Appendix
ESM Table 1 Baseline characteristics of the study population
Characteristic / ndata / Median [25th; 75th percentile] / Category / ncharacteristic (%)Age / 6,803 / 62 [57; 67]
Gender / 6,803 / Female / 3,866 (56.8 %)
Parental history of diabetes / 6,685 / 1,995 (29.8 %)
Diabetes in first-degree relative (adults, siblings and children) / 6,685 / 2,362 (35.3 %)
Smoking / 6,615 / Never / 3,426 (51.8 %)
Former / 2,115 (32.0 %)
Current / 1,074 (16.2 %)
Body mass index (kg/m2) / 6,798 / 26.9 [24.5; 29.7]
< 18.5 / 36 (0.5 %)
18.5 - 24.9 / 1,986 (29.2 %)
25.0 - 29.9 / 3,248 (47.8 %)
≥ 30 / 1,528 (22.5 %)
Systolic blood pressure (mmHg) / 6,711 / 140 [127; 150]
Total cholesterol (mg/dl) / 6,787 / 223 [189; 254]
HDL cholesterol (mg/dl) / 4,467 / 53 [43; 64]
Triglycerides (mg/dl) / 6,783 / 109 [77; 157]
HbA1c (%) / 6,803 / 5.6 [5.3; 5.8]
< 5.7 / 4,089 (60.1 %)
5.7 - 6.4 / 2,459 (36.1 %)
> 6.4 / 255 (3.7 %)
Fasting plasma glucose (mg/dl) / 6,803 / 90 [82; 99]
< 100 / 5,148 (75.7 %)
100 - 125 / 1,464 (21.5 %)
> 125 / 191 (2.8 %)
1
ESM Table 2 Comparison of different cut-offs for an HbA1c-defined pre-diabetes definition in the prediction of diabetes within five years and comparison to the prediction by IFG
Outcome / Lower cut-off for pre-diabetes / Pre-diabetes / No pre-diabetes / Prediction of occurrence within five yearsn a / cases / n a / cases / Sensitivity
(% [95% CI]) / Specificity
(% [95% CI]) / Positive predictive value (% [95% CI]) / Negative predictive value (% [95% CI])
Diabetes
FPG ≥ 100 mg/dl / 1,204 / 141 / 4,413 / 114 / 55.3 / [49.0; 61.5] / 80.2 / [79.1; 81.2] / 11.7 / [10.0; 13.7] / 97.4 / [96.9; 97.9]
HbA1c ≥ 5.7% / 2,023 / 172 / 3,538 / 76 / 69.4 / [63.2; 75.0] / 65.2 / [63.9; 66.4] / 8.5 / [7.3; 9.8] / 97.9 / [97.3; 98.3]
HbA1c ≥ 5.8% / 1,447 / 148 / 4,114 / 100 / 59.7 / [53.6; 65.8] / 75.6 / [74.7; 76.7] / 10.2 / [8.7; 11.9] / 97.6 / [97.1; 98.0]
HbA1c ≥ 5.9% / 1,023 / 129 / 4,538 / 119 / 52.0 / [45.8; 58.2] / 83.2 / [82.2; 84.2] / 12.6 / [10.6; 14.6] / 97.4 / [96.9; 97.8]
HbA1c ≥ 6.0% / 704 / 112 / 4,884 / 147 / 43.2 / [37.1; 49.5] / 88.9 / [88.0; 89.7] / 15.9 / [13.3; 18.8] / 97.0 / [96.5; 97.5]
a For analysis on FPG, subjects with possible diabetes by FPG levels > 125mg/dl, lost to follow-up after baseline or after 2-year follow up were excluded. For analysis on HbA1c, subjects with possible diabetes by HbA1c levels > 6.4%, lost to follow-up after baseline or after 2-year follow up were excluded.
2