High-dose fish oil AF risk limited to high-risk patients

Omega-3 fatty acids are among the most widely taken supplements in the world, and for years the conversation was about cardiovascular benefit. More recently, meta-analyses of randomised trials raised the opposite concern: that treatment might increase the risk of atrial fibrillation, an irregular heart rhythm associated with stroke and heart failure. Those earlier analyses, however, pooled at most eight trials, which is a thin base for a safety signal that could apply to millions of users. Whether risk depends on dose, on who is taking it, or on both, was the unresolved question.

The authors assembled a much larger evidence base, including both published and previously unpublished data. Eligible studies were randomised controlled trials of at least 500 mg/d of combined docosahexaenoic acid and eicosapentaenoic acid, with at least 12 months of treatment, participants aged 50 years or older, and where possible no known atrial fibrillation or atrial flutter at baseline. The primary outcome was new-onset atrial fibrillation. Their pre-stated hypothesis was that risk would depend jointly on dose, split above or below 1500 mg/d, and on background cardiovascular disease risk, with the two acting synergistically. In total, 35 trials contributing 37 data sets and 114 592 participants were pooled.

The signal was confined to one of the four dose-by-risk groups. Among participants at high cardiovascular risk treated with more than 1500 mg/d of EPA/DHA, atrial fibrillation risk was significantly increased, with a pooled odds ratio of 1.43 (95% CI 1.14-1.79) and an absolute risk difference of 0.8% (0.40%-1.1%). None of the other three combinations reached statistical significance: odds ratios were 1.07 for high risk with low dose, 1.06 for low risk with low dose, and 1.03 for low risk with high dose. In other words, neither high dose alone nor high baseline risk alone was enough to move the outcome.

This substantially extends the earlier meta-analyses that flagged omega-3 and atrial fibrillation, and reframes them: rather than a blanket hazard of fish-oil treatment, the pooled data point to an interaction between prescription-level doses and pre-existing cardiovascular risk. It also puts the risk in absolute terms, which the odds ratio alone obscures. Open questions remain, notably whether the finding holds in prospective trials designed to detect rhythm outcomes, how it balances against the cardiovascular and triglyceride benefits that motivate high-dose use, and whether detection of atrial fibrillation differed between trials.

The honest reading is that dose and context matter more than the supplement category. For adults taking modest amounts of EPA and DHA, this pooled analysis did not find a significant increase in atrial fibrillation, including among those at higher cardiovascular risk. The concern attaches to the high-dose end, above 1500 mg/d, in people who already have elevated cardiovascular risk, and even there the absolute difference was under one percentage point. High-dose omega-3 is frequently prescribed rather than self-selected, so anyone on such a regimen has a clinician to raise the trade-off with.