Therapy and antidepressants performed equally. Together they beat either alone.

Across 101 randomised trials covering 11,910 patients, psychotherapy and antidepressant medication produced statistically indistinguishable rates of response in adult depression — a relative risk of 0.99, with a confidence interval running from 0.92 to 1.08 [s1]. Combining the two beat either one on its own, by about a quarter in relative terms [s1].

That is the shape of the answer to the question people ask most often about depression treatment, and the interval around the head-to-head comparison is narrow enough to be informative rather than merely inconclusive.

What was pooled, and how

The analysis searched bibliographical databases for randomised trials in which a psychotherapy and a pharmacotherapy for depression were compared with each other, or in which the combination was compared with either one alone [s1]. The primary outcome was treatment response, defined as 50% improvement between baseline and endpoint; remission and acceptability were also examined, with acceptability defined as study drop-out for any reason [s1]. Depression in most of the included studies was moderate to severe [s1].

Combined treatment was more effective than psychotherapy alone (RR = 1.27; 95% CI 1.14–1.39) and more effective than pharmacotherapy alone (RR = 1.25; 95% CI 1.14–1.37) in achieving response at the end of treatment [s1]. Similar results were found for remission [s1]. And no significant difference was found between psychotherapy alone and pharmacotherapy alone (RR = 0.99; 95% CI 0.92–1.08) [s1].

The null result on the head-to-head comparison is the one worth dwelling on. A confidence interval of 0.92 to 1.08 around a relative risk is not a shrug — it is a reasonably tight bound that rules out a large advantage in either direction, on this outcome, in this population.

The acceptability finding, which points one way

Where the two approaches did separate was in how many people stopped. Combined treatment (RR = 1.23; 95% CI 1.05–1.45) and psychotherapy alone (RR = 1.17; 95% CI 1.02–1.32) were both more acceptable than pharmacotherapy [s1].

Acceptability here means drop-out for any reason, which bundles side effects, inconvenience, preference and life circumstances into a single number [s1]. It does not identify why people left. But it is a real, measured difference, and it is the kind of outcome that rarely reaches the summary version of this literature.

Whether it holds for harder cases

The authors examined chronic depression, treatment-resistant depression and baseline severity as potential moderators, and report that the results were similar for chronic and treatment-resistant depression [s1]. Their own conclusion is that the combination of psychotherapy and pharmacotherapy seems to be the best choice for patients with moderate depression [s1].

They are equally clear about what remains open: more research is needed on the long-term effects of treatments, including cost-effectiveness, on the impact of specific pharmacological and non-pharmacological approaches, and on the effects in specific populations of patients [s1].

That last point matters. This analysis compares categories — psychotherapy, pharmacotherapy, combination — not particular therapies or particular drugs. A companion network meta-analysis by the same group, covering 331 trials and 34,285 patients, examined the main types of psychotherapy against each other and found that individual psychotherapies did not differ significantly from one another, with the single exception of non-directive supportive counseling, which was less efficacious than all the others [s2].

The limits

The trials pooled here were largely conducted in people with moderate to severe depression, which means the result does not automatically transfer to milder presentations [s1]. Response, defined as a 50% symptom reduction, is a useful threshold but a crude one; two people can cross it and be in very different states. And a network meta-analysis inherits every weakness of its constituent trials — unblinding is unavoidable in psychotherapy research, drug trials carry the publication-bias problems that industry funding introduces, and neither problem is fixed by pooling.

The long-term picture is thinner than the acute one, which is why the authors single it out as the priority for further research [s1]. The companion analysis found that most psychotherapies still had significant effects at 12-month follow-up compared to care-as-usual, with problem-solving therapy showing somewhat higher long-term efficacy than some others [s2] — but that is one strand of evidence, not a settled account of what happens over years.

What a reader can take from this

The choice between talking therapy and medication for moderate to severe adult depression is not, on this evidence, a choice between a strong option and a weak one. On the main efficacy outcome they performed the same [s1]. Fewer people dropped out of psychotherapy and combined treatment than out of medication alone [s1]. And doing both outperformed doing either [s1].

Which of those facts should carry the most weight for a particular person depends on that person's history, preferences, access and tolerance for side effects — a judgment that belongs with them and a clinician. This article describes evidence and is not medical advice.

Sources

• A network meta-analysis of the effects of psychotherapies, pharmacotherapies and their combination in the treatment of adult depression — World Psychiatry, 2020-01-10
• Psychotherapies for depression: a network meta-analysis covering efficacy, acceptability and long-term outcomes of all main treatment types — World Psychiatry, 2021-05-18