Objective

To assess the correlates of risk of the different types of lipodystrophy and their modifications over time in a cohort of HIV-positive women receiving antiretroviral therapy (ART).

Methods

A consecutive series of HIV-infected women receiving ART was prospectively enrolled between 1 and 31 March 1998, and followed up for 2 years. Adipose tissue alterations (ATAs) and their variations over time were assessed by means of clinical observation and anthropometric measurements, and logistic regression analysis was used to describe the associated risk factors identified by univariate and multivariate analyses.

Results

One-hundred-and-seventeen of the 212 women (55.2%) developed ATAs during the 24 months of follow-up. Central adiposity was observed in 95 patients and peripheral lipoatrophy in 91, with 21 patients (9.9%) showing pure lipoatrophy, 26 (12.3%) pure fat accumulation and 70 (33%) combined forms. Only six of the 223 regional adipose tissue alterations identified in 74 patients during the first 12 months of the study had disappeared by month 24. Of the 43 patients who developed breast enlargement during the first 12 months, 11 (25.6%) showed a decrease in breast size during the second year of follow-up that was unrelated to changes in therapy or therapeutic success. The development of ATAs during the first 12 months of follow-up independently correlated with protease inhibitor (PI) use (OR 2.81, P=0.002) but, by the end of the second year of follow-up, the only factor significantly related to the development of ATAs was the overall duration of ART (OR 1.85, P=0.041). The use of PI significantly increased the risk of developing central adiposity during the first 12 months of the study (OR 2.27, P=0.002), whereas the only variable significantly influencing the risk at month 24 was HIV-infection due to intravenous drug use, which proved to be protective (OR 0.53, P=0.043). During the first 12 months of follow-up, the development of peripheral lipoatrophy was significantly associated with stavudine (OR 2.19, P=0.037) and PI use at enrolment (OR 2.27, P=0.023). At the end of the study, the variables associated with peripheral lipoatrophy were stavudine use at enrolment (OR 2.82, P=0.002), ART exposure for >1000 days at enrolment (OR 2.32, P=0.007), a CD4 cell count of >200/μl at enrolment (OR 2.89, P=0.002) and an age of >28 years (OR 1.91, P=0.036). The only factor significantly associated with an increased risk of breast enlargement during the first 12 months of follow-up was PI use (adjusted OR 2.51; 95% CI: 1.16–5.46, P=0.02); however, at month 24, none of the tested variables was associated with a significantly increased risk of this ATA.

Conclusions

ATAs (particularly central adiposity) are frequent in women treated with ART, and the different forms have different correlates of risk. Once they have become clinically evident, they generally tend to remain or worsen, and improve in only a small minority of cases. The considerable variations in adipomasty over time are apparently unrelated to changes in ART.