Binge Eating and Weight Loss Behaviors in Low-Income Adults at Risk for Food Insecurity

INTRODUCTION

Food insecurity, limited or uncertain access to sufficient food, is associated with diet-related chronic diseases including obesity, diabetes, and cardiovascular disease.¹ There is evidence that food insecurity is associated with disordered eating and dietary restraint behaviors among college students, bariatric candidates, postpartum women, and food pantry clients.² However, less is known about these behaviors among low-income adult primary care patients at risk for food insecurity and obesity. This study assessed the relationship of food insecurity with binge eating and weight loss behaviors among low-income community health center patients.

METHODS

This was a cross-sectional secondary analysis of a prospective cohort of adult (≥ 21 years old) health center patients in a study assessing the impact of a Medicaid food and housing support program.³ Participants completed three annual surveys; questions about binge eating and weight loss behaviors were included in the third survey cycle and utilized for this analysis.

Food insecurity in the last 12 months was measured using the 10-item U.S. Department of Agriculture Adult Food Security Survey module.⁴ Raw scores (range 0–10) were dichotomized into food secure (0–2, high/marginal food security) or food insecure (3–10, low/very low food security). Binge eating behavior was defined as answering yes to “any episodes of excessive overeating”;⁵ those answering yes were also asked if they felt “distressed about [their] episodes of excessive overeating”. Weight loss behavior was defined as answering yes to “tried to lose weight” during the last 12 months.

Associations between food security status and patient characteristics (income, insurance, body mass index [BMI] category, and Supplemental Nutrition Assistance Program use in past 12 months) were compared using t-tests and chi-square tests. Weight and height data for BMI were obtained through self-report (weight) and electronic health record data (height, weight if missing self-report). Logistic regressions were used to estimate the association between four outcomes (binge eating behavior, distress from binge eating, weight loss behavior, and concurrent binge eating and weight loss behaviors) and food insecurity, adjusting for age, gender, and BMI category. Post-estimation adjusted proportions and adjusted odds ratios (aOR) were reported. Results were considered significant if P < 0.05. Analyses were conducted using Stata (16.1, StataCorp LLC, College Station, TX).

RESULTS

Out of 1031 participants, 887 participants completed surveys (Table 1). Mean age (SD) was 46.3 (11.4) years, 78.0% were female, and 53.4% were food insecure. Compared to food secure participants, food insecure participants had lower income and higher obesity.

Table 1.

Cohort Characteristics of Study Participants by Food Security Status*

	Total	Food Secure	Food Insecure	P Value
Sample, n	887	413	474	–
Age in years, mean(SD)	46.3 (11.4)	46.1 (12.0)	46.4 (10.9)	0.65
Female, n(%)	692 (78.0)	321 (77.7)	371 (78.3)	0.83
Hispanic ethnicity, n(%)	479 (55.1)	209 (51.5)	270 (58.3)	0.10
Race, n(%)
Asian, NH/PI, or NA/NA†	30 (3.4)	10 (2.5)	20 (4.3)	0.03
Black	142 (16.3)	79 (19.5)	63 (13.6)
Mixed	110 (12.6)	41 (10.1)	69 (14.8)
Other‡	174 (20.0)	77 (19.0)	97 (20.9)
White	414 (47.6)	198 (48.9)	216 (46.5)
Income ≥ $30 000, n(%)	318 (48.7)	184 (59.0)	134 (39.3)	< 0.001
Insurance at baseline, n(%)
Medicaid	708 (79.8)	329 (79.7)	379 (80.0)	0.91
Commercial	179 (20.2)	84 (20.3)	95 (20.0)
SNAP use, n(%)§	517 (58.3)	223 (54.0)	294 (62.0)	0.016
BMI in kg/m2, n(%)‖
< 25	162 (18.4)	89 (21.6)	73 (15.5)	0.03
25 to < 30	287 (32.5)	141 (34.3)	146 (31.0)
≥ 30	433 (49.1)	181 (44.0)	252 (53.5)

*Missing data: Total: Hispanic ethnicity- 18, Race- 17, Income- 234, BMI- 5; Food Secure: Hispanic ethnicity- 7, Race- 8, Income- 101, BMI- 2; Food Insecure: Hispanic ethnicity- 11, Race- 9, Income- 133, BMI- 3

†NH/PI- Native Hawaiian/Pacific Islander; NA/NA- Native American/Native Alaskan

‡Many participants choosing Other self-described as “Hispanic” or “Latino”

§SNAP- Supplemental Nutrition Assistance Program

‖BMI- Body mass index

A higher proportion of food insecure compared to food secure participants reported binge eating behaviors (24.5% vs 17.8%; aOR 1.50 [95% CI 1.08–2.09], P = 0.02). Among those reporting binge eating (n = 202), a higher proportion of food insecure participants reported binge eating distress (86.5% vs 75.3%; aOR 2.09 [95% CI 0.98–4.45], P = 0.06) (Fig. 1). A lower proportion of food insecure compared to food secure participants reported weight loss behaviors (55.8% vs 63.3%; aOR 0.73 [95% CI 0.54–0.99], P = 0.04). Concurrent binge eating and weight loss behaviors were similar between both groups (17.5% vs. 14.6%; aOR 1.24 [95% CI 0.87–1.78], P = 0.23).

Figure 1.

Binge eating and weight loss behaviors by food security status, adjusted proportions

DISCUSSION

In this sample of low-income community health center patients, those experiencing food insecurity were more likely to report binge eating behavior. However, one in six people reported concurrent binge eating and weight loss behaviors, regardless of food security status, suggesting a concerning cycle of binging and dieting. One proposed mechanism behind the increased frequency of binge eating in food insecure patients is the feast-or-famine hypothesis, where those experiencing food insecurity may disproportionately eat more during periods of food availability.⁶ Many of those reporting binge eating experienced distress from this behavior, and although not statistically significant in this smaller subset, distress was higher among those who were food insecure. Weight loss behaviors were common among all participants, with more than half reporting trying to lose weight in the past year.

Limitations include that the study sample was from one northeastern urban health system and all participants had health insurance, which is not generalizable to all healthcare settings. Additionally, the eating behavior survey questions were not diagnostic for eating disorders. This study of low-income health center patients revealed a high prevalence of eating-related behaviors that may exacerbate diet-related chronic diseases related to food insecurity. These findings highlight the need for clinicians to recognize and address (e.g., screening, referral) potential disordered eating behaviors among low-income patients at high risk of food insecurity.

Abbreviations

BMI

Body mass index

aOR

Adjusted odds ratio

NH/PI

Native Hawaiian/Pacific Islander

NA/NA

Native American/Native Alaskan

SNAP

Supplemental Nutrition Assistance Program

Funding

National Institute of Diabetes and Digestive and Kidney Diseases, DK124145, Anne N Thorndike, National Heart, Lung, and Blood Institute, HL163073, Anne N Thorndike, T32HL098048, Jessica Cheng, Health Resources and Services Administration, 5T32HP42013, Yingfei Wu, National Heart, Lung, and Blood Institute, HL157763, Jessica L McCurley.

Data Availability

De-identified data used in this study can be made available upon request to the corresponding author.

Declarations:

Conflict of Interest:

All co-authors declare no conflicts of interest.