depresie

Vol 27, No.3, 2005

A journal ofmedical informationand internationalcommunicationfrom Servier

213

223

228

236

240

247

254

Contents continued overleaf...

84I S S U E

ISSN 0243-3397

Medicographia

EDITORIAL

WHY DO WE NEED NEW AND BETTER

ANTIDEPRESSANTS? POURQUOI AVONS-NOUS BESOIN D’ANTIDÉPRESSEURS

INNOVANTS ET PLUS EFFICACES ?

CHRONOBIOLOGICAL STRATEGIES FOR UNMET NEEDS

IN THE TREATMENT OF DEPRESSION

NEW PERSPECTIVES IN THE PATHOPHYSIOLOGY AND

TREATMENT OF AFFECTIVE DISORDERS: THE ROLE

OF MELATONIN AND SEROTONIN

SEVERE DEPRESSION: FROM DIAGNOSIS TO TREATMENT

SHORTCOMINGS OF CURRENT ANTIDEPRESSANT

THERAPIES

SEASONAL AFFECTIVE DISORDER: FROM DIAGNOSIS

TO TREATMENT

TRENDS IN DIAGNOSIS AND TREATMENT OF BIPOLAR

DISORDERS

S. MONTGOMERY, UNITED KINGDOM

A. WIRZ-JUSTICE, SWITZERLAND

M. HAMON, P.-A. BOYER

AND E. MOCAËR, FRANCE

J.-P. OLIÉ, FRANCE

S. H. KENNEDY, CANADA

D. WINKLER AND

S. KASPER, AUSTRIA

D. J. KUPFER AND

E. FRANK, USA

Unmet Needs in theTreatment of Depression

Vol 27, No.3, 2005 Medicographia

84I S S U E

M. P. DEVA, BRUNEI / T. PARTONEN, FINLAND /A. SOYKAN, TURKEY /A. C. ALTAMURA, ITALY / A. B. SMULEVITCH, RUSSIA /B. MILLET, FRANCE / J. SAIZ-RUIZ, SPAIN /V. N. VAHIA, INDIA

N. SARTORIUS,SWITZERLAND

R. EMSLEY, SOUTH AFRICA

C. RÉGNIER, FRANCE

I. SPAAK, FRANCE

CONTROVERSIAL QUESTION

WHAT ARE THE CORE SYMPTOMS OF DEPRESSION?

INTERVIEW

UNMET NEEDS IN DEPRESSION (AND OTHER MENTAL

DISORDERS)

FOCUS

THE IMPACT OF SLEEP DISORDERS ON THE COURSE

OF DEPRESSION

A TOUCH OF FRANCE

HYGIA VERSUS POLYMNIA: SOME FRENCH PAINTERS AND THEIR DISEASES

A TOUCH OF FRANCE

IN VAN GOGH’S FOOTSTEPS IN AUVERS-SUR-OISE

261

270

273

279

288

...Contents continued from cover page

Medicographiaa Servier publication

Editor in Chief: Jean-Philippe Seta, MD

Editorial Board: Laurence Alliot, PharmD;Christophe Charpentier, MD; WilliamGaussens, PharmD; Yves Langourieux, PhD; Didier Mochon, PharmD; AntoineMoukheiber, BSc Ph; Pascal Poullalié, MA;Frédéric Sesini, PharmD

Publisher: Laurence Alliot, PharmDProduction Manager: Noelle GuénotMedical Publications Editor:David Mason, MDProduction Editor: Iain Matheson, MB ChB Editorial Assistant: Judit SiklosiDesign & Layout: Myriam Bucquoit andBernard Crespin

Medicographia is published 4 times a yearand circulated throughout the world, inAfrica, the Americas, Asia, Australasia, andEurope, by Les Laboratoires Servier –22, rue Garnier – 92578 Neuilly sur SeineCedex – France, and printed by ImprimerieKapp Lahure Jombart, rue de l’Industrie Z.I.N°1 27025 Evreux Cedex

Printed in FranceDirecteur de la Publication:

Jean-Philippe Seta, MD

© 2005 by Les Laboratoires Servier

All rights reserved throughout the world andin all languages. No part of this publicationmay be reproduced, transmitted, or stored inany form or by any means either mechanicalor electronic, including photocopying, record-ing, or through an information storage andretrieval system, without the written permis-sion of the copyright holder. Opinions ex-pressed do not necessarily reflect the views ofthe publishers, editors, or editorial board.The authors, editors, and publishers cannotbe held responsible for errors or for any con-sequences arising from the use of the infor-mation contained in this journal.

Address for correspondence: Servier International / Medical PublishingDivision – 192, avenue Charles de Gaulle, 92578 Neuilly sur Seine Cedex – FranceTel. + 33 1 55 72 60 00 / Fax 68 88

Unmet Needs in theTreatment of Depression

MedicographiaA S e r v i e r p u b l i c a t i o n

212 MEDICOGRAPHIA, VOL 27, No. 3, 2005

Arabian Gul f (UAE)Ms D. EL ASSAADLes Laboratoires Servier Representative officeAPI World Tower 8th - Flat 801Sheikh Zayed Road - DubaiTel: 971 4 332 99 03

Argent inaDr F. MORIERServier Argentina S.A.Avenida Belgrano 14801093 Buenos Aires Tel: (54-11) 4383.5061

Austral iaMr P. TÊTEServier Laboratories Pty LtdServier House, 8 Cato StreetHawthorn, Victoria 3122Tel: (61-3) 88237333

Austr iaMr O. KIRSTServier Austria GmbHMariahilferstrasse 20/5 1070 WienTel: (43-1) 5243999

BelgiumMr A. CRULSServier Benelux SARiverside Business ParkBoulevard International No. 571070 BruxellesTel: (32-2) 5294311

Brazi lMs V. TOPPJIANLaboratorios Servier do Brasil LtdaEstrada dos Bandeirantes 4211 - JacarepaguaRio de Janeiro, RJ CEP: 22-775-113Tel: (55-21) 2142-1414

CanadaMr M. SUMPTERServier Canada Inc.235, boulevard Armand FrappierLaval, Québec H7V 4A7Tel: (1-450) 9789700

China (PRC)Mr D. CHINGServier InternationalBeijing OfficeRoom B 801 Han Wei PlazaNo. 7, Guanghua RoadChaoyang District, Beijing 100004People’s Republic of ChinaTel: 86 10 6561 0341

Czech Republ icM. F. D’HOUDAINServier SRO Praha City Center Klimentska 46, 110 02 Praha 1Tel: (420-2) 22118602

DenmarkMr P. VAN MUYLDERSServier Danmark A/SRoskildevej 39 A2000 FrederiksbergTel: (45-36) 442260

EcuadorDr J. LIBIOTLes Laboratoires ServierCasa de RepresentaciónAv. República de El Salvador 734 y PortugalEdif. Athos, of. 201, QuitoTel: (593.2) 2 252-522

EgyptDr T. KHANServier EgyptScientific Office67, El Horreya St., HeliopolisP.O.Box 123 HeliopolisTel: (20-2) 290 60 77

FinlandDr E. RAIMBAULTServier Finland OYAyritie 12C (3rd Floor)01510 VantaaTel: 358 9 279 8080

FranceMr F. SESINILes Laboratoires Servier 22, rue Garnier92578 Neuilly sur Seine CedexTel: 33 1 55 72 60 00

GermanyMr A. BRETHOUSServier Deutschland GmbHWestendstr. 170 80686 München 21Tel: (49-89) 5709501

GreeceMr A. LEKKOSServier Hellas181, Syngrou Avenue171 21 Nea SmyrniTel: (302-10) 9391000

Hong Kong Mr S. CHEUNGServier Hong KongRoom 1901, 19th Floor The Lee Gardens33 Hysan Avenue, Causeway BayTel: 852 2577 1922

HungaryMr F. PARISOTServier Hungaria KFTWestend OfficeB tower, 3rd floor Váci út 1-3, Budapest 1062Tel: (36-1) 238 77 99

IndiaMr J. Y. GALSerdia Pharmaceuticals (India Pvt) Ltd, Serdia HouseOff Dr S.S. Rao Road, ParelMumbai 400 012Tel: (91-222) 4162726

IndonesiaMr D. GORMANDPT Parvico BersaudaraMenara Kadin Indonesia 18th Floor JL. H.R. Rasuna Said Blok X-5 Kav. 2-3, Jakarta 12950Tel: (62-21) 57903940

Ire landMr A. VERNONServier Laboratories LtdBlock 2 West Pier Business Campus Old Dunleary RoadDun Laoghaire Co. DublinTel: (353-1) 2694488

I ta lyMr B. MILLETServier Italia SpAVia Luca Passi, 8500166 RomaTel: (39-06) 666081

JapanMr J. AMIRAULTNihon Servier Company LtdHongo MK Building 5F1-28-34, Hongo Bunkyo-Ku 113-0033 TokyoTel: (81-3) 5842 7111

KoreaMr F. DEBAILLON-VESQUEServier Korea, 12th Floor Hansol M Com Building 132-11 Seocho-Dong Seocho-Gu Seoul 137-070 Tel: (82-2) 34158510

LuxembourgDr S. BARSOUMServier Luxembourg3a, rue Guillaume Kroll1882 Luxembourg GasperichTel: (352) 493435

MalaysiaMr J. LEONGServier Malaysia SDN BHDLot 1301, Level 13 Uptown 22 Jalan SS21/37Damansara Uptown47400 Petaling JayaSelangor Darul EhsanTel: 603 772 638 66

MexicoDr T. CHAPUISLaboratorios Servier de México S.A. de C.V. Palmas 830 Lomas de ChapultepecC.P. 11000 - México D.F. Tel: (52-55) 52 02 33 59

MoroccoDr E. PRADEREServier Maroc Espace Jet Business ClassLotissement 16 19 Ettawfik la Colline 20190 Casablanca Tel: 212 22 97 48 49

NetherlandsMr L. G. LASSONNERYServier Nederland Farma BVEinsteinweg 82 2333 CB LeidenTel: (31-71) 5246700

PakistanMr J. CABANNESServier Research &Pharmaceuticals (Pakistan) PVT Ltd 65 Main Boulevard Gulberg, LahoreTel: (92-42) 5879500

Phi l ippinesMr P. MOLHANTServier Philippines, Inc.2 Orion Cor. Mercedes st.Bel Air Village - MakatiTel: 63-2 897 89 90 to 97

PolandMr P. ETORREServier PolskaUI. Sienna 7500-833 WarsawTel: (48-22) 820 14 00

PortugalDr A. MENDESServier PortugalAvenida Antonio Augusto de Aguiar, 128 1069-133 LisboaTel: (351-21) 3122000

Russian Federat ionDr F. GIRARDZAO Servier 2 Paveletskaya Pl., Bât. 1113054 Moscow - RussiaTel: (7-502) 937 07 00or (7-095) 937 07 00

Saudi ArabiaMr D. LEFAYLes Laboratoires ServierScientific Office401 Abdullatif Plaza Al Faisaliyah Dist.King Fahd Street P.O. Box 112744, Jeddah 21371Tel: 966 2 697 69 97

SingaporeDr D. KANDELAServier (S) Pte Ltd 510 Thomson Road#09-02 SLF BuildingSingapore 2981 35Tel: 65 6250 0377

Slovak Republ icMr M. ACCOLASServier Slovensko Spol S.r.o.Mostova 2 81102 BratislavaTel: 421 2 59204111

South Afr icaMr S. SPELLERServier Laboratories Pty LtdDevcon Road, 7 Autumn StreetRivonia 2128Tel: (27-11) 233 60 00

SpainMr P. FARALDOLaboratorios Servier SAAvenida de los Madroños 3328043 MadridTel: (34-91) 7489630

Switzer landMr S. NICOLLERATServier Suisse SA10, rue de la Bergère Case Postale 380 CH-1217 Meyrin 1Tel: (41-22) 7851480

Taiwan (ROC)Mr R. LOTaiwan Servier (S) Pte Ltd3rd Floor, No 168 Sung Chiang Road Taipei 104Tel: 886 2 2568 1380

Thai landDr G. LALANDEServier Thailand LtdPloenchit Center Building 15th Floor2 Sukhumvit Road, KlongtoeyBangkok 10110Tel: 66 2 656 8388

TurkeyMr B. PARRINServier Ilaç Ve Arastirma ASMaslak Giz Kule Meydan Sokak, No28 Kat: 22-2380650 Maslak IstanbulTel: (90-212) 3291400

United KingdomMr E. FALCANDServier Laboratories LtdGallionsWexham SpringsFramewood Road Wexham Slough SL3 6RJTel: (44-1753) 662744

United StatesMr Y. LANGOURIEUXServier International163, avenue Achille Peretti92578 - Neuilly sur Seine Cedex(France)Tel: (33-1) 55 72 60 00

VenezuelaMr R. SOLORZANOLaboratoires ServierEdificio Centro BerimerFinal Calle Vargas, PI-A2Boleita Norte Caracas 10062-ATel: (58-212) 2381155

VietnamMr P. LEQUEUXLes Laboratoires Servier162 Pasteur Street (9th Floor) District 1 Ho Chi Minh CityTel: 84 88 298 208

West IndiesMr M. LOUSTRICServier West Indies Ltd“Stafford House”The Garrison St Michael, BarbadosTel: (1-246) 4277667

I n t e r n a t i o n a l A d v i s o r y C o m m i t t e e

213Editorial – Montgomery MEDICOGRAPHIA, VOL 27, No.3, 2005

MAJOR DEPRESSION IS A SERIOUS DISORDER WHOSE LIFE-time prevalence of 16.2% in the USA1 is being increasingly echoedin contrasting cultures, eg, Japan and China. The apparent in-crease requires qualification: people may be readier to discuss

psychological problems than previously, and early studies were probably underestimates. In ad-dition, differential memory distorts retrospective estimates of lifetime prevalence. People forgetdistant episodes of depression and inflate current episodes: follow-up of patients hospitalized fordepression found that 25 years later only half could recall sufficient symptoms to justify a diag-nosis of major depression.2

Major depression is associated with substantial social and even physical dysfunction, signif-icantly more than some chronic medical conditions, eg, diabetes. Onset is typically in the teensand twenties and the course commonly recurrent or chronic, with depressive episodes occupy-ing 20% of postdiagnosis life. Unsurprisingly, therefore, major depressive disorder is now theleading cause of disability in those of active working age. Some 10% to 15% of depressed patientseventually commit suicide. A meta-analysis found a standard mortality rate, calculated by com-paring the suicide rate in a specific group with that in the general population, of 21.2 among de-pressives; mood disorders shorten life by 10 years from the combined effect of increased suiciderisk and increased physical illness.3

Not only do current antidepressants not “cure” the underlying condition, they are also onlymoderately effective in relieving symptoms during episodes. A National Institute of Mental Health(NIMH) follow-up of 431 patients seeking treatment found that 12% remained chronically de-pressed over 5 years, while 55% had suffered a relapse or recurrence, and only one third remainedhealthy4; after 15 years, 82% had had a recurrence, 6% remained chronically depressed, and only12% remained healthy. Given sufficiently long and careful follow-up, almost all those treated formajor depression will fail to recover, or suffer a recurrence.

Nevertheless, there is an overwhelming case for providing appropriate treatment for depres-sion, on both personal and socioeconomic grounds. Yet undertreatment remains widespread.Of the 17% with major depression identified in a European survey of 78 000 adults, 69% werereceiving no medical treatment, 43% had not even consulted a doctor, and under 8% were on anantidepressant.5 In addition, antidepressant treatment is often suboptimal: in a survey of 1 mil-lion patients in primary care, 89% were receiving a subtherapeutic dose of tricyclic antidepres-sant (TCA).6 Fortunately, underprescribing has become less common with newer antidepressantssuch as the selective serotonin reuptake inhibitors (SSRIs), which are only marketed at thera-peutic doses.

Poor treatment response is often due to poor patient compliance: 30% of primary care patientsnever fill their first prescription, while a further 25% to 33% stop treatment in the first month,and 62% fail to inform their physician accordingly. As a result, treatment duration in 55% to63% of cases is less than the minimum consensus recommendation of 6 months post remis-sion.7 However, even in well-conducted studies with various classes of antidepressant adminis-tered for the 6 to 8 weeks generally thought sufficient to show a drug/placebo difference, non-response rates average 30% to 40%.

E D I T O R I A L

Address for correspondence:Prof Stuart Montgomery,Imperial College School ofMedicine, University of London,London W13 8WH, UK(e-mail:[email protected])

Medicographia.2005;27:213-221.

Why do we need new and better antidepressants?

b y S . M o n t g o m e r y , U n i t e d K i n g d o m

Stuart MONTGOMERY, MDImperial College School of Medicine University of London London UNITED KINGDOM

Developments of antidepressantsThe presumed mechanism of action of both monoamine oxidase inhibitors (MAOIs) and TCAswas that they blocked the reuptake of the neurotransmitter monoamines, norepinephrine andserotonin (5-HT), thus increasing their availability in the synaptic cleft. Depression was pre-sumed due to subnormal neurotransmitter levels. Although oversimple and incomplete, thistheory provided a broad basis for the development of most subsequent antidepressants. Researchfocused on the design of compounds with greater selectivity for either monoamine, hence withfewer side effects. Maprotiline and reboxetine were examples of selective norepinephrine reup-take inhibitors (SNRIs). But most effort concentrated on the development of SSRIs, culminat-ing in fluoxetine, paroxetine, fluvoxamine, sertraline, citalopram, and escitalopram, which nowdominate the antidepressant market.

Several TCAs were found to act on 5-HT2 receptors: amitriptyline both blocked 5-HT re-uptake and was a 5-HT2 antagonist. This inspired the development of 5-HT2 antagonists suchas trazodone, followed by compounds more selective still, acting on 5-HT2 receptor subtypes,such as agomelatine, a 5-HT2C antagonist and melatonin M1/M2 agonist, and mianserin andmirtazapine, which are 5-HT2A and 5-HT2C receptor antagonists as well as α 2 antagonists and5-HT1A and 5-HT3 antagonists.

Superiority of particular antidepressantsClinical trials of antidepressants have failed to demonstrate a consistently superior response toactive treatment versus placebo.8 This is due, irrespective of the inherent (in)efficacy of the an-tidepressants themselves, to the inappropriate design and conduct of the studies, and some ma-jor theoretical and practical obstacles. One important failing has been inadequate power, whilea number of “unsuccessful” studies have never been published. Smaller studies demonstratingsignificant efficacy were often confined to severely depressed inpatients, whereas the move tomore realistic community-based studies recruited many patients with mild depression for whichdrug/placebo difference was difficult to demonstrate. Even the diagnostic criteria for major de-pressive disorder have proved inadequate, with particular regard to illness duration. Whereas theDiagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV) criteria require only2 weeks, compatible with a self-limiting disorder, investigators have long since increased the min-imum episode duration to 1 month to qualify for study inclusion.

Drug/drug difference is even more difficult to demonstrate. SSRIs may be safer and bettertolerated than TCAs, but they appeared significantly less effective than conventional clomipraminein one key study.9 Despite its methodological weaknesses, this study suggested that selectivityfor the serotonin system might not be universally beneficial. A plausible conclusion was thatselective antidepressants would only be effective in patients with the relevant monoamine defi-ciency, a theory bolstered by the reports from several subsequent studies that SNRIs such asvenlafaxine were more effective than SSRIs,10 which were themselves significantly better thanplacebo. However, this conclusion failed to secure consensus conviction, even with the backingof a meta-analysis, due to an array of methodological failings.

A rare opportunity for valid interdrug meta-analysis was afforded by the pooled studies ofthe SSRIs escitalopram and citalopram, in that all were of the same 8-week duration and usedthe Montgomery and Asberg Depression Rating Scale (MADRS): several meta-analyses acrossa variety of conditions reached the consensus conclusion that escitalopram was the more effec-tive agent.11

Early onset of responseA well-recognized shortcoming of antidepressants is that significant difference from placebo canonly be identified reliably some 6 weeks after initiating treatment. The delay is potentially dan-gerous since suicide risk is highest in the early treatment period when severity is greatest. Ab-sence of rapid positive feedback also encourages noncompliance. Earlier onset of effect wouldtherefore be welcome. Unfortunately, however, virtually no studies have investigated this aspectin individual antidepressants. Retrospective review of placebo-controlled studies with venlafax-ine indicates simply that response before 2 weeks is seen with early rapid escalation to high doses,while comparative studies have suggested that escitalopram induces a more rapid response thancitalopram.12

Efficacy of specific symptomsSome symptoms improve faster with one antidepressant than another.

E D I T O R I A L

214 Editorial – MontgomeryMEDICOGRAPHIA, VOL 27, No.3, 2005

◆ SleepAntidepressants with antihistaminic activity, eg, amitriptyline and mirtazapine, improve sleepearly in treatment, but at the cost of compromised daytime alertness and impaired concentra-tion. 5-HT2 antagonists, eg, nefazodone, improve sleep with no such penalty, as does the mela-toninergic agonist and selective 5-HT2C antagonist agomelatine.

◆ Pain and somatic symptomsSNRIs may be more effective in relieving pain syndromes. Thus, venlafaxine was superior to place-bo in diabetic neuropathy. In fact, SNRIs may be more effective on the somatic symptoms of de-pression generally.13 While these are not the core symptoms of the disorder, they may explainthe apparent advantage of SNRIs over SSRIs, particularly when remission is assessed using theHamilton Depression Rating Scale (HDRS) with its overrepresentation of somatic symptoms. Theadvantage may be less apparent when using a core symptoms scale such as the MADRS.

Discontinuation of medicationEarly studies with TCAs established that abrupt treatment withdrawal provoked a syndrome ofnausea, vomiting, headaches, giddiness, chills, weakness, and musculoskeletal pain in some pa-tients. Symptoms were transient, maximal in the first week and declined thereafter, rarely inter-fering with function.

A discontinuation syndrome was also reported with the SSRIs, whether withdrawal was grad-ual or abrupt. Its timing appeared dependent on the half-life of the compound concerned and anyactive metabolite. With fluoxetine, for example, the discontinuation syndrome may appear after4 weeks, whereas with paroxetine it is maximal in the first week. A consistent finding has been thatthe syndrome is more frequent with paroxetine than with either sertraline or fluoxetine,14 com-prising insomnia, dreaming, muscle aches, dizziness, chills, runny nose, nausea, and diarrhea.

The only antidepressant not associated with a discontinuation syndrome is agomelatine.15

This is a distinct advantage in that many patients receiving antidepressants are erratic and forgetto take their medication over a few days during a course of treatment and would therefore be ex-pected to suffer discontinuation symptoms.

SuicideSuicide risk is highest early in treatment and related to illness severity. The best predictor is ahistory of a previous attempt, although this is absent in over half of suicides. Long-term follow-up has shown that antidepressants reduce suicide risk 3-fold in unipolar major depression,16 astatistic confirmed by epidemiological studies in Swedish adults and American adolescents.

Being a rare event in clinical studies, suicide risk for an individual drug cannot be assessedeven from the large datasets contained in the submissions to drug regulatory authorities. How-ever, a recent meta-analysis of data on over 40 000 patients submitted to licensing authoritiesshowed nonsignificantly fewer suicides with SSRIs compared with placebo.17 A definitive com-parison would require the inclusion of millions of patients in placebo-controlled studies.

Results for suicidal thoughts recorded as an adverse event have been similar. Since suchthoughts are measured as part of the pivotal depression severity rating scales, they provide thebest basis for assessment. Several antidepressants, including imipramine, paroxetine, and esc-italopram, are significantly more effective than placebo in protecting against the emergenceof suicidal thoughts.18

SafetySide effects are a major cause of treatment noncompliance. TCAs are especially disadvantagedin this regard, having a broad range of pharmacologic actions in addition to their activity on thenorepinephrine and serotonin systems. Effects on muscarinic receptors produce anticholiner-gic reactions such as dry mouth, blurred vision, constipation, urinary hesitancy, and cognitivedeficits, while α1-adrenergic receptor effects are responsible for the dizziness and hypotensionthat are a particular problem in the elderly because of falls and fractures. Histamine effects causesedation and poor concentration. The cumulative result is poor tolerability and compromisedtreatment.

The absence of similar receptor effects explains why SSRIs should be better tolerated, with few-er premature treatment discontinuations.19 However, SSRIs have several characteristic serotoner-gic side effects; they produce unwanted gastrointestinal effects, for example, nausea and vomit-ing. Some SSRIs, eg, fluoxetine, cause increased agitation and nervousness, particularly at the

E D I T O R I A L

215Editorial – Montgomery MEDICOGRAPHIA, VOL 27, No.3, 2005

E D I T O R I A L

216 Editorial – MontgomeryMEDICOGRAPHIA, VOL 27, No.3, 2005

REFERENCES1. Kessler RC, Berglund P, Demler O, et al. The epidemiology ofmajor depressive disorder: results from the National Comorbid-ity Survey replication (NCS-R). JAMA. 2003;289:3095-3105.2. Andrews G, Anstey K, Brodaty H, Issakidis C, Luscombe G. Re-call of depressive episode 25 years previously. Psychol Med.1999;29:787-791.3. Angst F, Stassen HH, Clayton PJ, Angst J. Mortality of patientswith mood disorders: follow-up over 34-38 years. J Affect Disord.2002;68:167-181.4. Keller MB, Lavori PW, Mueller TI, et al. Time to recovery,chronicity, and levels of psychopathology in major depression:a 5-year prospective follow-up of 431 subjects. Arch Gen Psychi-atry. 1992;49:809-816.5. Lepine JP, Gastpar M, Mendlewicz J, Tylee A. Depression in thecommunity: the first pan-European study DEPRES (DepressionResearch in European Society). Int Clin Psychopharmacol.1997;12:19-30.6. Donoghue JM, Tylee A. The treatment of depression: prescrib-ing patterns of antidepressants in primary care in the UK. Br JPsychiatry.1996;168:164-168.7. CINP Task Force. Impact of neuropharmacology in the 1990s—strategies for the therapy of depressive illness. Eur Neuropsycho-pharmacol. 1993;3:153-156.8. Khan A, Khan S, Brown WA. Are placebo controls necessaryto test new antidepressants and anxiolytics? Int J Neuropsycho-pharmacol. 2002;5:193-197.9. Danish University Antidepressant Group. Paroxetine: a selec-tive serotonin reuptake inhibitor showing better tolerance, butweaker antidepressant effect than clomipramine in a controlledmulticenter study. J Affect Disord. 1990:18:289-299.10. Clerc GE, Ruimy P, Verdeau Pailles J. A double-blind compar-ison of venlafaxine and fluoxetine in patients hospitalized for ma-jor depression and melancholia. Int Clin Psychopharmacol.1996;9:139-143.

11. Einarson TR. Evidence based review of escitalopram in treat-ing major depressive disorder in primary care. Int Clin Psycho-pharmacol. 2004;19:305-310.12. Lepola U, Wade AG, Andersen HF. Do equivalent doses of es-citalopram and citalopram have similar efficacy? A pooled analy-sis of two positive placebo-controlled studies in major depressivedisorder. Int Clin Psychopharmacol. 2004;19:149-155.13. Detke MJ, Wiltse CG, Mallinckrodt CH, et al. Duloxetine in theacute and long-term treatment of major depressive disorder: aplacebo- and paroxetine-controlled trial. Eur Neuropsychophar-macol.2004;14:457-470.14. Montgomery SA, Huusom AK, Bothmer J. A randomisedstudy comparing escitalopram with venlafaxine XR in primarycare patients with major depressive disorder. Neuropsychobiol-ogy. 2004;50:57-64.15. Montgomery SA, Kennedy SH, Burrows GD, Lejoyeux M,Hindmarch I. Absence of discontinuation symptoms with agome-latine and occurrence of discontinuation symptoms with paroxe-tine: a randomized, double-blind, placebo-controlled discontin-uation study. Int Clin Psychopharmacol. 2004;19:271-280.16. Angst J, Angst F, Gerber-Werder R, Gamma A. Suicide in 406mood-disorder patients with and without long-term medication:a 40 to 44 years’ follow-up. Arch Suicide Res. 2005. In press.17. Gunnell D, Saperia J, Ashby D. Selective serotonin reuptakeinhibitors (SSRIs) and suicide in adults: meta-analysis of drugcompany data from placebo controlled, randomised controlledtrials submitted to the MHRA’s safety review. BMJ. 2005;330:385-388.18. Montgomery SA, Dunner DL, Dunbar GC. Reduction of suici-dal thoughts with paroxetine in comparison with reference antide-pressants and placebo. Eur Neuropsychopharmacol.1995;5:5-13.19. Anderson IM, Tomenson BM. Treatment discontinuationwith selective serotonin reuptake inhibitors compared with tri-cyclic antidepressants: a meta-analysis. BMJ.1995;310:1433-1438.

Keywords: antidepressant; tricyclic antidepressant; selective serotonin reuptake inhibitor; norepinephrine; melatonin

start of treatment. Their sexual side effects include anorgasmia in women and delayed ejacula-tion in men. There appears to be some toleration of the nausea, but less of the sexual side effects.

Some antidepressants have been shown to have fewer sexual side effects, eg, nefazodone, pos-sibly due to its 5-HT2C antagonist properties. Buproprion enjoys a similar advantage over SSRIs,although it not licensed as an antidepressant in Europe. Agomelatine appears relatively free ofsexual side effects: a specific study found it significantly superior to venlafaxine in patients in re-mission. There is also evidence that the 5-HT2 antagonist mirtazapine is largely unassociated withsexual dysfunction.

SNRIs are associated with all the SSRI side effects plus some norepinephrine-related effectssuch as dry mouth, constipation, and increased heart rate. Thus, high-dose venlafaxine is likelyto cause excess serotonergic effects. Venlafaxine is also associated with increased blood pressureand cholesterol, and this must be taken into account when prescribing. Duloxetine is licensed ata low dose specifically to reduce the risk of excess side effects. Both venlafaxine and duloxetineappear to be less well tolerated than SSRIs.

ConclusionDepression is a common and disabling disorder that places a substantial burden on patients, theirrelatives and friends, and society as a whole. It is also dangerous, being associated with increasedmorbidity and mortality, including from suicide. Depression shortens life by an average of onedecade. The available treatments are largely effective, but their use is compromised by poor tol-erability and low adherence. Very few depressed individuals receive treatment and, of those whodo, few are prescribed an adequate dose for long enough to secure a full response. There is a clearneed for treatments that are more effective, more rapidly effective, and better tolerated. Theywould reduce the number of patients who have a poor or inadequate response and leave fewerpatients with resistant depression. Poor compliance with medication and premature terminationof treatment currently compromise the chance of recovery. Antidepressants with a faster onsetof action would be expected to bring larger numbers of depressed patients to full remission. ❒

L A DÉPRESSION MAJEURE EST UNE MALADIE GRAVE DONT LAprévalence au cours de la vie atteint 16,2 % aux États-Unis1 et quis’approche progressivement de cette valeur dans des cultures trèsdifférentes, par exemple au Japon et en Chine. Cette augmentation

apparente nécessite d’être nuancée : les personnes semblent plus enclines à discuter de leursproblèmes psychologiques qu’auparavant et les études antérieures ont probablement sous-es-timé l’importance de cette maladie. En outre, la mémoire, sélective, tend vraisemblablement àfausser les estimations rétrospectives de la prévalence au cours de la vie. Les personnes ont ten-dance à oublier des épisodes de dépression lointains et à exagérer les épisodes actuels : le suivide patients hospitalisés pour dépression a permis de déterminer que, 25 ans plus tard, seulementla moitié des patients pouvait se rappeler de symptômes suffisants pour justifier un diagnosticde dépression majeure 2.

La dépression majeure est associée à un dysfonctionnement social et même physique im-portant, qui dépasse de manière significative celui provoqué par certaines affections médicaleschroniques, par exemple le diabète. Le déclenchement survient généralement au cours de l’ado-lescence et au début de l’âge adulte, et son déroulement est fréquemment récurrent ou chro-nique, avec des épisodes dépressifs occupant 20 % de la durée de vie qui suit le diagnostic initial.Par conséquent, il n’est pas surprenant de constater que le trouble dépressif majeur soit la prin-cipale cause d’incapacité observée pendant la vie active. Chez environ 10 à 15 % des patientsla dépression aboutit au suicide. Une méta-analyse a établi un taux de mortalité standard, cal-culé en comparant le taux de suicide dans un groupe spécifique avec celui de la population gé-nérale, égal à 21,2 chez les dépressifs ; les troubles de l’humeur raccourcissent la vie de 10 ansà cause de l’effet combiné d’une augmentation du risque de suicide et d’une aggravation des ma-ladies physiques 3.

Non seulement les antidépresseurs actuels ne « guérissent » pas la maladie sous-jacente,mais ils ne sont également que modérément efficaces pour soulager les symptômes au cours desépisodes dépressifs. Un suivi de l’Institut National de la Santé Mentale (National Institute ofMental Health, NIMH) effectué sur 431 patients en demande thérapeutique a montré que 12 %de ceux-ci restaient déprimés de manière chronique pendant 5 ans, tandis que 55 % d’entre euxavaient présenté une récidive ou une rechute, et seulement un tiers restait en bonne santé 4 ;après 15 ans, 82 % avaient subi une rechute, 6 % souffraient toujours d’une dépression chro-nique et seulement 12 % étaient restés en bonne santé. Avec un suivi suffisamment long et at-tentif, il apparaît que la presque totalité des patients traités pour dépression majeure ne gué-rissent pas, ou présentent une rechute.

Néanmoins, il y a beaucoup d’arguments en faveur de l’administration de traitements ap-propriés pour la dépression, pour des motifs tant personnels que socio-économiques. Pour-tant, jusqu’à présent, l’insuffisance de traitement reste très fréquent. Sur les 17 % de personnessouffrant de dépression majeure identifiées au cours d’une enquête européenne portant sur78 000 adultes, 69 % ne recevaient aucun traitement médical, 43 % n’avaient même jamaisconsulté un médecin, et moins de 8 % recevaient un antidépresseur 5. En outre, le traitement an-tidépresseur administré est souvent suboptimal : une enquête portant sur un million de patients

É D I T O R I A L

217Éditorial – Montgomery MEDICOGRAPHIA, VOL 27, No.3, 2005

Pourquoi avons-nous besoin d’antidépresseurs innovants et plus efficaces ?

p a r S . M o n t g o m e r y , R o y a u m e - U n i

de soins primaires a montré que 89 % d’entre eux recevaient une dose subthérapeutique d’an-tidépresseur tricyclique (ATC)6. Heureusement, le sous-dosage est devenu moins fréquent avecles nouveaux antidépresseurs, notamment les inhibiteurs sélectifs de la recapture de la séro-tonine (ISRS), qui sont uniquement commercialisés à des doses thérapeutiques.

La mauvaise réponse thérapeutique est fréquemment due à une mauvaise observance dutraitement par le patient : 30 % des patients de soins primaires ne se sont jamais fait délivrerleur première prescription, tandis que 25 à 33 % de patients supplémentaires interrompent leurtraitement au cours du premier mois, et 62 % n’en informent pas leur médecin. Par conséquent,la durée du traitement dans 55 à 63 % des cas est inférieure à la recommandation minimaleétablie de six mois après la rémission7. Cependant, même dans des études bien conduites portantsur différentes classes d’antidépresseurs administrés pendant les 6 à 8 semaines généralementconsidérées comme suffisantes pour démontrer une différence entre un médicament actif et unplacebo, les taux d’absence de réponse thérapeutique se situent en moyenne entre 30 et 40 %.

Le développement des antidépresseursLe mécanisme d’action invoqué pour les des inhibiteurs de la monoamine oxydase (IMAO) et desATC était le blocage de la recapture des neurotransmetteurs monoaminergiques, la noradré-naline et la sérotonine (5-HT), augmentant ainsi leur disponibilité dans la fente synaptique. Ladépression était supposée provenir de concentrations subnormales de neurotransmetteurs. Bienque simpliste et incomplète, cette théorie a fourni une base fructueuse pour le développementde la plupart des antidépresseurs découverts par la suite. Les recherches se sont concentréessur la découverte de composés présentant une sélectivité supérieure pour l’une ou l’autre desmonoamines, diminuant ainsi les effets indésirables. La maprotiline et la réboxétine sont desexemples d’inhibiteurs sélectifs de la recapture de la noradrénaline (ISRN). Mais les efforts lesplus importants se sont portés sur le développement des ISRS, et ont culminé avec la mise surle marché de la fluoxétine, la paroxétine, la fluvoxamine, la sertraline, le citalopram et l’esci-talopram qui dominent aujourd’hui le marché des antidépresseurs.

Il a été découvert que plusieurs ATC agissaient sur les récepteurs 5-HT2 : l’amitriptylinebloque à la fois la recapture de la 5-HT2 et exerce une action antagoniste sur les récepteurs5-HT2. Ce phénomène a inspiré le développement d’antagonistes des récepteurs 5-HT2 commela trazodone, suivie par des composés encore plus sélectifs agissant sur des sous-types de récep-teurs 5-HT2, comme l’agomélatine, un antagoniste des récepteurs 5-HT2C et la mélatonine, unagoniste des récepteurs de type M1 et M2, et la miansérine et la mirtazapine, qui sont des anta-gonistes des récepteurs 5-HT2A et 5-HT2C, ainsi que des antagonistes �2 et des antagonistes desrécepteurs 5-HT1A et 5-HT3.

Supériorité de certains antidépresseurs particuliersLes essais cliniques sur les antidépresseurs n’ont pas permis de démontrer une réponse constam-ment supérieure à un traitement actif par rapport à un placebo8. Cela est dû, quelle que soitl’(in)efficacité des antidépresseurs eux-mêmes, au schéma et au déroulement inappropriés desétudes, et à certains obstacles théoriques et pratiques majeurs. L’une des insuffisances impor-tantes a été une puissance inadéquate, tandis qu’un certain nombre d’études « sans succès »n’ont jamais été publiées. Des études de petite taille démontrant une efficacité significative ontsouvent été limitées à des patients hospitalisés souffrant de dépression sévère, tandis que le pas-sage à des études plus réalistes sur la population de ville a conduit au recrutement d’un grandnombre de patients souffrant de dépression légère pour lesquels une différence entre le médi-cament actif et le placebo a été difficile à mettre en évidence. Même les critères diagnostiques dutrouble dépressif majeur se sont avérés inadéquats, en particulier en ce qui concerne la duréede la maladie. Tandis que les critères du DSM IV [Diagnostic and Statistical Manual of MentalDisorders, 4th Edition] ne demandent que deux semaines, une durée compatible avec un troublespontanément résolutif, les investigateurs ont depuis longtemps augmenté la durée minimaled’un épisode à un mois pour justifier d’une inclusion dans ce type d’études.

La différence entre les médicaments est encore plus difficile à démontrer. Les ISRS peuventêtre plus sûrs et mieux tolérés que les ATC, mais ils sont apparus significativement moins effi-caces que le traitement classique par la clomipramine dans une étude importante9. Malgré sesfaiblesses méthodologiques, cette étude a suggéré que la sélectivité pour le système de la sé-rotonine pouvait ne pas s’avérer universellement bénéfique. Une conclusion plausible a été queles antidépresseurs sélectifs ne pourraient être efficaces que chez les patients présentant un dé-ficit significatif en monoamines, une théorie corroborée par des comptes rendus de plusieurs

É D I T O R I A L

218 Éditorial – MontgomeryMEDICOGRAPHIA, VOL 27, No.3, 2005

études ultérieures selon lesquelles des ISRN, comme la venlafaxine, se sont avérés plus efficacesque les ISRS10, qui ont eux-mêmes été significativement plus efficaces que le placebo. Cepen-dant, cette conclusion n’a pas permis d’établir une conviction consensuelle, même avec le sou-tien d’une méta-analyse, à cause d’un ensemble d’insuffisances méthodologiques.

Une rare occasion pour effectuer une méta-analyse médicamenteuse valide a été fourniepar les études groupées sur deux ISRS, l’escitalopram et le citalopram, dans la mesure où ellesont toutes deux été de la même durée de 8 semaines et ont utilisé l’échelle d’évaluation de la dé-pression de Montgomery & Asberg (Montgomery and Asberg Depression Rating Scale, MADRS) :plusieurs méta-analyses effectuées pour un certain nombre d’affections sont arrivées à la conclu-sion générale que l’escitalopram était la substance la plus efficace 11.

Déclenchement précoce de la réponseL’un des inconvénients bien connus des antidépresseurs est qu’une différence significative avecle placebo ne peut être mise en évidence de façon fiable qu’environ six semaines après le débutdu traitement. Ce délai est potentiellement dangereux, car les risques de suicide sont élevés dansla période précoce du traitement lorsque la sévérité est maximale. L’absence d’effet positif ra-pide décourage également l’observance thérapeutique. C’est la raison pour laquelle un déclen-chement plus précoce de l’effet thérapeutique serait bienvenu. Malheureusement, pratiquementaucune étude n’a analysé cet aspect sur les antidépresseurs examinés individuellement. Uneanalyse rétrospective d’études contrôlées par placebo portant sur la venlafaxine indique sim-plement qu’une réponse est observée avant deux semaines avec une augmentation précoce etrapide à des doses élevées, tandis que des études comparatives ont suggéré que l’escitalopraminduisait une réponse plus rapide que le citalopram12.

Efficacité sur des symptômes spécifiquesCertains symptômes s’améliorent plus rapidement avec un antidépresseur qu’avec un autre.

◆ SommeilLes antidépresseurs présentant une activité antihistaminergique, par exemple l’amitriptylineet la mirtazapine, améliorent le sommeil précocement au cours du traitement, mais au prixd’une altération de la vigilance diurne et de troubles de la concentration. Les antagonistes desrécepteurs 5-HT2, par exemple la néfazodone, améliorent le sommeil sans inconvénient de cetype, de même que l’agomélatine, un agoniste mélatoninergique et un antagoniste sélectif desrécepteurs 5-HT2C.

◆ Symptômes douloureux et somatiquesLes ISRN pourraient être plus efficaces pour soulager les syndromes douloureux. Ainsi, la ven-lafaxine a été supérieure à un placebo dans la neuropathie diabétique. En fait, les ISRN pour-raient être plus efficaces d’une manière générale sur les symptômes somatiques de dépression13.Bien qu’il ne s’agisse pas des symptômes centraux de ce trouble, ils peuvent expliquer l’avan-tage apparent des ISRN sur les ISRS, en particulier lorsque la rémission est estimée par l’échelled’évaluation de la dépression de Hamilton (Hamilton Depression Rating Scale) avec sa surre-présentation des symptômes somatiques. L’avantage peut apparaître de façon moins nette enutilisant une échelle des symptômes centraux, comme l’échelle MADRS.

Interruption du médicamentDes études antérieures sur les ATC ont établi que l’interruption brutale du traitement provo-quait un syndrome de sevrage se manifestant par des nausées, des vomissements, des maux detête, des vertiges, des frissons, une faiblesse et des douleurs musculosquelettiques chez certainspatients. Les symptômes étaient transitoires, atteignant une intensité maximale au cours de lapremière semaine puis ont déclinant par la suite, interférant rarement avec le fonctionnement.

Un syndrome de sevrage a également été observé avec des ISRS, que le retrait ait été pro-gressif ou brutal. Son apparition dépend de la demi-vie du composé considéré et de ses méta-bolites actifs. Avec la fluoxétine, par exemple, le syndrome de sevrage peut apparaître aprèsquatre semaines, tandis qu’avec la paroxétine, son intensité est maximale au cours de la pre-mière semaine. Les résultats montrent de façon constante que le syndrome était plus fréquentavec la paroxétine qu’avec la sertraline ou la fluoxétine14, et qu’il se manifestait par une insom-nie, une augmentation des rêves, des douleurs musculaires, des vertiges, des frissons, une rhi-norrhée, des nausées et une diarrhée.

É D I T O R I A L


Le seul antidépresseur qui n’est pas associé à un syndrome de sevrage est l’agomélatine15. Ils’agit d’un avantage important dans la mesure où de nombreux patients recevant des antidé-presseurs sont inconstants et oublient de prendre leur médicament pendant quelques jours aucours du traitement, et seraient, par conséquent, susceptibles de présenter des symptômes desevrage.

SuicideLe risque de suicide est maximal au début du traitement, et il est lié à la sévérité de la maladie.Le meilleur facteur prédictif est constitué par des antécédents de tentative de suicide, bien qu’ilssoient absents dans plus de la moitié des cas de suicides. Un suivi à long terme a montré que lesantidépresseurs réduisaient le risque de suicide d’un facteur 3 dans la dépression majeure uni-polaire16, une statistique confirmée par des études épidémiologiques effectuées chez des adultessuédois et des adolescents américains.

Dans la mesure où il s’agit d’un événement rare dans les études cliniques, le risque de sui-cide lié à un médicament individuel ne peut pas être évalué, même à partir des larges ensemblesde données nécessaires aux demandes d’autorisation de mise sur le marché remis aux autoritésréglementaires. Cependant, une récente méta-analyse effectuée sur des données portant sur plusde 40 000 patients soumises aux autorités sanitaires a montré une réduction non significativedu nombre de suicides avec les ISRS par rapport à un placebo17. Une comparaison décisive né-cessiterait l’inclusion de millions de patients dans des études contrôlées par placebo.

Les résultats concernant les pensées suicidaires enregistrées comme événement indésirableont été similaires. Ces pensées étant mesurées en utilisant les principales échelles d’évaluationde la sévérité de la dépression, ces dernières constituent la meilleure base pour l’évaluation. Plu-sieurs antidépresseurs, notamment l’imipramine, la paroxétine et l’escitalopram, sont signifi-cativement plus efficaces que le placebo dans la protection contre l’émergence des pensées sui-cidaires 18.

Sécurité d’emploiLes effets indésirables constituent une cause majeure de non observance du traitement. Les ATCprésentent un inconvénient particulier à cet égard, dans la mesure où ils exercent une large va-riété d’actions pharmacologiques outre leur activité sur les systèmes de la noradrénaline et dela sérotonine. Leurs effets sur les récepteurs muscariniques entraînent des réactions anticho-linergiques comme une sécheresse buccale, une vision trouble, une constipation, un retard à lamiction et des déficits cognitifs, tandis que les effets sur les récepteurs �1-adrénergiques sontresponsables de vertiges et d’hypotension, qui représentent un problème particulier chez les per-sonnes âgées à cause des chutes et des fractures. Les effets histaminiques provoquent une sé-dation et des troubles de la concentration. Au total, il en résulte une mauvaise tolérance et untraitement aléatoire.

L’absence d’effets similaires sur ces récepteurs fait que les ISRS devraient être mieux tolé-rés, et s’accompagner d’un moindre nombre d’interruptions prématurées du traitement19. Ce-pendant, les ISRS exercent plusieurs effets indésirables sérotoninergiques caractéristiques ; ilsentraînent des effets gastro-intestinaux indésirables, par exemple des nausées et vomissements.Certains ISRS, par exemple la fluoxétine, provoquent une augmentation de l’agitation et de lanervosité, en particulier au début du traitement. Leurs effets indésirables sexuels comprennentune anorgasmie chez la femme et un retard à l’éjaculation chez l’homme. Il semble que lesnausées puissent être partiellement tolérées, ce qui est moins le cas pour les effets indésirablessexuels.

Certains antidépresseurs ont montré qu’ils provoquaient moins d’effets indésirables sexuels,par exemple la néfazodone, ce qui pourrait être lié à ses propriétés antagonistes des récep-teurs 5-HT2C . Le buproprion présente un avantage similaire par rapport aux ISRS, bien qu’ilne soit pas commercialisé comme antidépresseur en Europe. L’agomélatine semble relativementexempte d’effets indésirables sexuels : une étude spécifique a montré sa supériorité significa-tive par rapport à la venlafaxine chez des patients en rémission. Il existe également des élémentsmontrant que la mirtazapine, un antagoniste des récepteurs 5-HT2, n’avait pratiquement au-cune association avec un dysfonctionnement sexuel.

Les ISRN sont associés à tous les effets indésirables des ISRS, auxquels s’ajoutent certainseffets liés à la noradrénaline, notamment la sécheresse buccale, la constipation et l’augmenta-tion de la fréquence cardiaque. Par conséquent, des doses élevées de venlafaxine sont suscep-tibles de provoquer un excès d’effets sérotoninergiques. La venlafaxine est également associée

É D I T O R I A L

220 Éditorial – MontgomeryMEDICOGRAPHIA, VOL 27, No.3, 2005

à une augmentation de la pression artérielle et du cholestérol, ce qui doit être pris en comptelors de la prescription. La duloxétine est autorisée à faible dose précisément pour réduire lesrisques d’effets indésirables excessifs. La venlafaxine et la duloxétine semblent être moins bientolérées que les ISRS.

ConclusionLa dépression est un trouble fréquent et invalidant qui impose un lourd fardeau aux patients, àleurs parents et amis, et à la société dans son ensemble. Elle est également dangereuse, dans lamesure où elle est associée à une augmentation de la morbidité et de la mortalité, y compris liéesau suicide. La dépression raccourcit la vie en moyenne de 10 ans. D’une manière générale, lestraitements disponibles sont efficaces, mais leur utilisation est compromise par une mauvaisetolérance et une observance insuffisante. Seul un très petit nombre d’individus déprimés re-çoivent un traitement, et parmi ceux-ci, il est rare qu’il soit prescrit à une posologie adéquate etpour une durée suffisante pour assurer une réponse complète. Il existe un besoin manifestepour des traitements plus efficaces, plus rapidement efficaces et mieux tolérés. Ils permettraientde réduire le nombre de patients présentant une réponse insuffisante ou inadéquate, et laisse-raient moins de patients souffrir d’une dépression réfractaire. La mauvaise observance et l’in-terruption prématurée du traitement réduisent actuellement les chances de guérison. Des an-tidépresseurs dont le déclenchement d’action serait plus rapide permettraient d’augmenter lenombre de patients déprimés bénéficiant d’une rémission complète. ❒

É D I T O R I A L


✦

The field of chronobiology studies 24-hour bio-logical clocks (the circadian system) and theirsynchronizers (“zeitgebers”) such as light, the

pineal hormone melatonin, food, activity, as wellas the factors regulating sleep.1 Light therapy hasarisenoutof this basic researchin circadianrhythms,whereas, in contrast, sleep deprivation (“wake ther-apy”) was established by astutely following up clin-ical observations.2 Chronotherapeutics can be de-fined as translating basic chronobiology researchinto valid treatments. The term is broad, and thetreatments subsumed under this heading will growas the field grows, and are of course not limited toaffective disorders (there is an important body ofevidence, for example, that the correct timing ofcancer treatments augments survival and dimin-ishes the often potent side effects).

The theme of meeting unmet needs in the treat-ment of depression is important and timely. Theonset of action of antidepressants is still not rapidenough, a proportion of patients do not respond,others have residual symptoms that predict relapse.Although medication based on classic neurotrans-mitter systems is still a prime focus, drug targetsother than monoamines are under intensive inves-tigation.3 Strategies promoting adjuvant therapyare on the increase, whether they encompass com-bination with other medications (eg, addition ofpindolol, of thyroid hormone) or psychological in-terventions (eg, cognitive behavioral therapy). Main-stream psychiatry is becoming more and moreeclectic, implementing a variety of approaches tohelp the individual patients. The question to be dis-cussed here is focused on why not also combine thechronobiological strategies of light therapy and/orwake therapy with psychopharmacological medica-tion? Light therapy has undergone widespread con-trolled randomized clinical trials, and wake thera-py has been so widely studied over decades that theefficacy data are strong. These nonpharmaceutical,biologically based therapies are not only powerfuladjuvants, but also antidepressants in their ownright.2,4

Why are we interested in biological rhythms?

One of the most striking clinical phenomena in af-fective disorders is the periodicity of recurrence—ranging from seasonal, as in winter depression, torapid cycling, which can be as short as 48 hours (re-viewed in 5). Other periodic phenomena are foundat the symptom level: diurnal variation of mood,

223Chronobiological strategies for unmet needs in the treatment of depression – Wirz-Justice MEDICOGRAPHIA, VOL 27, No. 3, 2005

U N M E T N E E D S I N T H E T R E A T M E N T O F D E P R E S S I O N

C hronobiological strategies may provide an effective means of address-ing some of the unmet needs in the treatment of depression, such as short-ening the latency of onset of antidepressant action, combating residual

symptoms, and preventing relapse in the long term. Light is the treatment ofchoice for winter depression (or seasonal affective disorder, SAD). Light thera-py given as an adjuvant to medication in major nonseasonal depression, as wellas in chronic and therapy-resistant depression, speeds up and potentiates clin-ical response. Light is also efficacious in bipolar depression; in these patients“dark therapy” (long nights) can diminish manic symptoms and stop rapid cy-cling. Total or partial sleep deprivation in the second half of the night (betterknown as “wake therapy”) induces marked improvement the following day.This amelioration can be maintained with concomitant treatment with antide-pressants, lithium, light therapy, sleep phase advance, or combinations thereof.Careful control of the light-dark cycle and of the timing of mealtimes, activity,and sleep may appear to be old-fashioned methods (“daily structures”) belong-ing to a long obsolete custodial psychiatry. However, these apparently simplemethods gain new validation when reconsidered within the framework of mod-ern chronobiology, since when appropriately timed, application of “zeitgebers”can aid treatment of affective disorders.Medicographia. 2005;27:223-227. (see French abstract on page 227)

Keywords: major depression; circadian rhythm; sleep deprivation; light therapy; melatonin

Anna WIRZ-JUSTICE, PhDCenter for Chronobiology University Psychiatric Hospitals Basel Basel, SWITZERLAND

Address for correspondence: Prof Dr Anna Wirz-Justice, Center for Chronobiology, University Psychiatric Hospitals, Wilhelm Klein Straße 27, CH-4025 Basel, Switzerland(e-mail: [email protected])

Chronobiological strategies for unmetneeds in the treatmentof depression b y A . W i r z - J u s t i c e , S w i t z e r l a n d

SELECTED ABBREVIATIONS AND ACRONYMS

5-HT serotonin5-HT2C serotonin receptor (subtype 2C)MDD major depressive disorderPVN paraventricular nucleus SAD seasonal affective disorder SCN suprachiasmatic nucleusSSRI selective serotonin reuptake inhibitor

(such as school or work schedules) may act direct-ly or indirectly on the SCN, since they determinethe timing of meals, sleep, physical exercise, andoutdoor light exposure. The circadian pacemakerhas inputs to sleep regulatory centers (eg, the raphenucleus, the dorsolateral hypothalamus [orexin],ventrolateral preoptic area),15 and sleep centers talkback to the clock.16 The timing and architecture ofsleep is considered to be a consequence of interac-tions between the circadian pacemaker and a home-ostatic process of rising sleep pressure, dependenton the duration of prior wakefulness, that declinesduring sleep (the “two-process model”).17

Mood is dependent on both time of day and time awake

This parsimonious two-process model has been ableto explain much of the physiology of sleep as well asof aberrant sleep-wake cycle behavior. Protocols de-veloped to analyze the contributions of circadianphase vs the sleep homeostat have provided fasci-nating information not only about sleepiness—asmight be expected—but also that mood is similarlyregulated by the two processes. This is shown veryclearly in the “forced desynchrony” protocol carriedout in healthy subjects.18 The circadian componentof mood follows the circadian rhythm of core bodytemperature rather closely. We wake up in not toogood a mood, but this improves throughout the dayto reach a maximum in the evening, and then mooddeclines during the night. The wake-dependentcomponent reveals that we are quite cheery after agood night’s sleep when sleep pressure is low, butthat thereafter mood declines monotonically withtime awake. If the temporal alignment between thesleep-wake cycle and the circadian pacemaker af-fects self-assessment of mood in healthy subjects,it might be expected that this is even more impor-tant for patients with depression. The phenomenonof diurnal mood variation as a characteristic of de-pressive state may indeed arise from phase relation-ships gone awry.

Diurnal mood variation can be manipulated byshifting or depriving sleep. The improvement aftera night’s wake therapy usually begins in the secondhalf of the night or the next day, suggesting thatstaying awake prevents the nocturnal plunge inmood.10 Furthermore, a phase advance of sleep tim-ing has been able to induce a day-by-day change indiurnal mood patterns over many weeks—evidencefor the profound effect of shifting phase relation-ships on mood (a more severe form of jet lag).19,20

Similar day-by-day changes in diurnal mood pat-terns have been found in a “forced desynchrony” ex-periment carried out in major seasonal depression.21

Shifting rhythms or sleep can be therapeutic

The above model helps to understand the change ofclinical state with time of day and after manipula-tions of sleep. The clinical findings, however, are theimportant point to be made—extending wakeful-ness is antidepressant. Wake therapy has been well

early morning awakening, and sleep disturbances.Abundant research has documented abnormal cir-cadian rhythms in biochemistry, neuroendocrinefunction, physiology, and behavior, often linked tochanges in affective state. These have been reviewedin detail elsewhere; the findings are not homoge-neous, even though a certain pattern appears char-acteristic of depression—there is increased vari-ability in day-to-day rhythms, decreased circadianamplitude, and circadian phase that is either ear-ly (advanced) or late (delayed).5-11 Bipolar disorderseems to be most clearly linked to abnormal orchanging circadian rhythm phase.6 In addition, al-terations in the sleep EEG in depression, althoughneither pathognomonic nor specific, display recog-nizable patterns of disturbance.12

Principles of circadian timing and sleep regulation

The biological timing system is schematically de-scribed in Figure 1. Circadian oscillators are foundin every organ and every cell—the so-called “pe-ripheral clocks.”13 A master pacemaker or biologi-cal clock in the suprachiasmatic nuclei (SCN) coor-

dinates these circadian rhythms in brain and body.1

The SCN is synchronized to the external light-darkcycle primarily by retinal light input. A specialized(“nonvisual”) retinohypothalamic tract provides di-rect neuronal connection to the SCN from novelphotoreceptors in the retinal ganglion cells thatmeasure illuminance.14 Nocturnal synthesis of thepineal hormone melatonin is driven by the SCN;melatonin also feeds back on melatonin receptorsin the SCN and thus melatonin belongs to the cat-egory of synchronizing agents or “zeitgebers” (lightbeing the major one). A serotoninergic pathwayfrom the raphe nucleus provides nonphotic input tothe SCN. Nonphotic zeitgebers such as exercise,sleep, or darkness are probably much weaker zeit-gebers than light on SCN function. Social zeitgebers

224 Chronobiological strategies for unmet needs in the treatment of depression – Wirz-JusticeMEDICOGRAPHIA, VOL 27, No. 3, 2005


Figure 1. Schematic representation of the circadian timing system. Light ( )is the major zeitgeber reaching the biological clock in the SCN via special-ized “circadian photoreceptors” in the retina ( ). A multisynaptic pathway tothe pineal gland via the paraventricular nucleus (PVN) drives the nocturnalsynthesis of melatonin and its suppression by light. Melatonin feeds back onreceptors in the SCN. The SCN also synchronizes the timing of peripheralclocks in other organs and cells, some of which have their own zeitgebers(eg, food for the liver clock). There are multiple connections from (and to)the SCN to areas of the brain involved in sleep regulation eg, the preopticarea, the dorsolateral and posterior hypothalamus, and the raphe nucleus.

Peripheralclocks

PinealSCN

Sleepregulating

centers

PVN

Melatonin

light and the use of central heating and air condi-tioning to control environmental temperature. Thisis clearly seen in the seasonality of birth (concep-tion) rates, that had a high spring peak in the 16thcentury, but declined to very low amplitude in the20th century, with a shift to an autumn peak.34 Psy-chiatrists have long remarked on seasonality in theirpatients’ symptoms, for example Esquirol, who not-ed that the peak admission rates to the Salpêtrière

hospital occurred in spring.35 What is not usuallyrecognized, is that not only depressive symptoms,but even response to placebo is seasonally modu-lated. The 10-day response rate to placebo in dou-ble-blind controlled trials of various antidepressantscarried out at the New York State Psychiatric Insti-tute was analyzed according to time of year (Fig-ure4, page 226).36 Three times higher response ratesoccurred in summer than in winter. In conclusion,many aspects of behavior, physiology, and neuroen-

established as a rapid treatment for depression forover 30 years, the response being particularly highin those patients who report daily mood swings.10

Modified protocols have been developed, such aswake therapy in the second half of the night, orphase advance of the sleep-wake cycle.10 These twomodifications emphasize the circadian factor—itbeing important to remain awake at a particulartime in order to prevent mood decline.

The main reason for the lack of enthusiasm forwake therapy as a treatment in everyday practice isthe equally rapid relapse following recovery sleep.A number of groups have taken up the challenge ofsearching for methods to prevent this: do not giveup wake therapy as a treatment just because its ef-fects don’t last long enough! In bipolar patients, thecombination with lithium appears to maintain an-tidepressant response.22-24 A number of differentmedications have been tried25; in particular, the useof SSRIs or light is recommended following one tothree episodes of wake therapy.26-29

How are circadian rhythms related to depression?

The basic question of how circadian rhythms arerelated to depression has not yet been answered.Genetic vulnerability and stress influence circadi-an rhythms and sleep patterns, leading to many ofthe symptoms characteristic of affective disorders.Circadian and seasonal rhythms involve the sameneurotransmitters postulated to be important fordepression—so that changes in one system haverepercussions on the other. For example, it is knownthat serotonin concentrations are highest in theSCN. The SCN also expresses high levels of melato-nin receptors, and exogenous melatonin is knownto be able to influence the phase and the period ofthe circadian clock. In humans, serotonin turnoverchanges markedly with time of day and year (Fig-ure 2),30-32 and light exposure rapidly simulates sero-tonergic function.32 Serotonin is also important insleep regulation, though its role appears complex.Prefrontal cortical serotonin has been linked withmood. These interrelationships have been concep-tualized in a dual model of circadian rhythms andserotonin in depression (Figure 3).33 The emphasisis on a system vulnerable to depression—whethergenetic, hormonal, dependent on light availabilityand light exposure—and, in parallel, the circadiansystem and its phase relationships with sleep andwith the outer world. Concurrent dysfunction canlead to major depression or its seasonal form. Cir-cadian abnormalities alone lead to certain forms ofsleep disorders (such as advanced or delayed sleepphase syndrome) without effects on mood. Seroton-ergic abnormalities alone lead to other serotonin-related illnesses (eg, obsessive compulsive disorder)again, without the mood disorder.

A digression on seasonality

Humans retain their capacity to undergo seasonalresponses, even though their extent has declinedin the last century since the invention of artificial



Figure 2. Average CNS serotonin turnover in vivo inhealthy subjects according to season of measurement.Redrawn from reference 32: Lambert GW, Reid C, Kaye DM, Jen-nings GL, Esler MD. Effect of sunlight and season on serotoninturnover in the brain. Lancet. 2002;360:1840-1842. Copyright ©2002, Elsevier Ltd.

Figure 3. Dual model of depression. Both a disturbance in the circadian timing systemand a neurobiological vulnerability (serotonergic [5-HT] hypofunction) are requiredfor the manifestation of major depressive disorder (MDD) or seasonal affective disorder(SAD).Redrawn from reference 33: Lam RW, Tam EM, Yatham LN, Shiah IS, Zis AP. Seasonal depression: thedual vulnerability hypothesis revisited. J Affect Disord. 2001;63:123-132. Copyright © 2001, Elsevier B.V.

MDD

SAD

Altered circadianamplitude

and/or phase relationship

Circadiansleep disorders

Other 5-HT–relateddisorders

5-HT vulnerabilitygenetic/gender/light exposure/

season

Summer Autumn Winter Spring

Estim

ated

CN

S se

roto

nin

turn

over

(pm

ol/m

il)

2000

1500

1000

500

0

trials of monotherapy with light.39 Double-blindplacebo-controlled studies have now shown thatlight therapy combined with an SSRI leads to morerapid (within 1 week) and more profound (by ca30%) improvement in patients with nonseasonalmajor depression.27,40 Recently, a study of light treat-ment in chronic depression (of greater than 2 yearsduration) yielded impressive results in this oftentreatment-resistant group.41 Thus, a new genera-tion of clinical trials supports the therapeutic effi-cacy of light, alone or in combination with medica-tion, for a variety of psychiatric disorders, and it isto be hoped that more will follow.

“More darkness” is a correlate of the above: pilotstudies suggest that the simple measure of promot-ing long nights (more rest, more sleep, no light) canstop rapid cycling in bipolar patients,42,43 or dimin-ish manic symptoms44—intriguing findings thatrequire replication.

Can chronotherapeutics provide new drugs as well?

Not only are the described chronotherapeutic ap-proaches efficacious antidepressants in themselves,but they also offer new models for pharmaceuticalresearch. Is part of the usefulness of light due to itszeitgeber function of stabilizing phase? Is part of itsefficacy due to serotonergic mechanisms? Giventhat the antidepressant properties of selective sero-tonin (5-HT) reuptake–inhibiting drugs are con-sidered to be related to the 5-HT2C receptor subtype,it is interesting that 5-HT2C receptor agonists in therat SCN mimic the effects of light.45 Serotonergicdrugs and melatonin improve entrainment. In thisrespect, the pharmacological profile of agomelatinefits the above model, as it is a melatonin receptortype 1 and 2 (MT1 and MT2) agonist with 5-HT2Cproperties.46 Melatonin itself has no antidepressantcharacteristics.

In summary, circadian rhythm and sleep researchhave led to nonpharmacological therapies of de-pression (light therapy, wake therapy) that can be—and should be—used in everyday practice.2,4 Therationale for attempting to resynchronize disturbedphase relationships between the clock and sleep isthe concomitant improvement in mood. Chrono-biological concepts emphasize the importance ofzeitgebers and provide psychopharmacology witha novel approach for developing “chronobiotic”drugs. ❒

docrine function are sensitive to season. One ex-ample is presented, that of the in vivo turnover ofcentral nervous system serotonin in healthy hu-mans—much higher in summer than in winter(Figure 2).32

More light!

As for the seasons, the annals of psychiatry aboundwith evidence that affective state can be modulatedby exposure to environmental light or darkness.37

The diagnosis of seasonal affective disorder (SAD)and the development of light therapy was based onneurobiological models of mammalian seasonali-ty—the first treatment in psychiatry to be ground-ed in basic research. Although light therapy wasinitially propagated by Kripke for nonseasonal de-pression,38 initial studies were too short in dura-tion to provide the convincing results that a singleweek of light therapy can now achieve in SAD: itis only now, 20 years on, that controlled long-termstudies of light at last have been and are being car-ried out in nonseasonal major depression, with ex-tremely promising results. For example, the needfor efficacious treatment of depression during preg-nancy without side effects on the fetus has led to

Figure 4. The 10-day response rate to placebo in dou-ble-blind controlled trials of various antidepressants.Each point is a single trial.Redrawn from reference 36: Terman M. On the question of mech-anism in phototherapy for seasonal affective disorder: considera-tions of clinical efficacy and epidemiology. In: Rosenthal NE, Ble-har MC, eds. Seasonal Affective Disorder and Phototherapy. NewYork, NY: Guilford Press; 1989:357-376. Copyright © 1989, Guil-ford Press.

226 Chronobiological strategies for unmet needs in the treatment of depression – Wirz-JusticeMEDICOGRAPHIA, VOL 27, No. 3, 2005


REFERENCES1. Klein DC, Moore RY, Reppert SM. Suprachiasmatic Nucleus:The Mind’s Clock. New York, NY: Oxford University Press; 1991.2. Wirz-Justice A, Benedetti F, Berger M et al. Chronotherapeu-tics (light and wake therapy) in affective disorders. Psychol Med.2005;35. Epub ahead of print March 10, 2005.3. Holden C. Future brightening for depression treatments. Sci-ence. 2003;302:810-813.4. Wirz-Justice A, Terman M, Oren DA, et al. Brightening depres-sion. Science. 2004;303:467-469.5. Wirz-Justice A. Biological rhythms in mood disorders. In:Bloom FE, Kupfer DJ, eds. Psychopharmacology: The FourthGeneration of Progress. New York, NY: Raven Press; 1995:999-1017.6. Wehr TA, Goodwin FK. Biological rhythms in manic-depressiveillness. In: Wehr TA, Goodwin FK, eds. Circadian Rhythms in Psy-chiatry. Pacific Grove, Calif: The Boxwood Press; 1983:129-184.

7. Wu JC, Bunney WE. The biological basis of an antidepressantresponse to sleep deprivation and relapse: review and hypothesis.Am J Psychiatry. 1990;147:14-21.8. Healy D, Waterhouse JM. The circadian system and the thera-peutics of the affective disorders. Pharmacol Ther.1995;65:241-263.9. Rosenwasser AM, Wirz-Justice A. Circadian rhythms and de-pression: clinical and experimental models. In: Redfern PH, Lem-mer B, eds. Physiology and Pharmacology of Biological Rhythms.Berlin, Germany: Springer Verlag; 1997:457-486.10. Wirz-Justice A, Van den Hoofdakker RH. Sleep deprivationin depression: what do we know, where do we go? Biol Psychiatry.1999;46:445-453.11. Boivin DB. Influence of sleep-wake and circadian rhythm dis-turbances in psychiatric disorders. J Psychiatry Neurosci. 2000;25:446-458.

SummerAutumn

WinterSpring

Plac

ebo

resp

onse

rat

e (%

)

40

30

20

10

0



34. Wehr TA. Photoperiodism in humans and other primates: ev-idence and implications. J Biol Rhythms. 2001;16:348-364.35. Esquirol E. Des Maladies Mentales. Paris, France: ÉditionsBaillière; 1838.36. Terman M. On the question of mechanism in phototherapyfor seasonal affective disorder: considerations of clinical efficacyand epidemiology. In: Rosenthal NE, Blehar MC, eds. SeasonalAffective Disorder and Phototherapy. New York, NY: GuilfordPress; 1989:357-376.37. Wehr TA. Seasonal affective disorders: a historical overview.In: Rosenthal NE, Blehar MC, eds. Seasonal Affective Disordersand Phototherapy. New York, NY: Guilford Press; 1989:11-32.38. Kripke DF, Mullaney DJ, Gillin JC, Risch SC, Janowsky DS.Phototherapy of non-seasonal depression. In: Shagass C, JosiassenRC, Bridges WH, Weiss KJ, Stoff D, Simpson GM, eds. BiologicalPsychiatry 1985, Proceedings of the IVth World Congress ofBiological Psychiatry. New York, NY: Elsevier; 1985:993-996.39. Epperson CN, Terman M, Terman JS, et al. Randomized clin-ical trial of bright light therapy for antepartum depression: pre-liminary find. J Clin Psychiatry. 2004;65:421-425.40. Martiny K. Adjunctive bright light in non-seasonal major de-pression. Acta Psychiatr Scand. 2004;110(suppl):1-28.41. Goel N, Terman M, Terman JS, Macchi MM, Stewart JW. Con-trolled trial of bright light and negative air ions for chronic de-pression: preliminary results. Psychol Med. 2005;35. In press.42. Wehr TA, Turner EH, Shimada JM, Lowe CH, Barker C, Leiben-luft E. Treatment of rapidly cycling bipolar patient by using ex-tended bed rest and darkness to stabilize the timing and dura-tion of sleep. Biol Psychiatry. 1998;43:822-828.43. Wirz-Justice A, Quinto C, Cajochen C, Werth E, Hock C. Arapid-cycling bipolar patient treated with long nights, bed rest,and light. Biol Psychiatry. 1999;45:1075-1077.44. Barbini B, Benedetti F, Colombo C, et al. Dark therapy formania: a pilot study. Bipolar Disord. 2005;7:98-101.45. Kennaway DJ. Light, neurotransmitters and the suprachi-asmatic nucleus control of pineal melatonin production in therat. Biol Signals Recept.1997;6:247-254.46. Lôo H, Hale A D’haenen H. Determination of the dose of ago-melatine, a melatoninergic agonist and selective 5-HT2C antag-onist, in the treatment of major depressive disorder; a placebo-controlled dose range study. Int Clin Psychopharmacol. 2002;17:239-247.

12. Benca RM, Okawa M, Uchiyama M, et al. Sleep and mood dis-orders. Sleep Med Rev.1997;1:45-56.13. Buijs RM, Kalsbeek A. Hypothalamic integration of centraland peripheral clocks. Nat Rev Neurosci. 2001;2:521-526.14. Berson DM, Dunn FA, Takao M. Phototransduction by retinalganglion cells that set the circadian clock. Science. 2002;295:1070-1073.15. Mignot E, Taheri S, Nishino S. Sleeping with the hypothala-mus: emerging therapeutic targets for sleep disorders. Nat Neu-rosci. 2002;5:1071-1075.16. Deboer T, Vansteensel MJ, Detari L, Meijer JH. Sleep states al-ter activity of suprachiasmatic nucleus neurons. Nat Neurosci.2003;6:1086-1090.17. Daan S, Beersma DGM, Borbély AA. Timing of human sleep:recovery process gated by a circadian pacemaker. Am J Physiol.1984;246:R161-R183.18. Boivin DB, Czeisler CA, Dijk DJ, et al. Complex interactionof the sleep-wake cycle and circadian phase modulates mood inhealthy subjects. Arch Gen Psychiatry. 1997;54:145-152.19. Wehr T, Wirz-Justice A, Goodwin F, Duncan W, Gillin J. Phaseadvance of the sleep-wake cycle as an antidepressant. Science.1979;206:710-713.20. Wirz-Justice A, Kräuchi K, Brunner DP, et al. Circadianrhythms and sleep regulation in seasonal affective disorder. ActaNeuropsychiatrica. 1995;7:41-43.21. Koorengevel KM, Beersma DGM, den Boer JA, van den Hoof-dakker RH. Mood regulation in seasonal affective disorder pa-tients and healthy controls studied in forced desynchrony. Psy-chiatry Res. 2003;117:57-74.22. Szuba MP, Baxter LRJ, Altshuler LL, et al. Lithium sustainsthe acute antidepressant effects of sleep deprivation: preliminaryfindings from a controlled study. Psychiatry Res.1994;51:283-295.23. Benedetti F, Colombo C, Barbini B, Campori E, Smeraldi E.Ongoing lithium treatment prevents relapse after total sleep de-privation. J Clin Psychopharmacol. 1999;19:240-245.24. Colombo C, Lucca A, Benedetti F, Barbini B, Campori E,Smeraldi E. Total sleep deprivation combined with lithium andlight therapy in the treatment of bipolar depression: replicationof main effects and interaction. Psychiatry Res.2000;95:43-53.25. Smeraldi E, Benedetti F, Barbini B, Campori E, Colombo C.Sustained antidepressant effect of sleep deprivation combinedwith pindolol in bipolar depression. A placebo-controlled trial.Neuropsychopharmacol. 1999;20:380-385.26. Benedetti F, Barbini B, Lucca A, Campori E, Colombo C,Smeraldi E. Sleep deprivation hastens the antidepressant actionof fluoxetine. Eur Arch Psychiatry Clin Neurosci. 1997;247:100-103.27. Benedetti F, Colombo C, Pontiggia A, Bernasconi A, FloritaM, Smeraldi E. Morning light treatment hastens the antidepres-sant effect of citalopram: a placebo-controlled trial. J Clin Psy-chiatry. 2003;64:648-653.28. Neumeister A, Goessler R, Lucht M, Kapitany T, Bamas C,Kasper S. Bright light therapy stabilizes the antidepressant effectof partial sleep deprivation. Biol Psychiatry.1996;39:16-21.29. Benedetti F, Colombo C, Serretti A, et al. Antidepressant ef-fects of light therapy combined with sleep deprivation are influ-enced by a functional polymorphism within the promoter of theserotonin transporter gene. Biol Psychiatry. 2003;54:687-692.30. Wirz-Justice A, Richter R. Seasonality in biochemical deter-minations: a source of variance and a clue to the temporal inci-dence of affective illness. Psychiatr Res.1979;1:53-60.31. Carlsson A, Svennerhold L, Winblad B. Seasonal and circa-dian monoamine variations in human brains examined postmortem. Acta Psychiatr Scand. 1980;61(suppl 280):75-85.32. Lambert GW, Reid C, Kaye DM, Jennings GL, Esler MD. Ef-fect of sunlight and season on serotonin turnover in the brain.Lancet. 2002;360:1840-1842.33. Lam RW, Tam EM, Yatham LN, Shiah IS, Zis AP. Seasonaldepression: the dual vulnerability hypothesis revisited. J AffectDisord. 2001;63:123-132.

STRATÉGIES CHRONOBIOLOGIQUES POUR LES BESOINSINSATISFAITS DANS LE TRAITEMENT DE LA DÉPRESSION

L es stratégies chronobiologiques peuvent représenter un moyen efficacepour faire face à certains besoins insatisfaits dans le traitement de la dé-pression. Ce sont : le raccourcissement du temps de latence qui précède

l’apparition de l’effet antidépresseur, la lutte contre les symptômes résiduelset la prévention de la rechute à long terme. La lumière est le traitement de choixpour la dépression hivernale (ou trouble affectif saisonnier, TAS). La lumino-thérapie, comme adjuvant au traitement dans la dépression non saisonnièremajeure comme dans la dépression chronique et résistante au traitement, accé-lère et potentialise la réponse au traitement. La lumière est aussi efficace dansla dépression bipolaire ; chez ces patients présentant ce trouble, le « traitementpar l’obscurité » (nuits longues) peut diminuer les symptômes maniaques etarrêter les cycles rapides. La privation totale ou partielle de sommeil dans la se-conde partie de la nuit (mieux connue sous le nom de « traitement par l’éveil »)induit une amélioration marquée le jour suivant. Cette amélioration peut êtremaintenue avec des traitements concomitants par les antidépresseurs, le li-thium, la luminothérapie, l’avance de phase de sommeil ou l’association de plu-sieurs de ces mesures. Un contrôle soigneux du cycle jour-nuit et de l’heure desrepas, de l’activité et du sommeil peut apparaître comme une méthode démo-dée (mise en place de « structures journalières ») appartenant à une obsolètepsychiatrie d’institutionnalisation. Cependant, ces méthodes apparemmentsimples retrouvent une nouvelle légitimation quand on les reconsidère à l’inté-rieur du cadre de la chronobiologie moderne, puisque l’utilisation bien réglée de« synchroniseurs », ou « zeitgebers » peut améliorer le traitement des troublesaffectifs.

✦

228 Pathophysiology and treatment of affective disorders: melatonin and serotonin – Hamon and others MEDICOGRAPHIA, VOL 27, No. 3, 2005

For almost 50 years, since the empirical dis-covery of the antidepressant properties of themonoamine oxidase inhibitors (MAOIs) and

tricyclics, the “monoamine hypothesis,” accordingto which depression involves imbalances in sero-tonergic, noradrenergic, and possibly dopaminer-gic function, has held sway over the concepts andtheories put forward to explain the pathophysiolo-gy of depression. More recently, however, other ap-proaches to the understanding of the central mech-anisms of antidepressant drugs have been explored,and have revealed the existence of important neu-robiological changes in response to chronic treat-ment with these drugs.


F or over 40 years, research into the pathogenesis of depression and thedevelopment of suitable drugs for effective treatment has been domina-ted by the monoamine hypothesis, which states that depression involves

imbalances in serotonin, norepinephrine, and possibly dopamine function. Al-though these monoamine neurotransmitters—the respective roles of which arereviewed—are unquestionably involved, the deficit in monoamines is only partof the story, and other events besides monoamine imbalance should be takeninto account. There is a clear need for more effective, better tolerated, and morerapidly acting antidepressant agents. The development of new antidepressantdrugs based on other mechanisms than the monoamine hypothesis holds muchpromise. Increasing attention is being focused on evidence that various affectivedisorders, including depression, exhibit abnormal patterns of circadian rhythms.This article highlights the impact of melatonin and its receptors on depression,and looks at a recently developed antidepressant with documented clinical ef-ficacy, agomelatine. This agent, characterized by melatonin agonist and 5-HT2C

receptor antagonist properties, heralds a new concept in the treatment of de-pression.Medicographia. 2005;27:228-235. (see French abstract on page 235)

Keywords: depression; monoamine hypothesis; neurotransmitter; circadianrhythm; melatonin; serotonin; agomelatine

Michel HAMON, PhDUMR 677 INSERM-UMPC, Faculté de Médecine Pitié-Salpêtrière Paris, FRANCE

Pierre-Alain BOYER, PhD Elisabeth MOCAËR, PhDIRIS, 6 place des Pléiades Courbevoie, FRANCE

▲

Address for correspondence: Prof Michel Hamon, UMR 677 INSERM-UMPC, Faculté de Médecine Pitié-Salpêtrière, 75634 Paris Cedex 13, FRANCE(e-mail: [email protected])

New perspectives in the pathophysiology and treatmentof affective disorders: the role

of melatonin and serotonin b y M . H a m o n , P. - A . B o y e r ,

a n d E . M o c a ë r , F r a n c e


5-HT 5-hydroxytryptamine (serotonin)BDNF brain-derived neurotrophic factorCREB cyclic (adenosine monophosphate) response

element binding (protein)CRF corticotropin-releasing factorECT electroconvulsive therapyHPA hypothalamo-pituitary-adrenocortical (axis)MAOI monoamine oxidase inhibitorMT melatonin (receptor) NE norepinephrineNK neurokinin (receptor)NMDA N-methyl-D-aspartateSCN suprachiasmatic nuclei SNRI selective norepinephrine reuptake inhibitorSSRI selective serotonin reuptake inhibitor

pathways, but of other pathways as well. In partic-ular, the “monoamine hypothesis” fails to explainthe 3 to 4 weeks’ delay in the therapeutic efficacyof MAOIs, tricyclics, SSRIs, and SNRIs. The tempo-ral mismatch between the rapid elevations in extra-cellular levels of monoamines induced by antide-pressant agents, and their slow onset of therapeuticaction reflects the initiation of adaptive mechanismsinvolving secondary or even tertiary changes inreceptor density, intracellular signaling, synaptictransmission, neuronal architecture, and neuroge-nesis.4

The aforementioned adaptive changes in mono-aminergic receptors were initially claimed to reflecttherapeutically relevant neurobiological changesinduced by long-term antidepressant treatment.Downregulationofcorticalβ1-adrenergicand5-HT2Aserotonergic receptors was long considered as a bi-ological marker of effective antidepressant therapywith tricyclics. This was confirmed by evidence ofpostmortem upregulation of these receptors in de-pressed patients having committed suicide.5 How-ever, other findings indicated that electroconvulsivetherapy (ECT) was associated with upregulation ofcortical 5-HT2A receptors, and that chronic treat-ment with most SSRI antidepressants was unableto downregulate cortical β1-receptors. Furthermore,functional changes after chronic administration ofat least some classes of antidepressants were evi-denced for other receptor types, especially 5-HT1Areceptors. In particular, there have been consistentreports of desensitization of 5-HT1A autoreceptorsin the dorsal raphe nucleus, with no changes inhippocampal 5-HT1A heteroreceptors, after chronictreatment with MAOIs, SSRIs, or NK1 receptor an-tagonists6,7 as well as after transcranial magneticstimulation8 and ECT.9 In contrast, hypersensitivi-ty of hippocampal 5-HT1A receptors with no changesin 5-HT1A autoreceptors has been found to occurin response to chronic administration of tricyclicantidepressants.10,11 Nevertheless, the “monoaminehypothesis” did not entirely account for the actionof the antidepressants, and researchers began look-ing for other possible mechanisms of action.

Other approaches to the understanding of the mechanism of action of antidepressant drugs

An important consideration was that of the possibleinvolvement of the hypothalamo-pituitary-adreno-cortical (HPA) stress axis: upregulation of gluco-corticoid receptors, notably in the hippocampus,was evidenced with various antidepressant drugsfollowing long-term administration. This changeis very probably linked to the therapeutic action ofthese drugs, since it contributes to restoring normalfunctioning of the depression-associated deficientnegative feedback control of the HPA axis, there-by decreasing excessive cortisol secretion back tonormal in depressive patients. High levels of gluco-corticoids are generally associated with a negativeinfluence upon mood, as well as deleterious struc-tural changes in the hippocampus, perhaps via re-duced brain-derived neurotrophic factor (BDNF)

The “monoamine hypothesis”

The monoaminergic pathways play a crucial role inthe control of mood and cognition, as well as of en-docrine secretions and chronobiotic rhythms, all ofwhich are disrupted in depressive states. It was thusvery logical to postulate a perturbation of mono-aminergic transmission in the etiology of depres-sive disorders. This has been consistently borne outby the findings from monoamine depletion stud-ies in patients, which confirmed the importance offunctionally competent monoaminergic pathwaysfor combating depressive states, as well as by theproven clinical efficacy of available treatments ofdepression in restoring the compromised activity ofthe corticolimbic monoaminergic pathways. Drugsfacilitating neurotransmission mediated by mono-amine neurotransmitters, especially serotonin (5-hydroxytryptamine, 5-HT) and norepinephrine (NE),generally exert a positive influence on mood. Byblocking the key enzyme in the catabolic pathwaysof both 5-HT and NE, MAOIs increase the concen-trations of these neurotransmitters in the synap-tic cleft at serotonergic and noradrenergic receptorlevel. In contrast, tricyclic antidepressants, whichhave also been shown to increase the concentra-tions of monoamines at the receptor level, act viablockade of the reuptake process of these neuro-transmitters from the extracellular space. Severalkey observations demonstrated that the resultingincrease in extracellular concentrations of both5-HT and NE entirely accounted for the antidepres-sant effects of tricyclics. Drugs were therefore de-signed to mimic the inhibitory effects of tricyclicsvia specific action on the 5-HT transporter (selec-tive serotonin reuptake inhibitors, or SSRIs) or theNE transporter (selective norepinephrine reuptakeinhibitors, or SNRIs), or both. These drugs, whichincreased extracellular concentrations of 5-HT, NE,or both, were shown to exert potent antidepressantactions, and are currently the drugs most widelyused in the treatment of major affective disorders.

Second-generation antidepressants are safer thanfirst-generation agents because SSRIs and SNRIsdo not generally interact with the various receptortypes (histaminergic, cholinergic, α-adrenergic, etc)on which tricyclic antidepressants act as antago-nists. Nevertheless, second-generation antidepres-sants offer little advantage in terms of efficacy, andthere remains a need for antidepressants that do notrequire several weeks of administration prior to fullexpression of clinical efficacy.1,2 Furthermore, thereis also a need to reduce undesirable side effects,such as sexual dysfunction, insomnia, and weightgain, which make the use of at least some of thesedrugs rather delicate, since they compromise pa-tient compliance and reduce drug efficacy.3

It is now well recognized that synaptic facilita-tion and augmentation of the levels and effects ofNE and 5-HT are only part of the explanation of theaction of antidepressants. The pure “monoaminehypothesis” is too restrictive, and depression can-not be simplistically ascribed to the dysfunction ofmonoaminergic systems only. Rather, it reflects acomplex disruption not only of monoaminergic

229Pathophysiology and treatment of affective disorders: melatonin and serotonin – Hamon and others MEDICOGRAPHIA, VOL 27, No. 3, 2005


havioral effects. In addition, the antidepressantshave been found to modulate the expression of var-ious factors involved in cell survival and growth,such as cyclic adenosine monophosphate responseelement binding protein (CREB), bcl-2, and mito-gen-activated protein (MAP) kinases.4,29,30

Finally, the complex pattern of mutual interac-tions between glutamatergic and monoaminergicnetworks plays a crucial role in the control of moodand cognition.31,32 Excessive corticolimbic gluta-mate release upon chronic exposure to stress maycontribute to the development of depressive states.33

In this regard, most studies have focused on therole of detrimental N-methyl-D-aspartate receptor(NMDA)–mediated alterations in the structure ofhippocampal neurons.4,31 A number of antidepres-sants, as well as ECT, exert regulatory actions on theNMDA receptor complex, and some NMDA receptorantagonists have antidepressant-like properties.34

Depression and abnormal circadian rhythms

In humans, the circadian pacemaker, or biologicalclock, is the site of the generation and entrainmentof circadian rhythms. It is located in the suprachi-asmatic nuclei (SCN) of the anterior hypothalamus.The great majority of physiological, metabolic, andbehavioral functions are controlled by the circadianpacemaker and are often used as circadian phasemarkers. The circadian pacemaker is sensitive tolight throughout the 24-h cycle and to the phase-shifting effects of various chemical and pharmaco-logical components, including melatonin, whichacts on the circadian clock through melatonin re-ceptors located in the SCN.35

A wide range of affective disorders, including uni-polar and bipolar depression, mania, seasonal af-fective disorder, and premenstrual dysphoric disor-der, are characterized by disorganization of internalrhythms.

Disruption of the circadian timing system canmanifest in several ways. The amplitude of an oscil-lation can be altered, notably through changes in:(i) the number of SCN neurons or their connectionswith other brain areas or peripheral targets; or (ii)the neuronal traffic to the effector systems. Thephase of the rhythms can also be altered, or theremight be a partial or complete failure to remaindriven by the prevailing light-dark cycle throughdisruption of neuronal input pathways that impose24-h rhythmicity on the intrinsically non–24-h,rhythmic SCN. A very large number of physiolog-ical variables showing circadian abnormalities indepressed patients are described in the literature.Alterations in circadian rhythms in body tempera-ture, hormone secretions, and vigilance state veryoften accompany endogenous depression,36-39 andunavoidable alterations in normal circadian rhythmscan trigger depressive episodes in human.40 Mostrhythms are phase advanced with respect to thesleep-wake cycle, diminished in amplitude, and/orshow day-to-day variability in entrainment. Al-though altered rhythmicity could be either a causeor an effect of an altered affective state, the high

synthesis, excessive glutamate release, and/or re-duced neuronal glucose uptake.4 In line with thesefindings, glucocorticoid synthesis inhibitors andglucocorticoid receptor antagonists have been re-ported to exert antidepressant-like effects.12,13

Consistent with the hypothesis of a causal rela-tionship between functional abnormalities in theHPA stress axis and affective disorders, several stud-ies have demonstrated elevated levels of cortico-tropin-releasing factor (CRF) in the cerebrospinalfluid of patients with moderate-to-severe depres-sion. Indeed, treatment with antagonists at CRFreceptors were shown to relieve behavioral deficitsin animal models of depression14,15 and to improvemood in depressed subjects.

Neurokinin type 1 (NK1) receptor antagonistshave been claimed to exert antidepressant activi-ty independent of monoamines in relevant animalparadigms as well as in humans, by inhibiting theaction of substance P, a tachykinin neuropeptidethat is localized in brain regions involved in the con-trol of stress and emotion.16

However, there is a possibility that both CRF andNK1 antagonists may also act, indirectly, throughmonoaminergic mechanisms. Indeed, tachykininNK1 receptor antagonists activate noradrenergicand dopaminergic pathways innervating the hip-pocampus and the frontal cortex, and long-termtreatments with these antagonists produce the sameadaptive changes in 5-HT neurotransmission (no-tably, functional desensitization of 5-HT1A auto-receptors, responsible for increased 5-HT tone) asthose observed after chronic treatment with MAOIsor SSRIs.17 In addition, recent investigations clear-ly demonstrated that CRF antagonists indirectlyenhanced the activity of serotonergic pathways4,18

although they do not, unlike tachykinin NK1 recep-tor antagonists, activate dopaminergic input to thecortex and may inhibit the activity of noradrener-gic neurons.19,20

Levels of BNDF are usually found to be increasedin response to long-term antidepressant drug treat-ment. This is probably related, at least in part, tothe negative influence of glucocorticoids on BDNFsynthesis.21,22 Actually, chronic intracerebroventric-ular infusion of BDNF exerts antidepressant-likeeffects via activation of its receptors, TrkB.23 Find-ings also suggest that progressive induction of BDNFand the resulting enhancement of neurogenesis arecausally related to the onset of antidepressant activ-ity.24 Experimental studies on stress and antidepres-sant treatment have recently implicated neurogen-esis in the etiology of major depressive disorders.25,26

So far, all chronic treatments aimed at alleviatingdepression, including ECT, have been shown tostimulate the proliferation of progenitor cells at theorigin of granular neurons in the dentate gyrus ofthe hippocampus.27 The precise functional signif-icance of these newly generated neurons in thepathophysiology of mood disorders is still disputed,but a recent study by Santarelli et al28 supports theidea that granule cell proliferation is directly relat-ed to the therapeutic action of antidepressant drugs,since suppression of this process by x-ray exposureresulted in loss of the antidepressant-related be-



Depression Rating Scale (HDRS) scores in a popu-lation of patients with symptomatic major depres-sion and another population in remission. Krahn etal59 also reported that successful ECT in patientswith major depression is associated with a decreasein 6-sulfatoxymelatonin urinary excretion.

However, other authors47,49 have unsuccessfullyattempted to find correlations between the clinicalmanifestation of depression, intensity of symptoms,and the course of the disease on the one hand, andmelatonin secretion disturbances on the other.Morerecently, Szymanska et al51 have shown that plas-ma melatonin levels do not differentiate patientsin terms of severity of the depressive symptoms, ab-normal levels being observed in patients with ma-jor depression as well as in patients in the remissionphase. Nevertheless, changes in urinary excretion of6-sulfatoxymelatonin have been reported to enablethe distinction between antidepressant respondersand nonresponders.60

In contrast, other authors hypothesize that theefficacy of antidepressants of different classes (de-sipramine, fluvoxamine, mianserin, venlafaxine) isattributable to the enhancement of melatonin secre-tion (which could involve monoaminergic mecha-nisms acting on the pineal gland61-63) and/or the in-hibition of the degradation of melatonin.64

Since melatonin has resynchronizing effects inboth animals and humans suffering from circadi-an rhythm disorganization, the exciting possibilitywas raised that a drug able to mimic the effects ofthis neurohormone and easily cross the blood-brainbarrier would be beneficial in the treatment of de-pression. This hypothesis led to the synthesis of ago-melatine.

Agomelatine: a new antidepressant with a novel

mechanism of action

Agomelatine (S 20098) is a potent ligand of mela-tonin receptors, with agonist properties at clonedhuman melatoninergic MT1 and MT2 receptor lev-el. In addition, it is also endowed with antagonistproperties at cloned human 5-HT2B and 5-HT2C re-ceptor level.

prevalence of circadian dysfunction in affective statessuggests that the circadian system holds importantclues regarding the etiology and treatment of affec-tive disorders.

Much less attention has been paid to variations incircadian amplitude than to those affecting phaseand period, but there has been increasing recogni-tion that blunted circadian amplitude may be oneof the main chronobiological abnormalities in de-pression.38,40,41 This suggests that novel treatmentsaddressing such abnormalities could have poten-tial value in the management of depression, andthat melatonin could be a target for antidepressanttherapy.

Melatonin: a target for antidepressant therapy

Melatonin (N-acetyl-5-hydroxytrytamine) is an en-dogenous neurohormone secreted by the pinealgland, whose circadian and nocturnal secretion iscontrolled by the SCN through β-adrenergic recep-tors. Melatonin is an endogenous synchronizer ofbiological rhythms in mammals and its secretionand actions are tightly related to seasonal and light-dark cycles.42-44 Melatonin’s ability to control cir-cadian rhythm synchronization has triggered nu-merous studies seeking to determine its role in thepathogenesis of various psychiatric disorders, in-cluding depression.45,46

Given the extensive perturbations of circadianrhythms in depressive disorders on the one hand,and the chronobiotic properties of melatonin onthe other, it was logical to expect that melatoninwould exhibit potential clinical benefits. Such hopeswere encouraged by reports of alterations of mela-tonin secretory patterns in various psychiatric dis-orders.46 Both decreases36,47,48 and increases37,39,49-51

in melatonin plasma levels have been evidenced indepressed patients. Both a familial vulnerability inthe endogenous melatonin signal in subjects proneto depression and an abnormal duration of the mela-tonin signal in subjects with current major depres-sion have been hypothesized.52

Preclinical studies have demonstrated that mela-tonin is active in several experimental models re-sponsive to antidepressant treatment (Table I).53

This antidepressant-like activity of melatonin hasbeen demonstrated, in particular, during repeatednighttime administration in a “chronic mild stress”paradigm with diurnal and sleep rhythm disrup-tions, in both C3H/He mice54 and rats.37,55 In addi-tion, melatonin, like effective antidepressants, isactive in the forced swimming test,37,56,57 and N-acetyltransferase “knocked-down” C57BL/6J mu-tant mice display a longer immobility time in thistest, regardless of circadian rhythm.58

Although most clinical studies have pointed outthe presence of melatonin secretion disturbancesin depressed patients, the link between melatoninand clinical depressive states remains unclear. Nev-ertheless, a relationship between melatonin levelsand depression has been reported by Souetre et al38

who found a correlation between low concentra-tions of circulating melatonin and the Hamilton



Melatonin 5-HT2C Agomelatineantagonist

Despair test -- -- + TypicalLearned helplessness -- -- + antidepressantBulbectomy -- ? + models

Chronic mild stress + ? + Circadian rhythmTransgenic model + ? + models

Isolated aggressive mice +/-- + + Stress orMarble burying test +/-- + + anxiety models

Table I. The antidepressant activity of melatonin, 5-HT2c antagonists, and agomelatine. Reproduced from reference 53: Lanfumey L, Boyer PA, Hamon M, Mocaër E. Antidepressant profile of themelatonin agonist and 5-HT2C antagonist melatonin (S 20098). Preclinical studies. Eur Neuropsychophar-macol. 2003;13(suppl 4):S274. Copyright © 2003, Elsevier B. V.

ments effectively reduced, to the same extent, thelocomotor hyperactivity caused by bulbectomy.74

To date, the model of chronic mild stress is con-sidered to be the most reliable model for identifyingthe antidepressant-like properties of drugs, becauseit focuses on anhedonia, one of the key symptomsof depression. In addition, antidepressants have tobe administered on a chronic basis in order to re-verse the induced behavioral deficits, in line withthe delay in the onset of therapeutic action of thesedrugs. In this paradigm, rats are subjected for sev-eral weeks to unpredictable minor stress sessionsthat cause behavioral alterations, which closely re-semble those most frequently observed in depressedpatients. In particular, rats develop anhedonia, asshown by their decreased voluntary intake of a su-crose solution. Chronic IP administration of agome-latine (10 and 50 mg/kg daily) was compared withtreatments with melatonin (10 and 50 mg daily) orthe classic antidepressants imipramine (10 mg/kgdaily) and fluoxetine (10 mg/kg daily), in this mod-el of depression. Agomelatine reversed anhedoniainduced by chronic mild stress, to a similar extentto that achieved by the other antidepressant drugs(Figure 1). Thus, agomelatine was found to be a po-tent and rapidly acting antidepressant in this mod-el. Once installed, the effect of agomelatine wasquite robust, with no withdrawal-induced relapse at1 week after cessation of agomelatine treatment.55

Interestingly, whether agomelatine was adminis-tered in the morning, ie, at the beginning of thelight period, or in the late evening, just prior to thestart of the dark period, had no influence on its ca-pacity to restore the rats’ preference for sucrose,suggesting that the antidepressant-like propertiesof this drug involve mechanisms independent ofcircadian rhythms. In contrast, under the same con-ditions, melatonin was found to partially restore therats’ preference for sucrose only when the neuro-hormone was injected just prior to the onset of thedark phase (Figure 1). This difference clearly indi-cates that the neurobiological mechanisms under-lying the antidepressant-like action of agomelatineare not shared by melatonin. Indeed, blockade ofmelatonin receptors by the MT1/MT2 receptor antag-onist S 22153 failed to reverse the antidepressant-like effect of agomelatine injected in the morning.55

Finally, alongside agomelatine’s specific proper-ties linked to its novel antidepressant profile of ac-tion, agomelatine was shown to exert a stimulatoryinfluence on cell proliferation within the subgran-ular layer of the dentate gyrus in rats. Chronic ad-ministration of agomelatine (40 mg/kg IP daily for21 days) significantly increased the number of cellslabeled by BrdU (a thymidine analog that is incor-porated in DNA synthesized during mitosis) in adultrat brain, within a circumscribed region of the hip-pocampus, the ventral part of the subgranular lay-er.75 A 3-week treatment with agomelatine (10 or50 mg/kg IP daily) also reversed the deficit in gran-ule cell proliferation that had been induced in adultrats from a mother subjected to repeated stress dur-ing gestation.76 With regard to cell proliferation inthe dentate gyrus, agomelatine therefore acts likeall other antidepressant therapies tested to date.

◆ Melatonin agonist properties of agomelatine◆ Action on circadian rhythmsBehavioral studies in rodents have shown that sys-temic administration of agomelatine could dose-dependently alter the functioning of the circadianclock. The chronobiotic properties of agomelatinewere first demonstrated by its ability to mimic theaction of melatonin in the synchronization of circa-dian rhythm patterns of locomotor activity, runningwheel activity, and body temperature in rodents.65,66

Agomelatine dose-dependently drives rhythmsby setting up a circadian pattern in free-runninganimals. This ability to synchronize rest-activityrhythms in free-running animals requires the in-tegrity of the SCN, lending further support to itsaction on melatonin receptors. In addition, agome-latine can reset a preexisting circadian rhythm fol-lowing a phase shift, since it has the ability to fasterreentrain rats following an 8-h phase advance of thelight-dark cycle in vivo. These effects are dose-de-pendent, up to a maximum usually obtained withoral administration of 8 to 10 mg/kg in rodents.67

Similarly, in aged animals, long-term treatmentwith agomelatine reduces the time needed for reen-trainment following a phase advance in the light-dark cycle.68 Finally, in aged animals with reducedresponsiveness to zeitgebers, agomelatine can re-store the sensitivity of the circadian clock to envi-ronmental synchronizers.69

◆ Efficacy of agomelatine in animal models of depressionInterest in agomelatine has recently increased dueto its potential use as a novel antidepressant agent,as demonstrated in a number of animal studiesusing well-validated animal models of depression(Table I).53 These findings support the potential useof agomelatine as an antidepressant in humans, andindeed, agomelatine has been shown to be clinical-ly effective in depressed patients.70,71

The antidepressant efficacy of agomelatine wasfirst investigated using the forced swimming test.Both acute and repeated treatment with agomela-tine (2, 10, and 50 mg/kg) showed an antidepres-sant effect in rats and mice, as evidenced by a sig-nificantly reduced duration of immobility at alldoses tested. Results following agomelatine admin-istration were comparable to those with imipramineand fluoxetine, whereas treatment with melatoninshowed no effect.72

The antidepressant efficacy of agomelatine wasfurther investigated in a large series of experiments.Results from the learned-helplessness model of de-pression also demonstrated agomelatine to be aneffective antidepressant. In this study, agomelatine(10, 50 mg/kg per os) was shown to be equivalentto imipramine in its ability to reverse the learningdeficit caused by exposure to aversive stimuli. Once-daily administration was superior to twice-dailyadministration of agomelatine, with the most sig-nificant antidepressant effects being noted with the10 mg/kg dose.73 The effects of agomelatine were alsostudied in the model of olfactory bulbectomizedrats. Ambulation scores were statistically similaramong animals treated with agomelatine (10 or50 mg/kg) or imipramine (10 mg/kg). Both treat-





pressant effects of agomelatine, is a question whichcannot be answered at present.

All new antidepressants for which new mecha-nisms of action are claimed should be investigatedfor the presence of possible interactions with mono-aminergic systems, since it is unlikely that antide-pressant efficacy can be achieved without recruit-ing monoaminergic mechanisms, either directlyor indirectly by actions upstream or downstream ofmonoaminergic neurons. Because of its 5-HT2C re-ceptor antagonist properties, both acute and chron-ic administration of agomelatine (10-40 mg/kg IP)produce concomitant dopamine and NE overflowin the frontal cortex.78 It can be assumed that thestimulatory effect of agomelatine on these mono-aminergic systems also contributes to its antide-pressant action, since these neurotransmitters mod-ulate cognitive-attentional performance and moodin the frontal cortex. On the other hand, 5-HT out-flow is unchanged by agomelatine treatment, whichindicates that mechanisms underlying its antide-pressant-like action are totally unrelated to thoseof tricyclics, SSRIs, and MAOIs,78,79 which all gen-erally elicit a marked increase in 5-HT outflow.

Conclusion

New antidepressants based on other mechanisms ofaction than the monoamine hypothesis are a pro-mising approach in the development of drugs withbetter tolerance and efficacy. One of the most ex-citing prospects in the treatment of depression hasbeen the development of the new antidepressantagomelatine, a compound that exhibits strong mela-tonin receptor agonist activity and 5-HT2C receptorantagonist properties. This unique combination ofpharmacological properties very probably explainsthe efficacy of agomelatine observed in all validat-ed animal models of depression. Agomelatine thusrepresents a new concept in the treatment of de-pression, and provides a strong incentive for thedevelopment of antidepressant drugs with mixedmechanisms of action encompassing monoaminer-gic as well as nonmonoaminergic systems. ❒

◆ 5-HT2C receptor antagonist propertiesof agomelatineAs mentioned above, agomelatine does more thansimply mimic the action of melatonin, and it hasbeen hypothesized that the melatonin agonist prop-erties of agomelatine may not, by themselves, besufficient to sustain its clear-cut clinical antidepres-sant efficacy.

In addition to the properties that led to agome-latine’s classification as a selective melatonin re-ceptor agonist, extensive studies have identified an-other mode of action for this drug. Using a classictest setting for 5-HT2C antagonists, agomelatine wasfound to antagonize the penile erections normallyinduced by 5-HT2C receptor activation in Wistar rats.Penile erections were induced by administration ofthe 5-HT2C receptor agonists mCPP and Ro-60-0175,and diminished, in a dose-dependent manner, whenagomelatine was added. In contrast, melatonin ad-ministration had no effect in this paradigm. It wastherefore concluded that this effect of agomelatinedid not involve its melatonin receptor agonist prop-erties, but, instead, that it was due to 5-HT2C recep-tor antagonism.77

Recent results from receptor binding and signal-ing studies have lent further support to the hypoth-esis of a secondary mode of action for agomelatine.Analyses of receptor binding in transfected Chinesehamster ovary (CHO) cells showed that agomela-tine acts as an antagonist at both the 5-HT2C and5-HT2B receptor subtype level. Agomelatine, there-fore, has a unique pharmacological profile, result-ing from stimulation of MT1 and MT2 receptors andblockade of 5-HT2C and 5-HT2B receptors.78 In vivoexperiments in relevant animal models (see above)have clearly demonstrated the reality of 5-HT2C re-ceptor blockade with agomelatine at doses used toelicit antidepressant-like effects. In contrast, verylittle is known to date regarding 5-HT2B receptorblockade and its possible behavioral consequencesin vivo (on food intake, sleep, etc). Thus, whether ornot blockade of the 5-HT2B receptor subtype, whichis expressed at low levels in the brain, contributes,to a more or less significant extent, to the antide-

Melatonin antagonist

Controls

Stress

Weeks of treatment

** P<0.01, *** P<0.001 relative to vehicle or drug treated control animals at week 0 # P<0.05, ## P<0.01, ### P<0.001 relative to drug-treated stressed animals at week 0

*****

***

*** ******

***

*** ***

***

**

Sucr

ose

inta

ke (g

)

18

16

14

12

10

8

6

40 1 2 3 4 5 6 7

VehicleMelatoninAgomelatine

+ Vehicle+ Melatonin+ Agomelatine

#

##

###

###

####

Figure 1. Agomelatine(morning treatment, 50 mg/kg IP) in chronicmild stress (effect on anhedonia). Reproduced from reference 55:Papp M, Gruca P, Boyer PA, Mocaer E. Effect of agomelatinein the chronic mild stress modelof depression in the rat. Neuro-psychopharmacology. 2003;28:694-703. Copyright © 2003, Elsevier Science, Ltd.



24. Malberg JE, Duman RS. Cell proliferation in adult hippocam-pus is decreased by inescapable stress: reversal by fluoxetine treat-ment. Neuropsychopharmacology. 2003;28:1562-1571.25. Jacobs BL, Van Praag H, Gage FH. Adult brain neurogenesisand psychiatry: a novel theory of depression. Mol Psychiatry.2000;5:262-269.26. Kempermann G, Kronenberg G. Depressed new neurons—adult hippocampal neurogenesis and a cellular plasticity hypoth-esis of major depression. Biol Psychiatry. 2003;54:499-503.27. Duman RS, Nakagawa S, Malberg J. Regulation of adult neu-rogenesis by antidepressant treatment. Neuropsychopharmacol-ogy. 2001;25:836-844.28. Santarelli L, Saxe M, Gross C, et al. Requirement of hippo-campal neurogenesis for the behavioral effects of antidepressants.Science. 2003;301:805-809.29. D’Sa C, Duman RS. Antidepressants and neuroplasticity. Bi-polar Disord. 2002;4:183-194.30. Manji HK, Moore GJ, Chen G. Lithium up-regulates the cy-toprotective protein Bcl-2 in the CNS in vivo: a role for neuro-trophic and neuroprotective effects in manic depressive illness.J Clin Psychiatry. 2000;61(suppl 9):82-96.31. Skolnick P, Legutko B, Li X, Bymaster FP. Current perspec-tives on the development of non-biogenic amine-based antide-pressants. Pharmacol Res. 2001;43:411-423.32. Zarate CA Jr, Du J, Quiroz J, et al. Regulation of cellular plas-ticity cascades in the pathophysiology and treatment of mood dis-orders: role of the glutamatergic system. Ann N Y Acad Sci.2003;1003:273-291.33. Reagan LP, Rosell DR, Wood GE, et al. Chronic restraint stressup-regulates GLT-1 mRNA and protein expression in the rat hip-pocampus: reversal by tianeptine. Proc Natl Acad Sci U S A.2004;101:2179-2184.34. Skolnick P. Antidepressants for the new millennium. Eur JPharmacol. 1999;375:31-40.35. Reppert SM, Weaver DR. Molecular analysis of mammaliancircadian rhythms. Annu Rev Physiol. 2001;63:647-676.36. Brown R, Kocsis JH, Caroff S, et al. Differences in nocturnalmelatonin secretion between melancholic depressed patients andcontrol subjects. Am J Psychiatry. 1985;142:811-816.37. Claustrat B, Chazot G, Brun J, Jordan D, Sassolas G. A chrono-biological study of melatonin and cortisol secretion in depressedsubjects: plasma melatonin, a biochemical marker in major de-pression. Biol Psychiatry. 1984;19:1215-1228.38. Souetre E, Salvati E, Belugou JL, et al. Circadian rhythms indepression and recovery: evidence for blunted amplitude as themain chronobiological abnormality. Psychiatry Res.1989;28:263-278.39. Thompson C, Franey C, Arendt J, Checkley SA. A compari-son of melatonin secretion in depressed patients and normal sub-jects. Br J Psychiatry. 1988;152:260-265.40. Wehr TA, Wirz-Justice A. Circadian rhythm mechanisms inaffective illness and in antidepressant drug action. Pharmaco-psychiatria. 1982;15:31-39.41. Wirz-Justice A. Biological rythms in mood disorders. In:Bloom FF, Kupfer DJ, eds. Psychopharmacology: The Third Gen-eration of Progress. New York, NY: Raven Press; 1995;999-1017.42. Borjigin J, Li X, Snyder SH. The pineal gland and melatonin:molecular and pharmacologic regulation. Annu Rev PharmacolToxicol. 1999;39:53-65.43. Reiter RJ. Pineal melatonin: cell biology of its synthesis andof its physiological interactions. Endocr Rev.1991;12:151-180.44. Sack RL, Lewy AJ, Hughes RJ. Use of melatonin for sleep andcircadian rhythm disorders. Ann Med. 1998;30:115-121.45. Brown GM. Light, melatonin and the sleep-wake cycle. J Psy-chiatry Neurosci. 1994;19:345-353.46. Pacchierotti C, Lapichino S, Bossini L, Pieraccini F, Castro-giovanni P. Melatonin in psychiatric disorders: a review on themelatonin involvement in psychiatry. Front Neuroendocrinol.2001;22:18-32.47. Beck-Friis J, Kjellman BF, Aperia B, et al. Serum melatoninin relation to clinical variables in patients with major depressivedisorder and a hypothesis of a low melatonin syndrome. Acta Psy-chiatr Scand. 1985;71:319-330.48. Wetterberg L, Aperia B, Gorelick DA, et al. Age, alcoholismand depression are associated with low levels of urinary mela-tonin. J Psychiatry Neurosci. 1992;17:215-224.49. Rubin RT, Heist EK, McGeoy SS, Hanada K, Lesser IM. Neu-roendocrine aspects of primary endogenous depression. XI. Serummelatonin measures in patients and matched control subjects.Arch Gen Psychiatry. 1992;49:558-567.50. Sekula LK, Lucke JF, Heist EK, Czambel RK, Rubin RT. Neu-roendocrine aspects of primary endogenous depression. XV: Math-ematical modeling of nocturnal melatonin secretion in major de-

REFERENCES1. Blier P. The pharmacology of putative early-onset antidepres-sant strategies. Eur Neuropsychopharmacol. 2003;13:57-66.2. Tamminga CA, Nemeroff CB, Blakely RD, Brady L, Carter CS,Davis KL et al. Developing novel treatments for mood disorders:accelerating discovery. Biol Psychiatry. 2002;52:589-609.3. Vida S, Looper K. Precision and comparability of adverse eventrates of newer antidepressants. J Clin Psychopharmacol.1999;19:416-426.4. Manji HK, Gottesman II, Gould TD. Signal transduction andgenes-to-behaviors pathways in psychiatric diseases. Sci STKE2003. 2003(207):e49.5. Stockmeier CA. Involvement of serotonin in depression: evi-dence from postmortem and imaging studies of serotonin recep-tors and the serotonin transporter. J Psychiatr Res.2003;37:357-373.6. Hensler JG. Differential regulation of 5-HT1A receptor-G pro-tein interactions in brain following chronic antidepressant ad-ministration. Neuropsychopharmacology. 2002;26:565-573.7. Le Poul E, Boni C, Hanoun N, et al. Differential adaptation ofbrain 5-HT1A and 5-HT1B receptors and 5-HT transporter in ratstreated chronically with fluoxetine. Neuropharmacology. 2000;39:110-122.8. Gur E, Lerer B, Dremencov E, Newman ME. Chronic repeti-tive transcranial magnetic stimulation induces subsensitivity ofpresynaptic serotonergic autoreceptor activity in rat brain. Neu-roreport. 2000;11:2925-2929.9. Ishihara K, Sasa M. Mechanism underlying the therapeuticeffects of electroconvulsive therapy (ECT) on depression. Jpn JPharmacol. 1999;80:185-189.10. Chaput Y, De Montigny C, Blier P. Presynaptic and postsynap-tic modifications of the serotonin system by long-term adminis-tration of antidepressant treatments. An in vivo electrophysiolog-ic study in the rat. Neuropsychopharmacology.1991;5:219-229.11. Tokarski K, Czyrak A, Mackowiak M, Wedzony K, Bijak M.Prolonged treatment with antidepressants increases the 5HT1A-mediated inhibition of hippocampal neurons without changingthe 5HT1A receptor binding. Acta Physiol Hung.1996;84:343-344.12. Belanoff JK, Kalehzan M, Sund B, Fleming Ficek SK, Schatz-berg AF. Cortisol activity and cognitive changes in psychotic ma-jor depression. Am J Psychiatry. 2001;158:1612-1616.13. Reus VI, Wolkowitz OM. Antiglucocorticoid drugs in thetreatment of depression. Expert Opin Investig Drugs. 2001;10:1789-1796.14. Griebel G, Simiand J, Serradeil-Le-Gal C, et al. Anxiolytic- andantidepressant-like effects of the non-peptide vasopressin V1b re-ceptor antagonist, SSR149415, suggest an innovative approachfor the treatment of stress-related disorders. Proc Natl Acad SciU S A. 2002;99:6370-6375.15. Griebel G, Simiand J, Steinberg R, et al. 4-(2-Chloro-4-methoxy-5-methylphenyl)-N-[(1S)-2-cyclopropyl-1-(3-fluoro-4-methylphenyl)ethyl]5-methyl-N-(2-propynyl)-1, 3-thiazol-2-amine hydrochloride (SSR125543A), a potent and selective cor-ticotrophin-releasing factor(1) receptor antagonist. II. Charac-terization in rodent models of stress-related disorders. J Phar-macol Exp Ther. 2002;301:333-345.16. Kramer MS, Cutler N, Feighner J, et al. Distinct mechanismfor antidepressant activity by blockade of central substance P re-ceptors. Science.1998;281:1640-1645.17. Froger N, Gardier AM, Moratalla R, et al. 5-hydroxytryptamine5-HT1A autoreceptor adaptive changes in substance P (neuro-kinin 1) receptor knock-out mice mimic antidepressant-induceddesensitization. J Neurosci. 2001;21:8188-8197.18. Reul JM, Holsboer F. Corticotropin-releasing factor recep-tors 1 and 2 in anxiety and depression. Curr Opin Pharmacol.2002;2:23-33.19. Lejeune F, Gobert A, Millan MJ. The selective neurokinin (NK)(1) antagonist, GR205,171, stereospecifically enhances mesocor-tical dopaminergic transmission in the rat: a combined dialysisand electrophysiological study. Brain Res. 2002;935:134-139.20. Millan MJ, Lejeune F, De Nanteuil G, Gobert A. Selectiveblockade of neurokinin (NK)(1) receptors facilitates the activityof adrenergic pathways projecting to frontal cortex and dorsalhippocampus in rats. J Neurochem. 2001;76:1949-1954.21. Smith MA, Makino S, Kvetnansky R, Post RM. Stress andglucocorticoids affect the expression of brain-derived neurotroph-ic factor and neurotrophin-3 mRNAs in the hippocampus. J Neu-rosci.1995;15:1768-1777.22. Nibuya M, Morinobu S, Duman RS. Regulation of BDNF andtrkB mRNA in rat brain by chronic electroconvulsive seizure andantidepressant drug treatments. J Neurosci. 1995;15:7539-7547.23. Siuciak JA, Lewis DR, Wiegand SJ, Lindsay RM. Antidepres-sant-like effect of brain-derived neurotrophic factor (BDNF).Pharmacol Biochem Behav.1997;56:131-137.

70. Lôo H, Hale A, D’haenen H. Determination of the dose of ago-melatine, a melatoninergic agonist and selective 5-HT2C antag-onist, in the treatment of major depressive disorder: a placebo-controlled dose range study. Int Clin Psychopharmacol. 2002;17:239-247.71. Emsley R. Efficacy and safety of agomelatine (25 mg/50 mg),a melatonergic and specific serotonergic antidepressant, in thetreatment of major depressive disorder: a randomised, doubleblind placebo-controlled study. Int J Neuropsychopharmacol.2004;7(suppl 1):S350-S351. 72. Bourin M, Mocaer E, Porsolt R. Antidepressant-like activityof S 20098 (agomelatine) in the forced swimming test in rodents:involvement of melatonin and serotonin receptors. J PsychiatryNeurosci. 2004;29:126-133.73. Bertaina-Anglade V, Mocaer E, Drieu-La-Rochelle C. Anti-depressant-like action of agomelatine (S20098) in the learnedhelplessness test. Int J Neuropsychopharmacol. 2002;5(suppl 1):1.74. Norman TR, Cranston I, Irons J. Effect of the novel antide-pressant agomelatine in the olfactory bulbectomised rat. Int JNeuropsychopharmacol. 2004;7(suppl):S461.75. Daszuta A, Banasr M, Hery M, Mocaer E. Agomelatine, a newantidepressant drug, increases cell proliferation in adult hippo-campus. Society for Neuroscience 33rd Annual Meeting of theSociety for Neuroscience, New Orleans, La, USA, November 8th-13th, 2003. 849.3.76. Morley-Fletcher S, Mairesse J, Mocaer E, Darnaudery M, Vil-tart O, Maccari S. Chronic treatment with agomelatine increaseshippocampal cell proliferation in prenatally stressed rats. Societyfor Neuroscience 33rd Annual Meeting of the Society for Neuro-science, New-Orleans, La, USA, November 8th-13th. 2003. 506.2077. Chagraoui A, Protais P, Filloux T, Mocaer E. Agomelatine(S 20098) antagonizes the penile erections induced by the stim-ulation of 5-HT2C receptors in Wistar rats. Psychopharmacology(Berl). 2003;170:17-22.78. Millan MJ, Gobert A, Lejeune F, et al. The novel melatonin ag-onist, agomelatine (S 20098, is an antagonist at 5-hydroxytryp-tamine 2C receptors, blockade of which enhances the activity offrontocortical dopaminergic and adrenergic pathways. J Phar-macol Exp Ther. 2003;306:954-964.79. Hanoun N, Mocaer E, Boyer PA, Hamon M, Lanfumey L. Dif-ferential effects of the novel antidepressant agomelatine (S 20098)versus fluoxetine on 5-HT1A receptors in the rat brain. Neuro-pharmacology. 2004;47:515-526.

pressives and normal controls. Psychiatry Res.1997;69:143-153.51. Szymanska A, Rabe-Jablonska J, Karasek M. Diurnal profileof melatonin concentrations in patients with major depression:relationship to the clinical manifestation and antidepressant treat-ment. Neuroendocrinol Lett. 2001;22:192-198.52. Tuunainen A, Kripke DF, Elliott JA, et al. Depression and en-dogenous melatonin in postmenopausal women. J Affect Disord.2002;69:149-158.53. Lanfumey L, Boyer PA, Hamon M, Mocaër E. Antidepressantprofile of the melatonin agonist and 5-HT2C antagonist melatonin(S 20098). Preclinical studies. Eur Neuropsychopharmacol. 2003;13(suppl 4):S274.54. Kopp C, Vogel E, Rettori MC, Delagrange P, Misslin R. Theeffects of melatonin on the behavioural disturbances induced bychronic mild stress in C3H/He mice. Behav Pharmacol.1999;10:73-83.55. Papp M, Gruca P, Boyer PA, Mocaer E. Effect of agomelatinein the chronic mild stress model of depression in the rat. Neuro-psychopharmacology. 2003;28:694-703.56. Overstreet DH, Pucilowski O, Retton MC, Delagrange P,Guardiola-Lemaitre B. Effects of melatonin receptor ligands onswim test immobility. Neuroreport. 1998;9:249-253.57. Shaji AV, Kulkarni SK. Central nervous system depressant ac-tivities of melatonin in rats and mice. Indian J Exp Biol.1998;36:257-263.58. Uz T, Manev H. Prolonged swim-test immobility of serotoninN-acetyltransferase (AANAT)-mutant mice. J Pineal Res. 2001;30:166-170.59. Krahn LE, Gleber E, Rummans TA, Pileggi TS, Lucas DL, Li H.The effects of electroconvulsive therapy on melatonin. J Electro-convuls Ther. 2000;16:391-398.60. Miller HL, Ekstrom RD, Mason GA, Lydiard RB, Golden RN.Noradrenergic function and clinical outcome in antidepressantpharmacotherapy. Neuropsychopharmacology. 2001;24:617-623.61. Durlach-Misteli C, Van Ree JM. Dopamine and melatonin inthe nucleus accumbens may be implicated in the mode of actionof antidepressant drugs. Eur J Pharmacol. 1992;217:15-21. 62. Franklin M, Clement EM, Campling G, Cowen PJ. Effect ofvenlafaxine on pineal melatonin and noradrenaline in the malerat. J Psychopharmacol. 1998;12:371-374.63. Palazidou E, Papadopoulos A, Ratcliff H, Dawling S, CheckleySA. Noradrenaline uptake inhibition increases melatonin secre-tion, a measure of noradrenergic neurotransmission, in depressedpatients. Psychol Med. 1992;22:309-315.64. Hartter S, Wang X, Weigmann H, et al. Differential effects offluvoxamine and other antidepressants on the biotransformationof melatonin. J Clin Psychopharmacol. 2001;21:167-174.65. Krauchi K, Cajochen C, Mori D, Graw P, Wirz-Justice A. Ear-ly evening melatonin and S-20098 advance circadian phase andnocturnal regulation of core body temperature. Am J Physiol.1997;272:R1178-R1188.66. Van Reeth O, Olivares E, Zhang Y, et al. Comparative effects ofa melatonin agonist on the circadian system in mice and Syrianhamsters. Brain Res. 1997;762:185-194.67. Redman JR, Guardiola-Lemaitre B, Brown M, Delagrange P,Armstrong SM. Dose dependent effects of S-20098, a melatoninagonist, on direction of re-entrainment of rat circadian activityrhythms. Psychopharmacology (Berl). 1995;118:385-390.68. Weibel L, Turek FW, Mocaer E, Van Reeth O. A melatoninagonist facilitates circadian resynchronization in old hamstersafter abrupt shifts in the light-dark cycle. Brain Res. 2000;880:207-211.69. Van Reeth O, Weibel L, Olivares E, Maccari S, Mocaer E, TurekFW. Melatonin or a melatonin agonist corrects age-related changesin circadian response to environmental stimulus. Am J PhysiolRegul Integr Comp Physiol. 2001;280:R1582-R1591.



PERSPECTIVES NOUVELLES DANS LA PHYSIOPATHOLOGIEET LE TRAITEMENT DES TROUBLES AFFECTIFS :

LE RÔLE DE LA MÉLATONINE ET DE LA SÉROTONINE

D epuis plus de 40 ans, la recherche sur la pathogenèse de la dépression etle développement de médicaments adaptés pour un traitement efficacea été dominée par l’hypothèse monoaminergique, qui stipule que la dé-

pression implique un déséquilibre en sérotonine, noradrénaline, et probable-ment en dopamine. Bien que ces neurotransmetteurs monoaminergiques – dontles rôles respectifs sont passés en revue – soient impliqués de façon indiscutable,le déficit en monoamines ne rend compte seulement que d’une partie des faits,et d’autres événements hormis le déséquilibre monoaminergique devraient êtrepris en compte. Il existe un besoin évident de médicaments antidépresseurs plusefficaces, mieux tolérés et d’action plus rapide. Le développement de nouveauxantidépresseurs dont le mécanisme d’action repose sur d’autres bases que l’hy-pothèse monoaminergique est très prometteur. Le fait que de nombreux troublesaffectifs, y compris la dépression, présentent des anomalies des rythmes circa-diens suscite une attention croissante. Cet article souligne l’impact de la méla-tonine et de ses récepteurs sur la dépression et présente un antidépresseur dé-veloppé récemment possédant une efficacité clinique démontrée, l’agomélatine.Ce produit, qui se distingue par des propriétés agonistes de la mélatonine et an-tagonistes des récepteurs 5-HT2C, représente un nouveau concept dans le trai-tement de la dépression.

✦

236 Severe depression: from diagnosis to treatment – Olié MEDICOGRAPHIA, VOL 27, No. 3, 2005

anxiety, pathologic personality traits, alcohol andpsychotropic substance abuse, concomitant phys-ical disease (organic brain disease, in particular),and the persistence of negative environmental fac-tors. Such factors can transform “difficult” or “re-sistant” depression into severe depression by virtueof its intra- and interindividual repercussions.

Age of depression onset may influence mood dis-turbance severity, especially as depressive symp-toms in children are often associated with otherdisturbances, such as attention deficit disorder oranxiety disorder. Adolescent depression typicallypresents as behavioral disturbance, which may notonly delay the diagnosis, but generate consequencesthat can be serious and irreversible.

As for the psychotic symptoms that may accom-pany the depressive episode, they reflect its sever-ity when mood-congruent (eg, delusional ideas ofguilt or ruin); when noncongruent (eg, delusionsof persecution or of telepathy with hallucinations),they herald a poor prognosis and are an indicationfor immediate first-line therapy with an antide-pressant+neuroleptic or electroconvulsive therapy(ECT). Depression associated with delusional hypo-chondriasis is a more complex presentation, with ahigh risk of developing into a chronic delusionaldisorder, especially if compounded by pathologicalpersonality traits.

High-intensity depression

This category is readily identified by symptom num-ber and intensity, producing high scores on theMontgomery-Asberg Depression Rating Scale(MADRS) or Hamilton Depression Rating Scale. Notonly low-intensity, but also high-intensity depres-sion are often observed to respond less well to old-er antidepressants.1 Studies have revealed that drugsindeed differ in this context, with some provingmore effective. Thus, certain psychiatrists remainconvinced that selective serotonin reuptake inhib-itors (SSRIs) are less potent antidepressants thanimipramine. However, this has never been proven.

The concept of severe depression has severalclinical connotations: (i) symptom intensi-ty: sadness, loss of interest, inhibition, and

gloomy thoughts can be evident to variable degrees,particularly if intense or threatening; the level of in-tensity or severity is readily measured using quan-titative scale scores; (ii) threat to life, whether fromthe physical impact of appetite loss, intense anxietyor cognitive disorganization with confusion,or fromthe power of suicidal ideas—both series of factorsdefine the highest emergency level: failure to mounta rapid medical response can have irreversible con-sequences, in particular in the elderly; (iii) severi-ty of the functional impact on daily life, notably onthe work and family level.

Severe depression also has outcome connota-tions. Certain characteristics compound the prog-nosis and/or treatment resistance: concomitant


S evere depression comes in a variety of presentations, from highly acuteintense depression to chronic refractory states complicated by predis-position or comorbidity. Current antidepressants are often ineffective in

depression with somatic or psychiatric comorbidity, psychotic symptoms, orimminent threat to life. Pharmacologic efficacy in severe depression is difficultto demonstrate in children and the elderly for methodological reasons. New an-tidepressants are as much needed as ever, among other reasons to advance ourunderstanding of the etiopathogenesis of depressive disease and the mecha-nism of antidepressant action: an acute increase in synaptic cleft monoaminelevels is a clearly inadequate guide either to understanding depression or tochoosing appropriate drug therapy. Improved tolerability, shortened time toeffect, and efficacy in relieving cognitive and somatic symptoms in emergency(life-threatening) situations are the criteria for evaluating any potential newantidepressant.Medicographia. 2005;27:236-239. (see French abstract on page 239)

Keywords: severe depression; childhood depression; depression in the elderly;antidepressant; treatment

Jean-Pierre OLIÉ, MDHospital-University DepartmentMental Health and TherapeuticsSainte-Anne HospitalParis, FRANCE

Address for correspondence: Prof Jean-Pierre Olié, Service Hospitalo-Universitaire, Santé Mentale et Thérapeutique, Hôpital Sainte-Anne, 7 rue Cabanis, 75684 Paris Cedex 14, France(e-mail: [email protected])

Severe depression: from diagnosis to treatment

b y J . - P. O l i é , F r a n c e

237Severe depression: from diagnosis to treatment – Olié MEDICOGRAPHIA, VOL 27, No. 3, 2005


Depressed patients with personality disorder dif-fer in several ways from those without: earlier-onsetmood disturbance, less social support, more oftensingle or divorced, more hospital admissions, andmore severe depressive symptoms with higher sui-cidality. The personality disorder also compoundstreatment resistance. An antidepressant that wouldnot only remain effective despite such comorbid-ity, but could also treat it, would have a definite ad-vantage.

Depression with psychotic symptoms

Mood-congruent and non–mood-congruent are notcategories cast in stone—they can be influenced bysociocultural factors.9 For some psychiatrists, theconcept of psychotic depression partly overlaps thatof endogenous depression. Should depression beviewed as a continuum, stretching from nondelu-sional depression (Kraepelin’s simple melancholia)to severe delusional depression or delusional melan-cholia? Or should delusional depression be consid-ered a specific category?10

Comparison between delusional and nondelu-sional depression reveals no differences in sex, ageof onset, number of episodes, or number of suicideattempts. However, cluster A personality distur-bances appear more frequent in delusional depres-sion, and cluster B disturbances more frequent inits nondelusional counterpart.11 Some psychiatristsfind delusional depression more frequent in bipo-lar than in unipolar patients.12 Delusions may in-crease the risk of suicide, in both bipolar and unipo-lar depressives.13 The risk of mood disturbance infirst-degree relatives may be higher in delusionalthan in nondelusional depression.14

Delusions and hallucinations reach critical levelsin the most serious forms of depression in whichobnubilation facilitates the emergence of terrifyinginterpretative visions and delusional ideas. Suchdelusional presentations have become rare, proba-bly aborted by psychotropic agents. Thus full-blownCotard syndrome, in which delusional negation oforgans or the entire body is combined with delu-sional ideas of eternal damnation and immortality,and the negation of family, space, and time, nowtends to be replaced by partial presentations, eg,delusions of damnation or individual organ nega-tion. No antidepressant has proved as effective indelusional depression as in its nondelusional coun-terpart.

Psychotic symptoms are significantly more fre-quent in adolescents15 and preadolescents,16 posinga differential diagnostic quandary with incipientschizophrenia and raising awkward managementproblems: older antidepressant alone, newer anti-depressant alone, or either in combination with anantipsychotic?

The concept of schizoaffective disorder introducedby Kasanin in 1933 raised classification problemsthat remain unresolved. In terms of prognosis, de-pressive schizoaffective disorder falls midway be-tween schizophrenia and major (psychotic) depres-sion.17 Here too, treatment is often mixed, combin-ing an antidepressant with an antipsychotic.

Imminently life-threatening depression

In this type of situation, ECT may be necessary, inparticular as malnutrition, dehydration, suicidal in-tent, and repeated attempted suicide all predict afavorable response.2 Endogenicity and melancho-lia, on the other hand, are poorer predictors, in con-trast to long-held belief.3 The best predictor of re-sponse to ECT is psychomotor retardation.4 In anyevent, ECT is mandatory in imminently life-threat-ening depression, in particular because of its rapidonset of effect and the risks of imipramine thera-py. This underlines the need to identify drugs thatcan act rapidly and safely in the (often malnour-ished) elderly.

The problem of time to antidepressant effect isclearly critical in severe depression, in particularwhen there is an imminent threat to life. At pre-sent, various chemotherapy strategies can be usedto lower suicide risk: restoration of the sleep-wakecycle using a hypnotic may help to relieve the de-pression and pressure of suicidal ideas; reductionof anxiety using an anxiolytic can prevent impulsivesuicide (in this context, benzodiazepines, with theirpotential for paradoxical disinhibitory reactions,are best avoided in favor of other types of anxiolyt-ic, eg, levomepromazine or cyamemazine); suici-dality may also respond to drugs such as lithium.

It is generally agreed that the efficacy of an anti-depressant can only be assessed after treatment for4 to 6 weeks. Certain effects may emerge earlier, ofcourse, heralding greater efficacy at 6 weeks.5 Tol-erability becomes crucial where the threat to life isphysical. Posttricyclic antidepressants marked asignificant advance in this regard. Safety—cardiac,hepatic, and neurologic—is mandatory. The newantidepressants are generally much easier to usethan their predecessors, although renutrition andgentle sedation often remain necessary to bridge thegap until the onset of antidepressant effect. Undersuch conditions, the ideal antidepressant is one withrapid onset of effect against anxiety and suicidality,combined with a high level of tolerability.

Depression with psychiatric comorbidity

Clinical depression can be associated with psycho-behavioral symptoms that reverse on effective anti-depressant treatment. They include obsessive symp-toms, eating disorders, irritability, and pathologicalpersonality traits.

Depression may also feature in the natural histo-ry of anxiety disorder or personality disorder. Theclassic example is that of panic disorder complicat-ed by a depressive episode: panic disorder can be di-agnosed in 20% to 30% of patients with depression;as a severity factor, it carries considerable psychoso-cial impact, with a high risk of suicide.6 However,the prognosis may be better if depression precedespanic disorder.7 Depression is often associated withgeneralized anxiety disorder: early-onset anxietydisorder may be a severity factor for anxiety and de-pression.8

238 Severe depression: from diagnosis to treatment – Olié MEDICOGRAPHIA, VOL 27, No. 3, 2005


of dementia, and even—for good measure—cor-rect the biological parameters that tend to declinewith age.

Treatment strategies

No one antidepressant has proved more effectivethan the others, including in severe depression. Cer-tain markers of endogenicity may predict a goodresponse to antidepressants: absence of mood re-activity to the environment, circadian variation,23

psychomotor retardation, anorexia, and weightloss.24 Resistance may simply reflect greater sever-ity.25 Various treatment strategies are available:◆ Increasing the dose of antidepressant (the ad-vantages of this approach were shown with ven-lafaxine26).◆ Using a different antidepressant, including amonoamine oxidase inhibitor (MAOI).27

◆ Combining two drugs with complementary phar-macologic effects, eg, an SSRI with an β-blocker(mianserin, mirtazapine).28

◆ Combining an antidepressant with lithium.29,30

◆ Combining an antidepressant with T3 thyroidhormone.31

◆ Combining an antidepressant with a partial sero-tonin 1A receptor agonist.28

ECT is the last resort, the only therapy that hasproved superior to an antidepressant.32 In severe de-pression the predictors of response to ECT are:◆ Presence of melancholic symptoms.33

◆ Threat to life (malnutrition, high suicide risk).34

◆ Psychotic symptoms.35

◆ Catatonic symptoms.3

◆ Pregnancy or severe postpartum depression.36

◆ Advanced age.37

This highlights the urgent need for new antide-pressant treatments that ideally:◆ Advance our understanding of the mechanismof mood-elevating activity: this requires the devel-opment of drugs with novel neurobiologic actions,given that the mere increase in monoamine levelsin the synaptic cleft is a far from adequate accountof antidepressant effect.◆ Can be tailored to a patient or a type of depres-sion; a current difficulty is our inability to predictwhich antidepressant will actually be best for a par-ticular patient or for a particular phase in thedisease.◆ Restore normal psychological and social func-tion by doing more than achieving symptom remis-sion.

If such drugs could be identified, today’s severedepression could be more easily treated and nodoubt prevented. ❒

Depression inchildren and adolescents

Although midterm outcome is often favorable, theprognosis of depression in the young is often poor,with 75% relapse at 5 years,18 thus considerably in-creasing the risk of depression in adulthood.19 Inaddition, comorbidity is high, with character or be-havioral traits that can interfere with the diagnosisof depression and also its treatment. Hostility bythe child or adolescent generates a vicious circle ofhostile communication that feeds, perpetuates, andeven amplifies the underlying disorder. This maytrigger risk-taking behavior.20 In addition, environ-mental factors can play a major role in facilitatingsuicidal behavior. A history of depressive episode inchildhood is a risk factor for attempted suicide inadulthood.21 There is also the diagnostic quandarymentioned earlier: how to differentiate betweenmood disorder on the one hand and depression inan incipient chronic schizophrenic on the other? Itis still difficult to demonstrate antidepressant effi-cacy in child and adolescent depression.

Depression in the elderly

In the elderly, it is mandatory to test for degenera-tive disease when depression coexists with psychosis,especially if there is no history of depression. Delu-sional hypochondriasis becomes more frequentwith age. Severe cases may present with confusion,or more commonly hallucinations that can be psy-chic, auditory-verbal, visual, or involve bodily sen-sations.

Cognitive and depressive symptoms are frequent-ly intertwined in this age group. Over 25% of de-mentia patients first present with altered mood.22

The term depressive pseudodementia is widely usedto reflect the depression/dementia interface. Thereare five main clinical variants: (i) depression withlow-grade cognitive impairment; (ii) depressive de-mentia; (iii) nondepressive dementia; (iv) demen-tia-induced depression; and (v) comorbidity.

In addition to increased suicide risk in the elder-ly, the two other factors that compound depressionin this age group are cognitive impairment, whichincreases risks such as dehydration, and organicbrain disease, which causes antidepressant resis-tance and heralds dementia. Because concomitantphysical ailments are so common, it is essential toselect an optimally tolerated antidepressant. Thedream drug in such cases is one that would remaineffective despite organic and psychiatric comorbid-ity, relieve cognitive impairment, and arrest the risk

REFERENCES1. Hardy P. [Concept of resistant depression (French)]. Encéphale.1986;12:191-196.2. De Carvalho W, Olie JP. Electroconvulsivothérapie. 97thFrench-speaking Psychiatrists and Neurologists Meeting, Biar-ritz, June 13-18, 1999.3. American Psychiatric Association. The Practice of Electrocon-vulsive Therapy: Recommendations for Treatment, Training, andPrivileging.2nd ed. Washington, DC: American Psychiatric Press,2001.4. Sobin C, Prudic J, Devanand DP, Nobler MS, Sackeim HA. Whoresponds to electroconvulsive therapy? A comparison of effective

and ineffective forms of treatment. Br J Psychiatry.1996;169:322-328.5. Nagayama H, Nagano K, Ikezaki A, Tashiro T. Prediction of ef-ficacy of antidepressant by 1-week test therapy in depression. J Af-fect Disord.1991;23:213-216.6. Coryell W, Endicott J, Winokur G. Anxiety syndromes as epi-phenomena of primary major depression: outcome and familialpsychopathology. Am J Psychiatry. 1992;149:100-107.7. Lydiard RB. Coexisting depression and anxiety: special diag-nostic and treatment issues. J Clin Psychiatry. 1991;52:48-54.8. Shores MM, Glubin T, Cowley DS, Dager SR, Roy-Byrne PP,

239Severe depression: from diagnosis to treatment – Olié MEDICOGRAPHIA, VOL 27, No. 3, 2005


DÉPRESSION SÉVÈRE : DU DIAGNOSTIC AU TRAITEMENT

L a notion de dépression sévère est polysémique, recouvrant aussi bien desdépressions particulièrement aiguës et intenses que des états difficiles àtraiter en raison d’un terrain particulier ou d’une comorbidité. Les dé-

pressions avec comorbidité somatique ou psychiatrique, avec symptômes psy-chotiques, avec risque vital imminent sont souvent peu efficacement traitéesavec les antidépresseurs aujourd’hui disponibles. Chez l’enfant et le sujet âgédes difficultés méthodologiques rendent difficiles la démonstration d’efficacitéd’un agent pharmacologique dans la dépression sévère. Ainsi de nouveaux an-tidépresseurs restent-ils nécessaires. Ils devraient aider à progresser dans laconnaissance étiopathogénique du trouble dépressif et du mécanisme d’actiondes antidépresseurs, l’augmentation aiguë de taux de monoamines dans la fentesynaptique ne pouvant, à l’évidence, suffire à guider notre compréhension dutrouble dépressif et le choix de nos prescriptions. L’amélioration de la tolérance,mais aussi le raccourcissement du délai d’action et l’efficacité dans des contextesd’urgence (risque vital) sur des symptômes tels que les symptômes cognitifs etsomatiques sont des objectifs à la lumière desquels doit être évaluée toute nou-velle molécule potentiellement antidépressive.

✦

Dunner DL. The relationship between anxiety and depression: aclinical comparison of generalized anxiety disorder, dysthymicdisorder, panic disorder, and major depressive disorder. ComprPsychiatry.1992;33:237-244.9. Olie JP, Hardy P, Akiskal HS et al. [Manic-depressive psychosis(French)]. In: Encycl Med Chir-Psychiatrie. Paris, France: Edi-tions Techniques; 4-1990, 37220 A (10):32 pp.10. Chardney DS, Nelson JD. Delusional and non delusional uni-polar depression. Further evidence for distinct subtypes. Am JPsychiatry.1981;138:328-333.11. Serretti A, Lattuada E, Cusin C, Gasperini M, Smeraldi E. Clin-ical and demographic features of psychotic and nonpsychotic de-pression. Compr Psychiatry. 1999;40:358-362.12. Endicott J, Nee J, Andreasen N, Clayton P, Keller M, CoryellW. Bipolar II. Combine or keep separate? J Affect Disord.1985;8:17-28.13. Roose SP, Glassman AH, Walsh BT, Woodring S, Vital-Herne J.Depression, delusions, and suicide. Am J Psychiatry.1983;140:1159-1162.14. Goldstein RB, Horwath E, Wickramaratne PJ, Wolk SI, War-ner V, Weissman MM. Familial aggregation of delusional depres-sion: re-examination in a recent family study. Depress Anxiety.1998;8:160-165.15. Carlson GA, Srober M. Affective disorder in adolescence: issuein misdiagnosis. J Clin Psychiatry.1978;39:59-66.16. Chambers WJ, Puig-Antich J, Tabrizi MA. Psychotic symp-toms in prepubertal major depressive disorders. Arch Gen Psy-chiatry. 1982;39:921-927.17. Coryell W, Lavori P, Endicott J, Keller M, Van Eedewegh M.Outcome in schizoaffective, psychotic and nonpsychotic depres-sion. Course during a six- to 24-month follow-up. Arch Gen Psy-chiatry.1984;41:787-791.18. Kovacs M, Feinberg TL, Crouse-Novak MA. Depressive dis-orders in childhood. I. A longitudinal prospective study of char-acteristics and recovery. Arch Gen Psychiatry.1984;41:229-237.19. Harrington R, Fudge H, Rutter M, Pickles A, Hill J. Adult out-comes of childhood and adolescent depression: II. Links with an-tisocial disorders. J Am Acad Child Adolesc Psychiatry.1991;30:434-439.20. Pesa JA, Cowdery JE, Westerfield RC, Wang M. Self-reporteddepression and risk-taking behaviors among Hispanic adoles-cents. Psychol Rep. 1997;81:235-243.21. Sechter D, Bonin B, Bertschy G, Vandel S, Bizouard P. [Pre-diction of suicide risk (French)]. Encephale.1991;17:361-364.22. Yesavage J. Differential diagnosis between depression anddementia. Am J Med. 1993;94(suppl 5A):23-28.

23. Maier W, Philipp M, Schlegel S, Heuser I, Wiedemann K,Benkert O. Diagnostic determinants of response to treatmentwith tricyclic antidepressants: a polydiagnostic approach. Psychi-atry Res. 1989;30:83-93.24. Joyce PR, Paykel ES. Predictors of drug response in depres-sion. Arch Gen Psychiatry. 1989;46:89-99.25. Kocsis JH. New issues in the prediction of antidepressant re-sponse. Psychopharmacol Bull. 1990;26:49-53.26. Charlier C, Pinto E, Ansseau M, Plomteux G. Venlafaxine: therelationship between dose, plasma concentration and clinical re-sponse in depressive patients. J Psychopharmacol. 2002;16:369-372.27. Thase ME, Kupfer DJ. Recent developments in the pharma-cotherapy of mood disorders. J Consult Clin Psychol.1996;64:646-659.28. Stahl SM. Essential psychopharmacology. Cambridge, UK:Cambridge University Press, 2000.29. De Montigny C, Grunberg F, Mayer A. Lithium induces rapidrelief of depression in tricyclic antidepressant drug nonrespon-ders. Br J Psychiatry. 1981;138:252-256.30. Januel D, Poirier MF, D’alche-Biree, Dib M, Olie JP. Multi-center double-blind randomized parallel-group clinical trial ofefficacy of the combination clomipramine (150 mg/day) plus lithi-um carbonate (750 mg/day) versus clomipramine (150 mg/day)plus placebo in the treatment of unipolar major depression. J Af-fect Disord.2003;76:191-200.31. Hatterer JA, Stanley M. Thyroid function in refractory depres-sion. In: Roose SP, Glassman AH, eds. Treatment Strategies forRefractory Depression. Washington, DC: American PsychiatricPress; 1990:169-192.32. Folkerts HW, Michael N, Tolle R, Schonauer K, Mucke S,Schulze-Monking H. Electroconvulsive therapy vs. paroxetine intreatment-resistant depression: a randomized study. Acta Psy-chiatr Scand. 1997;96:334-342.33. Abrams R, Vedak C. Prediction of electroconvulsive responsein melancholia. Convuls Ther.1991;7:81-84.34. Vanelle JM. [Predictive factors of response to electronarcosis(French)]. Encephale.1991;17:399-404.35. Petrides G, Fink M, Husain MM, et al. ECT remission rates inpsychotic versus nonpsychotic depressed patients: a report fromCORE. J ECT. 2001;17:244-253.36. Rabheru K. The use of electroconvulsive therapy in special pa-tient populations. Can J Psychiatry. 2001;46:710-719.37. Benbow SM. The role of electroconvulsive therapy in the treat-ment of depressive illness in old age. Br J Psychiatry.1989;155:147-152.

240 Shortcomings of current antidepressant therapies – Kennedy MEDICOGRAPHIA, VOL 27, No. 3, 2005

Failure to address compliance at the outset oftreatment and to reinforce its importance through-out treatment contributes to suboptimal outcomes.Relatively inexpensive and simple interventions canmake a difference. Patients who received a brief psy-choeducation program combined with antidepres-sant medication had a significantly higher responserate (74%) compared with those who received “usu-al care” (44%).3 These studies highlight the bene-fits of spending time with patients, exploring theirbeliefs about depression and antidepressants, andin providing simple “messages” about delayed on-set of antidepressant action, expected early and lateside effects, as well as the importance of continuingtreatment beyond response.

◆ Personality dimensionsPersonality variables also significantly influence awide range of treatment-related issues. Differentpersonality dimensions are easily assessed using aself-report measure such as the Eysenck Personali-

This chapter reviews patient variables that in-fluence treatment outcomes in Major Depres-sive Episode (MDE), before focusing on the

current limitations of today’s antidepressant options.

Treatment of depression: patient variables

◆ ComplianceCompliance reflects the positive or negative inter-actions among patient, drug, and prescriber vari-ables and is a significant, but frequently neglected,reason for poor treatment outcome. Results fromprimary care studies indicate that fewer than 50%of depressed patients continue to take their medi-cation after 12 weeks.1,2 The reasons cited in bothstudies are remarkably similar in content and fre-quency (Table I).


S uccessful treatment of a Major Depressive Episode (MDE) is influencedby many factors beyond the properties of a particular medication. Theseinclude the unique characteristics of each patient; the safety–tolerabil-

ity–effectiveness profile of the drug, and the interaction between patient andhealth care professionals. While newer antidepressants represent considerableadvances in safety and tolerability compared with tricyclic (TCA) and mono-amine oxidase inhibitor (MAOI) agents, the improvement in therapeutic bene-fit is not substantial. The delay to achieve antidepressant benefit, the percent-age of patients who do not reach response or remission criteria, the persistenceof unwanted side effects, and concerns about drug discontinuation emergenteffects are all shortcomings of current antidepressants. Medicographia. 2005;27:240-246. (see French abstract on page 246)

Keywords: antidepressant; major depressive disorder; personality; compliance; onset of action; side effect

Sidney H. KENNEDY, MD, FRCPCPsychiatrist-in-Chief University Health Network Professor, Department of PsychiatryUniversity of Toronto Toronto, CANADA

Address for correspondence: Sidney H. Kennedy, MD, FRCPC, University Health Network, 200 Elizabeth St, EN8-222, Toronto, ON M5G 2C4, Canada(e-mail: [email protected])

Shortcomings of current antidepressant therapies

b y S . H . K e n n e d y , C a n a d a


5-HT 5-hydroxytryptamine (serotonin)DESS discontinuation emergent signs and

symptomsEPQ Eysenck Personality QuestionnaireHRSD Hamilton Rating Scale for DepressionMADRS Montgomery-Asberg Depression Rating

ScaleMAOI monoamine oxidase inhibitorMDE Major Depressive Episode NE norepinephrineNEO-PI Neuroticism, Extraversion, and Openness

Personality InventorySNRI serotonin and norepinephrine reuptake

inhibitorSSRI selective serotonin reuptake inhibitorTCA tricyclic antidepressantTCI Temperament and Character Inventory

tween high scores and decreased compliance.13,14

This is well illustrated in the following case vignette,where Mr T’s high level of gregariousness both con-tributed to an early treatment response and earlydrug discontinuation.

◆ Comorbid personality disordersA related question is whether or not patients witha diagnosis of MDE and a comorbid personality dis-order have a worse outcome during antidepres-sant treatment. Using the Diagnostic and StatisticalManual of Mental Disorder, 3rd Edition, Revised

(DSM III-R) designation of cluster A(Paranoid, Schizoid, and Schizotypal),cluster B (Antisocial, Borderline, Histri-onic, and Narcissistic), and cluster C(Avoidant, Dependent, and Obsessive-Compulsive) personality disorder, Favaand colleagues15 examined the predic-tive value of personality disorder comor-bidity in patients who received fluoxe-tine. There were no differences in ratesof response between patients with orwithout a cluster A or C diagnosis, while

the presence of a cluster B diagnosis before treat-ment predicted a positive antidepressant response.Almost a decade later, Mulder and associates16 ex-amined the interaction between antidepressantclassselection (fluoxetine or nortriptyline), personalitydisorder (presence or absence) and drug response.Although the presence of a comorbid personalitydisorder did not have an overall effect on outcome,those patients with a cluster B personality had a sig-nificantly better response to fluoxetine comparedwith nortriptyline.

While others have reported less favorable out-comes for depressed patients with personality dis-orders during continuation therapy,17 the weight ofevidence suggests that such comorbidity does notadversely affect outcome.

ty Questionnaire (EPQ),4 Temperament and Charac-ter Inventory (TCI),5 or the Neuroticism, Extraver-sion, and Openness Personality Inventory (NEO-PI).6

Although few in number, several studies have eval-uated the influence of personality on help-seekingbehavior, symptom presentation, compliance, andultimately treatment outcomes. In most cases, neu-roticism and extraversion account for differencesin depressed patients.

For example, men in an urban setting who havea low level of neuroticism and an external attribu-tional style are less likely to seek help than men withhigher levels of neuroticism.7 With regard to symp-tom presentation, a high level of “harm avoidance”(equivalent to neuroticism) in both men and womenhas been shown to predict higher levels of severity,more physical symptoms including fatigue in de-pressed patients, while high “novelty seeking” (equiv-alent to extraversion) is associated with suicide at-tempts, impaired concentration, and alcohol abuse.8

The influence of neuroticism on side-effect re-porting was examined in a group of nonpsychiatrichealthy volunteers who received moclobemide orplacebo for 3 weeks under double-blind random-ized controlled conditions.9 The strongest predictorof side-effect reporting was baseline neuroticism,and this relationship increased over time in theplacebo group, but declined in the moclobemidegroup, again highlighting the importance of vari-ables beyond pharmacokinetic and pharmacody-namic properties of the drug.

The relationship between neuroticism and treat-ment response has also received considerable atten-tion. In a large primary care trial involving morethan 300 patients who met diagnostic criteria fordysthymia, Katon and colleagues10 reported that ahigh baseline level of neuroticism was associated

with failure to achieve remission, whether treat-ment involved paroxetine or problem-solving psy-chotherapy. Although Mulder11 broadly agreed withthis conclusion in the largest systematic review sofar reported, he also noted that personality pathol-ogy contributed least to outcome prediction in thebetter-designed studies.

Importantly, different personality variables caninfluence different aspects of the treatment processin opposing directions. This is particularly true ofextraversion. High baseline scores on extraversion(particularly the facet identified as gregariousness)predicted greater symptom reduction during an-tidepressant treatment,12 while others, using com-parable measures of extraversion (sensation seekingor novelty seeking), identified an association be-

241Shortcomings of current antidepressant therapies – Kennedy MEDICOGRAPHIA, VOL 27, No. 3, 2005


Mr T worked in the sales department of abrewing company. He was a particularly gre-garious individual. Over a 6-year period, hehad been seen by three different primary carephysicians. On each occasion, he presentedwith typical symptoms of a major depressiveepisode. For him, the most prominent symp-tom was losing interest in his normally activesocial life.

There was also a pattern to his antidepres-sant usage. Typically, he responded very quick-ly, often during the first week. He would re-gain interest in social activities and increasehis success in sales. At about the same time,he would start to miss doses of his antidepres-sant medication for days at a time and even-tually stop altogether.

Only after a detailed discussion about thepotential negative relationship between hisoutgoing personality style and medicationcompliance did Mr T agree to comply with an-tidepressant therapy and he has remained wellfor the past 2 years.

Demyttenaere et al,1 2001 Maddox et al,2 1994

Feeling better 55% Feeling better 35%Adverse events 23% Adverse events 30%Fear of dependence 10% Other reasons 17%Stigma 10% Physician instruction 15%Lack of effect 10% Lack of effect 15%

Table I. Reasons cited for discontinuation during 12 weeks of anti-depressant treatment.

tures. While most would agree that a diagnosis ofMDE with psychotic features warrants differenttreatment interventions, including combination an-tidepressant-antipsychotic medications or an earlyintervention with electroconvulsive therapy,25 thereis less consensus on the relevance of melancholia.In a natural practice setting, where patients com-pleted standard diagnostic and severity measures,melancholic patients had lower response rates toparoxetine, sertraline, and venlafaxine comparedwith nonmelancholic patients (48% vs 61%).26 Inthis study, there was no difference in rate of re-sponse among drugs in the melancholic subgroup.However, there is evidence that TCAs, in particularclomipramine, were more effective than SSRIs intreating melancholic depression (see Amsterdam27

for a review). In subsequent trials, venlafaxine wassuperior to placebo28 and fluoxetine29 in hospitalizedpatients with melancholia. Interestingly, Guelfi andcolleagues30 also reported similar outcomes whenmirtazapine was compared with venlafaxine in hos-pitalized severely depressed patients with melan-cholic features.

There are also differences between outpatient andinpatient groups in their response to TCA and MAOIantidepressants. In a meta-analysis, Thase and col-leagues31 reported superior response rates to MAOIantidepressants compared with TCAs among out-patients, while the converse was true for inpatients.Likewise, a group of Danish investigators, usingclomipramine as the active comparator, consistent-ly reported its superiority over the reversible in-hibitor of MAO-A, moclobemide,32 as well as twoSSRIs—paroxetine33 and citalopram34—in studiesinvolving depressed inpatients. However, a varietyof factors including suicidality, social supports, anddifferences in practice patterns across countries,influences the decision to admit a depressed patientto hospital.

In summary, shortcomings of current antidepres-sant therapies need to be considered in the contextof treatment adherence, distinct patient profiles in-cluding personality differences, and illness severi-ty. One or more of these variables often contributeto poor treatment outcome.

Unmet needs with currenttherapeutic choices

The widespread use of SSRIs to treat patients withdepression in primary care and specialist settingsrepresents a significant advance in managing oneof the most functionally disabling and frequentlycomorbid medical conditions.35,36 With greater safe-ty and tolerability compared with TCA and MAOIagents, compliance with SSRI and newer antide-pressants during acute and maintenance phases hasimproved, and the percentage of depressed patientswho are hospitalized for treatment continues todecline.37 Even so, there are still considerable lim-itations to SSRI and dual action (duloxetine, ven-lafaxine, bupropion, milnacipran, and mirtazapine)antidepressant agents.

The “therapeutic lag” between initiating treat-ment and achieving clinically meaningful effects

◆ Severity of illnessSince there is really no consensus on what consti-tutes “severe depression,” it is difficult to reach aconclusion about its impact on treatment outcome.A predetermined baseline score on the HamiltonRating Scale for Depression (HRSD)18 such as scor-ing 24 or higher at baseline on the 17-item HRSD,or a score above 28 at baseline on the Montgomery-Asberg Depression Rating Scale (MADRS),19 are fre-quently applied criteria for “severe depression.” Itmust be pointed out, however, that different symp-tom weighting on these scales (eg, there are 3 sleepitems on HRSD and 1 sleep item on MADRS) mayresult in some patients being considered “severe”on one scale, but not on another.

Relatively few studies have been designed to ex-amine the relationship between severity of symp-toms and preferential drug response, although itis generally believed that antidepressant treatmentsare less likely to prove superior to placebo in pa-tients with low pretreatment scores.20,21 Clinical im-pression that moclobemide is less effective in severedepression was initially supported in a global re-view of moclobemide trials, in which imipraminewas superior to moclobemide in the group withbaseline HRSD scores of 28 or higher,22 but not con-firmed in a subsequent analysis using the samedrug.23 Greater efficacy in more severely depressedpatients is a desirable characteristic of any antide-pressant. Khan and colleagues24 have also examinedthe relationship between severity of depressivesymptoms and response to a range of selective sero-tonin reuptake inhibitor (SSRI) antidepressants,venlafaxine, nefazodone, and placebo. Pretreatmentseverity was assessed using the 17-item HRSD, andpatients were categorized into 4 groups accordingto baseline scores. The drug-placebo difference frombaseline to end of treatment (effect size) was pro-gressively greater as baseline severity increased.

◆ Melancholic and psychotic featuresOther considerations that likely overlap, but are notsynonymous with severity, include inpatient statusand the presence of psychotic or melancholic fea-



Figure 1. Physician-reported preferences for antidepressants: the most significantdisadvantage of the most frequently prescribed antidepressant. Modified from reference 39: Wade AG. A holistic evaluation of attitudes to depression. Poster presentedat the DURG 16th Annual Meeting; London, UK, 2005. Copyright ©, Drug Utilisation Research Group(UK & Ireland).

Nil

Other

Inconvenient dosing

Poor long-term control

Side effects

Poor predictability of response

Slow onset of action

0 5 10 15 20 25 30 35 40 (%)

Weight gain and sexual dysfunction are probablythe most unacceptable long-term side effects formany patients. Since there are no randomized mul-tiple drug, placebo-controlled trials from which toderive comparative rates of side effects, it is neces-sary to rely on drug-placebo differences in side ef-fect reports as a crude method to compare differentantidepressants.◆ Selective serotonin reuptake inhibitor antide-pressants Greater tolerability and easy once-daily dosing havecontributed to the widespread adoption of SSRIs asfirst-line therapy. There are six SSRIs available totreat depression in most countries: citalopram, es-citalopram, fluoxetine, fluvoxamine, paroxetine, andsertraline. All of these agents block the reuptakeof serotonin, but with differing degrees of selectiv-ity and potency. Gastrointestinal side effects, espe-cially nausea, are common during the first 1 to 2weeks of treatment, but usually remit while othercentral nervous system effects including alterationsto sleep architecture may persist. Relative frequen-cies of side effects are reported in Table II.◆ Dual action antidepressantsVenlafaxine, milnacipran, and duloxetine inhibitboth 5-HT and norepinephrine (NE) reuptake, whilemirtazapine acts on both systems through indirectnoradrenergic mechanisms. The side effects asso-

can still be 3 to 4 weeks.38 Physicians across Europeidentified slow onset of action as the most problem-atic issue in current antidepressant therapy, butalso highlighted side effects and predictability ofresponse as significant shortcomings of currentagents (Figure 1).39 There is no overall increase inthe percentage of responders to SSRIs comparedwith TCAs,40 with a full third of patients in mostclinical trials showing an unsatisfactory responseand less than 50% achieving full symptomatic re-mission.26,41

◆ Therapeutic delayAntidepressants with a faster onset of action wouldoffer significant benefits to patients. A rapid im-provement in sleep would reduce the need for co-prescription of hypnotic medications and may de-crease suicidal ideation in the critical early weeksof treatment. Although serotonin reuptake occursalmost immediately after antidepressant adminis-tration, onset of antidepressant action is often 2 to3 weeks later. This coincides with the 2-to-4 weeklatency for adaptational downregulation of presy-naptic 5-HT1A receptors, which results in a net in-crease in 5-HT release to frontal cortex and limbicareas.42,43 This latency has also been linked to hippo-campal neurogenesis and other molecular changesin gene expression associated with antidepressanttreatment.44

Based on understanding of 5-HT1Aneurophysiology, pindolol, a 5-HT1Aantagonist and β-blocker, has beenadministered with various SSRIs toaccelerate and augment the effect ofSSRIs via blockade of 5-HT1A autore-ceptors. Although results have beeninconsistent, this may be partly ex-plained by differences in patient char-acteristics. An accelerated responsewas seen in previously untreated pa-tients, with a relatively short durationand low severity of depression and fewprior episodes (see review by Segraveand Nathan45). A further explanation ofthe inconsistent results obtained withpindolol relates to adequacy of the usu-ally prescribed dose to bind at 5-HT1Asites46). The extent to which mirtaza-pine and escitalopram may producean accelerated response is a matter forfurther exploration.47,48

◆ Side effectsSide effects with SSRIs and dual ac-tion agents, although different fromthose with TCAs, are nevertheless sub-stantial and frequently lead to prema-ture drug discontinuation. Anticholin-ergic and cardiovascular side effectsare most prevalent with TCAs, whilegastrointestinal and central nervoussystem side effects are more problem-atic with SSRIs. Dual action agents dif-fer in side effect profiles according totheir mechanism of action.



Incidence of side effectsAntidepressant

≥30% ≥10%

Citalopram None Drowsiness, sedation SweatingEscitalopram* Insomnia Tremor

Headache GI† distressAsthenia, fatigue Sexual disturbancesDry mouth

Fluoxetine Sexual disturbances Drowsiness, sedation Dry mouthInsomnia TremorDisorientation/confusion Orthostatic hypotension/Headache dizzinessAsthenia, fatigue GI distress

Fluvoxamine GI distress Drowsiness, sedation Dry mouthSexual disturbances Insomnia Constipation

Excitement, hypomania SweatingHeadache TremorAsthenia, fatigue

Paroxetine Sexual disturbances Drowsiness, sedation SweatingInsomnia TremorHeadache Orthostatic hypotension/Asthenia, fatigue dizzinessDry mouth GI distressConstipation

Sertraline GI distress Drowsiness, sedation Dry mouthSexual disturbances Insomnia Tremor

Excitement, hypomania Orthostatic hypotensionHeadache /dizziness

* Escitalopram is the stereoisomer of citalopram—comparable side effects to citalopram have been reported. † GI, gastrointestinal.

Table II. Frequently reported side effects across selective serotonin reuptake inhibitors (SSRIs). Modified from reference 53: Kennedy SH, Lam RW, Cohen NL, et al. Clinical guidelines for the treatment of depressivedisorders: IV medications and other biological treatments. Can J Psychiatry. 2001;46(suppl 1):38S-58S. Copyright ©2001, The Canadian Psychiatric Association.

When structured questionnaires areused to evaluate sexual dysfunction,30% to 50% of patients who are treat-ed with SSRIs or venlafaxine (compa-rable duloxetine data are not yet avail-able) report impairment of desire andorgasm on SSRIs, while bupropion,mirtazapine, and moclobemide causesignificantly less dysfunction (TableIV).49-51 A potential advantage of thetendency for SSRIs to delay ejacula-tion has been demonstrated in thetreatment of premature ejaculation.52

The development of future antidepres-sants with favorable effects on sexualfunction would be valued by depressedpatients and their partners.

◆ Weight gain across antidepressantclassesUntil recently, weight change receivedminimal attention as an outcome mea-sure in antidepressant studies, althoughthe tendency for TCA and MAOI agentsto cause substantial weight gain hasbeen recognized for many years. It isestimated that treatment with someTCAs resulted in a weight gain of ap-proximately 1 kg per month (reviewedin Kennedy et al53).

While the SSRIs are frequently as-sociated with modest reductions inweight during the acute phase of treat-ment, there is some evidence thatweight gain may occur during main-tenance treatment with some, but not

all, SSRIs.54 Moclobemide and bupropion SR do notappear to be associated with weight gain duringmaintenance treatment and for some patients maybe associated with weight reduction. Short-term tri-als suggest that venlafaxine is weight-neutral; how-ever, there is an absence of controlled data on long-term weight change.55

Among the currently available SSRI and dual ac-tion agents, mirtazapine is most likely to be associ-ated with weight gain during acute treatment. Thisreflects both histaminergic and 5-HT2C receptorantagonist properties. During maintenance treat-ment with mirtazapine, patients who have not ex-perienced weight gain in the acute phase are un-likely to start gaining weight, and those patients

ciated with duloxetine, milnacipran, and venlafax-ine are similar to those reported with SSRIs, whileweight gain and sedation are the commonest sideeffects associated with mirtazapine. Nefazodone hasweak noradrenergic effects as well as 5-HT2 and5-HT transporter blockade. It has been withdrawnfrom use in many countries due to rare reports ofliver toxicity. Trazodone is rarely used as an an-tidepressant, but also belongs in this group and isoften prescribed as a hypnotic in combination withanother antidepressant. Bupropion probably hasdual uptake blocking actions on the dopaminergicand noradrenergic systems. It may cause insomniaand overarousal. Other side effects are reported inTable III.

◆ Sexual dysfunction across antidepressantclassesDrug-induced sexual dysfunction appears to be morecommon in men than women, although levels ofsexual dysfunction are higherin untreated depressedwomen than in untreated depressed men.49 Preva-lence rates are lower when studies rely on sponta-neous self-report and are higher when specific ques-tions are asked about sexual desire, arousal, orgasm,and overall satisfaction. Rates of sexual dysfunctionwere as high or higher with TCA and MAOI agents,although this issue was usually minimized in thepresence of multiple side effects.



Incidence of side effectsAntidepressant

≥30% ≥10%

Bupropion None Insomnia ConstipationExcitement, hypomania SweatingHeadache TremorDry mouth GI* distressBlurred vision

Duloxetine GI distress Drowsiness, sedation ConstipationInsomnia Orthostatic hypotension/Dry mouth dizziness

Mirtazapine Drowsiness, sedation Asthenia, fatigue ConstipationDry mouth Blurred visionWeight gain (over 6 kg)

Nefazodone None Drowsiness, sedation Blurred visionDisorientation/confusion ConstipationHeadache Orthostatic hypotension/Asthenia, fatigue dizzinessDry mouth GI distress

Trazodone Drowsiness, sedation Asthenia, fatigue Orthostatic hypotension/Dry mouth dizziness

GI distress

Venlafaxine GI distress Drowsiness, sedation Dry mouthSexual disturbances Insomnia Constipation

Excitement, hypomania SweatingHeadache Orthostatic hypotension/Asthenia, fatigue dizziness

* GI, gastrointestinal.

Table III. Frequently reported side effects across dual action agents. Modified from reference 53: Kennedy SH, Lam RW, Cohen NL, et al. Clinical guidelines for the treatment of depressivedisorders: IV medications and other biological treatments. Can J Psychiatry. 2001;46(suppl 1):38S-58S. Copyright © 2001,The Canadian Psychiatric Association.

Incidence of sexual dysfunction

<10% 10%-30% >30%

Bupropion Citalopram FluoxetineMirtazapine Duloxetine FluvoxamineMoclobemide Venlafaxine ParoxetineNefazodone Sertraline

Table IV. Frequency of sexual dysfunction during an-tidepressant treatment. Modified from reference 61: Kennedy SH, Lam RW, Nutt DJ, ThaseME. Treating Depression Effectively: Applying Clinical Guidelines.London, UK: Martin Dunitz; 2004. Copyright 2004, Martin Dunitz.

agents have been observed most often when parox-etine or venlafaxine are stopped abruptly.57,58

In a double-blind, treatment interruption trial,patients who abruptly discontinued paroxetine andsertraline experienced significantly more discon-tinuation emergent signs and symptoms (DESS)compared with those who discontinued fluoxetine.57

The authors attributed differences in pharmacoki-netic profiles among SSRIs for the different rates ofDESS. Similarly, high rates of DESS have been re-ported with venlafaxine compared with escitalo-pram 1 week after discontinuation.58 The effects ofabrupt interruption of agomelatine, a new antide-pressant with a unique melatonin agonist profile,were compared with interruption of paroxetine. De-spite its short half-life of 2 hours or less,59 patientswho discontinued agomelatine experienced signif-icantly fewer DESS compared with those who dis-continued paroxetine.60

Conclusion

Advances in neurogenetic and neuroimaging tech-niques have contributed to understanding the com-plex etiopathology of depression. Future challengesin drug development involve better matches be-tween patient profiles and therapeutic targets, withan increased attention to drug tolerability and long-term safety. ❒

who initially gained weight are likely to maintaina plateau after the first few months of treatment.Nevertheless, for many depressed patients, the pros-pect of substantial weight gain during antidepres-sant treatment represents an unacceptable healthburden.

◆ Antidepressant discontinuation emergent effectsSince the mechanisms of action of reuptake inhib-itor antidepressants (SSRIs, SNRIs [serotonin andnorepinephrine reuptake inhibitors], and others) in-volve adaptive downregulation early in the treat-ment process, it is not surprising that drug discon-tinuation, particularly abrupt discontinuation,would result in unpleasant rebound effects. In gen-eral, the shorter the half-life of the drug, the greaterthe likelihood of discontinuation emergent symp-toms. Such occurrences are not limited to SSRIs ordual action antidepressants. Cholinergic reboundwas often observed following rapid discontinuationof amitriptyline or other TCAs with significant an-ticholinergic effects. Patients reported urinary fre-quency, headaches, hypersalivation, and diarrhea.Benzodiazepine withdrawal phenomena have alsobeen recognized for many decades and includesymptoms of insomnia, nausea, appetite loss, sweat-ing, and dysphoria.56 Emergent symptoms associat-ed with drug discontinuation of SSRIs or dual action



REFERENCES1. Demyttenaere K, Enzlin P, Dewe W, et al. Compliance with an-tidepressants in a primary care setting, 1: Beyond lack of efficacyand adverse events. J Clin Psychiatry. 2001;62(suppl 22):30-33.2. Maddox JC, Levi M, Thompson C. The compliance with anti-depressants in general practice. J Psychopharmacol.1994;8:48-53.3. Katon W, Von Korff M, Lin E, et al. Collaborative managementto achieve treatment guidelines. Impact on depression in primarycare. JAMA.1995;273:1026-1031.4. Eysenck SB, Eysenck HJ. An improved short questionnaire forthe measurement of extraversion and neuroticism. Life Sci.1964;305:1103-1109.5. Cloninger CR, Bayon C, Svrakic DM. Measurement of temper-ament and character in mood disorders: a model of fundamentalstates as personality types. J Affect Disord.1998;51:21-32. 6. Costa PT Jr, McCrae RR. Manual for the NEO Personality In-ventory. Odessa, Fla: Psychological Assessment Resources; 1985.7. Moller-Leimkuhler AM: Barriers to help-seeking by men: a re-view of sociocultural and clinical literature with particular ref-erence to depression. J Affect Disord. 2002;71:1-9.8. Grucza RA, Przybeck TR, Spitznagel EL, et al. Personality anddepressive symptoms: a multi-dimensional analysis. J Affect Dis-ord. 2003;74:123-130.9. Davis C, Ralevski E, Kennedy SH, et al. The role of personal-ity factors in the reporting of side effect complaints to moclobe-mide and placebo: a study of healthy male and female volunteers.J Clin Psychopharmacol. 1995;15:347-352.10. Katon W, Russo J, Frank E, et al. Predictors of nonresponse totreatment in primary care patients with dysthymia. Gen HospPsychiatry. 2002;24:20-27.11. Mulder RT. Personality pathology and treatment outcome inmajor depression: a review. Am J Psychiatry. 2002;159:359-371.12. Bagby RM, Joffe RT, Parker JDA, et al. Major depression andthe five-factor model of personality. J Personal Disord.1995;9:224-234.13. Cohen NL, Ross EC, Bagby RM, et al. The 5-factor model ofpersonality and antidepressant medication compliance. Can JPsychiatry. 2004;49:106-113.14. Ekselius L, Bengtsson F, von Knorring L. Non-compliancewith pharmacotherapy of depression is associated with a sensationseeking personality. Int Clin Psychopharmacol. 2000;15:273-278.15. Fava M, Bouffides E, Pava JA, et al. Personality disorder co-morbidity with major depression and response to fluoxetine treat-ment. Psychother Psychosom. 1994;62:160-167. 16. Mulder RT, Joyce PR, Luty SE. The relationship of person-

ality disorders to treatment outcome in depressed outpatients.J Clin Psychiatry. 2003;64:259-264.17. Peselow ED, Fieve RR, Difiglia C. Personality traits and re-sponse to desipramine. J Affect Disord. 1992;24:209-216.18. Hamilton MA. A rating scale for depression. J Neurol Neuro-surg Psychiatry. 1960;23:56-62.19. Montgomery SA, Åsberg M. A new depression scale designedto be sensitive to change. Br J Psychiatry. 1979;134:382-389.20. Paykel ES, Hollyman JA, Freeling P, Sedgwick P. Predictorsof therapeutic benefit from amitriptyline in mild depression: ageneral practice placebo-controlled trial. J Affect Disord.1988;14:83-95.21. Elkin I, Shea MT, Watkins JT, et al. National Institute of Men-tal Health Treatment of Depression Collaborative Research Pro-gram. General effectiveness of treatments. Arch Gen Psychiatry.1989;46:971-982.22. Angst J, Stabl M. Efficacy of moclobemide in different patientgroups: a meta-analysis of studies. Psychopharmacology (Berl).1992;106(suppl):S109-S113.23. Angst J, Amrein R, Stabl M. Moclobemide and tricyclic anti-depressants in severe depression: meta-analysis and prospectivestudies. J Clin Psychopharmacol. 1995;15(suppl 2):16S-23S.24. Khan A, Brodhead AE, Kolts RL, Brown WA. Severity of de-pressive symptoms and response to antidepressants and placeboin antidepressant trials. J Psychiatr Res. 2005;39:145-150.25. Schatzberg AF. New approaches to managing psychotic de-pression. J Clin Psychiatry. 2003;64(suppl 1):19-23.26. Kennedy SH, Eisfeld BS, Meyer JH, Bagby RM. Antidepres-sants in clinical practice: limitations of assessment methods anddrug response. Hum Psychopharmacol. 2001;16:105-114.27. Amsterdam JD. Selective serotonin reuptake inhibitor effica-cy in severe and melancholic depression. J Psychopharmacol.1998;12(3 suppl B):S99-S111.28. Guelfi JD, White C, Hackett D, et al. Effectiveness of venlafax-ine in patients hospitalized for major depression and melancho-lia. J Clin Psychiatry. 1995;56:450-458.29. Clerc GE, Ruimy P, Verdeau-Palles J. A double-blind compar-ison of venlafaxine and fluoxetine in patients hospitalized for ma-jor depression and melancholia. The Venlafaxine French Inpa-tient Study Group. Int Clin Psychopharmacol.1994;9:139-143.30. Guelfi JD, Ansseau M, Timmerman L, et al. Mirtazapine ver-sus venlafaxine in hospitalized severely depressed patients withmelancholic features. J Clin Psychopharmacol.2001;21:425-431.31. Thase ME, Trivedi MH, Rush AJ. MAOIs in the contemporary

49. Kennedy SH, Eisfeld BS, Dickens SE, et al. Antidepressant-induced sexual dysfunction during treatment with moclobemide,paroxetine, sertraline, and venlafaxine. J Clin Psychiatry.2000;61:276-281.50. Baldwin DS. Sexual dysfunction associated with antidepres-sant drugs. Expert Opin Drug Safety. 2004;3:457-470.51. Clayton AH, Pradko JF, Croft HA, et al. Prevalence of sexualdysfunction among newer antidepressants. J Clin Psychiatry.2002;63:357-366.52. Waldinger MD, Olivier B. Utility of selective serotonin re-uptake inhibitors in premature ejaculation. Curr Opin InvestigDrugs. 2004;5:743-747. 53. Kennedy SH, Lam RW, Cohen NL, et al. Clinical guidelines forthe treatment of depressive disorders: IV medications and otherbiological treatments. Can J Psychiatry. 2001;46(suppl 1):38S-58S.54. Fava M, Judge R, Hoog SL, et al. Fluoxetine versus sertralineand paroxetine in major depressive disorder: changes in weightwith long-term treatment. J Clin Psychiatry. 2000;61:863-867.55. Sussman N, Ginsberg DL. Weight effects of nefazodone, bu-propion, mirtazapine, and venlafaxine: a review of available evi-dence. Primary Psychiatry. 2000;7:33-48.56. Hallstrom C, Lader M. Benzodiazepine withdrawal phenom-ena. Int Pharmacopsychiatry. 1981;16:235-244.57. Rosenbaum JF, Fava M, Hoog SL, et al. Selective serotonin re-uptake inhibitor discontinuation syndrome: a randomized clin-ical trial. Biol Psychiatry. 1998;44:77-87.58. Montgomery SA, Huusom AKT, Bothmer J. A randomizedstudy comparing escitalopram with venlafaxine XR in primarycare patients with major depressive disorder. Neuropsychobiol-ogy. 2004;50:57-64.59. Lôo H, Hale A, D’haenen H. Determination of the dose ofagomelatine, a melatoninergic agonist and selective 5-HT2C an-tagonist, in the treatment of major depressive disorder: a place-bo-controlled dose range study. Int Clin Psychopharmacol.2002;17:239-247.60. Montgomery SA, Kennedy SH, Burrows GD, Lejoyeux M,Hindmarch I. Absence of discontinuation symptoms with agome-latine and occurrence of discontinuation symptoms with parox-etine: a randomized, double-blind, placebo-controlled discontin-uation study. Int Clin Psychopharmacol. 2004;19:271-280.61. Kennedy SH, Lam RW, Nutt DJ, Thase ME. Treating Depres-sion Effectively: Applying Clinical Guidelines. London, UK: Mar-tin Dunitz; 2004.

treatment of depression. Neuropsychopharmacology. 1995;12:185-219.32. Danish University Antidepressant Group (DUAG). Moclobe-mide: a reversible MAO-A-inhibitor showing weaker antidepres-sant effect than clomipramine in a controlled multicenter study.J Affect Disord.1993;28:105-116. 33. Danish University Antidepressant Group (DUAG). Paroxetine:a selective serotonin reuptake inhibitor showing better tolerance,but weaker antidepressant effect than clomipramine in a con-trolled multicenter study. J Affect Disord.1990;18:289-299.34. Danish University Antidepressant Group (DUAG). Citalopram:clinical effect profile in comparison with clomipramine. A con-trolled multicenter study. Psychopharmacology (Berl).1986;90:131-138.35. Murray CJ, Lopez AD. Alternative projections of mortality anddisability by cause 1990-2020: Global Burden of Disease Study.Lancet.1997;349:1498-1504.36. Kessler RC, Berglund P, Demler O, et al. The epidemiology ofmajor depressive disorder: results from the National Comorbid-ity Survey Replication (NCS-R). JAMA. 2003;289:3095-3105. 37. Montgomery SA, Kasper S. Side effects, dropouts from treat-ment and cost consequences. Int Clin Psychopharmacol.1998;13(suppl 2):S1-S5.38. Stahl SM, Nierenberg AA, Gorman JM. Evidence of early on-set of antidepressant effect in randomized controlled trials. J ClinPsychiatry. 2001;62(suppl 4):17-23.39. Wade AG. A holistic evaluation of attitudes to depression.Poster presented at the DURG 16th Annual Meeting; London, UK,2005. 40. Anderson IM. Selective serotonin reuptake inhibitors versustricyclic antidepressants: a meta-analysis of efficacy and tolera-bility. J Affect Disord. 2000;58:19-36.41. Thase ME, Entsuah AR, Rudolph RL. Remission rates duringtreatment with venlafaxine or selective serotonin reuptake in-hibitors. Br J Psychiatry. 2001;178:234-241.42. Artigas F, Perez V, Alvarez E. Pindolol induces a rapid im-provement of depressed patients treated with serotonin reuptakeinhibitors. Arch Gen Psychiatry. 1994;51:248-251.43. Blier P, de Montigny C. Current advances and trends in thetreatment of depression. Trends Pharmacol Sci.1994;15:220-226.44. Duman RS, Malberg J, Nakagawa S, D’Sa C. Neuronal plas-ticity and survival in mood disorders. Biol Psychiatry. 2000;48:732-739.45. Segrave R, Nathan PJ. Pindolol augmentation of selectiveserotonin reuptake inhibitors: Accounting for the variability of re-sults of placebo-controlled double-blind studies in patients withmajor depression. Hum Psychopharmacol.2005, Jan 12. (Epubahead of print).46. Rabiner EA, Bhagwagar Z, Gunn RN, et al. Pindolol augmen-tation of selective serotonin reuptake inhibitors: PET evidencethat the dose used in clinical trials is too low. Am J Psychiatry.2001;158:2080-2082. 47. Behnke K, Sogaard J, Martin S, et al. Mirtazapine orally dis-integrating tablet versus sertraline: a prospective onset of actionstudy. J Clin Psychopharmacol. 2003;23:358-364.48. Sanchez C, Bogeso KP, Ebert B, Reines EH, Braestrup C. Es-citalopram versus citalopram: the surprising role of the R-enan-tiomer. Psychopharmacology (Berl). 2004;174:163-176.



LIMITES DES TRAITEMENTS ANTIDÉPRESSEURS ACTUELS

L’ efficacité du traitement d’un Épisode Dépressif Majeur (Major Depres-sive Episode, MDE) est soumise à de nombreuses influences au-delà despropriétés intrinsèques de tel ou tel médicament. Ces influences com-

prennent les caractéristiques individuelles propres à chaque patient, le profilsécurité d’emploi/tolérance/efficacité du médicament, ainsi que l’interactionentre les patients et les soignants. Alors que les antidépresseurs récents ontpermis des avancées considérables en termes de sécurité d’emploi et de tolé-rance par rapport aux antidépresseurs tricycliques et aux inhibiteurs de la mo-noamine oxydase (IMAO), l’amélioration du bénéfice thérapeutique n’est pasprobante. Le délai pour l’obtention du bénéfice antidépresseur, le pourcentagede patients non-répondeurs ou de ceux qui ne présentent pas de rémission, lapersistance d’effets indésirables, et l’apparition d’effets émergents lors de l’ar-rêt du traitement soulignent les limites des antidépresseurs actuels.

✦

tients in spring and in fall. However, this finding wasnot much taken notice of until bright light thera-py (BLT) was proposed as a possible treatment forwinter depression by Lewy et al,3 followed by thefirst systematic study on seasonal affective disorder(SAD) by Rosenthal et al.4 Scientific research hasbeen encouraged ever since and has led to the pub-lication of more than one thousand studies on SADand BLT, not only because SAD is clinically impor-tant, but also because it represents an importantchronobiological paradigm.

Epidemiology

Numerous epidemiological studies have investi-gated the prevalence of SAD and of its subsyndro-mal form (s-SAD) in the general population (Fig-ure 1, page 248).2,5-21 In temperate climates, about2% to 5% of subjects seem to suffer from SAD, andthe rate of individuals with minor problems dur-ing the fall-winter period indicating s-SAD is evenhigher. A number of reports have documented anincrease in SAD prevalence with northerly lati-

Physiological functioning of life on earth hasover eons been influenced by the rising andsettingof the sun.Likewise, it has been known

since ancient times that seasonal changes in theenvironment affect humans’ physical and mentalhealth: Hippocrates of Kos (460-377 BC) and Are-taeus the Cappadocian (about 150 AD) proposed thehealing powers of sunlight for a range of somaticillnesses, but also for the state of melancholia.1 Themajority of humans seem to experience seasonalvariations of psychopathological symptoms to someextent.2 These seasonal changes are of clinical inter-est if their intensity leads to subjective sufferingand a reduction in psychosocial functioning. EmilKraepelin (1856-1926) was one of the first modernpsychiatrists to become aware of a rise in the inci-dence of affective episodes in manic-depressive pa-

247Seasonal affective disorder: from diagnosis to treatment – Winkler and Kasper MEDICOGRAPHIA, VOL 27, No. 3, 2005


S easonal affective disorder (SAD, fall-winter depression) is a subtype ofmajor depressive or bipolar disorder, and has been a subject of psychi-atric research for over two decades. The prevalence in the general pop-

ulation ranges between 2% and 5% in temperate climates, females are muchmore often afflicted by the syndrome. Atypical depressive symptoms with a re-versed vegetative symptomatology predominate during the depressive episodesin the darker time of the year, while about one quarter of patients experiencehypomanic (seldom manic) episodes during the successive spring/summer pe-riod. Bright light therapy (BLT) is generally accepted for depression in SAD, butpsychopharmacological treatment with antidepressants has likewise been foundto be effective and well tolerated. This article reviews the most important patho-physiological concepts and gives an overview of current diagnostic proceduresand treatment options for SAD.Medicographia. 2005;27:247-253. (see French abstract on page 253)

Keywords: seasonal affective disorder (SAD); winter depression; light therapy;antidepressant

Address for correspondence: Prof Siegfried Kasper, Department of General Psychiatry, Medical University of Vienna, Währinger Gürtel 18-20, A-1090 Wien, Austria(e-mail: [email protected])

Seasonal affective disorder:from diagnosis to treatment b y D . W i n k l e r a n d S . K a s p e r , A u s t r i a


5-HTTLPR serotonin transporter promoter repeat length polymorphism

BLT bright light therapyGSS Global Seasonality ScoreHDRS Hamilton Depression Rating ScalePMDD premenstrual dysphoric disorderSAD seasonal affective disorderSIGH-SAD Structured Interview Guide for the

HDRS SAD versionSPAQ Seasonal Pattern Assessment

Questionnaire S-SAD subsyndromal seasonal affective

disorder

Siegfried KASPER, MD

Dietmar WINKLER, MD

O Univ Prof Dr Dr hcDepartment of General Psychiatry Medical University of Vienna AUSTRIA

▲▲ ▲

Clinical picture inseasonal affective disorder

The prevalence of SAD is higher in female than inmale patients.24 In most clinical samples, womenoutnumber men by about 3.5:1 to 9:1,25 which is byfar higher than the 2:1 ratio commonly found innonseasonal depression.26 Psychiatric classificationdistinguishes between a winter type of SAD (fall-winter depression), and summer SAD,27,28 which israre and will not be further discussed in this report.According to most clinical and epidemiological stud-ies, the majority of patients experience their firstaffective episode in their late twenties or early thir-ties. The symptom pattern found in a clinical sam-ple of 610 SAD patients is shown in Figure 2.29 Sea-sonal depressive episodes share the core symptomswith nonseasonal major depression. Depressed moodand reduced drive (subjectively experienced as lossof energy) are reported by almost all patients. How-ever, most SAD patients display a reversed vegeta-tive symptom pattern and fulfill the criteria of theatypical feature specifier of DSM-IV-TR.30 Symptomsof atypical depression include hypersomnia, day-time fatigue, heaviness in the limbs (termed “lead-en paralysis”), increased appetite, changed eatinghabits (mostly cravings for carbohydrates or fattyfoods), weight gain, and interpersonal rejection sen-sitivity. In contrast, melancholic depression, whichis more often found in patients with nonseasonaldepression, is characterized by markedly depressedmood, complete loss of pleasure, reduced appetite,and, consecutively, loss of weight, insomnia (diffi-culty falling asleep, waking up several times duringthe night and in the early morning hours), feelingsof guilt, and psychomotor agitation or retardation.About three quarters of patients report increased ir-ritability,29 and many SAD patients suffer from anger

tude.22 In fact, the relationship between prevalenceand latitude seems to be more established in NorthAmerica.7 The overall prevalence of SAD is consis-tently higher in North America compared with Eu-rope. In contrast, prevalence rates have been foundto be unexpectedly low in Scandinavia, a fact thatmight be explained by a natural selection of the pop-ulation toward increased tolerance of winter dark-ness.11

SAD is quite frequent in psychiatric populations:Kasper and Kamo23 examined the rate of seasonaldepression in a sample of inpatients suffering frommajor depression. Their survey showed that 11.4%of patients suffered from a fall-winter pattern of feel-ing worse similar to that of patients with SAD.

248 Seasonal affective disorder: from diagnosis to treatment – Winkler and KasperMEDICOGRAPHIA, VOL 27, No. 3, 2005


Figure 1. Point prevalenceof seasonal affective disorder

(SAD) (dark grey bars) and subsyndromal seasonal

affective disorder (s-SAD)(light grey bars) in the gen-

eral population measured indifferent countries at differ-ent latitudes.2,5-21 Eagles et

al13 and Ozaki et al18 haveinvestigated specific popula-tions. The upper half of the

figure shows studies fromNorth America,2,5-8 the lower

half below the dotted lineshows mainly studies from

Europe,9-17 but also fromAsia18,21 and Australia.19,20

USA 64% [5]

USA 42%-50% [6]

USA 42.5% [7]

USA 40% [7]

USA 39% [7]

USA 39% [2]

USA 30%-49% [8]

USA 27% [7]

Finland 60%-70% [9]

Finland 60%-70% [10]

Iceland 62%-67% [11]

Sweden 55%-58% [10]

Sweden 54%-68% [12]

UK 57% [13]

Netherlands 53% [14]

Switzerland 47% [15]

Switzerland 47% [16]

Italy 39%-45% [17]

Japan 32%-42% [18]

Australia 10%-40% [19]

Australia 19% [20]

Philippines 15% [21]

0% 5% 10% 15% 20%

Figure 2. Distribution of symptoms in a cohort of 610 seasonal affective disorder (SAD) patientsfrom Austria and Ger-many. Data for generationof this bar chart are fromWinkler et al.29

Carbo

hydr

ate cr

aving

Increa

sed ap

petite

Redu

ced a

ppeti

te

Loss

of en

ergy

Depres

sed m

ood

Irrita

bility

Anxiet

y, inn

er ten

sion

Daytim

e fati

gue

Hypers

omnia

Loss

of lib

ido

Difficu

lties a

t work

100%

80%

60%

40%

20%

0%

difficulties during fall and winter. The diagnosticcriteria for s-SAD are presented in Table III.35 To ex-amine the severity of seasonal depression, a mod-ified version of the Hamilton Depression RatingScale (HDRS) has been recommended: the Struc-tured Interview Guide for the HDRS–SAD version(SIGH-SAD), which contains 29 items and coversnot only melancholic symptomatology of depres-sion, but also contains 8 items for atypical depres-

attacks, consisting of intense spells of inappropri-ate anger or rage.31 Furthermore, nearly half of allfemale SAD patients suffer from premenstrual dys-phoric disorder (PMDD).32 The severity of seasonaldepressive episodes ranges mostly between mild andmoderate. However, severe depression, occasional-ly with suicidal ideations requiring inpatient treat-ment, is possible. About one quarter of SAD patientssuffer from bipolar affective disorder with hypoman-ic or (rather seldom) manic episodes during springor summer in addition to depressive symptoms dur-ing the fall-winter period.

Diagnostic procedure

As SAD is either a form of recurrent major depres-sive or bipolar disorder with onset and remissionof the affective episodes at a specific time of the year,patients have to fulfill the diagnostic criteria of thesedisorders (see Table I 30 for the diagnostic criteriafor major depression). According to the Diagnosticand Statistical Manual of Mental Disorders, FourthEdition, Revised (DSM-IV-TR), it is possible to ap-ply a seasonal pattern specifier to diagnose SAD(Table II).30

The Global Seasonality Score (GSS), which iscalculated from the Seasonal Pattern AssessmentQuestionnaire (SPAQ),33,34 has frequently been usedto assess the degree of seasonal change in psycho-pathology in the past. According to the Rosenthalcriteria, patients have to obtain a GSS of at least 10to qualify for SAD. Furthermore, the overall degreeof impairment during the worst time of the yearhas to be at least “moderate” to establish the diag-nosis. This is the most important difference betweenSAD and s-SAD.35,36 Patients with the subsyndromalform report differences in several psychopatholog-ical items, however they experience no or only mild



◆ A. At least five of the following symptoms have been present during the same two-week period, nearly every day, and represent a change from previous functioning. At least one of the symptoms is either (1) depressed mood or (2) loss of interest or pleasure. NOTE: Do not include symptoms that are clearly due to a general medicalcondition, or mood-incongruent delusions or hallucinations

(1) Depressed mood (or alternatively can be irritable mood in children and adolescents)

(2) Markedly diminished interest or pleasure in all, or almost all, activities(3) Significant weight loss when not dieting or weight gain or decrease

or increase in appetite(4) Insomnia or hypersomnia (5) Psychomotor agitation or retardation(6) Fatigue or loss of energy(7) Feelings of worthlessness or excessive or inappropriate guilt (8) Diminished ability to think or concentrate, or indecisiveness (9) Recurrent thoughts of death (not just fear of dying), recurrent suicidal

ideation without a specific plan, or a suicide attempt or a specific plan for committing suicide

◆ B. The symptoms are not better accounted for by a mood disorder due to a general medical condition, a substance-induced mood disorder, or bereave-ment (normal reaction to the death of a loved one)

◆ C. The symptoms are not better accounted for by a psychotic disorder like schizoaffective disorder

Table I. Diagnostic criteria for a major depressive episodeaccording to DSM-IV-TR.30

Reproduced from reference 30: American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edition,text revision (DSM-IV-TR). Washington, DC: American PsychiatricPress; 2000. Copyright © 2000, American Pychiatric Press.

◆ A. Regular temporal relationship between the onset of major depressive episodes and a particular time of the year (unrelated to ob-vious season-related psychosocial stressors)

◆ B. Full remissions (or a change from depressionto mania or hypomania) also occur at a characteristic time of the year

◆ C. Two major depressive episodes meeting crite-ria A and B in last 2 years and no nonsea-sonal episodes in the same period

◆ D. Seasonal major depressive episodes substan-tially outnumber the nonseasonal episodesover the individual’s lifetime

Table II. Diagnostic criteria of the seasonal patternspecifier according to DSM-IV-TR.30

Reproduced from reference 30: American Psychiatric Association.Diagnostic and Statistical Manual of Mental Disorders, 4th edition,

text revision (DSM-IV-TR). Washington, DC: American Psychi-atric Press; 2000. Copyright © 2000, American Pychiatric Press.

◆ 1. A history of some difficulty during the winter months that had occurred on a regular basis (at least two consecutive winters)and had lasted for a sustained period of time (at least 4 weeks).Examples of these difficulties are decreased energy, decreasedefficiency at work (eg, concentration, completing tasks), de-creased creativity or interest in socializing, and change in eating habits (eg, eating more carbohydrates), weight (gainingweight), or sleep patterns (more sleep)

◆ 2. Subjects have to regard themselves as normal, ie, not sufferingfrom an illness or disorder

◆ 3. They have not sought medical or psychological help specificallyfor the above difficulties, nor has anyone else suggested that they do so

◆ 4. People who do not know them well do not recognize that they have a problem, or if they do, easily attribute it to circum-stances such as “flu” or “overwork”

◆ 5. The symptoms experienced by the subjects have not disrupted their functioning to a major degree, eg, calling in sick severaltimes per winter, or severe marital discord

◆ 6. No history of major affective disorder in wintertime

◆ 7. No serious medical illness

Table III. Diagnostic criteria for subsyndromal seasonal affective disorder(s-SAD) according to Kasper et al.35

Reproduced from reference 35: Kasper S, Rogers SL, Yancey A, Schulz PM, Skwerer RG,Rosenthal NE. Phototherapy in individuals with and without subsyndromal seasonal affec-tive disorder. Arch Gen Psychiatry. 1989;46:837-844. Copyright ©, 1989 JAMA & Archives.

mus.47 A further study was likewise able to showreductions in the availability of striatal dopaminetransporter binding sites in untreated depressedSAD patients.48 This study provides evidence thatthe dopaminergic systems may also be involved inthe pathogenesis of SAD. However, these findings donot explain whether lowered availability of mono-amine transporters in the synaptic cleft representsa primary defect or an attempt to overcome a stateof possible lowered dopamine/serotonin availabilityduring a depressive episode.

There is evidence that specific genetic factors maylead to increased vulnerability for developing SAD.Genetic association studies have examined the roleof several candidate genes: a number of surveys haveinvestigated the serotonin transporter promoter re-peat length polymorphism (5-HTTLPR), but the re-sults of these studies are inconsistent.49,50 One studyexamined the relationship between SAD and PMDDand found an association between the presence ofPMDD and family history and 5-HTTLPR long/shortallele-heterozygosity in females with SAD.51 PMDDand SAD may thus share some genetic vulnerabil-ity factors such as the 5-HTTLPR. A higher rate ofaffective disorders in relatives of patients with SADand PMDD may be indicative of higher genetic vul-nerability in this subgroup when compared with pa-tients with SAD alone. A further interest-deservingstudy investigated the association of the 5-HTTLPRwith the DSM-IV clinical subtypes of depression:Willeit et al52 were able to demonstrate that melan-cholic depression is associated with the 5-HTTLPRlong allele and atypical depression with the shortallele.

Apart from the 5-HTTLPR several other candidategenes have been investigated: guanine nucleotide-binding proteins (G-proteins) have been implicatedin affective disorders, with reports of altered signaltransduction and G-protein levels.53 The higher ac-tivity T-allele of the G-protein beta-3-subunit C825Tpolymorphism has been found to be associated withSAD54; however, other authors have failed to repli-cate these findings.55 The investigation of geneticvariations in clock-genes deserves specific attentionas SAD is thought to be a circadian rhythm disor-der: indeed, Johansson et al56 were able to demon-strate a significant association between the NPAS2471 Leu/Ser polymorphism and SAD, and with thePeriod3 647 Val/Gly polymorphism and diurnal pref-erences (morningness-eveningness tendencies). Inconclusion, the results of these genetic associationstudies have been most promising so far. However,each of these studies has to be considered prelim-inary and replicated in larger samples before defi-nite conclusions may be drawn.

Treatment of SAD

Even one of the oldest written resources of man-kind, the Bible, mentions the beneficial effects oflight: “Truly the light is sweet, and a pleasant thingit is for the eyes to behold the sun.” (Ecclesiastes11:7). From the first reports of SAD,4,57 bright arti-ficial light has been proposed as a most promisingtherapeutic method. Further studies58,59 have ac-

sive symptoms, which are frequently found in SAD.For many patients it takes years until the correctdiagnosis is made: the mean diagnostic latency ina German-speaking sample was approximately 10years after the first depressive episode.37 This is stillunacceptably long. However, since the early 1990s,the diagnostic latency has decreased by about 3years, probably due to education of medical profes-sionals and increased awareness of patients.

Pathogenesis of SAD

It is supposed that the shorter photoperiod duringfall and winter triggers the onset of depressive symp-toms in predisposed individuals. Light is one of themost important zeitgebers and is able to reliably al-ter the length and phase of the human circadianrhythm by affecting the main circadian pacemakerin the diencephalon, the suprachiasmatic nucleus.Scientific research has focused especially on thehormone melatonin (5-methoxy-N-acetyltrypt-amine), which is produced in the pineal gland andthought to mediate the photoperiod signal by theonset and duration of nocturnal secretion. Studiesmeasuring dim light melatonin onset have found aphase delay in SAD patients that was successfullyreversed with BLT.38-40 However, there is a subset ofSAD patients who do not display any circadian ab-normalities, so other mechanisms may also be in-volved in the pathogenesis of SAD.41

There is converging evidence that alteration ofmonoaminergic neurotransmitters, such as sero-tonin, norepinephrine, or even dopamine, may beinvolved in the pathophysiology of SAD. Tryptophandepletion has been used to selectively reduce theplasma tryptophan concentration by administrationof a mixtureof amino acids freeof tryptophan,whichis a precursor of serotonin. Studies using neuro-imaging techniques have been able to show thatserotonin synthesis is reduced for a few hours inthe human brain after this procedure.42 Additional-ly, two further studies have demonstrated that SADpatients in stable remission during summer time43

and after successful BLT44 experience a temporarydepressive relapse after tryptophan depletion. Toinvestigate the role of norepinephrine in SAD, Neu-meister et al45 conducted a study comparing the ef-fects of tryptophan depletion with catecholaminedepletion and sham depletion in SAD patients inremission with BLT: Both active depletion protocolsresulted in temporary depressive relapses provid-ing evidence that brain catecholaminergic systemsmay also be involved in the mechanism of action ofBLT. Interestingly, another study in SAD was ableto demonstrate that monoamine depletion led toalterations in several humoral and cellular immuno-logic parameters, suggesting a potential role of theimmune systeminthepathogenesisof thedisorder.46

Several neuroimaging studies have investigat-ed the importance of specific molecular targets inthe pathophysiology of SAD. Binding studies using123I-β-CIT and single photon emission computedtomography (SPECT) have found that depression inSAD is associated with reduced serotonin trans-porter availability in the thalamus and hypothala-



Conclusions

SAD is a rather frequent psychiatric disorder in thegeneral population,2 but it is inadequately identi-fied, with a diagnostic latency of about 10 years af-ter the first depressive episode. Besides, seasonalchanges in mood also impair social functioning and

knowledged BLT as the first-line treatment for SAD(Table IV). Light boxes emitting full-spectrum brightwhite light of an intensity of 2500 to 10 000 lux in adistance of 60 to 80 cm have been most frequentlyused for BLT. Patients have to be instructed to useBLT for half to 1 hour on a daily basis during thewhole fall and winter season. The duration of treat-ment can be gradually increased to 2 hours andmore per day in case of partial or insufficient re-sponse. It has been demonstrated that the antide-pressant effect of BLT is higher when it is used inthe morning.60,61 BLT has a very high acceptanceamong patients; side effects are mostly mild andtransient and rarely lead to discontinuation of ther-apy (Table V).62-67 The combination of BLT with oth-er nonpharmacological treatment approaches, suchas therapeutic sleep deprivation,68 has been recom-mended to increase its efficacy. In spite of the sig-nificance of BLT, it is also important to recognizethe limitations: Pjrek et al69 reviewed the clinicalcourse of 553 SAD patients and found that onlyabout half of them were sufficiently treated with BLTas monotherapy. Forty-nine percent of patients inthis study received psychopharmacological medi-cation and 35.4% had to be treated with antidepres-sants in addition to BLT.

Despite the clinical importance of antidepressantsand in contrast to the numerous reports of BLT,there have been relatively few studies on the effica-cy of psychopharmacological medication for SAD.70,71

To our knowledge, there is only one placebo-con-trolled trial on the selective serotonin reuptake in-hibitor (SSRI) sertraline72 that has been able to suc-cessfully confirm the superiority of the compoundunder consideration. Some other trials, like thestudy of Lam et al73 on fluoxetine or that of Ling-jaerde et al74 on moclobemide have shown a higherrate of responders or more rapid remission of atyp-ical symptoms, respectively. However, these resultsare not statistically significant due to small sam-ple sizes and a very high rate of placebo responders,which is a common methodological issue in sam-ples mostly consisting of patients with mild-to-mod-erate depression. Besides, there have been severalcase reports and open studies investigating the use-fulness of different antidepressant drugs, eg, citalo-pram,75 Hypericum extract,76,77 mirtazapine,78 andreboxetine.79 Table VI 80 summarizes potential indi-cations for pharmacological treatment of SAD.

While there is at least a scientific basis for thetreatment of SAD with a unipolar course of illness,a search of the available literature reveals a lack ofstudies on specific psychopharmacological treat-ment for bipolar SAD. Furthermore, few bipolarSAD patients are willing to accept treatment withphase prophylactic medication.69 However, in theabsence of controlled trials it seems reasonable toadapt guidelines for the treatment of nonseasonalbipolar affective disorder 81,82 for bipolar SAD.

Recently, there have been reports on the use oftranscranial magnetic stimulation (TMS) in patientswith SAD.83 While these preliminary data suggestthat TMS is effective and lacks major side effects,future studies have to evaluate the importance ofthis method in the clinical management of SAD.



◆ Intensity of BLT device

◆ Wavelength

◆ Distance from light source

◆ Duration

◆ Daytime

◆ During light therapy

◆ Latency to therapeutic response

2500 to 10 000 lux (measured at the level of the eyes of the patient)

Full spectrum of visible light (no ultraviolet component)

The patient should remain seated at about 60 to 80 cmfrom the light source

1/2 to 2 hours per day (depending on the intensity ofthe light source); from fall to spring on a daily basis

Morning BLT exceeds the efficiency of evening BLT

Patient can be active in a seated position, but should maintain the therapeutic distance to the light source

3 to 7 days until onset of antidepressant action

Table IV. Guidelines forthe use of bright lighttherapy (BLT) for sea-

sonal depression.

◆ Prior positive response to antidepressants or mood stabilizers

◆ Bipolar disorder (phase prophylaxis, treatment of acute manic or hypomanic episodes)

◆ Melancholic features specifier (predicts non-response to bright light therapy [BLT]80)

◆ Severe subtypes of depression (eg, psychotic)

◆ History of recurrent depression in the moderate-to-severe range (need for psychopharmaco-logic long-term treatment?)

◆ High suicide risk (assess the need for hospitali-zation, avoid tricyclic antidepressants to pre-vent overdose)

◆ Marked impairment in occupational functioning, in usual social activities, or in relationships with others

◆ Patient preference

◆ Intolerable side effects of BLT

Table VI. Indications forpharmacotherapyin patients withseasonal affectivedisorder (SAD).

◆ Common side effects62,63

– Headache– Eye strain, vision problems– Insomnia– Irritability– Agitation, restlessness– Nausea– Photophobia

◆ Rare side effects (casuistic reports)– Hypomania64

– Suicidal ideations65

– Menstrual disturbances66

– Retinal damage? (no certain proof until now)67

Table V. Adverse effects associatedwith bright lighttherapy.62-67



REFERENCES1. Winkler D. Geschichte der Psychiatrie [History of Psychiatry].In: Linden M, ed. Psychiatrie und Psychotherapie [Psychiatryand Psychotherapy]. 2nd fully revised ed. Vienna, New York:Springer: 2003:443-457.2. Kasper S, TA, Bartko JJ, Gaist PA, Rosenthal NE. Epidemiolog-ical findings of seasonal changes in mood and behavior. A tele-phone survey of Montgomery County, Maryland. Arch Gen Psy-chiatry.1989;46:823-833.3. Lewy AJ, Kern HA, Rosenthal NE, Wehr TA. Bright artificiallight treatment of a manic-depressive patient with a seasonal moodcycle. Am J Psychiatry. 1982;139:1496-1498.4. Rosenthal NE, Sack DA, Gillin JC, et al. Seasonal affective dis-order. A description of the syndrome and preliminary findingswith light therapy. Arch Gen Psychiatry. 1984;41:72-80.5. Booker JM, Hellekson CJ. Prevalence of seasonal affective dis-order in Alaska. Am J Psychiatry. 1992;149:1176-1182.6. Levitt AJ, Boyle MH. Latitude and the variation in seasonal de-pression and seasonality of depressive symptoms. Soc Light Treat-ment Biol Rhythms.1997;9:14. Abstract.7. Rosen LN, Targum SD, Terman M, et al. Prevalence of season-al affective disorder at four latitudes. Psychiatry Res.1990;31:131-144.8. Blazer DG, Kessler RC, Swartz MS. Epidemiology of recurrentmajor and minor depression with a seasonal pattern. The Nation-al Comorbidity Survey. Br J Psychiatry. 1998;172:164-167.9. Saarijarvi S, Lauerma H, Helenius H, Saarilehto S. Seasonalaffective disorders among rural Finns and Lapps. Acta PsychiatrScand. 1999;99:95-101.10. Hagfords C, Thorell LH, Arned M. Seasonality in Finland andSweden, an epidemiological study, preliminary results. Soc LightTreatment Biol Rhythms.1995;7:22. Abstract. 11. Magnusson A, Stefansson JG. Prevalence of seasonal affectivedisorder in Iceland. Arch Gen Psychiatry. 1993;50:941-946.12. Broman JE, Hetta J. Prevalence of seasonal affective disordersand related symptoms at two latitudes in Sweden. Soc Res BiolRhythms.1998;228:162. Abstract13. Eagles JM, Mercer G, Boshier AJ, Jamieson F. Seasonal affec-tive disorder among psychiatric nurses in Aberdeen. J Affect Dis-ord. 1996;37:129-135.14. Mersch PP, Middendorp HM, Bouhuys AL, Beersma DG, vanden Hoofdakker RH. The prevalence of seasonal affective disorderin The Netherlands: a prospective and retrospective study of sea-sonal mood variation in the general population. Biol Psychiatry.1999;45:1013-1022.15. Wicki W, Angst J, Merikangas KR. The Zurich Study. XIV.Epidemiology of seasonal depression. Eur Arch Psychiatry ClinNeurosci. 1992;241:301-306.16. Wirz-Justice A, Kräuchi K, Graw P, Schulman J, Wirz H. Sea-sonality in Switzerland: an epidemiological survey. Soc LightTreatment Biol Rhythms.1992;4:3. Abstract.17. Muscettola G, Barbato G, Ficca G, et al. Seasonality of moodin Italy: role of latitude and sociocultural factors. J Affect Disord.1995;33:135-139.18. Ozaki N, Ono Y, Ito A, Rosenthal NE. Prevalence of seasonaldifficulties in mood and behavior among Japanese civil servants.Am J Psychiatry. 1995;152:1225-1227.19. Murray G, Armstrong S, Hay D. Seasonal affective variationin Australia: disorder or preference. Soc Light Treatment BiolRhythms.1993;5:42. Abstract.20. Morrissey SA, Raggatt PT, James B, Rogers J. Seasonal affec-tive disorder: some epidemiological findings from a tropical cli-mate. Aust N Z J Psychiatry. 1996;30:579-586.21. Ito A, Ichihara M, Hisanaga N, et al. Prevalence of seasonalmood changes in low latitude area: Seasonal Pattern AssessmentQuestionnaire score of Quezon City workers. J Psychiatry Neu-rol. 1992;46:249.22. Pjrek E, Kasper S. Epidemiologie der saisonal abhängigen De-pression (SAD) und ihrer subsyndromalen Form (S-SAD) [Epi-demiology of seasonal-dependent depression (SAD) and its sub-syndromal form (S-SAD)]. In: Kasper S, Möller HJ; eds. Herbst-/Winterdepression und Lichttherapie [Fall-/Winter Depressionsand Phototherapy]. Vienna, New York: Springer; 2004:23-32.

23. Kasper S, Kamo T. Seasonality in major depressed inpatients.J Affect Disord. 1990;19:243-248.24. Lucht MJ, Kasper S. Gender differences in seasonal affectivedisorder (SAD) (German). Arch Women Ment Health.1999;2:83-89.25. Winkler D, Konstantinidis A. Klinisches Erscheinungsbild derHerbst-Winterdepression [Clinical picture of fall-winter depres-sion]. In: Möller HJ, ed. Herbst-/Winterdepression und Licht-therapie [Fall-/Winter Depression and Phototherapy].Vienna,New York: Springer; 2003:13-22.26. Weissman MM, Leaf PJ, Holzer CE, Myers JK, Tischler GL.The epidemiology of depression: an update on sex differences inrates. J Affect Disord. 1984;7:178-188.27. Wehr TA, Giesen H, Schulz PM. Summer depression: descrip-tion of the syndrome and comparison with winter depression. In:Rosenthal NE, ed. Seasonal Affective Disorders and Photother-apy. New York, NY: The Guilford Press; 1989:55-63.28. Wehr TA, Giesen HA, Schulz PM, Anderson JL, Joseph-Van-derpool JR, Kelly K, Kasper S, Rosenthal NE. Contrasts betweensymptoms of summer depression and winter depression. J AffectDisord. 1991;23:173-183.29. Winkler D, Praschak-Rieder N, Willeit M, et al. Seasonal af-fective depression in two German speaking university centers:Bonn, Vienna. Clinical and demographic characteristics (Ger-man). Nervenarzt. 2002;73:637-643.30. American Psychiatric Association. Diagnostic and StatisticalManual of Mental Disorders, 4th edition, text revision (DSM-IV-TR). Washington, DC: American Psychiatric Press; 2000.31. Winkler D, Pjrek E, Konstantinidis A, et al. Anger attacks inseasonal affective disorder. Psychother Psychosom.2004. Submit-ted for publication.32. Praschak-Rieder N, Willeit M, Neumeister A, et al. Prevalenceof premenstrual dysphoric disorder in female patients with sea-sonal affective disorder. J Affect Disord. 2001;63:239-242.33. Kasper S. Jahreszeit und Befindlichkeit in der Allgemein-bevölkerung. Eine Mehrebenenuntersuchung zur Epidemiolo-gie, Biologie und therapeutischen Beeinflußbarkeit (Lichtther-apie) saisonaler Befindlichkeitsschwankungen. Monographienaus dem Gesamtgebiet der Psychiatrie, Bd. 66. Berlin, Heidelberg,New York, Tokyo: Springer; 1991.34. Rosenthal NE, Bradt GH, Wehr TA. Seasonal Pattern Assess-ment Questionnaire. Washington, DC: National Institute of Men-tal Health; 1987.35. Kasper S, Rogers SL, Yancey A, Schulz PM, Skwerer RG,Rosenthal NE. Phototherapy in individuals with and without sub-syndromal seasonal affective disorder. Arch Gen Psychiatry.1989;46:837-844.36. Kasper S, Pjrek E. Diagnose und Behandlung der subsyn-dromalen SAD [Diagnosis and treatment of subsyndromal SAD].In: Kasper S, Möller HJ, eds. Herbst-/Winterdepression und Licht-therapie [Fall-/Winter Depression and Phototherapy]. Vienna,New York: Springer; 2004:33-40.37. Winkler D, Willeit M, Praschak-Rieder N, et al. Changes ofclinical pattern in seasonal affective disorder (SAD) over time ina German-speaking sample. Eur Arch Psychiatry Clin Neurosci.2002;252:54-62.38. Terman M, Terman JS, Quitkin FM, Cooper TB, Lo ES, Gor-man JM, Stewart JW, McGrath PJ. Response of the melatonin cy-cle to phototherapy for Seasonal Affective Disorder. Short note.J Neural Transm. 1988;72:147-165.39. Dahl K, Avery DH, Lewy AJ, et al. Dim light melatonin onsetand circadian temperature during a constant routine in hyper-somnic winter depression. Acta Psychiatr Scand. 1993;88:60-66.40. Sack RL, Lewy AJ, White DM, Singer CM, et al. Morning vsevening light treatment for winter depression: evidence that thetherapeutic effects of light are mediated by circadian phase shifts.Arch Gen Psychiatry. 1990;47:343-351.41. Lam RW, Levitan RD. Pathophysiology of seasonal affectivedisorder: a review. J Psychiatry Neurosci. 2000;25:469-480.42. Benkelfat C, Seletti B, Palmour RM, Hillel J, Ellenbogen M,Young SN. Tryptophan depletion in stable lithium-treated pa-tients with bipolar disorder in remission. Arch Gen Psychiatry.1995;52:154-156.

tion as well as an underlying genetic predispositionhave been implicated in the pathogenesis of SAD.Antidepressant psychopharmacotherapy is used forthe treatment of SAD depression similar to otherforms of depression; however, BLT is now recog-nized as the treatment option of choice. Future re-search should aim at investigating specific treat-ment options for SAD patients with a bipolar courseof illness. ❒

reduce patients’ performance at work.84 SAD, in itsmost frequent manifestation, fall-winter depression,is characterized in most cases by reverse vegetativesymptoms such as increased appetite, carbohydratecraving, hypersomnia, daytime fatigue, and loss ofenergy. However, the most important clinical featureis that patients feel worst in the fall-winter months,ie, in the time of light deficiency. Changes in hor-monal and monoaminergic neurotransmitter func-



66. Pjrek E, Winkler D, Willeit M, Konstantinidis A, Thierry N,Kasper S. Menstrual disturbances a rare side-effect of bright-lighttherapy. Int J Neuropsychopharmacol. 2004;7:239-240.67. Gallin PF, Terman M, Reme CE, Rafferty B, Terman JS, BurdeRM. Ophthalmologic examination of patients with seasonal affec-tive disorder, before and after bright light therapy. Am J Ophthal-mol.1995;119: 202-210.68. Praschak-Rieder N, Willeit M, Neumeister A, Hilger E, KasperS. Therapeutic sleep deprivation and phototherapy (German).Wien Med Wochenschr.1999;149:520-524.69. Pjrek E, Winkler D, Stastny J, Konstantinidis A, Heiden A,Kasper S. Bright light therapy in seasonal affective disorder—does it suffice? Eur Neuropsychopharmacol. 2004;14:347-351.70. Kasper S, Hilger E, Willeit M, et al. Drug therapy. In: Parto-nen T, Magnusson A, eds. Seasonal Affective Disorder: Practiceand Research. New York, NY: Oxford University Press; 2001:85-93.71. Hilger E, Praschak-Rieder N, Willeit M, et al. Pharmacother-apy of seasonal depression (German). Nervenarzt. 2002;73:22-31.72. Moscovitch A, Blashko CA, Eagles JM, et al. A placebo-con-trolled study of sertraline in the treatment of outpatients with sea-sonal affective disorder. Psychopharmacology. 2004;171:390-397.73. Lam RW, Gorman CP, Michalon M, et al. Multicenter, placebo-controlled study of fluoxetine in seasonal affective disorder. AmJ Psychiatry. 1995;152:1765-1770.74. Lingjaerde O, Reichborn-Kjennerud T, Haggag A, Gaertner I,Narud K, Berg EM. Treatment of winter depression in Norway: II.A comparison of the selective monoamine oxidase A inhibitor mo-clobemide and placebo. Acta Psychiatr Scand.1993;88:372-380.75. Wirz-Justice A, Van der Velde P, Bucher A, Nil R. Comparisonof light treatment with citalopram in winter depression: a longitu-dinal single case study. Int Clin Psychopharmacol.1992;7:109-116.76. Martinez B, Kasper S, Ruhrmann S, Möller HJ. Hypericum inthe treatment of seasonal affective disorders. J Geriatr Psychia-try Neurol. 1994;7(suppl):29-33.77. Kasper S. Treatment of seasonal affective disorder (SAD) withHypericum extract. Pharmacopsychiatry.1997;30(suppl):89-93.78. Hesselmann B, Habeler A, Praschak-Rieder N, Willeit M,Neumeister A, Kasper S. Mirtazapine in seasonal affective disorder(SAD): a preliminary report. Hum Psychopharmacol Clin Exp.1999;14:59-62.79. Hilger E, Willeit M, Praschak-Rieder N, Stastny J, Neumeis-ter A, Kasper S. Reboxetine in seasonal affective disorder: an opentrial. Eur Neuropsychopharm. 2001;11:1-5.80. Terman M, Amira L, Terman JS, Ross DC. Predictors of re-sponse and nonresponse to light treatment for winter depression.Am J Psychiatry. 1996;153:1423-1429.81. American Psychiatric Association. Practice Guideline for theTreatment of Patients With Bipolar Disorder [Revision]. Am JPsychiatry. 2002;159:1-50.82. Calabrese JR, Kasper S, Johnson G, et al. International Con-sensus Group on Bipolar I Depression Treatment Guidelines (Aca-demic highlights). J Clin Psychiatry. 2004;65:569-579.83. Konstantinidis A, Heiden A, Stastny J, et al. Rapid transcra-nial magnetic stimulation (r-TMS): a novel treatment option forseasonal affective disorder (SAD)? Eur Neuropsychopharm.2002;12(suppl):252.84. Booker JM, Roseman C. A seasonal pattern of hospital med-ication errors in Alaska. Psychiatry Res.1995;57:251-257.

43. Neumeister A, Praschak-Rieder N, Hesselmann B, et al. Effectsof tryptophan depletion in fully remitted patients with seasonalaffective disorder during summer. Psychol Med.1998;28:257-264.44. Neumeister A, Praschak-Rieder N, Hesselmann B, Rao ML,Gluck J, Kasper S. Effects of tryptophan depletion on drug-freepatients with seasonal affective disorder during a stable responseto bright light therapy. Arch Gen Psychiatry.1997;54:133-138.45. Neumeister A, Turner EH, Matthews JR, et al. Effects of tryp-tophan depletion vs catecholamine depletion in patients with sea-sonal affective disorder in remission with light therapy. Arch GenPsychiatry.1998;55:524-530.46. Stastny J, Konstantinidis A, Schwarz MJ, et al. Effects of tryp-tophan depletion and catecholamine depletion on immune pa-rameters in patients with seasonal affective disorder in remissionwith light therapy. Biol Psychiatry. 2003;53:332-337.47. Willeit M, Praschak-Rieder N, Neumeister A, et al. [123I]-beta-CIT SPECT imaging shows reduced brain serotonin transporteravailability in drug-free depressed patients with seasonal affectivedisorder. Biol Psychiatry. 2000;47:482-489.48. Neumeister A, Willeit M, Praschak-Rieder N, et al. Dopaminetransporter availability in symptomatic depressed patients withseasonal affective disorder and healthy controls. Psychol Med.2001;31:1467-1473.49. Rosenthal NE, Mazzanti CM, Barnett RL, et al. Role of sero-tonin transporter promoter repeat length polymorphism (5-HT-TLPR) in seasonality and seasonal affective disorder. Mol Psychi-atry.1998;3:175-177.50. Johansson C, Willeit M, Levitan R, et al. The serotonin trans-porter promoter repeat length polymorphism, seasonal affectivedisorder and seasonality. Psychol Med. 2003;33:785-792.51. Praschak-Rieder N, Willeit M, Winkler D, et al. Role of familyhistory and 5-HTTLPR polymorphism in female seasonal affec-tive disorder patients with and without premenstrual dysphoricdisorder. Eur Neuropsychopharmacol. 2002;12:129-134.52. Willeit M, Praschak-Rieder N, Neumeister A, et al. A polymor-phism (5-HTTLPR) in the serotonin transporter promoter geneis associated with DSM-IV depression subtypes in seasonal affec-tive disorder. Mol Psychiatry. 2003;8:942-946.53. Avissar S, Schreiber G, Nechamkin Y, et al. The effects of sea-sons and light therapy on G protein levels in mononuclear leuko-cytes of patients with seasonal affective disorder. Arch Gen Psy-chiatry.1999;56:178-183.54. Willeit M, Praschak-Rieder N, Zill P, et al. C825T polymor-phism in the G protein beta3-subunit gene is associated with sea-sonal affective disorder. Biol Psychiatry. 2003;54:682-686.55. Johansson C, Willeit M, Aron L, et al. Seasonal affective dis-order and the G-protein beta-3-subunit C825T polymorphism.Biol Psychiatry. 2004;55:317-319.56. Johansson C, Willeit M, Smedh C, et al. Circadian clock-relat-ed polymorphisms in seasonal affective disorder and their rele-vance to diurnal preference. Neuropsychopharmacology. 2003;28:734-739.57. Rosenthal NE, Sack DA, Carpenter CJ, et al. Antidepressanteffects of light in seasonal affective disorder. Am J Psychiatry.1985;142:163-170.58. Lam RW, Kripke DF, Gillin JC. Light therapy for depressivedisorders: a review. Can J Psychiatry. 1989;34:140-147.59. Terman M, Terman JS, Quitkin FM, McGrath PJ, Stewart JW,Rafferty B. Light therapy for seasonal affective disorder: a reviewof efficacy. Neuropsychopharmacology. 1989;2:1-22.60. Lewy AJ, Sack RL, Miller LS, Hoban TM. Antidepressant andcircadian phase-shifting effects of light. Science.1987;235:352-354.61. Terman JS, Terman M, Lo ES, Cooper TB. Circadian time ofmorning light administration and therapeutic response in winterdepression. Arch Gen Psychiatry. 2001;58:69-75.62. Kogan AO, Guilford PM. Side effects of short-term 10,000-luxlight therapy. Am J Psychiatry. 1998;155:293-294.63. Labbate LA, Lafer B, Thibault A, Sachs GS. Side effects in-duced by bright light treatment for seasonal affective disorder.J Clin Psychiatry.1994;55:189-191.64. Swiecicki L. Phototherapy for winter depression: report ofthree cases. Psychiatr Pol. 1993;27:667-672.65. Praschak-Rieder N, Neumeister A, Hesselmann B, Willeit M,Barnas C, Kasper S. Suicidal tendencies as a complication of lighttherapy for seasonal affective disorder: a report of three cases.J Clin Psychiatry.1997;58:389-392.

TROUBLE AFFECTIF SAISONNIER : DU DIAGNOSTIC AU TRAITEMENT

L e trouble affectif saisonnier (TAS, ou « blues de l’hiver ») est un sous-type de la dépression majeure ou du trouble bipolaire qui fait l’objet derecherches en psychiatrie depuis plus de 20 ans. La prévalence dans la

population générale varie entre 2 % et 5 % dans les pays à climat tempéré, lesfemmes étant beaucoup plus souvent touchées par le syndrome. Des symptômesdépressifs atypiques accompagnés d’une symptomatologie végétative inverséeprédominent pendant les épisodes dépressifs au cours de la période obscure del’année, tandis qu’un quart des patients environ présentant des épisodes hypo-maniaques (rarement maniaques) pendant la période printemps/été. La lumi-nothérapie est généralement admise comme traitement de la dépression du TAS,mais le traitement psychopharmacologique par les antidépresseurs s’est éga-lement montré efficace et bien toléré. Cet article passe en revue les conceptsphysiopathologiques les plus importants et donne un aperçu des méthodes dia-gnostiques actuelles ainsi que des options thérapeutiques pour les TAS.

✦

254 Trends in diagnosis and treatment of bipolar disorders – Kupfer and Frank MEDICOGRAPHIA, VOL 27, No. 3, 2005

Bipolar disorder (manic-depressive illness) isa chronically recurring subtype of mood dis-orders with considerable psychiatric and med-

ical comorbidity. As contrasted with the diagnosisand treatment of nonbipolar depression, which of-ten occurs in the primary care sector, the treatmentof bipolar disorder has often required great diag-nostic skill and sophisticated therapeutic interven-tions by specialists. Even among specialists bipolardisorder is routinely not recognized or misdiag-nosed.

Diagnostic characteristics and potential etiology

As shown in Figure 1, bipolar disorder can be char-acterized by varying degrees of symptomatology inseveral symptom domains: the manic mood and be-havior domain, the psychotic symptom domain, thedysphoric or negative mood and behavior domain,and the cognitive dysfunctioning domain. Not ev-ery one of the symptoms within a domain is pres-ent at the same time; however, it is important torecognize that symptoms from each of these do-mains may be present at the same time, along withneurovegetative symptoms including changes insleep, appetite, and energy. Indeed, some expertsargue that bipolar disorder always presents with asymptom picture that is mixed to a greater or less-er extent. Generally, however, the major symptomset in each of these domains leads to the diagnosisof a particular phase of the disorder, usually con-sidered to be mania with psychosis or withoutpsychosis, depression with or without psychosis, ora fully mixed state. The Diagnostic and StatisticalManual of Mental Disorders, 4th Edition (DSM-IV)nosology lays out criteria for the diagnosis of bipo-lar I disorder, bipolar II disorder, cyclothymic disor-der, and also bipolar disorder, not otherwise speci-fied (NOS). In a similar fashion, adiagnosisofbipolarspectrum disorder is included to subsume all pa-tients who meet DSM-IVcriteria for one of theabovediagnoses, as well as some patients who do not meetthe criteria for either mania or hypomania. Asshown in Figure 2, the age of onset for bipolar dis-order is much earlier than previously assumed. Ina 3000-person cohort who were self-identified ashaving bipolar disorder,1 the onset appears to peakbetween the ages of 15 and 19. This report is con-sistent with other new data sets suggesting that,with or without comorbid psychiatric disease, bipo-lar disorder first appears much more frequently nowin the teen years and in the twenties.

Given the often confusing range of symptoms aswell as the early onset, it is not surprising that bipo-lar disorder is often unrecognized or misdiagnosed,and consequently mismanaged. In one study, 69%of the 600 respondents were initially misdiagnosed.2

These patients waited almost 10 years to receive thecorrect diagnosis, with an average of 3.5 prior mis-diagnoses. Nonbipolar depression diagnoses weremade 60% of the time, followed by anxiety disorder.Failure to recognize and treat bipolar disorder addsto the huge clinical and human costs associatedwith this disorder.


B ipolar disorder (manic-depressive illness) is a chronically recurring sub-type of mood disorders that is associated with considerable psychiatricand medical comorbidity. Psychiatric comorbidity, such as alcohol or

drug abuse, anxiety, and panic disorder adversely influences outcomes. Medi-cal disease and medical risk factors are common in bipolar disorder and leadto even greater morbidity and mortality in this population. Associated with ahigh rate of suicide, bipolar disorder is a potentially lethal disease. In consid-ering appropriate interventions for bipolar disorder, it is easiest to group themas treatment of acute mania or mixed states, treatment of acute depression, andtreatment for the prevention of both manic and depressive recurrence. Mono-therapy, usually in the form of lithium, valproate, or one of the atypical anti-psychotic medications, is initially recommended for the treatment of maniaand mixed states. The treatment of acute depression is generally longer andmore difficult than mania. Many expert clinicians have recommended opti-mization of a mood stabilizer and, if this strategy is ineffective, adding an anti-depressant to the mood stabilizer. Mood stabilizer monotherapy was once con-sidered the ideal maintenance treatment; however, considerable attention hasnow been devoted to the conduct of controlled trials of atypical antipsychoticmedications in the maintenance phase. In considering future trends, the in-creased use of atypical antipsychotics and their possible efficacy, not only forstabilization of mania, but for depressive symptomatology, is an importantarea of further investigation. Furthermore, the use of agents with a regulatoryeffect on biological rhythms deserves further exploration.Medicographia. 2005;27:254-260. (see French abstract on page 260)

Keywords: bipolar disorder; atypical antipsychotic; mania; depression;maintenance treatment

David J. KUPFER, MD Ellen FRANK, PhDDepartment of Psychiatry University of Pittsburgh School of Medicine Western Psychiatric Institute and Clinic Pittsburgh, Pa, USA

▲

Address for correspondence: David J. Kupfer, MD, Department of Psychiatry, Western PsychiatricInstitute and Clinic, University of Pittsburgh School of Medicine, 3811 O’Hara Street, Pittsburgh, PA 15213, USA (e-mail: [email protected])

Trends in diagnosis and treatment of bipolar disorders b y D . J . K u p f e r a n d E . F r a n k , U S A

255Trends in diagnosis and treatment of bipolar disorders – Kupfer and Frank MEDICOGRAPHIA, VOL 27, No. 3, 2005


intervention. As highlighted recently,13 progress hasbeen made in understanding the role of geneticsin bipolar disorder and complemented by findingsfrom the neuroimaging field, both functional mag-netic resonance imaging (fMRI) and structuralimaging. First, studies repeatedly confirm that bi-polar disorder has extremely high heritability.14 Sec-ond, they demonstrate increased frequency of bipo-lar disorder in family members, but also of otherdisorders that are primarily psychotic disorders(schizoaffective disorder and schizophrenia).15 Spe-

cific chromosomal regions have been identified, butonly 5 out of 16 chromosomal regions currentlymeet criteria for significant linkage.13 Complement-ing this work on genetics, anatomical and MRI neu-roimaging studies have demonstrated functionalabnormalities of the medial prefrontal cortex.16 Neu-ropsychological dysfunction, including abnormal-ities in cognitive and memory activity,17,18 but par-ticularly in social cognition,19 have been found inindividuals with bipolar disorder. This is consistentwith work of others, including Phillips,20 who hasbeen particularly interested in alterations in emo-tional perception. Neurocognitive deficits and, per-haps, deficits in social cognition may represent, ifnot specific, at least consistent findings in bipolar

Another major feature of bipolar disorder is thelevel of psychiatric and medical comorbidity. Com-munity-based studies indicate that 60% to 70% ofindividuals with bipolar illness meet diagnostic cri-teria for a lifetime history of substance abuse or de-pendence.3,4 The risk for alcohol or drug abuse is 6to 7 times greater in patients with bipolar disorderthan would be expected in the general population.5

A more recent investigation from the Stanley Foun-dation Bipolar Network reports 2 to 3 times the pop-ulation rate of alcohol abuse in men with bipolardisorder and 7 times the frequency in women.6

Among other psychiatric comorbidities, anxiety,panic disorder, and panic spectrum disorder arecommon and adversely influence bipolar outcomes.Angst and colleagues have reported that the asso-ciation with panic is greater in patients with bipo-lar disorder than in patients with major depression(J. Angst, personal communication). Panic and oth-er forms of anxiety significantly delay remission ofacute bipolar episodes.7,8 The presence of somaticsymptoms in bipolar disorder, particularly cardio-vascular symptoms, is increased in the presence ofpanic disorder. McElroy and colleagues9 have high-lighted the frequency of eating problems in bipolardisorder, including symptoms of bulimia nervosaand, in particular, binge eating.

Associated with a high rate of suicide, bipolar dis-order is a potentially lethal disease.1,10,11 Medical dis-ease and medical risk factors are present in bipolardisorder and lead to even greater morbidity andmortality in this population. The most commonlyreported problems are cardiovascular disease, dia-betes, obesity, and thyroid disease. The accumula-tion of key medical risk factors related to excessivenicotine use, alcohol and drug use, concomitantanxiety, and eating disorders, lead to the early onsetof medical diseases with poor long-term outcomes.Furthermore, because patients with bipolar disor-der spend most of their time in the depressive phaseof the illness, there is often a loss of the disciplineand motivation required to reduce such medicalrisk factors. Katon12 has established the clear rela-tionship between depression and a host of negativehealth behaviors including smoking, poor diet, over-eating, and sedentary lifestyle and has shown thatdepression has a maladaptive effect on adherenceto medical regimens and direct adverse physiolog-ic effects including decreased heart rate variabilityand increased adhesiveness of platelets.

Bipolar I disorder is almost always treated in men-tal health settings, with patients viewing their psy-chiatric care as their most important form of med-ical care. This has led to a relative underrecognitionof and inattention to the many physical diseasesfrom which these patients suffer. Only recently hasa greater awareness of medical burden and medi-cal risk factors, stimulated by the introduction ofatypical antipsychotic medication for bipolar I dis-order with its attendant medical problems, led clin-icians and investigators to focus on these issues. To-day there is broad recognition that advances in thetreatment of bipolar disorder must go together withincreased medical risk factor assessment, ongoinglaboratory surveillance, and integrated treatment

Figure 2. Distribution of bipolar symptom onset by age.

Figure 1. Bipolar disorder symptom domains.

Bipolardisorder

Manic mood and behavior◆ Euphoria◆ Grandiosity◆ Pressured speech◆ Impulsivity◆ Excessive libido◆ Recklessness◆ Social intrusiveness◆ Diminished need for sleep

Psychotic symptoms◆ Delusions◆ Hallucinations

Dysphoric or negativemood and behavior◆ Depression◆ Anxiety◆ Irritability◆ Hostility◆ Violence or suicide

Cognitive symptoms◆ Racing thoughts◆ Distractibility◆ Disoganization◆ Inattentiveness

Age of first onset (y)

Perc

enta

ge

0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45+

30

25

20

15

10

5

0



Treatment paradigms in bipolar disorder

Once thought of as our “good prognosis” illness inpsychiatry, bipolar disorder has emerged in the lastdecade as one of our most significant treatmentchallenges. Thinking back to what it must havebeen like to try to manage this condition in the pre-lithium era, it is easy to understand how, once lithi-um was introduced, the field was lulled into a falsesense of complacency. In the late 1980s and early1990s, however, both naturalistic data27,28 and datafrom controlled trials (eg, Prien et al,29 1984) madeit clear that bipolar disorder was not a problemsolved. This led a relatively small group of investi-gators in the US and abroad to take a renewed in-terest in developing treatment strategies for bipolardisorder, both psychopharmacologic and psycho-therapeutic.30 As we have entered the 21st century,we have come to recognize the multiple factors thatcomplicate the process of achieving ideal outcomesfor those that suffer from bipolar disorder.

Our model for understanding these complexitiesis illustrated in Figure 3. According to our model,the early age of onset typical of bipolar disordermeans that early years spent in coping with the ill-ness often lead to social developmental/matura-tional deficits. Future experience of the illness often,then, takes place against a background of such defi-

disorder. The possible linkage of individual geneswith neuroanatomical findings and cognitive def-icits represents a potential cascade of processes deal-ing with gene-environmental interaction, similarto the research ongoing in the area of schizophre-nia.21

A second area of interest in bipolar disorder hasemerged from the long-standing research in bio-logical rhythms and bipolar disorder. Ehlers andcolleagues22 hypothesized that biological dysregu-lation, both at the level of neurobiology and also atthe level of social rhythms, may be responsible for acascade of events leading to episodes of mania anddepression and proposed a “social zeitgeber theo-ry,” to explain the onset of bipolar disorder.23 In sup-port of this concept, Malkoff-Schwartz and others24

demonstrated that events associated with changesin social routines were common prior to the onsetof a manic episode. The social rhythm disruptionconcept led to the design and implementation of apsychosocial intervention specifically tailored forthe treatment of bipolar disorder that has been de-veloped both from interpersonal psychotherapy25

and social rhythm theory.26 Consistent with this bi-ological rhythms perspective, a host of biologicalclock genes has been studied in both families andpatients with bipolar disorder, suggesting a vulner-ability in certain patient groups or groups at risk forbipolar disorder.

Array of bipolar disorder

symptoms

Psychiatric comorbidity◆ Panic◆ Ethyl alcohol◆ Substance disorders◆ Personality disorders

Dysphoric or negativemood and behavior◆ Depression◆ Anxiety◆ Irritability◆ Hostility◆ Violence or suicide

Back

grou

nd o

f soc

ial d

evel

opm

enta

l/m

atur

atio

nal d

efic

its

Medical comorbidity◆ Obesity◆ Thyroid disease◆ Diabetes◆ Cardiovascular

Cognitive symptoms◆ Racing thoughts◆ Distractibility◆ Disoganization◆ Inattentiveness

Limited educationalattainment

Unemploymentand poverty

Early onset ofillness

Poor treatmentadherence

Psychotic symptoms◆ Delusions◆ Hallucinations

Manic mood and behavior◆ Euphoria◆ Grandiosity◆ Pressured speech◆ Impulsivity◆ Excessive libido◆ Recklessness◆ Intrusiveness◆ Reduced need for sleep

Figure 3. A model of intervention points for improving treatment outcomes in bipolar disorder.



less ill manic patients, lithium, valproate, or carba-mazepine can be administered; for more severe ma-nia, atypical antipsychotics or valproate are rec-ommended. For mania associated with psychosis,atypical antipsychotics are preferred.

Monotherapy is initially recommended to addressmanic features.The use of lithium, valproate, as wellas of a number of atypical antipsychotics, includingolanzapine, risperidone, quetiapine, nefazodone, andaripiprazole is recommended. If patients do not im-prove on any of these drugs, a number of clinicalspecialists have suggested that the addition of olan-zapine, risperidone, or quetiapine to lithium or val-proate confers an additional effect, as does the addi-tion of valproate. The avoidance of major depressivesymptomatology as an outcome in the successful

resolution of manic symptoms could be helped byuse of antipsychotics, which may have antidepres-sant effects of their own. It is further recommend-ed that after acute treatment, on the way to stabi-lization and maintenance treatment, patients andfamily members receive psychoeducation, with par-ticular emphasis on recognition of early symptomsof mania. Many clinicians also further recommendthe availability of rescue medication, such as smalldoses of an atypical antipsychotic.

The treatment of acute depression is longer andmore difficult than that of mania, as shown in Fig-ure 4.34 Many expert clinicians have recommendedthe initial optimization of a mood stabilizer and, ifthis strategy is ineffective, adding an antidepressantto the mood stabilizer. If this second step does notwork, then there are a variety of recommendationsincluding adding a second mood stabilizer.35 Re-cently, it has been shown that an atypical antipsy-chotic plus an selective serotonin reuptake inhib-

cits making rational decision-making and copingwith difficult challenges more difficult for the in-dividual even during euthymic periods. The com-plexity of the disorder itself (as discussed earlier),involving as it does an ever-changing mix of psy-chotic symptoms, dysphoric or negative mood andbehavior, manic mood and behavior, and cognitivesymptoms makes treatment decision-making a con-stantly ongoing process in which no sooner is asymptom or group of symptoms effectively ad-dressed than others appear which require new treat-ment decisions. Further complicating the pictureare the psychiatric and medical comorbidities com-monly encountered in bipolar disorder: panic, al-cohol abuse, substance abuse, and personality dis-orders (particularly Cluster B disorders) as well asobesity, thyroid disease, diabetes, and cardiovascu-lar disease. Each of these comorbidities can haveimportant impact on treatment decisions. For ex-ample, we have observed that patients with paniccomorbidity are much more sensitive to medica-tion side effects, often necessitating lower startingdoses and slower titration of new treatments.8 Inthe area of medical comorbidity, those patients withthyroid disease or diabetes also warrant special clin-ical attention if their bipolar disorder treatment isto be benign with respect to these medical condi-tions. Finally, although treatment adherence is anissue in all of medicine, it seems to be a particularchallenge for patients with bipolar disorder. As Fig-ure 3 illustrates, multiple factors in the model in-cluding the poor judgment brought on by mania orsubstance use, the medication sensitivity broughton by panic comorbidity, and the desire to limitweight gain in an already obese individual, may allcontribute to the challenges to treatment adherencein this population. Thought of in this way, there are,then, a series of potential targets for improvingtreatment paradigms for bipolar disorder.

In considering appropriate interventions for bi-polar disorder, it is easiest to group them as treat-ment of acute mania or mixed states, the treatmentof acute depression, and the treatment for the pre-vention or prophylaxis of both mania and depres-sion in the future (maintenance treatment). Thetreatment of acute mania involves the followingaims: the improvement of manic symptoms, possi-bly requiring rapid tranquillization; the resolutionof manic symptomatology avoiding both the onsetof new depressive symptomatology or the unmask-ing of depressive symptomatology; and finally, theconsideration of mood stabilization and subsequenttreatment. With respect to rapid tranquillization,Cookson and others have demonstrated that hal-doperidol, as well as the new atypical antipsychotics,have immediate antimanic effects independent ofsedation.31 Other drugs often used in the treatmentof acute mania may not act as quickly and there-fore have delayed antimanic effects. The AmericanPsychiatric Association (APA) Guidelines suggestthat for mild mania, lithium or valproate or an atyp-ical antipsychotic can be utilized.32 For severe ma-nia, they recommend the immediate combinationof a mood stabilizer and an atypical antipsychotic.The recent British Guidelines33 conclude that for

Figure 4. Bipolar depressive and mixed episodes are difficult to treat.Reproduced from reference 34: Kupfer DJ, Frank E, Grochocinski VJ, et al. Stabilization in the treatmentof mania, depression and mixed states. Acta Neuropsychiatrica. 2000;12:110-114. Copyright © 2000,Blackwell Publishing.

1.0

0.8

0.0

0.2

0.4

0.6

Cum

ulat

ive

prop

ortio

n of

pat

ient

s re

mitt

ed

Weeks of preliminary phase treatment

100 20 30 40 50

DepressedMixed/cyclingManic

*P<0.0001 vs manic; **P<0.001 vs manic; and P<0.05 vs depressive



but not necessarily depressive episodes39; the datafor valproate for long-term treatment are minimaldespite its wide use40; the data for lamotrigine forlong-term treatment support a specific indicationfor the prevention of depressive as compared withmanic episodes.41,42 Recently, attention has been de-voted to the conduct of controlled trials of atypicalantipsychotics in the maintenance phase. The ma-jority of such controlled studies have been carriedout with olanzapine, with one recent trial of aripi-prazole.43 Studies of olanzapine demonstrate supe-rior efficacy to valproate; and comparisons of olan-zapine to lithium show an advantage of olanzapinewith respect to study completion.44,45 An olanzapineversus placebo study for relapse prevention demon-strated efficacy for olanzapine in preventing relapseinto mania as well as into depression.46 Finally, astudy comparing olanzapine plus lithium or val-proate versus placebo plus lithium or valproateshowed the advantage of the addition of olanzapineas compared with placebo, suggesting the possibil-ity of combined long-term treatment.47 Recently, a26-week trial comparing aripiprazole to placebo hasbeen reported demonstrating a significant advan-tage for this atypical antipsychotic.43

Maintenance treatment should include, at theminimum, patient management strategies33 con-sisting of five specific areas: establishment andmaintenance of a therapeutic alliance; education ofpatients and their families about the disorder; ef-forts to enhance treatment adherence; promotionof awareness of stresses, sleep disturbance, and ear-ly signs of relapse and regular patterns of activity;and finally, evaluation and management of func-tional impairment. Furthermore, targeted psycho-social interventions are now being studied in thetreatment of bipolar disorder that may be very ap-plicable to larger groups of patients suffering fromthis illness.26,48 Interestingly, these approaches gen-erally include strong emphasis on improving treat-ment adherence.

Future trends

In considering future trends, the increased use ofatypical antipsychotics and their possible utility,not only for stabilization of mania, but for depres-sive symptomatology, is an important area of fur-ther investigation. That monotherapy may not bethe optimal way to proceed with patients with bipo-lar disorder is becoming clearer, making it very like-ly that an increased emphasis will be placed on com-bination treatments in the next 5 years. It is hopedthat controlled clinical trials will be available toguide the clinician. With respect to the treatmentsspecifically for bipolar depression, it is likely thata novel serotonin receptor agonist or antagonistwill be used to treat bipolar depression. Some of theother novel treatments being suggested for unipo-lar depression may also be useful for bipolar depres-sion. In assessing the risk-benefit ratio of the useof any medication in bipolar disorder, it is importantto consider the level of medical risk factors that maymake patients more vulnerable to the adverse effectsof certain atypical antipsychotics and to the devel-

itor (SSRI) is effective in acute bipolar depression,36

leading to Food and Drug Administration (FDA) ap-proval of this combination for treatment of bipolardepression. In considering alternatives involvingmood stabilizers, various controlled trials have sug-gested that lithium may have antidepressant effica-cy. Similarly, lamotrigine may possess antidepres-sant efficacy in acute bipolar depression. Recentreviews of controlled trials of antidepressants versusplacebo37 have demonstrated a mild positive effectfor antidepressants, accompanied by a relatively lowswitch rate from depression to mania with antide-pressants (except for the tricyclics). Two clinical tri-als with quetiapine have shown efficacy in acutebipolar depression (J. R. Calabrese, personal com-munication). The failure to respond to a mood sta-bilizer (or an atypical antipsychotic) with or with-out an antidepressant leads the clinician to considersubstituting another antidepressant or mood sta-bilizer, as well as the use of electroconvulsive ther-apy (ECT) and/or adjunctive psychological interven-tions. Finally, a key unanswered question is howlong one should continue an antidepressant follow-ing resolution of a bipolar depressive episode. Noguidelines, similar to those available for the treat-ment of acute unipolar depression, exist.

When we review options for long-term treatment,one important issue is that we need to deal with theprevention of both manic and depressive episodes.Indeed, one could list the therapeutic objectives inthe following manner: to achieve symptom controlof all four domains that have been mentioned pre-viously; to prevent syndromal recurrence; and toreduce or oblate all subsyndromal symptomatologybecause this is often related to significant function-al impairment. Finally, it is important to minimizethe risk of cycling, which, when it occurs, makes itmore difficult to treat both the acute manic or de-pressive state and to manage the illness over thelong term.

Complicating the treatment of bipolar disorder,both acutely and long term, is the presence of con-siderable psychiatric and medical comorbidity.Therefore, a successful long-term treatment ap-proach will also assist in treating the psychiatricand medical comorbidities. Underscoring the im-portance of complete symptomatic relief is the asso-ciated functional recovery as manifested by returnto work, family, and schooling for younger people.

Another important consideration in maintenancetreatment is a reduction in overall mortality. Whilea great deal of appropriate attention has been de-voted to suicide prevention, bipolar patients also dieof other causes, especially medical disease.38 Con-sequently, a major goal of maintenance treatmentis to reduce overall mortality. This goal is compli-cated by the fact that adherence to treatments forboth medical and psychiatric disorders for patientswith bipolar disorder is very poor and, therefore,we must pay specific attention to treatment adher-ence principles.

In reviewing the possibilities of long-term treat-ment in bipolar disorder, at the present time wehave the following information: lithium appears tobe effective for the prevention of manic episodes,



neuroimaging techniques, the application of socialrhythm and social cognition strategies, as well asfurther examinations of temperament. Only by de-fining better phenotypes and endophenotypes, willwe be able to understand the pathogenesis of bipo-lar disorder. ❏

Funding/Support: National Institute of Mental Healthgrants MH-30915 and MH-29618; and The Common-wealth of Pennsylvania Department of Health grant ME-02385.

opment of medical diseases. Consequently, patientswith bipolar disorder should receive careful medi-cal monitoring throughout the period of treatment.

While we have considered the need for new drugtreatment targets, it is also clear that our under-standing of bipolar disorder will also be facilitatedby paying more careful attention to a better defini-tion of behavioral phenotypes as well as to identify-ing markers for intermediate endophenotypes. Thiswill require the use of pharmacogenomic strategies,

REFERENCES1. Kupfer DJ, Frank E, Grochocinski VJ, Cluss PA, Houck PR,Stapf DA. Demographic and clinical characteristics of individu-als in a bipolar disorder case registry. J Clin Psychiatry. 2002;63:120-125.2. Hirschfeld RM, Lewis L, Vornik LA. Perceptions and impactof bipolar disorder: how far have we really come? Results of thenational depressive and manic-depressive association 2000 sur-vey of individuals with bipolar disorder. J Clin Psychiatry. 2003;64:161-174.3. Goldberg JF. Bipolar disorder with comorbid substance abuse:diagnosis, prognosis and treatment. J Psychiatr Pract. 2001;7:109-122.4. Regier DA, Farmer ME, Rae DS, et al. Comorbidity of mentaldisorders with alcohol and other drug abuse. Results from theEpidemiologic Catchment Area (ECA) Study. JAMA. 1990;264:2511-2518.5. Kessler RC, McGonagle KA, Zhao S, et al. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the Unit-ed States. Arch Gen Psychiatry. 1994;51:8-19.6. Frye MA, Altshuler LL, McElroy SL, et al. Gender differences inprevalence, risk, and clinical correlates of alcoholism comorbid-ity in bipolar disorder. Am J Psychiatry. 2003;160:883-889.7. Feske U, Frank E, Mallinger AG, et al. Anxiety as a correlate ofresponse to the acute treatment of bipolar I disorder. Am J Psy-chiatry. 2000;157:956-962.8. Frank E, Cyranowski JM, Rucci P, et al. Clinical significance oflifetime panic spectrum symptoms in the treatment of bipolar Idisorder. Arch Gen Psychiatry. 2002;59:905-911.9. McElroy SL, Frye MA, Suppes T, et al. Correlates of overweightand obesity in 644 patients with bipolar disorder. J Clin Psychi-atry. 2002;63:207-213.10. Baldessarini RJ, Tondo L, Hennen J. Lithium treatment andsuicide risk in major affective disorders: Update and new findings.J Clin Psychiatry. 2003;64:44-52.11. Baldessarini RJ, Tondo L. Suicide risk and treatments forpatients with bipolar disorder. JAMA. 2003;290:1517-1519.12. Katon WJ. Clinical and health services relationships betweenmajor depression, depressive symptoms, and general medical ill-ness. Biol Psychiatry. 2003;54:216-226.13. Bebbington P. Recent findings in bipolar affective disorder(Editorial). Psychol Med. 2004;34:767-776.14. McGuffin P, Rijskijk F, Andrew M, Sham P, Katz R, Cardno A.The heritability of bipolar affective disorder and the genetic rela-tionship to unipolar depression. Arch Gen Psychiatry. 2003;60:497-502.15. Mortensen PB, Pedersen CB, Melbye M, Mors O, Ewald H.Individual and familial risk factors for bipolar affective disordersin Denmark. Arch Gen Psychiatry. 2003;60:1209-1215.16. Blumberg HP, Kaufman J, Martin A, et al. Amygdala andhippocampal volumes in adolescents and adults with bipolar dis-order. J Affect Disord. 2003;79:97-103.17. Ferrier IN, Stanton BR, Kelly TP, Scott J. Neuropsychologi-cal function in euthymic patients with bipolar disorder. Br J Psy-chiatry.1999;175:246-251.18. Deckersbach T, McMurrich S, Ogutha J, Savage CR, Sachs G,Rauch SL. Characteristics of non-verbal memory impairment inbipolar disorder: the role of encoding strategies. Psychol Med.2004;34:823-832.19. Pinkham AE, Penn DL, Perkins DO, Lieberman J. Implica-tions for the neural basis of social cognition for the study ofschizophrenia. Am J Psychiatry. 2003;160:815-824.20. Phillips ML. Understanding the neurobiology of emotion per-ception: implications for psychiatry. Am J Psychiatry. 2003;182:190-192.21. Weinberger DR. Biological phenotypes and genetic researchon schizophrenia. World Psychiatry. 2002;1:2-6.22. Ehlers CL, Frank E, Kupfer DJ. Social zeitgebers and bio-logical rhythms: a unified approach to understanding the eti-

ology of depression. Arch Gen Psychiatry. 1988;45:948-952.23. Ehlers CL, Kupfer DJ, Frank E, Monk TH. Biological rhythmsand depression: the role of zeitgebers and zeitstörers. Depression.1993;1:285-293.24. Malkoff-Schwartz S, Frank E, Anderson B, et al. Stressfullife events and social rhythm disruption in the onset of manicand depressive bipolar episodes: a preliminary investigation. ArchGen Psychiatry. 1998;55:702-707.25. Klerman GL, Weissman MM, Rounsaville BJ, Chevron ES.Interpersonal Psychotherapy of Depression. New York, NY; BasicBooks, Inc: 1984.26. Frank E, Swartz HA, Kupfer DJ. Interpersonal and socialrhythm therapy: managing the chaos of bipolar disorder. Biol Psy-chiatry. 2000;48:593-604.27. Harrow M, Goldberg JF, Grossman LS, Meltzer HY. Outcomein manic disorders: a naturalistic follow-up study. Arch Gen Psy-chiatry.1990;47:665-671.28. Markar HR, Mander AJ. Efficacy of lithium prophylaxis in clin-ical practice. Br J Psychiatry. 1989;155:496-500.29. Prien RF, Kupfer DJ, Mansky PA, et al. Drug therapy in theprevention of recurrences in unipolar and bipolar affective dis-orders: report of the NIMH Collaborative Study Group compar-ing lithium carbonate, imipramine, and a lithium carbonate-imi-pramine combination. Arch Gen Psychiatry.1984;41:1096-1104.30. Hyman SE. Goals for research on bipolar disorder: the viewfrom NIMH. Biol Psychiatry. 2000;48:436-441.31. Cookson JC, Moult PJ, Wiles D, Besser GM. The relationshipbetween prolactin levels and clinical ratings in manic patientstreated with oral and intravenous test doses of haloperidol. Psy-chol Med. 1983;13:279-285.32. American Psychiatric Association. Practice guideline for thetreatment of patients with bipolar disorder (revision). Availableat: http://www.psych.org/psych_pract/treat/pg/prac_guide.cfm.Assessed February 21, 2004.33. Goodwin GM. Evidence-based guidelines for treating bipolardisorder: Recommendations from the British Association forPsychopharmacology. J Psychopharmacol. 2003;17:149-173.34. Kupfer DJ, Frank E, Grochocinski VJ, et al. Stabilization inthe treatment of mania, depression and mixed states. Acta Neu-ropsychiatrica. 2000;12:110-114.35. Suppes T, Rush AJ, Dennehy EB, et al. Texas Medication Algo-rithm Project, Phase 3 (TMAP-3): clinical results for patients witha history of mania. J Clin Psychiatry. 2003;64:370-382.36. Tohen M, Vieta E, Calabrese J, et al. Efficacy of olanzapine andolanzapine-fluoxetine combination in the treatment of bipolar Idepression. Arch Gen Psychiatry. 2003;60:1079-1088.37. Gijsman HJ, Geddes JR, Rendell JM, Nolen WA, Goodwin GM.Antidepressants for bipolar depression: a systematic review of ran-domized, controlled trials. Am J Psychiatry.2004;161:1537-1547.38. Angst F, Stassen HH, Clayton PJ, Angst J. Mortality of patientswith mood disorders: follow-up over 34-38 years. J Affect Disord.2002;68:167-181.39. Geddes JR, Burgess S, Hawton K, Jamison K, Goodwin GM.Long-term lithium therapy for bipolar disorder: systematic re-view and meta-analysis of randomized controlled trials. Am JPsychiatry. 2004;161:217-222.40. Bowden CL, Calabrese JR, McElroy SL, et al. A randomized,placebo-controlled 12-month trial of divalproex and lithium intreatment of outpatients with bipolar I disorder. Divalproex Main-tenance Study Group. Arch Gen Psychiatry. 2000;57:481-489.41. Calabrese JR, Bowden CL, Sachs G, et al. A placebo-controlled18-month trial of lamotrigine and lithium maintenance treat-ment in recently depressed patients with bipolar I disorder. J ClinPsychiatry. 2003;64:1013-1024.42. Bowden CL, Calabrese JR, Sachs G, et al. A placebo-controlled18-month trial of lamotrigine and lithium maintenance treat-ment in recently manic or hypomanic patients with bipolar I dis-order. Arch Gen Psychiatry. 2003;60:392-400.



46. Tohen M, Bowden C, Calabrese J, et al. Olanzapine’s efficacyfor relapse prevention in bipolar disorder: a randomized double-blind placebo-controlled 12-month clinical trial. Bipolar Disord.2003;5(S1):89(p200).47. Tohen M, Chengappa KN, Suppes T, et al. Relapse preventionin bipolar I disorder: 18-month comparison of olanzapine plusmood stabiliser v. mood stabiliser alone. Br J Psychiatry. 2004;184:337-345.48. Miklowitz DJ, George EL, Richards JA, Simoneau TL, Sud-dath RL. A randomized study of family-focused psychoeducationand pharmacotherapy in the outpatient management of bipolardisorder. Arch Gen Psychiatry. 2003;60:904-912.

43. Keck PE, Sanchez R, Marcus RN, et al. Aripiprazole for re-lapse prevention in a 26-week placebo-controlled trial. Present-ed at: 157th American Psychiatric Association Annual Meeting;May 1-6, 2004; New York, NY. Abstract NR746.44. Tohen M, Ketter TA, Zarate CA, et al. Olanzapine versus dival-proex sodium for the treatment of acute mania and maintenanceof remission: a 47-week study. Am J Psychiatry. 2003;160:1263-1271.45. Tohen M, Marneros A, Bowden C, et al. Olanzapine versuslithium in relapse prevention in bipolar disorder: a randomized,double-blind, controlled, 12-month clinical trial. Am J Psychia-try. In press.

DIAGNOSTIC ET TRAITEMENT DES TROUBLES BIPOLAIRES : TENDANCES ACTUELLES

L e trouble bipolaire (psychose maniacodépressive) est un sous-type destroubles de l’humeur chroniquement récurrent associé à une considé-rable comorbidité psychiatrique et médicale. La comorbidité psychia-

trique, telle que l’abus d’alcool ou autres substances, l’anxiété et les troublespaniques influent défavorablement sur l’évolution. Une affection médicale etdes facteurs de risque médicaux sont fréquents au cours des troubles bipolaireset conduisent à une morbidité et une mortalité encore plus importantes danscette population. Associé à un taux élevé de suicide, le trouble bipolaire est unemaladie potentiellement létale. À la lumière des attitudes thérapeutiques adap-tées aux troubles bipolaires, il est plus facile de les regrouper en traitement pourles états maniaques ou mixtes aigus, traitement de la dépression aiguë et trai-tement préventif de la récurrence dépressive et maniaque. La monothérapie,habituellement administrée sous la forme de lithium, de valproate ou d’un mé-dicament antipsychotique atypique, est recommandée dans un premier tempspour le traitement des états maniaques et mixtes. Le traitement de la dépres-sion aiguë est généralement plus long et plus difficile que celui des états ma-niaques. De nombreux cliniciens experts ont recommandé l’utilisation optimiséed’un régulateur de l’humeur et, si cette stratégie n’est pas efficace, l’associa-tion d’un antidépresseur au régulateur de l’humeur. La monothérapie par ré-gulateur de l’humeur était considérée autrefois comme le traitement d’entre-tien idéal ; toutefois, l’intérêt se porte désormais de plus en plus sur des étudescontrôlées de médicaments antipsychotiques atypiques dans la phase d’entre-tien. L’utilisation croissante des antipsychotiques atypiques et leur possible ef-ficacité, non seulement pour stabiliser les états maniaques, mais aussi pour lasymptomatologie dépressive, constitue un important champ d’investigationpour les orientations ultérieures. Enfin, l’utilisation d’agents ayant une actionrégulatrice sur les rythmes biologiques mérite d’être explorée plus avant.

✦

261What are the core symptoms of depression? MEDICOGRAPHIA, VOL 27, No.3, 2005

Core symptoms of depressive illnessesDepression is characterized by the followingsymptoms, whether in developing countries orin other countries.◆ Lowering of mood, volition and outlook. Thepatient has a negative outlook, a low mood, afeeling of loss, of despondency. Everything is col-orless, negative, gray, and gloomy.◆ Inability to deal with normal work, play, andsocial obligations satisfactorily.◆ Preoccupation with bodily functions and phys-ical symptoms. Patients often want their doctorto prove they have a physical disease that is notthere.◆ Sleep disturbances. Patients usually reportearly morning awakening, but many experiencedifficulty in falling asleep.◆ Disturbances in eating. Patients show a lackof interest in food and therefore usually loseweight; occasionally, they eat too much and puton weight. ◆ In severe and prolonged depressive illnessesthere may be suicidal ideas, a stuporous state. ❒

T he problem with core symptoms is theavailability of acceptable guidelines andclassifications of depressive illnesses. Many

nonpsychiatrist doctors consider any patient whois distressed as depressed, especially if the patientis crying! Suicide attempts or parasuicides arealways equated by nonpsychiatrists as depression.Persons who are irritable and angry and easilyprovoked are often seen as either normal or psy-chotic, but seldom considered to be depressed.Such looseness in diagnosis leads to injudicioususe and noneffectiveness of antidepressant medi-cines. Thus, a review of the current concepts inrecognizing depressive illnesses and treating themis needed. The answer lies in a good history takingthrough a good interview of the patient. The coresymptoms include a history of at least 3 weeks(but usually 6 weeks) of low mood and negativethoughts, gloomy outlook on everyday topics,sleep patterns disturbed without reason, lack ofinterest in eating, some loss of weight, poor inter-est in work, recreation, sexual function. To pa-tients, everything looks gray, hopeless, useless,they have vague thoughts of leaving their job,going somewhere else, or even “ending it all.”

C O N T R O V E R S I A L Q U E S T I O N

1◆ M. P. Deva, Brunei

What are the core symptoms of depression?

M. Parameshvara DEVA, MDFRCPsych, FRANZCP, FAMMHon Professor of PsychiatryUPNG, SSB Hospital Kuala Belait, KA 1131BRUNEI(e-mail: [email protected])

put forward in favor of this. For instance, depres-sive disorders and generalized anxiety disordertend to be related to the same genetic factors andmay be interpreted as different manifestations ofthe same predisposition.1 This predisposition couldinvolve mutations in the so-called circadian clockgenes and result in errors in regulation of the cir-cadian pacemaker.2 This would lead to abnormalcircadian rhythms and alteration of the sleep-wake cycle, resulting in changes in core bodytemperature and short latencies to rapid eyemovement (REM) sleep,3-6 the outcome of whichis compromised mood regulation and ultimately,feelings of depression and anxiety. One way tolook at the question is to analyze the individualdomains that may show differing trajectories rath-er than the overall severity of depression reflectinga summation of these domains.7 Severe depres-sions are characterized by two symptoms onlythat commonly occur in both typical and atypi-cal depressive disorders. These key symptoms—

D epression is heterogeneous in terms ofetiology and phenotype. Depressions aremany; they vary according to profile and

intensity of symptoms, and course and cause ofillness. Symptoms may be melancholic, atyp-ical, psychotic, or mild; there may be one or nu-merous episodes; the course may be recurrentor chronic; the depression may be reactive or en-dogenous in nature. Thus, trying to pin down thecore of depression is like peeling an onion. Theessential characteristic of depression is the feel-ing of loss: the core symptoms of depressive dis-order are therefore lack of pleasure and sadness,in other words, anhedonia and depressed mood.These constitute the key criteria for a major de-pressive episode, not only in terms of diagnosis,but also of evolution. However, this point of viewhas been challenged by recent findings suggestinga need to focus on core pathways in the patho-genesis of illness rather than on manifestationsof individual disorders. Several arguments can be

2◆ T. Partonen, Finland

Timo PARTONEN, MDAssociate Professor ofPsychiatry at the Universityof Helsinki, andAcademy Research Fellow at the National Public HealthInstituteDepartment of Mental Healthand Alcohol ResearchHelsinki, FINLAND(e-mail:[email protected])


262 What are the core symptoms of depression?MEDICOGRAPHIA, VOL 27, No.3, 2005

9. Partonen T, Lönnqvist J. Seasonal affective disorder. Lancet.1998;352:1369-1374.10. Whooley MA, Avins AL, Miranda J, Browner WS. Case-find-ing instruments for depression: two questions are as good asmany. J Gen Intern Med. 1997;12:439-445.11. Timbremont B, Braet C, Dreessen L. Assessing depressionin youth: relation between the Children’s Depression Inventoryand a structured interview. J Clin Child Adolesc Psychol.2004;33:149-157.12. Teissedre F, Chabrol H. Detecting women at risk for post-natal depression using the Edinburgh Postnatal Depression Scaleat 2 to 3 days postpartum. Can J Psychiatry. 2004;49:51-54.13. Watson LC, Pignone MP. Screening accuracy for late-lifedepression in primary care: a systematic review. J Fam Pract.2003;52:956-964.14. Gabryelewicz T, Styczynska M, Pfeffer A, et al. Prevalenceof major and minor depression in elderly persons with mildcognitive impairment—MADRS factor analysis. Int J GeriatrPsychiatry. 2004;19:1168-1172.15. Rush AJ, Trivedi MH, Carmody TJ, et al. Self-reported de-pressive symptom measures: sensitivity to detecting changein a randomized, controlled trial of chronically depressed,nonpsychotic outpatients. Neuropsychopharmacology.2005;30:105-116.16. Scott J, Thorne A, Horn P. Quality improvement report:Effect of a multifaceted approach to detecting and managingdepression in primary care. BMJ. 2002;325:951-954.

REFERENCES1. Vollebergh WA, Iedema J, Bijl RV, de Graaf R, Smit F, Ormel J.The structure and stability of common mental disorders: theNEMESIS study. Arch Gen Psychiatry. 2001;58:597-603.2. Johansson C, Willeit M, Smedh C, et al. Circadian clock-re-lated polymorphisms in seasonal affective disorder and theirrelevance to diurnal preference. Neuropsychopharmacology.2003;28:734-739.3. Schulz H, Lund R. Sleep onset REM episodes are associatedwith circadian parameters of body temperature: a study in de-pressed patients and normal controls. Biol Psychiatry. 1983;18:1411-1426.4. Avery DH, Wildschiodtz G, Smallwood RG, Martin D, Rafael-sen OJ. REM latency and core temperature relationships inprimary depression. Acta Psychiatr Scand. 1986;74:269-280.5. Czeisler CA, Kronauer RE, Mooney JJ, Anderson JL, Allan JS.Biologic rhythm disorders, depression, and phototherapy: anew hypothesis. Psychiatr Clin North Am. 1987;10:687-709.6. Schwartz PJ, Rosenthal NE, Kajimura N, et al. Ultradianoscillations in cranial thermoregulation and electroencephalo-graphic slow-wave activity during sleep are abnormal in humanswith annual winter depression. Brain Res. 2000;866:152-167.7. Parker G, Roy K. The development of a six-item daily self-report measure assessing identified depressive domains. J AffectDisord. 2003;73:289-294.8. Sullivan PF, Kessler RC, Kendler KS. Latent class analysisof lifetime depressive symptoms in the national comorbiditysurvey. Am J Psychiatry. 1998;155:1398-1406.

flating scores on the routinely administered ratingscales.11-13 For example, depressive elderly personswith mild cognitive impairment tend to reportsadness, anhedonia, and pessimistic thoughts.14

Self-reports of depression may be of benefit interms of both expense and time, and a valuablesubstitute for clinician-based ratings in treatmenttrials of outpatients with major depressive disor-der, but no psychotic symptoms or cognitive im-pairment.15 Global ratings, however, seem to pro-duce less valid findings than specific item-basedratings. To ensure optimal control of the coresymptoms of depression, an antidepressant shouldhave mood-lifting and energizing effects, in otherwords it needs to achieve elation and activation.Rapid relief from anxiety and insomnia are nat-urally of great help for a good outcome. In thelong run, the patient will recover feelings of plea-sure and optimistic thoughts. However, it is neces-sary to realize that no treatment—not even themost potent one—stands a true chance of successif resources in terms of health-care professionalsare insufficient to ensure an adequate follow-upof patients suffering from depression.16 ❒

feelings of depression and loss of interest—arereported by 9 patients out of 10.8 Are these thecore symptoms of depression? Not necessarily,since in addition to these two symptoms, severetypical depression is characterized by difficultyin concentrating, insomnia, and loss of pleasure.Moreover, major depressive episodes with a sea-sonal pattern (seasonal affective disorder) arefrequently accompanied by atypical symptoms ofdepression such as prolonged sleep, weight gain,carbohydrate craving, and increased appetite.9

Nevertheless, a simple two-question screening toolinquiring about the presence of depressed moodand anhedonia may be just as effective as a morecomplex questionnaire.10 Despite its limitations,such a screening tool is certainly a useful, morepragmatic, and less time-consuming method fordetecting depression in the primary care setting.Specific groups of individuals with depressive dis-order, including children and adolescents, wom-en after delivery, and the elderly, need specificattention in clinical practice, since their clinicalpicture may differ, and therefore can be missed bythe usual case-finding instruments, thereby de-



disorders, thus decreasing the possibility of de-tecting differences in underlying etiopathology.In all, the drawbacks of the DSM system restrictthe number and power of the studies investigat-ing the heterogeneity of depressive disorders. Allsymptoms are the end-product or by-product ofpathological or functional changes in organs. Thebrain is an organ with highly specialized functionsassigned to different areas of the hemispheres.Unlike most other organs, expansion or a slightchange in the location of pathology can result invery different clinical presentations. In regard todepression, great variability of symptoms andhigh frequency of comorbid diseases could wellbe related to the extent of the areas affected. Todate, although much remains to be learned, we dohave some clues about the possible brain loca-tions that control some of functions that are af-fected in depression. For example, depressed moodpossibly originates in increased activity of thelimbic cortex, specifically, the anterior cingulatecortex and the limbic orbital cortex. Abnormalneuronal activity in the hypothalamus and otherlimbic areas could lead to anhedonia/loss of in-terest, decreased libido, and dysregulation of ap-petite and sleep. We also know some of the areasinvolved in other psychiatric disorders as well.For example, the hippocampal region is affectedin dementia, and mesolimbocortical dopaminer-gic pathways and the prefrontal cortex are af-fected in schizophrenia. I fully expect that, infuture, new syndromes will be defined accordingto the affected brain areas and circuits, and thatthis will lead to a thorough overhaul of currentdiagnostic criteria for psychiatric disorders. Inconclusion, although we still need to rely on“core” and other symptoms for the differentialdiagnosis of depression, we should also bear inmind that the current conceptualization of de-pression limits our understanding of depressionand how to treat it. Once the diagnosis of depres-sion has been established, all the symptoms andsigns presented by the patient should be consid-ered as “core” symptoms. Every single symptomshould receive clinical attention regardless ofseverity. We should rethink our current partiallysuccessful concept of medical treatment that con-siders depression as a whole. Instead, we shouldfocus on individual symptoms and new thera-peutic strategies to address them. ❒

REFERENCES1. Hamilton M. Development of a rating scale for primarydepressive illness. Br J Soc Clin Psychol. 1967;6:278-296.2. American Psychiatric Association. Diagnostic and Statisti-cal Manual of Mental Disorders. 4th ed. Washington, DC:American Psychiatric Association; 1994.3. Douki S, Tabbane K, Taktak MJ. Cross-cultural aspects ofdepressive disorders. Eur Psychiatry. 1997;12(suppl 2):124s.

D epression has been a long-standing focusof attention for writers, poets, philoso-phers, artists, and, eventually, scientists.

The theories proposed to explain depression haveshed light on the many facets of the disease andcontributed to increase our awareness, under-standing, and conceptualization of it. Thus, Freudstressed the importance of grieving in depression,in the absence of any obvious loss. Beck, a psy-choanalyst and the father of the Depression In-ventory that bears his name, started investigatingthe role of aggression directed toward self in de-pression, and ended up by formulating a cognitivetheory of depression. Interestingly, almost alltheories have emphasized two symptoms of de-pression; depressed mood and anhedonia/loss ofinterest. These symptoms are also included in thecore of Hamilton’s conceptualization of depres-sion as presented in the Hamilton Depression Rat-ing Scale.1 Eventually, the Diagnostic and Statis-tical Manual of Mental Disorders (DSM) approachhighlighted depressed mood and anhedonia/lossof interest as the cardinal features of depression.2

Since the third edition of the DSM in 1980, thedifferences among cultures have gradually sub-sided, mainly due to the widespread acceptanceof the conceptual framework enshrined in theDSM classification system. Recent studies indicatethat, although culture can affect the expressionof depressive mood and symptoms, there appearsto be a “core syndrome” of depression, regardlessof sociocultural context.3 Current classificationsystems consider depressed mood and anhedonia/loss of interest as the “core” symptoms of depres-sion. Classification systems relying on clinicalpresentation need to identify “core” symptoms todifferentiate individual “disorders,” which areaccordingly defined. However, major drawbacksare associated with the DSM approach regardingthe conceptualization of depression, and, as theetiopathology of depressive disorders is increas-ingly elucidated, the concept of “core” symptomstends to be replaced with etiologically relevantsymptom profiles. First of all, the DSM approachrelies heavily on “clinical” observations, and interms of etiology—the ultimate goal of any clas-sification—this hardly allows to go beyond syn-drome level. Secondly, the description of depres-sion is best adapted to middle-aged patientswithout comorbid psychiatric or physical disor-ders. Even when only anxiety disorders are takeninto account, at least 50% of depressive patientspresent with symptoms of anxiety that interactwith the symptomatology produced by depression.Last but not least, in describing a common “ma-jor depressive episode” syndrome, the DSM em-phasizes similarities more than differences among

3◆ A. Soykan, Turkey

Attila SOYKAN, MDAssociate ProfessorDivision of ConsultationLiaison PsychiatryDepartment of PsychiatryAnkara UniversitySchool of MedicineAnkaraTURKEY



in the same patient (this is more likely to happenin the most severe forms), but more often someaspects (besides depressed mood or loss of inter-est or pleasure) are predominant, particularly inthe early phase, where some patients can be “pau-cisymptomatic,” with a more prominent psychicor somatic or cognitive expression.2,3 However,along with the development of a full-blown majordepressive syndrome, all core or nuclear symp-toms tend to occur. Thus, depression is like therings that a pebble makes in a pond, which en-large progressively from the center (depressedmood and loss of interest), to progressively includemore and more symptoms (psychic, somatic, andcognitive). In other words, depression evolves froman initial “paucisymptomatological” presentationto the full-blown, multisymptomatic form usu-ally seen by the GP. Psychiatrists usually see pa-tients at a later stage, after weeks of suffering, ifthey are not adequately diagnosed and treated byGPs.4 It is of paramount importance that GPs beaware that patients with a monosymptomaticpresentation (depressed mood/anxiety and reducedinterest or anhedonia), even though the 2-weekcriterion is not fulfilled or other symptoms are notpresent, may present subclinical signs that canquickly develop into the core or nuclear symp-toms of a major depressive episode. The final pointwe must address is whether an antidepressantshould be active on the core symptoms of depres-sion.5,6 We favor a “holistic” approach to the psy-chobiology of depression taking into accountdysfunctions of the serotonergic (5-HT) or nora-drenergic (NA) pathways, as well as the fine reg-ulation of disturbed biological rhythms involvedin the pathogenesis of depression (such as thehypothalamopituitary axis [HPA], melatonin,etc), which are responsible for diurnal variationsin mood, sleep disorders, anxiety, neurovegetativesymptoms, etc.7 Future approaches should ad-dress not only the classic monoaminergic path-ways, but also other neurotransmitters and/orneuromodulators involved in the dysregulation ofbiological rhythms, which influence, at least inpart, the pathophysiology and outcome of majordepression.8,9 ❒

I t is well known that the word depression refersboth to a full-fledged clinical condition andto brief, mild downward mood swings that

we all experience as a part of daily living. In theclinical context, the term depression refers notsimply to a state of depressed mood, but to a syn-drome comprising mood disorders, psychomotorchanges, and various types of somatic (neuroveg-etative) dysfunction and cognitive impairment.1

According to the 4th edition of the Diagnostic andStatistical Manual of Mental Disorders (DSM-IV),all these changes may occur (they should be pre-sent for at least 2 weeks), but none except “de-pressed mood” and “loss of interest or pleasure innearly all activities,” are essential for the diagno-sis of Major Depression. DSM-IV and the Inter-national Statistical Classification of Diseases andHealth-Related Problems, 10th Revision (ICD-10)show some degree of divergence in their criteriafor depression, regarding type and number ofsymptoms. ICD-10 contains 10 items, in contrastto the 9 DSM-IV items (loss of self-esteem is in aseparate category from inappropriate guilt inICD-10). Also, in ICD-10, the severity of MajorDepressive Episodes is rated according to an in-cremental number of criteria: mild depressionstarts from 4 out of 10 symptoms, moderate de-pression from 6 out of 10 symptoms, and severedepression from 8 out of 10 symptoms. Finally, inICD-10, the diagnostic algorithm differs by re-quiring the presence of at least 2 out of the follow-ing 3 symptoms: depressed mood, loss of interest,and decreased energy for mild and moderate de-pressive episodes, and all 3 for severe depressiveepisodes. ICD-10 episodes with psychotic featuresexclude first-rank symptoms and bizarre delu-sions. In terms of core or nuclear symptoms, bothDSM-IV and ICD-10 include the following symp-toms: (i) psychic (eg, mood or anxiety or suicidalideation and guilt feelings); (ii) psychomotor (eg,retardation or agitation); (iii) somatic/neuroveg-etative (eg, sweating, tachycardia, dryness ofmouth); (iv) diurnal variation; (v) insomnia/hy-persomnia; (vi) loss of energy; and (vii) cognitivesymptoms (eg, difficulty concentrating and lossof memory). All these components can be present

4◆ A. C. Altamura, Italy

A. Carlo ALTAMURA, MDChairman Department ofPsychiatryCenter for Diagnosis andTreatment of Mood DisordersUniversity of MilanHospital Luigi SaccoMilan, ITALY(e-mail:[email protected])

long term treatment of recurrent depression: rationale, cur-rent methodologies and future directions. J Clin Psychiatry.1993;54(suppl 8):29-37. 6. Paykel ES. Treatment of depression: the relevance of researchfor clinical practice. Br J Psychiatry. 1989;155:754-763.7. Maes M, Meltzer HY. The serotonin hypothesis of major de-pression. In: Maes M, Meltzer HY, eds. Psychopharmacology:The Fourth Generation of Progress. New York, NY: RavenPress; 1994:933-944.8. Altamura AC, Carta MG, Carpiniello B, Piras A, Maccio MV,Marcia M. Lifetime prevalence of brief recurrent depression(results from a community survey). Eur Neuropsychophar-macol. 1995;5(suppl):99-102.9. Frank E, Prien RF, Jarret RB, et al. Conceptualization andrationale for consensus definitions of terms in major depres-sive disorder: remission, recovery, relapse, and recurrence.Arch Gen Psychiatry. 1991;48:851-855.

REFERENCES1. Altamura AC, Montresor C, Salvadori D, Mundo E. Doescomorbid subthreshold anxiety affect clinical presentation andtreatment response in depression? A preliminary 12-monthnaturalistic study. Int J Neuropsychopharmacol. 2004;7:1-7.2. Altamura AC. Anxious-depressive syndromes in the elderly.Assessment, clinical course and treatment. In: Racagni G,Smeraldi E, eds. Anxious Depression: Assessment and Treat-ment. New York, NY: Raven Press; 1987:209-216. 3. Lepine JP, Altamura AC, Ansseau M et al. Tianeptine andparoxetine in major depressive disorder, with a special focuson the anxious component in depression: an international, 6-week double-blind study. Hum Psychopharmacol Clin Exp.2001;16:219-227.4. Judd LL, Akiskal HS, Paulus MP. The role and clinical sig-nificance of subsyndromal depressive symptoms (SSD) in uni-polar major depressive disorder. J Affect Disord. 1997;45:5-17.5. Altamura AC, Percudani M. The use of antidepressants for



neurasthenic, autonomic and somatoform dis-orders, algias, sleep disturbances, mainly in theform of insomnia/hypersomnia), while the basicaffect manifestations (melancholy, psychomotordisturbances, ideas of guilt) are moderate. Inthese cases, the diagnosis of depression is madeeasier by evidence of a circadian rhythm, pes-simistic thoughts about the persons’ own healthand future, passive suicidal thoughts (desirabilityof an accident, sudden death, etc). The diagnos-tic criteria for seasonal depressions are based onchronobiological characteristics, as these typesof affective disorders manifest during the autumnand winter period. During spring and summer,depressions can be followed by a remission orhypomania. The peculiarities of the psychopatho-logical structure of seasonal depressions shouldbe taken into account. Two types of somatovege-tative symptom complex can be distinguished inseasonal depressions: hyperesthetic (increase inappetite, including appetency for carbohydrate-rich meals, increase in weight, hypersomnia) andanesthetic (reduction in appetite and taste, ac-com-panied by a nutrition deficit and consider-able loss of weight, insomnia with absence offeeling of sleep and reduction of sleep duration).The choice of antidepressants is based on thepsychopathological manifestations of depressionsas describe above. Effective control of typical de-pression with basic symptomatology is achievedwith the majority of current thymoanaleptics.Severe depression with agitation or marked psy-chomotor retardation requires use of potent psy-chotropic medications: injectable tricyclic anti-depressants are preferable. In mild or moderatedepressions, as well as atypical depressions, withpredominance of negative affectivity, the recentantidepressants are the drugs of choice. Com-pared with tricyclic antidepressants these aremuch more neurochemically selective and pro-duce fewer side effects (selective stimulators andinhibitors of serotonin reuptake, fluoxetine,paroxetine, etc; dual action antidepressants, in-cluding venlafaxine, mirtazapine, etc; selective,reversible monoamine oxidase A–type inhibitorslike pirlindole, or other antidepressants like tia-neptine). Finally, the acute treatment and pre-vention of seasonal affective disorder, which isassociated with dysfunction of the serotonergicsystem and pineal gland, as well as of other de-pressions characterized by abnormal biologicalrhythms, relies on new therapeutic strategiesinteracting with circadian rhythms and othertargets. ❒

T he concept of “depression” embraces a widerange of states of pathologically changedaffects (from typical melancholic forms to

atypical, masked, and seasonal, depressions). Thecore symptoms of depression determine the patho-logical course and the type of response to thera-py, and form the basis of the psychopathologicalstructure of depression. These manifestations ofdepression are also the main measures of compli-ance used to evaluate the action of psychotropicdrugs (target symptoms). The basic symptoms oftypical depression include positive affectivity man-ifestations (mentally oppressive depressive hyper-esthesia (eg, heartache, suffering) or somatic dis-orders, which determine the clinical presentationof typical (vital) depressions. These core diagnos-tic features include: (i) melancholy, ie, an uncer-tain (protopathic) feeling of intolerable pressurein the chest or epigastrium (precardiac or epi-gastric melancholy) accompanied by low spirits,despondency, despair, hopelessness; (ii) intellec-tual and motor retardation, ie, difficulties in con-centration and attention, delayed responses, slow-ness of movements, loss of spontaneous activity;(iii) pathological circadian rhythm, ie, moodswings during the day with maximum symptomseverity early in the morning; (iv) ideas of self-inferiority, ie, persisting thoughts of self-useless-ness and self-accusation, depreciation of personalsuccesses in the past, pessimistic thoughts aboutthe future; and (v) suicidal thoughts, ie, psycho-logically not derivable wish to die, thoughts thatlife is not worth living, suicidal ideation, occasion-ally an irresistible urge to commit suicide). Thediagnosis of atypical depressions is based on neg-ative affectivity manifestations, involving mentalalienation (persisting feeling of impoverishedand altered mental life, loss of motivation to anyactivity). The symptom complexes of negative af-fectivity include: (i) anaesthesia psychica dolorosa,ie, a painful feeling of loss of emotions, inabilityto feel love, hatred, empathy, anger; (ii) depres-sive devitalization, ie, loss of life instinct, instinctof self-preservation, vital inclinations (sleep, ap-petite, libido); (iii) apathy, ie, appetency deficitaccompanied by a loss of zest for life, lack of en-ergy, indifference to everything; (iv) dysphoria,gloominess, “anger attacks,” querulousness; and(v) anhedonia, ie, inability to experience plea-sure, joy, and delight. The difficulty of detectingmasked depressions (or “latent,” “alexithymic”),which are prevalent in general medical practice,is related to the predominance of neurotic symp-tom complexes (anxiety, phobic, hypochondriac,

5◆ A. B. Smulevitch, Russia

Anatoly B. SMULEVITCH,MDNational Mental HealthResearch CenterKachirskoe Shosse 34115522 MoscowRUSSIA



Statistical Manual of Mental Disorders (DSM-IV)required at least either depressed mood or loss ofinterest to meet the criteria for Major DepressiveEpisode.7 Cognitive symptoms (diminished abilityto think, concentrate, or take decisions) were alsogiven a major role. It was specified that depres-sive symptoms vary with the branch of medicineconcerned (in general medicine patients with de-pressive disorder report more pain and somaticillness than others) and age (predilections for ir-ritability and social withdrawal in children, im-pulsiveness and toxic substance use in adoles-cents, and cognitive impairment in the elderly).Disruption of circadian rhythm as a major symp-tom of depression appears only in the form ofsleep disturbance (maintenance insomnia, earlymorning awakening insomnia, or, conversely,hypersomnia).

Response to antidepressants and clinical profileAntidepressant response profiles appear to differwith the clinical characteristics of the depression:monoamine oxidase inhibitors in atypical de-pression (mood reactivity, increased appetite orbody weight, hypersomnia, impression of limbanesthesia, oversensitivity to rejection); tricyclicsand electroconvulsive therapy in melancholicdepression; selective serotonin reuptake inhibitorsin nonmelancholic depression.8 Depression withanxiety and sleep disturbance appears sensitiveto the more sedative antidepressants.

Neuroimaging and understanding of the patho-physiology of depressive symptomsThe key to future breakthroughs could lie in theability to identify specific symptoms with certainneuronal circuits and develop substances to targetthose circuits.9 The demonstration of structuresmediating the self-referential processing of emo-tional stimuli10 and of corticosubcortical circuitsinvolved in depression justifies further researchinto the connections between clinical symptomsand functional neurology.11,12 Depressive symp-toms may thus be divided into spheres:◆ Emotion: depressed mood, self-depreciation,guilt.◆ Motivation: loss of interest, loss of affect, psy-chomotor retardation.◆ Association: abnormal executor function, witheach sphere being concerned by a neighboring,but different, neuronal circuit.A pathophysiologic approach to depression bringsthe clinical targets of antidepressants into sharperfocus. Depression is a clinically mixed syndromewith core characteristics that are recognized byall clinicians. The next step is to establish clini-coneurophysiologic correlates that optimize thedeployment of our therapeutic arsenal. ❒

W hat is the epidemiologic and historicalbackground to a modern definition ofdepression?

Given its huge epidemiologic variation, with pointprevalences of Major Depressive Episode in thegeneral population over 1 year ranging from 0.6%in Taiwan through 5% in France to 10% in theUnited States, it is essential to define the coresymptoms of depression if we are to identify a nu-cleus of depression for use as a model in clinicalpsychiatric research and operating criteria forappropriate antidepressant treatment. The defini-tion of depression is grounded in Ey’s descriptionof melancholia as a state of “intense depressionexperienced with psychic pain and characterizedby psychomotor inhibition.”1 The concepts ofneurasthenia, neurotic depression, and anxietyand depression were then coined to characterizesyndromes falling outside this definition. Sincethese comprised symptoms such as insomnia,functional somatic complaints and anxiety, clin-icians were led to differentiate between “endoge-nous” and “psychogenic” depression—only forthis distinction to be swept aside after 1960 by thearrival of the first antidepressants that provedactive against all forms of depression, irrespec-tive of their cause. International classificationsystems subsequently jettisoned the concept ofendogenicity so dear to Kraepelin.2

Core symptoms of depressionHowever, a number of psychiatric manuals differ-entiate between endogenous depression (whetherunipolar or manic-depressive bipolar) and a muchmore mixed presentation extending from reac-tive depression to neurotic depression.3 Clinicalstudies have identified the following core symp-toms across a multiplicity of presentations 4,5:◆ Depressed mood (expressed in its most severeform as convictions of unworthiness, incurability,and guilt).◆ Inhibition or loss of élan vital (asthenia, globaldisinterest, loss of initiative, impaired intellectualperformance and motor function).◆ Somatic symptoms (anorexia, weight loss, con-stipation, sleep disturbance).◆ Anxiety and personality disturbances (irritabil-ity, impulsiveness, social withdrawal).◆ Ideas of death and suicidal thoughts (often as-sociated with depressed mood).◆ Delusional ideas (in the most severe forms ofdepression): mood-congruent (guilt, ruin, mourn-ing, misfortune, hypochondria) or mood-incon-gruent (ideas of influence and possession).6

The introduction of international criteria advancedthe understanding of depression by assemblingexperts who reanalyzed the literature data anddrew on new studies. It was thus no surprise when,in 1994, the 4th edition of the Diagnostic and

6◆ B. Millet, France

Bruno MILLET, MDService Universitaire dePsychiatrieCHS Guillaume Régnier108 avenue du GénéralLeclerc35011 Rennes CedexFRANCE(e-mail: [email protected])



7. American Psychiatric Association. Diagnostic and StatisticalManual of Mental Disorders (DSM-IV). 4th ed. Washington, DC:American Psychiatric Association; 1994.8. Esposito K, Goodnick P. Predictors of response in depres-sion. Psychiatr Clin N Am. 2003;26:353-365.9. LeDoux JE. Synaptic Self. New York, NY: Viking; 2001.10. Fossati P, Hevenor SJ, Graham SJ, et al. In search of theemotional self: an fMRI study using positive and negativeemotional words. Am J Psychiatry. 2003;160:1938-1945.11. Seminowicz DA, Mayberg HS, McIntosh AR, et al. Limbic-frontal circuitry in major depression: a path modeling meta-nalysis. Neuroimage. 2004;22:409-418.12. Mayberg HS. Positron emission tomography imaging indepression: a neural systems perspective. Neuroimaging ClinN Am. 2003;13:805-815.

REFERENCES1. Ey H, Bernard P, Brisset C. Manuel de Psychiatrie. Paris,France: Masson; 1963.2. Sartorius N. Classification: an international perspective.Psychiatr Ann. 1976;6:22-35.3. Lempérière T , Féline A, Gutman A, Adès J, Pilate C.Abrégé de Psychiatrie de l’Adulte. Paris, France: Masson;1977.4. Guelfi JD. Psychiatrie de l’Adulte. Paris, France: EditionsEllipses; 1985.5. Gelder M, Gath D, Mayou R. Oxford Textbook of Psychiatry.2nd ed. Oxford, UK: Oxford University Press; 1989.6. De Carvalho W, Cohen D. États dépressifs chez l’adulte. In:Olié JP, Poirier MF, Lôo H, eds. Les Maladies Dépressives. 2nded. Paris, France: Éditions Médecine Sciences Flammarion;2003:3-19.

▲

◆ Somatic: anorexia, asthenia, weight loss, sleepdisorders, pains, psychomotor retardation, etc.From a practical perspective, the high incidenceof depression has resulted in its being treated es-sentially at primary health care level. This patientpopulation has a lesser prevalence of dysphoriaor ideas of guilt, but a high incidence of fatigue,5

which has been proposed as the primordial symp-tom of major depression.6 We need to improve theearly diagnosis and recognition of depression inthis setting, for example, by using certain simpleepidemiological screening tools and evaluationscales (Primary Care Evaluation of Mental Disor-ders [PRIME-MD], Mini-Mental Status Examina-tion [MMSE], Zung Self-rating Depression Scale[SDS], etc).7,8 Physicians need to carry out appro-priately detailed clinical interviews that cover thefull psychopathological gamut of symptoms inorder to improve the detection rates of depression.Another factor that adds to the difficulties is thelack of stability in symptoms throughout thecourse of the depression. This was clearly shownin a recent study,9 with respect to both the con-tinuity of symptoms in successive recurrences andthe subtypes of depression. Variations based ongender have also been reported.10 Likewise, trans-cultural divergences have been pointed out whenstudying samples in different populations.11 How-ever, all of these studies suffer from having beenperformed on samples of severe patients, frequent-ly in hospital, which may not be very representa-tive of the usual setting in which depression oc-curs. One field of interest over recent years is thatof the importance of residual symptoms (fatigue,anxiety, sexual dysfunction, sleep disorders, etc),12

which may persist after partial relief from symp-toms following treatment of the episode. Theseresidual symptoms multiply the risk of relapse bythree, worsen social and occupational function-ing, favor chronic depression, and increase therisk of suicide. Although it is debated whetherthese residual symptoms are caused by comorbid-ity with personality or anxiety disorders, they are

T here are two possible presentations of de-pression:◆ Presence of one symptom (or a group of

symptoms) pathognomonic for depression, whichleads to an immediate diagnosis of depression; or◆ Absence of characteristic symptoms, in whichcase it is the ensemble of a more or less variableclinical picture that leads to diagnosis of thedisorder.The first presentation is the classic one in psychi-atry. Since Kurt Schneider,1 “vital sadness” hasbeen considered as a fundamental symptom ofdepression. Depressed mood is described as hav-ing a “different quality” than “normal sadness.”It is sadness that is nonreactive, internal, “cor-poralized,” persistent, not voluntarily changeable.For these authors, this symptom is the very coreof depression, it is depression itself. This type ofdepression is described as endogenous,2,3 and ischaracterized by inhibition, lack of emotional re-activity, and loss of interest and pleasure in things.The second presentation is increasingly at theforefront in current practice. Modern operatingclassifications have switched from the former en-dogenous (melancholy)/reactive (neurotic) dis-tinction to that between major depression anddysthymia. They postulate a sequential develop-ment of the different subtypes of depressionthroughout the subject’s evolution. These classifi-cations do not take into account the subtle phe-nomenological distinction of the “different quali-ty” of the depressive mood, which is so difficult todetect.4 It has been claimed that this devaluatesthe symptom, and blurs the limits between thesyndrome and the pathological entity and the dif-ferent types of mood disorder. However, the basicdiagnostic criteria continue to include an alter-ation of mood and a lack of pleasure, with variouscombinations of the following symptomatic com-plexes later completing the picture: ◆ Psychic: sadness, demoralization, lack of inter-est, negative thoughts, low self-esteem, decreasedattention and memory, etc, and

7◆ J. Saiz-Ruiz, Spain

Jerónimo SAIZ-RUIZ,MD, PhD Hospital Ramón y CajalHead, PsychiatricDepartment, andUniversity of AlcaláVice-Dean and Professor of PsychiatryCa Colmenar, km 9,128034 MadridSPAIN(e-mail:[email protected])

ical diagnoses are still lacking. Hopes have beenplaced on research into molecular genetics, neu-roimaging, and neuropsychology, and new dis-coveries are expected, which should lead to a bet-ter understanding and delimitation of depression,as well as to benefits in the field of antidepressanttreatment. ❒

now considered a prime target of antidepressanttreatment, with the aim of achieving total remis-sion and/or effective treatment of these symp-toms.13 Further research is necessary, in particu-lar through naturalistic studies, in order to gaina clearer picture of the clinical reality of depres-sion. Reliable biological markers to validate clin-



Med Care. 1990;28:239-250.8. Colon de Marti LN, Guzman Yunque FS, Guevara-RamosLM. Early detection of depression using the Zung Self-RatingDepression Scale. P R Health Sci J. 1997;16:375-379.9. Oquendo MA, Barrera A, Ellis SP, et al. Instability of symp-toms in recurrent major depression: a prospective study. AmJ Psychiatry. 200;161:255-261.10. Winkler D, Pjrek E, Heide A, et al. Gender differences inthe psychopathology of depressed inpatients. Eur Arch Psychi-atry Clin Neurosci. 2004;254:209-214.11. Fleck MP, Chaves ML, Poirier-Littre MF, Bourdel MC, Lôo H, Guelfi JD. Depression in France and Brazil: factorialstructure of the 17-item Hamilton Depression Scale in inpa-tients. J Nerv Ment Dis. 2004;192:103-110.12. Boulenger JP. Residual symptoms of depression: clinicaland theoretical implications. Eur Psychiatry. 2004;19:209-213.13. Menza M, Marin H, Opper RS. Residual symptoms in de-pression: can treatment be symptom-specific? J Clin Psychiatry.2003;64:516-523.

REFERENCES1. Schneider, K. Clinical Psychopathology. New York, NY: Gruneand Stratton; 19592. Mendels J, Cochrane C. The nosology of depression: the en-dogenous-reactive concept. Am J Psychiatry. 1968;124(suppl):1-11.3. Nelson JC, Charney DS. Primary affective disorder criteriaand the endogenous-reactive distinction. Arch Gen Psychiatry.1980;37:787-793.4. Carroll BJ. Problems with diagnostic criteria for depression.J Clin Psychiatry. 1984;45(7 pt 2):14-18.5. Suh T, Gallo JJ. Symptom profiles of depression among gen-eral medical service users compared with specialty mentalhealth service users. Psychol Med. 1997;27:1051-1063.6. Demyttenaere K, De Fruyt J, Stahl SM. The many faces offatigue in major depressive disorder. Int J Neuropsychophar-macol. 2005;8: 93-105.7. Magruder-Habib K, Zung WW, Feussner JR. Improvingphysicians’ recognition and treatment of depression in gen-eral medical care. Results from a randomized clinical trial.

disorder strive to find a physical cause for theirdiscomfort. Hence sleep disorders or vague somat-ic symptoms tend to be frequent presentations.5,9

Pain, particularly if the patient reports three si-multaneous pain symptoms, indicates depressivedisorder.4 Standardized classification systems likethe Diagnostic and Statistical Manual of MentalDisorders, 4th Edition, Text Revision (DSM-IV-TR)6

specify that besides reduced interest or sadness,at least four additional symptoms drawn from alist that includes vegetative symptoms, psycho-motor activity, cognitive symptoms, and suicidalideation, are a diagnostic requirement. Childrenand adolescents tend to describe their mood asirritable rather than sad.6 In addition to the con-text of primary mood disorders, symptoms ofdepression can occur in the context of obsessive,panic, eating, and generalized anxiety disorders,as well as drug and alcohol abuse. Symptomsand typical syndromes of depression are frequentin nonpsychiatric disorders 2 such as grief reac-tion, substance abuse, intoxication, or withdraw-al, and in purely physical disorders like diabetes,cancer, heart attack, stroke, and after use of pre-scription drugs for some of these disorders.4,10 Thetreating physician should thus determine whetherthere is a relationship between the symptoms ofthe primary physical illness, the effects of theadministered drugs, and affective symptoms. Thesequence of occurrence of symptoms, and thenature, intensity, diurnal fluctuations, past andfamily history of depression, help to diagnose

D epression with accompanying biological,behavioral, and cognitive changes is themost common form of mental disorder in

the community.1-4 A vigilant primary care physi-cian, the “gatekeeper” of health care,5 is crucialin early detection, prompt and effective treatment,adequate rehabilitation, and prevention of de-pression. Proverbially, the term “depression” isequated with a feeling of sadness.4 Variability inclinical manifestations, natural course, comorbid-ity, associated disability, or impairment, and itsimpact on the considerable underrecognition ofthe disorder is well documented.5 Guidelines fordiagnosis and detection of depression emphatical-ly state that not all clinically depressed patientsare sad and that many sad patients are not clin-ically depressed.4 Depression is not a normal re-action to life’s difficulties.2,4 Depressive disorderconsists of a group of symptoms and signs lasting2 weeks or more.6 Presence of pervasive sadnessor lack of interest in pleasurable activities or itsequivalents is essential for the diagnosis of de-pression.6 Diagnosis is easier when the patientmentions feeling sad or depressed.5 Apathy, anx-iety, irritability, and reduced interest or capacityfor pleasure or enjoyment may be experienced inaddition to or instead of sadness.4 Though affec-tive symptoms are cardinal, cognitive, behavioral,and somatic symptoms are integral to the clinicalpresentation.4,7,8 Social stigma, or viewing depres-sion as personal weakness, results in reluctanceto seek or accept the diagnosis.5 Victims of the

8◆ V. N. Vahia, India

Vihang N. VAHIA, MD Professor of Psychiatry (University of Bombay)G. S. Medical College & Cooper HospitalLilavati HospitalMumbai 400 054INDIA(e-mail: [email protected])

subtypes of depression.◆ Achieve complete remission, promote function-al recovery, prevent relapse.◆ Safe in overdose, pregnancy, lactation, and incardiac, respiratory, autoimmune, skin, hepatic,renal, and other systemic disorders.◆ Minimal drug interactions.◆ Established safety for long-term use in all agegroups. ◆ Absent or minimal adverse and emergent ef-fects on weight, hair growth, appetite, libido,and others. ◆ No diurnal sedation.◆ Easy to administer, preferably once dailystandard dose, no dose titration. ◆ Cost-effective, easily available.◆ Low risk of breakthrough mania. ❒

depression.4,5 Instruments like the DepressionScreening Questionnaires5 facilitate the diagno-sis of depression. Ultimately, the combination ofclinical awareness, vigilance, and diagnostic ori-entation of the primary care physician is criticalto early recognition of a depressive disorder.2,5

Once the presence of a depressive disorder is di-agnosed, a finer analysis is required to subtypethe disorder.4,10 Age, education, race, income, andmarital status are related to the outcome morethan the presentation.4 Psychosocial events mayinfluence the first few episodes, but do not havemuch relevance in subsequent episodes.4 An idealantidepressant drug is definitely on any clini-cian’s wish list.11 The ten tenets for an ideal an-tidepressant drug should be:◆ Rapid onset, broad spectrum of action in all



cal Manual of Mental Disorders. 4th Edition, Text Revision.Washington DC; American Psychiatric Association; 2000.7. Akiskal HS. Mood disorders: clinical features. In: Sadock BJ,Sadock VA, eds. Comprehensive Textbook of Psychiatry. 7thEdition, Vol 1. Philadelphia, Pa: Lippincott Williams & Wilkins;2000:1341. 8. Depression Research at National Institute of Mental Health.Depression Fact Sheet 1999. Available at http://www.loren-bennett.org/depfactsheet.htm. Accessed on 04/20/2005. 9. Greunberg AM, Goldstein RD. Mood Disorders: Depression.In: Tasman A, Kay J, Lieberman JA, eds. Psychiatry. Vol 2, 2ndedition. Indian Reprint. Bangalore, India: Panther PublishersPrivate Ltd; 2004:1217-1218.10. Greunberg AM, Goldstein RD. Mood Disorders: Depression.In Tasman A, Kay J, Lieberman JA, eds. Psychiatry. Vol 2, 2ndedition. Indian Reprint. Bangalore, India: Panther PublishersPrivate Ltd; 2004:1210-1213. 11. Lôo H. Antidepressants of the future: a clinician’s wish list.Medicographia. 2003;25:3-8. Editorial.

REFERENCES1. Ustun TB. The world-wide burden of depression in the 21stcentury. In: Weissman MM, ed. Treatment of Depression. Bridg-ing the 21st Century. Bombay, India: First Indian Edition;2003:35. 2. Murthy RS, Bertolote JM, Eppig Jordan J, eds. World HealthReport 2001. Geneva, Switzerland: World Health Organization;2001.3. Chisholm D, Sanderson K, Ayuso-Mateos JL, Saxena S. Re-ducing the global burden of depression population–level analy-sis of intervention cost-effectiveness in 14 regions. Br J Psy-chiatry. 2004;184:393-403.4. Depression Guidelines Panel. Depression in Primary Care.Vol 1. Detection and Diagnosis. Clinical Practice Guideline,Number 5. Rockville, MD. U.S. Department of Health and Hu-man Service, Agency for Health Care Policy and Research.AHCPR Publication No. 93-0550. April 1993.5. Wittchen HU, Beesdo K, Bittner A. Depression, an under-diagnosed disease. Medicographia. 2003;25:9-18.6. American Psychiatric Association. Diagnostic and Statisti-

✦

270 Unmet needs in depression (and other mental disorders) – SartoriusMEDICOGRAPHIA, VOL 27, No.3, 2005

significantly impaired and who requiremedical help necessary to prevent themfrom harming themselves or those aroundthem. These cases, however, representonly a relatively small proportion of thetotal numbers of people who are usingmental health services.

Are there any problems that make theassessment of needs defined in this waydifficult?

The main difficulty does not liein the assessment of needs and

demands, but in the fact that the needsand demands of the various stakeholdersin mental health care are not the same(see Figure).

What patients want overlapsonly partially with what familiesand society want. Patients, for ex-ample, may demand the best ofcare regardless of its cost. Fami-lies may be keen on preservingtheir prestige and avoiding em-barrassment from the unusualbehavior sometimes seen inmental illness (regardless of thetype of treatment) and wouldnot object to hospitalization farfrom home or other ways of hid-ing the patient. Governments

ify” them as having a mental disorder.They, however, request a service, get it,and feel better. It is therefore more use-ful not to equate prevalence data withneeds—were that done, there would alsobe the danger that every change of diag-nostic criteria would change the assess-ment of needs—but to define mentalhealth needs as demands for care madeby people who have a mental disorder forwhich the health services have an effec-tive response acceptable to the personsdemanding help and to the society inwhich they live. To the estimate of needsdefined in this way, it is necessary to addwork that should be done to look afterpersons whose mental disorder is of suchseverity that insight into their state is

Needs for mental health services havebeen traditionally expressed in termsof frequency of mental disorders. Isthis the best way of measuring needs?

No. Needs for mental health ser-vices cannot be equated with

the prevalence of mental disorders. Thereare mental disorders for which healthservices can do little and social servicesa great deal: in such cases, countingthese disorders as being mental healthservices needs is not relevant: they aresocial service needs. There are also peoplewho suffer from mental disorders thatdo not diminish their insight and whoprefer to deal with their problems them-selves, and should therefore not be count-ed as needing a mental health service.On the other hand, there are also peoplewho ask for help and advice of mentalhealth services although they do not havethe array of symptoms that would “qual-

I N T E R V I E W

M ental health needs should no longer be expressed interms of prevalence of mental disorders as such, but interms of demands for which an adequate response is

available to mental health services. Therefore, assessment of needsshould take into account not only what the patient wants, butalso the needs of families, doctors, and society. The main diffi-culty in the application of this definition is that the needs anddemands of each party may be conflicting or show only partialoverlap. What patients want, for example, is to be treated withdignity, to be free of symptoms, and to retain their source of in-come (jobs). Family needs include the preservation of their rep-utation and financial situation. Doctors want to be effective, avoidlitigation, be confident that the risk of suicide is low, and thattreatment will have a rapid onset of action and not be undulylong. Governments (voicing society’s demands) want to reducethe cost of care, while remaining in tune with their election plat-

form. Governmental policies contain the solution to some, butcertainly not all, health issues. Although the development of newand better medications is unquestionably useful, it is by no meansthe only solution to the needs for better mental health services.Knowledge about mental illnesses and ways of dealing with men-tally ill people should also be improved. Therapeutic advancesmust be accompanied by an adjustment of the health system andan active commitment of society to deal with mental disordersin the best way known.Medicographia. 2005;27:270-272. (see French abstract on page 272)

Keywords: depression; mental disorder; patient; family; society;government; health service; cultural setting

✦

Prof NormanSARTORIUS, MD, PhDFormer Director, WHO Division of Mental Health

Past President, World Psychiatric Association (WPA) and Association of European Psychiatrists (AEP)Geneva, SWITZERLAND

Address for correspondence: Prof Norman Sartorius, 14 Chemin Colladon, 1209 Geneva, Switzerland(e-mail: [email protected])

Needsexpressed by

patients

Needsexpressed by

families

Needs common to all

Needsexpressed by

society/government

UNMET NEEDS IN DEPRESSION(AND OTHER MENTAL DISORDERS)

I n t e r v i e w w i t h N . S a r t o r i u s S w i t z e r l a n d

I N T E R V I E W

271Unmet needs in depression (and other mental disorders) – Sartorius MEDICOGRAPHIA, VOL 27, No.3, 2005

And society’s needs?

In democratic societies, govern-ments are “spokespersons” for so-

ciety as a whole (or at least 51% of it).Governments, however, rarely have theluxury of fulfilling long-range plans andusually have to change them for politicalreasons. This also changes governments’demands on the health sector. Still, itseems that some demands are becominguniversal. Governments want health ser-vices to help them contain disruptions ofcommunities that they fear might arise ifpatients with mental illnesses were to beallowed to go about unattended; they wantto reduce the cost of care and keep it lowin all sectors with the exception of thosethat are politically very important andtherefore often very visible; they want tobuild and retain their image in harmonywith their election platform, which oftencontains the solution of some, but certain-ly not all, health problems; and they wantto avoid scandals which involve them.

What about unmet needs concerningtreatment?

Today’s array of medicationsand techniques that have been

shown to be effective and safe is consid-erable, certainly larger than ever before.The main challenge therefore does not liein the development of better treatments,but in the equitable use of those available.Stigma of mental illness reduces the pri-ority given to the development of effec-tive mental health services. Burnout ofpersonnel in institutions dealing withmental illness is on the increase, fueledby an excessive amount of work (withcontinuous reduction of personnel andother resources) and unrealistic expecta-tions. Ignorance about mental illness andways of dealing with people who have itis monumental. Also, the disintegrationof families and communities reduces theircapacity to deal with their ill members.

These are challenges that have to beaddressed if the currently available treat-ment techniques are to give maximumbenefit and satisfy the mental health ser-vice needs of the population.

Does this mean that we do not need toimprove treatment techniques?

No, improvement of therapy isvery useful, but it will not be

sufficient. Advances must go hand inhand with an adjustment of the healthsystem and an active commitment ofsociety to deal with mental disorders inthe best way known. ❒

their treatment; that they demand easyaccess—whenever necessary—to theservice providers; that they wish to befreed from symptoms that are severelydisturbing (some symptoms of mentaldisorders are not experienced as toomuch of a nuisance by the patient); thattheir disease—if it cannot be eliminat-ed—does not affect functions that theyconsider most important (for example,their thinking or their sexual function-ing); that they do not remain for evermarked by the disease; and that theirsources of income—eg, their job—isnot affected by the illness.

These examples of patients’ wishes andexpectations were found in a major pro-ject against stigma started a decade agoand currently under way in some 18countries. These findings may not bevalid everywhere, and it is therefore ex-tremely useful to carry out local studiesto find out what the patients’ needs are.

What about families’ needs?

Here, even more than with pa-tients’ needs, it is important

to carry out assessments in the culturalsetting in which the service is providedbecause the needs of families tend to beprofoundly affected by traditions andother cultural markers of a community.In many countries, families are reportedto go to extreme lengths to protect theirgood reputation; they try to avoid finan-cial crises because of one of its members’sickness; they will often have only a lim-ited amount of patience in waiting fortheir sick member to reassume his/herrole in the family; and they will insist onhaving a decisive word in decisions aboutthe patient’s treatment. They will alsooften refuse to take on—forever—theburden that a sick member imposes; forthe family as well as for the patient it isincomparably easier to fight if the endof the disease is in sight, even if verydistant.

And what about the doctors’s needs?

Doctors want to help their pa-tients, but also want to avoid

litigation; they want to be reassured thatthe risk of suicide in a patient they haveto treat is low; they want to see a rapidonset of treatment—not only becausethey want to help the patient, but alsobecause they want to see the good resultsof their work; they want to learn abouttechniques that can shorten the durationof their intervention; and they want to beable to refer away patients whose treat-ment is beyond their competence.

may want to reduce costs for care by focusing on certain groups of disordersand providing the least expensive service.All three would like to see the illness dis-appear, but in its presence will demandor do different things to resolve theproblems.

Can you give us another example illus-trating the relevance of this approachto the assessment of needs?

Another example is the oftenquoted desideratum that treat-

ment must be simple. The argument isthat this will allow less well-trained per-sonnel to apply the treatment and thatcompliance with the instructions of theservice provider to the patient will bebetter. The requirement that treatmentbe simple stems from priorities of thehealth service, which aims to simplifytraining and to delegate treatment re-sponsibility to less expensive staff. Pa-tients do not express that need. In fact,both the experience of traditional prac-titioners and of medical practice seem toindicate that patients are prepared tofollow very complex treatment regimensprovided that they have received instruc-tions about the treatment and that theybelieve that, if taken in the specified way,the treatment will help.

Another common example is the fami-ly’s demand to the health worker to finda treatment that will make the patient’ssleep/wake rhythm predictable and co-inciding with theirs. While a patient maynot feel the need to take sleeping pillsand might even prefer being awake for agreat part of the night, the family findsthis difficult to tolerate because of theirobligations during the day. Recognizingthe difference between what patients andfamilies need and taking it into accountwill require that doctors use their nego-tiating skills with both parties in orderto find a compromise solution. If they donot, they will lose the good will and con-fidence of one or both of the crucial par-ties involved in the treatment process—the patients and/or their families.

Tell us more about the needs thatpatients express?

R esearch on expressed needsand demands has only recently

begun to take on more serious dimen-sions. Until now, such studies were notseen as a priority and it was difficult tofind funding for them. What we do knowis that patients wish to be treated withdue respect to their dignity; that theywant to be involved in decisions about

I N T E R V I E W

272 Unmet needs in depression (and other mental disorders) – SartoriusMEDICOGRAPHIA, VOL 27, No.3, 2005

✦

de maintenir leur réputation et leur situation financière. Les mé-decins veulent être efficaces, éviter litiges et procès, être assurésque le risque de suicide est bas, que le traitement agira rapide-ment et ne s’étendra pas sur une durée trop longue. Les gouver-nements (représentant la société) cherchent à réduire les coûtsde santé, tout en restant à l’unisson de leurs plates-formes élec-torales. Si nul ne nie l’utilité de développer des médicamentsnouveaux et plus efficaces, ceci ne peut en aucun cas constituerla seule solution aux besoins des services de santé mentale. Laconnaissance des maladies mentales et la manière de soigner lespatients qui en souffrent doivent également être améliorées. Lesavancées thérapeutiques doivent s’accompagner des révisionsnécessaires des systèmes de santé et d’une implication active dela société dans la prise en charge des troubles mentaux.

BESOINS INSATISFAITS DANS LA DÉPRESSION (ET AUTRES TROUBLES MENTAUX)

L es besoins dans le domaine des maladies mentales ne de-vraient plus être exprimés en termes de prévalence destroubles mentaux en tant que tels, mais en termes de de-

mandes pour lesquelles il existe une réponse adéquate sur le plandesservices de santé mentale.L’évaluationdes besoinsdoitprendreen compte non seulement les souhaits du patient, mais égale-ment ceux exprimés par les familles, les médecins et la société.La difficulté principale liée à l’application de cette définition estque les besoins et les demandes de chaque partie peuvent être enconflit ou ne se recouvrir que de façon partielle. Ce que deman-dent les patients, par exemple, c’est d’être traités avec dignité,d’être soulagésde leurs symptômes, etd’être sûrsque leurs sourcesde revenus (emplois) ne soient pas affectées par leur maladie.L’évaluation des besoins des familles comprennent entre autres

sleep physiological processes may be intimately in-volved in the pathophysiological and recovery pro-cesses of major depression.2 Insomnia associatedwith depression can be categorized as follows: diffi-culty initiating sleep, difficulty maintaining sleep(or continuity disturbances), and early morningawakening. Difficulty initiating sleep is the leastspecific, and early morning awakening the mostspecific to major depression. Factors such as age,sex, severity of the episode, and comorbid anxietyplay a role in determining the incidence, severity,and nature of sleep disturbances associated withmajor depression.

Time course

Sleep abnormalities in depression are seen acrossthe age spectrum, from an early age, ie, in adoles-cence,3 to the elderly. The bidirectional risk rela-

This article looks at the pivotal importance ofdepression-associated sleep abnormalitiesthroughout their course: from diagnosis at

the early stage of depression to initial treatment,and to their long-term outcome.

The quality of life of depressed patients is dimin-ished by a variety of factors, sleep disturbances be-ing one of the most prominent among these. Veryfew patients with depression do not experience dis-turbance of sleep pattern in one form of another.Sleep disturbances in depression comprise insom-nia and, less commonly, hypersomnolence. Somepatients actually fluctuate between these two pat-terns of sleep disturbance during the course of anepisode of depression, reflecting the psychobiolog-ical heterogeneity of major depressive episodes.1

Whether as insomnia or hypersomnia, it has be-come clear that sleep dysregulation is more thanjust an epiphenomenon of depression. Thus, notonly is disturbed sleep one of the most commonconcomitants of a persistent depressed mood—ad-ditional associations include the following: (i) thepresence of sleep abnormality in depressed patientshas prognostic and outcome implications; (ii) themajority of antidepressants cause substantial chang-es in polysomnographic measures as well as in sub-jective and objective aspects of sleep; and (iii) var-ious sleep manipulations can alleviate or exacerbatesymptoms of depression in a subset of patients. Forthese reasons it has been suggested that changes in

273The impact of sleep disorders on the course of depression – Emsley MEDICOGRAPHIA, VOL 27, No. 3, 2005

F O C U S

A lmost all patients with major depression experience a dis-turbance of sleep pattern. Sleep disturbance is a core fea-ture of major depression, and is associated with important

prognostic and outcome implications. While sleep abnormalitieswith major depression are seen across the age spectrum, aging isassociated with more prominent insomnia. Disordered sleep pre-cedes the onset of a depressive episode, and the reemergence ofinsomnia may herald the onset of a new episode of recurrentdepression. While insomnia often resolves with successful treat-ment of the depressive episode, most polysomnographic sleep ab-normalities persist after recovery, suggesting that they are moretrait-like than state-like. Excessive daytime sleepiness and fa-tigue are consequences of insomnia, and can result in decreasedproductivity and an increased risk of accidents. Persistent sleepdisturbance is associated with significant risk of recurrence, anincreased risk of suicide, and poorer overall outcome. Most anti-depressants cause substantial changes in subjective aspects ofsleep as well as in polysomnographic measures. Tricyclic antide-pressants and the conventional monoamine oxidase inhibitors

suppress rapid eye movement (REM) sleep to varying degrees.Selective serotonin reuptake inhibitors (SSRIs) generally sup-press REM sleep, but are often associated with subjective reportsof insomnia. Insomnia in patients with major depression requirescareful assessment. Exclusion of primary insomnia, as well as ofgeneral medicaland substance-related causes, is important.Treat-ment strategies involve pharmacological and nonpharmacolog-ical interventions. Some clinicians prescribe an antidepressantwith sedative properties. Alternatively, a hypnotic can be copre-scribed. Neither of these approaches is without problems. Sleephygiene and other behavioral treatments may be helpful adjuncts.Medicographia. 2005;27:273-278. (see French abstract on page 278)

Keywords: sleep; insomnia; depression; major depression; antidepressant

✦

Robin EMSLEY, MB ChB, MMed, FCPsych, MD

Department of PsychiatryFaculty of Health SciencesUniversity of Stellenbosch

Cape Town, SOUTH AFRICA

THE IMPACT OF SLEEP DISORDERS ONTHE COURSE OF DEPRESSION

b y R . E m s l e y , S o u t h A f r i c a

Address for correspondence: Prof Robin Emsley, Department of Pyschiatry, Faculty of Health Sciences, University of Stellenbosch, PO Box 19063, Tygerberg 7505 South Africa (e-mail: [email protected])


HPA hypothalamic-pituitary-adrenal (axis)MAOI monoamine oxidase inhibitorN-REM non–rapid eye movement (sleep)REM rapid eye movement (sleep) SSRI selective serotonin reuptake inhibitor TCA tricyclic antidepressant

mentation during the night, early morning wak-ening with an inability to return to sleep, reducedsleep efficiency, and decreased total sleep time. ◆ Regarding abnormal sleep architecture, abnor-malities have been reported in the amounts anddistribution of non-rapid eye movement (N-REM)sleep stages throughout the night, including anincrease in the amount of light stage 1 sleep andreduction of deep, slow-wave (stages 3 and 4) sleep. ◆ REM sleep disturbances in patients include ashortening of REM latency (<65 minutes), a pro-longation of the first REM sleep period, and in-creased REM density (ie, increased total REM sleeptime), particularly in the first half of the night.14,15

Sleep disturbances can be verified with polysom-nography in 90% of patients with major depression.They are most pronounced in melancholic depres-sion16 and in bipolar and psychotic depression.17

The underlying mechanisms involved in thepathogenesis of sleep abnormalities in major de-pression are not known. In cases of early morningwakening and decreased REM latency there may bea phase advance of the sleep-wake cycle.18 Althoughnot fully elucidated, a recent study involving EEGsleep and regional metabolism assessments withpositron emission tomography reported interest-ing findings regarding the underlying neurobiologyof disturbed REM sleep in depression. Depressed pa-tients showed signs of altered function of limbic/anterior paralimbic, and prefrontal circuits duringthe REM sleep state, possibly reflecting an imbal-ance in monoaminergic/cholinergic function af-fecting not only the brainstem generation of REMsleep, but also the manner in which the forebrainresponds to the stimuli of REM sleep.19

Sleep disturbances seem to be more prominent inpatients with more severe depression, as suggestedby the finding that they were reported in 80% in-patients with depression (presumably more severe)compared with only 40% to 60% depressed outpa-tients. About 30% depressed patients actually havehypersomnia.13 When EEG sleep profiles were com-pared in the first 6 weeks of a depressive episode ina group of subjects with recurrent depression andcompared with their sleep profiles as measuredduring their previous episode at a later stage, it wasfound that REM sleep abnormalities were more pro-nounced earlier in the course of a depressive epi-sode.20 Deficiencies in sleep efficiency may explainwhy depressed patients often feel fatigued evenwhen they appear to sleep excessively. However, thediagnostic specificity of the sleep disturbances ispoor. It has been reported that no single sleep vari-able reliably distinguishes patients with major de-pression from healthy controls or from patientswith other psychiatric disorders.21

Clinical implications

Excessive daytime sleepiness and fatigue are un-avoidable consequences of insomnia, and are there-fore common features of patients suffering fromdepression.22 This in turn can result in impaired oc-cupational and social functioning, with decreasedproductivity and an increased risk of accidents.23

tionship between sleep and depression may be par-ticularly strong in older adults.4 Aging is associatedwith more prominent insomnia in depression, whilehypersomnolence is relatively more common ear-lier in life and may even be nearly as prevalent as in-somnia among younger depressed women.1 Thereis considerable evidence to suggest that disorderedsleep actually precedes the onset of a depressiveepisode. In a longitudinal epidemiological studyinvestigating the association between sleep distur-bance and psychiatric disorders in young adults,prior insomnia was a significant predictor of subse-quent major depression. On the basis of their find-ings, the authors suggested that complaints of 2weeks or more of insomnia nearly every night mightbe a useful marker of subsequent onset of a majordepressive episode.5 In the National Institute of Men-tal Health Epidemiologic Catchment Area study,10.2% and 3.2% of a community sample noted in-somnia and hypersomnia, respectively. The risk ofdeveloping new major depression was significantlyhigher in those who had insomnia compared withthose without insomnia, but this risk was much re-duced if the insomnia had resolved at a follow-upvisit.6 In subjects with preexisting major depressionthe reemergence of insomnia can similarly heraldthe onset of a new episode of recurrent depression.7

Considerable research attention has focused onwhether sleep changes occurring in major depres-sion are state-related (ie, emerge during the episodeonly) or trait-related (ie, whether they persist afterfull remission of the episode). Research to date hasproduced mixed results. On the one hand, improve-ment in subjective insomnia was found to be close-ly related to global improvement in depression,8 andit was originally thought to be a state marker. Thisis also supported by findings that hypothalamic-pituitary-adrenal (HPA) axis changes appear to bespecifically state-related.3 On the other hand, lon-gitudinal studies in fact indicate that most EEGsleep measures persist after episode recovery, sug-gesting that they are more trait-like.3,9,10 There areindications that the type of treatment may have aneffect on sleep in the postrecovery phase of depres-sion. A study reported quantitative differences inEEG sleep measures in subjects who had recoveredfrom a major depressive episode after treatmentwith interpersonal therapy and those who recov-ered after treatment with interpersonal therapy plusfluoxetine.11

Classification

Sleep EEG changes, together with HPA axis changesassociatedwith majordepressivedisorder are amongthe best replicated biological findings in psychia-try.12 Sleep disturbances in patients with major de-pressive disorders can be classified according topolysomnographic studies as: (i) difficulties initiat-ing and maintaining sleep; (ii) abnormal sleep ar-chitecture; and (iii) disruptions in the timing ofrapid eye movement (REM) sleep13:◆ Difficulties with sleep initiation and maintenanceinclude prolonged sleep latency (ie, sleep-onset in-somnia), intermittent wakefulness and sleep frag-

F O C U S

274 The impact of sleep disorders on the course of depression – EmsleyMEDICOGRAPHIA, VOL 27, No. 3, 2005

has the most prominent alerting effect on sleep.28

The clinical response to amitriptyline34 and clo-mipramine35 may be related to the degree of REMsleep suppression, with better outcome being as-sociated with a greater degree of REM suppression.One curious TCA is trimipramine. Although a clas-sic TCA, trimipramine shows unusual pharmaco-logical properties. Rather than reducing, it increas-es, REM sleep. This property has important clinicalimplications. Clinicians have for many years foundtrimipramine particularly useful in depressed pa-tients with prominent insomnia, and in additionit has been shown to be effective in treating non-depressed patients with primary insomnia.36

◆ Monoamine oxidase inhibitorsMonoamine oxidase inhibitors (MAOIs) have beenfound to also suppress REM sleep in patients withmajor depression.28 The onset of the suppressant ef-fect is delayed, however, and there is considerableassociated REM rebound upon withdrawal.37 Unlikein the case of TCAs, the antidepressant response tophenelzine treatment is not correlated with the de-gree of suppression of REM sleep.38 MAOIs also re-duce total sleep time and may decrease sleep effi-ciency.39 Different findings to the classic MAOIs havebeen reported with the reversible MAOI moclobe-mide. One study reported improved sleep continu-ity in depressed subjects, particularly during the in-termediate and late stages of drug administrationin a 4-week trial. This was accompanied by signif-icant improvements in the symptoms of depres-sion.40 A second study found an activating effectwith moclobemide, most marked during the earlyphase of treatment. The most noticeable effects wereonREM sleep, affecting polysomnographic andspec-tral sleep EEG parameters. A REM sleep habituationphenomenon was observed, and a slight REM sleeprebound effect occurred early during withdrawal.41

◆ Selective serotonin reuptake inhibitorsSerotonergic neurons play a critical role in modu-lating the onset and maintenance of sleep, and it isthought that insomnia in depression is linked todysfunction of central nervous system serotonergicsystems.31 Selective serotonin reuptake inhibitors(SSRIs) generally suppress REM sleep. However,they have variable effects on subjective reports ofeffects on sleep. Between 10% and 15% of patientstreated with an SSRI complain of insomnia duringthe early stages of treatment. As a consequence,clinicians often coprescribe concomitant hypnoticmedication. One source found that 35% patientsprescribed an SSRI also received a sedative/hypnot-ic.42 Studies have highlighted increased wakeful-ness with SSRIs, particularly on fluoxetine. A studyinvestigating the longer-term effects of fluoxetine inpatients with major depression who were treatmentresponders reported significantly reduced sleep ef-ficiency, decreased stage 2 sleep, increased stage 1sleep, prolonged REM latency, and a 3.4% reductionin REM time at 10 weeks of treatment. This studyshowed that these effects persist, as even after 30weeks of treatment there were still alerting effectson sleep.43 Paroxetine is also associated with an

There are other important clinical implicationsfor sleep disturbances in depression. Persistent sleepdisturbance is associated with significant risk of re-currence.6 Also, recurrent depression is associatedwith a more severe neurophysiologic substrate thanphenotypically similar single-episode cases.24 Twostudies have reported an association between sleepdisturbance and suicide. The first found that bothinsomnia and hypersomnia are associated with sui-cidal behavior in patients with major depression,25

and the second reported an association betweenpoor subjective sleep quality and suicidal behaviorin patients with major depressive disorder.26 Poly-somnographic sleep changes such as REM latencyhave also been reported to predict treatment re-sponse and the clinical course of illness.27

Effects of antidepressants on sleep

All antidepressants have some effects on sleep ar-chitecture.28 In fact, in early studies, suppression ofREM sleep was regarded as necessary for the anti-depressant action, based on a finding of a correla-tion between clinical response and REM suppres-sion as well as a temporal relationship between theonset of clinical response and REM suppression.However, this was later found not to be the case.1

Clinical trials investigating the effects of antide-pressants in patients with major depressive disor-der have reported varying effects of these agents onsleep latency and sleep efficiency, as well as on day-time somnolence.29 Some antidepressants have pro-found effects on sleep regulation, and others min-imal. Specific antidepressants also vary in theireffects from individual to individual. An antidepres-sant may produce a strong sedative effect in oneindividual, and no effect or even an activating effectin another. Clinicians have traditionally selectedan antidepressant with a more sedative profile fordepressed subjects with prominent insomnia, whilea more activating antidepressant may be prescribedin cases where hypersomnia is a significant prob-lem. It has also been suggested that the nature ofthe sleep disturbance at initial clinical presenta-tion may be relevant to the choice of antidepres-sant medications and the likelihood of experiencingtreatment-emergent side effects.30

◆ TricyclicsThe tricyclic antidepressants (TCAs) primarily in-hibit the reuptake of the neurotransmitter nore-pinephrine, and to a lesser extent serotonin.31 Theyhave additional effects on a variety of other recep-tors. Most of the TCAs are potent suppressors ofREM sleep, although some do so only moderately.32

Clomipramine and desipramine are the most po-tent in this regard. REM suppression is usually morepronounced during the first few hours of sleep andthere is some accommodation to these effects withcontinuation therapy.1 TCAs also prolong REM la-tency and decrease the total amount of REM sleeptime.13 Clomipramine, desipramine, and amitripty-line increase stage 1 sleep and decrease sleep effi-ciency over baseline levels.33 Clomipramine, withthe strongest serotonergic effects of all of the TCAs,

F O C U S


than depression (eg, anxiety disorder, mania, psy-chosis). There are various strategies for treating pa-tients with insomnia in major depression. First, anantidepressant with sedative properties can be pre-scribed. Mirtazapine is commonly used these days,although sedative TCAs such as amitriptyline andtrimipramine, trazodone, and nefazodone have alsobeen used. Patients successfully treated with anSSRI, but who experience persistent insomnia, maybe treated with a hypnotic such as zolpidem or zopi-clone. Alternatively, they may be switched to a se-dating antidepressant. Finally, combinations of an-tidepressants (ie, adding a sedative antidepressantat night to a patient being treated with an SSRI) canbe considered as a last resort in patients respondinginadequately to the above strategies.1 However, cau-tion is necessary with the TCAs because of poten-tially toxic effects due to drug-drug interactions,and combination of an SSRI with mirtazapine islikely to be safer.

In addition to pharmacologic management, sleep-hygiene and other behavioral strategies can be help-ful in treating insomnia associated with major de-pression. Activities that may be detrimental to op-timal nighttime sleep should be eliminated. Theseinclude things such as daytime napping, excessivealcohol consumption and brisk exercise in theevening, and excessive daytime caffeine intake.22

Other behavioral treatments include stimulus con-trol therapy (aimed at breaking the cycle of prob-lems commonly associated with initiating sleep),sleep restriction therapy, relaxation and biofeedbacktherapy, and chronotherapy (resetting the biolog-ical clock by progressively phase-delaying sleep).

Conclusion

Since insomnia associated with major depressioncauses subjective distress and is associated withmany risks, its optimal management must be animportant clinical goal. Clinicians have for a longtime been aware of the importance of the effects ofantidepressants on sleep when choosing a drug totreat major depression. Previously, such decisionswere based largely on clinical experience. Today,much research has been conducted, and we have abetter understanding of the specific effects of indi-vidual antidepressants on sleep regulation. Howev-er, there is still a lack of consensus on the manage-ment of insomnia associated with major depression.Consequently, long-term treatment of insomnia isdriven primarily by the individual choices of patientsand their clinicians.54 Clinicians are frequently facedwith a difficult choice when dealing with depressedpatients with prominent insomnia. Prescribing anantidepressant with an activating effect may exac-erbate the sleep disturbance, and coprescription ofa hypnotic goes against the basic principle of avoid-ing polypharmacy where possible, as well as intro-ducing the inherent risks of hypnotic therapy. Us-ing a sedative antidepressant is therefore often re-garded as a better option. However, this may resultin a spillover of daytime sedation, and also com-plicate the return of normal sleep-cycle in recover-ing patients. ❒

alerting effect44 as well as a significant REM reboundwith discontinuation.44 Fluvoxamine also appears tohave an alerting effect in patients with major de-pression. This effect was found to have a fast onset,being evident even after the first day or two of treat-ment.45 Citalopram did not appear to have any dele-terious effects on sleep continuity or wakefulnessduring the night in a study of patients over 6 weeksof treatment.46 Escitalopram is reported to cause in-somnia at a greater rate than placebo.47

◆ Other antidepressantsMirtazapine is a potent blocker of 5-HT2, 5-HT3 andhistamine receptors, and it has the strongest ini-tial sedative effects among the newer antidepres-sants. These effects cause daytime sedation in abouthalf of patients during the first 2 weeks of treat-ment,48 but they may be very useful for depressedpatients with prominent insomnia. In a study em-ploying low doses of venlafaxine, the effects on sleeparchitecture were similar to those of the SSRIs.49

Bupropion is generally considered to be an activat-ing antidepressant, and has been reported to short-en REM latency and increase REM sleep.50 Rebox-etine, a selective norepinephrine reuptake inhibitor,has been reported to induce sleep-EEG changessimilar to those after SSRIs by increasing intermit-tent wakefulness and decreasing REM time.51

Treatment options for insomnia associated with depression

Insomnia is a symptom with many possible under-lying causes. Careful assessment and establishmentof a differential diagnosis is necessary. It is impor-tant to rule out a primary sleep disorder as thecause of the insomnia as one of the first steps. A sig-nificant association has been reported between pri-mary sleep disorder and major depression. There isan increased risk of depression in patients with rest-less legs syndrome52 as well as sleep apnea,53 and ef-fective treatment of sleep apnea appears to improvedepression in these patients. A further importantreason for ruling out primary sleep disorders is be-cause some of the medications used to treat insom-nia associated with depression can actually exacer-bate the primary sleep disorders.1 Once identified,primary sleep disorders should be referred to a sleepspecialist. There are numerous other possible caus-es of insomnia. Insomnia may also be related to ageneral medical condition, whether due to pain,discomfort, or a more direct physiological mecha-nism. Examples of the latter include hyperthyroid-ism and pheochromocytoma. Insomnia may be re-lated to a side effect of medication, or to the use ofsome other substance. Both substance (eg, alcohol)intoxication and withdrawal may cause insomnia.In attempting to improve sleep patterns, patientswith insomnia may sometimes use medications in-appropriately. This may include the use of hypnoticsor alcohol to help them sleep at night, as well as caf-feine or other stimulants to combat excessive day-time fatigue. Alcohol dependence is often associat-ed with prominent, persistent insomnia. Insomniamay also be related to a psychiatric disorder other

F O C U S


F O C U S


REFERENCES1. Jindal RD, Thase ME. Treatment of insomnia asso-ciated with clinical depression. Sleep Med Rev. 2004;8:19-30.2. van Bemmel AL. The link between sleep and de-pression: the effects of antidepressants on EEG sleep.J Psychosom Res. 1997;42:555-564.3. Rao U, McCracken JT, Lutchmansingh P, et al. Elec-troencephalographic sleep and urinary free cortisol inadolescent depression: a preliminary report of changesfrom episode to recovery. Biol Psychiatry. 1997;41:369-373.4. Buysse DJ. Insomnia, depression and aging. Assess-ing sleep and mood interactions in older adults. Geri-atrics. 2004;59:47-51.5. Breslau N, Roth T, Rosenthal L, et al. Sleep distur-bance and psychiatric disorders: a longitudinal epi-demiological study of young adults. Biol Psychiatry.1996;39:411-4186. Ford DE, Kamerow DB. Epidemiologic study ofsleep disturbances and psychiatric disorders. An op-portunity for prevention? JAMA.1989;262:1479-1484.7. Perlis ML, Giles DE, Buysse DJ, et al. Self-reportedsleep disturbance as a prodromal symptom in recur-rent depression. J Affect Disord. 1997;42:209-212. 8. Casper RC, Katz MM, Bowden CL, et al. The patternof physical symptom changes in major depressive dis-order following treatment with amitriptyline or imi-pramine. J Affect Disord.1994;31:151-164. 9. Buysse DJ, Kupfer DJ, Frank E, et al. Electroen-cephalographic sleep studies in depressed outpatientstreated with interpersonal psychotherapy: II. Longi-tudinal studies at baseline and recovery. PsychiatryRes.1992;42:27-40. 10. Giles DE, Jarrett RB, Rush AJ, et al. Prospective as-sessment of electroencephalographic sleep in remittedmajor depression. Psychiatry Res.1993;46:269-284.11. Buysse DJ, Hall M, Begley A, et al. Sleep and treat-ment response in depression: new findings using pow-er spectral analysis. Psychiatry Res. 2001;103:51-67.12. Holsboer F. Neuroendocrinology of mood disor-ders. In: Bloom FE, Kupfer DJ, eds. Psychopharma-cology:The Fourth Generation of Progrress.New York,NY: Raven Press; 1995:957-969. 13. Armitage R. The effects of antidepressants on sleepin patients with depression. Can J Psychiatry. 2000;45:803-809.14. Armitage R. Microarchitectural findings in sleepEEG in depression: diagnostic implications. Biol Psy-chiatry.1995;37:72-84. 15. Bahro M, Riemann D, Stadtmuller G, et al. REMsleep parameters in the discrimination of probableAlzheimer’s disease from old-age depression. Biol Psy-chiatry.1993;34:482-486.16. Thase ME, Kupfer DJ, Fasiczka A, et al. Identify-ing an abnormal electroencephalographic sleep pro-file to characterize major depressive disorder. BiolPsychiatry.1997;1:964-973. 17. Dubovsky SL, Thomas M. Psychotic depression:advances in conceptualization and treatment. HospCommunity Psychiatry.1992;43:1189-1198. 18. Poland RE, McCracken JT, Lutchmansingh P, etal. Differential response of rapid eye movement sleepto cholinergic blockade by scopolamine in current-ly depressed, remitted, and normal control subjects.

and sleep electroencephalogram in depressed patientstreated with phenelzine. Arch Gen Psychiatry. 2001;58:268-276. 39. Kupfer DJ, Bowers MB Jr. REM sleep and centralmonoamine oxidase inhibition. Psychopharmacolo-gia. 1972;27:183-190. 40. Monti JM. Effect of a reversible monoamine oxi-dase-A inhibitor (moclobemide) on sleep of depressedpatients. Br J Psychiatry. 1989(suppl)6:61-6541. Minot R, Luthringer R, Macher JP. Effect of mo-clobemide on the psychophysiology of sleep/wake cy-cles: a neuroelectrophysiological study of depressedpatients administered with moclobemide. Int ClinPsychopharmacol. 1993;7:181-189. 42. Rascati K. Drug utilisation review of concomi-tant use of specific serotonin reuptake inhibitors orclomipramine with antianxiety/sleep medications.Clin Ther.1995;17:786-790. 43. Trivedi MH, Rush AJ, Armitage R, et al. Effects offluoxetine on the polysomnogram in outpatients withmajor depression. Neuropsychopharmacology. 1999;20:447-459. 44. Staner L, Kerkhofs M, Detroux D, et al. Acute, sub-chronic and withdrawal sleep EEG changes duringtreatment with paroxetine and amitriptyline: a dou-ble-blind randomized trial in major depression. Sleep.1995;18:470-477. 45. Kupfer DJ, Perel JM, Pollock BG, et al. Fluvox-amine versus desipramine: comparative polysomno-graphic effects. Biol Psychiatry. 1991;29:23-40. 46. van Bemmel AL, van den Hoofdakker RH, Beers-ma DG, et al. Changes in sleep polygraphic variablesand clinical state in depressed patients during treat-ment with citalopram. Psychopharmacology (Berl).1993;113:225-230. 47. Waugh J, Goa KL. Escitalopram: a review of its usein the management of major depressive and anxietydisorders. CNS Drugs. 2003;17:343-362. 48. Holm KJ, Markham A. Mirtazapine: a review of itsuse in major depression. Drugs. 1999;57:607-631. 49. Luthringer R, Toussaint M, Schaltenbrand N, etal. A double-blind, placebo-controlled evaluation ofthe effects of orally administered venlafaxine on sleepin inpatients with major depression. Psychopharma-col Bull.1996;32:637-646. 50. Nofzinger EA, Reynolds CF 3rd, Thase ME, et al.REM sleep enhancement by bupropion in depressedmen. Am J Psychiatry. 1995;152: 274-276. 51. Kuenzel HE, Murck H, Held K, et al. Reboxetineinduces similar sleep-EEG changes like SSRIs in pa-tients with depression. Pharmacopsychiatry. 2004;37:193-195. 52. Saletu B, Anderer P, Saletu M, et al. EEG mapping,psychometric, and polysomnographic studies in rest-less legs syndrome (RLS) and periodic limb movementdisorder (PLMD) patients as compared with normalcontrols. Sleep Med. 2002;3(suppl):S35-S42. 53. Millman RP, Fogel BS, McNamara ME, et al. De-pression as a manifestation of obstructive sleep ap-nea: reversal with nasal continuous positive airwaypressure. J Clin Psychiatry. 1989;50:348-351. 54. Jindal RD, Buysse DJ, Thase ME. Maintenancetreatment of insomnia: what can we learn from the de-pression literature? Am J Psychiatry. 2004;161:19-24.

Biol Psychiatry. 1997;41:929-938. 19. Nofzinger EA, Buysse DJ, Germain A, et al. In-creased activation of anterior paralimbic and executivecortex from waking to rapid eye movement sleep indepression. Arch Gen Psychiatry. 2004;61:695-702. 20. Kupfer DJ, Ehlers CL, Frank E, et al. EEG sleepprofiles and recurrent depression. Biol Psychiatry.1991;30:641-655.21. Benca RM, Obermeyer WH, Thisted RA, et al.Sleep and psychiatric disorders. A meta-analysis. ArchGen Psychiatry. 1992;49:651-658. 22. Doghramji K. Treatment strategies for sleep dis-turbance in patients with depression. J Clin Psychia-try. 2003;64(suppl 14):24-29. 23. Leger D, Guilleminault C, Bader G, et al. Medi-cal and socio-professional impact of insomnia. Sleep.2002;25:625-629. 24. Thase ME, Kupfer DJ, Buysse DJ, et al. Electro-encephalographic sleep profiles in single-episode andrecurrent unipolar forms of major depression: I. Com-parison during acute depressive states. Biol Psychia-try. 1995;38:506-515. 25. Agargun MY, Kara H, Solmaz M. Sleep distur-bances and suicidal behavior in patients with majordepression. J Clin Psychiatry. 1997;58:249-251. 26. Agargun MY, Kara H, Solmaz M. Subjective sleepquality and suicidality in patients with major depres-sion. J Psychiatr Res. 1997;31:377-381. 27. Buysse DJ, Kupfer DJ, Frank E, et al. Do electro-encephalographic sleep studies predict recurrence indepressed patients successfully treated with psycho-therapy? Depression. 1994;2:105-108. 28. Sharpley A, Cowen PJ. Effect of pharmacologictreatments on the sleep of depressed patients. BiolPsychiatry.1995;37:85-98. 29. Winokur A, Gary KA, Rodner S, et al. Depression,sleep physiology, and antidepressant drugs. DepressAnxiety. 2001;14:19-28. 30. Armitage R, Sussman N. Effects of fluoxetine onsleep architecture and quality of sleep in depressedpatients. Primary Psychiatry. 1997;4:34-37. 31. Thase ME. Treatment issues related to sleep anddepression. J Clin Psychiatry.2000;61(suppl 11):46-50.32. Vogel GW, Buffenstein A, Minter K, et al. Drug ef-fects on REM sleep and on endogenous depression.Neurosci Biobehav Rev. 1990;14:49-63. 33. Shipley JE, Kupfer DJ, Griffin SJ, et al. Compar-ison of effects of desipramine and amitriptyline onEEG sleep of depressed patients. Psychopharmacolo-gy (Berl).1985;85:14-22. 34. Gillin JC, Wyatt RJ, Fram D, et al. The relationshipbetween changes in REM sleep and clinical improve-ment in depressed patients treated with amitriptyline.Psychopharmacology (Berl). 1978;59:267-272. 35. Hochli D, Riemann D, Zulley J, et al. Is there a re-lationship between response to total sleep deprivationand efficacy of clomipramine treatment in depressedpatients? Acta Psychiatr Scand. 1986;74:190-192. 36. Berger M, Gastpar M. Trimipramine: a challengeto current concepts on antidepressives. Eur Arch Psy-chiatry Clin Neurosci.1996;246:235-239. 37. Dunleavy DL, Oswald I. Phenelzine, mood response,and sleep. Arch Gen Psychiatry. 1973;28:353-356. 38. Landolt HP, Raimo EB, Schnierow BJ, et al. Sleep

F O C U S


✦

L a plupart des patients atteints de dépression majeure ontune perturbation du tracé du sommeil. Le trouble du som-meil est une caractéristique essentielle de la dépression ma-

jeure et il est associé à d’importantes implications pronostiqueset évolutives. Alors que les anomalies du sommeil accompagnantune dépression majeure sont retrouvées à tout âge, le vieillisse-ment est associé à une insomnie plus marquée. Les troubles dusommeil précèdent l’installation de l’épisode dépressif et la réap-parition de l’insomnie peut annoncer le début d’un nouvel épi-sode de dépression récurrente. Alors qu’un traitement efficace del’épisode dépressif résout souvent l’insomnie, la plupart des ano-malies polysomnographiques du sommeil persistent après la gué-rison, suggérant qu’elles seraient plus liées au profil du maladequ’à l’épisode lui-même. Une fatigue et une somnolence diurnesexcessives sont les conséquences de l’insomnie, et peuvent provo-quer une baisse de productivité et un risque accru d’accidents. Untrouble du sommeil persistant est associé à un risque significa-tif de récurrence, un risque accru de suicide et une évolution glo-

bale plus mauvaise. La plupart des antidépresseurs provoquentdes changements substantiels des aspects subjectifs du sommeilainsi que des paramètres polysomnographiques. Les antidépres-seurs tricycliques et les inhibiteurs de la monoamine-oxydaseclassiques suppriment à des degrés divers le sommeil paradoxal(REM sleep). Les inhibiteurs sélectifs de la recapture de la séro-tonine (ISRS) suppriment généralement le sommeil paradoxal,mais sont souvent accompagnés de plaintes subjectives d’insom-nies. L’insomnie chez les patients atteints de dépression majeurerequiert une évaluation rigoureuse. Il est important d’exclure uneinsomnie primaire, ainsi que les causes médicales générales etcelles liées à une substance. Les interventions pharmacologiqueset non pharmacologiques font partie des stratégies thérapeu-tiques. Certains praticiens prescrivent un antidépresseur ayantdes propriétés sédatives. Autrement, un hypnotique peut être as-socié. Aucune de ces approches n’est dépourvue d’inconvénient.Une bonne hygiène du sommeil et d’autres traitements compor-tementaux peuvent être des auxiliaires utiles.

L’IMPACT DES TROUBLES DU SOMMEIL SUR L’ÉVOLUTION DE LA DÉPRESSION

279Hygeia versus Polymnia: some French painters and their diseases – Régnier MEDICOGRAPHIA, VOL 27, No.3, 2005

n the year 1700, Bernardino Ramazzini (1633-1714), doc-tor of medicine and philosophy, holder of the chair of prac-

tical medicine in Padua, published De Morbis ArtificiumDiatriba [Concerning Work-Related Diseases], the first

“modern” work devoted to occupational illnesses. A secondedition was published on the year of his death (with 12 ad-

ditional chapters), and was translated into French in 1777. Theauthor had ventured into several artists’ studios in the republicof Venice, and described many ailments induced by noxious sub-stances, unhealthy working conditions, and the repetition of un-usual movements.1

Ramazzini noted a marked similarity between the clinical manifestations of painters and those of metalworkers, “which differed only by their lesser intensity in painters.” Chapter 8 of his work was devoted topainters’ diseases such as “tremor, cachexia, black teeth, pallor, melancholy, and loss of sense of smell.”“Often, Ramazzini noted, when painting portraits, they endow their sitters with greater beauty and color

A T O U C H O F F R A N C E

ainters are not an exception, they also suffer from diseases. Some occur due to chance or ge-netics, others are related to their artistic activity (professional diseases). The repeated use of

organic solvents and pigments containing heavy metals or harmful substances have in the pastcaused a multitude of diseases due to intoxication (then not understood). The occurrence of disease in

an artist inevitably upsets his or her perception of the world; this upset can prove critical in painters whoexpress their impressions and feelings in their work. Psychiatric disturbance, loss of vision, neurologicaldisorders, articular ailments, and painful syndromes can transform, modify, or even interrupt a painter’soeuvre. Several French painters have been affected by diseases that gradually influenced their artistic work.For example: Édouard Manet (locomotor ataxia), Edgar Degas (loss of vision), Auguste Renoir (rheuma-toid arthritis), Vincent van Gogh (character disorders), Henri de Toulouse-Lautrec (pyknodysostosis),and Raoul Dufy (rheumatoid arthritis). However, the different histories show that disease was never anobstacle to the expression of true artistic genius; in the best case, the painter adapted to his condition byfinding new ways to express his art, or, in the worst case, died young while exploiting the time he had leftby an explosion of creativity. Medicographia. 2005;27:279-287. (see French abstract on page 287)

✦

Christian RÉGNIER, MDPraticien Attaché des Hôpitaux de ParisSociété Internationale d’Histoire de la Médecine9, rue Bachaumont 75002 Paris, FRANCE (e-mail: [email protected])

Hygeia versus Polymnia*

Some French painters and their diseases

b y C . R é g n i e r , F r a n c e

P* Hygeia, daughter of Asclepius, the Greek god of medicine, was renowned for her ability to prevent disease; in classical times shewas known as the goddess of health. Originally, Polymnia was the muse of lyrical poetry and glorified the gods and heroes. Shebecame the muse of painting in the 17th century when the painter Jacques de Stella portrayed her in his painting Minerva visit-ing the Muses (1640). Jacques de Stella wanted to show that the art of painting ought to be raised to the same rank as the otherintellectual/artistic disciplines.

17th centurywatercolor entitledThe Drip. A painter

(with his canvas un-der his arm) moveswith difficulty lean-

ing on a cane, hishands and back-

bone are deformed,he has a drip at the

end of his nose.© Private collection.

All rights reserved.

I


280 Hygeia versus Polymnia: some French painters and their diseases – RégnierMEDICOGRAPHIA, VOL 27, No.3, 2005

than nature has in fact endowed them, as they them-selves lack color and portliness.” Ramazzini also claimedthat the sedentary life that painters led and “the fantas-tic ideas that perturb them resulted in a sort of melan-choly genius.” He remarked on the “obnoxious smell oflatrines that permeated artists studios (…) they contin-ually breathe in these pernicious vapors that (…) disturbthe management of the natural functions and lead to allthe diseases we have noted.”1

A Danish study of 1988 confirmed these observations:painters who used bright colors (rich in metallic pig-ments) were frequently affected with articular disease.2

There were also the “lead colics” (known since ancientGreece) and observed in the 16th century by Jean Fernel,personal physician to Henry II, who described a painterof Angers who had the habit of sucking his paintbrushesto clean them.1

Edouard Manet’s ergotism Manet died on April 30, 1883 at the age of 51 years. Hisleft leg had been amputated several days previously be-cause of gangrene.

In 1880, Manet’s doctor, François Siredey (specialist inwomen’s diseases), advised him to start hydrotherapy totreat clinical signs of incipient locomotor ataxia. Werethese in fact manifestations of tertiary syphilis, which is characterized by disorders of coordination andequilibrium, paralysis, and proximal anesthesia of the lower limbs? Manet regularly attended shower ses-sions in the hydrotherapeutic establishment of Doctor Joseph Béni-Barde in Auteuil. “When the Béni-Bardeuses (thus he nicknamed the female attendants) see me running down the steps of the swimmingpool and laughing, I will have won—and that could be quite soon,” he wrote to his friend Antonin Proust,who had been a school pal at Rollina College when they were ten. When Proust became minister of arts,in 1880, he awarded Manet the Legion of Honor.3-5

In 1881, Manet was treated in Meudon at the Bellevue Clinic. These cures failed to provide the hoped-for benefits, so he switched to another doctor who prescribed rye smut (ergot), of which there then threeforms: Bonjean’s Ergotin, Yvon’s Ergotin, and Tanret’s Ergotinin. The results were spectacular, and Manetdecided to increase the dosage to what were horrifying levels according to his friend, the painter CamillePissarro.6 The clinical signs of ergotism soon set in: hallucinations, delirium, circulatory disorders of thelower limbs, and the famous Saint Anthony’s fire that burned his legs.

From 1881, Manet regularly used a walking stick, and the following year he had great difficulty gettingabout. Doctor Siredey warned Antonin Proust that Manet should avoid such blatant overdosing. Holi-

daying in Meudon in the summer of1882, Manet had become almost im-potent.4,7 At the end of march 1883,his left foot had become gangrenous.Manet’s surgeon René Marjolin anddoctors Siredey and de Berio, were infavor of amputation. On April 10, theycalled in Paul Tillot, a well-known sur-geon at Beaujon Hospital, and AristideVerneuil, professor of clinical surgeryat La Pitié Hospital, both members ofthe French Academy of Medicine, fora consultation at Manet’s home in

Man on Crutches, sketch by Édouard Manet (1878).© Metropolitan Museum of Arts. New York.

Ergot(rye smut).Engraving

taken fromthe Littrédictionary

(1865). © Coll. Inter-

Activités.

Spontaneous arterial gangrene, engraving taken from the atlas Patho-logical Anatomy of the Human Body,by Jean Cruveihlier (published between1830 and 1842 by J.-B. Baillière, Paris,France). © BIUM Paris.



Paris. The two surgeons decided that amputation of the left leg was necessary and urgent, and this wascarried out, one week later, at the painter’s home, under anesthesia with chloroform. However, his condi-tion did not improve. Racked with fever chills, Manet suffered stoically while the world of art came to paya final visit. A medical bulletin was regularly posted on the front door of his house.5,7 His sister-in-law,Berthe Morisot, described the two days he took to die as unbearable. On April 30, at 7 o’clock in the evening,Manet died in the arms of his son all the while complaining about pain in his absent leg.8

Ergot is a parasitic fungus, Claviceps purpurea, of the rye seed. It is known for its action on smooth mus-cle fibers (uterus, arteries), but when ingested in large amounts it causes ergotism. Known in the MiddleAges as Saint Anthony’s fire, the disease ultimately led to dry gangrene of the feet and hands. The FrenchLittré dictionary (1865) noted: “in the presence of gangrene, the strongest antiseptics must be used; am-putation is rarely successful.”9

Edgar Degas’s loss of vision: the art of adaptingTo the attentive observer, the creative evolution ofEdgar Degas (1834-1917) and the variety of tech-niques he used were clearly shaped by the courseof his eye ailment, a handicap that he adroitlymanaged to turn to his advantage. The painter andcritic André Lhote extolled “the second Degas, thatof the latter years, when his failing eyes forcedhim to embrace the essential,”10 and for AugusteRenoir, “Degas painted his best canvases when hecould no longer see.”11 Degas was treated by twoof the most famous ophthalmologists of his day,Edmond Landolt and Charles Abadie, but no med-ical record has yet been found. It was during theFranco-Prussian war that Degas, then aged 35,became aware of a loss of vision in his right eyeduring firing practice. His contemporaries, no-tably Edmond and Jules de Goncourt, noticed thatthe painter half shut his eyes when looking at dis-tant objects. This was diagnosed as myopia, a vagueterm, which at the end of the 19th century loose-ly applied to a great number of cases of “poor vi-sion.”12 In his study Les Yeux des Peintres [TheEyes of Painters] (1999), Philippe Lanthony, anophthalmologist who specializes in color visionat the Quinze-Vingts Hospital in Paris, has reser-vations about the diagnosis even though it wascompatible with Degas’s habit of painting subjectsclose up (and his horror of painting in the openair). Examination of the glasses prescribed by Landolt (now in the Orsay Museum) revealed only a smallcorrection of 1.5 diopters. The hundred or so portraits of Degas show him wearing glasses outside onlytwice and never when he is reading.13 According to Dr Lanthony, two diagnoses can be excluded: ambly-opia, which normally does not develop at a late age, and myopic maculopathy, in which the myopia ismuch more severe.12 The very special lighting and the representation of space in Degas’s work suggestmonocular vision; this defect in the perception of depth could also explain the frequent presence of mir-rors in his paintings.

In November 1871, writing to his friend James Tissot, Degas described a sudden loss of vision in his lefteye: I recently suffered, and still do, from a sudden weakening and disorder in my eyes. It began beside thelake at Chatou under a bright sun while painting a watercolor. And it made me lose three weeks, duringwhich I couldn’t read, travel, or even go outside.”13 The clinical details are precise: sudden loss of vision,a disorder resolving spontaneously in a few weeks, and absence of other clinical signs (pain, watery eyes,edema). Degas’s ophthalmic disorder developed progressively, with loss of sharpness of vision, photopho-bia (circa 1873), scotoma (circa 1884), and reduced color vision (circa 1895).

His eyes saw only one part of an object at a time; planes were cut up like a trellis (…) He saw some parts veryclearly while others were a complete blur (…). He told a young painter, Maurice Denis, I see your nose, butI do not see your mouth.12

Absinth (1876), canvas by Edgar Degas; has the artist depicted himself on the right of the drinker? All thetechniques used to compensate for his visual difficultiesare present: depth (with presence of the window), unusualviewing angle. © Musée du Louvre, RMN. Paris.



Despite considering himself nearly totally blind (he had his newspaper read to him by his housekeep-er), Degas nonetheless continued his social activities: visits to the Opera, journeys to Spain and Morocco(1889), participation in and organization of impressionist exhibitions. Similarly, his production of paint-ings was considerable up until 1895. Working in semidarkness, wearing glasses with strongly tinted lenses,Degas used magnifying glasses to examine photographs and drawings that he enlarged himself by succes-sive tracings and used for painting. His last portraits (Renoir, Mallarmé, Halévy, Rouart) were all paintedfrom enlarged photographs.12,13

Some people even went so far as to doubt whether Degas really had poor vision, like Claude Monet, whoconsidered him a “joker,” or the art dealer Ambroise Vollard who said, “Degas pretended not to see in or-der to avoid being obliged to recognize people.”14 Even his brother was a doubter: “He does not suffer fromany disease, except deafness.”13 However, Degas’s productivity and social activity declined after 1895; hethen began modeling statuettes, which gave him a better perception of three-dimensional objects thanksto the sense of touch. Degas noted: “My infirmities have prevented me from continuing to paint with oils,so I have to be satisfied with pastels.” By adopting this technique and using bright colors, he could carry

one, but he finally had to give up when he could no longer seethem.15 The diagnosis of retinochoroiditis was put forward by thepainter Maurice Denis who suffered from the same condition; to-day this hypothesis seems the most probable.12

Auguste Renoir: “one does not need hands in order to paint!”The first symptoms of polyarthritis struck the painter in 1901.His hands with their protruding bones beneath the wrinkled skinhad fingers that were twisted forward and wrapped in tight ban-dages to prevent the nails from penetrating the palms. The firstphalanx of the right thumb was bent back, having been dislocat-ed and badly reduced. It was separated from the paintbrush in-serted between thumb and forefinger by a thin corn plaster. It wasunder such conditions that Renoir painted during the last yearsof his life, without being the least discouraged, with the same pas-sion as in his youth, the same love of life, the same fervor, and thesame enjoyment.16 Renoir was treated with antipyrine, an analgesic

much in fashion at the start of the 19th century. He also took the waters at the spa towns of Bourbonne-les-Bains and Aix-les-Bains. He consulted the famous Doctor Gachet, physician and “protector” of HonoréDaumier, Vincent van Gogh, Édouard Manet, Armand Guillaumin, Paul Cézanne, Gustave Courbet, ClaudeMonet, Alfred Sisley, and Camille Pissaro.17

When his art dealer, Vollard, visited him in Cagnes, Renoir declared, “As you can see, hands are not need-ed for painting! The hand, it’s a lot of nonsense.”14 His son, the filmmaker Jean Renoir, noted in his mem-oirs that from 1911 on his father was not mobile, even with the aid of crutches. Contrary to legend, thepaintbrushes were not fixed to his hands, simply his palms were protected by a gauze bandage and tal-cum powder. Every day, Renoir was carried from his bed to his wheelchair and settled in front of his can-vas.11,18 Despite his physical sufferings and the difficulty of finding the ideal body position for applying thebrush, he painted right up to the end of his life without his work suffering from his deforming arthropa-thy. On the morning of his death, he painted a bouquet of anemones with shimmering colors.

A lover of bright tints (cinnabar, Naples yellow, ultramarine), Renoir naturally used numerous metallicpigments. A chain-smoker, he rolled his own cigarettes, unaware that each time he deposited the toxiccompounds of his paint-impregnated fingers on the cigarette paper. In fact, 0.05 mL of red cinnabar con-tains 100 to 150 mg of mercury sulfide, and repeated inhalation several times a day would be 100 timesmore toxic than the official toxic limit of mercury permitted in the air (0.01-0.1 g/m3). Painters ingestedmetallic pigments by smoothing their paintbrush with their mouth, and by drinking, eating, and smokingwithout washing their hands. The drying of clothes stained with paint and the burning of old canvasesin the stove also gave off toxic metallic vapors.2

In 1988, two doctors and two professors of the Copenhagen School of Fine Arts published in the Lanceta study that established a correlation between the onset of certain rheumatic diseases in painters and in-toxication by metallic pigments (cobalt, copper, manganese, lead, chromium) present in certain brightcolors. The Danish authors compared the use of colors in two paintings by Claude Monet with the paletteof Renoir. Thus, in the Skaters in the Bois de Boulogne (1868) and in the Woman with Umbrella (1875),Monet used very few bright colors such as blue, red, and yellow. The dominant colors were gray, olive green(iron silicates), ochre (iron oxide), blue gray and black (bone), which contain very little or no toxic heavy

In the movie Ceuxde Chez Nous (1917),

Sacha Guitry filmedPierre-Auguste Renoir,

whose fingers, deformedby polyarthritis, can be

clearly seen. © Ciné-mathèque Française.



metals. The same comparison can be made with Edgar Degas who used more pigments containing ironand carbon than his contemporary Renoir who was very exposed to heavy metals (mercury, arsenic, cad-mium, tin, antimony, cobalt). The authors showed that heavy metals that entered the organism inducedenzymatic inhibition, denaturation of proteins, or immunosuppression, all pathophysiologic processesimplicated in the onset of certain rheumatologic diseases.2

Vincent van Gogh: a victim of pathologists?Vincent van Gogh (1853-1890) was the son of a Calvinist preacher, and three of his paternal uncles wereart dealers. The life of the painter was punctuated with tragic adventures and dismal experiences, whichled to him killing himself with a bullet to the thorax on July 27, 1890 in Auvers-sur-Oise, where he was be-ing treated by Doctor Paul Gachet. His physician was a cultivated art collector and depressive freethinker,who was infatuated with psychiatry and a mediocre practitioner. van Gogh’s tragic end contributed tothe legend of the artist who had painted 44 self-portraits and kept up a regular correspondence with hisyounger brother Théodore, who died in the Utrecht asylum one year later.7,19,20

It all began with the “drama of the excised ear.” At Arles, on the evening of December 23, 1888, Vincentvan Gogh went up to his room in the Yellow House, sharpened his razor, and cut off the lower half of hisleft ear. He then covered his head with damp towels and a bonnet, wrapped the piece of ear in paper andwent to a brothel in the rue Bout d’Arles where he frequently visited a certain Rachel. He found the younggirl, offered her his grisly present and declared: this is a souvenir of me, look after it carefully. Rachel im-mediately fainted. The next morning, van Gogh was found unconscious at his lodgings, his head wrappedin bloodstained towels. Agitated and menacing when awoken, he was taken to Arles Hospital and put ina padded cell. The amputated ear was collected by a policeman, and the intern, Félix Rey rapidly sutured

SOME METALLIC PIGMENTS USED IN PAINTING

Known in Egypt around 8000 years BC, certain metallic salts (mercury, copper, arsenic, tin, iron)mixed with linseed oil, which forms nearly 80% of the pigment. From the Roman Empire to the Re-naissance, new metals were used as pigments (lead, antimony, cobalt, chrome, cadmium, manganese,aluminum).

◆ Cinnabar: a red vermilion color, derived from mercuric sulfide (HgS).◆ White lead: white pigment composed of lead carbonate (2PbCO3; Pb(OH)2).◆ Bright yellow: cadmium or lead sulfide (Pb CrO4).◆ Naples yellow: prepared from lead antimonate (Pb3 (SbO4)2). ◆ Violet: prepared from manganese ammonium phosphate and cobalt arsenate.◆ Verdigris: copper acetate.◆ Cobalt blue: composed of cobalt oxide and alumina (CoO; Al2O3).◆ Ultramarine: aluminum-based (Na6-10 Al6 Si6 O24 S2-4).◆ Oxide blue: prepared by exposing silver strips to a mixture of salt, vinegar, alkali, and alum.

The Yellow Room byVincent van Gogh

(Arles, 1888) where thedrama of the ear

unfolded. © Muséevan Gogh, Amsterdam.



the missing part back into place, and the wound healed without complications. On January 7, 1889, theintern noted in the discharge register: “A form of epilepsy characterized by hallucinations and episodes ofconfused agitation that were precipitated by alcohol excess.” During his stay in Arles, van Gogh had beenencouraged by Paul Gauguin to consume quantities of alcohol and to frequent the bordellos. Having goneto Provence in the hope of founding a painting workshop like the school at Pont-Aven (where Gauguin hadbeen a leading light), van Gogh had been profoundly disappointed by the hostility of Gauguin to the pro-ject. The two men had separated on bad terms a few days before the “drama of the severed ear,” whose ver-sions are as numerous (and contradictory) as the hypotheses explaining the event.17,19,21

On May 8, 1889, van Gogh was interned, at his own request, in the Saint-Paul-de-Mausole Asylum atSaint-Rémy-de-Provence. For one year he enjoyed the tranquillity and the comprehension of the nunsand personnel of the establishment. Appreciating the quality of the light and the glowing beauty of the set-ting (the asylum is a jewel of Provençal Romanesque art), van Gogh produced nearly 150 paintings anddrawings, including Irises and Vincent’s Room in Arles. This year spent in Saint-Rémy is considered oneof the important creative periods in the oeuvre of Vincent van Gogh.

The first diagnoses of the painter’s neuropsychiatric disorders were made by Doctors Rey and Urpar (Arles)and Peyron (Saint Rémi), who suggested epilepsy with visual and auditory hallucinations associated withacute mania with generalized delirium.22 Gachet spoke simply of unwise exposure to the sun plus chronicintoxication with turpentine and camphor (with which he impregnated his pillow in order to find sleep).Numerous other hypotheses by a host of doctors followed: cerebral tumor (Bader), psychomotor temporalepilepsy (Gastaut), schizoid syphilis (Eichbaum), precocious dementia (Bychowski), degenerative psychosis(Hutter), Ménière’s disease (Yasuda then Arenberg and colleagues), tuberculous meningoencephalitis(Dupinet), thujone intoxication (Arnold), schizophrenia (Kahn et Jaspers), melancholia (Michel).21-25

It seems that van Gogh was treated for psychomotor epilepsy with extracts of digitalis, and some believethat the intoxication could explain the visual disorders like dyschromatopsia and xanthopsia, and hencethe painter’s trademark yellow suns surrounded with circles and haloes, and the extraordinary yellow ofthe cornfields.26 In the two portraits of Doctor Gachet, van Gogh depicted a flower of the purple foxglove

(Digitalis purpurea). Other “pathobiographies” attribute the chro-matic particularities of the painter’s palette to cataract, corneal dys-trophy, or glaucoma (Maire).12,21 In a work devoted to van Gogh,Professor Henri-André Martin, painter, otorhinolaryngologist, hon-orary professor at the Lyon Faculty of Medicine, concluded:

It is wrong to want to understand everything and resolve van Goghlike an elementary mathematical problem. It would be unseemly towant to inflict a positivist analysis on his person and on his work.20

Henri-Marie de Toulouse-Lautrec: the misfortunes of the “little chap” The Montmartre painter Toulouse-Lautrec (1864-1901) suffered allhis life from his ungracious appearance. When about 8 years old,his weak health and multiple fractures had already given him aclose acquaintance with doctors and spas. A victim of all kinds ofabuses, he died aged 37, worn out and in despair.

His father, Alphonse de Toulouse-Lautrec Montfa, descendant of Raymond IV, Count of Toulouse, andhis mother Adèle Tapié de Céleyran were first cousins. Henri-Marie was born on November 24, 1864, inthe family mansion of Bosc situated on the ramparts of Albi. The baby seemed so healthy that he was nick-named “lou poulit” (patois for little beauty) and because of his insatiable appetite he was deemed worthyof the “legendary Lautrec stomach.” 27-29

In 1872, Toulouse-Lautrec studied at the Lycée Fontanes (now Condorcet) in Paris—where his palscalled him “little chap.” He then suddenly began to suffer from unexplained feverish attacks with severepains in the knees and hips. On the advice of doctors, he began taking thermal cures at Amélie-les-Bains,Plombières, Lamalou-les-Bains, Evian, Guyon, and Barèges, accompanied by his mother. In March 1877,he was subjected to an “electric brush treatment” that had previously cured his uncle Charles.30

In Albi, on May 30, 1878, Toulouse-Lautrec slipped, fell, and fractured his left thighbone. The family doc-tor set the limb with a plaster cast. After one month of immobilization and an equally long convales-cence, he could only walk lopsidedly, “like a duck,” as he noted in a letter to a friend. A Toulouse-Lautrecdoes not walk with a cane! roared his father when he happened to notice the instrument unworthy of thesporting reputation of the family (to the great despair of the count, his son had never mounted a horse).During the holidays in July 1879, at Barèges (Hautes-Pyrénées) while walking with his mother, he slippedinto a ravine and fractured his right thighbone; another immobilization, another convalescence. From

This small low chair(12 inches), purchased in

1876 in Bagnères deBigorre, is the chair from

which Toulouse-Lautrec felland began the series of acci-

dents contributing to hisdisability. © Albi. Maison

Natale de Toulouse-Lautrec.



that date on, his growth suddenly slowed. His lower and upper limbs ceased to grow even though his trunkgained 5 cm. When 20 years old, in 1884, Toulouse-Lautrec was only 5 feet tall, about one inch less thanthat required for military service (decree of November 30, 1872). The average height of conscripts in 1885was about 64 inches. He was already what would later be called: the “greatest dwarf of the century.” 28-30

In addition to his small size, Toulouse-Lautrec’s physical appearance was unsightly: he concealed hisvast head beneath a large hat and hid the hypoplasia of his lower jawbone with an abundant beard. He alsohad to wear glasses in order to paint. Yvette Guilbert, his favorite model, described him thus:

He was small in size, had an enormous brown head, a heavily colored face and black beard, greasy, oily skin, anose big enough for two people, with a mouth slashing the face from one side to the other, and pinkish-violet,flat and flabby lips that lined his fearful orifice.27,31

Toulouse-Lautrec sketched his muse as he saw her: because he had exaggerated her scrawniness andpointed nose, Yvette Guilbert revenged herself by calling him “the little monster,” without realizing it wasthe artist’s talent that would save her from oblivion.32

The etiology of the disease remains hypothetical. The consanguinity, the bone fragility, the late and un-equal bone growth, and the craniofacial deformations are the main characteristics of Toulouse-Lautrec’sdisease. After having eliminated the diagnoses of premature puberty and ricketts, biographers unani-mously class the disease among the genotypic chondrodystrophies. In the absence of x-rays and biologicalor genetic examinations, the retrospective diagnoses are as follows:

– osteogenesis imperfecta (Seedorf - 1949),27

– achondroplasia (Séjournet - 1955),29

– Morquio’s disease (Lévy - 1957),33

– multiple epiphyseal dysplasia (Flament - 1965),27

– pyknodysostosis (Marotteaux and Lamy - 1966),27,34 to-day this diagnosis is the unanimously accepted diagnosis.

Séjournet, the biographer of Toulouse-Lautrec, wrote,“Like Goya, like Ensor, he draws from his despair a forceagainst the destiny that isolates him. Thus to our eyesand forever, his paintings: Mélinite (an explosive), JaneAvril, cabaret dancer; La Goulue (The Glutton), LouisWeber, lead dancer at the Moulin Rouge; Grille d’Égout(The Sewer Grating), a dancer with terrible teeth; Valen-tin le Désossé (The Boneless One) male dancer; seem toecho the pathetic and derisory vortex of his own dancewith death.”29

Henri Toulouse-Lautrec takingthe pulse of his friend MauriceGuibert under the watchful eyeof his cousin Gabriel Tapié deCéleyran, who was an intern.Lautrec often said: If I wasn’t a painter, I would like to be adoctor, and his insatiable penciloften sketched the interns inthe staff waiting room or in thewards. In Paris he dined everyFriday evening at Drouant’sRestaurant in the company ofartists and doctors, includingGeorges Clemenceau and Henri Vaquez. From 1891 to1895, every Saturday morning,he accompanied his cousinGabriel (who was completinghis medical studies) to the department of Docteur ÉmilePéan in the Saint-Louis Hos-pital.32 © Albi. Maison Natalede Toulouse-Lautrec.

Henri de Toulouse-Lautrecin the company of Zidler,

manager of the Moulin Rougearound 1891. © Albi. Musée

Toulouse-Lautrec.

An Examination at the Faculty of Medicine. Oils on canvas(26� 32 inches) painted in July 1901, the year Toulouse-Lautrec died. © Albi. Musée Toulouse-Lautrec.



The Dufy “miracle” Raoul Dufy (1877-1953) suffered his first arthralgias in childhood. His rheumatoid polyarthritis devel-oped slowly from 1935 until 1948, when his functional disability was already well advanced. He was treat-ed with injections of gold salts, underwent vertebral manipulations, and was treated at spas in Spain. In1948, he suffered very intense articular pains in the wrists, knees, and ankles. However, he could still usehis left hand. He has not lost his subtle touch with lines or his fine strokes. “He draws with a pencil, andjust lightly strokes the paper, using his right hand as a support,” noted his friend Pierre Courthion, a well-known art critic.35

Since he used bright colors in his work, the painter was also a great consumer of metallic pigments.

Dufy, as everyone knows, is one of our greatest colorists. He has the magic touch. It is his favorite form ofexpression, and it takes precedence over all the other forms, which explains his novelty. Color gleams in hiswork, light spreads out with a richness, warmth, and life that are comparable, in their own way, to the colorsof Veronese and Rubens that Delacroix so admired. Everything is fresh, alive, clear, joyful as the spring in na-ture or youthfulness in life,

wrote Pierre Camo in 1947 in his Dufy the Enchanter.36 The following year, Jean Cocteau wrote a pamphleton the painter and reminisced on his intense physical suffering:

He had the appearance of a red devil always ready to emerge from a green box. Since then, we have each grownolder in or own corner and in our own way. Dufy became an invalid who fought, and this battle with evil re-placed in a way the youthful ardor.35

On April 11, 1950, the painter, almost bedridden, embarked for Boston to consult Doctors Homburgerand Bonner of the Jewish Memorial Hospital. Raoul Dufy was already famous on the other side of theAtlantic, and his photograph had appeared in Life Magazine. Homburger and Bonner had noted the de-formed joints and suggested he should come to the United States and be treated with a “revolutionary”therapy.36 Several days before his departure, Dufy wrote a poem that could be considered his artistic legacy:“Tomorrow, I sail to Boston in search of my hands…” He also painted a superb bunch of brightly coloredflowers entitled Cortisone (which harks back to the flowers painted by Auguste Renoir on the morning ofthe day he died). That same year, 1950, Philip Hench, Head of the Department of Rheumatic Disease atthe Mayo clinic in Rochester, received the Nobel Prize for Medicine or Physiology for his discovery of thehormones of the suprarenal cortex (cortisone and adrenocorticotropic hormone [ACTH]), together withtheir structures and biological effects. He shared the prize with the biochemists Edward Calvin Kendalland Tadeus Reichstein, professor in Zurich and Basel.

At the beginning of June 1950, Raoul Dufy began a series of injections of 100 mg per day of cortisoneacetate for 3 weeks. On his return to France at the end of July, he took cortisone by mouth at the dose

Yesterday the invitation arrived Drawing me to Dr Homburger And his hospital in America. Tomorrow I sail to Boston in search Of my hands. Forty years ago I painted with Braque

at l’Estaque. The cubist edges cut too deep into

my palette And I joined Matisse and became a Fauve. I covered my canvases only with the

colors that I could feel Avoiding those that I couldn’t see. Unsatisfied I traveled to Munich,

Normandy, Marseille Saw the terrible women of Avignon And found myself on the Riviera in Vence. Fifteen years ago the attacks began. Disfiguring my hands As if I were painting with leaden gloves.

I sought gold Injections. Underwent spinal

manipulations And made a pilgrimage to a Spanish spa. Yet my hands grew Still. I continued studies in my mind Waiting for the time When my brush could be awakened To decorate the canvas With arabesques and flourishes. I have long tried to capture The motion of race horses and regattas The movements of the orchestra As musicians chase the notes In concertos of yellow and red. I have painted Electricity With dynamos and electrons moving

at the speed of light. Craving more speed I released my canvases

To color silk scarves and dresses And finally saw my work sail with the

wind. Now my hands are splinted And I have become disjointed An old man whose hands will not obey

his heart. Rheumatism loosened Renoir’s brush

from his grasp And his son bound the bristles to the

hollow of his hand. But Dr Perles has assured me That the new remedy, cortisone and ACTH, Will release the bonds Of the arthritis that have held me in place. My only wish: to draw freehand Following the wind Flowers that beckon me And capturing them In a vase of Anemones.

POEM WRITTEN BY RAOUL DUFY IN CHANTILLYSEVERAL DAYS BEFORE HIS DEPARTURE FOR THE UNITED STATES IN APRIL 1950

of 100 mg twice a week.37,38 Dufy also received ACTH, but the form and dosage are not known. To expresshis thanks, Dufy painted the portrait of the American Doctor Freddy Homburger, his “savior.”

Despite the occurrence of secondary complications, Dufy benefited very rapidly from the revolutionarytreatment: he was again able to move around on his crutches, draw, squeeze the tubes of paint by himself,and renew work on abandoned canvases such as Amphitrite begun in 1935.14,37 Fully aware that he hadbenefited from a “therapeutic miracle,” Dufy wrote: “Is it a rebirth or a swan song?” Indeed, he was to suf-fer from many complications, such as abscess at the injection site, diarrhea, and severe stomachache.38

Raoul Dufy died on March 23, 1953, in his house at Forcalquier (Haute Provence Alps) of intestinalhemorrhage probably related to corticotherapy. ❒



REFERENCES1. Ramazzini B. De Morbis Artificum Diatriba [Des Maladies duTravail] 1700, translated into French by de Fourcroy A. Paris,France: AleXitère; 1990.2. Pedersen LM, Permin H. Rheumatic diseases, heavy-metalpigments, and the great masters. Lancet.1988;4:1267-1269.3. Manet J. Journal (1893-1899). Paris, France: Scala; 1987.4. Moreau-Nelaton E. Manet Raconté par Lui-Même. Paris,France: Plon; 1926.5.Proust A. Edouard Manet—Souvenirs.Paris,France:L’Échoppe;1996.6. Duret T. Histoire d’Edouard Manet et de son Œuvre. Paris,France: Hachette; 1906.7. Darragon E. Manet. Paris, France: Fayard; 1989.8. Perruchot H. La Vie de Manet. Paris, France: Gallimard; 1959.9. Littré E, Robin C. Dictionnaire de Médecine, de Chirurgie,de Pharmacie, des Sciences Accessoires et de l’Art Vétérinaire.Paris, France: J.-B. Baillière et fils; 1865.10. Lhote A. Peinture d’Abord. Paris, France: Denoël; 1942.11. Renoir J. Pierre-Auguste Renoir, Mon Père. Paris, France: Gal-limard; 1981.12. Lanthony P. Les Yeux des Peintres. Lausanne, Switzerland:L’Âge d’Homme; 1999.13. Boggs JS, Loyrette H, Pantazzi M, Tinterow G. Degas. Paris,France: Réunion des Musées Nationaux; 1988.14. Vollard A. Souvenirs d’un Marchand de Tableaux. Paris,France: Albin Michel; 1937.15. Rouart D. Degas à la Recherche de sa Technique. Geneva,Switzerland: Skira; 1988.16. Certigny H, Deslandres Y, Leprohon P. La Vie des GrandsPeintres Impressionnistes. Paris, France: Éditions du Sud; 1964.17. Porot D. Van Gogh ou le Hollandais Volant. Rueil-Malmaison:Ceigy; 1989.19. Lossouarn M. Médecine et Peinture. Rapports Historiques etActualisation. Medical Thesis No. 323; Brest, France. 1976.20. Van Gogh V. Lettres à Théo. Paris, France: Gallimard, L’Ima-ginaire; 1988.21. Martin HA. La Maladie de van Gogh: Le Mystère d’une FinTragique. Paris, France: Buchet-Chastel; 1994.

22. Lee TC. Ce que voyait van Gogh: une intoxication digitalique?JAMA (French edition).1981;3:1389-1392.23. Arnold WN. Vincent van Gogh and the thujone connection.JAMA.1988;260:3042-3044.24. Arenberg K, Countryman LF, Bernstein LH, Shambaugh GEJr. Van Gogh had Meniere’s disease and not epilepsy. JAMA.1990;264:491-493.25. Gastaut H. La Maladie de Vincent van Gogh Envisagée à laLumière des Conceptions Nouvelles sur l’Épilepsie Psycho-Mo-trice. Cahors, France: Coueslant; 1956.26. Michel FB. La Face Humaine de Vincent van Gogh. Paris,France: Grasset; 1999.27. Corcos M. Van Gogh, une étrange flamboyance. Gazette Med.1990;97:73.28. Darley P. La Pycnodysostose. Approche Diagnostique à Pro-pos d’une Nouvelle Observation et Revue de Littérature. La Ma-ladie de Toulouse-Lautrec? Medical Thesis. Dijon, France; 1978.29. Flament E. Henri de Toulouse-Lautrec. Pathologie de saCroissance. Hypothèses. Medical Thesis. Paris, France; 1966.30. Sejournet G. La maladie de Toulouse-Lautrec. Presse Med.1955;63:1866-1867.31. Desvoisins L. Henri de Toulouse-Lautrec, Essai d’Étude Cli-nique, ses Maladies, sa Mort. Medical Thesis, Montpellier, France;1958.32. Beaute G. Toulouse-Lautrec Vu par les Photographes. Lau-sanne, Switerland: Edita; 1988.33. Pasteur Valéry-Radot L. La médecine et les médecins dansl’œuvre de Toulouse-Lautrec. Presse Med.1951;59:1021-1022.34. Levy G. Réflexions sur la maladie de Toulouse-Lautrec. SemHop.1957;33:2691-2696. 35. Lamy M Marotteaux P. Les Chondrodystrophies Génoty-piques. Paris, France: L’expansion scientifique française; 1960.36. Courthion P. Raoul Dufy. Geneva, Switzerland: P. Cailler; 1951.37. Camo P. Dufy l’Enchanteur. Paris, France: Éditions Margue-rat; 1947.38. Cogniat R. Dufy. Paris, France: Flammarion; 1962.39. Homburger F, Bonner CD. The treatment of Raoul Dufy’s ar-thritis. N Engl J Med. 1979;301:669-673.

HYGIE CONTRE POLYMNIE : DES MALADIES DE QUELQUES PEINTRES FRANÇAIS

omme tous les hommes, les peintres ont souffert de maladies: les unes survenant selon les loisdu hasard ou de la génétique, les autres étant liées à leur activité artistique (maladies « profes-sionnelles »). L’emploi répété de solvants organiques et de pigments contenant des métaux lourdsou des substances délétères ont probablement généré quantité de pathologies par intoxication(alors inconnues). La survenue d’une maladie chez un artiste brouille inéluctablement sa percep-tion au monde; ce bouleversement prend toute son importance pour le peintre qui restitue ses im-

pressions et ses sentiments par des techniques plastiques. Troubles psychiatriques, malvision, affectionsneurologiques, maladies articulaires, syndromes douloureux constituent des évènements susceptibles detransformer, modifier ou interrompre une œuvre. Parmi les peintres français, plusieurs d’entre eux fu-rent affectés par des maladies qui influencèrent peu ou prou leur production artistique. On peut citer lesexemples d’Édouard Manet (ataxie locomotrice), d’Edgar Degas (troubles de la vision), d’Auguste Renoir(polyarthrite rhumatoïde), de Vincent van Gogh (troubles du caractère), d’Henri de Toulouse-Lautrec(pycnodysostose ?), de Raoul Dufy (polyarthrite rhumatoïde). Pour autant, les différentes pathobiogra-phies présentées montrent que la maladie ne fut jamais un obstacle à l’expression du génie artistique; aumieux, le peintre s’adaptait à sa pathologie en trouvant de nouveaux moyens pour exprimer son art, aupire, il disparaissait prématurément en profitant de sa courte vie pour faire exploser sa créativité.

✦

C

n the lovely cemetery in Auvers-sur-Oise stand two graves, leaning against the wall. They are cov-ered with a mantle of ivy giving the impression that there is in fact just one grave. Only the simple

inscriptions on the gravestones and the rather sad bunch of plastic sunflowers indicate that underthis little corner of soil and sand facing the gentle valleys of the Oise lie the remains of Vincent van

Gogh and his brother Theo. Auvers is only 30 kilometers north of Paris, and it was to this village set inthe peaceful countryside that Vincent van Gogh arrived by train on May 21, 1890. It was here, too, 70

days later, on July 29, that he died in the arms of his brother. During this short period of time, Vincentproduced 77 paintings, some of which feature amonghis most important works, as well as 30 drawings andan engraving.

Grief-stricken, Theo died six months later in Utrecht,on January 21, 1891. His widow, Johanna requested thatTheo’s remains be buried alongside his brother’s in thegraveyard of this charming French village. His asheswere in fact transferred to Auvers-sur-Oise in 1914.

288 In van Gogh’s footsteps in Auvers-sur-Oise – SpaakMEDICOGRAPHIA, VOL 27, No.3, 2005


ust 30 kilometers (19 miles) from Paris lies the pretty valley of the Oise where Vincent van Goghspent the last weeks of his life before committing suicide. After mutilating his ear on Decem-

ber 24, 1888, he entered the Saint-Paul-de-Mausole Asylum in Provence where he spent nearlyone year. He had only one wish: to flee the heat and blinding sunlight of the south for the milder

climes of the north. This he managed to do thanks to his brother Theo, who recommended him toa physician—Dr Gachet—who could look after him in the attractive little village of Auvers-sur-

Oise where he had a house. Vincent arrived on May 20, 1890 full of enthusiasm for this peaceful, verdantregion. He thought it was just the place he needed and threw himself into his work with zeal. But the respitedid not last long. After alternating between moments of great happiness and periods of profound anguish,Vincent finally shot himself in the chest on the afternoon of July 27, 1890, and died in the night two dayslater in the arms of his brother. During the 2 months he spent at Auvers, he did about 30 drawings, oneengraving, and 77 paintings, out of the 879 that have been attributed to him. The Portrait of Dr Gachet,The Church at Auvers, Daubigny’s Garden, and Wheat Field with Crows, are essential landmarks in thehistory of painting. Medicographia. 2005;27:288-295. (see French abstract on page 295)

J✦

Isabelle SPAAK37 rue des Plantes 75014 Paris, France(e-mail: [email protected])

In van Gogh’s footsteps in Auvers-sur-Oise

b y I . S p a a k , F r a n c e

View of the charming village of Auvers-sur-Oise, which liesto the north of Paris in the valley of the river Oise. It washere that Vincent van Gogh spent the last few months of hislife. The village itself and the surrounding countryside pro-vided van Gogh with inspiration for some of his most famouspaintings. The Oise region was in fact much frequented bypainters in the latter part of the 19th century. In particular,Cézanne painted a view of the village. © Moulu/Sunset.

I

Steeped in memories of its famous adopted son In the village, each alleyway, each meadow, and each empty lane,seems to recall the passage of one of the greatest painters of the19th century. “It was at this window on the first floor that Vincentpainted his only painting of the interior,” explains MaryvonneGrandfils, who shows visitors round the Auberge Ravoux, an un-pretentious building where the artist rented a garret. It was here,under the rooftops, after a night and day of agony, that he died inhis little iron bed. On the afternoon of July 27, 1890, in a field ofwheat and under a burning sun, van Gogh shot himself in thechest before dragging himself back to the inn to die.

Today, 125 years later, on a beautiful September afternoon,Madame Grandfils, the faithful and respectful guardian of thememory of this vulnerable person whose life she knows by heartsuggests we walk round the village and the surrounding country-side, before introducing us to the intimacy of the upstairs gar-ret that has been faithfully restored to give the impression thatVincent has just left it. “Make the most of the beautiful light, gowherever your feet lead you,” she advises. “Don’t hesitate to en-ter the little wood, you will find a quite charming walk.” She isright. The locality exudes calm and peace. Especially off-season,this pilgrimage is a must. One almost wonders whether van Gogh,a Dutchman, has not been adopted as a son of the village giventhe many tributes that are paid to him.

The village and the paintings: mirror imagesFirst of all there are the panels to which his paintings are fixed.They are to be found everywhere in the village, in the main street,and at vantage points. They show exactly where van Gogh’s can-

vases were painted, allowing the person viewing to contrast the reality with the artist’s interpretation,like an art history lesson in front of the original subject. The panels are located in front of the church, op-posite the town hall, and facing Daubigny’s Garden, and are all examples of the chromatic intensity andswirling touch of the “man suicided by society” as Antonin Artaud described him. The poet had been ina state of ecstasy before the work of van Gogh at an exhibition in the Musée de l’Orangerie in Paris in 1947.Artaud was fascinated by “the most genuine painter of all painters, who did not paint lines or forms, butinert natural things that seem to be in a state of violent convulsion.” It is this representation of the worldthat confronts the visitor each time he or she comes across one of these metallic panels and tries to workout the viewing angle and the position of the easel, and looks for differences between the painting and per-spective, until finally one is won over by the “bludgeon with which van Gogh never ceases to strike all formsof nature and objects.” In his paintings, reality is transformed by anguish, by the fundamental anxiety ofall beings confronted by an object. Even the most innocent view of a tumbledown cottage could conceal asecret. The church, peaceful on itshillock, seems twisted in the paint-ing by a mystical force. It vibrateswith an internal light against abackground of a tortured, almostblack, sky. The town hall proudlydisplaying its red, white, and blueflag for the 14th of July celebrationbecomes, in the hands of the painter,


289In van Gogh’s footsteps in Auvers-sur-Oise – Spaak MEDICOGRAPHIA, VOL 27, No.3, 2005

Self-Portrait With Straw Hat. 1887. This is one of severalself-portraits that van Gogh painted during his lifetime, andwas executed during his stay in Paris, the painter’s most prolific time in terms of output, two years before the famous incident when Vincent cut off his left ear. Oil on canvas, 54� 46 cm. Rijksmuseum Vincent van Gogh, Amsterdam(Vincent van Gogh Foundation).

Thatched Cottages at Cordeville.Painted in June 1890. On his arrival in

the Auvers region, Vincent van Goghwas immediately struck by the numberof buildings with thatched roofs in the

area, as they were becoming a rarity, and,indeed, he produced several canvases ofthatched cottages during his brief spell

there. Oil on canvas, 72� 91 cm.Musée d’Orsay, Paris. © AKG Images.

a sinister scene, the representation of infinite solitude in this square with its pennants drooping sadlybetween two spindly trees. This is the state of abandonment that the artist must have felt in this peace-ful village where he continued to fight against the demons he thought he had got away from after his yearat the Saint-Paul-de-Mausole Asylum in Provence.

A safe haven from depression?It was to escape from the blinding light of the south of France that he had taken refuge in this village in theclement Oise valley. But to no avail. Even the innocent Daubigny’s Garden—“one of my most sought-aftercanvases,” he wrote to Theo around July 23, 1890—was metamorphosed under his paintbrush into a seriesof disturbing convolutions among an overabundance of flowers. “Everywhere,” according to Artaud, “thelandscape paintings reveal their hostile tones, their anger at being disemboweled.” Above the village, it isfinally the Wheat Field with Crows that greets the visitor. The place has not changed. No building has cometo disfigure the plot of ploughed land. Even though we know today that this canvas is not the last one paint-ed by van Gogh, there is no denying that references to his suffering are everywhere in the dramatic portray-al of this landscape that at first glance seems innocuous. “There are vast fields of wheat beneath a torturedsky,” Vincent wrote to Theo, “and I did not need to look far to capture the extreme sadness and loneliness.”Each stroke of his paintbrush breathes disquiet, the aberrations of a lost man. Beyond the more or less fan-



Below: photograph showing the church at Auvers-sur-Oise made famous by the paint-ing by Vincent van Gogh. © Paul Cooper/Rex Features/Sunset. Right: van Gogh’s representation in hispainting The Church at Auvers, one ofhis best-known works. Painted June1890. The rather crooked church andthe dark swirling sky create asinister, violent effect. Oil on canvas, 94� 74 cm. Musée d’Orsay, Paris. © AKG Images.

Wheat Field With Crows. This picture, which van Gogh painted in July 1890, is considered by many to be his lastpainting before his suicide, although there is no evidence for this. This belief stemmed from the turbulent style inwhich he painted this piece with its darkening blue sky and the black of the crows, which were thought to portendan ill omen. The three paths in the painting leading to nowhere are believed by some to reflect the impasse vanGogh felt he had reached in his life. Oil on canvas, 50� 100.5 cm. Rijksmusem Vincent van Gogh, Amsterdam;Vincent van Gogh Foundation.© AKG Images.

ciful interpretations that this painting attracts, it nonetheless undoubtedly conceals a few symbolic keys in-cluding the three tracks that disappear in three different directions, all without an exit. Does this reflect thepoint of no return that van Gogh had reached? This is entirely possible. Then there are the crows painted,in the words of Artaud, “truffle black, rich banquet black.” Birds of ill omen that “did not open the door toposthumous glory,” but which allowed him to open “…the door…to an enigmatic and sinister afterlife.”

Though the fields have changed little since van Gogh knew them in the summer of 1890, the reputationof this tortured man has changed dramatically. It was a broken painter who stepped down onto the rail-way platform on May 20, 1890, a hard-up artist who, in his life, had sold only one single painting, The RedVineyard, for 400 francs, but whose works today are among the most expensive on the market.

The 427 million francs paid in New York on May 15, 1990, for the Portrait of Dr Gachet painted 100 yearsearlier, from June 3 to 5, 1890, seem indecent when one remembers the misery of the painter’s existence.The price makes a mockery of the measly 150 francs a month that Theo managed to scrape together as akind of allowance for his brother—a more than modest sum that the painter did not always receive andfor which he felt guilty, especially after Theo became the father of a son, Vincent, born on January 31, 1890,and was having difficulty feeding his little family.

The controversial Doctor GachetThe main figure during the painter’s last few weeks was Dr Gachet, a rather strange character. An eccen-tric practitioner, he befriended the numerous impressionists who lived in the Oise area and collected theirworks. He was also a painter in his spare time, and signed his few inferior works “P. van Ryssel.” Initially,Theo had hoped that a mutual friend, Camille Pissarro, would be able to look after his brother. However,married, with six children, and having just lost his mother, and himself unwell, Pissarro felt he could nottake on someone who needed a lot of attention and was often subject to violent attacks. So, instead, Theorecommended Dr Gachet who had a medical practice in Paris and a house in Auvers-sur-Oise, a favorite

spot for painters and Parisians insearch of green fields. Since April1889, Vincent had dreamed of leav-ing the asylum at Saint-Rémy-de-Provence where he had stayed forone year. He wanted to get back towork as quickly as possible afterhis last attack, and take advantageof a period of calm. “I am sure thatin the north I will recover rapidly,and will be well for at least sometime, even though I may relapsein a year or two.” He encouragedTheo to write to Gachet, and, tospeed things up, even suggestedwhat should go in the letter: “Mybrother would very much like to

meet you…, he hopes you will approve of his spending several weeks in your village where he wants todo some studies. He is sure he will get on well with you, and believes that his return to the north will ben-efit his illness, whereas a longer stay in the Midi would risk harming his health.”

On May 20, his wish was granted and he finally met Dr Gachet. After an exhausting journey, he had ar-rived in Paris three days earlier, on May 17, weighed down with about 60 pounds of luggage and impatientto see Theo again, and to meet Johanna, his sister-in-law, and of course little Vincent, now 6 months old.He made a good impression on Theo’s young wife, but could not stand the noise and bustle of the city forlong. After remaining shut in for a whole day in the apartment at 8, Cité Pigalle, while he contemplatedhis paintings spread out on the floor, he agreed the following day to visit his brother’s stock of paintings,where the works of Guillaumin, Gauguin, and Russel were piled up with his own. The two brothers alsovisited an exhibition at the Salon du Champ-de-Mars, then the time came for Vincent to leave for the coun-tryside exhausted. He finally arrived at Auvers with three paintings under his arm and a letter of intro-duction for the doctor whom he found “rather eccentric.” This first impression is even rather reserved,since he wrote to Theo that the doctor “appeared to be afflicted at least as seriously as I am.” Full of “black,black antiquities,” his large white house built at the foot of a chalk cliff did not impress Vincent either. Wasit by prudence, the need for independence, or for reasons of economy that he opted to take a room at theAuberge Ravoux that was cheaper, only 3.50 francs per day compared with 6 francs, and was further awayfrom Gachet’s house than the inn at Saint-Aubin that the doctor had recommended?



Village Street and Steps inAuvers With Figures. This

painting dates from late May1890, just after van Gogh’sarrival in Auvers-sur-Oise.

© AKG Images.

Solace in a frenzy of paintingsThough he was seduced by the village and its “many thatched roofs, which is becoming rare,” and bythe region “...it is extremely beautiful, a typical picturesque open countryside,” his relations with Gachetwere not what he had hoped for. “I cannot rely on Dr Gachet at all,” he wrote several weeks after his ar-rival. Yet, to begin with, he thought he could make a real friend of this old man who loved art and seemedto have known Manet, who had discovered the works of Monet and Renoir at the Impressionist Exhibitionof 1874, and who also had a long friendship with Pissarro. Cézanne regularly visited Dr Gachet’s housein 1873 during the year he spent at Auvers with his partner and son, and later Guillaumin. Having beenplunged into grief by the death of his wife in 1875, Gachet was a sad person “with a haggard face.” How-ever, van Gogh adapted to the doctor. He even followed some of his recommendations: going to bed at9 o’clock, getting up at 5 o’clock, and warding off the return of attacks by hard work. Gachet let Vincent usea printing press he had in his attic to etch a self-portrait he was working on. Vincent also lunched withGachet every Sunday and sometimes on Mondays or Tuesdays to finish a canvas of Marguerite at the Piano,

the doctor’s daughter, or of his daughter in theirgarden. He also decided to do a portrait of Gachet,and was only too happy to have a model who actu-ally wanted to pose. The painting shows the doctorwith “a white cap on his head, light-colored, andvery prominent, the hands in clear flesh tints, andwearing a blue tail coat, against a cobalt back-ground.” More than a portrait, it is a manifesto, akind of social icon, whereby van Gogh expressedthe misfortune that haunted this disillusioned andsad human being. The features transcend those ofjust the doctor. “I want,” explained Vincent, “to paintportraits that one century later people will view asapparitions. I now have a portrait of Dr Gachet bear-ing the heartbroken expression of our time.” The



The Auberge Ravoux,Auvers-sur-Oise.

Van Gogh spent the lastfew months of his life at

this inn, where he rentedan attic room, eking out

a meager existence onthe allowance that his

brother Theo was able togive him. The inn is today

an essential stop on theitinerary of the many

visitors to the village insearch of the places that

van Gogh made so famousthrough his paintings.

© Paul Cooper/RexFeatures/Sunset.

March 30, 1853: Birth of Vincent at Groot-Zundert, in the south of HollandMay 1, 1857: Birth of his brother Theodorus.Of the five brothers and sisters, Theo was theclosest to Vincent1869: Vincent is employed by Goupil & Co, acompany specializing in art and run by his un-cle in The Hague1872: Begins his lifelong correspondence withhis brother Theo1873: He joins the London branch of the Goupilcompany. First unhappy love affair with Ursula,the daughter of his landlady1874: He is sent to Paris by Goupil and returnsto London in December. Begins to lose interestin his work1875: Returns to ParisFrom 1876 to 1880: Sacked by Goupil for neg-ligence. Returns to England to become assis-tant to a clergyman. Goes to Amsterdam toprepare for entry to a seminary. Travels to themining region of Borinage to help the disad-vantaged miners. Wears himself out. His reli-gious fanaticism reaches its peak. He breaksoff all contact with his family during 18791880: Decides to devote himself to art

1881 to 1882: Falls in love with his cousinKee, who rejects him. Sinks into depression.Returns to The Hague and the studio of AntonMauve. Takes up with a prostitute, Sien, whobreaks with him in 1882. Theo pays Vincentan allowance of 150 francs/month 1883: Seeks solitude. Returns to his parents atNuenen1885: Sudden death of his father. Leaves forAntwerp in Belgium1886-1887: Theo and Vincent settle in Paris.His first attacks occur. Alcoholism. Frequentsthe boutique of Julien (“Père”) Tanguy, a dealerin paintings and artFebruary 1888: Settles in ArlesOctober 1888: Arrival of Paul GauguinDecember 1888: Severs his ear after a violentdispute with Gauguin1889: Vincent enters the asylum at Saint-Rémyand leaves it one year laterMarch 1890: Ten of van Gogh’s paintings areexhibited at the Salon des IndépendantsMay 20, 1890: Arrives at Auvers-sur-OiseJuly 27, 1890: Shoots himself in the chestJuly 29, 1890: Dies at dawn. He is 37 yearsold

MILESTONES

painting showed him “leaning against a red table on which thereis a yellow book and purple foxglove” (in homage to his interest inhomeopathy). Gachet loved the portrait so much that he imme-diately asked Vincent to paint another one for him. This secondpainting is today in the Musée d’Orsay in Paris, thanks to the do-nation made by Gachet’s son in 1949, while the first painting wassold by Christie’s.

The tragic climaxAt about this time, van Gogh’s reputation was starting to be estab-lished through the efforts of Theo. Several artists including Guil-laumin wanted to exchange canvases with him. In a very laudatory

article, the art critic Albert-Émile Aurier thanked him for a work painted at Saint-Rémy-de-Provence. Butthe praise fell on deaf ears, all Vincent wanted was for his brother and family to come and live near himin the country. On June 8, they all arrived after being invited to lunch by Gachet. It was a beautiful day,Vincent was in relaxed mood and played with his little nephew, showing him the hens and ducks. Thehappiness he dreamed of was within his reach. But not for long. Several days later, little Vincent becameill and nearly died. Anxious and himself short of money, Theo made it clear to his brother that it was be-coming more and more difficult to provide the monthly allowance. He invited him to come to Paris onJuly 6 to try to find a solution, and to introduce him to Albert-Émile Aurier, and the painters Émile Bernardand Toulouse-Lautrec. But Vincent could not stand the pressure. He cut short his visit that was meantto last several days and took the evening train back to Auvers the very same day. He was devastated. Was itbecause he had seen his paintings decaying in the squalid storeroom belonging to the art-dealer Julien(Père) Tanguy? Was it his fear for the life of little Vincent and for his mother exhausted after nights ofwatching over her son? Was it the prospect of being alone for several days while Theo and his family werein Holland? Was it, above all, the prospect of having to stop painting due to lack of money? In spite of aletter from Johanna that comforted and reassured him concerning little Vincent, he did not recover fromthe storm that was threatening his livelihood. He did not know how to react. He felt helpless confronted



Marguerite Gachet inthe Garden. Executed inJune 1890, this was oneof two paintings vanGogh did of Dr Gachet’sdaughter. Oil on canvas,46� 55.5 cm. Muséed’Orsay, Paris. © PhotoRMN - Gérard Blot.

Portait of Doctor Gachet. Van Gogh painted two versions of this portrait,at the request of the subject, both in June 1890. Dr Gachet, a painterhimself, was a friend of several of the Impressionists and invited themto stay with him Auvers-sur-Oise (Cézanne did a painting of his house).Vincent initially had high hopes of Dr Gachet, but came to realize thathe could not rely on him. In this portrait, van Gogh captured some ofthe melancholic, contemplative mood to which the doctor was prone. Onthe table can be seen a sprig of purple foxglove, the source of digitalis.Dr Gachet was the only male sitter that van Gogh painted in Auvers. Oil on canvas, 68� 57 cm. Musée d’Orsay, Paris. © AKG Images.

with problems of money and reproached his broth-er for leaving him stranded. “My own existence isthreatened at its very roots, my future too is totter-ing.” He could not see any way out, except to con-tinue working relentlessly “even though the paint-brush is falling from my hands.” On July 14, Theoleft Paris for Holland instead of taking several days’rest at Auvers as Vincent had so desperately hoped.Was it the coup de grâce for this abandoned soul?We will never know. Even though on his return Theoimmediately sent his brother a 50 franc note, noth-ing could stop Vincent from taking the fatal step.He left his little garret at the Auberge Ravoux after

lunch, took the street that runs through the village, climbed the hill, and there, alone, facing the fields,shot himself in the chest. It was July 27, 1890 and Vincent was just 37 years old. The innkeeper, seeing hehad not returned for supper, was starting to worry, when Vincent finally arrived and went straight upstairsto his room. It is in this little garret under the roof with its cracked walls, splashes of grayish paint, andfaithfully restored mildew that all visits to Auvers-sur-Oise end. It was here that Ravoux found the painterin agony. He summoned the doctor who lived opposite, then Dr Gachet, who administered first aid, butthey decided not to remove the bullet as it was lodged just below the heart. Theo was informed in Paris andarrived panic-stricken late the following morning. Despite Vincent’s strong constitution, he was writhing“in terrible agony.” He died on July 29 at 1:30 AM leaving an unfinished letter to his brother in which heexplained: “In my kind of work, I risk my life.” ❒



PRACTICAL GUIDE TO AUVERS-SUR-OISE

◆ How to get there By road: Take the A15, direction Cergy Pontoise, leavethe A15 at exit 7, direction Méry and Auvers. By train (about one hour). Leave from Gare Saint-Lazare.Change at Pontoise for Auvers. Information: Tourist office: 00 33 1 30 36 10 06, E-mail: [email protected]

◆ What to see The Château d’Auvers. Visit with spectacle “In the foot-steps of the impressionists.” Tel: 00 33 1 34 48 48 45. www.château-auvers.frAuberge Ravoux, called the Maison van Gogh.Tel: 00 33 1 30 36 60 60Musée Daubigny. Tel: 00 33 1 30 36 80 20Atelier Daubigny (workshop). Open only from Easter toAll Saint’s Day (November 1). Tel: 00 33 1 34 48 03 03Musée de l’Absinthe. Tel: 00 33 1 30 36 83 26

Photograph showing the gravestones of Vincent van Gogh andhis brother Theo, in the cemetery at Auvers-sur-Oise. Theo diedjust 6 months after Vincent’s death in Utrecht, the Netherlands,on January 21, 1891. His wife, Johanna, requested that Theo be buried alongside his beloved brother, which was finallyachieved in 1914 when his ashes were transferred to Auvers. © Rasmussen/Diaporama.

Vase With Irises Against a Yellow Background.This still life was painted in Saint-Rémy, in May 1890, just before Vincent left to live in Auvers-sur-Oise. He discovered the iris motif in the garden at the asylum at Saint-Rémy and produced several pictures with these flowers as the main subject during this period. Oil on canvas,92� 73.5 cm. Rijksmuseum Vincent van Gogh, Amsterdam. Vincent van Gogh Foundation.



est à trente kilomètres de Paris, dans une jolie vallée de l’Oise que le peintre Vincent vanGogh passera ses dernières semaines avant de se donner la mort. Après avoir été interné prèsd’un an à l’asile de Saint-Paul-de-Mausole, en Provence, suite à la mutilation de son oreille le24 décembre 1888, il ne rêve que d’une chose : fuir la chaleur et la lumière trop brutale du sudpour retrouver la douceur du nord. C’est chose faite grâce à son frère Théo qui le recommandeà un médecin – le Dr Gachet – qui pourra s’occuper de Vincent dans le joli petit village d’Auvers-

sur-Oise où il possède une maison. Le peintre arrive le 20 mai 1890 plein d’enthousiasme pour cette ré-gion paisible et verdoyante. Il pense avoir trouvé la paix et se remet au travail avec fougue. Mais, le répitest de courte durée. Alternant des moments de grands bonheur avec d’autres d’angoisse profonde, il finitpar se tirer une balle dans la poitrine le 27 juillet 1890 dans l’après-midi et rendra l’âme dans les bras deson frère dans la nuit du 29. Durant les deux mois qu’il passe à Auvers il aura réalisé quelques 30 dessins,une gravure et 77 tableaux, sur les 879 qu’on lui connaît. Essentiels dans l’histoire de la peinture mo-derne, « Le portrait du Dr Gachet », « L’Église à Auvers », « Le jardin de Daubigny » et « Champs de blé auxcorbeaux », font partie de ceux-là.

✦

C’SUR LES TRACES DE VAN GOGH À AUVERS-SUR-OISE

No idea was more disagreeable to van Goghthan Cézanne’s belief that he would produce “thepaintings of a madman.” And, in August 1889,while in the asylum at Saint-Rémy, Vincent wrotethat he hated nothing more “than having to aska doctor for permission to do a painting,” and wasconvinced “that if sooner or later I am more orless cured, it will be because I have cured myselfby working.” Though the genius of the painter isbeyond question, it is evident that this fragile per-son painted despite (or perhaps because of) themore or less clear-cut symptoms of his mentalillness manifested by the mutilation of his ear atArles in 1888, his repeated attacks, his confine-ment in the asylum at Saint-Rémy-de-Provencein 1889, and his suicide in Auvers in 1890. Ac-cording to a study by Doctor Jean-Marc Boulon,*Concerning the Clinical Case of Vincent vanGogh, the diagnoses of contemporary psychia-trists suggest “manic-depressive psychosis with

acute mania with hallucinations and delusionscomplicated by epileptic fits that are exacerbatedduring periods of undernutrition, overwork, orintoxication with absinth, digitalis, potassiumbromide (prescribed at Arles), camphor, and car-bon monoxide, as well as by his excessive addic-tion to coffee and smoking.” The cause of the bi-polar disorder should be looked for, as in allpatients with alternating periods of euphoria anddepression, in the family history and a probablegenetic origin. These periods of intense depres-sion with mutism, nonproductivity, and melan-cholia, that can even lead to suicide, alternatewith moments of hyperactivity (also called hypo-mania), thanks to which van Gogh was able tocreate his large number of paintings and to penhis voluminous correspondence.

*Association Saint-Paul-de-Mausole, Route des Baux BP 39, 13532 Saint-Rémy-de-Provence Cedex. Tel: 00 33 4 90 92 77 00.

WORK AS A TREATMENT FOR “MENTAL WEAKNESS”

General instruct ions◆ Manuscripts should be provided on word-processor disks(3.5-in, for IBM, IBM-compatible, or Apple computers) withone (1) hard copy (text and figures) printed on one side ofstandard-sized white bond paper, double-spaced, with 2.5-cmmargins. Pages must be numbered. Standard typed page =25 lines of 75 characters (including spaces) double-spaced,2.5-cm margins=a total of about 250 words per page.◆ All texts should be submitted in English. In the case oftranslations, the text in the original language should be in-cluded.◆ Provide black and white photograph(s) of author(s).◆ On the title page, provide: a title (concise and informative);full names of authors (first name, middle name initial, andlast name); highest academic degrees (in country-of-originlanguage); affiliations (names of department[s] and institu-tion[s] at the time the work was done); a short running title(no more than 50 letters and spaces), keywords (5-10); corre-sponding author’s complete mailing address and telephoneNo., fax No., and e-mail address; acknowledgments (on titlepage, or at end of main text).◆ Include an Abstract of 200-250 words (1 standard typedpage) for all texts except Editorials and replies to the Contro-versial Question.◆ Figures and Tables. Submit figures as two sets of clearglossy prints. Figures should be of good quality or profession-ally prepared, numbered according to their order, with properorientation indicated (eg, “top,” or “left”), and fully explicitlegends provided, not repetitive of text. All abbreviations usedshould be explained in the legends. As figures and graphs mayneed to be reduced or enlarged, all absolute values and statis-tics should be provided. Illustrations will be reproduced infull color only when clearly necessary, eg, images from nu-clear medicine or histology. Provide each table on a separatesheet, with title above and description below. All figures andtables should be cited in the text, with distinct numbering forfigures and tables.◆ Note that Editorials and Abstracts will be published in En-glish and French. Translations into French will be providedby the Publisher’s Editorial Department. ◆ Include Headings using a consistent style for the variouslevels of headings, to highlight key points and facilitate com-prehension of the text. The Editorial Department reserves theright to add or delete headings when necessary.◆ Abbreviations should be used sparingly and expanded atfirst mention. A list of selected abbreviations and acronymsshould be provided (or will be prepared by the Editorial De-partment) where necessary. ◆ Use Système International (SI) units.◆ Use generic names of drugs.◆ All references should be cited in the text and numberedconsecutively using superscript arabic numerals. Presenta-tion of the references should be based on the Uniform Requi-rements for Manuscripts Submitted to Biomedical Journals.Ann Intern Med. 1997;126:36-47 (“Vancouver style”). Theauthor-date system of citation is NOT acceptable. In press ref-erences are to be avoided. In the bibliography, titles of jour-

nals should be abbreviated according to the Index Medicus.All authors should be listed up to six; if there are more, onlythe first three should be listed, followed by “et al.” Wherenecessary, references will be styled by the Editorial Depart-ment to Medicographia copyediting requirements. Authorsbear total responsibility for the accuracy and completenessof all references and for correct text citation. Example of stylefor references:

1. Ouriel K, Geary K, Green RM, Geary JE, DeWeese JA. Factorsdetermining survival after ruptured abdominal aneurysm. J VascSurg. 1990;11:493-496.2. Darling RC, Brewster DC, Ottinger LW. Autopsy study of unop-erated abdominal aortic aneurysms: the case for early resection.Circulation. 1977;56(suppl II):II161-II164.3. Schulman JL. Immunology of influenza. In: Kilbourne ED,Alfade RT, eds. The Influenza Viruses and Influenza.Orlando, Fla:Academic Press Inc; 1975:373-393.

Speci f ic formats◆ Editorial: 1500 words or 6 standard typed pages. No abstractor illustrations should be included. A French translation ofthe Editorial will be provided by the Editorial Departmentand submitted to the author. ◆ Replies to the Controversial Question: 500 words or 2 stan-dard typed pages. References, if cited, should in no case exceed10. No abstract or illustrations should be included.◆ Theme - Focus - Update - Interview - Product/Institution-al article - Touch of France: Abstract: 250 words (1 standardtyped page); main text: 2500 words (10 standard typed pages).References: their number should not exceed 50. Illustrations(figures and tables): their number should not exceed 5 un-less clearly necessary.

Editor ia l process ing◆ Editorial style: All contributions to Medicographia will bestyled by the Editorial Department according to the specifi-cations of the current edition of the American Medical Asso-ciation Manual of Style, Williams & Wilkins.◆ Page proofs and editorial queries will be sent to the corre-sponding author for approval and should be returned with-in 5 days by fax AND express mail. If this deadline is not met,changes made by the Editorial Department will be assumedto be accepted by the author. Authors are responsible for allstatements made in their work, including changes made bythe Editorial Department and authorized by the author. Arti-cles and abstracts will be edited to required length or returnedto the author if specific requirements are not complied with.

Copyright◆ Copyright of articles will be transferred to the Publisher ofMedicographia. The Copyright Transfer Agreement must besigned by the main author and all coauthors and returned tothe Publisher. ◆ For reproduction of copyrighted work, it is the author’s re-sponsibility to obtain authorizations from the author(s) (in-cluding self) and the publisher(s) and provide copies of theseauthorizations with the manuscript.

Medicographia

I n s t r u c t i o n s f o r a u t h o r s

A S e r v i e r p u b l i c a t i o n

depresie

Documents

servier house

servier publication

sumpter servier canada

vernon servier laboratories

morier servier argentina

toppjian laboratorios

dublin tel

wien tel