The mean of fasting, 1-h, and 2-h plasma glucose levels is superior to each separate index in predicting diabetes

Abstract

AimsThe fasting, 1-h, and 2-h plasma glucose (PG) levels during oral glucose tolerance test represent different glucose metabolic functions. We examined whether averaging these PG indices (GLUM0.60.120) results in a better predictor of future type 2 diabetes (T2DM). Methods7533 participants were followed up biannually for 12 years. Hazard ratios (HRs), area under the curve (AUC) of the receiver-operating characteristic, and the net reclassification index (NRI) for T2DM were calculated to compare the discriminative ability of GLUM0.60.120 versus other PG indices. ResultsThe adjusted HRs and 95% confidence intervals for an increase in SD of GLUM0.60.120 was 2.50 (2.36–2.65) and 1.88 (1.73–2.04) in T2DM-free and normal glucose tolerance (NGT) participants, respectively. The AUC of GLUM0.60.120 was higher than that of fasting PG, 1-h, and 2-h PG values for T2DM-free (0.79 versus 0.67, 0.77, and 0.73) and NGT (0.73 versus 0.65, 0.72, and 0.61). The model using GLUM0.60.120 improved the classification of the models with fasting PG, 1-h, and 2-h PG values (NRI: 0.369, 0.272, and 0.282 for T2DM-free and 0.249, 0.131, and 0.351 for NGT participants with all p < 0.001). ConclusionsThe mean of fasting, 1-h, and 2-h PG levels predicts future T2DM better than each index.