n–3 Fatty acids and cardiovascular disease: navigating toward recommendations

¹ From the Institute of Human Nutrition (RJD and SRA) and the Department of Pediatrics (RJD), Columbia University, New York, NY

See corresponding article on page 5.

² Reprints not available. Address correspondence to RJ Deckelbaum, Institute of Human Nutrition, Columbia University, 630 West 168th Street, New York, NY 10032. E-mail: rjd20{at}columbia.edu.

Many studies in the scientific literature have shown a correlation between n–3 fatty acids (FAs) and cardiovascular disease (CVD), but translating the findings of these studies into intake recommendations for individuals and populations requires careful navigation. Two n–3 FAs, eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), are increasingly being recognized as important modulators of multiple biological pathways that affect health and disease (1-7). The essential n–3 FA precursor of EPA and DHA, -linolenic acid (ALA), is generally far less effective at inducing biological effects, in part because of its inefficient conversion to EPA and DHA in humans (8). Thus, maximal effects of n–3 FAs depend on the delivery of EPA and DHA rather than of ALA. EPA and DHA affect many pathways that could benefit CVD outcomes, including actions that lower blood concentrations of triacylglycerols, prevent thrombosis and cardiac arrhythmias, inhibit adverse inflammatory responses, and decrease platelet aggregation (7). EPA and DHA also reduce vascular smooth muscle hypertrophy, reduce endothelial dysfunction, and increase vascular reactivity and decrease cholesterol accumulation in the arterial wall (7). Surgically removed specimens from patients provided diets rich in EPA and DHA before undergoing carotid endarterectomy showed lower concentrations of inflammatory markers and other variables associated with plaque instability than did those from patients fed control diets (9).

Despite a substantial number of plausible pathways whereby n–3 FAs may reduce morbidity and mortality from CVD, a recent working group convened by the National Institutes of Health (Bethesda, MD) concluded " ... the body of evidence is consistent with the hypothesis that intake of omega-3 FA reduces CVD but that a definitive trial is needed" (10). In this issue of the Journal, Wang et al (11) take a large step forward in helping to resolve controversies related to the beneficial effects of n–3 FAs on CVD outcomes. The authors correctly propose that a meta-analysis of this topic is not possible because of the heterogeneity in "study designs, background diets, endpoint definitions, and baseline fish or n–3 FA intakes." Studies of the effects of n–3 FAs vary tremendously in terms of whether they provide n–3 FAs as ALA, as EPA combined with DHA (in various ratios), as EPA or DHA alone, as part of a triacylglycerol molecule, as ethyl esters, through the diet, or as supplements. Only a few trials have lowered intakes of n–6 FAs concomitant with increased intakes of n–3 FAs. Nevertheless, the authors have written a formidable review of this "basket" of n–3 FA trials and have been able to classify them into different groups according to how n–3 FA intake is linked to specific endpoints relating to CVD and also by differentiating between secondary- and primary-prevention studies. Because Wang et al measured the effect of n–3 FAs on actual CVD endpoints, they included only randomized controlled trials and prospective cohort studies that monitored patients for 1 y, and, in the case of the case-control studies, only those that actually documented intakes of fish oil and n–3 FAs. Of 842 articles reviewed by Wang et al, only 46 met their strict criteria. They reviewed many more studies for potential adverse effects of n–3 FAs; of particular importance is that no or very few complications were documented. In total, the evidence indicates that increased consumption of the n–3 FAs EPA and DHA, either through fish or supplements or both, reduced the rates of all-cause mortality, myocardial infarction, and sudden cardiac death. Evidence was strongest in secondary-prevention trials but was also present in primary-prevention studies. Similar to other biological outcomes, which indicate that EPA and DHA have positive effects on biological outcomes, evidence reported in the review by Wang et al indicates that ALA has no or only weak beneficial effects on diminishing the risk of CVD. However, because of the lack of available studies, the review could not address many important questions.

1. Will the effects of long-chain n–3 FAs be enhanced by a concomitant reduction in n–6 FAs? A recent analysis by Hibbeln et al (12) strongly supports this possibility. Although some attention has been focused on the ratios of n–3 to n–6 FAs, we suggest that the total amount of each type of FA also demands consideration.
   
2. What are the dose-response effects of EPA and DHA, and are tissue concentrations of these n–3 FAs critical to achieving biological effects? Which tissues need to be measured as the best surrogates for CVD outcomes? In most reported studies, plasma or tissue concentrations of n–3 FAs are not measured, and the use of these as surrogates has not necessarily been validated.
   
3. What are the basic molecular mechanisms underlying the wide array of effects of these potent CVD modulators? A recent study suggests that a metabolite of EPA is responsible for some biological effects (13). Through the identification of active n–3 FA metabolites and their downstream pathways, underlying mechanisms will be better elucidated.
   
4. Is DHA equally as effective as or more effective than EPA in reducing CVD endpoints, or is a combination of both required? A recent review by Mori and Woodman (14) suggests that, for many endpoints relating to CVD, DHA is equally as effective as or more so than EPA. Although improved CVD outcomes have been achieved by intakes of a combination of these 2 n–3 FAs in various ratios, "optimal" ratios are still to be determined.
   
5. Does an increase in the ratio of EPA to DHA reduce CVD equally in populations with high baseline intakes of EPA and DHA and in populations with low intakes? The analysis by Wang et al suggests that intervention trials show less CVD protection in some studies from Norway, Finland, and Japan—countries with high fish intakes (11).
   

Identification of the molecular pathways via which n–3 FAs act on specific CVD-related endpoints will help in planning future intervention studies. Data are still needed to determine critical periods in the life span when persons need to obtain adequate intakes of n–3 FAs for optimal effects. The available data are insufficient from which to make recommendations about the intake of one specific n–3 FA over another, but accumulating data reaffirm that the EPA and DHA precursor ALA is inefficient and much less efficacious than are EPA and DHA. Clearly, more evidence needs to be gathered regarding the primary prevention of CVD. However, even studies that have not shown a reduction in CVD risk have shown a reduction in all-cause mortality (11).

Note, however, that a recent meta-analysis of the effects of n–3 FAs concluded that "long chain and shorter chain omega-3 fats do not have a clear effect on total mortality, combined cardiovascular events, or cancer" (15). However, this analysis failed to account for many of the pitfalls articulated in this editorial and in the article by Wang et al (11). For example, studies of interventions with ALA were not separated from studies of EPA, DHA, or EPA+DHA from fish or from supplements. Also, the duration of many of the studies reviewed in the meta-analysis was 11 mo, and primary prevention studies were not separated from secondary-prevention studies.

With all the limitations in mind, and given the little to no risk associated with their consumption, the American Heart Association and several international health agencies recommend intakes of 1 g EPA+DHA/d for patients with known CVD and of 4–500 mg EPA+DHA/d (2 servings of oily fish/wk) for those without CVD (16). Despite concern about toxins in fish, proper selection and preparation of fish results in a low risk from toxins (17, 18), especially when compared with low intakes of EPA and DHA (18). What should we recommend today? We believe that the body of existing evidence is strong enough to suggest that in the United States, certainly, and in other countries where n–3 FA consumption is low, public health initiatives are needed to increase intakes of EPA and DHA.

ACKNOWLEDGMENTS

RJD and SRA contributed equally to this editorial, and neither author had a conflict of interest regarding n–3 fatty acids.

REFERENCES

Related articles in AJCN:

n–3 Fatty acids from fish or fish-oil supplements, but not -linolenic acid, benefit cardiovascular disease outcomes in primary- and secondary-prevention studies: a systematic review

Chenchen Wang, William S Harris, Mei Chung, Alice H Lichtenstein, Ethan M Balk, Bruce Kupelnick, Harmon S Jordan, and Joseph Lau
AJCN 2006 84: 5-17. [Full Text]