High-Dose Omega-3 Therapy Raises Atrial Fibrillation Risk

A measurable risk at higher doses

An updated meta-analysis of 35 randomized controlled trials involving 114,592 participants found that high-dose omega-3 fatty acid treatment increased the odds of atrial fibrillation compared with control treatment. The pooled odds ratio was 1.43, and the absolute risk difference was 0.8 percentage points.

The relative estimate is clinically notable, but the absolute result provides essential context. A difference of 0.8 percentage points corresponds to approximately eight additional atrial fibrillation events per 1,000 treated people over the trial periods. If the pooled estimate represented a causal and consistent effect in a comparable population, that would translate to roughly one additional event for every 125 people treated. That number should not be generalized across doses, formulations, treatment durations, or baseline risk levels.

The signal concerns pharmacologic omega-3 treatment, particularly at higher doses. It should not be extrapolated directly to eating fish, usual dietary intake, or every over-the-counter supplement. Nor does it establish that all omega-3 formulations carry identical risk: the evidence base includes products with different amounts and combinations of eicosapentaenoic acid, or EPA, and docosahexaenoic acid, or DHA.

How the evidence was assembled

The investigators pooled randomized trials comparing omega-3 fatty acid treatment with placebo or another control and evaluated atrial fibrillation events. Randomization reduces confounding and makes a treatment-related safety signal more credible than one derived only from observational data. The large combined sample also improves the ability to detect an uncommon adverse outcome.

No single follow-up duration applies to the pooled estimate because the analysis combined trials with different treatment and observation periods. The available summary for this brief did not provide a confidence interval or p-value for the high-dose estimate; the reported odds ratio and absolute risk difference therefore should be read alongside the full meta-analysis when precise statistical uncertainty is needed.

The finding also fits signals from major cardiovascular outcome trials. In REDUCE-IT, icosapent ethyl, a purified EPA product, reduced major ischemic cardiovascular events among statin-treated patients with elevated triglycerides, but atrial fibrillation or flutter requiring hospitalization occurred more often with active treatment. In STRENGTH, a high-dose EPA/DHA formulation did not reduce major adverse cardiovascular events compared with corn oil and was associated with more new-onset atrial fibrillation.

Those trials illustrate why atrial fibrillation risk cannot be considered apart from efficacy. A therapy that produces a demonstrated cardiovascular benefit may still have a favorable overall balance for appropriately selected patients despite a small arrhythmia risk. When cardiovascular benefit has not been established for a particular product or indication, the same safety signal weighs more heavily.

Applying the signal in cardiovascular care

The evidence supports reviewing why a patient is receiving omega-3 therapy, which formulation is being used, and whether its expected benefit is supported by clinical outcomes data. Prescription products studied for cardiovascular risk reduction are not interchangeable with mixed EPA/DHA supplements marketed for general wellness. Differences in composition, dose, purity, regulatory oversight, and trial evidence matter.

Baseline atrial fibrillation susceptibility also affects the absolute impact. A person with prior atrial fibrillation, atrial enlargement, heart failure, advanced age, hypertension, obesity, sleep apnea, or other arrhythmia risk factors may face a larger absolute increase than someone at low baseline risk, even if the relative treatment effect is similar. The meta-analysis does not provide an individualized risk calculator, however, and it cannot determine a precise threshold at which treatment becomes unfavorable.

For patients already receiving high-dose therapy, the finding supports medication reconciliation and a benefit-risk discussion rather than automatic discontinuation. Clinicians can confirm the indication, distinguish prescription therapy from nonprescription supplements, and ask about palpitations, irregular pulse, worsening exercise tolerance, dyspnea, dizziness, or unexplained fatigue. Evaluation of symptoms should follow usual clinical pathways; the meta-analysis did not test a special rhythm-monitoring strategy.

Policy and formulary decisions should likewise separate evidence-backed prescription indications from broad omega-3 use. The 0.8-percentage-point difference may appear small at the individual level, but it could produce a meaningful number of events when high-dose products are used across large populations. Coverage decisions should consider both demonstrated cardiovascular outcomes and arrhythmia harms for the specific formulation being reviewed.

Important uncertainties

The analysis inherits differences among its component trials. Formulations, doses, comparators, participant risk, follow-up, and methods for detecting atrial fibrillation were not uniform. Some trials may have captured only clinically apparent or hospitalized events, while others used more active surveillance. Unequal detection can influence pooled event rates.

Atrial fibrillation may also have been a secondary or safety outcome rather than a prespecified primary endpoint in individual trials. That can produce less consistent definitions and reporting. Trial-level pooling cannot fully identify which patients are most susceptible, and an overall or high-dose estimate should not be assumed to apply equally to EPA-only therapy, combined EPA/DHA therapy, or unregulated supplements.

Randomization supports a possible causal treatment effect, but the meta-analysis does not prove a single biological mechanism. It also does not show that the arrhythmia risk necessarily outweighs cardiovascular benefit. That judgment remains formulation- and indication-specific, with REDUCE-IT and STRENGTH demonstrating that products grouped under the omega-3 label can have materially different efficacy profiles.

Questions clinicians ask

Should high-dose omega-3 therapy be stopped because of this finding?

The evidence does not support automatic discontinuation. It supports reassessing the indication, formulation, demonstrated cardiovascular benefit, and individual atrial fibrillation risk. A modest absolute harm may be acceptable when a specific therapy has proven outcome benefits, but harder to justify when treatment is being used without a well-supported indication.

Does the result apply to over-the-counter fish oil?

Not directly. The meta-analysis concerns trial-tested omega-3 regimens, while nonprescription products vary in EPA/DHA content, purity, dose, and quality. The finding does reinforce that concentrated omega-3 products should not be assumed harmless, but it cannot assign a precise atrial fibrillation risk to every supplement.

How should the 0.8% absolute increase be explained?

It represents about eight additional atrial fibrillation events per 1,000 treated participants across the trial settings. That pooled average is more useful for counseling than the relative estimate alone, but an individual patient’s absolute risk may be higher or lower depending on baseline susceptibility and treatment duration.

Who may warrant closer review before or during treatment?

People with prior atrial fibrillation or substantial arrhythmia risk deserve particular attention, although the meta-analysis does not establish a dedicated monitoring protocol. Reviewing symptoms, medication purpose, formulation, and expected benefit is reasonable; rhythm testing should be guided by clinical presentation and established cardiovascular practice.

References

1. Effects of Omega-3 Fatty Acid Treatment on Risk for Atrial Fibrillation: An Updated Meta-Analysis of 35 Trials including 114 592 Individuals — PubMed, 2026 2. Cardiovascular Risk Reduction with Icosapent Ethyl for Hypertriglyceridemia — The New England Journal of Medicine, 2019 3. Effect of High-Dose Omega-3 Fatty Acids vs Corn Oil on Major Adverse Cardiovascular Events in Patients at High Cardiovascular Risk: The STRENGTH Randomized Clinical Trial — JAMA, 2020