Background:
Several lines of evidence suggest that supplementation of diet with omega-3 polyunsaturated fatty acids (ω-3) PUFA), commonly referred to as fish oils, may reduce blood pressure (BP). However, most clinical trials of ω-3 PUFA supplementation have been of insufficient size to detect relevant BP changes.
Methods:
We conducted a meta-analysis of 17 controlled clinical trials of ω-3 PUFA supplementation. To estimate an overall effect of ω-3 PUFA supplementation on BP, we calculated the net BP change in each trial (BP A in ω-3 PUFA group minus BP A in control group), which was then weighted according to the inverse of the variance.
Results:
In the 11 trials that enrolled normotensive individuals (n=728), ω-3 PUFA supplementation led to significant reductions of systolic BP (SBP) and diastolic BP (DBP) in two and one trials, respectively. In the six studies that enrolled untreated hypertensives (n=291), significant reductions of SBP and DBP were present in two and four trials, respectively. Weighted, pooled estimates of SBP and DBP change (mm Hg) with 95% confidence intervals were —1.0 (—2.0 to 0.0) and —0.5 (—1.2 to +0.2) in the trials of normotensives, and —5.5 (—8.1 to —2.9) and —3.5 (—5.0 to —2.1) in the trials of untreated hypertensives. In 13 of 17 studies, trial duration was less than 3 months. Doses of ω-3 PUFA tended to be high (average dose >3 g/d in 11 trials). The magnitude of BP reduction was greatest at high BP but was not significantly associated with dose of ω-3 PUFA. Side effects, most commonly eructation and a fishy taste, occurred more frequently in ω-3 PUFA participants than in control participants (28% vs 13%, P<.001).
Conclusions:
Our analyses indicate that diet supplementation with a relatively high dose of ω-3 PUFA, generally more than 3 g/d, can lead to clinically relevant BP reductions in individuals with untreated hypertension. However, use of ω-3 PUFA as antihypertensive therapy will require demonstration of long-term efficacy and patient acceptability of lower doses.(Arch Intern Med. 1993;153:1429-1438)