Table 2.
Cox Proportional Hazard Model of 15-Year Mortality based on Nine Early Frailty Transition Groups (1995–2013, N=1,171).
| Demographic characteristics | 15-Year Mortality HR(95%CI) |
|---|---|
| Age Group (years) | |
| 65-74 | Ref. |
| 75-84 | 1.90(1.61, 2.25)* |
| 85+ | 3.39(2.63, 4.37)* |
|
| |
| Gender | |
| Male | Ref. |
| Female | 0.69(0.60, 0.82)* |
|
| |
| Education (years) | |
| 0 | Ref. |
| 1-7 | 1.04(0.83, 1.29) |
| 8-11 | 1.34(1.01, 1.80)* |
| 12+ | 1.24(0.91, 1.69) |
|
| |
| Marital Status | |
| Unmarried | Ref. |
| Married | 0.87(0.74, 1.02) |
|
| |
| BMI Group | |
| 18.5-<25 | Ref. |
| <18.5 | 1.56(0.93, 2.63) |
| 25-<30 | 0.82(0.68, 0.98)* |
| ≥30 | 0.74(0.61, 0.90)* |
|
| |
| Comorbidity | |
| No | Ref. |
| Yes | 1.21(1.12,1.30)* |
|
| |
| Cognitive Impairment (MMSE<21) | |
| No | Ref. |
| Yes | 1.25(1.06, 1.48)* |
|
| |
| ADL Disabledˆ | |
| No | Ref. |
| Yes | 1.47(1.14, 1.91)* |
|
| |
| Smoking | |
| No | Ref. |
| Yes | 1.63(1.30,2.04)* |
| Nine Frailty Transition& | |
| Nonfrail-nonfrail (N=357) | Ref. |
| Nonfrail-prefrail (N=222) | 1.19(0.96, 1.47) |
| Nonfrail-frail (N=98) | 1.18(0.89, 1.56) |
| Prefrail-nonfrail (N=164) | 1.05(0.83, 1.34) |
| Prefrail-prefrail (N=140) | 1.20(0.93, 1.55) |
| Prefrail-frail (N=74) | 1.68(1.23, 2.28)* |
| Frail-nonfrail (N=27) | 0.96(0.53, 1.72) |
| Frail-prefrail (N=41) | 1.54(1.05, 2.28)* |
| Frail-frail (N=48) | 1.72(1.21, 2.44)* |
95% CI is significant
In Cox regression analysis, there was no violation of the proportionality assumption assessed by the significance of a term of the predictor associated with the logarithm of survival time. The HRs were controlled for the following factors: socio-demographics, BMI, ADL disability, cognitive impairment and comorbidities
ADL Disabled was defined as help needed or unable to perform ≥1 of the seven ADL activities from Katz ADL scale (walking, bathing, grooming dressing, eating, transferring, and toileting)34.
Abbreviations: HR (Hazard Ratio), CI (Confidence Interval), BMI (Body Mass Index), MMSE (Mini-Mental State Examination), ADL (Activity of Daily Living), SD (Standard Deviation).