Modelling of OGTT curve identifies 1 h plasma glucose level as a strong predictor of incident type 2 diabetes: results from two prospective cohorts

The relevance of the OGTT in predicting type 2 diabetes is unclear. We assessed the performance of 14 OGTT glucose traits in type 2 diabetes prediction.

We studied 2,603 and 2,386 Europeans from the Botnia study and Malmö Prevention Project (MPP) cohorts with baseline OGTT data. Over a follow-up period of 4.94 years and 23.5 years, 155 (5.95%) and 467 (19.57%) participants, respectively, developed type 2 diabetes. The main outcome was incident type 2 diabetes.

One-hour plasma glucose (1h-PG) was a fair/good predictor of incident type 2 diabetes in the Botnia study and MPP (AUC for receiver operating characteristic [AUC_ROC] 0.80 [0.77, 0.84] and 0.70 [0.68, 0.73]). 1h-PG alone outperformed the prediction model of multiple clinical risk factors (age, sex, BMI, family history of type 2 diabetes) in the Botnia study and MPP (AUC_ROC 0.75 [0.72, 0.79] and 0.67 [0.64, 0.70]). The same clinical risk factors added to 1h-PG modestly increased prediction for incident type 2 diabetes (Botnia, AUC_ROC 0.83 [0.80, 0.86]; MPP, AUC_ROC 0.74 [0.72, 0.77]). 1h-PG also outperformed HbA_1c in predicting type 2 diabetes in the Botnia cohort. A 1h-PG value of 8.9 mmol/l and 8.4 mmol/l was the optimal cut-point for initial screening and selection of high-risk individuals in the Botnia study and MPP, respectively, and represented 30% and 37% of all participants in these cohorts. High-risk individuals had a substantially increased risk of incident type 2 diabetes (OR 8.0 [5.5, 11.6] and 3.8 [3.1, 4.7]) and captured 75% and 62% of all incident type 2 diabetes in the Botnia study and MPP.

1h-PG is a valuable prediction tool for identifying adults at risk for future type 2 diabetes.