High-grade evidence backs long-term psychodynamic therapy for depression

Design

Pre-registered systematic review and meta-analysis of randomised trials, GRADE-rated

Population

16 trials, 1,992 participants receiving long-term psychodynamic psychotherapy for depression

Primary outcome

Reduction in depressive symptoms at end of treatment

Effect

Standardised mean difference 0.34 (0.20 to 0.47) vs all comparators, high-certainty; 0.27 vs active treatments, moderate-certainty

A pre-registered GRADE meta-analysis pooled 16 randomised trials and 1,992 participants of long-term psychodynamic psychotherapy — defined as at least nine months and 30 sessions — for depression. At the end of treatment it reduced depressive symptoms against all comparators (standardised mean difference 0.34, 95% CI 0.20 to 0.47), with effects generally sustained at follow-up.

The certainty is the notable part. GRADE rated the all-comparators result as high-certainty, and the comparison against active treatments as moderate-certainty (standardised mean difference 0.27). The larger effect seen against active controls specifically was very-low certainty, and risk of bias was a concern in most trials, so the firm claim is the modest one.

For practice, this supports offering or referring for long-term psychodynamic therapy as a real option in suitable patients with depression, particularly where shorter or pharmacological approaches have not held. It does not displace first-line treatments, but it gives the longer psychotherapy a firmer evidence base than it has usually been granted.

• Sixteen randomised trials, 1,992 participants; therapy of at least nine months and 30 sessions.
• Reduced depressive symptoms versus all comparators: standardised mean difference 0.34 (95% CI 0.20 to 0.47), high-certainty.
• Versus active treatments: standardised mean difference 0.27 (0.1 to 0.45), moderate-certainty.
• Effects were generally sustained at follow-up.
• Consider it for suitable patients, especially where shorter or drug approaches have not held.

Why it matters

Long psychodynamic therapy is often treated as poorly evidenced; a high-certainty synthesis gives it a firmer footing in the treatment conversation.

Don't overread it

The high-certainty finding is for a modest effect against all comparators; the larger advantage over active treatments was very-low certainty.

The statistics, in plain English

A standardised mean difference of about 0.3 is a small-to-moderate benefit; high-certainty means further research is unlikely to overturn it, while the larger effect against active controls is very-low certainty and should not be relied on.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.