This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation

Religious Participation and Mortality Risk Among the Oldest Old in China

Department of Sociology, University of Hawaii at Manoa.

Address correspondence to Wei Zhang, Department of Sociology, University of Hawaii at Manoa, 2424 Maile Way, Saunders 204, Honolulu, HI 96822. E-mail: weizhang{at}hawaii.edu

	Abstract

TOP Abstract Theoretical and Empirical... Methods Results Discussion References

 
Objectives. Rapid population aging in China calls for more research into social factors responsible for health and well-being among older adults. This article adds to this line of inquiry by examining the relationship between religious participation and mortality, as well as the potential pathways linking these factors and subgroup variations among oldest old Chinese.

Methods. Using two waves (1998 and 2000) of the Chinese Longitudinal Healthy Longevity Survey, I estimated Cox proportional hazards models for a nationwide sample of Chinese aged 80 to 105.

Results. Controlling for a wide range of covariates, I found religious participation to be significantly associated with lower risk of mortality for oldest old women and for individuals in poor health. Engaging in leisure activities and exercises partially accounted for this association.

Discussion. Findings suggest that (a) religious participation, associated with other socially integrated and cognitively stimulating activities, predicts mortality risk among oldest old Chinese; and (b) religious participation offers psychosocial resources that are likely to compensate for the increased mortality risk associated with disadvantaged socioeconomic conditions of certain vulnerable groups such as women and individuals in poor health.

Key Words: Religious participation • Mortality • Oldest old • China

CHINA, the most populous country in the world, is now aging at a rapid speed. The number of the oldest old population aged 80 and older, currently the largest of any nation in the world, may increase by more than 100 million by 2050 (Zeng, Liu, & George, 2003). China's population control program, the "one-child policy" introduced in 1980, worsened the situation by reducing the size of the younger population. This imbalance makes the care of older persons at both society and family levels increasingly difficult and calls for the recognition of social factors responsible for the well-being of older Chinese.

Religious participation stands out as one of the important health-promoting social factors worthy of exploration. Despite the recent decades of government constraint, China has historically been a country of multiple religious traditions, all of which have characteristics that are particularly relevant to health (e.g., Ai, 2000, 2006). A recent survey on religious beliefs in China (Wu, 2007) disclosed that approximately one fourth of the Chinese population, or more than 300 million Chinese, is involved in religious activities of one form or another. This makes it feasible to explore the religion–health association, a long neglected topic, in contemporary China. Specifically, this article examines the religion–mortality association, as well as the potential mechanisms linking these factors and the subgroup variations.

	THEORETICAL AND EMPIRICAL BACKGROUND

TOP Abstract Theoretical and Empirical... Methods Results Discussion References

 
Previous Research on Religion and Mortality in the United States
The past three decades have witnessed a growing body of U.S. research examining relationships between individual religious involvement and health (for reviews, see Ellison & Levin, 1998; Koenig, McCullough, & Larson, 2001). Increasing evidence suggests that multiple features of religiousness have beneficial effects on a wide range of health outcomes (Hummer, Ellison, Rogers, Moulton, & Romero, 2004; Koenig et al., 2001). In particular, more frequent service attendance is associated with a reduced risk of mortality (for reviews, see Hummer et al., 2004; McCullough, Larson, Hoty, Koenig, & Thoresen, 2000). Evidence comes from community-based epidemiologic surveys (e.g., Koenig et al., 1999; Lutgendorf et al., 2004; Oman & Reed, 1998; Strawbridge, Cohen, Shema, & Kaplan, 1997), nationwide surveys (Hummer, Rogers, Nam, & Ellison, 1999; Musick, House, & Williams, 2004), as well as clinical samples (Oxman, Freeman, & Manheimer, 1995; Pargament, Koenig, Tarakeshwar, & Hahn, 2001). Not surprisingly, many of these studies have focused on older adults (e.g., Helm, Hays, Flint, Koenig, & Blazer, 2000; Hill, Angel, Ellison, & Angel, 2005; Krause, 1998; Lutgendorf et al., 2004), among whom levels of religiousness tend to be high and for whom religion is an arguably important social resource.

Despite these recent advances in research on religion and mortality, investigators have only begun to examine this association among racial and ethnic minorities (Ellison, Hummer, Cormier, & Rogers, 2000; Hill et al., 2005). There is evidence indicating that religious involvement is associated with a reduction in the risk of mortality among African American adults (Ellison et al., 2000) and older Hispanics (Hill et al., 2005). However, little research has been done on the religion–mortality association in a non-Western setting. In this study, I hypothesized that participating in Eastern religions would also reduce the odds of death. This is the first hypothesis.

Religious Participation and Mortality in China
Most people in China practice a combination of folk religion, Confucianism, Taoism, and Buddhism. Different from Western congregational religions, Chinese religious traditions have long-standing and distinct psychosocial and behavioral functions on health and well-being. For instance, followers of Confucian teachings believe that virtues such as self-cultivation, chung (loyalty), shu (empathy), and jen (humaneness) are imperative to developing and maintaining deep and profound relationships with self, family, community, and state (Taylor, 1998). This greatly facilitates the formation of secular social ties, on the basis of which coreligionists are likely to engage in various leisure activities that are socially integrated or cognitively stimulating, such as collective exercises, playing cards, chase, or Mah-jongg (Ross & Zhang, 2008). Buddhism is more similar to Western religion with respect to spiritual and ritual practices that are likely to offer means to spiritual well-being, such as tolerance, self-control, harmony, inner peace, and wisdom (Ai, 2000). These faith-based attitudes, along with rituals such as prayer, may lead to positive expectations such as hope and optimism (Ai, Peterson, Tice, Bolling, & Koenig, 2004) that are likely to help vulnerable older Chinese such as women and individuals in bad health to cope effectively with traumatic life events. The spirit of Taoism underlines the mind–body connection and promotes holistic well-being through energy-based health practices and healing processes such as tai chi or qigong, which involve deep breath and meditation with movement (Ai, 2006).

In sum, Chinese religions have a potential impact on the health and well-being of older adults. They integrate essences from different religious traditions, build social as well as spiritual relationships, encourage an active lifestyle among participants, and create positive expectations. In this sense, China is a nation of religion, not by a Western definition, but in its own way. Many aspects of Chinese religions are likely to be associated with a reduction in the risk of mortality. This study hypothesized two sets of potential mediators linking Chinese religions and mortality: psychological resources indicated by optimism and autonomy, and health behaviors indicated by engaging in collective exercises and other leisure activities. This is the second hypothesis.

Religion and Mortality: Subgroup Variations
Findings on religion and mortality have become increasingly robust, but studies have only begun to examine the contingent effects of religious involvement on mortality, and the available evidence is far from conclusive. Rogers, Hummer, and Nam (2000) suggested a stronger religion–mortality association for younger adults, Ellison and colleagues (2000) showed gender differences in mechanisms underlying the religion–mortality link, and Oman and Reed (1998) indicated that religious attendance amplified the effects of social support on mortality. Whereas some studies have shown stronger religion–mortality links among women, and especially among older women (Koenig et al., 1999; Strawbridge et al., 1997), studies of general populations have not (Hummer et al., 1999; Musick et al., 2004). Virtually no evidence suggests that the religion–mortality association differs by health status.

To explain how the influence of religiosity on health and well-being conditioned on demographic factors, Schieman, Nguyen, and Elliot (2003) proposed two competing theories. One is called resource amplification theory: Religion is more valuable in the context of other resources such as education, income, and social activities. The other is called resource compensation theory: Religion is more important in the absence of other types of advantage. Supporting the resource amplification view, Schieman and colleagues suggested that the negative association between religiosity and mastery was stronger among those who were poorly educated. Their findings on religion and mental health shed light on the religion–mortality association. Alternatively, Ellison (1991) indirectly suggested the resource compensation view: The "existential certainty" was associated more positively with life satisfaction among those who were less educated. Similarly, Ross and Mirowsky (2006) disclosed the gender-conditioned association between education and psychological well-being: Women benefit more from education than men because resources substitute for one another, and the presence of education makes the absence of other resources less harmful.

Extending resource compensation or substitution theory into the religion–mortality scenario, one can reasonably argue that religiousness, and the resources derived from it, benefit longevity more for socioeconomically deprived groups such as women and individuals in poor health. For instance, Chinese women have historically lived in a patriarchal society where they have structurally and systematically been denied education and mentally stimulating occupations (National Research Council, 2000; Zhang, 2006). As a result, they face more economic dependency and are more socioeconomically disadvantaged. In this study, I examined how participating in religious activities may compensate for or amplify the disadvantaged socioeconomic positions for women and for individuals in poor health. This is the third hypothesis of this article.

	METHODS

TOP Abstract Theoretical and Empirical... Methods Results Discussion References

 
Data
Data to test these ideas came from the 1998 and 2000 Chinese Longitudinal Healthy Longevity Survey (CLHLS). CLHLS was conducted by randomly selecting one half of the cities and counties of China's 22 provinces where Han Chinese are the ethnic majority. The survey area covered more than 85% of the Chinese population. The baseline sample included 9,093 respondents aged 77 to 122 and had a response rate of 88%. In 2000, 4,831 respondents remained and were reinterviewed, whereas 3,368 respondents had died and 894 were unavailable for follow-up. This data set was ideal for testing the three proposed hypotheses because of its richness in variables tapping characteristics of religious participation, psychosocial resources, and a wide range of health behaviors. This study focused on 8,805 respondents aged 80 to 105 years old at baseline. I excluded those aged 106 and older due to insufficient information to validate their old age (Zeng & Vaupel, 2002). After listwise deletion of missing values, the analytic sample was reduced to 6,747.

Measurement
Mortality
Based on informant report, I recorded mortality from 1998 to 2000, yielding a roughly 2-year follow-up period. I created a binary variable to distinguish the oldest old who had died (sometime between 1998 and 2000) from those who were still alive at the 2000 interview. I defined survival time as the number of days between the baseline interview and the date of death or final interview, at which point surviving participants represented censored observations. Roughly 37% of the participants died over the 2-year study period.

Religious participation
Respondents were asked how often they participated in religious activities (1 = every day, 3 = never). I grouped participants indicating every day and sometimes were grouped into one category (1) and participants indicating never into another (0). More than 17% of the sample reported participating in religious activities at least sometimes.

Demographics
Demographic measures included age (in years), ethnicity (Han: 0 = minority, 1 = Han), gender (female: 0 = male, 1 = female), marital status (married: 0 = not married, 1 = married), education (literate: 0 = illiterate, 1 = literate), urban/rural residence (urban: 0 = rural residence, 1 = urban residence), and east/west residence (east: 0 = west residence, 1 = east residence). A majority of the sample was Han (93%), female (57%), rural (63%), illiterate (66%), and unmarried (82%).

Baseline health status
Respondents were asked to evaluate their overall health status (1 = very bad, 5 = very good). The mean level of self-reported health was 3.64, roughly approaching the "good" category.

Psychological resources
I used optimism and autonomy as the indicators of psychological resources. Respondents were asked the following: "How often do you look on the bright side of things?" and "How often do you make your own decisions?" Responses to these two questions were coded in an increasing (1–5) manner such that higher scores reflected more optimism and autonomy. A majority of the sample reported above average levels of optimism (M = 3.9) and of making their own decisions (M = 3.51).

Health behaviors
I incorporated three health risk factors, some of which have been used in previous studies (Hill, Ellison, Burdette, & Musick, 2007; Musick et al., 2004; Strawbridge, Shema, Cohen, & Kaplan, 2001), and some of which were highly relevant in the Chinese cultural setting (Ross & Zhang, 2008). A four-item index on engagement in leisure activities gauged how often (1 = never, 3 = every day) respondents (a) did gardening, (b) read newspapers or magazines, (c) watched television or listened to the radio, and (d) played cards or Mah-jongg. I measured exercise in response to the question "Do you exercise at present?" I measured smoking in response to the question "Do you smoke at present?" The answers were coded into yes = 1 and no = 0. The mean level of leisure activity engagement was around 1.34. Approximately 30% of the sample reported that they exercised, and 19% of the respondents smoked.

	RESULTS

TOP Abstract Theoretical and Empirical... Methods Results Discussion References

 
I first examined the sample attrition pattern. A binary variable marked those lost to the second interview and those who either were available for the second interview or had died. I then applied logistic regression to regress this binary outcome on the Wave 1 measures. Findings revealed that the sample attrition did not occur in a random manner: Participants who were lost between 1998 and 2000 were more likely to be Han, urban residents, and highly educated. However, religious participation and the proposed mediators did not predict the loss of sample. This kept enough variability in religious participation in the remaining sample and facilitated the follow-up analyses on the effects of religious participation on mortality.

I then applied Cox proportional hazards models to estimate mortality risk depending on religious participation and control, mediating, and moderating factors. Model 1 tested the religion–mortality association, adjusting for demographics and health status when appropriate. In Models 2 through 4, I added psychosocial mediators to see how much they could account for the religion–mortality association. I examined moderating effects by creating cross-product terms between religion and gender, and health status, with each cross-product term entering the main effects model separately.

Upon examination of the whole sample (N = 6,747; data not shown), the hazard ratio (HR) in the first model suggested that religious participation was significantly related to a lower risk of mortality (HR = 0.87, p =.002) over and above controls. Exercise and engagement were significant mediators, substantially reducing the significance level of religion (HR = 0.92, p =.16). Adding the cross-product terms into the final model (Model 4) suggested that the interaction effects for gender (HR = 0.81, p =.085) and health status (HR = 1.35, p =.009) were moderately to strongly significant.

To provide more insight into the moderating effects, I divided the sample into four subgroups: women, men, individuals in poor health, and individuals in good health. For each subgroup, I applied the same modeling strategies. Table 1 presents the hazard rate ratios from the Cox proportional hazards models for women and for individuals in poor health. For women, participating in religious activities significantly reduced the mortality risk over and above controls for demographics and health status (HR = 0.82, p =.006). Tests of mediating effects suggested that respondents who engaged little in exercise or other leisure activities were at an increased risk for mortality. By adding these mediators into Models 3 and 4, I found that the effects of religion diminished somewhat (HR = 0.87, p =.054). I found similar patterns for individuals in poor health (see the right-hand side of Table 1): Religious participation significantly reduced their mortality risk over and above controls and mediators (HR = 0.79, p =.009). By contrast, religious participation had no significant effects for men or for individuals in good health (data not shown). These results suggest that religious participation in China has significant and sizeable direct and indirect effects on mortality for women and for individuals in poor health.

	DISCUSSION

TOP Abstract Theoretical and Empirical... Methods Results Discussion References

In addition, the religion–mortality mediators exhibited unique features in China: A substantial portion of religious effect on mortality was mediated by health behaviors rather than by psychological resources such as optimism and autonomy. This indicates the dialectic nature of Eastern religions, which do not necessarily lead to positive psychological status as do Western religions. These findings suggest that religious participation in China might help older Chinese develop healthy lifestyles, stimulate social engagement, and thus reduce mortality risk. Not only do individuals participate in religious activities, they also help to produce them; by doing so, they greatly enhance their health and well-being (Musick et al., 2004).

The results of this investigation are intriguing; however, much remains to be explored. The first priority is measurement improvement. Future research should examine the content of religion and spirituality practice or beliefs rather than treat them as social or demographic components such as simple measures of education or income. For example, the measure of religious participation needs to be further decomposed to identify the precise facet of religion that is responsible for the health-protective effects. Does it reflect public attendance (similar to organizational participation in the United States) or private religious practice (such as praying at home)? Although my follow-up in-depth interviews in China suggested both, more well-developed survey questions gauging multiple features of religiousness are greatly needed. In addition, more waves of data covering a longer period of time, as well as more comprehensive samples covering younger generations, are likewise imperative. In this study, approximately 23% of the sample was lost due to item nonresponse. This high rate indicates the general low levels of education among oldest old Chinese. By sampling younger generations with more years of schooling, future researchers will likely overcome this problem.

	Acknowledgments

 
An earlier version of this study was presented at the 78th Annual Meeting of the Pacific Sociological Association, Oakland, CA, March 29 to April 1, 2007. The Chinese Longitudinal Healthy Longevity Survey (Zeng Yi and James W. Vaupel, principal investigators) was obtained from ICPSR. I would like to thank Christopher G. Ellison, Robert A. Hummer, Catherine E. Ross, Kenneth F. Ferraro, and three anonymous reviewers for their very helpful comments on earlier drafts, and I thank the managing editor and John Morris for their assistance with editing.

	Footnotes

 
Decision Editor: Kenneth F. Ferraro, PhD

Received for publication November 27, 2007. Accepted for publication May 12, 2008.

	References

TOP Abstract Theoretical and Empirical... Methods Results Discussion References

 

• Ai, A. (2000). Spiritual well-being, spiritual growth, and spiritual care for the aged: A cross-faith and interdisciplinary effect. Journal of Religious Gerontology, 11,3-28.
• Ai, A. (2006). Qigong. In M. S. Micozzi (Ed.), Fundamentals of complementary and integrative medicine (3rd ed., pp. 418–436). St Louis, MO: Elsevier Health Sciences.
• Ai, A., Peterson, C., Tice, T., Bolling, S., Koenig, H. (2004). Faith-based and secular pathways to hope and optimism subconstructs in middle-aged and older cardiac patients. Journal of Health Psychology, 9,435-450.[Abstract/Free Full Text]
• Ellison, C. (1991). Religious involvement and subjective well-being. Journal of Health and Social Behavior, 32,80-99.[Medline]
• Ellison, C., Hummer, R., Cormier, S., Rogers, R. (2000). Religious involvement and mortality risk among African American adults. Research on Aging, 22,630-667.[Abstract/Free Full Text]
• Ellison, C., Levin, J. (1998). The religion-health connection: Evidence, theory, and future directions. Health Education and Behavior, 25,700-720.[Abstract/Free Full Text]
• Helm, H., Hays, J., Flint, E., Koenig, H., Blazer, D. (2000). Does private religious activity prolong survival? A six-year follow-up study of 3,851 older adults. Journal of Gerontology: Medical Sciences, 55A,M400-M405.[Abstract/Free Full Text]
• Hill, T., Angel, J., Ellison, C., Angel, R. (2005). Religious attendance and mortality: A 8-year follow-up of older Mexican Americans. Journal of Gerontology: Social Sciences, 60B,S102-S109.[Abstract/Free Full Text]
• Hill, T., Ellison, C., Burdette, A., Musick, M. (2007). Religious involvement and healthy lifestyles: Evidence from the survey of Texas adults. Annals of Behavioral Medicine, 34,217-222.[Medline]
• Hummer, R., Ellison, C., Rogers, R., Moulton, B., Romero, R. (2004). Religious involvement adult mortality in the United States: Review and perspective. Southern Medical Journal, 97,1223-1230.[Medline]
• Hummer, R., Rogers, R., Nam, C., Ellison, C. (1999). Religious involvement and U.S. adult mortality. Demography, 36,273-285.[Medline]
• Koenig, H., Hays, J., Larson, D., George, L., Cohen, H., McCullough, M., et al. (1999). Does religious attendance prolong survival? A six-year follow-up study of 3,698 older adults. Journal of Gerontology: Medical Sciences, 54A,M370-M377.[Abstract]
• Koenig, H., McCullough, M., Larson, D. (2001). Handbook and religion and health. New York: Oxford University Press.
• Krause, N. (1998). Stressors in highly valued roles, religious coping, and mortality. Psychology and Aging, 13,242-255.[Medline]
• Lutgendorf, S., Ullrich, P., Russell, D., Tamara, B., Harris, T., Wallace, R. (2004). Religious participation, interleukin-6, and mortality in older adults. Health Psychology, 23,465-475.[Medline]
• McCullough, M., Larson, D., Hoyt, W., Koenig, H., Thoresen, C. (2000). Religious involvement and mortality: A meta-analytic review. Health Psychology, 19,211-222.[Medline]
• Musick, M., House, J., Williams, D. (2004). Attendance at religious service and mortality in a national sample. Journal of Health and Social Behavior, 45,198-213.[Medline]
• National Research Council. (2000). The aging mind: Opportunities in cognitive research. Washington, DC: National Academy Press.
• Oman, D., Reed, D. (1998). Religion and mortality among the community-dwelling elderly. American Journal of Public Health, 88,1469-1475.[Abstract/Free Full Text]
• Oxman, T., Freeman, D., Manheimer, E. (1995). Lack of social participation or religious strength and comfort as risk factors for death after cardiac surgery in the elderly. Psychosomatic Medicine, 57,5-15.[Abstract/Free Full Text]
• Pargament, K., Koenig, H., Tarakeshwar, N., Hahn, J. (2001). Religious struggle as a predictor of mortality among medically ill elderly patients: A 2-year longitudinal study. Archives of Internal Medicine, 161,1881-1885.[Abstract/Free Full Text]
• Rogers, R., Hummer, R., Nam, C. (2000). Living and dying in the USA: Health, behavioral, and social differentials of adult mortality. San Diego, CA: Academic Press.
• Ross, C., Mirowsky, J. (2006). Sex differences in the effect of education on depression: Resource multiplication or resource substitution. Social Science & Medicine, 63,1400-1413.
• Ross, C., Zhang, W. (2008). Education and psychological distress among elderly Chinese. Journal of Aging and Health, 20,273-289.[Abstract/Free Full Text]
• Schieman, S., Nguyen, K., Elliot, D. (2003). Religiosity, socioeconomic status, and sense of mastery. Social Psychology Quarterly, 66,202-221.
• Strawbridge, W., Cohen, R., Shema, S., Kaplan, G. (1997). Frequent attendance at religious services and mortality over 28 years. American Journal of Public Health, 87,957-961.[Abstract/Free Full Text]
• Strawbridge, W., Shema, S., Cohen, R., Kaplan, G. (2001). Religious attendance increases survival by improving and maintaining good health behaviors, mental health, and social relationships. Annals of Behavioral Medicine, 23,68-74.[Medline]
• Taylor, R. (1998). The religious character of the Confucian tradition. Philosophy East and West, 48,80-107.
• Wu, J. (2007, February 7). Religious believers thrice the official estimate: Poll. China Daily, p. 1.
• Zeng, Y., Liu, Y., George, L. (2003). Gender differentials of the oldest old in China. Research on Aging, 25,65-80.[Abstract/Free Full Text]
• Zeng, Y., Vaupel, J. (2002). Functional capacity and self-evaluation of health and life of oldest old in China. Journal of Social Issues, 58,733-748.
• Zhang, Z. (2006). Gender differentials in cognitive impairment and decline of the oldest old in China. Journal of Gerontology: Social Sciences, 61B,S107-S115.[Abstract/Free Full Text]

This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation