Summary: Major depression is the leading cause of disability worldwide. It has increased dramatically since 1990 with a huge burden for patients and healthcare system. Antidepressants are frequently used in routine practice in both developed and developing regions, however there is a long-lasting debate about their “true” efficacy and comparative effectiveness. We aimed to compare all licensed second-generation as well as four reference first-generation antidepressants against each other and with placebo in terms of efficacy and acceptability in the acute treatment of adults with unipolar major depression. The review is the update and extension of our previous network meta-analysis (Cipriani et al., Lancet 2009;373:746-58).

We identified 522 double-blind randomised controlled trials (116,572 patients) conducted between 1979 and 2016, and comparing agomelatine, bupropion, citalopram, desvenlafaxine, duloxetine, escitalopram, fluoxetine, fluvoxamine, levomilnacipran, milnacipran, mirtazapine, paroxetine, reboxetine, sertraline, venlafaxine, vilazodone, vortioxetine, nefazodone, trazodone, amitriptyline and clomipramine (which are included in the WHO List of Essential Medicines). All antidepressants were more efficacious than placebo, and some antidepressants were significantly better in terms of efficacy and/or acceptability than other active drugs. Estimated differences between antidepressants were smaller in placebo-controlled trials compared to head-to-head studies.

These results have important clinical implications which are not restricted to secondary care and should inform patients, physicians, guideline developers and policy-makers on the relative merits of the different antidepressants.