OBJECTIVES: The aim of the study was to assess the impact of the gain in body mass index (BMI) observed immediately after antiretroviral therapy (ART) initiation on the subsequent risk of cardiovascular disease (CVD) and diabetes.

METHODS: We analysed data from the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) cohort study. Outcomes were development of (i) CVD (composite of myocardial infarction/stroke/coronary procedure) and (ii) diabetes. The main exposure variable was change in BMI from ART initiation (pre-ART) to 1 year after initiation (continuous variable) in treatment-naïve individuals initiating ART with no history of CVD or diabetes (for respective outcomes). BMI [weight (kg)/(height (m))(2) ] was categorized as underweight (< 18.5), normal (18.5-25), overweight (25-30) and obese (> 30). Poisson regression models were fitted stratified for each pre-ART BMI category to allow for category-specific estimates of incidence rate ratio (IRR). Models were adjusted for pre-ART BMI and CD4 count, key known risk factors (time-updated where possible) and calendar year.

RESULTS: A total of 97 CVD events occurred in 43 982 person-years (n = 9321) and 125 diabetes events in 43 278 person-years (n = 9193). In fully adjusted analyses for CVD, the IRR/unit gain in BMI (95% confidence interval) in the first year of ART, by pre-ART BMI category, was: underweight, 0.90 (0.60-1.37); normal, 1.18 (1.05-1.33); overweight, 0.87 (0.70-1.10), and obese, 0.95 (0.71-1.28) (P for interaction = 0.04). For diabetes, the IRR/unit gain in BMI was 1.11 (95% confidence interval 1.03 to 1.21), regardless of pre-ART BMI (P for interaction > 0.05).

CONCLUSIONS: Short-term gain in BMI following ART initiation appeared to increase the longer term risk of CVD, but only in those with pre-ART BMI in the normal range. It was also associated with increased risk of diabetes regardless of pre-ART BMI.