A QUICK scan of the article by Kahn et al1 might lead the casual reader to conclude that antidepressant treatment does not reduce the risk of suicide. However, a careful analysis of the data shows that any such conclusion would be unfortunate and unsubstantiated, and that several important questions remain unanswered.
The article examines data from clinical trials that were submitted to the Food and Drug Administration as parts of 5 successful applications for indications in the treatment of depression. Among the 19 639 patients who participated in the trials, 34 committed suicide during the 4- to 8-week periods in which the trials were conducted. The authors, assuming a completely linear event-time relationship, converted this into an estimated annual rate of 757 suicides per 100 000 depressed patients treated with 1 of the 5 antidepressants. This suicide rate is nearly 70 times higher than the rate in the general population. The article indicates that this falls within the range of the estimated depressed patient suicide rates (between 275 per 100 000 and 1352 per 100 000) that the authors have found in the literature. Also, a community study in Sweden (the "Lundby" study) followed people for 25 years and found the annual suicide rate for men with a history of depression to be 650 per 100 000.2 In men whose depression caused severe impairment, the annual rate soared to 3900 per 100 000.