Endocrine Abstracts

Introduction: Plasma glucose concentration at 1h(1hPG) during an oral glucose tolerance test(OGTT) may be a better predictor of future diabetes mellitus(DM) than 2-h post load glucose concentration(2hPG). HIV-infected individuals may have differential risk of DM compared to the general population, and the optimal diagnosis algorithm for DM in HIV-infected persons remain unclear. We evaluated the agreement of the methods 1 hPG vs 2 hPG) for the diagnosis of pre-DM, as well as the CV risk, insulin resistance and β-cell function in patients with 1 hPG < 155 vs ≥ 155 mg/dl.

Methods: As part of a cross-sectional study, 225 Caucasian, non-institutionalized, HIV-1 infected adults under combined antiretroviral therapy (cART) were evaluated. Patients with DM diagnosis were excluded. To define pre-DM with 2 hPG we used 2019 ADA guidelines criteria; to define pre-DM in 1 hPG we used glucose ≥ 155 mg/dl. Differences between groups were tested by unpaired t-test. We used HOMA-IR, QUICKIand HOMA-β. Kappa coefficient was used to evaluate the concordance of two methods in pre-DM diagnosis.

Results: We included 225 patients (63.6% males; mean age: 45.3 ± 11.1 years) with baseline BMI: 24.7 (IQR 6.18) kg/m², waist circumference 91.0 (IQR 17.9) cm and hip circumference 94.0 (IQR 11.3) cm. The mean fasting plasma glucose (FPG) was: 91.4 ± 11.6 mg/dl, 1 hPG : 158.2 ± 43.4 mg/dl, 2 hPG : 123.1 ± 35.4 mg/dl and A1c:5.2 ± 0.4%. Patients with 1 hPG ≥ 155 mg/dl had higher FPG (95.2 mg/dl vs 87.3 mg/dl; P < 0.001), 2 hPG(138.9 ± 35.4 vs 105.9 ± 26.3; P < 0.001), A1c (5.3% vs 5.1%; P = 0.020) and C-Peptide levels (3.3 ng/ml vs 2.8 ng/ml; P = 0.039) than those with 1 hPG < 155 mg/dl but the difference in HOMA-IR, HOMA-β and QUICKI index between two groups was not significant. Patients with 2 hPG ≥ 140 mg/dl had not only higher FPG levels and at 1hPG, but also higher HOMA-IR (2.9 vs 2.1; P = 0.03) and C-Peptide levels (3.4 ng/ml vs 2.9 ng/ml; P = 0.027) than those with 2 hPG < 140 mg/dl. 52% of patients had pre-DM diagnosis accordingly 1 hPG criteria and only 27.6% had the same diagnosis with 2 hPG criteria. There was no statistically significant correlation between 1h-OGTTand lipid profile or blood pressure. The concordance correlation coefficient between the methods for diagnosis of pre-DM was 0.363, P < 0.001.

Conclusion: We observed that, in HIV-infected patients, 1 hPG criteria identified more patients with pre-DM than 2 hPG criteria. Although previous studies, performed in general population, have identified 1 hPG as a method associated not only with decreased insulin sensitivity but also with early signs of cardio-metabolic dysfunction, in our study these differences were not significant between two groups. Further investigation is needed to determine whether 1hPG should be considered as an adjunctive tool to predict dysglycemia and cardiometabolic dysfunction in setting of HIV infection.