Selective serotonin reuptake inhibitor antidepressants and the risk of suicide: a controlled forensic database of 14857 suicides
Isacsson G, Holmgren P and Ahlner J;
Commented by , 16 Jun 2005
Background
The Department of Health in the United Kingdom has stated that paroxetine should not be used for depression among children and adolescents, and the US Food and Drug Administration has warned about the possibility that all antidepressants may increase the risk of suicide among young people.
However, these initiatives are based on a rather small number of randomised trials showing that treatment with SSRIs was associated with increased risk of suicidal ideation and attempts compared to placebo. No completed suicide occurred in any of these trials.
Method
In Sweden, all (suspected) unnatural deaths are subject to a routine forensic medical investigation including a toxicological screening at the Department of Forensic Chemistry. In the study, the prevalence of antidepressants in toxicologic screenings was compared for all cases of suicide in Sweden and a control group of cases who after forensic investigation were judged to be natural deaths or accidents.
Results
Among a total of 14857 cases of suicide 20% were screened positive for antidepressants and analogously, 5.8% were screened positive among 26422 controls. SSRI as a class was associated with decreased risk of suicide (OR= 0.83 (99% CI: 0.77-0.90)).
More specifically, the OR for paroxetine was 0.87 (99% CI: 0.60-1.28). The OR for tricyclic antidepressants was 1.01 (99%CI: 0.90-1.14) and the OR for other newer antidepressants than SSRIs was 1.78 (99%CI: 1.46-2.16). Totally, younger individuals were not at higher risk of suicide than adults.
Among the 52 cases of suicide in the age group under 15 years, seven were positive for antidepressants but none for SSRIs. Among the 326 cases of suicide in the age group 15-19 years, 13 were positive for antidepressants - and among these 6 for SSRIs.
Discussion
Pharmacoepidemiological observational studies as the present are influenced by confounding of varies kind, especially confounding by indication: patients with depression are at greater risk of being treated with antidepressants and of committing suicide than individuals without depression. For this reason results should be interpreted with caution.
Nevertheless, the data do not present support for the view that SSRI’s are associated with increased risk of suicide compared to other antidepressants. Further, the low rates of positive screenings for antidepressants among suicide cases (13% in children, 4% in adolescents and 23% in the whole group) suggest that undertreatment still is a large problem.
It is crucial for patients and psychiatry as a medical discipline to try to elucidate the association between antidepressants and suicide. This is impossible to do in randomised trials for ethical reasons and as suicide is a random event. Therefore there is a need for more pharmacoepidemiological studies using other designs, which may help to reduce the effect of confounding by indication.