Long-term effects of fish oil on insulin resistance and plasma lipoproteins in NIDDM patients with hypertriglyceridemia.

OBJECTIVE: The aim of this study was to evaluate the long-term (6-month) effects of moderate fish oil supplementation on insulin sensitivity and plasma lipoproteins in NIDDM patients with hypertriglyceridemia. RESEARCH DESIGN AND METHODS: The study has been performed according to a randomized double-blind placebo-controlled design with a parallel group sequence. After a washout period of 4 weeks and a run-in period of 3 weeks, 16 NIDDM patients with hypertriglyceridemia (triglyceride [TG], 2.25-5.65 mmol/l) were randomly assigned to either fish oil (2.7 g/day eicosapentaenoic plus docosahexaenoic acid for 2 months, then 1.7 g/day for 4 more months) (n = 8) or placebo (n = 8). Diet and hypoglycemic drugs remained unchanged throughout the whole experiment. At baseline and after 6 months, insulin sensitivity was measured by euglycemic hyperinsulinemic clamp (insulin infused, 2.0 mIU.kg-1 body wt.min-1). At the same time, blood glucose control, fasting and postprandial serum insulin and nonesterified fatty acid (NEFA) concentrations, and fasting plasma lipoprotein concentrations were evaluated. RESULTS: In the group treated with fish oil compared with the baseline, there was: 1) a significant reduction in both plasma TG (2.92 +/- 0.23 vs. 3.85 +/- 0.32 [mean +/- SE] mmol/l, P < 0.001) and VLDL-TG (2.35 +/- 0.24 vs. 4.25 +/- 0.66 mmol/l, P < 0.01), without significant changes in blood glucose control; 2) a significant reduction in fasting NEFA concentrations (572 +/- 100 vs. 825 +/- 131 mumol/l, P < 0.01); and 3) a significant enrichment in long-chain omega-3 fatty acids of erythrocyte membrane phospholipids. In the placebo group, there were no changes in any of the variables analyzed. The insulin-mediated glucose uptake was unchanged in both groups (fish oil, 4.04 +/- 0.82 mg.kg-1.min-1 at baseline and 3.96 +/- 0.50 mg.kg-1.min-1 at 6 months; placebo, 3.51 +/- 0.62 mg.kg-1.min-1 at baseline and 4.09 +/- 0.49 mg.kg-1.min-1 at 6 months). CONCLUSIONS: In NIDDM patients with hypertriglyceridemia, moderate amounts of fish oil induce a long-term significant reduction in plasma triglycerides, VLDL triglycerides, and NEFA and a significant enrichment in the erythrocyte phospholipid content of long-chain omega-3 fatty acids, without deteriorating blood glucose control. However, this amount of omega-3 fatty acids was unable to improve insulin sensitivity in this group of patients.

Long-term effects of fish oil on lipoprotein subfractions and low density lipoprotein size in non-insulin-dependent diabetic patients with hypertriglyceridemia.

The effects of fish oil on lipoprotein subfractions and low density lipoprotein (LDL) size in non-insulin-dependent diabetes mellitus (NIDDM) patients with hypertriglyceridemia are unknown. To elucidate this, 16 NIDDM hypertriglyceridemic patients (plasma triglyceride 2.25- 5.65 mmol/l, plasma cholesterol < or = 7.75 mmol/l) were randomly assigned to a 6-month period with either moderate amounts of fish oil (n = 8) or placebo (n = 8) after 4 weeks of wash-out and 3 weeks of run-in. Diet and hypoglycemic treatment were unchanged throughout the experiment. LDL size were evaluated at baseline and after 6 months. Three VLDL and LDL subfractions were measured at the end of the two periods. The total lipid concentration of all very low density lipoprotein (VLDL) subfractions was lower at the end of fish oil treatment compared with placebo (large VLDL 124.3 +/- 19.7 mg/dl vs 156.7 +/- 45.5 mg/dl; intermediate VLDL 88.5 +/- 9.5 mg/dl vs 113.9 +/- 23.2 mg/dl; small VLDL 105.9 +/- 9.7 mg/dl vs 128.9 +/- 40.7 mg/dl) (mean +/- SEM), although the difference was not statistically significant. Moreover, at the end of the two treatments, the percentage distribution of VLDL subfractions was very similar (large 37.5 +/- 3.3% vs 37.6 +/- 2.6%, intermediate 27.6 +/- 0.9% vs 31.0 +/- 2.4%; small 34.9 +/- 3.7% vs 31.4 +/- 2.1%). Concerning LDL, no significant change in LDL size was observed after the two treatments (255.4 +/- 2.2 A vs 254.2 +/- 1.7 A, fish oil; 253.7 +/- 2.0 A vs 253.3 +/- 1.7 A, placebo). LDL subfraction distribution was also very similar (large 17 +/- 3% vs 17 +/- 2%; intermediate 62 +/- 3% vs 65 +/- 3%; small 21 +/- 3% vs 18 +/- 2%), at the end of the two periods, confirming the lack of effects on LDL size. In conclusion, our study indicates that in NIDDM patients with hypertriglyceridemia, fish oil does not induce any improvement in LDL distribution and LDL size despite its positive effects on plasma triglycerides.

Is the erythrocyte membrane fatty acid composition a valid index of skeletal muscle membrane fatty acid composition?

Recent studies suggest that insulin sensitivity is related to the fatty acid composition of phospholipids in skeletal muscle (SM) membranes. Since it is difficult to obtain muscle biopsies, it may be useful to have information on the fatty acid composition using more accessible cells such as erythrocytes. This would be possible only if the composition of erythrocyte and muscle membranes are very similar. Since no comparative data are available, we evaluated the phospholipid fatty acid composition of erythrocyte and SM membranes in 16 individuals, 10 nondiabetics (male to female ratio, 4:6; age, 50 +/- 11 years; body mass index, 27 +/- 5 kg/m2; mean +/- SD) and 6 type 2 diabetic patients (male to female ratio, 2:4; age, 64 +/- 5 years; body mass index, 27 +/- 4 kg/m2). All patients underwent abdominal surgery, during which a biopsy of the abdominal rectus muscle (50 to 100 mg) was obtained. Erythrocyte and SM phospholipid fatty acids were extracted and then methylated; the methyl fatty acids were separated and quantified by gas chromatography. Compared with erythrocyte membranes, muscle membranes showed a significantly higher proportion of omega-6 polyunsaturated fatty acid ([PUFA] 43.0% +/- 3.1% v29.7% +/- 1.6%, P < .001) and lower saturated fatty acid ([SFA] 41.1% +/- 1.5% v 43.4% +/- 1.2%, P < .001), monounsaturated fatty acid ([MUFA] 11.5% +/- 1.7% v 20.0% +/- 1.9%, P < .001), and omega-3 PUFA (3.8% +/- 0.6% v 7.4% +/- 1.0%, P < .001). The greatest increase involved linoleic acid (26.9% +/- 2.8% v 10.3% +/- 1.6%, P < .001), whereas lignoceric acid (0.8% +/- 0.2% v 5.0% +/- 0.6%, P < .001) and oleic acid (10.4% +/- 1.6% v 13.5% +/- 1.3%, P < .001) were significantly lower. These results show that erythrocyte and muscle membrane phospholipid fatty acids are significantly different. Therefore, data on SM membranes cannot be extrapolated on the basis of measures of erythrocyte phospholipid fatty acid composition.

Assay of erythrocyte membrane fatty acids. Effects of storage time at low temperature.

The study of the stability of saturated mono-, or polyunsaturated fatty acids, both esterified and not esterified, in plasma, circulating cells, and tissues is extremely important to validate the use of biological samples stored at low temperature in "biological banks", which are used for experimental, observational, dietary, or pharmacological studies. Since red blood cells are easily accessible cells, they are used as a marker of less-accessible tissues, especially in large-scale epidemiological studies. Data from the literature suggest that the addition of an antioxidant and the freezing of red blood cells do not cause any variation in the fatty acid composition for a period of 2-6 months up to 1 year. We evaluated the fatty acid concentration in red blood cells isolated from venous blood samples of one subject, preserved with butylated hydroxytoluene and N2 and stored at -80 degrees C for up to 2 years. Erythrocytes of venous samples of six subjects stored at -20 degrees C for 6 months without butylated hydroxytoluene and in the presence of air were used for comparison purposes. Our data demonstrate that a long storage time (2 years) does not significantly influence the erythrocyte fatty acid concentration when using very low temperatures (-80 degrees C) and antioxidants (butylated hydroxytoluene) in the presence of N2.

Very low density lipoprotein subfraction abnormalities in IDDM patients: any effect of blood glucose control?

Normolipidaemic insulin-dependent diabetic (IDDM) patients are characterized by an increase in the smaller VLDL particles, considered to be the most atherogenic. Since blood glucose control is one of the main regulators of lipid metabolism in diabetic patients, it could influence the shift in the distribution of VLDL subfractions towards smaller particles. To evaluate this possibility, VLDL subfractions, post-heparin lipoprotein lipase and hepatic lipase activities have been evaluated in male IDDM patients with either unsatisfactory blood glucose control (group 1, HbA1c > 8%, n = 18) or good blood glucose control (group 2, HbA1c < 8%, n = 16) and in 16 normoglycaemic individuals. The three groups were comparable for sex, age, body mass index, and plasma lipid levels. Three VLDL subfractions (large, Svedberg flotation unit (Sf) 175-400; intermediate, Sf 100-175; small, Sf 20-100) were separated by density gradient ultracentrifugation and analysed for cholesterol, triglyceride, and phospholipid levels. When compared to control subjects both groups of IDDM patients showed a clear shift in VLDL subfraction distribution with a significant increase in the proportion of small VLDL (group 1; 49 +/- 2%; p < 0.005; group 2: 51 +/- 3%, p < 0.01; control subjects 40 +/- 2%) (mean +/- SEM) in relation to total VLDL. By contrast, the absolute lipid concentration of small VLDL was higher only in group 1, compared to control subjects (35 +/- 4 vs 27 +/- 3 mg/dl, p = 0.05). Post-heparin hepatic lipase activity was significantly reduced in both IDDM groups (group 1: 254 +/- 19 mU/ml, p < 0.05; group 2: 202 +/- 19 mU/ml, p < 0.005; control subjects 317 +/- 31 mU/ml). In conclusion, normolipidaemic IDDM patients show an increase in the smallest VLDL, whatever their degree of blood glucose control. However, this abnormality may be clinically relevant only in patients with unsatisfactory blood glucose control, since absolute lipid concentration of these potentially atherogenic particles is only increased in this group.