- Design
- Systematic review and meta-analysis of cohort and case-control studies
- Population
- 12 studies, about 9.6 million people
- Primary outcome
- Ventricular arrhythmia or sudden cardiac death; atrial fibrillation
- Effect
- VA/SCD OR 1.34 (1.15 to 1.55), I² 90.5%; AF OR 1.75 (0.84 to 3.62)
This meta-analysis pooled 12 cohort and case-control studies (about 9.6 million people). Antidepressant use was associated with ventricular arrhythmia or sudden cardiac death (9 studies; OR 1.34, 95% CI 1.15 to 1.55), for both SSRIs and tricyclics. Heterogeneity was extreme (I² = 90.5%).
For atrial fibrillation there was no overall association (OR 1.75, 0.84 to 3.62; I² = 98.1%). Removing one dominant study left two small studies with OR 1.27 (1.07 to 1.50), which is too thin to rely on.
The authors themselves flag the central problem: depression is associated with cardiac death independently, so these studies cannot separate the drug from the illness. Read alongside the randomised trial data in today's practice-changer, the sensible conclusion is to weigh cardiac risk factors in drug choice rather than to withhold antidepressants.
- Do not withhold an indicated antidepressant because of this association alone
- Take a cardiac history — syncope, palpitations, family history of sudden death — before starting
- Be most careful with tricyclics in people with known heart disease
- Remember untreated depression also carries cardiac risk
Why it matters
It keeps cardiac history-taking part of every antidepressant start.
Don't overread it
An observational association with 90% heterogeneity is not evidence that antidepressants cause arrhythmia.
The statistics, in plain English
I² above 90% means the studies disagreed far more than chance would explain, so a single pooled number hides very different results. Confounding by indication means people prescribed the drug differ from those who are not — here, by having depression.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for psychiatry, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free