Page 40 - IJPS-1-1

Concordance and discordance in successful aging measures and mortality

Table 1. Sample distribution of variables in the study, CLHLS 2000–2008
Women Men
# of individuals 18,029 12,919
# of observations (2000–2008) 28,130 21,088
2000 5,536 4,029
2002 7,831 5,887
2005 7,012 5,337
2008 7,751 5,835
# of individuals died between the 2000 and 2011 waves 11,594 7,670

% OSA a 6.32 14.21
% SSA a 21.13 27.71
Concordance & discordance a
% Not-OSA, Not-SSA (Type I) 74.95 64.40
% Not-OSA, SSA (Type II) 18.73 21.39
% OSA, Not-SSA (Type III) 3.92 7.89
% OSA, SSA (Type IV) 2.40 6.32
Covariates a
% Ages 65-79 17.99 27.61
% Ages 80-89 23.34 33.60
% Ages 90-99 27.43 27.80
% Ages 100+ 31.24 10.98
% Urban 43.16 45.00
% Han ethnicity 93.57 94.08
% Currently married 15.39 47.54
% Coresidence with children 74.31 59.79
% Years of schooling (0) 84.75 37.94
% Years of schooling (1–6) 12.45 45.86
% Years of schooling (7+) 2.90 16.19
% White collar occupation 2.27 12.85
% Economic independence 14.17 40.34
% Current smoking 6.89 34.57
% Current alcohol consumption 11.71 32.19
% Doing regular exercise 22.55 37.70
Note: a, The percentage distribution was based on the pooled dataset among # of individuals from 2000 to 2008.

for men), rather than in objectively and subjectively good health (2.4% for women and 6.3%
for men).
Figures 2 and 3 present the Kaplan-Meier survival functions by age group for women
and men, respectively. It is apparent that those who were both OSA and SSA (i.e., Type IV)
have the highest survival probability across four age groups for both women and men,
while those who were both not-OSA and not-SSA (i.e., Type I) had the lowest survival
probability. For those who were OSA, the survival difference between subgroups of SSA
and not-SSA (i.e., Type IV versus Type III) in most cases is relatively small; for those who
were SSA, the survival difference between its subgroups of OSA and not-OSA (i.e., Type
IV versus Type II) is relatively pronounced. These results indicate that OSA likely has a
greater power in predicting subsequent mortality.

International Journal of Population Studies | 2015, Volume 1, Issue 1 34