A cohort study from Finland found that in a high risk populations, current use of an antidepressant was associated with increased risk of attempted suicide but a reduced risk of actual suicide. The authors note that the association between antidepressant and suicide is controversial, but that the rarity of completed suicide makes reliable study of the topic difficult. They suggest that prolonged (several years) follow-up of a high-risk population would give reliable estimates of the risk.

Previous attempted suicide is the most important risk factor for suicide, therefore they used a national register of hospital discharges to identify a cohort of people who had been hospitalised with a diagnosis of attempted suicide. They then related this to data from the national death register and prescription database to identify the risk of suicide and attempted suicide in the cohort, compared to use of different antidepressants and no antidepressant use.

The cohort included 15,390 patients, and mean duration of follow-up was 3.4 years. Over this period, there were 602 completed suicides, 7,136 attempted suicides, and 1,583 deaths in the study cohort. The analysis showed that antidepressant use was associated with an increased risk of attempted suicide (39%; P<0.001 ) compared to no use. Antidepressant use was, however, associated with a significantly lower risk of actual suicide (relative risk 0.91) compared to no use. Overall, there were no significant differences between the different drug groups, although analysis suggested a lower risk that the mean with fluoxetine and a higher risk with venlafaxine. Overall mortality was also significantly reduced by antidepressant use compared to no use (by around 31% to 41%). Analysis specifically of young people (age 10 - 19) produced similar results except for an increased mortality associated with paroxetine in this group (but note this is based on small numbers).

The authors conclude that in their high-risk population, current use of any antidepressants was associated with an increased risk of attempted suicide, but a reduced risk of actual suicide and a lower overall mortality compared to non-use. They note that as a non-randomised study, it has limitations, nevertheless it includes more suicides and attempted suicides than any previous study.