Background: Type 2 diabetes (T2D) is among the most frequent comorbidities in people with HIV (PWH) and occurs more often in PWH than in people without HIV. Polygenic risk scores (PRS) can be used to summarize the genetic risk for T2D, but it is unknown to what extent HIV-specific factors impact on or interact with genetic risk factors.

Methods: We performed a case control study using incidence density sampling to match participants with T2D to controls within the Swiss HIV Cohort Study (mean age 51.9, 73.8% male). Conditional logistic regression was used to evaluate risk factors for T2D, including two PRSs (for T2D and BMI), adjusting for age, sex, ethnicity, BMI, waist-hip ratio, ART-regimen, and CD4-cell count. Prediction accuracy was assessed by leave-one-out cross-validation and computing the area under the receiver operator characteristic (ROC) curve.

Results: Analyzing 687 cases and 2441 controls, we found a clear dose-response relationship between the PRS for T2D and T2D, with participants in the highest PRS-quintile having an over three-times larger risk of T2D (odds ratio, OR [95%-CI], 3.41 [2.61, 5.11]) compared to the first quintile. CD4-cell count at T2D diagnosis was not associated with T2D (1.01 [0.94, 1.09]). Compared to an NNRTI-based ART-regimen, an INSTI-based regimen was associated with an increased T2D-risk (2.45 [1.68, 3.58]), as was a PI-based regimen (1.70 [1.22, 2.37]).

Conclusion: We found that CD4-cell counts are not associated with the risk of T2D, and that neither CD4-cell count nor cumulative exposure to NNRTI or INSTI-class antiretrovirals modify this risk.