Changes in body mass index and clinical outcomes after initiation of contemporary antiretroviral regimens. AIDS
Bannister et al. for the EuroSIDA study group aimed to assess changes in body mass index (BMI) distribution during the period 2010–2019 and to investigate whether changes in BMI during this time of contemporary antiretroviral therapy (ART) are associated with long-term health outcomes (cardiovascular diseases [CVD], malignancies, diabetes mellitus [DM], all-cause mortality).
People with HIV (PWH) aged ≥18 years were included who started a new antiretroviral (baseline) during 2010–2019 with baseline and ≥1 follow-up BMI assessment available.
Overall, 6’721 PWH were included; 72.3% were male, median age 48 years (interquartile range [IQR] 40–55). At baseline, 8.4% were antiretroviral-naive, and 5.0% were underweight, 59.7% healthy weight, 27.5% overweight, and 7.8% were living with obesity. There was an 8.2% increase in proportion of overweight and 4.8% in obesity over the study period (median follow-up 4.4 years [IQR 2.6–6.7]). 100 CVDs, 149 malignancies, 144 DMs, and 257 deaths were observed with incidence rates 4.4, 6.8, 6.6, 10.6 per 1000 person-years of follow-up, respectively. Compared to stable BMI, >1 kg/m2 increase was associated with increased risk of DM (adjusted incidence rate ratio [IRR]: 1.96, 95% confidence interval [CI]: 1.36–2.80) and >1 kg/m2 decrease with increased risk of death (adjusted IRR: 2.33, 95% CI: 1.73–3.13). No significant associations were observed between BMI changes and CVD or malignancies.
In conclusion, the study observed some overall increase in proportions of PWH overweight or living with obesity by the end of our study period. An increase in BMI was associated with a higher rate of DM, consistent with previous research, and a decrease in BMI was linked to higher rate of all-cause mortality, which may be explained by declining health over a long period and potentially by differences in disease prevention interventions according to BMI. No significant association was found between BMI changes and CVD, although there was some indication of a potentially higher rate for decreased BMI. There was also no association with malignancies, but this included a variety of different cancers. The study authors determined that longer follow-up and additional events would be needed to confirm these results.