ABSTRACTPeople with HIV (PWH) have an increased risk for diabetes mellitus. Our objectives were to characterize the prevalence and incidence of diabetes in a cohort of people with HIV (PWH), and to evaluate both traditional and HIV-specific risk factors contributing to incident diabetes diagnoses. We conducted a retrospective study of a Southeastern US academic HIV clinic. All PWH age > 18 years of age who attended at least two clinic visits between 2008 and 2018 were evaluated to assess time

more »

... diabetes incidence. Laboratory, demographic, clinical, medication and diagnoses data were extracted from the Clinic EMR. Diabetes was defined when at least two of the following three criteria were met: (1) laboratory data consistent with a diagnosis as defined by the ADA SOC (Hgb A1C ≥ 6.5% and/or 2 glucose results >200 mg/dl (at least 30 days apart)), (2) diagnosis of diabetes in the EMR, or (3) exposure to diabetes medication. Time to Diabetes incidence was computed from the entire clinic population for each study year. Univariate Cox proportional hazard models were developed to evaluate associations between each baseline factor and time to DM. Multivariable Cox proportional hazard regression models with time-dependent covariates were created to evaluate the independent association between significant parameters from univariate analyses and time to incident DM. From 4113 PWH included in the analysis, we identified 252 incident cases of diabetes. In multivariable analysis, BMI classification, liver disease, steroid exposure, and use of Integrase Inhibitors were associated with incident diabetes. Additional associated factors included lower CD4 cell counts, duration of HIV infection, exposure to non-statin lipid lowering therapy, and dyslipidemia. Incident diabetes rates are increasing at an alarming rate among PWH. Diabetes prevalence increased from 8.8% in 2008 to 14% in 2018. Both traditional and HIV-related risk factors, particularly body weight, steroid exposure, and use of Integrase Inhibitors, were associated with incident diabetes. Notably, several of the risk factors identified are modifiable and should be targeted for intervention.