Marketing depression and making medicines work
Social Audit Ltd, P.O. Box 111, London NW1 8XE, UK
This version downloaded from http:\\www.socialaudit.org.uk
and posted on the Internet by kind permission of the Publishers:
IOS Press, Van Diemenstraat 94, 1013 CN Amsterdam, The
Netherlands (FAX: +33 20 620 34 19; email: firstname.lastname@example.org)
Medawar C: The Antidepressant Web - Marketing depression and making medicines
work, International Journal of Risk & Safety in Medicine
, 1997, 10
, 2, 75-126.
1. Abstract and introduction
This paper was conceived as chapters two and three of a still unwritten book,
and as a basis for discussion on a website and elsewhere. From early 1998,
revised and expanded versions of this text, plus feedback, will be posted on the
Internet; hence the numbered section headings, each denoting a separate
These chapters examine hard evidence relating to a wholly rhetorical andhypothetical question, “Do antidepressants work?” The reason for asking thisostensibly silly question is to provide the broadest possible framework forlooking at the meaning and values of medicine. Implicitly, the question alsoasks: what is better than nothing, and how much better are antidepressant drugsthan the placebos they are compared with in clinical trials?
Between the lines of the paper lie basic questions about the ethics, activities,performance and impact of the three main centres of power in medicine -government, professionals and the pharmaceutical industry. The underlyingissue is whether people who are miserably unfulfilled, sad, anguished ordepressed are in hands as safe as they might imagine or need.
There are no conclusions in this paper, but something of a hypothesis emergesfrom it: “depression” is fast developing as an iatrogenic disease and that,however much they are part of the solution, antidepressant drugs seem muchmore implicated in the problem of depression than is supposed. Moreover, thepublic’s opinion that antidepressants are drugs of dependence seems muchcloser to the reality than the exactly opposite orthodox view.
Similar scenarios go back a long way; they have repeatedly shown publicconcern to be justified and seen the medical establishment caught unawares.
Past experience in this field links risk of dependence to unaccountable power.
Is the same not true today?
Do antidepressants work?
Many people feel certain that
antidepressants have helped them and may even have saved their lives. They
might think it was naïve or dangerous even to ask whether antidepressantswork, and almost all health-care professionals would agree. By admitting thepossibility they might not, the question flies in the face of seemingly rock-solidmedical opinion and, whatever the answer, might promote loss of confidence inthe effectiveness of treatment. This could add to the problems of depressedpatients, put further demands on clinicians and health services and damagecommercial interests and reputations.
But the reason for asking if antidepressants work is not to try to prove that theydon’t; it is to review briefly the evidence that insists they do. Much of thisevidence is based on carefully conducted clinical studies and trials, but what isactually being tested? What is the “depression” these drugs treat ? What are“antidepressants” and what effects do they have? What does “work” reallymean, and how sound is the evidence they do? The answer to the centralquestion fundamentally depends on these and other matters of definition andinterpretation.
The question “Do antidepressants work?”
also provides a framework forthinking about a range of underlying issues, including the relationshipsbetween nature & nurture, sickness & health and benefit & risk. Questions alsoarise about business conduct and the roles of money and influence; aboutorganisational imperatives versus health goals; about the quality of science andthe basis of trust; and about the effectiveness of law and regulation.
Similar questions were addressed in an earlier study (Medawar, 1992) whichdiscussed the relationships between doctors, pharmaceutical companies,government and consumers - as reflected in the habitual prescribing ofdependence-producing drugs for anxiety, insomnia, depression and relatedproblems that go by a thousand other names. Over the past 200 years, doctorshave prescribed an almost uninterrupted succession of “addictive” drugs,always in the belief they would not cause dependence or that patients would bemainly responsible if they did. In the beginning were alcohol and opium, thenmorphine, heroin, and cocaine; alongside were chloral; numerous bromides,barbiturates and related compounds, and then a score of benzodiazepinetranquillisers. In their day, all these drugs were prescribed as sedatives formental distress, and except for alcohol, also as weaning treatments foraddiction to other drugs on the list.
The long-term efficacy of benzodiazepines proved largely an illusion, but onlyafter more than 20 years of extensive use. The reason most people stayed onthese drugs turned out to be they couldn’t readily stop taking them. They hadbecome dependent on them, in rather the same way that people get dependenton alcohol: usually very subtly and sometimes to disastrous effect. It is a long,sad story, perhaps not over yet.
If history were to be repeating itself, it would be both because and in spite of authoritative denials that any risk is involved. If antidepressants were in somesense drugs of dependence, but not recognised as such, it would increase theelement of risk and lead naturally to an over-estimation of their effectiveness aswell.
2.2. When antidepressants were first used
The first of the antidepressants in
use today came on the market about 40 years ago, so why does the questionarise now ? The short answer is that things changed recently with theintroduction of a new class of antidepressant, the Selective Serotonin ReuptakeInhibitors (SSRIs), exemplified by fluoxetine (Prozac). These drugs havestarted to overtake the more traditional antidepressants, the tricyclics and themonoamine-oxidase inhibitors (MAOIs). The tricyclics, in particular, havebeen the mainstay of drug treatment for years, but have lost some ground sincethe late 1980s when the SSRIs began to make their mark.
For all the differences between them, all these antidepressant have oneimportant thing in common and the evidence for it is overwhelming and hasnever been in dispute. When you carefully measure the effects of any of thesedrugs on whole populations, none proves more effective than any other intreating depression. Over the years, scores of different antidepressants havebeen tried, but patients generally respond (some very well, others less so) inabout 60% - 70% of cases. This compares with a typical 30% - 35% responserate reported with placebo.
The implications of this are fundamental to the analysis, first, because there isno difference in the quality
of response between antidepressants and placebo.
The difference is that active drugs can be expected to elicit that same responseapparently about twice as often as placebo; when placebos work in depression,they are as effective as the best active drugs. Secondly, when many differentdrugs elicit a similar and only partial response, it suggests lack of specificityand that the effects on depression are unlikely to result from these drugs’distinguishing chemical characteristics. One might conclude that the effects ofantidepressants are comparable to those of a strong placebo, a double-strengthplacebo to be precise.
There is, however, much discussion about the relative risks of differentantidepressants and about their benefits for particular patients, though suchdebates are ritual in any ‘new vs old’ drug evaluations and whenever clinicianshave a range of treatments to chose from. The same sort of debates were inprogress at about the time this story begins, when the following letter to TheLancet
was published in 1955. This was immediately before the advent of theoriginal antidepressants, the tricyclics and MAOIs. In those days, barbiturates+ amphetamines had taken over from opium as the bedrock of antidepressantdrug treatment and old habits took longer to die:
Sir,- Your annotation of May 21 does not mention opium. I think this is still avaluable drug in the treatment of minor depressive syndromes, many of themwith anxiety, which are so commonly seen in psychiatric practice . I havebeen prescribing (it) for many years. I have never seen a patient becomeaddicted to it (it is extremely unpleasant to take) and only once has a patientattempted to use it for ostensible suicidal purposes . Considering, too, theease with which patients may hoard barbiturate tablets and the frequency withwhich they are used in suicidal attempts, I think there is still much to be saidfor the old-fashioned opium mixture . It can be used safety for many weeksat a time and it goes well with the amphetamine group of drugs. I havegenerally found that it is only when insomnia is severe that it is necessary toadd a nocturnal barbiturate to this regime”.(Scottowe, 1955)
In the same year, iproniazid was introduced as a treatment for tuberculosis, andthis is also where part of the story begins. Iproniazid was found by chance to
have a marked effect on depressive symptoms in TB patients, so much so that itwas soon superseded by another less stimulating drug, (and finally withdrawnby the manufacturers in 1961, when found to cause liver damage):
“It was eventually displaced by isoniazid since iproniazid actually madesome of the patients feel ‘too well’ with the result that they failed to observeordinary precautions, overexerted themselves, or discontinued treatmentprematurely. Retrospectively it is evident that the drug not only relieveddepression but occasionally must have induced euphoria. In view of theexcessive good spirits of the patients it is strange that at the time theemotional reaction was regarded as a detrimental side effect and no one triedusing iproniazid for treatment of depression”. (Kline, 1964)
Iproniazid had been found to somewhat inhibit the effects of monoamineoxidase and the MAOIs were developed as compounds with a more potentinhibiting effect than iproniazid itself. This led to the development of ideasabout the biochemical basis of depression and about the actions ofantidepressant drugs. Because the enzyme, MAO, inactivated theneurotransmitter noradrenaline, it was first postulated that depression was dueto a deficiency of brain noradrenaline, and that mania resulted from excess.
Later it was proposed that depression resulted from a deficit of anotherneurotransmitter, serotonin (5-hydroxytryptamine or 5-HT). The scienceindicates otherwise, (Healy, 1987) but this view is still widely promoted andgenerally held.
By the end of the 1950s, four MAOIs were on the market. They were originallydescribed as “psychic energisers” but count as the earliest drugs still designatedand licensed for used as ‘antidepressants’. The main tricyclics (once known as“psychostimulants”), such as imipramine, came on the market a year or twolater; they were developed from work on antihistamines (classically recognisedas anti-allergy drugs).
Treatment of depression: the first 30 years
In the 1960s, the lack of
any defined, mass market for depression inevitably meant that pharmaceuticalcompanies were reluctant to try to develop drugs for it. Nevertheless, they hadbegun to see opportunities. Early on, one of the pioneers in this field publisheda small, helpful and hopeful volume, Recognising the Depressed Patient
(Ayd,1961; Raach, 1961) and “Merck Sharpe & Dohme bought 50,000 copies of itand distributed it not just to psychiatrists, but to family doctors and internistsand so forth”. (Ayd, 1996)
In those early days, no one knew how common depression was: “There wereno epidemiological studies worth a tinker’s damn. In fact, epidemiology as weknow it today in psychiatry didn’t exist then.” (Ibid) An important turningpoint came with the publication of a widely circulated estimate from the WHOthat “at least one hundred million people in the world . suffer from depressivedisorders amenable to treatment”. (Sartorius, 1974, 1978)
With these changes came new and different kinds of antidepressant drugs withconfident claims of effectiveness, plus more defined ideas about whatdepression was and how antidepressants worked. Two trends accelerated the
commitment to use drugs. One was the ascendancy of biological theories ofdepression over psychoanalytically-oriented views:
“If there is one central intellectual reality at the end of the twentiethcentury, it is that the biological approach to psychiatry - treating mentalillness as a genetically influenced disorder of brain chemistry - has beena smashing success. Freud’s ideas, which dominated the history ofpsychiatry for the past half century, are now vanishing like the lastsnows of winter” (Shorter, 1997).
The other factor was the decline in use of ECT, but not so much because of therisks (Drill, 1958, Pippard, 1992), nor because it was thought ineffective.
(APA, 1997) Medical texts tend to attribute the decline of ECT to publicresistance fuelled by misconceived portrayals, notably in the book (Kesey,1962) and film (Forman, 1975) One Flew Over the Cuckoo’s Nest.
However,the evidence that ECT treatment is sometimes poorly performed (Wise, 1997)and high costs may have also played some part. In the US, a single ECTsession is costed at £200-£500 (mainly the cost of anaesthesia) and a typicalcourse of treatment might be 6 - 12 sessions over several weeks. (APA, 1997)
Though drugs were usually cheaper and more convenient to use, their use hasalways been limited by poor compliance and unwanted effects. Patients usuallyexperienced uncomfortable rather than serious side effects, though there werealso significant risks. For example, recognition of a potentially dangerousinteraction between MAOIs and certain foods (eg cheese, yeast extracts)helped to promote the tricyclics, and later the tricyclics lost some ground to the“quadricyclics”, newer drugs promoted as safer in overdose. Very severedepression carries some risk of suicide, and it has often and long been arguedthat the greatest risk lies in not treating depression at all.
The scientific medical literature of the 1960s suggests that the originalantidepressants were given a rather cautious welcome, though this should beseen in the context of those times. In those days, the market was quite smalland the buzz in the journals (advertisements too) was mainly about anxiety,stress and insomnia. This was a huge and growing market, but strictly reservedfor the “tranquillisers”, and notably the benzodiazepines. Drugs like Librium(chlordiazepoxide) and Valium (diazepam) dominated, from 1960 and for thenext 30 years.
Nor did “depression” mean what it means today. Then, (endogenous)depression was exemplified by the mentally and physically immobilisedpatient, sitting with his head in his hands. This was well-recognised as aserious illness but it also carried quite a stigma; it was “not fashionable to bedepressed” (Kline, 1964). At the same time, most cases of “depression” werethought self-limiting: until the 1980s, the great medical textbooks and mostexperts emphasised that up to 80% of all cases of depression would curethemselves. If the implication was that depression often needed no drugtreatment, such views come close to heresy today:
“ . depression is, on the whole, one of the psychiatric conditions with thebest prognosis for eventual recovery with or without treatment. Mostdepressions are self-limited and the spontaneous or placebo-induced
improvement rate is often high. For example, in a series of nine controlledstudies on hospitalised patients, 57% of the patients given placebo therapyshowed improvement in two to six weeks.” (Cole, 1964)
“In the treatment of depression one always has as an ally the fact that mostdepressions terminate in spontaneous remission. This means that in manycases regardless of what one does the patient eventually will begin to getbetter.” (Kline, 1964)
“ . most depressed patients get better anyway and the patients who improveafter one has prescribed tablets have done so post hoc
but not necessarilyproper hoc.
” (Leyburn, 1967)
The physician “must also weigh the fact that perhaps 80% or more ofdepressed patients will eventually recover without treatment” (Byck, 1975). and “affective disorders have a very high rate of spontaneous remission,provided sufficient time passes” (Baldessarini, 1980).
Then as now it was recognised that a significant minority (around 25%) did notrespond to drug treatment. The standard response to “resistant depression”today would be to increase the dose and to prescribe other drugs, as well orinstead. In those days, resistant cases would usually be treated with electro-convulsive therapy (ECT); many experts believed this to be the most effectiveof all and some still do.
Less was known then about how antidepressant drugs worked and about thebiochemical rationales for using them and, in those days, psychodynamicunderstandings of depression held much greater sway (Lehmann, 1996;Shorter, 1997). Moreover, evidence had accumulated since the early 1960s of agulf between the advertised benefits of antidepressants and their actual effects,when assessed in controlled clinical trials. As a whole, the hard evidencelooked thin: it did suggest that the MAOIs and tricyclics could be distinguishedfrom placebo, but the difference was not great. This was the rather low opinionof one of the pioneers, a man still prominent in the field:
“The newer antidepressant drugs have now been used experimentally andclinically for approximately seven years. Their place in the physician’sarmamentarium is still far from clear, although many clinicians feel that thedrugs are useful and effective. However, controlled clinical trials of theseagents have not always led to unequivocally positive findings. Even whenthe findings have been favourable to the drugs under study, the differencesbetween the efficacy of the drug and a placebo have not been as great as onemight wish, or as one might have anticipated after reading published reportsof uncontrolled trials.” (Cole, 1964)
Soon after, the US National Institutes of Health reported the results of asystematic analysis of 490 studies published in 71 leading medical journalsbetween 1955 and 1966. The conclusion was that: “the methodology of drug
research is of more significance to the outcome of a clinical trial than is thedrug being studied . In well-designed studies, the differences between theeffectiveness of antidepressant drugs and placebo are not impressive”. (Smithet al., 1969) The effect of these original antidepressants on depression hasnevertheless become one of the main yardsticks for efficacy by which eachsuccessive generation of antidepressants has been proved.
Successive editions of a leading UK textbook on clinical pharmacology suggestthat the quality of such trials “has got only a little better since”; (Laurence,1966, 1974, 1980; Laurence & Bennett, 1987); indeed, low standards seemcommonplace today (Gore et al., 1992; Wise and Drury, 1996). Meanwhile,the number of tricyclic and related antidepressants proliferated, albeit to littleeffect. The 1970s and 1980s saw numerous attempts to manipulate drugmolecules, but antidepressant drug therapy “developed a bewilderingcomplexity” as a result. “None of these changes (has) produced anantidepressant that is more effective; approximately 80% of a heterogeneouspopulation will respond to adequate treatment with any tricyclic compound”,(Blackwell & Simon, 1988) and the same has proved true of the rest.
In time, the controversy quietened and antidepressant drug prescribing becameroutine, in spite of the uncertainties and probably because of them too. Onefactor which would have contributed to uncertainly was the complexity andcost of rigorous drug testing. Other factors would include the lack of evidentlybetter alternatives; the lower cost and convenience of drug treatment; the“rewarding” and “gratifying” results sometimes obtained; growing belief in thebiological basis of depression; the tendency to discount placebo and nocebofactors at work (Merry, 1972); confusion over the limitless opportunities fordiagnoses, with possibilities for always trying something new; and perhapsabove all, the intensity of drug promotion.
Given the essential similarities between the dozens of different drugs, attentionwas mainly focused on safety and on the prevalence of depressive conditions,and its many and subtle manifestations. These included phenomena labelled as“masked”, “smiling” and “hidden” depression; thus the many diagnosticuncertainties were simultaneously increased and also largely dispelled.
“I am sure many colleagues have shared with me the followingembarrassing experience. I prescribe a tricyclic drug for an outpatientwith a typical or “classical” endogenous depression. The patient returnsto see me three or four weeks later. She is very much better. When Iremind her of the importance of continuing drug therapy despite theimprovement, she smiles and says ‘Oh doctor, the tablets did not agreewith me, so I stopped taking them after the first two or three days’.”(Merry, 1972)
Treating depression: the 1990s
The first SSRI (zimeldine) was
introduced in 1980, but withdrawn soon afterwards when found to cause a verysmall but unacceptably high number of serious neurological and other
reactions. Next came fluvoxamine (Faverin/Luvox, Solvay), but it was nobreakthrough. At launch it was oversold (DTB
, 1988) and promoted for a widerange of somatic complaints which might (or might not) be linked todepression, including “aches and pains, agitation, anxiety, sleep loss, lowmood, dizziness, worry, sweating” etc. (Duphar, 1987) Also its adverseeffects had been underestimated and it ran into bad publicity in the lay media(Ferriman, 1988) after a warning about suspected adverse effects from theCommittee on Safety of Medicines (1988).
Fluoxetine was launched in 1988/89. Prozac
became a buy-word and the maindriving force behind the huge expansion of the depression market. Here is adrug immortalised by Woody Allen (as Valium was before it) and the subjectof overwhelming volumes of airtime and webspace, and countless miles ofprint. Ten popular books with “Prozac” in the title have been referred to in thispaper, but there are at least twice that number, in English alone. (Baker &Taylor, 1996)
Along with fluoxetine there are now several other SSRIs and related drugs andthe value of the world market (1997) is about £3bn a year. The table showshow Prozac and the others have, in the last five years, secured a 50% increasein the England market. (Department of Health, 1991-1995) The SSRIs haveyet not significantly eroded prescribing levels for other antidepressants (as theynow have in the US); the whole market dramatically expanded once theyarrived on the scene.
No of prescrip-
NHS spend on all
Cost of SSRIs (as
tions for all
for all SSRIs (as
% of total) in
% of total), in
in England (£m)
in England (£m)
Underlying the success of the SSRIs was the still widely-promoted theory thatdepression was in effect a serotonin deficiency disease. The thrust of themessage was that depression is as biological in origin as is lack of insulin forsomeone with diabetes - the implication being that drugs like Prozac might beconsidered almost as essential supplements for people with depression. Thoughstill strongly supported and promoted (See 3.6), the idea that depression hasmore complex and varied biological origins is said to be gaining ground(Delgado et al., 1992).
Genetic and biological factors can have an important role in depression, but thenotion that depression is basically caused by lack of brain serotonin (or some
simple imbalance with other neurotransmitters) is clearly problematic. It doesnot explain, for example, why drugs which have an immediate effect in raisingbrain serotonin levels nevertheless usually take at least a couple of weeks toexert an antidepressant effect. Nor would it explain why SSRIs have no moreeffect on depression than other antidepressants which hardly act on serotonin.
And how would one explain the lack of effect of antidepressants on the mostclear-cut cases of depression, the roughly one-quarter of all cases most resistantto treatment with drugs? Such theories are widely supported, but theirscientific basis is indeed questionable:
“ . far from these hypotheses being an unambiguous advance in thescientific understanding of mental illness, I have argued elsewhere(1987) that the monoamine hypotheses in particular were quitesimplistic; that they accounted for less of the clinical data and were asunscientific as the psychodynamic hypotheses before them, in that theyhave been in practice, incapable of disproof”. (Healy, 1990)
Several other recent developments helped the SSRIs to become established.
One was the belated recognition of the benzodiazepine (BDZ) dependenceproblem: in the late 1980s, new curbs on tranquilliser prescribing opened upthe market for medicines for anxiety, insomnia and the like. (CSM, 1988)Secondly, there were concerted professional initiatives to encourage bothpatients and doctors to recognise and treat depression more aggressively.
Thirdly, experts rewrote and transformed the treatment guidelines fordepression. Other experts formally redefined the condition known as“depression”, emphasising the need for prolonged treatment and linking itmore closely to the kinds symptoms for which BDZs had hitherto been almostexclusively used.
The definition of depression
The new, intensive focus on depression
as a widespread disease has been underpinned by the work of nosologists,specialists in classifying and defining illness. The foremost definitions ofdepression are those developed by panels of experts convened by the AmericanPsychiatric Association. The APA’s Diagnostic and Statistical Manual
wasfirst compiled in 1952 to assist the national census of mental disability, but hassince been transformed. The fourth edition, known as DSM-IV, was publishedin 1994 and is now internationally recognised as the prime definition of how torecognise depression and, implicitly, when and how to treat it. DSM-IVdefinitions are also closely linked to those in the WHO’s InternationalClassification of Diseases
(ICD-10) and arguably now drive them.
DSM-IV is in some sense a great achievement, each new edition representingdecades of development and years of expert work. The task is formidable andvery costly: establishing the ground rules demands feats of understanding,organisation and painstaking application, and great political skill would havebeen needed to secure anything like consensus and general acceptance. Andclearly the need for good definition is paramount. It is fundamental to commonunderstandings, good communication and effective diagnosis; lack of definitionincreases the risk of wishful, misguided thinking and unhelpful treatment andpractice.
However, what matters it is how useful the definitions are and to what effect on
health - and this depends on many different pluses and minuses, with muchjudgement needed about which is which. If DSM-IV were a fishing net, thequestion would be: what mesh size should be used to catch depressed fish butnot others? The mesh has been getting smaller over the years, but it this a goodor bad thing?
Diagnostic & Statistical Manual
Five editions of the DSM have produced a threefold increase in “diseaseentities”. What Hippocrates knew as melancholy is now identifiable in 300manifestations (including manic depression), detectable through the expressionof many commonplace symptoms and characterised by often familiarbehaviours. But how much does this explain ill-health and help doctors torelieve suffering, and has the time come for “National Depression ScreeningDays” (1997) to be extended beyond the US?
Perhaps the DSM classification offers convenient rather than convincingsolutions and has rationalised rather than reduced diagnostic chaos. Perhapslonger definitions make less sense, by directing towards a circumference ofblurry understandings, the more they elaborate the central point. In expandingdefinitions of “depression”, perhaps these guidelines have helped to promotesomething like hypochondriasis (DSM-IV, F45.2) as well:
“If people are educated to believe they are fundamentally fragile,always on the verge of mortal disease, perpetually in need of health-care professionals at every side, always dependent on an imagineddiscipline of ‘preventive’ medicine, there can be no limit to thenumbers of doctors’ offices, clinics, and hospitals required to meet thedemand . We are, in real life, a reasonably healthy people. Far frombeing ineptly put together, we are amazingly tough, durable organisms,full of health, ready for most contingencies. The new danger to ourwell-being, if we continue to listen to all the talk, is in becoming anation of healthy hypochondriacs, living gingerly, worrying ourselveshalf to death” (Thomas, 1979)
In authenticating more and more diagnoses, the DSM process has helped tolegitimise a dramatic increase in drug use (the dominant treatment mode) forconditions that become wider and wider in scope. That is the risk withelaborate definitions, especially when “diagnoses are made by countingsymptoms, prefer-ably those that are easily observable, and those that areeasily agreed upon by direct questioning of the patient”. (Van Praag, 1996)What kind of symptoms may signal a “major depressive episode”, forexample? The explanatory memorandum in DSM-IV brings to mind smallmesh and a wide net:
“The mood in a Major Depressive Episode is often described by theperson as depressed, sad, hopeless, discouraged, or ‘down in thedumps’ (Criterion A1). In some cases, sadness may be denied at first,but may subsequently be elicited by interview (eg by pointing out thatthe individual looks as if he or she is about to cry). In some individualswho complain of feeling ‘blah,’ having no feelings or feeling anxious,the presence of a depressed mood can be inferred from the person’sfacial expression and demeanour. Some individuals emphasise somaticcomplaints (eg bodily aches and pains) rather than reporting feelings ofsadness. Many individuals report or exhibit increased irritability .”
A2 “Loss of interest or pleasure is nearly always present, at least tosome degree. Individuals may report feeling less interested in hobbies. (eg a former avid golfer no longer plays, a child who used to enjoysoccer finds excuses not to practice).”
A3 “Appetite is usually reduced . (but) other individuals . may haveincreased appetite . there may be a significant loss or gain in weight.”
A4 The most common sleep disturbance associated with a MajorDepressive Episode is insomnia” (including “middle insomnia .
terminal insomnia” .and “initial insomnia”) and “less frequently,individuals present with oversleeping (hypersomnia) . Sometimes thereason that the individual seeks treatment is for the disturbed sleep”
A5 “Psychomotor changes include agitation (eg the inability to sit still.) . or retardation (eg slowed speech, thinking or bodymovements.)”
A6 “Decreased energy, tiredness and fatigue are common .”
A7 “The sense of worthlessness or guilt associated with a majorDepressive Episode may include unrealistic negative evaluations ofone’s worth or guilty preoccupations or ruminations over minor pastfailings.”
A8 “Many individuals report impaired ability to think, concentrate ormake decisions . They may appear easily distracted or complain ofmemory difficulties.
Importantly, the last of the identifiers (Criterion A9) suggests that “Frequently,there may be thoughts of death, suicidal ideation, or suicide attempts”. But it isnot a necessary condition for the diagnosis, and in other depressive states (eg“Depressive Disorder Not Otherwise Specified”) may not feature at all.
A formal diagnosis for Major Depressive Episode can be met by twoconditions. One relates to the severity and duration of the depressed state,though these might be inferred simply by reason of the patient going to thedoctor. In addition to depressed mood, the patient should also have at least fourticks in the remaining eight boxes (Criteria A2 to A9).
To this extent, the currency of “depression” has become debased over theyears, and this colours the question: “Do antidepressants work ?”
Nowadays,perhaps the most unifying definition of “depression” is that it is a condition tobe treated with antidepressant drugs. There may not be a lot to distinguishbetween the drugs, but there is no end of possibilities for prescribing them. Thetrend in definition has been to identify more and more people as “depressed”,
“The boundaries of what constitutes depression have been expandedrelentlessly outward. Depression as a major psychiatric illnessinvolving bleakness of mood, self-loathing, an inability to experiencepleasure and suicidal thoughts has been familiar for many centuries.
The illness has a heavy biological component. Depression in thevocabulary of post 1960s American psychiatry has become tantamountto dysphoria, meaning unhappiness, in combination with loss ofappetite and difficulty sleeping”. (Shorter, 1997)
The way in which depression is now formally defined has expanded the marketalso by effectively undermining a major instrument of regulatory control.
When drugs are licensed, by law they can be promoted only for quite strictlydefined indications - but as antidepressants have typically been indicated “forthe treatment of symptoms of depressive disease”, DSM-IV provides scope forgreat over-simplification. This is exemplified in the following, the completetext of a full-page advertisement for the leading SSRI: (Lilly, 1993)
“First line . for all nine symptoms of depression
, fluoxetine hydrochloride.”
Diagnosis - Anxiety or depression ?
As they have identified more
and more people who may be “depressed”, nosologists have helped to expandmarkets and sometimes create perfect niches for individual drugs. Alprazolam(Xanax, Upjohn) for “panic attacks” is a case in point. But what is the“depression market” and what does it reveal of “depression” itself ? Inparticular, are the diagnoses of “anxiety” and “depression” distinguishable ? To the extent they are not, it would have helped the SSRIs to get into themarket previously dominated by the BDZs.
It is universally accepted that “anxiety” is a major part of “depression” butthere is long-standing controversy about which is what. (Goldberg, 1995) Some believe they are variants of a single disorder; others hold they are distinctbut overlapping entities. However, it would be generally agreed that mostpatients with depression can be diagnosed as anxious too, and that “majordepression is a frequent secondary disorder associated with several of theanxiety disorders.” (Keller & Hanks, 1995). Thus, DSM-IV includes a formaldefinition of “Mixed Anxiety-Depressive Disorder” and, in practice,antidepressants and anxiolytics (usually BDZs) are often used interchangeablyor together at the same time. (Hale, 1997) Several well-controlled trialsindicate that benzodiazepines tranquillisers are often as effective asantidepressants in treating “depression”, just as antidepressants often work on
“anxiety” too (Rickels et al., 1993).
Around 1960, the markets for anxiety and depression were much more as one:mainly barbiturates for anxiety and barbiturates+amphetamines for depression.
The main concern would then have been to relieve the patient’s symptomseither by sedating or stimulating her - and being careful not to bring about asudden swing of mood in either direction. (Both depression and anxiety arediagnosed twice as often in women than men).
With the simultaneous arrival of the first antidepressants and the BDZs, astricter process of demarcation began. But this took time and uncertainty aboutthe relationship between anxiety and depression is evident in some of theearliest assessments of the MAOIs:
“The question must now arise whether MAOIs are really the antidepressantdrugs they are claimed to be or whether they act really more against anxiety,and perhaps as stabilisers of the autonomic nervous system. Such a questioninevitably brings up the controversial problem of what is meant by the terms‘depression’ and ‘anxiety’. Lewis (1934) believes that it is impossible toseparate clinically groups of depressions one from the other; he regardsanxiety often as a symptom of depression, and anxiety states and depressionas forming one long continuum of illness”. (Sargant & Dally, 1962)
“These findings suggest that the beneficial effect of phenelzine (an MAOI)in depressive illness is due more to a sedative action in relieving anxietythan to a specific antidepressant action.” (Hare et al., 1962)
The overlap between anxiety and depression is also prominent in an importantdiagnostic tool, the Hamilton Rating Scale for Depression. (Hamilton, 1967).
This scale, still the most widely used to screen patients entering clinical trials,includes many questions about anxiety. As a result “an effective anxiolyticagent may substantially reduce total scores and such reductions are then oftenuncritically interpreted as evidence of antidepressant efficacy” (Healy, 1991).
The same writer has also noted that “Hamilton himself did not see his scale asan instrument for measuring the severity of changes in a depressive illness.
Rather he saw it initially as a checklist of questions clinicians should be askingand observations they should be making. A great number of these questionsand observations concern anxiety.” (Healy, 1990)
Neither is the distinction between depression and anxiety too clear from longdominant theories about the biological basis of depression, and the role ofserotonin (5-HT). Though specifically identified as one of the keys todepression, serotonin is closely linked to “anxiety” too.
“ . a great number of new compounds, with relatively specific actions onthe 5-HT system . have begun to appear on the market. Are they anxiolyticor antidepressant or both ? The overview, above, of the behavioural effectsmediated through 5-HT receptors suggests that 5-HT has more to do withanxiety than depression. This however, is an issue that is likely to beconfounded greatly by the efforts of drug companies to market their
The evidence suggests some repositioning of “anxiety” through the promotionof “depression”, over most of the last ten years. Anxiety is now on the back-burner and depression has become the dominant “disease”. Then we wereanxious, now we are depressed. Valium out, Prozac in.
Standards for testing the efficacy of SSRIs
The definitive answer to
the question “Do antidepressants work?”
is the legal one. No one isallowed to market a medicinal drug without a license which, by law, can begiven only when the efficacy of a drug has been proved. Most countriesrequire this and have agencies to enforce the law; the “regulators” in the UKinclude the Medicines Control Agency (MCA), Committee on Safety ofMedicines (CSM) and the European Medicines Control Agency (EMEA).
The UK regulators work in strict secrecy which complicates evaluation oftheir work; (Medawar, 1996) however, it is clear that the scientific basis ofsome important assessments is slight. For example, given the overlapbetween “anxiety” and “depression”, it seems extraordinary that theregulators should have accepted evidence in which antidepressant efficacywas measured in trials where people took anxiolytics at the same time.
Why has this methodology slipped into the protocols of so many trials ? The welfare of patients would have been a factor and convenience andopportunism may also have played some part. But perhaps the driving forcewas unconscious bias. Its source may have been the conviction that itwouldn’t matter to give both drugs together, because antidepressant andanxiolytic drugs were quite different things. After thirty years of tightmarket segregation, such assumptions might have been made almost as amatter of course.
Much of the bedrock evidence put before the regulators appears suspect forthis reason. Publicly available data from the US Food & DrugAdministration (FDA) shows that fluoxetine (Prozac) was licensed in spiteof, rather than because of, clear-cut evidence of efficacy from controlledtrials:
The FDA relied on four pivotal studies designated as “adequate andwell-controlled trials which provided evidence of efficacy” offluoxetine. (FDA, 1988)
Of these four placebo-controlled trials, three permitted the use of “ofconcurrent psychotropic medication”, and one-quarter of the enrolledpatients (135/540) took benzodiazepines (or chloral) as well asfluoxetine.
If these 135 patients are excluded from the analysis in these three trials,fluoxetine does not show statistically significant efficacy over placebo(Breggin & Breggin, 1994).
The one study that did prohibit the use of other such medicines was alsothe only one of the four to find no statistically significant differencebetween fluoxetine and placebo.
One other standard trial procedure seems capable of wreaking havoc withefficacy evaluations. This is a feature of many protocols which is not onlyacceptable to the regulators, but also positively advocated by leadingauthorities. The principle author of the following recommendation is one of themost widely published experts on the SSRIs; he was also a member of theCommittee on Safety of Medicines at the time the main evidence on the SSRIswas being assessed (1987 -1992):
“Studies can be flawed by including too many inappropriate patients.
The inclusion of treatment-resistant patients, who are oftenconcentrated in inpatient studies, can reduce the likelihood of finding apositive result. Similarly placebo responders can confuse the picture.
The inclusion of a placebo treatment period before the entry severitycriteria is applied often helps to reduce this source of error”.
(Montgomery & Lambert, 1989)
In other words, to demonstrate the efficacy of an antidepressant (a positiveresult), one should first eliminate from the group of patients studied anyoneseverely enough depressed to be hospitalised, and anyone whose depressionreadily responds to placebo. Thus, inpatients were excluded from the pivotaltrials on fluoxetine, and all four protocols included a placebo washout period topre-screen unwanted subjects. The procedure involved measuring the Hamilton(HAM-D) rating for depression to identify possible patients for trials, thenputting everyone on placebo for a week and excluding from the trial anyonewhose HAM-D rating had dropped to below 80% of the original value, orbelow a specified HAM-D score.
But how can one use an active drug vs placebo model as the gold standard forefficacy, when permitting the pre-screening of patients to eliminate people whoreadily respond to placebo? This would lead (and has led) to grossunderestimation of the value of placebo treatment; even in the most severe andobvious cases of depression, it seems like screening a jury to extremes:
“A few years ago, we tried an experimental design in one of our studieswhich we hoped would eliminate ‘placebo reactors’ and increase oursensitivity in distinguishing between drugs. All depressed patients whoentered the hospital and were candidates for the study were first placedon a week of placebo treatment. At the end of the week, the psychiatristwas then asked to make a decision as to whether the patient should beadmitted to the study . We lost 50% of our potential sample, as thatnumber of patients had shown a degree of spontaneous improvementwhich would have confounded the effects of future treatment. Thetendency of depressed patients to improve spontaneously certainlycreates difficulties in the clinical evaluation of drugs.” (Hollister, 1972)
The increasing tendency (Senn, 1997) to exclude placebo responders mightexplain the apparent decline in the magnitude of the placebo response reportedin clinical trials, over the years. In their review of all properly controlledstudies of antidepressants, Smith et al., (1969) reported the median
improvement rate on placebo to be 46% (and for active drugs, 61%). Reportedresponse rates on placebo come closer to 33% today.
For licensing purposes, controlled trials to demonstrate the efficacy of SSRIstypically last about six weeks, though the minimum recommended period oftreatment in clinical practice is now of the order of six months. Since the1990s, longer-term studies have been conducted; they appear flawed to aboutthe same extent, but in different ways. See 3.7.
Blindness and placebo response in antidepressant drug trials
other problem has complicated the scientific evaluation of antidepressantefficacy and has proved hard to avoid. Fundamental to reliable evaluation is therequirement of blindness, and normally the double-blind procedure whenneither investigators nor patients know whether they are administering/takingeither the test drug or placebo. The FDA requires claims of efficacy to be basedon properly blinded studies, the aim being to limit the influence of wishfulthinking and bias.
The problem is that some side effects of antidepressants - particularly the olderones - are marked, distinctive and well known, often a give-away to patientsand investigators alike. Moreover, “many of the side effects of antidepressantsmimic depressive symptoms and it is often difficult to distinguish what is atreatment-emergent effect from the pre-existing depressive symptomatology”.
(Montgomery & Lambert, 1989) The problem of spontaneous unblinding hasbeen recognised for as long as antidepressant drugs have been around, and hasled some to suggest that the only reliable way of testing might be to comparethem with a non-inert placebo (eg atropine), or perhaps with an establishedantidepressant at a sub-therapeutic dose.
“Most antidepressant drugs cause side effects which are recognisable byexperienced investigators in a significant proportion of patients. Patientswho come into the consulting room for assessment, perhaps for the sixthtime and rather bored with the whole thing, but with their mouths so dry thatone can hear their tongues scraping and clicking about in their mouths, arelikely to be taking, say, amitryptyline, rather than the placebo.” (Leyburn,1967).
“The side effects of imipramine ensure that no trial can be conducted undercompletely ‘blind’ circumstances. In this study, 15 patients complained oftypical side effects and it was suspected they were taking imipramine; thesupposition was correct in 13 patients, who represented half of all those onthe drug.” (Porter, 1970)
There are two ways of looking at placebos. Typically, they are regarded asdummy drugs: they are pharmacologically inert but may trick patient anddoctor into thinking they are the real thing. Traditionally, placebos get a badpress: they are generally identified as non-treatments, inferior to active drugs;their use may involve ethical problems; they raise question marks about theaccuracy of diagnoses and the authenticity of illness; and no-one likes to feelfooled. As if by definition, placebos are regarded as less desirable things.
An alternative view would be that placebos often do work as effectively asanything else, and sometimes very powerfully, as one might expect with anypowerful form of suggestion. Even after pre-screening, about a third of allpatients with “major depression” consistently respond as well on placebo as onactive drug. This not only seems remarkable, in theory it also makes theplacebo much the superior treatment for those patients on grounds of bothsafety and cost.
Which view one holds about placebo effects would be much influenced byone’s understanding of the nature and origins of depression. Staunch advocatesof a biochemical basis of depression might be inclined to argue that it can’thave been a Major Depression if a sugar pill made the condition disappear.
Others might conclude that the pill served essentially as a token or symbol andthat simple interventions were sometimes enough to make depression go away.
Beyond this lie realms of magic and the unknown, and they are not necessarilyincompatible with good scientific sense. As Lewis Thomas used to say: “theonly solid piece of scientific truth about which I feel totally confident is that weare profoundly ignorant about nature”. Science was full of surprises for himand they delighted him:
“I was once told by a distinguished old professor of medicine, one of SirWilliam Osler’s bright young men, that it was his practice to pain gentianviolet over a wart and then assure the patient firmly that it would be gone ina week, and he never saw it fail. There have been several meticulous studiesby good clinical investigators, with proper controls. In one of these, fourteenpatients with seemingly intractable generalised warts on both sides of thebody were hypnotised, and the suggestion was made that all the warts onone side of the body would begin to go away. Within several weeks, theresults were indisputably positive; in nine patients all or nearly all the wartson the suggested side had vanished, while the control side had just as manyas ever.” (Thomas, 1979)
The strength of the placebo effect, and the sometimes dramatic responses thatare obtained, might partly explain why GPs usually prescribe tricyclicantidepressants at doses experts say are ineffective - ie no more effective thanplacebo. The fact that they do makes it seem all the more remarkable thatSSRIs apparently have no greater effect on depression than the traditionaldrugs they have begun to overtake.
Clinical advantages claimed for SSRIs
None of the SSRIs has any
more specific effect on “depression” than other drugs, some in use for 40 years.
This would not be for want of trying to prove a difference, since anymanufacturer who could demonstrate his drug was in fact more effective thanthe rest would sweep the board. So, have the SSRIs become so popular becausethey are safer or otherwise more acceptable than alternatives ? Most expertsbelieve so and many consider the advantages are great and worth the extra cost.
There are minor differences of emphasis, but the main message in promotionalmessages for SSRIs is of three main advantages, plus one. Just as newerantidepressants in their day were said to be more effective than tricyclics orMAOIs, the SSRIs are now claimed to have fewer unwanted effects thanalternatives; to be more acceptable to more patients (so fewer discontinuetreatment); and to be safer in overdose. The all-important net result is“evidence-based” claims that SSRIs give better therapeutic value for moneyover alternatives. That is no small claim, since a doctor with 100 patients onantidepressants at any one time could be costing the NHS between about £500and £30,000 per year, depending on the drugs prescribed.
The difference in cost between newer and older drugs is great for severalreasons, starting with the need to finance research into drugs for the future. Thepharmaceutical industry estimates the cost of bringing a new drug to market in1997 at about £200m-£250m. This implies that the total NHS drug bill wouldbe enough to pay the costs of developing only around twenty new chemicalentities each year.
A related reason for high costs is that the SSRIs are still under patent, so therecan be no competition from generic drugs. Therefore prices can be high inrelation to good alternative treatments - and partly kept high because there islittle significant price competition between the different SSRIs. This can beseen by looking at the US wholesale price (ie excluding the pharmacist’s mark-up) of the market leaders. (Medical Letter
Cost to pharmacist
30 days supply of:
$ US (1997)
With market leaders, comparable price differentials between major brandedand generic drugs tend to survive many years after expiry of patent life. Forwhatever reason, enough doctors believe their patients would suffer sufficientlyfrom taking an identical or near equivalent generic drug to justify their payingthrough the nose for an original brand. One may assume some enhancement ofplacebo effect with more expensive drugs, though it would be hard to know inthis case whether doctor or patient was more pleased.
The medical literature inevitably includes a broad spectrum of safety-relatedclaims, but the evidence overall does not suggest that SSRIs show any greatand decisive safety advantage over alternatives in day to day use. In pre-marketing controlled clinical trials of SSRIs, of the order of 15%-20% droppedout when suspected adverse effects became intolerable and in general practiceabout twice that proportion appear to quit within a month. (DSRU, 1993). Aftersix months, now the minimum recommended course, probably no more than
one-quarter to one-third of patients continue taking the SSRIs they started with;some switch to alternatives, others stop.
Evidence from controlled trials of the safety/efficacy of SSRIs compared withother antidepressants represents only a small proportion of all published drugassessments, but is the best available. Two independent meta-analyses, eachstarting with a careful search of the literature to identify all properly controlledtrials, have reached broadly similar conclusions - that SSRIs do have the edge onalternatives, but not by much. Results from 62 trials (mostly 4-6 weeks induration) showed a 54% drop-out rate with tricyclic antidepressants against a 49%drop out with SSRIs.
This suggests a decided (not decisive) advantage, as the overall difference “iscomparatively small and may not be clinically relevant” (Anderson & Tomenson,1995). Another analysis of 63 trials, including 16 which compared an SSRI witha non-tricyclic, showed that 3% fewer quit an SSRI because of side effects, withno difference in overall dropout rates nor for dropouts due to lack of efficacy(Song et al., 1993). The big picture is not dissimilar from the early days of thetricyclics (Kline, 1964) and many other drugs.
Little advantage for SSRIs is suggested by the flow of spontaneous (‘YellowCard’) reports of suspected adverse reactions to the Committee on Safety ofMedicines/ Medicines Control Agency (CSM/MCA). The actual numbers haveto be treated with great caution: many factors impede close interpretation offigures - not least that relatively few suspected adverse reactions are actuallyreported, even with serious and fatal reactions usually fewer than one in ten.
However, reports for the three main SSRIs, after less than ten years in use,approximates the total numbers of Yellow Cards reported for all prescribeddrugs in one year, and far exceed the numbers for supposedly moretroublesome antidepressants. (CSM/MCA, 1997) The percentage of all reportsattributable to one or another of these products is shown in the Table. Itsuggests that sertraline might possibly be a more agreeable starting point thanfluvoxamine, but probably otherwise indicate there is little to chose betweenany of them.
Market share, England,
% of ADR Yellow Card
1995 (% of SSRI sales
reports to March 1997 -
to NHS, by value)
Analysis of the overall safety profile is also complicated by a variety ofunhelpful methods of data presentation. For example, there is marked tendencyto over-differentiate between essentially related adverse effects in reports ofclinical trials. This confounds comparisons between drugs and seems to reducethe apparent frequency of many reported effects. Thus, nervousness, anxiety,agitation, restlessness and irritability might be listed individually as“infrequent“ adverse effects of treatment when, collectively, they might becounted as “frequent” manifestations of pretty much the same problem:
Percentage of patients in US pre-marketing trials reporting
(1) ‘anxiety’ or (2) ‘nervousness’, either on active drug or placebo
On active drug (total)
On placebo (total)
5.5 8.5 14
Venlafaxine 6 13 19
Fluvoxamine 5 12 17
Paroxetine 5 5.2 10
The true picture may be obscured also when reports of suspected side effectsfocus on the incidence but not the severity of reaction. The Table belowillustrates this problem in relation to the incidence of “headache” reported inpivotal studies in drug licence applications to the FDA. (Physicians DeskReference, 1996) These data imply no difference in severity of headache onactive drug or placebo. Nor indeed do they suggest any difference in incidence,though the consistently close correspondence of figures seems unreal.
Unblinding might well explain the pronounced placebo ‘tracking effect’ seenhere, as well as in the figures above:
Percentage of patients experiencing ‘headache’:
on active drug
The general picture for acute adverse effects is nevertheless reasonably clear, ifhard to predict. Commonly recognised SSRI side effects (ie usually affecting atleast 5% of patients) include agitation; anxiety; dizziness; headache; insomnia;nausea; nervousness; somnolence; drowsiness and tremor. Other regularlyreported effects (1% - 5% incidence) include loss of libido; sexual dysfunction;impaired concentration; confusion; abnormal dreaming and nightmares; andamnesia. In addition, around 1% of reports relate to aggression, hallucinations,
fatigue, malaise and depersonalisation. Characteristic of SSRIs is the broadspectrum of psychiatric and neurological side effects, resulting in over-stimulation in some cases and sedation in others.
Many users develop tolerance or otherwise adapt to such effects. However, theeffects of SSRIs on personality and cognitive and behavioural performance inlong-term use are not well understood. This was the issue that became centralin the benzodiazepine litigation. In the main legal action, claims for damageshad little to do with dependence as such; compensation was mainly sought inrelation to the alleged depersonalisation and related states resulting from theexcessive use which dependence had allegedly brought on. See 3.8.
Prozac, suicide and aggression
The complexities that underlie claims
made for the superior safety of SSRIs are also well illustrated in relation to riskof suicide, notably because fluoxetine and other SSRIs have proved safer inoverdose than tricyclic and other antidepressants. This has been highlighted asan important reason for prescribing the newer drugs: a Lilly-sponsoredsymposium concluded, for example, that for both legal and practical reasons,“it is difficult to justify the first line use of toxic antidepressants when saferalternatives are available” (Montgomery, 1994). However, it is not that simple:SSRIs are less toxic than tricyclics in overdose, but may not reduce the risk ofsuicide overall: “While it is accepted that fatal overdosage (with SSRIs) is lessof a problem, the overall incidence of death by suicide does not appear to havebeen reduced as patients have resorted to other means of suicide.” (Reynolds,1996)
Great emphasis has also been placed on the need to prevent suicide throughbetter recognition and treatment of depression. Failure to treat, andundertreatment, are regarded as major risk factors, and the perceived level ofrisk is high - “with a 15% risk of death from suicide with more severe forms ofdepression”. (National Depressive and Manic Depressive Association, 1997)However, this widely cited figure would be less relevant in general practice asit refers to the fate of patients hospitalised for depression. They would includemany resistant cases, people who hadn’t responded to drugs.
The essential proposition is that “depression probably precedes the largemajority of all completed suicides” (Paykel & Priest, 1992) and that the SSRIstreat “depression” most effectively of all. There is therefore some impliedlinkage between risk of suicide and low levels of brain serotonin. Indeed, thishas promoted much experimentation - including measurement of the levels ofthe principal metabolite of serotonin (5-HIAA) in the cerebrospinal fluid (CSF)of depressed, impulsive and aggressive patients, also of suicide victims. Theclinical significance of the overall findings is uncertain. Apart from the obviousdifficulties of distinguishing between cause and effect, no linkage was found inmost patients, but has been found with disorders other than depression:
“In brief, most authors conclude that a subgroup of depressed patients(35%) fall into a low CSF 5-HIAA group and that patients with low CSF 5-HIAA are more prone to impulsive, violent suicide. This finding has notbeen restricted to patients with depression but is also present in patients withother psychiatric illnesses (arsonists, some alcoholics and some
schizophrenics) who are suicidal or impulsive (Åsberg et al., 1987; Roy etal., 1990).” (Delgado et al., 1992).
By contrast, the following accounts emphasise the role of personalcircumstances and social factors in increasing the risk of suicide. The first is areflection dating from before the introduction of SSRIs, on “how to identifyand deal with the suicide-prone”; the other is a more recent account of theproblem as seen in the casualty department of St Mary’s Hospital, London. Thewriter of the first account is co-author of the second. He was also Chairman ofthe Defeat Depression Campaign and principal author of the Royal Colleges’guidelines for the treatment of depression. See 2.12, 2.13 below.
“If we want to pick out the person who will kill himself, many studieshave shown whom to look out for. The vulnerable patient is male ratherthan female, old rather than young, with a history of drug dependence,alcoholism or mental illness. He is childless and he is single, divorcedor widowed. He will be found living alone in a cheap hotel in a denselypopulated part of a big town. He gives a history of a broken home inchildhood, and recent break of routine (especially loss of job orretirement), and recent bereavement is common. He is likely to havehad some conflict with the law, to be geographically mobile, and to besuffering from physical illness .” (Priest, 1979)
A prospective study was conducted of all referrals to the emergencypsychiatric service of an inner-London hospital over one year. There were 53 individuals who presented with the specific and spontaneous complaint ofsuicidal ideation without any accompanying act of self-harm. The main diagnoses in this group were personality disorders (40%) and alcohol dependence (15%); only 13% were suffering from depressive illness.
Members of the group differed from the other 369 presenters to the service in that they were less likely to be accorded a diagnosis of a defined mentalillness, twice as likely to have a criminal record, and more likely to have aprevious history of deliberate self-harm. A quarter of the suicidalcomplainants were admitted to hospital following assessment.” (Hawley et al.,1992)
The question has also been raised, whether fluoxetine more than other SSRIsmight induce “suicidal ideation” and occasionally precipitate suicide attempts.
The manufacturers have denied it and regulatory authorities have agreed. Thereason the debate persists seems essentially to do with the difference betweenrisk and harm, there being good theoretical evidence of one but no compellingempirical evidence for the other. This could mean there was no problem or thatthe problem was rare, but would also reflect the many possible complicationsin research. Jick and colleagues, for example, found that the suicide rate withfluoxetine “seems to be substantially higher than that of the otherantidepressants”, but they concluded otherwise:
“. when the analysis was restricted to those without a history of havingfelt suicidal or who had only taken one antidepressant, the increasedrisk for those who took fluoxetine was reduced. We conclude that theincreased risk associated with fluoxetine in the current studies may beexplained by selection bias. Even after removing from the analysis
subjects with a history of being suicidal or taking multipleantidepressants, there may have been residual factors which reflected ahigher risk of suicide for subjects taking fluoxetine.” (Jick, et al., 1995)
This population-based study concluded that “the risk of suicide was notdetermined by the antidepressant prescribed”, and estimated the overallincidence to be one suicide per 1200 patient years. This would representthousands in a population of millions of users, but the role of the drug isuncertain and many factors might affect the numbers involved.
The research team at the centre of this controversy has acknowledged that “theoverwhelming preponderance of data indicate that these drugs are relativelysafe and of unquestionable value” and “have provided countless patients withundeniable relief”. Nevertheless, they suggest problems might be masked.
Following an extensive review, this team identified a range of clinicalmechanisms which might promote suicidal tendencies and concluded thus:
“Although antidepressants diminish suicidal ideation in many recipients,about as many patients experience worsening suicidal ideation on activemedication as they do on placebo. Furthermore, at least as many patientsattempted suicide on fluoxetine and tricyclic antidepressants as on placebo. These observations suggest that antidepressants may redistribute risk,attenuating risk in some patients who respond well, while possiblyenhancing risk in others who respond more poorly. Sophisticated studieswill need to be conducted to meaningfully explore this possibility”.
(Teicher, et al., 1993)
Has there been any discernible effect on suicide rates, since the start of the newwar on depression ? Suicide rates in the USA, (Mrela, 1997) where SSRIshave been most used, and in England (Department of Health, 1997) give noevidence of any national dose-response.
The possibility that fluoxetine (among other SSRIs) may trigger aggression andhostility has often been discussed, but the issue remains open to question.
, 1992) The many anecdotal reports of such effects are impossible toevaluate individually, as are the occasional reports of atrocities in which Prozacis alleged to have played some part, even when documented with great care.
(Cornwell, 1996) The complexities of analysis obscure almost everything, barthe feeling that it would be mad to assume they were all groundless.
Collectively, they add to the impression that all is not well, not that the courtswould be the place to establish what might have gone wrong. (Ibid.)
The popularity of Prozac and other SSRIs
The evidence from
clinical studies and from spontaneous reports of suspected adverse reactionscannot explain the explosive popularity of fluoxetine. Some of it can beattributed to the wealth of publicity in lay media - though most related to thealleged effects of Prozac in enhancing day to day living, rather than for“depression”. By 1994, the extent of publicity about Prozac as a feel-gooddrug, and references to “cosmetic psychopharmacology” and so on, hadreached such a pitch, the manufacturers decided to advertise their concern thatmuch of this “unprecedented publicity” had “trivialised the very serious nature
of the disease Prozac was specifically developed to treat - clinical depression”.
In a perfectly informed world, one might give no more credence to claims ofpersonality transformation than to anecdotal accounts of aggression andsuicidal ideation. But naturally eyebrows and/or expectations get raised when aUS West Coast clinical psychologist puts all of his 700 patients on Prozac(Toynbee, 1995), or when an East Coast psychiatrist tops the best seller list formonths with a book promoting the idea of Prozac as a key to happiness andgreater fulfilment in life, (Kramer, 1993) albeit with evidence befitting “notscience but soap”. (Medawar, 1994)
One cannot just dismiss accounts which suggest that, sometimes, Prozac anddrugs like it have astonishingly good effects, sometimes even when othertreatments have failed. Few if any drugs have attracted such publicity,including an array of feature articles with headlines like these:
“The Promise of Prozac .”; (Cowley et al., 1990) “That Prozac moment .”;(Rothman, 1994) “The pill of pills .”; (Nuland, 1994) “The cloud over bottledsunshine .”; (Doyle, 1994) “Power of the psychodrug .”; (James & Camden, 1993) “With millions taking Prozac, a legal drug culture arises .”; (Rimer, 1993) “Minddrug ‘no miracle’”; (Hunt, 1993) “Escape Capsule .”; (Bracewell, 1993) “The ProzacGeneration”; (Grant, 1994) “The Personality Pill.”; (Toufexis, 1994)
It is still not too clear what prompted this gush but, however exaggerated, itonly underlines an obvious point: if a placebo can sometimes produceastonishing effects, a potent drug surely can too. If a flank of warts can bothunderstand and respond to the suggestion they get lost, the only dangerousconclusion might be to assume that drugs like fluoxetine work exclusively andspecifically by pharmacological means. They clearly don’t.
Still, the idea that Prozac sometime works like magic has to be seen inperspective. Prozac has often proved good enough and sometimes veryeffective, but truly magic responses would be rare and neither are theypeculiar to fluoxetine. Shorter has recalled that when imipramine was firstgiven to depressed patients in 1955, “the response was ‘absolutelyincredible, so exciting’, electrifying both the hospital staff and the Geigyscientists who had been following this all with bated breath”. When it doeshappen, dramatic relief from bad depression must clearly seem astonishing,even without the hype:
“The language in which Kuhn reported the transformation is interesting,because it illustrates how resurrectionlike the recovery from depressioncan be, a recovery that each new generation of antidepressant drugsbelieves that it alone has achieved; witness the resurrectionist rhetoricaccompanying the introduction of the drug Prozac.” (Shorter, 1997)
The complexities involved in unravelling pharmacological effects from apossible placebo response come across quite well in the following account in
(Young and Depressed in America), the memoire of a youngwoman in treatment for “atypical depression”. (Wurtzel, 1996)
“It’s not just depression - it’s atypical depression. Who would have thoughtthey have a name to describe what is happening to me, and one thatpinpoints my symptoms so precisely ? In the book UnderstandingDepression, Donald F Klein MD and Paul H Wender MD characteriseatypical depressives as people who ‘respond positively to good things thathappen to them, are able to enjoy simple pleasures like food and sex, andtend to oversleep and overeat. Their depression, which is chronic rather thanperiodic and which usually dates from adolescence, largely shows itself inlack of energy and interest, lack of initiative, and a great sensitivity toperiodic - particularly romantic - rejection’. Those sentences perfectlydelineate my symptoms. I feel suddenly much less lonely . Enter Prozac,and suddenly I have a diagnosis. It seems oddly illogical: rather thandefining my disease as a way to lead us to fluoxetine, the invention of thisdrug has brought us to my disease .”
This successful treatment, following many which had failed, began with aconvincing diagnosis from a respected source, and was also linked to the nameof a special drug. Expectations of the new drug were high. The patient believedthat fluoxetine “is very pure in its chemical objectives” and that it “acts only onserotonin”. Her doctors had high hopes for it too: they were “completely gungho” about Prozac and “thought I was the perfect candidate for the drug” and“were all set to enrol me in a study that would have allowed me free treatmentand medical care .” But at first there were problems. The drug was slow to“kick in”, but her doctor urged her to persist: “I am so certain that thefluoxetine is going to help you really soon that I have just have to find a way tokeep you going through those next few days .”. The dosage was doubled butthen Wurtzel attempted suicide again:
“And then something just kind of changed in me. Over the next few days, Ibecame all right, safe in my own skin. It happened just like that. Onemorning I woke up, and I really did want to live, really looked forward togreeting the day, imagined errands to run, phone calls to return, and it wasnot with that feeling of great dread, not with the sense that the first personwho stepped on my toe as I walked through the square may well have drivenme to suicide. It was as if the miasma of depression had lifted off me, gonesmoothly about its business, in the same way that the fog in San Franciscorises as the day wears on. Was it the Prozac ? No doubt. Was it the catharticnature of going through a suicide attempt ? Probably. Just as I always saidthat I went down gradually and then suddenly, I also got up that way. Allthe therapy, all the travelling, all the sleeping, all the drugs, all the crying,all the missed classes, all the lost time - all of that was part of some slowrecovery process that came to the end of its tether at the same time that Ireached mine”.
Looking back, Wurtzel concluded that “the fact that Prozac in combinationwith other drugs has been, for the most part, a successful antidote” wasundoubtedly due to its effects on her body chemistry. It began with “years andyears of bad habits”, then “years and years of exogenous depression (a malaisecaused by external events)”. This “can actually fuck up your body chemistry so
much that you need a drug to get it working properly again”. This is how itseemed to work:
“It seemed that suddenly, sometime in 1990, I ceased to be thisfreakishly depressed person who had scared the hell out of people formost of my life with my mood swings and tantrums and crying spells,and I instead became downright trendy. This private world of loonybins and weird people that I had always felt I occupied had suddenlybeen turned inside out so that it seemed like this was one big ProzacNation, one big mess of malaise”
Being able to control severe and pervasive depression/anxiety is clearly a greatachievement, however accomplished and however such states arise.
The SSRIs arrived on the scene at the end of
the 1980s, just as benzodiazepine prescribing went into sharp decline becauseof concern about widespread dependence problems and the mass litigationarising from it. The companies marketing SSRIs of course wished to takeadvantage of this. Firmly labelling their products “antidepressants”, they setout to convince doctors of the value of their drugs and their advantages overanxiolytics.
“ . the temptation to market them (the SSRIs) as antidepressants is allbut irresistible. These compounds can be produced easily. They are farsafer than the earlier tricyclics and MAOIs. They are so safe that itbecomes a feasible proposition to take the current findings from socialpsychiatry and advise general practitioners that there are many moreuntreated depressives than was formerly thought; often conditionspresenting as anxiety stem from an underlying depression, and currentevidence suggests that antidepressants (in contrast to anxiolytics) need tobe taken chronically, in order to reduce the risk of relapse .” (Healy,1991)
The leadership in general practice and psychiatry did not need muchpersuading. In steering prescribers towards their drugs, the manufacturersenjoyed substantial support from a high-profile, professional initiative, whichthey in turn part funded. The “Defeat Depression” campaign was organised inthe UK (1992-97) by the Royal College of Psychiatrists (RCP) with the RoyalCollege of General Practitioners (RCGP) perhaps rather in tow. The thrust ofthe campaign was to explain depression and encourage people to recognise it;to persuade sufferers to come forward for treatment; and to emphasise that nostigma should attach to such a commonplace but distressing illness, a majorsocial problem as well. Only two years into the campaign, over three millionleaflets about depression had been circulated to the public and many otherinitiatives had been sponsored as well. (Royal College of Psychiatrists, 1992,1994, 1996)
The Defeat Depression campaign focused in particular on what the organisersbelieved were widely-held misconceptions. One concerned the public’s failureto recognise the value of drug treatment. Another was the general failure torecognise depression for the complex and hidden disease it may be. The launchof the Defeat Depression campaign was explained as a response to “the tragedythat, despite the availability of effective treatments, 70 per cent of sufferers gountreated”. In addition, there was the concern that depression, whenrecognised, was not treated aggressively enough: over the years, many surveyshad established that, as a general rule, GPs prescribe doses of drugs thatexperts consider ineffective. As GPs treat nine cases in every ten, this impliesthat most cases of depression are being treated with strong placebos. It wouldbe useful to know what exactly GPs are treating, and whether they appreciatesome things that experts don’t.
Perhaps by way of dissociating themselves from the BDZ debacle, theRCP/RCGP also addressed what they saw as a widespread, but mistaken beliefthat antidepressants were drugs of dependence. The Campaign’s first pressrelease was headlined, “Antidepressants not addictive .”, because a MORIpublic opinion poll commissioned by the Campaign had found that “78% of thepublic believe anti-depressants to be addictive”. “It is worrying”, said thelaunch press statement, “that people may fail to take the medicine in themistaken belief that it can cause dependence”. (RCP/RCGP 1992)
In unpublished correspondence, senior figures in both Colleges later explainedthat they saw no evidence of withdrawal problems (See 3.3) and mainly had inmind lack of evidence of antidepressant addiction and abuse. Essentially thesame points had repeatedly been made about the BDZs: (See 3.1)
“We have searched the literature and can find no reference to researchevidence that shows that (a) drug seeking behaviour or (dependence), or (b)rebound and withdrawal occur when prescribing antidepressant medication.” There is no street market in antidepressants. In fact it is our experiencethat it often difficult to get patients to take some initially, and to continue forthe recommended course length.” (McBride, 1992)
“The statement that antidepressants are not addictive is correct.
Antidepressant drugs do not result in drug-seeking behaviour, i.e. they donot have a market value, neither do they cause dependence in a technical useof the word.” Obviously a person who is still suffering from depressiveillness from whom the drug is then withdrawn would suffer a return ofdepressive symptoms that could have very serious consequences. This,however, is an indication of their efficacy not of dependence.” (Sims, 1992)
A former editor of the British Journal of Psychiatry
(published by the RCP)went further. Provoked by the suggestion that it seemed folly not to have testeddrugs like Prozac for their dependence potential (Medawar, 1994), he argued
that it was both mistaken and dangerous to have suggested that the question ofdependence arose at all: “It would be regrettable if serious depressive illness,often involving the risk of suicide, remained untreated through people beingmisinformed about the well-established properties of antidepressants .”.
“During the past 35 years, there has in fact been no evidence that anyantidepressants - whatever their structure - cause ‘addiction’ or‘dependence’. Medawar says there is ‘profound confusion’ over themeaning of these terms and, if so, he has certainly added to it. Diabetics aredependent on insulin and people with high blood pressure are dependent onhypotensives, in the sense they will become ill again if they stop taking thedrugs. Many sufferers from depression are in the same position, but this istotally different from the experience of people who take heroin or cocaine aseuphoriants.” (Freeman, 1994)
On this basis, the Defeat Depression Campaign emphasised the need forradically different standards of treatment. Fears of dependence weremisconceived and resulted from misunderstanding. In future, there should bemore prescribing for depression and at higher dosages than before, and seriousconsideration should be given to continuing treatment indefinitely.
Treatment guidelines and prescribing modes
Shortly after the
launch of the Defeat Depression Campaign, a Consensus Statement waspublished on recognition and management of depression in general practice.
One of the prime movers (Priest) was also the Chairman of the Campaign. Theconsensus statement and the Campaign were closely linked, though the authorsof the statement emphasised that their guidelines “do not necessarily reflect theofficial policy of either of the two colleges” (Paykel & Priest, 1992). Thestatement emerged from the proceedings of two conferences, each involvingabout 20 participants; no indication was given of any sources of support.
The statement began by referring to the tendency to prescribe tricyclics atineffective doses, noting also that “many newer compounds are less toxic inoverdose and have fewer side effects.” However, the major change was therecommendation to extend the duration of treatment. Previously, the generalidea had been to treat patients with antidepressants for several weeks and thenreduce the dose by about half and/or discontinue the drug. This was the advicegiven in the British National Formulary
until March 1996:
“Treatment should be continued for 2 weeks before suppression ofsymptoms can be expected and thereafter should be maintained at theoptimum level for at least another month before any attempt is made at dosereduction .”
By contrast, the recommendation in the Consensus Statement was to continuetreatment at the original dose, at least for a few months:
“. four to six months of antidepressant therapy after the initial treatmentphase is advocated to prevent relapse. There is no reason for a steepreduction to a ‘maintenance dose’, and drugs should be continued close to
the dose at which a clinical response was achieved, unless side effects makethis unacceptable .”
Over the years, increasing emphasis has been placed on the need for long-termor lifetime treatment, to prevent relapse. One turning point can be traced to theend of the 1980s, when the Committee on Safety of Medicines finally issuednew guidelines for prescribing benzodiazepines, to limit the risk ofdependence. (CSM, 1988) The new guidelines restricted BDZ prescribing to amaximum of a few weeks; before then they could be prescribed indefinitelyand often were. At the same time, the British National Formulary
deleted itslong-standing recommendation to consider prescribing BDZs andantidepressants together, immediately replacing it with a statement suggestingthe benefits of long-term use of antidepressants on their own:
“It may be appropriate during the early states of treatment to add a hypnoticto correct the sleeping pattern or an anxiolytic to allay anxiety or agitation.”(BNF
, No 18, September 1989)
“In recurrent depression, prophylactic maintenance therapy may need to becontinued for several years.” (BNF
, No. 19, March 1990)
The Consensus Statement developed this theme. Depression did tend to berecurrent and was potentially dangerous too, and the evidence showed thatantidepressants reduced the incidence of relapse. Patients more likely to relapsethan others, who would be candidates for long-term therapy, not only includedthose with a history of previous episodes of depression but also “patients wholack social support, and patients with continuing social difficulties (such asunemployment or dysharmony in interpersonal relationships)”. In addition:
“The patient clearly should be given as much information as possible indeciding whether to continue. Advice should include the facts thatantidepressants are not habit-forming or addictive and that a minimum offour months treatment is advised for classic depression to prevent relapse.
This will enable the patient better to make an informed choice aboutcontinuation with treatment.” (Paykel & Priest, 1992)
In spite of this emphasis on involving patients, the guidelines were clearlydesigned by professionals for professional use. They addressed the abidingpublic concern (Priest et al., 1996) that antidepressants were drugs ofdependence simply by denying it. The guidelines may or may not have beendesigned partly with fear of litigation in mind, but they hint of “defensivemedicine” and would certainly have reduced any risk. The chances of asuccessful action would be small, when depression was clearly identified as apotentially dangerous condition, yet under-recognised and undertreated; if theneed for and benefits of long-term antidepressant use were widely advocated;also if the most authoritative definition of “dependence” pointed irresistibly toproblems with the user, rather than to shortcomings of the drugs. See 3.5
The Consensus Statement went on to suggest that, even in less serious cases,drugs could prevent depression “for up to three years”. This (and other)recommendations represented a significant departure from the claims a
manufacturer would be allowed to make in the official prescribing referencedocument, the Data Sheet (UK) or Label (US). For example, the US Label forProzac says this:
“The efficacy of Prozac “was established in 5- and 6- week trials withdepressed outpatients . the antidepressant action of Prozac in hospitaliseddepressed patients has not been adequately studied . The effectiveness ofProzac in long-term use, that is, for more than 5 to six weeks, has not beensystematically evaluated in controlled trials .” (Lilly, 1996)
As well as advocating longer-term use, the Consensus Statement emphasisedthat doctors should be prescribing at higher dosages than most used. A majoranalysis of GP antidepressant prescribing (involving examination of 80,000NHS prescriptions written in early 1993) indicated a gulf betweenrecommended practice and normal prescribing regimes. It was found that 9 outof 10 prescriptions for tricyclics were for dosages below those recommended inthe Consensus Statement and that 4 out of 10 were for less than 6 months(Donoghue & Tylee, 1996). A similar picture emerged in the US: theConsensus Statement on the Undertreatment of Depression reported that “thevast majority of those treated with antidepressant medication are not prescribedan adequate dose for a long enough time”; taking into account the extent ofunderdiagnosis, “only about one in ten of those suffering from depressionreceived adequate treatment” (National Depressive and Manic DepressiveAssociation, 1997).
However, the perceived problem of low dose prescribing mainly concerned thetricyclics, the evidence suggesting that “newer antidepressants (lofepramineand the SSRIs) are prescribed comparatively well”. This finding has beeninterpreted as another reason for recommending general practitioners to useSSRIs. (Donoghue & Tylee, 1996) It also raises an issue directly relevant tothe question “Do antidepressants work?
” If SSRIs are prescribed attherapeutic doses, but tricyclics are used at doses which have effects ondepression comparable to placebos, why has there been no correspondingevidence of the superiority of SSRIs ?
Whatever the answer, “prescription of correct dosages and deciding uponappropriate maintenance therapy are difficult areas in antidepressantprescribing for both general practitioners and psychiatrists”. (Kerr, 1994) ForGPs in particular the problems start with the diagnosis. The originator of theHamilton Scale for Depression suggested that “an adequate interview willsurely not be less than half an hour” (Hamilton, 1967) much longer than theaverage GP would be able to spend. The situation may be comparable in theUS: a 1993 study by the Rand Corporation reportedly found that over half thephysicians surveyed wrote prescriptions after discussing depression withpatients for three minutes or less. (Wurtzel, 1996)
Dependence as an iatrogenic disease
There are dimensions to the risk
of dependence that go beyond the effects of drugs on the body, and notablythose relating to the question of the acceptability of risk. This is increasinglyseen as a matter for patients; their expectations and perceptions of benefit andrisk are central too. Naturally patients are very wary of dependence, but theyare not unresponsive to notions of benefit and risk: the more convinced apatient was of the value of continuing treatment, the lower any risk might seemto be. For example, the belief that an antidepressant provided the serotonin thatthe body needs could be expected to powerfully reinforce psychic dependenceon a drug, also to compound any risk of physical dependence by promotinglong-term use. Thus the real risk of dependence might increase, though thepatient might think it reduced. When someone believes in a drug and feels itworking well, dependence problems do not seem to arise.
Another important factor in determining the risk of drug dependence is thecommitment of those responsible to check well for evidence of problems, alsoto observe certain ground rules relating to openness and accountability. Thesignificance of this is well illustrated in the benzodiazepine (BDZ) dependencesaga - where the root problem was not about lack of data, but about thereluctance to seek out and reveal what there was. It was also about too manydoctors not heeding patients, nor seeing much need to; lack of criticalunderstanding; also perhaps too much of the conviction that meaning wellmeans doing good. (Medawar, 1992, 1996). Some doctors were of courseconcerned, but the focus early on was not specifically about some risk ofdependence; it was mainly about very high levels of prescribing (and repeats)and something of a diagnostic free-for-all.
“ ‘Well, Mrs Smith, I have listened to your story and examined you, and it seems tome you are a case of diazepam. You had better have some anxiety’. It seems to methat whether or not our patients are hooked on the drugs, the doctors are certainlyhooked on the diagnoses”. (Marinker, 1972)
The main counter-argument to this comes close to the justification often givenfor prescribing antidepressants today. This was the explanation given 20 yearsfor prescribing BDZs on such a scale (over twice levels of consumption today),and it had a decisive influence on national drug policy at the time:
“At first sight there appears to be a dramatic and disturbing increase in their use”(but) . “If we take relatively recent figures there is probably a level of significantpsychiatric morbidity in the population of some 20-30% of which about half isrecognised by general practitioners”. (in addition), “ ‘patients do not ordinarilyaccept psychotropic drug medication lightly.’ ” (More evidence is needed but, onthe best there is): “psychotropic drugs appear to be used conservatively by doctors(ie underused) rather than overused (and) their current consumption is notexcessive relative to the level of emotional morbidity in the community.” (Marks,1978)
With hindsight, it is clear that the evidence of benzodiazepine dependence wasthere all the time, though not revealed and/or obscured: blindness, bias and selfinterest all played their part, and wishful thinking too. Given the widespread
assumption (until the mid-1980s) that BDZs presented virtually no risk ofdependence, doctors rather assumed that people took them for years becausethey really worked.
Doctors also tended to think that, if patients got distressed when they tried tostop taking their drugs, it was a sign of relapse and the emergence ofunderlying illness. This reinforced the assumption that the drugs really workedand led to more prescribing and increased dependence. Much the same thinghad happened with other drugs prescribed for anxiety, insomnia anddepression, on many occasions this century, and in one before. (Medawar,1992)
The turning point in getting the BDZs recognised as drugs of dependence camein the early 1980s, just a year after the Committee on Review of Medicines(CRM) published the results of a “systematic review”. This had no discernibleeffect on prescribing behaviour and was a rather sloppy piece of work, muchinfluenced by a misconceived study by the former head of the UK subsidiary ofRoche, the leading manufacturer of BDZs (Marks, 1978). The CRM joined himin concluding that the incidence of dependence on benzodiazepines was about“5 - 10 cases per million patient months”, (Committee on Review ofMedicines, 1980) a figure which soon proved out by miles.
The turning point came with the publication of two controlled studies (Tyrer etal., 1981; Peturrson & Lader, 1981) which demonstrated that quite a highproportion of long-term BDZ users could be expected to experience significantwithdrawal symptoms when they tried to stop - enough to persuade many notto try. These studies were straightforward enough and presented quite modestevidence; it was still strong enough to prompt something of a retraction fromMarks (1983), if not from the CRM.
Though it was not much emphasised at that time, both studies alsodemonstrated the critical point that many BDZ withdrawal symptomspositively mimicked the symptoms of the conditions for which they wereusually prescribed. The most common symptoms of withdrawal were insomniaand extreme dysphoria (mainly anxiety, tension and depression). These wereeasily mistaken as evidence of a relapse, though some other symptoms (sensorychanges and impaired perceptions of movement) did help to distinguish thetrue withdrawal syndrome from any underlying condition.
These two studies began the process of legitimisation of many thousands ofpreviously overlooked complaints. It was a slow process, involving irresistiblepublic outcry; (Rantzen, 1985; Lacey & Woodward, 1985) then litigation (from1987) and finally some regulatory action. (CSM/MCA, 1988) By then it hadbecome clear that many long-term BDZ users recognised the dependenceproblem well before most doctors, and only after years of official denial thatsuch problems might exist. The error was great: in the early 1980s, officialestimates suggested that dependence might affect only a few dozen people;soon after, it seemed that between about 250,000 - 500,000 people might beinvolved (DTB
, 1985; Ashton, 1987; BMJ
“ . if the popular press and more recently the legal profession had not takenup arms against the overprescription of tranquillisers, the issue of
benzodiazepine dependence would still remain a medical curio only for thepages of medical journals. The media and lawyers have undoubtedly alteredprescribing practices, mostly for the better”. (Hallstrom, 1991)
This episode again reminded the medical establishment that dependence couldbe even more of a problem when withdrawal symptoms didn’t
appear. It took15 years even to speculate this might be so and another ten to begin to dosomething about it. The nub of it was explained in an aside cast by a pioneer inthe field, at a Roche-sponsored seminar held in 1976. Dr Leo Hollister’sinvestigations into the dependence liability of the BDZs, in the early 1960s,had convinced him there would be “a flood of reports of withdrawal reactions”for diazepam (Valium) and chlordiazepoxide (Librium), yet the flood nevercame:
“The probable reason is that patients abort these reactions early on becausethey think their original symptoms are returning, and they get back on thedrug. So we rarely see the full-blown picture.” (Hollister, 1977)
To this extent, the measure of a drug dependence problem is not so much thatsome people experience withdrawal symptoms, but that many others continuewith treatment to avoid them. With the BDZs, the focus was mainly on theproblem of the thousands who experienced sometimes vicious symptoms onwithdrawal - but the real problem was to do with the hundreds of thousands ofpeople who habitually suppressed them. Many still do.
Withdrawal problems with traditional antidepressants
MAOIs and the tricyclic antidepressants were introduced around forty yearsago, there was some concern about their dependence potential, as the labels“psychic energiser” and “psychostimulant” might imply. Given thatamphetamines were also widely prescribed for depression at that time (thoughnot then recognised as drugs of abuse), dependence on the new antidepressantswas never a major issue. Nevertheless, the clinical literature of the day didreflect some concern - for example: “In view of the euphoria sometimesproduced (by amitryptyline) there may be a small risk of addiction insusceptible individuals”. (Fullerton & Boardman, 1959)
Before long such concerns diminished. Confidence grew in the idea that thenew drugs had a specific action on depression; they were not pleasant totake; and it was also thought that problems could be contained by selectionof patients:
“Addiction to increasing the dose is not acquired, since raising the doseproduces unpleasant side-effects.” (Sargent, 1961)
“In view of the stimulant effects of the MAO inhibitors on verbal andpsychomotor behaviour, these drugs may be indicated in conditions inwhich slowing of thought and performance is more prominent than isdepression”. (Cole, 1964)
“That the antidepressants are not general euphoriants but act against a
specific biochemical type of depression is suggested by the fact that apatient’s condition may be completely unchanged by one antidepressantyet respond dramatically to another”. (Pare, 1965)‘
E m e r g i n g T r e a t m e n t s a n d T e c h n o l o g i e s O R I G I N A L Development and Validation of Stroke Risk Equation for Hong Kong Chinese Patients With Type 2 Diabetes The Hong Kong Diabetes Registry ILIN YANG, PHD RAMON R. LYU, PHD ING-YEE SO, FRCP DONALD D. YIN, PHD S troke is among the most common LICE P.S. KONG, FRCP CLIVE S. COCKRAM, MD HUNG-S
ARTÍCULOS DE REVISIÓN REV MED UNIV NAV Carcinoma microcítico de pulmón Carcinoma microcítico de pulmón J.L. Pérez-Gracia, J.M. López-Picazo, S. Martín-Algarra, S. Viteri, J. García-Foncillas, A. Gúrpide. Departamento de Oncología. Cínica Universitaria. Facultad de Medicina. Universidad de Navarra Correspondencia:José Luis Pérez-Gracia Departamento de OncologíaClínica U