Obesity following ART initiation is common and influenced by both traditional and HIV-/ART-specific risk factors

Abstract

Background: Obesity rates are increasing among HIV-infected individuals, but risk factors for obesity development on ART remain unclear.

Objectives: In a cohort of HIV-infected adults in Rio de Janeiro, Brazil, we aimed to determine obesity rates before and after ART initiation and to analyse risk factors for obesity on ART.

Methods: We retrospectively analysed data from individuals initiating ART between 2000 and 2015. BMI was calculated at baseline (time of ART initiation). Participants who were non-obese at baseline and had ≥90 days of ART exposure were followed until the development of obesity or the end of follow-up. Obesity incidence rates were estimated using Poisson regression models and risk factors were assessed using Cox regression models.

Results: Of participants analysed at baseline (n = 1794), 61.3% were male, 48.3% were white and 7.9% were obese. Among participants followed longitudinally (n = 1567), 66.2% primarily used an NNRTI, 32.9% a PI and 0.9% an integrase strand transfer inhibitor (INSTI); 18.3% developed obesity and obesity incidence was 37.4 per 1000 person-years. In multivariable analysis, the greatest risk factor for developing obesity was the use of an INSTI as the primary ART core drug (adjusted HR 7.12, P < 0.0001); other risk factors included younger age, female sex, higher baseline BMI, lower baseline CD4+ T lymphocyte count, higher baseline HIV-1 RNA, hypertension and diabetes mellitus.

Conclusions: Obesity following ART initiation is frequent among HIV-infected adults. Key risk factors include female sex, HIV disease severity and INSTI use. Further research regarding the association between INSTIs and the development of obesity is needed.

Figures

Figure 1.

Study participant flow chart. INI, Instituto Nacional de Infectologia Evandro Chagas.

Figure 2.

Factors associated with incident obesity after multivariable analysis. Note: date of ART initiation and end of follow-up date were accounted for in multivariable analysis. AZT, zidovudine; TDF, tenofovir. *Missing values imputed in the models. **Variables included in the models using restricted cubic splines.