Back

Optimal Pair Matching Combined with Machine Learning Predicts that Omega-3 Fatty Acid Supplementation Markedly Reduces the Risk of Myocardial Infarction in African Americans

Sun, S.; Hara, A.; Johnstone, L.; Hallmark, B.; Lu, E.; Watkins, J. C.; Thomson, C. A.; Schembre, S. M.; Sergeant, S.; Umans, J. G.; Yao, G.; Zhang, H. H.; Chilton, F. H.

2024-03-28 cardiovascular medicine
10.1101/2024.03.26.24304910 medRxiv
Show abstract

Conflicting results from clinical trials have contributed to a lack of consensus about cardioprotective effects of omega-3 (n-3) highly unsaturated fatty acids (HUFA). Although the VITAL trial did not demonstrate an overall benefit of n-3 HUFA supplementation on composite cardiovascular disease (CVD) and cancer outcomes, the substantial enrollment of African Americans (AfAm) afforded a unique opportunity for a post-hoc analysis of racial differences in the supplementation response. We employed propensity score matching to address potential covariate imbalances between AfAm and European American (EuAm) participants in VITAL (N=3,766 participants). Using Kaplan-Meier curves and two machine learning methodologies, we found that n-3 HUFA supplementation was significantly associated with a reduced risk of myocardial infarction (MI) exclusively in the AfAm subgroup, as evidenced by an odds ratio of 0.17 (95% CI [0.048, 0.59]). These findings indicate a potential cardioprotective benefit of n-3 supplementation in AfAm, specifically in reducing MI risks; a risk not identified in EuAm. Further investigation of n-3 HUFA effects through a hypothesis-driven randomized clinical trial among AfAm is needed to test a race-specific response that may inform recommendations for n-3 HUFA supplementation.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.