HAM-D17 Cut-off value | ROC-AUC | 95% Confidence Interval | Selected Scores of HRQOL |
---|
1 | 0.868 | 0.813–0.942 | BP, VT |
2 | 0.854 | 0.799–0.909 | BP,VT,SF,RE |
3 | 0.828 | 0.779–0.877 | VT,SF,RE |
4 | 0.848 | 0.808–0.887 | BP,SF,RE,MH |
5 | 0.864 | 0.831–0.898 | PF,BP,VT,SF,RE,MH |
6 | 0.852 | 0.819–0.885 | PF,VT,SF,RE,MH |
7 | 0.837 | 0.803–0.870 | PF,VT,SF,RE,MH |
8 | 0.844 | 0.813–0.875 | PF,BP,VT,SF,RE,MH |
9 | 0.837 | 0.806–0.868 | PF,BP,VT,SF, RE,MH |
10 | 0.834 | 0.803–0.864 | PF,BP,SF,RE,MH |
11 | 0.836 | 0.806–0.866 | PF,BP,VT,SF,RE,MH |
12 | 0.830 | 0.800–0.860 | PF,BP,SF,RE,MH |
13 | 0.820 | 0.789–0.850 | PF,BP,SF,RE,MH |
14 | 0.825 | 0.795–0.855 | PF,BP,SF,RE,MH |
15 | 0.820 | 0.790–0.850 | PF,BP,RE,MH |
16 | 0.813 | 0.782–0.843 | PF,BP,SF,RE,MH |
17 | 0.794 | 0.762–0.826 | PF,BP,SF,RE,MH |
18 | 0.790 | 0.758–0.822 | PF,BP,RE |
19 | 0.793 | 0.761–0.825 | PF,BP,SF,RE |
20 | 0.789 | 0.756–0.823 | PF,SF,RE |
21 | 0.792 | 0.758–0.826 | PF,SF,RE |
22 | 0.779 | 0.732–0.817 | PF,RE |
23 | 0.773 | 0.733–0.813 | |
- Abbreviation: HRQOL health-related quality of life, PF physical functioning, RP role physical, BP bodily pain, GH general health perceptions, VT vitality, SF social functioning, RE role-emotional, MH mental health
- The multivariate logistic regression with the HAM-D17 score as a dependent variable and with each of the eight domain scores of HRQOL (SF-36) as independent variables was conducted, and the maximal ROC-AUC of 0.864 was obtained. In addition, based on the cutoff value, by converting the HAM-D17 score into the binominal form of “0 or 1,” a multivariate analysis with HAM-D17 as a dependent variable and with each of the eight HRQOL domain scores (SF-36) as binominalized independent variables was performed. Then, the significantly (p < 0.05) associated scores of the eight HRQOL domains (SF-36) were identified. Specifically, after converting this cutoff score of HAM-D17 into the binominal form of “0 or 1” for HAM-D17 scores of ≥5 and ≤4, the domain scores of PF, BP, VT, SF, RE, and MH are significantly negatively correlated with HAM-D17