Selfrated Health and Mortality Risk in the Oldest Old in China Chinese Longitudinal Healthy Longevit

1
Self-rated Health and Mortality Risk in the
Oldest Old in China Chinese Longitudinal Healthy
Longevity Study

• Liu Yuzhi, Li Qiang
• Institute of Population Research,
  Center for Healthy Aging
• and Family Studies, Peking University,
  Beijing, China

2
Introduction
Risk factors
e.g. Socio-demographic
characteristics Engagement with life
Functional ability Disease burden
3
Introduction

• Many studies have suggested that self-rated
  health (SRH) is significantly associated with
  mortality risk in young people, adults and the
  young old (e.g. Benyamini Idler, 1999 Idler
  Benyamini, 1997). Moreover, early research have
  proposed that self-rated health reflect the
  effects of disease burden and functional ability
  on the mortality.
• Several research have pointed out that that older
  persons reported weaker associations between
  self-rated health and mortality compared with
  younger (Ilder Angel, 1990 Strawbridge
  Wallhagen, 1999).
• Meanwhile, previous studies have proposed that
  the oldest old (Fourth Age) is very different
  from the young old (Third Age e.g. Baltes
  Smith, 2003 Maier and Smith, 1999 Suzman, et
  al., 1992 Zeng et al., 2002).
• However, few studies investigated the
  relationship between self-rated health and
  mortality in the oldest old (Cai, 2004).

4
Research Questions

• Is there significant association between
  self-rated health and mortality risk in the
  Chinese oldest old?
• Is there the gender differences of the
  predictability of self-rated health for
  mortality risk?
• Is missing values (as the additional level) in
  self-rated health related to higher mortality
  risk?

5
Sample
Three waves of Chinese Longitudinal Healthy
Longevity Study. N7623 60.03 women 3 Age
cohort 8089 (n2962) 9099(n2651)100105
(n2010)
6
Models

• Hazard model with piecewise linear baseline
  intensity

the logarithm of the force of mortality at
attained age x
the log-hazard baseline with nodes at 85, 90, 95,
and 100 years of age
Self-rated health, the only time-varying
variable in the model
the coefficients of self-rated health
covariates in the model
coefficients of covariates.
7
Measures
Self-rated health How do you rate your health
at present?
Central Variable
Sex Type of residence Years of schooling
Occupation before 60 years old Marital Status
Caregivers while serious illness.
Socio-demographic characteristics
Engagement with life Number of activities
including housework,
reading, watch TV, religious
activities et al.
Covariates
Self-reported ADL MMSE Vision Hearing Bedridde
n.
Functional ability
Self-reported Number of illness
8
Results
Piecewise-Linear Baseline Log-Harzard of
Mortality in the Oldest Old
Log-Hazard
Age
9
Results
Mortality risk associated with Self-rated health
in the oldest old
Note plt. 1 plt.05 plt.01.
10
Results

Self-rated Health
Mortality risk
Risk factors
e.g. Socio-demographic
characteristics Engagement with life
Functional ability Number of illness
11
Results
Mortality risk associated with self-rated health
by gender
Note plt.05 plt.01 plt.001.
Gender Pattern is similar as the total pattern.
No significant gender difference was found in our
sample.
12
Discussion

• Self-rated health is a significant predictor of
  mortality risk in the oldest old in China with
  controls for socio-demographic characteristics,
  engagement with life, functional ability, and
  number of illness! Engagement with life and
  functional ability partly explain the
  relationship between self-rated health and
  mortality risk in the oldest old. The impact of
  number of illness can be negligible.
• No gender difference on the relationship between
  self-rated health and mortality risk is founded
  in our sample.
• Missing values in self-rated health is highly
  associated with mortality risk. And more,
  participants with incomplete data in the
  self-rated health have the highest mortality
  risks.

13
Follow-up Analyses

• Covariates time-varying variables.

14
Thank you