Abstract
Food insecurity is a significant public health issue that affects the physical and mental health of people of all ages. Higher levels of self-efficacy may reduce levels of food insecurity. In addition, acculturation is potentially an important factor for food insecurity among immigrant populations. The purpose of this study is to examine food insecurity associated with self-efficacy and acculturation among low-income primary care patients in the United States. A self-administered survey was administered in May and June 2015 to uninsured primary care patients (N = 551) utilizing a free clinic that provides free primary care services to low-income uninsured individuals and families in the United States. On average, participants reported low food security. Higher levels of self-efficacy were associated with lower levels of food insecurity. Higher levels of heritage language proficiency were related to lower levels of food insecurity. US-born English speakers, women, and unmarried individuals potentially have higher risks of food insecurity and may need interventions to meet their specific needs. Self-efficacy should be included in nutrition education programs to reduce the levels of food insecurity. Future studies should further examine why these groups have a high risk to better understand needs for interventions.
Introduction
F
Although poverty is positively associated with food insecurity, higher levels of self-efficacy, which refers to confidence in one's ability to change behaviors, 10 may reduce levels of food insecurity. Previous studies examined the association between food insecurity and self-efficacy among patients with diabetes. Food insecurity is related to low diabetes-related self-efficacy among adults with diabetes. 11 Higher levels of diabetes-related self-efficacy reduces barriers to health care access and utilization. 12 Other than patients with diabetes, food bank recipients who report high self-efficacy are less likely to experience food insecurity compared to those with low self-efficacy, 13 yet little is known about the association between food insecurity and self-efficacy among general low-income primary care patients. Because self-efficacy is related to intention to implement healthy eating behaviors, 14 it is important to increase knowledge about self-efficacy and food insecurity among low-income primary care patients who may experience food insecurity in order to develop effective health promotion programs.
In addition to self-efficacy, acculturation, which refers to the process of psychological and behavioral change as a result of contact with a new or different culture, 15 is potentially an important factor for food insecurity among immigrant populations. However, previous studies on food insecurity and acculturation show mixed results. Some studies suggest lower levels of acculturation are associated with food insecurity, 16,17 whereas other studies indicate higher levels of acculturation are related to food insecurity or poor health. 18,19 Acculturation has multidimensional aspects. 20 It would be helpful to understand what specific aspects of acculturation improve or worsen levels of food insecurity.
The purpose of this study is to examine food insecurity associated with self-efficacy and acculturation among low-income primary care patients. This study increases knowledge about how self-efficacy and acculturation affect food insecurity in order to develop effective intervention strategies to reduce food insecurity in a primary care setting. This study focuses particularly on low-income and uninsured patients who are utilizing a free clinic that provides primary care for the underserved, who also may have a high risk of food insecurity.
Methods
Overview
The current community-based research project was conducted at a free clinic in the Intermountain West. The clinic staff collaborated with the research team to develop the survey instrument, study protocol, participant recruitment strategies, and interpretation of study results. The clinic provides free health care services, mostly routine health maintenance and preventive care, for uninsured individuals who live below the 150th percentile federal poverty level and do not have access to employer-provided or government-funded health insurance. The clinic is staffed by 6 full-time paid personnel and more than 300 active volunteers, including approximately 60 volunteer interpreters. The clinic, which has been in operation since 2005, has no affiliation with religious organizations and is funded by non-governmental grants and donations. The clinic is open 5 days a week. The number of patient visits was 18,967 in 2013. The clinic does not ask patients to provide documentation of legal residency or citizenship and serves undocumented immigrants as well as US citizens and documented immigrants.
Study participants and data collection
Participants were aged 18 years old or older, speak and read English or Spanish, and were patients of the clinic. A bilingual translator translated English materials into Spanish. Another bilingual translator conducted back-translation from Spanish to English. The third bilingual translator checked accuracy of the translation. Participants were divided into 3 groups, namely US-born English speakers, non-US-born English speakers, and Spanish speakers, because these 3 populations have different sociodemographic characteristics, physical and mental health status, and needs for health care based on previous studies of free clinic patients. 21
Prior to data collection, the University of Utah institutional review board approved this study. Data were collected for 2 months, in May and June 2015. Recruitment occurred at the free clinic by distributing flyers to patients in the waiting room. If a potential participant expressed interest in participating in the study, he or she received a consent cover letter and a self-administered paper and pencil survey. Members of the study team were available to answer any questions while participants were taking the survey. Participants received a sample sunscreen or hand sanitizer (US$1 or less value) at the completion of the survey. The research assistants who collected surveys checked item nonresponses immediately after submission and asked the participant to complete missing parts if there were any, whenever it was possible. Only completed surveys were included in the analysis (N = 551).
Measures
Food insecurity
A 6-item food insecurity questionnaire from the U.S. Household Food Security Module 22 was used to measure levels of food insecurity. The first 2 questions were: “The food that (I/we) bought just did not last, and (I/we) didn't have money to get more,” and “(I/we) couldn't afford to eat balanced meals” in the last 12 months. The coding was 1 for often true or sometimes true and 0 for never true. The third question was “Did you ever cut the size of your meals or skip meals because there wasn't enough money for food?” in the last 12 months (yes coded as 1, no coded as 0). The fourth question was only for those who picked yes for the third question: “How often did it happen?” There were 3 choices for the fourth question: almost every month and some months but not every month were coded as 1, and only 1 or 2 months was coded as 0. The fifth question was “Did you ever eat less than you felt you should because there wasn't enough money for food?” (yes coded as 1, no coded as 0). The last question was “Were you ever hungry but didn't eat because there wasn't enough money for food?” (yes coded as 1, no coded as 0). The sum of the scores from all items was used for analysis. The score range is 0 to 6. Higher scores indicate higher levels of food insecurity. The assessment of the levels of food insecurity based on the scores is as follows: 0–1 high or marginal food security; 2–4 low food security; and 5–6 very low food security.
Self-efficacy
Self-efficacy was measured by the General Self-Efficacy Scale. 23 The scale has 10 items and uses a 4-point Likert scale (1 = not at all true, 4 = exactly true). Examples of items include “I can always manage to solve difficult problems if I try hard enough” and “I can usually handle whatever comes my way.” The scoring is based on a sum of all items (score range 0–40). Although there is no specific cutoff point to identify high or low levels of self-efficacy, the mean score of US American adults is 29.48. 24 This scale has been used in many countries and languages and its validity and reliability have been tested. 24 Cronbach alpha for this study population was 0.905.
Acculturation
Acculturation was measured using 5 of the subscales of the Abbreviated Multidimensional Acculturation Scale (ie, US identity, heritage culture identity, English language, heritage language, US competence), which has been tested for validity. 20 Participants were asked to provide answers for this scale if they are not US-born or have heritage culture (their culture of origin); for example, the culture of their birth country, the culture in which they have been raised, or any culture in their family background. US identity (6 items [eg, “I feel that I am part of US American culture”]) and heritage culture identity (6 items [eg, “I have a strong sense of being a member of my culture of origin”]) were measured using a 4-point Likert scale (1 = strongly disagree, 4 = strongly agree). English proficiency (9 items [eg, “How well do you speak English at school or work?”]), heritage language proficiency (5 items [eg, “How well do you speak your native language or language of your culture of origin?”]), and US competence (6 items [eg “How well do you know American national heroes?”]) were measured using a 4-point Likert scale (1 = not at all, 4 = extremely well). The scoring is based on the average score of each subscale. Cronbach alpha for this study population was 0.944 for US identity, 0.965 for heritage culture identity, 0.984 for English proficiency, 0.977 for heritage language proficiency, and 0.931 for US competence.
Food resources and purchase behaviors
Five questions were asked to describe food sources and purchase behaviors: (1) where a participant buys food most often (multiple answers [eg, large national chain supermarket, dollar store]), (2) what transportation a participant uses when he/she goes grocery shopping (multiple answers [eg, car, walk]); (3) how often a participant goes grocery shopping per week (less than once, once, twice, 3 times or more); (4) what free food resources a participant had used in the past year (multiple answers [eg, food stamps, food banks]); and (5) whether a participant had experienced a time when he or she was unable to buy lean meats and fresh produce because they were too expensive in the past month (yes or no). These questions were developed from health education programs at the clinic.
Demographic characteristics
Demographic questions included age, sex, race/ethnicity, education level, employment status, marital status, nativity, country of origin, length of years living in the US (non-US-born participants only), and number of years as a patient of the free clinic (2+ years or less).
Data analysis
Data were analyzed using IBM SPSS Statistics, version 22 (IBM Corp., Armonk, NY). The participants were divided into 3 modal groups for comparison based on nativity and English or Spanish proficiency: US-born English speakers, non-US-born English speakers, and Spanish speakers. Descriptive statistics were used to capture the distribution of outcome and independent variables. The 3 groups of the participants were compared using Pearson's chi-square tests for categorical variables and analysis of variance (ANOVA) for continuous variables. Prior to running ANOVA tests, equality of variance was tested for unbalanced sample sizes among the 3 groups. Post hoc analyses were conducted to assess the robustness of the ANOVA analyses and to confirm their findings. Multiple regression analysis was conducted to assess predictors of food insecurity, including self-efficacy, acculturation, and sociodemographic factors. Sociodemographic factors for the regression analysis were selected based on previous studies of free clinic populations. 21 Two separate regression analyses were performed: one for all participants and the other for those who completed the acculturation questionnaire because not all participants were non-US-born or considered themselves to have heritage culture. The acculturation measures were included only for the analysis of those who answered acculturation questions. Regression coefficients (standard errors) were used to obtain 95% confidence intervals.
Results
This study involved a convenience sample of 551 participants (164 US-born English speakers, 129 non-US-born English speakers, and 258 Spanish speakers). The average age was 44.37 and nearly 70% of the participants were women. Table 1 summarizes the sociodemographic and food-related characteristics of the participants. The participants were from 40 countries. Besides the United States, Mexico had the largest number of participants (n = 192, 34.8%), followed by Tonga (n = 19, 3.4%), and Peru (n = 18, 3.3%) (data not shown). Approximately half of the participants had been a patient of the clinic for 2 years or longer. Approximately 86% of the participants completed the acculturation questionnaire.
No. (%) or Mean (standard deviation).
N.S. = not significant.
P value denotes significance from Pearson's chi-square tests between categorical variables (for cell size ≥5 only), and analysis of variance tests for continuous variables comparing US-born English speakers, non-US-born English speakers, and Spanish speakers.
Large national chain grocery stores were the most common place to buy food among the participants and convenience stores the least (Table 1). US-born and non-US-born English speakers were more likely to buy food at a convenience store compared to Spanish speakers (P < 0.05), whereas non-US-born English speakers and Spanish speakers were more likely to buy food at an ethnic food store compared to US-born English speakers (P < 0.01). More than 80% of the participants used a car for grocery shopping. US-born and non-US-born English speakers were more likely to walk to a grocery store than Spanish speakers (P < 0.05). In terms of frequency of grocery shopping, once a week was most common, with US-born English speakers tending to go grocery shopping less than once or once a week and non-US-born English speakers and Spanish speakers commonly going grocery shopping once or twice a week. Approximately one quarter of the participants had received food stamps. US-born English speakers had higher use of food stamps and food pantries, for which legal immigrants living in the United States less than 5 years and illegal immigrants are likely ineligible, 25 than non-US-born English speakers and Spanish speakers. Spanish speakers had the highest percentage for a group that used food banks compared to the other 2 populations (P < 0.05). Nearly 60% of participants experienced inability to buy fresh food because of high cost.
Table 2 summarizes the descriptive statistics of food insecurity, self-efficacy, and acculturation, and the comparison among the 3 groups. On average, participants reported low food security. Seventy-five percent of participants were classified into low or very low food security (data not shown). US-born English speakers had higher levels of food insecurity, followed by non-US-born English speakers, and Spanish speakers. The average level of self-efficacy among the participants was similar to that of the general US American adult population. 28 Spanish speakers reported higher self-efficacy scores than US-born and non-US-born English speakers. US-born English speakers had significantly high scores for US identity, English language proficiency, and US competence. Spanish speakers and non-US-born English speakers had higher scores in heritage language proficiency compared to US-born English speakers. The negative kurtosis value for food security, which is not close to 0, suggests a flat distribution for the dependent variable of regression analysis.
N = 551. Mean (standard deviation).
N.S. = not significant.
P value denotes significance from analysis of variance tests comparing US-born English speakers, non-US-born English speakers, and Spanish speakers.
Higher scores indicate higher levels of food insecurity. Score range 0–6.
Higher scores indicate higher levels of self-efficacy. Score range 0–40.
Participants who answered the acculturation questions only. N = 472 (111 US-born English speakers; 126 non-US-born English speakers; 235 Spanish speakers).
Table 3 presents predictors of food insecurity. Because of the issue of a flat distribution of the dependent variable, robust regression (bootstrapping) was performed. Among all participants, the following factors were associated with higher levels of food insecurity: lower levels of self-efficacy, US-born English speakers, and unmarried. Among those who completed the acculturation questionnaire, the following factors were related to higher levels of food insecurity: lower levels of heritage language proficiency, unmarried, older age, and female sex. ANOVA tests were conducted to compare food insecurity among married women, unmarried women, married men, and unmarried men (data not shown). Unmarried women had the highest level of food insecurity (mean = 3.54, standard deviation = 2.03) while married men reported the lowest level of food insecurity (mean = 2.53, standard deviation = 2.03).
Multivariate multiple regression. P value denotes significance from multivariate regression analysis.
N.S. = not significant.
Reference category is Spanish speakers.
Discussion
This study examined food insecurity associated with self-efficacy and acculturation among low-income primary care patients utilizing a free clinic for the uninsured. On average, participants reported low food security. Seventy-five percent of participants reported having experienced food insecurity, while 14.3% of the US general population experienced food insecurity. 9 This study suggests 3 main findings that contribute to knowledge on food insecurity among general low-income primary care patients. First, higher levels of self-efficacy are associated with lower levels of food insecurity. Second, higher levels of heritage language proficiency are related to lower levels of food insecurity. Third, US-born English speakers, women, and unmarried individuals potentially have high risks of food insecurity.
The results of this study indicate that self-efficacy interventions may reduce food insecurity. Self-efficacy is positively related to adoption of healthy behaviors; for example, the increased purchase and consumption of fruits and vegetables for healthy living. 26 Self-efficacy also affects dietary quality. 27 High self-efficacy helps with reducing fats and carbohydrates within diet. 28 Psychological interventions that focus on self-efficacy in healthy diet education could help to promote healthy eating and reduce the risks for food insecurity.
The results of this study suggest that heritage language proficiency is associated with lower levels of food insecurity, which is not consistent with some previous studies. For example, Latino youths whose parents speak Spanish at home are more likely to experience food insecurity. 29 The association between acculturation and diet change is very complex because multiple factors affect diet among immigrants. 30 Immigrants experience some nutritional changes after migrating that may cause unhealthy transitions. 31 But, unhealthy nutritional changes among immigrants do not always occur after migration. Latino immigrants who migrated to the United States after 2000, when unhealthy foods became more available than before in a home country, most likely had already acquired an unhealthy diet before migration. 32 Previous studies on food insecurity and acculturation focus heavily on Latino, in particular Mexican, immigrants. A study including Asian/Pacific Islanders shows somewhat different results: limited English proficient Asian/Pacific Islanders tend to eat fruits and vegetables more often than non-Hispanic Whites and English-proficient Asian/Pacific Islanders. 33 The association between English or heritage language proficiency and food insecurity can be very complex. But, for individuals and families who speak their heritage language, heritage language proficiency may affect types of food and access to food-related resources compared to those who have heritage culture but are not proficient in a heritage language. Future studies should further examine the impact of heritage language proficiency on food insecurity to better provide health education to patients who have heritage culture and diverse racial/ethnic groups as different cultures might share uncommon patterns of healthy eating.
Furthermore, this study identified demographic factors that increase risks for food insecurity: US-born English speakers, women, and unmarried individuals. Previous studies of free clinic patients indicate that US-born English speakers reported higher levels of depression compared to non-US-born English speakers and Spanish speakers. 21 Food insecurity and depression are related. 6 The association between food insecurity and depression can be bidirectional. For example, maternal depression increases risks of food insecurity of the household. 34 Furthermore, pregnant women who experience food insecurity are more likely to experience prenatal depression. 35 Although this study did not include a depression measure, future research is necessary to examine if food insecurity among US-born English speakers is associated with depression and if healthy nutrition education helps them to reduce the levels of depression, given that a healthy diets reduces the risk for depressive symptoms. 36
In addition to US-born English speakers, women and unmarried individuals have higher risks for food insecurity than men or married individuals. Previous studies of food insecurity associated with sex and marital status focused mostly on women. Married women and unmarried female-led households are more likely to report higher levels of food insecurity than married men or married male-led households. 37 Single mothers have high risks for food insecurity and may receive benefits from psychological counseling to cope with difficulties related to food insecurity. 38 Additionally, women who experience food insecurity improve their diet if they choose healthy food based on their own preference, rather than choosing food generally considered healthy. 39 Although information about women and food insecurity has been available, few studies have examined unmarried men and food insecurity. The results of this study suggest marital status is an important factor for food insecurity; it is necessary to increase the knowledge base about marital status and food insecurity for men as well as women.
This study has limitations. The cross-sectional design of this study examines associations but does not assess causal relationships among variables. Patients who were not literate in either English or Spanish were not included in this study. The participants are from a convenience sample. There is no information about response rate; however, research assistants who collected surveys indicated that there were not many refusals among eligible patients. There are differences in the magnitude of regression coefficients of the variables between all participants and those with heritage culture, which suggests inconsistency. Although this study was conducted at 1 free clinic, the results can be valuable to other free clinic populations and increase knowledge about free clinic patients, who are significantly understudied, as the free clinic sample in this study shares common characteristics with other free clinics, such as: uninsured patients only, income requirements, 0% revenue from the government, no affiliation/independent, and volunteer providers. 40
Conclusion
Low-income, uninsured primary care patients utilizing a free clinic experience food insecurity. Self-efficacy should be included in nutrition education programs to reduce the levels of food insecurity. For those who have heritage culture, heritage language may be a useful resource to reduce risks for food insecurity. US-born English speakers, women, and unmarried individuals are high-risk populations for food insecurity among free clinic patients. Future studies should implement health education programs including self-efficacy and healthy lifestyle interventions, and evaluate the effectiveness of such programs to determine whether self-efficacy interventions actually reduce levels of food insecurity. The impact of heritage language proficiency on food insecurity and the implementation utilizing heritage language proficiency to promote food security should be examined. Finally, high-risk groups such as US-born English speakers, women, and unmarried individuals may need interventions that fit their specific needs. Future studies should further examine why these groups have a high risk to better understand their needs for interventions.
Footnotes
Author Disclosure Statement
Dr. Kamimura, Ms. Jess, Ms. Trinh, Ms. Aguilera, Mr. Nourian, Ms. Assasnik, and Ms. Ashby declared no conflicts of interest with respect to the research, authorship, and/or publication of this article.
The authors received the following financial support for this article: this project was funded by the Public Service Professorship, Lowell Bennion Community Service Center at the University of Utah.
Acknowledgments
The authors want to thank the patients who participated in this study and acknowledge the contribution of the staff and volunteers of the Maliheh Free Clinic. In addition, we thank Alla Chernenko, MA, Travis Dixon, Anthony Mills, Molly Pace, BS, Liana Prudencio, MS, Naveen Rathi, Monica Scott, Emily Stephens, Tamara Stephens, BS, Jennifer Tabler, MS, and Lindsey Wright, BS, for their help with data collection, data entry, and translation related to this study.