Do Dietary Omega-3 Fatty Acids Affect Coronary Disease as Measured by Angiography?

February 2000; Volume 2: 14-15

Abstract & Commentary

Source: von Schacky C, et al. The effect of dietary omega-3 fatty acids on coronary atherosclerosis. Ann Intern Med 1999;130:554-562.

Objective: To determine the effect of dietary intake of omega-3 fatty acids on the course of coronary artery atherosclerosis in humans.

Subjects: Two hundred twenty-three patients hospitalized for diagnostic coronary angiography at the University of Munich between 1992 and 1994 were randomized to receive fish oil capsules or a placebo. Patients were stratified by current lipid-lowering therapy, whether PTCA was done less than six months prior to randomization, and the presence of two or more cardiovascular risk factors (LDL greater than 150 mg/dL, current smoking, history of myocardial infarction in a first-degree relative younger than 60 years of age, and hypertension).

Methods: One hundred eleven patients were given daily placebo capsules containing 1 g of a fatty acid mixture reflecting the fatty acid composition of the average European diet. One hundred twelve patients were given capsules of fish oil high in omega-3 fatty acids. Recommended dosage was six capsules per day for the first three months, followed by three capsules a day for the next 21 months. Coronary angiography was performed prior to the treatment protocol and again after two years. The subjects were evaluated at months 1, 6, 12, and 18. Eighty in the placebo and 82 in the treatment group finished the study. Angiograms were reviewed simultaneously by a panel of three invasive cardiologists who were blinded to patient assignment and temporal order of films. A fourth blinded moderator documented the results of the panel. Additional analysis also utilized quantitative coronary angiography. Compliance was measured by questioning pill counts and analysis of erythrocyte phospholipid fatty acid composition (reflecting dietary fatty acid composition).

Results: Of the 80 patients in the placebo group with before and after angiograms, 35 showed changes in global scores; of 82 fish oil recipients with before and after angiograms, 35 showed changes in global scores. In an extended post hoc analysis of fewer patients and more coronary artery segments, only compliant patients with angiographic changes were included, and vessels with segments treated by PTCA were analyzed by the panel (the actual segments treated with PTCA were not analyzed). Of 48 pairs of films in the placebo group, 36 coronary artery segments showed mild progression, five showed moderate progression, and seven showed mild regression. Of 55 pairs of films in the fish oil group, 35 showed mild progression, four showed moderate progression, 14 showed mild regression, and two showed moderate regression (P = 0.041 for comparison between groups).

Quantitative angiography showed no significant differences in mean loss of minimal luminal diameter per patient. There was not a statistically significant difference between groups in the incidence of myocardial infarction and stroke. The authors conclude that dietary intake of omega-3 fatty acids modestly mitigates the course of coronary atherosclerosis in humans.

Funding: In part by the Bundesministerium fur Forschung und Technologic, Germany, through Gesellschaft fur Strahlenforschung (GSF. 07ERG03) and Deutsche Forschungsanstalt fur Luft-und Raumfahry (DLR. 01 EA 9501/7); Wilhelm Sander-Stuftung (93.032); Fundacion Federico; and the Deutsche Forschungsgemeinschaft (Scha 398). Pronova, A.S. Lysader, Norway provided the capsules and funds for monitoring.

Comments by William Benda, MD

Von Schacky et al have created a thorough, well-designed study to examine the benefit of dietary omega-3 fatty acids in heart health. If only life were as simple as their methodology.

Let’s begin by examining what works well in this study. The subjects were well chosen and the stratification variables well deliberated. The composition of both the placebo and the fish oil is thoroughly delineated. The blinding of the expert panel (and moderator) to all aspects of the results including the temporal order of films strives for thorough impartiality, and an appropriate statistical method for ordinal data was used. Examination of compliance to include erythrocyte phospholipid composition reveals extraordinary attention to an often-overlooked aspect of clinical studies, although a run-in period would make an even stronger case for compliance (which did not differ between groups). The funding source does not appear to represent a conflict of interest. The conclusions of the authors are modest to allow the results to be viewed with clarity by the general practitioner.

One significant question arises when we relate this controlled clinical trial to uncontrolled clinical practice. If we are investigating a supplement that is also present in commonly ingested foods, where are the dietary data? Most trials of this sort include information on the subjects’ nutritional history. The authors indirectly compare their trial to the Diet and Reinfarction Trial (DART), a larger trial of 2,033 men that showed a 29% reduction in overall mortality in survivors of a first myocardial infarction who consume actual fish rich in omega-3 fatty acids.1 The DART study looked at reduction in fat intake, an increase in the ratio of polyunsaturated to saturated fat, an increase in fatty fish intake, and an increase in cereal fiber. Only the fish intake was associated with reduction in mortality. Detailed dietary questionnaires were administered, the subjects received regular visits and telephone calls from dietitians, and smoking cessation was strongly encouraged. The GISSI-Prevenzione trial showed significant decrease in death, myocardial infarction, and stroke in 11,324 patients randomly assigned to omega-3 fatty acids, vitamin E, both, or neither (omega-3 fatty acids, but not vitamin E, significantly lowered risk).2 Food-frequency questionnaires were administered at months 0, 6, 12, 18, 30, and 42.

In the von Schacky trial overweight patients were "advised to restrict caloric intake," and all patients were "advised to avoid eating cholesterol-rich foods; no other dietary advice was given." The authors did follow the patients’ LDL/HDL/triglyceride levels, which may serve as surrogate markers for fat intake. Even in studies where diet is controlled as rigidly as possible, interference from variety in the patients’ intake occurs; in this trial such problems are not even addressed. It is risky to accept the conclusions in light of potential confounders. If the study were to be replicated it may be wise to incorporate thorough dietary investigation. It would also be a good idea to make the trial longer in duration. The von Schacky trial was two years shorter than the GISSI-Prevenzione trial; 18 months may be too brief to see dramatic results.

Let us also consider a philosophical point that goes beyond the intent of this study: The data collection follows the conventional path of examining patients after a disease is fully established. The test subjects are, as usual, sick but not too sick, and may represent only a small segment of the population for which fish or fish oil may be of use. Omega-3 fatty acids have wide-ranging anti-inflammatory effects (and atherosclerosis is considered an inflammatory event) that are not reflected in laboratory measurements of serum lipids. Perhaps the greatest benefit of dietary supplementation with omega-3 fatty acids is in prevention of atherosclerosis; not simply as an addition to the diet, but as a synergistic part of changes in nutrition, exercise, and lifestyle stressors.

Dr. Benda is a Fellow in the Program in Integrative Medicine at the University of Arizona in Tucson, AZ.


1. Burr ML, et al. Effects of changes in fat, fish, and fibre intakes on death and myocardial reinfarction: Diet and reinfarction trial (DART). Lancet 1989;334:757-761.

2. GISSI-Prevenzione Investigators et al. Dietary supplementation with n-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: Results of the GISSI-Prevenzione trial. Lancet 1999;354:447-455.