Increased abdominal visceral fat is associated with reduced bone density in HIV-infected men with lipodystrophy

Objective: To examine the relationship between bone density and changes in regional and whole body composition in HIV-infected men with and without li podystrophy. Design: Cross-sectional, observational study of HIV-infected men with and w ithout lipodystrophy and matched HIV-negative controls. Setting: Tertiary care academic medical institution. Patients: A total of 59 men, belonging to three different groups: HIV-posit ive men with lipodystrophy (n = 21), HIV-positive men without lipodystrophy (n = 20), and age-matched and body mass index-matched HIV-negative control s (n = 18). Methods: Bone density, markers of bone turnover and indices of calcium meta bolism were measured in all subjects. Quantitative computed tomography was used both to determine volumetric bone density of the spine and to quantify abdominal visceral fat. Dual energy X-ray absorptiometry was used to deter mine whole body composition and bone density. Statistical comparisons were performed according to lipodystrophy categorization and protease inhibitor exposure. Results: Men with HIV-associated lipodystrophy had reduced lumbar spine bon e density compared with both HIV-infected non-lipodystrophic men (mean +/- SD, 132 +/- 29 versus 154 +/- 30 mg/cm(3); P = 0.021 and HIV-negative contr ols (mean +/- SD 132 +/- 29 versus 148 +/- 18) mg/cm(3); P = 0.04). Lumbar spine bone density was reduced significantly in HIV lipodystrophy patients independently of protease inhibitor use. In an analysis among all HIV-infec ted subjects, increased visceral abdominal fat area was associated with dec reased lumbar spine bone density (r, -0.47; P = 0.002). The association bet ween visceral fat and bone density remained significant (P = 0.007) after c ontrolling for age, body mass index, lowest body weight, protease inhibitor use, and extremity fat in a multivariate regression model. Markers of bone turnover were not related to bone density or lipodystrophy status. Conclusions: Lumbar spine bone density is reduced in association with incre ased visceral fat in HIV-infected men with lipodystrophy. Further studies a re needed to determine the mechanisms of osteopenia in HIV lipodystrophy an d whether increased marrow fat occurs in such patients and affects bone den sity. (C) 2001 Lippincott Williams & Wilkins.