Eleven-year incident glucose disorders in treated HIV-infection. The St Vincentʼs HIV and Diabetes Study

OBJECTIVE: To determine the long-term incidence of glucose disorders in treated HIV infection, associations with traditional and HIV-specific risk factors. METHODS: Observational cohort of 104 men with treated HIV infection and without diabetes, aged 43 ± 8 years at baseline, with (mean ± SD) 11.8 ± 3.5 years follow-up. Ascertainment of glucose status by fasting glucose or, in a subset (n = 33), a 75 g oral glucose tolerance test by 10-12 years follow-up. A subset underwent sequential body composition measures (n = 58) to determine changes in total body and central abdominal adiposity. RESULTS: The cumulative incidence of glucose disorders was 48.1% (prediabetes 35.6%, diabetes 12.5%), with an incidence rate of 34.5/1000 years of patient follow-up (PYFU) (prediabetes: 24.3/1000 PYFU; diabetes: 10.2/1000 PYFU). Incident glucose disorders were independently associated with higher age (44.9 ± 8.4 vs. 41.1 ± 7.5 years, P = 0.027), baseline C-peptide (2.9 ± 1.3 vs. 2.4 ± 1.1 ng/ml, P = 0.019) and baseline 2-h glucose (135 ± 41 vs. 95 ± 25 mg/dl, P < 0.001). A prior AIDS-defining illness was independently associated with higher follow-up fasting glucose (108 ± 38 vs. 94 ± 16 mg/dl, P = 0.007). Abdominal fat gain over 2-4 years was associated with a 3.16-fold increased risk of incident glucose disorders (95% CI 1.30-7.68, P = 0.011). In a subgroup who underwent further oral glucose tolerance testing, 60% had a glucose disorder, the majority not detected by fasting glucose. CONCLUSION: Men with long-term treated HIV infection have high rates of incident glucose disorders associated with modest abdominal fat gain. Directed measures to prevent diabetes in this population are warranted.