Table 5

Cox proportional hazards models with all-cause dementia as the outcome variable

SCANSRUN DMC
UnivariateMultivariateUnivariateMultivariate
HR (95% CI)PHR (95% CI)PHR (95% CI)PHR (95% CI)
Model 1C=0.904 (0.022)C=0.914 (0.019)
 Apathy1.30 (1.01 to 1.68)0.0411.49 (1.05 to 2.11)0.0241.09 (1.06 to 1.12)<0.0011.05 (1.01 to 1.09)0.007
 Depression1.13 (0.70 to 1.85)0.6161.18 (0.59 to 2.37)0.6381.26 (0.89 to 1.77)0.1930.92 (0.64 to 1.31)0.629
 Age1.03 (0.99 to 1.09)0.1591.07 (1.00 to 1.14)0.0471.17 (1.12 to 1.23)<0.0011.17 (1.11 to 1.24)<0.001
 Education0.57 (0.30 to 1.07)0.080.92 (0.42 to 2.03)0.8440.43 (0.26 to 0.71)0.0010.67 (0.39 to 1.17)0.158
 PS0.10 (0.04 to 0.22)<0.0010.03 (0.01 to 0.12)<0.0010.16 (0.09 to 0.27)<0.0010.27 (0.14 to 0.54)<0.001
Model 2C=0.937 (0.027)
 Apathy1.11 (0.86 to 1.42)0.4361.53 (1.08 to 2.17)0.017
 Depression1.29 (0.67 to 2.46)0.4430.65 (0.10 to 4.38)0.654
 Age1.05 (0.97 to 1.14)0.2311.21 (1.02 to 1.45)0.033
 Education1.14 (0.71 to 1.82)0.5864.00 (1.50 to 10.70)0.006
 PS0.10 (0.04 to 0.23)<0.0010.00 (0.00 to 0.03)<0.001
  • Raw depression scores were log-transformed due to positive skew in all analyses.

  • PS, processing speed; RUN DMC, Radboud University Nijmegen Diffusion Tensor and Magnetic Resonance Cohort; SCANS, St. George's Cognition and Neuroimaging in Stroke.