Fasting hyperinsulinemia and changes in regional body composition in humanimmunodeficiency virus-infected women

A novel lipodystrophy syndrome (characterized by insulin resistance, hypert riglyceridemia, and fat redistribution) has recently been described in huma n immunodeficiency virus (HIV)-infected patients. However, investigation of the Lipodystrophy syndrome has generally been limited to men; and a compre hensive evaluation of insulin, lipids, and regional body composition has no t been performed in the expanding population of HIV-infected women. In this study, we assessed fasting insulin, Lipid levels, virologic parameters, an d regional body composition, using dual-energy x-ray absorptiometry, in a c ohort of 75 HIV-infected women (age, 25-46 yr), in comparison with 30 healt hy weight-matched premenopausal control subjects. HIV-infected women demons trated significant truncal adiposity (38.5 +/- 0.9 vs. 34.9 +/- 1.3%, P < 0 .05) hyperinsulinemia (15.9 +/- 1.5 vs. 7.5 +/- 0.6 mu U/mL, P < 0.001) and an increased insulin-to-glucose ratio (0.2 +/- 0.02 vs. 0.1 +/- 0.03, P < 0.001), compared with control subjects. Insulin and the insulin-to-glucose ratio were increased, even among HIV-infected patients with low body weight (<90% of ideal body weight) (insulin, 13.3 +/- 2.8 mu U/mL, P < 0.01 vs. c ontrol; insulin/glucose, 0.2 +/- 0.04, P < 0.01 vs. control). Insulin and t he insulin-to-glucose ratio were most significantly elevated among patients with increased truncal adiposity (insulin, 28.2 +/- 3.2 mu U/mL, P < 0.001 vs. control; insulin/glucose, 0.32 +/- 0.04, P < 0.001 vs. control). In co ntrast, no differences in insulin were seen in relation to protease inhibit or (PI) use. Similarly, HIV-infected women also demonstrated significant hy pertriglyceridemia (144 +/- 15 vs. 66 +/- 23 mg/dL, P < 0.01 vs, controls), which was present even among low-weight patients (148 +/- 32 mg/dL, P < 0. 001 vs. control) but was not related to truncal adiposity or PI usage. Thes e data demonstrate significant hyperinsulinemia and truncal adiposity in HI V-infected women. Our data suggest that these metabolic abnormalities occur at baseline in HN-infected women, independent of PI use. However, these da ta do not rule out a direct effect of PI therapy on fat metabolism or indir ect effects of PI therapy to further worsen glucose and lipid homeostasis i n association with weight gain and disease recovery.