In this study of community-dwelling older adults, being a black female was associated with the greatest risk of experiencing poor nutritional health. Black women were at highest nutritional risk, followed by black men, white women, and, last, white men. Specifically, black women were significantly more likely to report not having enough money to buy food, taking three or more medications, experiencing a recent significant change in weight, and having difficulty either shopping, cooking, or feeding one's self. Black women also were more likely to report experiencing a poor appetite and eating irregularly or skipping meals, although these associations were not statistically significant. The overall direction of these relationships, however, clearly indicates that black women were at increased risk for poor nutritional health regardless of the specific item used to assess poor nutrition. Additionally, black women were most likely to be socially isolated and to possess the lowest amounts of social support and capital. This relationship held across all measures and was statistically significant in regard to not having a reliable source of transportation (22.1%), to being limited in life–space to the room where one sleeps (24.9%), to limiting activities for fear of an attack (30.5%), and to not being married (79.5%). These findings are not new; much research over an extended period of time has documented the disadvantaged position that women and minorities experience in society across multiple domains, including health. Black women experience the cumulative disadvantage of being both female and black.
Despite black women's experience of being at greatest nutritional risk coupled with their being most socially isolated and possessing the lowest amount of social support and capital, of greatest note in this study are the findings that more indicators of social support and capital adversely affected nutrition risk for black men. Additionally, the indicators of social support and capital that adversely affected nutrition risk for black men were different from those that affected nutritional risk in other ethnic–gender groups. In fact, the other ethnic–gender groups were more similar than not, especially in regard to social isolation and lower income being associated with nutritional risk.
For black men, of most importance in this study are the findings that relate to social capital, including trust in community and regular religious participation. Both measures of trust in community, including limiting activities for fear of being attacked and experiencing discrimination within the past 6 months, were associated with increased nutritional risk only for black men. The threat of violence and the experience of discrimination is a very real one for black men who have lived their lives in a segregated south where threats may come from both the white community and the black community. Thus, older black men either may lack adequate community resources (such as grocery stores, nutrition services, etc.) because of the impoverished segregated communities they live in or their access to those resources that do exist may be curtailed because of the real or perceived fear and/or discrimination in attempting to do so. Others have made similar observations related to racism and discrimination and health overall (see e.g., Williams, 2000
). A limitation of our findings is that our measures of trust may be more accurately viewed as measures of distrust. An additional limitation of our measures of trust in community is that it is not known if participants experienced fear or discrimination within or outside their communities. Future research might more carefully distinguish between these in regard to nutritional risk.
Also at the level of social capital, not attending regular religious services was associated with higher nutritional risk for black men. Of particular note here is that black and white men were the groups least likely to report regular religious participation, 67.3% and 63.2%, respectively, while black women and white women were much more likely to report regular religious participation, 81.1% and 77.2%, respectively. Many eating activities take place within Southern churches, both black and white. While regular religious participation may lead to poorer nutritional outcomes such as obesity, regular religious participation may also provide opportunities to eat for those who may be undernourished. Additionally, women, traditionally the ones who prepare meals, are more likely to spend time at church events. Thus, black men who did not attend religious activities regularly may not have reaped the potential nutritional benefits of doing so. The same relationship may not have held for white men because the majority of them had wives who were still preparing meals, even though white men attended religious services less regularly than black men. The same relationship does not hold true for black men, because they were less likely to be married. Indeed, not being married (an indicator of social support) was a predictor for nutritional risk for black men only, lending further support to this interpretation of our findings.
Despite these unique associations for black men, there were similarities between the other groups. For example, for both black women and white men higher nutritional risk was associated with not having reliable transportation. Not having reliable transportation restricts access to resources that may be available to individuals either within or outside their communities. Other studies have also found that older women, particularly minority women, are less likely to drive or to stop driving earlier in life (Cape, 1987
; Kington, Reuben, Rogowski, & Lillard, 1994
; Marottoli et al., 1993
; Siegel, 1996
). Additionally, southern communities, and particularly, rural communities are less likely to have adequate public transportation so that older adults are more dependent on others (Arcury, Quandt, Bell, McDonald, & Vitolins, 1998
; Quandt & Rao, 1999
). Older white men are the group most likely to have driven throughout their lives and the group most likely to be responsible for transporting others, particularly their wives who never learned to drive. When these men are no longer able to drive, for any reason, they may be the group least likely to find others to transport them.
At the level of social support, for white women only there was an association between low levels of perceived social support and nutritional risk. These findings are not particularly unexpected because the instrument used to assess perceived social support tapped into perceptions of emotional types of support and whether individuals felt that significant others were aware of their needs or feelings and willing to help. Women have traditionally been the emotional caretakers in the family, both for spouses and children. It is not surprising, then, that those closest to these women would not be aware of their needs or feelings—as this would involve a reversal of roles, both the husband and wife roles and the parent and child roles.
For both black and white women and white men, lower levels of income were associated with higher levels of nutritional risk. This supports previous research, which has found that persons with lower socioeconomic status are more vulnerable to nutritional risk. Income is one measure of socioeconomic status that reflects spending power. In this study, education (also an indicator of socioeconomic status) did not predict nutritional risk for any group. Not having enough income may prevent persons from being able to obtain enough food to eat. Additionally, because indicators of social isolation were predictive of nutritional risk for all ethnic–gender groups in this study, persons with lower incomes may either not be able to afford to pay someone to grocery shop for them or to transport them to a grocery store.
Not unexpectedly, lower levels of independent life–space were associated with higher nutritional risk for all groups. This finding is not unexpected, as researchers have recently begun documenting the association between function and nutrition risk (Sharkey, 2002
; Lee & Frongillo, 2001
). Individuals may not be able to go outside of their homes to access community resources, and they may not be able to prepare or consume food within their own homes. Factors associated with social isolation, support, and capital, as well as reduced income, are especially important to consider in understanding nutritional risk of those with functional impairments compared to those without the same degree of impairment.
This research has several implications for nutritional policies directed towards older adults. First, clearly there is an opportunity to address nutritional risk in community-dwelling older adults by effecting changes in various aspects of community- and social life. The Elderly Nutrition Program (ENP), authorized by Congress under the auspices of the Administration on Aging under the Older Americans Act, is the largest program designed to coordinate community- and home-based nutrition services to older adults (Millen, Ohls, Ponza, & McCool, 2002
). As Wellman, Rosenzweig, and Lloyd (2002)
summarize, its original intent was to provide more than merely a meal. It was also intended to: decrease malnutrition, prevent physical and mental deterioration, promote health, reduce social isolation, link older adults to social and rehabilitative services, and provide low-cost, nutritionally sound meals. Food services (i.e., Meals on Wheels and congregate meals) are managed by State and Territorial Units on Aging and local Area Agencies on Aging and delivered by members of the local community (including volunteers). As Wellman (1999)
points out, however, funding has not increased to meet the demand for services; and only 7% of the high-risk population participates in the ENP, including approximately 25% of low-income and minority older adults. Wellman (1999)
recommends that programs need to more accurately assess those most in need of services. This paper provides some direction in considering what factors might be used in deciding who has the greatest need, particularly in the presence of health disparities.
Second, findings from this research suggest that not all programs or policies aimed at alleviating hunger in older adults will have the same impact because different factors affect nutritional risk for different groups. Older adults are a heterogeneous group whose diverse needs ought to be taken into account in this regard. For example, Wellman (1999)
also recommends that nutrition education, a component of the ENP that has not received the highest priority, needs to be developed and evaluated.
Lastly, this research found that what contributes most to nutritional risk is social isolation and lower income, particularly for black women, white women, and white men. This was indicated by several variables including not having adequate transportation and having limited independent life–space. For black men, however, measures of social support and capital were most important in predicting nutritional risk. Programs need to be targeted specifically to individuals who, for whatever reason, either do not have access to community food or food services or whose access is restricted, as in the case of black men. For example, initiatives need to focus on either providing transportation services so that individuals can access various food sources or bringing the food to participants. It is not enough to provide congregate meals, if persons are unable or afraid to get to the meal site. Because of the limitations of nutritional services currently available to high-risk older adults, different types of programs ought to be considered. For example, some communities are experimenting with providing older adults with several frozen meals at a single delivery. Frozen meals, groceries, and food commodities could also be mailed directly to the homes of homebound older adults. Such programs would not only directly improve nutrition, but indirectly impact on health, quality of life, and potentially enable older adults the means to maintain independence within the community.
The generalizability of findings from this study is limited by its reliance on data from older adults living in the southeastern region of the United States. This does not, however, limit the importance of the findings. One might expect that the more general findings of our study extend to other regions, as well. Future research might address these matters.