Diet supplementation with fish oils and blood pressure reduction: a meta-analysis
ACP J Club. 1994 Jan-Feb;120:9. doi:10.7326/ACPJC-1994-120-1-009
Appel LJ, Miller ER 3d, Seidler AJ, Whelton PK. Does supplementation of diet with 'fish oil' reduce blood pressure? A meta-analysis of controlled clinical trials. Arch Intern Med. 1993 Jun 28;153:1429-38.
To determine using meta-analysis whether supplementation of the diet with omega-3 polyunsaturated fatty acids (fish oils) reduces blood pressure (BP) in normotensive and hypertensive persons.
Relevant citations were retrieved by searching MEDLINE for English-language publications of controlled clinical trials evaluating the effectiveness of fish oil supplementation on BP. Bibliographies of review and original articles were also scanned.
Criteria for including trials in the meta-analysis were the existence of both an active intervention group and a control group; a concurrent control group receiving little or no fish oil; healthy study participants; no concurrent use of antihypertensive therapy; and results reported in such a way that the net BP difference between active and control groups and the standard error of the mean could be calculated.
Data on the sample size, participant characteristics, study design, type of fish oil intervention, eicosapentaenoic acid (EPA) dose, docosahexaenoic (DHA) dose, type of control intervention, side effects, and BP were extracted. Pooled weighted estimates of net BP change across studies were calculated.
17 clinical trials met the selection criteria. 11 trials were done with normotensive participants (n = 728) and 6 with untreated hypertensive (mean diastolic BP > 90 mm Hg) participants (n = 291). The mean dose of fish oil was ≥ 3 g/d in 11 trials. When data from all 17 studies were pooled, fish oil supplementation led to reductions in both systolic BP (SBP) and diastolic BP (DBP) (net change in SBP, -1.5 mm Hg; 95% CI, 2.4 to -0.6 mm Hg; net change in DBP, -1.0 mm Hg; CI, -1.6 to -0.4 mm Hg). The results were not consistent across studies (test for heterogeneity for DBP, P < 0.01; for SBP, P = 0.05). Weighted, pooled estimates of SBP and DBP change were -1.0 mm Hg (CI, -2.0 to 0.0 mm Hg) and -0.5 mm Hg (CI, -1.2 to 0.2 mm Hg), respectively, in trials with normotensive persons and -5.5 mm Hg (CI, -8.1 to -2.9 mm Hg) and -3.5 mm Hg (CI, -5.0 to -2.1 mm Hg), respectively, in trials with hypertensive participants. The tests of homogeneity for normotensive and hypertensive trials grouped separately were nonsignificant. The magnitude of BP reduction increased progressively by the level of BP and was not associated with fish oil dose.
Diet supplementation with omega-3 polyunsaturated fatty acids (> 3 g/d) reduced both systolic and diastolic blood pressure in untreated hypertensive participants. This reduction was not observed for normotensive persons.
Sources of funding: Johns Hopkins University School of Medicine; Pew Charitable Trust; The Rockefeller Foundation; National Institutes of Health.
For article reprint: Dr. L.J. Appel, Welch Center for Prevention, Epidemiology and Clinical Research, 600 North Wolfe Street, Carnegie 291, Baltimore, MD 21287-6231. FAX 410-955-0476.
Pharmacies, supermarkets, and health food stores sell fish oil capsules for prevention of cardiovascular disease. Has this therapy proved effective? These 3 meta-analyses and 1 other (1) provide some answers. Gapinski and colleagues reported that fish oil was associated with a significant reduction in angiographically defined restenosis after angioplasty. The studies by Appel and Morris and their colleagues showed that fish oil was associated with small reductions in BP in persons with hypertension and hypercholesterolemia.
Before fish oil supplementation is recommended routinely for such disorders, several issues should be considered. As is often the case, the numbers of participants involved in the reported trials have been small and have represented very selected groups. Few women, nonwhites, elderly persons, and patients with comorbidities have been studied. The trials assessing BP have been short; few have lasted more than 3 months. Most studies have compared fish oil supplementation, rather than high fish diets, with placebo capsules.
The doses that have been associated with positive effects have been on the order of 3.5 to 5 grams daily. This is equivalent to 8 to 10 fish oil capsules that are available commercially. (This is roughly equivalent to the following numbers of 100-gram servings of fish daily: mackerel, 1 to 2; salmon, 3 to 4; tuna, 6 to 10; and flounder, 15 to 25 [2, 3].) It is unclear whether benefits can be achieved through supplementations with small amounts of commercially prepared fish oil.
It is also unclear whether high doses of fish oil (> 6 g/d) have detrimental effects, such as higher restenosis rates or major bleeding risks. Even doses of 4 to 5 grams daily may produce side effects, such as elevated low-density lipoprotein cholesterol, belching, bad taste, and diarrhea (3). It may be difficult to achieve high compliance with the prescription of a large number of pills that are associated with socially undesirable side effects. Finally, the trials reviewed have assessed physiologic or intermediate outcomes, such as angiographic lesions and BP, rather than the clinical outcomes of angina or myocardial infarction.
Looking at fish consumption, rather than fish oil supplementation, several prospective studies and 1 randomized controlled trial have suggested that modest dietary fatty fish consumption is associated with decreases in all-cause and ischemic heart disease mortality (4, 5). The randomized trial evaluated 3 dietary interventions (low fat, high fatty fish, and high fiber) in 2033 men with recent myocardial infarction (5). The fatty fish diet consisted of at least 2 weekly portions of fish such as mackerel, herring, or sardines. 22% of participants in the fatty fish group also took very low doses of fish oil supplementation. The rate of all-cause mortality among the men on the fish diet decreased significantly from 12.8% to 9.3% (relative risk reduction, 29%). Ischemic heart disease mortality also fell significantly from 11.4% to 7.7% (relative risk reduction, 33%). No clinically significant side effects were seen with the fish diet, and serum cholesterol was not significantly increased at the end of the 2-year trial.
Although these 3 well-designed meta-analyses show that fish oil supplementation lowers BP in patients with hypertension and that it prevents restenosis after angioplasty, it is premature to adopt fish oil supplementation as routine clinical practice for several reasons. The information available may not be generalizable to most patients. The therapeutic window for fish oil is not well defined, and side effects occur. The long-term clinical outcomes are unknown. Finally, the feasibility of achieving compliance with multiple pills without adversely affecting compliance with other clinically proved regimens (e.g., diuretics for hypertension) is not clear.
Despite these deficiencies in our knowledge regarding fish oil supplementation, current findings are promising. Clinicians should look for results of ongoing multicenter trials that will better establish the safety, tolerability, and efficacy of palatable doses of fish oils. Whether fish oil supplementation is superior to dietary fish intake should also be evaluated. Meanwhile, it is prudent to routinely recommend modest dietary intake of fatty fish, especially for persons who have had a myocardial infarction.
Cynthia D. Mulrow, MD
Audie L. Murphy Memorial Veterans Hospital San Antonio, Texas, USA