Starting two antidepressants to prevent suicide: A potentially dangerous recommendation?

We agree with Horgan and Malhi (2018) that intensive contact is an effective means to combat suicide, but we were truly amazed about their recommendation to start two antidepressants. Meta-analyses of randomized controlled trials are considered the most credible and reliable evidence base that we have in medicine, so we will check what this research reveals about the suicide risk related to antidepressants in adults. All meta-analyses conducted thus far found either no statistically significant association between suicide risk (including both suicides and suicide attempts) and newer generation antidepressants or a statistically significant increase in suicide risk in antidepressant groups relative to placebo (e.g. Fergusson et al., 2005). We stress that no single meta-analysis conducted thus far found a significantly lower suicide risk in antidepressant groups relative to placebo recipients; hence, based on the most stringent evidence available, we must assume that newer generation antidepressants likely increase the suicide risk in both children and adults (Healy, 2006). Horgan and Malhi (2018) specifically recommend a combination of venlafaxine and mirtazapine. However, exactly these two drugs appear to convey the largest suicide risk in a dose-dependent manner according to a real-world analysis of a representative primary care database (Coupland et al., 2015).

The next question is, whether combining antidepressants is an effective means to attenuate depression symptoms after all. Horgan and Malhi (2018) cite just one small 6-week trial (n = 105) as evidence that antidepressant co-medication is effective in reducing depression symptoms, but fail to mention a much larger trial (n = 665) by Rush et al. (2011) which found that combined medications are no more effective than monotherapy both at 12-week and at 7-month follow-up. So, we think the jury is still out on this, but the evidence for combining antidepressants surely is not that conclusive and convincing as Horgan and Malhi (2018) assume.

In sum, there is a growing body of evidence from meta-analyses of randomized controlled trials (Healy, 2006) and from representative real-world pharmacoepidemiologic studies (Coupland et al., 2015) that newer generation antidepressants may increase the suicide risk, while there still is a lack of consistent evidence whether combining antidepressants is an effective strategy to reduce depression symptoms (Rush et al., 2011). We therefore contend that the recommendations of Horgan and Malhi (2018) to start two antidepressants to prevent suicide is at best empirically unsubstantiated and at worst conflicting with the literature and thus potentially dangerous.