The reasons for increased vascular disease in diabetic persons are not completely known, but certain risk factors are much more common in patients with diabetes, especially adult-onset diabetes (non-insulin-dependent diabetes mellitus). These risk factors include hypertension, hyperlipidemia, visceral obesity, and hyperinsulinemia and have been called the "deadly quartet" or "syndrome x" [4, 5]. Several approaches may be taken to the prevention of coronary disease and peripheral atherosclerosis in persons with diabetes. Better glucose control is certainly one answer, but it is sometimes difficult to achieve in the free-living, affluent, and sedentary U.S. population. Diet and drugs can be used to treat hypertension and hyperlipidemia. All of these measures are helpful but may not completely solve the problem of excessive vascular disease.

When evidence showed that Greenland Eskimos, because of their high intake of n-3 fatty acids, had a much lower incidence of coronary disease than Danes [6, 7], a question was posed. Would these fatty acids, found in fish and seal in the Eskimo diet, prevent coronary disease in Western populations? These n-3 fatty acids are the very long chained and highly polyunsaturated fatty acids: eicosapentaenoic acid (20:5) and docosahexaenoic acid (22:6). They are present only in food from the oceans and fresh waters: fish, sea mammals, and shellfish. Many studies then appeared showing that fish and fish oil preparations had many diverse actions that could alleviate vascular disease [8, 9]. These included mild blood pressure reduction, definite hypolipidemic effects, an antithrombotic action (achieved by inhibiting thromboxane A₂ synthesis in platelets), inhibition of cellular growth factors and monocyte migration, and enhancement of nitric oxide production by the endothelium.

It seemed reasonable, then, that fish oil be used in patients with non-insulin-dependent and insulin-dependent diabetes mellitus. Glucose homeostasis was achieved with no problems when fish oil was used in persons with insulin-dependent diabetes mellitus [10, 11], and one study actually showed less microvascular albumin leakage in the kidney [11]. However, in some studies, fish oil given to overweight patients with non-insulin-dependent diabetes mellitus caused mild glucose intolerance compared with a control period [12, 13]. As Heine [13] has reviewed, these adverse effects occurred in short-term experiments that had insufficient diet and weight control when 15 g or more of fish oil (90 kcal) was added to the usual diet [13]. A more recent study comparing fish oil and olive oil (15 g given randomly over 12 months) showed no difference in glucose homeostasis in persons who received fish oil compared with controls according to five criteria: fasting glucose level, hemoglobin A_1c level, plasma and urinary C-peptide levels, and 24-hour urinary glucose excretion [14]. At the same time, plasma triglyceride and very-low-density lipoprotein (VLDL) levels markedly decreased during the fish oil period.

The randomized, double-blind, placebo-controlled approach taken by Toft and colleagues, whose results are published in this issue [15], has provided more evidence for the absence of a deleterious effect of fish oil on glucose metabolism. Toft and colleagues administered n-3 fatty acids in a fish oil concentrate to 78 obese volunteers who had essential hypertension. Four 1-g fish oil capsules provided both eicosapentaenoic acid and docosahexaenoic acid in a total amount of 3.4 g/d. The placebo was 4 g of corn oil, which contained linoleic acid, an n-6 polyunsaturated fatty acid. After 16 weeks, fish oil had reduced the mean systolic blood pressure by 4.4 mm Hg and the mean diastolic blood pressure by 3.2 mm Hg. This mild hypotensive effect is similar to that shown in many other studies [8]. Fish oil did not alter glucose control, even in persons with mild glucose intolerance. The tests used were extensive: an oral glucose tolerance test, a hyperglycemic clamp technique, and a euglycemic hyperinsulinemic clamp technique. These tests measured insulin sensitivity and first- and second-phase insulin release.

At the same time, a mild hypolipidemic effect occurred, and plasma triglyceride and VLDL levels decreased significantly. The high-density lipoprotein cholesterol level was also higher in the fish oil group than in the corn oil group. The plasma cholesterol and low-density lipoprotein levels did not differ between the fish oil and corn oil groups.

The hypotriglyceridemic effects of fish oil, which have been so universally observed, result from a decrease in triglyceride synthesis in the liver and from an increase in the removal of triglyceride from the blood. This has been documented in several studies done in humans and animals [8, 16]. Because of these actions, fish oil has been useful in the treatment of hypertriglyceridemic patients [16, 17], especially those with the types of hyperlipidemia (V, IV, and IIb) so common in diabetic persons.

Both the epidemiologic evidence and the results of a subsequent clinical trial done in the United Kingdom suggest that two to three fish meals per week might provide considerable protection against coronary disease [8, 18]. The clinical trial of n-3 fatty acids involved several thousand men who had recovered from myocardial infarction. The men who ate fish at least twice a week for 2 years had a 29% reduction in total mortality, largely because they had fewer deaths from coronary disease [18]. Persons who could not eat fish were given three fish oil capsules per day to supply the desired amount of n-3 fatty acids.

Practical advice for the patient with diabetes would be that he or she should consume fish (prepared by poaching, grilling, baking, or broiling) two to three times per week. If eating fish is not possible, the consumption of two to three fish oil capsules (1 g each) per day would provide the equivalent amounts of eicosapentaenoic acid and docosahexaenoic acid.

Because the study by Toft and colleagues [15] indicates both that fish oil has no harmful effects on glucose tolerance and that it has beneficial effects on blood pressure and plasma lipid levels, fish or fish oil might now be considered useful in the prevention of vascular disease in patients with diabetes.