Abstract 43: Circulating Omega-3 Fatty Acid Levels and Total and Cause-specific Mortality: Prospective Evidence From 14 Cohorts in the Fatty Acids and Outcomes Research Consortium

Abstract

Background: Several recent studies have linked higher blood levels of the long-chain omega-3 polyunsaturated fatty acids (PUFAs) with lower risk of certain disease outcomes, in particular cardiovascular disease (CVD). Limited studies have evaluated associations between n-3 PUFA levels and total and cause-specific mortality. Methods: We meta-analyzed the association of circulating omega-3 PUFAs with total and cause-specific mortality from 15 distinct cohorts in 10 countries following a total of 36,840 individuals (baseline average age 66 yrs; 47% female; median follow-up 13.7 yrs), with 11,635 deaths. Total mortality, and 3 cause-specific mortalities (CVD, cancer or other), were separately predicted by levels of the long chain omega-3 PUFAs eicosapentaenoic acid (EPA), docosapentaenoic acid (DPA), docosahexaenoic acid (DHA) and the Omega-3 index (EPA+DHA), as well as levels of alpha-linolenic acid (ALA, an intermediate chain n-3 PUFA). Using cohort-specific fatty acid levels (range from 10 th to 90 th percentiles), each cohort fit a Cox regression model adjusting for demographics, health related covariates, and n-6 PUFA levels. Cohort-specific estimates were pooled with inverse-variance weighted meta-analysis. Results: Higher levels of EPA, DPA, DHA and the Omega-3 index were all associated with lower all-cause mortality [EPA meta-analysis Hazard Ratio (HR)=0.90 (95% CI: 0.85, 0.95), DPA HR=0.84 (95% CI: 0.77, 0.91), DHA HR=0.84 (95% CI: 0.75, 0.93), Omega-3 index HR = 0.84 (95% CI 0.77, 0.92) (Figure)], as well as for each cause-specific mortality (CVD, cancer and other; HRs from 0.85 to 0.92). ALA showed no association with total-mortality or cause-specific mortality [HRs from 0.97 to 1.00)]. Conclusions: We have found strong evidence of inverse associations between long-chain omega-3 PUFA levels and all-cause and cause-specific mortality. Further controlled studies are needed in order to determine the extent to which these observations represent a causal or coincidental relationships.