Reviews/Omega-3 Fatty Acids/Omega-3 raises atrial fibrillation risk
CV · OMEGA-CV-03 · v2
Several randomized trials of omega-3 supplements point the same way: people taking omega-3 tended to develop atrial fibrillation (an irregular heartbeat) slightly more often than those on placebo. In one post-heart-attack analysis [3] the increase was clear and statistically convincing, and a large elderly post-heart-attack trial [6] showed the same direction just short of the usual significance cutoff. But the biggest and best-designed trial [2], run in healthy older adults, found only a small difference that could have been chance. The signal also appears strongest at higher doses rather than across every regimen. The panel therefore treated the increased-risk claim as well supported and repeated across studies, while noting the flagship trial did not confirm it and the effect may depend on dose.
Every chronic-supplementation estimate in the corpus sits above 1.0: the large primary-prevention trial [2] reported incident AF in 469 (3.7%) vs 431 (3.4%) participants, hazard ratio 1.09 (95% CI 0.96-1.24; P=.19); the OMEMI main trial [6] reported 28 (7.2%) vs 15 (4.0%), HR 1.84 (0.98-3.45; P=0.06); and the OMEMI secondary analysis [3] reported 46 (11.9%) vs 24 (6.5%) with AF or 'micro-AF', HR 1.90 (1.16-3.11; P=0.011). [5], a within-arm STRENGTH sub-analysis, states that 'fish oils are associated with higher risk of incident atrial fibrillation (AF) than placebo' and found 126 (2.1%) of 6,000 ω-3 recipients developed AF. The direction is consistent and replicated, but significance is reached only in [3]. [7], a perioperative trial, found no effect (RR 0.98; p=0.922). The contested point: The division is whether a replicated directional signal, significant in only one post-hoc analysis [3] and merely trending in the main OMEMI trial [6] at P=0.06, outweighs a null result in the largest, best-powered primary-prevention trial [2] (HR 1.09; P=.19). A second fault line is dose: the clearest signals come from 1.8 g/d regimens [3][6] while the unqualified claim covers supplementation generally.
All 7 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Omega-3 Fatty Acids, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The largest and best-powered trial [2] (N=25,119, primary prevention) was non-significant (HR 1.09, P=.19); the strongest evidence against overreach is also the strongest evidence of a null.
The only significant result [3] is a post-hoc secondary analysis in a narrow elderly post-MI subgroup, combining AF with a novel 'micro-AF' endpoint that inflates event counts.
Where it remains uncertain. The largest primary-prevention trial [2] was non-significant (HR 1.09; P=.19), and the one significant result [3] rests on a narrow post-MI subgroup with a 'micro-AF' endpoint.
Aleatory
The significant and trending estimates rest on small counts: 70 AF/'micro-AF' events total in [3] and 43 in [6], giving wide confidence intervals (1.16-3.11 and 0.98-3.45) sensitive to chance.
Epistemic
The corpus pairs an 840 mg/d regimen [2] that was non-significant with 1.8 g/d regimens [3][6] showing the clearest signal, but no study directly compares doses, so the threshold at which risk rises is unknown.
Inferential
[3]'s significant result combines clinically detected AF with a novel 'micro-AF' endpoint (27 of 70 events) in a post-MI subgroup, so the inference depends on an endpoint not used elsewhere in the corpus.
The load-bearing papers from a corpus of 7, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Largest, best-powered trial; positive point estimate but non-significant, tempering the overall signal.
“For the EPA-DHA vs placebo comparison, incident AF events occurred in 469 (3.7%) vs 431 (3.4%) participants, respectively (hazard ratio, 1.09; 95% CI, 0.96-1.24; P = .19).”
Panel auditInternal validity: high · scope of inference: Adults aged 50+ without prior cardiovascular disease, cancer, or AF, taking 460 mg EPA / 380 mg DHA daily for primary prevention.Key limitations: relatively low EPA-DHA dose · primary-prevention population only · AF not the design's primary rhythm focus
Only statistically significant result; shows increased AF risk with supplementation in post-MI patients, though inflated by a novel 'micro-AF' endpoint.
“In the n-3 PUFA group 46 (11.9%) had AF/'micro-AF' (28 AF, 18 'micro-AF') and in the placebo group 24 (6.5%) had AF/micro-AF (15 AF, 9 micro-AF); HR 1.90 (95%CI 1.16-3.11), P = 0.011.”
Panel auditInternal validity: medium · scope of inference: Elderly (70-82 y) post-myocardial-infarction patients without prior AF, on 1.8 g/day EPA/DHA for two years.Key limitations: secondary/subgroup analysis · composite outcome includes 'micro-AF' · elderly post-MI population only · high dose
Corroborates a cross-trial AF signal in high-CV-risk patients receiving ω-3, though provides no internal placebo comparison.
“Several randomized trials have demonstrated that fish oils are associated with higher risk of incident atrial fibrillation (AF) than placebo”
Panel auditInternal validity: low · scope of inference: Patients at high cardiovascular risk receiving ω-3 carboxylic acid; identifies who among them develops AF, not whether omega-3 raises AF risk versus placebo.Key limitations: no comparator arm · risk-score development not a risk-comparison · internal validation only · conference abstract
3 of 7 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
A consistent, replicated directional signal with one significant RCT analysis supports the claim at Likely, held below Established by the non-significant flagship trial and apparent dose-dependence of the unqualified statement.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Dose-response for AF risk. The clearest signals come from 1.8 g/d [3][6] while the lower-dose trial [2] was non-significant, so the dose at which risk emerges is undefined and the unqualified claim cannot be scoped. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A randomized trial comparing multiple omega-3 doses against placebo with incident AF as a prespecified primary endpoint.
Open question: At what dose does chronic omega-3 supplementation begin to raise atrial fibrillation risk? Why it is still open: The non-significant trial [2] used 840 mg/d while the strongest signals came from 1.8 g/d regimens [3][6], and no study directly compares doses.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19