Sleep disorders in multiple system atrophy: a correlative video-polysomnographic study

Original article

Sleep disorders in multiple system atrophy: a correlative

video-polysomnographic study

Roberto Vetrugnoa,*, Federica Provinia, Pietro Cortellib, Giuseppe Plazzia,Enrico M. Lottia, Giulia Pierangelia, Carlotta Canalia, Pasquale Montagnaa

aDepartment of Neurological Sciences, University of Bologna, Via Ugo Foscolo 7, Bologna 40123, ItalybInstitute of Clinical Neurology, University of Modena and Reggio Emilia, Via Dal Pozzo, Modena 41100, Italy

Received 12 February 2003; received in revised form 17 June 2003; accepted 20 July 2003

Abstract

Objective: The reciprocal relation between breathing, heart and motor system abnormalities during sleep was studied in multiple system

atrophy (MSA) by means of video-polysomnographic recordings (VPSG).

Patients and Methods: Nineteen consecutive MSA patients underwent VPSG with scoring for sleep, respiratory abnormalities, heart (HR)

and breathing (BR) rates, and abnormal motor activities. A comparative analysis was performed versus 10 patients with obstructive sleep

apnoea syndrome (OSAS).

Results: All MSA patients displayed snoring, 42% stridor, and 37% OSAS. Mean sleep SaO2 was 92.7%, and lowest SaO2 86%. Patients

with stridor had a significant increase in BR from Wake to NREM and REM sleep, and higher HR during sleep. Respiratory muscles and

tibialis anterior EMG tonic activity was frequently found, more often in patients with stridor. All patients had REM sleep behaviour disorders

(RBD) and 88% periodic limb movements during sleep (PLMS). No OSAS patient had RBD or respiratory muscles and tibialis anterior tonic

activity.

Conclusions: MSA patients, especially those with associated stridor, commonly display impaired breathing and abnormal control of

respiratory and limb muscles during sleep. Breathing and motor abnormalities are often concomitant in the same patient, indicating a diffuse

impairment of sleep homeostatic integration that should be included within the diagnostic features of MSA.

q 2003 Elsevier B.V. All rights reserved.

Keywords: Multiple system atrophy; Sleep; Stridor; Snoring; Video-polysomnography

1. Introduction

Several sleep-related respiratory and motor disturbances

have been reported in multiple system atrophy (MSA).

Subjective sleep complaints reported by patients or relatives

include insomnia, excessive daytime sleepiness (EDS),

snoring or other respiratory noises, and motor restlessness

while asleep, sleep talking or overt violent behaviours [1].

Video-polysomnographic recordings (VPSG) demonstrate

abnormal sleep architecture, respiratory disturbances such

as nocturnal alveolar hypoventilation [2–4], obstructive and

central sleep apnoeas [5–8] and nocturnal stridor [9–13]

and motor abnormalities including periodic limb move-

ments during sleep (PLMS) [14] and REM sleep behaviour

disorder (RBD), the latter often forerunning the disease

[15–19].

The relation between abnormal motor control and sleep-

related breathing abnormalities in MSA, and the occurrence

in the same patient of sleep-related breathing and motor

abnormalities have not been systematically analysed, except

by questionnaire analysis [1], which may be open to

criticism when not checked against VPSG recordings.

We performed VPSG recordings in 19 consecutive MSA

patients, monitoring respiration, heart (HR) and breathing

rate (BR) and limb muscular EMG activity, to define the

extent and reciprocal relation of sleep-related autonomic

and motor disturbances in MSA.

1389-9457/$ - see front matter q 2003 Elsevier B.V. All rights reserved.

doi:10.1016/j.sleep.2003.07.002

Sleep Medicine 5 (2004) 21–30

www.elsevier.com/locate/sleep

* Corresponding author. Tel.:þ39-051-644-2225; fax:þ39-051-644-2165.

E-mail address: [email protected] (R. Vetrugno).

2. Patients

Nineteen consecutive patients (13 men and 6 women)

fulfilling criteria for clinically probable MSA [20] were

admitted for diagnostic purposes in our Neurological

Department, which has an interest in sleep disorders.

Patients were not selected on the basis of sleep, respiratory

or motor problems. Thirteen of these patients had probable

cerebellar type MSA (MSA-C) and six had probable

parkinsonian type MSA (MSA-P) [20]. Nine patients died

during follow-up with a median survival of 9.2 years (range

4–20). Especially in view of the respiratory and skeletal

muscle EMG findings (see later), 10 consecutive age- and

mean body mass index (BMI)-matched patients with

isolated obstructive sleep apnoea syndrome (OSAS) and

without stridor were also investigated. The latter patients

were free from autonomic and neurological disturbances,

diabetes mellitus or medications.

3. Methods

A history of sleep patterns and disorders was obtained

from each patient according to a structured interview and

checked with interviews of close relatives. Patients under-

went VPSG including EEG (C3-A2, O2-A1, Cz-A1), right

and left EOG, surface EMG from submental, intercostalis

(electrodes placed at least 2 cm apart on the second anterior

intercostal space lateral to the sternum), diaphragm

(electrodes at the seventh/eighth intercostal space on the

anterior axillary line, on the right to reduce cardiac artefact),

right and left tibialis anterior and extensor carpi radialis

muscles, tracheal microphone, oro-nasal airflow, thoracic

and abdominal respirogram by means of gauges placed at

the level of the axilla and just superior to the iliac crest,

respectively, ECG, arterial blood pressure by means of

Finapress and oxyhaemoglobin saturation (SaO2) by means

of ear oxymeter. Endoesophageal pressure was additionally

recorded by an esophageal balloon in three patients, all with

stridor (patients 12, 16, 17 in Table 1). Data were acquired

on a Grass polygraph with a paper speed of 10 mm/s

and synchronized video recording. Video records were

analysed for sleep, breathing, SaO2 and abnormal move-

ments, including EMG activity in the mylohyoideus,

intercostalis, extensor carpi radialis and tibialis anterior, in

30-s epochs.

Sleep data were staged according to standard methods

[21], and sleep structure and sleep efficiency were

evaluated. An abnormal breathing event during sleep was

defined as apnoea (a complete cessation of airflow lasting

10 s or more) or hypopnoea (a discernible reduction in

respiratory airflow accompanied by a decrease of at least 4%

Table 1

Single patient data (top) and averaged synthesis (bottom)

Patients Diagnosis Disease onset

(initial complaints)

(years/months)

RBD onset

(years/months)

Stridor onset

(years/months)

VPSG

(years/

months)

Death

(years/

months)

Duration versus

onset (years/

months)

Duration versus

stridor (years/

months)

(1) Male MSA-C 59 (RBD þ C) 59 – 62 65 6 –

(2) Male MSA-P 47 (RBD þ AF) 47 – 66 67 20 –

(3) Male MSA-C 62 (RBD þ C) 62 – 64 – – –

(4) Female MSA-C 72 (C) 75 – 75 76 4 –

(5) Male MSA-P 50 (P) 53 – 53 – – –

(6) Male MSA-P 60 (RBD þ AF) 60 – 70 – – –

(7) Male MSA-C 45 (RBD) 45 – 50 56 11 –

(8) Male MSA-C 58 (C) 60 – 60 – – –

(9) Male MSA-C 59 (C) 65 – 66 67 8 –

(10) Female MSA-C 58 (RBD þ C) 58 – 61 – – –

(11) Male MSA-P 41 (RBD þ AF) 41 – 44 49 8 –

(12) Female MSA-P 53 (RBD þ stridor) 53 53 57 – – –

(13) Female MSA-C 49 (RBD) 49 51 54 – – –

(14) Female MSA-C 52 (AF) 55 54 55 – – –

(15) Male MSA-P 66 (RBD þ AF) 66 71 71 73 7 2

(16) Male MSA-C 53 (C) 58 59 59 – – –

(17) Female MSA-C 57 (RBD þ C) 57 61 62 – – –

(18) Male MSA-C 54 (RBD) 54 57 58 63 9 6

(19) Male MSA-C 49 (C) 53 56 58 59 10 3

MSA patients–mean

(range)

54/11 (41–72) 56/3

(41–75)

60/3

(44–75)

63/10

(49–76)

9/2

(4–20)

Without stridor (range) 55.6 (41–72) 56.9 (41–75) – 61 (44–75) 63.4 (49–76) 9.6 (4–20)

With stridor (range) 54.1 (49–66) 55.7 (49–66) 57/9

(51–71)

59.3 (54–71) 65 (59–73) 8,8 (7–10) 3/8

(2–6)

Disease duration is expressed regarding either to the first symptom(s) (versus onset) or to the appearance of stridor (versus stridor). AF, autonomic failure;

C, cerebellar; P, parkinsonian; RBD, REM sleep behaviour disorder.

R. Vetrugno et al. / Sleep Medicine 5 (2004) 21–3022

in SaO2) [22–24]. The average number of apnoea–

hypopnoea episodes per hour of sleep was the respiratory

disturbance index (RDI). As a criterion for OSAS, we chose

an RDI above 10, which is an accepted international scoring

standard [24–26].

Respiratory noise was inspiratory when synchronous

with down-deflection of oral and thoracic-abdominal

respiratory traces representing inspiration and with inter-

costal EMG burst activity; expiratory if synchronous with

up-deflection of respiratory traces and positive excursion of

endoesophageal pressure. Stridor was defined as a strained,

high-pitched, harsh respiratory sound on audio–video

monitoring. Paradoxical breathing (PB) was identified as a

phase-out of thoracic as opposed to abdominal respiratory

traces [24]. Sleep time with snoring or with stridor was

arbitrarily defined as the number of epochs with at least 50%

of respiratory movements associated with snoring or stridor,

and the percentage of time of snoring (sleep time with

snoring/total sleep time) and percentage of time of stridor

(sleep time with stridor/total sleep time) were then

calculated. Intercostalis EMG activity was evaluated and

limb EMG activity was scored for PLMS according to

Coleman’s criteria [27]: a series of four or more consecutive

movements lasting 0.5–5 s with an intermovement interval

of 4–90 s. The PLMS index (number of PLMS per hour of

sleep) was calculated. EMG activity of tonic type (in the

shape either of continuous muscle recruitment pattern or of

repetitive uninterrupted motor unit potentials at rates greater

than 5–7 impulses per second [28]) on the tibialis anterior,

mylohyoideus, and intercostalis muscles was analysed and

scored whenever it was present in at least two consecutive

epochs. Given its high frequency among MSA patients (see

later) and to ensure reliability, it was arbitrarily scored as

pathological when present in at least one third of total VPSG

epochs. PLMS and other simple or complex motor events

during sleep, in particular REM behavioral disorders

(RBD), the latter defined according to Mahowald and

Schenck [29], were checked against the video recordings.

Mean HR and mean BR were calculated for every 5 min

of noise and/or artefact-free epochs.

Ten OSAS patients were subjected to the same

procedures as the MSA patients. Endoesophageal pressure

was obtained in all of them.

Statistical analysis was done with non-parametric tests of

significance (the Wilcoxon test for continuous variables and

the Mann –Whitney test for nominal variables). We

considered statistically significant P values of less than

0.05 ðP , 0:05Þ:

4. Results

4.1. MSA: subjective complaints and clinical findings

All patients complained of autonomic symptoms and 15

had mild–moderate symptomatic postural hypotension

(dizziness, syncope and visual disturbances). The most

frequent sleep-related subjective complaints were abnormal

violent motor behaviours during sleep and respiratory

noises.

Eight out of the 19 patients (42%), six with MSA-C

(three men) and two with MSA-P (one man) complained of

nocturnal stridor upon admission. Mean BMI was 25.9 (24.1

in patients without stridor). Remarkably, RBD was reported

as the initial (or among the initial) complaints of disease in

12 patients, and in three patients isolated RBD had preceded

other symptoms by 2–15 years. Diurnal and nocturnal

irregular respiration, e.g. gasp, snoring and respiratory

stridor, were a complaint in 10 patients. None of these

symptoms was, however, the cause for referral, evaluation

or admission.

Mean age at disease onset was slightly lower in patients

with stridor (54.1 versus 55.6 years), and disease duration

shorter in patients with stridor (8.8 versus 9.6 years).

Patients with stridor had RBD earlier than patients without

stridor (55.7 versus 56.9 years, Table 1). These differences

between MSA with and without stridor were not significant.

4.2. MSA: VPSG findings

4.2.1. Sleep structure

Sleep structure was characterized by increased NREM

sleep stages 1–2 (63%, n.v. 45–55%), decreased stages 3–4

(12%, n.v. 15–25%) and a normal amount of REM sleep

(25%, n.v. 20–25%) [30]. Sleep efficiency was reduced

(52%; n.v. . 85%). Except for sleep latency, there were no

differences in sleep structure between patients with and

without sleep-related respiratory disturbances.

4.2.2. Respiratory abnormalities

All patients displayed inspiratory noise during sleep:

snoring in 100%, stridor in 42%. The mean sleep RDI was

10.6 in the whole group of MSA patients, with significant

differences between NREM (mean RDI 11.9) and REM

sleep (mean RDI 6.3, P ¼ 0:01, Table 2). Only seven MSA

patients had a sleep RDI higher than 10, thus fulfilling the

criterion adopted in this study for a diagnosis of OSAS

[24–26]. Of these seven patients, four did not display

nocturnal stridor, and could thus be categorized as OSAS-

only MSA patients; three patients with stridor had

concomitant OSAS (Table 2).

The three patients with stridor (patients 12, 16, 17,

Tables 1 and 2) who had endoesophageal pressure

measurements showed mean negative values during sleep

ranging from 24 to 224 cm H2O (mean 214 cm H2O,

n.v. $ 2 4 cm H2O) [31], suggesting increased sleep-

related respiratory effort.

For the whole group of patients, mean SaO2 throughout

sleep remained at 92.7%, with the lowest values at 82%.

SaO2 was not especially reduced during REM sleep. Mean

SaO2 during sleep remained practically the same in patients

with and without stridor and there was no significant mean

R. Vetrugno et al. / Sleep Medicine 5 (2004) 21–30 23

SaO2 difference between MSA with and without OSAS

(Table 2).

PB (Fig. 1) was found in about half of the patients, more

frequent in patients with stridor (Table 2). Expiratory noise

was present in 74% of the patients, accompanied in all with

expiratory intercostalis EMG activation. All patients with

stridor showed expiratory intercostalis EMG activation

associated with concomitant slightly positive endoesopha-

geal pressure (Fig. 2). Stridor persisted also during relaxed

wakefulness in six out of the eight (75%) patients with

nocturnal stridor.

4.2.3. Breathing and heart rates

The mean BR while awake in the whole group of MSA

patients was 19 ^ 4 cycles/min, with increases during

NREM and REM sleep (Table 3). Though slight,

differences between REM, Wake and NREM sleep were

significant ðP , 0:05Þ: The increased rates during sleep

were, however, due to the presence of the stridor subgroup:

patients with stridor had BR comparable to patients

without stridor during Wake but higher rates during both

NREM and REM sleep when compared to patients without

stridor and to Wake values (the latter differences being

statistically significant, P , 0:05), whereas patients with-

out stridor showed significantly lower BR during NREM

versus REM sleep ðP , 0:05Þ but not versus Wake

(Table 3).

Mean HR was 70 ^ 10 beats/min during Wake, decreas-

ing during NREM and REM sleep for the whole group of

patients, differences between Wake and NREM and

between NREM and REM being significant (P , 0:05;

Table 3). There were no significant differences between

patients with and without stridor in mean HR during the

various Wake and sleep stages. However, patients with

stridor showed no significant decrease during NREM and

higher HR during REM sleep compared to Wake, while

patients without stridor had significantly lower HR between

Wake and NREM sleep (P , 0:05; Table 3).

Oscillations of HR and arterial pressure related to the

respiratory events, whether snoring, sleep apneas or stridor,

were notably blunted in MSA patients especially when

compared with the OSAS controls (Fig. 3). No effort,

however, was made to quantify these changes.

4.2.4. Motor activity, PLMS, RBD

Intercostalis EMG tonic activity in the shape of

subcontinuous muscle recruitment patterns and repetitive

motor unit potential discharges (Fig. 4) was scored in most

of MSA patients during Wake, decreasing but still present

during NREM and even during REM sleep (Table 4).

Tibialis anterior tonic EMG activity (continuous EMG

activity and repetitive motor unit potential discharges) was

recorded in 17 out of 19 patients with MSA (6 with stridor

and 11 without stridor) again more prevalent among patients

Table 2

Sleep-related quantitative respiratory and PLMS findings: single patient data and averaged synthesis

Patients Stridor Sleep Eff

(%)

PB Sleep RDI NREM

RDI

REM

RDI

Mean SaO2

(%)

Lowest

SaO2 (%)

Mean NREM

SaO2 (%)

Mean REM

SaO2 (%)

Time

snoring

(%)

Time

stridor

(%)

PLMS

index

(1)a – 38.1 Yes 15a 16.9 0 91a 88 91 90 30.4 – 24.3

(2) – 72.5 Yes 6.9 6.9 7.1 90.7 86 90.6 91.8 70.5 – 60.8

(3) – 17.7 4.0 0 7.1 95.8 94 96 95.8 32 – 48

(4) – 30.9 8.4 6.4 11 90.4 88 91.4 89.6 20.7 – 41.3

(5)a – 76.7 17.8a 18.5 14.1 93.3a 90 93.4 93 39.7 – 76.1

(6) – 73.4 1.5 1.8 0 93.6 93 93.6 92 46 – 26

(7)a – 85.4 21a 26.9 5.6 92.6a 90 92.5 93.9 14.9 – 24.9

(8) – 51.7 Yes 6.7 8.7 0 95.5 94 95.5 95 90 – 19.2

(9)a – 18.7 17.6a 22.2 3 95.2a 92 95.1 96 33 – 14.5

(10) – 67.9 Yes 1.5 2.4 0 92.6 88 92.6 92 46 – 28

(11) – 51.2 3 3.5 0 94.3 88 94.3 94 48 – 16

(12) Stridor 46.5 Yes 4.3 4 8.6 90 86 91 89 22.2 67 26

(13) Stridor 39.3 Yes 8 8.3 6.7 94.7 92 94.7 95 28.7 36.3 33.3

(14) Stridor 78.9 5.2 5.2 5.2 94.7 90 94.7 95 2.9 34.1 43

(15)a Stridor 73.8 Yes 16.1a 16.6 10 89.2a 82 89.3 87.8 23 23 NA

(16)a Stridor 26.5 Yes 15.5a 16.9 10 89.3a 88 89.2 90 56.3 40 NA

(17) Stridor 63.5 Yes 5.2 6.1 4.5 89.2 86 89.1 89.3 5.7 93 38

(18) Stridor 51.7 7 9 1 94.6 92 94.6 95 48.8 47 34

(19)a Stridor 29.0 Yes 36.9a 45.6 25.1 94.5a 92 94.7 94.1 57.7 26.9 20

Mean 52.3 53% 10.6 11.9 6.3 92.7 92.8 92.5 37.7 33.7

Without stridor 53.1 36% 9.4 10.4 4.4 93.2 93.3 93 42.8 34.4

With stridor 51.2 75% 12.3 14 8.9 92 92.1 91.9 30.6 45.9 32.3

Sleep Eff, sleep efficiency; RDI, respiratory disturbance index; SaO2, oxyhaemoglobin saturation; PB, paradoxical breathing; PLMS, periodic limb

movement during sleep. NA, not available.a Marks patient with concomitant OSAS (RDI . 10).

R. Vetrugno et al. / Sleep Medicine 5 (2004) 21–3024

during Wake, but still present during NREM and also REM

sleep. Both intercostalis and tibialis anterior EMG activity

were more frequent in patients with stridor. PLMS were

present in 88% of patients and in all patients with stridor,

with varying but always abnormal PLMS index (n.v. # 5,

Tables 2 and 4).

All MSA patients showed continuous EMG activity

and repetitive motor unit potential discharges in the

submental mylohyoideus muscle while awake and during

sleep, with phasic or tonic increase during REM sleep.

Excessive chin muscle tone persisted throughout sleep

(Table 4). EMG twitching during REM sleep could be

synchronous or asynchronous with complex limb move-

ments. Sometimes bursts of EMG activity on the

mylohyoideus muscle could recur synchronously with

brief central apnoeas during wake and sleep.

RBD in the form of complex motor behaviours such as

limb and body jerks or complex vigorous movements were

recorded in all patients (Table 4).

4.3. Comparative analysis of MSA versus OSAS patients

Ten OSAS patients had a mean of BMI 25.9 (range

25.1–26.8), a mean age of 56.5 years (range 54–61) and

a mean disease duration of 4 years (range 1–10); all

these values were not significantly different from our

MSA patients.

Sleep structure showed prevalent stages 1–2 NREM

sleep (67%, n.v. 50%), and reduced stages 3–4 NREM

(17%, n.v. 25%) and REM sleep (16%, n.v. 25%) with sleep

efficiency of 83% (n.v. . 85%). Mean RDI was 47 (range:

21–58, 38 during NREM and 55 during REM sleep), and

mean SaO2 86%, dropping to 81% during REM sleep. Mean

sleep-apnea related endoesophageal pressure negative value

was 230 cm H2O (range 220 to 260 cm H2O), values

significantly worse than our MSA patients ðP , 0:05Þ:

Typical oscillations of HR and arterial pressure

accompanied each apnoea event (Fig. 3). None of the

OSAS patients showed expiratory intercostalis EMG

activity or tonic intercostalis or tibialis anterior EMG

activity and/or repetitive motor unit potentials discharges

either while awake or during sleep, whereas 60% of them

displayed PLMS. All of the OSAS patients showed

physiological atonia without abnormal motor behaviour

during REM sleep (Table 4).

5. Discussion

5.1. Wake–sleep architecture and parameters

Our patients with MSA had abnormal sleep structure,

with reduced NREM deep sleep and decreased sleep

efficiency. Though these findings probably relate at least

Fig. 1. NREM sleep recording (excerpts) in an MSA-C patient (no. 16) shows inspiratory and expiratory noise. Hypopnoeas and paradoxical breathing, marked

by the asterisk (*) are present with slightly positive endoesophageal pressure and minimal SaO2 variability. EOG, electro-oculogram; mylohyoid.,

mylohyoideus muscle; ECG, electrocardiogram; microph., microphone; Interc., intercostalis muscle; Thor., thoracic; Abd., abdominal; resp., respiration;

intraoesoph. press., endoesophageal pressure; system. art. press., systemic arterial pressure; SaO2, oxyhaemoglobin saturation.

R. Vetrugno et al. / Sleep Medicine 5 (2004) 21–30 25

in part to a first night effect, the decreased sleep efficiency,

striking in MSA compared to the OSAS patients who were

subjected to the same procedures, seems to indicate that lack

of sleep is characteristic of MSA, and cannot be attributed

entirely to the disturbing effects of our investigational

procedures.

Lack of sleep is also a potentially relevant finding when

trying to explain the prevalence of sleep-related respiratory

abnormalities in MSA. Indeed sleep disruption may

promote sleep apnea. Such a possibility cannot be

discounted in our patients, even though a comparison of

the sleep efficiency with RDI did not disclose any obvious

correlation in our MSA patients (Table 2).

5.2. Respiratory abnormalities during sleep

Significant findings in our MSA sample were the high

rates of respiratory disturbances observed during sleep.

Inspiratory noise was universal, while noise upon expiration

was found in nearly 4/5 of patients. More relevantly, stridor

was found in nearly half of our patients, while, based on our

adopted diagnostic criterion for OSAS (RDI above 10), 37%

had OSAS. Remarkably, patients spent more than 1/3 or

nearly 1/2 of their sleep time snoring or with stridor. OSAS

and stridor were not always coincidental, since four patients

had OSAS only in the absence of stridor. Stridor also

persisted also during relaxed wakefulness in 75% of

patients. These sleep-related respiratory disturbances in

MSA remained generally unassociated with marked SaO2

changes, especially when MSA was compared to OSAS.

Mild (93.3%) O2 desaturation, however, occurred during

sleep in MSA even unassociated with OSAS or stridor,

indicating that sleep-related O2 desaturation may be

intrinsic to the disease. SaO2 was only slightly worsened

in MSA associated with OSAS (mean SaO2 at 92.1%) or

Fig. 2. NREM sleep recording (excerpts) in an MSA-C patient (no. 17) with inspiratory stridor (microph.). Abbreviations as per Fig. 1. Note the additional

intercostalis (interc.) expiratory activation with concomitant slightly positive endoesophageal pressure, the normal thoracic and abdominal respiratory traces

and the absence of significant SaO2 changes.

Table 3

Mean breathing and heart rates during Wake, NREM and REM sleep

Wake

(mean ^ SD)

NREM sleep

(mean ^ SD)

REM sleep

(mean ^ SD)

Mean breathing rate

MSA patients (no. 19) 19 ^ 4 20 ^ 6 21 ^ 5

Without stridor (no. 11) 19 ^ 2 18 ^ 3 19 ^ 3

With stridor (no. 8) 19 ^ 5 23 ^ 8 23 ^ 9

Mean heart rate

MSA patients (no. 19) 70 ^ 10 67 ^ 11 69 ^ 11

Without stridor (no. 11) 69 ^ 10 65 ^ 10 66 ^ 10

With stridor (no. 8) 72 ^ 11 71 ^ 13 73 ^ 12

SD, standard deviation.

R. Vetrugno et al. / Sleep Medicine 5 (2004) 21–3026

stridor (mean SaO2 at 92.6%). Noticeably, our MSA

patients did not demonstrate worsened SaO2 during REM

sleep and the mean RDI observed during NREM sleep was

higher than during REM sleep, an unusual finding when

compared to the opposite pattern typical of OSAS and

confirmed also in our cohort of OSAS controls. Except for

further emphasizing that REM sleep is distinctly abnormal

in MSA, these latter findings remain difficult to explain.

Finally, sleep-related respiratory disturbances, in particular

stridor, were more frequent in MSA type C.

Fig. 3. Polygraphic recordings in MSA-C patient (no. 16, panel A) show markedly blunted systemic blood pressure oscillations related to the respiratory events

when compared to OSAS (panel B). Abbreviations as per Fig. 1. Time scale is identical in the two panels.

Fig. 4. NREM sleep recordings (excerpts) in MSA type C patient (patient no. 3) showing subcontinuous motor unit potential discharges in the intercostalis and,

to a lesser extent, tibialis anterior muscles EMG. Abbreviations as per Fig. 1. Delt, deltoid muscle; tib, tibialis anterior.

R. Vetrugno et al. / Sleep Medicine 5 (2004) 21–30 27

5.3. Heart and breathing rates

When considering BR and HR, patients with stridor,

already tachypnoic during Wake, had a significant increase

in BR from Wake to NREM and REM sleep. Moreover,

their HR did not significantly decrease in NREM and

actually increased in REM sleep. These findings, striking in

the face of the characteristic fixed pulse rate in MSA,

indicate impaired sleep state-related heart and respiratory

regulation [32–34] in MSA with stridor. In agreement with

the autonomic denervation typical of MSA, however, the

oscillations in HR and blood pressure associated with the

respiratory events were markedly blunted. Abnormal

circadian variations in HR with decreased correlation

between HR and blood pressure have already been reported

in MSA [35]. Our BR findings are also in good agreement

with Isono et al. who reported increased respiratory

frequency during sleep compensating for a reduction of

tidal volume in MSA with stridor [36].

5.4. Motor abnormalities

Major findings of our study were also the abnormal

motor patterns involving limb and respiratory muscles

during sleep. Sustained EMG activity in the intercostalis

muscle, especially during wake but persisting throughout

sleep, was present in MSA patients, in particular those with

stridor. Together with intercostalis expiratory EMG acti-

vation and PB, these findings indicate a pathological sleep-

related respiratory muscle pattern. EMG tonic activity was

not, however, restricted to muscles related to respiration but

also involved skeletal limb muscles such as the tibialis

anterior, in which both tonic discharges of motor unit EMG

potentials and more complex patterns of PLMS were found.

This respiratory and limb motor overactivity throughout

wake and sleep was always more prevalent in MSA with

stridor. Finally, RBD was found in all of our MSA patients.

We entertained the hypothesis that the increased EMG

activity could be due to increased respiratory effort. None of

the 10 OSAS patients, however, in whom airflow restriction

and respiratory effort as measured by endoesophageal

balloon, were clearly more severe, presented sustained

tonic type intercostalis and tibialis anterior EMG activity

during wake or sleep; none had RBD or abnormal lack of

atonia during REM sleep. This clearly shows that the

abnormal sustained EMG activity of respiratory and limb

muscles in MSA does not simply relate to increased

respiratory effort. We think that it reflects central overactive

and disordered motor output, which could best be defined as

a kind of dystonia. Recent EMG data indeed suggest that

stridor in MSA is not simply due to denervation of laryngeal

muscles but rather to abnormal motor control similar to

dystonia [37]. We also consider that the universal

presence of RBD in MSA fits this interpretation of the

EMG findings, reflecting dysfunction in inhibitory brain-

stem pathways [38,39].

5.5. Final considerations

Our data show that disordered breathing and abnormal

motor control during sleep are common and often associated

features in MSA. The clinical relevance of these findings is

underlined by their high prevalence in our cohort of patients,

who were not expressely selected for respiratory or motor

disturbances during sleep. Respiratory and motor sleep-

related abnormalities should rightly be incorporated within

the useful diagnostic markers of the disease [20]. In MSA

with stridor in particular, abnormal breathing patterns and

motor abnormalities during sleep represent a constant

feature found concomitantly in all of our patients.

Our study allows for some pathophysiological consider-

ations on the relevance of stridor in MSA. Stridor is related

to a poorer prognosis in MSA, but how exactly this happens

remains unclear. Recordings in nine MSA patients with

stridor demonstrated airflow limitation only during inspira-

tion, associated with laryngeal narrowing and phasic

inspiratory thyro-arytenoid (TA) muscle EMG activity

[36]. Since TA is reflexively activated in normal subjects

by external resistive loads, stridor in MSA was considered

to result from hyperactivity of the TA uncontrasted by the

denervated posterior crico-arytenoid muscles, which are

the vocal cord abductors. Indeed CPAP, by abolishing the

resistive load, abolished the abnormal TA activity and

stridor [36]. The posterior crico-arytenoid muscles were not,

however, specifically searched for denervation in this study

[36]. Stridor in MSA is thus associated with laryngeal

narrowing, at variance with OSAS, in which obstruction in

airflow occurs at several levels in the pharynx [40–43].

Table 4

EMG activity in MSA and OSAS patients expressed as percentage of patients having such activity during wake, NREM sleep and REM sleep

Intercostalis EMG

activity (%)

Tibialis anterior EMG

activity (%)

PLMS (%) Mylohyoideus EMG activity RBD (%)

Wake NREM REM Wake NREM REM Wake NREM REM

MSA patients (no. 19) 89 79 37 76 65 24 88 100 100 100 100

Without stridor (no. 11) 91 64 27 73 55 18 82 100 100 100 100

With stridor (no. 8) 87 100 50 83 83 33 100 100 100 100 100

OSAS patients (no. 10) 0 0 60 Atonia in REM sleep 0

TA, tibialis anterior; PLMS, periodic limb movement during sleep; RBD, REM sleep behaviour disorder.

R. Vetrugno et al. / Sleep Medicine 5 (2004) 21–3028

SaO2 and endoesophageal pressure values during sleep

were not, however, particularly severe in MSA with stridor,

and mild when compared to OSAS. In our opinion, stridor

just represents a harbinger of severe changes in autonomic

centres controlling respiration and other vital functions,

accounting for death in MSA. Recently, impaired chemo-

sensitivity to hypoxia was reported as an early marker and a

risk factor for respiratory failure and sudden death in MSA

patients, independently of any mechanical airway obstruc-

tive factors [44]. Our patients had a mild degree of O2

desaturation during sleep even in the absence of stridor or

OSAS. If the impaired chemoceptor response to hypoxia

makes MSA patients unable to compensate for worsening

ventilatory insufficiency, even slight additional O2 desatura-

tion due to the combined effects of sleep and stridor or

OSAS may then prove fatal. These considerations beg the

still unanswered question of the type, timing and extent of

assisted ventilation needed to overcome the sleep-related

respiratory disturbances and the risk for sudden death

in MSA.

Acknowledgements

A. Laffi gave invaluable secretarial help, and A. Collins

corrected the English manuscript. Supported by MURST

ex-60% 2000, MURST ex-40% cofin 2000 prot.

MM06244347_004 and MURST ex-40% cofin 1999 prot.

9906037938 grants.

References

[1] Ghorayeb I, Yekhlef F, Chrysostome V, et al. Sleep disorders and their

determinants in multiple system atrophy. J Neurol Neurosurg

Psychiatry 2002;72:798–800.

[2] Chokroverty S. Sleep, breathing and neurological disorders. In:

Chokroverty S, editor. Sleep disorders medicine. Boston, MA:

Butterworth-Heineman; 1994. p. 295–335.

[3] Williams A, Hanson D, Calne DB. Vocal cord paralysis in the Shy-

Drager syndrome. J Neurol Neurosurg Psychiatry 1979;42:151–3.

[4] Chester CS, Gottfried SB, Cameron DI, Strohl KP. Pathophysiological

findings in a patient with Shy-Drager and alveolar hypoventilation

syndromes. Chest 1988;94:212–4.

[5] Castaigne P, Laplane D, Autret A, et al. Shy-Drager syndrome with

disturbances of the respiratory rhythm and consciousness. A propos of

an anatomo-clinical case. Rev Neurol 1977;133:455–6.

[6] Guilleminault C, Lehrman AT, Forno L, Dement WC. Sleep apnoea

syndrome: states of sleep and autonomic dysfunction. J Neurol

Neurosurg Psychiatry 1977;40:718–25.

[7] McNicholas WT, Rutherford R, Grossman R, et al. Abnormal

respiratory pattern generation during sleep in patients with autonomic

dysfunction. Am Rev Respir Dis 1983;128:429–33.

[8] Chokroverty S, Sachdeo R, Masdeu J. Autonomic dysfunction and

sleep apnea in olivopontocerebellar degeneration. Arch Neurol 1984;

41:926–31.

[9] Teravainen H, Udd B. Vocal cord paralysis in the Shy-Drager

syndrome. Acta Neurol Scand 1982;66:505–7.

[10] Martinovits G, Leventon G, Goldhammer Y, Sadeh M. Vocal cord

paralysis as a presenting sign in the Shy-Drager syndrome. J Laryngol

Otol 1988;102:280–1.

[11] Sadaoka T, Kakitsuba N, Fujiwara Y, et al. Sleep-related breathing

disorders in patients with multiple system atrophy and vocal fold

palsy. Sleep 1996;19:479–84.

[12] Wenning GK, Tison F, Ben-Shlomo Y, et al. Multiple system atrophy:

a review of 203 pathologically proven cases. Mov Disord 1997;12:

133–47.

[13] Kakitsuba N, Sadaoka T, Kanai R, et al. Peculiar snoring in patients

with multiple system atrophy: its sound source, acoustic character-

istics and diagnostic sygnificance. Ann Otol Rhinol Laryngol 1997;

106:380–4.

[14] Wetter TC, Collado-Seidel V, Pollmacher T, et al. Sleep and periodic

leg movement patterns in drug-free patients with Parkinson’s disease

and multiple system atrophy. Sleep 2000;23:361–7.

[15] Tachibana N, Kimura K, Kitajima K, et al. REM sleep without atonia

at early stage of sporadic olivopontocerebellar atrophy. J Neurol Sci

1995;132:28–34.

[16] Tison F, Wenning GK, Quinn NP, Smith SJ. REM sleep behaviour

disorder as the presenting symptom of multiple system atrophy.

J Neurol Neurosurg Psychiatry 1995;58:379–80.

[17] Plazzi G, Corsini R, Provini F, et al. REM sleep behaviour disorders in

multiple system atrophy. Neurology 1997;48:1094–7.

[18] Plazzi G, Cortelli P, Montagna P, et al. REM sleep behaviour

disorders differentiates pure autonomic failure from multiple system

atrophy with autonomic failure. J Neurol Neurosurg Psychiatry 1998;

64:683–5.

[19] Olson EJ, Boeve BF, Silber MH. Rapid eye movement sleep

behaviour disorders: demographic, clinical and laboratory findings

in 93 cases. Brain 2000;123:331–9.

[20] Gilman S, Low PA, Quinn A, et al. Consensus statement on diagnosis

of multiple system atrophy. J Neurol Sci 1999;163:94–8.

[21] Rechtschaffen A, Kales A. A manual of standardized terminology,

techniques and scoring system for sleep stages of human subjects. Los

Angeles: Brain Information Service, Brain Research Institute; 1968.

[22] Guilleminault C, van den Hoed J, Mitler MM. In: Guilleminault C,

Dement C, editors. Sleep apnea syndromes. New York: Alan R. Liss;

1978. p. 1–12.

[23] Block AJ, Boysen PG, Wynne JM, Hunt LA. Sleep apnea, hypopnea

and oxygen desaturation in normal subjects. A strong male

predominance. N Engl J Med 1979;300:513–7.

[24] Kryger MH. Monitoring respiratory and cardiac function. In: Kryger

M, Roth T, Dement W, editors. Principles and practice of sleep

medicine, 3rd ed. Philadelphia, PA: W.B. Saunders; 2000. p.

1217–30.

[25] Polysomnography task force, American Sleep Disorders Association

Standards of Practice Committee, Practice parameters for the

indications for polysomnography and related procedures. Sleep

1997;20:406–22.

[26] American Academy of Sleep Medicine, Sleep related breathing

disorders in adults: recommendations for syndrome definition and

measurement techniques in clinical research. Sleep 1999;22:667–89.

[27] Coleman RM. Periodic movement in sleep (nocturnal myoclonus) and

restless legs syndrome. In: Guilleminault C, editor. Sleeping and

waking disorders: indications and techniques. Menlo Park, CA:

Addison-Wesley; 1982. p. 265–95.

[28] Kimura J. Techniques and normal findings. In: Kimura J, editor.

Electrodiagnosis in diseases of nerve and muscle: principles and

practice, 2nd ed. Philadelphia, PA: F.A. Davis; 1989. p. 227–48.

Chapter 12.

[29] Mahowald MW, Schenck CM. REM sleep parasomnias. In:

Kryger M, Roth T, Dement W, editors. Principles and practice of

sleep medicine, 3rd ed. Philadelphia, PA: W.B. Saunders; 2000.

p. 724–37.

[30] Carskadon MA, Dement WC. Normal human sleep: an overview.

In: Kryger M, Roth T, Dement W, editors. Principles and practice

R. Vetrugno et al. / Sleep Medicine 5 (2004) 21–30 29

of sleep medicine, 3rd ed. Philadelphia, PA: W.B. Saunders; 2000.

p. 15–25.

[31] Petit JM, Milic-Emili G. Measurement of endoesophageal pressure.

J Appl Physiol 1958;3:481–5.

[32] Zemaityte D, Varoneckas G, Sokolov E. Heart rhythm control during

sleep. Psychophysiology 1984;21:279–89.

[33] Hornyak M, Cejnar M, Elam M, et al. Sympathetic muscle nerve

activity during sleep in man. Brain 1991;114:1281–95.

[34] Okada H, Iwase S, Mano T, et al. Changes in muscle sympathetic

nerve activity during sleep in humans. Neurology 1991;41:1961–6.

[35] Kitae S, Murata Y, Tachiki N, et al. Assessment of cardiovascular

autonomic dysfunction in multiple system atrophy. Clin Auton Res

2001;11:39–44.

[36] Isono S, Shiba K, Yamaguchi M, et al. Pathogenesis of laryngeal

narrowing in patients with multiple system atrophy. J Physiol 2001;

536:237–49.

[37] Merlo IM, Occhini A, Pacchetti C, Alfonsi E. Not paralysis, but

dystonia causes stridor in multiple system atrophy. Neurology 2002;

58:649–52.

[38] Benarroch EE. New findings on the neuropathology of multiple

system atrophy. Auton Neurosci 2002;96:59–62.

[39] Benarroch EE, Schmeichel AM. Depletion of cholinergic mesopon-

tine neurons in multiple system atrophy: a substrate for Rem

behaviour disorders? Neurology 2002;58(Suppl. 3):A345. Abstract.

[40] Lugaresi E, Coccagna G, Cirignotta F. Polygraphic and cineradio-

graphic aspects of obstructive apneas occurring during sleep:

physiopathological implications. In: Von Euler C, Lagercrantz H,

editors. Central nervous control mechanisms in breathing. Oxford:

Pergamon Press; 1979. p. 495–501.

[41] Fujita S, Conway W, Zorick F, Roth T. Surgical correction of anatomic

abnormalities in obstructive sleep apnea syndrome: uvulo-palato-

pharyngo-plasty. Otolaryngol Head Neck Surg 1981;89:923–34.

[42] Lugaresi E, Cirignotta F, Montagna P. Pathogenic aspects of snoring

and obstructive apnea syndrome. Schweiz Med Wochenschr 1988;

118:1333–7.

[43] Kuna S, Remmers JE. Anatomy and physiology of upper airway

obstruction. In: Kryger M, Roth T, Dement W, editors. Principles and

practice of sleep medicine. Philadelphia, PA: W.B. Saunders; 2000. p.

840–58.

[44] Tsuda T, Onodera H, Okabe S, Kilichi Y, Itoyama Y. Impaired

chemosensitivity to hypoxia is a marker of multiple system atrophy.

Ann Neurol 2002;52:367–71.

R. Vetrugno et al. / Sleep Medicine 5 (2004) 21–3030