Table 1.
Predictability of individual variables for death.
Table 2.
Composition of the KLoSHA frailty index.
Table 3.
Comparisons of demographic, anthropometric, and laboratory data in examinees between dead or alive during the follow-up period.
Figure 1.
The distributions of each KFI-component variable by KFI and fractional polynomial prediction plots.
The shaded area denotes a 95% confidence interval for the prediction curve. (A) 12-SPPB score (B) K-ADL score (C) K-IADL score (D) K-MMSE score (E) albumin (g/dL).
Figure 2.
The distributions of KFI by age and fractional polynomial prediction plots plotted by sex, (A) female, (B) male.
The shaded area denotes a 95% confidence interval for the prediction curve.
Table 4.
Unadjusted and adjusted hazard ratio for the mortality according to the frailty status from each frailty index.
Table 5.
Comparisons of predictability for mortality by frailty status from each frailty index.
Figure 3.
ROC curves for (A) functional decline, (B) hospitalization.
(A) The area under the curve is 0.937 for KFI, 0.704 for CHS, and 0.565 for SOF; for the difference of each ROC curve, p = 0.001 for KFI and CHS, p<0.001 for KFI and SOF, p = 0.069 for CHS and SOF. (B) The area under the curve is 0.543 for KFS, 0.560 for CHS, and 0.505 for SOF; for the difference of each ROC curve, p = 0.639 for KFI and CHS, p = 0.374 for KFI and SOF, p = 0.180 for CHS and SOF.
Table 6.
Impact of frailty status from each frailty index on subsequent functional decline.
Table 7.
Impact of frailty status from each frailty index on following experience of hospitalization.