Newcastle-Ottawa quality assessment scale for studies meeting inclusion criteria.

A score of 0 indicates “no”; a score of 1 or 2 indicates “yes”. Studies scoring ≥67% were considered high methodological quality, 34%–66% were considered moderate methodological quality, and ≤33% were considered low methodological quality. Given that the distribution of possible confounders in randomised controlled trials is related to chance alone, randomised controlled trials were not assessed for confounding.ART, antiretroviral therapy; BMI, body mass index; NA, not applicable.