, or substance use disorder


the AASM (1999) also states that there arecommon pathogenetic mechanisms forobstructive apnoea syndrome, centralapnoea syndrome, Cheyne–Stokes breathingand sleep hypoventilation syndrome. It ismore preferable to discuss each of theseseparately, although they could be placedunder the common denominator of ‘‘sleepdisordered breathing syndrome’’.

Further reading

N American Academy of Sleep Medicine.The AASM Manual for the Scoring ofSleep and Associated Events: Rules,Terminology and Technical Specifica-tions. Westchester, American Academyof Sleep Medicine, 2007.

N The International Classification of SleepDisorders. Second edition. Diagnostic andCoding Manual. Westchester, AmericanAcademy of Sleep Medicine, 2005.

N Task Force of the American Academy ofSleep Medicine. (1999). Sleep-related breath-ing disorders in adults: recommendations

for syndrome definition and measure-ment techniques in clinical research.Sleep; 22: 667–689.

N Banno K, et al. (2007). Sleep apnea:clinical investigations in humans. SleepMed; 8: 400–426.

N Bloch KE, et al. Central sleep apnoea.In: Palange P, Simonds A, eds. ERSHandbook of Respiratory Medicine.Sheffield, European Respiratory Society,2010; pp. 410-413.

N De Backer W. Obstructive sleep apnoea/hypopnoea syndrome. In: Palange P,Simonds A, eds. ERS Handbook ofRespiratory Medicine. Sheffield, EuropeanRespiratory Society, 2010; pp. 410-413.

N Muir JF. Hypoventilation syndromes.In: Palange P, Simonds A, eds. ERSHandbook of Respiratory Medicine.Sheffield, European Respiratory Society,2010; pp. 410-413.

N Strollo PJ, Jr, et al. (1996). Obstructivesleep apnea. N Engl J Med; 334: 99–104.

N Young T, et al. (1993). The occurrence ofsleep-disordered breathing among middle-aged adults. N Eng J Med; 328: 1230–1235.


Obstructive sleep apnoeasyndrome

Johan Verbraecken

OSAS is highly prevalent and represents anincreasing part of clinical respiratorypractice in developed countries. OSAS isnow recognised as one of the most commonchronic respiratory disorders in adults, withonly asthma and possibly COPD having ahigher prevalence. Questions about risks,diagnosis and treatment options are ofimportance to both the clinician and tohealthcare policy makers. As the medicalcommunity and the general public havebecome more aware of the relationshipsamong snoring, excessive daytimesleepiness, cardiovascular disease andOSAS, physicians are seeing an increasingnumber of patients with these problems.Excessive daytime sleepiness is one of themajor symptoms. Untreated OSA hasdangerous health consequences. SevereOSA is associated with an increased risk of

death from any cause in middle-aged adults,especially men. OSA is closely related toincreases in body weight and, as a result,there is a tendency to produce or worsenupper airway obstruction during sleep.Upper airway collapse is multifactorial.Upper airway size, which is significantlyaffected by obesity and craniofacialcharacteristics, is known to play a majorrole. The pathophysiological basis of OSASis also partly genetic in origin, to whichacquired factors, such as obesity, alsocontribute. Increased upper airwaycollapsibility and impaired neuromuscularresponse both contribute, as well aspharyngeal neuropathy, ventilatoryinstability and described more recently,possible fluid shift towards the pharynx. It isobvious that OSAS has different phenotypesand that there are substantial sex differencesand important ageing effects. Moreover, thecondition carries significant morbidity and isassociated with an increased risk ofhypertension, arrhythmias, myocardialinfarction, stroke, metabolic andneurobehavioural consequences.

Further reading

N Decramer M, et al. (2011). Chapter 1.Prevention. In: European Respiratory Road-map. Sheffield, European RespiratorySociety; pp. 9–26.

N Levy P, et al. (2011). Sleep apnoea syn-drome in 2011: current concepts and futuredirections. Eur Respir Rev; 20: 134–146.

N McNicholas WT, et al. (2010). Sleepapnoea: Introduction. Eur Respir Monogr;50: vii–ix.

Key points

N OSAS is now recognised as one of themost common chronic respiratorydisorders in adults.

N Physicians are seeing an increasingnumber of patients with OSAS.

N Untreated OSA has dangerous healthconsequences.

N Upper airway collapse ismultifactorial.

N There are substantial sex differencesand important ageing effects.


Obstructive sleep apnoea:epidemiology and risk factors

Johan Verbraecken

The landmark report which clarified theprevalence of OSAS in the general middle-aged population showed that OSAS affects,4% of males and 2% of females, but theprevalence is increasing because of therecent obesity epidemic. Common riskfactors of OSAS are obesity, sex, ageing, andcomorbidities, such as enlarged tonsils andadenoids and craniofacial abnormalities.

Obesity

Approximately 80% of OSAS patients areobese and obesity is a recognised risk factorfor OSAS. A very close relationship has beenobserved between body weight change andAHI: a 10% weight gain has been shown topredict an approximate 32% increase in AHI;a 10% weight loss predicts a 26% decreasein the AHI; and a 10% weight gain predicts asix-fold increase in the odds of developingmoderate-to-severe OSA. Obesity causesairway narrowing as a result of excess fat inthe (peri and para) pharyngeal tissues.

Age

Several papers have shown a higherprevalence of OSAS in the elderly. The SleepHeart Health Study showed that 25% ofmales and 11% of females in the 40–98-yr-old age group had an AHI .15 events?h-1.However, daytime symptoms may be lesscommon with increasing age. The influenceof male sex and BMI on OSAS tends to wanewith age and the overall prevalence of OSASseems to plateau after age 65 yrs. Incontrast, the age distribution of patientsfirst diagnosed with OSAS generally peaks atage 50 yrs.

Sex

Epidemiological studies have shown thatOSAS is much more prevalent in males. Areferral bias and sex differences in clinicalpresentation may have resulted in moremales than females being diagnosed. Maleswith OSAS are more likely to havesymptoms of snoring, witnessed apnoeas orsleepiness. Females with OSAS have moresymptoms of depression or morningheadache. Recent evidence providesrationale for clinicians to consider thelikelihood of OSAS in obese females with ahistory of insomnia, depression orhypothyroidism.

Other comorbidities

A number of other factors may contribute to,and can be considered as, risk factors forOSAS. These are summarised in table 1.

Key points

N OSAS is highly prevalent with anestimated prevalence of at least 4% inmales and 2% in females.

N Obesity has a very close relationshipwith AHI.

N The influence of male sex and BMI onOSAS tends to wane with age.


Table 1. Factors promoting upper airway (UA) collapse.

Abnormal anatomy of the UA

Skeletal factors

Maxillary and/or mandibular hypoplasia or retroposition

Hyoid position (inferior displacement)

Soft tissue factors

Increased volume of soft tissues

Adenotonsillar hypertrophy

Macroglossia

Thickened lateral pharyngeal walls

Increased fat deposition

Pharyngeal inflammation and/or oedema

Increased vascular volume

Increased muscle volume

Pharyngeal muscle factors

Insufficient reflex activation of UA dilator muscles

Impaired strength and endurance of pharyngeal dilators

Pharyngeal compliance

Increased UA collapsibility

Sensory function

Impaired pharyngeal dilator reflexes

Impaired mechanoreceptor sensitivity

Lung volume dependence of UA cross-sectional area

Increased below functional residual capacity

Ventilatory control system factors

Unstable ventilatory control

Increased ventilatory responses and loop gain

Sex factors

Male influences

Centripetal pattern of obesity

Absence of progesterone

Presence of testosterone

Weight

Obesity causing peripharyngeal fat accumulation

Reproduced from Verbraecken et al. (2009), with permission from the publisher.


Further reading

N Banno K, et al. (2007). Sleep apnea:clinical investigations in humans. SleepMed; 8: 400–426.

N Peppard PE, et al. (2000). Longitudinal studyof moderate weight change and sleep-disordered breathing. JAMA; 284: 3015–3021.

N Verbraecken J, et al. (2009). Upper airwaymechanics. Respiration; 78: 121–133.

N Young T. (1996). Sleep-disordered breath-ing in older adults: is it a condition distinct

from that in middle-aged adults? Sleep;19: 529–530.

N Young T, et al. (2002). Sleep Heart HealthStudy Research Group. Predictors ofsleep-disordered breathing in community-dwelling adults: the sleep Heart HealthStudy. Arch Intern Med; 162: 893–900.

N Young T, et al. (1996). The gender bias insleep apnea diagnosis. Are women missedbecause they have different symptoms?Arch Intern Med; 156: 2445–2451.


Obstructive sleep apnoea:pathophysiology

Johan Verbraecken and Wilfried De Backer

The narrowing or occlusion of the upperairway (UA) during sleep has beenattributed to several factors. An abnormalanatomy of the UA, pathological andinsufficient reflex activation of UA dilatormuscles and increased collapsibility of thepassive UA have all been demonstrated tooccur and contribute to the UA collapse. UAdysfunction in OSA is a progressiveevolution, starting with snoring and flowlimitation, that can lead to mechanicaltraumata that progressively injure the UAtissues. Recently, it has been shown that UAcollapse occurs during the terminal phase ofthe expiration preceding the apnoea.

Craniofacial and upper airway morphology

Several studies have correlated skeletaldimensions with the prevalence and severityof OSAS, with anatomical variations that

decrease internal skull volume being foundto correlate with OSAS. Reported differencesare a retroposition of the mandible or themaxilla, micrognathia and differences inhyoid bone position. These findings areconfirmed by the high prevalence of OSA inpatients with craniofacial disorders.

In addition to narrowing of the airway by thelateral pharyngeal walls caused by oedemaand fat deposition, enlargement of thetonsils, uvula or tongue can contribute tothe occlusion of the UA during sleep. Thelatter are not the classical risk factor forOSAS in adults (although they are inchildren), but it can be expected thatincreased soft tissue sizes will decreaseairway lumen dimensions and increase therisk of OSAS.

Moreover, recent evidence hasdemonstrated that a fluid shift from thelower limbs to the upper airways takes placein OSA at night, contributing to UA oedema.With detailed MRI techniques, it has beendemonstrated that the volume of the tongueand the lateral walls are an independent riskfactor for sleep apnoea, while the variationsin the velopharyngeal area during therespiratory cycle are greater in apnoeicpatients than in controls. However,structural narrowing of the UA at onespecific location is unlikely to be a majorcause of OSA. Studies have shown thatthe UA collapse is not restricted to oneplace, but is a dynamic phenomenonstarting at a particular level and spreadingcaudally. UA obstruction involves more thanone specific site of the upper airway in themajority of patients.

Key points

N OSAH patients are characterised by acompromised upper airway anatomy,involving more than one specific sitein most patients, with abnormalcollapsibility, leading to a passivenarrowing at the end of expiration.

N OSAH patients have a sleep-inducedloss of compensatory mechanisms(high upper airway muscle activity,negative pressure upper airway reflex).

N Ventilatory control (loop gain) andsleep instability (arousal threshold)contribute to upper airway instability.


Function and dysfunction of upper airwaymuscles

Upper airway dilator muscle activation inOSAS patients is quite adequate and evenintensified during wakefulness. This ismainly the result of reflex activationprovoked by negative intrapharyngealpressures that are more pronounced inOSAS patients due to the smaller airway.Several other mechanisms are involved,however, including a decrease in lungvolume and a high neuromodulator rate(acetylcholine, adenosine, ATP, nitric oxide,norepinephrine, orexin, serotonin,substance P, thyrotropin-releasing hormoneand vasopressin). This reflex is quite activeduring wakefulness but declines significantlyduring sleep. Particularly during NREMsleep, this negative pressure reflex issubstantially diminished or lost completely.Studies of genioglossal muscle activitysuggest that patients with OSAS have amuch greater reduction in the genioglossalEMG than normal subjects. Due to the lossof this compensatory reflex activation of theUA dilator muscles, the UA of OSASpatients may narrow significantly or collapseduring inspiration when asleep, with thedevelopment of flow limitation. Thesemechanisms have been described as the‘‘balance of forces’’ model. It is likely that acombination of UA mechanical loads anddisturbances in neuromuscular mechanismsaccount for the pathogenesis of OSAS. Forexample, in a group of OSA subjects, one-third of the variability in OSAS severity wasascribed to mechanical loads, suggestingthat neuromuscular mechanisms accountedfor the remaining two-thirds. However, thereis increasing evidence that the collapse ofthe UA occurs during expiration.

Furthermore, it has been convincinglyshown that the UA behaves like a Starlingresistor, making the collapse independent ofthe suction force brought about by thediaphragm, but rather dependent on thebalance between the UA pressure and thetissue pressure at the collapsible site. Theairway remains patent, regardless of theexcessive pressure applied, as long as thecritical pressure of positive end-expiratory

pressure (Pcrit) remains low relative to thepressure upstream of the collapsiblesegment (Pu). Closure of the upper airwayoccurs when Pu falls below the surroundingtissue pressure (Pcrit). In the model of theStarling resistor, maximal flow (V9max)becomes a function of the pressure gradientand the resistance in the segment upstreamto the collapsible segment (Ru), such thatV9max5(Pu-Pcrit)/Ru. The collapse of the UAthen finally occurs during expiration when,due to the absence of dilator muscle, Pcrit

exceeds the upstream pressures. Prolongedexpiratory time, as occurs during centralapnoeas, therefore predisposes to collapse.

Central respiratory control mechanisms

Both arousal threshold and loop gain (as amarker of the degree of respiratory controlsystem instability) may contribute to apnoeapresence and severity. An obstructed UA canreopen due to a reflex, without arousal, ifchemical drive is allowed to reach athreshold, but this is often pre-empted by alow arousal threshold. The relation betweenchemical and arousal thresholds, as well asthe lung-to-carotid circulation time and therate of rise of chemical drive during theobstructive event determine the magnitudeof ventilatory overshoot at the end of anevent and, by extension, whether initialobstructive events will be followed by stablebreathing, slow evolving hypopnoeas withoccasional arousals or repetitive events.OSAS patients may have a higher chemicaldrive (hypercapnic ventilatory response) thatcan contribute to the increased loop gain.Moreover, cyclic changes in arterial CO2

around the CO2 threshold for activation ofUA motor neuron activity could lead to animbalance of forces acting on the pharyngealairway and favour closure.

Further reading

N Ciscar MA, et al. (2001). Magneticresonance imaging of the pharynx inOSA patients and healthy subjects. EurRespir J; 17: 79–86.

N Dempsey JA, et al. (2010). Pathophy-siology of sleep apnea. Physiol Rev; 90:47–112.


N Fogel RB, et al. (2001). Genioglossalactivation in patients with obstruc-tive sleep apnea versus control sub-jects. Mechanisms of muscle control.Am J Respir Crit Care Med; 164:2025–2030.

N Mezzanotte WS, et al. (1992). Wakinggenioglossal electromyogram in sleepapnea patients versus normal controls (aneuromuscular compensatory mechan-ism). J Clin. Invest; 89: 1571–1579.

N Schellenberg JB, et al. (2000). Physicalfindings and the risk for obstructive sleepapnea. The importance of oropharyngealstructures. Am J Respir Crit Care Med; 162:740–748.

N Schwab RJ, et al. (2003). Identification ofupper airway anatomic risk factors forobstructive sleep apnea with volumetricmagnetic resonance imaging. Am J RespirCrit Care Med; 168: 522–530.

N Verbraecken J, et al. (2009). Upper airwaymechanics. Respiration; 78: 121–133.

N Vos WG, et al. (2010). Correlationbetween the severity of sleep apnea andupper airway morphology in pediatric andadult patients. Curr Opin Allergy ClinImmunol; 10: 26–33.

N Younes M. (2003). Contributions of upperairway mechanics and control mechan-isms to severity of obstructive apnea. Am JRespir Crit Care Med; 168: 645–658.


Obstructive sleep apnoea:clinical aspects

Stefan Andreas

OSA is a disease of developed countries,where chronic excess of calorie intake canlead to accumulation of body fat. Clearly,obesity is the main risk factor for OSA. Forthe consequences of OSA discussed in‘‘Obstructive sleep apnoea: consequences’’,it should be kept in mind that obesity itselfincreases the risk of cardiovascular diseaseand is a cause of daytime sleepiness. Forpneumologists, with their clear focus on theburden of cigarette smoking, it should alsobe mentioned that smoking not onlyincreases the risk of cardiovascular diseasebut is also a risk factor for snoring and OSAitself, although the latter association has notbeen firmly investigated (Franklin et al.,2004). Conversely, physical activity andweight reduction ameliorate not onlycardiovascular disease, but also OSA(Peppard et al., 2004). The growing obesityepidemic should remind us that preventionis key in most common diseases andcertainly is in OSA.

As OSA can be considered a risk factor forcardiovascular disease, a historyencompassing arterial hypertension, other

cardiovascular risk factors, heart diseaseand stroke is mandatory in the assessmentof OSA. Similarly, respiratory diseases suchas COPD or restrictive lung disease shouldbe searched for, as they aggravate OSA andare amenable to treatment.

Symptoms

Symptoms of OSA can be divided intosymptoms recognised by the bed partnerand symptoms experienced by the patientthemselves. Snoring with periods of silenceand choking can make the diagnosis, ifwitnessed and reported by the bed partner.Typically, these symptoms are more obviousin the supine position or after alcoholconsumption. Upper airway dysfunction inOSA has a progressive evolution, startingwith snoring, and mechanical trauma due tosnoring and flow limitation progressivelyinjure the upper airway tissues. The patientthemselves will mostly experience the morediscreet and less specific daytimesymptoms, such as sleepiness, falling asleepinadvertently during the day, impairedvigilance and mood disturbance, mainlydepression. A dramatic consequence ofdaytime sleepiness is traffic accidents,which occur about four times more often inOSA than in controls (Tregear et al., 2010).

Questionnaires are helpful in daily practiceto supplement oral history. The BerlinQuestionnaire is only one example, with thefollowing questions being predictive of OSA(Netzer et al., 1999).

N Has your weight changed?N Do you snore?N Snoring loudness.N Snoring frequency.

Key points

N Obesity is the main risk factor forOSA.

N Questionnaires are helpful in dailypractice to supplement oral history.

N Every patient suspected of OSAshould undergo an examination of theupper airways.


N How often have your breathing pausesbeen noticed?

N Are you tired during wake time?N Have you ever fallen asleep while driving?N Do you have high blood pressure?

Since daytime sleepiness is important, aspecific questionnaire, such as the ESS,should be used. The ESS comprises onlyeight items on a four-point scale.Questionnaires and clinical impressionalone, however, have limited sensitivity andspecificity, and the diagnosis is usuallyobtained by polygraphy or PSG. Other moreobjective tests to evaluate daytimesleepiness, like the MSLT or the MWT, arecovered in detail in ‘‘Assessment of daytimesleepiness’’, later in this book. Althoughsleepiness is difficult to quantify, it isimportant to decide whether CPAPtreatment should be instituted to treatdaytime sleepiness. Furthermore, the effectof CPAP on objective measures such asblood pressure clearly depends on thedegree of daytime sleepiness (see‘‘Obstructive sleep apnoea: treatment’’).

Generally, it should be kept in mind thatmost research has been performed in 50–60-yr-old, clearly overweight male patientswith moderate-to-severe OSA. Thus, thesymptoms and neurobehavioural and

cardiovascular sequelae of OSA discussedlater mainly apply to this cohort. The femaleOSA phenotype is less archetypical: malebed partners are less likely than females toreport snoring and apnoeas to a medicalprofessional. Furthermore, symptomsexperienced by female OSA patients are lessspecific and include insomnia, headacheand mood disturbance. Thus, female OSApatients are less likely to be diagnosed andtreated than males. Similarly, the elderly(.70 yrs of age) will present with lessspecific history and daytime symptoms(Martinez-Garcia et al., 2009).Unsurprisingly, comorbidities have a largeimpact on daytime sleepiness and quality oflife in the elderly.

Physical examination

On physical examination, heart rate, bloodpressure and smoking status must bedetermined. Since obesity is the maintreatable cause of OSA (Foster et al., 2009),measures of obesity, such as weight, height,BMI or neck circumference, must beevaluated. For example, a neckcircumference .48 cm indicates asubstantial risk of having OSA. There aresimple clinical prediction models that allowestimation of the likelihood of OSA usingthese simple measures in addition to sex

Table 1. Symptoms in OSA.

Night-time symptoms Daytime symptoms

N Witnessed apnoeas

N Snoring

N Nocturnal choking

N Disturbed unrefreshing sleep

N Thirst during the night

N Nocturnal diuresis, enuresis

N Nocturnal sweating

N Impotence

N Excessive salivation

N Gastro-oesophageal reflux

N Excessive daytime sleepiness

N Fatigue

N Morning dry mouth

N Morning headache

N Difficulty concentrating

N Irritability, mood changes

Reproduced from Tkacova et al. (2010).


and age. However, while these predictionmodels have reasonable sensitivities, theyhave relatively low specificities. Thus, adefinite diagnosis should be reached by anobjective sleep study.

As in other diseases, a general examinationis useful to ensure pulmonary,cardiovascular, neurological, skeletal andfacial abnormalities, hypothyroidism, orstorage diseases are not overlooked.However, it is more likely to find sequelae ofoverweight and OSA, such as heart failure.

Every patient suspected of OSA shouldundergo an examination of the upperairways. This does not necessarily have to beperformed by an ENT specialist. However,the dentures, the pharyngeal arch and thetonsils must be visualised. Albeit veryuncommon, abnormalities such asmacroglossia (in acromegaly) or tumoursshould be recognised. The degree ofpharyngeal narrowing can be quantified, e.g.by the Malapati score. Strikingly, impairednasal breathing can favour OSA and, moreimportantly, can make treatment of OSA bya nasal mask impossible. Thus, nasalbreathing should be checked, typically forthe right and left nostril separately.

Craniofacial conditions, such asretrognathia, mandibular hypoplasia,micrognathia and inferior displacement ofthe hyoid, can contribute to upper airwayocclusion. Individuals with rare hereditarysyndromes, such as Alpert, Down, Hunter,

Prader–Willi and Pierre Robin syndromes,are also more likely to develop OSA.

Further reading

N Foster GD, et al. (2009). A randomizedstudy on the effect of weight loss on obstruc-tive sleep apnea among obese patientswith type 2 diabetes: the Sleep AHEADstudy. Arch Int Med; 169: 1619–1626.

N Franklin KA, et al. (2004). The influenceof active and passive smoking on habitualsnoring. Am J Respir Crit Care Med; 170:799–803.

N Martinez-Garcia MA, et al. (2009).Obstructive sleep apnea has little impacton quality of life in the elderly. Sleep Med;10: 104–111.

N McNicholas WT, et al., eds. (2010). SleepApnoea. Eur Respir Monogr 50.

N Netzer NC, et al. (1999). Using the BerlinQuestionnaire to identify patients at riskfor the sleep apnea syndrome. Ann IntMed; 131: 485–491.

N Peppard PE, et al. (2004). Exercise andsleep-disordered breathing: an associa-tion independent of body habitus. Sleep;27: 480–484.

N Tregear S, et al. (2010). Continuouspositive airway pressure reduces risk ofmotor vehicle crash among drivers withobstructive sleep apnea: systematicreview and meta-analysis. Sleep; 33:1373–1380.

N Tkacova R, et al. (2010). Clinical presen-tations of OSA in adults. Eur RespirMonogr; 50: 86–103.


Obstructive sleep apnoea:consequences

Stefan Andreas

Neurobehavioural (sleepiness, cognitiveand psychiatric)

The concept of ‘sleeping on a problem’ isold and well known. Indeed, recent evidenceshows that memory reprocessing duringsleep is important in consolidating andforming our declarative, procedural andcognitive memories, and brain processes(Stickgold, 2005). Thus, it comes as nosurprise that OSA, with consecutive sleepdisturbance, has detrimental effects oncerebral function. This is most striking inchildren and adolescents, where the brain isstill developing. Studies using positronemission tomography (PET) or functionalMRI indicate sleep loss is the primary causeof neurocognitive deficits (mainly a basalslowing of information processing), more sothan hypoxaemia (Verstraeten, 2007).

Whether there are not only functional, butalso anatomical, sequelae of OSA is moredifficult to establish, since the pervasiveeffects of decreased alertness on highercognitive functioning mimic cerebraldamage due to hypoxia. Moreover,measures of daytime sleepiness are notassociated with objective measures ofapnoea burden, such as the AHI,hypoxaemia or sleep parameters. However,

since OSA causes arterial hypertension andis a cardiovascular (CV) risk factor (seelater), a high apnoea burden can causecerebrovascular disease, with stroke beingthe most obvious.

To better understand the effects of OSA onneurobehavioural aspects, treatment studiesare of importance. A recent meta-analysisconcluded that CPAP is clearly effective inreducing objective and subjective daytimesleepiness. Furthermore, some cognitivefunctions, particularly executive functionslinked to daytime sleepiness, were improvedwith CPAP. Depression occurs in 40% ofOSA patients, but this is not related to OSAseverity and CPAP has no consistent effecton depression. Further well-designed, long-term studies are needed to better delineatethe negative effects of OSA on complexcerebral functions, such as cognition andmood. At least for younger patients withgood treatment compliance, CPAP willnormalise all aspects of brain function.

OSA is common in patients with stroke.CPAP treatment is inherently difficult inpatients following stroke and two randomisedtrials failed to show meaningful improvementof physical or cognitive function in thosepatients (Bradley et al., 2009).

Cardiovascular

CV disease (CVD) is the main cause ofmorbidity and mortality in the ageingwestern population. Older people sufferparticularly from CVD. CVD encompassescoronary artery disease (CAD), HF andcerebrovascular disease/stroke. Althoughmortality due to CAD has decreased slightlyin the past decade, CAD remains the mostcommon cause of death, with an incidence

Key points

N OSA has detrimental effects oncerebral function.

N OSA has multiple negative effects onthe CV system.

N OSA affects the metabolic syndrome.


of ,400 myocardial infarctions per 100,000person-yrs between the ages of 36 and64 yrs. The prevalence of HF is ,2%, risingto .10% in subjects .70 yrs of age. Themean age of HF patients in the communityis 75 yrs. About half of the patientspresenting with the clinical syndrome of HFhave a normal left ventricular ejectionfraction. These cases are defined as HF withpreserved ejection fraction (HFprEF) andleft ventricular diastolic dysfunction isconsidered to be a common underlyingpathology. HFprEF typically occurs in theelderly and is associated with classical riskfactors, such as arterial hypertension,diabetes mellitus and atrial fibrillation. Theprognosis of HFprEF patients, oncehospitalised for HF, is similarly poor to thatin systolic HF, but has, in contrast tosystolic heart failure, not improved duringthe past few decades. Similar to diastolicHF, cerebrovascular disease and stroke aremore common in the elderly. The main riskfactor for stroke is arterial hypertension.

OSA occurs frequently in patients with CAD,HF and stroke. Several studies on patientswith CAD who are slightly overweight yieldedan incidence of OSA between 30% and 50%(Luthje et al., 2008). A high incidence ofOSA has also been found in patients witharterial hypertension, stroke and HF(Bradley et al., 2009). OSA and CVD sharecommon risk factors, such as obesity, malesex, smoking and advanced age.

In recent decades, research has revealedmultiple negative consequences of OSA onthe CV system. The pathophysiologicalinteraction between OSA and CVD iscomplex, and comprises neural, humoral,mechanical, haemodynamic and circadianrhythm components. Oxidative stress andendothelial dysfunction are mainmechanisms leading to CVD in OSA.Another important effect of OSA is anincrease in sympathetic activity, whichpersists during the day. Well-controlledstudies have shown effects of OSA ondifferent aspects of early vasculardysfunction. These comprise vascularresponses to different stimuli, arterialstiffness, pulse wave velocity, carotid intima

media thickness, carotid plaques andalterations of the coronary arteries. All thesealterations are ultimately related to clinicallyovert CV disease in different vascular beds.

Nowadays, OSA is accepted as anindependent risk factor for arterialhypertension (Haentjens et al., 2007). Arterialhypertension is a risk factor for CAD as well asHF, especially HFprEF, and is a strong riskfactor for stroke. This association makes itdifficult to prove a causal relation betweenOSA and CVD independent of arterialhypertension in epidemiological studies.

Large epidemiological cross-sectional andlongitudinal studies have described anassociation between OSA and CVD morbidityand mortality, independently of confounderssuch as obesity. Furthermore, a dose–response relationship has been established.Independent of other risk factors, such asobesity and arterial hypertension, thelikelihood of dying from CV causes doubles inpatients with moderate/severe OSA. This hasbeen shown for a variety of major CV end-points, such as CV death, acute myocardialinfarction, stroke, occurrence of HF,hospitalisation for CVD, left ventriculardiastolic dysfunction, complex ventriculararrhythmias and sudden cardiac death.

The most convincing evidence for OSAhaving negative CV effects and being animportant CV risk factor stems fromtreatment trials: CPAP therapy markedlyreduces nocturnal sympathetic activation,thereby reducing heart rate, blunting bloodpressure surges, and improving vascularstiffness and left ventricular diastolicfunction. Furthermore, CPAP therapy has apositive effect on oxidative stress,endothelial dysfunction, leptin levels,interleukins, tumour necrosis factor (TNF)-a, C-reactive protein (CRP), plateletaggregability and fibrinogen levels in well-controlled studies. CPAP reduces not onlynocturnal but also diurnal blood pressure inOSA (Haentjens et al., 2007). These effectsare more pronounced in patients with goodtreatment adherence, high initial AHI,arterial hypertension and daytimesleepiness, and are also present in patientswith CVD, such as HF.


In conclusion, current data suggest thatOSA increases the risk of developing CVDand that its treatment has the potential todiminish such risk. This has beenrecognised in guidelines for the treatment ofarterial hypertension and HF.

Metabolic

The metabolic syndrome is tightly associatedwith obesity and its sequelae. According to a2009 joint statement, metabolic syndrome isdefined as the presence of three or more ofthe following: waist circumference o88 cmin females or o102 cm in males; high bloodpressure (BP) (systolic o130 and/ordiastolic o85 mmHg) or antihypertensivetreatment; high fasting blood glucose (o5.6 mmol?L-1) or antidiabetic treatment;high triglycerides (o1.7 mmol?L-1) or lipid-lowering treatment; reduced high-densitylipoprotein (HDL)–cholesterol(,1.3 mmol?L-1 in females or ,1 mmol?L-1 inmales) or lipid-lowering treatment. TheWorld Health Organization definition isslightly more complex and emphasises thedirect measurement of insulin resistance.Unsurprisingly, the metabolic syndrome is astrong CV risk factor.

White adipose tissue is a major endocrineand secretory organ, and plays a key role inthe metabolic syndrome. Adipose tissuereleases a large number of peptides, such asleptin and other factors, collectively termedadipocytokines. Harmful free fatty acids andangiotensin II are also released andcontribute to inflammation and negative CVeffects. In obesity, the secretory activity ofadipose tissue is increased and altered. Areason for the altered adipocytokine releaseis assumed to be relative hypoxia in clustersof adipocytes that become distant from thevasculature as adipocyte cell size increases.These negative effects are mostly caused byintra-abdominal or visceral adipose tissuerather than subcutaneous adipose tissue.

Sleep impacts on adipocyte activity and themetabolic syndrome. In healthy subjects,experimental sleep restriction caused insulinresistance, reduced leptin and increasedghrelin plasma concentrations, andincreased appetite (Levy et al., 2009). This is

explained by a pro-inflammatory staterelated to stress and neurohumoralactivation. In a broader context, the current,well-received ‘selfish brain’ theory holds thatchronic stress alters the way the brainsecures its own energy supply. Thereby,energy uptake from the environment ischronically increased in most individuals,ultimately leading to obesity. However, inepidemiological studies, a causalrelationship between sleep duration andobesity is difficult to prove and, indeed, isnot established at present.

The mechanisms through which OSA mayworsen metabolism are complex, andencompass sympathetic and neurohumoralactivation, glucose homeostasis,inflammation, and oxidative stress mainlydue to chronic intermittent hypoxia. Cross-sectional studies estimate a five-fold risk ofthe metabolic syndrome in OSA patientscompared with controls. Accordingly,markers of the metabolic syndrome, such asleptin, are increased and correlate with OSAseverity. However, due to the strongconfounding effect of obesity, causalitycannot be derived. A recent review (Levyet al., 2009) concluded that despite theabundance of cross-sectional evidence forthe link between OSA and abnormal glucosecontrol, further well-designed longitudinaland interventional studies are needed toprove causality.

As in hypertension or CVD, interventionstudies using CPAP therapy could help todisentangle causality. In OSA patientswithout type 2 diabetes mellitus and, thus,mostly without obesity, CPAP improvedglucose control and insulin sensitivity.However, in the typical overweight OSApatient, CPAP treatment does notsignificantly affect the metabolic status.Similarly, the effects of CPAP on themetabolic syndrome in type 2 diabeticpatients with OSA are controversial.Furthermore, there is no convincingevidence that visceral fat decreasesfollowing CPAP treatment. These studiesresemble findings in CVD, where the effectsof CPAP are most striking in young,otherwise healthy male OSA patients.


While discussing the effects of OSA on themetabolic syndrome, we should be aware ofthe fact that at least three large, well-controlled randomised trials showed aclinically meaningful reduction in bodyweight of ,20 kg with intensive counsellingand a low-energy diet. The reduction in bodyweight was accompanied by a clearimprovement of OSA in .60% of patients.Furthermore, all aspects of the metabolicsyndrome are strikingly improved by weightreduction. Long-term treatment studies arenow needed to validate weight loss as aprimary treatment strategy for OSA.

Further reading

N Bradley TD, et al. (2009). Obstructivesleep apnoea and its cardiovascularconsequences. Lancet; 373: 82–93.

N Haentjens P, et al. (2007). The impact ofcontinuous positive airway pressure onblood pressure in patients with obstruc-tive sleep apnea syndrome: evidencefrom a meta-analysis of placebo-con-trolled randomized trials. Arch Int Med;167: 757–764.

N Levy P, et al. (2009). Sleep, sleep-disordered breathing and metabolic con-sequences. Eur Respir J; 34: 243–260.

N Luthje L, et al. (2008). Obstructive sleepapnea and coronary artery disease. SleepMed Rev; 12: 19–31.


N Stickgold R. (2005). Sleep-dependentmemory consolidation. Nature; 437:1272–1278.

N Verstraeten E. (2007). Neurocognitiveeffects of obstructive sleep apnea syn-drome. Curr Neurol Neurosci Rep; 7(2):161–166.


Central sleep apnoea andCheyne–Stokes respiration

Winfried Randerath

Awareness of sleep-related breathingdisorders has grown rapidly in recent years.Sleep-related breathing disorders have beenacknowledged as independent risk factorsfor cardiovascular disorders. They aresignificantly associated with increasedmorbidity and mortality. Therefore, thediagnosis and treatment of OSA and CSA isof major interest in cardiac patients.

CSA is defined by recurrent cessations ofairflow and simultaneous reduction of thebreathing effort (American Academy ofSleep Medicine, 2005). In contrast to OSA,

ventilatory impulses generated by the brainstem are lacking in CSA. The typical findingsof paradoxical breathing and flattening ofthe inspiratory flow curve in OSA do notcharacterise CSA. CSA syndromes includeprimary CSA, CSR, CSA due to high-altitudeperiodic breathing and CSA caused by brainstem lesions or drugs and substances. CSR,or periodic breathing, is characterised byrecurrent apnoeas and hypopnoeas and acrescendo–decrescendo pattern of flow andeffort. The cycle length averages 60–90 s inCSR, while it is shorter in other forms of CSA(Tkacova et al., 1997).

Further reading

N American Academy of Sleep Medicine(2005). International Classification ofSleep Disorders. 2nd Edn. Westchester,American Academy of Sleep Medicine.

N Tkacova R, et al. (1997). Left ventricularvolume in patients with heart failure andCheyne–Stokes respiration during sleep.Am J Respir Crit Care Med; 156: 1549–1555.

Key points

N CSAs are characterised by a lack ofventilatory impulses generated by thebrain stem.

N Paradoxical breathing and flatteningof the inspiratory curve do notcharacterise CSAs.


Central sleep apnoea/Cheyne–Stokes respiration:epidemiology and risk factorsWinfried Randerath

It is estimated that 10% of breathingdisturbances during sleep are of centralorigin. CSA is more prevalent in the elderly.However, this might be associated with theincreasing relevance of cardiovasculardiseases in this group, which predisposesthem to CSA. A large cohort study of thegeneral population showed that the male sexis another risk factor for CSA (Bixler et al.,2001). There are differences in the apnoeathreshold between the sexes; women have alower apnoeic threshold, which stabilisesrespiration (Skatrud et al., 1983).

CSR is a subgroup of CSA. It is characterisedby a periodic shift of overshooting andundershooting of the ventilation, whichresults in the typical pattern of periodicbreathing (crescendo-decrescendo pattern).

The most relevant risk factors for centralbreathing disturbances during sleep arecardiovascular diseases, such as arterialfibrillation, heart failure and stroke (Solin etal., 1999; Lanfranchi et al., 2003). Javaheri etal. (1998) showed a prevalence of 51% ofsleep-related breathing disorders in stableHF, mostly central. Sin et al. (1999) analysedthe data of 450 consecutive patients with HFand found that male sex, atrial fibrillationand hypocapnia were significantlyassociated with CSA. Cerebral ischaemic

events are frequently accompanied bybreathing disturbances during sleep (Parraet al., 2000). Within the first 72 h .70% ofthe patients showed an AHI .10 events?h-1

and 28% an AHI .30 events?h-1. Whileduring the initial phase 26% of the patientsshowed CSA, the amount was reduced to 7%after 3 months.

Other risk factors include chronic medicationor use of opioids. They are associated withcentral apnoeas, periodic breathing andatactic breathing (Wang et al., 2005).

Further reading

N Bixler EO, et al. (2001). Prevalence ofsleep-disordered breathing in women:effects of gender. Am J Respir Crit CareMed; 163: 608–613.

N Javaheri S, et al. (1998). Sleep apnea in 81ambulatory male patients with stableheart failure. Types and their prevalences,consequences, and presentations. Circula-tion; 97: 2154–2159.

N Lanfranchi PA, et al. (2003). Central sleepapnea in left ventricular dysfunction:prevalence and implications for arrhyth-mic risk. Circulation; 107: 727–732.

N Parra O, et al. (2000). Time course ofsleep-related breathing disorders in first-ever stroke or transient ischemic attack.Am J Respir Crit Care Med; 161: 375–380.

N Sin DD, et al. (1999). Risk factors forcentral and obstructive sleep apnea in450 men and women with congestiveheart failure. Am J Respir Crit Care Med;160: 1101–1106.

N Skatrud JB, et al. (1983). Interaction ofsleep state and chemical stimuli insustaining rhythmic ventilation. J ApplPhysiol; 55: 813–822.

Key points

N The risk of CSA is increased in theelderly and in males.

N Cardiovascular disorders and the useof opioids predispose to CSA.


N Solin P, et al. (1999). Influence ofpulmonary capillary wedge pressure oncentral apnea in heart failure. Circulation;99: 1574–1579.

N Wang D, et al. (2005). Central sleepapnea in stable methadone main-tenance treatment patients. Chest; 128:1348–1356.


Central sleep apnoea/Cheyne–Stokes respiration:pathophysiologyWinfried Randerath

CSA due to hypoventilation

Central apnoeas may appear in patients withincreased or decreased ventilation.Hypoventilation syndromes includediseases with reduced central respiratorydrive or the inability to translate breathingimpulses into thoracic movements. Theseinclude disorders of the central nervoussystem such as inflammatory or ischaemicdiseases of the brain stem, neuromusculardiseases (e.g. amyotrophic lateral sclerosis)or thoraco-skeletal disorders (e.g.kyphoscoliosis). Central apnoeas appearduring sleep in these patients as theventilatory drive and muscle function aremore severely reduced during sleepcompared with the wake state.

CSA due to hyperventilation

The pathophysiology of CSA due tohyperventilation is not yet fully understood.Nevertheless, several pathophysiologicalaspects can be described including the loopgain of the ventilatory response, the apnoeathreshold and the instability of therespiratory control systems during sleep.

During wakefulness, ventilation is primarilyregulated by behavioural factors. In contrast,during sleep, it is mainly influenced bymetabolism, the production and eliminationof CO2. While an increase in PaCO2

(hypercapnia) stimulates ventilation,breathing is diminished during hypocapnia.NREM sleep physiologically reducesventilatory drive and V9E. Thus, CSA appearspredominantly during NREM sleep. Theinfluence of the carbon dioxide level isalleviated during REM sleep as muscleactivity and arousability are reduced.

Loop gain

The ventilatory control system can becompared to a loop gain known fromengineering. The actual PaCO2 is influencedby any disturbance of respiration, e.g.stimulation of ventilation by noise, pain orcortical impulses. V9E can be described asthe plant gain, where increases or decreasesin V9E result in variations in CO2. Thesechanges are measured at the feedback gain,represented by the chemoreceptors in theventilatory system. The perception of thesevariations can be influenced by a circulatorydelay in cardiovascular disorders. Thechemoreceptors stimulate the respiratorycontrol system in the brain stem, thecontroller gain, which urges the plant gain(lungs and thorax) to change ventilation.

Key points

N Central disturbances due tohyperventilation and hypoventilationshould be discriminated.

N Central disturbances due tohypoventilation appear in diseaseswith reduced respiratory drive orfailure of thoracic movements.

N Central disturbances due tohyperventilation depend onovershooting of the ventilation,changes of the apnoea threshold andincreased chemosensitivity.

N Central disturbances are closelyrelated to HF, atrial fibrillation andbrain infarction.


A strong response to breathing disturbancescharacterises a high loop gain. It results inovershooting of ventilation, a shift betweenhyperventilation and hypocapnia on the onehand and apnoeas and hypercapnia/hypoxiaon the other. As muscle activity and atendency to arousal are reduced during REMsleep, ventilatory overshoot is more oftenpresent during NREM sleep.

Apnoea threshold

The extent of V9E is determined by the PaCO2.If the prevailing PaCO2 falls below a definitelevel, breathing ceases. This level of thePaCO2 is called the apnoea threshold. Duringnormal breathing the PaCO2 exceeds theapnoea threshold. Whenever the distancebetween the actual PaCO2 and the apnoeathreshold is narrowed, the probability of anapnoea increases. On the one hand, theapnoea threshold has been shown to beelevated in CSA, without relevant change inPaCO2. On the other hand, hyperventilationbrings PaCO2 closer to the apnoea threshold,which is a typical finding in patients withunstable breathing. In this case, smallvariations in ventilation lead to oscillationsin PaCO2 above and below the apnoeathreshold resulting in a pattern of periodicbreathing. Chemical irritation of thepulmonary tissue receptors increases thecentral ventilatory drive in animals. Thisreflex activation is transmitted by vagalafferent nerves. Interstitial inflammatoryprocesses can stimulate irritant receptors inhumans. However, this irritation is muchmore common in HF patients induced bypulmonary congestion. Therefore,pulmonary congestion, represented by anincrease of the pulmonary wedge pressureas a measure of left ventricular failure,correlates with the AHI and the ventilatorydrive. Consecutively, improvement ofcardiac function reduces hyperventilationand the AHI.

Hypoxic and hypercapnic ventilatoryresponse

The sensitivity of the peripheral and centralchemoreceptors also influences the stabilityof the respiratory control system. Theventilatory response to hypercapnia is

significantly increased in CSA patientscompared with healthy subjects, HF patientswithout breathing disturbances and HFpatients with OSA. Therefore, hyperreactivityof the chemoreceptors is a typical marker,not of HF in general, but only in patientswith associated central breathingdisturbances. From a clinical point of view,hyperreactivity of the chemoreceptors leadsto a higher increase of V9E in response to agiven PaCO2, resulting in hypocapnia.Hypocapnia is followed by a reduction ofventilation, apnoea and hypoxia. A viciouscircle of overshooting and undershooting isgenerated.

The role of delayed circulation time in thepathophysiology of central breathingdisturbances is unclear. Based on animaltrials it has been suggested that a delayedperception of blood gas changes mightinduce decelerated reactions of theventilation. However, it is difficult totranslate these finding to humans.

Other factors seem to play a minor role inthe pathophysiology of CSA, includingcentral chemoreception. The cerebral bloodflow differs between healthy subjects andCSA patients. Hypercapnia increases thecerebral blood flow, but the vasoreactivity isdiminished in CSA patients. The variationsof the cerebral blood flow physiologicallycounterbalance changes of the H+concentration in the cerebral fluid. If thesereflexes are dampened, cerebral alkalosismight be intensified leading to a reductionof the ventilatory drive and central apnoea.In contrast, increases of cerebral acidosisinduce ventilatory overshoot andhyperventilation and thus unstable breathing.

Arousals

Arousals restore the waking state of non-chemical control of ventilation, lower theCO2 set point and increase ventilatoryreactivity in healthy persons and in OSA. ThePaCO2 level during sleep is higher comparedwith the new set point in the sleep/waketransition. As a consequence, ventilation israpidly elevated during an arousal.Moreover, arousals increase the tone of theupper airway muscles. These factors


normalise the patency of the upper airwaysand terminate obstructive breathingdisturbances. However, in central apnoea,arousals perpetuate the elevated loop gain.Due to the higher sensitivity of thechemoreceptors, increases of the ventilationaggravate the overshoot and thereforeincrease breathing disturbances.

Further reading

N Bradley TD, et al. (1992). Central sleepapnea. Clin Chest Med; 13: 493–505.

N Crowell JW, et al. (1956). Basic oscillatingmechanism of Cheyne–Stokes breathing.Am J Physiol; 187: 395–398.

N Hanly P, et al. (1993). Pathogenesis ofCheyne–Stokes respiration in patients withcongestive heart failure. Relationship toarterial PCO2. Chest; 104: 1079–1084.

N Lorenzi-Filho G, et al. (2002). Relation-ship of carbon dioxide tension in arterialblood to pulmonary wedge pressure inheart failure. Eur Respir J; 19: 37–40.

N Naughton M, et al. (1993). Role ofhyperventilation in the pathogenesis ofcentral sleep apneas in patients withcongestive heart failure. Am Rev RespirDis; 148: 330–338.

N Phillipson EA. (1978). Control of breath-ing during sleep. Am Rev Respir Dis; 118:909–939.

N Solin P, et al. (1999). Influence ofpulmonary capillary wedge pressure oncentral apnea in heart failure. Circulation;99: 1574–1579.

N Solin P, et al. (2000). Peripheral andcentral ventilatory responses in centralsleep apnea with and without congestiveheart failure. Am J Respir Crit Care Med;162: 2194–2200.

N Wellman A, et al. (2003). Respiratorysystem loop gain in normal men andwomen measured with proportional-assist ventilation. J Appl Physiol; 94:205–212.

N Xie A, et al. (2002). Apnea-hypopneathreshold for CO2 in patients with con-gestive heart failure. Am J Respir Crit CareMed; 165: 1245–1250.

N Xie A, et al. (2005). Cerebrovascularresponse to carbon dioxide in patientswith congestive heart failure. Am J RespirCrit Care Med; 172: 371–378.

N Xie A, et al. (1994). Interaction ofhyperventilation and arousal in the patho-genesis of idiopathic central sleep apnea.Am J Respir Crit Care Med; 150: 489–495.

N Yu J, et al. (1998). Stimulation of breath-ing by activation of pulmonary peripheralafferents in rabbits. J Appl Physiol; 85:1485–1492.

N Yumino D, et al. (2008). Central sleepapnea and Cheyne-Stokes respiration.Proc Am Thorac Soc; 5: 226–236.


Central sleep apnoea/Cheyne–Stokes respiration:clinical aspectsWinfried Randerath

The typical clinical findings of patients withOSA are daytime sleepiness, neurocognitivedeficits and limitations in daily life. Thesesymptoms of disrupted sleep may alsocharacterise CSA. However, often they arenot noticed and mentioned spontaneouslyby the patient. Subjective measures ofdaytime sleepiness often show lesslimitation in CSA patients compared with

controls, while CSA patients perform worsein objective tests. This discrepancy may bedue to the overwhelming symptoms of theunderlying disease. Thus, central breathingdisturbances cannot be excluded if patientsdo not report daytime symptoms. Carefulquestioning, additional tests such as drivingsimulators, OSLER test or MWT and PSGare mandatory.

There are no clinical signs which allow fordifferentiation between CSA and other typesof breathing disturbances during sleep.

Further reading

N Arzt M, et al. (2006). Sleepiness andsleep in patients with both systolic heartfailure and obstructive sleep apnea. ArchIntern Med; 166: 1716–1722.

N Hastings PC, et al. (2006). Symptomburden of sleep-disordered breathing inmild-to-moderate congestive heart failurepatients. Eur Respir J; 27: 748–755.

Key points

N CSA patients often do not reportsleep-related symptoms.

N Objective measures of sleepinessshow limitations in CSA comparedwith controls.

N CSA cannot be differentiated fromother breathing disturbances basedon clinical findings.


Central sleep apnoea/Cheyne–Stokes respiration:consequencesStefan Andreas

Neurobehavioural (sleepiness, cognitiveand psychiatric)

While there is convincing evidence that OSAhas detrimental neurobehaviouralconsequences, as alluded to in the sectionof this book on ‘‘Obstructive sleep apnoea:consequences’’, there is no such evidencefor CSA in HF. Indeed, in a large, randomcommunity sample, patients with HF hadless daytime sleepiness (as measured bylower ESS scores) than controls, despitereduced total sleep time (Arzt et al., 2006).Similarly, other cross-sectional studiesfound no correlation between ESS and AHIin HF. However, these studies showedcorrelations between ESS and the severity ofHF. This resembles a number of otherdiseases that are accompanied bydaytime sleepiness independently of sleepand breathing.

For the treatment of CSA, most studiesfound less stage 1 and more stage 2 sleepwith active treatment. However, slow-wavesleep, REM sleep and arousals wererelatively unaffected by treatment. Thus, it isnot clearly established whether CSAprofoundly disturbs sleep or only affectsconventional sleep scoring. Some controlledtrials have found an improvement indaytime sleepiness with successful

treatment of CSA while others have not. Tomake things more complex, cause and effectare complicated by the fact that HF itselfcauses daytime sleepiness.

The absence of a clear causal associationbetween CSA and neurobehaviouralsequelae is not unexpected. Patients withHF are mostly elderly and, thus, their sleepis more disturbed compared to youngerpatients. Similarly, HF itself, older age orother conditions affecting general healthimpair subjective and objective sleep. Thisassociation is related to sympatheticactivation and oxidative stress. As discussedin more detail later, CSA, unlike OSA, doesnot cause striking sympathetic activation.

Cardiovascular

When considering the effects of CSA oncardiovascular (CV) disease, it should beconsidered that this is a very different storyto the effects of OSA on CV disease for tworeasons. First, patients with CSA alreadyhave severe HF, since HF is the main causeof CSA (see the section of this book on‘‘Central sleep apnoea/Cheyne–Stokesrespiration: pathophysiology’’). Secondly,the oscillation of ventilation, heart rate,oxygen saturation, etc. around a meancharacteristic for CSA is clearly distinct fromobstructive apnoeas with negative pressureswings and consecutive hypoxia. A failingheart with reduced ejection fraction,diastolic dysfunction and increased fillingpressure is more vulnerable to stressorssuch as increased blood pressure (afterload)or sympathetic activation compared with ahealthy heart. This is the reason why OSA isespecially detrimental in patients withestablished HF.

Key points

N There is no clear association of CSAwith neurobehavioural sequelae.

N In contrast to OSA, the CV sequelae ofCSA are not well established.


For CSA, the CV consequences are lessclear, and controversy remains as to whetherCSA contributes to morbidity and mortalityin CHF patients or whether it is simply anepiphenomenon. While some studies havedescribed an association between CSA andmortality independent of confounders, otherstudies were not able to replicate thisobservation (Pinna et al., 2009). Since CHFis a complex disease related to CSA by avariety of mechanisms (see ‘‘Central sleepapnoea/Cheyne–Stokes respiration:pathophysiology’’) this does not comeas a surprise.

This might be due to the fact that CSA,unlike OSA, is not characterised by negativeintrathoracic pressure swings. Furthermore,although there are clearly oxygendesaturations during CSA, the averageoxygen saturation is still in the normal rangein most patients due to the periods withhyperventilation with concomitant highoxygen saturations. Indeed, CSA in CHFpatients seems not to cause the strikingsympathoexcitation and daytime sleepinessthat occur with OSA (Spicuzza et al., 2003).Likewise, noradrenaline spill-over and brainnatriuretic protein (BNP) were associatedwith the severity of HF but not CSA in CHFpatients (Mansfield et al., 2003).

Important insights into possible CVsequelae of CSA in CHF can also be gainedfrom therapeutic interventions. A largerandomised controlled trial using CPAP totreat CSA was clearly negative for the mainend-point, transplantation-free survival,while surrogate end-points (noradrenalineand ejection fraction) were still positive(Bradley et al., 2005). However, CPAPreduced CSA by only ,50%. Other largerandomised controlled trials using moreeffective treatment modalities are stillrunning, and might give insights into

whether there is a causal associationbetween CSA and CV disease (see ‘‘TreatingCSA’’). For the time being, it is fair toconclude that the sequelae of CSA on the CVsystem are far less detrimental than those ofOSA. When considering treatment of HF, itshould be kept in mind that successfultreatment of HF also improves central andobstructive apnoeas. HF management iscomplex, and comprises pharmacologicaland electrophysiological therapy. Indeed,cardiac resynchronisation therapy on top ofconventional medical therapy has shown aclear improvement of left ventricularfunction, exercise capacity and CSA in anumber of well-controlled studies.

Further reading

N Arzt M, et al. (2006). Sleepiness andsleep in patients with both systolic heartfailure and obstructive sleep apnea. ArchIntern Med; 166: 1716–1722.

N Bradley TD, et al. (2005). Continuouspositive airway pressure for central sleepapnea and heart failure. N Engl J Med; 353:2025–2033.

N Mansfield D, et al. (2003). Raised sympa-thetic nerve activity in heart failure andcentral sleep apnea is due to heart failureseverity. Circulation; 107: 1396–1400.


N Pinna GD, et al. (2009). Pathophy-siological and clinical relevance of sim-plified monitoring of nocturnal breathingdisorders in heart failure patients. Eur JHeart Fail; 11: 264–272.

N Spicuzza L, et al. (2003). Autonomicmodulation of heart rate during obstruc-tive versus central apneas in patients withsleep-disordered breathing. Am J RespirCrit Care Med; 167: 902–910.


Hypoventilation syndromes/chronic respiratoryinsufficiency in sleepViliam Donic and Zoltan Tomori

Hypoventilation syndromes

Sleep-related hypoventilation syndromes(SRHS) form part of the SDB spectrum, inaddition to OSA and CSA. They arecharacterised by elevated PaCO2

(.45 mmHg) during sleep ordisproportionately increased PaCO2 relativeto levels in wakefulness (Muir, 2010). Thereis accompanying arterial oxygendesaturation during sleep, usuallydemonstrated by an SaO2 ,90% for morethan 5 min with a nadir of 85% or lower(Caples et al., 2005). This ventilatoryderangement is associated with excessivedaytime sleepiness, tiredness, morningheadaches and poor sleep quality as a resultof frequent interruption of sleep withrepetitive arousals and CO2 retention.According to the International classificationof Sleep Disorders (ICSD-2; AmericanAcademy of Sleep Medicine, 2005) andother sources, hypoventilation disorders canbe divided into three major categories: a)hypoventilation syndromes; b) comorbidrespiratory disorders; and c) comorbidnonrespiratory disorders.

Pathophysiology of ventilatory changesduring sleep

Hypoventilation and chronic respiratoryinsufficiency Hypoventilation and chronicrespiratory insufficiency caused by a rangeof pathologies are characterised byhypercapnia and hypoxia, which worsenduring sleep. The ventilatory response tohypercapnia and hypoxemia decrease duringsleep. During REM sleep in healthy people,general muscle hypotonia develops, with theexception of the diaphragm. This increasesthe collapsibility and decreases the patency

of the upper airway, increasing resistance toventilation. The accessory respiratorymuscles and intercostal muscles are alsoaffected, which impacts adversely onrespiratory mechanics. The resulting shallowbreathing increases dead space ventilation.As a consequence of all these changes, gasexchange worsens during sleep, particularlyin neuromuscular disease, chest wall andchronic obstructive airway diseases. Thedecreased tidal volume and V9E result inalveolar hypoventilation, hypoxaemia andcarbon dioxide retention. These changestrigger sleep fragmentation, frequentarousals and daytime sleepiness, fatigueand possible cognitive impairment (DeBacker, 1995). If this sequence is notaddressed, over time deteriorating alveolarventilation–perfusion with increasedpulmonary vascular resistance can causepulmonary hypertension, with consequentright ventricular impairment and corpulmonale. Early diagnosis and effectivetreatment are crucial to prevent these severecomplications.

Respiratory-related sleep disruption isgenerally not significant until thediaphragmatic function is markedlyimpaired. In this situation, for instance inDuchenne muscular dystrophy,hypoventilation and oxygen desaturationoccur, initially during REM sleep, buthypoventilation progresses into NREM sleepas muscle strength deteriorates.

Central hypoventilation syndromes

Congenital central alveolar hypoventilationsyndrome (CCHS or Ondine’s curse)represents a failure of central respiratorydrive. Hypoventilation begins in infancy and


it is worse during sleep than duringwakefulness (see also the section of thisbook on ‘‘Sleep disordered breathing inchildren’’). Persons afflicted with CCHSclassically suffer from hypoventilation, whichmay culminate in respiratory arrest duringsleep. CCHS is exhibited typically from birthor in infancy, and should be distinguishedfrom rare central hypoventilation disordersacquired as a consequence of severe brainor spinal trauma (e.g. after an automobileaccident or stroke, or as a complication ofneurosurgery or meningo-encephalitis).CCHS reflects a diffuse dysregulationprocess within the autonomic system andaffected individuals may haveHirschsprung’s disease or cardiacarrhythmias and develop tumours of neuralcrest origin. Mutations in the homeoboxgene PHOX2B were identified as the causeof CCHS in 2003. More than 90% of CCHSpatients have mutations that add extraalanines to the polyalanine tracts in thePHOX2B gene. These are called polyalaninerepeat mutations (PARMs); the remaining,10% of cases have a non-PARM missense,nonsense or frameshift mutation (Weese-Mayer et al. 2009). It is notable that the non-PARM mutations tend to occur de novo andresult in a severe phenotype with therequirement for continuous ventilatorysupport. A relationship between PARMgenotypes and phenotypic expression is alsobecoming evident. The genotype is denotedwith reference to the normal 20-pair alaninetract (expressed as 20/20 in individualswithout mutation). Individuals with 20/27and 20/33 genotypes frequently requireventilatory support. Late-onset cases withmild hypoventilation may present withrespiratory failure after a general anaestheticor ingestion of respiratory depressants orfollowing a chest infection, and are morelikely to have the 20/24 or 20/25 genotypes.Such cases may be adequately treated withnocturnal NIV, in contrast to more severecongenital variants.

About 1 in 200,000 live-born children haveCCHS. Severe cases develop life-threateningepisodes of hypoventilation or apnoea withcyanosis, usually in the first months of life.In typical congenital cases, medical

examination should exclude lesions of thebrain, heart and neuromuscular disorders,and investigation will demonstrate animpaired ventilatory response tohypercapnia and decreases in PaO2.Polysomnography shows that hypoventilationis most marked during NREM sleep (animportant differential point from patients withneuromuscular disorders in whomhypoventilation worsens in REM sleep). Inmore marked cases, however, hypoventilationis also present progressively during othersleep stages and even in wakefulness. Mostpeople with CCHS do not survive infancy,unless they receive ventilatory assistance,tracheostomy and lifetime mechanicalventilation at least during sleep, butprognosis has improved with early recognitionand care. This improvement in outcome hasbeen aided by the creation of nationalregistries and the formation of a network ofspecialised centres in some countries, whichprovide optimal multidisciplinary surveillanceand management.

Several other genetic syndromes can beassociated with central hypoventilation.They include myelomeningocele withArnold–Chiari malformation, skeletaldysplasia, Prader–Willi syndrome andinborn errors of the metabolism such aspyruvate dehydrogenase deficiency andcarnitine deficiency.

Sleep-related hypoventilation or hypoxaemiamay occur also as a result of pulmonaryparenchymal or vascular pathology. This iscaused by interstitial lung disorders such asinterstitial pneumonitis or sickle cellhaemoglobinopathies.

There is also a rare category of idiopathicsleep-related nonobstructive alveolarhypoventilation. This refers to decreasedalveolar ventilation resulting in sleep-relatedarterial oxygen desaturation andhypercapnia in patients with normalmechanical properties of the lungs.

Obesity hypoventilation syndrome

OHS is an increasingly recognised disorder,characterised by symptoms present in thepreviously described Pickwickian syndrome.Daytime hypercapnia (PaCO2 .45 mmHg) is


a cardinal sign of OHS, reflecting reducedventilation during sleep that persists duringwakefulness, hence the definition of OHS isobesity (BMI .30 kg?m-2 accompanied bySDB and waking hypercapnia. Other causesof hypoventilation as described in thischapter should be excluded. In the majorityof OHS cases, obstructive apnoeas are seen,but others have an AHI of ,5 events?h-1 andexperience pure hypoventilation asconfirmed by an increase in PaCO2 duringsleep of 10 mmHg above waking value. Amild restrictive ventilatory defect iscommonly seen on pulmonary functiontesting, due to increased thoracic impedance.Respiratory muscle strength measurementsshow variable results due either to normalmuscles working at a mechanicaldisadvantage or to a degree of weakness.Compared with normocapnic OSA patients,OHS patients have a worse quality of life andgreater propensity to pulmonary hypertension(Mokhlesi et al., 2007).

Obese normocapnic subjects and OHSpatients have similar hypercapnic ventilatoryresponse (HCVR), measured by theocclusion pressure technique, and in bothvalues are higher than in normal subjects.However, OHS patients are unable toincrease HCVR in response to a hypercapnicchallenge compared with obesenormocapnic subjects. Obesity, geneticcauses, SDB and leptin resistance arethought to explain the blunted HCVR. Earlywork suggests that serum leptin levels maydecrease following OHS therapy with bi-levelNIV, accompanied by an increase in HCVR.

Neuromuscular disorders

Neuromuscular disorders are diseasescaused by impairment of the motor unitcomprising the motor neuron, nerve root,myoneural junction and muscle. Sleep-related hypoventilation or hypoxaemia resultfrom neuromuscular disease or chest walldisorders. They include myasthenia, acidmaltase deficiency, Duchenne musculardystrophy, amyotrophic lateral sclerosis(ALS), myotonic myopathy, poliomyelitis,high spinal injuries with tetraplegia andrespiratory paralysis, and spinal muscularatrophy. OHS, kyphoscoliosis and sequelae

of thoracoplasty for tuberculosis and otherdiseases belong to the class of chest walldisorders. Chronic respiratory muscle failureusually develops slowly over a period ofyears. ALS and Creutzfeldt–Jacob diseasecan progress rapidly toward death, butfamilial dysautonomia and other autonomicdisorders can have a slower evolution.Poliomyelitis can alter central and peripheralrespiratory function, being evident asmuscle atrophy and restricted ventilation.Hereditary/sensory polyneuropathies andneuromuscular junction impairment (e.g.congenital and acquired myasthenia) alsoresult in disorders of breathing. Musculardisease (myotonic dystrophy) may manifestwith periods of alveolar hypoventilationresulting from additional impairment ofcentral respiratory drive. Other myopathies(Duchenne muscular dystrophy, maltasedeficiency, myopathy and myotonicdystrophy) also cause hypoventilationresulting in progressive muscle weakness.High spinal injuries may cause tetraplegiaand respiratory paralysis, as well as spinalmuscular atrophy.

Orthopnoea is a common complaint in ALSand can compound sleep disruption. DuringREM sleep, there is a post-synapticinhibition of somatic motor neurons or evenloss of tone in ribcage and other accessoryrespiratory muscles. The diaphragm isrelatively unaffected. The breathingdisorders manifest with prolongedhypoventilation or central and obstructiveapnoeic episodes, depending on the type ofneuromuscular impairment. In slowlyprogressive neuromuscular disorders,respiratory failure usually advances in threestages. In stage 1, alveolar hypoventilationdevelops during only REM sleep; in stage 2,it is present during REM and NREM sleep;and in stage 3, it develops during both sleepand wakefulness. The gradual transitionfrom stage 1 to stage 3 respiratory failure ischaracteristic of Duchenne musculardystrophy, in which there is a progression inrespiratory muscle weakness over a decadeor so. In other disorders (congenitalmuscular dystrophy) respiratoryinsufficiency is present, but varies accordingto disease subtypes. In other diseases,


respiratory muscle weakness may beextremely severe, with presentation of stage3 respiratory failure, e.g. rapid-onsetGuillain–Barre syndrome.

Nocturnal hypoventilation is found in asubset of COPD patients, but does notusually occur unless FEV1 is ,1 L, and isless likely in pure emphysema patientsunless they are in end-stage disease.Increased work of breathing, overloadedrespiratory mechanics, inadequate nutritionresulting in respiratory muscle weakness,concurrent OSA and genetically lowventilatory drive are precipitating factors.Nocturnal hypoventilation in cystic fibrosisis described in the section of this book on‘‘Comorbid respiratory disorders in children’’.

Further reading

N American Academy of Sleep Medicine.(2005). International Classification ofSleep Disorders. 2nd Edn. Westchester,American Academy of Sleep Medicine.

N Caples SM, et al. (2006). Central sleepapnea, hypoventilation syndromes andperiodic breathing disorders. In:Randerath WJ, et al. Sleep Apnea. Basel,Karger; pp. 180–191.

N De Backer WA. (1995). Central sleepapnoea, pathogenesis and treatment: anoverview and perspective. Eur Respir J; 8:1372–1383.

N Mokhlesi B, et al. (2007). Recentadvances in obesity hypoventilation syn-drome. Chest; 132: 1322–1336.

N Muir JF. Hypoventilation syndromes. In:Palange P, Simonds AK, eds. ERSHandbook of Respiratory Medicine.Sheffield, European Respiratory Society,2010; pp. 414–416.

N Piper AJ. (2007). Current perspectives onthe obesity hypoventilation syndrome.Curr Opin Pulm Med; 13: 490–496.

N Weese-Mayer DE, et al. (2009).Congenital central hypoventilation syn-drome from past to future: model fortransitional autonomic medicine. PaediatrPulmonol; 44: 521–535.


Nonrespiratory sleepdisorders

Renata L. Riha

Nonrespiratory sleep disorders comprise alarge proportion of the problems presentingto any sleep department. It is essential to beable to identify them and either to treatthem appropriately or to refer them for morespecialised assessment. Many of thesedisorders will require more extensiveinvestigations, including extended EEGmontage and overnight monitoring usingvideo telemetry.

Disorders such as chronic insomnia areamong the most common within thecommunity and can overwhelm services ifthey are not appropriately managed. Anumber of simple strategies outlined inthis chapter can be used in dealing withpatients with insomnia. Furthermore,

recognition of the impact that psychiatricillness can have on sleep is vital. Depressionor other affective disorders can compromisesleep states and can also mimic sleepdisorders such as those classified under thecentral hypersomnias.

This section of the handbook focuses on thenosology, diagnosis and management ofnonrespiratory disorders of sleep. Thereader is referred to the InternationalClassification of Sleep Disorders, 2ndEdition, Diagnostic and Coding Manual(American Academy of Sleep Medicine,2005). It provides an extensive and succinctclassification of both respiratory and non-respiratory conditions and can also be usedas a guide in conjunction with the overviewpresented here.

Further reading


N Winkelman JW, et al. (2008). AmericanAcademy of Sleep Medicine Case Book ofSleep Medicine (ICSD-2). Westchester,American Academy of Sleep Medicine.

Key points

N It is essential that respiratory sleepphysicians can recognise non-respiratory sleep disorders.

N Psychiatric illness can have a majorimpact on sleep, and its effects canmimic those of sleep disorders.


Insomnia

Dirk Pevernagie

The essence of insomnia is a predominantcomplaint of dissatisfaction with eitherduration or quality of sleep. Insufficientsleep duration may be due to difficultiesinitiating sleep (DIS), difficultiesmaintaining sleep (DMS), or awakening tooearly in the morning. Nonrestorative sleep(NRS) denotes a feeling of beingunrefreshed after a sufficiently long andconsolidated nocturnal sleep period.

Supposedly, NRS pertains to decreasedsleep quality, and is therefore believed to bea manifestation of insomnia.

Diagnosis

To establish a diagnosis of insomnia, threefundamental criteria must be met: 1)recurring difficulties with sleep; 2) adequateopportunities for sleep; and 3) impairmentof daytime functioning.

A poor night of sleep is a nuisance thateveryone experiences sooner or later.Mostly, it is a transient situation thatrequires no medical attention. When sleepdisturbances are persistent and occur onmost nights of the week, a diagnosis ofinsomnia can be made.

Adequate time allocation and environmentalamenities conducive to obtaining optimalsleep are important, and must be consideredin the differential diagnosis of insomnia. Ifthese conditions are not met, the sleepdisturbance may be ascribed to externalfactors or inadequate sleep hygiene or may belabelled as insufficient sleep syndrome.Correction of these restrictive factors shouldresult in improvement of sleep.

Another obligatory diagnostic criterion is thepresence of daytime dysfunction, whichresults from too little sleep or an inferiorquality of sleep. Daytime complaints consistof fatigue, impaired cognitive abilities(e.g. decreased memory and concentration)and mood disturbances (e.g. irritability anddysphoria). Because these symptoms maypersist for prolonged time periods, chronicinsomnia may lead to complications such associofamilial disharmony and lost workproductivity. Insomniacs are typically

Key points

N Insomnia consists of a predominantcomplaint of dissatisfaction witheither duration or quality of sleep.Difficulties with sleep are recurrent,occur despite adequate opportunitiesfor sleep, and are associated withimpairment of daytime functioning.

N Chronic insomnia occurs in ,10% ofthe adult population. Psychiatricillness and SDB are the mostcommon comorbidities.

N Insomnia is considered a disorder,whether it is primary (withoutco-existing disease) or comorbid(associated and interacting withmedical or psychiatric disease).

N Assessment of insomnia is based oncareful history, clinical examination, atwo-week sleep/wake diary and, ifneeded, additional tests such as PSGand actigraphy.

N Insomnia that is comorbid with SDBis a particular condition, characterisedby more severe symptoms.


preoccupied with their inability to sleep.Their health-related quality of life isimpaired. Because of the daytimedysfunction, insomnia may increase the riskof accidents at work or on the road.Treatment with hypnosedative drugs,especially the longer-acting ones, may causeresidual drowsiness during the daytime andthus jeopardise safety aspects of workingand driving.

The symptoms of daytime dysfunctionshould improve with adequate treatment ofinsomnia. This concept is essential tocorroborate the diagnosis, because daytimeimpairment may be due to other medicalconditions, e.g. other primary sleep disordersand somatoform disorders, such as chronicfatigue syndrome and fibromyalgia.

Subjects who have curtailed nocturnal sleep,but who perform normally during thedaytime, do not suffer from insomnia: theyare referred to as habitual short sleepers.

Primary, secondary and comorbid insomnia

Insomnia may be a solitary manifestationwithout any demonstrable associatedmedical or mental illness. This condition,named primary insomnia, is classified as anindependent disorder. Often, however,insomnia is an epiphenomenon of anailment that is associated with mental orbodily discomfort. Lying awake or waking upis then the consequence of a nuisance thatnegatively interferes with sleep. Accordingly,insomnia is a very common complaint inchronic medical or psychiatric diseases.

Sleeplessness due to obvious causes isreferred to as secondary insomnia. In thiscase, insomnia is a symptom rather than anautonomous condition. Then again, chronicinsomnia may play a contributing role in thepathogenesis or aggravation of mental andsomatic diseases. In other words, therelationship between the sleep disorder andthe associated illness may be bidirectional.Often, it is not clear from the historywhether insomnia has preceded theassociated disease or vice versa. Theentwined co-existence between disorderedsleep and the associated diseasemanifestations is termed comorbid

insomnia. Insomnia is in this situation botha symptom and a disorder that calls forspecific treatment in addition to themanagement of the associated medical orpsychiatric disorder. The controversy ofcause–consequence relationships will beresolved with the future publication of theDiagnostic and Statistical Manual of MentalDisorders (DSM-5). The use of the term‘insomnia disorder’ will be proposedwhenever diagnostic criteria are met,whether or not there are co-existingpsychiatric, medical, or other sleep disorders.

Insomnia may co-exist with other primarysleep disorders such as sleep apnoea,narcolepsy and restless legs syndrome. Theresulting clinical picture may be complexand difficult to tackle diagnostically andtherapeutically. Comorbid insomnia in SDBis an enigma par excellence, becausetreatment of insomnia with hypnotics mayhave detrimental effects on breathing,whereas application of CPAP for sleepapnoea may aggravate insomnia.

In this section, some essentials ofprevalence, pathogenesis, classification,assessment and relation with SDB areelaborated. The treatment options forinsomnia are described in Module 6 ofthis handbook.

Prevalence

The prevalence of insomnia inepidemiological studies depends highly onthe diagnostic criteria used and on thetarget population. While in population-based surveys up to 30% of adults report atleast one of the symptoms of insomnia(DIS, DMS, waking up too early and NRS),adding the diagnostic criterion of daytimedysfunction reduces the prevalence toapproximately 10%. An increased prevalenceof insomnia is associated with the femalesex and older age. The most commoncomorbidities are psychiatric illness andSDB. Current estimates indicate that 40% ofall insomnia patients have a co-existinganxiety or depression disorder, andinsomnia is a diagnostic symptom ofthese conditions.


Pathophysiology

Evidence is accumulating that insomnia isassociated with inappropriately elevatedphysiological arousal. This state, known ashyperarousal, appears to be presentthroughout the 24-h period. It has beenshown that insomnia is associated withincreased metabolic rates of brain and body,elevated heart rates, augmented secretion ofstress hormones, increasedelectroencephalographic activity in the high-frequency spectrum, and activation of thesympathetic nervous system during sleep.Therefore, chronic insomnia may beassociated in the long term with increasedrisk for cardiovascular diseases anddepression. Animal research and humanbrain metabolic studies in insomnia modelssuggest that brain sites controlling sleep areactivated simultaneously with arousal systems.

Hyperarousal comprises physiological,cognitive and emotional features. Thecognitive arousal model implies that worriesabout problems in daily life negativelyinterfere with the ability to sleep. Once thesleep difficulties are settled, the focus ofworrying shifts from life events to concernsabout sleep itself and about the daytimeconsequences of not getting enough sleep.The experience of daytime impairmentreinforces the patients’ yearning for goodsleep and incites them to try to regaincontrol over their sleep. Paradoxically, thisintention further stimulates arousalmechanisms and impedes mental relaxationthat is needed to fall asleep. From aphysiological point of view, excessiveattempts to obtain sleep are associated withincreased somatised tension, e.g. increasedmuscular tone. Frustration and anxiety whenlying awake are typical emotionalexpressions of heightened arousal ininsomnia. The recurrent problem ofdisturbed sleep leads the insomnia patientinto a vicious spiral, sustained byconditioned hyperarousal, dysfunctionalbeliefs about sleep and learnt maladaptivebehaviour, as a consequence of which sleepis further disabled.

In a time-honoured model explaining thedynamic development of insomnia over

time, three pathogenetic groups of factorsare implicated (fig. 1). Predisposing factorsare trait characteristics that may begenetically determined. They are static overtime within individual patients. Thesefactors relate to proneness to insomnia andencapsulate the risk for developinginsomnia in the pre-morbid phase.Precipitating factors are acute occurrencesor stressors that trigger the onset ofinsomnia. These triggers are related tostressful life events (e.g. professional andsocial mishaps and also health-relatedincidents). The third group (theperpetuating factors) refers to the cognitiveand behavioural issues that constitute theself-perpetuating mechanism of chronicinsomnia. Precipitating and perpetuatingfactors are non-static, and changes in theirrelative intensities will modulate theexpression of chronic insomnia over thecourse of time. This 3P model provides thefundamentals on which the nosologicalclassification and the cognitive behaviouraltreatment strategies for insomnia are built.

Nosological classification

Several types of insomnia are listed in theInternational Classification of SleepDisorders (ICSD-2). The first four entitiesbelong to the group of primary insomnia.

Adjustment (acute) insomnia has a suddenonset and is caused by an obvious stressfulevent. The sleep disturbance is expected tobe of short duration (not longer than3 months), and usually remits with thedisappearance of the precipitating factor.

Psychophysiological insomnia is theprototype of insomnia that is maintained bythe influence of perpetuating factors, i.e.hyperarousal and learnt maladaptivebehaviour. Wakefulness is promoted bynegative anticipation, muscle tension,frustration and ruminating thoughts. Indivi-duals are typically preoccupied about sleepdifficulties and their adverse consequenceson performance during the daytime.

Paradoxical insomnia refers to a complaintof severe insomnia that is not on a par withobjective evidence of sleep disturbance.Typically, patients claim to have been awake


for most of the night, while PSG shows littleor no abnormalities in the hypnogram.Formerly, this condition was known as sleepstate misperception. It is currently unknownwhether the complaint is based on thepatient’s inadequate awareness of beingasleep, or whether there is an, as yet,unidentified intrinsic disturbance of sleep.Because the suspicion of unawareness maycompromise the patient’s credibility, theterm misperception is no longer used.

Idiopathic insomnia is diagnosed when theinability to sleep persists from early life andno major periods of spontaneous sustainedremission have occurred. Idiopathic insomniais a prototype of a sleep disturbance that isdue to predisposition (constitutional factors),with little or no contribution of precipitatingor perpetuating factors.

Insomnia due to mental disorder refers tosleeplessness that occurs in the course of apsychiatric disease. This insomnia isthought to be caused by that disease and issevere enough to cause distress and towarrant special treatment.

A diagnosis of inadequate sleep hygiene ismade when voluntary behaviour impairsadequate sleep. Such activities may induce

arousal or directly interfere with sleep.Common examples are irregular sleep/wakescheduling, use of nicotine or caffeine orengaging in activities that are not conduciveto sleep (working on the computer, watchingexciting movies, etc.).

Insomnia due to a drug or substanceimplies an inappropriate use of prescriptionmedicine, recreational or illicit drugs.Typically, the insomnia will remit followingdiscontinuation of the substance.

Insomnia due to a medical condition is aterm used in the context of a comorbidmedical disorder. Although complaintsabout sleep are often present in variousmedical conditions, this diagnosis isrestricted to patients in whom insomnia iscaused by the medical disorder and isassociated with significant distress thatrequires specific medical care.

Insomnia that is not related to substanceabuse or known physiological conditions(nonorganic insomnia) or that is related tounspecified physiological conditions(organic insomnia) are the last entities ofthe ICSD-2 insomnia classification. Thesediagnoses are typically used when furtherassessment is required to identify

0

Inso

mni

a in

tens

ity %

Pre-morbid Acute insomnia Early insomnia Chronic insomnia

PerpetuatingPrecipitatingPredisposing

Figure 1. The Spielman model or 3P model of insomnia (see text for explanation). The threshold linedelineates the level above which insomnia is severe enough to become a prominent complaint. AU:arbitrary units. Reproduced from Spielman et al. (1991), with permission from the publisher.


particular disease contexts, or when thepatient doesn’t meet the criteria for aspecific sleep disorder.

Assessment of insomnia

The evaluation of insomnia is founded onclassical clinical principles, including adetailed patient history and examination, withspecial attention to the sleep and wakingfunctions. In addition, an inventory is made ofcurrent medical, psychiatric and medication-or substance-related comorbidities.Additional technical investigations may beperformed where appropriate.

First, the primary complaint is addressed.Attention is paid to the type of sleepdisturbance, whether DIS, DMS, earlyawakening or NRS. Although the complaintmay be limited to only one of thesesymptoms, a combination of symptoms isoften extant and the complexity of theclinical presentation may change over time.Details about the various features ofinsomnia must be carefully noted. Past andcurrent issues that precipitate or perpetuateinsomnia are inquired, as well as previoustherapies and the responses to treatment.Second, the circumstances and activities ofthe pre-sleep period are scrutinised. Sleep-incompatible activities such as watching TV,computer use, eating in bed, etc. aredocumented. Elements of the sleepingenvironment that are favourable or adverseto sleep must be identified. Also, knowingthe state of mind with respect to beingrelaxed versus aroused at bedtime is helpfulin understanding the situational intricaciesthat affect sleep. Third, the day-to-dayvariability of insomnia, the extremedifferences between good and bad nightsand the evolution over time are importantvariables that must be recorded. Sleeplatency, number of awakenings, wakefulnessafter sleep onset, total sleep duration andtaking naps during the daytime are variablesthat can be assessed on average by taking aretrospective history. To obtain anapproximation of the nightly variability ofthese parameters, a prospective assessmentwith sleep/wake diaries should be carriedout. Although insomnia disorder is not

defined by cut-off criteria, indicators forrelevant insomnia are an average sleeplatency exceeding 30 min, wake after sleeponset exceeding 30 min, sleep efficiency,85%, and/or total sleep time ,6.5 hours.Fourth, nocturnal symptoms and behavioursare reviewed. The bed partner should beinvolved in this part of the history, as thepatient may be unaware of the issues. Suchinformation may be useful to furtherdocument the factors associated withinsomnia (e.g. anxiety), but also to discoversymptoms that may point to co-existentprimary sleep disorders (e.g. SDB andparasomnias). Finally, questions on daytimeperformance should be asked. The answersmay provide clues to potential causes andconsequences of insomnia. Informationshould be obtained on fatigue, somnolence,napping, work schedules andcircumstances, coping with work, copingwith social life, lifestyle, quality of life, mood,cognitive function and mutual influencebetween insomnia and comorbid disorders.

The sleep history is complemented with ageneral medical and psychiatric history anda record of currently used drugs andsubstances. A general physical examinationis carried out. In the further workup, the useof specific questionnaires for sleep quality,insomnia, sleepiness, psychologicalassessment and quality of life should beconsidered. The use of a 2-week sleep/wakediary (fig. 2) is standard. It is an excellentmeans to supplement the oral history and tocharacterise the different insomniacomponents (complaint type, duration,frequency, severity, daytime dysfunction,effects of consuming food, beverages,natural stimulants and hypnoticmedication). Moreover, the sleep/wake diarymay disclose circadian rhythm sleepdisorders that are co-existent with, or thatexclude, the diagnosis of insomnia.Actigraphy for a period of 2–4 weeks maybe indicated when the information fromthe sleep/wake diary is doubtful orincomplete. Laboratory testing and PSG arenot routinely indicated in the evaluation ofinsomnia but may be appropriate in patientsin whom comorbid medical or sleepdisorders are suspected.


Management of insomnia

The multifactorial mechanisms involved inthe initiation and maintenance of insomniahave provided the connecting points for themultimodal treatment methods that haveemerged in recent decades. The differenttreatment modalities are based onpharmacological and nonpharmacologicalinterventions. A comprehensive review ofinsomnia treatment is included in Module 6.

Special interest for the respiratory sleepphysician: insomnia as a comorbidcondition in SDB

While the relationship between insomniaand SDB (mostly OSA) has received littleattention in the medical literature, recent

evidence points out that these two disordersco-occur in a substantial number ofpatients. Insomnia and SDB show a highprevalence in the general population and arethe two most common sleep disorders.Because of inconsistencies in assessmenttechniques, target populations and studydesign, reliable epidemiological figures onthe combined occurrence of insomnia andSDB are not available. From the existingdata, it is estimated that 39–58% of patientswith an established SDB diagnosis alsosuffer from insomnia, and, conversely, 29–67% of insomnia patients have an AHIexceeding 5 events?h-1. Data regarding theinfluence of age and sex on the combinedappearance of these conditions areconflicting.

Figure 2. The sleep/wake diary should be recorded during a period of at least 2 weeks. This figure shows a1-week log. For each day, the following variables are derived from the diary: bedtime, sleep latency (time tofall asleep following bedtime; hatched bars), awakenings and duration of each awakening (hatched bars),periods of sleep with duration (black bars), time out of bed and duration (empty bars), wakefulness aftersleep onset (the sum of all awakenings and time out of bed), time in bed (time from bedtime to gettingout of bed), total sleep time (time in bed minus (sleep latency plus wakefulness after sleep onset)), sleepefficiency (total sleep time divided by time in bed, expressed as a percentage), time out of bed, nap times(frequency, times and durations). The diary may also include reports of perceived sleep quality, daytimefunctioning, medications, and caffeine and alcohol consumption for each 24-hour period.


In fact, SDB and insomnia are dissimilarfrom a clinical and pathophysiological pointof view. From a clinical perspective, thesedisorders can be modelled as paradoxical(orthogonal) conditions because insomniais associated with increased arousal andalertness, whereas in SDB excessive daytimesleepiness is the most characteristic feature.It could be plausible that both disorders areunrelated to each other and, given the highprevalence, their co-occurrence is due tocoincidence. This would imply that insomniaand SDB should be managed separately.

Recent studies, however, shed a differentlight on the SDB–insomnia relationship. Itappears that the co-occurrence of bothdisorders brings about additive negativeeffects in terms of increasedsymptomatology. The situation in whichsymptoms of SDB are aggravated by theco-existence of insomnia is referred to, atleast by some authors, as ‘‘SDB-plus’’. Theclinical presentation is actually worse thanwhat can be expected from the singlecontribution of each individual disorder.Additional negative effects aredemonstrated with respect to subjectivesleep quality, sleep consolidation and totalsleep time. While DIS may be present, DMSseems to be more problematic.Furthermore, these patients suffer fromgreater functional impairment during thedaytime. Sleepiness and neurocognitivefunctions are worse and there are higherlevels of mood disturbance. SDB-pluspatients also endure more worrying anddysfunctional beliefs about sleep, includingnegative anticipation towards healthconsequences in the long term.

Whether SDB intensifies pathogeneticfactors of insomnia, and vice versa, iscurrently unknown. However, insomnia andSDB share arousal from sleep as a commonfeature in their respectivepathophysiologies. In SDB, short arousalsterminate episodes of obstructed breathing,whereas insomnia is characterised by longerperiods of awakening from sleep. It may behypothesised that co-occurrence ofinsomnia and SDB may influence thepredominant arousal patterns that occur in

each of the constituent conditions.Increased levels of sleep deprivation may inreturn contribute to increased upper airwaycollapsibility, thus inducing a higher numberof respiratory events. There may be a role forsympathetic, metabolic and neuro-endocrine mechanisms that typically takeeffect in SDB, to also induce or aggravateinsomnia disorders. In summary, reciprocalfeedback mechanisms could be operationalin the co-occurrent SDB–insomnia situation,thus explaining the detrimental additions tothe clinical profile.

To assess the diverse clinical dimensions ofSDB-plus, a careful diagnostic approach isneeded. PSG, which is not routinelyrecommended for the assessment ofinsomnia, is mandatory in SDB-pluspatients to demonstrate the combination ofSDB and reduced total sleep time. PSG willclearly show the patterns of DIS and DMS inthe hypnogram. Polygraphy (PG), lackingneurophysiological leads, is obviously anunsuitable method to assess the features ofinsomnia. Moreover, the AHI will beunderestimated with PG, becausewakefulness is not taken into account. Thediagnosis of insomnia should be madeappropriately, in line with the assessmenttechniques described previously. Thisimplies the mastering by the managingphysician of pertinent knowledge and skillsthat extend beyond the limits of therespiratory sleep domain. Finally, theimportance of an integrated,interdisciplinary approach to SDB-plus mustbe emphasised.

Further reading

N American Academy of Sleep Medicine.(2005). The International Classification ofSleep Disorders. 2nd Edn. AmericanAcademy of Sleep Medicine, Westchester.

N Bonnet MH, et al. (2010). Hyperarousaland insomnia: state of the science. SleepMed Rev; 14: 9–15.

N Luyster FS, et al. (2010). Comorbidinsomnia and obstructive sleep apnea:challenges for clinical practice andresearch. J Clin Sleep Med; 6: 196–204.


N Morin CM, et al. (2003). Insomnia: aclinical guide to assessment and treat-ment. 1st Edn. New York, KluwerAcademic/Plenum Publishers.

N Riemann D, et al. (2010). The hyperar-ousal model of insomnia: a review of theconcept and its evidence. Sleep Med Rev;14: 19–31.

N Roth T. (2007). Insomnia: defini-tion, prevalence, etiology, and con-sequences. J Clin Sleep Med; 3: Suppl. 5,S7–S10.

N Schutte-Rodin S, et al. (2008). Clinicalguideline for the evaluation and manage-ment of chronic insomnia in adults. J ClinSleep Med; 4: 487–504.

N SpielmanAJ,etal. (1991).Thevariednatureofinsomnia. In: Hauri PJ, ed. Case studies ofinsomnia.NewYork,PlenumPress;pp. 1–15.

N Stone KC, et al. (2008). Nonrestorativesleep. Sleep Med Rev; 12: 275–288.

N Wickwire EM, et al. (2010). Insomnia andsleep-related breathing disorders. Chest;137: 1449–1463.


Parasomnia and movementdisorders during sleep

Renata L. Riha

Parasomnias are disorders resulting inabnormal events during the night such assleepwalking, sleep terrors and bruxism.They can be divided into primaryparasomnias, arising directly from sleep andspecific to a particular stage of sleep, andsecondary parasomnias, which arise frompathology arising during sleep, such asnocturnal epilepsy, nocturnal leg crampsand sudden cardiac death.

Primary parasomnias can arise from eitherREM or NREM sleep but some are notconfined to any particular sleep stage.

Table 1 lists what are considered to benormal events during NREM and REMsleep, generally requiring only reassurance,and also lists abnormal events arisingduring these sleep stages, which mayrequire further investigation and that areconsidered to be potentially more disruptive.

Other primary parasomnias includenocturnal groaning (catathrenia), bruxism,enuresis, rhythmic rocking movementdisorder, sleep talking and sleep eating.These parasomnias can occur throughout allsleep stages.

A number of factors can trigger parasomniasor increase the frequency of occurrence,including: fever, alcohol, sleep deprivation,physical activity, emotional stress andmedications. Parasomnias can also beworsened by pregnancy and menstruationas well as SDB. There is often an interactionbetween genetic and environmental factorsto cause these problems. Bizarre sleep-related activities can occur in the generalpopulation with no underlying comorbiditiesor known predisposition to parasomniasand usually do not warrant furtherinvestigation or treatment. However,medical advice should be sought ifbehaviours are potentially violent/injurious(forensic cases too), extremely disruptive tohousehold or family, result in complaints ofexcessive daytime sleepiness or areassociated with medical, psychiatric orneurological symptoms and signs.

When diagnosing parasomnias, a history ofthe behaviour is taken and it is helpful if bedpartners or witnesses to the behaviour arealso present. PSG with video recording withextended montage may be useful. Events donot always occur on the night of the study,which may thus need to be repeated.Telemetry over a number of nights may beuseful in conditions more diagnosticallychallenging, including frontal lobe epilepsy.Referral to a specialised sleep centre isadvisable in less clear-cut cases as well as inforensic cases, particularly in respect ofaccurate diagnosis.

REM-related disorders

Of the REM-related parasomnias, the mostimportant diagnostically is REM behaviourdisorder (RBD).

Key points

N Parasomnias occur commonly.

N Abnormal nocturnal behaviour canarise out of NREM or REM sleep.

N Complex or difficult cases requirespecialist investigation andmanagement.


Classification RBD was first described bySchenck et al. (1986), and is classified asidiopathic or secondary. The existence ofidiopathic RBD is currently under debate,but it has been used by sleep specialists todescribe RBD not associated with, orpreceding, other disorders. Secondary RBDcan be related to neurodegenerativedisorders and tauopathies such asAlzheimer’s disease, corticobasilardegeneration and progressive supranuclearpalsy; other neurological disorders such asGuillain–Barre syndrome, stroke and spino-cerebellar ataxia; sleep disorders such asnarcolepsy; medications such as selectiveserotonin re-uptake inhibitorantidepressants and b-blockers; andwithdrawal states from alcohol or drugs ofabuse. However, secondary RBD is mostclosely associated with a-synucleinopathies.Longitudinal follow up of RBD patients hasshown that 38–65% develop asynucleinopathy between 10 and 29 yrs fromdiagnosis. The most frequentsynucleinopathy is Parkinson’s disease (PD)but Lewy body disease (LBD) and multi-system atrophy (MSA) may also occur.Likewise, 70% of patients with MSA, 40% ofpatients with LBD and 15–33% of patientswith PD will have RBD. Patients should beappraised of the link between disorders.

Prevalence and presenting features RDB isthought to have a prevalence of 0.38–0.5%

in Western populations (Frauscher et al.,2010). RBD occurs in both sexes, with themajority of patients presenting with thedisease after the age of 50 yrs. It ischaracterised by abnormal and often violentbehaviours emerging during REM sleep,usually in the context of dream mentation.Behaviours can include talking, punching,kicking, flinging or hitting out and can resultin involuntary self-harm or harm to a bedpartner. Between 33% and 65% of RBDpatients have reported a sleep-related injury,ranging from bruises to subduralhaematomas (Boeve, 2010). The violence inRBD relies on proximity of physical contactand can be directed inadvertently at a bedpartner in the context of the dream.

Diagnosis RBD is diagnosed from asuggestive history and supportive PSG. Thediagnostic features on PSG comprise loss ofnormal EMG atonia characterised bysustained muscle activity during REM sleepand excessive transient muscle activity ineither the chin or limb EMG. Use of videoduring PSG recording is helpful in recordingany behaviour.

Pathophysiology The pathophysiology ofRBD is unknown, but is being extensivelyresearched using human and animalmodels. Therapeutic approaches have thusfocused on symptomatic treatment of thedisorder, with the primary aim being tominimise sleep disruption and prevent

Table 1. Normal and abnormal events during NREM and REM sleep.

NREM sleep REM sleep

Normal events Normal events

Sleep starts (hypnic myoclonus) Hypnagogic/hypnopompic hallucinations

Exploding head syndrome Sleep paralysis

Explosive tinnitus

Abnormal events Abnormal events

Confusional arousals Dream anxiety attacks (nightmares)

Excessive fragmentary myoclonus REM behaviour disorder

Sleep-related eating disorder REM-related painful erections

Sleep sex

Sleep terrors

Sleepwalking


violent behaviour that may result insignificant harm.

Management Apart from modifying thebedroom environment (Level A evidence),the two most frequently recommendedtreatments for RBD comprise clonazepamand melatonin. Clonazepam is consideredthe treatment of choice by clinicians forpatients with RBD irrespective ofcomorbidities. However, clonazepam doesnot work in all patients with RBD and cancause unacceptable side-effects that includeexcessive daytime somnolence, excessivemuscle relaxation in the elderly, poor sleepquality, cognitive impairment, depressedmood and worsening of concurrent sleepapnoea. A recent randomised, controlledtrial has shown that melatonin is alsoeffective in RBD (Kunz et al., 2010). Apartfrom clonazepam and melatonin, there islittle evidence to support the use of othermedications. Referral to a specialist sleepservice is recommended for diagnosis andmanagement advice.

Non-REM-related disorders

The focus in this section is on disordersof arousal arising out of slow-wave sleep.They are among the most frequentlyencountered parasomnias.

Classification and epidemiology Parasomniasarising from slow-wave sleep are classifiedin the International Classification of SleepDisorders, 2nd Edition (ICSD2) as arousaldisorders. Disorders that fall under thissubheading include: sleepwalking,confusional arousals, sleep terrors, sleep-related eating disorder and sexsomnia.Arousal disorders occur in up to 20% ofchildren and up to 4% of adults (Hublinet al., 2003). Confusional arousals,sleepwalking and sexsomnias have all beenassociated with forensic behaviour. Sleeptalking can occur throughout all stages ofsleep.

Pathophysiology There is often a strongfamily history suggesting a geneticpredisposition. Arousal disorders can beexacerbated by alcohol, sleep fragmentationand co-occurrence with other sleepdisorders (namely sleep apnoea).

Dreamlike mentations can occur duringsleepwalking and sleep terrors in adults thatdo not require significant executivefunctioning of the frontal cortex. Currentopinion suggests that disorders of arousalare triggered largely by physiologicalchanges occurring during the arousal staterather than by any dream activity precedingthe state per se.

Only one case report exists of an imagingsession in a sleepwalker, which showed theposterior cingulate cortex (part of theemotional processing network) and theanterior cerebellum being activated duringsleepwalking (Bassetti et al., 2000).

The most common psychiatric disordersassociated with sleep disorders aredepression and anxiety with comorbid mooddisorders found in 30.4 % of night terrorsand 14.6 % of sleepwalkers (Mahowaldet al., 2005).

Sleepwalking (somnambulism) Sleepwalkingconsists of a series of complex behavioursthat are initiated in slow-wave sleep andresult in walking during sleep. Thebehaviour may terminate spontaneously orthe sleepwalker may return to bed.Occasionally, there is also inappropriatebehaviour such as smoking, cooking andpreparing food. Falls and injuries may occurif walking into dangerous situations, e.g. outof the door onto the street. Rarely,aggressive behaviour occurs, often as aresponse to attempts to restrict thesleepwalker’s mobility; homicide or suicidehas been reported. Sleepwalking occursalways or often in 1–3% of children andoccasionally in 6–29% of children. It willusually disappear after adolescence. Inadults, sleepwalking occurs always or oftenin 0.1–0.6% and occasionally in 0.9–3% ofthe population. Sleepwalking is hereditaryand about 85% of adult sleepwalkerssleepwalked in childhood.

Night terrors Night terrors are characterisedby a sudden arousal from slow-wave sleepwith a piercing scream or cry, accompaniedby manifestations of intense fear. The attackusually resolves spontaneously. Nightterrors occur always or often in 3% of


children and occasionally in 3.5–15% ofchildren. They usually resolve duringadolescence and less than 1% of adultsexperience them.

Diagnosis Arousal disorders are diagnosedfrom a suggestive history and supportivePSG. Use of video during PSG recording ishelpful in recording any behaviouralmanifestations of the disorder and multiplenight studies may be necessary. ExtendedPSG montage is recommended to excludenocturnal seizures, dissociation and otherparasomnias. Collateral history from a bedpartner, household member and familymember is important. A family history ofarousal disorders should always be askedabout. Attention should also be paid to anypotential aggravating factors for the behaviour.

Management If the behaviours occurinfrequently, then no specific treatment isusually necessary apart from simple safetyprecautions in the bedroom and house. Ifthe behaviours are disrupting relationships,potentially dangerous, or occur veryfrequently, then a number of medicationscan be trialled. Clonazepam and thebenzodiazepine class of drugs are generallyfirst line, followed by tricyclic antidepressants.Psychotherapy, progressive relaxation,hypnosis and cognitive behavioural therapyare often employed in long-term managementwith less drug-responsive cases, or thoseparticularly associated with stress etc. Referralto a specialist sleep unit is recommended inmore difficult cases to establish diagnosisand assist with management.

Nocturnal seizures

After stroke and Alzheimer’s disease,epilepsy is one of the most commonlyoccurring neurological conditions inindustrialised countries.

Epilepsy is the term applied to a number ofseizure disorders which are characterised byabnormal, uncontrolled electrical dischargesin the brain manifesting as motor, verbal orexperiential phenomena, which areundesirable and often harmful.

The importance of epilepsy in the context ofsleep medicine is the recognition that it is

common and can co-exist with or evenmimic sleep disorders. Sleep deprivationcan increase the incidence of seizure activityand, conversely, seizures can affect thesleep/wake cycle.

There is increasing evidence that OSAco-exists in epilepsy: in 10% of unselectedadults with epilepsy, 20% of children withepilepsy and 30% of drug-resistant epilepsypatients. Seizures and parasomnias can co-exist in the same patients. Arousal disorderscan co-exist with frontal lobe epilepsy,making diagnosis difficult. Nocturnal frontallobe epilepsy, which can occur exclusively insleep, manifests with bizarre/dramaticbehaviours, which if not recognised (oftenthe EEG is unhelpful) can make treatmentand management difficult (Derry et al., 2009).

RBD has been documented as a comorbidityin up to 12% of elderly epilepsy patients.

Management A careful history withcollateral allows for differentiation ofpossible diagnoses. This should be followedby video-PSG employing a full scalpmontage. Multiple studies may be necessaryif events are infrequent and sometimescontinuous video-EEG telemetry may benecessary. Treatment regimens are variableand related to the type of seizure disorderencountered. Referral to a specialist sleepunit with experience in epileptology isrecommended to establish diagnosis andassist with management.

Periodic leg movement disorder and‘restless legs’ syndrome

Restless legs syndrome (RLS) is asensorimotor-related sleep disorder, highlyprevalent but potentially under-recognisedthroughout the world.

RLS is an unpleasant sensation in the legs,but can also affect the arms, occurring duringthe day but usually worsening at night.

Diagnosis The essential diagnostic criteriafor RLS were established in 2002 by theInternational RLS Study Group (IRLSSG)and comprise the following: 1) an urge tomove the legs, often accompanied by anunpleasant sensation in the legs or other


body parts; 2) symptoms aggravated by rest;3) symptoms alleviated by movement; and 4)symptoms worse in the evening or night withan urge to move the legs, usuallyaccompanied by an unpleasant sensation inthe legs (other body parts may be involved, inaddition to the legs). The differentialdiagnosis includes periodic limb movementdisorder, positional discomfort at night,nocturnal leg cramps, arthritis, akathisia, andperipheral neuropathy and anxiety disorders.

Classification and epidemiology RLS isclassified into two main types: 1) primaryRLS, which usually occurs earlier in life witha peak incidence at age 20 yrs, having astrong genetic component (40–60% of allpatients with RLS have a positive familyhistory); 2) secondary RLS, which developslater in life and is associated with a medicaldisorder, such as diabetes mellitus and iron-deficiency anaemia. The general populationprevalence of RLS ranges from 1–15%. MostRLS studies, including the large geneticstudies, do not have PSG confirmation ofRLS (manifested by periodic limb movements)and bias is inherent in changes in diagnosticcriteria over the years (Yeh et al., 2011).

Pathophysiology There is strong evidence fora genetic component in primary RLS. Severallarge genome-wide association studies haveshown an increased frequency of varioussusceptibility loci which, however, have notbeen confirmed to be of functionalsignificance to date. Some evidence pointsto dopaminergic mechanisms and potentialcentral nervous system iron deficiency. Riskfactors for development of RLS include:pregnancy, low iron levels, poor overallhealth, increased age, PD and end-stagerenal disease.

Management Sleep studies are not alwaysnecessary if the history is clear. Full ironstudies should always be performed.Maintaining a ferritin level of .50 mg?L-1

may abate symptoms secondary to irondeficiency. Severity of symptoms can beassessed using the IRLSSG rating scale (10-point scale). If PSG is performed, patientsare found to have an elevated periodic limbmovement index. Periodic limb movementsare rhythmic extensions of the big toe and

dorsiflexion of the ankle with occasionalflexions of the knee and hip. Periodic limbmovements are scored only if they are partof a series of four or more consecutivemovements lasting 0.5–5 s with an intervalof 4–90 s. A periodic limb movement indexof greater than 5 per hour is consideredabnormal but the significance outwithsymptoms of RLS is debatable.

RLS can be very difficult to treat,augmentation of symptoms can occur aswell as tachyphylaxis, requiring alteration ofdrug regimens. A combination ofpharmacological treatments is often a usefulstrategy. Dopamine agonists are thetreatment of choice. Benzodiazepines,anticonvulsants, dopaminergic agents,bupropion and folate and ironsupplementation can also be effective.Nonpharmacological methods such aslifestyle changes are potentially usefuladjuncts, e.g. cessation of caffeine intake.

Further reading

N Aurora RN, et al. (2010). Best practiceguide for the treatment of REM sleepbehavior disorder (RBD). J Clin SleepMed; 6: 85–95.

N Bassetti C, et al. (2000). SPECT duringsleepwalking. Lancet; 356: 484–485.

N Boeve BF. (2010). REM sleep behaviordisorder: updated review of the corefeatures, the REM sleep behaviordisorder-neurodegenerative disease asso-ciation, evolving concepts, controversies,and future directions. Ann N Y Acad Sci;1184: 15–54.

N Derry CP, et al. (2009). NREM arousalparasomnias and their distinction fromnocturnal frontal lobe epilepsy: a videoEEG analysis. Sleep; 32: 1637–1644.

N Hublin C, et al. (2003). Genetic aspectsand genetic epidemiology of parasom-nias. Sleep Med Rev; 7: 413–421.


N Kemlink D, et al. (2009). Replication ofrestless legs syndrome loci in three Euro-pean populations. J Med Genet; 46: 315–318.


N Kunz D, et al. (2010). A two-part, double-blind, placebo-controlled trial of exogen-ous melatonin in REM sleep behaviourdisorder. J Sleep Res; 19: 591–596.

N Mahowald MW, et al. (2007). Pathophy-siologic mechanisms in REM sleep beha-vior disorder. Curr Neurol Neurosci Rep; 7:167–172.

N Manni R, et al. (2010). Comorbiditybetween epilepsy and sleep disorders.Epilepsy Res; 90: 171–177.

N Oudiette D, et al. (2009). Dreamlike

mentations during sleepwalking and sleep

terrors in adults. Sleep; 32: 1621–1627.

N Siclari F, et al. (2010). Violence in sleep.

Brain; 133: 3494–3509.

N Yeh P, et al. (2011). Restless legs

syndrome: a comprehensive overview on

its epidemiology, risk factors, and treat-

ment. Sleep Breath; [Epub ahead of print

doi: 10.1007/s11325-011-0606-x].


Hypersomnia

Renata L. Riha

Hypersomnias of central origin comprise anumber of entities which are listed intable 1. The best studied, and in somerespects defined, primary hypersomnia isnarcolepsy with or without cataplexy.Secondary causes for hypersomnia shouldalways be excluded prior to making adiagnosis of primary hypersomnia, whichincludes narcolepsy with or withoutcataplexy, recurrent hypersomnia, Kleine–Levin syndrome and idiopathichypersomnia. Psychiatric and organicdisorders, including head injury, can alsolead to excessive daytime sleepiness. Withrare disorders, referral to a specialised sleepclinic for confirmation of diagnosis andmanagement advice is always advisable.

Narcolepsy

Presenting features The term narcolepsy wasfirst used in 1880 by Dr Jean-Baptiste-Edouard Gelineau, to describe a patient withsleep attacks, excessive daytime sleepinessand episodes of muscle weakness triggeredby emotions. Table 2 outlines the classicalsymptoms of narcolepsy.

Narcolepsy has a prevalence of 25–50 per100,000 people with no sex preponderance.The age of onset is variable, from earlychildhood to mid-50s. Epidemiologicalstudies have shown two peaks of onset atages 15 and 36 yrs (Longstreth et al., 2007).It is often unrecognised and many patientsmay be labelled with other neurological orpsychiatric disorders before a diagnosis ismade, causing a delay in diagnosis of up to6 yrs from the onset of symptoms(Longstreth et al., 2007).

Narcolepsy can exist either with or withoutcataplexy, or as a consequence of a medicalcondition (secondary narcolepsy) (table 3).

Cataplexy is virtually pathognomonic of thedisease, with all voluntary skeletal musclesbeing affected. However, presentation canrange from only slight weakness of the arms,legs or facial muscles to severe episodes ofweakness. Most patients will be aware ofsymptom onset and can generally adopt asafe posture at the onset of the attack,making injury rare. Other features ofnarcolepsy to be aware of, in addition tothose listed in table 2, are an increasedtendency of developing REM behaviourdisorders (RBD): an association is seen in7–36% of patients and is higher in patientswith cataplexy (Nightingale et al., 2005).

Pathogenesis A canine model for narcolepsywas first described in 1973 and has beenused in narcolepsy research subsequently.In Doberman Pinschers and LabradorRetrievers, the condition is autosomalrecessive with complete penetrance. Thisobservation led to the discovery of amutation in the gene coding for thehypocretin-2 receptor, suggesting thathypocretin may be involved in the diseasepathogenesis (Lin et al., 1999).

Hypocretin Hypocretins (also known asorexins) were described only in 1998 (deLecea et al., 1998). Two forms of hypocretinare currently known to exist (hypocretin-1and -2), produced by a specific set ofneurons in the lateral hypothalamus thatproject to the olfactory bulb, cortex,thalamus, hypothalamus and parts of the


brainstem. Projections in the brainstem, tothe laterodorsal and pedunculopontinetegmental nuclei, and pontine reticularformations are considered important forsleep regulation. Recent studies suggestthat hypocretin is important to maintainvigilance, by exerting excitatory effects onthe monoaminergic and cholinergic systemsin the cortex and hypothalamus (Tsujinoet al., 2009). An imbalance in both of thesesystems may lead to narcolepsy. Bothanimal and human studies have supportedthis hypothesis. Cerebrospinal fluid (CSF)studies of patients with narcolepsy show lowor undetectable levels of hypocretin, andpost mortem studies have identifiedhypocretin-ligand deficiency in thenarcoleptic brain (Peyron et al., 2000). Inaddition, patients with narcolepsydemonstrate reductions in their baselineenergy homeostasis with resulting obesityand type-II diabetes mellitus, both partlyregulated by hypocretin (Nishino et al.,2001). This suggests that the metabolicchanges are directly linked to thepathophysiology of the condition, ratherthan being related to excessive sleepinessor inactivity.

One important clinical marker fornarcolepsy is the presence of altered humanleukocyte antigen (HLA) gene complexes.For example, one specific marker (HLADQB1*0602) is found in 90% of patients

with narcolepsy with cataplexy, but is onlypresent in 40% with narcolepsy alone(Mignot, 1998). The clinical relevance ofthese abnormalities is unclear, with variableassociations between ethnic groups andsubtypes of narcolepsy. The evidence for anautoimmune process is currently notdefinite, although case reports of narcolepsycommencing after streptococcal infectionsor vaccinations are highly suggestive.

The pathogenesis of cataplexy remainsunclear at the present time.

Diagnosis Although sleepiness is difficult todefine objectively, a clear history ofsubjective sleepiness with an eyewitnesswhere possible (for the cataplexy), willgreatly assist with diagnosis.

A note of caution! Many cases have beendescribed of non-organic, or ‘‘pseudo’’narcolepsy, often in association withdrug-seeking behaviour. The InternationalClassification of Sleep Disorders 2ndEdition (ICSD2) criteria suggest thatnarcolepsy with cataplexy ‘‘should, whereverpossible’’ be confirmed with PSG andMSLTs. In the absence of cataplexy, a PSGand MSLT are mandatory to confirm thediagnosis of narcolepsy.

Recent guidelines (Littner et al., 2005;Billiard et al., 2006) support the use ofobjective and subjective monitoring of sleep

Table 1. Hypersomnias of central origin (ICSD II classification).

Primary hypersomnias Secondary hypersomnias

Narcolepsy with cataplexy Narcolepsy due to medical conditions e.g. headinjury

Narcolepsy without cataplexy Menstruation-related hypersomnia

Idiopathic hypersomnia with long sleep time Behaviourally-induced insufficient sleepsyndrome

Idiopathic hypersomnia without long sleeptime

Hypersomnia due to medical condition, drug orsubstance

Recurrent hypersomnia Physiological (organic) hypersomnia, nototherwise specified

Kleine–Levin syndrome (malespredominantly)

Hypersomnia not due to substance or knownphysiological condition


duration at least 1 week prior to in-labinvestigation (using sleep diary andactigraphy), followed by an overnight PSGprior to daytime testing using the MSLT.Sleep deprivation and medications canadversely influence the outcome of the MSLT.

Overnight PSG may demonstrate featuresconsistent with narcolepsy, including shortREM latency, sleep fragmentation,increased stage 1 sleep and excessive slow-wave sleep towards the end of the night.Urinalysis for drugs commonly abused isalso helpful, as these drugs can alsoinfluence the outcome of the MSLT andovernight PSG.

On the MSLT, a sleep latency ,8 min withtwo sleep onset in REM periods (SOREMPs)(REM periods within 15 min of sleep onset)is considered to be consistent withnarcolepsy (fig. 1) (Billiard et al., 2006).However, recent studies have shown that15% of cataplexy patients will not haveSOREMPs or shortened sleep latency onMSLT (effectively a negative test) (Mignotet al., 2002), while patients with OSA maypresent with MSLT data that is consistentwith narcolepsy (Aldrich et al., 1997).

Table 2. Diagnostic criteria for narcolepsy with cataplexy#.

Symptoms present for at least 3 months

Excessive daytime sleepiness

Repeated naps or lapses into sleep of short duration (,1 h) occurring almost daily

Refreshing sleep with recurrent sleepiness within 1–2 h

Sleep attacks in socially unacceptable situations

History of cataplexy

Sudden loss of muscle tone provoked by strong emotion

Short duration lasting seconds to minutes

Immediate and complete recovery

No impairment of consciousness or memory and no respiratory or ocular muscle involvement

Associated features (not always present)

Sleep paralysis

Hypnagogic hallucinations

Automatic behaviour

Nocturnal sleep disruption

Hypersomnia not explained by another sleep disorder, psychiatric disorder or medical illness

Investigations

PSG and MSLT results: mean sleep latency of f8 min with o2 SOREMPs following at least6 h of sleep prior to the test

CSF hypocretin-1 levels f110 ng?L-1, or less than one-third reference range

SOREMPs: sleep onset in REM periods; CSF: cerebrospinal fluid. #: adapted from ICSD-2.

Table 3. Common causes of secondary narcolepsy.

Lesions affecting the hypothalamus

Head trauma

Multiple sclerosis

Myotonic dystrophy

Parkinson’s disease

Prader–Willi syndrome

Paraneoplastic encephalitis

Niemann–Pick type C disease

Sarcoidosis

Coffin–Lowry syndrome


HLA DQB1*0602 analysis can also beundertaken despite its low specificity(Dauvilliers et al., 2007). CSF hypocretinlevels can also be checked; low CSFhypocretin levels (,110 pg?mL-1 or one-thirdof the mean control value) are consistentwith narcolepsy. CSF hypocretin levels arenever this low in patients without narcolepsyand cataplexy (Dauvilliers et al., 2007).

Cataplexy cannot at present be measuredobjectively and the diagnosis is largelybased on eyewitness history. Unfortunately,narcolepsy without cataplexy often does nothave hypocretin deficiency and this canconfuse the diagnosis (Dauvilliers et al.,2003). Other causes of hypersomnia shouldbe considered, including OSAS, idiopathichypersomnia, recurrent hypersomnia,insufficient sleep syndrome andhypersomnia associated with depression.

Management Stimulants remain thecornerstone for treating the excessivedaytime sleepiness of narcolepsy and aresummarised in table 4. Although there islittle evidence supporting their use,antidepressants are used for the treatmentof cataplexy. In recent trials, sodium oxybatehas also been shown to improve cataplexy insome patients.

Importantly, all anti-narcolepsy medications,while improving the symptoms of sleepinessdo not necessarily return the patient to fulland sustained alertness.

Scheduled naps to maintain wakefulness areoften promulgated as a form of treatment,but reviews suggest that this has onlylimited use in patients who remain sleepydespite stimulants (Wise et al., 2007). Iftolerance develops, a drug holiday with

REM

Nap 4 15:19:35

W1234

REM

Nap 3 13:21:35

W1234

REM

Nap 2 11:18:05

W1234

REM

Nap 1 09:22:35 09:43:05

11:38:35

13:42:05

15:40:05

W1234

Figure 1. MSLT showing four sleep onset REM periods with a multiple sleep latency of 0.8 min. Black:REM sleep; green: N2 sleep; yellow: N1 sleep.


recommencement at a lower stimulant doseis effective.

Safety and efficacy profiles should beconsidered in detail while prescribing alldrugs for narcolepsy and cataplexy and thepatient kept under review.

The issues of pregnancy, use of stimulantsin the context of comorbidities such aspsychoses and hypertension, and risksrelated to driving and certain forms ofemployment, pose challenges which benefitfrom specialised input. There is also anincreased incidence of psychological andpsychiatric comorbidity in narcolepsyusually related to diagnostic delay,hypersomnolence/cataplexy if severe andoccasionally to medications used inits treatment.

Idiopathic hypersomnia

Bedrich Roth, a Professor of Neurology atPrague, was among the first to describe

idiopathic hypersomnia over 40 yrs ago,noting the features to be excessive daytimesomnolence unaccompanied by cataplexy(Roth, 1976). A characteristic of this disorderis ‘sleep drunkenness’ or difficulty achievingfull alertness during wakefulness despiteadequate sleep.

The ICSD-2 classification distinguishesbetween idiopathic hypersomnia with andwithout long sleep time.

Idiopathic hypersomnia with long sleep timeis characterised by complaints of excessivesleepiness occurring daily for at least3 months. Patients have prolonged sleep ofmore than 10 h, which is subjectively andobjectively documented. Waking up at theend of this period is difficult and naps arenot refreshing.

There are no secondary causes to explain thehypersomnia, including psychiatric, lifestyleand medical conditions (Billiard, 2007).

Table 4. Commonly prescribed treatment for narcolepsy and cataplexy.

Pharmacological compound Proposed mode of action

Stimulants

Amphetamine Increases monoamine release (dopamine,noradrenaline, serotonin); blocks monoaminere-uptake and monoamine oxidase at highdoses

Methylphenidate Blocks monoamine uptake at doses lower thanamphetamine

Modafinil Low binding affinity for dopamine receptor;mode of action currently unknown

Other

Sodium oxybate (salt of GHB) Possibly acts via GABA-b or via specific GHBreceptors; decreases dopamine release

Anti-cataplectic medication Antidepressants

Venlafaxine Noradrenaline and serotonin selective re-uptake inhibitor

Fluoxetine Selective serotonin re-uptake inhibitor (highdoses sometimes required)

Imipramine Monoaminergic uptake inhibitors, less selectivefor noradrenaline, dopamine and serotonin;anitcholinergic effects; active metabolites

Clomipramine

Protryptiline

GABA: c-aminobutyric acid; GHB: c-hydroxybutyrate.


Idiopathic hypersomnia without long sleeptime likewise needs to be present for at least3 months on presentation, but the patientreports normal nocturnal sleep (.6 h but,10 h) objectively and subjectivelydocumented. There is generally no difficultyin waking up in the morning (Billiard, 2007).

In both forms of idiopathic hypersomnia,nocturnal PSG is unremarkable for any othersleep-related pathology. An MSLTperformed after adequate PSG is considereddiagnostic for the condition if the meansleep latency is ,8 min and there are fewerthan two SOREMPs (Billiard, 2007).

The epidemiology of true idiopathichypersomnia is unknown; it presentsgenerally at a younger age and it isconsidered to be rare. Many cohortspublished in the literature have not takeninto account all possible comorbidities,including psychiatric and organic, and somestudies have failed to convincingly excludecircadian rhythm disturbances.

In contrast to narcolepsy, there are generallyno laboratory abnormalities associated withidiopathic hypersomnia and there isgenerally no family history. Recent studieshave suggested a genetic predisposition(Trakada et al., 2003). Treatment comprisesstimulants, as for narcolepsy, but responseto medication is generally variable and notalways satisfactory. Management has to betailored to the individual.

Further reading

N Aldrich MS, et al. (1997). Value of themultiple sleep latency test (MSLT) forthe diagnosis of narcolepsy. Sleep; 20:620–629.

N Billiard M, et al. (2006). EFNS guidelineson management of narcolepsy. Eur JNeurol; 13: 1035–1048.

N Billiard M. (2007). Diagnosis of narco-lepsy and idiopathic hypersomnia. Anupdate based on the International classi-fication of sleep disorders, 2nd edition.Sleep Med Rev; 11: 377–388.

N Dauvilliers Y, et al. (2003). CSF hypocretin-1 levels in narcolepsy, Kleine-Levin

syndrome, and other hypersomniasand neurological conditions. J NeurolNeurosurg Psychiatry; 74: 1667–1673.

N Dauvilliers Y, et al. (2007). Narcolepsywith cataplexy. Lancet; 369: 499–511.

N de Lecea L, et al. (1998). The hypocretins:hypothalamus-specific peptides with neu-roexcitatory activity. Proc Natl Acad SciUSA; 95: 322–327.

N American Academy of Sleep Medicine(2005). International Classification ofSleep Disorders. 2nd Edn. Westchester,American Academy of Sleep Medicine.

N Lin L, et al. (1999). The sleep disordercanine narcolepsy is caused by a muta-tion in the hypocretin (orexin) receptor 2gene. Cell; 98: 365–376.

N Littner MR, et al. (2005). Practice para-meters for clinical use of the multiplesleep latency test and the maintenance ofwakefulness test. Sleep; 28: 113–121.

N Longstreth WT Jr, et al. (2007). Theepidemi-ology of narcolepsy. Sleep; 30: 13–26.

N Mignot E, et al. (2002). The role ofcerebrospinal fluid hypocretin measure-ment in the diagnosis of narcolepsy andother hypersomnias. Arch Neurol; 59:1553–1562.

N Mignot E. (1998). Genetic and familialaspects of narcolepsy. Neurology; 50:Suppl. 1, S16–S22.

N Morrison I, et al. (2011). Diagnosingnarcolepsy with cataplexy on historyalone: challenging the InternationalClassification of Sleep Disorders (ICSD-2) criteria. Eur J Neurol; 18: 1017–1020.

N Nevsimalova S. (2009). Narcolepsy inchildhood. Sleep Med Rev; 13: 169–180.

N Nightingale S, et al. (2005). The associa-tion between narcolepsy and REMbehavior disorder (RBD). Sleep Med; 6:253–258.

N Nishino S, et al. (2000). Hypocretin(orexin) deficiency in human narcolepsy.Lancet; 355: 39–40.

N Nishino S, et al. (2001). Low cerebrosp-inal fluid hypocretin (Orexin) and alteredenergy homeostasis in human narco-lepsy. Ann Neurol; 50: 381–388.

N Peyron C, et al. (2000). A mutation in acase of early onset narcolepsy and ageneralized absence of hypocretin pep-tides in human narcoleptic brains. NatMed; 6: 991–997.


N Roth B. (1976). Narcolepsy and hyper-somnia: review and classification of 642personally observed cases. Schweiz ArchNeurol Neurochir Psych; 119: 31–41.

N Trakada G, et al. (2003). Family aspectsof idiopathic hypersomnia (polysympto-matic form). Sleep; 26: Suppl. A350.

N Tsujino N, et al. (2009). Orexin/Hypocretin: a neuropeptide at the inter-face of sleep, energy homeostasis, andreward system. Pharmacol Rev; 61: 162–176.

N Wise MS, et al. (2007). Treatment ofnarcolepsy and other hypersomnias ofcentral origin. Sleep; 30: 1712–1727.

N Xyrem International Study Group. (2005).A double-blind, placebo-controlled studydemonstrates sodium oxybate is effectivefor the treatment of excessive daytimesleepiness in narcolepsy. J Clin Sleep Med;1: 391–397.

N Young TJ, et al. (2006). Hypersomnias ofcentral origin. Chest; 130: 913–920.


Circadian rhythm disorders

Renata L. Riha

Circadian rhythmicity has beendemonstrated in all human cells andregulates integral physiological processessuch as core body temperature andhormone secretion, including cortisol,prolactin and growth hormone (Sack et al.,2007a). There is significant interindividualvariation as a result of age, sex andmorningness/eveningness preference. Thelatter reflects an individual’s preferences fortiming activity and correlates with timing ofthe individual’s circadian pacemaker.Morning types, in contrast to evening types,schedule sleep earlier and experience earlierpeaks in alertness and performance duringthe ‘24-h period’ (Sack et al., 2007a; Hofstraet al., 2008).

The differences between morningness andeveningness are partly the result ofgenotype. Several genes have been identifiedin animal studies affecting the expression ofautoregulatory translation–transcriptionfeedback loops, including the Period genes

(PER1, PER2 and PER3), the CLOCK geneand two cryptochrome genes (CRY1 andCRY2). Deletion or mutation of these genesleads to abnormalities in circadian rhythmwhen tested under constant conditions (Dijket al., 2009, 2010; Viola et al. 2007; Wulffet al., 2009). Several human families havebeen identified with specific abnormalitiesin these genes affecting circadianrhythmicity. However, there is currently noevidence to support routine genetic testingto establish the basis of these disorders ingeneral clinical practice (Sack et al., 2007a;Hofstra et al., 2008; Sack et al., 1992;Morgenthaler et al., 2007).

Recent, exhaustive reviews and guidelineshave been published by the AmericanAssociation of Sleep Medicine (AASM) onthe use of actigraphy and other tools forevaluating the various circadian rhythmdisorders (summarised in table 1)(Morgenthaler et al., 2007).

There is insufficient evidence currently torecommend routine measurement ofcircadian rhythm utilising tools other thanactigraphy and sleep logs/diaries in themajority of disorders. However, dim lightmelatonin onset (DLMO) and core bodytemperature have proven useful indiagnosing and assessing free runningcircadian disorder (FRD), which affects upto 50% of blind individuals (Sack et al.,2007b; Skene et al., 2007).

Recommendations published on the therapyof circadian disorders highlight the fact thatthere is still much to be learned about theintrinsic pathophysiology governing themendogenously, and the degree to whichexogenous factors determine their onset,

Key points

N DSPS is typified by an exaggerated‘night owl’ pattern of sleep and wake-fulness; ASPS is the opposite of DSPS.

N FRD is common in blind individuals.

N In irregular sleep/wake syndrome,there is an absence of a well-definedsleep/wake cycle and there is no majorsleep period.

N Shift work disorder is common in 24-hsocieties and can lead to significantmorbidity.


Table 1. Summary of evaluation tools and indications for their use in circadian rhythm disorders.

Evaluation tool Data collection method Factors masking signal Application

DLMO 1) Melatonin concentrationin plasma or saliva

2) 6-sulfatoxymelatonin inurine

3) Mid-sleep melatonin(supplementary)

4) Must be collected in dimlight (50 lx)

Posture; exercise; sleep;sleep deprivation; caffeine;medications e.g. NSAIDs

and b-blockers; age?

Useful in FRD

CBT Thermometer Posture; physical activity;meals; ambient

temperature; sound;humidity; bright light; age

Useful in FRD

Cortisol levels Cortisol concentration inserum and saliva

Stress; light; ageing; high-protein meals; sleep

architecture

Useful in FRD

Questionnaires 1) MEQ2) CTI

3) MCTQ

MEQ assesses preferencesof timing of activity, and

does not assess work andweekend/holidays

separately; MCTQ collectsinformation on actual

timing of daily sleep andactivities

Useful to assessmorningness/

eveningness butnot routinely

recommendedfor diagnosis

Actigraphy Actimetry sensor worn onnondominant wrist tomeasure gross motor

activity; increased movementduring wakefulness and

reduced movement duringsleep

Must be worn continuouslyfor o5 days

Can be taken off; lessreliable if sleep is

fragmented e.g. limbmovement disorders,

parasomnias, insomnia;restless legs syndrome

Indicatedroutinely to

diagnose andassess response

to therapy inDSPS, ASPS;

optional use inISWS, FRD,SWD, JLD

Sleep log/sleep diary

A variety of sleep diaries areavailable

Failure to complete diaryon daily basis in

accordance with actualsleep/wake times

Indicatedroutinely to

assess presenceof and response

to therapy inISWS, DSPS,ASPS, FRD,SWD, JLD

PSG Electrophysiological dataderived from EEG, EMG,

EOG, ECG, pulse oximetry,markers of respiratory effort

First night effect;insufficient sleep; technical

issues

Indicated toexclude other

concurrent sleepdisorders e.g.

sleep apnoea ifsuspected onclinical history

NSAID: nonsteroidal anti-inflammatory drug; CBT: core body temperature; MEQ: Morningness–Eveningness

Questionnaire; CTI: Circadian Type Inventory; MCTQ: Munich Chronotype Questionnaire; ISWS: irregular

sleep/wake syndrome; SWD: shift work disorder; JLD: jet lag disorder.


course and response to treatment (Sacket al., 1992, 2007a). All circadian rhythmdisturbances are at least partially, if not fully,amenable to planned sleep schedules, timedlight exposure (except the unsighted) andtimed melatonin administration (Sack et al.,1992, 2007a).

Sleep/wake rhythm disorders

Delayed sleep phase syndrome Delayed sleepphase syndrome (DSPS) is typified by anexaggerated ‘night owl’ pattern of sleep andwakefulness. Sleep onset and waking timesare delayed by 3–6 h compared withconventional sleep/wake times. The personwith DSPS is generally sleepy and ready togo to bed between 02:00 and 06:00 h andgenerally gets up at 10:00–13:00 h. Sleepitself is normal. The sleep/wake patternmust be present for o3 months for thediagnosis to be made. Exclusion ofdepression and other medical andpsychiatric conditions is important to thediagnosis. DSPS can occur in children andwell into old age, being most common inadolescents and young adults. About 7% ofyoung adults and adolescents are estimatedto have DSPS (Sack et al., 1992;Morgenthaler et al., 2007). Presentingcomplaints are generally of sleep-onsetinsomnia (delay in falling asleep into theearly hours of the morning) and severedifficulty rousing in the morning at sociallyappropriate times (e.g. for school andemployment). Sufferers may complain ofdaytime tiredness, which may lead toimpairment of school and job performance.Tiredness is greatest in the morning, withincreasing alertness during the day.Behaviourally induced DSPS is a subtypewhich is frequently seen, particularly inyounger patients.

A combination of factors most likely resultsin DSPS. There are genetic factors, asdiscussed above, associated with thecircadian clock genes that consitute ahereditary component in some cases. Subtleproblems in the regulation of circadianrhythm and with sleep recovery ability mayexist. Behaviour plays a strong role inreinforcing DSPS.

A combination of techniques is generallyused to make the diagnosis, includingactigraphy, PSG, sleep diaries and a historyof the complaint (table 1). A number ofdifferent methods can be used to treatDSPS; these include bright light therapy,changing bedtimes gradually(chronotherapy), use of melatonin andreinforcing behavioural changes. Specialistsupervision should be available at all timesto assist with these treatments.

Advanced sleep phase syndrome Advancedsleep phase syndrome (ASPS) is theopposite of DSPS. Sleep is normal but thesleep/wake schedule runs ,3 h earlier thanaverage. Irresistible sleepiness occurs in thelate afternoon and bedtime becomes earlyevening. Waking time is generally between02:00 and 05:00 h. The causes, diagnosisand treatment of ASPS are similar to thoseof DSPS but targeted at different parts of theday and circadian rhythm. About 1% ofmiddle-aged adults have ASPS but it isgenerally not considered such a sociallyincapacitating problem as DSPS (Hofstraet al., 2008). People with ASPS oftengravitate towards jobs that suit their extremeearly morning habits.

Free-running circadian disorder FRD isthought to result from a circadianpacemaker that has no stable phaserelationship to the 24-h cycle (fig. 1). Theproblem is rare in sighted people and,generally, will result from DSPS that has‘escaped’ any form of entrainment and isoften behaviourally induced. This disorder iscommon in the totally blind with somestudies showing 50% of subjects havingfree-running melatonin rhythms (Skene et al.,2007). Investigation is as for other circadiansleep disorders. Long-term melatoninadministration is the treatment of choice inthe totally blind (Sack et al., 2007b) and maybe supplemented by bright light therapy insighted individuals. Behavioural andpsychological factors predisposing to ormaintaining FRD should be addressed andmanaged appropriately.

Irregular sleep/wake syndrome Irregularsleep/wake syndrome is a rare disorder,commonly occurring in association with


neurological dysfunction,institutionalisation (e.g. for dementia),schizoaffective disorders and psychomotorretardation. There is an absence of a well-defined sleep/wake cycle and there is nomajor sleep period. Patients generallycomplain of insomnia or excessivesleepiness in association with irregular sleepbouts/naps. The total sleep time over the 24-h period is generally appropriate to theperson’s age and requirements. Diagnosisand treatment are similar to those for othercircadian rhythm disorders, using acombination of pharmacological, light-related and behavioural strategies.Outcomes are variable (Sack et al., 1992;Morgenthaler et al., 2007).

Shift work and sleep

Sleep and other health problems arisingfrom working schedules outside normaldaylight hours can result in shift workdisorder. Shift work problems result fromthe interaction of social/domestic factors,circadian rhythm factors and sleep factors.This disorder is common in every societywhere the service or production sector of theeconomy needs to be maintained 24 h?day-1.Although many individuals seem to copewith shift work, people most vulnerable todeveloping problems include: those aged.50 yrs; those with a history of sleepdisorders (e.g. sleep apnoea andnarcolepsy); those with epilepsy, diabetes or

Thu 06 NovFri 07 NovSat 08 NovSun 09 Nov

00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00

Mon 10 NovTue 11 NovWed 12 NovThu 13 NovFri 14 NovSat 15 NovSun 16 NovMon 17 NovTue 18 NovWed 19 NovThu 20 NovFri 21 NovSat 22 NovSun 23 NovMon 24 NovTue 25 NovWed 26 NovThu 27 NovFri 28 NovSat 29 NovSun 30 NovMon 1 DecTue 2 DecWed 3 DecThu 4 DecFri 5 DecSat 6 DecSun 7 DecMon 8 DecTue 9 DecWed 10 DecThu 11 DecFri 12 DecSat 13 Dec

Mon 15 DecSun 14 Dec

Tue 16 Dec

Figure 1. Actigraphy showing FRD.


heart disease; and those with psychiatricillness, digestive problems, or a history ofalcohol or substance misuse. Other factorsimpairing tolerance to shift work include aheavy domestic work load, working two ormore jobs and ‘morning-type’ orientation(Sack et al., 1992; Morgenthaler et al., 2007).

Management Drawing up an appropriateshift work roster is crucial to maintainingnot only optimal performance but also inminimising problems of coping with shifts.Types of shifts detrimental to well-beinginclude: 12-h shifts that involve criticalmonitoring tasks, heavy physical labour orexposure to harmful substances; more thanfour 12-h shifts in a row; first shift startingearlier than 07:00 h; split shifts with tooshort breaks between shifts; complicatedschedules making planning ahead difficult;weekly rotations; more than five late nightshifts in a row without a break; andbackward rotating hours (Sack et al., 2007a;Morgenthaler et al., 2007).

Impact on overall health Shift workers havean increased risk of heart disease andhypertension, and many have problems withtheir digestive tracts. Disrupted family andsocial roles are also a side-effect. Permanentshift workers have an increased risk ofaffective disorders, particularly depressionand substance misuse/abuse. Their sleep isnever entirely normal.

Sedative and stimulant medication, appliedjudiciously, can sometimes be useful toregulate the sleep/wake pattern. ‘Night owls’tend to do better than ‘morning larks’. Aftermiddle-age, shift work becomes moredifficult to cope with due to changes in sleeppatterns with ageing. Those with seriousmedical conditions should consider leavingshift work (Sack et al., 2007a; Morgenthaleret al., 2007).

Jet lag

The body clock adjusts slowly in response toabrupt changes in environmental time cues.No change is probably more abrupt thanthat of air travel across several time zones.Jet lag comprises a constellation ofsymptoms including: inappropriate daytimetiredness in the new time zone; insomnia;

early awakening; disturbed sleep patterns;and impaired concentration and alertnessduring the day. Other symptoms includeloss of appetite, inappropriate toilet timesand excessive urination at night (Sack et al.,2007a).

Jet lag is worse when travelling eastwards,as the day is shortened to ,24 h. Travellingwestwards has the opposite effect. It iseasier for most people to phase-delay, ratherthan attempt to sleep at an earlier time thanusual (this occurs when one travels east).Travelling westwards generally leads toquicker adaptation to local sleep times. Theseverity of symptoms also depends on otherfactors, including the number of time zonescrossed, the length of the flight and theability of the circadian system to adapt tochanges. Approximately one-third of peoplewho fly do not experience any significanteffects of jet lag.

Treatment Melatonin may, if takenappropriately, assist with decreasing thesymptoms of jet lag. The usual dose trialledis 5 mg. Melatonin can cause reducedalertness and induce sleepiness, thecommonest side-effects being sleepiness,headache and nausea. Strategies forreducing jet lag have been extensivelypublished (Sack et al., 2007a).

Further reading

N Dijk DJ, et al. (2009). Light, sleep, andcircadian rhythms: together again. PLoSBiol; 7: e1000145.

N Dijk DJ, et al. (2010). PERIOD3, circadianphenotypes, and sleep homeostasis.Sleep Med Rev; 14: 151–160.

N Hofstra WA, et al. (2008). How to assesscircadian rhythm in humans: a review ofliterature. Epilepsy Behav; 13: 438–444.

N Horne JA, et al. (1976). A self-assessmentquestionnaire to determine morning-ness–eveningness in human circadiansystems. Int J Chronobiol; 4: 97–110.

N Morgenthaler TI, et al. (2007). Practiceparameters for the clinical evaluation andtreatment of circadian rhythm sleep dis-orders. An American Academy of SleepMedicine report. Sleep; 30: 1445–1459.


N Sack RL, et al. (1992). Circadian rhythmabnormalities in totally blind people:incidence and clinical significance. J ClinEndocrinol Metab; 75: 127–134.

N Sack RL, et al. (2007a). Circadian rhythmsleep disorders: part I, basic principles,shift work and jet lag disorders. AnAmerican Academy of Sleep Medicinereview. Sleep; 30: 1460–1483.

N Sack RL, et al. (2007b). Circadian rhythmsleep disorders: part II, advanced sleepphase disorder, delayed sleep phasedisorder, free-running disorder, and irre-gular sleep-wake rhythm. An American

Academy of Sleep Medicine review. Sleep;

30: 1484–1501.

N Skene DJ, et al. (2007). Circadian rhythm

sleep disorders in the blind and their

treatment with melatonin. Sleep Med; 8:

651–655.

N Viola AU, et al. (2007). PER3 polymorph-

ism predicts sleep structure and waking

performance. Curr Biol; 17: 613–618.

N Wulff K, et al. (2009). Sleep and circadian

rhythm disturbances: multiple genes and

multiple phenotypes. Curr Opin Genet

Dev; 19: 237–246.


Psychiatric aspects of sleep

Chiara Baglioni and Dieter Riemann

Sleep difficulties are a pervasive problem formany patients suffering from psychiatricconditions. This is especially true fordepression and anxiety disorders, but alsoholds for schizophrenia, eating disorders,alcoholism and bipolar disorder. Sleepdifficulties can be present as objectivelymeasured alterations in sleep architecture,as well as insomnia, hypersomnia, delayedsleep phase, reduced sleep need,nightmares and nocturnal panic attacks(Harvey, 2011). The most frequentlyencountered symptoms are, however, those

consistent with the definition of insomnia inthe Diagnostic and Statistical Manual ofMental Disorders (DSM) IV (AmericanPsychiatric Association, 2000) as difficultiesin initiating or maintaining sleep ornonrestorative sleep accompanied bydecreased daytime functioning lasting for atleast 4 weeks. Historically, insomniasymptoms have been conceptualised as aconsequence of psychopathology. However,both clinical and research evidencedemonstrates that this is not the case.Insomnia is now considered as anindependent diagnostic entity, which sharesunderlying psychophysiological factors withmany psychiatric conditions and whichmight play an important role in thecausation and maintenance ofpsychopathology.

Sleep difficulties and depression

Impairment of sleep has been associatedwith affective disorders, and especiallymajor depression. Up to 80% patients withdepression have sleep complaintsconsistent with insomnia. Moreover,symptoms of insomnia often present beforethe first onset of a depressive episode andmay persist into remission and recovery,even after adequate treatment ofdepression. The close link between the twoconditions suggests that they are notrandomly associated. More than 40 studieshave been published evaluating the questionof whether insomnia is a predictor of thedevelopment of depression (reviewed inRiemann, 2009 and Baglioni et al., 2010a).A recent meta-analysis summarisedquantitatively the results of the availablelongitudinal epidemiological studiesbetween 1980 and 2010 (Baglioni et al.,

Key points

N Sleep difficulties consistent with thesymptoms of insomnia are a clinicalpredictor of major depression andmay play an important role in thecausation and maintenance ofpsychopathology in general.

N Sleep is important for the maintenanceof adaptive emotion regulation andreactivity and alterations of theemotional system have been found inpatients with insomnia compared withhealthy controls.

N The addition of psychological treatmentfor insomnia in the standardisedintervention protocols of manypsychiatric conditions could improvethe efficacy of these interventions.

N Chronic insomnia occurs in ,10% ofthe adult population. Psychiatricillness and SDB are the mostcommon comorbidities.


2011), showing that non-depressed peoplewith insomnia have a two-fold increased riskof developing depression, compared withpeople with no sleep difficulties (fig. 1).Subgroup analysis indicates that the effectof insomnia in predicting subsequentdepression is similar in children,adolescents, working-age individuals andthe elderly.

Sleep difficulties and other psychiatricdisorders

In 1992, Benca et al. conducted a meta-analysis of studies that measured sleep in arange of psychiatric disorders: affectivedisorders, anxiety disorders, schizophrenia,borderline personality disorder, eatingdisorders, alcohol abuse, dementia,insomnia and narcolepsy. Data for theanalysis were obtained from studies thatscored sleep EEGs through PSG followingstandard criteria. Sleep variables consideredincluded total sleep time, sleep onset

latency, sleep efficiency, REM latency,lengths in minutes and in percentages ofREM sleep and slow-wave sleep, REMdensity, REM activity and duration of thefirst REM period. The authors found thatalthough sleep alterations were present in allconditions, no impairment in any singlesleep variable had an absolute specificity forany particular psychiatric disorder.Consistently, Harvey (e.g. Harvey et al., 2011)advanced the hypothesis that sleepdisturbances could represent amechanistically transdiagnostic process –that is, a contributing factor to the onset andmaintenance of a range of disorders – in thecontext of psychiatric disorders bycontributing to their development andmaintenance. Such a process can operate atdifferent levels: biological, psychological andemotional, social and contextual. Dataconsistent with the transdiagnostichypothesis include: the reciprocalrelationship between insomnia and

Combined odds ratio

Vollrath et al. 1989Breslau et al. 1996

Weissman et al. 1997

Chang et al. 1997

Foley et al. 1999Mallon et al. 2000

Johnson et al. 2000

Roberts et al. 2002Hein et al. 2003

Perlis et al. 2006

Morphy et al. 2007

Roane et al. 2008Jansson-Fröjmark et al. 2008

Buysse et al. 2008

Cho et al. 2008

Kim et al. 2009

Szklo-Coxe et al. 2010

0.00 2.00 4.00

Odds ratio

6.00

FIGURE 1. Odds ratio of developing depression for people with insomnia versus those with no sleepingdifficulties. (Based on data from Baglioni et al., 2011.)


emotional regulation; genes known to beimportant in the generation and regulationof circadian rhythms that have been linkedto a range of psychiatric disorders; and theinterplay between sleep/circadian biologyand neurotransmitter systems known to beimportant across a range of psychiatricdisorders, such as the dopamine andserotonin systems (see also Wulff et al., 2010).

Sleep and emotions

In recent years, there has been increasedinterest in the relationship between sleepand emotions. Studies on sleep deprivationhave shown enhanced emotionalphysiological responses to negative stimuli,suggesting an important role of sleep for themaintenance of adaptive emotion regulationand reactivity (e.g. Yoo et al., 2007). Withrespect to primary insomnia, recentaetiological theories suggest that peoplepresenting with this disorder arecharacterised by heightenedpsychophysiological arousal, which could bephysiological, cognitive and emotional(reviewed in Riemann et al., 2010).Alterations in the emotional system havebeen described by two patterns of subjectiveexperience of emotions. The cognitivemodel of insomnia (Harvey, 2002) describesthe heightened cognitive activityexperienced by people with insomnia asexcessively negatively toned. Thepsychobiological model of insomnia (Espie,2002) suggests that insomnia ischaracterised by strong positive andnegative emotions. Recent experimentaldata based on self-reported questionnairessuggest that people with insomniaexperience more negative emotions thangood sleepers both in general and close tosleep time (reviewed in Baglioni et al.,2010a). Very few studies have evaluated thephysiological correlates of emotionalprocesses in primary insomnia (e.g. Baglioniet al., 2010b). Moreover, the relationshipbetween positive emotions and sleep is stillnot understood. Norlander et al. (2005;described in Baglioni et al, 2010a) foundthat good sleep quality was associated withintense positive emotions independently ofthe intensity of negative emotions. This

could mean that positive emotions have aprotecting value for sleep, and specificallyfor the subjective perception of it. However,further investigation is needed.

Psychological intervention for insomniacomorbid with other psychiatric disorders

Cognitive behavioural therapy for insomnia(CBT-I) is a multicomponent interventionincluding behavioural, cognitive, andeducational components. Taylor et al.(2007) published a pilot study evaluatingthe efficacy of CBT-I in a sample of 10patients with both depression and insomnia.The study showed that psychologicaltreatment for insomnia was efficacious notonly for sleep symptoms, but that it had anameliorating effect on depressivesymptoms. Consistent with these results,Manber et al. (2008) found that CBT-I in 30patients with symptoms of both insomniaand depression is efficacious andguarantees a better treatment outcome inthis population than standardantidepressive treatment alone. In a recentrandomised controlled trial (Edinger et al.,2009), 81 adults with chronic primaryinsomnia (n540) or comorbid insomniaassociated predominantly with mixedpsychiatric disorders (n541) were assignedto either CBT-I or sleep hygiene education(SHE). CBT-I included strategies directed to:a) educate patients about sleep regulationprocesses (sleep education model); b)correct patients’ misperceptions about sleep(cognitive module); c) associate the bedwith sleep and to reinforce a regular sleep/

wake schedule (stimulus control behaviouralmodule); and d) make the patients stay inbed only for the time she or he slept (sleeprestriction behavioral module). SHEincluded the sleep education model togetherwith the delivery of a list ofrecommendations, for example eliminatingcaffeine and alcohol in the evening orengaging in moderate exercise. CBT-I wasequally efficacious for primary and comorbidinsomnia. These findings, althoughpreliminary, suggest that the addition ofpsychological treatment for insomnia instandardised intervention protocols of manypsychiatric conditions could improve the


efficacy of these interventions. Moreover,considering the longitudinal associationbetween insomnia and depression and thetransdiagnostic hypothesis, treatingsleeping difficulties at an early stage couldinterrupt the sequential process thatgradually reduces the quality of life of peoplewith chronic primary insomnia and ends inthe development of symptoms ofpsychopathology. Therefore, there isincreasing interest in developing newtreatment algorithms for the psychologicaltreatment of insomnia, which are easy andaccessible for the general population. Forinstance, stepped-care models or internetprograms have been proposed and found tobe efficacious (reviewed in Morin et al., 2012).

Further reading

N American Psychiatric Association.(2000). Diagnostic and StatisticalManual of Mental Disorders. 4th Edn,Text Revision. Arlington, AmericanPsychiatric Association.

N Baglioni C, et al. (2011). Insomnia as apredictor of depression: a meta-analyticevaluation of longitudinal epidemio-logical studies. J Affect Disorders; 135:10–19.

N Baglioni C, et al. (2010b). Psychophy-siological reactivity to sleep-related emo-tional stimuli in primary insomnia. BehavRes Ther; 48: 467–475.

N Baglioni C, et al. (2010a). Sleep andemotions: a focus on insomnia. SleepMed Rev; 14: 227–238.

N Benca RM, et al. (1992). Sleep andpsychiatric disorders. A meta-analysis.Arch Gen Psychiat; 49: 651–668.

N Edinger JD, et al. (2009). Cognitivebehavioral therapy for patients withprimary insomnia or insomnia associatedpredominantly with mixed psychiatricdisorders: a randomized clinical trial.Sleep; 32: 499–510.

N Espie CA. (2002). Insomnia: conceptualissues in the development, persistence,and treatment of sleep disorders inadults. Annu Rev Psychol; 53: 215–243.

N Harvey AG, et al. (2011). Sleep distur-bance as transdiagnostic: considerationof neurobiological mechanisms. ClinPsychol Rev; 31: 225–235.

N Harvey AG. (2002). A cognitive model ofinsomnia. Behav Res Ther; 40: 869–893.

N Manber R, et al. (2008). Cognitivebehavioral therapy for insomnia enhancesdepression outcome in patients withcomorbid major depressive disorder andinsomnia. Sleep; 31: 489–495.

N Morin CM, et al. (2012). Chronic insom-nia. Lancet; 379: 1129–1141.

N Riemann D. (2009). Does effective man-agement of sleep disorders reducedepressive symptoms and the risk ofdepression? Drugs; 69: 43–64.

N Riemann D, et al. (2010). The hyperar-ousal model of insomnia: a review of theconcept and its evidence. Sleep Med Rev;14: 19–31.

N Taylor DJ, et al. (2007). A pilot study ofcognitive-behavioral therapy of insomniain people with mild depression. BehavTher; 38: 49–57.

N Wulff K, et al. (2010). Sleep and circadianrhythm disruption in psychiatric andneurodegenerative disease. Nat RevNeurosci; 11: 589–599.

N Yoo SS, et al. (2007). The humanemotional brain without sleep – a pre-frontal amygdala disconnect. Curr Biol; 17:877–878.


Sleep history

Silke Ryan and Walter T. McNicholas

Chronic disturbances in sleep and sleep/wakepatterns are very common, althoughfrequently under-reported by affectedsubjects. In addition, medical and otherhealthcare professionals are generally lessaware of sleep disorders and theirmanagement than of conditions arisingduring wakefulness. The assessment of sleepdisorders requires an understanding ofnormal sleep and how it may alter in differingcircumstances, and also with disorders thataffect the nature of sleep. This knowledgeneeds to guide the patient interview, and athorough sleep and medical history is the firstand most important step in the diagnosis andmanagement of sleep disorders.

A sleep history involves more than adescription of a patient’s sleep and isactually a sleep/wake history with anevaluation of the entire 24-h period. Thisinvolves recording of alertness andtiredness, occupation and leisure hours, inaddition to details of rest and sleep.Particular attention should be paid to the

frequency, type and time of onsets ofsymptoms. A potential relationship toexternal factors, such as environmental,social or medical influences, also needs tobe understood and evaluated, and everyhistory must include details on medication,caffeine, alcohol and illicit drug use.

The first step in obtaining a history is toassess the patient’s complaint or reason forseeking attention, i.e. primary insomnia,excessive daytime sleepiness (EDS) orabnormal breathing behaviour ormovements during sleep. Furthermore, thehistory should establish when the problembegan and whether there was anyrelationship to an external factor.

Detailed questionnaires have beendeveloped to cover important questions thatneed to be asked and that can serve as aroad map for planning the direction of theinterview. It often includes a scale ofalertness, such as the ESS. Completion of asleep diary for 1–2 weeks may giveimportant indications of sleep habits, sleephygiene and daytime symptoms, and isespecially useful in the assessment ofinsomnia or daytime sleepiness of unknowncause. This typically includes the recordingof bedtime, time asleep, nocturnalawakenings, rising time, daytime naps andconsumption of substances that may affectalertness.

The patient often has little or no awarenessof the problem and, therefore, it is highlydesirable to conduct an interview with thepatient’s bed partner or caregiver or, in thecase of a child, with the parent or teacher.The bed partner can provide vitalinformation regarding sleep behaviour,

Key points

N A detailed longitudinal sleep history isthe most critical part of theassessment of subjects with sleepdisorders.

N A sleep history involves assessment ofthe whole 24-h sleep/wake pattern.

N The elicitation of a collateral history isoften crucial in the assessment of asubject with a sleep disorder.


abnormal movements or breathingproblems during sleep, and can also providevaluable independent input about eventsduring wakefulness.

While many components of the sleep historyare common to all sleep complaints, somedisorders require special questions, and thephysician should be constantly formulatingand testing hypotheses of diagnosis as thehistory evolves. One must also bear in mindthat there are often multiple causes of asleep-related problem.

The most important aspects of the sleephistory are detailed below.

Details of the patient’s sleep

It is important to establish the regularity ofthe sleep/wake pattern, including preferredbed and rising time. Irregularity of thispattern, such as with varying work shifts, atweekends or during holidays, should benoted and, where relevant, compared to thepremorbid sleep cycle. Particularly in thecase of insomnia, note should also be takenof the sleep environment. Noise,temperature and brightness of the bedroom,and comfort of the bed can frequentlyinfluence sleep quality, and the patient maybe unaware of these links.

The time of sleep onset should be recordedand if sleep latency is prolonged, potentialreasons should be sought with the patient.Activities before bed and any behaviourwhile awake in bed, such as reading orwatching television, should be described, asthese can affect sleep onset.

Furthermore, problems with sleepmaintenance need to be ascertained. Theseinclude recording of the number ofawakenings with an attempt to determinethe causes of arousals, which could includeexternal factors, such as restless partners,noise, nightmares, or medical causesincluding dyspnoea, leg jerking, nocturia oranxiety. Multiple causes of pain ordiscomfort can also arouse patients,including arthritis, fibromyalgia, restlesslegs or angina. The length of time beforereturning to sleep and an estimation of totaltime spent asleep should be evaluated.

Finally, events at sleep termination areimportant in the overall evaluation. Theseinclude the time of awakening, whether theawakening is spontaneous and how tired orrefreshed the patient feels. Early morningsymptoms, such as headache, dry mouthand sore throat, could point to obstructivesleep apnoea.

Specific nocturnal symptoms

Respiration Snoring is among the mostcommon nocturnal symptoms and,frequently, the primary reason for a patientand partner to seek medical attention. Tohelp distinguish ‘simple’ snoring fromsnoring as part of OSAS, one shouldascertain the intensity, duration andfrequency of snoring, as well as the sleepingposition associated with snoring, in additionto potential association with alcohol orsedative medication. The bed partner shouldbe carefully interviewed about frequency andlength of apnoeas as well as the position inwhich they occur. Apnoeas may also beassociated with jerking movements,suggesting an associated arousal, and thepatient may report episodes of waking with achoking sensation in the throat. Nocturnaldyspnoea can be a symptom of variousdiseases and further details, such aswheeze, association with apnoeas or bodyposition when dyspnoea occurs, may help tonarrow the differential diagnosis.

Movements Abnormal movements duringsleep could potentially point to periodic limbmovement disorder, epileptic seizures orvarious parasomnias. Any history of restlesslegs syndrome should be elucidated.Information about behaviour during dreams,such as shouting, flailing the arms or kickingthe legs, should be sought. If abnormal motorbehaviour is present, the time of events aswell as the occurrence of any injuries to thepatient or others should be noted.

Specific daytime symptoms

Excessive daytime sleepiness Sleepiness is avery common complaint and is experiencedin a number of somatic, psychiatric andprimary sleep disorders, but also occursphysiologically in the absence of sufficientlyrestorative sleep periods. EDS should be


distinguished from mental or physicalfatigue, which usually has an organic causeor may be related to insomnia. EDS severitycan be gauged by frequency of occurrenceand the type of situation in which the patientfalls asleep. EDS is likely to be more severe ifsleep occurs despite stimulatingcircumstances, such as while talking, eatingor on exertion, and if it occurs frequently andat any time during the day. Sleepiness whiledriving should be characterised in terms ofthe time and distance before lapses ofalertness occur, and whether motor vehicleaccidents or near misses have occurred as aconsequence. The duration and frequency ofnaps should be noted and whether or notthey are restorative.

Motor and sensory symptoms Abnormalmovements during the daytime due toepilepsy or a primary movement disordermay be related to unusual movementsduring sleep and, therefore, should bedetermined. Cataplexy, i.e. the suddenbilateral loss of muscle strength due toemotion, especially laughter, is a classicalsymptom of narcolepsy. Any history oftransient paralysis or hallucinations mayfurther support this diagnosis but it shouldbe noted that these manifestations couldalso occur as part of other conditions.

An overview of specific nocturnal anddaytime symptoms and their possibleimplications is provided in table 1.

Further reading

N Chokroverty S. (2010). Overview of sleep &sleep disorders. Ind J Med Res; 131: 126–140.

N Engleman HM, et al. (2004). Sleep. 4:Sleepiness, cognitive function, and qualityof life in obstructive sleep apnoea/hypop-noea syndrome. Thorax; 59: 618–622.

N Falloon K, et al. (2011). The assessmentand management of insomnia in primarycare. BMJ; 342: d2899.

N Ferini-Strambi L. (2007). RLS-like symp-toms: differential diagnosis by historyand clinical assessment. Sleep Med; 8:Suppl. 2, S3–S6.

N Hirshkowitz M. (2004). Normal humansleep: an overview. Med Clin North Am;88: 551–565.

N Johns MW. (1991). A new method formeasuring daytime sleepiness: the Ep-worth sleepiness scale. Sleep; 14: 540–545.

N Jones BE. (2000). Basic mechanisms ofsleep–wake states. In: Kryger MH, et al.,eds. Principles and Practice of Sleepmedicine. Philadelphia, WB Saunders;pp. 134–154.

N Phillips B. (2004). Movement disorders:a sleep specialist’s perspective. Neurology;62: Suppl. 2, S9–S16.

N Schlosshan D, et al. (2004). Sleep. 3:Clinical presentation and diagnosis of theobstructive sleep apnoea/hypopnoea syn-drome. Thorax; 59: 347–352.

N Silber MH, et al. (2004). Assessing thepatient with a sleep disorder. In: SilberMH, et al, eds. Sleep Medicine in ClinicalPractice. London, Taylor & Francis; p. 31–42.

Table 1. Important nocturnal and daytime symptoms and their possible implications.

Symptom Possible implications

Snoring ‘Simple’ snorer, UARS, OSA

Nocturnal gasping/choking OSA, asthma, gastro-oesophageal reflux, panic attacks

Early morning headaches OSA, CO2 retention, insomnia

EDS Extrinsic: sleep deprivation, shift work sleep disorder, drug-related hypersomniaIntrinsic: sleep-disordered breathing, narcolepsy, idiopathichypersomnia, RLS and PLMD, Circadian rhythm sleep disorder

Loss of strength with emotion Cataplexy

Pre-sleep leg discomfort RLS, PLMD

UARS: upper airway resistance syndrome; RLS: restless legs syndrome; PLMD: periodic limb movement

disorder.


Differential diagnosis

Walter T. McNicholas

Hypersomnolence, also referred to asexcessive daytime sleepiness (EDS), isdefined as the inability to stay awake andalert during the waking periods of the day,resulting in lapses into drowsiness or sleep.Thus, EDS is a condition that is associatedwith an increased tendency to fall asleepduring daytime activities where alertness isa normal feature. EDS is more likely to occurin boring monotonous situations that do notrequire active participation, such aswatching television and reading a book, butcan also occur in potentially dangeroussituations, such as while driving oroperating machinery. When very severe, EDSmay result in episodes of automaticbehaviour, in which subjects have nomemory of events that they have performed.In children, EDS may present as restlessness,inattention and hyperactivity. Chronichypersomnolence is arbitrarily defined asbeing present for at least 3 months.

The causes of EDS vary widely, ranging frombehavioural factors such as insufficient orfragmented night-time sleep, medical

disorders associated with sleepfragmentation such as OSAS, use ofsedative medications, circadian rhythmdisturbances that result in a mismatch ofthe sleep/wake pattern, and neuronaldegeneration within the brain structures thatmodulate the sleep/wake rhythm, such asnarcolepsy. The range of disordersassociated with EDS are summarised intable 1. EDS is a central symptomassociated with several sleep disorders,including OSAS and narcolepsy, in additionto neurological or other medical disorderssuch as Parkinson’s disease, hepatic failureand hypothyroidism. The most commoncause of EDS is behavioural where anindividual persistently fails to obtainsufficient sleep to maintain a normal level ofalertness and wakefulness, and this probablyexplains why some normal people recordhigh sleepiness scores on subjectivesleepiness scales such as the ESS. The mostcommon medical cause of EDS is OSAS,although not all OSAS patients report thissymptom. The presence of EDS is aprerequisite for the diagnosis of some sleepconditions such as narcolepsy but not ofothers such as OSAS. In some disorders,chronic EDS may represent the principal andmost disabling complaint. It can also be thefirst to occur, such as in narcolepsy,depression or Parkinson’s disease.

Chronic EDS may be ignored orunderestimated by affected patients simplybecause it is a subjective feature whichdepends on the patient’s ability todistinguish normal from abnormalbehaviour. Individuals may denyexperiencing EDS but report other relatedsymptoms such as tiredness, poor memory

Key points

N EDS is commonly the result of poorlifestyle habits associated withinsufficient sleep time.

N Proper evaluation of EDS requires acareful history, preferably with thepartner present.

N Subjective sleepiness correlates poorlywith objective tests of sleepiness suchas the MSLT and the MWT.


and concentration, in addition to lack ofenergy, weakness or fatigue. EDS maypresent in different ways, which may provideinsight into the underlying cause. One is astate of sleepiness that is recognised by thepatient, thus allowing him/her to fightagainst the symptom. Patients experience aconstant pressure towards falling asleep andalso difficulty with remaining awake. Theytend to fall asleep and/or take frequent shortnaps at inappropriate times and/or settings,especially when situations are notstimulating, such as watching television andreading. In more severe cases, the sleepinessdevelops while driving, but affected patientscan usually recognise the problem and pullover to rest and/or take a nap.

The other presentation of EDS is much lesscommon and consists of sudden onset ofsleep episodes. These episodes are abrupt,brief and unexpected and have beenreported to occur during active situationssuch as driving, eating, talking, walking, orwhile speaking on the telephone. It isunclear whether these episodes, oftentermed ‘‘sleep attacks’’, constitute a uniqueentity or they are merely an extrememanifestation of hypersomnolence. Theyhave been classically described in patients

with narcolepsy and more recently inParkinson’s disease, particularly in thosepatients taking dopaminergic replacementtherapy. However, sleep attacks have alsobeen described rarely in severe OSAS,typically associated with severehypersomnolence.

It is also necessary to differentiate EDS fromother forms of "fatigue". This is not alwayseasy as the terms "sleepiness" and "fatigue"are often used interchangeably. However,the definition of sleepiness as theoccurrence of actual dozing off makes itpossible to distinguish situations where thepatient feels like sitting or lying and doingnothing, from situations where the patientwould like to do things but is just unable todo so because he/she dozes off, especiallywhen the situation is boring. Tirednesswithout sleepiness is rarely a symptom ofOSAS, and is seen more in other conditionssuch as chronic fatigue syndrome ordepression. However, females may be morelikely than males to report fatigue as asymptom of OSAS. These considerationsstress the need for specific questions duringthe history taking: vague questions such as"Are you sleepy during the day?" are likely tobe unreliable. Situational questions such as"Do you fall asleep reading or watchingtelevision, as a passenger in car, or whileattending meetings?’’ provide a moreaccurate picture of true sleepiness.Questions should be adapted to thepatients’ usual activities, and should alsotake into account the fact that patientsengaged in active, outdoor, occupations areless likely to report sleepiness than patientswith more passive, indoor, occupations.

The clinical evaluation of EDS may bedifficult since subjective sleepiness scalesmay be unreliable, either because of thepatient’s lack of perception of the true extentof the problem, or alternatively, a desire tominimise the problem because of concernabout possible occupational consequencesof EDS, such as driver licensing. Collateralhistory from the spouse or another closerelative may be helpful, and it is importantto distinguish EDS from other relatedsymptoms such as fatigue, lack of energy,

Table 1. Causes of EDS/hypersomnolence.

OSAS

Other sleep-related breathing disorders

Narcolepsy

Behaviourally-induced insufficient sleep

Depression

Idiopathic hypersomnia (with/withoutincreased sleep duration)

EDS due to a medical disorder: infections,metabolic disturbances, endocrinopathies,Parkinson’s disease, myotonic dystrophy

Drug or substance induced

Recurrent hypersomnia (e.g. Kleine–Levinsyndrome)

Circadian rhythm disorders

Shift work sleep disorder

Chronic fatigue syndrome


asthenia, poor concentration and tiredness.Several objective tests of sleepiness areavailable, most importantly the MSLT andMWT (see also ‘‘Assessment of daytimesleepiness’’).

The multiple sleep latency test

The MSLT is considered by most specialiststo be the standard objective measure of EDSand is the most widely used and extensivelypublished. The MSLT is intended tomeasure physiological sleep tendency understandardised conditions in the absence ofexternal alerting factors and is based on thepremise that the degree of sleepiness isreflected by the sleep latency. The test istypically performed during the day afterovernight PSG to evaluate the possiblepresence of a cause of EDS (e.g. OSAS), andto document an adequate quantity andquality of sleep on the night preceding theMSLT. The standard clinical MSLT consistsof four or five nap opportunities performedat 2-h intervals across a patient’s usualperiod of wakefulness during whichstandard PSG recordings are obtained toobjectively identify sleep stage during therecordings. Strict guidelines should befollowed regarding meals, ambientenvironment and surrounding noise andother relevant factors to ensure a consistentand reproducible test.

Sleep onset is determined by the time fromlights out to the first epoch of any stage ofsleep in a 30-s epoch. Sleep onset is definedas the first epoch containing more than 15 sof cumulative sleep in a 30-s epoch. Thepatient is permitted to sleep for 15 min afterthe first epoch scored as sleep. This is toassess sleep continuity and the earlyoccurrence of REM sleep. REM sleep latencyis taken as the time from the first epoch ofsleep to the beginning of the first epoch ofREM sleep. Sleep onset REM periods(SOREM) are defined as episodes of at least15 s of REM sleep within a 30-s epoch.Conditions that may adversely affect thevalidity of the MSLT include: 1) insufficientsleep during nights prior to the test; 2)performance of the MSLT at a time differentthan the patient’s usual time of wakefulness;3) excessive noise or temperature extremes;

4) use of medications that could alter sleeparchitecture, such as antidepressants, wakepromoting agents and benzodiazepines; and5) stimulating or upsetting events betweennap opportunities.

While the MSLT is a detailed and complextest, there is some debate as to whethersleep latency is a true measure of EDS.Precise normative ranges for mean sleeplatency are not well defined and manyfactors may influence it, such as motivation,age, external stimuli, medication, previoussleep quality and sleep time, medicaldisorders and psychological factors.Nonetheless, a mean sleep latency .10 minis generally accepted as normal. Thediagnostic sensitivity of the MSLT for thediagnosis of narcolepsy has been estimatedat ,60%, while the diagnostic specificitywhen two or more SOREM are present is,95%. Mean sleep latency of ,3 min istypically seen in narcolepsy, and whencombined with the occurrence of one ormore SOREM provides a diagnosticaccuracy close to 100%. The MSLT is notroutinely indicated for the evaluation ordiagnosis of OSAS or assessment ofresponse to treatment of OSAS. Mean sleeplatency in OSAS is typically in the region of7 min, although it varies with the severity ofthe disorder. However, the MSLT may beindicated in OSAS patients that continue toexperience EDS despite correct treatment ofthe underlying condition.

Maintenance of wakefulness test

The MWT is a variant of the MSLT in which,rather than attempt to sleep, the patient isasked to remain awake as long as possible,and the test assesses ability to resist theurge to fall asleep during soporificcircumstances. Thus, MWT is not adiagnostic test for the degree of sleepinessbut rather a test of the ability to remainawake. MWT can be used to assess theefficacy of a particular treatment once thediagnosis has been reached previously, andis also used to assess the patient fitness todrive and to fly, or ability to return to workand, obviously, to quantify the extent of EDS.This is to assess an individual’s ability toremain awake when his or her ability to


remain awake constitutes a public orpersonal safety issue. The MWT is subject tomore motivational factors than the MSLT.

Similar to MSLT, four nap opportunities areperformed at 2-h intervals after an overnightPSG. The patient is asked to try to remainawake with the eyes open, but is not allowedto use extraordinary measures such aswalking, singing or talking. The latency tothe first three consecutive 30-s epochs ofsleep (or to the first epoch of sleep as in theMSLT) is measured for each trial and themean sleep latency is calculated. The patientis awakened when sleep onset is scored.

A variant of the MWT is the OSLER test,where the subject is asked to press a switchin response to a repetitive flashing lightplaced in front of the face, which flashes forone second in every three. Sleep onset isidentified when the subject fails to respondto seven successive light flashes.

There is a poor correlation between MSLT andMWT, and also between MSLT/MWT andsubjective scores of sleepiness such as the ESS.This lack of correlation between tests impliesthat measures of sleepiness may be testspecific, thus indicating that there is more thana single component to EDS.

Further reading

N Ancoli-Israel S, et al. (2006). Insomniaand daytime napping in older adults.J Clin Sleep Med; 2: 333–342.

N Bodkin CL, et al. (2011). Office evaluationof the "tired" or "sleepy" patient. SeminNeurol; 31: 42–53.

N Chellappa SL, et al. (2009). Chronobio-logy, excessive daytime sleepiness anddepression: is there a link? Sleep Med; 10:505–514.

N Coelho FM, et al. (2011). Testing sleepi-ness and vigilance in the sleep laboratory.Curr Opin Pulm Med; 17: 406–411.

N Dauvilliers Y. (2006). Differential diag-nosis in hypersomnia. Curr NeurolNeurosci Rep; 6: 156–162.

N De Cock VC, et al. (2008). Sleep dis-turbances in patients with parkinsonism.Nat Clin Pract Neurol; 4: 254–266.

N Engleman HM, et al. (2004). Sleep. 4:sleepiness, cognitive function, and qualityof life in obstructive sleep apnoea/hypop-noea syndrome. Thorax; 59: 618–622.

N Fong SY, et al. (2005). Comparing MSLTand ESS in the measurement of excessivedaytime sleepiness in obstructive sleepapnoea syndrome. J Psychosom Res; 58:55–60.

N Kothare SV, et al. (2008). The clinical andlaboratory assessment of the sleepy child.Semin Pediatr Neurol; 15: 61–69.

N Ohayon MM. (2008). From wakefulnessto excessive sleepiness: what we knowand still need to know. Sleep Med Rev; 12:129–141.

N Pigeon WR, et al. (2003). Distinguishingbetween excessive daytime sleepinessand fatigue: toward improved detectionand treatment. J Psychosom Res; 54: 61–69.

N Sullivan SS, et al. (2008). Multiple sleeplatency test and maintenance of wakeful-ness test. Chest; 134: 854–861.

N Thorpy MJ. (2005). Which clinical condi-tions are responsible for impaired alert-ness? Sleep Med; 6: Suppl. 1, S13–S20.

N Zeman A, et al. (2004). Narcolepsy andexcessive daytime sleepiness. BMJ; 329:724–728.


Questionnaires on sleep

Brian D. Kent and Walter T. McNicholas

OSAS is a highly prevalent but markedlyunder-diagnosed condition. Valuable clinicalinformation can be obtained by carrying outa careful history and examination ofindividuals suspected of having thedisorder. However, these are susceptible toreporter and observer bias, and may notproduce objective, reproducible findings.Furthermore, accurate risk stratification andclinical triage of patients referred to sleepservices is highly desirable, as the currentdiagnostic capacity of sleep laboratoriesacross the developed world is greatlyexceeded by the number of subjectsawaiting investigation. Consequently, effortshave been made to standardise clinicalhistory using questionnaires, potentiallyallowing both for the identification ofsubjects with a high likelihood of SDB and

for triage according to theirsymptomatology. These questionnaires canbe broadly divided into three groups – thoseassessing sleepiness, those assessing pre-test probability of OSAS and thoseassessing sleep quality.

Questionnaires assessing sleepiness

Excessive daytime sleepiness (EDS) is oneof the hallmarks of OSAS. Sleepiness can beobjectively measured by using in-laboratoryinvestigations, such as the MSLT and theMWT or modifications thereof. However,these investigations are highly resourceintensive, require expert interpretation andare inconvenient for the patient.

The Epworth Sleepiness Scale Probably themost widely used questionnaire in clinicalsleep medicine is the ESS, a simple andpractical tool designed to assess thesubject’s likelihood of falling asleep across arange of everyday activities (Johns, 1991).Eight separate situations are addressed,with the patient asked to score theirpropensity for falling asleep in each situationfrom 0 to 3 (table 1), with 0 representing nochance of dozing and 3 a high chance ofdozing. A total score of o10 across the eightquestions is indicative of EDS and shouldprompt further investigation. Caution shouldbe exercised in interpreting the ESS score,however, as there are conflicting data on itscorrelation with objective sleepinessmeasured by the MSLT (Sullivan et al., 2008;Punjabi et al., 2003) and its utility in theprediction of the presence or severity ofOSAS remains uncertain (Gottlieb et al.,1999; Kapur et al., 2005). In particular, anormal ESS score does not exclude thepresence of significant SDB.

Key Points

N The ESS is used to subjectivelyquantify sleepiness across a range ofeveryday situations, with a cumulativescore of .10 suggesting EDS.

N The Berlin Questionnaire has highspecificity for identifying subjects withmoderate–severe OSAS, but relativelylow sensitivity.

N The STOP and STOP-Bangquestionnaires have high sensitivityfor identifying subjects with sleepdisordered breathing, but relativelylow specificity.

N Sleep quality can be subjectivelyassessed by the PSQI.


The Stanford Sleepiness Scale In contrastwith the ESS, the Stanford Sleepiness Scale(SSS) measures current, rather thaneveryday, sleepiness (Hoddes et al., 1973).Subjects choose one of seven statementsdescribing their level of sleepiness, rangingfrom 1 (‘feeling active, vital, alert, or wideawake’) to 7 (‘no longer fighting sleep, sleeponset soon, having dream-like thoughts’).Selection of statements 4–7 at a time whenthe subject should be feeling alert isconsidered consistent with excesssleepiness. The key advantage of the SSS isits ability to be administered on severaloccasions throughout the day and night,allowing a dynamic picture of the subject’ssymptomatology to be built up. Moreover,as a research tool, it is well validated againstobjective measures of sleepiness. Conversely,it is unable to differentiate betweenindividuals who are sleepy due to simplesleep deprivation and those who have anunderlying sleep disorder, thereby limiting itsutility in everyday clinical practice.

Screening questionnaires for OSAS

As discussed earlier, reliance on the clinicalhistory and examination as a screening toolfor OSAS may lead to reporter and observerbias. In particular, patients may considerlongstanding symptoms to be a normal

variant, while the absence of the typicalclinical phenotype does not exclude adiagnosis of SDB. This has led to thedevelopment and attempted validation of anumber of screening questionnaires forOSAS. Perhaps the best studied of these arethe Berlin Questionnaire, and STOP andSTOP-Bang questionnaires.

The Berlin Questionnaire The BerlinQuestionnaire classifies subjects as low- orhigh-risk for OSAS based on their responsesin three clinical categories – snoring history,daytime sleepiness and history ofhypertension or obesity (table 2) (Netzeret al., 1999). To be considered at high riskfor OSAS, patients need to respondpositively in two of the categories assessed.The Berlin Questionnaire has been validatedin primary care, sleep clinic and surgicalpopulations, and provides a high degree ofspecificity for the diagnosis of moderate-to-severe SDB, albeit with a relatively lowdegree of sensitivity (Abrishami et al., 2010).In a sentinel study assessing its use in aprimary care setting, Netzer et al. (1999)reported a sensitivity of 54% and aspecificity of 97% for identifying subjectswith a respiratory disturbance index (RDI) of.15 events?h-1, with a positive predictivevalue of 97%.

The STOP questionnaire Developed in anattempt to screen surgical patients forundiagnosed OSAS, the STOP questionnairecomprises four questions (Chung et al.,2008). These address the presence orabsence of snoring, daytime tiredness,observed apnoeas during sleep and a historyof high blood pressure. Positive answers totwo or more of these questions categorisesthe patient as at high risk for underlyingOSAS. The sensitivities of STOP with AHIs.5, 15 and 30 events?h-1 were 65.6%, 74.3%and 79.5% respectively in a validation cohortof predominantly overweight and obesesurgical patients (Chung et al., 2008).However, this was at the cost of reducedspecificity compared with the BerlinQuestionnaire.

The STOP-Bang questionnaire A refinementof the STOP questionnaire, incorporatingadditional questions on BMI, age, neck

Table 1. Situations assessed by the ESS.

Sitting and reading

Watching television

Sitting inactive in a public place

As a passenger in a car for an hour without abreak

Lying down to rest in the afternoon whencircumstances permit

Sitting and talking to someone

Sitting quietly after a lunch without alcohol

In a car, while stopped for a few minutes intraffic

Each situation is scored from 0 to 3, with 0

representing no chance of dozing and 3 a high

likelihood of dozing. The upper limit of normal

of the sum of the scores is generally considered

to be 10.


circumference and sex (table 3), has beenshown to improve its sensitivity significantly.When applied to .4,000 community-basedparticipants in the Sleep Heart Health Study,STOP-Bang had a sensitivity of 87% in theidentification of subjects with moderate–severe SDB (Silva et al., 2011). Moreover, itmay also have a role in predicting severity ofunderlying OSAS (Farney et al., 2011),thereby facilitating triage of patientsattending sleep clinics.

Questionnaires assessing sleep quality

SDB is associated with poor subjective sleepquality. Longitudinal assessment of sleepquality may be performed in patients with OSASto gauge symptomatic response to treatment.The best studied instrument in this field is thePittsburgh Sleep Quality Index (PSQI).

Pittsburgh Sleep Quality Index The PSQI is aself-administered questionnaire thatmeasures sleep quality over a 1-month timeperiod. It consists of 19 questions on sleepquality, symptoms and hygiene, which arecombined into seven component scores,each rated 0–3, giving a global score of 0–21(Buysse et al., 1989). A score of o5 may besuggestive of some form of underlying sleepdisorder, but PSQI scores do not correlatesignificantly with formally measured on PSGsleep abnormalities (Buysse et al., 2008).Consequently, it has utility in assessingsubjective quality of sleep, but is deficient asan objective measure of this.

Further reading

N Abrishami A, et al. (2010). A systematicreview of screening questionnaires forobstructive sleep apnea. Can J Anaesth;57: 423–438.

Table 2. The Berlin Questionnaire.

Questions Answers Scoring

Category 11) Do you snore?2) If yes, how loud is it?3) How often do you snore?4) Has your snoring ever bothered

other people?5) Has anyone noticed you stop

breathing during your sleep?Category 2

6) How often do you feel tired orfatigued after your sleep?

7) During your waking time, do youfeel tired, fatigued or not up to par?

8) Have you ever nodded off or fallenasleep while driving a vehicle?

9) If yes, how often does this occur?Category 3

10) Do you have high blood pressure?

Items 1, 4, 8, 10Yes (1), no or don’t

know (0)Item 2

Slightly louder thanbreathing (0)

As loud as talking (0)Louder than talking (1)

Can be heard inadjacent room (1)

Items 5, 6, 7, 9Nearly every day (1)3–4 times a week (1)1–2 times a week (0)

1–2 times a month (0)Never (0)

Category 1Positive if total score o2

pointsCategory 2

Positive if total score o2points

Category 3Positive if answer to item

10 is yes or if BMI.30 kg?m-2

High risk of OSAS: o2categories positive

Low risk of OSAS: ,2categories positive

Table 3. The STOP-Bang questionnaire.

1) Do you snore loudly?

2) Do you often feel tired, fatigued or sleepyduring the daytime?

3) Has anyone observed you stop breathingwhile you sleep?

4) Do you have or are you being treated forhigh blood pressure?

5) BMI o35 kg?m-2?

6) Age o50 yrs?

7) Neck circumference o40 cm?

8) Male?

High risk of OSAS: answering yes to three or

more items; low risk of OSAS: answering yes to

fewer than three items.


N Buysse DJ, et al. (2008). Relationshipsbetween the Pittsburgh Sleep QualityIndex (PSQI), Epworth Sleepiness Scale(ESS), and clinical/polysomnographicmeasures in a community sample. J ClinSleep Med; 4: 563–571.

N Buysse DJ, et al. (1989). The PittsburghSleep Quality Index: a new instrument forpsychiatric practice and research.Psychiatry Res; 28: 193–213.

N Chung F, et al. (2008). STOP question-naire: a tool to screen patients forobstructive sleep apnea. Anesthesiology;108: 812–821.

N Farney RJ, et al. (2011). The STOP-Bangequivalent model and prediction of sever-ity of obstructive sleep apnea: relation topolysomnographic measurements of theapnea/hypopnea index. J Clin Sleep Med;7: 459B–465B.

N Gottlieb DJ, et al. (1999). Relation ofsleepiness to respiratory disturbanceindex: the Sleep Heart Health Study.Am J Respir Crit Care Med; 159: 502–507.

N Hoddes E, et al. (1973). Quantification ofsleepiness: a new approach. Psychophysio-logy; 10: 431–436.

N Johns MW. (1991). A new method for mea-suring daytime sleepiness: the Epworthsleepiness scale. Sleep; 14: 540–545.

N Kapur VK, et al. (2005). Sleepiness inpatients with moderate to severe sleep-disordered breathing. Sleep; 28: 472–477.

N Netzer NC, et al. (1999). Using the BerlinQuestionnaire to identify patients at riskfor the sleep apnea syndrome. Ann InternMed; 131: 485–491.

N Punjabi NM, et al. (2003). Predictors ofobjective sleep tendency in the generalpopulation. Sleep; 26: 678–683.

N Silva GE, et al. (2011). Identification ofpatients with sleep disordered breathing:comparing the four-variable screening tool,STOP, STOP-Bang, and Epworth Sleepi-ness Scales. J Clin Sleep Med; 7: 467–472.

N Sullivan SS, et al. (2008). Multiple sleeplatency test and maintenance of wakeful-ness test. Chest; 134: 854–861.


Predisposing factors

Maria R. Bonsignore

In the clinical assessment of patients withsuspected SDB, anatomical and functionalpredisposing factors should be carefullyassessed. In addition, several diseases mayincrease the risk of developing SDB throughvarious mechanisms. Table 1 summarisespredisposing conditions that should alertthe physician and suggest SDB as a possiblediagnosis.

General predisposing factors

Age OSA occurs at all ages and itsprevalence increases with ageing (Lindberg

et al., 2010). However, the clinicalsignificance of asymptomatic OSA in elderlysubjects is unclear (Launois et al., 2007).

Sex Men are more susceptible to SDB thanwomen, but the sex difference becomesattenuated in old age (Lindberg et al., 2010),suggesting that hormonal factors may playsome role. Women are relatively resistant toupper airway collapse during sleep (Lin et al.,2008; Kirkness et al., 2008). Centralapnoeas are also less frequent in womenthan in men due to differences in ventilatorycontrol (lower apnoea threshold andventilatory overshoot after apnoeas inwomen) (Lin et al., 2008).

Race Although data are scarce, Asians andAfrican-Americans appear to be at higherrisk of developing OSA than Caucasians(Villaneuva et al., 2005), a finding at leastpartly explained by differences in craniofacialstructures (Lee et al., 2010).

Anatomical abnormalities

Several conditions can decrease thedimensions and patency of the upper airwayby affecting bony structures or soft tissuesof the face and neck. Clinically, in anypatient with suspected SDB, it is advisableto record the Mallampati score, a simpleway to evaluate the patency oforopharyngeal airways (Nuckton et al.,2006). Craniofacial structure and upperairway morphology are geneticallydetermined and explain racial differences inprevalence of OSA. Increased fat depositionin the neck is a common finding in OSApatients (Schwab et al., 2003), and neckcircumference correlates with OSA severity.The following conditions predispose todevelopment of OSA.

Key points

N Several conditions predispose todevelopment of SDB.

N Anatomical and functional factors arecentral to the pathogenesis of upperairway obstruction, alone or inassociation with other pathologicalconditions.

N Obesity is the risk factor mostfrequently found in OSA patients,sometimes associated with endocrinedisorders.

N Other diseases also predispose toSDB and should be carefullyinvestigated in the clinical assessmentof patients with suspected OSA.

N Chronic HF predisposes to Cheyne–Stokes breathing during sleep,through complex mechanismsinvolving central and peripheralcontrol of breathing, but is oftenassociated also with OSA.


Nasal obstruction This can be associatedwith deviation of the nasal septum orhypertrophy of turbinates. Classically, it isconsidered a predisposing factor for upperairway collapse through increased nasalresistance and development of markedlynegative pressure during inspiration.According to a recent meta-analysis nasalobstruction increased the risk of snoring butnot of OSA (Kohler et al., 2007).

Micrognathia A small retropositionedmandible, such as in patients with PierreRobin syndrome, reduces theretropharyngeal area considerably (Gaudetteet al., 2010).

Hyoid bone displacement The hyoid bone isan anchoring site for pharyngeal muscles.Its downward displacement alters theposition of the tongue, favouring upperairway collapse (Gaudette et al., 2010).

Hypertrophy of uvula, soft palate and tonsils A‘‘crowded’’ oropharynx can be found in OSApatients, in part secondary to fat depositionin the neck associated with obesity and inpart resulting from snoring-associatedvibrations and inflammation of soft tissues(Gaudette et al., 2010).

Macroglossia In both adults and children, anenlarged tongue is a predisposing factor ofSDB. In adults, macroglossia is a feature ofacromegaly, but is also common in obesepatients; in paediatric patients, macroglossiacan be found in Down’s syndrome,mucopolysaccharidosis type VI and othergenetic diseases (Gaudette et al., 2010).

Upper airway shape and length Profoundalterations in the shape of upper airwayshave been documented in OSA patients, inparticular an elliptical shape of the airwaywith increased anteroposterior dimension

Table 1. Predisposing factors to be considered in the diagnosis of SDB.

General Ageing

Males

Asian or African-American race

Anatomical abnormalities Nasal obstruction

Micrognathia

Hyoid bone displacement

Hypertrophy of uvula, soft palate and tonsils

Macroglossia

Upper airway shape and length

Functional factors Factors affecting upper airway collapsibility(drugs and alcohol)

Airway inflammation

Instability of respiratory drive

Associated diseases Obesity

Congenital syndromes (Pierre Robin, Down,etc.)

Endocrine disorders

Heart failure

Stroke

Chronic kidney disease (CKD)

Gastro-oesophageal reflux disease


and lateral narrowing (Schwab et al., 2003;Gaudette et al., 2010). Increased upperairway length also predisposes to collapse(Malhotra et al., 2002). Such alterations arenot easily appreciable clinically, and requireimaging studies.

Functional factors

Increased upper airway collapsibility OSA ischaracterised by increased collapsibility ofthe oropharynx, which lacks rigid supportstructures. The Starling resistor is theclassical model to study collapsible tubes,and the critical closing pressure (Pcrit) isdefined as the pressure inside the airway atwhich collapse occurs (Gaudette et al., 2010;McKay et al., 2010). Pcrit is affected byseveral factors, such as age, sleep stage,position, presence of obesity, oedema of theairways and reflexes activated by negativepressure. Upper airway dilator musclesactively counteract airway collapse(Gaudette et al., 2010; McKay et al., 2010).Drugs such as benzodiazepines(Guilleminault, 1990) or alcohol (Peppardet al., 2007) negatively affect respiratorydrive and upper airway collapsibility, andworsen SDB. During clinical assessmentpatients with suspected SDB should beasked specific questions regarding the useof such substances.

Airway inflammation Inflammationcontributes to upper airway alterations inOSA patients and is caused by mechanicaltrauma associated with snoring or apnoeas,and oxidative stress associated with hypoxia-reoxygenation. Factors that can worsen theinflammatory milieu in OSA patients includesmoking, allergy or systemic inflammation(see Gaudette et al., 2010 for more details).Such factors should also be activelyinvestigated in the clinical assessment ofpatients with SDB.

Instability of respiratory drive The complexresponse to obstructive apnoeas duringsleep is associated with sleepfragmentation, as the arousal is necessary tore-establish airway patency. The shift fromwakefulness to sleep is associated withreduced respiratory drive, i.e. the responseof the respiratory centre to chemical and

mechanical stimuli. The opposite occurswith arousal (i.e. a shift from sleep towakefulness). These state changes areresponsible for the apnoeas which mayphysiologically occur in normal subjectswhile falling asleep, but may also contributeto the cyclic occurrence of respiratory eventsin patients with SDB (Gaudette et al., 2010;McKay et al., 2010).

There are two pathophysiologicalmechanisms involved in instability of therespiratory drive, the apnoeic threshold andan increased loop gain (Gaudette et al.,2010; McKay et al., 2010). Briefly, theapnoeic threshold defines the level of PaCO2

(usually 1–2 mmHg below the eupneicwaking level) at which ventilation stopsduring NREM sleep. If arousal from sleep isassociated with hyperventilation,hypocapnia ensues and apnoea developswhen sleep is re-established.

The loop gain is a complex mechanismdefining the response of the respiratorycentre to a change in CO2. The larger thisresponse, the higher the instability inrespiratory drive. The reader is referred toGaudette et al. (2010) and McKay et al.(2010) for a comprehensive description ofthis topic. See also the articles on CSA inchapter 2 of this book. Gaudette et al. (2010)summarises the role of an unstablerespiratory drive in the pathogenesis ofsleep apnoea. Such a mechanism is centralto explain central apnoeas, but can also beimportant in OSA. Clinical markers of anunstable respiratory drive are scarce; somepatients may report difficulty in maintainingsleep due to frequent awakenings with senseof suffocation due to apnoeas. Arterial bloodgases during wakefulness can show atendency towards hypocapnia. Occurrenceof central or mixed apnoeas during anocturnal recording is also suggestive of anunstable respiratory drive.

Associated diseases

Several diseases which increase the risk ofdeveloping SDB have been recognised.Some congenital syndromes predispose toSDB due to craniofacial or upper airwayalterations (Pierre Robin and Down’s


syndrome, among others). They are usuallyof paediatric pertinence, but can occasionallybe seen in adults referred for sleep studies.

Obesity is by far the most frequent diseasepredisposing to OSA, especially in subjectsshowing central fat distribution (Gaudetteet al., 2010). Obesity explains, at least inpart, the increased frequency of SDB foundin many other diseases, such as Down’ssyndrome or hypothyroidism. The closerelationship between obesity and OSA isconfirmed by the changes in OSA severityassociated with weight changes.

Endocrine disorders can be associated withSDB (Attal et al., 2010). Approximately 10%of obese patients with suspected OSA showsubclinical hypothyroidism. Acromegalymodifies upper airway structure by affectingboth soft tissues and bones, and about two-thirds of acromegalic patients show OSA.Diabetes is often associated with bothcentral and obstructive events during sleep(Rasche et al., 2010). The polycystic ovarysyndrome also carries a high risk for OSA(Nitsche et al., 2010).

Chronic HF predisposes to central apnoeas,and especially to Cheyne–Stokes breathing(McKay et al., 20110; Naughton et al., 2010).OSA is also frequent and should besuspected in overweight/obese patients withchronic HF with a clinical history of snoring(Naughton et al., 2010). Identification ofthese patients is important, since untreatedOSA increases cardiovascular risk.

Stroke is often associated with SDB (Ramaret al., 2010). According to epidemiologicalstudies, OSA often precedes stroke andcould be a risk factor. Many patients showsevere OSA after a stroke, but CPAPtreatment is problematic in these patientsand often not accepted. Central apnoeasand CSR have been shown to occur quitecommonly in the acute phase of stroke, butspontaneously resolve with time and rarelyneed treatment, unless concomitant chronicHF is found. It should be noted that strokecan be associated with several other sleepdisorders, such as insomnia, hypersomnia,circadian rhythm disturbances and periodicleg movements during sleep or restless leg

syndrome. Therefore, patients who havepreviously had a stroke should be carefullyinvestigated for sleep disorders.

Chronic kidney disease may predispose toOSA through the mechanisms of volumeoverload and nocturnal shift of fluid fromperipheral tissues (Mavanur et al., 2010).

The role of gastro-oesophageal refluxdisease (GERD) in OSA is debatable.Improvement of OSA was shown in subjectswith documented pathological proximal pH-metry after proton pump inhibitortreatment, but no randomised controlledtrial is available on this topic (Karkos et al.,2009). Given the high prevalence of bothOSA and GERD, it is advisable to enquireabout GERD symptoms in patients withsuspected OSA.

Further reading

N Attal P, et al. (2010). Endocrine aspects ofobstructive sleep apnea. J Clin EndocrinolMetab; 95: 483–495.

N Gaudette E, et al. (2010). Pathophysio-logy of OSA. Eur Respir Monogr; 50: 31–50.

N Guilleminault C. (1990). Benzodiaze-pines, breathing, and sleep. Am J Med;88: 25S–28S.

N Karkos PD, et al. (2009). Reflux andsleeping disorders: a systematic review.J Laryngol Otol; 123: 372–374.

N Kirkness JP, et al. (2008). Contribution ofmale sex, age, and obesity to mechanicalinstability of the upper airway duringsleep. J Appl Physiol; 104: 1618–1624.

N Kohler M, et al. (2007). The role of thenose in the pathogenesis of obstructivesleep apnoea and snoring. Eur Respir J;30: 1208–1215.

N Launois SH, et al. (2007). Sleep apnea inthe elderly: a specific entity? Sleep MedRev; 11: 87–97.

N Lee RW, et al. (2010). Differences in cranio-facial structures and obesity in Caucasianand Chinese patients with obstructive sleepapnea. Sleep; 33: 1075–1080.

N Lin CM, et al. (2008). Gender differencesin obstructive sleep apnea and treatmentimplications. Sleep Med Rev; 12: 481–496.

N Lindberg E. (2010). Epidemiology of OSA.Eur Respir Monogr; 50: 51–68.


N Malhotra A, et al. (2002). The malepredisposition to pharyngeal collapse:importance of airway length. Am J RespirCrit Care Med; 166: 1388–1395.

N Mavanur M, et al. (2010). Sleep disor-dered breathing in patients with chronickidney disease. Indian J Med Res; 131:277–284.

N McKay LC, et al. (2010). Physiology andneural control of breathing during sleep.Eur Respir Monogr; 50: 1–16.

N Naughton MT, et al. (2010). Sleep apnoeain chronic heart failure. Eur RespirMonogr; 50: 396–420.

N Nitsche K, et al. (2010). Obstructive sleepapnea and metabolic dysfunction inpolycystic ovary syndrome. Best PractRes Clin Endocrinol Metab; 24: 717–730.

N Nuckton TJ, et al. (2006). Physicalexamination: Mallampati score as an

independent predictor of obstructivesleep apnea. Sleep; 29: 903–908.

N Peppard PE, et al. (2007). Association ofalcohol consumption and sleep disor-dered breathing in men and women.J Clin Sleep Med; 3: 265–270.

N Ramar K, et al. (2010). The relationshipbetween sleep disorders and stroke.Postgrad Med; 122: 145–153.

N Rasche K, et al. (2010). Obstructive sleepapnea and type 2 diabetes. Eur J Med Res;15: Suppl. 2, 152–156.

N Schwab RJ, et al. (2003). Identification ofupper airway anatomic risk factors forobstructive sleep apnea with volumetricmagnetic resonance imaging. Am J RespirCrit Care Med; 168: 522–530.

N Villaneuva AT, et al. (2005). Ethnicity andobstructive sleep apnoea. Sleep Med Rev;9: 419–436.


Diagnostic algorithms

Josep M. Montserrat, Ferran Barbe and Juan Fernando Masa

The gold standard for the diagnosis ofOSAH is attended PSG, but this is time-consuming and expensive. The prevalence,morbidity and mortality of OSAH, as well asthe increasing awareness of the problem inboth the medical community and thegeneral population, have increased thedemand for consultations and diagnosticstudies in recent years. There is, therefore, agrowing interest in alternative diagnosticmethods and approaches. The AmericanAcademy of Sleep Medicine (AASM)recommends home diagnosis with portablemonitoring devices in patients with a highpre-test probability of moderate-to-severeOSAH but no significant comorbidities(Collop et al., 2007). A negative resultshould lead to a PSG. Of the four systemsfor OSAH evaluation available (level 1: sleeplaboratory PSG; level 2: unattended PSG;level 3: respiratory polygraphy; level 4: one ortwo respiratory variables), the AASM’sminimum recommendation proposed theuse of a type 3 portable monitoring devicethat includes airflow, thoracoabdominal

bands and oximetry (respiratorypolygraphy). Different nationalrecommendations also endorse thisstrategy, including the excellent Frenchguidelines (Societe de Pneumologie deLangue Francaise, 2010). Nowadays,however, some other important aspectsmust be taken into account, as the wholediagnostic management picture is about tochange in several important ways and, infact, some changes can already be noted.The most significant of these are theinvolvement of different fields or levels ofmedicine, the search for strategies thatguarantee cost-effectiveness, thedevelopment of telemedicine (Masa et al.,2011a, b) and, finally, the current re-examination of the real role of the diagnosticprocedure for level 4 (one or two respiratoryvariables) (Thornton et al., 2012). Beforemoving onto a description of the differentstrategies applied to diagnosticmanagement, a few points need to beconsidered.

1. At present, the typical population inneed of OSAH evaluation is quitedifferent from that of a few years ago(table 1). This change in the populationaffects diagnosis strategies becausesome patients are more difficult todiagnose and, therefore, full PSG ismandatory in a significant number ofcases.

2. Most patients, at least in somecountries, are controlled by the sleepclinic of a fully equipped referencehospital (Hernandez et al., 2007), whichmeans that only a minority arecontrolled in primary medicine (fig. 1a).It is important to consider this aspect

Key Points

N There has been a shift from laboratoryPSG to home portable monitoring inpatients with a high pre-test probabilityof OSAS.

N Two groups of patients need to beevaluated: i) those with symptoms;and ii) those with risk factors for OSAS.

N A network approach is likely to be thebest care model for managing thegrowing sleep medicine caseload.


because when a disease is prevalent, allmedical levels must be involved andequipped with the appropriatetechnology to screen or diagnosepatients, or at least severe ones.

3. Finally, as a consequence of points 1and 2, and the prevalence of thedisease, a network system is needed forthe management of these patients(fig. 1b), because a network optimisesthe use of knowledge emerging frominteractivity. In other words, the wholenetwork is more than the sum of itsindividual parts.

All three of these points needed to be takeninto consideration in our approach to thediagnostic management of OSAH patients.

Regardless of the strategy chosen for OSAHdiagnosis (as discussed later), figure 2 showsa summary of the various aspects and stepsthat must always be followed in every case. Ascan be seen at the top of figure 2, two groupsof patients need to be evaluated.

1. Those patients with symptoms (we planto treat the symptoms).

2. Those patients with risk factors (weplan to treat the risk factors) for OSA,

Table 1. ‘‘Typical’’ and actual characteristics of people who need OSAH evaluation.

‘‘Typical’’ patients Actual patients

‘‘Pickwickian syndrome’’Obesity

Heavy snoringPauses in breathing

Systemic hypertension

Elderly patients, childrenPregnancy

FibromyalgiaUPPP

Maxillomandibular surgeryBariatric surgery

Cardiovascular patientsNonclinic population

Mild-to-moderate symptomsNeurological and psychiatric diseasesMyopathies and metabolic diseases

UPPP: uvulopalatopharyngoplasty.

NetworkOptimal conditions

Cost-effective

Non-networkNonoptimal conditions

Not cost-effective

Reference hospital

Nonreference hospitals

Specialists

Family medicine

Nurses

At present Futurea) b)

Figure 1. a) Left: the current situation. Most patients are managed by the reference hospital. Right: themost reasonable follow-up approach when a disease is common. All the medical levels should be involvedand when sleep apnoea is mild, family medicine should be implicated. b) Intercommunication betweenthe different levels is the optimal way of working. As mentioned in the text, a network optimises the use ofknowledge emerging from interactivity. Orange: reference hospitals; light blue: nonreference hospitals;dark blue: family physicians; red: other medical levels.


despite the absence of classicalsymptoms. In this case, someuncertainty exists over aspects of thetreatment. Of course, these patientsneed to be evaluated, but the effects oftreatment with CPAP have not been fullydemonstrated in every part of thesituation presented in figure 2. Inpatients with refractory hypertension, itis clear that they can be asymptomaticand that treatment of OSA with CPAPsignificantly improves blood pressure

(Barbe et al. 2010). In contrast, the roleof CPAP treatment in all spectrums ofpatients with asymptomatic OSA andmetabolic syndrome is not clear atpresent (Sharma et al., 2011; Lam et al.,2012). In these cases, more long-terminterventional trials are still needed(Lam et al., 2012).

Therefore, the key point in deciding whethera sleep study is performed is the history ofsymptoms obtained by the physician. This

Patients that need to be evaluated

Symptoms Snoring Witnessed apnoeas Nocturia Daytime somnolence or fatigue Unrefreshing sleep

Patients at high risk Hypertension (resistant) Obesity (BMI >35 kg·m-2), especially in professional drivers Before bariatric surgery Pulmonary hypertension Metabolic disturbances and arrhythmias (AF) Stroke and cardiac failure Myopathies Some respiratory disease with hypercapnia

OSAHclinical assessment

Sleep study

RP

Clear sleep apnoea symptoms without known comorbidity

Symptoms with known comorbidity or suspicion of nonrespiratory sleep disorders

Intensity of the symptoms Somnolence that clearly disturbs social or work life Somnolence while driving Severe episodes of chokingCompressive history and examination Broad ENT examination Spirometry in smokers Basic blood test with glycosylated Hb Check for reflux

Differential diagnosis Sleep hygiene Sleep restriction Narcolepsy RLS Insomnia Depression Drugs that induce somnolence Different neurological disease

PSGIf discordance between

symptoms and RP

Figure 2. The three steps that should always be followed when sleep apnoea is suspected: detection(patients to be evaluated); clinical assessment for the differential diagnosis and assessment of the clinicalseverity of the disease; and, finally, the sleep study, which should be undertaken on the basis of a pre-testpossibility of sleep apnoea and the presence or otherwise of known comorbidity. AF: atrial fibrillation; RLS:restless legs syndrome; Hb: haemoglobin; RP: respiratory polygraphy.


will also determine the type of test chosenand the urgency of its application. In thecase of high-risk patients with no majorsymptoms or no symptoms at all, a sleeptest is mandatory in patients withnoncontrolled hypertension or heart diseasewith an ejection fraction ,40%, or in thosewho are truncally obese professional drivers.

The next step is the clinical assessment ofOSAH, as shown in figure 2. First of all, agood differential diagnosis should bereached. Table 2 shows the basic entitiesthat any professional working with OSAHneeds to know: differential diagnosis, how todiagnose, management and the basicfeatures of treatment, as well as differentialdiagnosis taking into account other causesof somnolence and daytime fatigue; these

are the major OSAH symptoms but they aresometimes confusing as they can havedifferent causes, which must also be wellunderstood. The most important of thesealternatives are insufficient sleep time anddepression. Moreover, the severity of thesymptoms must be checked. As regardssomnolence, in addition to the classicquestionnaires, which of course must beused, two simple questions need to beasked: whether sleepiness inducesdrowsiness that hinders work or social life(especially in the morning) and whether thepatient experiences somnolence whiledriving. Choking episodes during the nightcan also be very stressful. On the basis of allthese data, the appropriate type of sleepstudy can be chosen along the lines shownin figure 2. This procedure is advocated by

Table 2. Basic entities in OSAH.

Daytime somnolence Bad sleep hygiene or insufficient sleepShift workSleep apnoeaDepressionNarcolepsy

Narcolepsy Cataplexy: sudden and transient episode of loss of muscle tonetriggered by emotions

Restless legs syndrome Need to move the legs to stop uncomfortable leg sensationsthat improve with movingDisturbs sleep and can induce insomnia and somnolence

Depression Lack of interest and pleasure in daily activities, insomnia orexcessive sleeping, lack of energy or inability to concentrateNegative thoughts of death or even suicide

Insomnia Difficulty initiating (20 min) or maintaining sleep with theperception of poor-quality sleepCan produce tiredness or somnolence during the day or be asymptom of depression

Somnolence induced by drugs Antianxiety drugs, some antidepressants or antihistamines ornarcoticsOthers such as pramipexole

Neurological diseases Infections, tumours, Steiner myopathy, stroke etc.

Other sleep disorders

REM behaviour disorders Singular behaviours during sleep, usually violentIn some cases, will result in injury to the bed partnerIn some cases, precedes brain degenerative diseases(Parkinson’s)

Classical parasomnias Somnambulism, nocturnal terrors, nightmares and the basicchronobiological entities


several sleep societies, including the SpanishSleep Network (Lloberes et al., 2011).

Moving more specifically onto the differentdiagnosis strategies on offer, figure 3summarises the various possibilities,illustrating the maxim that when a disease iscommon, almost all medical levels must beinvolved. This means a networkingapproach, in which each patient is managedat the appropriate medical level, such as areference or nonreference hospital or afamily physician. All these levels probablyneed to be involved, especially during follow-up, but also in the diagnosis strategy.Figure 3a shows the classic diagnosis

management model. Some groups use onlyfull PSG for OSAH diagnosis, although anumber of centres use respiratorypolygraphy when patients show a highpossibility of having OSAH but do notpresent any notable comorbidity. All theseprocedures, however, are performed inreference hospitals and, therefore, as can beseen in table 1, these centres areovercrowded. Figure 3b shows what wouldperhaps be the most adequate approach todiagnosis management nowadays, withnonreference hospitals handling asignificant number of patients, always inaccordance with the guidelines of figure 2.When full PSG is needed, patients should be

SAHS suspicionFamily physicians or

other specialist

SAHS suspicionFamily physicians or other specialist

Sleep unitNonference hospital

RP

Sleep unitReference hospital

Multidisciplinary teamFully equipped

c)

b)

a)

Family physicians or other specialistSimplified devices

Special patients (known diseases, insomnia, depression, or suspicion of neurological entities)

Some hospitals perform full PSG in all patients

Sleep unitNonreference hospital

RP

Sleep unitReference hospital

Multidisciplinary teamFully equipped

Difficultpatients

Can diagnose and manage a significant number of patients

SAHS suspicion Sleep unitReference hospital

High pre-testprobability

Low pre-testprobability Full PSG

RP

Figure 3. A summary of the various diagnostic strategies on offer. Although this will depend on thecharacteristics of each centre or country, (b) is probably the most well-balanced. Reference andnonreference hospitals are involved in two types of sleep test. a) is the most classical approach. It is onlyperformed in reference centres and sometimes all patients are studied via a full PSG. Finally, c) representswhat could happen in the future when a family physician or nonhospital specialist may use simpletechnology to play an important role in screening and even diagnosis. In the latter case, in particular, thiswill occur when sleep medicine has become well understood and incorporated into the daily routine of non-hospital physicians. RP: respiratory polygraphy.


sent to a reference hospital. Finally, figure 3cshows what could, and ideally will, happenin the near future if we want to reach all ourpatients. Even family physicians coulddiagnose very typical patients, probably withthe use of simplified devices. One importantfactor, as mentioned above, and as pointedout recently by Masa et al. (2001a, b), iscost-effectiveness. The study of Masa et al.(2001a, b) focused on patients withsuspected OSAH due to the presence ofsnoring, observed apnoeas, sleepiness (ESSscore .10) or unrefreshing sleep, whileexcluding patients with other suspectedsleep disorders or severe and unstable heartdisease, or those who were unable to set uprespiratory polygraphy by themselves. Theauthors’ main conclusions were that a homerespiratory polygraph connected to ahospital via a telemetric procedure canexclude and confirm the diagnosis of OSAHand that home respiratory polygraphy hasthe same diagnostic efficacy as PSG, at halfthe cost (or even less). Therefore, asatisfactory cost-efficiency could be attainedby following this strategy.

Lastly, another important question is how toapproach the level 4 procedure (one or tworespiratory variables) in OSAH diagnosis.Once sleep medicine is more widelyunderstood, however, all the proceduresmentioned above will be complementary,because networking is cost-effective andessential, as it is the only way to reach mostpatients. Accreditation in sleep medicinerepresents a huge advance. Medicalpreparation and knowledge that will allow agood clinical evaluation to be made is thebest sleep test of all.

Further reading

N Barbe F, et al. (2010). Long-term effect ofcontinuous positive airway pressure inhypertensive patients with sleep apnea.Am J Respir Crit Care Med; 181: 718–726.

N Collop NA, et al. (2007). Clinical guide-lines for the use of unattended portablemonitors in the diagnosis of obstructivesleep apnea in adult patients. PortableMonitoring Task Force of the AmericanAcademy of Sleep Medicine. J Clin SleepMed; 3: 737–747.

N Dempsey JA, et al. (2010). Pathophysiologyof sleep apnea. Physiol Rev; 90: 47–112.

N Duran J, et al. (2001). Obstructive sleepapnea–hypopnea and related clinical fea-tures in a population-based sample ofsubjects aged 30 to 70 yr. Am J Respir CritCare Med; 163: 685–689.

N Hernandez L, et al. (2007). Managementof sleep apnea: concordance betweennonreference and reference centers. Chest;132: 1853–1857.

N Lam JC, et al. (2012). Obesity, obstructivesleep apnoea and metabolic syndrome.Respirology; 17: 223–236.

N Lloberes P, et al. (2011). Diagnosis andtreatment of sleep apnea–hypopnea syn-drome. Spanish Society of Pulmonologyand Thoracic Surgery. Arch Bronconeumol;47: 143–156.

N Marin JM, et al. (2005). Long-termcardiovascular outcomes in men withobstructive sleep apnoea-hypopnoea withor without treatment with continuouspositive airway pressure: an observationalstudy. Lancet; 365: 1046–1053.

N Masa JF, et al. (2011a). Therapeuticdecision-making for sleep apnea andhypopnea syndrome using home respira-tory polygraphy: a large multicentricstudy. Am J Respir Crit Care Med; 184:964–971.

N Masa JF, et al. (2011b). Effectiveness ofhome respiratory polygraphy for thediagnosis of sleep apnoea and hypop-noea syndrome. Thorax; 66: 567–573.

N Montserrat JM, et al. (2001). Effective-ness of CPAP treatment in daytimefunction in sleep apnea syndrome: arandomized controlled study with anoptimized placebo. Am J Respir Crit CareMed; 164: 608–613.

N Sharma SK, et al. (2011). CPAP for themetabolic syndrome in patients withobstructive sleep apnea. N Engl J Med;365: 2277–2286.


N Societe de Pneumologie de LangueFrancaise. (2010). Recommandation pourla pratique Clinique. Syndrome d’apneeshypopnees obstructive du sommeil del’adulte. [Recommendations for clinicalpractice. Obstructive sleep apnea hypop-nea syndrome in adults]. Rev Mal Respir;27: 806–833.

N Teran-Santos J, et al. (1999). The associa-tion between sleep apnea and the risk oftraffic accidents. Cooperative Group Burgos-Santander. N Engl J Med; 340: 847–851.

N Thornton AT, et al. (2012). AASM criteriafor scoring respiratory events: interactionbetween apnea sensor and hypopneadefinition. Sleep; 35: 425–432.


Quality of life

Maria R. Bonsignore

Quality of life is defined as ‘the overall stateof well-being that individuals experience asassessed by subjective and objectivemeasures of functioning, health andsatisfaction with the important dimensionsof their lives’ (Reimer et al., 2003).Assessment of quality of life in patients withSDB or other sleep-related pathologicalconditions is an important issue, since itallows us to appreciate the real effects of thedisease on several domains of patients’ lives(Reimer et al., 2003; Weaver, 2001).

Quality of life includes many complex items,such as the level of satisfaction related tohousing conditions or current job. In the

medical field, analysis is usually restricted tohealth-related quality of life (HRQoL), i.e.the aspects more strictly correlated with thehealth status of patients with regard to theeffects of their disease. The instruments forHRQoL assessment are usually in the formof questionnaires developed according tothe most common complaints associatedwith poor health in different domains of life(Reimer et al., 2003; Weaver, 2001).

Evaluating HRQoL helps to understand theeffects of a given disease on patientscompared with healthy individuals from thegeneral population, but is even moreimportant in evaluating the effects oftreatment. An effective treatment for acertain disease should not only normalise orimprove a chosen disease indicator but alsocause tangible improvement in the daily lifeof patients, since the latter is the real reasonthat leads patients to seek medical care.Hence, large randomised clinical studiescommonly include evaluation of HRQoLamong their outcomes, since a positiveeffect of treatment on HRQoL is of utmostimportance for patients and for assessmentof cost/benefit ratios of healthcare.

There are several questionnaires availablefor evaluation of HRQoL, but the maindistinction is between ‘generic’ and ‘disease-specific’ instruments. The former are usefulto assess HRQoL in heterogeneouspopulations and compare the effects ofdifferent disease states; however, they canbe relatively insensitive to the effects of aspecific disease or its treatment. Conversely,disease-specific questionnaires have beendeveloped based on complaints reported byaffected patients and provide valuable, in-depth information on the effects of the

Key points

N HRQoL measures the impact of adisease on several domains of life ofthe patients and is an importantoutcome in the evaluation oftreatment.

N HRQoL can be measured by genericand disease-specific instruments,which are generally in the form ofquestionnaires. Both types have beenused in patients with OSAS or othersleep disorders.

N OSAS significantly affects HRQoL andCPAP treatment causes improvementin several domains.

N The effects of OSAS on HRQoL arelimited in elderly patients, in whomHRQoL appears more linked tooccurrence of comorbidities thanto SDB.


disease on different HRQoL domains(Reimer et al., 2003; Weaver, 2001). Severalsuch instruments have been developed forOSA, as well as restless legs syndrome,narcolepsy and insomnia. In 2001, theWorld Health Organization developed theInternational Classification of Functioning,Disability and Health (ICF) and an ‘ICF CoreSet for Sleep Disorders’ is currently understudy in order to identify the best existinginstruments or develop new ones accordingto a standardised method to be used atlarge in the context of a commonbiopsychosocial framework (Gradinger et al.,2011). Due to the large number of genericand specific instruments available andspace limitations, this chapter willsummarise available evidence for HRQoL inpatients with OSAS. The interested reader isreferred to the review by Gradinger et al.(2011) for further information regarding 115questionnaires used to evaluate HRQoL inpatients with OSAS or other sleep disorders.

Generic instruments for HRQoLassessment

The most commonly used genericinstrument for assessment of HRQoL is theMedical Outcomes Study 36-item short-formhealth survey (SF-36) (Ware et al., 1992). It isa self-administered questionnaire coveringeight dimensions of health (physicalfunctioning, role limitation due to physicalproblems, role limitation due to emotionalproblems, social functioning, mental health,energy/vitality, body pain and general healthperception). The SF-36 does not includequestions on sleep, but is widely used andvalidated in different languages. In Europe,the Nottingham Health Profile (NHP) is alsopopular, has the advantage of being easilyadministered and includes sleepassessment (Hunt et al., 1980).

One major problem with generic HRQoLinstruments is that they are insensitive indetecting subtle differences, especially insubjects with very good or very poor HRQoL(so-called ‘ceiling’ and ‘floor’ effects). Theirwidespread use, however, has produced alarge amount of data in the last few decades,which represent solid backgroundinformation on many disease states.

Despite their limited sensitivity in theassessment of sleep complaints, suchquestionnaires have been used to measureHRQoL in the general population andanalyse the effects of OSAS and other sleepdisorders. In the Sleep Heart Health Study(SHHS), mild-to-moderate SDB wasassociated with decreased vitality, whereassevere SDB was associated with low HRQoLassessed by SF-36, with overall scoressimilar to those associated with otherchronic disease states (Baldwin et al., 2001).More recent data from the SHHS suggestthat sleepiness, difficulty in maintainingsleep (DIMS) and low HRQoL are foundespecially in patients with SDB occurring inNREM sleep, while REM-predominantevents seem not to exert detrimental effectson HRQoL (Chami et al., 2010).

In clinical series of OSAS patients, thesubscales ‘physical functioning’, ‘generalhealth’ and ‘role limitation due to physicalproblems’ of SF-36 were shown to correlatewith measures of sleep fragmentation(Goncalves et al., 2004). However, PSGmeasures of OSAS severity correlated weaklywith self-reported measures obtained by theSF-36 (Goncalves et al., 2004; Weaver et al.,2004). Finally, in elderly patients, HRQoL wasunaffected by OSAS and its reduction wasbest explained by occurrence of comorbidities(Martinez-Garcia et al., 2009).

As for the effects of CPAP treatment,Bennett et al. (1999) were the first todemonstrate a positive effect of CPAP on SF-36. More recently, a meta-analysisdocumented improvements in treated versusuntreated OSAS patients in physicalfunction, body pain, energy vitality and thephysical component summary of SF-36 andin NHP part 2 (Jing et al., 2008).

Disease-specific instruments for HRQoL insleep disorders

Disease-specific instruments have beendeveloped in the last two decades toaddress the problem of HRQoL in patientswith sleep disorders (Reimer et al., 2003;Weaver, 2001).

1. The Functional Outcome of SleepQuestionnaire (FOSQ) is a 30-item self-


administered instrument (Weaver et al.,1997). It provides information ondifficulties encountered by patients inperforming selected tasks and rolesbecause of excessive daytimesleepiness (EDS). Domains exploredare: activity level, vigilance, intimacyand sexual relationships, generalproductivity, and social outcome. TheFOSQ has two major advantages, thefirst being the possibility of a ‘did notengage’ answer for activities notperformed by the patient for reasonsother than sleepiness. The secondadvantage is that it explores theintimacy/sexual domain, which is oftendisregarded by other questionnaires.The FOSQ performs well in theassessment of treatment-inducedchanges and a short 10-item versionhas been recently introduced (Chasenset al., 2009).

2. The Calgary Sleep Apnea Quality of LifeIndex (SAQLI) is a 45-itemquestionnaire exploring the followingdomains: daily functioning, socialinteractions, emotional functioning,symptoms and treatment-relatedsymptoms (Flemmons et al., 2002). Itis administered by a trained inteviewer;

3. The Quebec Sleep Questionnaire(QSQ) is a 32-item, self-administeredquestionnaire, similar to the SAQLI(Lacasse et al., 2004).

In general, at least for research purposes,different instruments are used at the sametime, such as the combination of a genericwith a disease-specific questionnaire, inorder to encompass multiple aspects ofHRQoL, often with the addition of the ESSand the Beck Depression Inventory (BDI).Depression is a common finding in patientswith sleep disorders and may independentlycontribute to worsening HRQoL (Harris etal., 2009).

From the point of view of the clinician, whilesimpler and more effective instruments arecurrently under development (Gradinger etal., 2011), it is important to carefully andsympathetically listen to patients’complaints regarding not only physical

symptoms, but also emotional and socialdisturbances. The physician should be awareof the concept and utility of HRQoLassessment in daily practice and be able toemploy validated instruments to objectivelydocument HRQoL before and aftertreatment. Some data also indicate that bedpartners of OSAS patients improve theirquality of life when CPAP treatment isstarted, suggesting that the HRQoLimplications of OSAS treatment extend tofamily members of the patients. Finally,assessment of new treatments must includeHRQoL evaluation, as only treatments withpositive impact on HRQoL are really ofbenefit for the patients.

Further reading

N Baldwin CM, et al. (2001). The associationof sleep-disordered breathing and sleepsymptoms with quality of life in the SleepHeart Health Study. Sleep; 24: 96–105.

N Bennett LS, et al. (1999). Health status inobstructive sleep apnea: relationship withsleep fragmentation and daytine sleepi-ness, and effects of continuous positiveairway pressure treatment. Am J RespirCrit Care Med; 159: 1884–1890.

N Chami HA, et al. (2010). Sleepiness,quality of life, and sleep maintenance inREM versus non-REM sleep-disorderedbreathing. Am J Respir Crit Care Med;181: 997–1002.

N Chasens ER, et al. (2009). Developmentof the FOSQ-10: a short version of thefunctional outcomes of sleep question-naire. Sleep; 32: 915–919.

N Flemmons WW, et al. (2002). Measure-ment properties of The Calgary SleepApnea Quality of Life Index. Am J RespirCrit Care Med; 165: 159–164.

N Goncalves MA, et al. (2004). Obstructivesleep apnea syndrome, sleepiness, andquality of life. Chest; 125: 2091–2096.

N Harris M, et al. (2009). Obstructive sleepapnea and depression. Sleep Med Rev; 13:437–444.

N Hunt SM, et al. (1980). A quantitativeapproach to perceived health status: avalidation study. J Epidemiol CommunityHealth; 34: 281–286.


N Jing J, et al. (2008). Effect on quality of lifeof continuous positive airway pressure inpatients with obstructive sleep apneasyndrome: a meta-analysis. Lung; 186:131–144.

N Lacasse Y, et al. (2004). A new standar-dised and self-administered quality of lifequestionnaire specific to obstructivesleep apnea. Thorax; 59: 494–449.

N Martinez-Garcia MA, et al. (2009).Obstructive sleep apnea has little impacton quality of life in the elderly. Sleep Med;10: 104–111.

N Reimer MA, et al. (2003). Quality of life insleep disorders. Sleep Med Rev; 7: 335–349.

N Ware JE Jr, et al. (1992). The MOS 36-itemshort-form health survey (SF-36). I.

Conceptual framework and item selec-tion. Med Care; 30: 473–483.

N Weaver EM, et al. (2004). Polysomno-graphy vs self-reported measures inpatients with sleep apnea. ArchOtolaryngol Head Neck Surg; 130: 453–458.

N Weaver TE, et al. (1997). An instrument tomeasure functional status outcomes fordisorders of excessive sleepiness. Sleep;20: 835–843.

N Weaver TE. (2001). Outcome measure-ment in sleep medicine practice andresearch. Part 1: assessment of symp-toms, subjective and objective daytimesleepiness, health-related quality of lifeand functional status. Sleep Med Rev; 5:103–128.


Surgical and anaesthesia riskassessment

Maria R. Bonsignore

OSA patients may encounter problemswhen undergoing general anaesthesia andsurgical procedures (Vasu et al., 2012).Retrospective studies in surgical patientshave reported increased respiratorycomplications in patients with a pre-operative oxygen desaturation index (ODI).5 events?h-1 (Hwang et al., 2008) or a highincidence of perioperative complications,especially occurrence of periods of oxygendesaturation, in OSA compared with non-OSA patients (Liao et al., 2009). In a verylarge retrospective study, Memtsoudis et al.(2011) reported a high incidence ofperioperative pulmonary complications inOSA patients undergoing orthopaedic orgeneral surgical procedures, includingaspiration pneumonia, acute respiratorydistress syndrome (ARDS) and intubation/mechanical ventilation.

The common association of OSA withobesity makes its prevalence especially highin patients undergoing bariatric surgery.Such surgical procedures are increasinglycommon due to the rising prevalence ofmorbid obesity. Screening of these patientsby cardiorespiratory monitoring orsimplified devices is nowadays standardpractice before surgery, and OSA is usuallydetected and treated pre-operatively.Anaesthesiologists are aware of OSA in thispopulation and can adopt preventivestrategies to control perioperative risk.

The real problem is that OSA is largelyunderdiagnosed in the general population,and screening procedures cannot beextensively applied to all patientsundergoing surgery. Therefore,anaesthesiologists need simple and reliabletools to recognise patients at risk and

Key points

N The high prevalence of unrecognisedOSA in the general population andthe increased OSA-associatedperioperative risk have alertedanaesthesiologists to the need todetect patients at high risk of OSA.

N Short clinical questionnaires areuseful in identifying patients at highrisk of OSA, and several protocolshave been developed for theirperioperative management.

N OSA patients may undergoproblems at intubation (difficultairways) or extubation (slow orinsufficient recovery of upper airwaymuscle tone); the most frequentlyencountered problem is thedevelopment of hypoxaemia in bothimmediate and delayed post-surgical periods.

N Use of opioids for analgesia mayprecipitate hypoventilation and/orcentral apnoeas and should beavoided. Day surgery proceduresshould also be avoided in patientsat risk.

N Patients with known OSA treatedwith CPAP do not show increasedrisk and should continue CPAP usein the perioperative period.

N Post-operative use of fixed orautomatic CPAP in high-risk patientsnot previously identified and treatedfor OSA is not recommended.


prevent OSA-related perioperativecomplications. This is especially important ifthe burden of OSA comorbidities isconsidered. A recent prospectiveobservational study estimated that about 24%of patients undergoing elective non-upperairway surgery are at high risk of OSA, and thelarge majority of them are undiagnosed at thetime of surgical evaluation.

For these reasons, practice guidelines onperioperative management of OSA patientswere published by a Task Force of theAmerican Society of Anesthesiologists (ASA)in 2006. Recent updates and algorithms tobe used in the perioperative period werepublished in 2010 (Seet et al. 2010;Adesanya et al., 2010).

Questionnaires for clinical screening canfacilitate pre-operative identification of OSApatients (table 1 (see also the‘‘Questionnaires on sleep’’ section of thisbook)). The ASA checklist is different fromthe others, as it considers the patient profilebut also the need for light or deepanaesthesia according to the plannedsurgery. The Berlin questionnaire and theSTOP-Bang score are excellent in identifyingpatients with an AHI .30 events?h-1, but areless useful in mild disease. The STOP-Bangscore is increasingly popular, since it isbased on four simple questions related toSnoring, Tiredness, Observed apnoea, andblood Pressure, to which BMI, Age, Neckcircumference, and male sex information isadded by the clinician; all questions haveyes/no answers. Other questionnaires arealso used (table 1). The reader is referred toa recent meta-analysis for a comprehensivereview of clinical screening instruments insurgical patients (Ramachandran et al., 2009).

The most common problems encounteredin OSA patients are:

1. Intubation could be difficult due toupper airway anatomy or excess fatdeposits in the neck and occipitalregion. A prospective study in obesesubjects undergoing bariatric surgeryreported no intubation problems whena ramp position was used during theintubation manoeuvre. The ramp

position is obtained by elevating thehead, upper body and shoulders abovethe chest, in order to improve laryngo-scopic view (Neligan et al., 2009).

2. Difficult mask ventilation (DMV) undergeneral anaesthesia was recently foundto predict OSA in a prospective pilotstudy (Plunkett et al., 2011).

3. General anaesthesia causes markedhypotonia of upper airway muscles,thereby increasing the risk ofperioperative complications. OSApatients are prone to develop upperairway problems at relatively mildlevels of anaesthesia, which can befurther worsened by depression ofrespiratory drive with lowresponsiveness to chemical stimuli. Allthese factors concur in increasing therisk of oxygen desaturations afterextubation in the immediate post-operative period, and in the followingdays, when a rebound in REM sleepoccurs post-surgery.

4. Use of opioid medications foranalgesia may favour occurrence ofhypoventilation and hypercapnia orcentral apnoeas post-operatively, andshould be limited in high-risk patients,with preference for short-action drugs.

The management strategy for patients athigh risk of OSA, in view of surgicalprocedure, is not yet evidence-based. Thecurrent attitude is to reserve special care topatients identified as high-risk patients orpatients with known OSA. In these high-riskpatients without a previous diagnosis ofOSA, it has been suggested to delay electivesurgery in order to perform diagnosticstudies and start CPAP treatment pre-operatively. Patients on CPAP with a goodcompliance to treatment have been reportedto show a limited risk compared with OSApatients not previously diagnosed orshowing a low compliance to CPAPtreatment. These findings are possiblyexplained by the positive effects of CPAP onupper airway oedema and stability.

Precautions should be implemented in theperioperative management of OSA patients.During surgery, monitoring of bloodpressure and arterial blood gases may be


indicated in high-risk patients. Extubationshould be performed in the semi-recumbentposition to improve airway patency,preferentially when the patient is awake andmuscle tone is fully re-established.Immediately after surgery, respiration andoxygen saturation should be closelymonitored for a sufficient time, in order todetect respiratory complications.Occurrence of recurrent respiratory eventsin the immediate post-operative period hasbeen recognised to be associated with anincreased risk of oxygen desaturation in thesubsequent period, warranting furthermonitoring during the post-surgical periodin the hospital ward. For these reasons, day-surgery procedures are not recommended inhigh-risk patients, especially becauseevidence-based information is not yetavailable on its safety.

Similarly, the use of CPAP is recommendedpost-operatively in patients already ontreatment, but evidence regarding theopportunity to use CPAP in surgical patientswho had not been previously identified asOSA patients is poor. It is worth noting thatuse of opioids for pain control may favouroccurrence of central events, which are noteffectively prevented by CPAP. A similarreasoning can be applied to the use ofautomatic CPAP devices in the post-operative setting, which is not currentlyrecommended.

In summary, increased awareness of theperioperative complications associated withpreviously unrecognised OSA has promptedanaesthesiologists to routinely use standardquestionnaires to detect high-risk patients

and adopt preventive strategies to limitperioperative risk. The most frequentlyreported complication in OSA patients ispost-operative hypoxaemia, but difficultairway control or excessive loss of upperairway muscle tone can also occur.Measures to increase safety include: pre-operative screening for OSA in high-riskpatients and delay of surgery in order tostart CPAP pre-operatively; use of short-acting anaesthetic drugs and extubationwhen the patient is awake, and avoidance ofopioids for post-operative analgesia.Observation for an extended time post-operatively is useful to detect patients whoneed prolonged monitoring after surgery.

Further reading

N Adesanya AO, et al. (2010). Perioperativemanagement of obstructive sleep apnea.Chest; 138: 1489–1498.

N Chung F, et al. (2008a). Validation of theBerlin Questionnaire and AmericanSociety of Anesthesiologists checklist asscreening tools for obstructive sleepapnea in surgical patients. Anesthesio-logy; 108: 822–830.

N Chung F, et al. (2008b). STOP question-naire. A tool to screen patients forobstructive sleep apnea. Anesthesiology;108: 812–821.

N Chung F, et al. (2008c). Patients withdifficult intubation may need referral tosleep clinics. Anesth Analg; 107: 915–920.

N Finkel KJ, et al. (2009). Prevalence of undi-agnosed obstructive sleep apnea amongadult surgical patients in an academicmedical center. Sleep Med; 10: 753–758.

Table 1. Clinical screening tools for OSA in surgical patients.

Validated Self-administered Reference

Berlin questionnaire Yes Yes Chung et al., 2008a

ASA checklist Yes No Gross et al. 2006; Chung et al., 2008a

STOP-Bang Yes Yes Chung et al., 2008b

SACS Yes No Flemons et al., 1994

PSAP score Yes No Ramachandran et al., 2010

SACS: Sleep apnoea clinical score; PSAP: perioperative sleep apnoea prediction.


N Flemons WW, et al. (1994). Likelihoodratios for a sleep apnea clinical predictionrule. Am J Respir Crit Care Med; 150:1279–1285.

N Gross JB, et al. (2006). Practice guide-lines for the perioperative managementof patients with obstructive sleep apnea:a report by the American Society ofAnesthesiologists Task Force on Peri-operative Management of patients withobstructive sleep apnea. Anesthesiology;104: 1081–1093.

N Hwang D, et al. (2008). Association ofsleep-disordered breathing with post-operative complications. Chest; 133:1128–1134.

N Kapur V, et al. (2002). Underdiagnosis ofsleep apnea syndrome in U.S. commu-nities. Sleep Breath; 6: 49–54.

N Liao P, et al. (2009). Postoperativecomplications in patients with obstruc-tive sleep apnea: a retrospective matchedcohort study. Can J Anesthesiol; 56: 819–828.

N Memtsoudis S, et al. (2011). Perioperativepulmonary outcomes in patients withsleep apnea after noncardiac surgery.Anesth Analg; 112: 113–121.

N Neligan PJ, et al. (2009). Obstructivesleep apnea is not a risk factors fordifficult intubation in morbidly obesepatients. Anesth Analg; 109: 1182–1186.

N Plunkett AR, et al. (2011). Does difficultmask ventilation predict obstructive sleepapnea? A prospective pilot study to identifythe prevalence of OSA in patients withdifficult mask ventilation under generalanesthesia. J Clin Sleep Med; 7: 473–477.

N Ramachandran SK, et al. (2009). A meta-analysis of clinical screening tests forobstructive sleep apnea. Anesthesiology;110: 928–939.

N Ramachandran SK, et al. (2010).Derivation and validation of a simpleperioperative sleep apnea predictionscore. Anesth Analg; 110: 1007–1015.

N Seet E, et al. (2010). Management ofsleep apnea in adults – functional algo-rithms for the perioperative period:Continuing Professional Development.Can J Anesth; 57: 849–864.

N Vasu TS, et al. (2012). Obstructive sleepapnea syndrome and perioperative com-plications: a systematic review of theliterature. J Clin Sleep Med; 8: 199–207.


Comorbidity assessment

Maria R. Bonsignore

Comorbidities are frequent in OSA patientsand contribute to increased health costs inthe five years preceding the diagnosis of

OSA (Smith et al., 2002). Suchcomorbidities include: hypertension,congestive HF, cardiac arrhythmias,coronary artery and peripheral arterialdisease, COPD and depression. The highprevalence of cardiovascular comorbiditieshighlights the potential role of OSA inaccelerating establishment and progressionof atherosclerotic lesions (Drager et al.,2011; Grote et al., 2010). Moreover, thecommon association of OSA with obesityincreases the risk of type II diabetes,dyslipidaemia and metabolic syndrome.Since OSA may independently contribute tothe pathogenesis of metabolic disorders(Bonsignore et al., 2012), a modern clinicalapproach to OSA patients should aim atassessing and correcting major comorbiditiesthat might contribute to the increasedcardiovascular morbidity and mortality insevere untreated OSA (Marin et al., 2005).

Systemic hypertension

Although about 50% of OSA patients areknown to be hypertensive at diagnosis, asshown by several studies in both clinicalseries and the general population(Bonsignore et al., 2010), assessment ofblood pressure profile during sleep andwakefulness should be routinely obtained inthe diagnostic workup of all patients. It hasbeen convincingly shown that maskedhypertension, i.e. normal office bloodpressure but increased 24-h blood pressurelevel, is frequent in untreated OSA patients.Baguet and coworkers obtained ambulatoryblood pressure monitoring (ABPM) for 24 hin 111 OSA patients, and found that normalblood pressure, masked hypertension, andoffice hypertension each accounted forabout one-third of cases, whereas ‘‘white

Key points

N Assessment of comorbidities shouldbe part of the diagnostic workup ofpatients with OSA.

N Since untreated OSA is associatedwith increased cardiovascularmorbidity and mortality, special careshould be given to assesshypertension, cardiac hypertrophy andother cardiovascular diseases, such ascoronary or vascular problems.

N Assessment of COPD in smokers isadvisable, since co-existence of OSAand COPD (overlap syndrome)worsens respiratory function.

N Depression is common and oftenoverlooked in OSA patients and mayimprove with OSA treatment.

N Metabolic abnormalities represent animportant clinical topic. Assessmentof the metabolic syndrome by clinicalcriteria is simple and inexpensive,and allows risk stratification in theclinical setting.

N There are no formal guidelines on theassessment of comorbidities in OSApatients, and information in thissection should be consideredsuggestive, not mandatory, and basedon current knowledge and medicalcommon sense.


coat’’ hypertension, i.e. high office bloodpressure with normal 24-h blood pressurelevel, was rarely observed (Baguet et al.,2008). Therefore, ABPM should be obtainedin OSA patients with normal office bloodpressure in order to detect unknownhypertension. ABPM should also beobtained in known hypertensivesundergoing treatment, to check whetherblood pressure is adequately controlled bytreatment. OSA is frequently associated withresistant hypertension, defined as lack ofnormalisation of blood pressure despite useof three or more anti-hypertensivemedications. Finally, because poorlycontrolled blood pressure is associated withincreased organ damage (left ventricularhypertrophy, vascular dysfunction, renaldamage), detecting abnormal bloodpressure values should prompt thephysician to extend assessment, and tomonitor return towards the normal range ofcardiovascular variables after initiation ofCPAP treatment.

Cardiac structure and function

Several studies have assessed the role ofOSA in causing cardiac hypertrophy withvariable results, possibly explained by theconcomitant effects of hypertension andobesity. Interestingly, echocardiographicstudies suggest that cardiac hypertrophyinvolves not only the left but also the rightventricle (Baguet et al., 2008; Hanly et al.,1992), indirectly suggesting a role of OSA inpulmonary hypertension. Therefore,echocardiography appears clinically usefuland informative, although its use in OSA

patients is not evidence-based or currentlyrecommended by any guideline.

A dose–response effect between OSAseverity, in particular the degree of nocturnalhypoxaemia, and degree of left ventricularhypertrophy, has been found in studies inOSA patients free from clinically evidentcardiovascular abnormalities (Baguet et al.,2009). OSA appeared to be associated withcardiac remodelling and altered diastolicfunction, and to exert an additive effect tothat of increased blood pressure in patientswith both OSA and hypertension (Drageret al., 2007).

Most of the studies on the effects of CPAPon cardiac abnormalities found significantimprovement of cardiac function, especiallydiastolic, after OSA treatment (Noda et al.,1995; Shivalkar et al., 2006). A recent long-term follow-up study documented aprogressive improvement in cardiacremodelling in OSA patients on CPAPtreatment for a year and good compliance totreatment, i.e. nightly use .4.5 h (Colishet al., 2012). These results are in line withthose obtained by a randomised controlledstudy in patients with an average nightly useof CPAP .6 h (Arias et al., 2005). However,unchanged systolic and diastolic functionswere found in patients with poor complianceto CPAP treatment (Akar Bayram et al., 2009).In patients with HF and OSA, there is evidencethat CPAP treatment improves left ventricularejection fraction (Kasai et al., 2011).

Periodic assessment with ECG andechocardiography might also be indicated inOSA patients with atrial fibrillation who fail

Table 1. Diagnosis of the metabolic syndrome according to NCEP-ATP III criteria.

Elevated waist circumference In Caucasian men: .94 cm, women .88 cm

Elevated triglycerides o150 mg?dL-1, or treatment

Reduced HDL-cholesterol f40 mg?dL-1 in men, f50 mg?dL-1 in women, ortreatment

Elevated blood pressure Systolic o130 mmHg and/or diastolic o85 mmHg

Elevated fasting glucose o100 mg?dL-1, or treatment

A clinical diagnosis of the metabolic syndrome is based on occurrence of at least three of the criteria listed in

the table. Modified from Alberti et al., 2009 with pemission from the publisher.


to maintain sinus rhythm after catheterablation (Ng et al., 2011). Weaker evidencesuggests that CPAP treatment may alsoprotect against recurrence of atrialfibrillation in these patients.

Coronary artery disease and peripheralarterial disease

Clinical history should focus on previous orcurrent symptoms of coronary artery disease(CAD) or peripheral arterial disease (PAD)and their treatment, since patients with OSAand atherosclerotic disease may significantlybenefit from OSA treatment. There is clinicaland experimental evidence that intermittenthypoxia and inflammation worsen vascularfunction and accelerate progression ofatherosclerosis. Obesity, however, isassociated with multiple metabolicabnormalities that worsen the circulatinglipid profile, and intermittent hypoxia couldcontribute to worsening of insulin resistanceand plasma lipids (Bonsignore et al., 2012).Although definite clinical evidence is stillmissing, recognising patients potentially athigh risk for cardiovascular complications isa major task of the clinicians dealing withOSA patients. Care should be taken toinsure optimal treatment of such conditions.

Chronic obstructive pulmonary disease

Since COPD and OSA are known tosynergistically worsen respiratory function(‘‘overlap syndrome’’), smoking habitsshould be carefully investigated andspirometry and lung volume measurementsshould be obtained in patients suspected ofhaving both diseases. Both OSA and COPDinvolve inflammatory activation, but little isknown about the mechanisms at play whenthey co-exist in the same patient(McNicholas, 2009). Some studies suggestthat pulmonary hypertension is more severein the overlap syndrome. Nocturnal oxygensaturation should be monitored duringfollow-up, as these patients may requiresupplemental oxygen during sleep besidesCPAP to maintain oxygen saturation .90%.

Depression

The frequent association between OSA anddepression has been known for a long time,

but a clear cause–effect relationshipbetween these two conditions has not beenshown yet, possibly because of themultifactorial pathogenesis of both diseases(see Ejaz et al., 2011 for review). Depressionhas been evaluated by several tools indifferent studies; its prevalence is generallyhigher (up to five-fold) in OSA patientscompared with the general population. Poorsleep quality and sleepiness may play a role,as may nocturnal hypoxaemia and cognitivedysfunction. The effects of CPAP treatmenton depression are also unclear. Use of self-administered questionnaires to assessdepression and anxiety is advisable inpatients with OSA, together with carefulenquiry on the use of antidepressant drugsor benzodiazepines, since such drugs mayexert detrimental effects on upper airwaymuscle tone during sleep.

Obesity and metabolic disorders

Central (or visceral) obesity is the mostfrequent comorbidity associated with OSAand its evaluation should be part of theroutine management of OSA patients(Bonsignore et al., 2012; Lam et al., 2010).BMI, neck circumference and waist-to-hipratio should be routinely measured in OSApatients during clinical assessment, andhistory should include questions aboutdiabetes or other endocrinological disorders(i.e. thyroid disease, polycystic ovarysyndrome, etc.) associated with increasedfat deposition.

Because OSA patients often show disturbedglucose metabolism, fasting blood glucoseand insulin should be routinely measured,and the homeostasis model assessment(HOMA) index calculated as a measure ofinsulin resistance. Patients without knowndiabetes should undergo an oral glucosetolerance test, and glycosylatedhaemoglobin should be measured in allpatients since it appears as a sensitive indexof abnormal glucose metabolism in OSA.Similarly, plasma lipids (total cholesterol,triglycerides and high density lipoprotein(HDL)-cholesterol) are simple routinemeasurements to be included in theassessment of OSA patients. Finally,because of the frequent occurrence of


hepatic steatosis, liver function should alsobe investigated by ultrasound andmeasurement of hepatic enzymes, althoughthe latter are not sensitive for the diagnosisof non-alcoholic fatty liver disease (NAFLD).

The metabolic syndrome, according to thesimple clinical criteria National CholesterolEducation Program (NCEP)-Adult TreatmentPanel III (ATP III) (table 1), should beassessed in all patients as a surrogatemarker of cardiovascular and metabolic risk,since most studies to date havedocumented progressive metabolicworsening with OSA severity (Bonsignoreet al., 2012; Lam et al., 2010).

The role of CPAP treatment in reversingmetabolic abnormalities is unclear(Bonsignore et al., 2012; Lam et al., 2010).Most studies to date have failed to showimprovement in insulin resistance ormetabolic variables after short-term OSAtreatment. It is possible that glycosylatedhaemoglobin is a more sensitive marker formetabolic improvement after CPAP, butmore studies are necessary to obtain furtherevidence on this point. It is likely that weightloss during CPAP treatment or complianceto CPAP may modulate metabolic variablesin OSA patients, complicating the analysis

of metabolic changes independently linkedto OSA. It is also worth underlining thatmost randomised controlled studies on theeffects of CPAP on metabolism were short-term, and changes may require a longer timeto become evident. In addition, thecharacteristics of the patients (i.e. thepercentage of diabetic or obese subjects inthe different studies) probably contribute tovariability of results.

Summary

Assessment of comorbidities is increasinglyimportant in the management of OSApatients. The evaluation of these subjectsshould aim at correcting not only SDB, butalso the potential sources of increased risk,especially in young patients. Table 2suggests a possible chart to use in thediagnostic workup and during follow-up; inthe absence of clinical guidelines, repetitionof tests during follow-up should be limitedto patients with initial abnormal values (inparentheses in the table) in order to limitcosts. A full assessment of OSA patients atdiagnosis is crucial to define the besttherapeutic options for OSA, and motivatethe patient to lose weight or adhere toCPAP treatment.

Table 2. Assessment of comorbidities.

Comorbid condition Tests When to do them

Systemic hypertension Office blood pressure, ABPM Diagnosis (follow-up)

Cardiac hypertrophy,atrial fibrillation

Echocardiography, ECG Diagnosis (follow-up)

Atherosclerosis (coronary orperipheral arterial disease)

ECG, Doppler Diagnosis (follow-up)

COPD Spirometry, lung volumemeasurements, arterial blood gases

Diagnosis (follow-up)

Depression Questionnaires Diagnosis (follow-up)

Obesity BMI, neck circumference, waist-to-hipratio, plasma lipids, hepatic enzymes,

liver ultrasound


Diabetes Fasting blood glucose, insulin,HOMA index, oral glucose tolerance

test, glycosylated haemoglobin(HbA1c)


Metabolic syndrome NCEP-ATP III Diagnosis (follow-up)


Further reading

N Akar Bayram NA, et al. (2009). Effects ofcontinuous positive airway pressure ther-apy on left ventricular function assessedby tissue Doppler imaging in patientswith obstructive sleep apnoea syndrome.Eur J Echocard; 10: 376–382.

N Alberti KG, et al. (2009). Harmonizingthe metabolic syndrome: a joint interimstatement of the International DiabetesFederation Task Force on Epidemiologyand Prevention; National Heart, Lung, andBlood Institute; American Heart Associa-tion; World Heart Federation; InternationalAtherosclerosis Society; and InternationalAssociation for the Study of Obesity.Circulation; 120: 1640–1645.

N Arias MA, et al. (2005). Obstructive sleepapnea syndrome affects left ventriculardiastolic function. Effects of nasal con-tinuous positive airway pressure in men.Circulation; 112: 375–383.

N Baguet JP, et al. (2008). Masked hyper-tension in obstructive sleep apnea syn-drome. J Hypertens; 26: 885–892.

N Baguet J-P, et al. (2009). Early cardiovas-cular abnormalities in newly diagnosedobstructive sleep apnea. Vasc Health RiskManag; 5: 1063–1073.

N Bonsignore MR, et al. (2010). Sleepapnoea and systemic hypertension. EurRespir Monogr; 50: 150–173.

N Bonsignore MR, et al. (2012). The adiposetissue in obesity and obstructive sleepapnoea. Eur Respir J; 39: 746–767.

N Colish J, et al. (2012). Effects of contin-uous positive airway pressure on cardiacremodeling as assessed by cardiac bio-markers, echocardiography, and cardiacMRI. Chest; 141: 674–681.

N Drager LF, et al. (2007). Obstructivesleep apnea, hypertension, and theirinteraction on arterial stiffness and heartremodeling. Chest; 131: 1379–1386.

N Drager LF, et al. (2011). Obstructive sleepapnea: an emerging risk factor for ather-osclerosis. Chest; 140: 534–542.

N Ejaz SM, et al. (2011). Obstructive sleepapnea and depression: a review. InnovClin Neurosci; 8: 17–25.

N Grote L, et al. (2010). Early atherosclero-sis and cardiovascular events. Eur RespirMonogr; 50: 174–188.

N Hanly P, et al. (1992). Ventricular functionin snorers and patients with obstructivesleep apnea. Chest; 102: 100–105.

N Kasai T, et al. (2011). Obstructive sleepapnea and heart failure: pathophysiologicand therapeutic implications. J Am CollCardiol; 57: 119–127.

N Lam JC-M, et al. (2010). Diabetes andmetabolic aspects of OSA. Eur RespirMonogr; 50: 189–215.

N Marin JM, et al. (2005). Long-termcardiovascular outcomes in men withobstructive sleep apnoea-hypopnoea withor without treatment with continuouspositive airway pressure: an observationalstudy. Lancet; 365: 1046–1053.

N McNicholas WT. (2009). Chronicobstructive pulmonary disease andobstructive sleep apnea. Am J Respir CritCare Med; 180: 692–700.

N Ng CY, et al. (2011). Meta-analysis ofobstructive sleep apnea as predictor ofatrial fibrillation recurrence after catheterablation. Am J Cardiol; 108: 47–51.

N Noda A, et al. (1995). Cardiac hypertrophyin obstructive sleep apnea syndrome.Chest; 107: 1538–1544.

N Shivalkar B, et al. (2006). Obstructivesleep apnea syndrome: more insights onstructural and functional cardiac altera-tions, and the effects of treatment withcontinuous positive airway pressure. J AmColl Cardiol; 47: 1433–1439.

N Smith R, et al. (2002). What are obstruc-tive sleep apnea patients being treated forprior to this diagnosis? Chest; 121: 164–172.


Polysomnography

Renata L. Riha

The term ‘polysomnography’ (PSG) was firstintroduced by Holland et al. in 1974. Itcomprises the recording, analysis andinterpretation of multiple, simultaneousphysiological parameters that are used inthe diagnosis of sleep disorders. PSG isessential to the understanding of normaland abnormal sleep.

In 1968, Rechtschaffen and Kales conveneda panel of experts to agree on a standardisedmanual for recording and scoring the sleepEEG. This manual then became the ‘‘goldstandard’’ for scoring the sleep EEG until itwas superseded in 2007 by the AmericanAcademy of Sleep Medicine (AASM) manualfor the scoring of sleep and associatedevents. This manual updates the ‘R&K’,incorporating biological principles as well aspublished evidence on interpretation of thevarious signals used in PSG.

The polysomnogram is a summary output ofelectrophysiological signals integrating sleepsignals, respiratory signals, cardiovascularsignals and movement. The recording ofsleep states requires acquisition of three mainmeasures: the EEG, EOG and EMG.

Respiration is monitored by oronasal flow aswell as respiratory effort, movement andsnoring. The effects of respiration are alsomonitored using oximetry and sometimescarbon dioxide monitoring, usuallytranscutaneously. The ECG monitors heartrate and rhythm. Movement is recordedusing EMG on the tibialis muscles and thebody position sensor. These are theminimum requirements for a simple PSG.More elaborate montages are used,particularly in the investigation ofparasomnias and nocturnal seizure disorders.

Sleep EEG

The recording of electrical potentialsgenerated by the cortex and deep structures,especially the thalamus, is the primarypurpose of the EEG. The relative potentialdifference between two recording electrodesforms the basis of the trace. One electrodeis negative compared with the other. Notethat negative charges are upwardlydeflecting. The left-sided channels on thehead are odd-numbered and the rightchannels even-numbered. The 10–20international electrode placement developedfor the monitoring of EEG signals in theinvestigation of epilepsy is the standardmontage from which sleep montage isderived. The R&K sleep montage initiallyused the following placements (fig. 1).

N C4–A1N C3–A2N EOG on the left eye referenced to A1N EOG on the right eye referenced to A1N EMG mentalis

The AASM (2007) electrode placement is asfollows (fig. 2).

N F4–M1 (best for d waves)N C4–M1 (best for capturing spindles)N O2–M1 (best for capturing a rhythm)

Additional electrode placement includes F3–M2, C2–M2 and O1–M2. The general rule isto read from the left cortical channels withthe other side as a back-up.

The EOG

The EOG records the corneoretinal potentialdifference. Note that it is not a recording ofthe movement of eye muscles.


The electrodes at the outer canthus of theright eye (ROC) are offset by 1 cm above thehorizontal plane and, likewise, the LOC atthe outer canthus of the left eye is offset by1 cm below the horizontal plane. The ROC isusually referenced to A1 and the LOC isgenerally referenced to A2. Additionalelectrodes can be placed; for instance, whenthe patient is undergoing multiple sleeplatency testing, when the capture of REMsleep is particularly important.

The EMG

Up to three EMG electrodes can be placedon the face to allow for an alternativeelectrode should one malfunction. The toneof the mentalis and submentalis muscles ismonitored, and is essential to the diagnosisof REM sleep. An electrode can sometimesbe placed over the masseter muscle torecord bruxism.

Sensors and associated monitoring

A variety of electrodes can be used to recordthe EEG, EOG and EMG. For the EEG, typesthat can be used include gold-cup, silver-cupand disposable electrodes. Collodion-soakedgauze has been one method usedextensively in the past to attach electrodesto the scalp.

When applying sensors, the aim is to keepimpedances ,5 kV. The ECG is acquired usingelectrodes placed in standard configuration.

Measurement of rib cage and abdominalmovements can be undertaken using avariety of techniques. Inductiveplethysmography is one such technique andexamines the change in cross-sectional area

of thoracic and abdominal compartments(two-compartment model of thoracic cage)using two bands: one placed around thethorax and one around the abdomen. Ineach coil, there is an inductive band whoseelectromagnetic properties depend on thearea enclosed by the band. This allows forassessment of the magnitude of ahypopnoea or apnoea. However, the majorpractical problem arises from positioning ofthe bands on the body. Thoracoabdominalbands can be used to semiquantitativelyassess the magnitude of the inspiratoryeffort in upper airway obstruction, andanalysis is undertaken of the amplitude ofswing movements of thoracic andabdominal compartments, and comparisonof synchrony between both movements. Themeasurement is semiquantitative. Othertechnology used to measurethoracoabdominal movement includesstrain gauges, piezoelectric transducers andpneumatic bands. Oesophageal monitoring

A1

A1 Right eye–A1

Left eye–A1E1E2

C4-A1

EMG

A2

A2

C4 C3

Figure 1. Placement of electrodes for sleep stagingaccording to R&K (1968).

Leftside

Rightside

Back

Front Vertex

Figure 2. AASM (2007) electrode placement.

Table 1. Example of an instruction series prior tocommencing the PSG recording.

1) With eyes open, look straight ahead for30 s

2) With eyes closed, look straight ahead for30 s

3) Holding head still, look to the left andright; up and down

4) Holding your head still, slowly blink youreyes five times

5) Grit your teeth; clench your jaw

6) Inhale and exhale

7) Hold your breath for 10 s

8) Flex your right foot; flex your left foot


is quite invasive and is not used routinely inthe assessment of sleep disorderedbreathing on a clinical level.

Various techniques have also been developedin order to assess flow. For a very long time,thermistors and thermocouples were usedalmost exclusively in measuring flow. Theycomprise small sensors whose electricalcharacteristics (resistance and voltage)depend on their temperature. The device isplaced close to the airway opening and flow isdetected by measuring the variation in theelectrical properties of the sensor as exposed

to the temperature changes in flow at theairway (room air temperature duringinspiration and at 37uC at expiration). Thesignal is semiquantitative, has a poordynamic response and there is a nonlinearrelationship between electrical signal andairflow dependent on exact position at theairway opening. These devices cannotaccurately quantify magnitude of flow but theycan detect apnoeas. In many respects, theyhave been superseded by nasal prongs, whichrecord pressure at the nostrils. Theequipment comprises conventional prongsconnected to a pressure transducer. Airflow

Table 2. Wake stage (stage W).

Recording Characteristics

EOG Voluntary control, rapid eye movements or none, blinks, slow rolling eyemovements when drowsy

EEG Eyes open: low voltage, mixed frequency; eyes closed: rhythmic a, prominentin occipital, attenuates with attention

a waves 8–13 Hz, seen mostly in occipital channels, mainly associated withwakefulness, also associated with microarousals, 10–20% of patients will nothave a detectable a rhythm on EEG

EMG High tonic activity; voluntary movement

Table 3. Stage N1 sleep (stage 1 sleep).


EOG Slow rolling eye movements

EEG Low voltage, mixed frequency; may be h rhythm 2–7 Hz, up to 50–75 mVrange; vertex sharp waves (200 mV)

h waves 4–7 Hz; low voltage, mixed frequency backgrounds; often appear as sharpvertex waves; mostly associated with stage 1

EMG Tonic activity, slight decrease compared with waking

Table 4. Stage N2 sleep (stage 2 sleep).


EOG Slow rolling eye movement occasionally near sleep onset

EEG Low voltage, mixed frequency

Sleep spindles Bursts of 12–14 Hz activity, o0.5 s long, no amplitude requirement, mostlyassociated with stage 2

K complexes Sharp negative wave followed by positive component, o0.5 s long, noamplitude requirement, mostly associated with stage 2, maximal over frontalderivations

EMG Tonic activity, low level


EOG

EOG

EEG α waves

EEG

EMG

Figure 3. Wake stage (stage W). Note a-waveson EEG.

Table 5. Stage N3 sleep (stages 3 and 4).


EOG None, reflects EEG

EEG d rhythm .20% of epoch, high amplitude waves (.75 mV), low frequency(f2 Hz), prominent in frontal regions, sleep spindles may persist but notnecessary for scoring

d waves ,4 Hz, .75 Hz in amplitude, must be ,2 Hz to qualify for N3, mostlyassociated with N3

EMG Tonic activity, low level

Table 6. Stage R sleep (REM sleep).


EOG Phasic REMs

EEG Low voltage, mixed frequency, sawtooth waves, h activity, slow a activity

Sawtoothwaves

2–6 Hz serrated bursts of activity, maximal centrally and precede REMs

EMG Tonic suppression, phasic twitches

Phasic muscleactivity

Bursts of EMG activity lasting ,0.25 ms, detected on chin EMG, anteriortibialis, EOG–EEG leads

EOG

EOG

EEG

Spindle K complexEEG

EMG

Figure 4. Stage N2 sleep (stage 2 sleep).

EOG

EOG

EEG

δ wavesEEG

EMG

Figure 5. Stage N3 sleep (stages 3 and 4). Noteslow, high-amplitude d-waves.

EOG

EOG

EEG

REMs

Phasic twitch

Sawtooth wavesEEG

EMG

Figure 6. Stage R sleep (REM sleep). REMs: rapideye movements.


turbulence at the nostrils induces pressurethat is directly related to the magnitude offlow. There is an excellent dynamic response.The relationship between pressure and flow isnonlinear and may result in overestimation offlow magnitude. Unfortunately, thedrawbacks are lack of detection of mouthbreathing, poor signal with a blocked noseand increased resistance with prongs ofinadequate size. However, they are vastlysuperior to thermistors and thermocouplesand are recommended for detecting airflow inmodern montages.

Pulse oximetry is used in monitoringevidence of desaturation. However, no two

oximeter makes are alike and the oximeter isthe subject of continual technologicalevolution. Interestingly, no minimumtechnical specification or standardisation ofsignal processing is available internationallyor nationally for oximeters. The minimumstandard criteria set by AASM are a samplerate of 25 Hz with averaging of three values.A resolution of 0.1% is desirable but notspecified. Oximetry is discussed in greaterdetail later.

Measurement techniques

Sensitivity, filter settings and amplificationsneed to be checked carefully prior to

LOC–Mix R256 µV

ROC–Mix L256 µV

Mix L–Cz256 µV

Mix R–Cz256 µV

Cz–Pz128 µV

Pz–Cz128 µV

EMG1–EMG264 µV

ECG1–ECG22.05 µV

Figure 7. A 30-s page from a PSG showing an arousal from N2 sleep.

Table 7. Definitions of respiratory events.

Event Definition

Obstructiveapnoea

The complete cessation of flow for o10 s with continued respiratory effortthroughout the apnoea

Hypopnoea Minimum 30–50% reduction in thoracoabdominal movement for o10 s(definitions vary and may include 3–4% desaturations with or withoutarousals)

Central event The complete cessation of respiratory movement and flow for 15 s (someauthors use 10 s)

Mixed apnoea The complete cessation of flow for o10 s with respiratory effort initiallyabsent, but returning midway through the apnoea


ECG2.5 mV

EMG31.3 µV

EOG(L)250 µV

EOG(R)250 µV

EEG250 µV

THOR RES×1

ABDO RES×1

Nasal airflow

×1Airflow

×1

LEG(R)20 mV

LEG(L)20 mV

SOUND×1

EMGdia-phragm 62.5 µV

POSITION

SpO2

100

50

FBLR

Figure 8. 5-min page of PSG showing obstructive respiratory events (ObA) accompanied by desaturationsand arousals. THOR RES: thoracic respiratory band; ABDO RES: abdominal respiratory band; F: front;B: back; L: left side; R: right side.

ECG2.5 mVEMG

15.6 µVEOG(R)250 µVEOG(L)250 µV

EEG250 µV

THOR RES×1

ABDO RES×1

Nasal pressure

×1

Airflow×1

LEG(R)20 mV

LEG(L)20 mV

SOUND×1


POSITION

SpO2

Ptc,CO2mmHg

100

100

50

50

FBLR

Figure 9. 5-min page of PSG showing hypopnoeas (Hyp) accompanied by desaturations on the oxygensaturation trace. THOR RES: thoracic respiratory band; ABDO RES: abdominal respiratory band; F: front;B: back; L: left side; R: right side.


recording the PSG (see Spriggs, 2008). Therecording speed of the EEG is generally10 mm?s-1 with epochs of sleep historicallyscored as 30 s of sleep per page. After theelectrode and monitor applications arecomplete, physiological calibrations to assessproper electrode functioning are imperative. Anumber of instruction series are available thatshould be performed prior to the start of everyPSG (table 1). This testing ensures thatelectrodes are placed appropriately and alsoallows for baseline data to be collected forcomparison during sleep scoring.

Scoring the sleep EEG

Each sleep stage is assigned a 30-s epoch.Each epoch is assigned a single state orsleep stage. The sleep stage that comprisesthe greater part of the epoch is assigned.Events can be described within or acrossepochs. The sleep stages are shown intable 2–6 and figures 3–6 with the stage

designated according to the AASM (2007)scoring manual and the R&K (1968)equivalent in brackets. Please consult theAASM and R&K manuals as detailed andexhaustive guides on scoring sleep andassociated events.

EEG arousals

An arousal appears in the EEG as a briefinterruption of sleep continuity. The currentAASM definition is an abrupt shift in EEGfrequency, for instance from h to a or.16 Hz. An arousal should be o3 sduration and should be preceded by o10 sof stable sleep. In stage R, the EMG will beelevated with respect to the arousal.Arousals in non-REM sleep do not require arise in EMG tone to be scored as such. EEGarousals are scored in any stage of sleepexcept wakefulness using the central andoccipital EEG. Arousals cannot be scored onchanges in submental EMG alone.

ECG2.5 mV

EMG15.6 µV

EOG(R)250 µV

EOG(L)250 µV

EEG250 µV

THOR RES×1

ABDO RES×1

Nasal pressure

×1

Airflow×1

LEG(R)20 mV

LEG(L)20 mV

SOUND×1


POSITION

SpO2

100

50

FBLR

Figure 10. 10-min page of PSG showing central apnoeas showing a Cheyne–Stokes respiratory pattern.THOR RES: thoracic respiratory band; ABDO RES: abdominal respiratory band; F: front;B: back; L: left side; R: right side.


Scoring respiratory events

While sleep staging has been standardisedsince 1968 and again more recently in 2007,an international attempt to standardiserespiratory events recorded during sleep andother movement disorders was onlypublished in 1999 and again in 2007.Table 7 details the definitions and types ofsleep related respiratory events commonlyrecorded. Figures 8–10 represent the mostcommonly scored respiratory events.

Please consult the AASM (2007) manual asa detailed and exhaustive guide on scoringrespiratory and associated events.

The severity of sleep apnoea is classifiedusing the AHI, which comprises the totalnumber of apnoeas and hypopnoeas scoredduring sleep divided by the total sleep time(in hours) as recorded using the EEG. Sleepapnoea severity is classified as mild if theAHI is 5–15 events?h-1, moderate if the AHIis .15 but ,30 events?h-1 and severe if theAHI is .30 events?h-1. However, the AHImust be interpreted in the context of actualsleep efficiency (the less sleep time, thesmaller the denominator, with artificialelevation of the AHI) and in the context ofage and sex. The older individual may havegreater number of events with no adverseimpact on health at all.

Scoring movement in sleep

Movements in sleep encompass a largevariety of phenomena, including periodiclimb movements, excessive fragmentarymyoclonus, bruxism, rhythmic rockingmovement disorder and REM sleepbehaviour disorder.

Please consult the AASM (2007) manual asdetailed and exhaustive guide on scoringsleep associated movements.

Evaluating the ECG

In general, during PSG, a single modifiedECG lead II is used with torso placement,but this can be modified as necessarydepending on the type of study beingundertaken. Sinus tachycardia in sleep isdefined as .90 beats?min-1 for adults andbradycardia as ,40 beats?min-1.

Please use Hampton (2008) as a detailed andexhaustive summary of normal and abnormalECG traces, and the AASM (2007) manualas an exhaustive guide on cardiac rulesduring sleep.

Interpretation of PSG

PSG is used in the diagnosis of sleepdisorders, including obstructive sleepapnoea/hypopnoea syndrome, narcolepsy,periodic limb movement disorders and

Table 8. Normative data for sleep stages across the lifespan

Age

20–29 yrs 30–39 yrs 40–49 yrs 50–59 yrs .60 yrs

TST min 374.9 375.8 370.2 366.6 348.8

Sleep efficiency % 94.4 94.4 90.2 90.4 85.8

Sleep latency min 6.3 10.0 8.4 6.1 8.2

Awakenings n 6.3 4.7 8.4 9.7 12.3

Stage R % TIB 22.2 23.1 20.4 20.9 16.4

Stage N1 % TIB 3.0 2.5 4.3 4.7 4.0

Stage N2 % TIB 50.5 52.8 54.6 56.7 57.6

Stage N3 % TIB 18.8 16.1 10.9 8.1 7.7

Sleep efficiency is defined as (TST6100)/TIB. Sleep latency is defined as the time to initial sleep onset from

TIB. TST: total sleep time; TIB: time in bed. Modified from Hirshkowitz (2004).


parasomnias. It is also important for theexclusion of sleep disorders and can beuseful in patients who have sleep statemisperception. PSG is also an essentialresearch tool, providing information on theduration and amount of sleep, the pattern ofsleep, the quality of sleep, and any abnormalbehaviour during sleep.

Once the PSG has been scored, theconsolidated output is formatted in astandardised fashion and should depictboth the hypnogram (summary of sleepstages), and the summary scoring andrecording of events, oxygen saturations,body positions, etc.

Figure 11 depicts a normal PSG andfigure 12 a PSG of someone with severeOSA. When interpreting and reporting the

PSG, the general method is to first look atthe hypnogram and comment on sleeparchitecture, cycling of stages, paucity ofsleep stages and fragmentation of sleep.Table 8 gives an indication of thedistribution of sleep stages across thelifespan. The second stage of interpretingthe PSG is to assess whether there are anyabnormalities in the respiratory, ECG andmovement traces, coupled with anyabnormalities in oximetry or transcutaneouscarbon dioxide monitoring, if relevant. Bodyposition during sleep can affect the degreeof sleep disordered breathing recorded andshould be commented on if relevant (e.g.fig. 12 shows severe sleep apnoea). Moredetailed examination of the raw data, e.g.with extended montage or with video, isadvisable in the case of parasomnias andsuspected seizures.

1234

120

+10

+10

+5

+5+5+5+5+5

50

100

1EpochhTime

01211

241 3612 3

4814

6015

7216

8417

5:36:1322:36:13

RLB

F

20

Heart ratebeats·min-1

PLMs

Leg movements

Sound

Central apnoea

Body Position

Arousals

REMMOV AWK

SpO2 %

Obstructive apnoeaMixed apnoeaHypopnoeaUnsure

Figure 11. PSG report in a person without a sleep disorder, demonstrating normal sleep architecture, sleepprogression and sleep cycling. PLM: periodic limb movement.


Pulse transit time

Pulse transit time (PTT) is based onmeasuring the time required for the arterialpulse wave to travel between two points inthe arterial tree: from the moment when thepulse leaves the aortic valve (R wave onECG) to time when it reaches the vessels inthe finger as identified by pulse oximetry. Allthat is required to undertake PTTmeasurements is a pulse oximeter, ECGleads and a computation unit. Pulse wavespeed depends on the vessel stiffness andstiffness is in turn determined by bloodpressure. During airway obstruction,increased swings in the intrathoracicpressure modulate blood pressure andinduce parallel changes in PTT. IndividualPTT values do not correlate with absolutevalues of pleural pressure; however,oscillations estimate swings of pleuralpressure that occur during the obstructivebreaths with high sensitivity and specificity.PTT can thereby be used with some success

in estimating the severity of sleepdisordered breathing.

Further reading

N American Academy of Sleep Medicine.(1999). Sleep-related breathing disordersin adults: recommendations for syn-drome definition and measurement tech-niques in clinical research. The Report ofan American Academy of Sleep MedicineTask Force. Sleep; 22: 667–689.

N American Academy of Sleep Medicine.The AASM Manual for the Scoring ofSleep and Associated Events: Rules,Terminology and Technical Specifica-tions. 1st Edn. Westchester, AmericanAcademy of Sleep Medicine, 2007.

N Bohning N, et al. (2010). Comparability ofpulse oximeters used in sleep medicinefor the screening of OSA. Physiol Meas; 31:875–888.

N Grigg-Damberger MM. (2009). TheAASM scoring manual: a critical apprai-sal. Curr Opin Pulm Med; 15: 540–549.

RW1234

+10

+5

+5

+5+5+5+5

55

100

1EpochhTime

01211

2412

3613

6015

03.0002.0001.0000.0023.00

22:28:31

RLB

F

PLMs

Snoring

Central apnoea

Body Position

Hypnogram

SpO2 %

Obstructive apnoeaMixed apnoeaHypopnoeaUnsure

4814

03:28:31

Figure 12. PSG report consistent with severe sleep apnoea. The hypnogram shows an absence of short-wavesleep (SWS), poor sleep cycling and sleep fragmentation. There is severe arterial oxygen desaturationassociated with repetitive obstructive events. PLMs: periodic limb movements.


N Hampton JR. The ECG Made Easy. 7thEdn. Edinburgh, Churchill LivingstoneElsevier, 2008.

N Hirshkowitz M. (2004). Normal humansleep: an overview. Med Clin North Am;88: 551–565.

N Moser D, et al. (2009). Sleep classifica-tion according to AASM and Rechts-chaffen & Kales: effects on sleep scoringparameters. Sleep; 32: 139–149.

N Rechtschaffen A, et al. A Manual ofStandardised Terminology, Techniques,and Scoring System for Sleep Stages ofHuman Subjects. Bethesda, NationalInstitute of Neurological Disease andBlindness, 1968.

N Ruehland WR, et al. (2009). The newAASM criteria for scoring hypopneas:

impact on the apnea hypopnea index.Sleep; 32: 150–157.

N Silber MH, et al. (2007). The visual

scoring of sleep in adults. J Clin Sleep

Med; 3: 121–131.

N Sleep disorders Atlas Task Force of the

American Sleep Disorders Association.

(1992). EEG arousals: scoring rules and

examples: a preliminary report from the

Sleep disorders Atlas Task Force of the

American Sleep Disorders Association.

Sleep; 15: 173–184.

N Spriggs W. (2008). Essentials of polysom-

nography. Burlington, Jones & Bartlett.

N The Atlas Task Force. (1993). Recording

and scoring leg movements. Sleep; 16:

748–759.


Assessment of daytimesleepiness

Renata L. Riha

Assessment of daytime sleepiness

The causes of excessive daytimesomnolence (EDS) are many and varied andnot all will be attributable to sleep disorders.However, there are a few very specific sleepdisorders that can be diagnosed with the aidof daytime tests of sleepiness, such asnarcolepsy and idiopathic hypersomnolence.Conversely, it is often useful to assessresponse to treatment, irrespective of thetype of disorder, using a test for ability tomaintain wakefulness.

Multiple sleep latency testing

MSLT is designed to objectively assesssleepiness and is used for the assessmentand diagnosis of disorders of excessive

somnolence and daytime sleepiness. Thecentral premise of the MSLT is that thesleepier a subject is, the more quickly theywill fall asleep. The test is carried out understandardised laboratory conditions. Theinitial guidelines for conducting this test(originally designed for confirming adiagnosis of narcolepsy) were developedand published in 1986 (Carskadon et al.,1986) with the most recent review of theguidelines in 2005 (Littner et al., 2005).Table 1 summarises the instruction seriesfor conducting an MSLT and the changesbetween the two sets of guidelines with time.

The patient wears a standard EEG montageand is asked to lie down and sleep in adarkened room on four to five separateoccasions. Each nap opportunity is 20 minlong and 2 h apart over the course of theday. Sleep is recorded and the time taken tofall asleep is averaged over the number ofnaps taken.

Figures 1 and 2 show a PSG performed thenight before daytime testing of sleepinessusing the MSLT in a narcoleptic patient.

Potential drawbacks of the MSLT includeindividual clinical interpretation ofinstructions which may lead to variation inthe conducting and reporting of the test.This has major implications for diagnosis,disease classification and treatment (seeCarskadon et al., 1986 for more detail).

Maintenance of wakefulness test

The MWT works on the opposite premise ofthe MSLT, in that it measures the ability tostay awake for a defined period of time in astandardised stimulus-free environment.The ability to maintain wakefulness should

Key points

N The Multiple Sleep Latency Test(MSLT) is used to objectively testpropensity to sleepiness in disordersof daytime somnolence.

N The Maintenance of Wakefulness Test(MWT) measures ability to stayawake for a defined period in astandardised environment.

N The OSLER Wake Test is abehavioural version of the MWT inwhich subjects respond to a timedlight stimulus.

N Although they are usefuldiagnostically, no test of sleepinesscan reliably predict sleepiness andperformance in real-life situationswith certainty.


Table 1. Guidelines for the MSLT#.

General considerations

1-2 weeks of sleep diaries preceding MSLT

1 night PSG on habitual schedule

Consideration of drug schedule

Withdrawal of drugs that influence sleep latency

Withdrawal of drugs that influence REM sleep

Urinary drug screen

Loose comfortable street clothing

A minimum of four tests at 2-h intervals beginning 1.5–3 h after wake-up

Quiet, dark, temperature-controlled room

No alcohol or caffeine

Montage

Standard montage

C3–A2 or C4–A1

Right and left horizontal (or oblique) EOG

Mental/submental electromyogram

Strongly recommended

O1–A2 or O2–A1

A vertical EOG

ECG

Optional (respiratory indicators)

Respiratory flow

Respiratory sounds

Patient series

30 min – smoking should be ceased

15 min – vigorous exercise should be ceased

10 min – remove shoes and loosen restrictive clothing

5 min – hooked up in bed

30 s – ask patient to lie quietly, close their eyes and try and fall asleep

Calibration series

Impedances

Calibration

Stanford Sleepiness Scale

Study time schedule

20 min if no sleep

NREM sleep 15 min after sleep onset

REM sleep 15 min after sleep onset

Sleep onset

1st epoch of any stage of sleep"

Interpreting MSLT

,5 min pathological daytime sleepiness

10–20 min normal ranges

5–10 min diagnostic grey area

#: adapted from Carskadon et al. (1986) and Littner et al. (2005); ": in 1986 guidelines: three consecutive

epochs of stage 1 sleep or one epoch of any other sleep stage.


be assessed in the context of the clinicalhistory. A number of protocols forundertaking the MWT are available, but therecommended one uses the 40-min protocolover four periods during the day. In contrastto the MSLT, onset of sleep is defined asthree continuous epochs of stage 1 sleep orone epoch of any other stage of sleep. It isunusual to see REM sleep on this test. For adetailed and exhaustive discussion of thistest please consult Littner et al. (2005).

Non-EEG-based tests of vigilance

There are a number of computer-basedassessments of sustained visual attentionwhich measure the reaction time taken to

press a button in response to a stimulusthat is presented on a screen at randomintervals over a given time period. One ofthe simplest tests to this effect is theOSLER test.

The original OSLER test is the simplest testto assess daytime sleepiness and is abehavioural version of the MWT. The subjectsits in a darkened room; a small LED is litfor 1 s in every three, and the subject isrequired to respond each time. Softwarecontrols the LED, stores the response andalerts the technician when the subject failsto respond. Consecutive misses mean sleep(Bennett et al., 1997). Figure 3 shows theresults of an OSLER test for a sleep subject.These tests are practical and well-validated

+5

70

100

REMW1234

R

LFP

+5+5+5

CnA

Time 12 am 1 am 2 am 3 am 4 am 5 am 6 am 7 am

Hrs 01

1121

2241

3361

4481

5601

6721

7841

8961Epoch

Hypnogram

Body position

PLM

Snorng

23:00.11 07:00.11

Sp,O2

CbAMx.AHyp

Unscored

+10

+5

Figure 1. PSG for a narcoleptic patient; note the early onset REM-sleep and excessive sleep fragmentation.


W1234

REM

09:25:04

Nap 1

W1234

REM

11:25:04

Nap 2

W1234

REM

13:25:04

Nap 3

W1234

REM

15:20:34

09:45:04

11:45:04

13:45:04

15:40:34

Nap 4

Figure 2. MSLT in narcolepsy showing a sleep latency of ,8 min and four sleep-onset REM periodspathognomonic for the condition.

7Misses

Interval 00:40:00 Time min

Run 4, start 16:00:00

Run 3, start 14:00:00

Run 2, start 12:00:00

Run 1, start 10:00:00

6

4

2

0

7Misses

6

4

2

0

7Misses

6

4

2

0

7Misses

6

4

2

00

0:4

0:0

0

00

:35:

00

00

:30

:00

00

:25:

00

00

:20

:00

00

:15:

00

00

:10

:00

00

:05:

00

00

:00

:00

Figure 3. OSLER test showing a mean sleep latency of 21 min. The subject fell asleep on three of four runs.


as alternatives to more complex testing,but are insufficiently standardised regardingnormative population data. None of thetests discussed in this section can be usedunequivocally to predict sleepiness inreal-life situations.

Further reading

N Bennett LS, et al. (1997). A behaviouraltest to assess daytime sleepiness in obstruc-tive sleep apnoea. J Sleep Res; 6: 142–145.

N Carskadon MA, et al. (1986). Guidelinesfor the multiple sleep latency test(MSLT): a standard measure of sleepi-ness. Sleep; 9: 519–524.

N Littner MR, et al. (2005). Practice para-

meters for clinical use of the multiple

sleep latency test and the maintenance of

wakefulness test. Sleep; 28: 113–121.


Cardiorespiratory monitoringduring sleep

Maria R. Bonsignore and Juan Fernando Masa

Full nocturnal PSG in the sleep laboratory isthe gold standard for diagnosis of sleepdisorders, including OSA. However, full PSGis not free from limitations (Kuna, 2010),

since: 1) the signals recorded are notcalibrated and scoring is based onrecognition of qualitative patterns; 2)scoring of sleep and respiratory events showsome problematic points, even after therevision of criteria by the American Academyof Sleep Medicine (AASM) (2005); 3) theresults of PSG correlate poorly with clinicalsymptoms of OSA and measures ofcardiovascular risk; and, consequently, 4)there is no evidence-based threshold for AHIor other variables to be used in comparativestudies between different diagnostic ortherapeutic management strategies (Kuna,2010). In addition, the in-hospital PSG is notcarried out under ‘normal’ sleep conditions,with subsequent alterations in sleep qualityand/or frequency of apnoeic events. Full PSGentails a high cost in both equipment andpersonnel, and its limited availability withrespect to the increasing demand for sleeptests causes long waiting lists.

A full PSG is also indicated for CPAPtitration, further increasing the time betweendiagnosis of OSA and institution oftreatment. These problems have led to theacceptance of ‘split-night’ studies (Epsteinet al., 2009), in which the first part of thePSG is used for diagnosis and the secondpart to titrate CPAP in patients with severeOSA (AHI .20–40 events?h-1). Suchprocedures are reasonably accurate(Khawaja et al., 2010).

Alternative ambulatory methods have beendeveloped for more efficient use ofhealthcare resources. For diagnosis of OSA,several devices are currently available to beworn at home by patients during sleep.Sleep monitoring devices are classifiedaccording to completeness of the signals

Key points

N Ambulatory management of OSApatients has been shown to be aseffective as traditional, laboratory-based management in symptomaticpatients with severe OSA freefrom comorbidities.

N Portable monitoring devices shouldinclude respiratory signals with thesame sensors as those recommendedfor PSG; however, devices based ondifferent technology (i.e. PAT) havebeen found to reliably identify OSA.

N Limited monitoring (1–3 signals) isnot recommended by currentguidelines; however, according torecent data, identification of OSA ispossible by using nasal pressure and/or oximetry signals.

N Actigraphy has been suggested as apossible tool to better define sleeptime, but is not recommended bycurrent guidelines and more studiesare necessary to ascertain its efficacyin ambulatory management of OSA.

N Monitoring of PtcCO2 provides usefulinformation in patients withhypoventilation during sleep.Technological advancement havemade measurement of PtcCO2 easierand more reliable than in the past.


recorded and presence of a technicianovernight (table 1). Type 1 monitoring is thecomplete PSG recording in the sleeplaboratory monitored by a technician, whiletype 2–4 monitoring devices provideunattended recordings from full PSG (type2), to four- to eight-channel (type 3) or one-to three-channel (type 4) recordings.

Type 3 cardiorespiratory monitoring duringsleep is increasingly used in the diagnosis ofOSA. Underdiagnosis of OSA in the generalpopulation and the rising prevalence ofobesity worldwide have increased thenumber of patients referred to sleep centresfor suspected sleep apnoea. In addition, thecontinuing search for cost-effectivediagnostic and therapeutic approachesmakes type 3 portable devices very attractiveto deal with a very large number ofsuspected OSA patients.

A recent survey among physicians workingin academic sleep centres in 22 Europeancountries revealed that type 3cardiorespiratory polygraphy was used in.70% of them at home or in the sleeplaboratory. Type 4 devices were found to beused for diagnosis of OSA in five countries.Cardiorespiratory polygraphy was also usedin a third of the centres to checkeffectiveness of treatment (i.e. in patientsundergoing domiciliary CPAP titration by anautomatic device) (Fietze et al., 2011).

In the USA, the use of portable devices fordiagnosis of OSA is less widespread (Collopet al., 2007; Ahmed et al., 2007). The AASMrecommends that portable type 3cardiorespiratory monitoring for thediagnosis of OSA be used only in patientswith a high clinical probability of OSA andwithout comorbidities (Collop et al., 2007).Similar recommendations were recentlyissued by the Canadian Thoracic Society(Canadian Sleep Society et al., 2010;Fleetham et al., 2011). Recent dataindicating that clinical results are notinferior to those of classic OSAmanagement in the sleep laboratory mightchange such restrictive use in the future (seelater). Several problems remain to beaddressed by future research studies beforeuse of respiratory polygraphy can beextended to additional groups of patients,such as subjects with comorbidities (Ayappaet al., 2004).

As for type 4 devices, recording of airflow bya nasal cannula, alone or in combinationwith pulse oximetry, is a possible screeningtool in patients with high clinical probabilityof OSA (Ayappa et al., 2004; Grover et al.,2008). The nasal pressure signal andanalysis of the flow limitation pattern allowreliable confirmation or exclusion of theoccurrence of OSA in patients with highclinical probability of obstructive events(Ayappa et al., 2004). More recently, nasal

Table 1. Sleep monitoring devices (Collop et al., 2007.).

Type Characteristics Pros Cons

1 Full in-laboratoryovernight PSG#, attended

High diagnostic yield High cost, long waiting lists

2 Full ambulatory overnightPSG#, unattended

Feasible, can be used forin-hospital studies

Possible loss of signals, up to20% of recordings inadequate,more studies needed

3 4–8-channel" polygraphy Respiratory polygraphyused in many centres

Lack of standardisation andvariability of signals recorded

4 1–3-channel recording(pulse oximetry plus 1–2other channels)

Used as inexpensivescreening tests for severeOSA or in obese patients

Not recommended by AASM

#: EEG, EMG and EOG channels, plus cardiorespiratory and body position monitoring channels, etc.; ":

usually cardiorespiratory monitoring and position channels as in type 1–2 devices.


pressure recordings for three nights at homewere found to be accurate diagnostic toolsin patients with a high clinical probability forOSA (Rofail et al., 2010a). Such devices arepromising, since they are simple,inexpensive and well tolerated by patients.However, further validation is needed inorder to optimise the choice of patients inwhom they provide clinically useful results.

Validation studies of portablecardiorespiratory monitoring devices carrysome relevant problems. As recentlyunderlined by Kuna (2010), the diagnosticyield of a type 3 portable device can varyaccording to the setting where it is used, i.e.in a sleep laboratory or home setting. Sleepapnoea assessment using a portable devicein the laboratory was closer to results of PSGthan those obtained in a patient’s home(Kuna, 2010). This finding may, in part, besecondary to the resolution of problems byan experienced technician in the sleeplaboratory setting compared with the homesetting. Another relevant factor, however, isrepresented by the night-to-night variabilityin SDB, which further complicates theassessment of reproducibility of resultswhen different devices/settings are compared.

Polygraphy scoring: settings, trends andpitfalls

Figures 1 and 2 show typical examples ofeight-channel respiratory polygraphicrecordings. Home monitoring with type 3monitoring devices for the diagnosis of OSAshould be performed according to currentAASM guidelines (Epstein et al., 2009;Collop et al., 2007), which state the following:

N Use of portable monitoring for thediagnosis of OSA should be limited topatients with a high clinical pre-testprobability of moderate-to-severe OSAand free from significant comorbidities.

N Patients with suspected sleep disordersother than OSA should undergo full PSG,since validation studies of portablemonitoring in these disorders are lacking.

N Use of portable monitoring for diagnosisof OSA may be appropriate in patientswho cannot undergo full PSG because ofimmobility, safety or critical illness.

N Portable monitors can be used to assessthe effects of OSA treatment.

N The minimum of signals to be recordedby portable monitoring devices includeairflow, respiratory effort and bloodoxygenation. Sensors should be the sameas those used in in-laboratory PSG.

N The equipment should be applied byexperienced personnel, and raw datashould be displayed and manually editedby the reader; a sleep-certified physicianshould revise the results.

N Negative or technically inadequate resultsin patients with a high clinical probabilityof OSA should prompt in-laboratory PSG.

Commercially available type 3cardiorespiratory monitoring devices caninclude different signals. Pulse oximetry andrespiratory signals (respiratory effortmonitored by inductive plethysmography,airflow and snoring) should be present, asrecommended by the AASM guidelines(Collop et al., 2007; Redline et al., 2007).Other signals that can be included inportable monitoring devices are heart rate orcardiovascular changes associated withrespiratory events (Penzel et al., 2010). Anexample of the latter technology is pulsearterial tonometry (PAT), which has beenused to diagnose sleep apnoea incombination with pulse rate, oximetry andactigraphy, with promising results (Pittmanet al., 2004). The PAT signal undergoesattenuation when sympathetic-mediatedvasoconstriction occurs at the end ofapnoeas and such signals are used torecognise respiratory events. The PAT signalhas recently been analysed to deriveinformation on sleep stages and found toshow a moderate agreement with standardPSG; interestingly, OSA severity did notaffect the results (Hedner et al., 2011).

Current research is also investigating thepossibility of diagnosing OSA by derivingrespiratory events from electrocardiographicrecordings obtained during sleep(Babaeizadeh et al., 2011); such techniques,however, involve complex algorithms andneed validation in large clinical populations.

A major problem in assessing the validity ofcommercially available devices for the


diagnosis of OSA is that they vary in the typeand number of sensors. Moreover, lack ofstandardisation makes it difficult tocompare results obtained with differentdevices or analyse results of differentstudies by meta-analyses.

Portable monitoring devices have been usedin ambulatory OSA management protocols,including CPAP titration. In these studies,traditional in-laboratory management, basedon full diagnostic PSG and manual CPAPtitration during PSG, has been compared toan ambulatory protocol, using portablemonitoring devices and automatic positiveairway pressure (APAP) ventilators at hometo establish the pressure level needed by theindividual patient. Some recent randomisedcontrolled trials tested whether clinicaloutcomes were comparable betweenpatients studied by type 3 portablemonitoring or in-laboratory PSG-basedmanagement. Since CPAP titration andfollow-up after CPAP initiation are the final

objective of an effective managementstrategy, such studies assessed whether theresults of both procedures were comparable.These studies, summarised in table 2,indicate that ambulatory management is notinferior to the traditional diagnostic pathwayused for OSA diagnosis and treatment(Mulgrew et al., 2007; Berry et al., 2008;Antic et al., 2009; Kuna et al., 2011b; Masaet al., 2011a). Interestingly, both inclusioncriteria and diagnostic methods variedamong the studies (table 2). Masa et al.(2011a) enrolled patients with moderateclinical OSA severity and found that portablemonitoring showed better diagnostic cost-effectiveness than PSG. In addition, theyalso tested the feasibility and results of atelemedicine application, concluding that itbore similar overall costs to the ambulatorymanagement, since the costs of informaticswere compensated for by the costssustained by the patients to travel to andfrom the laboratory (fig. 3) (Masa et al.,2011b). Future studies can be expected to

08:0002:00 03:00 04:00 05:00 06:00 07:0001:00

Obstructive s

Activity

Snore

SpO2 %

Pulsebeats·min-1

Desat10080

1401201008060

ButtonSupine

Moving

LeftProneRight

Upright

high

low

low90

3060

Central s90

3060

Mixed s90

3060

highSnore

Figure 1. Respiratory polygraphy. Summary graphics regarding total recording time in a patient with OSA.From top to bottom: body position, activity (moving and continuous signal), oxygen saturation (eventsand continuous signal), pulse (lower and higher values of heart rate), snoring (events and continuoussignal), and obstructive, central and mixed events with their duration in seconds (s). Desat.: desaturation.


further focus on the diagnostic yield andcosts associated with widespread use ofinformation technology applied toambulatory management of OSA patients.

The recent Canadian Thoracic Society 2011guideline update reviewed evidence-baseddata regarding outcomes of portablemonitoring and PSG-based management ofOSA patients. Although the AHI valuesobtained by portable monitoring devicesand PSG were different, no difference wasfound between the two managementstrategies for: residual sleep apnoea,sleepiness and quality of life; CPAPadherence; CPAP pressure; patientsatisfaction; or neurocognitive function(Fleetham et al., 2011).

Despite these promising results, Pietzschet al. (2011) reported that ambulatorymanagement was not advantageouscompared with in-laboratory PSG. Thisfinding was due to the need to repeatrecordings whenever ambulatory tracingswere technically unsatisfactory or resultednegative in patients with high clinicalprobability of OSA, requiring furtherassessment. False negative tests in patientswith high pre-test probability for OSA have

been estimated to occur in 17% of cases(Collop et al., 2007). The results of Pietzschet al. (2011) have been critically reviewed byAyas et al. (2011), with special emphasis onthe multiple assumptions to be consideredin modelling studies, especially sinceevidence-based data derived fromrandomised controlled studies are notavailable. This is an area of major interestthat is in a state of continuous evolution,and more studies can be expected toaddress and refine cost analysis ofambulatory systems for diagnosis of SDB inthe near future.

Interpretation of cardiorespiratorypolygraphy

There are some problems in theinterpretation of cardiorespiratorypolygraphy (Kuna, 2010; Collop et al., 2007;Ahmed et al., 2007; Kuna et al., 2011b),which need to be known by the physicianusing portable devices instead of in-laboratory full PSG. These problems concernscoring of apnoeas and hypopneas and theassessment of OSA severity.

N The exact duration of sleep and the AHI,i.e. the number of events divided by sleep

Obstructive (35.00 s)

Desaturation (50.60 s) [11]

Position (0–379 cm)

Snore pressure (0–262 mbar/cm)

Pulse (50–125 bpm)

Abdomen (1.12 mV/cm)

Thorax

Nasal flow

7:12:40 7:12:50 7:13:00 7:13:10 7:13:20 7:13:30 7:13:40 7:13:50 7:14:00 7:14:10 7:14:20 7:14:30 7:14:40 7:14:50 7:15:00 7:15:10 7:15:20

Nasal PDG (5.46 mbar/cm)

SpO2 (70–100%)Desaturation (50.60 s) [11]

Snore

Obstructive (45.50 s)

Figure 2. 3-min respiratory polygraphic recording in a patient with OSA. Note the obstructive apnoeas,consequent desaturations and their durations in seconds (s). Signals include, from top to bottom: nasalpressure (cannula), nasal flow (cannula), thoracic movement (inductive band), abdominal movement(inductive band), oxygen saturation (pulse oximeter), heart rate (pulse oximeter), snoring (nasal cannula)and body position (sensor).


time, cannot be assessed withoutrecording of the signals used for sleepscoring (EEG, EOG and EMG). Whenportable monitoring devices are used todiagnose OSA, the frequency ofrespiratory events is reported as therespiratory disturbance index (RDI), i.e.the number of events divided byrecording time. This leads to a systematicunderestimation of OSA severity basedon the frequency of respiratory eventscompared with the results obtained byfull PSG (fig. 4).

N The AASM Practice Parameters guidelineindicates two possible ways to scorehypopneas (AASM, 2005). The firstrequires a 30% decrease in airflow for

o10 s associated with a 4% oxygendesaturation. The alternative definition ofhypopnea is a 50% decrease in chest wallmovement or airflow associated with a3% oxygen desaturation and/or detectionof an EEG arousal (Redline et al., 2007).While hypopneas of the first type can bedetected by cardiorespiratory monitoring,those of the second type cannot, due tothe lack of the sleep signals (EEG, EOGand EMG), leading to furtherunderestimation of the number of events.The AASM guidelines do not mention useof a nasal cannula and carefulassessment of the flow limitation pattern,which may provide an accurate estimateof obstructive events (Kuna et al., 2011).

Table 2. Ambulatory management of OSA: effects on functional outcomes/therapeutic decision-making.

Ref. Study Patients Outcome Arms Results

Mulgrew et al.,

2007

RCT, single

centre

68 patients

with high

clinical

probability of

moderate-to-

severe OSA

AHI on CPAP,

ESS, SAQLI

score, CPAP

adherence at

3 months

Parallel design:

home oximetry/

APAP (n533);

hospital PSG/CPAP

(n535)

No difference at

3 months in AHI on

CPAP, ESS, SAQLI

score; CPAP adherence

better in the ambulatory

managed group

Berry et al.,

2008

RCT, Veterans

Affairs Medical

Center

88 patients

with EDS and

high clinical

probability of

OSA

CPAP

adherence,

ESS, FOSQ at

6 weeks

Parallel design:

home 4-channel

monitoring#/APAP

(n545); hospital

PSG/CPAP (n543)

Similar CPAP adherence

at 6 weeks, mean

nightly use; decrease in

ESS, FOSQ; CPAP

satisfaction in both

groups

Antic et al.,

2009

RCT, 3

academic sleep

centres,

Australia

195 patients

with suspected

moderate-to-

severe OSA

ESS, CPAP

adherence at

3 months

Parallel design:

home oximetry/

APAP (n5100);

hospital PSG/CPAP

(n595)

Similar CPAP adherence

at 3 months, mean

nightly use; decrease in

ESS; lower cost (the

home programme was

led by nurses)

Kuna et al.,

2011b

RCT, 2

Veterans

Affairs Medical

Centers

223

consecutive

patients with

suspected OSA

FOSQ and

CPAP

adherence at

3 months

Parallel design:

home RP/APAP

(n5113); hospital

PSG/CPAP

(n5110)

Home management not

clinically inferior to the

laboratory management

of OSA

Masa et al.,

2011a

Multicentre

RCT, Spanish

Sleep Network

348 patients

with

intermediate-

to-high

probability of

OSA

Therapeutic

decision on

CPAP

prescription

Cross-over design:

home RP/hospital

PSG

RP performed well in

severe OSA (AHI

o30 events?h-1), less

well in moderate OSA

(AHI 15–30 events?h-1)

RCT: randomised controlled trial; RP: respiratory polygraphy. #: Watch-PATTM 100 (Itamar Medical Inc.,

Norwood, MA, USA).


N A third problem regards the type of signalto detect airflow to be used incardiorespiratory monitoring. For optimaldetection of changes in airflow, the AASMguideline (2005) recommends use of twosensors, i.e. the nasal pressure signal andthe thermistor oronasal flow signal. Nasalpressure is more sensitive for detectionof the typical pattern of progressiveinspiratory flow limitation, while thesignal derived by thermal sensor candetect respiratory events occurring duringoral breathing. Therefore, somedifferences in diagnostic yield betweendifferent devices may depend on the typeof airflow signal used, and the physicianshould be aware of this possibility.

Oximetry

Pulse oximetry is commonly recorded duringthe night in patients with suspected OSA inthe context of full PSG or cardiorespiratorypolygraphy. The AASM recommends using

devices that average the signal over 3–4 s(AASM, 2005; Epstein et al., 2009; Collopet al., 2007).

Pulse oximetry has limited diagnostic yieldwhen used as the only signal recordedduring sleep, for the following reasons:

N Occurrence of oxygen desaturations doesnot provide any information on the typeof the associated event (central orobstructive apnoeas or hypopneas).

N In patients with high oxygen saturationduring wakefulness, respiratory eventsmight be insufficient to cause significantdrops in oxygen saturation, due to theirposition on the flat portion of the oxygendissociation curve (i.e. nonobese sleepapnoea patients).

N The reliability of the pulse oximeter signaldepends on adequate perfusion of thesite where the arterial oxygen saturation(SpO2) measurement is obtained. Lowperfusion states (as in critically illpatients) are often associated with poorreadings and abnormally low SpO2 values.

N The results of oximetry vary when comparingdifferent devices (Zafar et al., 2005).

Therefore, pulse oximetry alone is notrecommended by any guideline on thediagnosis of SDB. Recent reports, however,have provided new algorithms for analysis ofSpO2, alone or coupled with other signals,which could considerably improve thediagnostic yield of oximetry. In addition,some studies suggest that the number of3% oxygen desaturation events (ODI3%)obtained by pulse oximetry recordings forthree nights may be used to diagnose OSAwith acceptable accuracy, with a diagnosticyield similar to nasal pressure recordings(Rofail et al., 2010b).

Pulse oximetry is commonly used in patientson home NIV to check effectiveness oftreatment. Unfortunately, it is not specific,since fast or slow oxygen desaturations mayoccur for several reasons (Janssens et al.,2011). Therefore, it should only beconsidered as a screening tool to identifypatients who need more sophisticatedreassessment (Janssens et al., 2011).

800

600

700

500

500

300200

100

0

Cos

t in

200

9 €

PSG NontelemetricHRP

TelemetricHRP

Figure 3. Cost per patient in a cohort of 348subjects (Masa et al., 2011). PSG was performedin hospital and respiratory polygraphy at homewith two cost approaches: patients obtained therespiratory polygraphy device at the hospital, wenthome and returned the following morning(nontelemetric); or the devices were moved by atransport company from home to home withtelemetric transmission of data files from home tohospital (telemetric). The costs of homerespiratory polygraphy (HRP) were adjusted forthe same diagnostic efficacy as PSG. Note thatnontelemetric and telemetric HRP had similarcosts, which were approximately half of the costsof PSG. Data are presented as mean¡SD.


Actigraphy

Actigraphy records body movements usingaccelerometers and is largely used inepidemiological studies to measure sleepduration and analyse circadian sleepdisorders. Actigraphy has been proposed asan additional signal to be recorded duringhome sleep monitoring in order to solve theproblem of the lack of an objectivelymeasured sleep period. This could allow amore precise evaluation of RDI, since thenumber of events recorded could be dividedby the actual sleep duration instead of by theentire recording time. The 2007 AASMpractice guideline did not find evidence forsufficient accuracy of actigraphy devices forsuch purposes (Collop et al., 2007).Actigraphy together with PAT and oximetryrecordings provided acceptable results forambulatory diagnosis of OSA using acommercially available device (Ayas et al.,2003). Some recent studies suggest that

actigraphy can be used instead of PSG toshow improved sleep efficiency and reducedfragmentation in OSA patients treated withCPAP (Otake et al., 2011). New models toanalyse actigraphy data may provide clinicallyuseful algorithms (Wang et al., 2011).

A recently developed device with somesimilarity to actigraphy is a noncontactbiomotion sensor allowing detection of bodyand respiratory movements during sleep;initial results appear promising (De Chazalet al., 2011).

Nocturnal capnography

Transcutaneous carbon dioxide tension(PtcCO2) has been used for the last 30 yearsto noninvasively and continuously monitorventilation in children and adults in differentclinical contexts, including sleep. Recenttechnological advancements have made thesensors smaller and more reliable comparedwith the first commercially available models(Eberhard, 2007). Carbon dioxide diffusesthrough the skin and this process isenhanced by warming the skin to 42–45uC;most devices measure PtcCO2 by measuringchanges in pH (or optical absorption in thenear-infrared light, as recently proposed).Technology advancements have providedsensors with faster responses and less needfor calibration or membrane change, makingit possible to record PtcCO2 for longerperiods of time with minimum discomfort tothe patient and no need to change theapplication site after 4 h, as in the oldermodels (Kelly et al., 2011). PtcCO2

measurements have been used in neonataland paediatric settings since theirintroduction, but have recently gainedapplication in adults, especially in operatingrooms and intensive care settings, in sleeplaboratories and for home monitoring ofNIV. Both PtcCO2 and SpO2 can be recordedby the most recently developed devices.

PtcCO2 values show excellent correlation withPaCO2 values obtained by arterial blood gas(ABG) sampling with an acceptable delay (afew minutes) of the transcutaneouscompared with the direct measurement(Janssens et al., 2011), although largedeviations (.10 mmHg) have been

39

37

38

36

35

34

33

32

Even

ts·h

-1

PSG AHI RP RDI RP AHI

Figure 4. In PSG, the number of apnoeic episodesis divided by the total sleep time (TST) to obtainthe AHI. In respiratory polygraphy (RP), thenumber of apnoeic episodes is divided by therecording time, which is normally greater than theTST, to obtain the RDI. Accordingly, in a cohort of348 patients where PSG and RP were performedsimultaneously in hospital, the PSG AHI was4 events?h-1 higher than the RP RDI. However, ifthe number of RP apnoeic events is divided by theTST obtained from PSG, the resultant index (RPAHI) is close to the PSG AHI. This calculationemphasises that the habitual underestimation of RPRDI is mainly attributable to the use of recording timeinstead of sleep time, as in the PSG.


reported in patients on NIV with initial PaCO2

.60 mmHg (Kelly et al., 2011). Somestudies have shown that a drift occurs in thePtcCO2 signal over time, and suggestedobtaining ABG samples at the beginningand end of the recording in order to correctthe results; future devices may automaticallyadjust for such drift. However, mostclinicians agree that the major advantage ofPtcCO2 monitoring is the possibility ofassessing the trend of carbon dioxide valuesover time, besides estimating absolutePaCO2 values.

A clinically important application of PtcCO2

monitoring is the detection ofhypoventilation in patients with OHS oramyotrophic lateral sclerosis (ALS),especially during sleep. PtcCO2 monitoringeliminates the need to obtain ABG samplesbefore and after sleep and avoids falsenegative results due to rapid normalisationof ventilation after awakening.

PtcCO2 can be used to monitor the effects ofNIV in acute settings, thus limiting thenumber of ABG samples to be drawn toassess the effectiveness of NIV (Storre et al.,2007). In the home setting, it can yield veryuseful information to check the effectivenessof NIV and help resolve problems (Janssenset al., 2007). Although SpO2 and PtcCO2

monitoring do not provide information onthe cause of the disturbances during NIV(i.e. leaks, insufficient pressure, upper airwayobstruction, etc.), their use has been suggestedin order to identify patients who need furtherassessment to optimise ventilation duringsleep (Janssens et al., 2007).

However, hypocapnia is a major risk factorfor the development of central sleepapnoeas and Cheyne–Stokes breathing inpatients with heart failure. In these patients,CPAP was shown to decrease AHI andincrease PtcCO2, suggesting that continuousrecording of PtcCO2 during sleep may beuseful in these patients (Naughton et al.,1994). With regard to patients with OSA,some studies have shown that increasingPtcCO2 during sleep occurs in patients withsevere obesity and oxygen desaturations,and predicts hypercapnia at awakening inOSA patients (Chin et al., 1997). It should be

pointed out that AASM guidelines do provideindications for use of PtcCO2 monitoring fordiagnosis and treatment of SDB.

In summary, monitoring of PtcCO2 should beconsidered a useful adjunct to diagnosticand therapeutic management of acute andchronic hypoventilation conditions. Newdevices for use in adults are easy to use andreliable. Conversely, measurement of end-tidal carbon dioxide is less practical and notcurrently recommended for use in the sleeplaboratory or at home (Janssens et al., 2011).

Further reading

N Ahmed M, et al. (2007). Portable moni-tors in the diagnosis of obstructive sleepapnea. Chest; 132: 1672–1677.


N Antic NA, et al. (2009). A randomizedcontrolled trial of nurse-led care forsymptomatic moderate-severe obstruc-tive sleep apnea. Am J Respir Crit CareMed; 179: 501–508.

N Ayappa I, et al. (2004). Comparison oflimited monitoring using a nasal-cannulaflow signal to full polysomnography insleep-disordered breathing. Sleep; 27:1171–1179.

N Ayas NT, et al. (2003). Assessment of awrist-worn device in the detection ofobstructive sleep apnea. Sleep Med; 4:435–442.

N Ayas NT, et al. (2011). The demise ofportable monitoring to diagnose OSA?Not so fast!. Sleep; 34: 691–692.

N Babaeizadeh S, et al. (2011).Electrocardiogram-derived respiration inscreening of sleep-disordered breathing.J Electrocardiol; 44: 700–706.

N Berry RB, et al. (2008). Portable monitor-ing and autotitration versus polysomno-graphy for the diagnosis and treatment ofsleep apnea. Sleep; 31: 1423–1431.

N Canadian Sleep Society, et al. (2010).Canadian Sleep Society/Canadian ThoracicSociety position paper on the use ofportable monitoring for the diagnosis ofobstructive sleep apnea/hypopnea in adults.Can Respir J; 17: 229–232.


N Chin K, et al. (1997). Changes in thearterial PCO2 during a single night’s sleepin patients with obstructive sleep apnea.Intern Med; 36: 454–460.

N Collop NA, et al. (2007). Clinical guide-lines for the use of unattended portablemonitors in the diagnosis of obstructivesleep apnea in adult patients. J Clin SleepMed; 3: 737–747.

N De Chazal P, et al. (2011). Sleep/wakemeasurement using a non-contact bio-motion sensor. J Sleep Res; 20: 356–366.

N Eberhard P. (2007). The design, use, andresults of transcutaneous carbon dioxideanalysis: current and future directions.Anest Alag; 105: Suppl., S48–S52.

N Epstein LJ, et al. (2009). Clinical guidelinefor the evaluation, management andlong-term care of obstructive sleep apneain adults. J Clin Sleep Med; 5: 263–276.

N Fietze I, et al. (2011). Management ofobstructive sleep apnea in Europe. SleepMed; 12: 190–197.

N Fleetham J, et al. (2011). CanadianThoracic Society 2011 guideline update:diagnosis and treatment of sleep dis-ordered breathing. Can Respir J; 18: 25–47.

N Grover SS, et al. (2008). Automateddetection of sleep disordered breathingusing a nasal pressure monitoring device.Sleep Breath; 12: 339–345.

N Hedner J, et al. (2011). Sleep stagingbased on autonomic signals: a multi-center validation study. J Clin Sleep Med;7: 301–306.

N Janssens J-P, et al. (2011). Nocturnalmonitoring of home non-invasive ventila-tion: the contribution of simple toolssuch as pulse oximetry, capnography,built-in ventilation software and auto-nomic markers of sleep fragmentation.Thorax; 66: 438–445.

N Janssens JP, et al. (2001). Is continuoustranscutaneous monitoring of PCO2 (TcPCO2) over 8 h reliable in adults? RespirMed; 95: 331–335.

N Kelly A-M, et al. (2011). Agreementbetween arterial and transcutaneousPCO2 in patients undergoing non-invasiveventilation. Respir Med; 105: 226–229.

N Khawaja IS, et al. (2010). Diagnosticaccuracy of split-night polysomnograms.J Clin Sleep Med; 6: 357–362.

N Kuna ST, et al. (2011a). An official ATS/AASM/ACCP/ERS workshop report:Research priorities in ambulatory man-agement of adults with obstructive sleepapnea. Proc Am Thorac Soc; 8: 1–16.

N Kuna ST, et al. (2011b). Noninferiority offunctional outcome in ambulatory man-agement of obstructive sleep apnea. Am JRespir Crit Care Med; 183: 1238–1244.

N Kuna ST. (2010). Portable-monitor test-ing: an alternative strategy for managingpatients with obstructive sleep apnea.Respir Care; 55: 1196–1215.

N Masa JF, et al. (2011a). Therapeuticdecision-making for sleep apnea andhypopnea syndrome using home respira-tory polygraphy: a large multicentricstudy. Am J Respir Crit Care Med; 184:964–971.

N Masa JF, et al. (2011b). Effectiveness ofhome respiratory polygraphy for thediagnosis of sleep apnoea and hypop-noea syndrome. Thorax; 66: 567–573.

N Mulgrew AT, et al. (2007). Diagnosis adinitial management of obstructive sleepapnea without polysomnography. A ran-domized validation study. Ann InternMed; 146: 157–166.

N Naughton MT, et al. (1994). Effects ofcontinuous positive airway pressure oncentral sleep apnea and nocturnal PCO2 inheart failure. Am J Respir Crit Care Med;150: 1598–1604.

N Otake M, et al. (2011). Monitoring sleep–wake rhythm with actigraphy in patientson continuous positive airway pressuretherapy. Respiration; 82: 136–141.

N Penzel T, et al. (2010). Ambulatorydiagnosis of OSA and new technologies.Eur Respir Monogr; 50: 136–149.

N Pietzsch JB, et al. (2011). An integratedhealth-economic analysis of diagnosticand therapeutic strategies in the treat-ment of moderate-to-severe obstructivesleep apnea. Sleep; 34: 695–709.

N Pittman SD, et al. (2004). Using a wrist-worn device based on peripheral arterialtonometry to diagnose obstructive sleepapnea: in-laboratory and ambulatory vali-dation. Sleep; 27: 923–933.

N Redline S, et al. (2007). Reliability andvalidity of respiratory event measurementand scoring. J Clin Sleep Med; 3: 169–200.


N Rofail LM, et al. (2010a). The utility of

single-channel nasal airflow pressure

transducer in the diagnosis of OSA at

home. Sleep; 33: 1097–1105.

N Rofail LM, et al. (2010b). Comparison

between a single-channel nasal airflow

device and oximetry for the diagnosis

of obstructive sleep apnea. Sleep; 33:

1106–1114.

N Storre JH, et al. (2007). TranscutaneousPCO2 monitoring during initiation of non-invasive ventilation. Chest; 132: 1810–1816.

N Wang J, et al. (2011). Measuring the impactof apnea and obesity on circadian activitypatterns using functional linear modeling ofactigraphy data. J Circadian Rhythms; 9: 11.

N Zafar S, et al. (2005). Choice of oximeteraffects apnea–hypopnea index. Chest; 127:80–88.


Treatment of obstructivesleep apnoea

Johan Verbraecken, Arie Oksenberg, Marie Marklund and An Boudewyns

Therapeutic strategies for patients withOSAS may be categorised into three generalgroups: behavioural, medical and surgical.Treatment decisions should be based on theeffect of OSAS on daytime symptoms,cardiovascular and metabolic functionrather than on the absolute number of

episodes of apnoea or hypopnoea. The finalgoal is to establish a stable oxygen curveand ventilatory pattern, abolish snoring,eliminate sleep fragmentation due to upperairway (UA) collapse and enhance alertnessduring the daytime.

Patient education

All patients with OSAS should be counselledabout the potential benefits of therapy andthe hazards of going without treatment. Therelative value of avoiding factors thatincrease the severity of UA obstruction, suchas the use of alcohol, sedatives andhypnotics, and weight gain, should bediscussed. Pharyngeal fat deposits lead to adecrease in pharyngeal patency andunderline the risk factor of obesity. Weightloss is recommended in .80% of patientswith OSAS and leads to a decrease in criticalclosing pressure (Pcrit), and consequentlydecreases the severity of OSA.Unfortunately, only 5% of overweightpersons are able to lose weight andmaintain weight loss. This problem isexacerbated when patients are sleepy duringthe daytime, tend to snack to stay awake andare too tired to exercise. Most patients wholose weight experience an improvement inapnoea, but it is difficult to predict theamount of improvement associated withloss of a specified amount of weight. Basedon a large epidemiological study, it wasfound that a 10% weight loss was associatedwith a 26% decrease in AHI. Some patientsmay be substantially better after a weightloss of only a few kilograms, but others maycontinue to have symptoms despite asignificant weight loss (50–70 kg). Dietintervention studies report a body weightreduction of 3.4–27.2 kg (mean¡SD -15¡7 kg),

Key points

N All patients with OSAS should becounselled about the potentialbenefits of therapy and the hazards ofgoing without treatment.

N Patients with supine-position OSAtend to have a lower AHI, to beyounger and to be less obese.

N In mild-to-moderate OSA, oralappliances have a similar efficacy asCPAP therapy in terms of changesin AHI.

N Successful UA surgery is mostcommonly defined as a .50%reduction in AHI and a post-operativeAHI ,20 events?h-1.

N Bariatric surgery is indicated inindividuals with a BMI o40 kg?m-2 orwith a BMI o35 kg?m-2 withimportant comorbidities (arterialhypertension, diabetes and OSA).

N No specific drug is able toconsistently reduce the severity ofOSA by more than 50%.

N OSA patients should be screened forcardiovascular and metaboliccomplications.


a change in BMI of 1.6–9.4 kg?m-2

(mean¡SD-4.7¡2.5 kg?m-2) and a decreasein AHI from 42¡25 to 24¡15 events?h-1

(mean change -23¡13 events?h-1 or -45%).A partial improvement in AHI can beexpected in 39% of patients, while a curefrom OSA is observed in 23% of cases.Some patients being treated with CPAPfind it easier to lose weight than beforetreatment, probably because of increasedactivity during the day or decreasedingestion of snacks to increase alertness.Others report a weight gain, which couldbe related to altered energy expenditure atnight. However, weight loss takes time,and only a minority of patientssuccessfully maintain it. As a primarytreatment, weight loss should be targetedtowards patients with mild-to-moderateOSAS, especially if there is no interest inother options. Alcohol selectively reducesUA muscle tone, and increases thefrequency of respiratory events duringsleep. Alcohol also prolongs apnoea orhypopnoea by delaying arousal and is amajor source of calories.

In spite of the widespread recommendationof CPAP (and other therapeutic modalities)in the management of OSA, there areconcerns about its continued acceptanceamong people who have to use it in the longterm. Side-effects, and psychological andother social factors may impede sustainedand successful treatment. Improvedadherence can be obtained with increasedpatient education, which guides treatmentuses and sensitises patients to the benefitsof treatment and the risks of nontreatment.Three different educational and behaviouralstrategies for improving CPAP use in adultswith OSA have been described. An intensive,supportive intervention with additionalnursing input (including CPAP education athome and involvement of partners, andCPAP titration in the sleep centre followedby additional home visits) improves averageCPAP usage over a limited period of time insymptomatic OSA (+0.59 h per night), whilegroup education sessions can improvecompliance with CPAP therapy. Short-courseeducation at titration (e.g. a 15-minvideotape session) does not alter objective

CPAP compliance. Cognitive behaviouraltherapy is highly successful in terms ofobjective compliance and leads to thelargest increase in average machine usage(+2.92 h per night), and also increases thenumber of subjects using CPAP for .6 hper night. Overall, patient education willincrease the patient’s self-efficacy, treatmentcommitment and motivation for CPAPnightly use.

Positional treatment

Conservative treatment strategies alsoinclude the use of a lateral sleeping position.Studies have shown that the frequency ofapnoea and hypopnoea is greater with asupine sleeping position. Clinical experienceand observational studies suggest that thepatients exhibiting a large decrease in AHI inthe lateral position compared with thesupine position tend to have a lower AHI,and be younger and less obese. This therapyis most adequate for patients with OSASwho have most of their breathingabnormalities concentrated in the supineposture and, while sleeping in the lateralpostures, the number of these abnormalitiesis significantly reduced and, in some cases,is totally eliminated. Originally, this therapywas suggested for OSAS patients with anAHI in the supine posture that is double orgreater than that in the nonsupine postures.Recently, this has been refined to a moreclinically oriented definition in which theAHI in the nonsupine posture (obstructiveor other types of abnormalities) should be ata nonpathological level (,5 events?h-1 or, atmost, ,10 events?h-1). This therapy couldalso be advisable for patients who did notsucceed with CPAP, an oral device or anysurgical treatment, and mainly havebreathing abnormalities in the supineposture. Also, it is suitable for patientswithout OSAS but with snoring that isconfined mainly to the supine posture.Positional therapy is not recommended forpatients who for any reason (shoulderproblems or any other physical disabilitythat interferes with their sleep in the lateralposition) cannot avoid the supine postureduring sleep. It is also inadequate forpositional OSA patients who prefer to sleep


in the supine posture, but is mainlyinadequate for nonpositional OSA patientswho have many breathing abnormalities inthe supine and lateral postures as well. Forthese patients, CPAP is the treatment ofchoice, since they also suffer from a moresevere disease. This therapy is also not anoptimal solution for positional OSA patientswho continue to snore loudly and perhapsalso have events of flow limitation whilesleeping in the lateral postures. Differentmethods have been designed for positionaltherapy: verbal instructions, the classicaltennis ball technique, the positional alarm, atriangular pillow with space to place an armand a soft vest attached to a board placedunder the pillow. A small individual cohortstudy investigated the effects of elevatingthe upper body by 60u and found amoderate reduction of AHI. There are nodata comparing the different devices withthe exception that verbal instructions seemto be less effective than the positional alarm.As described above, patients exhibiting alarge decrease in AHI in the lateraldecubitus position compared with thesupine position tend to have a lower AHI,and be younger and less obese. Accordingly,patients with a clear improvement in AHIwith positional therapy tended to beyounger, have a lower AHI and be lessobese. It is not possible to extract from thedata whether AHI, age or obesity is the bestpredictor of treatment success. It is morelikely that AHI, age and obesity are mutuallyinterrelated. Cardiovascular sequelae, suchas endothelial or left ventricular function,were not investigated following positionaltherapy, while one study showed a decreasein blood pressure. In some studies,positional therapy is inferior, while in otherstudies CPAP is comparable to positionaltherapy. Overall, positional therapy can yieldimportant reductions in AHI, but only inclearly selected positional OSA patients. Inthe very few follow-up studies, long-termcompliance with positional therapy waspoor. Thus, it can be concluded thatpositional therapy can be used in individualcases, but sleep studies are recommendedto document individual success. Long-termcompliance has to be secured by regularfollow-up studies. Follow-up of OSA patients

using positional therapy is imperative sinceincrease in weight may convert positionalOSA patients into nonpositional OSApatients, and for them, CPAP is consideredthe treatment of choice.

Oral devices

Oral appliances for the treatment of snoringand sleep apnoea include mandibularadvancement devices (MADs) and the lesscommon tongue-retaining devices. MADsrepresent the main non-CPAP therapy forpatients with OSAS. They are recommendedfor patients with mild-to-moderate OSA andfor those with more severe disease who donot tolerate CPAP. These devices aim toincrease the UA size and reduce the risk ofsleep apnoeas and snoring in patients withOSAS. The UA is widened, particularly in itslateral dimension. The pharyngeal fat padsrelocate laterally from the airway and thetongue base muscles move anteriorly. Thisleads to a reduction in pharyngealcollapsibility.

A wide variety of appliances is available,differing both in construction and in themanner in which they alter the oral cavity.With adjustable devices, it is easy for thepatient or the dentist to change themandibular positioning in order to achievethe desired effects. The monoblock deviceshave to be adjusted at a dental laboratory.The variety of device designs may explainsome of the variability in outcome betweenpatients and studies. The appliances areworn only during sleep and are generally welltolerated. Not all patients have a clinicallymeaningful response to oral appliances.

The treatment effect of MADs has to beverified in a renewed sleep study with thedevice in all patients with OSA, sincepatients may have a suboptimal treatmentresponse. One major limitation of MADtreatment is its dependence on oral healthand the fact that it takes some time tobecome accustomed to the device. Side-effects from the treatment, such as pain inthe teeth and jaws, are generally mild andtransient. Patients therefore need anadaptation period that may sometimes lastfor months before they become accustomed


to the device. In the longer term, bitechanges become more common, but theseare usually minor and do not disturbpatients who are satisfied with the treatmentoutcome in terms of snoring and daytimesymptoms. Patients may continue usingtheir devices for many years, although thetreatment needs to be followed up in termsof side-effects and effectiveness.

Promising prediction methods based onthe mechanism of MADs and theirinfluence on UA structures have beenproposed. Various imaging techniques havebeen used to visualise the individualincrease in UA size during mandibularadvancement. The velopharyngeal area iscentral to this modification process, whichleads to reduced pharyngeal collapsibilityand sleep apnoea. Good responders have alarger increase in velopharyngeal airwaysize compared with poor responders. Withnasendoscopy, it has been shown that theairway is still open during Muller’smanoeuvre and mandibular advancementin good responders compared with a clearcollapse in poor responders. A model ofthe UA using combined UA imaging andcomputational fluid dynamics can be usedto evaluate the influence of MADs on UAvolume and resistance, and has potentialfor the prediction of treatment success inthe future.

The degree of mandibular advancement isan important modulator of the treatmentoutcome, since there is a dose-dependenteffect on nocturnal oxygenation andpharyngeal collapsibility. Mandibulartitration is, therefore, a key procedure whenit comes to obtaining optimal effects onOSAS with the device. A small advancementproduces a suboptimal treatment effect,while too large an advancement producesmore side-effects. Moreover, there are alsostudies showing a reduction of efficacy whenadvancing the mandible too much. It isbelieved that in some cases, the airwaydimension changes from a wide lateraldiameter to a narrow lateral diameter iftitration is overdone. It is possible that theexact degree of mandibular advancement isof less importance for patients with

mild-to-moderate disease compared withpatients with more severe OSAS. A mouthopening of 4–14 mm has not been found toinfluence the treatment outcome in terms ofsleep apnoea, although patients preferredthe device with a smaller opening. There iscurrently no agreement regarding the bestmethod to measure and define the degree ofmandibular repositioning in the individualpatient. A custom-made device titrated froman initial 50% of maximum mandibularadvancement can be recommended.

Self-reported short-term compliance withMAD treatment ranges from 76% to 95% ofthe patients. Some studies observe a highernightly compliance with MADs than withCPAP, while other studies describe a similarnightly compliance. Compliance monitorshave been introduced into the market andare being evaluated. Objective compliancemonitoring of MAD use is currently notwidely available and, generally, onlysubjective compliance reporting has beenrelied upon in previous studies. In thelonger term, patients usually continue to usetheir MADs, although compliance appearsto decrease slightly with time. After 1 yr,76% of the patients were still on treatmentand, after 2–4 yrs, about half the patientswere still using their devices. The treatmenteffect of MADs on AHI is usually stable ordecreases slightly in patients who have agood initial treatment outcome. About aquarter of the patients discontinue MAD orCPAP treatment, either because of side-effects or ineffective treatment. Closecollaboration between physicians anddentists is necessary to ensure optimalpatient selection, and to avoid any alterationof dental occlusion or temporomandibularjoint discomfort. The treatment must befollowed up and the device adjusted orexchanged in relation to the outcome.

Surgery of the UA

A significant proportion of patients is unableto tolerate or adhere to CPAP therapy.Therefore, many patients turn to surgicaloptions as a salvage treatment. Surgery canalso be considered as a first-line treatmentin mild OSA and in patients with moderate-to-severe disease for whom other


noninvasive treatments have failed. Surgicaltreatment of OSA aims to correctanatomical abnormalities in the UA,contributing to its collapse during sleep.Various surgical modifications of the UAhave been proposed to manage and, insome cases, treat OSA. Their efficacy,however, remains questionable.

Since the results of UA surgery in unselectedOSAS patients are poor, it was suggestedthat one should first try to identify the site ofUA obstruction and subsequently correct theanatomical abnormalities at this particularsite. Patients may be identified withretropalatal collapse, retrolingual collapse orboth, and the lateral pharyngeal walls maybe involved. At present, sleep endoscopy isthe most commonly used technique in dailypractice to evaluate the site(s) of UAobstruction during sleep. Several studiesdemonstrate that there are many sites in theUA, which may contribute to pharyngealcollapse during sleep. Therefore, theconcept of one obstruction site/one surgicaltreatment has been replaced by multilevelsurgery addressing several airway segments.

Various surgical techniques might becombined to correct abnormalities at thenose, soft palate and tongue base includingsoft tissue surgery and/or skeletalframework surgery. The addition ofhypopharyngeal surgery improves clinicaloutcome in many OSA patients. The resultsof isolated nasal surgery for OSAS are poorand it is impossible to predict who willrespond on clinical grounds. However, nasalsurgery might significantly improve CPAPand mandibular repositioning appliance(MRA) tolerance in those patients sufferingfrom chronic nasal obstruction. Theselection of the appropriate treatments for aparticular patient becomes more complexand a thorough knowledge of the differentsurgical options, their results and possiblecomplications is required. Some authorshave concluded that there is little role for UAsurgery in the management of OSASpatients because high-evidence-level studiesare not available, but the body of evidence isgrowing with several recent controlled level 1trials. Surgical treatment for snoring and

OSAS should never be performed without apre-operative PSG and a post-operativesleep study is mandatory in OSAS patients.OSAS patients have an increasedanaesthetic risk at the time of induction(difficult intubation) and upon awakening(respiratory compromise). A discussionbetween the surgeon and theanaesthesiologist about the pre- and post-operative airway management is, therefore,required. The results of surgical proceduresfor OSAS should be objectively documentedby PSG 3–6 months post-operativelybecause of the discrepancy betweensubjective and objective results. Appropriateclinical follow-up is required since resultstend to deteriorate in the long term.

Considerable clinical experience with thevarious UA surgical procedures has beenobtained over the past few years. Theinterpretation of surgical results reported inthe literature is not always straightforwardbecause of different criteria for success,limited numbers of patients and, often, ashort follow-up time. The lack of aconsensus on the definition of OSAS andsurgical success seriously hampers theinterpretation of surgical results andcomparisons among different treatmentoptions. The criteria employed by Sher et al.(1996) are a reduction of the respiratorydisturbance index (RDI) of o50% and apost-operative RDI ,20 events?h-1 or areduction of the apnoea index (AI) ofo50% and a post-operative AI,10 events?h-1. These criteria are frequentlyemployed in the current literature on OSAsurgery but their validity might bequestioned given the recent data that evenmild forms of OSA might be associatedwith cardiovascular morbidity. Other PSGvariables have been suggested, such as theoxygen desaturation index, the amount ofsleep fragmentation, cardiac events, andpatient-based factors, such as the degree ofdaytime sleepiness or obesity, which areprobably better related to the long-termhealth consequences of OSA.

Surgical procedures, listed according to thesite of intervention and their indications, arelisted in table 1.


Bariatric surgery

Obesity is one of the most important riskfactors for the development of OSA. ,70%of OSA patients are obese, i.e. exhibit a BMIo30 kg?m-2. In morbidly obese patients, i.e.a BMI o40 kg?m-2, the prevalence of OSAvaries between 40% and 90%. The severityof OSA is also greater than in less obesepatients. Weight reduction can, therefore, bean important strategy in the management ofOSA. Even mild weight reduction can lead to

an improvement in sleep and breathingpattern, and decrease hypersomnolence.Bariatric surgery is an effective means toachieve major weight loss and is indicated inindividuals with a BMI o40 kg?m-2 or thosewith a BMI o35 kg?m-2 with importantcomorbidities (arterial hypertension,diabetes and OSA) and in whom dietaryattempts at weight control have beenineffective. Bariatric surgery is not an optionin mild-to-moderate OSA. The aims ofbariatric surgery are to reduce caloric intake

Table 1. Main otolaryngological and maxillofacial surgical treatments for OSA and their indications.

Interventionsite

Main procedures Main indications

Nose SeptoplastyTurbinoplastyValvuloplastyFESS

For subjects with nasal obstruction, mainlyto improve symptoms or feasibility of CPAPor MRA

Nasopharynx Adenoidectomy Adenoid hypertrophy

Oropharynx TonsillectomyUvulopalatopharyngoplastyUvulopalatal flapLateral pharyngoplastySoft palate RFITTPalatal implantsLaser-assisteduvulopalatopharyngoplasty

Tonsillar hypertrophy; in adults, mainly withother proceduresRetropalatal obstructionRetropalatal obstructionLateral pharyngeal wall collapseAlone, mainly for snoringRetropalatal obstruction in mild OSANot recommended for OSA

Hyoid bone Hyoid myotomy and suspensionHyoid myotomy withthyroidpexia

No longer usedHypopharyngeal obstruction

Tongue GlossectomyTongue RFITT

Marked tongue hypertrophy; rarely usedtodayModerate macroglossia and retrolingualobstruction, mainly in mild-to-moderate OSA

Maxillae/mandible

Mandible osteotomy with GAMaxillomandibular advancementDistraction osteogenesis

Hypopharyngeal obstruction; rarelyperformed aloneMandibular deficiency, severe OSA withobstruction at multiple sitesMainly for children with craniofacialabnormalities

Larynx Epiglottoplasty Obstruction at the epiglottic level

Trachea Tracheostomy Emergency situations; other treatments notfeasible in severe OSA

Multisite Variably combined procedures Obstruction at multiple sites in moderate-to-severe OSA

FESS: functional endoscopic sinus surgery; RFITT: radiofrequency interstitial thermotherapy; GA:

genioglossus advancement. Modified from Marrone et al. (2010).


and to alter the hormonal milieu involved innutrient absorption. Procedures can beclassified as restrictive or malabsorptiveinterventions, or a combination of both. Therestrictive procedures have the intention torestrict food intake by reducing the totalstomach capacity, such as gastric banding,the gastric balloon and gastric sleeveresection (gastroplasty). A purelymalabsorptive procedure is thebiliopancreatic derivation technique, alsocalled the Scopinaro technique. Due tosubstantial malabsorption problems, thistechnique is no longer used. Currently, theRoux-en-Y gastric bypass is the mostcommonly available, and combinesrestrictive and malabsorptive effects.Overall, gastric bypass is more efficaciousthan gastric banding or gastroplasty, butwith higher complication rates. At present,these techniques can be performed as alaparoscopic procedure. As with other majorsurgical interventions, considerablemorbidity and mortality have beendescribed. Along with deep venousthrombosis and impaired functional status,OSA itself is one of the three conditionsassociated with a risk of major adverseoutcome during the peri- and post-operativeperiod of bariatric surgery. Leakage at thestapler line is the most feared immediatepost-operative complication and occurs in1–5% of cases. In gastric banding, frequentvomiting can lead to slipping of the bandwith obstruction and semiurgentreintervention as a consequence (in 10% ofpatients). In addition, erosion of the gastricwall occurs in 1–4% of these cases.

Studies have shown the beneficial effect ofbariatric surgery on OSA. The ERS TaskForce on Non-CPAP Therapies in OSAreported a change in BMI of 10–24.4 kg?m-2,with a decrease in AHI of 44¡22 events?h-1

(-77%) after bariatric surgery. 94 (34.2%)patients were identified who showed apartial improvement in AHI, while 177(64.4%) patients were cured of OSA. Inanother meta-analysis, a reduction in AHI of34 events?h-1 was reported and 85.7% of thecases were cured. Positive results have alsobeen reported for metabolic consequencesand comorbid conditions. Unfortunately,

despite the usual massive weight lossobtained with bariatric surgery, cure of OSAis not systematically obtained and asubstantial number of patients have tocontinue CPAP. This reflects a significantinterindividual variability and emphasisesthe need for ongoing clinical and PSGfollow-up of these patients.

The optimal time post-surgery to reassessOSA severity is uncertain. The nadir ofweight loss reported following bariatricsurgery occurs after 1 yr and, subsequently,weight is regained.

Reassessment of AHI away from this nadirmay underestimate the benefit of bariatricsurgery, while reassessment at the nadirmay overestimate the long-term benefit. Thelack of randomised controlled trials in thisfield has been argued to be because it isunethical to randomise patients at futurerisk due to their comorbidities to nonsurgicaltreatment. This point of view, however,neglects the increased peri- and post-operative risks following bariatric surgery.

Drug treatment

Although there may be a predisposingairway abnormality, it is changes inrespiratory drive, airway tone or surfaceforces that cause airway closure duringsleep. Pharmaceutical agents might reducesleep apnoea by increasing respiratory drive,changing sleep structure (in particular,suppressing REM sleep), increasing upperairway muscle tone, changing respiratoryand cardiovascular reflexes that mayperpetuate apnoeas, and reducing surfaceforces that encourage closure of the upperairway. Several attempts have been made toidentify an effective pharmacologicaltreatment in patients with OSA and severalpotential targets have been considered.Most studies in this area have applied drugsthat are already in use in other medicalconditions. Since OSA comprises aspectrum of phenotypes (sex, age of onset,body composition, craniofacialabnormalities and comorbid conditions)that require specific approaches, we cannotexpect that a single drug therapy will fit allOSA patients. The challenge will be to


identify various subgroups of OSA patientsthat selectively respond to a specificpharmacological intervention. Based on theERS Task Force report on Non-CPAPtherapies, 24 studies of sufficient quality canbe identified, reporting the effects of 24drugs and recruiting between them 413subjects. Most studies have addressed thedirect interventional effects on OSA.However, no currently described drugconsistently reduces the severity of thecondition by more than 50%.

Protriptyline is a tricyclic antidepressant thatinhibits re-uptake of serotonin andnoradrenaline. It might reduce theproportion of REM sleep in people withREM-predominant OSA and increase theairway tone mediated by serotonin acting ongenioglossus via the hypoglossal nerve.There is no impact on the respiratory indicesbut there was an improvement in daytimesymptoms in two out of three trials,presumably due to a nonspecific alerting effectof the drug. Adverse side-effects, such as drymouth and urinary symptoms, are commonlyreported. Protriptyline does not have any placein the routine treatment of OSA.

Specific serotonin re-uptake inhibitors, suchas paroxetine, have been investigated aspossible treatments for OSA. In a smallsingle-night study, paroxetine (40 mg) hadno impact on AHI compared with placebo,in severe OSA. By contrast, in a study with 6-week treatment arms, paroxetine (20 mg)was shown to reduce AHI to 23.3 events?h-1,compared with 30.3 events?h-1 for placebo.There was a positive impact on respiratoryevents in non-REM, but not REM, sleep andno improvement in daytime symptoms.

Mirtazapine is another drug withantidepressant activity that acts as anagonist of some serotonin receptors and canalso increase serotonin secretion. Thismight increase serotonergic tone to thehypoglossal nerve, which could beparticularly helpful during REM sleep.Reductions in the AHI were reported in onestudy but could not be reproduced in twomulticentre trials, while many participantsreported side-effects of sleepiness andweight gain.

Acetazolamide inhibits carbonic anhydrase,producing a metabolic acidosis thatincreases ventilatory drive. It was shown toreduce AHI in a study of 10 subjects butthere was no positive impact on daytimesymptoms after 1 week of treatment. Thesubjects who had responded best to thedrug were offered a more prolonged trial oftreatment but only one could tolerate it inthe long term. Therefore, it has no role in theroutine management of OSA.

Among the other drugs that have beentrialled, naltrexone, theophylline andaminophylline have been shown to reducethe number of respiratory events overnight,but this was at the expense of sleepcontinuity and total sleep time, which makesthem unsuitable agents for the treatment ofOSA. Intranasal steroids, as a singleintervention, are also not recommended fortreatment of adult OSAS.

At the present time, there is no evidencethat any drug is likely to benefit anunselected patient with OSA. It is likely thatbetter characterisation of the predominantmechanisms of OSA in individual patientswill lead to better results and this also needsfurther study. It is recommended that noneof the drugs investigated to date is used totreat OSA.

Recognition and initial treatment ofcomorbidities

It is recognised that OSA patients oftenpresent with cardiovascular risk factors atdiagnosis, which gives them a predicted 10-yr cardiovascular risk of .30% according toFramingham studies. 46% havedyslipidaemia, 33% are obese, 11% haveclinical diabetes mellitus, 4% have glucoseintolerance, 11% have establishedcardiovascular consequences and 68% havearterial hypertension. This multi-comorbidityof OSA lends itself to a multidisciplinaryapproach, incorporating dieticians,psychologists, physiotherapists, sleepspecialists, cardiologists and surgeons. Ifsleep apnoea is associated with severeobesity, lowered energy intake and/orgreater energy consumption are required.Pharmacotherapy can be considered, but at


present, only the lipase inhibitor orlistat isavailable and its long-term effects arelimited. The best weight loss results withpharmacotherapy are obtained whenmedication is used as an adjunct to anintensive diet and lifestyle-changeprogramme, and the effects of thesestrategies are additive. Overall, the targetweight does not need to be the normalweight, which is usually an unrealisticobjective for severely obese patients.Arresting the accumulation of weight canalready be a significant initial result. Themorbidly obese can be referred to a bariatricsurgeon. Added metabolic benefits may beachieved with weight reduction, includingimprovements in insulin resistance, high-density lipoprotein cholesterol, and visceraland subcutaneous abdominal fat, which areassociated with a reduction in cardiovascularrisk. Moreover, the association between OSAand traffic accidents or metabolic disorderscould broaden the scope of the indication ofCPAP treatment, because physicians could betempted to opt for treatment with CPAP,despite the absence of OSA-related symptoms.

Overall, cardiovascular comorbidities arerelated to OSA severity, including AHI andoxygen desaturation severity indices. This istrue, both in general and clinicalpopulations, regarding arterial hypertension,nocturnal arrhythmias, ischaemic heartdisease and stroke. This is also found withregard to vascular subclinical markers, suchas carotid intima-media thickness and pulsewave velocity. Additional outcomes could be

considered, such as attention deficits andearly cardiovascular changes (table 2).Regarding this latter point, it would seemappropriate to use subclinical cardiovascularmarkers, since they have been demonstratedas predictors of future cardiovascularmorbidity. CPAP treatment could beindicated when OSA is associated withcardiovascular disorders, regardless of thepresence of related symptoms. This is ofparamount importance in mild orasymptomatic OSA, given its frequency.Identification of cardiovascular phenotypesmay be helpful in difficult therapeuticdecisions. Effects of other therapies, like oralappliances, on cardiovascular and metabolicoutcomes remain to be further evaluated.

Although CPAP is effective regardingsleepiness, daytime functioning and bloodpressure, it is also obvious that mostchronic consequences of OSA may not befully reversed by CPAP alone. The degree ofblood pressure reduction achieved bytreatment in hypertensive OSA variesbetween the therapies, with a b-blockersbeing the treatment of choice in one study.However, antihypertensive drugs are farmore effective than CPAP at controllingblood pressure, although there is abeneficial effect when combining CPAP withthe angiotensin II receptor antagonistvalsartan compared with CPAP alone. Inaddition, residual excessive daytimesleepiness is relatively prevalent in OSA,despite adequate CPAP treatment, whichmay require use of wakefulness stimulants,

Table 2. Screening for cardiovascular (CV) and metabolic consequences.

Undiagnosed hypertension Clinical and ambulatory blood pressuremeasurement

Early atherosclerotic processes PWVIMTCT angiography

Diastolic dysfunction Transthoracic echocardiography

Global evaluation of CV risk (obesity,diabetes, dyslipidaemia)

GlycaemiaHbA1cCholesterolTriglycerides

PWV: pulse wave velocity; IMT: intima-media thickness; CT: computed tomography; Hb: haemoglobin.


after excluding any specific cause ofsuboptimal objective adherence to CPAP,such as ill-fitting PAP masks, insufficientsleep, poor sleep hygiene or other sleepdisorders, (e.g. narcolepsy, periodic legmovements, restless legs syndrome anddepression), and improving CPAP duration,if feasible. Last but not least, sleep apnoeamay be associated with depression, whichmust be treated appropriately usingmodern antidepressants.

In conclusion, comorbidities are highlyprevalent in OSA and should be screened forin order to stratify these patients to theoptimal treatment modality. Moreover,outcome of OSA treatment has not only tobe evaluated in terms of AHI, but also to berated in terms of sleep improvement,wellbeing, improved cognitive function,reduction of comorbid metabolic orcardiovascular disorders, and modificationof the risk of developing cardiovascularcomplications in OSA. It is also obvious thatmost chronic consequences of OSA may notbe fully reversed by CPAP alone and thatcombined therapies, associating CPAP anddisease-modifying drugs, should be furtherevaluated. This will broaden the spectrum oftherapeutic choice, but also emphasises theneed for specific training for physicians toenable them to make the right choices.

Further reading

N Boudewyns A, et al. (2006). Surgicaltreatment for obstructive sleep apnea.Prog Respir Res; 35: 167–173.

N Chang CL, et al. (2010). Weight losstreatment for OSA: medical and surgicaloptions. Eur Respir Monogr; 50: 302–320.

N Epstein LJ, et al. (2009). Clinical guidelinefor the evaluation, management andlong-term care of obstructive sleep apneain adults. J Clin Sleep Med; 5: 263–276.

N Hedner J, et al. (2010). Pharmacologicalmanagement of sleep-disordered breath-ing. Eur Respir Monogr; 50: 321–339.

N Hui DSC, et al. (2000). Effects ofaugmented continuous positive airwaypressure education and support on com-pliance and outcome in a Chinesepopulation. Chest; 117: 1410–1416.

N Kiely JL, et al. (2000). Cardiovascular riskfactors in patients with obstructive sleepapnoea syndrome. Eur Respir J; 16: 128–133.

N Levy P, et al. (2010). Outcomes of OSAand indications for different therapies.Eur Respir Monogr; 50: 225–243.

N Marklund M, et al. (2012). Non-CPAPtherapies in obstructive sleep apnoea:mandibular advancement device therapy.Eur Respir J; 39: 1241–1247.

N Marrone O, et al. (2010). Upper airwaysurgery. Eur Respir Monogr; 50: 286–301.

N Montserrat JM, et al. (2007). Diagnosticand therapeutic approach to nonsleepyapnea. Am J Respir Crit Care Med; 176: 6–9.

N Oksenberg A, et al. (1998). The effect ofbody posture on sleep-related breathingdisorders: facts and clinical implications.Sleep Med Rev; 2: 139–162.

N Oksenberg A. (2005). Positional andnon-positional obstructive sleep apneapatients. Sleep Med; 6: 377–278.

N Pepin JL, et al. (2010). Comparison ofcontinuous positive airway pressure andvalsartan in hypertensive patients withsleep apnea. Am J Respir Crit Care Med;182: 954–960.

N Randerath WJ, et al. (2011). Non-CPAPtherapies in OSA. Eur Respir J; 37: 1000–1028.

N Sher AE, et al. (1996). The efficacy ofsurgical modifications of the upper airwayin adults with obstructive sleep apneasyndrome. Sleep; 19: 156–177.

N Smith CE, et al. (2009). Patient educationcombined in a music and habit-formingintervention for adherence to continuouspositive airway (CPAP) prescribed forsleep apnea. Patient Educ Couns; 74:184–190.

N Smith I, et al. (2009). Educational,supportive and behavioural interventionsto improve useage of continuous positiveairway pressure machines for adults withobstructive sleep apnoea. CochraneDatabase Syst Rev; 2: CD007736.

N Strollo PJ, et al. (1996). Obstructive sleepapnoea. New Engl J Med; 334: 99–104.

N Wang W, et al. (2011). Effects of patienteducation and progressive muscle relaxa-tion alone or combined on adherence tocontinuous positive airway pressure treat-ment in obstructive sleep apnea patients.Sleep Breath; [Epub ahead of print DOI:10.1007/s11325-011-0600-3].


Positive airway pressuretreatment

Anita K. Simonds

Principles and machines

The fundamental cause of OSA is upperairway collapse. Positive airway pressuretherapy (PAP) is used to redress thiscollapse, working as a pneumatic splint. PAPcan be delivered constantly at a fixed levelthroughout inspiration and expiration(continuous positive airway pressure,CPAP), as a variable pressure via an auto-adjusting device (APAP), or as a bi-levelpressure support system that provides ahigher level of inspiratory and expiratorysupport sufficient to maintain airwaypatency (box 1). CPAP may also improvefunctional residual capacity (by recruitingalveoli), reduce the work of breathing andimprove haemodynamics in patients withheart failure, by reducing preload andafterload. An increase in lung volume hasrelevance, as it has been suggested that thisvolume change associated with PAP pullsdown on the trachea and stretches theairway open. Alternatively, it may reflexivelyincrease upper airway dilator muscle tone

(Gordon et al., 2005). The relativeimportance of these mechanisms is unclear,but upper airway stenting seems to be thedominant factor. By preventing recurrentcollapse and trauma throughout the night,CPAP use may secondarily reduce upperairway oedema, and this combined withimprovement in sleep quality may improveupper airway reflexes.

PAP machines For CPAP, eithercompressors or, more usually, blowerdevices, are used to deliver bias flow. Flowrates of 20-60 L?min-1 are required to deliverpressures of up to 20 cmH2O. To ensure aconstant pressure throughout therespiratory cycle, flow is increased duringinspiration and reduced during expiration.CPAP is delivered to the patient via aninterface, and each circuit requires anexhalation option to minimise rebreathing.

A wide range of portable CPAP systems isavailable. Ideally, these should be reliable,small, portable and inexpensive and shouldfunction as noiselessly as possible. Theyshould provide compliance data, and mostproduce additional data on leaks, flow, andrespiratory events. It should be noted thatrespiratory events as determined by theCPAP machine algorithm may not be directlycomparable to events detected by othersleep monitoring systems.

Many machines provide a gradual rampedincrease in CPAP to the preset level in thefirst 5–20 min of use. Some individuals,particularly those beginning treatment orwho require high CPAP levels, find thishelpful. Others find the ramp unnecessaryand prefer to ’get on with’ thepredetermined pressure as soon as the

Key points

N Auto-adjusting CPAP may be as effectiveas PSG titration in establishingovernight pressure requirements.

N Use of CPAP for at least 4 h per nightis required to control sleepiness. Usefor longer periods, e.g. 5–6 h, may berequired to control cardiovascularvariables and produce furthercognitive benefit.

N CPAP adherence may be improvedwith a problem-solving approach.


mask is in place. Expiratory pressure releaseapproaches are employed in an attempt toimprove patient comfort. Here, the flowgenerator detects the start of inhalation anddecreases motor speed to reduce thepressure level during expiration. Thesemodifications might help some patientstolerate therapy better empirically, but havenot been evaluated in crossover trials.

Autotitrating positive airway pressure(APAP) devices have been developed, inwhich pressure is varied throughout thenight with the aim of matching patientrequirements better. A meta-analysis oflong-term fixed-level versus autotitratingdevices, has shown that, compared tostandard CPAP, APAP is associated with areduction in mean pressure. However, APAPand standard CPAP were similar in terms ofpatient adherence, their ability to eliminaterespiratory events and their ability toimprove subjective sleepiness. Given thatAPAP is more costly than standard CPAP,APAP should not be considered first-linechronic therapy in all patients with OSA.However, APAP may be useful in othersituations, e.g. home titrations (see below)or detection of mouth leak, or in certainsubgroups of patients with OSA who requirehigh pressures. Crucially, when consideringAPAP devices, attention should be paid tothe type of breathing event the algorithmresponds to (flow limitation, snoring,apnoea, etc.), whether the device respondsto the severity of the event, whether itresponds pre-emptively, the ability to detectand respond to leak, the rate of response,and how it deals with ‘false’ events such asswallowing, coughing, arousal from sleep orcentral apnoeas – plus, of course, whetherthe algorithm has been validated in clinicalcircumstances. It is probable that responseto flow limitation is more effective thanresponse to apnoeas. Short-term use of

autotitrating devices to determine anappropriate fixed-level pressure setting isconsidered below.

PAP interfaces

A significant part of the improvement ofCPAP delivery over the past decade can beattributed to improved mask design andcomfort. Broadly, interfaces can be classifiedas: a) nasal; b) oro-nasal; c) full-face(covering last part of face/head); or d) oral.The advantages and disadvantages of eachare given in table 1. Oro-nasal interfacesminimise problems from mouth leak, as leakpromotes increased flow, which can increasenasal resistance, in turn leading tosymptomatic rhinitis, which may limitadherence. While heated humidification mayreduce nasal symptoms and aid compliancein some patients, trials of humidificationhave produced equivocal results forenhancing adherence in a general CPAPpopulation. Leaks from around the interfaceare addressed by ensuring optimal mask fitand by design features that includeinflatable or mouldable mask cushions.There is good evidence that mask leaks notonly limit the efficiency of ventilation butalso impair sleep quality. Furthermore, noisyair leaks can affect sleep quality in partners.Most nasal and oro-nasal CPAP interfacesare vented, such that dead space is flushedby continuous flow through the circuit. Deadspace is also a vital determinant of efficacy.It is important to consider effective (ordynamic) dead space, rather than staticdead space, as the flow through the interfaceand position of vents are vital. For example,although the Total face mask (PhilipsRespironics, Best, The Netherlands), whichcovers the whole face, has a large staticdead space, effective dead space is verysmall as exhalation ports are closelyadjacent to the nares. A further complicationof interfaces is pressure sores or discomfort,particularly affecting the bridge of the nose,forehead and maxillary region. Whilealternation of interfaces may help (e.g.swapping to nasal plug system in somepatients and in children), customised maskshave been shown to be superior tocommercial interfaces in reducing pressure

Box 1. Terminology of positive airway pressure (PAP).

Continuous positive airway pressure (CPAP)Autoadjusting positive airway pressure(APAP)Bi-level positive airway pressureExpiratory positive airway pressure (EPAP)


area problems. In addition, a problematicadverse effect of long-term CPAP or NIV inchildren is facial skeletal maldevelopment,usually mid facial hypoplasia. Newer,smaller nasal interfaces can help but aswitch to a customised interface may be thebest solution (Fauroux et al., 2005). Facialdevelopment should be assessedradiologically and photographically every1–2 years in children requiring long-term CPAP.

PAP titration

While detailed PSG-directed manual titrationof CPAP has a long track record, recent workhas shown that PSG is not necessary toestablish a therapeutic CPAP level. Clinicalalternatives include:

N Algorithmic derivation of CPAP basedon AHI or SaO2 dip rate and neck size,and/or BMI.

N Use of autotitrating devices for a singlenight or longer periods.

N Empirical (educated ‘guesstimate’) e.g.10 cmH2O

Algorithms are attractive and several havebeen validated (Stradling et al., 2004;Miljeteig et al., 1993). See Box 2.

PSG titration versus algorithm or unattendedautotitrating device In a large clinical trial(n.300) of newly diagnosed OSA patientswith AHI .30 events?h-1 and mean ESS of15, randomised to standard PSG titration,unattended home autotitrating device oralgorithm-generated (Miljeteig–Hoffsteinmethod) determination of CPAP level, Masaet al. (2004) found similar improvement inAHI, subjective sleepiness and compliancein all three groups over 12 weeks’ follow-up.A further study in CPAP-naıve patientsshowed similar outcomes in terms ofefficacy, symptom control and adherence,

Table 1. Advantages and disadvantages of interfaces for CPAP.

Interface Advantages Disadvantages

Nasal mask Allows feeding, speechEasier for coughingCan be less claustrophobicLess risk of aspiration ofvomit, or gastric distension

Mouth leaksCan be a problem to use inedentulous patientsLimited efficacy and may causeside-effects in patients withnasal obstruction, rhinitis andnose bleeds

Oro-nasal or full-face mask No leak from mouthMay be more stable pressureduring sleepHelpful in patients with nasalobstruction

May be claustrophobic(although total face mask mayreduce this)Can be associated withaspiration of vomit (but rare instable patients)May be more likely to causegastic distension

Nasal plugs Less surface area in contactwith faceCan be used to vary pressureeffects on face and allowhealing of nasal bridgepressure sores, and reducemid-facial hypoplasia inchildrenCan be helpful inclaustrophobic patients

Mouth leaksCan be more unstable fit thanfull-face mask


when comparing algorithm-designed fixedlevel CPAP followed by self-adjustment athome with standard titration. West et al.(2006) compared three methods of CPAPdelivery over 6 months. In this study of 98patients with an SaO2 .4% dip rate of.10 events?h-1 and ESS of .9 wererandomised to: 1) Autotitration CPAP(Autoset Spirit; ResMed, San Diego, CA,USA) throughout the study period; 2)Autotitration CPAP (Autoset Spirit) for1 week followed by remaining period withCPAP fixed at 95th centile of pressure in firstweek; and 3) fixed CPAP level set usingOxford algorithm. Outcome measuresassessed at 1 and 6 months were ESS,OSLER wake test, 24-h blood pressure, SF-36 health status score, sleep apnoea-relatedquality of life and CPAP compliance. Theauthors found no difference in any of thesevariables, nor in the CPAP monitoring databetween the groups. The 95th centilepressure levels in the 6-month autotitrationand 1-week autotitration groups were higherthan in the algorithm-generated pressuregroup, although mean pressure was lower inthe 6-month autotitration group.

In an earlier investigation comparingpressure determined by 1 month ofautotitrating CPAP, 4 nights of autotitratingCPAP, Oxford algorithm or standard fixed-pressure (10 cmH2O) CPAP, Stradling et al.(2004) found no disadvantage to using thealgorithm, and indeed the empirical10 cmH2O setting was reasonably effective.

All of these studies confirm that complexlaboratory PSG manual titration does notoffer advantages over autotitrating devicesor simple algorithm set-up. This is clearly animportant finding from a cost-efficiencyviewpoint, and useful for those managingclinical pathways. It should be noted that

the patients studied had relatively severe,symptomatic OSA: the results cannot bereadily translated to individuals with mildOSA, although they should at leastprovide a useful guide to set-up in themajority of patients.

PAP adherence

Optimal CPAP use has previously beendefined as use for .4 h for .70% of nights.However, a linear relationship has beenshown between CPAP use and improvementin ESS up to 5 h, and improvement inmemory was significantly greater in thoseusing CPAP for .6 h per night comparedwith ,2 h (Weaver et al., 2007). Therelationship between CPAP ‘dose’ andsubjective and objective sleepiness andfunctional outcomes is shown in figure 1.Barbe et al. (2010) found the reductions inblood pressure were seen only with CPAPused for .5.6 h per night in hypertensive

Box 2. Algorithms for setting CPAP level.

Oxford algorithmStradling et al., 2004

CPAP level 5 .4% oxygen saturation dip rate6 0.048) + (neck size (inches#) 6 0.325) + 2.1

Miljeteig–Hoffstein methodMiljeteig et al., 1993

CPAP level 5 (0.16 6 BMI) + (0.13 6 neckcircumference) + (0.04 6 AHI) - 5.12

#: If neck size measured in cm, multiply CPAP level by 0.128.

0.7

0.3

0.6

0.5

0.4

0.2

0.1

0

Nor

mal

val

ues

%

Nightly CPAP use h

≤2 >2–<4 ≥4–<5 ≥5–<6 ≥6–<7 ≥7

▲▲

◆◆

◆◆ ◆

◆

■

■ ■■

■

■

▲

▲▲ ▲

FIGURE 1. Dose of CPAP and effects on sleepinessand functional outcomes. Cumulative proportionof participants obtaining normal threshold valueson the ESS (squares), MSLT (triangles), andFunctional Outcomes of Sleep Questionnaire(diamonds). Reproduced from Weaver et al.(2007), with permission from the publisher.


OSA patients. Theoretically it is possiblethat differing periods of use determinedifferent outcomes, including cognitive,autonomic and cardiovascular short- andlong-term effects. Data so far would suggestthat at least 5–6 h use is necessary. There isa variety of predictors of compliance relatedto the patient, equipment, initiation oftherapy and social environment (Sawyeret al., 2011). Not surprisingly, severity ofOSA and high symptom burden areassociated with better adherence, as is thepresence of a supportive partner. Upperairway side effects, increased nasalresistance, depression, claustrophobia andpassive coping personality type limit useand should be specifically addressed. Long-term adherence to CPAP can be predictedwithin the first 3 nights’ use, so attentiondirected to educational assistance andproblem solving at the outset is vital.

PAP problem solving A step-wise,collaborative approach with the patient isbest. After discussion with the patient todetermine acceptance, understanding of the

technique and limiting factors, usefulinformation can be obtained from downloaddata from the device. A variety of patterns ofuse can be seen but use every night for shortinterrupted periods often indicates interface,leak and upper airway problems, whileerratic and only occasional night use mayindicate limited understanding of therationale of therapy, inappropriate pressuresetting, poor buy-in to the concept oftherapy, insomnia or noise problems. Aguide to problem solving is given in figure 2.

CPAP use and comorbidity In individualswho remain hypersomnolent orsymptomatic despite satisfactory adherenceand demonstration of effective delivery oftherapy on titration study, a further cause forsleep disruption may be present. Relativelycommon causes include restless legssyndrome, inadequate sleep opportunityand conditions such as arthritis, poorasthma control or back pain, which mayfragment sleep. Additional intrinsic sleepdisorders should be explored by PSG(Simonds, 2007).

Poor control of SDB

Check device performance, review diagnosis. Are obstructive events controlled? Consider further titration and pressure settings change. Are leaks responsible? Does patient have additional hypoventilation?

Problems with rationale, concept, acceptability

Mask discomfort, rhinitis, upper airway dryness, pressure sores, leaks

Consider different interface, humidification, education on mask fit

Consider patient education and psychological support, more intensive home support

Discuss with patientExamine patterns of useExamine AHI, leaks

Sleep problems,persistent somnolenceInsomnia, sleep initiation, sleep maintenance difficulty, noise

Consider other pathology e.g. restless legs (assess with more detailed sleep study)Consider course of short-acting sedative e.g. zopiclone to aid sleep initiation

Consider bi-level ventilation if hypercapnic

Review download data from PAP device

FIGURE 2. Adherence to PAP and problem solving.


Alternatives to CPAP

Bi-level positive airway pressure ventilatorsUse of inspiratory pressure support (IPAP)combined with expiratory positive airwaypressure (EPAP) is indicated in patients withupper airway obstruction coupled withalveolar hypoventilation. In practice, this iscommonly seen in patients with obesityhypoventilation syndrome (Piper et al., 2011)or neuromuscular disorders such asDuchenne muscular dystrophy, whereobstructive apnoeas may be seen in theteenage years only to be followed bynocturnal hypoventilation as diaphragmstrength progressively deteriorates. Acombination of OSA and nocturnalhypoventilation may also be seen in someCOPD patients, and hypopnoea andhypoventilation are more commonly seenthan pure obstructive apnoeas in childrenwith OSA. As a rule, bi-level pressuresupport is required in patients with diurnalhypercapnia (PaCO2 .7 kPa) or nocturnalhypoventilation not controlled by CPAP,and may be successful in patients withhigh pressure requirements who fail totolerate CPAP.

CPAP versus ASV and bilevel ventilation in HFpatients with SDB Over 50% of patients withchronic HF have SDB and, contrary toprevious belief, this is true not just inpatients with end-stage disease but those inNew York Heart (NYHA) classificationgroups II–IV. CPAP in patients with OSAproduces improvement in symptoms,quality of life and left ventricular ejectionfraction. However, in HF patients withpredominantly central sleep apnoea, arandomised controlled trial (Bradley et al.,2005) showed no advantage of CPAP –indeed there was excess mortality in theCPAP group, although post-hoc analysisshowed that the latter occurred in patientsin whom CPAP did not suppress respiratoryevents (Arzt et al., 2005). Identifying HFpatients with CSA is difficult, as their ESS isusually normal. Adaptive servo ventilation(ASV) varies IPAP delivery to smooth outcentral apnoeas and subsequenthyperpnoeas in the Cheyne–Stokes patternof breathing, while concomitant obstructive

events are dealt with by setting a fixed levelof CPAP. Short-term trials have shown thatASV may be more effective than CPAP inCSA. The approach is being evaluated in theEuropean multicentre SERVE-HF trial, andmanagement of CSA in HF is discussedfurther later in this chapter.

Mandibular advancement splints (MAS) arean important alternative to CPAP in mildand moderate OSA, and may be used inpatients who fail to tolerate CPAP. MAS areconsidered in the next section of this book.

Expiratory positive airway pressure (EPAP) Arecent development is the application ofnasal valves designed to increase expiratoryresistance and thus generate a level ofpositive airway pressure on expiration whichis proportional to respiratory flow. In earlytrials, the device has been shown to reduceAHI and daytime subjective sleepinesscompared to sham control and was welltolerated. Since no flow occurs during anapnoea, the device cannot control eventsthat have already occurred, and mouth leakmay be a limiting factor. Further work iscontinuing to explore the role of EPAP inOSA therapy in comparison to CPAP andalternative approaches.

Further reading

N Barbe F, et al. (2010). Long-term effect ofcontinuous positive airway pressure inhypertensive patients with sleep apnea.Am J Resp Crit Care Med; 181: 718–726.

N Bradley TD, et al. (2005). Continuouspositive airway pressure for central sleepapnea and heart failure. New Engl J Med;353: 2025–2033.

N Fauroux B, et al. (2005). Facial sideeffects during noninvasive ventilation inchildren. Int Care Med; 31: 965–969.

N Gordon P, et al. (2005). Positive airwaypressure for obstructive sleep apnoea/hypopnoea syndrome. Thorax; 60: 68–75.

N Masa JF, et al. (2004). Alternative meth-ods of titrating continuous positive air-way pressure. A large multicenter study.Am J Respir Crit Care Med; 170: 1218–1224.


N Miljeteig H, et al. (1993). Determinants ofcontinuous positive airway pressure levelfor treatment of obstructive sleep apnea.Am Rev Respir Dis; 147: 1526–1530.

N Piper AJ, et al. (2011). Obesity hypoventi-lation syndrome. Am J Resp Crit Care Med;183: 292.

N Sawyer AM, et al. (2011). A systematicview of CPAP adherence across agegroups: clinical and empiric insights fordeveloping CPAP adherence interven-tions. Sleep Med Rev; 15: 1–14.

N Simonds AK. (2007). Continuous positiveairway pressure therapy for sleep disor-dered breathing in adults. In: SimondsAK, ed. Non-Invasive Respiratory Support.A Practical Handbook. London, HodderArnold; pp. 291–322.

N Stradling JR, et al. (2004). Relativeaccuracy of algorithm-based prescrip-tion of nasal CPAP in OSA. Respir Med;98: 152.

N Weaver TE, et al. (2007). Relationshipbetween hours of CPAP use and achiev-ing normal levels of sleepiness and day-time functioning. Sleep; 30: 711–719.

N West SD, et al. (2006). Comparison ofthree ways to determine and deliverpressure during nasal CPAP therapy forobstructive sleep apnoea. Thorax; 61:226–231.

N Arzt M, et al. (2005). Association of sleep-disordered breathing and the occurenceof stroke. Am J Respir Crit Care Med; 172:1447–1451.


Treatment of central sleepapnoea

Winfried Randerath

Optimal treatment for patients with CSA isstill under investigation. However, a cleardefinition of the disease and diagnosis ofunderlying diseases forms the basis of anytherapeutic approach.

Central apnoea associated withhypoventilation

Central apnoea associated withhypoventilation appear in patients withinsufficient respiratory drive or reduced V9E

due to neuromuscular or thoracoskeletaldisorders. For these patients, NIV is thetreatment of choice. It normalises V9E andmay also improve survival (Aboussouanet al., 2001; Bourke et al., 2006).

CSA associated with hyperventilation

There is currently a lack of evidence provinga survival benefit under treatment of CSAassociated with hypoventilation. However,many studies have shown that survival ofpatients with cardiovascular disorders issignificantly reduced when associated withcentral breathing disturbances (Yuminoet al., 2008). In addition, CSA is a marker ofpoor outcome in patients with stroke andrenal failure. Therefore, it can hardly beaccepted from a clinical point of view toleave CSR in patients with cardiovasculardiseases or central breathing disturbancesin patients with stroke or renal failureuntreated. In addition, it is unclear to whatdegree central breathing disturbances impairquality of life or daytime functioning inpatients with underlying cardiovasculardiseases or organ failure. Therefore, diagnosisand evaluation of treatment efficacy duringtitration and follow-up must be based on PSGanalysis of central respiratory disturbances inthese patient groups.

The overwhelming majority of CSA isassociated with the aforementioneddisorders, so the first step of treatment isthe abolition of these causes. This includespharmaceutical or surgical improvement of

Key points

N NIV is the therapy of choice in centraldisturbances associated withhypoventilation, if there are nosufficient options to treat theunderlying disease in those patients.

N The therapeutic approach to patientswith central disturbances associatedwith hyperventilation begins withoptimal treatment of any underlyingdisease, including pharmacological orinterventional cardiac or cerebraloptions.

N There is limited evidence showingeffectiveness of treatment with oxygenor pharmaceutical agents(theophylline, acetazolamide).

N After optimal treatment of underlyingdiseases, the application of positiveairway pressure, especially CPAP,should be trialled. The evidence for bi-level therapy is limited.

N There is growing evidence thatadaptive servo ventilation normalisesventilation, improves cardiac functionand quality of life in CSA patients ifCPAP fails.


cerebral blood flow, therapy of inflammatoryor tumorous brain diseases orpharmacological or interventional treatmentof cardiac disorders. Few non-randomisedstudies have investigated the effect ofcardiac drugs in HF patients with breathingdisturbances. ACE inhibitors and b-blockershave been shown to improve cardiacfunction and reduce CSA in small groups ofthese patients. Cardiac resynchronisation(CRT) therapy effectively improves severeleft ventricular systolic dysfunction. Therapywith atrial synchronised bi-ventricularpacemakers reduced the duration andnumber of central disturbances. If centraldisturbances remain after optimal treatmentof the underlying disease, strategiesinterfering with hypoxia, apnoeic threshold,chemosensitivity or cardiac or lung mechanicsmay be discussed (Brack et al., 2012).

However, it is unclear whether therapygenerally improves survival or quality of lifein patients with primary CSA. Thus, theindication for treatment should be based onboth PSG parameters such as breathingdisturbances and sleep profile on the onehand, and impairment of quality of life onthe other.

Oxygen

The application of oxygen may improve leftventricular function and reduce the reflexactivation of the peripheral chemoreceptors.Several studies have shown that the oxygendesaturation index can be normalised underoxygen supply. However, AHI has only beenreduced by about 50% and sleepparameters, clinical symptoms andcognitive impairment have not beensufficiently influenced in patients withchronic HF and CSR (Staniforth et al., 1998;Krachman et al., 1999). Finally, there areconflicting results on the influence of oxygenon sympathetic activity (Staniforth et al.,1998). Based on these data, oxygen supplycannot be recommended as a first linetherapy in CSA patients.

Application of CO2

Central apnoeas emerge when the prevailingCO2 level falls below the apnoeic threshold.The application of CO2 raises its blood level

clearly above the threshold and thusstabilises respiration. It has been shown thatelevation of the CO2 level by externalapplication or by enlargement of dead spaceis able to normalise ventilation in CSApatients (Xie et al., 1997; Andreas et al.,1998; Lorenzi-Filho et al., 1999). However,the application of CO2 might increasesympathetic activity (Andreas et al., 1998).Nevertheless, this treatment option cannotbe recommended until safety concerns ofgas application have been resolved.Moreover, long-term prospective data areneeded to define clinical outcome andadverse effects.

Pharmaceutical options

Acetazolamide and theophylline have beenconsidered to improve central respiratorydisturbances by influencing the respiratorycontrol system in the brain stem.Acetazolamide is a mild diuretic andrespiratory stimulant. A randomised,double-blind, cross-over, placebo-controlledshort-term study showed a reduction ofcentral disturbances by half, but failed tonormalise the disturbances (Javaheri, 2006).In a single-blind, randomised placebo-controlled study, theophylline reduced thetranscutaneous PtcCO2 indicating thestimulation of ventilation (Andreas et al.,2004). However, there are concerns that thismight aggravate the loop gain. Therefore,pharmaceutical treatment cannot currentlybe recommended for patients with CSA.

Positive airway pressure treatment

Positive airway pressure (PAP) can beapplied constantly (CPAP), varying betweeninspiration and expiration (bi-level) orcontinuously changing on a breath-by-breathbasis (ASV). PAP options are commonlyconsidered to maintain airway patency andenhance alveolar ventilation. Consecutively,ventilation/perfusion mismatches in thelungs improve and the intrapulmonary gasreservoir enlarges, which reduces thevariation of gas proportions. Moreover, PAPinfluences cardiac function mechanically.Positive pressure application may reducethe venous return to the heart and thecardiac index in healthy persons. In contrast,


PAP reduces pleural pressure swings andthus left ventricular afterload in HF patients.Moreover, it reduces the work of breathingand therefore the oxygen consumption ofthe respiratory muscles. These effectsimprove heart function (Kaneko et al., 2003).

Several studies have shown that CPAPreduces AHI by about 50% but does notnormalise respiration. Ongoing therapymight lead to additional improvement after3 months in some individuals (Arzt et al.,2009). A larger study (the CanPAP trial)confirmed the reduction of respiratorydisturbances by half, improvements ofminimum oxygen saturation and leftventricular function, but failed to show asurvival benefit over 3 years. Interestingly, apost-hoc analysis of the CanPAP data showedthat optimal suppression of respiratorydisturbances was associated withimprovement of survival (Bradley et al.,2005; Arzt et al., 2007 and 2008).

There are limited data on the use of bi-levelPAP in central breathing disturbances.Small-sized studies in CPAP non-respondersshowed a better reduction of centralrespiratory disturbances. However, there is alack of clinical data comparing bi-level PAPwith other treatment options. Long-termtrials and data on cardiovascular parametersare lacking. Therefore, bi-level PAP has notbeen proven to be superior to CPAP (Dohiet al., 2008).

Adaptive servo ventilation

ASV applies two different pressure levelsduring expiration and inspiration, similar tobi-level treatment. However, the inspiratorypressure changes continuously to overcomecentral disturbances and CSR. Thedifference between inspiratory andexpiratory pressure defines the tidal volume,i.e. the pressure support. By elevating theinspiratory pressure, tidal volume isincreased, which is needed during thehypoventilation periods in CSR. In contrast,during hyperventilation, the differencebetween inspiration and expiration isminimised because no additional pressuresupport is needed. Thus, the algorithms

overcome the overshooting andundershooting of ventilation in CSR.

The expiratory pressure is set to eliminateupper airway obstructions. It can be fixedbased on manual titration. The most recentalgorithms adapt the expiratory pressureautomatically according to the prevailinglevel of obstruction. In addition, all devicesapply mandatory breaths in case of centralapnoeas.

ASV has been compared with oxygen, CPAPand bi-level PAP in a small-sized randomisedstudy in 14 patients, and proved to be superiorto all other treatment options (Teschler et al.,2001). ASV has shown to improve chronic HFsympathetic activity and daytime performance(Pepperell et al., 2003). There is growingevidence that ASV is superior to CPAP interms of left ventricular function andcompliance (Kasai et al., 2010; Philippe et al.,2006). Most of these studies have beenperformed in patients with almost pure CSA/CSR. However, in patients with co-existingOSAS and CSA/CSR, ASV has been shown toimprove obstructive disturbances aseffectively as CPAP but to suppress centralapnoeas and hypopnoeas more sufficiently(Arzt et al., 2008; Randerath et al., 2009).

In recent years, other complicated breathingpatterns have been discussed intensively.These include central breathing disturbancesin patients using opioids and CSA emergingunder CPAP therapy. Data on the optimaltherapy of opioid-induced sleep apnoea arecontroversial. While a retrospective analysisfound no difference between CPAP and ASV,a prospective small sized case series showedsuperiority of ASV by using higher pressuresupport (Farney et al., 2008; Javaheri et al.,2008). Finally, ASV was compared to CPAP,CPAP plus oxygen, bi-level PAP inspontaneous and spontaneous/timed modein patients with complicated breathingpatterns (CSA/CSR/CPAP emergent CSA).ASV normalised breathing disturbances in allsubgroups and was more effective than theother treatment options.

Despite the growing evidence on the efficacyof adaptive servoventilation in patients withcentral breathing disturbances, there is still


a lack of data on survival. Early long-termdata show that ASV improves centraldisturbances better than CPAP (Randerathet al., 2012). Moreover, two large sized studieson long-term outcome have been started.

In conclusion, the clinical pathway generallystarts with optimising the treatment of anyunderlying disorder. In patients with cardiacdisorders and CSR or CSA, a CPAP trial isrecommended for two reasons. Firstly, CPAPsufficiently suppresses central respiratorydisturbances in about 50% of patients.Secondly, the CPAP titration determines theexpiratory pressure necessary to overcomeupper airway obstruction. If CPAP does notnormalise CSA/CSR, a switch to ASV shouldbe considered. As survival is significantlyreduced in patients with HF and CSA/CSR,we generally recommend sufficient positivepressure treatment (CPAP or ASV). Oxygenor bi-level PAP cannot generally berecommended in advance to ASV. As CSA/CSR is a marker of poor prognosis in strokeand renal failure, we also recommendpositive pressure treatment in thesepatients. Patients with primary CSA shouldbe treated if they suffer from clinicalsymptoms, cognitive impairment or daytimesleepiness. Finally, in patients with centralapnoea associated with hypoventilation, NIVis the treatment of choice if no causaltherapeutic options are available.

Further reading

N Andreas S, et al. (2004). Differentialeffects of theophylline on sympatheticexcitation, hemodynamics, and breathingin congestive heart failure. Circulation;110: 2157–2162.

N Andreas S, et al. (1998). Treatment ofCheyne–Stokes respiration with nasaloxygen and carbon dioxide. Eur Respir J;12: 414–419.

N Arzt M, et al. (2007). Suppression ofcentral sleep apnea by continuous positiveairway pressure and transplant-free sur-vival in heart failure: a post hoc analysis ofthe Canadian Continuous Positive AirwayPressure for Patients with Central SleepApnea and Heart Failure Trial (CANPAP).Circulation; 115: 3173–3180.

N Arzt M, et al. (2009). Time course ofcontinuous positive airway pressureeffects on central sleep apnoea inpatients with chronic heart failure. JSleep Res; 18: 20–25.

N Bourke SC, et al. (2006). Effects of non-invasive ventilation on survival and qual-ity of life in patients with amyotrophiclateral sclerosis: a randomised controlledtrial. Lancet Neurol; 5: 140–147.

N Brack T, et al. (2012). Cheyne–Stokesrespiration in patients with heart failure:prevalence, causes, consequences andtreatments. Respiration; 83: 165–176.

N Bradley TD, et al. (2005). Continuouspositive airway pressure for central sleepapnea and heart failure. N Engl J Med; 353:2025–2033.

N Dohi T, et al. (2008). Bi-level positiveairway pressure ventilation for treatingheart failure with central sleep apnea thatis unresponsive to continuous positiveairway pressure. Circulation; 72: 1100–1105.

N Farney RJ, et al. (2008). Adaptive servo-ventilation (ASV) in patients with sleepdisordered breathing associated withchronic opioid medications for non-malignant pain. J Clin Sleep Med; 4: 311–319.

N Javaheri S, et al. (2008). Adaptive pres-sure support servoventilation: a noveltreatment for sleep apnea associatedwith use of opioids. J Clin Sleep Med; 15:305–310.

N Javaheri S. (2006). Acetazolamideimproves central sleep apnea in heartfailure a double-blind, prospective study.Am J Respir Crit Care Med; 173: 234–237.

N Kasai T, et al. (2010). Effect of flow-triggered adaptive servo-ventilation com-pared with continuous positive airwaypressure in patients with chronic heartfailure with coexisting obstructive sleepapnea and Cheyne-Stokes respiration.Circ Heart Fail; 3: 140–148.

N Krachman SL, et al. (1999). Comparisonof oxygen therapy with nasal continuouspositive airway pressure on Cheyne–Stokes respiration during sleep in conges-tive heart failure. Chest; 116: 1550–1557.

N Lorenzi-Filho G, et al. (1999). Effects ofinhaled carbon dioxide and oxygen onCheyne–Stokes respiration in patientswith heart failure. Am J Respir Crit CareMed; 159: 1490–1498.


N Mansfield DR, et al. (2004). Controlledtrial of continuous positive airway pres-sure in obstructive sleep apnea and heartfailure. Am J Respir Crit Care Med; 169:361–366.

N Pepperell JC, et al. (2003). A randomizedcontrolled trial of adaptive ventilationfor Cheyne–Stokes breathing in heartfailure. Am J Respir Crit Care Med; 168:1109–1114.

N Philippe C, et al. (2006). Compliance withand effectiveness of adaptive servoventi-lation versus continuous positive airwaypressure in the treatment of Cheyne–Stokes respiration in heart failure over asix month period. Heart; 92: 337–342.

N Randerath W, et al. (2009). Longtermefficacy of adaptive servo-ventilation

(ASV) in patients with co-existingobstructive sleep apnoea (OSAS) andCheyne–Stokes respiration (CSR). A ran-domised CPAP-controlled trial. EurRespir J; 34: 38s.

N Staniforth AD, et al. (1998). Effect ofoxygen on sleep quality, cognitive func-tion and sympathetic activity in patientswith chronic heart failure and Cheyne–Stokes respiration. Eur Heart J; 19: 922–928.

N Teschler H, et al. (2001). Adaptivepressure support servo-ventilation: anovel treatment for Cheyne–Stokesrespiration in heart failure. Am J RespirCrit Care Med; 164: 614–619.

N Yumino D, et al. (2008). Central sleepapnoea and Cheyne–Stokes respiration.Proc Am Thorac Soc; 15: 226–236.


Treatment of hypoventilation/chronic respiratoryinsufficiencyAnita K. Simonds

Treatment algorithms

The management plan should include thesteps in box 1.

The considerations below relate to particularpathophysiological causes of nocturnalhypoventilation, which also influencetreatment choices.

Obesity hypoventilation syndrome

In the long term, weight loss is clearlycrucial, coupled with treatment ofunderlying SDB. In patients withpredominant OSA, CPAP is the treatment ofchoice. In those with nocturnalhypoventilation, CPAP may be effective inabout 50–80% of patients, but this ispredominantly in those with mildhypoventilation. Some algorithms use a

cut-off point of daytime PaCO2 .7.3 kPa or50 mmHg to indicate requirement fornocturnal NIV rather than CPAP. A variety ofapproaches have been applied (Masa et al.,2001; Piper et al., 2008). A randomisedcontrolled trial (RCT) by Piper et al. (2008)found that treatment choice of CPAP or NIVcould be best ascertained after an initialnight of CPAP to see whether CPAP aloneeffectively controlled AHI, SaO2 and PtcCO2

overnight; however, more severe patientswere excluded. Masa et al. (2001) found thatpatients with OHS treated with eitherpressure- or volume-cycled NIV hadimproved arterial blood gas tensions andthis gain was similar to that found inpatients with kyphoscoliosis treated withNIV. Marked hypercapnia during the dayand acute ventilatory decompensationfavour the use of NIV. In these patients withventilatory decompensation, NIV mayincrease hypercapnic ventilatory drive anddecrease leptin resistance or increase leptinlevels in those with low baseline levels.Janssens et al. (1998) found a 5-year survivalrate of 88% in OHS patients using nocturnalNIV, and in a further retrospective analysis2-year survival was 92% and 5-year survival70%. In contrast, there is a reported 40%mortality over 50¡25 months in OHSpatients who refused NIV. This outcomemay be influenced by other factors such aspoor adherence to other therapies. Bi-levelNIV titrated appropriately will controlhypoventilation, and expiratory positiveairway pressure (EPAP) should be titrated todeal with upper airway collapse andobstructive apnoeas/hyponoeas. EPAP mayalso recruit basal alveoli and improvefunctional residual capacity. Up to 50% ofOHS patients may require supplemental

Key points

N Symptomatic nocturnalhypoventilation is usually the triggerfor elective initiation of NIV.

N OHS patients with mild nocturnalhypoventilation and/or OSA can bemanaged with CPAP; those withsevere nocturnal hypoventilationwarrant NIV.

N Overnight monitoring of simplemeasures of oximetry andtranscutaneous CO2 tension plusventilator download data can beused to assess the adequacy ofventilatory support.


oxygen therapy and this should be entrainedinto ventilator circuit to increase SaO2 to.90%. It is often possible to reduce ordiscontinue oxygen therapy and continue onNIV alone following acute ventilator failureor as the patient loses weight. A suggestedalgorithm is given in figure 1.

Bariatric surgery is appropriate in morbidlyobese patients, and guidelines exist foreligibility for weight-reducing surgery insome countries. Patients should use CPAPor NIV in the peri-operative period andcareful liaison should be carried out with thesurgical and anaesthetic team. There is noevidence that use of positive airway pressuretherapy in the post-operative periodincreases the risk of surgical dehiscence.Weight following bariatric surgery isassociated with improvements in thecomplications of obesity such as type IIdiabetes and metabolic syndrome.Corrections of obstructive events overnightare seen, associated with improvement indiurnal ventilation. Weight loss achievedusing high-intensity diets can improve OSA,but there is less evidence in patients withnocturnal hypoventilation

Neuromuscular/skeletal disorders

Broadly speaking, neuromuscular and chestwall disorders can be categorised as static orslowly progressive (e.g. congenital musculardystrophies), or markedly progressive (e.g.motor neurone disease). While manyconditions are genetic in aetiology, genetherapy has not yet reached clinical practiceand few diagnoses have specific therapies.

Box 1 Steps in the management plan

N Identification of high-risk cases

of nocturnal hypoventilation(see table 1)

N Regular evaluation with sleep studies

including assessment of PaCO2

N Treatment of reversible causes

e.g. weight reduction for obesity

N Initiation of ventilatory support in

neuromuscular and chest wall diseasepatients at the development ofsymptomatic nocturnalhypoventilation

N Supplemental oxygen therapy if SaO2

, 90% despite optimum controlof PaCO2

N Follow-up to assess adequacy of

ventilatory control

Table 1 Risk of ventilatory decompensation.

Risk Disorder

High Spinal muscular atrophy type I (in infancy)MND/ALSDuchenne muscular dystrophy (age .12 years)High spinal cord lesions (above C5)Early onset nemaline myopathy (in infancy)X-linked centronuclear/myotubular myopathy (in infancy)Merosin deficient congenital muscular dystrophy (in early childhood)Congenital scoliosis associated with ostegenesis imperfecta, neurofibromatosis

Medium Morbid obesityLimb girdle muscular dystrophy variantsEarly-onset scoliosis (at age ,5 years) with vital capacity ,50% predicted andproximal curvatureSpinal muscular atrophy type IIOverlap syndrome (COPD plus OSA) with FEV1 ,1.5 L

Low Spinal muscular atrophy type IIIEmphysema patients on LTOT (unless end-stage disease)Interstital lung disease (unless end-stage disease)


The exceptions to this include medicaltherapies for congenital and acquiredmyasthenia and immunosuppressivetherapy for polymyositis.

NIV is the treatment of choice for nocturnalhypoventilation and in patients withscoliosis shows preferable results tolong-term oxygen therapy (LTOT). An RCT

and case series (Ward et al., 2005; Mellieset al., 2003) show that NIV should beintroduced at the time of nocturnalhypoventilation before the development ofdaytime hypercapnia. Large cohort studiesshow that NIV improves survival, improvesarterial blood gas tensions during the day,controls nocturnal hypoventilation andimproves quality of life in conditions such

Hypercapnia in an obese subject

pH

>7.35<7.35

No

Severity criteria(enoephalopathy, shockor NIV contraindication)

Yes

Endotrachealintubation

NIV(acute setting)

Endotrachealintubation

NIV(acute setting)

>50 mmHg >50 mmHg

NIPPV± O2

CPAP trial

NocturnalSaO2

(or PG/PSG)

Remainedhypercapnic

Continuousdesaturation

No YesSaO2 dips (or obstructive events if PG)

Persistedelevated PaCO2

IncreaseEPAP

NIV± O2

PaCO2

PG/PSG(1)

OSANo OSA

ContinueCPAP

Switch to NPPV

IncreaseIPAP

O2 addition(or increase)

Yes No

FIGURE 1. Management of nocturnal hypoventilation in obese patients. PG: respiratory polygraphy.Reproduced from Veale (2008), with permission from the publisher.


as Duchenne muscular dystrophy and spinalmuscular atrophy. It would be difficultethically to carry out RCTs in these groups.In an RCT of NIV in motor neurone disease/

amyotrophic lateral sclerosis (MND/ALS)(Bourke et al., 2006), NIV produced asurvival advantage of approximately7 months and improved quality of life. In asubgroup with severe bulbar MND/ALS,mortality did not decrease, but there weregains in quality-of-life measures associatedwith SDB. This raises the valid point thatNIV can be used to palliate symptoms ofnocturnal hypoventilation, with the sole aimof symptom control, rather than to extendlife in progressive conditions. This approachhas also been used in infants with nocturnalhypoventilation secondary to severe type Ispinal muscular atrophy (Chatwin et al.,2011). Importantly, each case should beindividually assessed with the purpose ofsetting measurable treatment goals forNIV. A suggested respiratory managementplan for MND/ALS patients is shown infigure 2.

Tracheostomy ventilation is indicated inpatients with severe bulbar disease,recurrent aspiration pneumonia, 24-hourventilator dependency (although somepatients prefer to use 24-hour NIV), if NIVfails to control nocturnal hypoventilation, orthere are difficult to control secretions orupper airway/interface problems.

Respiratory symptoms, SDB, orthopnoea, FVC <70%, PaCO2 >6 kPa

Counselling, anticipatory plan of ventilatory support at diagnosis

Ventilatory supportdeclined

Ceiling of non-invasive support (patient choice)

Trial of NIV/coughassist

Daytime NIV plus cough assistance ± PEG

Withdrawal ofventilation

Disease progression

Disease progression

Disease progression

Tracheostomy - IPPV

Palliative care

FIGURE 2. Suggested respiratory management algorithm for motor neurone disease/ALS. Reproducedfrom Simonds (2007), with permission from the publisher.

Box 2. Indications for initiating nocturnal NIV in COPD.

N In patients with recurrent acute

exacerbations responding to NIV(.2 per year) and PaCO2 .50 mmHgbut ,54 mmHg

N In those with symptomatic

hypoventilation and diurnal PaCO2

.55 mmHg

N Or in those with PaCO2 .50 mmHg

but ,54 mmHg with nocturnaldesaturation ,88% for o5 min whilereceiving oxygen therapy ato2 L?min-1


SpO2

SpO2

THO

ABD

FLO

PRE

SpO2

THO

REM sleep Awake

ABD

FLO

PRE

(i) Leaks (ii) Insufficient pressure support

(iii) Leaks

06:0005:3005:0004:3004:0003:3003:0002:30

02:30 04:0002:00 03:3001:30 03:0000:30 01:00

10090807060

(ii) Residual obstructive events

(i) Decrease of ventilatory command

SpO2SpO2

THO

ABD

FLO

PRE

THO

ABD

FLO

PTT

PRE

SpO2

SpO2

THO

ABD

FLO

PTT

PRE

100

90

85

898990 919294

94 96 9107 840711

Figure 3. Evaluating adequacy of ventilatory support and problem-solving. Oximetry traces from a patienton overnight NIV. (a) shows recurrent episodic dips in SpO2. (b) shows persistent periods of desaturation.The further data below each example show thoracoabdominal activity, flow and ventilator pressuredemonstrating, how the underlying cause of the events can be determined, e.g. leaks, inadequate pressure,upper airway obstruction or desynchrony with ventilator. Reproduced from Janssens et al. (2011), withpermission from the publisher.


Ventilatory support in CCHS is consideredin Module 8, ‘‘Management of SDBin children’’.

Nocturnal hypoventilation in obstructivelung disease

The majority of COPD patients who fulfilcriteria for LTOT tolerate oxygen therapy atnight and do not develop significanthyercapnia during sleep.

However significant nocturnalhypoventilation occurs in a proportion ofpeople with severe COPD and will inevitablybe present in those with daytimehypercapnia. While NIV is strongly evidence-based therapy for acute acidotic hyper-exacerbations of COPD, the only guidelinesfor use of nocturnal NIV in stable COPDpatients were published over a decade ago(Consensus Conference, 1999) and basedon consensus conferencerecommendations. There have been severalRCTs of NIV versus LTOT plus NIV in stablehypercapnic COPD patients with nocturnalhypoventilation. The most recent showedincreased survival with NIV, but health-related quality-of-life measures were notenhanced (McEvoy et al., 2009). Further trialsare due to report soon. Pragmatically, someteams now use the indications shown inbox 2 for initiation of nocturnal NIV in COPD.

In cystic fibrosis and bronchiectasis, airwayand parenchymal lung disease causepredominant ventilation/perfusion gas-exchange problems and alveolarhypoventilation causing nocturnal SDB arefairly late-stage phenomena. As indicated inmodule 8, worsening hypoventilation can beseen in REM sleep due to reduction inintercostal and postural muscle tone,increased work of breathing and decreasedventilatory reponses to hypercapnia andhypoxia. In patients with markedhypercapnia, NIV may reduce symptoms,reduce failure to thrive and result in fewerinfective exacerbations. In two RCTs,improved gas-exchange was seen, with noeffect on lung function. NIV has beensuccessfully used to bridge end-stage cysticfibrosis patients with chronic ventilatoryfailure to transplantation.

Assessing adequacy of control of nocturnalhypoventilation and follow-up

Bi-level positive pressure ventilation is themost commonly used form of ventilatorysupport, although volume-preset ventilatorshave a long track record in neuromuscularpatients. Set-up of the ventilator usingclinical guidance has been shown to be aseffective as detailed set-up using complexphysiological measures such asoesophageal pressure (Fauroux et al., 2004).Volume-assured modes such as AVAPS havenot been shown to produce major long-termadvantages over standard bi-level pressuresupport in OHS (Storre et al., 2006; Murphyet al., 2012), and produced equivalentcontrol of nocturnal ventilation inneuromuscular patients (Jaye et al., 2009).Monitoring of overnight oximetry andtranscutaneous CO2 tension can be used toassess control of hypoventilation. Janssenset al. (2011) have demonstrated how thecombination of this information with datadownloaded from the ventilator can be usedto solve common ventilator-relatedproblems (figure 3).

Further reading

N Bourke SC, et al. (2006). Effects of non-invasive ventilation on survival and qual-ity of life in patients with amyotrophiclateral sclerosis: a randomised controlledtrial. Lancet Neurol; 5: 140–147.

N Chatwin M, et al. (2011). Outcome ofgoal-directed non-invasive ventilation andmechanical insufflation/exsufflation inspinal muscular atrophy type I. Arch DisChild; 96: 426–432.

N Consensus Conference. (1999). Clinicalindications for noninvasive positive pres-sure ventilation in chronic respiratoryfailure due to restrictive lung disease,COPD, and nocturnal hypoventilation – aconsensus conference report. Chest; 116:521–534.

N Fauroux B, et al. (2004). Setting ofnoninvasive pressure support in youngpatients with cystic fibrosis. Eur Respir J;24: 624–630.


N Janssens JP, et al. (2008). NIV andchronic respiratory failure secondary toobesity. Eur Respir Monogr; 41: 251–264.

N Janssens JP, et al. (2011). Nocturnalmonitoring of home non-invasive ventila-tion: the contribution of simple toolssuch as pulse oximetry, capnography,built-in ventilator software and auto-nomic markers of sleep fragmentation.Thorax; 66: 438–455.

N Jaye J, et al. (2009). Autotitrating versusstandard noninvasive ventilation: a ran-domised crossover trial. Eur Respir J; 33:566–573.

N Masa JF, et al. (2001). The obesityhypoventilation syndrome can be treatedwith noninvasive mechanical ventilation.Chest; 119: 1102–1107.

N McEvoy RD, et al. (2009). Nocturnal non-invasive ventilation in stable COPD: arandomised controlled trial. Thorax; 64:561–566.

N Mellies U, et al. (2003). Long-termnoninvasive ventilation in children andadolescents with neuromuscular disor-ders. Eur Respir J; 22: 631–636.

N Murphy PB, et al. (2012). Volume targetedversus pressure support non-invasive

ventilation in patients with super-obesityand chronic respiratory failure: a rando-mised controlled trial. Thorax; [Epubahead of print DOI: 10.1136/thoraxjnl-2011-201081].

N Piper AJ, et al. (2008). Randomised trialof CPAP vs bilevel support in the treat-ment of obesity hypoventilation syndromewithout severe nocturnal destauration.Thorax; 63: 395–401.

N Simonds AK. (2007). Home ventilation inprogressive disorders, quadriplegia andpalliative non-invasive ventilation. In:Simonds AK, ed. Non-invasive RespiratorySupport: a practical handbook. London,Hodder Arnold; pp. 193–207.

N Storre JH, et al. (2006). Average volume-assured pressure support ventilation inobesity hypoventilation. A randomisedcrossover trial. Chest; 130: 815–821.

N Veale D. (2008). Respiratory complica-tions of obesity. Breathe; 4: 210–223.

N Ward SA, et al. (2005). Randomisedcontrolled trial of non-invasive ventilation(NIV) for nocturnal hypoventilation inneuromuscular and chest wall diseasepatients with daytime normocapnia.Thorax; 60: 1019–1024.


Treatment of nonrespiratorysleep disorders

Dirk Pevernagie

Because of their diverse nature, there is nogeneric approach to the treatment ofnonrespiratory sleep disorders. Variousmedical and nonmedical conditions areimplicated in disturbed sleep. The

International Classification of SleepDisorders 2nd Edition (ICSD-2) lists .70nonrespiratory nosological items in varioussections, including insomnia, hypersomniaof central origin, circadian rhythm sleepdisorders, parasomnias, sleep-relatedmovement disorders and other categories ofimpaired sleep. Treatment of thisheterogeneous group of sleep disorders isbased on nonpharmacological approachesas well as the prescription of agentsbelonging to particular drug classes. Sincetherapy must be tailored to the individualpatient’s needs, choices between differentpharmacological and nonpharmacologicaltreatment options, or combinations ofthese, are to be made. Pharmacologicaltherapy requires adjustment of the choice,dose and regimen of drugs in keeping withthe treatment response. Finally, patients maysuffer from a combination of sleep disorders.Concurrent sleep disorders call for anintegrated therapeutic approach based oncombining different treatment modalities.

Proficient management of sleep disorders isbased on detailed knowledge of sleep and itsdisturbances. The scope of this chapter isnot to elaborate on the intricacies of theentire spectrum of pathological sleep, asthis would require a thorough review of allthe nosological items listed in the ICSD-2.The aim is, rather, to provide a generaloverview of treatment principles innonrespiratory sleep medicine. Therespiratory physician should have a broadknowledge of the available treatmentoptions and strategies, but is not compelledto take the lead in the management ofnonrespiratory sleep problems.Nevertheless, the respiratory physician must

Key points

N Due to the heterogeneous nature ofthe various sleep disorders, nogeneric approach to treatment can beproposed. In each of the majornosological categories,pharmacological andnonpharmacological treatmentmodalities are available.

N The mainstay of treatment ofinsomnia is nonpharmacological. Astepwise treatment plan is proposed.Treatment with hypnotic drugs isappropriate as a primary step insituational insomnia. In chronicinsomnia, hypnotics may be added asadjuvant treatment to cognitivebehavioural therapy.

N The mainstay of treatment ofhypersomnia of central origin is basedon prescription of stimulant drugs.Other medications are available forcataplexy, one of the core symptomsof narcolepsy.

N RLS and PLMD are responsive totreatment with medications belongingto various pharmacological classes.Currently, DA agents are the first-lineof treatment. Opioids may be used insevere cases of RLS.


be able to handle comorbid issues (e.g. sleepapnoea complicated by insomnia). Treatmentof sleep patients requires a multidisciplinarysetting. In sleep medicine, diagnostic andtherapeutic management is founded oncooperation between professionals with adifferent specialty background. As respiratorydisturbances are among the most prevalentsleep disorders, there is a prominent role forthe respiratory physician in thismultidisciplinary process. This role, however,implies appropriate knowledge of the relevantrespiratory and nonrespiratory aspects ofsleep and its disorders.

Rather than split this chapter into sectionson pharmacological andnonpharmacological therapy, bothtreatment aspects will be discussed underfour main themes, namely insomnia,hypersomnia of central origin, sleep-relatedmovement disorders and miscellaneousconditions. The content and relative role ofeither treatment option will be discussed ineach of these categories. Where appropriate,the special interest for the respiratoryphysician will be indicated.

Insomnia

In line with the many components that playa role in the causation and maintenance ofinsomnia, several treatment methods haveemerged. The various factors that play a rolein the pathogenesis of insomnia arereviewed in chapter 3. The usefulness ofdifferent therapeutic modalities depends onthe contributing factors and the time courseof the illness. Whether insomnia is acute orchronic and whether insomnia is associatedwith medical or psychiatric comorbiditieswill influence the management strategy. Thetreatment may be general in scope ortargeted to specific adverse factors, such associoprofessional stressors, environmentaltriggers, maladaptive behaviour orunderlying disease. Prescription ofhypnotics is common and accepted practicefor acute insomnia. For chronic insomnia,the emphasis is on a nonpharmacologicalapproach. The use of hypnotic drugs isadjuvant in this setting. Diagnosis andtreatment of underlying causes is pertinentand should be considered whenever

appropriate. The treatment goals are toimprove duration and quality of sleep,and to improve daytime dysfunction dueto insomnia.

Nonpharmacological treatmentNonpharmacological treatment of insomniacomprises a stepped care approach. Theincremental steps include generalrecommendations, specific (tailored)recommendations, and structuredpsychological and behavioural interventions(e.g. cognitive behavioural therapy (CBT)).Furthermore, the level of support may bescaled up progressively, starting with self-support, then group therapy at the nursinglevel and individual treatment administeredby specifically trained professionals(medical psychologists).

At all times, irrespective of whetherinsomnia is primary, secondary orcomorbid, the patient should be providedwith proper general information. Explainingthe nature of insomnia and the principles ofits management is a starting point.Recommendations on good sleep hygieneare a default step in the treatment plan(table 1). Adherence to these instructionsoften results in better coping with theproblem, and in some patients, no additionaltherapeutic measures are required.

For unremitting chronic insomnia,psychological and behavioural approachesare the treatment of choice. Some of the basicmethods are briefly explained here; for moredetailed information, see Further Reading.

Stimulus control is a structural behaviouralapproach aimed at turning the negativefeelings associated with going to bed into apositive anticipation. Patients are taught tobreak with habits that are adverse to sleep.The sleeping environment is re-establishedas a stimulus for good sleep.

Sleep restriction is based on curtailing timein bed. Opportunities for sleep such asnapping are limited. This intervention firstinduces additional sleep deprivation. Thedrive to sleep will thus increase and,eventually, sleep will be more consolidated.When sleep continuity improves, time in bedis progressively extended. Ultimately, there


should be sufficient sleep time for thepatient to feel rested during the day.

As insomnia is associated with increasedsomatised tension, relaxation training isindicated in some patients. Severaltechniques have been described, includingalternating tensing and relaxing of musclegroups, concentrating on abdominalbreathing, and guided imagery.

CBT for insomnia (CBT-I) is the collectivename for a combination of theaforementioned techniques with apsychotherapeutic method to restoreappropriate cognitive pathways. The aim isto identify disbeliefs about sleep and to turnthese cognitions into positive and realisticconcepts. There is ample scientific evidencethat CBT-I is beneficial, and that the positiveeffects on sleep quality are durable.

Pharmacological treatment In contrast tononpharmacological treatment, there are noestablished guidelines on pharmacotherapyfor chronic insomnia. While effectiveness ofhypnotic medications has been confirmed inrandomised controlled trials (RCTs), none ofthese studies has been designed to assessthe superiority of one medication overanother in terms of efficacy and safety.Neither is there any systematic studycomparing hypnotics withnonpharmacological interventions.

Hypnotic drugs are the first-line treatment inthe acute setting, i.e. in patients with

situational insomnia. For this indication,hypnotics are prescribed to be taken nightlyfor a couple of weeks. Some caveats must beborne in mind when hypnotics areconsidered for treatment of chronicinsomnia. While successful in the shortterm, the effect of hypnotic drugs may wearoff over time. As habituation takes effect,incremental doses may be necessary to keepup the initial pharmacotherapeutic results.Therefore, hypnotics carry an intrinsic risk oftolerance, and physical and psychologicaldependence. The latter problem may occurespecially in patients with a history of drugor alcohol abuse. Increased sleeplessness isa frequent complaint following abruptdiscontinuation of hypnotic treatment. Thiscondition, known as rebound insomnia, maybe accompanied by other symptoms ofwithdrawal, e.g. anxiety and agitation. Thesedrawbacks are the prime reasons whyhypnotics may not be suitable for long-termuse. One way to circumvent the problem oftolerance is to prescribe hypnotics forintermittent use, and to restrict intake tothree times per week, at maximum.

Many hypnotics, especially those with longelimination half-lives, may have a carry-overeffect and cause unwanted sedation in themorning. Long-acting hypnotics should notbe prescribed in patients who drive motorvehicles. Short-acting drugs avoid thiscomplication but may be inefficacious atcontrolling sleep maintenance insomnia.There are several accounts of inappropriate

Table 1. Principles of sleep hygiene: recommendations for the insomnia patient.

N Limit time in bed for sleeping; lying awake in bed weakens sleep

N Get out of bed when you are not feeling sleepy

N Improve environmental conditions of the bedroom (noise, light and temperature)

N Take a light snack before going to bed but avoid consuming large beverages

N Avoid too much alcohol and stimulants (e.g. nicotine and caffeine)

N Schedule stressful activities so that they occur a long time before bedtime; unwind and

relax in the hours prior to going to sleep

N Do not check your alarm clock at night

N Restrict the use of sleeping pills


behaviour in individuals who were awakenedat night after taking sleeping pills. In theelderly, confusional arousals and falls areserious side-effects. Inability to rapidlyregain consciousness upon awakening is anintrinsic disadvantage of hypnotictreatment. Moreover, hypnotics may causeanterograde amnesia, rendering someindividuals unable to recall their nocturnalbehaviour. There are concerns over impairedcognition and a tendency towardsdeveloping depression with chronic use.

Some hypnosedative drugs inhibitrespiratory neurons in the brainstem andthereby suppress the drive to breathe. Thisis especially relevant in OSA patients, inwhom hypnotics increase the incidence andduration of disturbed breathing events.Unless these patients are adequately treatedwith CPAP, the use of hypnotics (particularlybenzodiazepines (BZDs)) iscontraindicated. Finally, hypnotics shouldnot be used in pregnancy, during lactationor by patients with hepatic failure.

As a rule of thumb, in insomnia, the lowesteffective dose of a single hypnotic drugshould be prescribed for the shortestamount of time. Indeed, the primaryintention of prescribing hypnotic medicationis to achieve symptomatic control. Incontrast with nonpharmacological treatmentmodalities, drug treatment fails to addressthe causative mechanisms of insomnia,including precipitating factors, dysfunctionalbeliefs and maladaptive behaviour patternsthat tend to sustain the insomnia process.Therefore, hypnotic drug prescription is tobe considered an adjuvant treatment toother therapeutic interventions. In allinstances where hypnotics are prescribed forlong-term use, the option to taper anddiscontinue the sleep medication should bereassessed on a regular basis.

Hypnotic agents have several effects onsleep structure. Most often, there is adecrease of sleep latency (SL) and numberof nocturnal awakenings (#A) versus anincrease in total sleep time (TST) and sleepefficiency (SEF). Often, stage 2 NREM sleep(N2) is increased. Effects on deep sleep

(N3) and REM sleep may be variable anddepend on the class of drugs selected.

When choosing a specific drug among thedifferent available medications, theprescriber should take into account the keycharacteristics of that particular agent,including drug class, mode of action, andpharmacokinetic and pharmacodynamicproperties. In this chapter, hypnotics areclassified into two general categories: BZDreceptor ligands and non-BZD sedative drugs.

BZDs are a group of compounds that sharea common structure of a benzene ring fusedwith a diazepine ring. These drugs bind to aspecific locus of the c-aminobutyric acid(GABAA) receptor complex, called the BZDreceptor. Activation of the BZD receptorsincreases the effect of GABA-ergicneurotransmission. BZDs have anxiolytic,hypnosedative, anticonvulsant, amnesic andmuscle-relaxant properties. They may beprescribed for various medical andpsychiatric indications. While a sedatingeffect is a prominent and common feature,not all BZDs are labelled for the specifictreatment of insomnia. Based on theirpharmacokinetic profiles, BZDs are dividedinto three categories pertaining to durationof action: long-, intermediate- and short-acting (table 2). BZDs decrease SL and #A,increase TST, SEF and N2, and tend todecrease N3 and REM. In general, BZDs aresafe and efficacious in the treatment ofsituational insomnia. When prescribedcautiously and restrictively, BZDs also havea place in the treatment of chronic insomnia(e.g. on an intermittent prescription basis).

BZD receptor ligands that lack a BZDchemical structure have been introduced inthe last two decades. These agents arecalled non-BZD BZD receptor agonists(NBBRAs) or ‘Z drugs’, because theirgeneric names begin with a ‘Z’ (zolpidem,zopiclone and zaleplon; table 2). They are,to date, the most extensively studiedhypnotics. They have a similar effect onsleep architecture as the classical BZDs. Thecombination of efficacy with reasonablesafety promotes this class of drugs as aplausible first-line pharmacotherapy forinsomnia. However, despite acceptable


Tabl

e2.

Hyp

notic

s.

Co

mp

ou

nd

Cla

ssM

od

eo

fac

tio

nH

ypn

oti

cd

ose

mg

tmax

min

t1/2

hD

ura

tio

no

fac

tio

nM

etab

olis

mS

ide-

effe

cts

Am

itri

pty

line

TC

AIn

hib

itio

no

fN

Aan

d5H

Tre

up

take

;a

1,M

1an

dH

1re

cep

tor

anta

go

nis

m

25–

100

.12

010

–28

Lo

ng

Hep

atic

;ac

tive

met

abo

lite(

s)C

ard

iac

arrh

yth

mia

s;an

tich

olin

erg

ican

dan

tih

ista

min

icef

fect

s;ag

gra

vati

on

of

RLS

Flu

raze

pam

BZ

DG

AB

A-e

rgic

15–

304

7–10

0Lo

ng

Hep

atic

;ac

tive

met

abo

lite(

s)G

ener

icsi

de-

effe

cts

of

BZ

D

Gab

apen

tin

Str

uct

ura

lan

alo

gu

eo

fG

AB

AP

rob

ably

GA

BA

-erg

ic30

0–

60

018

05–

9In

term

edia

teR

enal

clea

ran

ceR

esid

ual

effe

cts

inth

em

orn

ing

;d

izzi

nes

s;at

axia

;h

ead

ach

e;ra

rely

leu

kop

enia

Lo

rmet

azep

amB

ZD

GA

BA

-erg

ic1–

212

012

Inte

rmed

iate

Hep

atic

;in

acti

vem

etab

olit

e(s)

Gen

eric

sid

e-ef

fect

so

fB

ZD

Mid

azo

lam

BZ

DG

AB

A-e

rgic

7.5–

1530

–9

02–

3S

ho

rtH

epat

ic;

acti

vem

etab

olit

e(s)

Gen

eric

sid

e-ef

fect

so

fB

ZD

Mir

taza

pin

eN

ora

dre

ner

gic

and

spec

ific

sero

ton

iner

gic

anti

dep

ress

ant

a1,a

2,

5HT

1,5H

T2

and

H1

rece

pto

ran

tag

on

ism

15–

306

0–

190

13–

40

Lo

ng

Hep

atic

;ac

tive

met

abo

lite(

s)R

esid

ual

effe

cts

inth

em

orn

ing

;w

eig

ht

gai

n;

agg

rava

tio

no

fR

LS

Nit

raze

pam

BZ

DG

AB

A-e

rgic

5–10

120

30–

40

Lo

ng

Hep

atic

;in

acti

vem

etab

olit

e(s)

Gen

eric

sid

e-ef

fect

so

fB

ZD

Ram

elte

on

Mel

ato

nin

Ian

dII

rece

pto

rag

on

ist

Un

kno

wn

84

51–

3S

ho

rtH

epat

ic;

acti

vem

etab

olit

e(s)

No

nsp

ecif

ic

Qu

etia

pin

eD

iben

zoth

iaze

pin

ean

tip

sych

oti

ca

1,M

1,H

1,5H

T2

and

D2

rece

pto

ran

tag

on

ism

25–

506

0–

120

7In

term

edia

teH

epat

ic;

inac

tive

met

abo

lite(

s)S

edat

ion

;ag

gra

vati

on

of

RLS

Tem

azep

amB

ZD

GA

BA

-erg

ic10

–30

507–

11In

term

edia

teH

epat

ic;

inac

tive

met

abo

lite(

s)G

ener

icsi

de-

effe

cts

of

BZ

D

Tra

zod

on

eH

CA

5HT

2an

da

1re

cep

tor

anta

go

nis

m50

–10

06

0–

120

5–9

Lo

ng

Hep

atic

;ac

tive

met

abo

lite(

s)R

esid

ual

effe

cts

inth

em

orn

ing

;ca

rdia

car

rhyt

hm

ias;

ort

ho

stat

ich

ypo

ten

sio

n;

rare

lyp

riap

ism

(no

td

ose

rela

ted

)

Tri

azo

lam

BZ

DG

AB

A-e

rgic

0.1

25–

0.2

5012

02–

6S

ho

rtH

epat

ic;

inac

tive

met

abo

lite(

s)G

ener

icsi

de-

effe

cts

of

BZ

D


overall safety, anterograde amnesia,confusional nocturnal behaviour anddependency have been reported with someNBBRAs in some susceptible patients. TheNBBRAs are less likely than the usual BZDsto cause rebound phenomena afterdiscontinuation. These advantages arecounterbalanced by the higher cost oftreatment in comparison with the older BZDs.

Several psychotropic drugs outside the classof BZD receptor agonists have potentsedating side-effects. Among these agentsare antidepressants, neuroleptics,antiepileptics, melatonin receptor agonists,antihistaminics and nonprescriptionsubstances. Depending on the clinicalcontext, the sedative action may be adverseor beneficial. Sedation is an undesirableside-effect when treatment is aimed atmobilising and stimulating the patient.When given in nightly dose, sedation mayimprove sleep and may be helpful tomanage comorbid insomnia. Therefore, asingle psychotropic agent may serve a dualpurpose of treating the underlying disorderand controlling insomnia as a symptom atthe same time. However, the dosage neededto improve sleep is generally much lowerthan the dosage required to observe anytreatment effect on the disorder for whichthe medication is registered. In depression,for instance, a low-dose sedativeantidepressant at bedtime is not sufficientto improve depressive symptoms. Toachieve this goal, the low-dose hypnoticantidepressant should be combined withtherapeutic dosages of other antidepressantsadministered during the daytime. It isincreasingly common practice, however, toprescribe some of these agents for thesymptomatic treatment of chronic insomniairrespective of its cause. While there isinsufficient medical evidence to support theoff-label use of non-BZD psychotropic agents,empirical results provide no motive todiscourage this practice. Moreover, theseagents are not prone to tolerance to the sameextent as classical BZDs. Some of thecommonly used medications are describedbelow. Their pharmacological profiles areshown in table 2.Ta

ble

2.C

ontin

ued

Co

mp

ou

nd

Cla

ssM

od

eo

fac

tio

nH

ypn

oti

cd

ose

mg

tmax

min

t1/2

hD

ura

tio

no

fac

tio

nM

etab

olis

mS

ide-

effe

cts

Zal

eplo

nN

BB

RA

GA

BA

-erg

ic5–

1012

01

Sh

ort

Hep

atic

;in

acti

vem

etab

olit

e(s)

Gen

eric

sid

e-ef

fect

so

fB

ZD

Zo

lpid

emN

BB

RA

GA

BA

-erg

ic10

30–

180

2–3

Sh

ort

Hep

atic

;in

acti

vem

etab

olit

e(s)

Gen

eric

sid

e-ef

fect

so

fB

ZD

Zo

pic

lon

eN

BB

RA

GA

BA

-erg

ic7.

54

5–12

05–

7In

term

edia

teH

epat

ic;

inac

tive

met

abo

lite(

s)G

ener

icsi

de-

effe

cts

of

BZ

D

Do

seto

be

red

uce

din

the

eld

erly

and

tho

sew

ith

hep

atic

or

ren

alfa

ilure

;p

reg

nan

cyan

dla

ctat

ion

are

con

trai

nd

icat

ion

s.t m

ax:

tim

eto

max

imu

mco

nce

ntr

atio

no

fth

ed

rug

;t1

/2:

hal

f-lif

e;T

CA

:tr

icyc

lican

tid

epre

ssan

t;H

CA

:h

eter

ocy

clic

anti

dep

ress

ant;

NA

:n

ora

dre

nal

ine;

5HT

:5-

hyd

roxy

tryp

amin

e(s

ero

ton

in);

M:

mu

scar

inic

;H

:h

ista

min

e.


Trazodone is a heterocyclic antidepressant,with weak antidepressive but strongsedative effects. This agent increases TST,SEF and N3, and has variable effects onREM sleep. Trazodone has little potential fortolerance or dependence.

Amitriptyline is a classical tricyclicantidepressant with sleep-enhancingproperties. This agent increases TST, SEFand REM latency, and has variable effects onN3. Cardiac, antihistaminic andanticholinergic side-effects limit the use ofamitriptyline and other tricyclicantidepressants. In low doses, however,these side-effects are minimal, whereassleep-promoting effects are still present.

Mirtazapine is a noradrenergic and specificserotoninergic antidepressant. Effects onsleep are similar as those observed withtrazodone. Due to a substantialantihistaminic action, weight gain is acommon side-effect.

Ramelteon is a potent melatonin agonist.The hypnotic action is limited to decreasingSL and probably results from an effect onthe suprachiasmatic nucleus. This agent isnot commercialised in Europe.

In the class of neuroleptics (antipsychotics),among other agents, quetiapine is a drugthat can be used for treating insomniaoutside the context of psychotic and bipolardisorders. It is known to increase TST, SEF,N2 and N3, whereas SL and #A are decreased.

Gabapentin belongs to a class ofanticonvulsant medications. This agent alsohas hypnotic and analgesic properties.Effects on sleep include an increase in N3and a reduction of #A. In addition, beneficialeffects have been observed on restless legssyndrome, periodic limb movements(PLMs) and parasomnias of N2.

Other agents are available to treat insomnia.This group of miscellaneousnonprescription drugs includes melatonin,valerian, antihistaminics and chloralhydrate. Prescription of melatonin isdiscussed later. Use of valerian,antihistaminics and chloral hydrate is not

recommended and they are not furtherreviewed in this chapter.

Special interest for the respiratory physicianComorbid insomnia in OSA is a prevalentclinical problem that presents a particulartherapeutic challenge to the patient andphysician. The prescription of sleep-promoting drugs may worsen SDB and theapplication of nasal CPAP may worseninsomnia. The resolution of this enigmarequires a multidisciplinary approach.Depending on the relative severity of eitherof the components, treatment for insomniaand OSA may be initiated simultaneously orsequentially. In this category of patients,there is a prominent role fornonpharmacological treatment of insomniaand for intensive nursing support tooptimise adherence to CPAP treatment.Prescribing hypnotics without CPAPtreatment is considered unsafe.

Hypersomnia of central origin

Sleepiness is a natural phenomenon. It maybe a normal physiological manifestation (e.g.occurring at bedtime), or it may be induced incertain conditions (e.g. sleep deprivation) orin the context of certain diseases (e.g. sleepapnoea and narcolepsy). Excessive daytimesleepiness (EDS), also known ashypersomnia, is a term reserved for a clinicalcondition in which a subject experiencesrecurrent, abnormal lapses of alertness andhas a strong tendency to fall asleepinvoluntarily. This condition entails secondaryproblems due to impaired performance inprivate, professional and social areas of life.Lost productivity, increased risk of causingpersonal injury and public damage, andsignificantly decreased quality of life aresequelae that are regularly observed inpatients suffering from EDS. Hypersomnia isa common symptom in several sleepdisorders, including sleep disturbances dueto medical or psychiatric causes, sleepapnoea syndromes and hypersomnias ofcentral origin. In this chapter, the discussionof treatment of excessive sleepiness will belimited to the latter group of disorders.

Hypersomnias of central origin are listed inthe third section of the ICSD-2 and are


reviewed in chapter 3 of this handbook. It isa collective name for disorders in which EDSis a primary complaint that is notattributable to disturbed nocturnal sleep ora circadian rhythm misalignment. Othersleep disorders may co-exist, but must betreated appropriately prior to establishing adiagnosis in this category.

Narcolepsy is one of the main entities in thecategory of the central nervous system(CNS) hypersomnias. This disease is knownto result from the loss of hypocretin-producing neurons in the hypothalamus.Most often, narcolepsy is associated withparoxysmal episodes of muscle atonia,called cataplexy. Cataplexy is typicallyelicited by sudden emotion. Some variantsof narcolepsy lack this symptom. Cataplexyis a disabling manifestation that requiresspecial therapeutic attention.

In contrast with narcolepsy, thepathogenesis of other disorders in thisgroup is still unknown. Recurrenthypersomnias and idiopathic hypersomnia(IHS) are believed to be caused by atemporary or permanent dysfunction of thesleep/wake regulating mechanisms in theCNS. Behaviourally induced insufficientsleep syndrome and hypersomnias due tomedical causes constitute a miscellaneousgroup of conditions that are primarilyrelated to the effect of external factors onnocturnal sleep and, as a consequence, ondaytime alertness.

The principles of treating hypersomnia ofcentral origin encompass improvement ofdaytime alertness and the control ofassociated features such as cataplexy.Stimulants are the cornerstone in thetreatment of EDS. Nonstimulating drugs areused for the treatment of cataplexy. Thissymptom may fully remit with propertreatment. In contrast, it proves verydifficult, if not impossible, to restore anormal level of alertness in hypersomniapatients with the administration of thecurrently available stimulating agents.

Nonpharmacological treatment Behaviourallyinduced insufficient sleep is the mostpertinent differential diagnosis in

hypersomnia of central origin, especially inyoung people. It may prove difficult toestimate the extent of sleep deprivation inthese patients. Nevertheless, sleepextension (i.e. prolonging the nocturnalsleep period) should be recommendedwhenever this diagnosis is suspected. Witha successful outcome, the need for treatmentwith stimulating drugs may be obviated.

While nonpharmacological treatment fornarcolepsy and other CNS hypersomnias ispart of an integrative therapeutic approach,evidence of its effectiveness is limited.Patients are advised to adhere to regularbedtimes, to take care to get sufficient night-time sleep and to schedule ‘strategic’ napsduring the daytime, at least when shortbouts of sleep have a restorative effect. InIHS, for instance, naps are often notrefreshing and may even cause additionaldrowsiness. These patients should abstainfrom napping during the daytime. Neither isthere strong evidence to recommendbehavioural treatment for cataplexy. Patientsshould avoid circumstances that elicitstrong emotions, as cataplexy is precipitatedby these conditions. Avoiding situationalemotions will certainly reduce the number ofattacks, but the downside of this ‘remedy’ isthat patients tend to withdraw from sociallife. This is obviously an undesirableoutcome. Therefore, lifestyle issues seem tohave minimal impact and pharmacologicaltreatment is mandatory in most of the patients.

Pharmacological treatment Stimulatingagents are key to the treatment of EDSassociated with narcolepsy and IHS.Amphetamines and methylphenidateincrease the release and block the reuptakeof monoamines and, thus, produce centraland peripheral sympathomimetic effects. Atthe cognitive level, they increase alertnessand psychomotor activity. Modafinil is anatypical nootropic and stimulant agent thatis chemically different from theamphetamines and has a weaker wake-promoting effect. This substance has an asyet undisclosed mode of action. While RCTsare lacking regarding the use ofmethylphenidate and amphetamines innarcolepsy and IHS, available evidence


suggests that these agents are capable ofenhancing alertness to at least two-thirds ofnormal levels, whereas with modafinil, thiseffect is less pronounced. Side effects arehighly variable among patients and are dose-dependent. Methylphenidate andamphetamines are associated with classicalmonoaminergic side-effects, includinginsomnia, loss of appetite, tremor,irritability, headache, palpitations andelevated blood pressure. Within therecommended dose range, the risk oftolerance and dependence is low innarcolepsy and IHS, and there is no need toschedule ‘drug-free holidays’. In higher thanrecommended dose ranges, patients shouldbe monitored carefully with respect todeveloping mental problems andhypertension. Modafinil has been studiedmore extensively than the other stimulants.It has a good benefit-to-risk ratio with fewerand milder side-effects. Because of its safetyand long-acting profile, modafinil has beenrecommended by the American Academy ofSleep Medicine as first-line treatment fornarcolepsy. Therapeutic efficacy is usuallypreserved in the long term. The higher costof treatment, however, is an obviousdisadvantage. The aforementioned agentsare listed in table 3.

The only drug that is labelled for thetreatment of cataplexy is sodium oxybate(pharmaceutical name for c-hydroxybutyrate; GHB). It is administered intwo doses: at bedtime and once again after3–4 h of sleep. With chronic administration,sodium oxybate improves sleep continuityand reduces cataplectic attacks as well asEDS. Because sodium oxybate has powerfulCNS depressant effects, it should not beused in combination with alcohol orsedative drugs, and is contraindicated inSDB and respiratory failure. Elimination ofthis agent involves conversion to succinicsemialdehyde that is subsequentlytransformed to succinate. Followingmetabolisation of succinate in the Krebscycle, the end product is carbon dioxide. Themode of action is unknown. GHB belongs tothe class of hypnotics. It can be abused,however, as a recreational drug. GHB as anillicit compound is known for significant Ta

ble

3.St

imul

ants

.

Co

mp

ou

nd

Ind

icat

ion

Mo

de

of

acti

on

Do

sem

gP

oso

log

yd

ose

sp

erd

ay

tmax

ht1

/2h

Du

rati

on

of

acti

on

Met

abo

lism

Sid

e-ef

fect

s

Dex

tro

amp

het

amin

eE

DS

Incr

ease

sth

ere

leas

ean

din

hib

its

the

reu

pta

keo

fN

Aan

dd

op

amin

e

20–

60

2–3

37–

14In

term

edia

teH

epat

ican

dva

riab

lere

nal

clea

ran

ceG

ener

icsi

de-

effe

cts

of

psy

cho

stim

ula

nts

Met

hyl

ph

enid

ate

ED

SIn

crea

ses

the

rele

ase

and

inh

ibit

sth

ere

up

take

of

NA

and

do

pam

ine

10–

60

3–4

22–

4S

ho

rtH

epat

ic;

inac

tive

met

abo

lite(

s)G

ener

icsi

de-

effe

cts

of

psy

cho

stim

ula

nts

Mo

daf

inil

ED

SU

nkn

ow

n10

0–

40

01–

22–

310

–12

Lo

ng

Hep

atic

and

min

imal

ren

alcl

eara

nce

;in

acti

vem

etab

olit

e(s)

Hea

dac

he

mo

stfr

equ

entl

y;ra

rely

toxi

cep

ider

mal

nec

roly

sis

So

diu

mo

xyb

ate

Cat

aple

xyan

dE

DS

Un

kno

wn

4.5

–9

.02

0.5

–2

0.5

–1

Sh

ort

Hep

atic

Nau

sea,

hea

dac

he,

dro

wsi

nes

s,co

nfu

sio

nal

aro

usa

l,co

nst

ipat

ion

Do

seto

be

red

uce

din

the

eld

erly

and

tho

sew

ith

hep

atic

or

ren

alfa

ilure

;p

reg

nan

cyan

dla

ctat

ion

are

con

trai

nd

icat

ion

s.t m

ax:

tim

eto

max

imu

mco

nce

ntr

atio

no

fth

ed

rug

;t1

/2:

hal

f-lif

e.


associated morbidity and mortality. Whilesodium oxybate should be prescribedcautiously and its use should be monitored,the abuse potential in narcolepsy seems tobe low.

Tricyclic antidepressants and selectiveserotonin reuptake inhibitors are effective(but prescribed off label) in the treatment ofcataplexy and other manifestations ofnarcolepsy, including sleep paralysis andhypnagogic hallucinations. Inhibition ofnoradrenaline (NA) reuptake presumablymediates the effect of antidepressants oncataplexy. At present, there is no evidence toaccept that particular antidepressants aremore powerful than others in controllingcataplexy, despite different propertiesregarding NA reuptake inhibition. A reviewof antidepressants prescribed for cataplexyis outside the scope of this chapter.

A disturbance of maintaining sleep is afrequent complaint in narcoleptic patients.Sleep disruption in narcolepsy may requiretreatment with hypnotic drugs. Whencataplexy is also present, sodium oxybate isthe treatment of choice.

Special interest for the respiratory physicianResidual hypersomnia is a problem that isencountered in 5–10% of well-treated OSApatients. Hypersomnias of central origin, asdescribed above, should be considered inthe differential diagnosis. If this conditionremains unexplained after appropriateclinical investigation, symptomatictreatment with stimulants should beconsidered. In the USA and some Europeancountries, modafinil is approved for treatingOSA patients with persistent EDS despiteadequate CPAP treatment. However, carefulmonitoring is required and modafinil use isnot recommended in individuals withuncontrolled hypertension, ischaemic heartdisease or epilepsy.

Sleep-related movement disorders

The most prominent nosological items in theICSD-2 section on sleep-related movementdisorders are restless legs syndrome (RLS)and PLM disorder (PLMD). While theseentities often co-occur, it is uncertainwhether they share a common

pathophysiological background. RLS is aneurological disorder with sensory and motormanifestations. The diagnosis is based on atetrad of symptoms: there is an irresistibleurge to move the legs due to an unpleasant‘creepy’ sensation, this urge increases uponresting, movement temporarily relieves theunpleasant sensation and these symptomsare worse in the evening and in the initial partof the night. RLS is relevant for sleep, as itmay be associated with difficulties initiatingsleep (DIS) or difficulties maintaining sleep(DMS). PLMD is associated with recurrentstereotyped leg movements that may inducearousals and, thus, disrupt sleep. Treatmentof PLMD is largely empirical, as well-designed RCTs are lacking. At present, it isnot clear whether PLMD outside the contextof RLS is clinically relevant, and whether thereis an indication to treat. Neither are there anyguidelines with respect to choosing theprimary outcomes that should guide thetreatment strategy. In contrast, the treatmentof RLS is well documented, as many RCTs arenow available. Guidelines for the treatmentand follow-up of RLS have been published.Therefore, the therapeutic options that followdeal mainly with the management of RLS.

Nonpharmacological treatment RLS patientsshould avoid smoking and consuming foodand beverages that contain CNS-stimulatingsubstances, such as alcohol, caffeine andchocolate. Some patients may benefit fromaerobic exercise, manipulating the legs, orapplying warm or cold water to the legs. Inaddition, good sleep hygiene should bemaintained to avoid the development ofinsomnia as a comorbid factor. Thesemeasures are often insufficient and cannotreplace pharmacological treatment inmoderate-to-severe RLS.

Pharmacological treatment Several classes ofdrugs are available for the treatment of RLS,including mineral supplements, BZDreceptor agonists, dopaminergic (DA)agents, opiate agonists and anticonvulsants.Only DA agonists have been the subject ofextensive RCTs.

RLS may be a clinical manifestation of irondeficiency. When serum ferritin levels are,50 mg?L-1, the patient should be checked


for concealed loss of blood and ironsupplementation should be commenced.Magnesium sulfate can temporarily relievesymptoms of RLS, but there are no data onlong-term benefits.

Clonazepam, a long-acting BZD, is cited inthe older literature as effective for treatingRLS. There are no convincing studies toshow that clonazepam and other BZDssufficiently control RLS symptoms. BZDsshorten sleep onset and suppress thearousals resulting from movements ratherthan the movements themselves. Symptomsof discomfort as well as PLMs usuallypersist despite the use of these agents.Therefore, BZDs are not considered first-linetreatment of RLS.

DA agents are the best studied and mostsuccessful drugs for treatment of RLS (andPLMD). Levodopa improves the symptomsof RLS and reduces the number of PLMs.The use of this medication is limitedbecause of rapid emergence of worseningsymptoms in the hours prior toadministration (augmentation; see later).Dopamine agonists are now considered thefirst-line treatment for RLS. The older ergotmedications (e.g. pergolide) have beenabandoned because of rare but seriousadverse effects, including the developmentof pleuropulmonary fibrosis or cardiacvalvulopathy.

Nonergot preparations are preferred andcurrently three dopamine agonists have alabel for RLS in this category. Pramipexole isa full dopamine agonist with high affinity forthe D3 receptor subtype. This agent is veryeffective in controlling RLS and PLM.Ropinirole has a similar pharmacodynamicand clinical profile. Rotigotine is a D3 and D2

dopamine agonist that has been developedfor transdermal administration. It may beused to treat night-time and daytimesymptoms of RLS. Mild skin reactions maybe seen at the application site of the patch.These compounds are highly effective andhave mild (mostly transient) side-effects,including nausea and drowsiness. Recentevidence suggests that these dopamineagonists are also safe and efficacious in thelong term. However, this optimistic

expectation needs confirmation byappropriately designed prospective RCTs.

Rarely, RLS patients on DA agents developdisturbances of impulse control, and showan increased desire for gambling and sexualinteractions. In up to one-third of thepatients, the use of DA agonists may belimited by augmentation. This dose-dependent phenomenon is characterised bythe unpleasant sensation and urge to moveoccurring earlier during the day, being moresevere and spreading to the upper limbs. Incases of severe augmentation, treatmentwith DA agonists should be discontinuedand agents from another class of drugsshould be prescribed temporarily.

Opioids have also proven therapeuticefficacy and durability, but are restricted tothe treatment of severe RLS. Patients whoare unresponsive to other medications orwho have developed augmentation arecandidates for opioid therapy. While there isa low risk of tolerance in RLS patients,opioids should not be prescribed toindividuals with a previous history of alcoholor substance abuse. These agents carry arisk for respiratory depression during sleepin susceptible patients.

Gabapentin improves RLS symptoms atdoses between 300 and 2,400 mg a day.This medicine is an alternative choice forpatients with mild symptoms or forreplacement of DA agents whenaugmentation becomes an issue. Moreover,gabapentin may be prescribed as an add-onwhen night-time symptoms are severe orpain is a prominent complaint.

Some frequently used agents from the fourdrug classes described here are presentedin table 4.

Special interest for the respiratory physicianWhen reading PSG recordings, it may provedifficult to distinguish hypopnoeas fromPLM events. Hypopnoeas are temporaryreductions in airflow terminated by a fewhyperpnoeic breaths. In PLMD, the limbmovements are often accompanied by atemporary increase in amplitude of tidalbreathing. The (normal) breathing betweenPLM events may be mistakenly identified as


hypopnoeas. Careful reading of thepolysomnogram is of primeimportance for making the correctdiagnosis of SDB versus PLMDand, therefore, for choosingappropriate treatment.

Miscellaneous sleep disorders

Parasomnias Parasomnias aresubdivided into disorders of NREMand REM sleep. NREM parasomniasare primarily disorders of arousal:patients awaken partially from deepsleep and demonstrate strangebehaviours before returning to sleep.Conservative measures are themainstay of managing thesedisorders. Patients should avoid astressful lifestyle and should securethe bedroom environment againstself-inflicted injury. Drug treatment isempirical. REM sleep parasomnias arecharacterised by vivid dreamexperiences or unusual dreamenactment (see section on REM sleepbehaviour disorder). Antidepressantsmay be indicated for these disorders,because of their REM sleep-suppressive properties. There is aspecial interest for the respiratoryphysician, because some parasomniamanifestations may be triggered bydisturbed breathing events. In thiscase, the underlying respiratorydisorder should be treated beforeconsidering other treatment options.

Delayed sleep phase syndrome Amongthe different circadian rhythm sleepdisorders, jet lag disorder, shift workdisorder and delayed sleep phasesyndrome (DSPS) are the mostprevalent clinical entities. DSPS is acondition in which the patients havedifficulties in falling asleep at normalbedtimes and waking up in themorning: the nocturnal sleep phase isshifted by o2 h relative toconventional or socially acceptedtimes. Because the patients tend tooversleep, significant sociofamilialdisharmony may ensue. Thetreatment is based on a combinationTa

ble

4.D

rugs

used

totr

eat

RLS

and

PLM

D.

Co

mp

ou

nd

Mo

de

of

acti

on

Do

sem

gD

ose

sp

erd

aytm

axh

t1/2

hM

etab

olis

mS

ide-

effe

cts

Clo

naz

epam

BZ

D0

.5–

21

1–4

30–

40

Hep

atic

;ac

tive

met

abo

lite(

s)G

ener

icsi

de-

effe

cts

of

BZ

Ds

Co

dei

ne

Op

ioid

rece

pto

rag

on

ist

10–

301

1–2

3–4

Hep

atic

;ac

tive

met

abo

lite(

s)G

ener

icsi

de-

effe

cts

of

op

ioid

s

Gab

apen

tin

An

tico

nvu

lsan

t30

0–

60

01

35–

9R

enal

clea

ran

ceR

esid

ual

effe

cts

inth

em

orn

ing

;d

izzi

nes

s;at

axia

;h

ead

ach

e;ra

rely

leu

kop

enia

Met

had

on

eO

pio

idre

cep

tor

ago

nis

t5–

40

11.

5–3

19–

55H

epat

ic;

inac

tive

met

abo

lite(

s)G

ener

icsi

de-

effe

cts

of

op

ioid

s

Pra

mip

exo

leD

2d

op

amin

eag

on

ist

(aff

init

yfo

rD

3

rece

pto

rsu

bty

pe)

0.1

25–

0.5

00

1–2

1–3

8–12

Ren

alcl

eara

nce

Gen

eric

sid

e-ef

fect

so

fd

op

amin

eag

on

ists

Ro

pin

iro

leD

2d

op

amin

eag

on

ist

(aff

init

yfo

rD

3

rece

pto

rsu

bty

pe)

0.2

5–1.

00

1–2

1.5

6H

epat

ic;

acti

vem

etab

olit

e(s)

Gen

eric

sid

e-ef

fect

so

fd

op

amin

eag

on

ists

Ro

tig

oti

ne

D3,

D2

do

pam

ine

ago

nis

t1–

3T

ran

sder

mal

1–3

5–7

Hep

atic

;in

acti

vem

etab

olit

e(s)

Gen

eric

sid

e-ef

fect

so

fd

op

amin

eag

on

ists

Do

seto

be

red

uce

din

the

eld

erly

and

tho

sew

ith

hep

atic

or

ren

alfa

ilure

;p

reg

nan

cyan

dla

ctat

ion

are

con

trai

nd

icat

ion

s.tm

ax:

tim

eto

max

imu

mco

nce

ntr

atio

no

fth

ed

rug

;t1

/2:

hal

f-lif

e.


of nonpharmacological measures andprescription of melatonin. An interventionwhereby a shifting sleep schedule isprescribed is called chronotherapy: sleepmay either be systematically delayed oradvanced, until sleep is synchronised withthe desired bedtime. Light therapy in themorning (10,000 lux for 30 min) may besupplemented with formulations ofmelatonin, which is a natural hormonesecreted by the pineal gland. In a dose rangeof 0.5–3 mg, administered o3 h before theendogenous dim-light melatonin onset,melatonin may advance sleep. In addition,melatonin has weak hypnotic properties butis not often powerful enough to beprescribed as a hypnotic drug.

Sleep in somatoform disorders Manypsychosomatic and somatoform disorders(chronic fatigue syndrome, fibromyalgia,burn-out, etc.) are accompanied by sleepcomplaints, including insomnia,hypersomnia and nonrestorative sleep. Insome patients, a primary sleep disorder canbe diagnosed. Appropriate treatment of theunderlying sleep disturbance should resultin an improvement of the psychosomaticcomplaints. Often, however, the findingsfrom the diagnostic work-up are nonspecificor no obvious sleep disturbance can bedemonstrated. In this case, no therapeuticrecommendations can be made. At present,there are no compounds available thatwould improve the ‘intrinsic quality’ of sleep.

Further reading

N American Academy of Sleep Medicine.(2005). The International Classification ofSleep Disorders. 2nd Edn. Westchester,American Academy of Sleep Medicine.

N Billiard M, et al. (2006). EFNS guidelineson management of narcolepsy. Eur JNeurol; 13: 1035–1048.

N Buysse D. (2011). Clinical pharmacologyof other drugs used as hypnotics. In:Kryger MH, et al., eds. Principles andPractice of Sleep Medicine. 5th Edn. StLouis, Elsevier-Saunders; pp. 492–509.

N Espie CA. (2009). ‘‘Stepped care’’: ahealth technology solution for deliveringcognitive behavioral therapy as a first lineinsomnia treatment. Sleep; 32: 1549–1558.

N Hening WA, et al. (2004). An update onthe dopaminergic treatment of restlesslegs syndrome and periodic limb move-ment disorder. Sleep; 27: 560–583.

N Mendelson W. (2011). Hypnotic medica-tions: mechanisms of action and phar-macologic effects. In: Kryger MH, et al,eds. Principles and Practice of SleepMedicine. 5th Edn. St Louis, Elsevier-Saunders; pp. 483–491.

N Montplaisir J, et al. (2011). Restless legssyndrome and periodic limb movementsduring sleep. In: Kryger MH, et al, eds.Principles and Practice of Sleep Medicine.5th Edn. St Louis, Elsevier-Saunders;pp. 1026–1037.

N Morgenthaler T, et al. (2006). Practiceparameters for the psychological andbehavioral treatment of insomnia: anupdate. An American Academy of SleepMedicine report. Sleep; 29: 1415–1419.

N Morin CM, et al. (2006). Psychologicaland behavioral treatment of insomnia:update of the recent evidence (1998-2004). Sleep; 29: 1398–1414.

N Schutte-Rodin S, et al. (2008). Clinicalguideline for the evaluation and manage-ment of chronic insomnia in adults. J ClinSleep Med; 4: 487–504.

N Wise MS, et al. (2007). Treatment ofnarcolepsy and other hypersomnias ofcentral origin. Sleep; 30: 1712–1727.


Medicolegal and economicaspects of sleep disorders

Gimbada B. Mwenge and Daniel Rodenstein

Sleep is a constitutive component of humanlife. Both sleep quality and duration arefactors that influence health and wellbeing.Quantitative and qualitative sleep deficitslead to a deterioration of health (Mageeet al., 2011) and impaired cognitive function.According to a recent review of the literature(Grandner et al., 2010), the ideal duration ofsleep would be ,7 h. 7-h sleepersexperience the lowest risks for all-causemortality, whereas those with shorter orlonger sleep duration have significantlyhigher mortality risk.

It has been reported that 20% of thepopulation suffer from sleep disorders inEurope. The most frequently found sleepdisorders are insomnia (6–15%), OSA (5%)and restless legs syndrome (RLS) (4%). Todate, the deleterious effects of sleepdisorders on health have been welldemonstrated only in OSA (it promotes

hypertension, weight gain and, probably,type 2 diabetes mellitus, cardiovascularcomplications and an increased risk ofmotor vehicle accidents (MVAs)) (George,2007). OSA is characterised by objectivephysiopathological alterations, such asintermittent hypoxia, which could plausiblybe linked to the pathogenesis of thosecomplications (Ryan et al., 2005). However,insomnia, RLS and excessive daytimesleepiness (EDS) are subjective complaintsand frequently lack any objective correlate.

Sleep disorders, like any chronic disease,may carry a significant economic cost. Costscan be direct or indirect. Direct costs aredefined as hospital care, cost of drugs, costof medical care and cost of care at otherinstitutions, whereas indirect costs aredefined as the value of lost production,including sick leave, short- or long-termdisability and early mortality. This chapterdescribes the medicolegal aspects of themost frequent sleep disorders and theireconomic consequences.

To evaluate the cost–benefit, cost–utilityand cost-effectiveness of insomniatherapies, and even to compare them toeach other, we need a common tool ofmeasure: the quality-adjusted life-year(QALY). QALYs are the most commonoutcome measure in cost–utility analyses ofhealthcare programmes. The QALY offers astraightforward procedure for combining thetwo most important outcomes of healthcareprogrammes, quality of life (change in thehealth status as perceived by the patient,due to a disease or to a medicalintervention) and quantity of life (theduration of this change), into one singlemeasure. QALYs can be separated into two

Key points

N To evaluate the cost–benefit, cost–utility and cost-effectiveness ofinsomnia therapies, and even tocompare them to each other, we needa common tool of measure: the QALY.

N OSA is associated with an increasedrisk of MVAs; driver licensing lawsvary greatly within the EU with regardto OSA.

N Pharmacological intervention for RLShas favourable cost–utility balances.

N Insomnia is a cause of workplacedisability.


components: survival time and the quality oflife during this survival. This score rangesfrom zero (representing death) to one(representing perfect health) for 1 yr. Utilityrepresents the subjective weight of healthimprovement as perceived by the patient,who attributes a value to this change. Thismay differ from what a physician may thinkabout the change in health status of thepatient, but utilities are determined bypatients, not doctors. Thus, a patient wholives for 1 yr in a health status with a utilityof 0.5 accumulates only 0.5 QALYs (i.e.1 yr 6 0.50 utility). However, as insomniais defined by subjective scales, the QALY is,in this case, strictly subjective, especiallysince there is no change in survival ininsomniacs. When assessing treatments,once QALYs are determined for each of thedifferent treatments being compared, onemust determine their cost-efficiency. Thisratio is simply the cost divided by the outputexpressed in QALYs.

The advantage of this complex concept isthat one can compare the QALYs bought perEuro in different diseases or in a givendisease by different treatments. This canhelp one to decide whether to allocateresources in a more efficient way, byspending money in health sectors where thegain in QALYs will be maximal, or indeciding to stop the reimbursement of amedical intervention because a second onewill cost less and grant the same QALYs, orcost the same but obtain more QALYs.

It is generally assumed that when the cost-effectiveness ratio of a medical interventionis below a given threshold, the interventionis worth paying for. In international terms,the threshold is assumed to lie between twoand three times the per capita gross domesticproduct of a given country. Therefore, theeconomic status of the country has a definiteimpact on the type of treatments that can beafforded in that country. This is aboutJ30,000 per QALY in Europe.

Obstructive sleep apnoea

OSA is the most common type of SDB, andis characterised by periodic complete orpartial upper airway obstruction causing

sleep fragmentation and repetitivehypoxaemia, which leads to poor,nonrestorative sleep, EDS, neurocognitiveimpairment, arterial hypertension,myocardial infarction, stroke and increasedrisk of MVAs. The main severity parameter isthe AHI, the number of complete (apnoeas)or incomplete (hypopnoeas) interruptions inbreathing per hour slept. It is also welldocumented that OSA is associated withearly mortality.

The degree of daytime sleepiness isgenerally not as extreme as in narcolepsy,but it is consistently described to be in theupper limits of what is considered normal,or already within the boundaries of EDS.Moreover, sleepiness (quantified by theESS) is not always perceived by the subject.Indeed, studies show that the average ESS innarcolepsy is o15, whereas it is around 11–12 in OSA. The prevalence of OSA isestimated at 5% of adults according tovarious studies (Young et al., 1993).However, professional drivers are believedto have an increased prevalence of OSA(Gurubhagavatula et al., 2004). AnAustralian report found, in a sample ofcommercial lorry drivers, that 15% had OSA(assessed by questionnaires and PSG),whereas a study from Brazil found a quarterof lorry drivers to be at risk for OSAaccording to the Berlin questionnaire(Moreno et al., 2004). A previous reportfrom the USA found that 10% of long-distance lorry drivers had .30 falls inoxygen per hour of sleep (Stoohs et al.,1995). However, there is a lack of uniformityin the way the driving licence regulationsconsider OSA in their medical annexesworldwide and even within the EuropeanUnion (EU).

Professional performance and work disabilityOSAS leads to a certain degree ofneurocognitive impairment, which couldaffect professional activities.

Mulgrew et al. (2007) studied patients withSDB referred for PSG. They found worklimitations (assessed by a multidimensionaloccupational questionnaire) to besignificantly more present in the most sleepyquartile of patients compared with the least


sleepy quartile. This was independent of theseverity of OSA itself and seemed to dependsolely on the degree of daytime sleepiness. Itshould be mentioned that the sleepier groupwas indeed extremely sleepy, much morethan is usual for OSA patients.

Engleman et al. (2000) studied cognitiveperformance in SAHS. They suggested thatdeficits broadly worsen with diseaseseverity, with large average values forattention and executive functions. Theyconcluded that sleepiness and hypoxaemiamight contribute to cognitive deficits.

Finally, through a meta-analysis ofrandomised placebo-controlled crossoverstudies of CPAP treatment involving 98SAHS patients (AHI .5 events?h-1 and twoor more symptoms), Engleman et al. (2000)showed trends towards better cognitiveperformance on CPAP than on placebo,although the enhancements were small.

More recent studies have assessed theinvolvement of OSA in work disabilitywithout assessing sleepiness or cognitiveperformances.

Omachi et al. (2009) performed PSG in agroup of 153 employees and they found 83subjects with sleep apnoea (AHI.5 events?h-1). They found prospectivelythat the combination of OSA and EDScontributes to work disability, but that OSAitself contributes to recent work disability inpeople who have OSA but do not have EDS.

Those data were confirmed in another studyconducted in Finland (Sjosten et al., 2009)in a total of 766 employees, which foundthat employees with OSA (both females andmales), compared with controls, had aparticularly pronounced risk (two-foldincrease) of long-term disability caused byinjury or mental disorders.

Finally, accidents in the workplacerepresent a significant economic burden foremployers and a significant health risk foremployees. OSA may represent a causalfactor in work accidents.

Lindberg et al. (2001) showed that subjectswith snoring and daytime sleepiness had an

increased risk for occupational accidents(retrieved from a national database) duringa 10-yr follow up, with an adjusted oddsratio of 2.2 (95% CI 1.3–3.8). This increasedrisk took into account differences in age,BMI, smoking, alcohol dependence, years atwork, blue-collar job, shift work, andexposure to noise, organic solvents, exhaustfumes and whole-body vibrations. Neithersnoring alone nor daytime sleepiness alonewere associated with an increasedoccupational risk.

Krieger et al. (1997) showed that in the12 months after initiation of CPAP, bothoccupational and domestic accidents andnear-missed accidents decreased; thenumber of patients declaring an accident ora near-missed accident also decreased.

Increased MVA risks and legal aspectsPatients with OSA have been shown to havemore MVAs than drivers from the generalpopulation. OSA represents a risk factor forMVA estimated at three- to seven-fold thatof the general population. The first reportsof an increased number of MVAs in patientswith sleep apnoea were published in the late1980s (George et al., 1987). Many studieshave been published since then, usingdifferent methodologies, fromepidemiological studies to cohort studies,case–control studies and follow-up studies.Several investigators relied on patientresponses (subjective data), whereas manyothers turned to official databases of MVAs(in general, these record more seriousaccidents involving personal injury orsignificant financial consequences). Severalarticles have reported on the effects oftreatment (usually CPAP) on MVA rate,comparing, in a cohort of patients, the rateof MVAs for a given time before and afterthe institution of therapy. Patients werefrequently compared with a control sample,either limited in number and matched to thepatient group on several aspects orincluding all of the population in a givengeographical area (an island or state). Inmost studies, many factors that couldinfluence the comparisons were accountedfor in the calculations. These included, forinstance, distance driven per year, visual


problems, medications and alcoholconsumption, BMI, smoking, work scheduleand comorbidities. Most studies have triedto assess the possible relationships betweenthe severity of the disease, either throughthe AHI or the desaturation index, or theseverity of the disease-related sleepiness,usually assessed with the ESS, or someindex of sleep fragmentation, such as theMovement Arousal Index, on the one hand,and the rate of MVAs on the other hand.Results are concordant: there is anincreased risk of MVAs in patients with OSA,either for MVAs in general or for seriousMVAs only, and this excess is cancelled outwhen patients start and remain on treatmentwith CPAP. The average increase in riskacross all studies is in the range of three-foldwith respect to the general population.

A meta-analysis was performed comparingthe risks of MVAs in all medical conditionsreported in the literature. The summary ofthis meta-analysis is that most medicalconditions confer a risk increased between1.2- and two-fold with respect to a healthypopulation (meaning that the diseaseincreases the MVA rate by 20–100%). OSAhad the highest increased risk, with a relativerisk of 3.71, which is second only to age andsex as a general risk factor for MVAs.

In some countries, the increased risk ofMVAs has prompted a specificconsideration of OSA in the legislation fordriving licences. For instance,questionnaires for driving licence applicantsmay include questions on OSA symptoms;patients diagnosed but untreated may beconsidered unable to drive or somerestrictions may be applied to their driving,whereas treated patients may recover fullpermission to drive. Directive 91/439/CEE,published on July 29, 1991, is the legal basisfor the common rules for driving licences inthe EU. It includes an annex (Annex III) thatdeals with all medical aspects, i.e. specificdiseases that may impair driving abilitiesand that need a specific assessment toevaluate whether it is safe to allow thecandidate to drive, either as they are orproviding remedial measures are adopted(for instance, to wear glasses to correct a

visual defect, to stop driving for a specifiedperiod after a heart attack, etc.). Within theEU, there is still a lack of uniformityconcerning driving licence regulations.Drivers with undiagnosed OSA may freelydrive across Europe, crossing bordersbetween member states with different rules.Annex III does not mention OSA, nor doesthe next modification of the directive, whichwill come into effect in 2013, even thoughOSA is probably the disease with the greatestrisk of MVAs (Barbe et al., 2007). Beyondabiding by directive 91/439/CEE, individualstates are free to institute more stringentrules. When it became evident that OSA was aserious risk factor for MVAs, some medicalexperts started national lobbying actions and,sometimes, succeeded in modifying theirnational driving licence regulations. However,this is far from being generalised. Moreover,the EU has extended its geographicalboundaries far beyond what they were in 1991.Therefore, the present situation resembleschaos more than a rational plan. A surveypublished in 2008 and concerning 25European countries showed that OSA wasmentioned in the Annex III of the nationalrules in 10 countries, whereas it is notmentioned at all in the 15 other ones(Engleman et al., 2000). Since free movementis a cardinal rule within the EU, drivers fromcountries where untreated OSA does notconstitute a limitation to drive can freely drivein countries where there are restrictions fordrivers with untreated OSA. This applies bothto private and commercial drivers.

It appears wise to continue lobbying actionsto include the disease representing the mostserious risk for MVAs in Annex III of theEuropean directive. All the more so, as thisis a treatable disease, and that it has beenshown time and again that compliantpatients treated with CPAP normalise theirMVA risk (Barbe et al., 2007).

Economic consequences Studies from variouscountries have shown that healthcare costsare higher in patients with OSA, and thatthey increase over time until a diagnosis ismade and decrease after the start ofeffective treatment. Costs are related to theseverity of the disease, with more severe


patients costing more than less affectedones (Tarasiuk et al., 2005).

It seems beyond doubt that OSA increaseshealth expenditure during the silent phase ofthe disease, i.e. when OSA is already presentbut has not been diagnosed (this is alsoreferred to as the pre-clinical phase of adisease). Several well-conducted studieshave consistently shown that healthcarecosts are increased by 50–100% in patientswho will be diagnosed with OSA in thefuture, with respect to the generalpopulation. This is true not only in middle-aged adults, but also in children, and inolder males and females (Tarasiuk et al.,2004). These increased costs have beenanalysed in some studies, and may beattributable to cardiovascular disease,digestive problems and metabolic disease inadults, whereas in children, the excess costsare mainly due to ENT and respiratoryconditions (Smith et al., 2002; Tarasiuket al., 2005).

Once the disease has been diagnosed, atreatment may be proposed. The applicationof an efficient treatment in OSA results in asignificant reduction in healthcare costs inthe months and years that follow. Thesereductions narrow the difference betweenthe costs of patients (that start to decrease)and those of the general population (thatcontinue to increase as time goes by, thoughat a much slower rate). The treatment doesnot completely cancel the difference, butattenuates it by about half. Again, thisappears to be true at all ages, from infancyto old age (Smith et al., 2002).

The economic impact of MVAs related toOSA was assessed in a theoretical study bySassani et al. (2004). Based on all availablepublished evidence, they calculated the totalnumber of MVAs attributable to OSA in theUSA. They then estimated the total cost dueto these accidents (medical andnonmedical, including lost work days,vehicles repairs, etc.). This resulted in a totalcost of US$15.9 billion. The authorsestimated then the cost of diagnosing andtreating all patients in the country, and theimpact of this policy on MVAs. They usedvery conservative estimates (for instance, a

20% diagnostic yield and a 70% therapysuccess rate), yet they concluded that thenet result would be a very significant savingof US$7.9 billion.

In the UK, a study (Guest et al., 2008)evaluated the cost-effectiveness of thetreatment of OSA with CPAP. The authorsincluded the direct medical costs (includingthe costs of diagnosing and treating thedisease), the preventive effect of treatmenton cardiovascular and cerebrovascularaccidents, and the costs related to MVAs.They calculated that treating OSA is cost-effective provided the treatment lasts foro2 yrs. The cost-effectiveness will improvethereafter as the cost remains stable (thediagnostic process does not need to berepeated and the CPAP machine has alreadybeen bought) but the benefits increase yearafter year. Considering the costs that wouldhave been incurred if the treatment had notbeen started, for instance, if an untreatedpatient has suffered a myocardial infarction,the authors calculated that the treatment ofOSA was not only cost-effective but it waseven cost-saving after 13 yrs of treatment.This means that if treatment lasts for 13 yrs,the entire health costs incurred by thepatient will be less than the cost of theirtreatment for OSA during those 13 yrs;society would save money (Guest et al.,2008). In other words, they estimated that:at year 1, the cost per QALY for CPAPcompared with no CPAP is expected toexceed £20,000; after 2 yrs, the expectedcost per QALY gained is £10,000 or less;and after 11 yrs, CPAP is the most cost-effective treatment. For instance, after14 yrs, the cost per QALY gained comparedwith no treatment is £1,620 (£4,123 to£259). Other reports coming from Canadaor Spain have also found this treatment tobe cost-effective (Banno et al., 2009;Wittmann et al., 2004). Ayas et al. (2006) inCanada performed a cost–utility analysiscomparing CPAP with no treatment for thetreatment of patients with moderate-to-severe OSA. The base-case analysis was ofpatients aged between 25 and 54 yrs whowere newly diagnosed with moderate-to-severe obstructive sleep apnoea/hypopnoeasyndrome (OSAHS), classified as having an


AHI o15 events?h-1. In this study, theauthors took into account CPAP compliance.A compliance rate of 70% was assumed. InSpain, Mar et al. (2006) performed a cost–utility analysis comparing CPAP with notreatment for the treatment of patients withOSA. Table 1 summarises these findings,which compare very favourably with otherpublicly funded therapies, such as primaryprevention of cardiovascular events usingcholesterol-lowering therapy ($54,000 perQALY to $1.4 million per QALY gained)(Prosser et al., 2000).

These values compare very favourably withother publicly funded therapies, such asprimary prevention of cardiovascular eventsusing cholesterol-lowering therapy ($54,000per QALY to $1,4 million per QALY gained)(Prosser et al., 2000).

Restless legs syndrome

RLS is characterised by an irresistible urgeto move the legs while resting (Innes et al.,2011). This urge is usually accompanied byuncomfortable sensations (creeping orburning) in the legs. Symptoms begin orworsen during periods of inactivity, arepartially or totally relieved by the movement,and are worse in the evening or at night. Theprevalence of RLS varies from 4 to 29% inthe general population. RLS can be primaryor secondary: primary RLS often appearsbefore the age of 20 yrs and has a familialcomponent; secondary RLS is usuallyassociated with disorders such as peripheralneuropathy. There is a higher incidence of

RLS in pregnant women, patients undergoingdialysis, type 2 diabetes patients, patientswith low iron levels and Parkinson’s diseasepatients (Karroum et al., 2008).

Medical and economic consequencesDiagnosis of RLS is based on subjectivereports. The real impact of RLS on sleep ispoorly understood. It is deemed to causeinsomnia, hypersomnia or EDS. RLS couldalso be associated with OSA; in this case,the treatment of OSA by CPAP may decreaseor remove RLS (Delgado Rodrigues et al.,2006). However, objective measures, suchas PSG, do not always show sleepabnormalities, and the presence of periodiclimb movements (PLMs) of sleep isinconstant and not considered as a formaldiagnostic criterion. However, thissyndrome is much discussed and is a sourceof controversy.

Only one study performed PSG in RLSpatients (a small simple size of 27 patients;Kallweit et al., 2009). 10 of the patientsreported EDS assessed by the ESS; they hadlower sleep latency on the MSLT and longertotal sleep time, but no differences betweenPLMs. Further studies are needed to betterunderstand this sleep disorder, which seemsnot to lead to objective sleep impairment.

RLS is a treatable disease. Once RLS isdiagnosed, medication is given (usuallydopaminergic agonists). These medicationshave proven effective, with improvement inquality of life (Giorgi et al., 2006).

Table 1 Cost-effectiveness studies of CPAP treatment.

Guest et al. (2008) Ayas et al. (2006) Mar et al. (2006) Tousignant et al.(1994)

Study design Compared CPAP,auto-CPAP and no

treatment

CPAP versus notreatment



OSAS patients n 120 99 46 19

Cost-effectivenessratio per QALYlimited to 5 yrs

£1400 US$3354 (rangeUS$1064–9715)

J7861 US$9792

AHI events?h-1 .30 .15 41.3¡14.6 67.6¡24.3

ESS 12 Mean¡SD 13.8¡5.8

Country UK Canada Spain Canada


Some authors have tried to evaluate theeconomic burden of this disease. It has beenreported that RLS patients have meanproductivity losses of 1 day per week, andthat treatment and rehabilitation cost aboutUS$228 million per year (Allen et al., 2011).Another study, conducted in Germany, foundthat direct and indirect medical costs over a 3-month period amounted to about J2,090 perpatient (Dodel et al., 2010). Another recentreview of the economics of RLS confirmed itsimpact on patient quality of life and equatedits burden to that observed in patients withother chronic conditions.

Among the drugs used to treat RLS, somehave been assessed for their cost–utility. Inthe UK, the incremental cost per QALY forpramipexole (a dopaminergic agonist usedto treat RLS) was £3,349 versus no treatmentand showed a cost-saving profile of £92against ropinirole. In Sweden, pramipexoleproduced cost savings of SEK 2,381 versus notreatment and SEK 3,564 versus ropinirole.QALY gains in both countries were 0.095versus no treatment and 0.007 versusropinirole. These results compare well withUK cost-effectiveness thresholds of £20,000per QALY and are cost saving for Sweden.One-way and probabilistic sensitivityanalyses showed results to be robust.

Insomnia

Epidemiology Insomnia is a subjectiveperception of insufficiency of sleep durationor quality. The diagnosis and assessmentare based on subjective reports (sleepquestionnaires or scales) and sleep diaries.It can be defined as a complaint ofprolonged sleep latency, difficulty in fallingasleep or maintaining sleep, ornonrestorative sleep accompanied bysignificantly impaired daytime functioning inthe absence of a specific physical, mental orsubstance-related cause. The complaintshave to last for o4 weeks to be diagnosedas insomnia. Insomnia can occur acutely orbecome a chronic disorder when it occurs atleast three times per week, usually foro3 months’ duration. The AmericanPsychiatric Association Diagnostic andStatistical Manual of Mental Disorders(DSM) classifies insomnias into primary

insomnia (PI), insomnia related to a medicalor mental disease, and secondary insomniarelated to the intake or substance abuse/dependency (American Academy of SleepMedicine, 2005).

Insomnia is the most prevalent sleepdisorder, especially in the elderly, with aprevalence of 10–20% of the population inEurope and Canada, and it is mostlydiagnosed in women. There is no relationshipbetween insomnia and sleep duration.

Medical consequences Many studies basedon questionnaires and surveys or reviewshave reported cognitive impairment inpatients with PI (Fulda et al., 2001;Shekleton et al., 2010). However, a newmeta-analysis has shown that theirperformance was comparable to that ofnormal sleepers for other aspects ofattention (alertness, divided attention,sustained attention and vigilance),perceptual and psychomotor processes,verbal functions, procedural memory andsome aspects of executive functioning(verbal fluency and flexibility), as well asgeneral cognitive functioning (e.g. memoryand concentration). This casts doubts onthe real impact of insomnia on cognitivefunctions, which could be much less thanpreviously thought.

Patients suffering from insomnia commonlyreport poor sleep and significant reductionsin sleep duration and quality of life.

Insomnia has also been related to impairedimmune function, elevated BMI and type 2diabetes (Motivala, 2011; Vgontzas et al.,2009). A population sample study of 953patients in Canada that compared goodsleepers with insomniacs showed that thelatter have a higher rate of prescriptionmedications, most notably to treat insomnia(which is understandable), mood andanxiety disorders. But in terms of mortality,patients with insomnia without underlyingcomorbidities appear to have no survivalrisk, as long as they refrain from long-termuse of sleeping pills.

Economic consequences According to variousstudies, the use of healthcare resources iswidely increased in insomnia patients.


For instance, in a study conducted in arandomly selected sample of 948 adults inQuebec, Canada, the total cost of insomniawas estimated at CA$6.6 billion, while in theUSA, direct and indirect costs associatedwith insomnia may exceed US$100 billionannually. A variety of treatments areavailable for the management of chronicinsomnia, including cognitive–behaviouraltherapy (CBT), which is believed to be themost effective treatment of PI, hypnoticssuch as Z-drugs, and alternative therapies(e.g. off-label medicines, alcohol, herbalremedies and alternative activities), whichare supported by very little evidence. CBTimproves sleep quality, reduces hypnoticdrug use and improves health-related qualityof life at a favourable cost (Morgan et al.,2004) even among long-term hypnotic userswith chronic sleep difficulties. Z-drugs, suchas eszopiclone, have been shown to improvequantitative sleep and quality of life, andreduce insomnia severity and work limitations(Krystal et al., 2008; Walsh et al., 2007).

A study conducted in 2004 by Morgan et al.in the UK found that the mean cost perQALY at 6 months was £3,418 in 209patients with chronic insomnia treated byCBT. Another study conducted to assess thecost-effectiveness of long-term treatmentwith eszopiclone compared with placebotherapy found that the incremental cost perQALY gained was approximately US$9,930including productivity gains and US$36,894excluding productivity gains in theeszopiclone group. This cost-effectivenesswas also confirmed in a more recent study,which found a cost per QALY of US$4,800for eszopiclone versus placebo.

Work disability A cohort study (Sivertsenet al., 2009) conducted in 6,599 workingsubjects over 4 yrs investigated thecontribution of insomnia versus sleepduration to later long-term work disability,where disability pension was chosen as theoutcome. Insomnia was a strong predictorof permanent work disability (OR 4.56) andthis effect remained significant aftercontrolling for sleep duration and otherpossible confounders (OR 1.88), but thesecorrelations were not found when they

considered only short sleepers. AnotherCanadian study conducted in 900 patientsfound that patients with insomnia are morefrequently absent from work, while somereported having experienced reducedproductivity compared with good sleepersand higher rates of non-motor-vehicleaccidents. However, as insomnia is asubjective perception, one might wonderwhether the perception of lower profitabilityin the workplace is not affected by the sameperception of cognitive dissatisfaction withdaytime functioning.

Legal aspects The ability to drive isinfluenced by many factors, such asattention span, concentration ability,reaction time, good vision and socialbehaviour on the road. All these conditionscan be reduced by poor self-perception,depression and the use of psychotropicmedications.

The literature is very sparse concerninginsomnia and MVA risks, and the few dataavailable are controversial. In one studymentioned previously, which comparedgood sleepers and insomniacs, the authorsdid not retrospectively find a higher risk ofMVAs in insomniacs. A more recent studyconducted by Williamson et al. (2011)revealed that daytime fatigue and sleepinessas a consequence of insomnia are of directsafety concern, as they can affect accidentrisk. Similarly, one retrospective studyconducted in a population of patients whohad already had one or more accidentsfound an odds ratio .5 in patients withinsomnia. Although a strong correlation hasbeen demonstrated between the risk ofaccidents in short sleepers (Marshall et al.,2004) and subjects with EDS, we cannotextrapolate from these data to insomniacs.Moreover, people suffering from insomniaare not often short sleepers nor do they havereal EDS confirmed by objective tests, so itis difficult to incriminate those parametersin explaining the accident risk.

However, as previously stated, insomniapatients consume more psychoactive drugsthan good sleepers and, therefore, couldhave a higher risk of vehicle accidentsbecause of the treatment of insomnia, rather


than from the insomnia itself (Smink et al.,2010). In Germany, driving licenceregulations refer specifically to people withchronic insomnia (Fromm, 2008). Followingthe text of this directive, unaccompaniedpeople with insomnia no longer havepermission to drive when they suffer frommeasurably pronounced somnolence duringthe daytime. As a condition of obtaining andkeeping a driving licence, the person has toundergo continuous checks of their disease.People with special driving licence classes,i.e. for heavy goods vehicles, buses andtaxis, are regularly screened and have to beexamined by a doctor if they have typicalsymptoms of insomnia. To our knowledge,this German law seems to be unique.

Conclusion

We have reviewed in this chapter threefrequent sleep complaints or disordershaving medicolegal and economicimplications. We have purposely leftnarcolepsy aside due to its rarity and well-known consequences on working capacity,driving ability and quality of life.

Sleep disorders, whether completelysubjective or characterised by objectivefindings, may result in some impairment insocial and professional behaviour, as well asin the risk of MVAs. In addition, bythemselves or through the cost oftreatments, they represent an economicburden that needs to be assessed whenconsidering a disease as a whole.

Further reading

N American Academy of Sleep Medicine.(2005). The International Classification ofSleep Disorders. 2nd Edn. Westchester,American Academy of Sleep Medicine.

N Allen RP, et al. (2011). Prevalence anddisease burden of primary restless legssyndrome: results of a general populationsurvey in the United States. Mov Disord;26: 114–120.

N Ayas NT, et al. (2006). Cost-effectivenessof continuous positive airway pressuretherapy for moderate to severe obstruc-tive sleep apnea/hypopnea. Arch InternMed; 166: 977–984.

N Banno K, et al. (2009). Expenditure onhealth care in obese women with andwithout sleep apnea. Sleep; 32: 135–136.

N Barbe F, et al. (2007). Effect of contin-uous positive airway pressure on the riskof road accidents in sleep apnea patients.Respiration; 74: 44–49.

N Delgado Rodrigues RN, et al. (2006).Outcome of restless legs severity aftercontinuous positive air pressure (CPAP)treatment in patients affected by theassociation of RLS and obstructive sleepapneas. Sleep Med; 7: 235–239.

N Dodel R, et al. (2010). Health economicburden of patients with restless legssyndrome in a German ambulatory set-ting. Pharmacoeconomics; 28: 381–393.

N Engleman HM, et al. (2000). Cognitivefunction in the sleep apnea/hypopneasyndrome (SAHS). Sleep; 23: Suppl. 4,102–108.

N European Parliament. (2003). Directive2003/59/EC of the European Parliamentand of the Council of 15 July 2003 on theinitial qualification and periodic trainingof drivers of certain road vehicles for thecarriage of goods or passengers, amend-ing Council Regulation (EEC) No 3820/85and Council Directive 91/439/EEC andrepealing Council Directive 76/914/EEC.Off J Eur Union; L266: 4–17.

N Fromm IE. (2008). The revision of theregulations concerning driving licences(FeV) to combat the dangers of sleepapnoea and insomnia in road traffic.Pneumologie; 62: 387–391.

N Fulda S, et al. (2001). Cognitive dysfunc-tion in sleep disorders. Sleep Med Rev; 5:423–445.

N George CF, et al. (1987). Sleep apnoeapatients have more automobile acci-dents. Lancet; 2: 447.

N George CF. (2007). Sleep apnea, alert-ness, and motor vehicle crashes. Am JRespir Crit Care Med; 176: 954–956.

N Giorgi L, et al. (2006). Efficacy andtolerability of ropinirole in patients withrestless legs syndrome and a baselineIRLS total score . or 5 24 points – datafrom the ropinirole clinical trial pro-gramme. Curr Med Res Opin; 22: 1867–1877.

N Grandner MA, et al. (2010). Mortalityassociated with short sleep duration: theevidence, the possible mechanisms, andthe future. Sleep Med Rev; 14: 191–203.


N Guest JF, et al. (2008). Cost-effectivenessof using continuous positive airwaypressure in the treatment of severeobstructive sleep apnoea/hypopnoea syn-drome in the UK. Thorax; 63: 860–865.

N Gurubhagavatula I, et al. (2004). Occupa-tional screening for obstructive sleepapnea in commercial drivers. Am JRespir Crit Care Med; 170: 371–376.

N Innes KE, et al. (2011). Prevalence ofrestless legs syndrome in North Americanand Western European populations: asystematic review. Sleep Med; 12: 623–634.

N Kallweit U, et al. (2009). Excessive day-time sleepiness in idiopathic restless legssyndrome: characteristics and evolutionunder dopaminergic treatment. EurNeurol; 62: 176–179.

N Karroum E, et al. (2008). Le syndromedes jambes sans repos. [Restless-legssyndrome]. Rev Neurol (Paris); 164: 701–721.

N Krieger J, et al. (1997). Accidents inobstructive sleep apnea patients treatedwith nasal continuous positive airwaypressure: a prospective study. Chest; 112:1561–1566.

N Krystal AD, et al. (2008). Long-termefficacy and safety of zolpidem extended-release 12.5 mg, administered 3–7 nightsper week for 24 weeks, in patients withchronic primary insomnia: a 6-month,randomized, double-blind, placebo-con-trolled, parallel-group, multicenter study.Sleep; 31: 79–90.

N Lindberg E, et al. (2001). Role of snoringand daytime sleepiness in occupationalaccidents. Am J Respir Crit Care Med; 164:2031–2035.

N Magee CA, et al. (2011). Short and longsleep duration are associated with pre-valent cardiovascular disease in Australianadults. J Sleep Res; [Epub ahead of printDOI: 10.1111/j.1365-2869.2011.00993.x].

N Mar J, et al. (2003). The cost-effectivenessof nCPAP treatment in patients withmoderate-to-severe obstructive sleepapnoea. Eur Respir J; 21: 515–522.

N Marshall NS, et al. (2004). Abnormalsleep duration and motor vehicle crashrisk. J Sleep Res; 13: 177–178.

N Moreno CR, et al. (2004). High risk forobstructive sleep apnea in truck driversestimated by the Berlin questionnaire:prevalence and associated factors.Chronobiol Int; 21: 871–879.

N Morgan K, et al. (2004). Psychologicaltreatment for insomnia in the regulationof long-term hypnotic drug use. HealthTechnol Assess; 8: 1–68.

N Motivala SJ. (2011). Sleep and inflamma-tion: psychoneuroimmunology in thecontext of cardiovascular disease. AnnBehav Med; 42: 141–152.

N Mulgrew AT, et al. (2007). The impact ofobstructive sleep apnea and daytime sleepi-ness on work limitation. Sleep Med; 9: 42–53.

N Omachi TA, et al. (2009). Obstructivesleep apnea: a risk factor for workdisability. Sleep; 32: 791–798.

N Prosser LA, et al. (2000). Cost-effective-ness of cholesterol-lowering therapiesaccording to selected patient character-istics. Ann Intern Med; 132: 769–779.

N Rodenstein D. (2008). Cost-B26 actionon sleep apnoea syndrome. Driving inEurope: the need of a common policy fordrivers with obstructive sleep apnoeasyndrome. J Sleep Res; 17: 281–284.

N Ryan S, et al. (2005). Selective activationof inflammatory pathways by intermittenthypoxia in obstructive sleep apnea syn-drome. Circulation; 112: 2660–2667.

N Sassani A, et al. (2004). Reducing motor-vehicle collisions, costs, and fatalities bytreating obstructive sleep apnea syn-drome. Sleep; 27: 453–458.

N Shekleton JA, et al. (2010). Searching forthe daytime impairments of primaryinsomnia. Sleep Med Rev; 14: 47–60.

N Sivertsen B, et al. (2009). Insomnia andlong sleep duration are risk factors forlater work disability. The HordalandHealth Study. J Sleep Res; 18: 122–128.

N Sjosten N, et al. (2009). Obstructive sleepapnoea syndrome as a predictor of workdisability. Respir Med; 103: 1047–1055.

N Smink BE, et al. (2010). The relationshipbetween benzodiazepine use and trafficaccidents: a systematic literature review.CNS Drugs; 24: 639–653.

N Smith R, et al. (2002). What are obstruc-tive sleep apnea patients being treated forprior to this diagnosis? Chest; 121: 164–172.

N Stoohs RA, et al. (1995). Sleep and sleep-disordered breathing in commercial long-haul truck drivers. Chest; 107: 1275–1282.

N Tarasiuk A, et al. (2004). Adenoton-sillectomy in children with obstructive sleepapnea syndrome reduces health care utili-zation. Pediatrics; 113: 351–356.


N Tarasiuk A, et al. (2005). Determinantsaffecting health-care utilization in obstruc-tive sleep apnea syndrome patients. Chest;128: 1310–1314.

N Tousignant P, et al. (1994). Qualityadjusted life years added by treatmentof obstructive sleep apnea. Sleep; 17: 52–60.

N Vgontzas AN, et al. (2009). Insomniawith objective short sleep duration isassociated with type 2 diabetes: apopulation-based study. Diabetes Care;32: 1980–1985.

N Walsh W, et al. (2007). Nightly treatmentof primary insomnia with eszopiclone

for 6 months: effect on sleep, qualityof life, and work limitations. Sleep; 30:959–968.

N Williamson A, et al. (2011). Investigatingthe relative effects of sleep deprivationand time of day on fatigue and perfor-mance. Accid Anal Prev; 43: 690–697.

N Wittmann V, et al. (2004). Health carecosts and the sleep apnea syndrome.Sleep Med Rev; 8: 269–279.

N Young T, et al. (1993). The occurrenceof sleep-disordered breathing amongmiddle-aged adults. N Engl J Med; 328:1230–1235.


Development of breathing andsleep and physiopathology ofapnoea in the first years of lifeGary Cohen, Miriam Katz-Salamon and Ha Trang

The scope of this section is to describe thedramatic changes that occur in the firstmonths and years of life regarding sleeporganisation and breathing control, and touse this as a background to explain thepathophysiology of apnoea and apparentlife-threatening events (ALTEs) in infants.Circadian sleep/wake rhythms are alreadypresent in utero, observed during the lasttrimester of gestation. These fetal rhythmsare lost at birth but re-appear during the firstweeks and months of life. The organisationof sleep/wake rhythms is driven by both anendogenous biological clock and numerousexternal factors such as day/night variationsand parental/social factors. These may atleast in part underlie the large variabilityobserved with regard to sleep developmentand characteristics among individuals.

Development of sleep in the first years of life

Sleep and wakefulness states can beidentified in utero from the beginning of thethird trimester of gestation. In humannewborns, sleep is divided into twodifferentiated states: active sleep (AS) andquiet sleep (QS). AS is characterised by

low-voltage and rapid EEG, inhibited EMGand burst of saccadic and rapid eyemovements, associated with rapid andirregular respiratory and heart rates. Incontrast, QS is characterised by high-voltageand slow EEG, inhibited EMG and rare eyemovements, associated with slow andregular respiratory and heart rates. Thus, ASshares some features with the future REMsleep and QS shares some features with thefuture NREM sleep, sleep states that areobserved in adults.

In the first month of life, sleep duration lastsnearly 16–17 h?day-1, structured upon anultradian rhythm with alternate diurnal andnocturnal sleep. Newborns begin to sleep inAS, which occupies a high percentage ofsleep (50% of total sleep time). Sleep cycleswith alternate AS and QS last approximately50 min. No early- and late-night differencesin AS/QS distributions are observed.

Significant changes in sleep characteristicsand organisation are observed in the first6 months of life, with development of themain features of adult sleep. Beyond the firstfew months of life, the proportion of AS/REM sleep decreases significantly whereasQS/NREM sleep increases and differentiatesinto stages 1, 2 and 3 sleep with specific EEGwaveforms. Sleep onset no longer occurs inREM sleep after age 6 months. Organisationof sleep following a circadian rhythm isestablished by age 1 yr.

With increasing ages, mean duration of totalsleep time per day decreases progressively(average 14–15 h at age 6 months, 13–14 hat 1 yr, 12–13 h at 3 yrs, and 11 h at 6 yrs).Diurnal sleep time shows greater reductionthan nocturnal sleep time. The

Key points

N Breathing irregularities are commonin infancy but resolve spontaneouslyby 3–6 months.

N Breathing and cardiovascular controlare inextricably linked and should beevaluated together.

N ALTEs are not benign and may impacton long-term outcome.


disappearance of diurnal naps between 2and 6 yrs of age is associated withsubstantial reorganisation of nocturnalsleep. Sleep cycles that include REM andNREM periods last longer (55 min at age3 months, to 75 min at age 2 yrs). Thepercentage of REM sleep decreases rapidlyafter age 9 months. REM sleep latencyincreases progressively (15 min at age3 months, 70 min at age 2 yrs, 143 min at6–7 yrs). The duration and percentage ofREM sleep increase progressively overnight,being longer in the second half of the nightcompared to the first.

Parental/social factors and day/nightvariations play a key role in the developmentof the circadian rhythm of sleep.

Development of control of breathing

The purpose of breathing is to match theoxygen supply and demand and eliminatecarbon dioxide – a precarious task in infancywhen the metabolic rate is high, bodyoxygen stores are low, and breathing isirregular. Specialised chemoreceptorsfacilitate this task by constantly monitoringoxygen and carbon dioxide levels in theblood and tissues. Functionalchemoreceptors are not required to initiateor maintain breathing at birth (neurogenic/brain activity is sufficient to achieve this),but over subsequent days-weeks their input(neural ‘‘drive’’) becomes progressivelymore important for sustaining breathingrhythm, especially during sleep.Chemoreceptor dysfunction consequentlyunderpins and exacerbates sleep-relatedbreathing disorders.

Carbon dioxide Carbon dioxide stimulatesthe peripheral (carotid body) and central(brainstem) chemoreceptors at all ages. Theventilatory response (VR) of the term-borninfant and adult are more-or-less equivalent,so CO2 sensitivity is normally well developedif not fully mature at birth. The infant curveis however displaced to the left (it shiftsrightwards with age), which means minutevolume (V9E) in infancy is greater at anyPaCO2; this reflects a two-fold greater (weightadjusted) metabolic rate/CO2 production.For the infant born pre-term, the VR does

appear to increase after birth, althoughwhether this is due to an increase in CO2

chemosensitivity or improved lungcompliance and ribcage stability is uncertain.

Hypoxia Hypoxia stimulates mainly theperipheral chemoreceptors. Newborns areknown to be less sensitive and responsive tohypoxia. Two events – an increase in carotidbody chemosensitivity, and the waning ofhypoxia’s depressant actions on the brain –transform this situation. The first process israpid, and is normally well advanced by thesecond post-natal day; it augments mainlythe amplitude of the VR. The second processoccurs slowly over weeks to months andincreases the duration of the response, i.e.the increase in V9E/hyperpnoea lasts longer;this largely explains why the responsebecomes less ‘‘biphasic’’ and more‘‘monophasic’’ with age.

Asphyxia Asphyxia refers to a low O2–highCO2 state; it can develop for a variety ofreasons, e.g. central apnoea (no breathingefforts), or airway obstruction at the level ofthe mouth and nose, pharynx or larynx.Tolerance to asphyxia is particularly low. Itcan quickly lead to organ damage, braindepression, coma and death and poses aparticular danger in infancy because itdevelops rapidly (for the reasons givenabove), and because sensitivity to hypoxia(although not CO2) is weak. Normally (inadults) when both stimuli occur together,cardiorespiratory activation and arousal isdramatically heightened. This‘multiplicative’ interaction (which occurswithin the carotid body and brainstem) isweak at birth but gradually develops as thecarotid chemoreceptors reset, although thetime course is uncertain.

The upper airway The upper airway musclesare the accessory muscles of breathing. Theyare normally activated in parallel with thediaphragm and intercostal muscles toregulate airway calibre and resistance.During inspiration, the pharyngeal andlaryngeal abductors tense and dilate theairway to lower resistance and facilitate lunginflation. During expiration, the adductorsconstrict the larynx (in particular) to slowexpiration and maintain lung volume above


passive relaxation volume. The tone andphasic activity of these muscles isinfluenced by drive from thechemoreceptors and airway and laryngealsensors that monitor pressure, flow and pH.Upper and lower airway muscle coordinationimproves with age; if poorly coordinated, itmay cause obstruction or significantlyreduce airflow, especially whenchemoreceptor drive is strong.

Breathing and sleep During sleep,behavioural drives are lost or suppressedand tonic drive from the chemoreceptorshelps sustain a normal breathing rhythm.Consequently, chemoreceptor dysfunction isoften unmasked and breathing irregularitiesaccentuated, particularly during QS. The rateand depth of breathing fall at sleep onset;both become more irregular and variableduring REM sleep and reach a nadir duringNREM sleep (QS). Responsiveness to CO2

and hypoxia is also attenuated during QSand falls even further during REM sleep,accentuated by chest wall instability due tomuscle atonia. Short respiratory pauses aremore frequent, particularly during REMsleep. Premature and some full-term infantsexhibit bouts of periodic breathing (cycles ofregular breathing efforts separated bypauses lasting 3–10 s). The cyclical hypo-and hyperventilation may cause significanthypo/hypercapnia and intermittentdesaturation. Periodic breathing is commonearly on but decreases dramatically infrequency and duration with age as thechemoreceptors and central rhythmgenerators mature. Persisting immaturity ineither or both may accentuate periodicbreathing and apnoea.

Plasticity Development in general does notfollow a fixed trajectory but is moulded byexperience. Unusual perinatalcircumstances (e.g. pre-term birth, fetalexposure to drugs such as nicotine,repetitive apnoea, asthma or lung disease)can trigger long-lasting changes inchemoreception, reflecting structural orfunctional ‘reprogramming’ of thechemoreceptors themselves, of processingcentres within the brainstem, or of both.Exposure to chronic or repetitive

intermittent hypoxia and/or hypercapnia, forexample, can depress or enhance ventilatoryresponsiveness, depending on thedevelopmental stage at exposure.Circumstances that alter, delay or otherwisereprogramme chemoreflex developmentmay exacerbate cardiorespiratory failure andvulnerability to asphyxia.

Physiopathology, diagnosis and assessmentof apnoea and apparent life-threateningevents

Developmental aspects of respiratory rhythm inhealth The breathing pattern of the newborninfant is often characterised by apnoeas ofvarying duration and frequency, and byperiodic breathing (cycles of 3–5 regularbreaths separated by pauses). Breathingirregularities are evident in almost all infantsborn extremely pre-term (,28 weeks) andabout 30% of those born at full term. Thepropensity for apnoea and periodicbreathing in infancy has been ascribed to aweak response to hypoxia and hypercapnia,and/or a low CO2 apnoeic threshold. Mostbreathing irregularities resolvespontaneously by age 3–6 months due tomaturation of central rhythm generators andchemoreceptors, the two systems essentialfor maintaining a normal breathing rhythmduring sleep.

Aetiology of breathing irregularities Apnoea isa symptom rather than a disease. Breathingirregularities occur in diseases of the centralnervous system, lungs, muscles,metabolism, upper airways, etc. Apnoea maybe aggravated by many factors, e.g.immature or abnormal chemoreception,lung disease, upper airway dysfunction,systemic infection, intracranial haemorrhage,hypo- or hyperthermia, glucose/electrolyteimbalance, anaemia, gastro-oesophagealreflux and patent ductus arteriosus.

Clinical diagnosis is complicated by theintricate interaction between central andperipheral control mechanisms. Centralmechanisms control respiratory rhythm, aswell as the diaphragm and accessory musclesof breathing. Weak or unstable central drivecan therefore result in central, obstructive ormixed apnoea as well as irregular breathing.


Similarly, upper airway, lung and respiratorymuscle disease may further modify centraldrive via altered sensory feedback, andexacerbate central irregularities.

Clinical aspects of apnoea Control ofbreathing is inextricably linked withcardiovascular control, particularly duringsleep. Dysfunction in one affects the otherand vice versa, and can rapidly lead to adangerous or fatal downward spiral ofevents (see below). In clinics, the twosystems should be carefully evaluated intandem. Thus, duration and frequency ofcentral, obstructive and mixed apnoea mustbe linked with concomitant changes inblood gases as well as in heart rate andblood pressure. Brief apnoea is oftenregarded as physiological, but if frequentand associated with significant blood gasand circulatory changes (e.g. bradycardia)may compromise health (fig. 1).

Apparent life-threatening events An ALTE is‘an episode that is frightening to theobserver due to combination of apnoea,colour change (cyanotic or pallid), change inmuscle tone (marked limpness), choking, orgagging’. The mixture of breathing,circulatory and neuromuscular symptomsmakes separating coexisting and causativefactors difficult. ALTEs occur morefrequently in infants with pronouncedperiodic breathing or apnoea and are oftenheralded by repetitive apnoea, desaturationsand slow circulatory changes in thepreceding hours. Clinical managementshould focus on excluding seriousunderlying causes such as seizures, cardiacdiseases, bacterial infection, metabolicdisorders and central hypoventilation.Subsequent investigation should includecardio-respiratory function, blood gases,polysomnography/polygraphy, and(preferably) long-term home monitoring.

Figure 1. The sequence of apnoea, bradycardia and desaturation. a) Physiological apnoea, no desaturationor bradycardia. b) Short apneoa with rapid desaturation due to airway closure, low functional residualcapacity, ventilation/perfusion mismatch, intrapulmonary shunts and tachycardia. c) Apnoea withbradycardia due to vagal and/or trigeminal inhibition during swallowing, obstruction or reflux.


ALTEs should not be ignored; they causesevere hypoxia and are unlikely to be benign,even if the link with sudden infant deathsyndrome (SIDS) is controversial. If thereare no serious underlying conditions, therisk of SIDS is certainly low, but thelong-term outcome for these infants is yet tobe determined.

Further reading

N Carroll JL. (2003). Developmental plasti-city in respiratory control. J Appl Physiol;94: 375–389.

N Cherniack NS. (2006). Cardiopulmonaryintegration: the dark side. Respiration; 73:733–734.

N Cohen G, et al. (2005). Development ofchemoreceptor responses in infants.Respir Physiol Neurobiol; 149: 233–242.

N Darnall RA. (2010). The role of CO2 andcentral chemoreception in the control ofbreathing in the fetus and the neonate.Respir Physiol Neurobiol; 173: 201–212.

N Dewolfe CC. (2005). Apparent life-threa-tening event: a review. Pediatr Clin NorthAm; 52: 1127–1146.

N Gauda EB, et al. (2004). Maturation ofperipheral arterial chemoreceptors inrelation to neonatal apnoea. SeminNeonatol; 9: 181–194.

N Hunt CE, et al. (2004). Cardiorespiratoryevents detected by home memory mon-itoring and one-year neurodevelopmentaloutcome. J Pediatr; 145: 465–471.

N Katz-Salamon M. (2004). Delayed che-moreceptor responses in infants withapnoea. Arch Dis Child; 89: 261–266.

N Lorch SA, et al. (2011). Epidemiologyof apnea and bradycardia resolutionin premature infants. Pediatrics; 128:e366–e373.

N Poets CF, et al. (1991). Arterial oxygensaturation and breathing movementsduring the first year of life. J Dev Physiol;15: 341–345.

N Praud JP. (2010). Upper airway reflexes inresponse to gastric reflux. Paediatr RespirRev; 11: 208–212.


Sleep disordered breathing inchildren

Ha Trang and Anita K. Simonds

Obstructive sleep apnoea syndrome

OSAS is a public health problem affecting2–5% of the paediatric population. Partial ortotal obstruction of the upper airway resultsin obstructive apnoeas and hypopnoeasduring sleep, with or without associateddesaturation, hypercapnia and sleepfragmentation. It is thought that there is acontinuum between the normal state andOSAS, which is the complete clinicalpresentation of the disorder. Intermediatepresentations include:

1. primary snoring, which is chronicsnoring without associated apnoea orarterial desaturation; and

2. upper airway resistance syndrome(UARS), characterised by inspiratoryrespiratory efforts during sleep identifiedusing oesophageal pressure, but withouttypical apnoea or hypopnoea.

It is currently thought that primary snoringmay not have adverse effects on health. Incontrast, UARS, which is difficult to identifyin current practice, may progress towardscomplications if undiagnosed.

Epidemiology Prevalence of OSAS inchildren is estimated to range from 0.7% to10.3%, with a peak incidence between 5 and10 yrs of age. An estimate of 6.4%prevalence in children 1–6 yrs of age and of3.7% for those 7–12 yrs of age was reportedin one study. Prevalence of OSAS isunknown in infants and toddlers.Nevertheless, all studies have shown thatOSAS occurs equally in females and males.

Approximately 8–10% of children snoreregularly. However, not all children whosnore have OSAS. Infants can snore withoutapnoea, as shown in a case–control studyincluding 400 patients.

There are no large controlled longitudinalstudies relating to the natural history ofOSAS in children. It is not known whetherchildhood OSAS evolves into adult OSAS, orwhether there are two different diseases.Only one case series reported that asubgroup of infants who had previouslypresented spells and obstructive apnoeaswere diagnosed with OSAS at 5 yrs of age.

Physiopathology Factors underlyingdevelopment of OSAS in children areprimarily associated with reduced patency ofthe upper airway. The main causes areenlarged tonsils and/or adenoids. However,the pathophysiological mechanisms aremultiple and intertwined, as follows.

N Abnormal craniofacial skeleton, rangingfrom major malformations (such as Pierre

Key points

N Childhood OSAS is under-recognised.

N Snoring is not predictive of thepresence of OSAS.

N PSG is an important tool fordiagnosing OSAS.

N Paediatric rules for scoring sleep andrespiration of PSG should be used.

N Paediatric criteria for OSAS shouldbe used.

N Hypoventilation is more severe inNREM sleep than in REM sleepin CCHS.


Robin syndrome, Crouzon syndrome,Apert syndrome or achondroplasia) tomilder abnormalities not falling within alabelled syndrome. The latter may be amoderate hypoplasia of the middle third ofthe face with an arched palate and narrownasal cavity, or a narrow oropharyngealcavity, with or without a retrognathia and alingual retrusion.

N Abnormal craniofacial and pharyngealsoft tissue, resulting in reduction of theupper airway size: enlargement oflymphoid tissues (e.g. tonsils andadenoids), cervical fat accumulation(e.g. obesity), increased upper airwaycollapsibility, nasal obstruction (Rapp,2003) etc.

N Neuromuscular dysfunction, in thebroadest sense of the term, includingimpaired neurological control of breathingor upper airway muscle tone, etc.

Moreover, genetic and environmentalfactors have been identified in thepredisposition to OSAS. The abnormalitiesare often numerous and complex. It is notalways easy to determine whichabnormalities are causes and which areconsequences of OSAS.

Clinical presentation Clinical presentation ofOSAS differs with age. Indeed, symptomsare dominated by respiratory difficulties atnight in all age groups, with snoring duringsleep, breathing pauses, laboured breathingor mouth breathing. Infants and youngerchildren (1–6 yrs of age) often struggle toremain awake at sleep times and takeunusual positions, such as prone with theneck extended during sleep. The parentscomplaint of sleep problems (restless sleep,frequent body movements, night sweats andfrequent nocturnal awakenings) orsecondary enuresis. Morning sluggishnessand occasional daytime fatigue are reported.Sometimes, nonrespiratory symptoms canbe misleading: impaired attention orconcentration, hyperactivity and behaviouralproblems all often manifested as alteringacademic performance.

In older children (pre-adolescent andadolescent), loud snoring and breathingdifficulties are common. They may report

daytime hypersomnolence, resulting indaytime naps, irritability, and impairedconcentration and attention.Hypersomnolence is a difficult symptom toassess in children, as typical subjectiveassessments using questionnaires may benot applicable in this age group. Occasionally,patients may report morning headaches.

Clinical assessment should include ageneral paediatric evaluation and acomprehensive examination of craniofacialsegments and oropharynx (see chapter 4).

Confirmation of OSAS

Polysomnography is an important diagnostictest PSG is an important tool that helps:

1. to confirm or exclude OSAS;2. to determine the severity of the disorder

and therefore intervene in treatmentand treatment strategy decisions; and

3. to exclude other possible causes ofsleepiness (Aurora et al., 2011).

PSG should include monitoring ofneurological parameters EEG, EOG, chinand leg EMG and ECG) and of respiratoryparameters (nasal air pressure transducer,oronasal thermal sensor, chest andabdominal respiratory movements, pulseoximetry, and carbon dioxide tension (PCO2)).

It is of importance to use paediatric scoringand diagnostic criteria of sleep studies whenmanaging children suspected of OSAS. Areview of sleep and respiratory rules for PSGin children has been released recently by theAmerican Academy of Sleep Medicine(Redline et al., 2007; see also ‘‘Diagnostictechniques in children’’). Obstructiveapnoeas or hypopnoeas are consideredwhen longer than two respiratory periods.Typically, obstructive apnoeas andhypopnoeas may occur only during REMsleep or be predominant in REM sleep.There is no international consensus on thePSG criteria for OSAS in children. However,an apnoea index .1 event?h-1 or an AHI.1.5 events?h-1 is considered abnormal inchildren and used as a cut-off value inmany studies.


Meanwhile, young children, typically infantsor toddlers, may demonstrate sustainedobstruction of breathing during sleep withhypercapnia and/or hypoxaemia, but withouttypical apnoea or hypopnoea. This pattern istermed as obstructive hypoventilation and isdefined as sleep-related hypoventilationassociated with partial airway obstruction.

Diagnosis of UARS is not performed inroutine practice as it requires invasivemeasurement of oesophageal pressure.Further evaluation of noninvasivetechniques of diagnosis is needed.

In summary, OSAS is confirmed by thefinding of either an elevated apnoea–hypopnoea index (.1.5 events?h-1) orobstructive hypoventilation in conjunctionwith the appropriate history and associatedclinical features.

Are there alternative diagnostic procedures toPSG for diagnosis of OSAS? An evidence-based review has recently been published forrespiratory indications for PSG in children(Wise et al., 2011). Snoring is highlyassociated with OSAS and, although thepresence of snoring does not always implyOSAS, its intensity is correlated with theseverity of OSAS. Conversely, the absence ofsnoring does not exclude OSAS. Breathingpauses may be unrecognised by parents asthey are predominant in REM sleep, whichoften occurs late at night.

A limited number of paediatric sleepquestionnaires are available; some exploreonly respiratory symptoms (Brouillettequestionnaire) whereas the others exploremany domains, such as sleep quality,hypersomnolence and behaviour. Somesleep questionnaires are generally inaccordance with PSG results, but are notable to discriminate children with primarysnoring from those with OSAS or measurethe severity of OSAS. The ESS, which iswidely used in adults, has been modifiedand adapted for children by changing thenature of diurnal activities. However, thisquestionnaire has not yet been validated inlarge paediatric populations and noconsensus exists on the cut-off scorein children.

Nocturnal home oximetry is typicallyconsidered positive when the graph showsbursts of desaturations occurring duringone or many periods of 10–30 minovernight. Positive oximetry may berelatively specific for OSAS, but negativeoximetry does not exclude the disorder.Oximetry may be used as a screening tool inselected populations, but there is noevidence supporting the idea that oximetryalone can replace PSG for OSAS diagnosis.

Respiratory polygraphy is a system forovernight recordings of respiratoryparameters without neurologicalparameters; thus, it does not allowdetermination of sleep periods and sleepstages. No studies have assessed thevalidity of home respiratory polygraphy forthe diagnosis of OSA in children. Twostudies have compared respiratorypolygraphy findings to PSG, both attendedat hospital. Both show high sensitivity todetect OSAS whenever the latter is definedas apnoea–hypopnoea index .5 events?h-1.

The validity of unattended home PSG fordiagnosing OSAS in children has not beenstudied. When performed in children aged5–12 yrs, 91% of unattended home PSGs areinterpretable. Sensors were installed byprofessionals at home 1 h before bedtime.Children were supervised by parents, ofwhom 46% said they had not slept at night.Airflow is the signal most frequently lost.

In summary, at this time, there is noevidence supporting the idea that theambulatory diagnostic techniques availablehave an acceptable degree of diagnosticaccuracy to replace PSG. There is a need forassessing these techniques in largepaediatric populations.

Associated complications OSAS adverselyaffects neurocognition, behaviour, the heartand circulation, and growth andmetabolism, resulting in increasedmorbidity in affected children.

The effects of OSAS on neuropsychologicaland cognitive functions have caught muchattention. Early studies reported learningdifficulties or school problems. A largenumber of studies have established that


neurocognitive deficits are highly prevalentin children with OSAS. The latterdemonstrate memory and attentionimpairment, executive function impairmentand lower intellectual quotient comparedwith controls. They may present behaviouraldisorders also, with various degrees ofhyperactivity and aggressiveness,sometimes evolving to attention deficithyperactivity disorder. These disordersresult in deterioration in learning abilitiesand school performance. Importantly,neurocognitive deficits are found to be morepronounced in children with more severeOSAS. Moreover, they are found not only inchildren with identified OSAS, but also inthose with primary snoring, i.e. withoutidentified apnoeas or hypopnoeas. Finally,tonsillectomy significantly improvescognition and behaviour of children.Different factors underlying brain deficitshave been suggested: hypoxaemia,especially repeated hypoxaemic swingsovernight that may cause ischaemic injury invulnerable brain regions including thecerebellum and hippocampus, and sleepfragmentation. Therefore, a number ofstudies show that OSAS is associated withcognitive and behavioural disorders inchildren .2 yrs of age. No studies areavailable on this aspect when OSAS occursin infants or toddlers.

In early cases of OSAS, cor pulmonale withheart failure was a presenting symptom.Now, pulmonary hypertension may be morecommonly asymptomatic. Recent studiesshow that children with OSAS may havesystemic hypertension (both diurnal andnocturnal systolic and diastolic bloodpressure). More discrete abnormalities canbe observed: loss of wake–sleep modulationof blood pressure (which is autonomicnervous system-mediated), left ventricularremodelling or endothelial abnormalities.Cardiovascular abnormalities correlate withdesaturation index, suggesting the role ofrepeated hypoxaemic episodes.

Failure to thrive was a presenting symptomof childhood OSAS in early reports.Resolution of OSAS produces significantgrowth rebound. It has been hypothesised

that failure to thrive may be due to reducedrelease of growth hormone (caused by sleepfragmentation) or increased energyexpenditure during sleep related to repeatedinspiratory efforts against upper airwayobstruction. Finally, it is unclear whetherOSAS is associated with insulin resistancein children.

Conclusion OSAS is common in children,but still under-recognised and undertreated.There are a number of unaddressedquestions on the pathophysiology,complications, diagnosis and treatment ofOSAS in children. There is a need forcontrolled, longitudinal, large-scale studiesin these fields in order to improveunderstanding and care of OSAS in children.

Sleep hypoventilation syndromes

Sleep hypoventilation syndromes include awide variety of disorders that affect centralcontrol of breathing and/or respiratorymechanics. Alveolar hypoventilation is dueto inadequate ventilation and results inincreased PaCO2 and concomitanthypoxaemia. Whatever the causativemechanisms, treatment aims to provideventilation support by use of NIV and totarget the primary disease when possible.

Central congenital hypoventilation syndrome(CCHS) is the result of congenital failure ofautonomic control of breathing. Theincidence of CCHS is estimated to be one in200,000 (Trang et al., 2005). More than90–92% of patients with CCHS harbourPHOX2B mutations, the most frequentbeing polyalanine expansions. Treatmentis currently supportive, by lifelongventilatory support.

CCHS typically presents at birth. Neonateswith CCHS demonstrate repeated apnoeas,cyanosis during sleep, bradycardia anddespite severe hypoxaemia, fail to increasetheir breathing. PSG shows severe alveolarhypoventilation during sleep with shallowbreathing and a slow respiratory rate.Hypoventilation is most severe during sleep,and especially during NREM sleep, a sleepstate during which control of breathingdepends nearly exclusively upon the centralcarbon dioxide level. The hallmark feature of


CCHS is absent or markedly reduced centralhypercapnic ventilatory responses. Aventilatory deficit persists throughout life.With increasing age, most patients canbreathe spontaneously while awake.However, 5–10% of them also requireventilatory support during daytime.

Milder CCHS can be diagnosed duringchildhood or even adulthood, being typicallymanifested by cyanotic apnoea spells andunexplained convulsions resistant totreatment with a normal EEG, and the abilityto hold breath for prolonged periods. Thesepresentations have been associated withshort expansion mutations containing 25alanine repeats (Weese-Mayer et al., 2010).

Neuromuscular and related conditionsContrary to CCHS, ventilatory drive isusually well preserved in neuromusculardisorders. Instead, alveolar hypoventilationoccurs as weak respiratory muscles cannotwithstand the work of breathing. InDuchenne muscular dystrophy, congenitalmuscular dystrophies and myopathies,inspiratory and expiratory muscle weaknessusually progress in tandem. In spinalmuscular atrophy, expiratory muscleweakness may initially outstrip expiratorymuscle weakness. Scoliosis and upperairway obstruction can add to themechanical load. Hypoventilation occursfirst in REM sleep and then progresses toNREM sleep, and finally to diurnalhypoventilation if the vicious cycle is notaddressed. A forced vital capacity of ,60%predicted and raised serum bicarbonatelevel are simple screening tools fornocturnal hypoventilation, and probablymore useful than direct measurements ofrespiratory muscle strength (Hull et al.,2012). Cough peak flow is a useful measureof cough efficacy. Children over ,10 yrs ofage with a cough peak flow of ,270 L?min-1

are at risk of difficulty clearing secretionsand have an increased rate of chestinfections. Sleep studies should be carriedout annually in children with a vital capacityof ,60% pred, as well as in those who havelost ambulation or who have any symptomsof nocturnal hypoventilation such as poor

sleep quality, frequent awakenings, morningheadaches, concentration problems orsleepiness during the day, and anorexia atbreakfast. The frequency of sleep studiesshould increase if symptoms progress orlung function changes rapidly. At aminimum, PaO2 and transcutaneous/end-tidal PCO2 should be monitored; respiratoryevents can be better characterised intocentral and obstructive by multichannelrespiratory monitoring. Arousals can eitherbe detected by full PSG or by measurementof autonomic surrogates, such as heart ratevariation. Oximetry alone is not sufficient tofully exclude nocturnal hypoventilation, butan overnight mean SpO2 of o93% in a childwho has slept well and who is entirelyasymptomatic makes significant nocturnalhypoventilation, requiring ventilatorysupport, unlikely (Hull et al., 2012). Thetreatment for children with symptomaticnocturnal hypoventilation is positivepressure ventilation, usually NIV.

Further reading

N Aurora RN, et al. (2011). Practice para-meters for the respiratory indications forpolysomnography in children. Sleep; 34:379–388.

N Hull J, et al. (2012). British ThoracicSociety guideline for respiratory manage-ment of children with neuromuscularweakness. Thorax; [Epub ahead of printDOI: 10.1136/thoraxjnl-2012-201964].

N Redline S, et al. (2007). The scoring ofrespiratory events in sleep: reliability andvalidity. J Clin Sleep Med; 3: 169–200.

N Trang H, et al. (2005). The FrenchCongenital Central HypoventilationSyndrome Registry: general data, pheno-type, and genotype. Chest; 127: 72–79.

N Weese-Mayer D, et al. (2010). An officialATS clinical policy statement: congenitalcentral hypoventilation syndrome. Am JRespir Crit Care Med; 181: 626–644.

N Wise MS, et al. (2011). ExecutiveSummary of respiratory indications forpolysomnography in children: an evidence-based review. Sleep; 34: 398A–398AP.


Comorbid respiratorydisorders in children

Anita K. Simonds

In children with chronic lung diseaseassociated with limited pulmonary reserve,the normal physiological changes in sleepcan cause significant gas exchangeabnormalities. Furthermore, these episodescan fragment sleep, as can problems withcough, wheeze, nasal blockage, chest wallasynchrony and side-effects from medication(Gaultier, 2000), with consequent adverseeffects on daytime function.

Asthma

Sleep disturbance is common in childrenwith asthma; both objective and subjectivemeasures confirm decreased total sleeptime, and increased nocturnal awakenings.Reduced slow-wave sleep and increaseddaytime sleepiness may occur. In a survey of8–9-yr-old children, 32.6% of those withasthma reported sleep disturbance at leastonce a week and one-third experiencednocturnal cough once a week or more

(Meijer et al., 1995). Circadian variation inlung function including reduction in peakflow during sleep, decreased lung volume,and possible allergen exposure to dust mitesin bedding, are factors. Nocturnal asthma isassociated with impaired attention andcognitive performance in the day whichimprove on optimising asthma control.Enhancement of mood and improvement indaytime behaviour can also be achievedafter interventions to improve nocturnalasthma control (Stores et al., 1998).However, some deficits may remain inchildren with well-controlled asthma. This islikely to be related to the fact that therapies,such as corticosteroids and theophylline,cause sleep disruption and there are casereports of insomnia as a side-effect ofleukotriene inhibitors. The situation iscomplicated by the fact that asthma is anindependent risk factor for SDB in children.Redline et al. (1999), showed that a historyof asthma was associated with an odds ratioof 3.83 (95% CI 1.39–10.55) for SDB, whereasthe odds ratio for persistent wheeze was 4.71(95% CI 1.3–16.76), with both ratiosadjusted for race and obesity. Theunderlying mechanisms are speculative.OSA/H syndrome is associated with swingsin intrathoracic pressure which may provokereflux and increased bronchialhyperreactivity, and heighten cholinergictone. Furthermore, improved asthmacontrol was seen in asthmatics with OSA/Htreated with CPAP therapy. The dynamic islikely to be complex and bidirectional, assleep deprivation can worsen OSA/H, andboth asthma and OSA/H are associated withactivation of pro-inflammatory cascades.

Key points

N Chronic respiratory disorders such ascystic fibrosis and interstitial lungdisease can be associated withworsening gas exchange overnight.

N Asthma and rhinitic symptoms canfragment sleep.

N Asthma is an independent risk factorfor SDB.

N Improvement in control of asthmaand rhinitis usually leads to animprovement in sleep quality.


Allergies

Rhinitis Allergic rhinitis, nasal blockage,postnasal drip and sinusitis can cause poorsleep quality and delayed sleep initiation. Ina community based study of 7-yr-olds,snoring frequency and loudness showed aseasonal variation, peaking in spring andsummer, suggesting that both allergens andviral aetiology play a role. Allergy is stronglyassociated with SDB (Storms, 2008), in onestudy of children who habitually snored,39% had evidence of allergy which is aroundthree times the norm. Intranasalcorticosteroids are first-line therapy formoderate-to-severe and perennial allergicrhinitis. Antihistamines reduce sneezing andrhinorrhoea but may not be as effective asintranasal corticosteroids at reducing nasalobstruction. In an open study of childrenwith allergic rhinitis, arousal frequency wassignificantly reduced by intranasalbudesonide with reported improvements insleep quality and rhinitis symptoms. Allergycontrol measures are also likely to bebeneficial. Any child with persistent rhinitisshould be closely examined for adenoidaland tonsillar hypertrophy.

Cystic fibrosis Nocturnal arousals, poor sleepefficiency and decreased REM sleepduration are associated with disease severityin cystic fibrosis (CF), and worsen at thetime of infective exacerbations. Sleepdisturbance is related not only to chroniclung disease but also to the frequentpresence of chronic sinusitis and nasalpolyps. Nocturnal hypoxaemia in CF hasbeen demonstrated for decades.Desaturation is most pronounced duringREM sleep due to a combination ofventilation/perfusion imbalance andhypoventilation. A decrease in tonicdiaphragm and intercostal muscle toneleads to a decrease in end-expiratory lungvolume which in turn causes closure ofdependent areas of lung (Muller et al.,1980). In addition, REM sleep is associatedwith a decrease in ventilatory drive, andarousal mechanisms. While SDB is seen inadolescents and young adults with moreadvanced CF lung disease, there is only aweak correlation between the extent of

desaturation during sleep and daytimeoxygenation or pulmonary function. Arterialblood gas tensions during wakefulness are areasonable guide in that CF patients with adiurnal PaO2 ,8 kPa (60 mmHg) spend60% of total sleep time with SaO2 ,90%,whereas those with diurnal PaO2 .9.3 kPa(70 mmHg) spent only 20% of sleep timewith SaO2 ,90%. Arousals during sleep maybe related to SDB, but more prolongedepisodes are often related to coughingspasms. In general, the main form of SDB isnocturnal hypoventilation; rarer obstructiveevents may be due to polyps or rhinosinusitis.

The indications for oxygen therapy orventilatory support during sleep in CF arenot precisely defined. Oxygen therapy isused in those with diurnal hypoxaemia. Inthose with nocturnal hypoxaemia alone, itmay improve cognitive function and school/work attendance, but most of these studieshave been in young adults. Significant CO2

retention is unlikely in hypoxaemic patientswho are normocapnic during the day, butwith more advanced disease markednocturnal hypoventilation with high CO2

levels can be seen. In this situation the useof nocturnal NIV may bridge the patient totransplantation, or can be used to enableoptimal oxygenation without worseningPaCO2, and achieve a reduction in associatedsymptoms such as morning headaches andsleep fragmentation.

Interstitial lung disease Chronic interstitiallung disorders are rare in childhood. As arule, chronic hypoxaemia during the dayworsens during sleep with the nadiroccurring in REM sleep. Oxygen flow ratemay, therefore, need to be increasedovernight. Progressive hypoventilation maybe seen in advanced lung disease. This canbe a challenging situation to address inthose with obliterative bronchiolitiscomplicating transplantation. Hereprofound desaturation and hypercapnia canbe seen due to extreme ventilation/perfusionmismatch and hypoventilation. NIV andsupplemental oxygen therapy may reducesymptoms of SDB, but are unlikely to reducedaytime breathlessness. When setting theventilator care must be taken not to induce


high inspiratory airway pressures as therisk of barotrauma and pneumothoraxis increased.

Bronchopulmonary dysplasia/chronic lungdisease A US study showed that 10% ofinfants weighing 1,500 g at birth developedbronchopulmonary dysplasia (BPD),whereas this rose to over 80% in thoseweighing ,800 g. Hypoxaemia during theday can worsen during sleep due to furtherderangements in pulmonary mechanics andhypoventilation. This can be seen even inthose with SaO2 .93% on air duringwakefulness. REM sleep-relateddisturbances cause sleep disruption andwhen supplemental oxygen therapy isprovided sleep quality and duration mayimprove through lengthening of REM sleepperiod. Chest wall asynchrony during sleepand desaturation have been associated withchanges in heart rate variability andsignificant impairment of right ventricularfunction with a lesser effect on leftventricular function. In infants with BPD,ventilatory control including hyperoxicventilatory response can be abnormal orabsent (Katz-Salamon et al., 1995). This isprobably the consequence of repeatedepisodes of hypoxaemia, but adds a furtherlayer of vulnerability and to the risk ofsudden infant death.

In BPD and chronic lung disease the aim isto increase mean SaO2 to .92% overnightand normalise as far as possible.Transcutaneous CO2 monitoring will identify

the small subset with nocturnalhypoventilation and CO2 retention thatmight benefit from NIV.

Further reading

N Gaultier C. (2000) Effects of breathingduring sleep in children with chronic lungdisease. In: Loughlin GM, Carroll JL,Marcus CL, eds. Sleep and Breathing inChildren. New York, Marcel Dekker; pp.661–689.

N Katz-Salamon M, et al. (1995). Bluntedperipheral chemoreceptor response tohyperoxia in a group of infants withbronchopulmonary dysplasia. PediatrPulmonol; 20: 101–106.

N Meijer GG, et al. (1995). Frequency ofnocturnal symptoms in asthmatic chil-dren attending a hospital out-patientclinic. Eur Respir J; 8: 2076–2080.

N Muller NL, et al. (1980). Mechanism ofhemoglobin desaturation during rapid-eye-movement sleep in normal subjectsand patients with cystic fibrosis. Am RevRespir Dis; 121: 463–469.

N Redline S, et al. (1999). Risk factors forsleep-disordered breathing in children.Assocations with obesity, race andrespiratory problems. Am J Respir CritCare Med; 159: 1527–1532.

N Stores G, et al. (1998). Sleep andpsychological disturbance in nocturnalasthma. Arch Dis Child; 78: 413–419.

N Storms W. (2008). Allergic rhinitis-induced nasal congestion:its impact onsleep quality. Prim Care Resp J; 17: 7–18.


Nonrespiratory conditions inchildren

Ha Trang, Michel Lecendreux and Eric Konofal

Children present a large range of sleepproblems, including SDB and, even morefrequently, nonrespiratory disorders, such ashypersomnias, insomnias and parasomnias.These sleep disorders adversely affect healthin children and may differ greatly with regardto diagnosis and management. Acomprehensive understanding of specificphysiopathological mechanisms andtherapeutic strategies is required in order tooptimise diagnosis and management ofthese patients. Indeed, clinicians should beable to identify these nonrespiratory sleepdisorders, which may be diagnosed as aprimary sleep disorder or be associated withSDB in a number of cases. Here we willfocus on narcolepsy, which is one of thewell-studied hypersomnias, and attentiondeficit hyperactivity disorder (ADHD), whichis one of the most common sleep-relatedbehavioural disorders.

Narcolepsy and other hypersomnias

Hypersomnias of central origin are a rarecause of excessive daytime sleepiness (EDS)but should be considered in children andadolescents as well as in adults. Narcolepsy,with or without cataplexy, remains the mostwell studied of the primary hypersomnias.Narcolepsy is a chronic and disablingdisorder affecting sleep and wakefulness,which is characterised by EDS, sudden sleepepisodes and attacks of muscle atoniamostly triggered by emotions (cataplexy)(Challamel et al., 1994). Narcolepsy withcataplexy is a rare disorder with a prevalenceestimated at 56 per 100,000 in the generalpopulation. The prevalence of the disorderin children remains unknown; however,based on studies conducted in the USA, thedisorder could affect 0.02–0.05% of thepaediatric population (Longstreth et al., 2009).

Narcolepsy is a lifelong disorder, theoccurrence of which during childhood isfrequent. The occurrence of the disorderduring childhood and adolescence shouldbe taken into consideration. Narcolepsy inchildren and adolescents is stillunderdiagnosed and is often mistaken in itsonset for other diseases or even neglected.Based on clinical experience, young patientsaffected by the disorder often show dramaticand abrupt impairment in their social skillsand academic performance due to EDS,fatigue and lack of energy. Symptoms maystart abruptly and sometimes verydramatically, with the occurrence of suddenand complete cataplexy, or progressivelyand insidiously, with EDS, weight gain orprecocious puberty being the onlysymptoms over weeks or months; this leadsto difficulties in recognising the condition or

Key points

N Nonrespiratory sleep disorders aretreatable in children.

N Excessive daytime sleepiness is nota common symptom in childrenwith SDB.

N Excessive daytime sleepiness is acommon symptom in narcolepsy.

N Children with iron deficiency candevelop sleep disorders such as RLSor PLM syndrome.

N Sleepiness or drowsiness iscommonly reported in children withADHD with predominant inattention.


making the diagnosis difficult at an earlystage. Indeed, if cataplexy is the most specificsymptom of the disorder, it is usually not thefirst symptom to appear. EDS appears initiallyin the majority of patients and cataplexy maymanifest within 1–5 yrs.

For instance, in our own cohort (Paris,France), all of the patients (100%)complained of EDS, which was the firstsymptom to occur (97% of the cases)besides weight gain. Clear-cut cataplexy wasreported or observed in 82% of the cases atthe first evaluation. Furthermore, hypnologichallucinations and sleep paralysis werereported in 33% of the subjects.

Narcolepsy is believed to be caused by theselective loss of a population ofhypothalamic neurons producing theneuropeptide hypocretin-1 (Nishino et al.,2000). Since 1983, it has been demonstratedthat narcolepsy is associated with thehuman leukocyte antigen (HLA) allele of theHLA-DQ gene, HLA-DQB1*0602. Becauseof this close HLA association, the disorderhas been suggested to be autoimmune. InAugust 2010, the Medicinal Product Authority(MPA) in Sweden announced a specialinvestigation regarding narcolepsy followingthe influenza virus A H1N1 outbreak. Duringthe summer of 2010, the MPA received, intotal, 22 reports of narcolepsy as an adversereaction after (A)H1N1 vaccination. Thereports concern children aged 12–16 yrswhere symptoms were compatible withnarcolepsy, were diagnosed after thoroughmedical investigation, and had occurred 1–2 months after vaccination. This particulartopic is currently under investigation andcould add to the understanding of theautoimmune hypothesis of narcolepsy(Dauvilliers et al., 2010).

To date, the treatment of narcolepsy hasbeen essentially symptomatic, with noprospect, at the present time, of a definitivecure or even remission for subjects who areobliged to take psychotropic medication ona long-term basis. Considering therepercussion of the disorder and itsdramatic consequences in children and/oradolescents, medication is often required ata very early stage of the disease and, in

many cases, soon after the diagnosis isconfirmed. Given the long-term nature ofthe treatment, the disease must bediagnosed and assessed as precisely aspossible before medication-based therapy isinitiated (Lecendreux et al., 2008).

For those children who will receivepharmacotherapy from a very young age, butalso in adolescents, training and educationare crucial to enhance the benefit of thetreatment. Having to be treated on a dailybasis represents a challenge for manyyoungsters, who may be reluctant to takemedication every day. Most medications areprescribed during the daytime period andgiven during main meals. However, recenttreatments, such as sodium oxybate,although not yet approved in children andadolescents, are given at bedtime and 4 hafter sleep onset, requiring specific andoperant strategies from the child as well asefficient supervision from the parents (Aranet al., 2010).

Although no studies demonstrating theireffectiveness are available, measures aimedat enabling one or more daytime sleepperiods are generally recommended. One totwo routine naps of 20–30 min increasedaytime wakefulness and psychomotorperformance. Nutritional advice, regularmeal times and physical activity should alsobe encouraged at an early stage in childrenand adolescents, in order to avoid weightgain and to help maintain regular growth.Counselling or brief psychotherapy is oftenrequired to enable the child to accept theloss of their previous healthy state andprogressively accept the reality of a disablingchronic condition.

Idiopathic hypersomnia (IHS) IHS is aprimary hypersomnia. IHS is a rare andpoorly defined sleep disorder, an essentialfeature of which is chronic EDS despiteundisturbed nocturnal sleep. IHS is difficultto diagnose in children and adolescents,where the condition seems to fluctuateacross time and developmental stages. IHSin children may be defined as fulfilling thefollowing criteria: chronic EDS for.6 months; absence of cataplexy; no otherevident cause of EDS, such as sleep apnoea,


periodic limb movements (PLMs) andinsufficient sleep syndrome; delayed sleepphase syndrome; and ADHD. IHS mayinitiate during childhood or adolescence andpersist into adulthood, and should thereforebe considered as a disabling condition thatrequires appropriate diagnosis and adequatemedication (Janackova et al., 2011).

Recurrent hypersomnia Recurrenthypersomnia is characterised by periodichypersomnia episodes that last from a fewdays to several weeks, and which recurweeks or months apart. Cognitive andbehavioural disturbances are common,including confusion, feelings of unreality,hallucinations, binge eating, hypersexuality,irritability and aggressive behaviour (Arnulfet al., 2008). The best characterised form ofrecurrent hypersomnia is the Kleine–Levinsyndrome, which is mostly described inadolescent males. In adolescent females,hypersomnia may occur in association withthe menstrual cycle, often within the first fewmonths after menarche.

ADHD and other behavioural disorders

ADHD ADHD is a commonneurodevelopmental disorder, with a highprevalence of 3–10% in children(Lecendreux, 2011; Biederman, 2005).ADHD is typically characterised bydevelopmentally inappropriate symptoms ofinattention, hyperactivity and impulsivity,with onset before the age of 7 yrs andimpaired functioning in two or moresettings (e.g. at school and at home).

Laufer et al. (1957) were the first clinicians toreport that sleep disturbances couldrepresent a significant source of distress forchildren with ADHD and their parents.Wender (1975) further speculated that thesechildren had ‘an increased frequency ofsleep difficulties: difficulty in falling [asleep]and remaining [asleep], and early awakening’.

Because any sleep disorder that results inshort sleep duration, fragmented or disruptedsleep, or EDS can cause problems with mood,attention and behaviour, symptoms of sleepdisturbances may mimic those of ADHD inchildren incorrectly diagnosed with thecondition (Cortese et al., 2006).

The most frequent sleep-related symptomsobserved in children with ADHD aredifficulty falling asleep, bedtime resistance,night awakenings, restless sleep anddifficulty awakening in the morning.Difficulties in settling to sleep and delayedsleep onset have been investigated inchildren with ADHD through the use ofparental reports or questionnaires(Biederman, 2005; Cortese et al., 2006).Disruptive night awakenings, caused byabnormal activity during sleep, PLMsyndrome or SDB, have been reported instudies using actigraphy, video monitoringand PSG. In cases of ADHD where theresults from a number of PSG tests or otherobjective studies remain inconclusive, anddifficulties in sleep maintenance, low sleepefficiency and alterations in sleeparchitecture do not seem indicative ofspecific sleep disturbances, there are reportsof sleep-onset delays, increased motoractivity and deficits in alertness.

Management of sleep problems maysignificantly improve the quality of life ofchildren with ADHD. Indeed, some studieshave showed that sleep disturbances worsenADHD symptoms, associated mood disordersor both; therefore, treatment of comorbidsleep disorders and interventions targeted atensuring adequate sleep may substantiallyimprove daytime ADHD symptoms.

Restless legs syndrome (RLS) RLS ischaracterised by uncomfortable legsensations with an irresistible urge to movethe legs. Because children have limitedability to describe their subjectivesymptoms, diagnosis of RLS may be moredifficult in this age group. A set of criteriaspecific for this population has beenproposed by the International Restless LegsSyndrome Study Group (Allen et al., 2003).

The relationship between RLS and ADHDwas recently examined (Allen, 2004). Severalhypotheses have been proposed to explainsuch an association with ADHD or ADHD-like symptoms. Indeed, RLS-associatedsleep disturbance may causeinattentiveness, moodiness, and‘paradoxical overactivity’, mimickingsymptoms of ADHD. Alternatively,


idiopathic ADHD and RLS can be comorbidconditions (Allen, 2004); individuals withRLS, and some of those with ADHD, mightshare a common dopamine dysfunction.

It is now accepted that the dopamine systemplays a central role in the pathophysiology ofRLS (Allen, 2004; Cervenka et al., 2006).Iron is a cofactor in the biosynthesis ofdopamine, and it has been speculated thatRLS symptoms could result from a primarydysregulation of iron metabolism in thebrain. Moreover, there is an emerging bodyof literature regarding iron deficiency inchildren with restlessness, overactivity andinattention. Furthermore, RLS symptomsmay exacerbate, or cause later onset of, thesymptoms of ADHD during the lifetime ofthe child. Children with RLS can developbedtime opposition, probably because theyassociate bedtime with the occurrence ofunpleasant RLS sensations. One mayrecommend that RLS be systematicallysought in patients with ADHD or ADHD-like symptoms, and that in patients withADHD, symptoms of RLS should beassessed also.

Cross-over between hypersomnias,behavioural disorders and SDB in children

In contrast to adults with SDB, EDS is not acommon complaint in the general paediatricpopulation with SDB. However, thissymptom may be more prevalent in somespecific populations, such as in obesechildren, whether caused by obesity per se ordue to the associated SDB.

The relationship between SDB and ADHD ismore speculative. Despite somemethodological limitations, several studieshave suggested an association betweensymptoms of SDB and ADHD. A meta-analysis, which included studies usingrigorous criteria for ADHD, recently showedthat mild SDB may indeed be associatedwith full ADHD (Cortese et al., 2006).

Sleep disorders encountered in childreninclude a large range of diseases that arenow more fully investigated. One shouldbear in mind that the potential associationsof different comorbid sleep disorders with

SDB in this age group, and subsequentdifficulties in recognition and in management,should always inform treatment.

Further reading

N Allen R. (2004). Dopamine and iron inthe pathophysiology of restless legssyndrome (RLS). Sleep Med; 5: 385–391.

N Allen RP, et al. (2003). Restless legssyndrome: diagnostic criteria, specialconsiderations, and epidemiology. Areport from the restless legs syndromediagnosis and epidemiology workshop atthe National Institutes of Health. SleepMed; 4: 101–119.

N Aran A, et al. (2010). Clinical and thera-peutic aspects of childhood narcolepsy-cataplexy: a retrospective study of 51children. Sleep; 33: 1457–1764.

N Arnulf I, et al. (2008). Le syndrome deKleine-Levin. [Kleine-Levin syndrome:state of the art.]. Rev Neurol (Paris); 164:658–668.

N Biederman J. (2005). Attention-deficit/hyperactivity disorder: a selective over-view. Biol Psychiatry; 57: 1215–1220.

N Cervenka S, et al. (2006). Support fordopaminergic hypoactivity in restless legssyndrome: a PET study on D2-receptorbinding. Brain; 129: 2017–2028.

N Challamel MJ, et al. (1994). Narcolepsy inchildren. Sleep; 17: Suppl., S17–S20.

N Cortese S, et al. (2006). Sleep andalertness in children with attention-defi-cit/hyperactivity disorder: a systematicreview of the literature. Sleep; 29: 504–511.

N Cortese S, et al. (2006). ADHD andinsomnia. J Am Acad Child AdolescPsychiatry; 45: 384–385.

N Dauvilliers Y, et al. (2010). Post-H1N1narcolepsy-cataplexy. Sleep; 33: 1428–1430.

N Janackova S, et al. (2011). Idiopathichypersomnia: a report of three adolescent-onset cases in a two-generation family.J Child Neurol; 26: 522–525.

N Konofal E, et al. (2001). High levels ofnocturnal activity in children withattention-deficit hyperactivity disorder: avideo analysis. Psychiatry Clin Neurosci; 55:97–103.

N Laufer MW, et al. (1957). Hyperkineticbehaviour syndrome in children. J Pediatr;50: 463–474.


N Lecendreux M, et al. (2008). Narcolepsieavec cataplexie chez l’enfant: particular-ites cliniques et approches therapeu-tiques. [Narcolepsy with cataplexy in thechild: clinical evaluation and therapeuti-cal management.]. Rev Neurol (Paris);164: 646–657.

N Lecendreux M, et al. (2011). Prevalence ofattention deficit hyperactivity disorderand associated features among childrenin France. J Atten Disord; 15: 516–524.

N Longstreth WT Jr, et al. (2009).

Prevalence of narcolepsy in King

County, Washington, USA. Sleep Med;

10: 422–426.

N Nishino S, et al. (2000). Hypocretin

(orexin) deficiency in human narcolepsy.

Lancet.; 355: 39–40.

N Wender PH. (1975). The minimal brain

dysfunction syndrome. Ann Rev Med; 26:

45–62.


Clinical assessment inchildren

Maria Pia Villa

Sleep history and clinical assessment ofan infant

Sleep problems during infancy (rangingfrom 1 to 12 months of age) are among themost prevalent problems presented topaediatricians, in terms of disorders ofinitiating and maintaining sleep. Insomniaduring infancy may be subdivided into twomain categories: behavioural insomnia andinsomnia related to diseases. Behaviouralinsomnia commonly occurs in 20–30% ofinfants, and, if left untreated, bedtimeproblems and night awakenings may resultin behavioural, emotional, and learningdifficulties, persisting into the preschool andschool-aged years. History should focus on

the infant’s sleeping environment, bedtimeroutines and parental expectations.Insomnia due to medical conditions ismainly caused by infant colic; otitis;gastrointestinal problems, such asregurgitations, vomiting, diarrhoea,abdominal cramps and bloating that may bea manifestation of gastrooesophageal reflux;and cow’s milk allergy or lactose intolerance.Information about sleeping arrangements, ifthe baby sleeps in the parent’s room or bed,or in an infant crib in a separate room, thebaby’s position while sleeping (on his/herback, side or belly) and environmentaltobacco smoke exposure are very important.Bed routine should be investigated in orderto understand if good sleep hygiene ispresent. Is it crucial to know the sleep/wakepattern of the baby, and if the baby hasdifficulty falling asleep, or whether frequentand prolonged night-time awakenings arereported by caregivers. Physical examinationof an infant with sleep disorders shouldinclude an initial visual examination of thebaby, noting any dysmorphic features orobvious malformations. Height, weight andhead circumference measurements shouldbe reported on appropriate growth charts.Signs of respiratory infections, or ear andurinary tract infections should beinvestigated.

If a suspected apparent life-threateningevent (ALTE) has occurred, history shouldfocus also on the skin colour at the time ofthe event, duration of the event, sleepposition and type of sleeping arrangement(chair, lounge, crib, car seat, bed), clothing,presence of abnormal movements, includingabnormal eye movements, muscle tone,presence of blood or bloody fluid at the

Key points

N Insomnia during infancy can beclassified into: behavioural insomniaand insomnia related to a specificmedical condition.

N An ALTE is a frightening andunexpected change in an infant’sbreathing behaviour that needs acareful evaluation.

N PSG is indicated when OSAS orcongenital central alveolarhypoventilation syndrome, ALTE,or sleep-related hypoventilationare suspected.

N On physical examination signscommonly associated with SDB,such as ‘‘adenoid facies’’, mayresult from oral breathing due toenlarged adenoids.


mouth or nose, vomiting, the relationship ofthe event to feeding and the degree ofresuscitation required. Finally, relevant pastmedical history (especially prematurity),immunisation status, family history ofsudden unexplained death in infancy orlater, exposure to smoke and metabolicdisease should be investigated. Heart andrespiratory rate and oxygen saturation mustbe recorded. Moreover, a careful neurologicexamination and neurobehaviouralassessment (e.g. head lag, posturing, motorabilities, eye tracking, social smile) shouldbe performed.

Sleep history and clinical assessment ofa child

Sleep disturbances such as bedtimestruggles, delayed sleep onset and multiplenight awakenings in children andadolescents are common and frequentlyassociated with both emotional andbehavioural difficulties in children andadolescents (such as anxiety, mooddisorders and attention deficit hyperactivitydisorders). Sleep disruption affects daytimefunctioning, with a significant impact on thefamily’s quality of life, and the presence ofSDB needs to be ruled out since the daytimeconsequence and neurobehaviouralsymptoms are the same as paediatricinsomnia (table 1). For this reason, adetailed clinical history and assessment of achild referred for sleep problems is relevant.Important points in the sleep history shouldinclude a complete sleep diary, recordingsleep patterns (bed-onset and sleep-onsettime, number of nocturnal awakenings) onweekdays, weekends and changes duringsummer. Nocturnal symptoms suggestive ofgastro-oesophageal reflux, asthma, or pain,presence of snoring or witnessed apnoeas(pauses in breathing) need to beinvestigated. The frequency and patterns ofsnoring, whether it is seasonal or related toany specific factor, such as upper airwayinfection, are of interest. On physicalexamination signs commonly associatedwith SDB need to be evaluated, such asobesity or being overweight (assessed byelevated age-appropriate BMI) andstructural features, known as ‘‘adenoid

facies’’, characterised by high arched andnarrow hard palate, increased facial heightand crossbite, which may result from oralbreathing due to enlarged adenoids. Nasalairflow should be assessed by holding a coldspatula or tongue depressor under the noseand asking the child to breathe. Misting ofthe cold metal indicates airflow. A lack ofairflow may indicate the presence ofnasopharyngeal obstruction resulting fromenlarged adenoids or nasal obstruction. Thenasal cavity may be examined, using anappropriate light, for the presence of rhinitis,obstructive polyps, enlarged turbinates ordeviated septum. When examining themouth, tonsillar size is usually graded from0 (in the case of small tonsils) to 4 (in thecase of enlarged tonsils obstructing at least75% of the lateral airway dimension).Moreover, Mallampati score, based onvisualisation of the airway on opening of themouth with tongue protruded, is helpful inidentifying a small airway, such as thepresence of macroglossia, and of a narrow

Table 1. Daytime and night-time symptoms suggestiveof SDB.

Night time symptoms

Snoring

Apnoeas

Nocturnal sweating

Nightmares

Sleep talking

Bruxism

Bedwetting

Sleepwalking

Restless sleep

Frequent arousals

Oral breathing

Daytime symptoms

Sleepiness

Hyperactivity

Inattention

Headache

Learning difficulties

Oral breathing


and arched palate. Looking at the bite, acrossbite or significant overjet or overbiteare signs of maxillary or mandibulardeficiency. Finally, a patient’s profile mayreveal retrognathia or micrognathia. Thepresence of glue ear may co-exist withadenoid hypertrophy. The blood pressureshould also be taken. Hypertensionsuggests complications of SDB.

Indications for respiratory and non-respiratory investigation

Further investigation should be performedon the basis of the history and physicalexamination: sleep questionnaires(investigating sleep pattern and symptomsof several sleep disorders) and sleep diary,are usually completed by the parents.

An objective assessment of sleep may beobtained using actigraphy, if sleep rhythmdisorders or restless legs syndrome and/orperiodic limb movement disorders aresuspected. An actigraph is an accelerometerthat records motion, generally worn on thenon-dominant wrist. The actigraph providesa noninvasive means to assess patterns ofactivity that reflect sleep/wake cycles acrossseveral consecutive days and nights.

PSG is indicated when the clinicalassessment suggests the diagnosis of OSASor congenital central alveolarhypoventilation syndrome, ALTE, sleep-related hypoventilation due toneuromuscular disorders or chest walldeformities. At this time, attended PSG isthe gold standard for the diagnosis of SDBas there is not sufficient evidence to suggestthe use of other diagnostic tools, such asunattended complete PSG or simple

cardio-respiratory recording. For the samereasons, even if home oximetry findings maybe highly specific for OSAS when it ispositive, it cannot alone replace PSGrecordings. When the sleep laboratory is notavailable, appropriate questionnaire,medical examination and home videooximetry is appropriate.

Video-PSG recording with an extended EEGchannel should be performed if a nocturnalseizure is suspected or if parasomnias,abnormal movements during sleep, arereported, and/or excessive daytimesomnolence is not related to a sleep apnoeasyndrome. The MSLT (that is used tomeasure the time elapsed from the start of adaytime nap period to the first signs ofsleep, also called sleep latency) should beperformed in all cases reporting daytimehypersomnolence, in particular, whennarcolepsy is suspected.

Further reading

N Aurora RN, et al. (2011). Practice para-meters for the respiratory indications forpolysomnography in children. Sleep; 34:379–388.

N Brodsky L. (1989). Modern assessment oftonsils and adenoids. Pediatr Clin NorthAm; 36: 1551–1169.

N Guilleminault C, et al. (2007). Adenoton-sillectomy and obstructive sleep apnoeain children: a prospective survey.Otolaryngol Head Neck Surg; 136: 169–175.

N Hall KL, et al. (2005). Evaluation andmanagement of apparent life-threateningevents in children. Am Fam Physician; 71:2301–2308.


Diagnostic techniques inchildren

Stijn Verhulst and Wilfried De Backer

Polysomnography

The following recommendations are madeaccording to the American Academy of SleepMedicine (AASM) Manual for the Scoring ofSleep and Associated Events (Iber et al., 2007).

Sensors

N EEG: the EEG electrode position isdetermined by the international 10–20system. The recommended derivationsare F4-M1, C4-M1 and O2-M1. Backupelectrodes should be placed at F3, C3, O1

and M2 to allow display of F3-M2, C3-M2

and O1-M2 if electrodes malfunctionduring the study. Alternative acceptablederivations are FZ-CZ, CZ-OZ and C4-M1

with backup electrodes at Fpz, C3 and O1.An initial EEG sensitivity of 7 mV?mm-1 isappropriate but the sensitivity oftenneeds to be adjusted in infants andyounger children to 10 or 15 mV?mm-1. Inthe last case, portions of the sleep

recording should be reviewed using7 mV?mm-1 to display and recognise lowvoltage faster frequencies.

N EOG: the recommended EOG derivationsare E1-M2 and E2-M2. E1 is placed 1 cmbelow the left outer canthus and E2 1 cmabove the right outer canthus. For smallerchildren, this distance could be reducedto 0.5 cm. Alternative derivations are E1-Fpz and E2-Fpz (the eye electrodes are nowplaced 1 cm below and lateral to the outercanthus of the left and right eye). Thealternative derivations record thedirection of eye movements.

N EMG: three electrodes should be placedto record chin EMG. One is placed in themidline 1 cm above the inferior edge ofthe mandible and the other two areplaced 2 cm below the inferior edge ofthe mandible 2 cm to the left and right ofthe midline. For smaller children, thisdistance should be reduced to 1 cm.

N A thermistor or nasal air pressuretransducer should be used todetect apnoea.

N A nasal air pressure transducer withoutsquare root transformation of the signalshould be used for detection ofhypopnoeas. A thermistor, end-tidalcarbon dioxide tension (PCO2) monitor orsummed calibrated inductanceplethysmography can also be used.

N Acceptable sensors for detection ofrespiratory effort are oesophagealmanometry or calibrated or uncalibratedinductance plethysmography.Oesophageal manometry is rarely used indaily practice because it is too invasivefor use in children.

Key points

N The scoring of sleep in infants andchildren is dependent on age andcentral nervous development.

N As infants and children express ahigher breathing rate, adult scoringrules for respiratory events do notapply to children. Specific scoringrules need to be used.

N Full PSG is the gold standard. Limiteddata exists on the usefulness ofscreening methods such as oximetryand cardiorespiratory monitoring.


N Pulse oximetry should be used with amaximal signal averaging time of 3 s.

N Transcutaneous or end-tidal PCO2

monitoring is used for assessingalveolar hypoventilation.

Scoring EEG in an infant/child Thesepaediatric sleep scoring rules can be used toscore sleep and wakefulness in childrenaged 2 months post-term and older. Sleep isdivided into the following stages in children:stages W (wakefulness), N1, N2, N3 and R.Stages N1, N2 and N3 are sleep stages ofNREM sleep. Stage R is a sleep stage ofREM sleep. Sleep is scored according to 30-sepochs commencing at the start of thestudy. A stage is assigned to each epoch. Iftwo or more stages co-exist during a singleepoch, the stage that comprises the greatestpart of the epoch should be assigned.

N Stage W: .50% of the epoch should havea rhythm or the age-appropriatedominant posterior rhythm over theoccipital region. The dominant posteriorrhythm is a dominant reactive EEGrhythm over the occipital regions inrelaxed wakefulness with eyes closed,which is slower in infants and youngchildren and attenuates with eye openingor attention. The frequency is 3.5–4.5 Hzwhen first seen in infants aged3–4 months post-term, 5–6 Hz by age5–6 months and 7.5–9.5 Hz by 3 yrs ofage with an amplitude of .50 mV. Inepochs without visually discernible arhythm or age-appropriate dominantposterior rhythm, stage W can be scoredwhen there are eye blinks at a frequencyof 0.5–2 Hz, reading eye movements orirregular conjugated eye movementsassociated with normal or high chinmuscle tone.

N Stage N1 can be scored if the posteriorrhythm is attenuated and replaced by lowamplitude, mixed-frequency activity for.50% of the epoch. In subjects who donot generate a posterior rhythm, stage N1is defined with the onset of EEG activityin the range of 4–7 Hz with slowing ofbackground frequencies by o1 Hz fromthose of stage W, vertex sharp waves,slow eye movements, rhythmic anterior h

activity, hypnagogic hypersynchrony or adiffuse or occipital-predominant, high-amplitude rhythmic 3–5 Hz activity.Hypnagogic hypersynchrony consists ofparoxysmal bursts or runs of diffuse highamplitude sinusoidal, 75–350-mV,3–4.5-Hz waves, which begin abruptly andare usually widely distributed, but oftenmaximally expressed, over the central,frontal or fronto-central scalp regions.

N Stage N2 starts if one or both of thefollowing occur during the first half of theepoch or the last half of the previousepoch: one or more K complexes notassociated with arousals or one or moretrains of sleep spindles. One cancontinue to score epochs with low-amplitude, mixed-frequency EEG activitywithout K complexes or sleep spindles aslong as it is preceded by K complexes notassociated with arousals or sleep spindles.

N Stage N3 can be scored when at least20% of an epoch consists of slow-waveactivity. Sleep spindles may persist inthis stage.

N Stage R can be scored in epochs with allof the following: a low-amplitude, mixed-frequency EEG, low chin EMG tone andrapid eye movements. One can continueto score epochs if the EEG continues toshow low-amplitude, mixed-frequencyactivity without K complexes or sleepspindles and if the chin EMG remains low.

N A major body movement is defined asmovement and muscle artefactsobscuring the EEG for more than half ofthe epoch to the extent that the sleepstage cannot be determined. An epochwith a major body movement should bescored as stage W if a rhythm is present,or if no a rhythm is discernible but anepoch scored as stage W either precedesor follows the epoch. Otherwise, theepoch should be scored as the samestage as the epoch that follows it.

N An arousal is defined as an abrupt shift ofEEG frequency including a, h and/orfrequencies .16 Hz (but not spindles)that lasts at least 3 s with at least 10 s ofstable sleep preceding the change.Scoring of an arousal during REM sleeprequires a concurrent increase insubmental EMG lasting at least 1 s.


The terms quiet and active sleep are alsoused when scoring infant sleep. Quiet sleepis characterised by trace alternant or high-voltage slow-wave activity. Trace alternant isan EEG pattern in which 3–8-s bursts ofmoderate-to-high voltage slow wavesintermixed with sharply contouredwaveforms alternate with 4–8 s intervals ofattenuated mixed-frequency EEG activity.High-voltage slow-wave activity consists ofcontinuous moderately rhythmic 50–150-mV,0.5–4-Hz slow activity without the burstingactivity of trace alternant. Active sleepconsists of either low-voltage 5–6-Hz EEGactivity called activite moyenne or a mixtureof high- and low-voltage activity including dactivity called mixed. The AASM task forcerecommends that sleep in infantso2 monthspost-term should be scored as NREM andREM, because all the EEG and PSG features ofREM sleep are present at this age and quietsleep, if not NREM sleep by this age, is at leastnot REM sleep. This generally holds for mostfull-term infants as well.

Sleep is undifferentiated before 32 weeksconceptional age (CA). At approximately32 weeks CA, rapid eye movements andphasic muscle twitches identify active sleepwhile quiet sleep is associated with thepresence of far fewer movements.Recognisable EEG patterns of active andquiet sleep appear around 34 weeks CA. By36 weeks CA, all the EEG and behaviouralcorrelates of wakefulness, active and quietsleep are clearly recognisable, although alarge proportion of sleep is scored asindeterminate sleep. This is defined whenmixtures of more than two sleep/wake statesare seen within an epoch. The percentage ofindeterminate sleep declines rapidly after36 weeks CA. Non-EEG correlates are oftenneeded in the distinction between quiet andactive sleep in young infants. Active sleep istypically associated with irregularrespiration, chin EMG atonia, transientmuscle activity and rapid eye movements.Regular respiration, no or vertical eyemovements and preserved chin EMG are allassociated with quiet sleep.

The various factors comprising NREM sleepneed to develop over time in infants. Sleep

spindles are present in infants aged2–3 months post-term or older, while Kcomplexes are usually present in infantsaged 4–6 months or older. Slow-waveactivity usually develops by 4–5 monthspost-term. On average, NREM sleep can bescored as stage N1, N2 and N3 by5–6 months post-term or older. In view ofthe variability of sleep in infants, fourpossible scenarios are described.

1. If all epochs of NREM sleep contain norecognisable sleep spindles, Kcomplexes or slow-wave activity, scoreall epochs of NREM sleep as stage N.

2. If some epochs of NREM sleep containsleep spindles or K complexes, scorethese epochs as stage N2. If there isno recognisable slow-wave activity inthe other epochs, score as stage N.

3. If some epochs of NREM sleep contain.20% of slow-wave activity, scorethese epochs as stage N3. If there areno recognisable K complexes or sleepspindles in the other epochs, score asstage N.

4. If NREM is sufficiently developed andsome epochs contain sleep spindles orK complexes and other epochs containsufficient slow-wave activity, thenscore NREM sleep according to thepaediatric rules.

Scoring breathing in an infant/child Thepaediatric rules can essentially be used inchildren f18 yrs of age. However, a sleepspecialist can decide to use the adult rulesin a child aged 13 yrs or older depending onthe clinical context. The following definitionsare used.

N Obstructive apnoea: the event lasts for atleast two missed breaths or the durationof two breaths as determined by thebaseline breathing pattern. The event isassociated with a .90% fall in the signalamplitude for o90% of the entirerespiratory event compared with thepre-event baseline amplitude, and theevent is associated with continued orincreased respiratory effort throughoutthe entire period.

N If an apnoea, as defined previously, isinitially associated with absent respiratory


effort followed by resumption of inspiratoryeffort before the end of the apnoea, then itis classified as a mixed apnoea.

N A central apnoea is associated withabsent inspiratory effort throughout theentire duration of the event with a periodof at least 20 s or of at least two missedbreaths, but associated with arousal,awakening or o3% desaturation. Centralapnoeas occurring after a snore, sigh,respiratory event or arousal are normalphenomena in children and are notscored unless associated with arousal,awakening or o3% desaturation.

N Hypopnoea is defined as a o50% fall inamplitude of the nasal pressure oralternative signal with the duration of atleast two missed breaths (or the durationof two breaths as determined by thebaseline breathing pattern). The fall inamplitude must last for o90% of theentire respiratory event and the event isassociated with an arousal, awakening oro3% desaturation.

N Respiratory effort-related arousal (RERA):there is a discernible fall in the amplitudeof signal from a nasal pressure sensorbut ,50% compared with the baselinelevel with a flattening of the nasalpressure waveform. Furthermore, it isaccompanied by snoring, noisy breathing,elevation in PCO2 or visual evidence ofincreased work of breathing which lasts atleast two breath cycles.

N Sleep-related hypoventilation is definedwhen CO2 levels are .50 mmHg asmeasured by transcutaneous or end-tidalCO2 sensors for .25% of total sleep time.

N Periodic breathing is defined if there aremore than three episodes of centralapnoea lasting .3 s separated by nomore than 20 s of normal breathing.

Normative data in an infant/child Neonatesexpress a high percentage of REM sleep,accounting for 50–60% of total sleep time.This percentage is even higher in prematureinfants. Infants can also enter sleep in REM.Total sleep time for neonates rangesbetween 16 and 20 h per day and theyexpress no day–night differentiation. A sleepcycle typically lasts 40 min. In the first yearof life, total sleep time decreases to 14 h per

day and day–night differentiation appearsfrom the age of 1 month onwards. Themajority of infants show a longerundisrupted sleep time of 3–4 h at 3 monthsand sleep ‘‘through the night’’ around9 months of age. They also have two napslasting 2–4 h. During the first months of life,the various sleep stages develop and sleepcan be completely scored at an average ageof 6 months. The percentage of REM sleepreaches 30% at around the age of 1 yr.

In the preschool period, total sleep timefurther diminishes to 12 h per day. Thesechildren usually have one nap per day, whichdisappears between the ages of 3 and 5 yrs.The percentage of REM sleep decreases(to 20–25%) and stage N3 increases. Thesleep cycle length slowly increases towardsadult levels.

School-age children have an average totalsleep time of 11 h and adolescents typicallyhave 9–10 h of sleep. REM latencyincreases. The percentage of stage N3increases until adolescence after which itsteadily decreases to adult levels.

In healthy infants, central apnoeasfrequently occur but are of short durationand are not followed by bradycardia oroxygen desaturation. REM sleep isassociated with a higher incidence of centralapnoeas. Obstructive and mixed apnoeasare less frequently seen. The optimaldefinition of an obstructive event,particularly hypopnoeas and RERAs has notbeen established in infants. Most studieshave reported only obstructive and mixedapnoeas, using thermistor-based studies.Obstructive apnoea or mixed apnoeaappears to be more common in prematureinfants and decreases in frequency over thefirst year of life. Studies have shown that, ingeneral, the obstructive apnoea index isbelow 1 event?h-1 in the first year of life bothin full-term and in pre-term infantsGuilleminault et al., 1979; Hoppenbrouwerset al., 1993). Obstructive events are associatedwith a much greater decline in oxygenationand heart rate compared with central apnoeasof equal lengths. During the first year of life,the number of central apnoeas decreases.Periodic breathing normally disappears in the


first 6 months of life. The median baselineoxygen saturation during sleep in a term infantat birth is approximately 98% and the medianlow is 83%. Short desaturations can beobserved in infants during periodic breathing,following normal respiratory pauses andduring REM sleep.

During childhood and adolescence, centralapnoeas can still occur although at a lowerfrequency that decreases with age.Obstructive and mixed events are rarelyseen. Mean saturation during sleep isapproximately 97% with a mean nadir o90%.

Interpretation and report of PSG in an infant/child Studies have shown that a single-nightPSG is sufficient to diagnose sleep apnoeain children. However, when sleeparchitecture is the primary outcome, twonights are often necessary to overcome thefirst-night effect. A PSG report shouldindicate the equipment and sensors used.Furthermore, it should indicate whether thechild is using any medication. It should startat the child’s regular bedtime. Time ofwaking up in the morning should beindicated so that the total time in bed can becalculated. The reported sleep parametersare: time in bed, total sleep time, sleepefficiency, sleep onset latency, wake aftersleep onset, REM latency, time andpercentage of total sleep time spent in stageN1, N2, N3 and REM, movement time,periodic limb movements and arousal index.Abnormal EEG readings should also bereported. Respiratory variables includenumber of central, obstructive and mixedapnoeas, hypopnoeas, RERAs and thepresence of snoring. Sleep-stage distributionand position-related distribution of theserespiratory events should be noted. Thenumber of desaturations with SaO2 nadirshould be recorded as well as the averagesaturation during sleep and a saturationhistogram. Similarly, transcutaneous or end-tidal PCO2 levels should also be recorded.Average, minimum and maximum heart rateand respiratory rate should also be calculated.

Diagnostic thresholds for OSA in childrenare based on normative PSG values inchildren. It is important to note that thesestudies differ in patient selection criteria and

sometimes use different definitions for thevarious respiratory events. Therefore, sleepcentres should keep this in mind whenselecting their diagnostic criteria. OSA inchildren is commonly diagnosed with anobstructive apnoea index o1 event?h-1 and/or an obstructive AHI o2 events?h-1.

Cardiorespiratory monitoring in children

Home-based and portable monitoringstudies offer a potentially very interestingapproach to study sleep and breathing inchildren in their home environment.Although a number of validation studies anda consensus statement have been publishedin the adult literature, data are scarce in thepaediatric field. The findings in adultscannot be extrapolated to children becauseof the marked differences in frequency andseverity of breathing pauses during sleep.One pilot study in 11 adolescents withDuchenne muscular dystrophy showed thatportable monitoring correctly identified thethree adolescents with an abnormal sleepstudy (Kirk et al., 2000). In spite of the lackof validation studies, several researchpapers have been published that have useda portable diagnostic system to study sleepand breathing in children. Therefore, noclear recommendations can be made on theclinical usefulness and cost-benefit ratio ofportable monitoring in the diagnosis of SDBin children.

Oximetry

Oximetry is used to detect sleep-relateddesaturations and hypoxaemia and is a veryeasy technique suitable for home monitoring.However, normal sleep oximetry does not ruleout OSA, because many obstructive eventsare not associated with desaturation inchildren. Therefore, it is only useful for theidentification of the subset of patients withmore severe OSA with marked desaturationsduring sleep. Furthermore, no large-scalestudies have been performed to identify otherpossible limitations. A validation studyreported a positive predictive value of ,100%in children with a history of snoring. However,the negative predictive value was very low(Brouillette et al., 2000). Therefore, in childrenwith a suggestive clinical history and physical


examination but with a normal oximetry, a full-night PSG is still necessary. Finally, oximetrydoes not rule out the contribution of centralapnoeas or other respiratory diseases tohypoxaemia. This also requires PSG. In recentyears, studies have emerged that havecombined oximetry with ECG signal andderived signals such as the pulse transit time,which could also allow the detection ofarousals and showed promising results.

Assessment of daytime sleepiness

MSLT and MWT: scoring rules and normativedata The MSLT is a valid and widely usedtest to measure the tendency to fall asleepand offers an objective measure of daytimesleepiness. MSLT is performed after a fullnight PSG that should show at least 6 h ofsleep. MSLT is performed by allowing fivenap opportunities lasting 20 min each at 2-hintervals. Sleep stages are scored accordingto standard rules. Parameters of interestinclude sleep latency defined as the timefrom lights out to stage N1. Sleep-onsetREM periods (SOREMPs) are defined byREM onset within 15 min of sleep onset. Themean sleep latency over the five naps is anobjective index of severity of sleepiness.

Normative values in children are less welldefined than in adults, but normal school-aged children typically remain awakethroughout the nap opportunity, ordemonstrate mean sleep latency in the15–20 min range. Prepubertal children havelonger mean sleep latency while adolescentshave the shortest latency. Two studies founda mean sleep latency .20 min inprepubertal children (Gozal et al., 2001;Palm et al., 1989). Therefore, 30-min naps inchildren have been proposed on a researchbasis. A sleep latency of 5–10 min indicatesmoderate daytime sleepiness, whereasf5 min indicates severe sleepiness. Thepresence of two or more SOREMPscombined with low mean sleep latency isconsidered diagnostic for narcolepsy.However, the physician should remain alertfor other factors that might disrupt sleeparchitecture including OSA, periodic limbmovement disorders or insufficient sleep.Further limitations of MSLT are the lack ofnormative data in paediatric patients, the

last nap effect (increased arousal because ofthe anticipation of almost going home) andthe influence of patient motivation.

The MWT is indicated to assess the efficacyof treatment in patients. It can be consideredas the opposite test of MSLT and measuresthe ability to stay awake for a defined time. Itconsists of four trials of a 40-min testprotocol performed at 2-h intervals with thefirst trial beginning 1.5–3 h after the patient’susual wakeup time. It has an identicalrecording montage as MSLT. Normal valuesof MWT are .30 min and the test isabnormal when it is ,12 min. Paediatricnormative values are not available.Motivation to stay awake is very important fora reliable interpretation of this test.

Further reading

N American Thoracic Society. (1996).Standards and indications for cardiopul-monary sleep studies in children. Am JRespir Crit Care Med; 153: 866–878.

N Brouillette RT, et al. (2000). Nocturnalpulse oximetry as an abbreviated testingmodality for pediatric obstructive sleepapnea. Pediatrics; 105: 405–412.

N Gozal D, et al. (2001). Objective sleepi-ness measures in pediatric obstructivesleep apnea. Pediatrics; 108: 693–697.

N Grigg-Damberger M, et al. (2007). Thevisual scoring of sleep and arousal ininfants and children. J Clin Sleep Med; 3:201–240.

N Guilleminault C, et al. (1979). Mixed andobstructive sleep apnea and near miss forsudden infant death syndrome: 2. Com-parison of near miss and normal controlinfants by age. Pediatrics; 64: 882–891.

N Hoppenbrouwers T, et al. (1993).Obstructive apnea, associated patternsof movement, heart rate, and oxygenationin infants at low and increased risk forSIDS. Pediatr Pulmonol; 15: 1–12.

N Hunt CE, et al. (1999). Longitudinalassessment of hemoglobin oxygensaturation in healthy infants during thefirst 6 months of age. CollaborativeHome Infant Monitoring Evaluation(CHIME) Study Group. J Pediatr; 135:580–586.


N Iber C, et al. (2007). The AASM Manualfor the Scoring of Sleep and AssociatedEvents: Rules, Terminology and TechnicalSpecifications. Westchester, AmericanAcademy of Sleep Medicine.

N Kahn A, et al. (1982). Effects of obstruc-tive sleep apneas on transcutaneousoxygen pressure in control infants, sib-lings of sudden infant death syndromevictims, and near miss infants: compar-ison with the effects of central sleepapneas. Pediatrics; 70: 852–857.

N Katz ES, et al. (2002). Night-to-nightvariability of polysomnography in childrenwith suspected obstructive sleep apnea. JPediatr; 140: 589–594.

N Kirk VG, et al. (2003). Comparison ofhome oximetry monitoring with labora-tory polysomnography in children. Chest;124: 1702–1708.

N Kirk VG, et al. (2000). Sleep-disorderedbreathing in Duchenne muscular dystro-phy: a preliminary study of the role ofportable monitoring. Pediatr Pulmonol;29: 135–140.

N Littner MR, et al. (2005). Practice para-meters for clinical use of the multiplesleep latency test and the maintenance ofwakefulness test. Sleep; 28: 113–121.

N Loughlin GM, et al. (2000). Sleep andBreathing in Children: A DevelopmentalApproach. New York, Marcel Dekker, Inc.

N Marcus CL, et al. (1992). Normal poly-somnographic values for children andadolescents. Am Rev Respir Dis; 146:1235–1239.

N Montgomery-Downs HE, et al. (2006).Polysomnographic characteristics in nor-mal preschool and early school-agedchildren. Pediatrics; 117: 741–753.

N Palm L, et al. (1989). Sleep and wakeful-ness in normal preadolescent children.Sleep; 12: 299–308.

N Scholle S, et al. (2011). Normative valuesof polysomnographic parameters in child-hood and adolescence: cardiorespiratoryparameters. Sleep Med; 12: 988–996.

N Traeger N, et al. (2005). Polysomno-graphic values in children 2–9 years old:additional data and review of the litera-ture. Pediatr Pulmonol; 40: 22–30.

N Uliel S, et al. (2004). Normal polysomno-graphic respiratory values in children andadolescents. Chest; 125: 872–878.

N Verhulst SL, et al. (2006). First nighteffect for polysomnographic data inchildren and adolescents with suspectedsleep disordered breathing. Arch DisChild; 91: 233–237.

N Verhulst SL, et al. (2007). Referencevalues for sleep-related respiratory vari-ables in asymptomatic European childrenand adolescents. Pediatr Pulmonol; 42:159–167.


Management of sleepdisordered breathing inchildrenAthanasios Kaditis, Maria Pia Villa, Anita K. Simonds and Ha Trang

Untreated OSA may be related to delayedsomatic growth, learning problems, daytimesleepiness, hyperactivity and adverse effectsfrom the cardiovascular system such ashypertension. Thus, it is of importance thatrespiratory disorders during sleep are

diagnosed and treated in order to preventthe development of complications.

Apnoeas in infants

Apnoea of prematurity and OSA unrelated toprematurity are the most common SDBpatterns in infancy that have been described.Apnoea of prematurity (gestational age,37 weeks) has been defined as pauses inairflow of at least 20 s in duration, or shorterbut accompanied by oxygen desaturation ofhaemoglobin and bradycardia. Apnoeas maybe of the central, obstructive or mixed typeand they usually disappear by 36–40 weekspostconceptional age.

It is unclear whether apnoea of prematurityhas long-term adverse effects, such asunfavourable neurodevelopmentaloutcomes, because of the many otherconfounding factors in the early pre-termperiod. After secondary causes of apnoeaare excluded (i.e. infection, intracranialhaemorrhage, hypoxaemic-ischaemicencephalopathy, seizure disorder, anaemia),potential therapeutic interventions are pronepositioning with neck elevation by 15u,administration of methylxanthine(theophylline, caffeine), CPAP and nasalintermittent positive pressure ventilation(table 1). Doxapram, kinaestheticstimulation and red blood cell transfusion incases of anaemia are among theinterventions that have been used for thetreatment of apnoea of prematurity withunclear benefit.

OSA not related to prematurity may be theresult of a variety of pathogeneticmechanisms, i.e. congenital craniofacialdeformities, nasal or laryngeal obstruction(choanal atresia, congenital nasal pyriform

Key points

N A variety of interventions are used forthe treatment of OSA in infancy, suchas methylxantines,adenotonsillectomy, orthodonticappliances, CPAP and tracheostomy.

N Adenotonsillar hypertrophy is afrequent cause of OSA in otherwisehealthy children, but fewer than 30%of them achieve a normal AHI(,1 event?h-1) afteradenotonsillectomy.

N Oral appliances and functionalorthopaedic devices are effective incases of maxillary constriction ormandibular retrusion withassociated OSA.

N Long-term home ventilation is requiredfor children with central hypoventilation.

N NIV in neuromuscular disordersshould be initiated when symptomaticnocturnal hypoventilation develops.

N NIV improves quality of life, morbidityand mortality in many stable orslowly progressive neuromusculardisorders, and may be considered topalliate symptoms in other moreprogressive conditions.


aperture stenosis, laryngomalacia, vocalcord paralysis, subglottic stenosis),neurological disorders (spinal muscularatrophy, cerebral palsy), gastro-oesophagealreflux and adenotonsillar hypertrophy (age.6 months). Acute viral upper respiratoryinfection may deteriorate the clinical pictureof all the above conditions temporarily.Endoscopic examination of the upper andlower airway is an important diagnostic toolto determine the exact site(s) of obstruction.

The severity of upper airway obstruction ininfancy requiring treatment has not beendefined. Moreover, there is not alwaysconsensus regarding the type of therapyindicated in each case, but a stepwisetreatment approach moving from the less tothe more invasive therapeutic intervention issummarised in table 1 according to thecause of airway obstruction.Adenotonsillectomy for OSA in infants has ahigher frequency of post-operativecomplications and residual SDB comparedwith the same procedure in older children. Itis accompanied by appreciable accelerationin somatic growth. The procedure has beenperformed in infants as young as 5 months

old, but most ENT surgeons would operateon children older than 1 yr. CPAP may be auseful stabilising measure but it is frequentlytolerated poorly and it is associated with sleepdisruption. In cases of severe laryngomalacia,removal of the lateral edges of the epiglottisand of redundant mucosa covering thearytenoids along with incision of thearyepiglottic folds can relieve upper airwayobstruction effectively (supraglottoplasty).Tracheostomy is applied to secure the upperairway while awaiting more permanenttherapeutic solutions.

Congenital craniofacial deformities cover awide and complex range of diseases, inparticular affecting newborns and infantsand resulting in narrowing of the airway dueto midface hypoplasia (e.g.craniosynostosis) and/or mandibularhypoplasia (e.g. Pierre Robin sequence).Congenital craniofacial deformities can beregarded as a model for sleep apnoea ininfancy, and their high association with OSAsuggests that PSG should be performedpromptly even if nocturnal symptoms arenot reported. Cleft lip and/or palate (CL/P) isa common congenital malformation

Table 1. A stepwise treatment approach for OSA in infants.

1 For history of prematurity (,37 weeks gestational age): exclude infection, intracranialhaemorrhage, hypoxaemic-ischaemic encephalopathy, seizures, anaemia and then useprone positioning and/or methylxanthine treatment; for mandibular hypoplasia: use pronepositioning or insert a nasopharyngeal tube (e.g. Pierre Robin sequence); for gastro-oesophageal reflux: antireflux medications.

2 For adenotonsillar hypertrophy (usually age .6 months): adenotonsillectomy (usually age.12 months).

3 For congenital mandibular hypoplasia (syndromic or non-syndromic): intraoral orthodonticappliance.

4 CPAP if: 1) history of prematurity (,37 weeks gestational age) and no response to pronepositioning or methylxanthine therapy; 2) OSA and major craniofacial deformities duringthe waiting period for a craniofacial procedure; and 3) residual OSA after a craniofacialprocedure.

5 For laryngomalacia: supraglottoplasty; for mandibular hypoplasia: surgical tongueadvancement (i.e. glossopexy via tongue–lip adhesion) or mandibular distractionosteogenesis; for midface hypoplasia: midface advancement (¡distraction osteogenesis).

6 Tracheostomy if treatment modalities 1–5 are not adequate or to secure the upper airwaywhile awaiting surgical intervention.

Use nocturnal PSG or oximetry at each step in order to move to the next step if residualOSA is present.


affecting approximately two per 1,000newborns worldwide. The Pierre Robinsequence is a well-defined subgroup of theCL/P population, characterised additionallyby micrognathia, glossoptosis and upperairway obstruction and it has a prevalence ofone per 8,500 newborns. Pierre Robinsequence is often part of other Mendeliansyndromes such as Stickler’s,velocardiofacial and Marshall’s syndromes,or as part of carpomelic dysplasia which is arare skeletal dysplasia.

The severity of clinical manifestations inpatients with Pierre Robin sequence mayvary from 1) normal breathing and feeding;to 2) normal breathing but with feedingdifficulties; and 3) respiratory failure withneed for intervention and feeding difficultiesnecessitating nutrition via a nasogastric tube.The most severely affected infants with PierreRobin sequence may suffer from intermittenthypoxaemia, hypercapnia, cor pulmonale,failure to thrive, neurodevelopment delay oreven sudden death.

Current treatment options for OSAassociated with Pierre Robin sequenceinclude: 1) prone positioning; 2) insertion ofa nasopharyngeal tube (commonly usedintervention); 3) use of an intraoralorthodontic appliance; 4) surgical tongueadvancement (i.e. glossopexy via tongue–lipadhesion); 5) mandibular distractionosteogenesis; and 6) tracheostomy. Infantswith minor manifestations can be managedby placement in the prone position, takingadvantage of the effect of gravity on movingthe tongue forwards. Use of glossopexy hasbeen declining, since it does not alwaysachieve complete resolution of airwayobstructive symptoms. An effective,noninvasive and safe treatment modalityinvolves implementation of an intra-oralorthodontic appliance with velar extension.Craniofacial procedures for mandibularhypoplasia are discussed later.

Craniosynostosis is a congenital disorderaffecting one in 2,500 newborns that ischaracterised by premature fusion of thecalvarial sutures. Premature fusion restrictsnormal growth of the skull, brain and face,necessitating surgical correction. In about

40% of cases, craniosynostosis is part ofsyndromes such as Apert, Crouzon, Pfeiffer,Muenke or Saethre–Chotzen. Almost 50% ofchildren with Apert, Crouzon or Pfeiffersyndrome develop OSA during the first 6 yrsof life. Increased risk for OSA is usually theresult of midface hypoplasia, but otherfactors such as adenotonsillar hypertrophyor even mandibular hypoplasia may alsocontribute to upper airway obstruction.Surgical procedures for the treatment ofmidface hypoplasia are discussed later.

OSA

The following treatment options areavailable.

Tonsillectomy and/or adenoidectomy and otherENT procedures Surgical excision of thehypertrophic pharyngeal and palatine tonsils(adenoidectomy and tonsillectomy,respectively) is the standard treatment forOSA in childhood, which reduces upperairway resistance and the tendency of thepharyngeal airway to collapse. It is unknownwhether adenoidectomy only, tonsillectomyonly, partial tonsillectomy or totaladenotonsillectomy should be performed toachieve complete resolution of OSA. Indeed,no randomised, controlled trials on theefficacy of adenotonsillectomy have beenpublished, although such studies are inprogress. A retrospective, multicentre studyhas used PSG indices as primary end-pointsto evaluate the efficacy of adenotonsillectomyin children with OSA and without associatedanomalies (e.g. genetic syndromes orneuromuscular disorders). Postoperatively,significant improvements in AHI, respiratoryarousal index and oxygen saturation ofhaemoglobin nadir have been demonstrated.However, a normal AHI (,1 episode?h-1) wasachieved in fewer than 30% of all studiedsubjects. Potential reasons for residual OSA(AHI o1 episode?h-1) postoperatively includenasal mucosa inflammation (e.g. allergicrhinitis), co-existent obesity, craniofacialabnormalities or neuromuscular disorders.Age greater than 7 yrs is an additional riskfactor for postoperative persistence of OSA.

The operation may also be accompanied bybeneficial effects on cognition, attention,


behaviour and quality of life. It should benoted that a subgroup of children exhibitrapid weight gain after adenotonsillectomyand ultimately partial recurrence of OSA,most likely due to improved food intake,reduced nocturnal work of breathing orhormonal changes. Adenotonsillectomy isnot devoid of complications such asanaesthetic risks, postoperative pain, airwaycompromise and bleeding and in very rarecases even death.

In addition to adenotonsillectomy,uvulopalatopharyngoplasty has been appliedto children with Down syndrome or cerebralpalsy and to subjects with major craniofacialanomalies. Only case reports and small caseseries on the efficacy of the procedure havebeen published. Therefore, its usefulness inthe management of paediatric OSA is highlycontroversial. Other surgical interventionsthat may decrease nasal resistance, such asradio-frequency treatment of the inferiorturbinates, have not been studiedsystematically and they are not recommended.

Intranasal corticosteroids administered for4–8 weeks to children with mild OSA andadenoidal hypertrophy improve symptomsof nasal obstruction and PSG indices. Theiruse results in reduction of the upper airwayresistance and severity of obstructive SDBvia a decrease in adenoidal tissue volume.Topical corticosteroids have also been usedwith benefit for those children with residualOSA after adenotonsillectomy.

Craniofacial procedures Craniofacialprocedures are part of the treatment plan forcongenital craniofacial deformities (mainlymidface and/or mandibular hypoplasia)which affect respiration and feeding.Mandibular distraction osteogenesis hasrecently acquired an important position inthe management protocol of mandibularhypoplasia. With this procedure, the tongueand the suprahyoid muscles are movedforwards and the size of the pharyngealairway increases due to gradual lengtheningof the mandible.

In patients with Apert, Crouzon or Pfeiffersyndromes and moderate-to-severe OSA,neither tonsillectomy nor adenoidectomy

have a significant effect on respiration, whilemidface advancement appears to be thetreatment of choice. Co-existent anatomicalmalformations of the nasopharynx arecommon, causing functional airwayobstruction at this level. Midfaceadvancement usually improves respirationin the short term, but long-term efficacy ofthe procedure, defined as no need forrespiratory support for at least 2 yrspostoperatively, is not as clear as wasthought previously. Relapse of OSA aftersurgery may be due to overcorrection of themidface hypoplasia with change ofmalocclusion from class III to class II.Moreover, collapse of the pharyngeal airwayis a functional abnormality which does notnecessarily improve after the anatomiccorrection achieved by midface advancement.

Orthodontic procedures This section isfocused on the management of mildercraniofacial abnormalities predisposing toOSA compared with the profoundcongenital craniofacial deformitiesdiscussed previously. A narrow upper airwayaccompanied by maxillary constriction andmandibular retrusion is a commonphenotype related to paediatric OSA.Mandibular retroposition predisposes tocollapse of the upper airway during sleepand it is associated with posteriordisplacement of the tongue base whichresults in further narrowing of the upperairway and a high-arched (ogival) palate(fig. 1). The reversibility of these craniofacialabnormalities by means ofadenotonsillectomy for co-existentadenotonsillar hypertrophy has yet to beestablished, and orthodontic treatmentbased on oral appliances is a validadditional treatment for paediatric OSA.

Oral appliances and functional orthopaedicdevices have been used in children to shiftthe mandible forwards, to enlarge the upperairway and to improve respiration. Rapidmaxillary expansion (RME) is a dento-facialorthopaedic procedure for the treatment ofconstricted maxillary arches and maxillarytransverse deficiencies, commonly used inyoung patients (.4 yrs of age). Suchpatients usually display unilateral or bilateral


posterior crossbite and anterior dentalcrowding. The distance between the lateralwalls of the nasal cavity and the nasalseptum is often reduced, leading toincreased nasal resistance to airflow andconsequently to nasal respiratory difficulties.RME may relieve nasal obstruction byincreasing the transverse dimensions of themaxilla, which in turn widens the nasal cavity.

The efficacy of RME has been demonstratedin non-obese children suffering from nasalbreathing and OSA, but without enlargedtonsils or adenoids. 4 months aftercompletion of the orthodontic treatment, allchildren had normal anterior rhinometry anda significant decrease in AHI. The effects ofRME were also evaluated in another groupof non-obese children with OSA and dentalmalocclusion i.e. ogival palate associatedwith deep bite, retrusive bite or crossbite.RME was applied for 12 months. By the endof treatment, AHI decreased significantlycompared with baseline, reaching normalvalues in most patients. An improvement inclinical symptoms was reported by parents,such as less snoring, oral breathing, sleepapnoeas, and daytime sleepiness andtiredness. Beneficial effects of RME persistedeven 2 yrs after its completion. In summary,results of these studies suggest thatorthodontic therapy should be encouraged inpaediatric OSA, and that its earlyimplementation may permanently modifynasal breathing and respiration, therebypreventing obstruction of the upper airway.

Oropharyngeal exercises derived fromspeech therapy (myofunctional therapy) areanother important treatment modality.Oropharyngeal exercises may be an effectivetreatment option for children with OSA,because they adjust physiological breathingand eliminate oral breathing, both of whichare involved in upper airway muscle functionand airway patency.

CPAP treatment Children with OSA andobesity, those with residual disease afteradenotonsillectomy or sleep apnoeicchildren with cerebral palsy or craniofacialabnormalities are candidates for CPAP,which is usually effective in amelioratingapnoeas and hypopnoeas. During sleep, thechild receives continuous airflow deliveredby the CPAP ventilator via a mask (nasalmask in most cases or facial mask) in orderto maintain positive airway pressure.Compliance with this type of treatment canbe improved dramatically by gradualacclimatisation and introduction of themask as the child sleeps, as well as bycontinued gentle and intensive familysupport. The pressure level that is necessaryto ‘‘stent’’ the pharyngeal airway so as toprevent the collapse of its walls is determinedin the sleep laboratory prior to prescribing theCPAP ventilator. The recommended startingpressure level during the titration procedureis 4 cmH2O and the goal of treatment withCPAP is to lower AHI below 5 episodes?h-1. Inpatients with OSA and concomitant alveolarhypoventilation due to neuromuscular

Figure 1. Examples of common craniofacial phenotypes of children with OSA. 1: a high-arched palate andoral breathing; 2: a narrow maxilla, retrognathia; 3: hypotonic lips; and 4: increased lower facial height.


disorder or chronic lung disease, noninvasive,bi-level, positive pressure ventilation can beused instead of CPAP.

Lifestyle management A balanced diet andregular physical exercise should beimplemented in all children with OSA andespecially in those who are obese. Limitedevidence indicates that weight loss by asmuch as 35% in children with OSA andsevere obesity can be accompanied by animprovement in the severity of intermittentupper airway obstruction during sleep.However, it should be noted that such adegree of weight reduction is difficult toachieve and sustain. Avoidance of exposureto cigarette smoke and indoor allergens areadditional measures that could reduceupper airway inflammation, resistance toairflow and the tendency for pharyngealairway collapse.

Indications

There are no long-term, prospective studiesto evaluate whether OSA symptoms,abnormal PSG findings, OSA-relatedmorbidity or any of their potentialcombinations are indications for treatment.AHI .5 episodes?h-1 in paediatric patientsthat reflects moderate-to-severe intermittentupper airway obstruction during sleep,decreases after adenotonsillectomy insubjects with adenotonsillar hypertrophy.Although AHI .5 episodes?h-1 has beenclearly related to an increased risk formorbidity, even an AHI ,1 episode?h-1 withno apnoeas or gas-exchange abnormalities(primary snoring) has been associated withexcessive daytime sleepiness and learningproblems. When treatment decisions for mildOSA (AHI 1–5 episodes?h-1) are made, itshould be taken into consideration that SDBin certain children may persist in adolescence.

Despite the low methodological quality ofpublished evidence on the value of treatingobstructive SDB and the absence ofconsensus between practitioners, someindications for therapeutic intervention canbe summarised as follows:

1. Moderate-to-severe OSA (AHI.5 episodes?h-1) irrespective of thepresence of morbidity.

2. Mild OSA (AHI 1–5 episodes?h-1), butwith OSA-related morbidity (e.g.enuresis, inadequate somatic growth,poor academic performance,inattention, hyperactivity, pulmonaryhypertension, or systolic or diastolicblood pressure .95th percentile forsex, age and height).

3. OSA and increasing BMI percentileespecially in pre-adolescent males(increased risk for persistent OSA).

4. OSA associated with neuromusculardisorders or craniofacial anomalies(increased risk for development ofpulmonary hypertension).

There are no data supporting the treatmentof children with primary snoring or upperairway resistance syndrome.

Treatment of OSA may be followed by: 1)improvement in quality of life; 2) appreciableweight and height gain; 3) resolution ordecrease in the frequency of enuresis; 4)decrease in systemic blood pressure andpulmonary artery pressure and reversal ofcor pulmonale; 5) less daytime sleepiness,hyperactivity and aggression; and 6)reduction in healthcare utilisation.

As discussed previously, the availabletreatment options for OSA include: anti-inflammatory medications and weightcontrol, adenotonsillectomy, orthodonticdevices, CPAP, craniofacial procedures andtracheostomy. If a child with OSA iscandidate for treatment, therapeuticinterventions should address allabnormalities predisposing to upper airwaydysfunction. For example, a child withadenotonsillar hypertrophy and retrusion ofthe mandible may benefit from thecombination of adenotonsillectomy andapplication of an orthodontic device. Nolarge, randomised controlled trials havebeen conducted to define the order ofimplementation of available treatmentmodalities. At present, a stepwise treatmentapproach from the less to the more invasivetherapeutic interventions is proposed(table 2). After completion of one step, thepatient should be re-evaluated and the nexttherapeutic option should be applied to treatresidual OSA.


Sleep hypoventilation syndromes

Treatment options Either due to central or toperipheral causes, hypoventilation is alwaysmost severe during sleep (see ‘‘Sleep-disordered breathing in children’’).Ventilatory support aims to provide optimalventilation, i.e. normal PaO2 and normalPaCO2. Long-term home setting ventilatorysupport allows children to live and grow upat home with families. Positive pressureventilation via tracheostomy, bi-levelpositive pressure ventilation via nasal orfacial mask and respiratory pacing are themain options for these patients. Negativepressure ventilation is rarely used.

Positive pressure ventilation via tracheostomyCommercially available positive pressureventilators are adequate for a long-termhome setting, equipped with battery securityand easily portable. However, not allpositive pressure ventilators are fitted forinfants and young children who requiredelivery of small tidal volumes. Particularattention should be paid to maintain aventilator circuit appropriate for the child’ssize and able to deliver heated andhumidified air to the patient.

Management of tracheostomy should becarefully carried out at initial presentationand during regular follow-up at leastannually at a centre with extensiveexperience in the care of these children. It isrecommended to use a tracheostomycannula smaller than the airway diameter inorder to avoid traumatic lesions of theairway wall, but enable adequate mechanicalventilation during sleep.

Noninvasive bi-level positive pressureventilation Noninvasive bi-level positivepressure ventilation (NIV) is delivered via anasal mask or facial mask using bi-levelpositive pressure ventilators. Theseventilators use a blower to produce avariable continuous flow and cancompensate for leaks occurring around themask or in case of infection-related reducedlung compliance. The timed mode ensures aminimal number of breaths delivered toinfants and young children who are unableto trigger the ventilator. Some ventilators arenow equipped with a volume preset-pressure mode. Mask ventilation has beenassociated with mid-face hypoplasia whenused in infants or young children. Close

Table 2. A stepwise management approach for children with OSA who have indications for treatment.

1 Nasal corticosteroids for adenoidal hypertrophy and AHI ,5 episodes?h-1; may also be usedfor residual OSA following adenotonsillectomy

2 Adenotonsillectomy for adenotonsillar hypertrophy

3 Orthodontic devices for craniofacial abnormalities (e.g. oral appliances for mandibularretroposition or rapid maxillary expansion for constricted maxillary arches and maxillarytransverse deficiencies). Orthodontic treatment may be used for residual OSA afteradenotonsillectomy (AHI .5 episodes?h-1)

4 Nasal CPAP if AHI .5 episodes?h-1 and: 1) residual OSA after adenotonsillectomy; 2) OSArelated to obesity; 3) OSA related to neuromuscular disorders or craniofacial abnormalitiesand unresponsive to measures 1–3; 4) major congenital craniofacial deformities whileawaiting a craniofacial procedure

5 Craniofacial surgery if treatment modalities 1–4 are not adequate. Craniofacial proceduresinclude mandibular distraction osteogenesis for micrognathia and midface advancement(¡ distraction osteogenesis) for midface hypoplasia

6 Tracheostomy if treatment modalities 1–5 are not adequate or to secure the upper airwaywhile awaiting surgical intervention

Weight control based on a balanced diet and regular exercise is recommended for allchildren with OSA and especially for those who are overweight.

Use nocturnal PSG or oximetry at each step in order to move to the next step if residualOSA is present.


follow-up by a multidisciplinary team,including a paediatric orthodontist ormaxillofacial surgeon is highly recommended.

Negative pressure ventilation Negativepressure ventilators apply a negativepressure outside the chest and abdomenwith a shell or a cuirass, allowing expansionof the chest and effective ventilation inchildren and adolescents.

Respiratory pacing Respiratory pacinggenerates breathing by electricallystimulating the patient’s diaphragm. Abattery-operated external device producestrains of electrical pulses that aretransformed into a radiofrequency signaland transmitted to internally implantedreceivers. The latter are connected toelectrodes either inserted around thephrenic nerves (phrenic nerve pacing) ordirectly implanted into the diaphragmmuscle (diaphragm pacing). Phrenic nervepacing requires a normally functioningphrenic nerve-diaphragm axis. Thistechnique has been shown to provideefficient ventilator support. In contrast,direct diaphragm pacing is a newertechnique that needs to be more extensivelyassessed in adults and is still scarcely usedin children in a research setting at this time.The surgical process and initiation of thepacing system should be made by amultidisciplinary paediatric team with extensiveexpertise in diaphragm pacing in children.

Indications

Central hypoventilation Ventilatory supportis a life support in patients with congenitalcentral hypoventilation syndrome (CCHS) orcentral hypoventilation secondary todifferent causes (see ‘‘Sleep-disorderedbreathing in children’’). Hypoventilation ismost severe during sleep, but may occuralso during daytime in some patientsseverely affected.

The main objective of ventilatory support incentral hypoventilation is to provide optimalventilation 24/7 beginning in the first days ofdiagnosis of these patients in order topreserve optimal neurocognitive outcome.In patients with CCHS, use of positivepressure ventilation via tracheostomy is

recommended in the first several years oflife, allowing early discharge to home(American Thoracic Society). After 6–8 yrsof age, most children with CCHS whorequire sleep ventilation only can besuccessfully transitioned from positivepressure ventilation via tracheostomy toNIV. It should be noted that directcomparisons between noninvasive positivepressure and invasive (trachestomy-delivered) ventilation have not been carriedout in CCHS.

NIV has been reported as an initialtreatment performed in some infants withCCHS. However, again there are no data orsystematic studies on the long-termoutcome in patients receiving NIV as initialand sole mode of ventilation.

One must keep in mind that oxygenadministration without ventilatory support isinadequate in patients with centralhypoventilation as it increases oxygensaturation, but worsens hypoventilation,subsequently producing pulmonaryhypertension or coma.

Respiratory pacing may be recommended toactive children with central hypoventilationwho require ventilatory support day and night.These children may combine different modesof ventilatory support, receiving positivepressure ventilation during night-time andusing their respiratory pacers during daytime,so that they are totally free from the ventilatorduring their diurnal activities.

Regardless of the mode of ventilation,patients with central hypoventilation requirecontinuous monitoring with pulse oximetryand end-tidal carbon dioxide tensionduring nighttime.

Neuromuscular disorders The probability ofdeveloping nocturnal hypoventilation inneuromuscular diseases depends on thediagnosis, age and extent of respiratorymuscle weakness. For example, in type 1spinal muscular atrophy (SMA), respiratoryinsufficiency is likely in the first few monthsof life, whereas in Duchenne musculardystrophy, SDB is not usually seen until thelate teenage years. Sleep studies should be apart of the regular follow-up, and for most


authorities the indication is to start NIV inthe presence of symptomatic nocturnalhypoventilation, even if daytime PaCO2 isnear normal. Successful NIV use isassociated with resolution of symptomssuch as poor sleep, headaches,concentration/cognitive difficulties, andimportantly results in decreased morbidityand increased survival. In children withsevere progressive respiratory failure (e.g.due to type I SMA) NIV may be used topalliate symptoms and facilitate dischargehome. The choice of ventilator is important.As indicated above, bi-level pressuresupport is most commonly used, butattention to ventilator performance such astrigger function is critical. The risk ofmidfacial hypoplasia induced by NIV usecan be reduced by the use of customisedmasks, rotation of interfaces and applicationof nasal plug interfaces. Many children withneuromuscular disease can be managedwith NIV throughout the course of theirillness. Ventilator settings should bechecked regularly with overnight monitoringincluding SpO2 and PCO2 control, plus theuse of ventilator downloads to assessoptimal synchrony with the ventilator. Thecombination of NIV with cough assistanceusing mechanical in-exsufflation to controlbronchial secretions can help reduce chestinfections and their complications. Thecough in-exsufflator has not been subject torandomised controlled trials but clinicalbenefit is evident from case series. The useof cough assistance and NIV has reducedthe need for tracheostomy ventilation inneuromuscular disease.

Current indications for tracheostomyventilation include:

N Severe bulbar weakness leading toaspiration

N Upper airway problemsN Near 24-h ventilator dependencyN Failure to control ventilation with

noninvasive modeN Intractable interface problemsN Patient/family choice

References

N Arens R, et al. (2010). Sleep, sleepdisordered breathing, and nocturnalhypoventilation in children with neuro-muscular diseases. Paediatr Respir Rev; 11:24–30.

N Anderson IC, et al. (2010). Prevalence andseverity of obstructive sleep apnea andsnoring in infants with Pierre Robinsequence. Cleft Palate Craniofac J; 48:614–618.

N Bhattacharjee R, et al. (2011).Adenotonsillectomy outcomes in treat-ment of obstructive sleep apnea inchildren: a multicenter retrospectivestudy. Am J Respir Crit Care Med; 182:676–683.

N Chen ML, et al. (2005). Diaphragmpacers as a treatment for congenitalcentral hypoventilation syndrome. ExpertRev Med Devices; 2: 577–585.

N Kaditis A, et al. (2012). Algorithm for thediagnosis and treatment of pediatricOSA: a proposal of two pediatric sleepcenters. Sleep Med; 13: 217–227.

N Katz ES, et al. (2012). Obstructive sleepapnea in infants. Am J Respir Crit CareMed; 185: 805–816.

N Villa MP, et al. (2011). Mandibularadvancement devices are an alternativeand valid treatment for pediatric obstruc-tive sleep apnea syndrome. Sleep Breath;[Epub ahead of print DOI: 10.1007/s11325-011-0595-9].

N Villa MP, et al. (2011). Efficacy of rapidmaxillary expansion in children withobstructive sleep apnea syndrome: 36months of follow-up. Sleep Breath; 15:179–184.

N Weese-Mayer D, et al. (2010). An officialATS clinical policy statement: congenitalcentral hypoventilation syndrome. Am JRespir Crit Care Med; 181: 626–644.

N Zhang L, et al. (2008). Intranasal corti-costeroids for nasal airway obstruction inchildren with moderate to severe adenoi-dal hypertrophy. Cochrane Database SystRev; 3: CD006286.

N Zhao J, et al. (2011). Apnea of prematur-ity: from cause to treatment. Eur J Pediatr;170: 1097–1105.


Index

Aaccidents motor vehicle see motor vehicle accidents (MVAs) workplace, increased risk in OSA patients 191acetazolamide therapy central sleep apnoea 165 obstructive sleep apnoea 154acetylcholine sweat-spot test 14achondroplasia 206acid maltase defi ciency 50acromegaly, SDB association 96, 98actigraphy 143 in children 220 circadian rhythm evaluation 74, 75 free-running circadian disorder 76 insomnia assessment 57active sleep (AS) 200 polysomnography 223acute insomnia 55adaptive servo ventilation (ASV) 162, 165, 166–167“adenoid facies” 219adenoid hypertrophy 219, 220, 229adenotonsillectomy 152, 230–231 infants 229ADHD see attention defi cit hyperactivity disorder (ADHD)adherence, CPAP therapy 160–161adipose tissue 37adjustment insomnia 55advanced sleep phase syndrome (ASPS) 75, 76age eff ect on sleep cycle 3–4 as obstructive sleep apnoea risk factor 26, 95algorithms diagnostic 100–105 nocturnal hypoventilation management neuromuscular/skeletal disorders 172 obese patients 169, 171 Oxford 160 for setting CPAP level 159–160allergic rhinitis, in children 211Alzheimer’s disease 62, 64ambulatory blood pressure monitoring (ABPM) 115–116American Academy of Sleep Medicine (AASM) electrode placement 120, 121 home diagnosis recommendations 100, 137 hypopnoea scoring 141 manual for scoring of sleep 120, 126 sleep disordered breathing defi nitions 21, 23aminophylline 154amitriptyline 180, 182amphetamines hypersomnia treatment 183–184 narcolepsy 71 side-eff ects 184amyotrophic lateral sclerosis (ALS) 50, 144, 170, 172anaesthesia risk assessment, obstructive sleep apnoea 111–113

ANS (autonomic nervous system) 13 sleep-related changes see under sleep anticonvulsant drugs, insomnia treatment 180, 182antidepressants cataplexy management 70, 71, 185 insomnia management 180, 181, 182 OSA treatment 154antihistaminics 182antihypertensives 155antipsychotics, insomnia treatment 180, 182APAP (autotitrating positive airway pressure) devices 139,

141, 157, 158Apert syndrome 206, 230, 231apnoea central see central apnoea defi nition 21 in infants see under infants obstructive see obstructive apnoea presentation 22apnoea of prematurity 228apnoea threshold 97 central sleep apnoea 43apnoea–hypopnoea index (AHI) 21 upper airway anatomy and 7, 8apparent life-threatening events (ALTEs) 203–204 defi nition 203 sleep history 218–219appetite regulation, sleep and 18arousal disorders 63–64 classifi cation 63 diagnosis 64 epidemiology 63 pathophysiology 63 treatment 64, 187arousals central sleep apnoea pathophysiology 43–44 EEG 126 in children 222arterial blood gas (ABG) sampling 143, 144arterial hypertension see hypertensionasphyxia, infants 201ASPS (advanced sleep phase syndrome) 75, 76asthma, in children 210ASV (adaptive servo ventilation) 162, 165, 166–167atherosclerosis, assessment in OSA patients 117, 118, 155atrial fi brillation CSA patients 40 OSA patients, assessment 116–117, 118attention defi cit hyperactivity disorder (ADHD) 213, 215 restless legs syndrome and 215–216 sleep disordered breathing and 216autonomic nervous system (ANS) 13 sleep-related changes see under sleep autonomic ‘stress tests’ 14autotitrating positive airway pressure (APAP) devices 139,

141, 157, 158

237

Bbariatric surgery OSA patients 111, 112, 152–153 benefi cial eff ect 153 techniques 153Beck Depression Inventory (BDI) 109behavioural disorders, paediatric OSA patients 208behavioural insomnia, infants 218benzodiazepine therapy arousal disorders 64 eff ect on sleep structure 179 insomnia 179–181 restless legs syndrome 186Berger, Hans 1Berlin Questionnaire 32–33, 92, 93, 112, 113bi-level positive airway pressure ventilation 157, 158, 162 central sleep apnoea treatment 165, 166 nocturnal hypoventilation management in obstructive

lung disease 173, 174 sleep hypoventilation syndrome management neuromuscular/skeletal disorders 171–173 obesity hypoventilation syndrome 169, 171 sleep hypoventilation syndrome management in children

234–235 central hypoventilation 235 neuromuscular disorders 236biliopancreatic derivation technique 153blood pressure monitoring 14–15 ambulatory 115–116breathing control of see breathing, control of monitoring 120 periodic see periodic breathing during sleep see under sleep sleep history 85 at wake/sleep transition see wake/sleep transition see also entries beginning respiratorybreathing, control of 6–7 development 201–202 asphyxia and 201 breathing and sleep 202 carbon dioxide and 201 hypoxia and 201 plasticity 202 sleep and 202 upper airway 201–202 obstructive sleep apnoea syndrome pathophysiology 30 during sleep 6–7 at wake/ sleep transition 7breathing, sleep disordered see sleep disordered breathing

(SDB)bronchiectasis 174bronchopulmonary dysplasia (BPD) 212Brouillette questionnaire 207

CCAD see coronary artery disease (CAD)Calgary Sleep Apnea Quality of Life Index (SAQLI) 109capnography, nocturnal 143–144carbon dioxide application, central sleep apnoea 165carbon dioxide sensitivity, infants 201cardiac function, assessment in OSA patients 116–117, 118cardiac hypertrophy, assessment in OSA patients 116, 118cardiac resynchronisation therapy 165cardiac structure, assessment in OSA patients 116–117, 118cardiorespiratory monitoring children 225 OSA patients see under obstructive sleep apnoea

syndrome (OSAS) cardiorespiratory polygraphy 137 interpretation 140–142, 143 scoring 138–140cardiovascular changes, sleep-related see under sleepcardiovascular disease (CVD) central sleep apnoea and 40, 46–47, 98 obstructive sleep apnoea and 35–37, 98, 155cataplexy, narcolepsy with 67, 68, 69, 70, 86 management 70, 71, 183, 184–185 prevalence 213 symptoms 213–214 see also narcolepsycentral apnoea in childhood/adolescence 225 defi nition 224 healthy infants 224 see also central sleep apnoea (CSA)central hypoventilation syndromes 7, 48–49 management 235 see also congenital central hypoventilation syndrome

(CCHS)central respiratory event, defi nition 124central sleep apnoea (CSA) 21, 22–23, 39–47 clinical aspects 22–23, 45 consequences 46–47 cardiovascular 46–47 neurobehavioural 46 defi nition 21, 39 obstructive sleep apnoea vs 39 pathophysiology 21, 42–44 apnoea threshold 43 arousals 43–44 hyperventilation 42 hypoventilation 42 hypoxic and hypercapnic ventilatory response 43 loop gain 42–43 prevalence 40 risk factors 40 see also sleep disordered breathing, predisposing factors symptoms 22–23, 45 treatment 164–167 adaptive servo ventilation 162, 165, 166–167

238

carbon dioxide 165 CSA associated with hyperventilation 164–165 CSA associated with hypoventilation 164 oxygen 165 pharmaceutical options 165 positive airway pressure treatment 47, 165–166cephalometry 8cerebrospinal fl uid (CSF) analysis, narcolepsy 68, 69chemosensitivity 6, 7 sleep-related reduction 10, 11chest wall disorders 50, 170 see also neuromuscular/skeletal disordersCheyne–Stokes respiration (CSR) 23, 39, 40, 98 see also central sleep apnoea (CSA)children clinical assessment 218–220 child 219–220 indications for further investigation 220 infant 218–219 physical examination 218, 219 development control of breathing see under breathing, control of sleep development in fi rst years of life 200–201,

224–225 diagnostic techniques 221–226 assessment of daytime sleepiness 226 cardiorespiratory monitoring 225 oximetry 207, 220, 225–226 polysomnography see polysomnography (PSG) nonrespiratory sleep disorders 213–216 ADHD 215 hypersomnias of central origin 213–215 restless legs syndrome 215–216 SDB and 216 sleep disordered breathing see sleep disordered breathing

in children see also infantschloral hydrate 182cholinergic neurons 1, 2chronic kidney disease 98chronic lung disease, in children 212chronic obstructive pulmonary disease (COPD) assessment in OSA patients 117, 118 nocturnal hypoventilation 51, 174 indications for initiation of nocturnal NIV 172, 174chronic respiratory insuffi ciency, nocturnal hypoventilation

and 48 management 174circadian rhythm disorders 74–78 evaluation tools 74, 75 jet lag see jet lag disorder (JLD) shift work disorder see shift work disorder (SWD) sleep/wake rhythm disorders 76–77 advanced sleep phase syndrome 75, 76 delayed sleep phase syndrome 75, 76 free-running circadian disorder see free-running

circadian disorder (FRD)

irregular sleep/wake syndrome see irregular sleep/wake syndrome (ISWS)

treatment 74, 76circadian rhythmicity 74 eff ect on metabolic regulation 17, 18Circadian Type Inventory (CTI) 75CLOCK gene 74clomipramine, cataplexy management 71clonazepam therapy arousal disorders 64 REM behaviour disorder (RBD) 63 sleep-related movement disorders 186, 187codeine therapy, sleep-related movement disorders 187cognitive behavioural therapy, CPAP compliance and 148cognitive behavioural therapy for insomnia (CBT-I) 82,

178, 196cognitive performance, OSA patients 191 paediatric 207–208comorbid insomnia see under insomniacongenital central hypoventilation syndrome (CCHS) 7,

48–49, 208–209 management 235continuous positive airway pressure (CPAP) therapy

157–162 adherence 160–161 algorithms for setting level 159–160 alternatives 162 central sleep apnoea 47, 165–166 adaptive servo ventilation vs 162, 166, 167 children 232–233 assessment of facial development 159 infants 229 comorbidity and 161 interfaces 158–159 machines 157–158 obesity hypoventilation syndrome 169 obstructive sleep apnoea syndrome 102 alternatives 162 cardiovascular eff ects 36, 116 children 232–233 cost-eff ectiveness 193–194 infants 229 metabolic eff ects 37, 118 neurobehavioural eff ects 35 patient education 148 positional therapy vs 149 quality of life eff ects 108, 109 surgical patients 112, 113 weight eff ects 148 principles 157 problem solving 161 titration 159–160 unattended autotitrating device 159–160COPD see chronic obstructive pulmonary disease (COPD)core body temperature, circadian rhythm evaluation 74, 75coronary artery disease (CAD), OSA patients 35–36 assessment 117, 118

239

corticosteroids, intranasal, OSA management in children 231

cortisol 17, 18cortisol levels, circadian rhythm evaluation 75cost-eff ectiveness CPAP therapy 193–194 restless legs syndrome treatment 195costs, sleep disorders 189 see also healthcare costscosts, healthcare see healthcare costsCPAP therapy see continuous positive airway pressure

(CPAP) therapycraniofacial disorders congenital 229–230, 232 obstructive sleep apnoea syndrome and 29, 34, 205–206 management 230, 231craniofacial procedures 231craniosynostosis 229, 230critical closing pressure see under upper airwayCrouzon syndrome 206, 230, 231cystic fi brosis (CF) 174, 211

Ddaytime dysfunction, insomniacs 53, 54daytime sleepiness assessment in children 226 see also excessive daytime sleepiness (EDS)daytime symptoms, sleep history 85–86delayed sleep phase syndrome (DSPS) 75, 76, 187 treatment 187–188depression Beck Depression Inventory 109 insomnia and 80–81 psychological intervention 82–83 obstructive sleep apnoea syndrome and 35 assessment 117 treatment 156development control of breathing see under breathing, control of sleep 200–201, 224–225dextroamphetamine 184diabetes, SDB association 98 assessment in OSA patients 117, 118diagnostic algorithms 100–105dietary interventions, obstructive sleep apnoea 147–148, 152diffi culties initiating sleep (DIS) 53diffi culties maintaining sleep (DMS) 53dim light melatonin onset (DLMO) 74, 75diurnal sleep, young children 200–201dopamine agonists, restless legs syndrome management

65, 186dopaminergic neurons 1Down’s syndrome 96, 97–98, 231driving licence regulations insomnia patients and 197 OSA patients and 191

drug treatment central sleep apnoea 165 hypersomnias of central origin 183–185 narcolepsy 70, 71, 183–185, 214 insomnia 178–182, 195 obstructive sleep apnoea 153–154 restless legs syndrome 65, 185–187, 195 see also specifi c drugsDuchenne muscular dystrophy cardiorespiratory monitoring 225 hypoventilation 48, 50, 162, 170, 209 treatment 171–172, 235–236

EECG (electrocardiography) 120 evaluation of ECG traces 127 OSA patients 116, 118echocardiography, OSA patients 116, 118economic impacts insomnia 195–196 obstructive sleep apnoea 192–194 restless legs syndrome 195education see patient educationelderly central sleep apnoea risk 40 sleep cycle 4electrocardiography see ECG (electrocardiography)electroencephalography see sleep EEGEMG (electromyography) 121 paediatric 221 sleep stages 122, 123emotions, sleep and 82endocrine disorders, SDB-associated 98 diabetes see diabetes endoscopy, sleep 151EOG (electrooculography) 120–121 paediatric 221 sleep stages 122, 123epiglottoplasty, obstructive sleep apnoea 152epilepsy 64Epworth Sleepiness Scale (ESS) 33, 84, 91–92 paediatric 207ergot preparations, restless legs syndrome management 186excessive daytime sleepiness (EDS) 182 assessment 88–90, 131–135 in children 226 MSLT see multiple sleep latency test (MSLT) MWT see maintenance of wakefulness test (MWT) OSLER test 90, 133, 134 test limitations 135 causes 87, 88, 131 CSA patients 46 diff erential diagnosis 87–90 OSA patients 35, 87, 88 assessment 33, 89 treatment 185

240

presentation 88 questionnaires 91–92 Epworth Sleepiness Scale see Epworth Sleepiness Scale

(ESS) Stanford Sleepiness Scale 92 sleep history 85–86 see also hypersomnias of central originexpiratory positive airway pressure (EPAP) 162 obesity hypoventilation syndrome management 169

Ffacial development, assessment in children on long-term

CPAP therapy 159failure to thrive, OSA patients 208fatigue, excessive daytime sleepiness vs 88fl uoxetine, cataplexy management 71fl urazepam 180free-running circadian disorder (FRD) 76 evaluation 74, 75, 77full-face CPAP mask 158, 159Functional Outcome of Sleep Questionnaire (FOSQ)

108–109

GGABAergic neurons 2, 3gabapentin therapy insomnia 180, 182 sleep-related movement disorders 186, 187galanin 3gastric bypass 153 see also bariatric surgerygastro-oesophageal refl ux disease (GERD) 98Gélineau, Jean-Baptise-Edouard 67gender see sexgenioglossus, reduced activity 11ghrelin 17, 18glossopexy 230glucose metabolism OSA patients 37 assessment 117, 118 sleep and 18glutaminergic neurons 2growth hormone (GH) 17, 18Guillain–Barré syndrome 51, 62

Hhealth-related quality of life (HRQoL) defi nition 107 questionnaires 107–109 disease-specifi c 108–109 generic 108healthcare costs insomnia 195–196 obstructive sleep apnoea 192–194

restless legs syndrome 195heart failure central sleep apnoea and 40, 43, 46, 47, 98 CPAP vs adaptive servo ventilation 162 obstructive sleep apnoea and 36heart failure with preserved ejection fraction (HFprEF) 36heart rate variability (HRV) analysis 14, 15high spinal injuries 50histaminergic neurons 1, 2HLA DQB1*0602, narcolepsy association 68, 70, 214home monitoring, obstructive sleep apnoea 100 in children 207human leukocyte antigen gene complex, narcolepsy associa-

tion 68, 70, 214hyoid bone displacement 96hyoid bone surgery, obstructive sleep apnoea 152hyperarousal 55hypercapnic ventilatory response central sleep apnoea 43 obesity hypoventilation syndrome 50 during sleep 10, 11hypersomnias of central origin 67–72 in children 213–215 idiopathic see idiopathic hypersomnia (IHS) narcolepsy see narcolepsy primary 67, 68 recurrent see recurrent hypersomnia secondary 67, 68 sleep disordered breathing and 216 treatment 182–185 nonpharmacological 183 pharmacological 183–185 see also excessive daytime sleepiness (EDS)hypertension OSA patients 18, 25, 32, 36 assessment 115–116, 118, 155 paediatric 208 treatment 155 pulmonary 48, 50hyperventilation, central sleep apnoea pathophysiology 42 treatment 164–165hypnogram typical overnight sleep pattern 3 see also polysomnography (PSG)hypnotics 178–182 contraindications 179 eff ect on sleep structure 179hypocapnia 40, 43, 97, 144hypocretin (orexinergic) neurons 1, 2hypocretins, narcolepsy and 67–68, 69, 70, 183, 214hypopharynx 7hypopnoea defi nition 21, 124, 224 periodic limb movement disorder vs 186–187 presentation 22hypothyroidism, SDB association 98

241

hypoventilation syndromes 21, 23, 48–51 central see central hypoventilation syndromes characteristics 23, 48 in children see under sleep disordered breathing in

children neuromuscular/skeletal disorders see neuromuscular/

skeletal disorders obesity hypoventilation syndrome see obesity

hypoventilation syndrome (OHS) treatment see under nocturnal hypoventilation see also nocturnal hypoventilationhypoxaemia, post-operative 113hypoxic ventilatory response central sleep apnoea 43 infants 201

Iidiopathic hypersomnia (IHS) 71–72, 183 in children 214–215 narcolepsy vs 72 treatment nonpharmacological 183 pharmacological 183–185 see also hypersomnias of central originidiopathic insomnia 56imipramine, cataplexy management 71infants apnoea 202–203 aetiology 202 clinical aspects 203 diagnosis 202 management 228–230 see also apparent life-threatening events (ALTEs);

obstructive sleep apnoea syndrome in children clinical assessment 218–219 periodic breathing see periodic breathing sleep cycle 200, 201, 224 sleep development 200–201, 224–225 sleep problems 218 see also children; sleep disordered breathing in childreninfl ammation, airway 9, 97insomnia 52, 53–59, 80, 195 adjustment (acute) 55 assessment 57–58 behavioural, in infants 218 comorbid 54, 56 therapeutic aspects 182 daytime dysfunction 53, 54 diagnosis 53–54, 195 due to drug/substance 56 economic consequences 195–196 emotions and 82 epidemiology 195 idiopathic 56 legal aspects 196–197 medical consequences 195

nonorganic 56–57 nosological classifi cation 55–57 organic 56–57 paradoxical 55–56 pathophysiology 55, 56 perpetuating factors 55, 56 precipitating factors 55, 56 predisposing factors 55, 56 prevalence 54, 195 primary 54, 55–56 psychiatric disorders and 54, 56, 80–83 depression 80–81 psychological intervention 82–83 psychophysiological 55 rebound 178 SDB and 54, 58–59 secondary 54, 56 Spielman (3P) model 56 treatment 58, 177–182 comorbid OSA and 182 economic impact 196 nonpharmacological 82–83, 177–178, 196 pharmacological 178–182, 196 work disability 196insulin resistance 117, 118insulin sensitivity, eff ect of sleep deprivation 18intermittent hypoxia, eff ect on sympathetic activity 17International Classifi cation of Functioning, Disability and

Health (ICF) 108International Classifi cation of Sleep Disorders (ICSD-2) hypersomnia 68 idiopathic 71 insomnia 55–57 parasomnias 63 sleep disordered breathing 21interstitial lung disease, in children 211–212intranasal corticosteroids, OSA management in children 231iron defi ciency, restless legs syndrome and 65, 185–186, 216irregular sleep/wake syndrome (ISWS) 76–77 evaluation 75

Jjet lag disorder (JLD) 78 evaluation 75 treatment 78

KK complexes 17, 122, 123 in children 222, 223Kleine–Levin syndrome 67, 68, 215

Llateral sleeping position, obstructive sleep apnoea therapy

148–149legal aspects

242

increased motor vehicle accident risk in OSA patients 191 insomnia patients and vehicle accidents 196–197leptin 17, 18levodopa therapy, restless legs syndrome 186Lewy body disease 62lifestyle management, OSA patients 147–148, 152, 155 children 233liver function, assessment in OSA patients 118loop gain 97 central sleep apnoea pathophysiology 42–43lormetazepam 180lung disease chronic, in children 212 interstitial, in children 211–212 obstructive, nocturnal hypoventilation 174 see also specifi c diseaseslung volumes, infl uence on upper airway geometry 8

Mmacroglossia 96magnesium sulfate 186maintenance of wakefulness test (MWT) 89–90, 131, 133 in children 226Mallampati score 219mandibular advancement devices (MADs) 149–150, 162,

231mandibular distraction osteogenesis 231mandibular plane–hyoid (MP–H) distance, measurement 8Marshall syndrome 230maxillofacial surgery, obstructive sleep apnoea 152MCTQ (Munich Chronotype Questionnaire) 75Medical Outcomes Study 36-item short-form health survey

(SF-36) 108melatonin therapy jet lag 78 REM behaviour disorder (RBD) 63metabolic disorders assessment in OSA patients 117–118, 155 see also specifi c disordersmetabolic syndrome defi nition 37 diagnosis 116 obstructive sleep apnoea syndrome and 37–38 assessment 116, 118methadone therapy, sleep-related movement disorders 187methylphenidate, side-eff ects 184methylphenidate therapy, hypersomnia 183, 184 narcolepsy 71micrognathia 96microneurography 16midazolam 180midface advancement 231Milijeteig–Hoff stein method 159, 160mirtazapine 154, 180, 182mixed apnoea, defi nition 124, 223–224modafi nil, hypersomnia treatment 183, 184

narcolepsy 71 OSA patients 185Morningness–Eveningness Questionnaire (MEQ) 75motor neurone disease (MND) 170, 172motor symptoms, sleep history and 86motor vehicle accidents (MVAs) increased risk in OSA patients 189, 190, 191–192 economic impact 193 legal aspects 192 insomnia patients and 196–197movement disorders, sleep-related periodic limb movement disorder see periodic limb move-

ment disorder (PLMD) restless legs syndrome see restless legs syndrome (RLS) treatment 185–187movements scoring in sleep 127 sleep history 85Muenke syndrome 230multi-system atrophy 62multiple sleep latency test (MSLT) 89, 131, 133, 134 in children 220, 226 disadvantages 131 guidelines 132 idiopathic hypersomnia 72 narcolepsy 68, 69, 70, 133, 134 technique 131Munich Chronotype Questionnaire (MCTQ) 75muscle sympathetic nerve activity (MSNA), eff ect of inter-

mittent hypoxia 17MWT see maintenance of wakefulness test (MWT)myasthenia 50myotonic myopathy 50

Nnaltrexone 154narcolepsy 67–71, 183 with cataplexy see cataplexy, narcolepsy with in children 213–214 diagnosis 68–70 multiple sleep latency test 68, 69, 70, 133, 134 idiopathic hypersomnia vs 72 pathogenesis 67, 214 hypocretins 67–68, 69, 70, 183, 214 prevalence 67, 213 secondary 67, 69 symptoms 67, 69, 213–214 treatment 70–71, 214 nonpharmacological 183 pharmacological 70, 71, 183–185, 214 see also hypersomnias of central originnasal CPAP mask 158, 159nasal obstruction 96 clinical assessment in children 219nasal plugs, CPAP interface 158, 159nasal pressure, measurement 22, 137–138, 142

243

in children 221, 224nasal surgery, obstructive sleep apnoea 151, 152nasopharyngeal surgery obstructive sleep apnoea 152 see also adenotonsillectomynasopharynx 7NBBRAs (non-benzodiazepine receptor agonists) 179, 181negative pressure ventilation, sleep hypoventilation

syndrome management in children 234, 235neural regulation breathing during sleep 6–7 NREM and REM sleep 1, 2, 3 wakefulness 1, 2neurobehavioural sequela central sleep apnoea 46 obstructive sleep apnoea 35neurobiology, sleep 1–4neurokinin-1 receptor (NK1R) 6neuromuscular/skeletal disorders 50–51 in children 209 management 235–236 management 170–174 algorithm 172 in children 235–236newborns sleep cycles 200 see also infantsNHP (Nottingham Health Profi le) 108‘night owl’ pattern of sleep 76, 78night terrors 63–64nitrazepam 180nocturnal capnography 143–144nocturnal hypoventilation central sleep apnoea associated with pathophysiology 42 treatment 164 in children 208, 209 chronic respiratory insuffi ciency and see chronic respira-

tory insuffi ciency identifi cation of high-risk cases 170 in obstructive lung disease 174 COPD see under chronic obstructive pulmonary

disease (COPD) cystic fi brosis 174, 211 treatment 169–174 algorithms 169 in children 234–236 evaluation 173, 174 management plan 170 in neuromuscular/skeletal disorders 170–172, 174 in obese patients 169–170, 171 in obstructive lung disease 174 see also hypoventilation syndromesnocturnal seizures 64nocturnal symptoms, sleep history 85, 86non-benzodiazepine receptor agonists (NBBRAs) 179, 181nonrespiratory sleep disorders 52

in children see under children see also specifi c disordersnoninvasive ventilation (NIV) negative pressure ventilation, sleep hypoventilation

syndrome management in children 234, 235 positive pressure see positive airway pressure (PAP)

therapy nonorganic insomnia 56–57nonrestorative sleep (NRS) 53noradrenergic neurons 1, 2Nottingham Health Profi le (NHP) 108NREM sleep abnormal behaviour see arousal disorders neural regulation 1, 2, 3 normal vs abnormal events 61, 62 polysomnography 122, 123 in children 222, 223 respiratory mechanics and ventilation 10, 11 stages 3 see also wake/sleep transitionNREM–REM cycle see sleep cycle

Oobesity adipocyte activity 37 obstructive sleep apnoea and 26, 27, 32, 98, 152 assessment 117–118 surgical considerations 111, 112 upper airway eff ects anatomy 7, 8 critical closing pressure 9 see also bariatric surgery; metabolic syndromeobesity hypoventilation syndrome (OHS) 23, 49–50 treatment 169–170, 171obstructive apnoea defi nition 21, 124, 223 infants 224 see also obstructive sleep apnoea syndrome (OSAS)obstructive lung disease nocturnal hypoventilation 174 see also specifi c diseasesobstructive sleep apnoea syndrome (OSAS) 21–22, 25–38,

190 cardiorespiratory monitoring 136–144 actigraphy 143 interpretation of cardiorespiratory polygraphy 140–142 nocturnal capnography 143–144 polygraphy scoring 138–140 portable devices 137, 138, 139, 140, 141 pulse oximetry see oximetry central sleep apnoea vs 39 in children see obstructive sleep apnoea syndrome in

children clinical aspects 32–34, 103 comorbid insomnia see under insomnia comorbidity assessment 115–118, 155

244

cardiac structure and function 116–117, 118 chronic obstructive pulmonary disease 117, 118 coronary artery disease and peripheral arterial disease

117, 118 depression 117, 118 obesity and metabolic disorders 117–118 systemic hypertension 115–116, 118 comorbidity management 154–156 consequences 25, 32, 35–38, 189 cardiovascular 35–37 in children 207–208 metabolic 37–38 neurobehavioural 35 defi nition 21 diagnosis 21, 23, 100–105 in children 206–207, 220 steps 101–103 strategies 104–105 epilepsy and 64 hypertension and see under hypertension management 147–156 ambulatory 139–140, 141 bariatric surgery 111, 112, 152–153 comorbidities 154–156 CPAP see continuous positive airway pressure (CPAP)

therapy drug treatment 153–154 economic impact 193 network system 101 oral devices 149–150, 162 patient education 147–148 perioperative 112–113 positional 148–149 reference hospitals 100, 101 upper airway anatomy as predictor of response to

therapy 9–10 upper airway surgery 150–152 weight loss 147–148, 152, 155 medicolegal and economic aspects 190–194 economic consequences 192–194 increased motor vehicle accident risk 189, 190,

191–192 professional performance and work disability 190–191 pathophysiology 21, 25, 29–30, 205–206 autonomic nervous system assessment and 17 central respiratory control mechanisms 30 craniofacial morphology 29, 205–206 see also craniofacial disorders function/dysfunction of upper airway muscles 30 upper airway morphology 29, 206 see also sleep disordered breathing, predisposing factors perioperative complications 111–112 perioperative management 112–113 physical examination 33–34 pre-operative screening 112, 113 prevalence 26, 190 in children 205

questionnaires 92–93 Berlin Questionnaire see Berlin Questionnaire Epworth Sleepiness Scale see Epworth Sleepiness Scale

(ESS) pre-operative screening 112, 113 quality of life assessment see under health-related qual-

ity of life (HRQoL) STOP-Bang questionnaire 92–93, 112, 113 STOP questionnaire 92 risk factors 26–27 age 26, 95 obesity 26, 27, 32, 98, 152 sex 26, 27 upper airway abnormalities 27, 96–97 see also sleep disordered breathing, predisposing factors surgical and anaesthesia risk assessment 111–113 symptoms 32–33, 103 in children 206 see also excessive daytime sleepiness (EDS) “typical” vs actual patients 100, 101 upper airway anatomical abnormalities and 7–8, 27, 29,

96–97 as predictor of therapeutic interventions 9–10 surgery 150–152 upper airway and 96–97 critical closing pressure 9, 30, 97 infl ammation 9, 97 infl uence of lung volumes on geometry 8 refl ex activation responses 8obstructive sleep apnoea syndrome in children 205–208 associated complications 207–208 clinical presentation 206 diagnosis 206–207, 220 epidemiology 205 infants 228–230 management 230–234 indications 233 in infants 228–230 stepwise approach 229, 233, 234 pathophysiology 205–206occupational accidents, OSA patients 191oesophageal pressure, measurement 22 children 221Ondine’s curse see congenital central hypoventilation

syndrome (CCHS)opioids central sleep apnoea risk and 40, 112, 113 eff ect on respiratory rhythm-generating neurons 6 restless legs syndrome treatment 186, 187oral devices, obstructive sleep apnoea treatment 149–150 in children 231–232orexinergic neurons 1, 2orexins (hypocretins), narcolepsy and 67–68, 69, 70, 183,

214organic insomnia 56–57orlistat 155oronasal CPAP mask 158, 159

245

oropharyngeal exercises 232oropharyngeal surgery obstructive sleep apnoea 152 see also tonsillectomyoropharynx 7orthodontic procedures, OSA management in children

231–232orthopnoea, amyotrophic lateral sclerosis 50OSLER test 90, 133, 134otolaryngological surgery, obstructive sleep apnoea 151, 152Oxford algorithm 160oximetry 124, 137, 142 in children 207, 220, 225–226oxygen therapy central sleep apnoea 165 cystic fi brosis 211 obesity hypoventilation syndrome 169–170

PPAP see positive airway pressure (PAP) therapyparadoxical insomnia 55–56parafacial respiratory group (pFRG) 6, 7parasomnias 61–65 defi nition 61 diagnosis 61 primary 61 non-REM-related see arousal disorders REM-related see REM behaviour disorder (RBD) secondary 61 nocturnal seizures 64 treatment 187 trigger factors 61 see also restless legs syndrome (RLS)Parkinson’s disease 62, 69, 88, 103, 109PARMs (polyalanine repeat mutations) 49paroxetine 154patient education obstructive sleep apnoea treatment 147–148 sleep hygiene 82Period genes 74periodic breathing 202, 203, 224–225 defi nition 224periodic limb movement disorder (PLMD) 65, 85, 86, 185 hypopnoeas vs 186–187 treatment 185, 186, 187perioperative complications, obstructive sleep apnoea

patients 111–112perioperative management, obstructive sleep apnoea patients

112–113peripheral arterial tonometry (PAT) 15–16, 138peripheral artery disease (PAD), assessment in OSA patients

117, 118, 155Pfeiff er syndrome 230, 231pharmacological treatment see drug treatmentPhox2b 6, 7

PHOX2B gene, mutations 49physical examination children with sleep disorders 219 infants with sleep disorders 218 obstructive sleep apnoea 33–34Pickwickian syndrome 49 see also obesity hypoventilation syndrome (OHS)Pierre Robin syndrome 34, 96, 97, 205–206, 230Pittsburgh Sleep Quality Index (PSQI) 93plasticity, development of breathing control 202poliomyelitis 50polyalanine repeat mutations (PARMs) 49polycystic ovary syndrome, SDB association 98polygraphy (PG) cardiorespiratory see cardiorespiratory polygraphy insomnia 59polysomnography (PSG) 120–129 arousal disorders 64 in children 220, 221–225 interpretation and reporting 225 normative data 224–225 obstructive sleep apnoea 206–207 scoring breathing 223–224 scoring EEG 222–223 sensors 221–222 circadian rhythm disorders 75 ECG evaluation 127 EEG see sleep EEG EMG see EMG (electromyography) EOG see EOG (electrooculography) hypersomnia idiopathic 72 narcolepsy 68, 69 instruction series 121, 126 interpretation 127–129 in infants/children 225 limitations 136 measurement techniques 124, 126 nocturnal seizures 64 normal PSG 128 obstructive sleep apnoea 100, 102, 104–105, 128 in children 206–207 portable monitoring vs 138, 139–140, 142, 143 REM behaviour disorder 62 restless legs syndrome 65 scoring EEG arousals 126 movement in sleep 127 respiratory events see respiratory events sleep EEG see sleep EEG SDB-plus patients 59 sensors and associated monitoring 121–122, 124 in children 221–222portable monitoring devices, obstructive sleep apnoea 137,

138, 139, 140, 141positional treatment, obstructive sleep apnoea 148–149

246

positive airway pressure (PAP) therapy 157–162 adherence 160–161 autotitrating devices 139, 141, 157, 158 bi-level see bi-level positive airway pressure ventilation central sleep apnoea 165–166 continuous see continuous positive airway pressure

(CPAP) therapy expiratory see expiratory positive airway pressure (EPAP) interfaces 158–159 machines 157–158 principles 157 titration 159–160positive pressure ventilation via tracheostomy 234, 235, 236post-operative hypoxaemia 113pramipexole, restless legs syndrome treatment 186, 187 cost 195pre-ejection systolic period (PEP) 15preBötzinger Complex (preBötC) 6, 7primary insomnia 54, 55–56 see also insomniaprofessional performance, OSA patients 190–191protriptyline adverse eff ects 154 cataplexy treatment 71 OSA treatment 154PSQI (Pittsburgh Sleep Quality Index) 93psychiatric disorders, insomnia and see under insomniapsychological interventions, insomnia 82–83, 177–178psychophysiological insomnia 55psychosomatic disorders, treatment of sleep disturbances

188pulmonary congestion 43pulmonary hypertension 48, 50pulse oximetry see oximetrypulse transit time (PTT) 15, 129pulsus paradoxus 15

Qquality-adjusted life-year (QALY) 189–190 CPAP therapy 193–194 restless legs syndrome treatment 195quality of life 107–109 assessment see health-related quality of life defi nition 107Quebec Sleep Questionnaire (QSQ) 109questionnaires 84, 91–93 circadian rhythm assessment 75 OSAS screening 92–93 Berlin Questionnaire see Berlin Questionnaire STOP-Bang questionnaire 92–93, 112 STOP questionnaire 92 paediatric 207 quality of life assessment see under health-related quality

of life (HRQoL) sleep quality assessment 93 sleepiness assessment 91–92

Epworth Sleepiness Scale see Epworth Sleepiness Scale (ESS)

Stanford Sleepiness Scale 92quetiapine 180, 182quiet sleep (QS) 200 breathing and 202 polysomnography 223

Rrace, as SDB predisposing factor 95ramelteon 180, 182rapid maxillary expansion (RME) 231–232rebound insomnia 178recurrent hypersomnia 67, 69, 88, 183 in children 215 see also hypersomnias of central originrefl ex activation, upper airway dilator muscles see under

upper airwayrelaxation training, insomnia treatment 178REM behaviour disorder (RBD) 61–63 classifi cation 62 diagnosis 62 pathophysiology 62–63 presenting features 62 prevalence 62 treatment 63, 187REM sleep neural regulation 1, 2, 3 normal vs abnormal events 61, 62 polysomnography 123 in children 222 respiratory mechanics and ventilation 10–11 see also sleepresistant hypertension 116respiration monitoring 120 sleep history 85 see also breathingrespiratory disturbance index (RDI) 92, 141, 143, 151respiratory drive, instability 97respiratory eff ort-related arousals (RERAs) 22, 224respiratory events defi nitions 124 scoring 125–126, 127 in children 223–224respiratory mechanics, during sleep 10–11respiratory pacing 235respiratory polygraphy (RP), obstructive sleep apnoea 100,

102, 104, 105 in children 207respiratory rhythm, developmental aspects 202restless legs syndrome (RLS) 64–65, 85, 86, 185, 194 attention defi cit hyperactivity disorder and 215–216 in children 215–216 classifi cation 65, 194 diagnosis 64–65, 185

247

economic consequences 195 epidemiology 65, 194 iron defi ciency and 65, 185–186, 216 medical consequences 194 pathophysiology 65, 216 treatment 65, 185–187 economic impact 195 nonpharmacological 185 pharmacological 65, 185–187, 195retrotrapezoid nucleus (RTN) 6, 7rhinitis, in children 211RME (rapid maxillary expansion) 231–232ropinirole, restless legs syndrome treatment 186, 187Roth, Bedrich 71rotigotine, restless legs syndrome treatment 186, 187Roux-en-Y gastric bypass 153R–R interval analysis 14, 153

SSaethre–Chotzen syndrome 230SAQLI (Calgary Sleep Apnea Quality of Life Index) 109Scopinaro technique 153SDB see sleep disordered breathingsecondary insomnia 54, 56 see also insomniaseizures, nocturnal 64selective serotonin re-uptake inhibitors cataplexy treatment 71, 185 OSA treatment 154‘selfi sh brain’ theory 37sensory symptoms, sleep history and 86serotonergic neurons 1, 2sex, as SDB predisposing factor 95 central sleep apnoea 40 obstructive sleep apnoea 26, 27SF-36 (Medical Outcomes Study 36-item short-form health

survey) 108shift work disorder (SWD) 77–78 evaluation 75 health impacts 78 management 78SIDS (sudden infant death syndrome) 204skeletal disorders see neuromuscular/skeletal disorderssleep arousal from 1 autonomic and cardiovascular changes 13 obstructive sleep apnoea 17 physiological data 16–17 autonomic and cardiovascular changes, assessment 13–16 autonomic ‘stress tests’ 14 blood pressure monitoring 14–15 heart rate variability analysis 14, 15 methodological issues 13–14 microneurography 16 OSA patients 17

peripheral arterial tonometry 15–16, 138 pulse transit time 15 breathing during 6–11 control of 6–7 developmental aspects 202 pathophysiology of ventilatory changes 48 respiratory mechanics and ventilation 10–11 development in fi rst years of life 200–201, 224–225 diaries see sleep logs/diaries disorders see sleep disorders emotions and 82 history see sleep history hygiene see sleep hygiene logs see sleep logs/diaries metabolic changes 17–18 appetite regulation 18 glucose metabolism 18 neurobiology 1–4 non-REM see NREM sleep questionnaires see questionnaires REM see REM sleep requirements 3, 4Sleep Apnea Quality of Life Index (SAQLI) 109sleep apnoea central see central sleep apnoea (CSA) obstructive see obstructive sleep apnoea syndrome

(OSAS) sleep attacks 88sleep cycle 3 eff ect of ageing 3–4 infants 200, 201, 224sleep deprivation, eff ects 3 appetite regulation 18 glucose metabolism 18sleep disordered breathing (SDB) in children see sleep disordered breathing in children CSA see central sleep apnoea defi nitions 21–24 diagnostic algorithms 100–105 insomnia and 54, 58–59 OSA see obstructive sleep apnoea predisposing factors see sleep disordered breathing,

predisposing factors quality of life and see health-related quality of life

(HRQoL) SDB-plus 59 sleep-related hypoventilation see hypoventilation

syndromes sleep disordered breathing in children 205–209 clinical assessment 219–220 comorbid nonrespiratory disorders 216 comorbid respiratory disorders 210–212 allergic rhinitis 211 asthma 210 bronchopulmonary dysplasia 212 chronic lung disease 212

248

cystic fi brosis 211 interstitial lung disease 211–212 obstructive sleep apnoea see obstructive sleep apnoea

syndrome in children sleep hypoventilation syndromes 208–209 congenital central hypoventilation syndrome see

congenital central hypoventilation syndrome (CCHS)

management 234–236 neuromuscular and related conditions see

neuromuscular/skeletal disorders see also childrensleep disordered breathing, predisposing factors 95–98 age 95 anatomical abnormalities 95–97, 96 congenital syndromes 97–98 hyoid bone displacement 96 hypertrophy of uvula, soft palate and tonsils 96 macroglossia 96 micrognathia 96 nasal obstruction 96 upper airway shape/length 96–97 associated diseases 96, 97–98 chronic kidney disease 98 endocrine disorders 98 gastro-oesophageal refl ux disease 98 heart failure 98 obesity 98 stroke 98 functional factors 96, 97 airway infl ammation 97 increased upper airway collapsibility 97 instability of respiratory drive 97 race 95 sex 95sleep disorders medicolegal and economic aspects 189–197 costs 189 see also healthcare costs QALY assessment see quality-adjusted life-year (QALY) nonrespiratory 52 in children see under children psychiatric aspects 80–83 depression 80–81 psychological interventions 82–83 role of emotions 82 in somatoform disorders, treatment 188 see also sleep disordered breathing (SDB); specifi c disorders‘sleep drunkenness’ 71sleep EEG 120 arousals 126 in children 222 electrode placement 120, 121 historical aspects 1 paediatric 221 scoring 222–223 scoring 122, 123, 126

in infant/child 222–223 see also polysomnography (PSG)sleep endoscopy 151Sleep Heart Health Study (SHHS) 108sleep history 84–86 children 219–220 daytime symptoms 85–86 infants 218–219 insomnia assessment 57 nocturnal symptoms 85, 86 questionnaires see questionnaires sleep/wake pattern 85sleep hygiene inadequate 56 recommendations for insomnia patients 88, 177, 178sleep hygiene education (SHE) 82sleep logs/diaries 84 circadian rhythm evaluation 74, 75 insomnia assessment 57, 58sleep monitoring devices, classifi cation 136–137sleep-onset REM periods (SOREMPs) 89 defi nition 89 idiopathic hypersomnia 72 narcolepsy 69sleep quality, questionnaires 93sleep questionnaires see questionnairessleep-related hypoventilation defi nition 224 see also hypoventilation syndromes; nocturnal

hypoventilationsleep-related movement disorders periodic limb movement disorder see periodic limb

movement disorder (PLMD) restless legs syndrome see restless legs syndrome (RLS) treatment 185–187sleep requirements 3, 4sleep restriction eff ects 4 insomnia treatment 177–178sleep spindles 122, 123 in children 222, 223sleep stages normative data across lifespan 127 polysomnography 122, 123, 126sleep state misperception 56sleep/wake rhythm disorders see circadian rhythm disorderssleepiness, daytime assessment in children 226 see also excessive daytime sleepiness (EDS)sleepwalking 63snoring 32, 85, 86 in children 205, 207 clinical assessment 219sodium oxybate 70, 71, 184–185soft palate, hypertrophy 96somatoform disorders, treatment of sleep disturbances 188somnambulism 63

249

Spielman model of insomnia 55, 56spinal injuries 50stage 1 sleep, polysomnography 122 in children 222stage 2 sleep, polysomnography 122, 123 in children 222stage 3/4 sleep, polysomnography 123 in children 222stage R sleep polysomnography 123 in children 222 see also REM sleepstage W (wake stage), polysomnography 122, 123 in children 222Stanford Sleepiness Scale (SSS) 92Stickler’s velocardiofacial syndrome 230stimulants, hypersomnia treatment 183–184 narcolepsy 70, 71stimulus control, insomnia treatment 177STOP-Bang questionnaire 92–93, 112, 113STOP questionnaire 92stroke 98 central sleep apnoea and 40, 98 obstructive sleep apnoea and 35, 36, 98sudden infant death syndrome (SIDS) 204surgery, bariatric see bariatric surgerysurgical risk assessment, obstructive sleep apnoea 111–113α-synucleinopathy 62systemic hypertension see hypertension

Ttemazepam 180tetraplegia 50theophylline therapy central sleep apnoea 165 obstructive sleep apnoea 154thermistors 122, 142, 2213P (Spielman) model of insomnia 55, 56tongue size measurement 8 sleep apnoea and 29, 96tongue surgery, obstructive sleep apnoea 152tonsillectomy 152, 230–231 in infants 229tonsils, hypertrophy 96, 229Total face mask 158tracheostomy obstructive sleep apnoea 152 in infants 229 sleep hypoventilation syndromes in children 234, 235, 236trazodone 180, 182triazolam 181tricyclic antidepressants cataplexy treatment 185 insomnia treatment 180, 182 OSA treatment 154

Uupper airway 7–10 anatomy 7–8, 9 in obstructive sleep apnoea see under obstructive sleep

apnoea as predictor of therapeutic interventions 9–10 critical closing pressure 8–9 obstructive sleep apnoea syndrome 9, 30, 97 in development of breathing control 201–202 eff ect of mandibular advancement 150 examination, obstructive sleep apnoea 34 infl ammation 9, 97 infl uence of lung volumes on geometry 8 refl ex activation of dilator muscles 8 in obstructive sleep apnoea 30 surgery, in obstructive sleep apnoea 150–152 three-dimensional reconstruction 8, 9 volume, men vs women 8upper airway resistance syndrome (UARS) 22, 205uvula, hypertrophy 96uvulopalatopharyngoplasty 152, 231

Vvalerian 182valsartan 155venlafaxine 71ventilation during sleep 10–11 see also breathingventilatory response (VR) hypercapnic see hypercapnic ventilatory response hypoxic see hypoxic ventilatory response infants 201ventrolateral pre-optic area (VLPO) neurons 3ventrolateral respiratory column (VRC) 6, 7video-polysomnography, in children 220

Wwake/sleep transition 3 breathing changes in lung mechanics 10 control of 7wake stage see stage W (wake stage)wakefulness, neural regulation 1, 2‘wakefulness drive to breathe’ 7weight loss, OSA patients 147–148, 152, 155work disability insomnia patients 196 OSA patients 190–191

Zzaleplon 179, 181zolpidem 179, 181zopiclone 179, 181250

ERS Sleep Handbook

Documents

pedunculopontine tegmental nuclei

dorsal raphe nuclei

germany randerath klinik

online cme test

children ha trang

sleep laboratory school

obstructive sleep apnoea

european respiratory society