AIMS/HYPOTHESISWe hypothesised that loss of peripheral fat in HIV patients would result in decreased plasma adipocytokines, in particular adiponectin, and that this decrease would be associated with changes in VLDL, IDL and LDL apolipoprotein B kinetics.METHODSPlasma adiponectin, leptin and other cytokines were measured in uninfected control subjects (n=12) and three HIV-positive groups comprising treatment-naïve patients (n=15) and patients on triple antiretroviral therapy containing protease inhibitors (PI, n=15) or non-nucleoside reverse transcriptase inhibitors (NNRTI, n=25). VLDL, IDL and LDL apolipoprotein B kinetics were measured with an infusion of [1-(13)C] leucine. Regional body fat was measured with a dual energy X-ray absorptiometry scan. Insulin resistance was calculated using homeostasis model assessment (HOMA).RESULTSAdiponectin (median [interquartile range]) was reduced in the treatment-naive (5.4 microg/ml [4.7-8.5]), PI (5.0 microg/ml [3.3-6.4]) and NNRTI (5.0 microg/ml [3.1-6.7]) groups compared with controls (9.7 microg/ml [6.9-13.3]) (p<0.05). In all subjects adiponectin correlated positively with HDL-cholesterol levels, the VLDL, IDL and LDL apolipoprotein B fractional clearance rates, and with the limb fat:lean body mass ratio (all p<0.01). Adiponectin correlated negatively with plasma triglyceride levels and HOMA (p<0.001). In a linear regression model that included HOMA, adiponectin was an independent predictor of VLDL and HDL-cholesterol levels and the IDL fractional clearance rate. TNF was higher in treatment-naive and PI subjects, and soluble TNF receptor superfamily, members 1A and 1B (previously known as TNF receptors 1 and 2) was higher in PI patients than in control subjects (p<0.05).CONCLUSIONS/INTERPRETATIONAdiponectin levels are significantly reduced in treated and untreated HIV patients and are predictive of VLDL and IDL apolipoprotein B fractional clearance rates. Adiponectin may have a direct effect on lipoprotein metabolism, which may be independent of insulin.