LONGITUDINAL CHANGES IN BODY-COMPOSITION MEASURED WITH A VARIETY OF METHODS IN PATIENTS WITH AIDS

We test the hypothesis that human immunodeficiency virus (HIV)-related weight loss is accompanied by inappropriately large losses of fat-fre e mass (FFM). Our secondary aims were to examine whether FFM increases during weight gain and to compare several techniques for measuring FF M change. FFM was measured at intervals averaging 5 months in 21 AIDS patients by means of skinfold thickness (SF), dual-energy x-ray absorp tiometry (DEXA), total body water (TBW), and bioelectrical impedance u sing the equation of the manufacturer of the equipment (BIA(EZComp)) a nd a published prediction equation (BIA(Segal)). The FFM content of we ight loss was similar for SF (57%), DEXA (60%), TBW (55%) and BIA(EZCo mp) (65%), but the result from BIA(Segal) (78%) was higher. The result s were close to predicted starvation values apart from the results wit h BIA(Segal), which were significantly higher than predicted values. W eight gain was also composed of a large proportion of FFM. There were large intermethod differences in measurements of absolute FFM, but for measuring changes in FFM, the bias between SF, DEXA, and TBW was mini mal. The results of BIA vary with the prediction equation used. In thi s group of patients with the acquired immune deficiency syndrome (AIDS ), weight loss was composed of a large proportion of FFM, but in gener al this is compatible with undernutrition as the underlying cause and does not support the hypothesis of excessive FFM catabolism in HIV dis ease. SF, DEXA, TBW, and BIA(Segal) show reasonable agreement for meas uring body composition changes. This information should be considered in the design of future intervention studies for HIV-related wasting.