Household Food Insecurity Is Associated with Self-Rated Health Status among Single Mothers in Alabama’s Black Belt Counties

Abstract

This study was conducted in Alabama’s Black Belt Counties to examine the association between household food insecurity and self-reported health status. Data were collected from 400 households to measure household food insecurity and self-reported general health status using the U.S. Food Security Module. In bivariate analyses, household food insecurity was significantly associated with health status as some mothers from food insecure households were significantly more likely to rate their health as fair/poor. In regression analysis model controlling for potentially confounding variables, household food insecurity was still associated with poor self-reported health status. Food intake of some household members was reduced, and their eating patterns disrupted at times because the household lacked money for obtaining food. Policy changes to increase economic resources and access to federal food programs are needed to reduce household food insecurity in this region. Gendered experiences in the context of consequences of poverty should not be ignored.

Introduction

Household food insecurity, defined as the “limited or uncertain availability of nutritionally adequate and safe foods, or limited or uncertain ability to acquire food in socially acceptable ways to feed the whole household” (Coleman-Jensen et al., 2019; Zekeri, 2019) continues to affect millions of American families in the rural south. Household food insecurity is a social problem that disproportionately affects households in rural areas.

In previous research of single mothers in the Black Belt counties, those who experienced food insecurity had significantly lower levels of income, less education, and were more likely to be African Americans and report higher depressive symptoms than women from food secure households (Zekeri 2019; 2010; 2016; Zekeri et al., 2016). The aim of the present study is to determine if household food insecurity is a specific correlate of self-reported health status. The central hypothesis guiding this analysis is that food insecurity among household headed by poor single mothers is associated with their health status. This is suggested because household food insecurity is embedded within the context of poverty which is likely to produce anxiety and fear that may take a toll on physical health.

Several studies have examined the impact of food insufficiency as measured by a scale derived from the National Health and Nutrition Examination Survey III on health status among adults in urban areas (Dixon, Winkleby, Household Food Insecurity Is Associated with Self-Rated Health Status among Single Mothers in Alabama’s Black Belt Counties

Andrew A. Zekeri, B.Sc, M.Sc., Ph.D.& Radimaer 2001; Heflin, Siefert & Williams, 2005; Nelson et al., 2001; Olson, 1999; Roe, 1990; Siefert et al., 2001, 2004; Vozoris & Tarasuk, 2003). In these urban studies, self-rated health status is associated with food insufficiency. However, reports of the relationship between household food insecurity and health status in rural areas are limited. Despite its potential impact on health and well-being, surprisingly little research has been done on the relationship between household food insecurity and health status among poor female-headed families in the rural south. Therefore, as an extension of my previous research (Zekeri, 2019; 2011; 2010), the goal of this study is to examine the association between household food insecurity and self-reported health in this poverty-stricken region of Alabama.

Method

The research design combined quantitative and qualitative methods. This research was conducted as part of a larger project, Food Insecurity in Poor, Female-Headed Families in Five of Alabama Black Belt Counties. My prior research from the same data set (Zekeri, 2010; 2016; 2019) focused on food insecurity and depression. In the present investigation, my focus is on household food insecurity and self-reported health status. The survey was designed to measure food insecurity status at the household level. The research is based on a survey, in-depth interviews, and a workshop, and was conducted in collaboration with key informants that helped the researcher gain trust and entry to this community. A sample of 400 households located in five Alabama Black Belt Counties (Bullock, Dallas, Macon, Lowndes, and Wilcox) that were receiving welfare and/or food stamp benefits agreed to participate in the study. The sample for this study, 400 single mothers was drawn from a list of over 1,000 families in a five-county area of Alabama Black Belt that participated in previous studies carried out by the author (Zekeri, 2010; 2016; 2019).

These counties have many African Americans and individuals living below the poverty level. Also, these are counties where I had previously developed ties with many of the community informants and residents. Households were selected using random stratified sampling methodology. The randomly selected single mothers were located at previous addresses, as well as through community informants and local churches. Once the selected mothers were located, I made personal home visits requesting participation in the study. Most of these single mothers were low-income individuals that had previously helped me understand their world - how they perceived it, coped with it, and interacted with it—and their opinions about electronic delivery of food stamps with the Benefit Security Card (Zekeri, 2010; 2016).

Alabama’s Black Belt

Alabama’s Black Belt, the site for this study, is an ideal setting for examining the prevalence and adverse health consequences of food insecurity. Among the poorest places in the United States, this region is characterized by its high concentration of African American residents. Regionally, the Alabama Black Belt’s poverty rates as of 2019 compare to that of the Appalachian mountain region, where the poor are predominantly White, the Rio Grande Valley/Texas Gulf Coast, where the poor are largely Latino, and the reservations of the Southwest, where the poor are largely Native Americans. The Black Belt region suffers from poor employment opportunities, chronic unemployment, limited educational attainment, poor health, high concentration of single parents, and heavy dependence on public assistance programs (; 2019; Zekeri et al., 2016).

The Study Variables

Food insecurity was measured using a structured questionnaire based on CFSM and the other questionnaires constructed and used in previous studies, all investigations were approved by the Human Subject Participants Review Committee at Tuskegee University. The CFSM scale is based on respondents’ answers to a series of 18 questions regarding behaviors and experiences known to characterize households that are having difficulty meeting food needs. Face-to-face, in-home, structured interviews were conducted. Consent to participate in the study was obtained from all participants. The interviews lasted approximately 90 minutes. For the present analysis, food security status is a dichotomous variable (food insecure and food secure).

Dependent Variable

Self-reported health is the dependent variable. Self-reported health is the respondent’s subjective assessment of her general health. Single mothers were asked to rate their overall health at the time of the of the survey with a standard five-category item for self-rated health, with values ranging from excellent (1) to poor (5). Self-rated health has been shown to be a reliable, valid measure of health, and it is predictive of subsequent functional decline (Idler, Russell& Davis, 2000). It is a valid and reliable measure of general physical well-being. It predicts mortality net of chronic and acute disease, physician assessment made by clinical exam, physical disability, and health behaviors (Idler & Benjamin 1997).

Key Independent Variable

The independent variable of interest is a binary variable that indicates whether a household is food insecure or not. I collected questionnaire data in face-to-face, in-home, structured interviews with single mothers who were heads of the household. The interviews lasted about 90 minutes. Household food insecurity was measured using a structured questionnaire (based on the USDA’s Food Security Core Module). The study was approved by Tuskegee University Committee on Human subjects.

Control Variables

To empirically distinguish the influence of household food insecurity and other variables, respondent’s age, race (black or white), educational attainment (years of completed schooling), unemployment, and annual earnings known to be associated with health status in previous research (Heflin et al., 2005; Siefert et al., 2000, 2001, 2004; Stuff et al., 2004; Vozoris & Tarasuk, 2003) are included in the model. Household income (in dollars), and age of single mothers (in years) were included as continuous variables. Race is a dummy variable coded 1 for African Americans. Education level was a dummy variable coded 1 for those with a four-year college degree. Unemployment is also a dummy variable coded 1 if the single mother is unemployed. Taken together, these sociodemographic and work characteristics provide a basic outline of the respondent’s social position.

Data Analysis

The analysis employs multiple regression methods using SPSS 14.0. First, health is regressed on food insecurity to determine any statistically significant associations. Then, an expanded form of regression analysis examines effect estimates (regression coefficients) of food insecurity and the control variables on health status.

Results

Using the USDA’s Food Security Scale, 36% of the households were classified as food insecure (Table 1). The prevalence of food insecurity in this region contrasts sharply with the national figure, where roughly 11.1% suffer from food insecurity in 2018. Most of the households are African Americans (70%) and 32% had no education beyond high school. Regarding household income, 52.2% earned less than $15,000 and 36.5% had no employment.

Table 1.

Percentages for Dependent and Independent Variables^a

Dependent	Percent
Health Status
Poor	8.3
Fair	29.0
Very Good	29.6
Excellent	7.9
Independent Variables
Prevalence of Food Insecurity
Food Secure	64.0
Food Insecure	36.0
Race/Ethnicity
African American	70.0
White	28.0
Hispanic	2.0
Educational Attainment
Did not complete high school	31.5
Completed high school or equivalent	32.7
Some college of post high school training	23.5
Completed a college degree	10.2
Income
Under $15,000	52.2
$15,000 to $34,0000	47.5
Employed (including part-time)	59.3
Unemployed	36.5

^aSome percentage scores do not sum to 100% because missing data are not reported.

Bivariate Analysis

Bivariate analysis showed a statistically significant association between household food insecurity and health status (see Table 2). This finding is consistent with some previous research (for example, see Heflin et al., 2005; Olson, 1999, 2005; Siefert et al., 2001, 2004; Stuff et al., 2004; Vozoris & Tarasuk 2003) that found that the association between household food insecurity and health status is statistically significant. In the bivariate analysis, food insecurity alone explained about 3.6% of the variation in health status.

Table 2.

Standardized coefficients of regression analyses of health status among households

	Self-Reported Health Status
	Model 1	Model 2
Household food insecurity	−.201 *	−.147 *
Age		.035
Income		.019
Full-time employment		.237 *
Racea (Black=1)		−.035
Education attainment
Four-year college degree		.161 *
Adjusted R 2	.036	.130

^*p <.05;

^aWhite persons are the reference group

Multivariate Analysis

Table 2 presents the results from regression analysis predicting self-rated health status. Furthermore, and confirming some previous studies (e.g., Siefert et al., 2001, 2004; Stuff et al., 2004; Vozoris & Tarasuk, 2003), model 2 in Table 2 indicated that among the predictor variables, food insecurity is still a statistically significant predictor of health status. Model 2 demonstrated that educational attainment and employment have large and statistically significant relationships to general health. Having a four-year college degree and being gainfully employed are related to better self-reported health as shown by the relatively large parameter estimates. Thus, consistent with previous literature (e.g., Dowd & Zajacova, 2007), higher levels of education and employment predict more health-optimistic rating. Income, in contrast, is unrelated to ratings net of other variables. This finding is contrary to the results of other researchers that found low income and neighborhood poverty are linked to poor self-reported in rural women (Kobetz et Al., 2003; Zekeri, 2019) Taken together, the results provided support for the hypothesis: Household food insecurity is associated with health status.

Discussion

In this study, I have assessed whether household food insecurity is associated with self-rated health status. Similar to some previous studies, household food insecurity is a significant predictor of self-rated health. The association remained statistically significant after controlling for potentially cofounding variables, a finding that supports past evidence (e.g., see Heflin et al., 2005; Olson et al., 2004; Siefert et al., 2000, 2001, 2004; Stuff et al., 2004; Tarasuk & Beaton, 1999; Vozoris& Tarasuk, 2003). In a random sample of 724 single women, who were welfare recipients in Northern Michigan, Siefert et al. (2001) found that food insufficiency was significantly associated with poor or fair self-reported health. The current study advances our knowledge of this relationship by focusing on the link between household food insecurity and self-rated health in rural Alabama where access to health treatment can be even more difficult to obtain than in urban areas. Moreover, the majority of the previous studies used a single item measure of food insufficiency, while this study used the U.S. Food Security Module scale to measure food insecurity allowing researchers to be more confident in the reliability and validity of the finding (Alaimo et al., 1998, 2001; Heflin et al., 2005; Siefert et al., 2001, 2004). Race was not a significant predictor of health in this sample from the rural south. Further research is needed to ascertain whether this finding persists in other similar regions.

In sum, the association of household food insecurity with self-reported health, regardless of causal direction, shows the precarious situations household in rural areas face. Beyond food problems, households struggle with health problems. The findings highlight the need to prevent household food insecurity and ensure that all rural households are adequately fed to improve health and social well-being. It is also crucial to note that measures of education, earnings, race, and age commonly regarded as key predictors of health are included in the model. The production of health is clearly very complex, but I argue that food insecurity can have a meaningful impact on well-being, along with other factors considered here and a multitude of unmeasured influences. In poor rural areas such as Alabama’s Black Belt, a number of obstacles to health care and health care access could also contribute to poor health status. Alabama’s Black Belt faces a unique combination of factors that create disparities in health care not found in urban communities such as economic factors, cultural differences, and educational attainment.

The future direction of food insecurity research must go beyond just monitoring food insecurity to linking it with medical related outcomes such as health status. Education, employment, and food insecurity are all related health outcomes. Alabama’s Black Belt faces unique challenges in each of these areas.

The study has some limitations. First, the cross-sectional nature of the data makes it impossible to draw causal inferences from the relationships in the model. Further tests of the model should utilize longitudinal data, including a more comprehensive measure of household food insecurity to ascertain the true nature of the associations reported here. Reliance on self-reported health rather than formal diagnosis based upon structured medical interview may be seen as another limitation. Clearly, the ideal study would include screening with self-reported measure of health followed up with a structured clinical interview. The significance of self-reported health, however, should not be underestimated. Research examining food insecurity in relation to objective measure of health is required to confirm the findings of this study.