Abstract

Aims: To evaluate the predictive ability of 2-h post-load glucose level in addition to fasting and 1-h glucose levels in predicting the risk of type 2 diabetes.

Methods: We examined a prospective population-based cohort study of 654 subjects without type 2 diabetes at baseline. All subjects underwent an oral glucose tolerance test (OGTT), with measurement of glucose at 0, 60, and 120 min at baseline, and after 12 years in a follow-up survey. We evaluated the predictive properties of fasting, 1- and 2-h post-load glucose levels by comparing the areas under the receiver-operating characteristic (ROC) curve.

Results: We found that 2-h glucose concentration in the prediction model with fasting and 1-h glucose levels did not significantly increase the predictability of type 2 diabetes compared to a model only including fasting and 1-h glucose levels (AUC 0.83 vs. AUC 0.82, respectively; p = 0.23). The area under the ROC curve was the largest for 1-h glucose level (AUC 0.81), compared to fasting (AUC 0.71; p < 0.01) and 2-h glucose levels (AUC 0.72; p = 0.01).

Conclusions: Adding 2-h glucose to the model with fasting and 1-h glucose levels did not improve the predictability of new onset type 2 diabetes.