Background:  HIV-associated lipodystrophy is characterized by central fat accumulation, or subcutaneous fat atrophy, or abnormalities in glucose tolerance or lipid metabolism. Interventions that improve glucose tolerance have been postulated to reduce visceral fat. The use of metformin in the treatment of lipodystrophy has been studied in HIV-infected persons with impaired glucose tolerance. However, metformin has not been well studied in HIV-infected persons with lipodystrophy and normal glucose tolerance.

Methods:  HIV-infected persons with lipodystrophy, defined by self-reported increase in abdominal girth and waist-hip ratio ≥0.95 in men and ≥0.85 in women, and normal glucose tolerance (fasting glucose ≤100 mg/dL) were randomized to receive metformin 1500 mg daily or placebo for 24 weeks in a prospective, double-blind trial. The primary endpoint was appendicular fat mass measured by DEXA after 24 weeks. Secondary endpoints included visceral adipose tissue measured by single-slice CT scan and serum lipid profile.

Results:  We randomized 48 participants to the metformin (n = 25) and placebo (n = 23) groups. Of all participants, 56% were male and mean age was 42 years. Mean fasting glucose in the metformin and placebo groups was 94.0 vs 97.2 mg/dL (p = 0.73). In each group, 52% of participants received protease inhibitors. Mean CD4 count in the metformin and placebo groups was 370 vs 420 cells/mm³ (p = 0.61). The percentage of participants in each group with an undetectable viral load was similar. Metformin use over 24 weeks was associated with a decrease in appendicular fat mass compared to placebo (–686.0 vs 161.0 kg, p = 0.03). After adjusting for age, height, and baseline appendicular fat mass, a trend towards decreased appendicular fat mass remained (–614.0 vs 95.3 kg, p = 0.12). There was no significant change in visceral adipose tissue after 24 weeks of metformin vs placebo (–22.2 vs –3.85 cm², p = 0.17). Metformin did not significantly change triglycerides, LDL, or HDL after 24 weeks compared to placebo.

Conclusions:  This randomized placebo-controlled trial failed to show a benefit of metformin on fat redistribution or dyslipidemia in HIV-infected persons with lipodystrophy and normal glucose tolerance. There was an unexpected trend toward reduction in appendicular fat mass. Metformin should be used with caution in HIV-associated lipodystrophy, and if used, should be reserved for persons with impaired glucose tolerance.