Dentate gyrus growth belongs to ECT, not to antidepressants in general

Design

Prospective multisite non-randomised study with serial structural MRI and linear mixed-effects models

Population

450 participants: 269 with major depressive disorder (CBT 106, pharmacotherapy 83, rTMS 39, ECT 41) and 181 healthy

Primary outcome

Hippocampal dentate gyrus volume

Effect

Significant overall increase driven by the ECT group; volume change correlated only with the core depressive symptom dimension

Four standard treatments for major depression were followed with serial MRI in 269 patients and 181 healthy controls: cognitive behavioural therapy, pharmacotherapy, repetitive transcranial magnetic stimulation and electroconvulsive therapy. Dentate gyrus volume was the pre-specified primary outcome, measured at baseline, at the end of a treatment course and six months later.

Volume rose across the patient group as a whole, but the rise was driven by the 41 patients who had ECT. Change in dentate gyrus volume correlated with improvement in the core depressive dimension alone — not anxiety, not somatic symptoms, not insomnia, and not any cognitive domain. Over six months the trajectory rose and then fell, in parallel with the core symptoms.

The neurogenesis story has been told for two decades as the common final pathway of antidepressant action. This says it is not shared: CBT, drugs and rTMS produced clinical improvement without it. That makes MRI-detectable dentate plasticity a candidate mechanism specific to ECT, and a reason to stop invoking it when explaining how an SSRI works.

• Do not use hippocampal neurogenesis as the explanation when counselling a patient about starting an antidepressant.
• The cognitive effects of ECT did not track dentate volume; treat memory complaints on their own terms.
• Volume gain was not permanent — it fell back over six months alongside symptoms.
• Treatment allocation was not randomised, so the ECT group was the most severely ill by design.
• Nothing here changes who should be offered ECT; it changes what is claimed about why it works.

Why it matters

It removes a mechanism that has been attributed to every antidepressant treatment and assigns it to one.

Don't overread it

Non-randomised allocation means the ECT effect cannot be cleanly separated from ECT patients' greater baseline severity.

The statistics, in plain English

This was prospective but not randomised, so the four treatment groups differed systematically in severity — the ECT group unavoidably the sickest, which is also where the largest volume change appeared. The correlation with the core depressive dimension and not with anxiety, somatic or cognitive measures is what makes the finding interesting: a non-specific effect of getting better would have moved all the dimensions together.

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.