Diabetes risk in <scp>HIV</scp> infection and antiretroviral therapy: A systematic review and meta‐analysis of prospective cohort evidence

Abstract Objectives To quantify the incidence rate (IR) of diabetes in people living with HIV (PLWH), assess disparities relative to HIV‐negative individuals and clarify the associations of antiretroviral therapy (ART) exposure and treatment duration with diabetes incidence. Methods We searched PubMed up to July 20, 2025, for prospective cohort studies including PLWH aged ≥18 with ≥6 months of follow‐up and reported diabetes incidence. Two reviewers independently assessed study quality using the Newcastle‐Ottawa Scale and extracted data. A random‐effects model estimated pooled incidence rates and subgroup differences. Meta‐regression was performed to evaluate study‐level associations between ART duration and diabetes incidence. Results Thirty‐one studies contributed 1 230 314 person‐years (PY) of follow‐up. The IR in PLWH was 12.98 cases per 1000 PY (95% CI: 10.97–14.98). No significant difference was observed between PLWH (12.97 cases per 1000 PY) and HIV‐negative individuals (12.67 cases per 1000 PY; p = 0.9597), although this comparison was based on a limited number of studies. However, incidence among PLWH on ART (14.05 cases per 1000 PY) was significantly higher than in those not on ART (7.42 cases per 1000 PY; p = 0.0179). Each year of ART exposure was associated with an increase of 0.354 cases per 1000 PY in univariable meta‐regression ( p = 0.008). In multivariable meta‐regression, the association between ART duration and diabetes incidence remained significant after adjustment for mean age ( p = 0.038), while age was also independently associated with higher incidence ( p &lt; 0.001). Conclusions Diabetes incidence in PLWH is substantial and varies across regions and diagnostic definitions. Evidence comparing PLWH with HIV‐negative individuals remains limited. ART exposure and longer ART duration are associated with higher diabetes incidence, highlighting the importance of routine metabolic screening in long‐term HIV care.