The magnitude and heterogeneity of antidepressant response in depression: A meta-analysis of over 45,000 patients

To determine the relative variability and magnitude of symptomatic improvement in antidepressant-treated individuals compared to placebo-treated individuals, and to investigate moderating factors.Multiple databases and previous publications were searched through February 2019 to identify all randomized controlled trials comparing placebo and antidepressants in acute treatment of depression. Primary outcome was relative variability of change in symptom severity in antidepressant-treated individuals compared to placebo-treated patients quantified using the coefficient of variation ratio (CVR).Of 9389 identified records, 134 were found to be eligible (total n = 46,646). Antidepressant-treated patients showed a significantly greater magnitude (g = 0.28, 95% CI 0.25-0.30, p.0001) and lower variability (CVR = 0.94, 95% CI 0.93-0.95, p.0001) of change in symptom severity relative to placebo-treated patients. Compared to placebo antidepressant-related improvement was more uniform in older studies (z = 3.01, p = .003) and in studies where antidepressants showed greater efficacy (z = -7.21, p.0001). | Imipramine, moclobemide, amitriptyline and mirtazapine showed significantly lower CVR than several other antidepressants. However, no difference in CVR exists between multiple and single-neurotransmitter profile antidepressants (z = -0.01, p = .99).There is lower variability and greater magnitude of change in symptom severity with antidepressant treatment relative to placebo. This is not consistent with our hypothesis that there are distinct sub-groups of treatment-responsive and treatment-resistant patients with major depression. Our results in-stead suggest that antidepressants show a relatively uniform effect.