This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Social Contact, Socioeconomic Status, and the Health Status of Older Malaysians

^a Sealy Center on Aging, University of Texas Medical Branch, Galveston, TX
^b Department of Preventive Medicine and Community Health, University of Texas Medical Branch at Galveston

Correspondence: Z. Helen Wu, PhD, Sealy Center on Aging, University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77555-0860. E-mail: zhwu{at}utmb.edu.

Decision Editor: Vernon L. Greene, PhD

	Abstract

TOP Abstract The Stress-Buffering Hypothesis Low SES as a... The Malaysian Setting Methods Results Discussion References

 
We tested the applicability of the stress buffering hypothesis in a developing country setting with data from the Senior Sample of the Malaysian Family Life Survey-2. Using ordered logistic regression methods, we examined whether having daily contact with adult children moderates the effect of low socioeconomic status (SES; conceptualized as a chronic stressor) on self-assessed health status. We found that low SES is associated with poorer health for all three ethnic groups—Malay, Chinese, and Indian. Further, for Malays and Chinese, we found that the negative effects of low SES on health tend to be stronger for older people with less frequent contact with adult children than for those who have daily contact. These results provide general support for the buffering model and suggest that, as found in developed countries, active intergenerational relationships in developing country settings may have protective effects on the health of older people experiencing chronic stressors.

Key Words: Intergenerational relations • Social support • Chronic stressor

Recent and projected increases in the proportion and number of elderly persons in many developing countries have drawn attention to issues concerning the well-being of this potentially vulnerable age group. The pace of population aging is projected to be relatively rapid in several modernizing Asian societies, where formal institutions to provide economic and health-related assistance are not well developed (Kinsella and Taeuber 1992; Neville 1992). Policy makers and researchers have expressed concerns that the social and economic changes accompanying modernization may weaken the traditional support system for elderly persons—the extended family—and place the expanding group of elders in jeopardy (e.g., Phillips 1992; United Nations 1992).

High levels of family interaction and exchange, especially between adult children and aging parents, have been documented in aging East and Southeast Asian nations (Andrews, Esterman, Braunack-Mayer, and Rungie 1986; Chen and Jones 1989). A generally accepted truism motivating much of this research is that stronger intergenerational relationships enhance the health and well-being of the older generation. In fact, the presence of social support fairly consistently has been found to be associated with more positive health outcomes in developed country settings (House, Landis, and Umberson 1988). For the past decade, research efforts in the social support and health literature have been focused on untangling the causal pathways linking support to morbidity and mortality. Given the presumably vital importance of family relationships to elderly persons in modernizing societies, it is crucial that we begin to more carefully examine the linkages between social support and health in these settings as well.

	The Stress-Buffering Hypothesis

TOP Abstract The Stress-Buffering Hypothesis Low SES as a... The Malaysian Setting Methods Results Discussion References

 
Using data from the Senior Sample of the second Malaysian Family Life Survey (MFLS-2), we tested the utility of the stress-buffering hypothesis in the Malaysian setting. That hypothesis suggests that the presence of social support will protect (or buffer) an individual from the negative health effects associated with various life stressors. We conceptualized low socioeconomic status (SES) as a chronic stressor that may negatively influence subjective health status, and we used level of contact with adult children as the measure of a social resource that may buffer that relationship. In other words, the model would predict that active family ties will protect an elderly individual from the health-damaging effects of stressors associated with low SES, whereas a more isolated individual who experiences the same stressors will be adversely affected. The stress-buffering hypothesis has been examined in a number of developed countries and has received general support, although results have varied across types of stressors, measures of social support, and health outcomes (Thoits 1995).

Preston and Taubman 1994 have advocated blending the economic and social psychological approaches to health studies to better understand socioeconomic differentials in health. Economic approaches focus on "goods and services" in the production of health, whereas social psychological approaches incorporate social stressors and relationships as additional determinants of health. Conceptualizing low SES as a chronic stressor (involving both absolute material deprivation and relative status in the social hierarchy) and invoking social psychology's buffering hypothesis is a small step toward merging the two perspectives. Furthermore, using subjective health status as our outcome measure allows us to capture both the physical and emotional toll that low SES may take over a lifetime.

Su and Ferraro 1997 have reported that social integration with family and friends has direct positive effects on the subjective health assessments of elderly Malaysians, even controlling for the negative effects of diminished functional health. We investigated the role that involvement with adult children may play in moderating the negative effects of a chronic stressor (low SES) on subjective health status.

	Low SES as a Chronic Stressor

TOP Abstract The Stress-Buffering Hypothesis Low SES as a... The Malaysian Setting Methods Results Discussion References

 
The bulk of the literature focusing on the interrelationships among stressors, coping resources (such as social relationships), and health has focused on the stress generated from acute life events such as death of a spouse or job loss. Recently, a growing number of studies have addressed the impact of chronic (or longer term) stressors on health (Aneshensel 1992; Thoits 1995). Sources of chronic stressors have been identified in all areas of daily life (e.g., employment, family relationships, residence), and these stressors fairly consistently have been found to worsen both physical and mental health (Aneshensel 1992; Thoits 1995). Many of the factors identified as chronic strains are closely associated with SES and stem from a relative or absolute lack of resources, power, or prestige. For example, such factors as material deprivation, barriers to achievement of life goals, goal-striving stress, and general social and economic hardship have been termed chronic strains (Aneshensel 1992).

A large body of literature has documented that SES is inversely related to health. Furthermore, a gradient in health status is routinely observed over the entire SES hierarchy, meaning no simple threshold effect (e.g., below vs above the poverty line) is apparent (see reviews by Adler et al. 1994; Marmot, Bobak, and Smith 1995). A variety of hypotheses have been offered to explain this gradient, focusing on SES differences in such factors as lifestyle and health behaviors, health care resources and utilization, environmental and occupational exposures, and exposure to stressors and the availability of coping resources. Certainly, these hypotheses are not mutually exclusive, and more than one factor may be relevant in the development of specific diseases. At least part of the general effect of SES on health, however, appears to be due to the strains associated with one's position in the SES hierarchy (Adler et al. 1994).

Researchers have begun to explore the relationship of SES to health outcomes among older persons in Asian societies. Financial strain has been found to increase depressive symptoms among elderly Japanese (Krause, Jay, and Liang 1991) and to increase psychological distress among elderly Chinese (Krause and Liang 1993). In two studies of Taiwanese elders, the authors consider the possibility that strong and pervasive family relationships in modernizing societies might mask the effect of SES on health and mortality (presumably through the family's protective or buffering roles). These studies, however, demonstrate that education is negatively related to both the risks of onset of functional limitations (Zimmer, Liu, Hermalin, and Chuang 1998) and mortality (Liu, Hermalin, and Chuang 1998), even with controls for social relationships. Education is also positively associated with self-ratings of health status among older Malaysians (Su and Ferraro 1997).

	The Malaysian Setting

TOP Abstract The Stress-Buffering Hypothesis Low SES as a... The Malaysian Setting Methods Results Discussion References

 
Past and recent declines in fertility and mortality have led to projections that the elderly population will be the fastest growing segment of the Malaysian population in the coming decades. The percentage of the population aged 65 or older is projected to increase from 3.8% in 1991 to 7.8% by 2021 (Leete 1996). These national figures mask some ethnic diversity in the experience of population aging. Census figures for 1991 show that 58.3% of the population was Malay, 29.4% Chinese, 9.5% Indian, and 2.8% other ethnic groups (Leete 1996). Population aging will be observed for all three of the main ethnic groups, but the pace will be slower among the Malays, due to their relatively smaller declines in fertility.

The fertility and mortality declines that are fueling population aging in Malaysia have occurred in the context of dramatic social and economic modernization. The gross domestic product has grown markedly in recent decades, and poverty levels have been reduced in all regions of the country. Among the changes occurring are higher education levels among younger generations, increased rural-to-urban migration, and greater work opportunities outside agriculture. Many of the economic and educational opportunities accompanying modernization, however, have not been available to the current cohort of older persons. Malaysians have expressed concern that these social and economic changes may disrupt the traditional family support network for elderly persons (United Nations 1992).

Although intergenerational relationships may be in flux, both coresidence and reliance on family assistance are still common patterns for elderly Malaysians. Various surveys have shown that roughly three quarters of elderly persons are living with younger family members and more than half of the aged rely on children or grandchildren for their main source of material support (Andrews et al. 1986; Chen and Jones 1989; DaVanzo and Chan 1994). Women are more likely than men to be coresiding with adult children and to be economically dependent upon their children. Elderly Indians are more likely to live with an adult child, and elderly Malays are least likely. One reason for the relatively lower rates of coresidency among the Malays is that housing costs are lower in the rural areas, and Malays traditionally have been a rural population. Controlling for coresidency, the elderly Chinese are most likely to receive financial assistance from grown children and least likely to receive transfers of services.

Family and household structure differs in important ways across the three main ethnic groups. Household structure in Malaysian communities is predominantly nuclear, although household composition is quite fluid, and often complex, as members move in and out depending on need. Even when adult children do not coreside with parents, they often live nearby. The Malaysian kinship system is generally bilateral, with some areas adhering to a matrilineal system in patterns of postmarital residence and inheritance. In contrast, the Chinese adhere to a patrilineal kinship system in which extended or stem families are the ideal. Extended families are also more common among the Indians, but the particular kinship systems vary in this minority population according to the part of India from which the immigrants originated (see Kling 1995, and Noor, Tan, Tey, and Rohani 1985).

	Methods

TOP Abstract The Stress-Buffering Hypothesis Low SES as a... The Malaysian Setting Methods Results Discussion References

 
Malaysian Family Life Survey-2
We used data from the Senior Sample of the Malaysian Family Life Survey-2 (MFLS-2), conducted in 1988-89 by RAND and the National Population and Family Development Board (NPFDB) of Malaysia. The Senior Sample includes 1,357 respondents aged 50 or older and is a representative sample of the older population in Peninsular Malaysia. The sample had a very high response rate, with 97% of selected elders completing the questionnaire. Indian households were oversampled (at twice the rate of the other ethnic groups) to obtain a sufficient sample size. Sample weights are available in the data. (For details on the sampling procedure, see Haaga, DaVanzo, Peterson, Peng, and Ann 1993.) We excluded 10 respondents listed as other ethnicity (not Malay, Chinese, or Indian) and one respondent with missing data, for an effective sample size of 1,346. Unweighted sample sizes for the three ethnic groups are: .

Measures
Table 1 reports the weighted percentage distributions by ethnicity for each of the measures used in the analyses.

Daily Contact With Adult Children.
We assessed social contact with a dichotomous measure indicating whether or not the respondent has daily contact with at least one adult child. Daily contact is defined as either being coresident with an adult child or having daily visits with a non-coresident child. This two-category measure of social support is rather simple, but it does address a key issue in the Malaysian setting, namely the availability and proximity of adult children. Children who coreside with or live near aging parents are able to provide a range of supportive services, including transfers of money and material goods, assistance with household chores, sick care, and companionship. The measure of daily contact is preferable to a simple coresidence measure because many adult children (especially among the Malays) live in houses that are separated from their parent's house, but in the same compound.

Socioeconomic Status.
We use two measures of socioeconomic status—literacy status and main occupation during prime working years. Both measures are intended to assess social and economic standing throughout adulthood, rather than simply indicating current economic status. This approach is particularly important when studying an elderly population whose members may have experienced recent declines in personal income due to retirement or disability. Literacy status is a two-category variable indicating whether or not the respondent is literate (can read or write) in any language. Literacy status is preferable to a measure of years of schooling for this age cohort, because most respondents had very little formal education. Respondents' main occupations during their working years are grouped as follows: professional or managerial, clerical, agricultural, labor or skilled occupation, and missing or not employed. The missing/not employed category is predominantly female and appears to be primarily women who do not report ever being in the labor force.

Functional Status.
The items used to create a physical limitations scale were adapted from the Medical Outcomes Study (Ware, Sherbourne, and Davies 1992). Respondents were asked whether or not their health limited them in any way in: (a) performance of vigorous activities (e.g., lifting heavy objects, doing hard work); (b) performance of moderate activities (e.g., moving a table, doing home repairs); (c) walking uphill or climbing stairs; (d) bending or stooping; (e) walking to a nearby house (100 meters); or (f) eating, dressing, bathing, or using the toilet. Affirmative responses were coded 1 and summed to form a scale from 0–6. Although the frequency distribution of the scale is reported in three categories, the scale is treated as continuous in the regression analyses.

Demographic characteristics.
We stratified our analyses by ethnic group (Malay, Chinese, Indian). Our analyses controlled for sex, age group (50–59, 60–69, and 70 or older), and marital status (currently married vs widowed, divorced, separated, or never married).

Analyses
We used ordered logistic regression methods to assess the relationship of socioeconomic status and social contact to our three-category measure of health status. The cumulative logit model is based on the cumulative distribution probabilities of the categories of health status, instead of on their individual probabilities. In other words, our models estimate the probability of being in good or fair versus poor health and the probability of being in good versus fair or poor health. One coefficient is estimated for each independent variable, because it is assumed that the effect of any variable on moving from category k to k+1 is the same for all k. For example, the model assumes that the effect of gender is the same for both health comparisons (see Agresti 1990).

We estimated three hierarchical models to test for the posited buffering effects of social contact. First, to identify the bivariate relationship between SES and health status, we predicted the probability of being in poorer versus better health with the individual SES measures. Second, we adjusted for potential confounders by estimating the effects of the SES measures on health status controlling for age group, sex, marital status, and physical limitations. Third, we estimated the full model stratified by the daily contact with adult children measure and examined differences in the effects of SES between the strata. All analyses were stratified by ethnic group and were performed on the weighted data.

Support for the buffering hypothesis will be found if the effects of SES on health status vary by levels of social contact. Specifically, we expected to find that the negative effects of low SES on health status are stronger for individuals without daily contact with adult children than for individuals with daily contact. To assess whether the SES effects vary significantly between the strata of social contact, we conducted a statistical test of the homogeneity of the parameter estimates. The logit-based test of homogeneity of the odds ratios is based on a weighted sum of the squared deviations of the stratum-specific log-odds ratios from their weighted mean, where the weights are the inverse of the variance of the log-odds ratios (Hosmer and Lemeshow 1989, pp. 73–74). This test statistic has a chi-square distribution with degrees of freedom (df) equal to one less than the number of strata (or 1 df in our analyses).

One potential problem in our analyses is the possible endogeneity of our stratification variable—daily contact with adult children. Specifically, health status and social contact may be reciprocally related in that persons in poor health may require care by adult children and thus may be in daily contact. To examine this, we conducted an endogeneity test using number of living children as an instrument for daily contact in a two-step logistic regression model as proposed by Bollen, Guilkey, and Mroz 1995. The results of this test suggested that social contact was not endogenous in the model. Furthermore, by controlling for functional status in the models, we are essentially examining that portion of self-assessed health status that reflects more general subjective well-being and not physical conditions that may limit self-care.

	Results

TOP Abstract The Stress-Buffering Hypothesis Low SES as a... The Malaysian Setting Methods Results Discussion References

 
Ethnic Differentials in Health, SES, and Social Contact
Health status and SES patterns differ across the ethnic groups, whereas distributions on the other variables are more similar (see Table 1 ). Only one third of Indian respondents assess their health status as good, compared to 48% of the elderly Malays and 41% of the Chinese. Ethnic differences in functional status are similar, with Indian respondents reporting significantly more physical limitations

We also examined the bivariate associations among some of the key measures within ethnic groups (results not shown). We found that contact with adult children does not differ significantly by SES for the Malaysian and Indian respondents; however, among the Chinese, persons of lower SES report more frequent contact than do higher SES persons (e.g., 45.2% of literate Chinese have daily contact compared to 55.0% of illiterate Chinese). In addition, we found that the association between daily contact and health status is not significant for any of the ethnic groups.

The Relationship of SES to Health Status
The measures of SES exhibit the expected bivariate relationship with health status for all three ethnic groups (unadjusted odds ratios [OR] greater than 1), but the strength and significance of the associations vary by ethnicity (see Table 2 ). Controlling for only the demographic factors (results not shown) tends to weaken, but not eliminate, the observed associations. Adding functional status to the model removes the effect of SES on health status, with the exception of the relationship between literacy and health status among the Chinese. Differences across occupational groups are most marked for the Malays and Indians, whereas literacy status appears to be a more relevant measure of SES among the Chinese.

A similar pattern is observed among the Chinese respondents, although, for this group, the relevant SES measure is literacy status rather than occupation. Illiterate persons are more likely to report worse health than are literate persons, but the strength of the effect is significantly weaker for Chinese with daily contact than for those with less frequent contact

The relationship of SES to health status among the Indian respondents is less clear. Indians who occupied professional or managerial positions exhibit a health advantage relative to other occupational groups in both social contact groups. The smaller sample size for the Indians may hinder our ability to adequately test the hypothesis for this population, or the buffering hypothesis may simply be inappropriate for understanding the health status of this group.

Controlling for physical limitations in the model means we are examining that part of self-assessed health status that reflects more general well-being. In order to assess the effects of SES on global health status, we also estimated models without including physical limitations as a control variable (results not shown). The results for these models produced even stronger confirmation of the buffering hypothesis for the Malaysian population (i.e., the differences in the effects of occupation were even larger between the two contact strata). Among both the Chinese and the Indians, the models produced results consistent with the buffering hypothesis, although differences across the strata were not significant.

The effects of age group, functional status, and sex are generally consistent across the models, with older age and worse functional health predicting poorer self-assessed health, and sex having no significant effect. Marital status differences in self-assessed health status are noteworthy. Among Malaysian respondents who have daily contact with adult children, marital status is not significantly associated with health status. In contrast, among persons with less frequent contact with adult children, married persons are significantly less likely than unmarried persons to report being in worse health. It appears that unmarried persons experience a health disadvantage in the absence of frequent interaction with adult children, whereas the close proximity of adult children may offset this disadvantage.

	Discussion

TOP Abstract The Stress-Buffering Hypothesis Low SES as a... The Malaysian Setting Methods Results Discussion References

 
In sum, the stress-buffering hypothesis receives general support in the Malaysian context. Having adult children available and proximate protects older persons who have experienced a lifetime of low socioeconomic status from the potentially health-damaging effects of that status. These results tend to hold whether we examine health status as a global (physical and emotional) measure or whether we control for physical limitations and restrict our focus to health status as more of an indicator of general well-being.

Although the strength of the relationship varies by ethnicity, the general pattern is observed for both the Malays and the Chinese suggesting that the utility of the hypothesis in the Malaysian setting may not vary by ideal family type (nuclear vs stem or extended). The relevant indicator of SES, however, does vary by ethnicity. Occupational status appears to be the best SES predictor of health status for the Malays and Indians, whereas literacy status is a stronger predictor for the Chinese. Historically, most of the Malaysian population has resided in the rural areas and engaged in agricultural or fishing occupations. Elder Malays who held professional or managerial occupations represent an elite and privileged group. The Chinese, in contrast, have been predominantly urban dwellers involved in industry and trade. In the urban environment, being literate (and having the education that reflects) may have led to higher paying and higher prestige jobs within occupational groups.

Our reliance on a dichotomous measure of social contact does not permit us to specify the mechanism underlying the buffering relationship. The daily contact with adult children variable simply indicates the availability of children and does not identify what particular aspects of parent–child interaction may be protective. Children living with or near aging parents can more readily provide companionship and emotional support, in addition to transfers of needed goods or services. Having children proximate may also mean having grandchildren nearby, and grandchildren can be another important source of companionship and household services. The MFLS-2 does include data on intergenerational transfers of time, money, and goods, but only for inter-household transfers, not intra-household transfers. The available transfers data indicate that most older adults (65.3%) received money or goods in the past year from non-coresident children, but only a minority (20.0%) received time transfers (assistance with household or business activities). These figures suggest that parent–adult child proximity is not required for monetary transfers, but that proximity is likely to be a crucial determinant of the level of time assistance and emotional support that adult children can provide. Thus, adult children can buffer multiple pathways between low SES and health outcomes. Adult children can offset a lack of health-related resources by transferring money and goods. The younger generation can also protect against some of the emotional strain associated with low SES by providing companionship and offering respect to the aging parent.

Another limitation of our analyses is our use of cross-sectional data. Sorting out the causal ordering of the relationships among social resources, SES, and health is complicated. We have attempted to circumvent this issue in part by using measures of SES that reflect status throughout adulthood and are prior to current health status. In fact, among the respondents who reported any physical limitations (67.6%), the majority stated that the limitation had begun within the past 5 years. Furthermore, the most common limitations were in performing vigorous and moderate activities, conditions that likely would not necessitate coresident assistance. Only 2.6% of respondents reported difficulty performing the activities of daily living, and just 10.9% stated they had difficulty walking 100 meters. Thus, only a minority of these older adults may have a real need for help with self-care or household tasks, suggesting that parent–adult child coresidence is not typically in response to poor parental health.

Our findings suggest that the importance of adult children to the health of aging parents is not limited to simply providing sick care or resources, but that the simple presence of children in parents' lives also may provide some protection from ill health for persons facing the chronic stressor of low SES. Coupled with the findings of Su and Ferraro 1997 regarding the direct health benefits of social integration, these results emphasize the importance of either keeping the extended family strong or providing older adults with alternative opportunities for social interaction. In fact, in regional meetings on the status of older Asians (United Nations 1996, United Nations 1997), researchers and policy makers have addressed the benefits of integration both within the family and within the community.

	Acknowledgments

 
An earlier version of this article was presented at the 1996 annual meeting of the Population Association of America. This study was funded in part by Grant AG12544 from the National Institute on Aging.

Received for publication August 28, 1997. Accepted for publication January 7, 2000.

	References

TOP Abstract The Stress-Buffering Hypothesis Low SES as a... The Malaysian Setting Methods Results Discussion References

 

• Adler N. E., Boyce T., Chesney M. A., Cohen S., Folkman S., Kahn R. L., Syme S. L., 1994. Socioeconomic status and health: The challenge of the gradient. American Psychologist 49:15-24. [Medline]
• Agresti A., 1990. Categorical data analysis John Wiley & Sons, New York.
• Andrews G. R., Esterman A. J., Braunack-Mayer A. J., Rungie C. M., 1986. Aging in the western Pacific World Health Organization, Manila.
• Aneshensel C. S., 1992. Social stress: Theory and research. Annual Review of Sociology 18:15-38.
• Bollen K. A., Guilkey D. K., Mroz T. A., 1995. Binary outcomes and endogenous explanatory variables: Tests and solutions with an application to the demand for contraceptive use in Tunisia. Demography 32:111-131. [Medline]
• Chen A., Jones G., 1989. Ageing in ASEAN: Its socio-economic consequences Institute of Southeast Asian Studies, Singapore.
• DaVanzo J., Chan A., 1994. Living arrangements of older Malaysians: Who coresides with their adult children?. Demography 31:95-113. [Medline]
• Haaga J., DaVanzo J., Peterson C., Peng T. N., Ann T. B., 1993. The Second Malaysian Family Life Survey: Overview and technical report RAND, Santa Monica, CA.
• Hosmer D. W., Jr. Lemeshow S., 1989. Applied logistic regression John Wiley & Sons, New York.
• House J. S., Landis K. R., Umberson D., 1988. Social relationships and health. Science 241:540-544. [Abstract/Free Full Text]
• Kinsella K., Taeuber C.M., 1992. An aging world II. U.S. Bureau of the Census, International Population Reports, P25, 92-3 U.S. Government Printing Office, Washington, DC.
• Kling Z., 1995. The Malay family: Beliefs and realities. Journal of Comparative Family Studies 26:43-66.
• Krause N., Jay G., Liang J., 1991. Financial strain and psychological well-being among the American and Japanese elderly. Psychology and Aging 6:170-181. [Medline]
• Krause N., Liang J., 1993. Stress, social support, and psychological distress among the Chinese elderly. Journal of Gerontology: Psychological Sciences 48:P282-P291.
• Leete R., 1996. Malaysia's demographic transition: Rapid development, culture, and politics Oxford University Press, New York.
• Liu X., Hermalin A. I., Chuang Y. L., 1998. The effect of education on mortality among older Taiwanese and its pathways. Journal of Gerontology: Social Sciences 53B:S71-S82. [Abstract]
• Marmot M., Bobak M., Smith G. D., 1995. Explanations for social inequalities in health. Amick B. C., Levine S., Tarlov A. R., Walsh D. C., , ed.Society & health 172-210. Oxford University Press, New York.
• Neville W., 1992. The dynamics of population ageing into the twenty-first century: ASEAN and selected countries of Pacific Asia. ASEAN Economic Bulletin 9:4-21.
• Noor, L., Tan, B. A., Tey, N. P., & Rohani, A. R. (1985). Ethnicity and fertility in Malaysia. Research notes and Discussion Paper No. 52. Institute of Southeast Asian Studies.
• Phillips D. R., 1992. Ageing in east and south-east Asia Edward Arnold, London.
• Preston S. H., Taubman P., 1994. Socioeconomic differences in adult mortality and health status. Martin L. G., Preston S. H., , ed.Demography of aging 279-318. National Academy Press, Washington, DC.
• Su Y., Ferraro K. F., 1997. Social relations and health assessments among older people: Do the effects of integration and social contributions vary cross-culturally?. Journal of Gerontology: Social Sciences 52B:S27-S36. [Abstract]
• Thoits, P. A. (1995). Stress, coping, and social support processes: Where are we? What next? Journal of Health and Social Behavior, Extra Issue, 53–79.
• United Nations1992. Population ageing: Review of national policies and programmes in Asia and the Pacific. Asian Population Studies Series No. 109 United Nations Publications, New York.
• United Nations1996. Lifelong preparation for old age in Asia and the Pacific United Nations Publications, New York.
• United Nations1997. International and regional mandates on ageing. Source book on ageing (Volume II) United Nations Publications, New York.
• Ware J., Sherbourne C., Davies A., 1992. Developing and testing the MOS 20-item short form health survey: A general population application. Stewart A., Ware J. E., , ed.Measuring functioning and well-being: The Medical Outcomes Study approach 277-290. Duke University Press, Durham, NC.
• Zimmer Z., Liu X, Hermalin A., Chuang Y.L., 1998. Educational attainment and transitions in functional status among older Taiwanese. Demography 35:361-375. [Medline]

This article has been cited by other articles:

				M. O. Rahman and A. J. Barsky Self-Reported Health Among Older Bangladeshis: How Good a Health Indicator Is It? Gerontologist, December 1, 2003; 43(6): 856 - 863. [Abstract] [Full Text] [PDF]

				H. P. Greenwald and W. L. Beery Reducing Isolation among Inner-City Elders: An Outcome Evaluation Health Promot Pract, July 1, 2001; 2(3): 233 - 241. [Abstract] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation