Abstract
We evaluated the prevalence and correlates of household food insecurity (HFI) in participants attending weight management and bariatric surgery programs (WMP). We surveyed participants (N=298) residing with a child and romantic partner from two WMP offering bariatric surgery and medical weight management. We assessed HFI using the Hunger Vital Sign and conducted multivariate logistic regression to assess correlates of HFI. HFI was present in 29.2% of participants. There was increased odds of HFI in participants with lower income and those with 3 or more children living at home. There was no association with sex, race, program type, education, or BMI. HFI is highly prevalent among participants of WMP. WMP should screen all participants for food insecurity to address this potential barrier to a healthy lifestyle.
INTRODUCTION
Household food insecurity (HFI), a household without consistent access to enough food for an active, healthy life,(1) shares many risk factors and associated health outcomes with obesity. In 2017, 12.5% of households in the United States (US) were food insecure, with higher rates among households with children (17%).(1) Adults with HFI are more likely to have a poor diet, decreased physical activity, diabetes, hypertension, and cardiovascular disease.(2) Children with HFI are more likely to have iron deficiency anemia, poor diet, and poor academic performance.(3) Similar to obesity, risk factors for HFI include low household income, racial/ethnic minority, and living in the Southern US. Additionally, HFI is more likely for households with children or headed by a single adult and for individuals with poor physical/mental health and changes in housing or income stability.(4)
Those living in food insecure households may rely more on less expensive, high-energy foods,(1) which can lead to increased energy intake and obesity. Among adults, especially women, HFI has been linked to excess weight accumulation leading to obesity.(4) Very little is known, however, about the prevalence of HFI in adults attempting weight loss. One study of a pre-surgical bariatric population at a single center in Central Pennsylvania found a HFI prevalence of 27.6%,(5) indicating that adults with obesity attempting weight loss may be at especially high risk for HFI. HFI rates in medical weight management programs have not been studied, to our knowledge. Therefore, we aimed to describe the prevalence and correlates of HFI in a population of adult patients attending a weight management program (WMP) at two centers for either bariatric surgery or medical weight management.
METHODS
Study Design and Participants
This was a secondary data analysis of a cross-sectional study performed at two tertiary care adult WMPs in the United States located in North Carolina and Ohio between May and July 2017. Adults attending a visit were included if they were over the age of 18 years, could read and speak English, and had a romantic partner and at least one child aged 2–18 living at home with them the majority of the week (≥ 4 days), as this study focused on the family system impacting adult weight loss.(6) Participants were excluded if they were attending an initial intake visit. We recruited participants from the bariatric surgery programs and the medical weight loss programs (in which lifestyle, pharmacological, and meal replacement interventions are used) at both sites. Informed consent was obtained from all individual participants included in the study. At the beginning of their visit, a research assistant consented participants who then completed paper questionnaires during or after their visit. The institutional review boards of Wake Forest School of Medicine and The Ohio State University approved the study protocol.
Study Instruments and Measurements
HFI was assessed with the validated 2-item Hunger Vital Sign (HVS) screener: “Within the past 12 months we worried whether our food would run out before we got money to buy more” and “Within the past 12 months the food we bought just didn’t last and we didn’t have money to get more.” Response options included often, sometimes, or never. Participants screened positive if they answered sometimes or often on either question.(7) The gold standard for assessing HFI is the 18-item Food Security Survey Module which classifies participants as high food security (no problems or anxiety about food), marginal food insecurity (anxiety about accessing adequate food, but variety and quantity not substantially reduced), low food insecurity (reduced quality and variety, but normal quantity), and very low food insecurity (food intake is reduced by one or more family member).(1) A positive screen on the HVS captures both food insecurity and marginal food insecurity (α=0.86). The HVS has a sensitivity of 97% and specificity of 81.9% compared to the 18-item screener.(7)
Participants reported demographics within the questionnaire including age, sex, race (categorized in this analysis as non-Hispanic white or other), highest completed education, household income, and the number of children living at home. Participants also reported their most recent height, weight, and the type of WMP they were participating in (categorized as bariatric surgery program, all other medical weight management programs).
Statistical Analysis
We used Pearson’s chi-square tests and independent sample T-tests to examine the association of HFI with participant demographics. Logistic regression analysis was performed examining correlates of HFI including sex, race, income, education, number of children, bariatric surgery program (yes, no), and participant BMI. Because female and not male sex has been associated with increased weight gain among individuals in food insecure households,(8) we also examined sex as a possible effect modifier; however, there were no significant differences so sex was considered a possible confounder in the final analyses.
RESULTS
Participants (N=298) predominantly identified as white (75%) and female (85%), with an average age of 41.2 (SD=7.1) years. The majority (89%) completed at least some college, and 66% had an annual household income of at least $60,000. Most participants had obesity (22% Class 1, 20% Class 2, and 43% Class 3), Mean(SD) BMI was 39.6(10.6) and about half of participants (54%) were in a bariatric surgery program (Table 1).
Table 1:
Demographics of Participants
| Total N=298 N (%) | Food Secure N=211 N (%) | Food Insecure N=87 N (%) | P-value | |
|---|---|---|---|---|
| Female Sex | 252 (84.6) | 175 (82.9) | 77 (88.5) | 0.2 |
| Non-Hispanic White Race | 217 (74.3) | 154 (74.4) | 63 (74.1) | 0.9 |
| Bariatric Surgery Program | 156 (54.0) | 106 (52.2) | 50 (58.1) | 0.4 |
| Income | <0.001 | |||
| <$20,000 | 17 (5.8) | 4 (1.9) | 13(15.1) | |
| $20–39,999 | 46 (15.7) | 17 (8.2) | 29(33.7) | |
| $40–59,999 | 38 (13.0) | 22 (10.6) | 16 (18.6) | |
| $60–99,999 | 83 (28.3) | 66 (31.9) | 17 (19.8) | |
| $100,000 or more | 109 (37.2) | 98 (47.3) | 11 (12.8) | |
| Education | <0.001 | |||
| High school or less | 32 (10.8) | 12 (5.7) | 20 (23.0) | |
| Some college | 83 (27.9) | 49 (23.2) | 34 (39.1) | |
| Associate degree | 36 (12.1) | 24 (11.4) | 12 (13.8) | |
| Bachelor’s degree | 86 (28.9) | 72(34.1) | 14 (16.1) | |
| Master’s degree or higher | 60 (20.1) | 53 (25.1) | 7 (8.1) | |
| Number of children | <0.001 | |||
| 1 | 67 (22.6) | 55 (26.2) | 12 (13.8) | |
| 2 | 122 (41.1) | 96 (45.7) | 26 (29.9) | |
| 3+ | 108 (36.4) | 59 (28.1) | 49 (56.3) | |
| Weight Status | 0.7 | |||
| Healthy Weight | 7 (2.4) | 4 (1.9) | 3 (3.6) | |
| Overweight | 37 (12.7) | 27 (13.0) | 10 (11.9) | |
| Class 1 Obesity | 64(21.9) | 49 (23.6) | 15 (17.9) | |
| Class 2 Obesity | 58(19.9) | 42 (20.2) | 16 (19.1) | |
| Class 3 Obesity | 126 (43.2) | 86(41.4) | 40 (47.6) | |
| BMI (mean (SD)) | 39.6 (10.6) | 39.1 (9.9) | 40.6 (12.1) | 0.1 |
Almost one in three participants (29.2%) reported HFI. Of those participants with HFI, 30% only responded affirmatively to the first question that they were worried that food would run out, 5% only responded affirmatively to the second question that the food had not lasted and they did not have money to buy more, while the majority (65%) answered affirmatively to both.
In bivariate analyses HFI was more common among participants with lower income and education levels, and those with an increased number of children living at home (Table 1). In multivariate analyses participants with lower income categories had increased odds of HFI (OR 13.33 [95% CI 3.12–57.00] for those making <$20,000, 10.80 [3.70–31.50] for those making $20–39,999, and 5.31 [1.81–15.54] for those making $40–59,999) compared to those making and $100,000 or more. Participants also had increased odds of HFI if there were three or more children living at home (3.33 [1.36,8.15]). Odds of HFI were not different by program type (bariatric surgery or medical weight management) or participant BMI.
DISCUSSION
HFI was present in 29.2% of adult participants in two weight management and bariatric surgery programs. HFI was more common among low-income households and households with three or more children, although it was present among all income and education levels. Program type and participant BMI were not associated with HFI. Our results were similar to published data by Price and colleagues who noted a marginal HFI rate of 27.6% and found that HFI was more common in low-income households in a pre-bariatric surgery sample in Pennsylvania.(5)
The rate of HFI observed in these WMP populations is much higher than the national rate of 17% for households with children, and is higher than the overall rate of HFI in the local communities where our WMPs are located (Columbus, Ohio=16.5%; Winston-Salem, NC=15.5%).(9) The reasons for this should be explored in future qualitative studies. Inclusion criteria for our study comprised having at least one child 2–18 years of age living in the home, and HFI is known to be higher in households with children. Some participants were using meal replacements, and it is possible that this additional expense was financially straining and contributed to HFI. It is also possible that due to participation in a WMP, participants were making changes to their diet, such as eating increased fruits and vegetables, which were more expensive. Future research should explore methods to connect participants in WMPs with nutrient dense and affordable healthy foods, such as through partnerships with community gardens or discounts at food retail stores.
Despite the strengths, including a large sample of both bariatric and medical WMP participants at two institutions, our study was limited by lack of information regarding many factors known to be important to HFI such as nutritional assistance benefits, recent changes in employment or income, disability status, and overall physical/mental health. Participants self-reported height and weight, which could have led to over or under-estimation of BMI. Our participants were primarily non-Hispanic white, female, and highly educated, which although similar to many bariatric and medical WMP programs,(10) may limit generalizability to all patient populations. Additionally, all of our participants were parents of children living at home the majority of the week, a population known to have higher levels of HFI, limiting the generalizability of our findings to adults without children living at home.
HFI was present in 29.2% of participants in two weight management and bariatric surgery programs, a rate more than twice than the national average. Providers working in WMPs should screen all patients for HFI, as this is a population that is likely at high risk for HFI regardless of income or education levels. For patients undergoing bariatric surgery, providers should discuss during the preoperative period if patients will have access to appropriate food, supplements, and vitamins after surgery. All WMPs should consider interventions that address factors that both HFI and obesity have in common, such as providing access to fresh fruits and vegetables, offering suggestions for preparing healthy family meals on a limited budget, and assisting eligible participants in registering for food assistance programs. Future studies should assess HFI longitudinally, throughout a patient’s participation in the WMP, and the effect that HFI may have on patient’s and family’s long-term outcomes.
Table 2:
Correlates of food insecurity
| OR (95% CI) | |
|---|---|
| Female Sex | 1.18 (0.46–3.01) |
| Non-Hispanic White Race | 1.17 (0.56–2.43) |
| Bariatric Surgery Program | 0.63 (0.31–1.25) |
| Income | |
| <$20,000 | 13.33 (3.12–57.00)*** |
| $20–39,999 | 10.80 (3.70–31.50)*** |
| $40–59,999 | 5.31 (1.81–15.54)** |
| $60–99,999 | 2.07 (0.80–5.35) |
| $100,000 or more | ref |
| Education | |
| High school or less | 3.87 (0.96–15.65) |
| Some college | 2.50 (0.85–7.39) |
| Associate degree | 2.61 (0.76–8.98) |
| Bachelor’s degree | 1.13 (0.37–3.47) |
| Master’s degree or higher | ref |
| Number of children | |
| 1 | ref |
| 2 | 1.70 (0.67–4.32) |
| 3+ | 3.33 (1.36–8.15)** |
| BMI | 0.99 (0.96–1.03) |
Grant Support:
Research assistant support for the project described was provided by the National Center for Advancing Translational Sciences (NCATS), National Institutes of Health, through Grant Award Number UL1TR001420. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Footnotes
Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of a an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.
Callie Brown: no conflict of interest
Joseph Skelton: no conflict of interest
Deepak Palakshappa: no conflict of interest
Keeley Pratt: no conflict of interest
“All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.”
References:
- 1.Coleman-Jensen A, Rabbitt M, Gregory C, Singh A. United States Department of Agriculture Household Food Security in the United States in 2017. 2018.
- 2.Gundersen C, Ziliak JP. Food insecurity and health outcomes. Health affairs. 2015;34(11):1830–9. [DOI] [PubMed] [Google Scholar]
- 3.Cook JT, Frank DA, Berkowitz C, Black MM, Casey PH, Cutts DB, et al. Food insecurity is associated with adverse health outcomes among human infants and toddlers. The Journal of nutrition. 2004;134(6):1432–8. [DOI] [PubMed] [Google Scholar]
- 4.Franklin B, Jones A, Love D, Puckett S, Macklin J, White-Means S. Exploring mediators of food insecurity and obesity: a review of recent literature. Journal of community health. 2012;37(1):253–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Price JA, Zickgraf HF, Rigby A. Food insecurity in a pre-bariatric surgery sample: prevalence, demographics and food shopping behaviour. Public Health Nutrition. 2019:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Brown CL, Pratt K, Martin S, Hulshult H, Skelton JA. Weight Control Practices in Children of Parents Participating in Weight Management Programs. Childhood Obesity. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hager ER, Quigg AM, Black MM, Coleman SM, Heeren T, Rose-Jacobs R, et al. Development and validity of a 2-item screen to identify families at risk for food insecurity. Pediatrics. 2010;126(1):e26–32. doi: 10.1542/peds.2009-3146. [DOI] [PubMed] [Google Scholar]
- 8.Seligman HK, Berkowitz SA. Aligning programs and policies to support food security and public health goals in the United States. Annual review of public health. 2019;40:319–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Food Insecurity in The United States: Feeding America; 2019. [September 30, 2019]. Available from: https://map.feedingamerica.org/.
- 10.Kizy S, Jahansouz C, Downey MC, Hevelone N, Ikramuddin S, Leslie D. National trends in bariatric surgery 2012–2015: demographics, procedure selection, readmissions, and cost. Obesity surgery. 2017;27(11):2933–9. [DOI] [PubMed] [Google Scholar]