Introduction: Many studies have shown an inverse association of omega-3 polyunsaturated fatty acid (PUFA) from fish intake and supplementation with coronary heart disease (CHD) mortality. However, findings on the relationship of omega-3 PUFA with nonfatal CHD and stroke are less consistent. Higher doses or prolonged intake may be needed to affect non-fatal cardiovascular events. We hypothesized that among Veterans, greater baseline fish intake is associated with a lower risk of nonfatal CHD and stroke, particularly among those who regularly use omega-3 supplements.
Methods: The Million Veteran Program (MVP) is an ongoing nation-wide longitudinal cohort study of U.S. Veterans with self-reported survey, biospecimen, and electronic health record data. Participants who were free of CHD or stroke at the time of the baseline food frequency questionnaire were included in this analysis. Separate Cox proportional hazard models were used to assess the associations of fish intake with non-fatal incident CHD and stroke, defined from electronic health records using validated algorithms. Fish intake (serving=3-5 oz.) within the past year was categorized as <1 serving/month, 1-3 servings/month, 1 serving/week, 2-4 servings/week, 5-6 servings/week, and 1+ serving/day. We assessed whether omega-3 fatty acid supplementation (yes/no) modifies these relationships by testing its interaction with fish intake. Multivariable models included demographics and known CHD and stroke risk factors: age, sex, race, smoking status, alcohol use, physical activity, education, overall dietary pattern, BMI, family history of CHD or stroke, and baseline diabetes, hypertension, and use of statins and fibrates.
Results: Among 508,699 MVP participants to date, 210,180 (mean age 66±12 years, 92.1% male) were free of CHD or stroke and had data from the food frequency questionnaire at baseline. Median fish intake was 1 serving/week and 21.7% (45,552 out of 210,180) regularly took omega-3 supplements. Over a median follow-up of 2.9 years, there were 5,991 and 4,244 incident cases of non-fatal CHD and stroke, respectively. Using <1 fish serving/month as the reference, multivariable adjusted HR (95% CI) for CHD were 0.99 (0.91-1.07) for 1-3 servings/month, 1.06 (0.98-1.15) for 1 serving/week, 1.06 (0.96-1.16) for 2-4 servings/week, 1.26 (1.03-1.53) for 5-6 servings/week, and 0.96 (0.68-1.34) for 1+ serving/day, (p-trend=0.08); corresponding values for stroke were 0.95 (0.87-1.05) for 1-3 servings/month, 0.96 (0.87-1.06) for 1 serving/week, 0.99 (0.88-1.11) for 2-4 servings/week, 1.05 (0.83-1.34) for 5-6 servings/week, and 1.23 (0.87-1.74) for 1+ serving/day, (p-trend=0.60). Intake of omega-3 fatty acid supplements did not modify the fish-CHD/stoke relations (p-interaction=0.32 and 0.82, respectively).
Conclusions: We observed no consistent association between fish intake and non-fatal CHD or stroke.