Talk therapy or medication for depression: what is the evidence fight actually about?

Is talk therapy or medication better for depression? Researchers are still arguing about it, and this lesson sets out what each side points to.

Start with where sources agree. Psychiatric medication is the most frequently prescribed therapy for major depression, but psychotherapy may be effective alone or combined with medication. Wikipedia's summary of the evidence says antidepressants are in general as effective as psychotherapy for major depression, and that more recent research suggests cognitive behavioural therapy can perform as well as antidepressants in moderate to severe depression. The same source notes that significantly more patients drop out of antidepressant treatment than psychotherapy, likely because of side effects.

Now the fight. Antidepressants consistently outperform placebo in meta-analyses, but the difference is modest, and it is not clear that statistical superiority amounts to clinical benefit. Proponents answer that the main rating scale (the HDRS) is not suited to measuring drug action and that the threshold for "clinical significance" is arbitrary. Irving Kirsch and Joanna Moncrieff propose that antidepressants work largely through placebo mechanisms. Even the size of the placebo puzzle is argued over: Kirsch estimated that spontaneous remission and regression to the mean account for only about 25% of the improvement on antidepressants, while Michael Hengartner has argued they might account for most of it. Critics of antidepressant trials also point to industry involvement and publication bias.

Two scientists even drew opposite conclusions from similar results: one saw little evidence to support prescribing antidepressants to any but the most severely depressed unless alternatives failed; the other said the "glass" is far from full but not completely empty, and noted that psychotherapy, the first-line alternative, does not have superior efficacy. A 2022 meta-analysis by Leichsenring and colleagues found the effects of both psychotherapies and psychiatric drugs "limited", despite half a century of research.