Abstract
AIMS/HYPOTHESIS: The IDF has proposed 1 h plasma glucose (1 h PG) as a diagnostic test for type 2 diabetes. This study evaluated the utility of 1 h PG in diagnosing type 2 diabetes, compared with fasting plasma glucose (FPG), 2 h plasma glucose (2 h PG), HbA(1c) and the combination of HbA(1c) plus FPG. METHODS: Analyses were conducted using data from five independent cohorts: KoGES, CATAMERI, GENFIEV, PLIS (follow-up) and TULIP (follow-up). Type 2 diabetes was defined according to ADA criteria (FPG ≥7.0 mmol/l [≥126 mg/dl], 2 h PG ≥11.1 mmol/l [≥200 mg/dl] or HbA(1c) ≥48 mmol/mol [≥6.5%]) or IDF criteria (1 h PG ≥11.6 mmol/l [≥209 mg/dl]). Area under of the receiver operating characteristic curves (AUC-ROCs) were used to assess the performance of 1 h PG relative to FPG and HbA(1c), individually and in combination, for diagnosing diabetes. Random-effects meta-analyses were applied to pooled data to summarise the overall diagnostic accuracy across studies. RESULTS: Cohort-specific analyses demonstrated consistently higher AUCs for 1 h PG in KoGES (AUC 0.96 vs 0.88; Δ 0.08; sensitivity 84.2 vs 77.0; specificity 98.6 vs 87.0), CATAMERI (AUC 0.98 vs 0.86; Δ 0.12; sensitivity 75.0 vs 69.4; specificity: 98.4 vs 78.9), GENFIEV (AUC 0.97 vs 0.89; Δ 0.08; sensitivity 89.5 vs 69.4; specificity 100.0 vs 88.3), PLIS follow-up (AUC 0.98 vs 0.76; Δ 0.22; sensitivity 94.9 vs 46.8; specificity 100.0 vs 92.3) and TULIP follow-up (AUC 0.98 vs 0.83; Δ 0.15; sensitivity 90.2 vs 90.2; specificity 100.0 vs 65.0) compared with FPG plus HbA(1c) (all p<0.001). Meta-analysis of five cohorts (N=11,968) revealed superior diagnostic performance of 1 h PG compared with FPG plus HbA(1c), with pooled AUCs (95% CI) of 0.97 (0.96, 0.98) vs 0.85 (0.82, 0.88). CONCLUSIONS/INTERPRETATION: These findings support the superior utility of the IDF-recommended 1 h PG vs FPG, 2 h PG, HbA(1c) and FPG plus HbA(1c) for diagnosing type 2 diabetes.