To analyze the association between chronic hepatitis C virus (HCV) and cytomegalovirus (CMV) infections with type 2 diabetes in HIV-infected patients.Methods:
HIV-1-infected patients enrolled in ICONA, a prospective cohort study involving 42 tertiary care centers in Italy, were selected with the following characteristics: for the diabetes incidence analysis, all patients with available CMV IgG results (first available test = baseline) and without type 2 diabetes were followed until onset of type 2 diabetes, last available clinical follow-up, death or September 30, 2014, whichever occurred first; for the prevalence analysis, all ICONA patients were analyzed at their last follow-up visit. Main outcome measures were the new onset of type 2 diabetes (incidence analysis) and the prevalence of type 2 diabetes at last follow-up.Results:
During 38,062 person-years of follow-up (PYFU) in 6505 individuals, we observed 140 cases of incident type 2 diabetes (Incidence rate 3.7, 95% CI: 3.1 to 4.3, per 1000 PYFU). In a multivariable Poisson regression model, HCV-antibody (Ab)+/HCV RNA+ patients [adjusted relative rate versus HCV-Ab negative 1.73 (95% CI: 1.08 to 2.78)] but not HCV Ab+RNA− or CMV IgG+ patients, had a higher risk of diabetes. Among 12,001 patients, 306 (2.5%) prevalent cases of type 2 diabetes were detected. HCV Ab+RNA+ status was independently associated with prevalent diabetes (adjusted Odds Ratio vs HCV Ab− 2.49; 95% CI: 1.08 to 5.74), whereas HCV-Ab+/HCV RNA− and CMV IgG+ status were not.Conclusion:
In HIV-infected individuals, active HCV replication but not prior HCV exposure or latent CMV infection is associated with incident and prevalent type 2 diabetes.