Diabetes, mortality and the burden of comorbidities in people living with HIV (PLHIV), and the effect of exposure to protease inhibitors - a single center study in Midwestern United States

Background: Patients with both HIV and diabetes might be at greater risk of metabolic and other adverse outcomes. We planned to evaluate the prevalence of various comorbidities, and mortality in PLHIV with and without diabetes, and to evaluate the association between exposure to different classes of antiretrovirals and risk of diabetes.

Methods: We conducted a retrospective cross sectional period prevalence study, using HERON research data repository at Kansas University Medical Center (KUMC), and enrolled all adults with ICD-9 diagnosis of HIV from 2000-2015. Demographics, comorbidities, and medication prescription data was abstracted from the database. The association between exposure to different classes of antiretrovirals and diagnosis of diabetes was evaluated by multivariate logistic regression. The association between time to diabetes diagnosis and time to death was evaluated using multivariate Cox regression.

Results: Of 1816 adults with HIV seen at KUMC, from 2000-2015, 80.9% were male, 56.9% were white, 9.0% were Hispanic and 59.8% were prescribed antiretrovirals. At HIV code entry, mean age was 47.5 years and mean BMI was 26.7 kg/m². The period prevalence of diabetes was 14.7% (267/1816). This prevalence increased with increasing age at HIV code entry (up to 30.4% in those >65 years). PLHIV and diabetes were more likely to have neuropathy (18.4% vs 1.7%), hypertension (76.8% vs 25.1%), chronic kidney disease (32.6% vs 8.2%), ischemic heart disease (28.1% vs 7.6%), hypothyroidism (17.2% vs 4.6%), cerebrovascular disease (24% vs 7.2%), hypercholesterolemia (23.2% vs 7.0%), vitamin D deficiency (32.2% vs 10.8%), hyperglyceridemia (14.2% vs 6.5%), Hepatitis C (16.9% vs 7.8%), depression (45.3% vs 27.2%), and COPD (38.2% vs 22.1%) compared to those without diabetes. Exposure to protease inhibitors was associated with 2.9 times increased risk of diabetes (95% CI 1.3-2.7) compared to other antiretroviral classes exposure, and 2.5 times increased risk compared to treatment naïve patients. Diabetes was associated with 2.37 times higher hazard of death (95% CI 1.55-3.63).   

Conclusion: Over a 15 year period, the prevalence of diabetes in this cohort of PLHIV seen at KUMC is significantly higher than reported in the literature, and is associated with increased morbidity and mortality. More awareness is needed among physicians.