Therapy, Medication, or Both for Depression: Reading the Comparative Evidence

For most adults with major depression, the strongest comparisons find that psychotherapy alone and antidepressant medication alone perform about equally well on average, while combining the two tends to do better than either by itself. In the largest network meta-analysis on this question, published in World Psychiatry in 2020, combined treatment achieved response more often than psychotherapy or medication alone, and the two monotherapies showed no reliable gap between them. Those numbers describe groups, not individuals. The averages sit on top of wide person-to-person variation, the underlying trials carry real design limits, and the defensible reading of the evidence is that it supports offering more than one reasonable option rather than declaring a single winner.

How a head-to-head comparison gets built

The cleanest evidence starts with a randomized trial that assigns patients to one treatment or another and measures who improves. A 2013 meta-analysis in World Psychiatry pooled the direct head-to-head trials and found psychotherapy and pharmacotherapy broadly comparable for depression, with small differences that depended on the specific therapy and drug. Pooling many such trials increases precision, but few studies compare every option against every other option.

Network meta-analysis fills that gap. It combines direct comparisons (A versus B) with indirect ones (A versus C and B versus C) to rank treatments that were rarely or never tested against each other. That extra reach depends on an assumption called transitivity: the trials being linked have to be similar enough in patients, severity, and outcome definitions that the indirect math is fair. When that assumption is shaky, the rankings look more certain than they really are. Most of these analyses define response as a 50 percent reduction in symptom scores and remission as dropping below a clinical cutoff, so how a study defines success shapes what "better" means.

What effect sizes and number needed to treat mean

Trials rarely turn on a single dramatic number, so researchers summarize the gap between groups as an effect size. A standardized mean difference, often reported as Hedges' g, expresses that gap in units of variability: roughly 0.2 is small, 0.5 is moderate, and 0.8 is large. The 2014 World Psychiatry meta-analysis of adding psychotherapy to medication found a pooled g near 0.43, a small-to-moderate advantage for the combination over medication alone.

Number needed to treat translates that into something more concrete. It estimates how many people you would treat with the better option for one additional person to benefit who would not have benefited otherwise. The 2014 analysis put the number needed to treat around four, meaning that for roughly every four patients given combined treatment instead of medication alone, one extra person reached a good outcome. Lower numbers signal a larger practical effect. Relative measures tell a parallel story: in the 2020 network analysis, combined treatment raised the chance of response by about a quarter compared with either monotherapy.

Where combination tends to pull ahead

The 2020 network meta-analysis, drawing on more than one hundred trials and roughly twelve thousand patients, found combined treatment more effective than psychotherapy alone and than pharmacotherapy alone for achieving response, with similar patterns for remission. It also looked at acceptability, measured by how many people dropped out, and combined treatment and psychotherapy were better tolerated on that metric than medication alone. The combination generally shows its clearest advantage in moderate-to-severe, chronic, or recurrent depression, where the added structure and two mechanisms may reinforce each other. Major guidelines reflect this pattern. The American Psychological Association's 2019 depression guideline recommends offering either psychotherapy or a second-generation antidepressant as a starting point and supports pairing cognitive-behavioral or interpersonal therapy with medication when a combined approach is chosen.

The limits of saying one option beats another

Several features of this literature should slow anyone tempted to declare a decisive winner. Psychotherapy trials cannot be blinded the way a pill-versus-placebo trial can, and the comparison group matters enormously: measuring therapy against a waitlist tends to inflate its apparent effect compared with an active control. Researcher allegiance, where investigators favor the treatment they developed or prefer, has been shown to nudge results. Publication bias, small samples, and uneven risk of bias across trials all widen the true uncertainty around any tidy point estimate.

There is also the gap between the average patient and the person in front of a clinician. An effect size near 0.43 and a number needed to treat near four describe a meaningful but modest edge, not a guarantee, and they say nothing about which specific individual will respond to which approach. Side effects, cost, access, prior experience, and personal preference all belong in the decision, and the acceptability data suggest many people stay in treatment longer with a psychological component. This article is educational and not medical advice; treatment choices belong in a conversation with a qualified clinician who knows the individual case.

Read carefully, the comparative evidence is reassuring rather than deflating. It says that people facing depression usually have more than one route with genuine support behind it, that combining approaches is a reasonable and often stronger choice for more serious presentations, and that claims of one universal best treatment tend to outrun what the data can show.