Skip to main content

Table 2 Added value of health indicators to mortality prediction based on age and sex

From: Do changes in frailty, physical functioning, and cognitive functioning predict mortality in old age? Results from the Longitudinal Aging Study Amsterdam

Model

Health Indicator

Five-year Mortality

(220 died)

Ten-year Mortality

(466 died)

AUCa (95% CI)

P-valueb

AUCa (95% CI)

P-valueb

Demographic variables

Age and sex

72.8 (69.0 – 76.5)

 

77.7 (74.8 – 80.5)

 

 + Current status

 + Frailty Phenotype

 + 1.6 (0.6 – 2.6)

0.001

 + 0.8 (0.0 – 1.6)

0.050

 + Frailty Index

 + 3.3 (1.6 – 4.9)

 < 0.001

 + 1.8 (0.5 – 3.1)

0.005

 + Grip strength

 + 1.0 (0.3 – 1.7)

0.005

 + 0.6 (0.0 – 1.2)

0.040

 + Walking speed

 + 0.7 (-0.0 – 1.4)

0.060

 + 0.4 (-0.1 – 0.8)

0.100

 + MMSE

 + 1.2 (0.2 – 2.3)

0.020

 + 1.1 (0.4 – 1.8)

0.002

 + Three-year change

 + Frailty Phenotype

- 0.0 (-0.2 – 0.1)

0.400

- 0.1 (-0.2 – 0.0)

0.200

 + Frailty Index

 + 1.1 (0.1 – 2.1)

0.030

 + 0.5 (-0.3 – 1.3)

0.200

 + Grip strength

 + 0.3 (-0.0 – 0.7)

0.090

 + 0.2 (-0.1 – 0.5)

0.100

 + Walking speed

 + 0.4 (-0.1 – 0.9)

0.090

 + 0.1 (-0.2 – 0.4)

0.500

 + MMSE

 + 0.1 (-0.3 – 0.6)

0.600

- 0.0 (-0.4 – 0.3)

0.900

 + Current status + Three-year change

 + Frailty Phenotype

 + 1.9 (0.6 – 3.1)

0.003

 + 0.9 (0.0 – 1.9)

0.050

 + Frailty Index

 + 3.2 (1.6 – 4.8)

 < 0.001

 + 1.7 (0.5 – 3.0)

0.008

 + Grip strength

 + 0.9 (0.1 – 1.6)

0.020

 + 0.5 (-0.1 – 1.1)

0.100

 + Walking speed

 + 0.6 (-0.2 – 1.3)

0.100

 + 0.4 (-0.2 – 0.9)

0.200

 + MMSE

 + 1.2 (0.2 – 2.3)

0.020

 + 1.2 (0.5 – 1.9)

0.001

  1. aAUC Discrimination ability as measured using Area Under the Curve in percentages
  2. bP-values indicate the significance of the difference between a model containing age and sex plus a health indicator and the reference model that is age and sex only given the null hypothesis that the health indicator has no added discrimination ability