Table 1—

Baseline characteristics according to serum creatinine levels and multivariate logistic regression models of the incidence of type 2 diabetes

TotalSerum creatinine (mg/dl)
0.40–0.600.61–0.700.71–0.800.81–0.900.91–1.60
n8,5706992,1013,0461,837887
Age (years)47.8 ± 4.248.1 ± 4.047.7 ± 4.247.7 ± 4.247.7 ± 4.148.0 ± 4.1
Height (cm)168.9 ± 5.6167.6 ± 5.7168.6 ± 5.7169.1 ± 5.5169.2 ± 5.4169.4 ± 5.6
Weight (kg)66.8 ± 9.463.9 ± 9.265.3 ± 9.767.0 ± 9.068.1 ± 9.169.6 ± 9.3
BMI (kg/m2)23.4 ± 2.922.7 ± 2.923.0 ± 3.023.4 ± 2.823.8 ± 2.724.2 ± 2.9
Fasting plasma glucose (mg/dl)97.6 ± 9.597.7 ± 9.897.3 ± 9.497.6 ± 9.597.8 ± 9.298.1 ± 9.5
Daily alcohol consumption (g ethanol)25.2 ± 21.929.0 ± 24.227.4 ± 22.625.2 ± 21.723.7 ± 21.320.2 ± 19.5
Drinking habit categories
    Nondrinkers15.514.713.315.616.818.6
    Light drinkers19.315.618.018.920.424.5
    Moderate drinkers35.033.935.634.435.336.1
    Heavy drinkers30.135.833.131.127.520.9
Smoking habit
    Nonsmokers21.512.718.221.324.929.5
    Past smokers22.414.317.822.727.129.2
    Current smokers56.173.064.056.048.041.3
Walk to work
    0–10 min19.818.018.218.922.023.3
    11–20 min52.452.152.254.649.052.2
    ≥21 min27.929.929.626.529.024.5
Regular leisure-time physical activity18.110.015.817.921.723.1
Family history of diabetes12.912.712.612.913.612.1
Incidence of type 2 diabetes877 (10.2)103 (14.7)223 (10.6)283 (9.3)176 (9.6)92 (10.4)
Total model
    Crude odds ratio (95% CI)1.69 (1.32–2.15)1.16 (0.96–1.40)1.001.04 (0.85–1.26)1.13 (0.88–1.45)
    Multiple-adjusted odds ratio (95% CI)*1.91 (1.44–2.54)1.32 (1.07–1.63)1.001.04 (0.83–1.30)1.01 (0.76–1.35)
Stratified analysis according to median BMI
    BMI ≤23.31 kg/m24,286
Crude odds ratio (95% CI)1.90 (1.31–2.76)1.27 (0.94–1.71)1.001.11 (0.79–1.57)0.99 (0.61–1.62)
        Multiple-adjusted odds ratio (95% CI)*1.96 (1.28–2.98)1.37 (0.98–1.91)1.001.13 (0.78–1.65)0.91 (0.53–1.56)
    BMI >23.31 kg/m24,284
Crude odds ratio (95% CI)1.78 (1.28–2.46)1.20 (0.94–1.52)1.000.95 (0.75–1.22)1.06 (0.79–1.42)
        Multiple-adjusted odds ratio (95% CI)*1.85 (1.26–2.73)1.29 (0.98–1.71)1.000.99 (0.75–1.32)1.05 (0.75–1.48)
  • Data are means ± SD, %, or n (%) unless otherwise indicated. In multiple logistic regression analysis, age, daily alcohol consumption, fasting plasma glucose, and serum creatinine showed a nonlinear association with the incidence of type 2 diabetes in all models. Therefore, we fit the models by using these variables categorized for easy understanding.

  • *

    * The multiple logistic regression model was adjusted for age (40–44, 45–49, or 50–55 years), BMI, fasting plasma glucose (<100, 100–109, or 110–125 mg/dl), daily alcohol consumption (nondrinkers, light drinkers [0.1–16.3 g ethanol/day], moderate drinkers [16.4–42.6 g ethanol/day], or heavy drinkers [42.7–115.0 g ethanol/day]), smoking habit (nonsmokers, past smokers, or current smokers), the duration of the walk to work (0–10, 11–20, or ≥21 min), regular leisure-time physical activity, and family history of diabetes.