Background:  Adiponectin and leptin are adipose-derived proteins with important roles in energy and glucose metabolism. Low adiponectin and high leptin concentrations are associated with obesity and HIV lipodystrophy. Although hypoadiponectinemia has been associated with use of some protease inhibitors (PI) and nucleoside reverse transcriptase inhibitors (NRTI), including ritonavir (RTV) and stavudine (4dT) among HIV-infected adults, no data are available in HIV-infected children and adolescents.

Methods:  Height, weight, and adiponectin and leptin (by radioimmunassay) were measured in 97 HIV-infected children and adolescents. Measures of glucose, cholesterol, triglycerides, hepatic enzymes, ART use, and Tanner stage were obtained from medical charts.

Results:  Mean age was 15.0 years (range 3 to 24 years), 34 (35%) were children <13 years of age, 50.5% were male, 52 (54%) were Latino, 22 (23%) were African American, 18 (19%) were Caucasian, and 5 (5%) were Asian. Mean body mass index was 21 kg/m² (range 14 to 41 kg/m²); 23% were Tanner I and 56% were Tanner stage V; 70% used ART, including lamivudine (3TC) (41%), zidovudine (AZT) (28%), nelfinavir (NFV) (26%), 4dT (22%), efavirenz (EFV) (22%), lopinavir/ritonavir (LPV/r) (16%) and RTV (11%). Mean CD4 percentage was 28%, and 55% had viral loads <400 copies/mL. Mean adiponectin levels were 12.74 µg/mL (SD 6.3). In univariate analysis, adiponectin was inversely associated with body mass index, age, and pubertal status (p <0.001), and was marginally associated with female gender (p = 0.06) and alanine aminotransferase (ALT) (p = 0.056). Adiponectin was significantly lower among 4dT (10.63 vs 13.32 µg/mL, p = 0.049) and 3TC users (11.4 vs 13.7 µg/mL, p = 0.031), after adjusting for body mass index, Tanner stage, and sex, but did not differ by PI nor NNRTI use. Multivariable analysis showed that 4dT (β coefficient:  –4.02, p = 0.019) or 3TC use (β coefficient:  –2.47, p = 0.033) were independently associated with lower adiponectin concentration after adjusting for body mass index, sex, Tanner stage, ALT, and concomitant PI or NNRTI use. The mean leptin concentration was 6.37 ng/mL (SD 6.7). Leptin was associated with female sex, body mass index, age, and Tanner stage (p <0.02) in univariate analysis. However, multivariable regression showed body mass index (β coefficient 0.87, p <0.0001) and female gender (β coefficient 6.06, p <0.0001) were the only covariates associated with higher leptin levels.

Conclusions:  Hypoadiponectinemia was associated with 4dT and 3TC use, while leptin concentration was not associated with ART use in this cohort of HIV-infected children and adolescents.