Chung M, Ethan M Balk, Thomas A Trikalinos, Adams Gp, Christopher W Halladay, Lin L, Stewart A. Robertson, Lichtenstein Ah, Lau J, Dale W. Steele, Smith Bt, Langberg, and Yip A
Background The effect and association of omega-3 fatty acids (n-3 FA) intake and biomarker levels with cardiovascular (CV) clinical and intermediate outcomes remains controversial. We update prior Evidence Reports of n-3 FA and clinical and intermediate CV disease (CVD) outcomes. Objectives Evaluate the effect of n-3 FA on clinical and selected intermediate CV outcomes and the association of n-3 FA intake and biomarkers with CV outcomes. The n-3 FA under review include eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), docosapentaenoic acid (DPA), stearidonic acid (SDA), and alphalinolenic acid (ALA). Data sources MEDLINE®, Embase®, the Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and CAB Abstracts from 2000 or 2002 to June 8, 2015, and eligible studies from the original reports and relevant existing systematic reviews. Review methods We included randomized controlled trials (RCTs) of any n-3 FA intake compared to no, lower, or other n-3 FA intake with an outcome of interest conducted in healthy adults, those at risk for CVD, or those with CVD. We also included prospective observational studies of the association between baseline n-3 FA intake or biomarker level and followup outcomes. We required 1 year or more of followup for clinical outcomes and 4 weeks for intermediate outcomes (blood pressure [BP] and lipids). Results From 11,440 citations (from electronic literature searches and existing systematic reviews), 829 abstracts met basic eligibility criteria; 61 RCTs and 37 longitudinal observational studies (in 147 articles) were included. Most RCTs and observational studies had few risk-of-bias concerns. Total n-3 FA: There is low strength of evidence (SoE) of no association between total n-3 FA intake and stroke death or myocardial infarction. There is insufficient evidence for other outcomes.Marine oils, total: There is moderate to high SoE that higher marine oil intake lowers triglycerides (Tg), raises high density lipoprotein cholesterol (HDL-c), and lowers the ratio of total cholesterol to HDL-c but raises low density lipoprotein cholesterol (LDL-c); also that higher marine oil intake does not affect major adverse CV events, all-cause death, total stroke, sudden cardiac death, coronary revascularization, atrial fibrillation, or BP. There is low SoE of associations between higher marine oil intake and decreased risk of CVD death, coronary heart disease (CHD), myocardial infarction, ischemic stroke, and congestive heart failure (CHF). There is low SoE of no association with CHD death or hemorrhagic stroke. There is insufficient evidence for other outcomes.Marine oil FA individually: There is low SoE of no associations between EPA or DHA intake (separately) and CHD, and between EPA or DPA and atrial fibrillation. There is low SoE of no association between EPA biomarkers and atrial fibrillation, but moderate SoE of no effect of purified DHA supplementation on BP or LDL-c. There is insufficient evidence for other specific marine oil FA and outcomes.ALA: There is moderate SoE of no effect of ALA intake on BP, LDL-c, HDL-c, or Tg. There is low SoE of no association between ALA intake or biomarker level and CHD, CHD death, atrial fibrillation, and CHF. There is insufficient evidence for other outcomes.Other n-3 FA analyses: There is insufficient evidence comparing n-3 FA with each other or for SDA.Subgroup analyses: Nineteen of 22 studies found no interaction of sex on any effect of n-3 FA. Likewise, 19 of 20 studies found no differential effect by statin co-use. Within 16 studies evaluating diabetes subgroups, 2 found statistically significant beneficial effects of n-3 FA in those with diabetes but not in those without diabetes, but no test of interaction was reported. Conclusions The 61 RCTs mostly compared marine oil supplements with placebo on CVD outcomes in populations at risk for CVD or with CVD, while the 37 observational studies mostly examined associations between various individual n-3 FA and long-term CVD events in generally healthy populations. Compared with the prior report on n-3 FA and CVD, there is more robust RCT evidence on ALA and on clinical CV outcomes; also, by design there are newly added data on associations between n-3 FA biomarkers and CV outcomes. However, conclusions regarding the effect of n-3 FA intake on CV outcomes or associations with outcomes remain substantially unchanged. Marine oils statistically significantly raise HDL-c and LDL-c by similar amounts (≤2 mg/dL), while lowering Tg in a dose-dependent manner, particularly in individuals with elevated Tg; they have no significant effect on BP. ALA has no significant effect on intermediate outcomes. Limited data were available from RCTs on the effect of n-3 FA on clinical CVD outcomes. Observational studies suggest that higher marine oil intake (including from dietary fish) is associated with lower risk of several CVD outcomes. No clear differences in effects or associations were evident based on population, demographic features, or cointerventions. Future RCTs would be needed to establish adequate evidence of the effect of n-3 FA on CVD outcomes or to clarify differential effects in different groups of people. However, future trials are unlikely to alter conclusions about the effects of n-3 FA supplementation on intermediate cardiovascular outcomes (BP, LDL-c, HDL-c, or Tg).