Page 82 - IJPS-2-2

Self-rated health and interviewer-rated health: differentials in predictive power for mortality among subgroups of Chinese elders

where rural/urban socioeconomic inequality is substantial. Some studies reported substantial differ-
ences between rural and urban residents in China in terms of cognitive procedure behind
self-reported disability and the measure’s power to predict mortality (Purser, Feng, Zeng et al., 2012;
Feng, Hoenig, Gu et al., 2010).
Gender differentials have long been found as one major source of health and mortality disparity in
old age (e.g., Arber and Cooper, 1999; Case and Paxson, 2005; Kaneda, Zimmer, Fang et al., 2009;
Verbrugge, 1985). According to the review by Deeg and Bath (2003), gender differences in elderly
health have multiple manifestations: women and men suffer different health problems in later life;
given a particular health problem, women tend to develop functional limitations while men are more
likely to die; elderly women and men do not perceive health in the same manner due to the different
social conditions such as marital status and SES; and women and men may differ in their sensitivity
to physical symptoms of illness, and thus they may rate their health differently even if they have the
same illnesses. Health behaviors between women and men are also distinctive. Women are more
willing and have a greater motivation to engage with health-related information (Stefan, 2013).
Women also tend to have a different mortality trajectory and different trend in disability over time
than men do (Kaneda, Zimmer, Fang et al., 2009; Zimmer, Hidajat, and Saito, 2015). In self-repor-
ting health, studies show that men emphasize health-oriented domains while women tend to empha-
size family and social relationship domains, and men often choose healthy age peers as a referent
while the age peers of women are more likely to be in poor health (Benyamini, 2008; Deeg and
Kriegsman, 2003). Thus, it is important to examine gender differences in SRH and how they affect
the association between SRH and mortality.
The cognitive procedure of self-rating health and the predictive power of SRH for mortality may
also vary by age and marital status. Previous studies have shown age differences in SRH because
physical health domains such as functional limitation and chronic disease are more important for
SRH assessment of older people, whereas young people tend to highlight health behaviors (Krause
and Jay, 1994; Shadbolt, 1997). Unlike objective dimensions of health such as physical and cogni-
tive functioning, self-reported health may not decline sharply with age (Zeng, Feng, Gu et al., 2016).
Frail elderly individuals may report relatively good self-rated health status because they have
adapted to chronic conditions (Groot, 2000). As a result, the predictive power of SRH for mortality
may decline with age (Zajacova and Woo, 2016). The self-rating of health also varies by marital sta-
tus. Studies have repeatedly shown that married people report better health compared to the unmar-
ried (e.g., Waite and Gallagher, 2000; Zhu and Gu, 2010; Verbrugge, 1979) and the predictive power
of SRH for mortality is greater among the married than among the unmarried (Zheng and Thomas,
2013).
The rating procedure and predictive power of IRH may also differ across these subgroups, though
the current literature is relatively limited in this regard. This is because interviewers may incorporate
respondents’ reported health information into evaluation when such an assessment is performed at
the end of interview (Feng, Zhu, Zhen et al., 2016), and because interviewers’ assessments are also
susceptible to subjectivity (Brissette, Leventhal, and Leventhal, 2003; Feng, Zhu, Zhen et al., 2016).
Some research has indeed shown that factors such as respondents’ socioeconomic status affect not
only SRH, but also affect physicians’ and interviewers’ reporting of health outcomes (Smith and
Goldman, 2011). It is therefore valuable to examine the predictive power of IRH for mortality by
gender, age, and other socioeconomic factors, especially given the recent interest in this important
measure.
To better understand global assessments of health and their ability to predict mortality, it is neces-
sary to examine IRH relative to SRH in different subgroups. This study examines the predictive
power of IRH and SRH for three-year mortality in different subpopulations among Chinese elders.
China is an aging giant, holding the largest elderly population in the world now and for a foreseeable
future. Based on UNPD statistics (2015), in the year 2010 8.2% of China’s 1.34 billion population

76 International Journal of Population Studies | 2016, Volume 2, Issue 2