Table 3.
AUC | SE | Sig | CI (95%) | Cut‐ offa | Sensitivity | Specificity | |||
---|---|---|---|---|---|---|---|---|---|
Lower Bound | Upper Bound | ||||||||
MCI vs. HS | MMSE | 0.906 | 0.03 | 0.0001b | 0.847 | 0.965 | 27.5 | 0.82 | 0.984 |
25 (OH)D3 | 0.765 | 0.04 | 0.0001b | 0.681 | 0.849 | 27.5 | 0.902 | 0.541 | |
1,25(OH)2 D3 | 0.5 | 0.05 | 1.000 | 0.396 | 0.604 | — | — | — | |
Mild AD vs. HS | MMSE | 0.993 | 0.01 | 0.0001b | 0.979 | 1 | 27.5 | 1 | 0.984 |
25 (OH)D3 | 0.815 | 0.04 | 0.0001b | 0.736 | 0.895 | 25.5 | 0.927 | 0.574 | |
1,25(OH)2 D3 | 0.41 | 0.06 | 0.124 | 0.293 | 0.527 | — | — | — | |
Moderate AD vs. HS | MMSE | 1 | 0 | 0.0001b | 1 | 1 | 22 | 1 | 1 |
25 (OH)D3 | 0.812 | 0.04 | 0.0001b | 0.725 | 0.899 | 27.5 | 0.943 | 0.541 | |
1,25(OH)2 D3 | 0.472 | 0.06 | 0.645 | 0.351 | 0.593 | — | — | — | |
Severe AD vs. HS | MMSE | 1 | 0 | 0.0001b | 1 | 1 | 17 | 1 | 1 |
25 (OH)D3 | 0.911 | 0.03 | 0.0001b | 0.857 | 0.966 | 20.5 | 0.906 | 0.787 | |
1,25(OH)2 D3 | 0.493 | 0.07 | 0.913 | 0.367 | 0.62 | — | — | — | |
Mild AD vs. Moderate AD | MMSE | 1 | 0 | 0.0001b | 1 | 1 | 20.5 | 1 | 1 |
25 (OH)D3 | 0.494 | 0.07 | 0.929 | 0.363 | 0.626 | — | — | — | |
1,25(OH)2 D3 | 0.566 | 0.07 | 0.327 | 0.436 | 0.695 | — | — | — | |
Severe AD vs. Mild AD | MMSE | 1 | 0 | 0.0001b | 0 | 0 | 15.5 | 1 | 1 |
25 (OH)D3 | 0.661 | 0.06 | 0.019b | 0.538 | 0.785 | 21.5 | 0.341 | 0.969 | |
1,25(OH)2 D3 | 0.566 | 0.07 | 0.333 | 0.433 | 0.7 | — | — | — | |
Severe AD vs. Moderate AD | MMSE | 1 | 0 | 0.0001b | 1 | 1 | 10.5 | 1 | 1 |
25 (OH)D3 | 0.675 | 0.07 | 0.014b | 0.546 | 0.804 | 21.5 | 0.314 | 0.969 | |
1,25(OH)2 D3 | 0.515 | 0.07 | 0.836 | 0.374 | 0.655 | — | — | — | |
Total AD vs. HS | MMSE | 0.997 | 0.003 | 0.0001b | 0.992 | 1 | 27.5 | 1 | 0.984 |
25 (OH)D3 | 0.843 | 0.03 | 0.0001b | 0.782 | 0.904 | 20.5 | 0.741 | 0.787 | |
1,25(OH)2 D3 | 0.455 | 0.05 | 0.327 | 0.366 | 0.543 | — | — | — |
AUC, area under the ROC curve; CI, confidence interval; SE, standard error of AUC.
Cutoff values at which optimal balance of sensitivity and specificity can be obtained according to Youden's index; Youden's Index can be calculated as the sum of sensitivity plus specificity minus 1 for all possible cutoff points.
Sig, significance of AUC.