NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) Committee on Population; Martin LG, Soldo BJ, editors. Racial and Ethnic Differences in the Health of Older Americans. Washington (DC): National Academies Press (US); 1997.

Cover of Racial and Ethnic Differences in the Health of Older Americans

Racial and Ethnic Differences in the Health of Older Americans.

Show details

5How Health Behaviors and the Social Environment Contribute to Health Differences Between Black and White Older Americans

and .


Since data have been collected in the United States on racial differences in health status, blacks have been found to be at increased risk for almost every poor health outcome from most causes of morbidity to mortality and disability. Such inequalities in health are documented for men and women from birth to old age. Even crossovers in very old age in which blacks have been shown to have a survival advantage are now viewed with renewed skepticism (see Elo and Preston, and Manton and Stallard, in this volume). Furthermore, while health status has improved over the last decades for all Americans, the gains have been greater for whites than for blacks, producing an even larger health disparity for blacks in the last decade or two. For instance, during the last 30 years, life expectancy at age 65 increased 2.4 years for white males and only 1.5 years for black males. The corresponding increases for women are 3.1 years for white women and 2.5 years for black women (U.S. Department of Health and Human Services, 1993).

The reasons offered to account for these disparities range from genetic and selection factors to environmental exposures and differential access to medical care. Rather than explore this multitude of possibilities, this paper explores the extent to which health-damaging and health-promoting behaviors explain black-white differences in health status. In addition, we have taken the perspective that while behaviors are de facto performed by individuals, individual behaviors occur in a social context. They are heavily influenced by the larger social structure. For instance, laws regulating the consumption and taxation of alcohol and cigarettes lead directly to altered patterns of consumption. Most behaviors, in fact, vary across social strata. In this paper we are specifically interested in the extent to which there are differences in the distribution of health behaviors or social networks between blacks and whites in the United States and whether this differential distribution is related to underlying differences in socioeconomic status.

The paper is divided into several sections. First we review the evidence on the distribution of health-damaging and health-promoting behaviors among blacks and whites. We take a rather broad perspective on such behaviors, reviewing those that are traditionally called risk behaviors, such as cigarette and alcohol consumption, as well as health-promoting activities, such as physical exercise and maintaining social community ties.

In the second section we examine the distribution of health-damaging and health-promoting behaviors in a large study of older men and women, the New Haven Established Populations for the Epidemiologic Study of the Elderly (EPESE). In that section, we examine both traditional health behaviors and cardiovascular risk factors as well as social conditions related to social networks and socioeconomic status.


Comparisons of the health practices of black and white older men and women should not be made without the following considerations:

  1. It is now becoming evident that knowledge and information are not the only determinants of behavior. A corollary is that a particular behavior does not necessarily reflect lack of information. People who are aware that exercise helps reduce the risk of heart disease may nonetheless maintain a sedentary lifestyle (Oldridge, 1982). Many cigarette smokers, aware of the association between coronary artery disease or lung cancer and tobacco, continue to smoke (Rigotti et al., 1994). Data about racial differences in knowledge of particular health risks must be considered in the context of other factors that influence the ability or desire to act on that knowledge. Poverty, poor access to medical care, perceived powerlessness and frustration, peer pressure, and differential access to alcohol, tobacco, and food all can impede the adoption of health-enhancing practices (Braithwaite and Lythcott, 1989; Blendon et al., 1989; Rogers, 1992).
  2. Even when knowledge and information are adequate, other factors on the health care side of the equation may lead to behavior that is associated with poor health outcomes. Both race and social position have been shown to correlate with the quality and type of screening and therapeutic recommendations that health care providers give their patients (Burstin et al., 1992; American Medical Association Council on Ethical and Judicial Affairs, 1990). Analyzing data from the 1988-1990 Behavioral Risk Factor Surveillance System, Giles et al. (1993) showed that adults consulting a physician for preventive care were less likely to be screened for hypercholesterolemia if they were black or had not completed high school than if they were white or more educated. Black adults who needed treatment for elevated blood cholesterol were also less likely to be treated. That difference persisted after adjustment for socioeconomic status. In a meta-analysis of studies on behavior of health care providers, Hall et al. (1988) found that social class correlated positively with the information the provider gave the patient: patients of higher social position received superior care, and whites received better care than blacks. Hall et al. found that health care providers were also more likely to engage in “positive talk” with their white patients. That communication style, characterized by reassurance, approbation, and encouragement, correlated positively with patient adherence to treatment.
  3. A related phenomenon is that a person's health practices partially reflect the attitudes of his or her health care provider. A clinician's appraisal of a patient's ability to modify health-affecting behavior may cause the clinician to lower his or her expectations for favorable outcomes (Hall et al., 1988). This practice does not necessarily reflect client-centered care. Rather, it may create a self-fulfilling prophecy: with lowered expectations, a health care provider may make less effort to promote health-enhancing behaviors and thereby help to perpetuate the clients' health-damaging behavior. This relationship between provider attitudes and patient practices has been evident in clinical settings, particularly when practitioners have been unsuccessful in helping their patients lose weight, stop smoking, or control their blood sugar. Discouraged by such failure, clinicians can become more accepting of the adverse behaviors of other patients, particularly those from the same racial, social, or ethnic group.

In this section we examine the influence of (and interaction between) race and socioeconomic status on the following attitudes, behaviors, and social environments of black and white older men and women: (1) perceived health and health behavior; (2) diet, smoking, alcohol consumption, and physical exercise; (3) prevalence and control of diabetes and hypertension; and (4) social networks and support. We selected these risk factors because of the recognition that their effects on many chronic diseases, especially cardiovascular and cerebrovascular disease, are interrelated.

Self-Appraisal of Health and Health Behavior

Many investigators suggest that it is important to study preventive health practices of older adults in order to elucidate the behavioral determinants of successful aging (Lubben et al., 1989b; Stults, 1984; Rowe, 1987; Kane, 1985). Others explore racial differences in health practices in an attempt to understand why blacks experience excess morbidity and mortality (Rogers, 1992; Mutchler and Burr, 1991; Duelberg, 1992). Comparisons of health-protective attitudes and behaviors among blacks and whites are also made for the purpose of explaining the possible crossover of their mortality rates (Ford et al., 1990). In a study of the interaction between the extent to which adults worry about heart disease, Ransford (1986) found that at all educational levels, blacks reported greater concern about heart disease. Surprisingly, those concerns translated into more health-protective behavior (exercise, changed dietary habits, or smoking cessation) only among blacks who had less than a high school education. In a survey examining the relationship between socioeconomic characteristics and health beliefs, Weissfeld et al. (1990) found that blacks and those of lower socioeconomic status (measured by a composite of education and income level) placed a higher value on healthful habits than others. After adjustment for race, the association between perceived health threats and socioeconomic status diminished. Ford et al. (1990) reported the results of a cohort study that demonstrated that blacks over age 74 considered themselves less healthy (including mentally) than did whites. A survey by Mutchler and Burr (1991) demonstrated a similar disparity in self-appraisal of health that persisted after adjustment for socioeconomic status. Lubben et al. (1989a) found a similarity between the health practices of black and white Medicaid recipients in California over age 65. Whites were more likely to maintain their desired body weight, but there were no significant racial differences in smoking, exercising, alcohol intake, or maintenance of social networks.

Thus, these studies would suggest, albeit not conclusively, that blacks, especially older blacks, are concerned about chronic disease, appraise their health status as worse than whites, and place an equal or higher value on health-protective behaviors. These findings reinforce the theory that behavior change does not rest on knowledge and awareness alone.

Health Behaviors

Diet and Exercise

The behavioral aspects of obesity have an impact on the difference in disease patterns between blacks and whites. Obesity is associated with hypertension, non-insulin-dependent diabetes mellitus, and osteoarthritis (Pi-Sunyer, 1993), all of which are more prevalent among blacks (National Center for Health Statistics, 1990). It is now believed that weight loss and maintenance are probably best achieved through a combination of dietary control and increased physical activity (DiPietro et al., 1993). For this reason, we will discuss issues related to diet and exercise together.

As many as 60 percent of black women are overweight (Kumanyika, 1987). The higher prevalence of obesity in older black women is related to their higher baseline body mass index in middle age (Williamson et al., 1991) and to their being less likely to lose weight during those years (Kahn et al., 1991). That black women express a positive attitude about their weight more often than white women is a behavioral element that is felt to influence the potential success of dietary interventions designed for them. Although studies about weight perceptions have generally been done in younger populations (Kumanyika et al., 1993; Rucker and Cash, 1991), Stevens et al. (1994) found that elderly black women also have a greater acceptance of being overweight than their white counterparts. Socioeconomic status was not included in that analysis. In a study of dietary patterns of centenarians and adults who were in their sixties and eighties, Johnson et al. (1992) found that three times as many blacks as whites reported a desire to eat more nutritiously. In that study, blacks also reported larger fluctuations in their adult weight. Again, we see that blacks maintain a desire to eat more nutritiously although they tend to be more overweight.

Comparisons of activity levels in elderly racial groups have more often included examination of the effects of socioeconomic status than have studies of dietary patterns (Folsom et al., 1990; Sheridan et al., 1993; Kaplan et al., 1987). Numerous studies have shown that blacks exercise less than whites and engage is less leisure time physical activity (Burke et al., 1992; Washburn et al., 1992; Heath and Smith, 1994): 19 percent of black versus 29.9 percent of white women over age 65 exercise or play sports regularly. The comparative proportions of black and white men who so exercise is 25.1 percent to 36.3 percent. In the cohort above 65 years old, 38.4 percent of black versus 44.9 percent of white women walk for exercise, while for men the rates are 42.6 percent and 52.4 percent, respectively. Folsom et al. (1990) reported data from the Minnesota Heart Survey that demonstrated that the largest difference in leisure time physical activity between blacks and whites occurred in those with a high school education or less. Using data from the National Health Interview Survey (NHIS) and adjusting for education and income, Duelberg showed that these racial differences in exercise patterns could not be entirely explained by socioeconomic status. In a study of how education level and race were associated with risk factors for coronary artery disease among younger men and women, Sheridan et al. (1993) found that lack of regular exercise was more common in blacks, even after adjustment for education. Kaplan et al. (1987) reported that decreased physical activity was a mortality risk factor independent of socioeconomic position in Alameda County elders.

Cigarette Smoking

The prevalence of cigarette smoking among older Americans is complex and is related to cohort effects and different norms among men and women. Population-based estimates provide valuable information about smoking patterns in all blacks and whites, as well as in elderly subgroups. Some general observations are that (1) overall, whites are heavier smokers than blacks; (2) black men smoke more and black women smoke less than their white counterparts; and (3) smoking rates are higher and cessation rates are lower in persons with less education (Centers for Disease Control [CDC] 1994).

In the 1991 NHIS (CDC, 1994), the prevalence of smoking was higher in blacks than whites (29.1% vs. 25.5%), although a higher percentage of blacks reported having never smoked (56.2% vs. 48.9%). Also in that survey, 50 percent of respondents over age 65 had never smoked, compared with 67 percent in 1965. This decrease in the percentage of elderly abstainers was not consistent among all gender and racial subgroups. Whereas abstention rates increased among elderly men, they decreased in women. Moreover, between 1965 and 1991, the net increase in the percentage of those who had never smoked was twice as high for older blacks as for whites (10.4% vs. 5.1%). Despite these trends, smoking rates are higher for elderly black men than for their white counterparts. In women, the rates are higher in whites than in blacks. These differences are illustrated by NHIS data (National Center for Health Statistics, 1993), which show that in 1987, 30 percent of elderly black men were smokers, compared with 16 percent of white men. Whereas 13.9 percent of older white women in that survey smoked, only 11.7 percent of black women did. Although the racial differences in the percentage of smokers among both genders was a bit lower in the 1990 NHIS (National Center for Health Statistics, 1993), the overall trends were identical.

Smaller studies have also described the epidemiology of smoking in older adults. Sheridan et al. (1993) performed a cross-sectional analysis of cardiovascular risk factors in urban South Carolina residents. In his sample, the prevalence of smoking in elderly black and white men and women was very similar to that found in the population-based studies. Other investigators have described similar differences in the tobacco use of black and white elders. For example, in a study of hypertensives, Svetkey et al. (1993) found mean lifetime cigarette consumption higher in whites than blacks. Among Medi-Cal recipients in the study by Lubben et al. (1989b), blacks were 1.66 times less likely to have ever smoked than whites. Finally, Aronow and Kronson (1991) demonstrated a higher smoking rate (9% vs. 7%) in African Americans in their analysis of coronary risk factors in the elderly. Additional research is needed to identify possible correlates (including social class) of tobacco use for older blacks and whites. Smoking cessation confers health benefits on the elderly who have already developed smoking-related cardiovascular or pulmonary disease (CDC, 1990). If low social class explains a large share of the described black-white differences in smoking behavior, strategies to promote quitting should be directed toward a subset of elders of low socioeconomic position, rather than toward those in either racial category.

Alcohol Use

Most studies of alcohol consumption show either no differences by race or increased consumption among white men and women. In the 1990 NHIS (National Center for Health Statistics, 1993), whites displayed a higher overall prevalence of alcohol consumption than did blacks; a similar comparison was also evident in the cohort aged 65 and older: 9 percent of white men and 3.4 percent of white women reported drinking an ounce or more of ethanol in the 2 weeks prior to their interview, compared with 1.9 percent and 0.4 percent of black men and women, respectively. Although the research by Lubben et al. (1989b) on the health of poor elderly adults in California did not find a racial difference in alcohol patterns, a cross-sectional investigation by Molgaard et al. (1990) did. In that study of San Diego County adults age 45 and older, the overall prevalence of drinking was higher in whites than in blacks (79% vs. 50%). For adults 65 and older, drinking rates were also higher in whites (74% vs. 49%). However, those percentages represent only the self-reported prevalence of drinking after age 40. Therefore, direct inferences about current age-specific drinking patterns could not be made.

Health Practices of Elderly Diabetics and Hypertensives

Exercise, dietary control, and adherence to therapy all influence morbidity associated with hypertension and diabetes. Exercise enhances weight reduction with resultant improvement in glucose tolerance in diabetics (Ruderman et al., 1990). However, control of high blood sugars alone does not attenuate diabetics' risk for coronary artery disease. Management of co-morbid conditions such as hypercholesterolemia, hypertension, and cigarette smoking is an essential aspect of diabetic care. Although diabetes is twice as prevalent in older black Americans than in whites (National Center for Health Statistics, 1990), as previously noted, African Americans are less likely to engage in a level of physical activity that might help modulate this condition. This behavioral tendency was demonstrated by Spangler and Konen (1993), who conducted a survey of type I and type II diabetics and found that white diabetics were more likely not only to engage in high levels of exercise, but also to be former smokers. Cowie et al. (1993), evaluating participants in the National Health and Nutrition Survey (NHANES) II, showed that racial differences in the risk for non-insulin-dependent diabetes could not be explained by differences in social status but were strongly correlated with obesity. In a descriptive study of the health behaviors of a small sample of elderly black diabetics in San Francisco, Reid (1992) found that older patients relied more on traditional (rather than folk) medical care than a similar group of younger people. In that study, elders with multiple co-morbid conditions took better care of themselves and adhered more to treatment.

Exercise also has beneficial effects on blood pressure control. Like studies of diabetics, examinations of the health practices of elderly hypertensives often focus on behavioral determinants of obesity (diet, exercise) and salt intake (Svetkey et al., 1993; Spangler and Konen, 1993). Svetkey's study of hypertensive EPESE participants in North Carolina demonstrated that elderly black hypertensives actually had lower salt intake than whites. Lower salt consumption was inversely associated with hypertension risk. In that study, adjustment for socioeconomic status reduced the association between hypertension and race, but did not eliminate it. Having a low-sodium diet was also characteristic of the blacks in Johnson's report on the nutritional habits of centenarians. The authors in both studies hypothesized that black patients, aware of their disproportionate risk of high blood pressure, listen more to messages to moderate their salt intake. Another indicator of such adherence can be found in data from the National Center for Health Statistics (1993), which reveal that a slightly higher percentage of elderly black women and men with high blood pressure also reported that they were taking antihypertensive medications. Sheridan et al. (1993) reported the results of a cross-sectional survey that explored the association between ethnicity and educational level with several behavioral risks for cardiovascular disease in adults between the ages of 20 and 64. Black respondents were less likely than whites to know their own blood pressure or cholesterol level or to accurately indicate what the desirable levels of these risk factors were. These racial differences persisted after adjustment for educational attainment. Since this study did not examine the prevalence of diseases that increase the risk of cardiovascular disease, data about hypertensives and what proportion of them adhered to therapy were not reported. Age-specific data on that topic might clarify whether the observed health practices of elders have been consistent throughout their adult lives and whether the differences in health behaviors between blacks and whites increase or decrease with age. Including social position in such analyses would help illuminate how much it influences those differences.

Health promotion initiatives based on most of the findings presented in this paper would usually reflect secondary or tertiary prevention efforts given the high prevalence of chronic disease in older age groups. How much impact the initiation of cardiovascular risk reduction efforts for these age groups will ultimately have on their mortality will become clearer as prospective studies exploring this issue are done. Most work of this kind has been in younger adults. However, additional studies on the effect that race and socioeconomic status have on health practices of the elderly not only are important for understanding the determinants of successful aging but also can provide insight into the assumption that improvement in education and income level in blacks may be associated with the adoption of more health-enhancing practices and, ultimately, improved morbidity and mortality. If health differences persist in the context of blacks' and whites' having congruent behaviors and seemingly greater economic equality, there is more reason both to examine the influence of racism on overall health and to explicate the role that genetic predisposition plays in certain diseases such as diabetes and hypertension.

Social Networks and Social Support

There is a large body of evidence documenting the role of social networks and social support in influencing survival and life expectancy (Berkman, 1995, House et al., 1988). Among the elderly, social relationships, especially those that are explicitly defined as supportive, are important not only in reducing mortality risk but also in delaying institutionalization and in maintaining functional independence. Studies to date suggest that the important components of network structure or support are the size of networks, the availability of support, and its perceived adequacy. Furthermore, it seems to be more important that the individual identify someone or some group or organization as fulfilling a specific role than that a specific type of person (i.e., kin, non-kin) fill a role. This suggests that older men and women have a number of options that will be equally “protective” in providing supportive relationships. In other words, one doesn't have to be married to experience decreased mortality risk if other intimate ties are available. Most studies indicate that it is only in the absence of any close ties that risks become elevated.

While the major aim of this section is to explore differences in network structure and social support among older black and white men and women, it is important to recognize that social networks assume forms that reflect the broad structure of society and an individual's position in that social structure (Craven and Wellman, 1973; Tilly, 1972; Bultena, 1969). Socioeconomic status, level of urbanization, and geographic and occupational mobility are all associated with network structure. For instance, in the Alameda County study (Berkman and Breslow, 1983), analysis of individual components of the Social Network Index indicated that people of lower socioeconomic status were less likely to be married and to be affiliated with church and voluntary organizations than others. Only with respect to close friends and relatives do lower class men and women compare favorably. Several ethnographers have documented the relative extensiveness of the working class person's contacts with friends and extended family and have described it as a major resource in the face of other considerable challenges and stresses (Stack, 1974; Liebow, 1967). Thus, it is important to review the evidence that black-white differences in networks and support may be related to social and economic conditions, which may vary between ethnic groups. Only in some studies have the social and economic conditions been thoroughly discussed.

The Alameda County study, conducted in the mid-1960s and 1970s, suggests that there are very small differences between blacks and whites in the percentage of men and women who appear to be isolated on the Social Network Index, a composite index of marital status, contact with friends and relatives, and church and group membership (Berkman and Breslow, 1983). Among men, 38 percent of blacks and 32 percent of whites score in the top two categories of isolation, and 46 percent of black women compared with 44 percent of white women score in the same range. Of more interest are variations within subcategories of the index.

In this area, blacks were less likely to be married or to report membership in voluntary organizations than whites but they were more likely to belong to church groups and to have more contacts with friends and relatives. The strong role of the church in the life of black Americans and the importance of the extended family and three-generation networks have been extensively noted in many studies, most of which are qualitative ethnographic studies (George, 1988; Taylor and Chatters, 1986).

In one of the most thorough and insightful reviews of social networks and support among older black Americans, Taylor (1988) provides both a theoretical overview of a model of family support relationships and a review of non-kin sources of support. The reader is referred to this paper for a more extensive review of the literature than is provided here. Among the more notable studies in this newer wave of research on black Americans is the National Survey of Black Americans. This survey studied a nationally representative cross-sectional sample of the adult black population living in the continental United States that was based on the 1970 census. The study resulted in 2,107 interviews with a response rate of 69 percent. Among a subsample (n = 581) of older blacks, 67 percent of respondents said that their families were very close in their feelings for one another, and only 8 percent stated that their families were either not too close or not close at all (Chatters et al., 1985, 1986). Of helper networks, 56 percent were composed exclusively of immediate family and another 33 percent were composed of immediate family and others. Neither family income nor education was associated with network composition. There were, however, regional variations, with southern blacks having larger helper networks than those in other regions of the United States. Since most studies have been based in communities rather than on national samples, it will be important to consider whether regional variations may account for differences in network structure and support between black and whites.

Finally, there is a consistent body of evidence indicating that older blacks are less likely to live alone than older whites. While older blacks are less likely to be married, they are more likely to live in extended households with children and grandchildren. Three-generation households or households whose members do not belong to a nuclear family may have particular advantages for old people, especially those with limited economic resources or functional limitations. When families live as units that share resources and tasks, the individuals in those families may experience benefits that they would not obtain alone or in small nuclear families. At the same time, multigenerational households, if they are bound together only for financial or logistic reasons, may place strains on older people that have not been well identified (George, 1988).

As is suggested by George, there are more similarities in the maintenance of social relationships in late life between blacks and whites than there are differences. The major similarities between blacks and whites are that (1) black and white elders maintain frequent contact with families, especially their children, and (2) both black and white women tend to have more extensive contacts with family members than do men. Both blacks and whites have a similar hierarchy by which support is provided to older members of the family. This principle of hierarchy developed by Cantor (1979, 1983) suggests that spouses are the first choice of source of support, followed by children (especially daughters) if spouses are unavailable. More distant relatives are the source of support if members of a nuclear family are unavailable. George reviews multiple studies indicating that this hierarchy applies to both racial groups.

On the other hand, most studies indicate that older blacks both receive and provide more social support than do older whites.


In this section we examine differences in health and social behaviors among older black and white men and women. Data are analyzed from the New Haven EPESE program, which is based on a probability sample of 2,806 noninstitutionalized men and women living in New Haven, Connecticut, in 1982. Details of the sampling frame have been reported elsewhere (Berkman et al., 1986). This study has the advantage of including (1) a very heterogeneous sample in terms of race, socioeconomic status, and health status and (2) a broad array of health behaviors and social conditions. Since it is a community sample of older men and women in an urban area of the Northeast, the associations found in this cohort may not be generalizable to other areas of the United States. In fact, our earlier review suggested that there may be substantial geographic variations in patterns of behaviors and social conditions.

Briefly, the sampling frame included samples drawn from three housing strata reflecting the most common types of housing for those aged 65 years and older: (1) public elderly housing, which is age and income restricted; (2) private elderly housing, which is age restricted; and (3) private community houses and apartments. In public housing, all eligible persons were interviewed, whereas women were subsampled in both the private and the community strata. The overall response rate was 82 percent.

Table 5-1 shows the distribution of a broad range of risk factors, including both standard risk factors such as smoking, alcohol consumption, and prevalence of chronic diseases (e.g., diabetes and hypertension) and aspects of social networks and support. Neither blacks nor whites have a consistently greater prevalence of high-risk behaviors. For instance, older black men and women are less likely to have ever smoked cigarettes than are whites, but among people who have ever smoked, whites are significantly more likely than blacks to have quit smoking. Though data are not shown for cohort effects for both blacks and whites, younger old men and women (i.e., those in their sixties) are more likely to be current smokers than are older generations.

TABLE 5-1. Percentage of Blacks and Whites With Various High-Risk Health and Social Behaviors by Gender: New Haven EPESE (N = 2,806), 1982, Weighted Percentages.


Percentage of Blacks and Whites With Various High-Risk Health and Social Behaviors by Gender: New Haven EPESE (N = 2,806), 1982, Weighted Percentages.

Black and white men show no differences in body weight. Substantial differences in weight are observed, however, between black and white women. Almost 50 percent of black women fall into the upper tercile of body mass assessed by the Quetelet Index compared with 28 percent of white women (p ≥ .01).

With regard to alcohol consumption, the New Haven data are consistent with many other reports indicating that older blacks are less likely to consume alcohol than are whites. Among black women, only 21 percent report any alcohol consumption compared with 49 percent of white women and 50 percent and 67 percent of black and white men, respectively. This lower prevalence of alcohol consumption and, to a lesser extent, cigarette consumption among black Americans, especially women, in part reflects the strong cultural value placed on abstinence in that older generation. Since norms regarding these behaviors have changed over time, we may expect to see substantially different patterns of use among new cohorts of Americans as they enter older ages.

As discussed in the earlier section, blacks tend to have higher levels of chronic conditions that lead to greatly increased mortality risks. In the New Haven EPESE study, the prevalence of diabetes and hypertension among black and white men is similar, but black women have a higher prevalence than all other groups. The prevalence of self-reports of physician-diagnosed hypertension is 68 percent among black women compared with 47 percent among white women (p ≥ .001). The prevalence of diabetes is also higher in black women compared with white women (20% vs. 12%), although this difference is not statistically significant.

A very balanced picture comes into view when we examine the distribution of aspects of social network and support. First we examine a summary measure of social ties composed of four components: marital status, contacts with friends and relatives, church membership, and group membership. Contacts with friends and relatives is a summary measure of the number of close friends and relatives the respondent identified. Group membership is assessed by an affirmative response to participation in any groups (e.g., voluntary, public service, self-help, community, or work related). The percentage of people who scored zero on this summary scale, indicating social isolation, is very similar across racial and gender groups. There are no statistically significant differences among these groups. When each component of the measure is examined individually, a different picture emerges. For instance, both white men and women are more likely to be married than their black counterparts (p ≥ .001). Black men and women, on the other hand, are more likely than their white counterparts to belong to a church group. Almost 60 percent of black men belong to a church versus 46 percent of white men (p ≥ .01). The respective percentages for women are 67 percent versus 53 percent (p ≥ .001). Contact with friends and relatives is very similar among all groups. Only among women are racial differences in group involvement in voluntary activities revealed, with black women being less likely to report voluntary group membership than white women (p ≥ .001) or men of either race.

We now turn to an examination of differences between older blacks and whites in the availability of social support. We report here on two domains. The availability of instrumental support is assessed by a response of yes to the question, “When you need some extra help, can you count on anyone to help you with daily tasks like grocery shopping, house cleaning, cooking, telephoning, giving you a ride?” The availability of emotional support is assessed by a response of yes to the question, “Can you count on anyone to provide you with emotional support (talking over problems or helping you make a difficult decision)?” There are no differences in the degree to which blacks and whites report the availability of emotional and instrumental support though, in general, men are less likely than women to report having support. Overall, these data indicate that there are not substantial differences in the extent to which blacks and whites maintain social ties and obtain social support. While there are differences in the degree to which blacks and whites engage in specific relationships, overall differences balance out. Since most data have indicated that it is overall levels of social ties or emotional support that are critical to well-being rather than absolute levels on any single one domain or type of relationship, it is unlikely that the extent to which blacks and whites maintain social networks accounts for differences in health status between older blacks and whites.

The above percentages do not take into consideration the possibility that differences between races may be the result of underlying differences between blacks and whites in socioeconomic status. Indeed, many recent reports have suggested that race per se does not reflect a meaningful biological grouping but rather a social grouping defined by social and economic circumstances. To the extent that differences between blacks and whites reflect underlying differences in socioeconomic status or position, we would expect that controlling statistically for socioeconomic status would diminish differences between blacks and whites in health and social behaviors.

Table 5-2 shows the results of individual multiple logistic models in which we examined the odds that black men and women would maintain a specific behavior or have a specific condition compared with whites. For each behavior, we conducted two analyses: one in which we controlled for age and a second in which we controlled for age and education, the strongest indicator we have of socioeconomic status in this study. In all analyses, we used the Professional Software for Survey Data Analysis (SUDAAN) (SAS Inc., Cary, N.C.) to adjust for possible design effects of the stratified sampling design that was used to construct the New Haven EPESE cohort. In 1982, we designed the cohort by stratifying on type of housing and we oversampled men. The SUDAAN software computes sampling variances for parameter estimates using the Taylor series approximation method. It is important to use design-based variance estimates with these data (Freeman et al., 1992).

TABLE 5-2. Odds Ratios (ORs) for High-Risk Health and Social Behaviors for Black and White Older Men and Women (Results of Weighted Multiple Logistic Regressions, New Haven EPESE Baseline, 1982, Data).


Odds Ratios (ORs) for High-Risk Health and Social Behaviors for Black and White Older Men and Women (Results of Weighted Multiple Logistic Regressions, New Haven EPESE Baseline, 1982, Data).

Table 5-2 reveals that some differences between blacks and whites in the prevalence of health-promoting and health-damaging behaviors reflect underlying distributions of age and, more importantly, socioeconomic status. For instance, while the odds of being a current or past smoker are less for blacks than for whites, once we adjust for covariates, especially education, these differences become nonsignificant. On the other hand, even adjustment for age and education does not reduce differences in alcohol consumption. Black women are about a third as likely to drink alcohol as white women, and black men are half as likely to drink as white men. When we control for age and especially for education the odds of being hypertensive or diabetic are reduced, while for black women the odds of being overweight remain substantial.

The availability of social networks and social support patterns are consistent with earlier bivariate analyses. While there are no significant differences in the level of social ties when we look at the summary measure, there are differences in the specific domains even after adjustment for age and educational level. Specifically, black men and women are about twice as likely to be unmarried as whites.

For instance, the odds ratios (ORs) for black men are 2.29 with 95 percent confidence intervals (CIs) not overlapping 1 (CI = 1.58, 3.33).

Black men are also somewhat less likely than white men to have few contacts with friends and relatives (OR = 1.28; 95% CI = 1.02, 1.62). After adjusting for age and education, however, blacks—both men and women—are much more likely to belong to church than are whites.

These multivariate analyses indicate that black-white differences in most health behaviors are only slightly reduced by controlling for education. Greater reductions were seen for the risks associated with a history of hypertension or diabetes when we controlled for education. Thus, part of the explanation for increased rates of these important causes of morbidity and mortality among black women may be their lower socioeconomic status.


In this paper we have examined differences between black and white older men and women in a broad array of health-promoting and health-damaging behaviors. Overall, we find inconsistent patterns, with blacks' having a higher prevalence of some behaviors and chronic conditions and a lower prevalence of other conditions. With regard to social networks and support, while there are differences in the degree to which older black Americans maintain some type of ties when compared with whites, there are no overall differences in summary measures of social ties or social support. Finally, when the level of education is introduced as control for socioeconomic status, differences between blacks and whites are occasionally reduced, but the strongest associations remain significant.

Our results suggest that the mortality disadvantage experienced by older black Americans is unlikely to be the sole result of their high-risk practices or behaviors. In addition, differences in the extent to which blacks and whites maintain close and effective social ties are minimal. Other conditions more closely linked to social class and environmental exposures, as well as stressful experiences related to discrimination, are worthy of serious investigation.


  • American Medical Association Council on Ethical and Judicial Affairs. Black-white disparities in health care. Journal of the American Medical Association. 1990;263(17):2344–2346. [PubMed: 2182918]
  • Aronow WS, Kronson I. Prevalence of coronary risk factors in blacks and whites. Journal of the American Geriatric Society. 1991;39:567–570. [PubMed: 1828074]
  • Berkman LF. The role of social relations in health promotion. Psychosomatic Medicine. 1995;57(3):245–254. [PubMed: 7652125]
  • Berkman LF, Berkman CS, Kasl S, Freeman DH Jr., Leo L, Ostfeld AM, Cornoni-Huntley J, Brody JA. Depressive symptoms in relation to physical health and functioning in the elderly. American Journal of Epidemiology. 1986;124:372–388. [PubMed: 3740038]
  • Berkman LF, Breslow L. Health and Ways of Living: The Alameda County Study. New York: Oxford University Press; 1983.
  • Blendon RJ, Aiken LH, Freeman HE, Corey CR. Access to medical care for black and white Americans. A matter of continuing concern. Journal of the American Medical Association. 1989;261(2):278–281. [PubMed: 2909026]
  • Braithwaite RL, Lythcott N. Community empowerment as a strategy for health promotion for black and other minority populations. Journal of the American Medical Association. 1989;261(2):282–283. [PubMed: 2909028]
  • Bultena GL. Rural-urban differences in the familial interaction of the aged. Rural Sociology. 1969;34(1):5–15.
  • Burke GL, Savage PJ, Manolio TA, Sprafka JM, Wagenknecht LE, Sidney S, Perkins LL, Liu K, Jacobs DR. Correlates of obesity in young black and white women: The CARDIA study. American Journal of Public Health. 1992;82(12):1621–1625. [PMC free article: PMC1694535] [PubMed: 1456336]
  • Burstin HR, Lipsitz SR, Brennan TA. Socioeconomic status and risk for substandard medical care. Journal of the American Medical Association. 1992;268(17):2383–2387. [PubMed: 1404794]
  • Cantor MH. The informal support system of New York's inner city elderly: Is ethnicity a factor? In: Gelfand DE, Kutzik AJ, editors. Ethnicity and Aging: Theory, Research and Policy. New York: Springer; 1979. pp. 153–174.
  • Cantor MH. Strain among caregivers: A study of experience in the United States. Gerontologist. 1983;23:597–604. [PubMed: 6662373]
  • Centers for Disease Control. The surgeon general's 1990 report on the health benefits of smoking cessation. Morbidity and Mortality Weekly Report. 1990;39:RR12. [PubMed: 2120576]
  • Centers for Disease Control. Surveillance for selected tobacco use behaviors—United States, 1900-1994. Morbidity and Mortality Weekly Report CDC Surveillance Summaries. 1994;43:3. [PubMed: 7969014]
  • Chatters LM, Taylor RJ, Jackson JS. Size and composition of the informal helper networks of elderly blacks. Journal of Gerontology. 1985;40(5):605–614. [PubMed: 3875644]
  • Chatters LM, Taylor RJ, Jackson JS. Aged blacks' choices for an informal helper network. Journal of Gerontology. 1986;41(1):94–100. [PubMed: 3510245]
  • Cowie CC, Harris MI, Silverman RE, Johnson EW, Rust KF. Effect of multiple risk factors on differences between blacks and whites in the prevalence of non-insulin dependent diabetes mellitus in the United States. American Journal of Epidemiology. 1993;137(7):719–732. [PubMed: 8484363]
  • Craven S, Wellman B. The network city. Sociological Enquiry. 1973;43:57–88.
  • DiPietro L, Williamson DF, Caspersen CJ, Eaker E. The descriptive epidemiology of selected physical activities and body weight among adults trying to lose weight: The Behavioral Risk Factor Surveillance System survey, 1989. International Journal of Obesity. 1993;17:69–76. [PubMed: 8384167]
  • Duelberg SI. Preventive health behavior among black and white women in urban and rural areas. Social Science and Medicine. 1992;34(2):191–198. [PubMed: 1738872]
  • Folsom AR, Cook TC, Sprafka JM, Burke GL, Norstead SW, Jacobs DR. Differences in leisure time physical activity levels between blacks and whites in population-based samples: The Minnesota Heart Survey. Journal of Behavioral Medicine. 1990;14(1):1–9. [PubMed: 2038041]
  • Ford AB, Haug MR, Jones PK, Roy AW, Folmar SJ. Race-related differences among elderly urban residents: A cohort study, 1975-1984. Journal of Gerontology: Social Sciences. 1990;45(4):163–171. [PubMed: 2365977]
  • Freeman DH, Livingston MM, Leaf PJ, Berkman LF. Sampling strategies for studying older populations. In: Wallace R, Woolson R, editors. The Epidemiologic Study of the Elderly. New York: Oxford University Press; 1992. pp. 46–59.
  • George LK. Social participation in later life: Black-white differences. In: Jackson JS, editor. The Black American Elderly: Research on Physical and Psychosocial Health. New York: Springer; 1988. pp. 99–126.
  • Giles WH, Anda RF, Jones DH, Serdula MK, Merritt RK, DeStefano F. Recent trends in identification and treatment of high blood cholesterol by physicians: Progress and missed opportunities. Journal of the American Medical Association. 1993;269(9):1133–1138. [PubMed: 8240474]
  • Hall JA, Roter DL, Katz NR. Meta-analysis of correlates of provider behavior in medical encounters. Medical Care. 1988;26(7):657–675. [PubMed: 3292851]
  • Heath GW, Smith JD. Physical activity patterns among adults in Georgia: Results from the 1990 Behavioral Risk Factor Surveillance System. SouthernMedical Journal. 1994;87(4):435–439. [PubMed: 8153767]
  • House JS, Landis KR, Umberson D. Social relationships and health. Science. 1988;241(4865):540–545. [PubMed: 3399889]
  • Johnson MA, Brown MA, Poon LW, Martin P, Clayton GM. Nutritional patterns of centenarians. International Journal of Aging and Human Development. 1992;34(1):57–76. [PubMed: 1737661]
  • Kahn HS, Williamson DF, Stevens JA. Race and weight change in U.S. women: The roles of socioeconomic and marital status. American Journal of Public Health. 1991;81(3):319–323. [PMC free article: PMC1405001] [PubMed: 2036117]
  • Kane RL. Prevention and the elderly: Risk factors. Health Services Research. 1985;19:945–1006. [PMC free article: PMC1068847] [PubMed: 3918961]
  • Kaplan GA, Seeman TE, Cohen RD, Knudsen LP, Guralnik J. Mortality among the elderly in the Alameda County Study: Behavioral and demographic risk factors. American Journal of Public Health. 1987;77(3):307–312. [PMC free article: PMC1646902] [PubMed: 3812836]
  • Kumanyika S. Obesity in black women. Epidemiologic Reviews. 1987;9:31–50. [PubMed: 3315719]
  • Kumanyika S, Wilson J, Guilford-Davenport M. Weight-related attitudes and behaviors of black women. Journal of the American Dietetic Association. 1993;93:416–422. [PubMed: 8454809]
  • Liebow E. Tally's Corner: A Study of Negro Streetcorner Men. Boston: Little, Brown; 1967.
  • Lubben JE, Weiler PG, Chi I. Gender and ethnic differences in the health practices of the elderly poor. Journal of Clinical Epidemiology. 1989;42(8):725–733. [PubMed: 2760663]
  • Lubben JE, Weiler PG, Chi I. Health practices of the elderly poor. American Journal of Public Health. 1989;79(6):731–734. [PMC free article: PMC1349632] [PubMed: 2658629]
  • Molgaard CA, Nakamura CM, Stanford EP, Peddecord KM, Morton DJ. Prevalence of alcohol consumption among older persons. Journal of Community Health. 1990;15(4):239–251. [PubMed: 2212094]
  • Mutchler JE, Burr JA. Racial differences and health care utilization in later life: The effect of socioeconomic status. Journal of Health and Social Behavior. 1991;32:342–356. [PubMed: 1765625]
  • National Center for Health Statistics. Current Estimates from the National Health Interview Survey, 1989. Hyattsville, MD: Public Health Service; 1990.
  • National Center for Health Statistics. Health, United States, 1990. Hyattsville, MD: Public Health Service; 1991.
  • National Center for Health Statistics. Health Promotion and Disease Prevention, United States, 1990. Hyattsville, MD: Public Health Service; 1993.
  • Oldridge NB. Compliance and exercise in primary and secondary prevention of coronary heart disease: A review. Preventative Medicine. 1982;11:56–70. [PubMed: 7079247]
  • Pi-Sunyer FX. Medical hazards of obesity. Annals of Internal Medicine. 1993;119(7):655–660. [PubMed: 8363192]
  • Ransford HE. Race, heart disease, worry and health protective behavior. Social Science and Medicine. 1986;2(12):1355–1362. [PubMed: 3738559]
  • Reid BV. It's like you're down on a bed of affliction: Aging and diabetes among black Americans. Social Science and Medicine. 1992;34(12):1317–1323. [PubMed: 1529369]
  • Rigotti NA, McKool KM, Shiffman S. Predictors of smoking cessation after coronary artery bypass graft surgery: Results of a randomized trial with 5-year follow-up. Annals of Internal Medicine. 1994;120(4):287–293. [PubMed: 8291821]
  • Rogers RP. Living and dying in the U.S.A.: Sociodemographic determinants of death among blacks and whites. Demography. 1992;29(2):287–303. [PubMed: 1607053]
  • Rowe JW. Human aging: Usual and successful. Science. 1987;237:143–149. [PubMed: 3299702]
  • Rucker C, Cash T. Body images, body size perceptions, and eating behaviors among African American and white college women. International Journal of Eating Disorders. 1991;12:291–299.
  • Ruderman N, Apelian AZ, Schneider SH. Exercise in therapy and prevention of type II diabetes. Diabetes Care. 1990;13(11):1163–1168. [PubMed: 2261837]
  • Sheridan DP, Hornung CA, McCutcheon EP, Wheeler FC. Demographic and educational differences in smoking in a tobacco growing state. American Journal of Preventive Medicine. 1993;9(3):155–159. [PubMed: 8347366]
  • Spangler JG, Konen JC. Predicting exercise and smoking behavior in diabetic and hypertensive patients. Archives of Family Medicine. 1993;2(Feb.):149–155. [PubMed: 8275184]
  • Stack CB. All Our Kin. New York: Harper and Row; 1974.
  • Stevens J, Kumanyika SK, Keil JE. Attitudes toward body size and dieting: Differences between elderly black and white women. American Journal of Public Health. 1994;84(8):322–325. [PMC free article: PMC1615469] [PubMed: 8059896]
  • Stults BM. Preventive health care and the elderly. Western Journal of American Medicine. 1984;141:832–845. [PMC free article: PMC1011218] [PubMed: 6395498]
  • Svetkey LP, George LK, Burchett BM, Morgan PA, Blazer DG. Black/white differences in hypertension in the elderly: An epidemiologic analysis in central North Carolina. American Journal of Epidemiology. 1993;137(1):64–73. [PubMed: 8434574]
  • Taylor RJ. Aging and supportive relationships. In: Jackson JS, editor. The Black American Elderly: Research on Physical and Psychosocial Health. New York: Springer; 1988. pp. 259–281.
  • Taylor RJ, Chatters LM. Church-based informal support among elderly blacks. The Gerontologist. 1986;26:637–642. [PubMed: 3803954]
  • Tilly C. An interactional scheme for analysis of communities, cities, and urbanization. Department of Sociology, University of Michigan; Ann Arbor: 1972. Unpublished manuscript.
  • U.S. Department of Health and Human Services. Chartbook on Health Data on Older Americans: United States, 1992. Vital and Health Statistics. 1993;3:29.
  • Washburn RA, Kline G, Lackland DT, Wheeler FC. Leisure time physical activity: Are there black/white differences? Preventative Medicine. 1992;21(1):127–135. [PubMed: 1738764]
  • Weissfeld JL, Kirscht JP, Brock BM. Health beliefs in a population: The Michigan blood pressure survey. Health Education Quarterly. 1990;17(2):141–155. [PubMed: 2347692]
  • Williamson DF, Kahn HS, Byers T. The 10-y incidence of obesity and major weight gain in black and white U.S. women aged 30-55 y. American Journal of Clinical Nutrition. 1991;53(suppl.):1515–1518. [PubMed: 2031480]
Copyright © 1997, National Academy of Sciences.
Bookshelf ID: NBK109839
PubReader format: click here to try


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (3.0M)

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to pubmed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...