Further research should include the use of nonhuman primate or other animal models to study the effects of fish oil on thrombosis and vascular injury as well as the ability of fish oil to inhibit atherogenesis. The possibility that fish oil will hasten regression of established atherosclerosis must be explored in both animals and man. We think it is still premature to conduct large-scale, exceedingly expensive clinical trials of fish oils in humans with coronary artery or peripheral vascular disease, since its optimal preparations and dosages (eg, whole oil, modified oil, purified fatty acids) are still unknown. However, smaller clinical trials focusing on the use of fish oil in highly selected human populations with vascular injury such as following coronary artery transluminal angioplasty or saphenous vein bypass grafting, should be encouraged.
Last, the potential benefits and possible toxicities of combining fish oil with other antithrombotic or antiatherosclerotic agents should be explored. For example, the use of antiplatelet agents or anticoagulants such as warfarin in conjunction with fish oil might be considered, with close observation for hemorrhagic or other side effects. review
1. Sufficient information is now available to recommend an expansion of studies which examine the effects offish oil or its components on the development and regression of atherosclerosis and on platelet-vascular interactions in experimental or clinical vascular disease.
2. A recommendation for the use of fish oil or purified omega-3 fatty acids in atherosclerotic vascular disease or thrombosis cannot be made at this time. There are no prospective controlled clinical trials on which to base a judgment as to efficacy or toxicity.
3. These agents should not be used in the treatment or prophylaxis of human vascular disease except as a part of controlled clinical studies.
4. The inclusion of fish in the diet, especially as a substitute for foods containing large amounts of saturated fatty acids, should not be harmful and might be beneficial.