Share this article on:

Social and Demographic Predictors of Nutritional Risk: Cross-sectional Analyses From the UAB Study of Aging II

Buys, David R. PhD, MSPH, CPH; Kennedy, Richard E. MD, PhD; Williams, Courtney Phillips MPH; Brown, Cynthia J. MD, MSPH; Locher, Julie L. PhD, MSPH

doi: 10.1097/FCH.0000000000000180
Original Articles

Social factors may disparately affect access to food and nutritional risk among older adults by race and gender. This study assesses these associations using the Mini Nutritional Assessment among 414 community-dwelling persons 75+ years of age in Alabama. Descriptive analyses on the full sample and by African American men, African American women, white men, and white women showed that mean scores for the full Mini Nutritional Assessment differed by groups, with African American men and African American women having the highest nutritional risk. Multivariable analyses indicated that social factors affect nutritional risk differently by race and gender. Nutritional risk interventions are warranted for older adults.

Department of Food Science, Nutrition and Health Promotion, Mississippi State University, Mississippi State (Dr Buys); Division of Gerontology, Geriatrics and Palliative Care (Drs Kennedy, Brown, and Locher and Ms Williams), Department of Health Care Organizations and Policy, University of Alabama at Birmingham (Dr Locher); and Atlanta Geriatric Research, Education, and Clinical Center, Veterans Affairs Hospital, Birmingham, Alabama (Dr Brown).

Correspondence: David R. Buys, PhD, MSPH, CPH, Department of Food Science, Nutrition and Health Promotion, Mississippi State University, PO Box 9805, Mississippi State, MS 39762 (

The present study was supported by the National Institute on Aging Award nos. R01 AG16062, “Mobility Among Older African Americans and Whites—The UAB Study of Aging” (C.J.B.), and K07 AG043588, “Translational Nutrition and Aging Research Academic Career Leadership Award” (J.L.L.); the Agency for Healthcare Research and Quality Training grant T32 HS013852 (D.R.B.); and the Mississippi Agricultural and Forestry Experiment Station (D.R.B.).

The authors declare no conflict of interest.

FOOD insecurity and nutritional risk matter for older adults' health and well-being because as people age, they become unique ecological actors1–3 whose nutritional status may be differentially affected by various elements of social structure.4 Understanding what social factors put older adults at risk for poor nutrition is especially important so that those indicators can be targeted in tailored interventions.5 , 6 Furthermore, these factors may affect members of society differently, based on their race and gender status. Previous work by Locher et al4 found that specific social and demographic factors mattered differently for nutritional risk by race and gender among a group of adults 65+ years of age living in the Deep South, namely, 5 rural and urban counties in Alabama, and using the DETERMINE Checklist.7 Specifically, that work found that African American women had the highest levels of nutritional risk and that they were particularly susceptible to limited transportation and low community mobility as risk factors for nutritional risk. In general, for older adults, nutritional risk can be affected by social isolation, social support, and social capital, which are the factors of interest for this research.

Social isolation matters for older adults and particularly for their nutritional well-being because those who need help or assistance with meals, including with shopping and food preparation, may not be able to access such help. Locher et al showed that among African American and white women and men, lower life-space mobility (the ability to move about one's home and community) was associated with increased nutritional risk. For African American women and white men, not having adequate transportation, also a possible indicator of social isolation, was associated with increased nutritional risk. In addition, in a recent study of older adults in Lebanon, social isolation, as measured by the Lubben Social Network Scale, was predictive of malnutrition,8 indicating that these phenomena are not exclusive to the Deep South or even to the United States.

Social support consists of assistance on which people draw, the frequency of contact with other individuals, or the perceived value of that support.9 Social support is particularly critical for older adults who are experiencing illness and who have nutritional needs.10 In the Locher et al study in 2005, social support in the form of marital status mattered for African American men. For white women, the perception of a low level of social support was associated with increased nutritional risk.

Social capital includes the networks, norms, and trust that facilitate coordination and cooperation for the shared benefit of individuals and institutions.11 , 12 Furthermore, the concept highlights that individuals may both receive benefits and experience liabilities of such connections due to the expectations others have of them. Individuals with longer periods of residence at their address may have greater connections in both quantity and intensity with their neighbors; also, persons who regularly attend religious services are likely to have more frequent contact with a constant group of people and have investments in those people from whom they can draw support. Those who perceive risks in neighborhood safety may be less likely to have ties in their communities, limiting their social capital, and those who are veterans have access to resources as part of their networks such as the Veterans Health Administration.4 All of these may impact the access that individuals have to nutrition-related resources. In the previous study by Locher et al, social capital mattered greatly for African American men; specifically, not attending religious services regularly, restricting activities for fear of being attacked, and perceived discrimination were all associated with increased nutritional risk. In addition, for African American women and white women and white men, lower income was associated with increased nutritional risk.4

The work reported earlier by Locher et al was conducted using the DETERMINE Checklist, which was developed by the American Dietetic Association (now the Academy of Nutrition and Dietetics), the American Academy of Family Physicians, and the National Council on Aging as an educational tool to increase awareness of the importance of nutrition in older populations.7 It includes 10 items that reflect nutritional risk and has also been used as a screening tool. In contrast, the Mini Nutritional Assessment (MNA) was specifically developed as a nutrition screening and assessment tool that identifies geriatric patients 65 years and older who are malnourished or at risk of malnutrition, downstream outcomes of food insecurity.13–16 Several measures of the MNA may also be reasonable markers of food insecurity including changes in food intake, involuntary weight loss, number of meals consumed each day, self-perception of malnutrition, and protein consumption. This study uses a similar approach as the 2005 article by Locher et al,4 but now the focus is on an older population and measures nutritional risk using the MNA, which is better suited for nutritional risk assessment. Specifically, this article examines how social factors affect older adults' nutritional risk in the Deep South, namely, 5 rural and urban counties in Alabama, and what disparities there may be between these outcomes by race and gender subpopulations.

Back to Top | Article Outline



Data are from the baseline assessment of the University of Alabama at Birmingham (UAB) Study of Aging II, a cohort study of 419 community-dwelling persons 75+ years of age from Alabama.17 Participants were recruited from former cohort members in the UAB Study of Aging I (1999-2008) or the State of Alabama Long Term Care Needs Assessment (2002)18 , 19; they were required to be living in one of 17 counties from which participants had been previously recruited, able to communicate on the telephone to set an appointment for an in-home interview, and able to answer questions by themselves. Baseline assessments were conducted in patients' homes by trained interviewers. Each visit lasted approximately 2 hours, including obtaining informed consent. A follow-up telephone interview was scheduled a minimum of 2 weeks following the in-home assessment. Enrollment lasted from June 2010 through August 2011. The study protocol was approved by UAB institutional review board.

Back to Top | Article Outline

Data collection

Nutritional risk

Nutritional risk was assessed with the MNA. The MNA is a well-validated assessment designed to measure either risk of or presence of overt malnutrition.13 , 20 The tool consists of a screener with 6 items and a full questionnaire with 18 items, each of which is rated with 0 to 3 points; the maximum risk score one could attain would be 30 points on the full assessment. Persons with more than 24 points are “well-nourished,” persons with 17 to 23.5 points are at risk of malnourishment, and persons with fewer than 17 points are malnourished. The MNA incorporates multiple domains, including functional status, lifestyle, diet, and self-perception of health; the tool has been shown to be predictive of adverse clinical events and mortality among older hospitalized patients as well as those in the community.15 , 16 , 21–23 Questions on the MNA used to assess nutritional risk are included in the Appendix. The full assessment was used for the final analysis.

Back to Top | Article Outline

Social isolation

Social isolation was assessed using measures of rural versus urban living status, transportation difficulty, and life-space mobility. Rural versus urban living status was assigned on the basis of whether individuals lived in urban or rural counties. Those in counties with fewer than 21 000 persons were rural; those in counties with 21 001 or more persons were classified as urban. Transportation difficulty was assessed with the following question: “Over the past month, have you had any difficulty getting transportation to where you want to go?” Responses included “no difficulty,” “a little difficulty,” and “a lot of difficulty.” Those indicating “no difficulty” were assigned a value of 0; others were assigned a value of 1. Finally, the life-space mobility assessment was used to determine mobility; the assessment indicates the distance a person moved in the previous 4 weeks and whether assistance was required from an individual or device.24 Possible zones of movement include through one's bedroom, home, yard, neighborhood, town, and out of town. The range of scores is from 0 to 120, with higher scores reflecting greater mobility.

Back to Top | Article Outline

Social support

Social support was measured by accounting for marital status and the Lubben Social Network Scale. Marital status was assessed with the question, “Are you now married, or are you widowed, separated, divorced, or have you never been married?” Persons currently married were considered married; others were considered not married. The Lubben Social Network Scale includes 10 items that assess individuals' living arrangements, size of social networks, reciprocal social supports, as well as frequency and emotional closeness of social contact with family and friends.25 , 26 Each item is scored on a 5-point Likert scale, with higher scores indicating less social support.

Back to Top | Article Outline

Social capital

Social capital was assessed using the number of years an individual had lived at his or her address, religious participation, neighborhood safety, and veteran status. Number of years an individual lived at one's address reflects one's embeddedness in his or her community; this variable was categorized as 0 to 5, 6 to 10, and 10+ years. Religious participation was assessed with the following question: “How often do you attend church or other religious meetings?” Responses included “more than once a week,” “once a week,” and “a few times a month,” which were coded as regular religious participation, and “a few times a year,” “once a year or less,” and “never,” which were coded as not engaging in regular religious participation. Neighborhood safety was assessed using the following question: “How safe from crime do you consider your neighborhood to be?” with responses including “extremely safe,” “quite safe,” “slightly safe,” and “not at all safe.” Persons indicating “slightly safe” and “not safe at all” were classified as “not feeling safe in their neighborhood.” Finally, affiliation with the Veterans Affairs (VA) system was coded for persons who indicated having a VA doctor versus having never seen a VA doctor.

Back to Top | Article Outline


Covariates included age, education, and income. Age was included as continuous variable, and education was included as a categorical variable: 6th or less grade; 7th to 11th grades; high school graduate; and more than high school education. Income was also assessed categorically, with $12 000 or less, $12 000 to $49 999, and $50 000 or more.

Back to Top | Article Outline

Statistical analysis

Descriptive analyses were conducted using SAS 9.4 on the full sample and separately for African American men, African American women, white men, and white women for independent variables and the MNA. Ordinary least squares regression analyses were used to examine correlates of the MNA full scores.

Back to Top | Article Outline


This study included 60 African American men, 84 African American women, 114 white men, and 156 white women. Participants' mean age was 81.6 years. Table 1 includes the descriptive statistics of the sample, including all individual items from the MNA.



Table 1 also includes the predictor variables that comprise the social isolation, social support, and social capital constructs, as well as the control variables. Among components of the MNA, there were statistically significant differences between race and gender groups in involuntary weight loss, mobility status, psychological stress, neuropsychological problems, body mass index (BMI), number of full meals per day consumed, and subjective health status in comparison with others.

Table 2 shows that mean scores for the full MNA differed by groups, with African American men and women having the same level of risk, the highest among the 4 subgroups, followed by white women and then white men.



Differences were also found in the percentage of those at nutritional risk or malnourished according to the MNA screener, with the highest proportion of African American men at risk, followed by African American women, white men, and white women. For the full MNA, the highest proportion was African American men and then African American women, white men, and white women at risk with no weight loss, at risk with weight loss, or malnourished.

Finally, Table 3 shows multivariable analyses assessing the joint impact of all social factors on nutritional risk for each race-gender group. Among the social isolation variables, transportation difficulty mattered for white women and life-space mobility mattered across all demographic subgroups. In this sample, social support did not appear to have any significant effect on nutritional risk for any group, although for social capital, frequent religious attendance was associated with lower nutritional risk for African American men and veteran status mattered for white men such that those with veteran status had a lower nutritional risk score. Furthermore, for white men, lower education was associated with more nutritional risk.



Back to Top | Article Outline


Review of findings

This study provides important updates to findings from the Locher et al study in 2005; although older African American women have high levels of nutritional risk, they are not at highest risk, and, in fact, older African American men have equal or greater risk, depending on the assessment and cut points used. Scoring of the full MNA shows that African American men and women had equal risk. Using the full MNA cut points and categories, 61.5% of African American men and 55.8% of African American women were at nutritional risk. Based on the 6-item MNA screener, there were even greater proportions of these subpopulations at nutritional risk, with 81.6% of African American men and 66.6% of African American women being at risk.

In the bivariate analyses, African American men were more likely to have experienced involuntary weight loss (39.9%) and have neuropsychological problems including mild or severe dementia (51.6%). African American women were more likely to be homebound (17.8%), suffer from psychological distress (82.1%), have a BMI of 23 or greater (91.6%), eat just 1 or 2 meals per day (54.6%), or rate their health status as not as good as others (10.7%). In addition, African American men were the most likely to have not finished high school (65%) whereas African American women were the most likely to report having transportation difficulty (13.1%), being not married (82.1%), most frequently attending religious services (88.1%), having the lowest level of life-space mobility (mean = 51.8), and making $12 000 per year or less (36.9%). White men were the most likely to be veterans (33.3%).

With regard to the full multivariable regression model, African American men were sensitive to transportation difficulty and life-space mobility as predictors of nutritional risk. African American women were only sensitive to life-space mobility. White men were sensitive to life-space mobility, veteran status, and education, and white women were sensitive to transportation difficulty and life-space mobility. In fact, all race and gender groups were sensitive to life-space mobility, indicating that the distance a person moved in the previous 4 weeks mattered for his or her likelihood of being at nutritional risk. Life-space mobility is a proxy for social participation,24 , 27 so this finding suggests that the less social participation an individual has, the more likely he or she is to have difficulty with nutrition-related needs.

Back to Top | Article Outline

Interpretation of findings

The population with the greatest level of nutritional risk differed from that of the Locher et al study in 2005; however, the discrepancies may be due to differences inherent in the DETERMINE Checklist and the MNA. They may also be different because the average age of the population is 5 years in the current study. This may indicate that the MNA is more sensitive to nutritional risk or that risk increases with age differentially for these subpopulations. It is curious that while African American men were only statistically more sensitive to 2 markers of nutritional risk, African American women were more sensitive to 5 markers of nutritional risk; African American men were still more likely to be at nutritional risk. That so few variables emerged as significant factors associated with nutritional risk could indicates that, in fact, it is the collective burden of social isolation, limited social support, and social capital that matters more than any one particular variable for these older adults.

Back to Top | Article Outline

Relevance for food insecurity and obesity

The World Health Organization defines food security as existing “when all people at all times have access to sufficient, safe, nutritious food to maintain a healthy and active life.”28 Food insecurity, in contrast, exists “whenever the availability of nutritionally adequate and safe foods or the ability to acquire acceptable foods in socially acceptable ways is limited or uncertain.”29 These conditions stem from poverty or other factors, not voluntary nutritional modifications such as fasting. While in the general population, poverty is the most likely factor to cause food insecurity in the general population, among older adults, it may be caused by physical impairments or disabilities that limit their ability to procure, prepare, and/or consume the food.30 , 31 As noted in the introduction, several indicators that are part of the MNA are adequate proxies or downstream outcomes of food insecurity. These include food intake, involuntary weight loss, number of meals consumed each day, self-perception of malnutrition, and protein consumption; on all of them, there were higher proportions of African Americans performing poorly and on all except the change in food intake, there were statistically significant differences. The relationship between race, gender, and obesity status was not as straightforward, however; obesity rates were highest for African American women, followed by white men, African American men, and white women. For all of these, it was clear that obesity status was characteristic of the majority of the group, raising the question of how relevant the food insecurity-obesity paradox32 may be generally for older adults. Given that the proxy markers of food insecurity were so high among African American women and that they also had the highest rates of obesity, the authors suggest that of the groups observed in this study, they may be the most likely to be not only affected by this paradox of having food insecurity but also being obese at the same time.

Back to Top | Article Outline


This study suggests that addressing nutritional risk in community-dwelling older adults should be a priority of social services programs, particularly in the Deep South. Nearly 50% of those in the study were identified as being at risk, and 2 subpopulations had even higher percentages of persons at risk, with some even being identified as malnourished. It is especially important to note that overweight and obesity status should not be used as an indicator that malnutrition is not present; in fact, in this study, more than 80% of the participants had a BMI above the highest cut point. Furthermore, targeted screening and interventions at those with the greatest risk may be warranted, and based on our study results, these efforts should be directed to older African American men and women. There was not substantial nuance among the factors associated with nutritional risk, and life-space mobility limitation was associated across the subpopulations, indicating that tailored interventions may be difficult to achieve. Services that provide transportation, direct nutritional assistance, social participation and connectedness, and other efforts to curb nutritional risk should be continued and expanded. Such services include those delivered under the Older Americans Act such as Meals-on-Wheels, congregate meals, in-home services, transportation, legal services, elder abuse prevention, and caregiver support. Either independently or in conjunction with each other, these programs help older adults remain independent, living in their homes and communities as long as possible, saving families and taxpayers millions of dollars in expensive institutional care. Such programs are administered through states and in partnership with local Area Agencies on Aging. Findings from this study suggest that there is no single intervention that can be the focus of nutrition risk prevention and that multicomponent approaches must be embraced and continued. In addition, this study suggests that social ties and resources inherent in people's communities, which are independent of government services, are important for their nutritional well-being.

Back to Top | Article Outline

Limitations and strengths

While the generalizability of these findings should be limited to those in the southeastern United States, the approach used herein can be replicated in other studies. Furthermore, these findings, in concert with other important work going on in nutrition and aging, can be used to emphasize the importance of interventions that seek to minimize nutritional risk among both older adults generally and by different subpopulations. In addition, we recommend that other studies with more generalizable samples may seek to replicate these findings.

Back to Top | Article Outline


1. Lawton MP, Nahemow L. Ecology and the aging process. In: Eisendorfer C, Lawton MP, eds. The Psychology of Adult Development and Aging. Washington, DC: American Psychological Association; 1973:619.
2. Lawton MP. Social ecology and the health of older people. Am J Public Health. 1974;64(3):257–260.
3. Lawton MP. Environment and Aging. Belmont, CA: Brooks/Cole; 1986.
4. Locher JL, Ritchie CS, Roth DL, Baker PS, Bodner EV, Allman RM. Social isolation, support, and capital and nutritional risk in an older sample: ethnic and gender differences. Soc Sci Med. 2005;60(4):747–761.
5. Oria M, Cappelucci K, Rodgers A, Rapporteurs AV. Meeting the Dietary Needs of Older Adults: Workshop in Brief. Washington, DC: The National Academies Press; 2016.
6. Campbell AD, Godfryd A, Buys DR, Locher JL. Does participation in home-delivered meals programs improve outcomes for older adults? Results of a systematic review. J Nutr Gerontol Geriatr. 2015;34(2):124–167.
7. Posner BM, Jette AM, Smith KW, Miller DR. Nutrition and health risks in the elderly: the nutrition screening initiative. Am J Public Health. 1993;83(7):972–978.
8. Boulos C, Salameh P, Barberger-Gateau P. Social isolation and risk for malnutrition among older people. Geriatr Gerontol Int. 2017;17(2):286–294.
9. Hooyman N, Kiyak HA. Social Gerontology: A Multidisciplinary Perspective [With Access Code]. Needham Heights, MA: Allyn and Bacon; 2007.
10. Jung SE, Hermann JR, Bishop A. Impact of nutritional risk on self-care capacity: social support as a source of protection for community-dwelling older adults living in a rural area. J Frailty Aging. 2013;2(3):145–149.
11. Buys DR, Mohan RP. Social capital and structural challenge in health services delivery: a comparative case study. Int J Contemp Sociol. 2012;49(2):217–237.
12. Putnam RD. Bowling Alone: The Collapse and Revival of American Community. New York, NY: Simon & Schuster; 2001.
13. Guigoz Y, Vellas B, Garry PJ. Mini Nutritional Assessment: a practical assessment tool for grading the nutritional state of elderly patients. Facts Res Gerontol. 1994;(suppl 2):15–60.
14. Vellas B, Guigoz Y, Garry PJ, et al The Mini Nutritional Assessment (MNA) and its use in grading the nutritional state of elderly patients. Nutrition. 1999;15(2):116–122.
15. Vellas B, Villars H, Abellan G, et al Overview of the MNA—its history and challenges. J Nutr Health Aging. 2006;10(6):456–463; discussion 463–465.
16. Kaiser MJ, Bauer JM, Rämsch C, et al Validation of the Mini Nutritional Assessment Short-Form (MNA-SF): a practical tool for identification of nutritional status. J Nutr Health Aging. 2009;13(9):782–788.
17. Brown CJ, Kennedy RE, Lo AX, Williams CP, Sawyer P. Impact of emergency department visits and hospitalization on mobility among community-dwelling older adults. Am J Med. 2016;129(10):1124.e9–1124.e15.
18. Allman RM, Sawyer P. Charting the Course: State of Alabama Long Term Care Needs Assessment. Birmingham, AL: UAB Center for Aging; 2004.
19. Allman RM, Sawyer P, Roseman JM. The UAB Study of Aging: background and insights into life-space mobility among older Americans in rural and urban settings. Aging Health. 2006;2(3):417–429.
20. Rubenstein LZ, Harker JO, Salva A, Guigoz Y, Vellas B. Screening for undernutrition in geriatric practice: developing the Short-Form Mini-Nutritional Assessment (MNA-SF). J Gerontol A Biol Sci Med Sci. 2001;56(6):M366–M372.
21. Kaiser MJ, Bauer JM, Rämsch C, et al Frequency of malnutrition in older adults: A multinational perspective using the Mini Nutritional Assessment. J Am Geriatr Soc. 2010;58(9):1734–1738.
22. Kaiser MJ, Bauer JM, Uter W, et al Prospective validation of the modified Mini Nutritional Assessment Short-Forms in the community, nursing home, and rehabilitation setting. J Am Geriatr Soc. 2011;59(11):2124–2128.
23. Bauer JM, Kaiser MJ, Anthony P, Guigoz Y, Sieber CC. The Mini Nutritional Assessment®—its history, today's practice, and future perspectives. Nutr Clin Pract. 2008;23(4):388–396.
24. Baker PS, Bodner EV, Allman RM. Measuring life-space mobility in community-dwelling older adults. J Am Geriatr Soc. 2003;51(11):1610–1614.
25. Lubben JE. Assessing social networks among elderly populations. Family Community Health. 1988;11(3):42–52.
26. Lubben JE, Gironda M, Lee A. Refinements to the Lubben Social Network Scale: the LSNS-R. Behav Meas Lett. 2002;7:2–11.
27. Peel C, Baker PS, Roth DL, Brown CJ, Bodner EV, Allman RM. Assessing mobility in older adults: the UAB Study of Aging Life-Space Assessment. Phys Ther. 2005;85(10):1008–1019.
28. World Health Organization. Food Security. Geneva, Switzerland: Trade, Foreign Policy, Diplomacy and Health, World Health Organization; 2012. http:// Accessed October 13, 2012.
29. Economic Research Service, US Department of Agriculture. Food insecurity in the US: measurement. http:// Published 2012. Accessed July 18, 2015.
30. Lee JS, Frongillo EA Jr. Factors associated with food insecurity among U.S. elderly persons: importance of functional impairments. J Gerontol B Psychol Sci Soc Sci. 2001;56(2):S94–S99.
31. Lee JS, Frongillo EA. Nutritional and health consequences are associated with food insecurity among U.S. elderly persons. J Nutr. 2001;131(5):1503–1509.
32. Dinour LM, Bergen D, Yeh MC. The food insecurity-obesity paradox: a review of the literature and the role food stamps may play. J Am Diet Assoc. 2007;107(11):1952–1961.
Back to Top | Article Outline

APPENDIX: Mini Nutrition Assessment Questions





*In this study, all participants were community-dwelling, so they did not have the opportunity to score any differently than the prescribed score of 1. If they had been a mixed population of hospitalized, institutionalized, and community-dwelling, there would have been more room for variation on this score.


food security; nutritional risk; obesity; older adults; social factors

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved