Lagerberg et al. [1] reported on the risk of suicidal behavior before and after the initiation of SSRIs, based on the national Swedish register. This allowed a precise estimation of the course of suicide risk, which was highest before SSRI prescription and declined afterwards. However, because there was no control group, the effect of SSRIs remains unknown. There could be the same or even a better course without treatment or a different treatment. Suicidality, just like depression, is an episodic condition, so risk reduction over time could be fully explained by regression to the mean and spontaneous remission. Lagerberg et al. correctly state that there are “no strong grounds to claim that the risk reduction immediately following SSRI initiation is causal”. However, some parts of the text may be read as a causal interpretation, for example, “This suggests that, in the first year of treatment, SSRI medication does not bring the risk of suicidal behavior to baseline”. “Another statement that might be interpreted as causal claim is in the conclusion: The results suggest that, overall, clinicians should be reassured that SSRI initiation is not associated with an increased risk of suicidal behaviour”. Moreover, they point out problems with the representativeness of randomized controlled trials (RCT) and confounding in observational studies, thereby suggesting that their design is better to assess the effect of SSRIs on suicide risk.

Lagerberg et al. also did not discuss important research opposing the assumption that SSRIs reduce suicide risk, including a study by Björkenstam et al. which used the same Swedish register data, but with a case-crossover control design [2]. Björkenstam et al. found an increased risk for suicide at initiation of SSRI, especially in the second week after initiation. Lagerberg et al. dismissed this study because of confounding by indication, but according to Björkenstam et al.: “the descriptive question on whether there is an increased risk of suicide in the early phase of treatment, which is of great clinical relevance, is not confounded by indication”. In light of the increased risk especially in the second week, Lagerberg et al. should have reported their results week-wise, at least for the first month.

Log in or create a free account to read this content

Gain free access to this article, as well as selected content from this journal and more on nature.com

Access through your institution

or

Sign in or create an account
Continue with Google
Continue with ORCiD