Omega-3 Fatty Acid Biomarkers and Incident Atrial Fibrillation

October 18, 2025
| No Comments

Comment:

comment here

Summary:

Clinical Bottom Line

This large-scale, pooled analysis of observational studies provides moderate-quality evidence that higher levels of omega-3 fatty acids from habitual dietary intake are not associated with an increased risk of atrial fibrillation (AF) and may, in fact, be modestly protective. Specifically, while eicosapentaenoic acid (EPA) levels showed no association, higher levels of docosahexaenoic acid (DHA) and docosapentaenoic acid (DPA) were associated with a roughly 10% lower risk of developing AF. This finding suggests that current dietary guidelines recommending fish consumption for cardiovascular health are safe with respect to AF risk. Importantly, these results from long-term, diet-derived omega-3 levels stand in contrast to recent clinical trials where high-dose supplementation, particularly in high-risk patients, was linked to an increased risk of AF.

Results

Summary of Results

This analysis pooled data from 54,799 participants across 17 prospective cohort studies, identifying 7,720 new cases of AF over a median follow-up of 13.3 years. After adjusting for numerous potential confounding factors, the key findings were:

 

  • EPA: No association with incident AF. The hazard ratio (HR) was 1.00 (95% CI: 0.95-1.05). A hazard ratio of 1.00 means there was no difference in the rate of developing AF between individuals with higher versus lower EPA levels.

  • DPA: Associated with a 11% lower risk of incident AF (HR: 0.89; 95% CI: 0.83-0.95).

  • DHA: Associated with a 10% lower risk of incident AF (HR: 0.90; 95% CI: 0.85-0.96).

  • Total EPA + DHA: Associated with a 7% lower risk of incident AF (HR: 0.93; 95% CI: 0.87-0.99).

Assertive Critical Appraisal

Certainty of Evidence (GRADE Framework)

The overall certainty of this evidence is best characterized as low to moderate. The evidence is drawn from prospective cohort studies, which inherently carry a higher risk of bias than randomized controlled trials (RCTs), primarily due to the potential for unmeasured confounding. However, the study has several strengths that increase confidence in the findings: a very large sample size providing precise estimates, a consistent direction of effect for DHA and DPA across multiple cohorts, and a rigorous, pre-specified analytical plan that harmonized definitions and statistical models across all included studies.

Heterogeneity

The authors reported the I² statistic to quantify the percentage of variation across studies due to real differences rather than just chance.

  • For DPA, heterogeneity was very low (I²=0%), indicating the results were highly consistent across the different cohorts.

  • For EPA (I²=52.2%) , DHA (I²=47.5%) , and EPA+DHA (I²=60.7%), there was moderate heterogeneity. This suggests that while the overall trend points in one direction, the magnitude of the association varied between the included studies for reasons the authors could not fully identify through subgroup analyses. This warrants some caution in interpreting the precise magnitude of the pooled effect.

Publication Bias

The study’s design as an individual participant data meta-analysis from a research consortium (FORCE) is a major strength that significantly reduces the risk of publication bias. Rather than relying only on published results (which may be biased towards positive findings), the investigators obtained raw data from 17 different cohort studies and conducted new, standardized analyses. This approach is more likely to provide an unbiased estimate of the true association.

 

Risk of Bias in Included Studies

As this meta-analysis is based on observational studies, the principal risk of bias is confounding. The authors made a robust effort to mitigate this by adjusting for a wide range of pre-specified potential confounders in their statistical models, including demographics, lifestyle factors (smoking, alcohol, physical activity), BMI, and multiple prevalent health conditions like hypertension, diabetes, and heart disease. While it is impossible to completely rule out residual confounding from unmeasured factors, the extensive adjustments performed strengthen the validity of the findings.

 

Special Consideration for Pooled Results

It is important to note that while the overall pooled result for DHA and DPA is statistically significant, some of the individual included studies did not find a significant effect on their own (as seen in the forest plots in Figure 1). This scenario often occurs when individual studies are underpowered (too small) to detect a modest but real protective effect. By combining their data, the meta-analysis increases the overall statistical power, which can reveal a consistent trend that was not apparent in the smaller studies. This demonstrates the value of this pooled analysis in providing a more precise and powerful estimate of the association.

 

Reporting Quality Assessment (PRISMA)

This study does not include a PRISMA flow diagram, which is typically expected for a systematic review. However, this omission is justifiable because this is an individual participant data meta-analysis from a pre-existing consortium, not a review based on a systematic literature search. The authors are transparent about which 17 cohorts were included and why, fulfilling the key PRISMA principle of transparently reporting the study selection process.

Research Objective

To determine the prospective association between blood or adipose tissue biomarker levels of the omega-3 fatty acids EPA, DPA, and DHA and the risk of developing new-onset atrial fibrillation.

  • Population: 54,799 adults without prevalent AF from 17 prospective cohorts in North America, Europe, Asia, and Africa.

  • Exposure: Higher levels of circulating or tissue omega-3 fatty acids.

  • Comparator: Lower levels of the same omega-3 fatty acids.

  • Outcome: Incident (new-onset) atrial fibrillation ascertained via ECG, diagnostic codes, or medical record review.

Study Design

This was an individual participant-level data meta-analysis. Investigators from 17 different prospective cohort studies shared their original participant data. The coordinating center then conducted a de novo analysis on each dataset using a single, harmonized, and pre-specified statistical plan. The results (hazard ratios) from each cohort were then pooled together using an inverse-variance weighted meta-analysis to generate an overall summary estimate of the association.

 

Setting and Participants

The analysis included a total of 54,799 participants from 17 cohorts across 21 nations. The mean age of participants was 63 years, and about half were women. Over follow-up, 7,720 participants developed atrial fibrillation.

Bibliographic Data

  • Title: Omega-3 Fatty Acid Biomarkers and Incident Atrial Fibrillation

  • Authors: Qian F, Tintle N, Jensen PN, et al., on behalf of the Fatty Acids and Outcomes Research Consortium (FORCE).

  • Journal: Journal of the American College of Cardiology

  • Year: 2023

  • DOI: 10.1016/j.jacc.2023.05.024

This AI-generated analysis is for informational and research purposes only and is not a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of a qualified health provider with any questions you may have regarding a medical condition.

Original Article:

Full text pdf: Omega-3 Fatty Acid Biomarkers and Incident Atrial Fibrillation

CVD, Fish Oil-EFA