Skip to main content
. 2021 Jan 10;21:38. doi: 10.1186/s12877-020-01990-1

Table 2.

Results of sex-specific competing-risks regression a, b, c, d

Men Women
Age 0.995 1.014
(0.956–1.036) (0.985–1.044)
Age2/100 1.000 1.000
(1.000–1.000) (1.000–1.000)
Years of education 0.827** 0.800**
(0.721–0.950) (0.724–0.884)
Years of education2/100 1.675 1.785#
(0.823–3.407) (0.948–3.362)
Current worker 0.980 1.010
(0.943–1.018) (0.972–1.049)
Longest occupation (Ref. Professional)
 Clerical 1.171 1.313
(0.850–1.614) (0.862–1.999)
 Manual 1.516** 1.619*
(1.142–2.013) (1.048–2.501)
 Agriculture/forestry/fishery 1.336# 1.811**
(0.967–1.845) (1.156–2.837)
 Domestic worker 2.127** 1.697*
(1.199–3.773) (1.118–2.575)
Single 0.966 1.007
(0.909–1.026) (0.977–1.038)
Home ownership 1.037 0.980
(0.978–1.100) (0.944–1.018)
Group activity 0.970 1.015
(0.935–1.007) (0.987–1.045)
Current smoker 0.979 1.017
(0.944–1.014) (0.969–1.067)
Alcohol consumption 1.011 0.958*
(0.975–1.049) (0.925–0.993)
Exercise 0.989 0.999
(0.954–1.024) (0.972–1.026)
Hearing impairment 0.971 1.045
(0.906–1.040) (0.981–1.112)
Diabetes 1.005 0.951
(0.926–1.090) (0.870–1.040)
Hypertension 0.992 0.977
(0.953–1.033) (0.948–1.007)
Stroke 0.989 1.029
(0.888–1.103) (0.919–1.151)
Observations 1958 2371

a Sub-hazard ratio with 95% confidence intervals obtained by robust standard errors in parentheses

b ** p< 0.01, * p< 0.05, # p< 0.10

c Participating waves are included in the models

d Age, working status, being single, home ownership, group activity, and health-related variables are treated as time-varying covariates by multiplying the logarithm of analysis time