eNS Drugs 1 (3): 223-231.

1994
ADVERSE EFFECTS 1172-7047/94/0003-o223/S04.50/0

© Adis International Limited. All rights reserved.

The Fluoxetine and Suicide

Controversy
A Review of the Evidence
David Healy
Academic Sub-Department of Psychological Medicine, North Wales Hospital,
Denbigh, Clwyd, Wales

Contents
Summary ......... . 223
1. A Review of the Evidence. 223
2. Toxicity Indices . . . . . . . 224
3. Catecholamines and Suicide. 225
4. Serotonin and Suicide . . . . . 226
5. Databases versus Case Reports 227
6. A Plausible Mechanism? .... 228
7. Conclusions and Management Recommendations 229

Summary Evidence is emerging that a range of psychotropic drugs may precipitate

akathisia and/or panic reactions in predisposed patients. The use of fluoxetine has
hitherto been the most notable example of this occurrence. These reactions may
foster the genesis of suicidal ideation in a small proportion of patients. At present,
it is not clear what biological mechanism may underlie this finding. The serotonin
(5-hydroxytryptarnine; 5-HT) system may be involved in these reactions.
The best management of such reactions will involve counselling patients be-
forehand about the possibility of these reactions, stopping treatment with the
agents if such a reaction is suspected, or adding an agent with 5-HTIA antagonistic
properties (e.g. propranolol) to the treatment regimen.

In February 1990, a report by Teicher and col- 1. A Review of the Evidence

leagues appeared in the American Journal of Psy-
chiatry claiming that 6 patients who had been tak- The article by Teicher et al.[1] caused consider-
ing the selective serotonin (5-hydroxytryptamine; able controversy, probably for 3 reasons.
5-HT) reuptake inhibitor (SSRI), fluoxetine, had First, the newer generation of antidepressants,
become intensely suicidal.[1] The present article re- such as fluoxetine, were specifically designed to be
views this and other evidence in an attempt to answer less toxic in overdose than older agents. Therefore,
the question: can fluoxetine lead to the emergence of they were presumed to be less likely to cause fatal-
suicidal ideation? In addition, the appropriate man- ities through suicide. In contrast, the observations
agement for such an occurrence is discussed. of Teicher and colleagues suggested that these
224 Healy

agents somewhat perversely cause suicide fatali- Table I. Major criticisms[2.41 01 the data provided by Teicher et al. ll1
concerning Iluoxetine-indu ed suicide ideation
ties, albeit for a different reason.
Secondly, this report appeared to go against ob- No patients had major depression
servations that suggested that antidepressants act- Some patients had a diagnosis 01 borderline personality disorder
(a condition associated with prominent suicidal ideation and
ing on the noradrenergic system were more liable
impulsive behaviour)
to lead to suicide than those acting primarily on the
No patients were treated with Iluoxetine alone
serotonergic system. Thirdly, the findings con-
No patients were treated with Iluoxetine at currently
tradicted research that points to the existence of a recommended dosages
serotonergic deficit in individuals who commit sui- All patients were or had recently been receiving other medications
cidal acts. All patients had a history 01 suicide ideation or attempted suicide
There was immediate criticism of the article by
Teicher et al.[l] The main criticisms are listed in
table I. Subsequently, a meta-analysis of clinical with these 'unsafe' compounds, mianserin was
trials involving 3065 patients[S] and a retrospective found to be a relatively safe agent. This was despite
analysis of 1017 patients[6] who had received the fact that it induced agranulocytosis with a
fluoxetine was performed. These analyses sug- marginally higher frequency than other anti-
gested that fluoxetine was unlikely to lead to the depressants (a claim that has not been fully sub-
emergence of suicidal ideation. stantiated).
Nevertheless, a series of case reports involving A considerable part of the initial appeal of the
1 to 6 patients have continued to emerge implicat- SSRIs lay in their relative safety in overdose.
ing fluoxetine in the generation of suicidal ide-
While there have been insufficient prescriptions as
ation,l7-I3] In addition to these case reports, a re-
yet to construct fatal toxicity indices for each of the
analysis of the data presented by Fava and
SSRIs, these drugs have been specifically designed
Rosenbaum[6] was carried out. This has suggested
to be safe in overdose. As a result, it can be ex-
that the data can be interpreted such that fluoxetine
pected that their toxicity index will be low.l I8 ]
does indeed lead to the emergence of suicidal ide-
A standing comment in presentations over re-
ation.[I4] The frequency of suicide ideation in pa-
cent years has been that if the tricyclic antidepres-
tients receiving fluoxetine was somewhat higher
sants were to be developed now, they would not be
than that in recipients of tricyclic antidepres-
approved because of their toxicity. This sugges-
sants.l I4 ]
tion has led to the proposal that any programme
aimed at suicide reduction should encourage the
2. Toxicity Indices replacement of tricyclic antidepressants as first-
In the mid 1980s, the use of mianserin was being line treatments for depression, with SSRIs. How-
questioned because of reports of drug-induced ever, some have argued that in practice this would
agranulocytosis. As part of a programme aimed at make very little difference to the rates of suicide in
'saving' the agent, considerable research was done general practice.l I9 ,20]
on the number of deaths that occurred per million It is against this background that the questions
prescriptions of antidepressants. In the course of raised by the fluoxetine controversy must be
this and other work, the notion of a fatal toxicity judged. When considering the issue, a number of
index was developed.lIS-I7] authorities have pointed to the clear benefits
This index indicated that a number of com- gained from the use of fluoxetine in terms of its
monly prescribed tricyclic antidepressants, partic- relative safety in overdose and in patients with a
ularly desipramine, dothiepin and amitriptyline, range of other medical conditions, such as cardio-
were associated with a greater risk of death from vascular disease.[2-4] Such safety is clearly desir-
overdose than other antidepressants. Compared able for most depressed patients. However, this

© Adls International limited. All rights reserved. eNS Drugs 1 (3) 1994
Fluoxetine and Suicide 225

property is unlikely to carry much legal weight if 3. Catecholamines and SuiCide

some individuals are inevitably led to take their
own or others' lives as a direct consequence of re- As early as the mid 1960s, it was suggested that
ceiving fluoxetine. The fact that pertussis vaccines monoamine oxidase inhibitors (MAOIs) might be
reduce the overall likelihood of brain damage from more likely than tricyclic antidepressants to lead to
suicide.
the pertussis virus does not prevent legal action in
It was thought that this might occur via a
the event of vaccine-induced brain damage.
MAOI-induced increase in catecholamine [dopa-
Moreover, it has been pointed out in media re-
mine, noradrenaline (norepinephrine) and adrena-
ports that antidepressants are often developed us-
line (epinephrine)] levels. This effect might in-
ing clinical trials conditions that differ substan-
crease drive and arousal more rapidly than the
tially from those found in the real world. It can be
more sedative tricyclic antidepressants. Activating
argued that the clinical trials performed in the de- patients in this way might enable them to act on
velopment of the early tricyclic antidepressants their latent suicidal intentions. This idea never re-
(some of which reached the market place without ceived any clear empirical support; however, it has
any clinical trial programme at all) were biased. never fully gone awayJ4,25] Indeed, there have
This was because they focused heavily on inpatient been suggestions that the newly launched revers-
samples of severely depressed patients. However, ible monoamine oxidase A-selective agent,
criteria for entry into current antidepressant trials moclobemide, might be particularly likely to in-
specify that patients should not be suicidal, should duce suicidal ideation and lead to suicide at-
not have any co-existing medical conditions, tempts,l26,27]
should not be receiving any other treatment, and More specifically in 1988, Damluji and Fergu-
should not have bipolar disorders. Increasingly, the son[28] reported the occurrence of desipramine-in-
patients who enter such trials are recruited from duced dysphoria and suicidal ideation in 4 patients.
primary care settings and have depressive episodes As desipramine has a preferential effect on cate-
of relatively recent onset and relatively mild to cholamine rather than serotonin reuptake inhibi-
moderate severity. tion, this appeared consistent with the ideas noted
Improvements have been made in the conduct above.
of clinical trials, and there is now a broader defini- These ideas also found some support from a trial
tion of drug-induced adverse events. Despite this, by Rouillon and colleagues.[29] These investiga-
it can be argued that the use of sophisticated post- tors studied the long term outcome of patients
maintained on maprotiline, a catecholamine re-
marketing surveillance methods is the only accu-
uptake inhibitor, compared with those on placebo.
rate way of assessing the possibility of suicidal ide-
They found that maprotiline alleviated symptoms
ation induced by the newer agents. Available
of depression and reduced the risk of relapse more
post-marketing surveillance data from individual
effectively than placebo. However, use of the agent
countries do not suggest that any of the SSRIs are also led to a greater number of individuals attempt-
particularly liable to cause aggressive, impulsive ing and completing suicide during the follow-up
or suicidal actsJ21-23] If it is later established that year.
they or other antidepressants do cause such effects, These data have interfaced with ideas that sero-
the question would arise as to whether this is a tonin is implicated in the control of impulsive
direct or indirect effect. behaviourspO,31] and that SSRIs might be associ-
Teicher et al.[1] appear to have initially believed ated with reduced rates of impulsivity and at-
that the effects they described arose in a manner tempted or completed suicide.[32] It is clear that
that was not open to psychosocial intervention but SSRIs have effects on appetitive behaviours that
later have perhaps modified their positionJ24] appear loosely linked to an 'impulse system';

© Adis International Limited. All rights reserved. eNS Drugs 1 (3) 1994
226 Healy

hence the potential usefulness of SSRIs in eating and Shopsin[38] reported that the majority of 14
and sexual disorders)33,34] Unfortunately, the field studies of CSF 5-HIAA levels that had been per-
of impulse disorders remains unstandardised, and formed by 1979 did not support the conclusion that
it is not clear which disorders qualify as impulse depression was associated with reduced 5-HIAA
disorders. As a result, it is not known whether levels.
SSRIs have any specific effect on impulsivity. This did not prevent investigators at the Na-
tional Institute of Mental Health predicting respon-
4. Serotonin and Suicide siveness to antidepressants based on alterations in
CSF 5-HIAA levels. It was hypothesised that
In 1973, Van Praag and colleagues[35] claimed agents that supposedly increase serotonergic func-
that probenecid reduced cerebrospinal fluid (CSF) tion, such as SSRIs, should be more effective in
levels of 5-hydroxyindoleacetic acid (5-HIAA), depressed patients with reduced levels of CSF 5-
the metabolite of serotonin, in depressed individu-
HIAA. It followed that agents that preferentially
als more than it did in nondepressed individuals.
act on catecholamine reuptake might be expected
This finding found some support from other inves-
to be of greater benefit in individuals with reduced
tigators,l36] Asberg and colleagues[37] also re-
catecholamine metabolism,l45] However, this has
ported a reduction of 5-HIAA levels in the CSF of not been found to be the case,l39,40,46,47]
depressed patients. This latter finding, however,
Quite apart from these negative findings, in gen-
was not clearcut and was not associated with any
eral, studies of 5-HIAA have not been vigorously
one category of depression. Subsequent investiga-
controlled methodologically. They have not ade-
tion has suggested that reduced CSF 5-HIAA levels
is not a feature of depressive illness.[38-40] quately controlled for the effects of physical activ-
ity[48-50] or alcohol intake, [5 1,52] both of which are
Nevertheless, investigation in this area contin-
clearly implicated in suicidal or aggressive states.
ued. In 1981, Traskman, Asberg and colleagues
found that depressed patients as a group do not Indeed, the basic question of whether 5-HIAAlev-
display a particular reduction of 5-HIAA levels. els in CSF taken from lumbar samples reveal any-
However, they claimed that there was a disorder of thing more than the state of serotonin metabolism
serotonin metabolism in those who are depressed in the lumbar spinal cord[53] has not been settled.
and suicidaU4I] Subsequently, the same group[42] Efforts to approach the problem by looking at
reported that the correlation between reduced 5- other aspects of serotonin metabolism have not
HIAA levels and suicidal behaviour was one that clarified the situation. Evidence in favour of 5-HT2
cut across diagnostic categories. receptor and [3H]-imipramine binding site abnor-
In supporting this claim, the investigators pre- malities in post mortem brains of suicide victims
sented data suggesting that altered 5-HIAA meta- has been cited,l3,44,54] Unfortunately, few if any of
bolism might predict future suicide attempts. Oth- these findings have been replicated. [47] In addition,
ers have pointed to correlations of 5-HIAA levels just as with work on 5-HIAA, post mortem brain
with previous suicide attempts or aggressive research is subject to many confounding method-
acts,l43] Still others have accounted for heteroge- ological variables. The contaminating influence of
neity in the data by appealing to distinctions be- such factors has not been well controlled for.
tween intentional and impulsive acts and distinc- Despite all these caveats, the idea that there is
tions between self-injury and suicide proper.[44] an abnormality of serotonin metabolism in suicidal
It would appear that initial claims regarding re- states has gained hold. Indeed, in recent months
duced 5-HIAA levels in the CSF were readily ac- reports of cholesterol-lowering agents possibly
cepted by the biological psychiatry community. In precipitating aggressive or suicidal acts have been
contrast, subsequent failure to replicate the origi- greeted with interest. There have been a number of
nal findings have not been. For example, Annitto proposals linking this effect with an association

© Adls International Limited, All rights reserved, eNS Drugs 1 (3) 1994
Fluoxetine and Suicide 227

between suicide and serotonin deficits, by means Table II. Criteria required for a case report of a drug-induced
of an effect of lipid lowering on serotonergic me- adverse effect to be considered reliable l61 ]

tabolism.l 55 -60] As Owens has noted)581 however, • There should be a close temporal relationship between
administration of the drug and occurrence of the adverse effect
the existence of a biologically plausible explana-
tion for this finding may prevent further indepth • The adverse effect should not be a manifestation of the
underlying illness
analysis of the situation. In a broadly similar man-
• The individual should not be taking other medications
ner, were the existence of an association between
• The effect should be clearly documented
reduced serotonin levels and suicide to be con-
• Discontinuation of the drug should lead to cessation of the
firmed, far from solving the problem, this finding
adverse effect
would risk compromising the analysis of the
• The adverse effect should reemerge on readministration of the
fluoxetine and suicide data. compound
Even if the findings of altered serotonin meta-
bolism in suicidal states were to be replicated, it is
not clear that treatment with SSRIs would produce
general marking down of HRSD scores as patients
or reduce suicidal ideation. Even when depressed
in a trial improve. Even in patients in whom suici-
individuals with reduced 5-HIAA levels are se-
dal ideation may have emerged or increased, there
lected for treatment with SSRIs, responses to treat-
may be a simultaneous improvement in other as-
ment do not appear to be particularly enhanced or
compromised.[39,40,46,47] pects of their mental state. This effect would have
masked increased ratings on item 3.
In addition, patients have a natural bias to min-
5. Databases versus Case Reports imise cognitive dissonance and so would be un-
In reply to the case reports of fluoxetine- likely to volunteer information regarding emerging
induced suicidality, Beasley and colleagues suicidal ideation. This is particularly true if they
scrutinised the Eli Lilly database for evidence of were experiencing improvements in other aspects
increased suicidality in patients receiving fluox- of their mental state. Unless the investigators were
etine.l 5] No such evidence has been found. These suspicious of the induction of suicidal ideation, it
data from several thousand patients, and the evi- is unlikely that they would have picked up such
dence that fluoxetine reduces suicidal ideation, emerging tendencies.
must on any scientific scale outweigh the dubious Case reports are clearly an unreliable form of
evidence of a handful of case reports. information. Several criteria have been proposed
There are, however, problems with the use of to enhance the validity of conclusions that may be
such a database. The evidence collected as part of drawn from such reports (see table II). To these we
the clinical trial programme for fluoxetine in the might add the criterion that there should be a clin-
early 1980s did not specifically focus on the ques- ically plausible mechanism that might make sense
tion of suicidal ideation. There was no reason to of the claim of an adverse effect.
have such a focus. Accordingly, scales that might While the initial reports by Teicher et al.[l] of
be particularly sensitive to variations in mental suicidal ideation after fluoxetine administration
state and associated induction of suicide were not were clearly compromised on many of these cri-
used. The primary variable used in the database teria, subsequent reports[7-11] have not been.
was ratings on item 3 of the Hamilton Rating Scale Across all reports there is a broad correspondence
for Depression (HRSD). This is a insensitive item, regarding the time of emergence of suicidal ide-
particularly when it comes to rating possibly emer- ation - about 10 to 14 days after initiation of
gent suicidal ideation. fluoxetine treatment. It is clear from many reports
Furthermore, clinical investigators will tend to that some individuals were prescribed fluoxetine
mark down any scores on this item in line with a only and in currently recommended doses. The ef-

© Adls International limited. All rights reserved. eNS Drugs 1 (3) 1994
228 Healy

fects have been clearly documented. Although sui- receiving fluoxetine, it is not clear that the aka-
cidal ideation is a manifestation of depressive ill- thisia referred to in the above case reports has in-
ness, intense suicidal ideation of the form reported volved this type of effect. The states being referred
is less commonly encountered. Furthermore, it cer- to might also be labelled agitation or dysphoria
tainly would not be expected to emerge frequently rather than akathisia. Unfortunately the phenome-
in individuals taking placebo. nology of such drug induced restlessness/dys-
With this particular adverse effect, there are phoria is not well clarified.l68-701.
problems with a rechallenge. Few investigators or There are 2 mechanisms, possibly related,
patients would be anxious to revisit a compound whereby drug-induced 'akathisia' may lead to sui-
that had caused suicidal ideation. Rothschild and cidal ideation. First, it would seem likely that some
Locke[71, however, did just this in 3 patients. They depressed individuals might misattribute the de-
found that whilst suicidal ideation cleared on dis- velopment of drug-induced dysphoria or agitation
continuation of fluoxetine, it reemerged with to a worsening of their mental state. Seeing this as
readministration of the agent. However, readmini- a worsening of their depressive illness despite drug
stration of fluoxetine was not done in a double- treatment, they may conclude that their case must
blind manner, and hence these reports are open to be hopeless.l 341
some bias. Creaney et a1.[11] found reemergence of An alternative mechanism has recently
suicidal ideation in 2 individuals who had experi- emerged. In 1989, Weissman and colleagues[711 re-
enced these effects after receiving fluoxetine when ported an association between suicidality and
they were rechallenged with SSRIs other than panic disorder. In their analysis of findings derived
fluoxetine. This suggested that the problem of a
from the epidemiologic catchment area study,
possible induction of suicidal ideation was not spe-
panic disorder appeared particularly likely to be
cific to fluoxetine.
associated with both suicide attempts and com-
pleted suicides. In response to criticism from oth-
6. A Plausible Mechanism? ers,£72- 741 these authors have revised their esti-
The induction of suicidal ideation by fluoxetine mates downwards of how likely panic is to lead to
seems to be in opposition to prevailing knowledge suicide.l751 However, they still argue that the data
of the causes of suicide and mechanism of action point strongly toward an association between panic
of SSRIs. However, a possible mechanism mediat- and suicide, particularly when the symptoms of
ING such an effect can be postulated. panic occur during an episode of depression.
It has commonly been suggested that individu- Indeed, panic may provide a final common
als who experienced fluoxetine-induced suicidal mechanism through which a number of unrelated
ideation had developed akathisia.l1. 7,1l,62, 631Inde- states lead to suicidal ideation. For example,
pendent of such reports, there have also been case Creaney and colleagues[1l1 reported 1 patient who
reports implicating fluoxetine in the development became suicidal following a dissociative reaction
of akathisia.l 641 Furthermore, fluoxetine seems to to SSRIs. This reaction was characterised by pro-
be associated with a higher incidence of ex- nounced depersonalisation and derealisation. Such
trapyramidal adverse effects than other antidepres- a mechanism may furthermore underline instances
sants.l651 It has also been noted that the induction of paradoxical aggression which have been re-
of akathisia by antipsychotics may lead to a precip- ported following administration of a variety of an-
itation of suicidal attempts.[66,67] tidepressants. [761
The phenomenon induced by fluoxetine may Against this background, the notion that the pre-
not be accurately described as 'akathisia'. Aka- cipitation of agitation by an antidepressant might
thisia traditionally refers to a visible restlessness lead to suicidal ideations/suicide attempts seems
and pacing. While this can be observed in patients less remarkable. In addition, it has recently been

© Adls International limited. All rights reserved. eNS Drugs 1 (3) 1994
Fluoxetine and Suicide 229

recognised that persons who are anxious are par- sponses to psychotropic agents for a range of mu-
ticularly likely to develop a 'jitteriness' syndrome tant genes. If the genes were found to be abnormal,
when treated with antidepressants.l 77 ] family members could be invited to undergo a dou-
Power and Cowen[l8] have argued that such ad- ble-blind challenge with the psychotropic drug of
verse effects may be related to a serotonin particular concern in their relative. Such a method
behavioural syndrome. This can be blocked by pro- might lead to conclusive findings relatively rap-
pranolol, a ~-adrenoceptor blocking agent that also idly, and might in due course lead to the estab-
has 5-HTIA antagonistic properties. As a conse- lishment of diagnostic tests to determine whether
quence, propranolol might be particularly effective there are particular psychotropic drugs that certain
at alleviating any restlessness/akathisia caused by individuals should avoid.
SSRIs. Indeed, Rothschild and Locke[7] reported
that propranolol appeared to relieve the akathi- 7. Conclusions and Management
sialsuicidal ideation precipitated by re-exposure to Recommendations
fluoxetine.
In the opinion of this author, the volume of case
If suicidal or homicidal ideation develops in pa-
reports and other studies is sufficient to demonstr-
tients taking SSRIs or antipsychotics as a conse-
ate that antidepressants and antipsychotics may in-
quence of the development of akathisia or panic
duce suicidal ideation in certain individuals under
and aggravated by a process of misattribution or
certain conditions. These reports must be set
other cognitive process, such states would not be against a general recognition that the peak time for
directly drug-induced. Whether or not such states suicide attempts is shortly after individuals start
occur is still a matter that needs to be determined treatment with a new antidepressant or antipsy-
conclusively. Nevertheless, there appears to an in- chotic. Therefore, it seems a reasonable proposi-
creasing consensus that these states arise via some tion that, in some patients, taking a new psy-
secondary indirect process rather than directly. As chotropic agent may lead to the emergence of
such, it is likely that the burden of responsibility suicidal ideation.
for drug-induced injury, should any cases be con- The fluoxetine controversy, for a number of rea-
tested, is likely to shift from the makers of SSRIs sons, has served to highlight this issue. Some of the
to the prescribers and to issues of clinical practice. case reports regarding fluoxetine-induced suicidal
The induction of significant akathisia or panic ideation do appear sufficiently well documented
during antidepressant treatment is infrequent and and detailed to sustain an argument that fluoxetine
unpredictable. Furthermore, there is a range of may lead to the emergence of suicidal ideation.
variables that might confound the interpretation of However, this effect is uncommon.
causation. Therefore, large scale clinical trials are Suicidal ideation associated with fluoxetine is
unlikely to settle the question of whether and how probably not a direct effect of fluoxetine, but is
often such reactions occur. A more promising line mediated through the induction of akathisialagita-
of investigation might be to investigate genetic fac- tion/panic. This may occur through changes in 5-
tors influencing pharmacokinetic and behavioural HTIA or 5-HT2 receptors. The effects may be suc-
responses to psychotropic drugs. cessfully blocked by propranolol.
Brunner and colleagues[78] have recently iden- Inappropriate reactions to akathisia, agitation or
tified a point mutation in a structural gene for panic may also playa significant part in any sub-
monoamine oxidase A. This mutation is associated jective distress that emerges. Any deleterious ef-
with an increased incidence of impulsive aggres- fects of such reactions will be prevented by a clear
sion, arson, attempted rape and exhibitionism. awareness on the part of the therapist and patient
Thus, it might be possible to investigate those in- that such reactions may develop. Patients should
dividuals who have significantly unusual re- be instructed to report any such effects promptly,

© Adis International Limited. All rights reserved. eNS Drugs 1 (3) 1994
230 Healy

so that potential misattributions by the patient may 18. Power AC, Cowen PJ. Fluoxetine and suicidal behaviour. Some
clinical and theoretical aspects of a controversy. Br J Psychi-
be dealt with. atry 1992; 161: 735-41
Such reactions to fluoxetine are rare, and undue 19. Donovan S, Freeman H. Deaths related to antidepressants: only
consideration. J Drug Dev 1990; 3: 113-20
concern over them may lead to more suicides by
20. Kelleher MJ, Daly M, Kelleher MJA. The influence of antide-
virtue of patients giving up treatment. It is this au- pressants in overdose on the increased suicide rate in Ireland
thor's opinion that acknowledging that such reac- between 1971 and 1988. Br J Psychiatry 1992; 161: 625-8
21. Ahmad SR. Fluoxetine "not linked to suicide". Lancet 1991;
tions may occur with all psychotropic drugs, and 338: 875-6
developing management guidelines for such reac- 22. Committee on Safety of Medicines. Safety of fluoxetine (Pro-
tions, is essential for all parties involved in the zac): comparisons with fluvoxamine (Faverin). Curr Probl
June 1992
therapy. 23. Consumer's Association. Fluoxetine, suicide and aggression.
Drug Ther Bull 1992: 35-6
24. Teicher MH, Glod C, Cole JO. Dr Teicher and associates reply.
Acknowledgement Am J Psychiatry 1991; 148: 1260-2
25. Benkert 0, Hippius H, editors. Psychiatrische phar-
Dr Healey acts as a consultant for Eli Lilly. makopherapie. Berlin: Springer, 1980: 5
26. Danish University Antidepressant Group. Moclobemide: a re-
References versible MAO-A-inhibitor showing weaker antidepressant ef-
1. Teicher MH, Glod C, Cole JO. Emergence of intense suicidal fect than clomipramine in a controlled multicentre study. J
preoccupation during fluoxetine treatment. Am J Psychiatry Affect Disord 1993; 28: 105-16
1990; 147; 207-10 27. Fahy TJ. Side effects of moclobemide in depressed patients
2. Goldblatt MJ, Schatzberg AF. Does treatment with antidepres- refractory to other treatments. Ir J Psych Med 1993; 10: 24-7
sant medication increase suicidal behaviour? Int Clin Psy- 28. Damluji NF, Ferguson JM. Paradox of the worsening of depres-
chophannacoI1991;6:219-26 sive symptomatology caused by antidepressants. J Clin Psy-
3. Mann 11, Kapur S. The emergence of suicidal ideation and chophannacol 1988;8: 347-9
behaviour during antidepressant phannacotherapy. Arch Gen 29. Rouillon F, Phillips R, Serrurier D, et al. Rechutes de depression
Psychiatry; 1991: 48, 1027-33 unipolaire a et efficacite de la maproteline. Encephale 1989;
4. Molcho A, Stanley M. Antidepressants and suicide risk: issues 15: 527-34
of chemical and behavioural toxicity. J Clin Psychophannacol 30. Depue RA, Spoont MR. Conceptualising a serotonin trait: a
1992; 12 Suppl. 2: 13S-8S behavioural dimension of constraint. Ann NY Acad Sci 1986;
5. Beasley CM, Dornseif BE, Bosomworth JC, et al. Fluoxetine 487: 47-62
and suicide: a metanalysis of controlled trials of treatment for 31. Spoont MR. Modulatory role of serotonin in neuroinformation
depression. BMJ 1991; 303: 685-92 processing: implications for human psychopathology. Psy-
6. Fava M, Rosenbaum JE Suicidality and fluoxetine: is there is a chol Bull 1992; 112: 330-50
relationship? J Clin Psychiatry 1991; 52: 108-11 32. Montgomery SA. Anxiety and depression. Petersfield: Wright-
7. Rothschild AJ, Locke CA. Re-exposure to fluoxetine after seri- son Biomedical Publishing Ltd., 1990: 118-9
ous suicide attempts by 3 patients: the role of akathisia. J Clin 33. Beaumont G, Healy D. The place of clomipramine in the devel-
Psychiatry 1991; 52: 491-3 opment of psychophannacology. J Psychophannacol1993; 7:
8. Wirshing WC, Van Putten T, Rosenberg J, et al. Fluoxetine, 378-88
akathisia and suicidality: is there a causal connection? Arch 34. Healy D. Psychiatric drugs explained. London: Mosby Year
Gen Psychiatry 1992; 49: 580-1 Books, 1993
9. King RA, Riddle MA, Chappell PB, et al. Emergence of self 35. Van Praag HM, Korf J, Schut E. Cerebral monoamines and
destructive phenomena in children and adolescents during depression: an investigation with the probenecid technique.
fluoxetine treatment. J Am Acad Child Adolesc Psychiatry Arch Gen Psychiatry 1973; 28: 827-31
1991; 30: 179-86 36. Goodwin FK, Post RM. CSF amine metabolites in affective
10. Masand P, Gupta S, Dwan M. Suicidal ideation related to illness. J Psychiatr Res 1974; 10: 320-30
fluoxetine treatment. N Engl J Med 1991; 324: 420 37. Asberg M, Bertilsson L, Tuck D, et al. Indoleamine metabolism
II. Creaney W, Murray I, Healy D. Antidepressant induced suicidal in the CSF of depressed patients before and during treatment
ideation. Hum Psychophannacol1991; 6: 329-32 with nortriptyline. Clin Phannacol Ther 1973; 14: 277-83
12. Dasgupta K. Additional case of suicidal ideations associated 38. Annitto W, Shops in B. Neuropharmacology of mania. In:
with fluoxetine. Am J Psychiatry 1990; 147: 1570 Shopsin B, editor. Manic illness. New York: New York Aca-
13. Hoover CE. Additional cases of suicidal ideations associated demic Press, 1979: 105-64
with fluoxetine. Am J Psychiatry 1990; 146: 1570-1 39. Gerner RH, Fairbanks L, Anderson G, et al. CSF neurochemis-
14. Brewerton TD. Fluoxetine induced suicidality, serotonin and try in depressed, manic and schizophrenic patients compared
seasonality. Bioi Psychiatry 1991; 30: 190-6 with that of normal controls. Am J Psychiatry 1984; 141:
15. Pinder R. The benefits and risks of antidepressant drugs. Human 1533-40
PsychophannacoI1988;3:73-86 40. Healy D. The structure of psychophannacological revolutions.
16. Cassidy SL, Henry JA. Fatal toxicity of antidepressant drugs in Psychiatr Dev 1987; 4: 349-76
overdose. BMJ 1987; 295: 1021-4 41. Traskman L, Tybring G, Asberg M, et al. Monoamine metabo-
17. Henry JA. The safety of antidepressants. Br J Psychiatry 1992; lites in the CSF and suicidal behaviour. Arch Gen Psychiatry
160: 439-41 1981; 38: 631-6

© Adis International limited. All rights reseNed. eNS Drugs 1 (3) 1994
Fluoxetine and Suicide 231

42 Traskman-Bendz L, Asberg M, Bertilsson L, et aI. CSF mono- 61. Edwards G. Antidepressants and murder. Psychiatr Bull 1992;
amine metabolites of depressed patients during illness and 16: 537-9
after recovery. Acta Psychiatr Scand 1984; 69: 333-42 62. Chouinard G. Fluoxetine and pre-occupation with suicide. Am
43. Brown GL, Linnoila MI. CSF serotonin metabolite (5-HIAA) J Psychiatry 1991; 148: 1258-9
studies in depression, impulsivity and violence. J Clin Psy- 63. Opler LA. Fluoxetine and pre-occupation with suicide. Am J
chiatry 1990; 51: 31-41 Psychiatry 1991; 148: 1259
44. Van Praag HM. Biological suicide research: outcome and lim- 64. Lipinski JF, Mallya G, Zimmerman P, et aI. Fluoxetine - induced
itations. Bioi Psychiatry 1986; 21: 1305-23 akathisia: clinical and theoretical implications. J Clin Psychi-
45. Maas J, Koslow S, Davis J, et aI. Biological component of atry 1989; 50: 339-42
the NIMH clinical research branch collaborative pro- 65. Baldwin D, Feinberg N, Montgomery S. Fluoxetine, fluvoxam-
gramme on the psychobiology of depression. Psychol Med ine and extra-pyramidal tract disorders. Int Clin Psy-
1980; 10: 759-77 chopharmacoI1991;6:51-8
46. Maas J, Koslow S, Katz M, et al. Retreatment neurotransmitter 66. Drake RE, Ehrlich J. Suicide attempts associated with akathisia.
levels and response to tricyclic antidepressants drugs. Am J Am J Psychiatry 1985; 142: 499-501
Psychiatry 1984; 141: 1159-71 67. Shear MK, Frances A, Weiden P. Suicide associated with
47. Healy D, Paykel ES. Neurochemistry of mood disorders. In: akathia and depot fluphenazine treattnent. J Suicide 1983; 3:
Asbury AK, McKhann GM, McDonald WI, editors. Dis- 235-6
eases of the nervous system. Philadelphia: Saunders, 1992; 68. Van Putten T. The many faces of akathisia. Compr Psychiatry
I: 815-30 1975; 16: 43-7
48. Green AR, Grahame-Smith DG. 5-hydroxy and other indoles in 69. Kahn EM, Munetz MR, Davis MA, et aI. Akathisia: clinical
the CNS. In: Iversen LL, Iversen SD, Snyder SH, editors. phenomenology and relationship to tardive dyskinisia.
Handbook of psychopharmacology, Vol. 3. New York: Ple- Compr Psychiatry 1992; 33: 233-7
num Press, 1975: 169-245 70. Singh MM. Dysphoric response to neuroleptic treattnent and
49. Post ER, Kotin J, Goodwin F. Psychomotor activity in CSF schizophrenia and its prognostic significance. J Dis Ner Syst
amine metabolites in affective illness. AmJ Psychiatry 1973; 1976; 191-6
130: 67-72 71. Weissman MM, Klerman GL, Markowitz JS, et aI. Suicidal
50. Meek JL, Neff MH. Is CSF the major source for the removal of ideation and suicide attempts in panic disorder and attacks. N
5-hydroxy indole acetic acid from the brain? Neuropharma- Engl J Med 1989; 321: 1209-14
cology 1973; 12: 497-503 72. Beck AT, Steer RA, Sanderson WC. Dr Beck and colleagues'
51. Davis VE, Cashaw JL, Hoff JA, et al. Identification of 5- reply. Am J Psychiatry 1992; 149: 1412-3
hydroxyptophol as a serotonin metabolite in man. Proc Soc
73. Fava GA, Grandi S, Savron G, et aI. Panic disorder and suicidal
Exp Bioi Med 1966; 122: 890-2
ideation. Am J Psychiatry 1992; 149: 1412
52. Feldstein A, Hoagland H, Freeman H, et aI. The effect of ethanol
74. Montgomery SA. Anxiety and depression. Petersfield: Wright-
ingestion on serotonin-C 14 in man. Life Sci 1967; 6: 53-61
son Biomedical Publishing Ltd., 1990: 26-7
53. Bulat M. Some criteria for the study of biochemical processes
75. Johnson J, Weissman MM, Klerman GL. Panic disorder, co-
in CNS: analysis of cerebrospinal fluid. Clin Neuropharmacol
morbidity and suicide attempts. Arch Gen Psychiatry 1990;
1984; 7 Suppl.l: 153S-4S
47: 805-8
54. Golden RN, Gilmore JH, Corrigan MH, et aI. Serotonin, suicide
and aggression: clinical studies. J Clin Psychiatry 1991; 52 76. Rampling D. Aggression of paradoxical response to tricyclic
Supp1.l2: 61-9 antidepressants. Am J Psychiatry 1978; 135: 117-8
55. Engleberg H. Low serum cholesterol and suicide. Lancet 1992; 77. Pohl R, Yeragali VK, Balon R, et aI. The jitteriness syndrome
338: 727-9 in panic disorder treated with antidepressants. J Clin Psychi-
56. Conroy RN. Low cholesterol and violent death. Irish J Psych atry 1988; 49: 100-4
Med 1993; 10: 67-70 78. Brunner HG, Nelen M, Breakefield XO, et al. Abnormal behavi-
57. Hawton K, Cowen P. Low serum cholesterol and suicide. Br J our associated with a point mutation in the structural gene for
Psychiatry 1993; 162: 818-25 monoamine oxidase A. Science 1993; 262: 578-80
58. Owens D. Low serum cholesterol and suicide. Br J Psychiatry
1993; 162: 818-25
59. Bond A. Low serum cholesterol and suicide. Br J Psychiatry
1993; 162: 818-25 Correspondence and reprints: Dr David Healy, Academic
60. Elliott M. Low serum cholesterol and suicide. Br J Psychiatry Sub-Department of Psychological Medicine, North Wales
1993; 162: 818-25 Hospital, Denbigh, Clwyd LL16 5SS, Wales.

© Adis International Limited. All rights reserved. eNS Drugs 1 (3) 1994