To compare the incidence of dyslipidemia in people with HIV receiving integrase inhibitors (INSTI) versus boosted protease inhibitors (PI/b) and nonnucleoside reverse transcriptase inhibitors (NNRTI) within RESPOND consortium of prospective cohorts.
Participants were eligible if they were at least 18 years, without dyslipidemia and initiated or switched to a three-drug antiretroviral therapy (ART)-regimen consisting of either INSTI, NNRTI, or PI/b for the first time, between 1 January 2012 and 31 December 2018. Dyslipidemia was defined as random total cholesterol more than 240 mg/dl, HDL less than 35 mg/dl, triglyceride more than 200 mg/dl, or initiation of lipid-lowering therapy. Poisson regression was used to determine the adjusted incidence rate ratios. Follow-up was censored after 3 years or upon ART-regimen discontinuation or last lipid measurement or 31 December 2019, whichever occurred first.
Overall, 4577 people with HIV were eligible (INSTI = 66.9%, PI/b = 12.5%, and NNRTI = 20.6%), 1938 (42.3%) of whom were ART-naive. During 1.7 (interquartile range, 0.6–3.0) median years of follow-up, 1460 participants developed dyslipidemia [incidence rate: 191.6 per 1000 person-years, 95% confidence interval (CI) 182.0–201.7]. Participants taking INSTI had a lower incidence of dyslipidemia compared with those on PI/b (adjusted incidence rate ratio 0.71; CI 0.59–0.85), but higher rate compared with those on NNRTI (1.35; CI 1.15–1.58). Compared with dolutegravir, the incidence of dyslipidemia was higher with elvitegravir/cobicistat (1.20; CI 1.00–1.43) and raltegravir (1.24; CI 1.02–1.51), but lower with rilpivirine (0.77; CI 0.63–0.94).
In this large consortium of heterogeneous cohorts, dyslipidemia was less common with INSTI than with PI/b. Compared with dolutegravir, dyslipidemia was more common with elvitegravir/cobicistat and raltegravir, but less common with rilpivirine.