Antidepressants and suicide: what is the balance of benefit and harmBMJ 2004; 329 doi: https://doi.org/10.1136/bmj.329.7456.34 (Published 01 July 2004) Cite this as: BMJ 2004;329:34
All rapid responses
In the article “Antidepressants and suicide: what is the balance of
benefit and harm,” the authors argue:
“Concern is growing that serotonin reuptake inhibitors (SSRI’s) may
precipitate suicidal behavior especially in children.”
“Reassuringly, although antidepressant prescribing in Britain has
more than doubled in the past 15 years, population suicide rates have
“If the risks of SSRI associated suicidal behaviour seen in children
were to apply to suicide in adults, the number of ‘antidepressant induced
suicides would be small enough to be masked by currently favourable
This argument ignores the question of whether any risk of suicide
associated with the use of antidepressants could be even further reduced
consistent with the aggressive use of these drugs to prevent suicide.
Paroxetine, one of the SSRI’s mentioned in the article is metabolized
by the cytochrome P450 (CYP) 2D6 genotype and differences in this genotype
can result in elevated plasma paroxetine levels in the blood.1 These same
differences in CYP2D6 genotype have been known to cause life-threatening
serotonin syndrome in some patients taking paroxetine3 and serotonin
syndrome is associated with suicidality.4
Pharmacokinetic testing is now available for use in clinical
psychiatric practice to test for genotypic differences in patients with
different CYP2D6 genotypes.5 It would seem that the use of these
diagnostic tests could result in a more efficacious use of SSRI’s to
prevent suicide and thereby further reduce its incidence in the population
of suicide-prone patients.
1. Gunnell D, Ashby D. Antidepressants and suicide: what is the
balance of benefit and harm. BMJ Jul 3 2004; 329:34-38
2. Sawamura K, Suzuki Y, Someya T. Effects of dosage and CYP2D6-
mutated allele on plasma concentrations of paroxetine. Eur J Clin
Pharmacol. 2004 Oct; 60(8):553-7.
3. Sato A, Okura Y, Minagawa S, et al. Life-threatening serotonin
syndrome in a patient with chronic heart failure and CYP2D6*1/*5. Mayo
Clin Proc. 2004 Nov; 79(11):1444-8.
4. Serfaty MA, McCluskey S, Eccleston D. Extreme suicidality
following serotonin syndrome. Br J Psychiatry. 1995 Sep: 166(3):401-402.
5. de Leon J, Armstrong SC, Cozza KL. Clinical guidelines for the use
of pharmacogenetic testing for CYP450 2D6 and CYP450 2C19. Psychosomatics.
2006 Jan-Feb; 47(1):75-85.
Competing interests: No competing interests
In the context of the current debate on the efficacy of
antidepressants the following may be of interest.Ireland's suicide rate
has risen threefold since antidepressants became available with the
greatest increase occurring following the introduction of the SSRIs.
Additionally, admission rates (first and readmission rates) for depressive
disorders to psychiatric units and hospitals have shown no decline (unlike
those for schizophrenia) nor has there been any decrease in length of stay
for those hospitalised for depression over these time periods.
Competing interests: No competing interests
Drs Gunnell and Ashby’s paper makes it clear that the argument that
SSRI antidepressants could not be causing suicide because national suicide
rates have been declining coincident with increasing SSRI use is not a
strong argument. It is quite possible to envisage a drug induced problem,
even though overall national rates of suicide have fallen.
However, the figures put forward in Gunnell and Ashby, indicative of
a rising prescription rate since the launch of these drugs, may be
misleading, in that a rising prescription rate does not mean a rising rate
of new users. In order to establish what a rising prescription rate means
it is necessary to have a mechanism to translate prescriptions into
patients. We hope to submit such a model for peer review later this year.
Our modelling at present indicates, as does a moment’s reflection, that
the bulk of rising prescription rates stems not from an increasing number
of new users but rather from an accumulating number of long term users of
This point is important in that the risk of suicide with SSRIs has
been linked primarily to the early weeks of treatment. If this is the
case, then any increase in suicides from increasing SSRI use in Britain
will have occurred in the years from 1989 through to 1996, after which our
model shows that the number of new patients starting on SSRIs stabilises,
and the contribution of SSRIs to national suicide rates should remain at
some constant level. Estimates as to the numbers of patients adversely
affected should also be based primarily on those starting treatment, while
those benefiting might be expected to be concentrated among longer term
users, provided the effects of withdrawal on inducing suicide are not too
great. This latter issue has not been studied in randomised trials,
although the relative risk of a suicidal act in the post-treatment phase
of the recently posted paroxetine studies in children appears to be of the
order of 4.3 times greater on drug compared to placebo.
Graham Aldred & David Healy MD FRCPsych
Gunnell D, Ashby D (2004). Antidepressants and Suicide: what is the
balance between benefit and harm. British Medical Journal 329, 34-38.
GA has no competing interests in the sense that these are usually understood.
In recent years DH has had consultancies with, been a principal investigator or clinical trialist for, been a chairman or speaker at international symposia for or been in receipt of support to attend meetings from :
Astra-Zeneca Boots/Knoll Pharmaceuticals Eli Lilly Janssen-Cilag Lorex-Synthelabo Lundbeck Organon Pharmacia & Upjohn Pierre-Fabre Pfizer Rhone-Poulenc Rorer Roche SmithKline Beecham Solvay
He has been an expert witness for the plaintiff in seven legal actions involving SSRIs and have been consulted on a number of other attempted suicide, suicide and suicide-homicide cases following antidepressant medication, in the majority of which he has offered the view that the treatment was not involved. He has also been an expert witness for the NHS in a series of LSD (46) and ECT (1) cases.
Competing interests: Astra-ZenecaBoots/Knoll PharmaceuticalsEli LillyJanssen-CilagLorex-SynthelaboLundbeckOrganonPharmacia & UpjohnPierre-FabrePfizerRhone-Poulenc RorerRocheSmithKline BeechamSolvay
In my response above I stated
"these NNTs are not negligible even if they are smaller than one would like" whereas I meant to say that "these NNTs are not negligible even if they are larger than one would like"
Competing interests: No competing interests
I hope I am not the only person who is a little uncomfortable with the balance in Gunnell and Ashby's clinical review of the risks versus benefits of SSRIs.1 I am sure that this is an excellent area for review, as there is much misinformation on this topic. However, I am not sure how much this article clears the air.
Regarding the issue of risk, unless I am mistaken, the authors taken adverse "suicide related event" data reported by the MHRA (largely over-arousal, suicidal thoughts and self-harm) to explicitly mean completed suicide. They state themselves that no suicides actually occurred in the data released by the MHRA in children and adolescents. Yet they ignore this and multiply an estimate of the total number of completed suicides per year in those taking antidepressants by the relative risk (incorrectly stated as "odds ratio" in table 1) of suicide related events to calculate the excess number of suicides attributable to antidepressants. Even if both figures that they quote were correct, the final figure would be the excess of deliberate self-harm in the worst case and over-arousal in the best case (but more likely a heterogeneous composite effect). Taking the worst case, there is actually good data linking deliberate self harm with suicide and the study with the longest follow up showed a 13% risk of suicide after deliberate self-harm over 37 years.2 This would translate into an excess of 30 possible cases not 233 for men as stated and 20 cases rather than 255 for women (see table 2). Of course, even this smaller number would be a concern if real and not vastly outweighed by beneficial effects in the long term (both in terms of treating the depressive syndrome and reducing complications therein) which appears to be more likely.3
Regarding the issue of benefits of SSRIs, Gunnell and Ashby state that most SSRIs seem to be ineffective for childhood depression. Others disagree.4 The Whittington meta-analysis quoted by the authors shows a number needed to treat (NNT) for response totalling 5 for fluoxetine, 10 for sertraline and 20 for paroxetine, even allowing for unpublished reports.5 If we consider depression to be a serious condition, these NNTs are not negligible even if they are smaller than one would like. The authors also state that evidence of beneficial effects of antidepressants on suicide is hard to find. This is certainly true is one only looks at short term RCTs. However, there is evidence from much longer term studies.6 For example Winokur's group (1989)7 found that in 1076 patients with depression followed for between two and thirteen years, the standardized mortality ratio was highest in those patients who received either inadequate or no treatment. If one expands the consideration to include suicidal behaviour there is some evidence for prevention of deliberate self-harm after first presentation using an SSRI.8
The point I am making is that we are in danger of throwing the baby out with the bathwater if we stop using drugs with rare serious adverse effects and do not use anything in their place, leaving the patient essentially untreated. Clearly more research evidence is need about the benefits and risks of SSRIs but it may be sobering to remember that less than one in ten patients who are depressed in the community receive adequate doses of antidepressants, regardless of their suicide risk.9 It is important therefore to present a balanced view.
1 Gunnell D, Ashby D. Antidepressants and suicide: what is the balance of benefit and harm. BMJ 2004; 329: 34-38.
2 Suominen K et al. Completed Suicide After a Suicide Attempt: A 37-Year Follow-Up Study. Am J Psychiatry 2004; 161:563-564.
3 Isacsson G. Suicide prevention - a medical breakthrough? Acta Psychiatr Scand 2000;102:113-117.
4 Varley CK. Psychopharmacological treatment of major depressive disorder in children and adolescents
JAMA 2003; 290 (8): 1091-1093.
5 Whittington CJ, Kendall T, Fonagy P, et al. Selective serotonin reuptake inhibitors in childhood depression: systematic review of published versus unpublished data. LANCET 2004; 363 (9418): 1341-1345.
6 Moller HJ. Suicide, suicidality and suicide prevention in affective disorders. Acta Psychiat Scand 2003; 108: 73-80.
7 Black DW, Winokur G, Mohandoss E, Woolson RF, Nasrallah A. Does treatment influence mortality in depressives? A follow-up of 1076 patients with major affective disorders. Ann Clin Psychiatr 1989; 1, 165-173.
8 Verkes R, van der Mast R, Hengeveld M, et al. Reduction by paroxetine of suicidal behaviour in patients with repeated suicide attempts but not major depression. Am J Psychiatry 1998;155:543-7.
9 Suominen KH. Isometsa ET. Henriksson MM. Ostamo AI. Lonnqvist JK. Inadequate treatment for major depression both before and after attempted suicide. American Journal of Psychiatry. 1998;155(12):1778-80.
Competing interests: No competing interests
your questioning BMJ commentary 3 July is as puzzling as Gunnell & Ashby's "Antidepressants: what is the balance of benefit and harm" p34-38. Why do neither your comment nor Gunnell & Ashby mention lithium? How can it be ignored when, after 150 years of medicinal use, it has been acknowledged in Europe, UK Canada and now USA for at least 30 years as the best?
The German psychiatric and Medicines Council heads (Muller- Oerlinghausen et al) noted recently, 'suicide-related mortality among patients with affective disorders is approximately 30 times higher, and overall mortality 2 to 3 times higher, than suicide-related mortality in the general population. Lithium demonstrated possibly specific antisuicidal effects apart from its prophylactic efficacy: it significantly reduces the high excess mortality of patients with affective disorders. Rational treatment strategies most likely would demand that prescription rates be about 10 times higher(1).
'Recent findings indicate that lithium markedly reduces suicide risk in major affective disorders; comparing lithium with carbamazepine and amitriptyline, during 2.5 years randomized prospective long-term study (N = 378), of the nine suicides and five attempted suicides, none took place during lithium treatment(2). 'Findings support the view that lithium has a specific antisuicidal effect over and above its prophylactic benefit'(3).
'A recent international multi-centre trial is compatible with long- term lithium treatment extending survival of patients suffering from affective disorders to match the general population. A similar reduction of mortality was found in Canadian patients, In patients given lithium for two years or longer (n = 641), both suicide and cardiovascular mortality were the same as, or only slightly higher than, in the general population. In addition to its ability to prevent recurrences, prophylactic lithium treatment appears capable of reducing both the excess suicide risk and excess cardiovascular mortality of affective illness(3)'.
Goodwin(USA) 'as reviewed elsewhere in this issue, massive data have accumulated on antisuicide effect of lithium(4).
Coppen(UK) in 1994 reported 15 year follow-up of 103 patients: 'compliance with the regimen was high; with <2% suicide rate in the group--one of the patients had discontinued taking lithium some months before the suicide(5)'.
as Mogens Schou (Denmark) summed up in 1994: 'after 40 years, I hoped that lithium could be rejected as something better emerged... That has not happened.. current alternatives may be better for some, but we still await something radically superior to lithium.(6a). Mortality of manic depressive patients is two to three times higher than that of the general population, a highly significant difference. In studies reviewed, the patients' mortality during lithium treatment was not significantly higher or only moderately higher than in the general population. After discontinuation of lithium the mortality was again significantly higher. The number of patients attempting suicide was 6-15 times lower and the number of patients completing suicide was 3-17 times lower when they were on lithium than they were not. Similar observations have not been reported for prophylactic treatment with other mood stabilizers'(6b)
Baldessarini and Tondo(Harvard) 'reviewed published reports on long- term lithium treatment (1970-1996),and in 360 patients with DSM-IV bipolar disorder who entered into lithium maintenance monotherapy after 1970. Lithium is unmatched in research support for long-term clinical effectiveness against morbidity and mortality associated with depression or mania in bipolar I and II disorders. Data evaluated herein did not support suggestions that benefits of lithium have been exaggerated in the past or have been lost recently'(7)
Norman Rosenthal(8) noted 'their findings(7) are in accord with my own clinical experience that lithium is effective and well tolerated, especially at moderate serum lithium concentrations. One factor that may be responsible for lithium’s fall from grace among clinicians and patients is its status as an “orphan” or “poor relative.” Since lithium is cheap and unpatented, no wealthy drug company has any interest in demonstrating its merits as a mood stabilizer. Finally, lithium is regarded as old news and, as such, is less appealing to psychiatric researchers eager to make their mark on the field. Without the public relations dollars that boost its richer relatives into the public eye in favorable ways, lithium lacks their cachet. Lithium works. Psychiatrists and patients should not snub it because of orphan status'(8).
Refs: (1) Muller-Oerlinghausen B, Berghofer A, Ahrens B: The antisuicidal and mortality-reducing effect of lithium prophylaxis: Can J Psychiatry. 2003 Aug;48(7):433-9
(2)Thies-Flechtner K, Muller-Oerlinghausen B, Seibert W, Walther A, Greil W.Effect of prophylactic treatment on suicide risk in patients with major affective disorders. Data from a randomized prospective trial. Pharmacopsychiatry. 1996 May;29(3):103-7.
(3)Ahrens B, Grof P, Moller HJ, Muller-Oerlinghausen B, Wolf T Extended survival of patients on long-term lithium treatment.Can J Psychiatry. 1995 Jun;40(5):241-6.
(4)Goodwin FK.Anticonvulsant therapy and suicide risk in affective disorders. J Clin Psychiatry. 1999;60 Suppl 2:89-93;
(5)Coppen A. Depression as a lethal disease: prevention strategies.J Clin Psychiatry. 1994 Apr;55 Suppl:37-45.
(6a)Schou M: First International Conference on Bipolar Disorder, Pittsburgh, Pa, June 23, 1994.
(6b)Schou M: The effect of prophylactic lithium treatment on mortality and suicidal behavior: J Affect Disord. 1998 Sep;50(2-3):253-9
(7) Baldessarini RJ, Tondo L : Does lithium treatment still work? Evidence of stable responses over 3 decades Arch Gen Psychiatry 2000 Feb;57(2):187-90
(8)Norman E. Rosenthal.Lithium: An Orphan Drug: Arch Gen Psych:2001: 58:973
Competing interests: No competing interests