Table 3 —

Sensitivity, specificity, PSEP, and shrinkage analysis for measures of insulin sensitivity (expressed Mffm/I) using three different methods

CutoffSensitivitySpecificityPSEP
Method A
 Insulin12.2 mU/l0.570.820.42
 BMI29.3 kg/m20.560.760.33
 TAG1.5 mmol/l0.530.750.29
 Insulin, BMI, and TAG0.630.820.46
  Validation estimates0.640.810.46
  95% CI0.53–0.730.74–0.880.40–0.52
 Insulin and TAG0.620.840.40
  Validation estimates0.640.830.49
  95% CI0.53–0.730.74–0.880.44–0.52
Method B
 Insulin12.1 mU/l0.560.840.45
 BMI30.0 kg/m20.520.780.31
 TAG1.6 mmol/l0.530.790.34
 Insulin, BMI, and TAG0.610.850.50
  Validation estimates0.450.910.50
  95% CI0.32–0.60.85–0.960.45–0.55
 Insulin and TAG0.610.840.49
  Validation estimates0.490.920.51
  95% CI0.32–0.60.86–0.950.47–0.54
Method C
 Insulin12.0 mU/l0.710.690.39
 BMI30.0 kg/m20.690.640.33
 TAG1.5 mmol/l0.630.580.20
 Insulin, BMI, and TAG0.650.790.45
  Validation estimates0.520.83
  95% CI0.0–0.650.72–1.0
  • Method A, ISI and predictors analyzed continuously; method B, categorical ISI (≤6.3 M · mU−1 · l−1) and continuous predictors; method C, categorical ISI and predictors. Cutoffs have been derived from the equations for each variable.