Food Insecurity is Associated with Increased Weight Loss Attempts in Children and Adolescents

Abstract

Household food insecurity (HFI) has been associated with adverse childhood outcomes and shares many common risk factors with obesity. Half of adolescents with overweight or obesity are actively attempting to lose weight. We aim to evaluate whether HFI is associated with weight loss attempts and unhealthy weight loss control practices in children. We examined cross-sectional data of children ages 8–15 years old from the National Health and Nutrition Examination Survey. Attempted weight loss was more common among children with very low food security (OR 1.50, 95% CI 1.09, 2.07). Children with a healthy weight with very low food security had increased odds of attempting weight loss (OR 1.51, 95% CI 1.00, 2.26) but there was no association in children with overweight or obesity. Very low food security was also associated with unhealthy weight control practices (OR: 1.42, 95% CI: 1.04, 1.93). Physicians should counsel all children and adolescents on healthy and unhealthy weight loss behaviors, regardless of weight or food security status.

Introduction

The consequences of living in a food-insecure household, defined by the United States Department of Agriculture (USDA) as a household without consistent access to enough food for an active and healthy life, have been well-studied over the past two decades.¹ Food insecurity is associated with adverse childhood outcomes such as worse physical health, cognitive scores, and mental health.^2–5 An estimated 13.9% of US households with children – 6.0 million children – lived in food-insecure households in 2018.¹ While the association between childhood obesity and food insecurity is unclear, the two conditions share common risk factors, such as poverty and low socioeconomic status.^6,7 Previous studies have shown evidence that food insecurity is associated with unhealthy eating patterns and lifestyle practices (e.g. increased consumption of low-cost, calorically-dense foods), which may lead to increased chances of developing obesity.^6,8,9

Obesity affects over 17% of children and adolescents, and studies have shown approximately half of adolescents with overweight or obesity actively attempt to lose weight.^10,11 Children report various motivations for attempting weight loss such as desiring to improve at sports, teasing from other children, or because their peers are attempting weight loss.¹² Children and adolescents in all weight categories report unhealthy weight control practices, for example the use of severely restrictive diets, induced vomiting, and the use of laxatives.^12–14 In addition to serious medical complications from disordered eating behaviors, a recent study showed that adolescents who use unhealthy weight control practices have greater increases in body mass index (BMI) over a 10-year period compared to adolescents of the same weight without use of these practices.¹⁵ Furthermore, household food insecurity (HFI) has been associated with eating behaviors that can contribute to childhood obesity, such as increased consumption of sugar-sweetened beverage.¹⁶ Adolescents with HFI are more likely to engage in unhealthy behaviors (e.g. skipping breakfast, smoking), but it is unclear if weight loss is the motivation driving these behaviors. Although numerous studies have examined the association between food insecurity and obesity, there is limited data on the relationship between food security status and weight management in children and adolescents. The objective of this study was to evaluate whether HFI is associated with self-reported weight loss attempts and unhealthy weight control practices in a nationally representative sample of children and adolescents in the United States.

Methods

Study setting and design

We examined pooled cross-sectional data of children ages 8–15 years from the National Health and Nutrition Examination Survey (NHANES) between 2005–2012. We limited to this age group because NHANES did not collect weight loss behaviors among those under age 8, and those 16 and older responded to different questions. Following 2012, detailed weight loss efforts were not collected. Participants who were missing food insecurity data (N=37), missing BMI (N=106) or pregnant (N =3) were excluded for a total unweighted sample size of 6,077.

Measures

The primary exposure of interest was HFI, which was assessed using the USDA 18-item Household Food Security Survey Module. NHANES uses the established scoring system for responses to the survey. Households were categorized as having high food security (0 affirmative responses; no indication of food access problems), marginal food security (1–2 affirmative responses; anxiety over food sufficiency), low food security (3–7 affirmative responses; individuals reduce the quality of their diet but not food intake) and very low food security (≥8 affirmative responses; multiple indications of disrupted eating patterns and reduced food intake). Participants with low or very low food security (FS) are considered to have HFI. For this study, we utilized the 4-category household food security status (High, marginal, low, and very low FS).

The primary outcomes were self-reported weight loss attempts and unhealthy weight control practices. Participants were asked “In the past year, how often have you tried to lose weight?” and answered “never,” “sometimes,” or “a lot.” Participants who answered “sometimes” or “a lot” were then prompted to answer “How often did you try to lose weight through the following behaviors” and were given 20 weight loss methods to choose from; participants could select more than one method.¹⁴ Response options included “never,” “sometimes,” or “a lot.” For this study, we defined unhealthy weight control practices as affirmative responses (“sometimes” or “a lot”) to “skipped meals” or “starved to lose weight.”

Statistical analysis

BMI percentiles were standardized for age and sex by growth charts from the Centers for Disease Control and Prevention and categorized into healthy weight (<85^th percentile, including underweight), overweight (85^th to <95^th percentile), and obesity (≥95^th percentile).¹⁷ Participant characteristics were compared using t-tests and Pearson chi-square tests as appropriate with p values of <0.05 considered statistically significant. We used separate multivariate logistic regression models to test the association between HFI and weight loss attempts and weight control practices, controlling for age, sex, race/ethnicity (white, Black, Hispanic, other), poverty status, and weight class (healthy weight vs. overweight/obesity). Participants were considered below the poverty level if their reported household income was less than 100% of the federal poverty level. We also separately examined weight status as a possible effect modifier, using separate models (i.e., fully-interacted) for healthy weight and overweight/obesity. All analyses accounted for the complex survey design of NHANES by applying sample weights, clustering, and the primary sampling unit. Analyses were completed using Stata 15.0 (StataCorp, College Station, TX). The Wake Forest School of Medicine Institutional Review Board deemed this study of publically available, de-identified data exempt from human subjects research.

Results

HFI was present in 19.2% of participants (12.9% with low FS and 6.3% with very low FS). Low and very low FS were more common in children with black and Hispanic race/ethnicity and in children who are living below the federal poverty level (Table 1). Half (49.1%) of participants reported attempted weight loss in the past year. Attempted weight loss was more common among children with low (53.6%) and very low (58.9%) FS compared to children with high FS (p<0.0001). Utilization of unhealthy weight control practices was reported in 18.4% of the participants. Children living in low (23.4%) or very low (26.3%) FS households were significantly more likely to report unhealthy weight control practices than children in high FS households (15.9%, p<0.0001).

Table 1.

Demographics by food security status, adjusted for complex survey design to represent the U.S. population.^a

Demographics:	Total N=6,077	High food security N=3,681	Marginal food security N=786	Low food security N=1,087	Very low food security N=523	p-valueb
Age 8–11, %	48.5	48.3	50.6	48.0	47.5	0.69
Male Sex, %	50.9	51.6	46.3	48.4	55.8	0.03
Race, %
White	57.7	66.1	37.7	36.0	39.7	<0.001
Black	14.2	11.2	22.2	21.3	21.0
Hispanic	20.0	14.2	33.7	36.4	29.9
Other	8.1	8.5	6.4	6.3	9.4
Below Poverty Level, %	21.7	12.2	35.8	47.4	52.7	<0.001
Weight Status, %						<0.001
Overweight/obesity	37.8	33.4	41.8	42.1	40.4

^aThese represent unadjusted N’s and adjusted proportions.

^bp-value is based on chi-square test and denotes a difference in percentage of individuals in each weight class as a function of age, sex, income, or race, respectively.

In multivariable models we found that children living in very low FS households were significantly more likely to report attempting weight loss in the past year than children living in high FS households (OR:1.50, 95% CI: 1.09, 2.07; p=0.01) (Table 2). This was seen among children with healthy weight (OR: 1.51, 95% CI: 1.00, 2.26, p<0.05), but was not significant among children with overweight/obesity. Children in marginal (OR: 1.33, 95% CI: 1.08, 1.65, p=0.01) and very low FS (OR: 1.42, 95% CI: 1.04, 1.93; p<0.05) households also had a significantly higher odds of reporting unhealthy weight control practices than children in high FS households. WIC/SNAP eligibility and household size were analyzed as potential confounders but were not significantly associated.

Table 2:

Multivariable analysis evaluating association between food security and weight loss behaviors

Weight Loss Attempts
	Total		Healthy Weight		Overweight/Obesity
	OR (95% CI)	p-value	OR (95% CI)	p-value	OR (95% CI)	p-value
High FS	Ref		Ref		Ref
Marginal FS	1.01 (0.77, 1.31)	0.96	1.04 (0.71, 1.52)	0.83	0.93 (0.60, 1.45)	0.75
Low FS	1.06 (0.86, 1.30)	0.58	1.09 (0.85, 1.40)	0.51	0.99 (0.62, 1.60)	0.98
Very low FS	1.50 (1.09, 2.07)	0.01	1.51 (1.00, 2.26)	<0.05	1.48 (0.90, 2.43)	0.12
Unhealthy Weight Control Practices
	Total		Healthy Weight		Overweight/Obesity
	OR (95% CI)	p-value	OR (95% CI)	p-value	OR (95% CI)	p-value
High FS	Ref		Ref		Ref
Marginal FS	1.33 (1.08, 1.65)	0.01	0.98 (0.67, 1.43)	0.91	1.63 (1.19, 2.22)	0.003
Low FS	1.15 (0.90, 1.46)	0.25	1.07 (0.77, 1.49)	0.67	1.21 (0.88, 1.65)	0.23
Very low FS	1.42 (1.04, 1.93)	<0.05	1.45 (0.88, 2.39)	0.15	1.39 (0.99, 1.95)	0.06

Abbreviation: FS, food security

Discussion

In this nationally representative sample of children 8–15 years of age, we found a significant association between very low FS and self-reported attempted weight loss. Children in very low FS households had a higher odds of reporting unhealthy weight practices than children in high FS households.

This data suggests that children living with very low FS are more likely to report attempting weight loss within the past year compared to children living with all other levels of FS, even if they have a healthy weight. It is possible that this association was found because children with very low FS report attempting weight loss as a coping mechanism for having inconsistent access to food or potential pressure from parents to “eat less.” Alternatively, because parents are worried about eating too much food and the family cannot afford more, older children and adolescents may believe it is because of the need to lose weight. Further research is necessary to evaluate these theories. These results support the importance of education on healthy weight management among all children, not just children with overweight or obesity. Providers should consider HFI when developing weight management strategies for their patients.

There are several limitations to this study. This was a secondary analysis of survey data, so the primary outcome variables were self-reported and conclusions about causality cannot be drawn. Additionally, response options for weight control practices were limited and undefined, so it is difficult to determine the motivation behind some of the practices (such as skipping meals) or if there were other unhealthy practices that were not available options. Further, we cannot account for any effect of the parents’ weight status in this analysis given the survey data restrictions.

Conclusion

We found that very low FS is significantly associated with increased likelihood of attempting weight loss in the past 12 months among children with healthy weight. Additionally, very low FS was significantly associated with utilizing unhealthy weight control practices. Future studies should evaluate the motivations that drive children living with very low FS to attempt weight loss and how this might differ from the motivations of children living with different levels of FS. Health care providers should educate all children and adolescents on healthy weight management behaviors, regardless of their weight or food security status.