Abstract
Objective:
To examine dual food and energy hardship and internalizing and externalizing behavior problems in 9 year-old children.
Methods:
We conducted a cross-sectional analysis of the Fragile Families and Child Wellbeing Study, a prospective national urban birth cohort, when the children were 9 years-old. Maternal reported “food hardship” (ever hungry and/or ever received free food) and “energy hardship” (ever unable to pay utility bill and/or utility shut-off) within the past year, and child behavior using the Child Behavior Checklist/6–18 years were assessed. Multiple logistic regression analyses estimated associations between individual and dual food and energy hardship and child behavior problems, adjusting for a priori covariates (child sex, health insurance, maternal sociodemographic characteristics, poverty, reported health, attention deficit hyperactivity disorder, depressive symptoms, smoking, and substance and alcohol abuse).
Results:
Approximately 10% of households reported dual food and energy hardship. Children experiencing dual food and energy hardship had 3 times greater odds of withdrawn/depressed behaviors (adjusted odds ratio [AOR]=2.8, 95% CI:1.4–5.5), 3 times greater odds of somatic complaints (AOR=3.2, 95% CI:1.5–6.9), and 4 times greater odds of rule breaking behavior (AOR=3.7, 95% CI:1.5–9.2) in the borderline/clinical range compared to children with no hardship, and had 4 times greater odds of borderline/clinical range somatic complaints (AOR=4.2, 95% CI: 1.7–10.3) compared to children with only energy hardship.
Conclusions:
Children experiencing dual food and energy hardship have greater odds of coexisting internalizing and externalizing behaviors after controlling for possible confounders. Providers can consider screening and resource referrals for these addressable hardships alongside behavior assessments in the clinical setting.
Keywords: food hardship, energy hardship, poverty, child behavior, screening
INTRODUCTION
Poverty has been associated with poor health and behavior problems in children.1,2 Material hardships exacerbating hunger or thermal discomfort may contribute to physiologic stress on child behavior and development.3,4 Material hardships may also disrupt child development through parental stress, maternal depression, and negative parenting.5–7 Stress and poor parenting contribute to child behavioral maladjustment. Problem behaviors have been associated with poor academic motivation, which can in turn limit children’s social-emotional learning and relationship skillfulness development.8,9
Food hardship and energy hardship represent two important forms of material hardship for low-income low-resource families who experience challenges in meeting basic needs of nutrition and utilities for lighting, cooking, heating and cooling.10 Food hardship is a measure of hunger, uncertainty about one’s next meal, and receiving free food out of necessity.11 It serves as a proxy for food insecurity, defined by the U.S. Department of Agriculture (USDA) as, “a limitation and/or uncertainty of having available or acquiring adequate food for a household.”12
Energy hardship is defined as an inability to pay the electric or gas utility bill. It is an approximation of energy insecurity, a novel construct that refers to “the inability to adequately meet household energy needs.”4 Energy insecurity has physical, economic and behavioral dimensions and has been measured as the experience of either threatened or actual utility shut-off or delivery refusal, use of cooking stove as heat source, or forgoing heating or cooling due to inability to pay utilities.13 Energy insecurity is associated with decreased food- and healthcare-related expenditures, thermal discomfort, and potentially unsafe heating or cooling methods, which can lead to adverse social, environmental and health outcomes.4,13,14
Food insecurity has been associated with negative child development and behavior outcomes.3,6,15–20 Likewise, energy insecurity has been associated with cognitive and developmental impairments and poor educational outcomes among children.4,13,14 Although economic disadvantage can be difficult for families to overcome, food hardship and energy hardship are important health-related social factors that may also be addressable. Pediatricians of disadvantaged children with behavior problems have an opportunity to attend to addressable needs as part of primary care.
Although general associations of poverty and food insecurity with health outcomes have been established, few studies have examined food with energy hardship and their association with the social-behavioral development of elementary school-aged children.10,18,21 To address this gap, our study focused on household dual food and energy hardship and coexisting behavior problems in 9 year-old children from a national cohort of socioeconomically vulnerable families. We hypothesized that children who experience dual food and energy hardship will exhibit behavior problems compared to similar children who experience neither or only one hardship.
METHODS
Study Design and Participants
We conducted a cross-sectional analysis of the Fragile Families and Child Wellbeing Study (FFCWS), which follows a birth cohort of 4,898 children who were born between 1998 and 2001 in over 20 large cities (populations > 200,000) to “fragile families” at greater risk of breaking up and living in poverty than more traditional families.22 Children were identified via a stratified random sampling frame with a purposeful oversampling of non-marital births. Data were collected from parents postpartum and subsequently via in-person or telephone interviews and mailed surveys when focal children were ages 1, 3, 5, 9, and 15 years-old. Data from the biological mothers’ 9-year interviews, conducted via telephone or in-person from 2007–2010, were used to capture material hardships and child behaviors directly after the onset of the Great Recession. A total of 3,508 mothers (75% of the eligible sample) completed 9-year interviews with complete information on food and energy hardship. Sample sizes for final analyses ranged from 3,094 to 3,135 due to missing items at random for outcome variables. Mothers who did not participate in the 9 year-interviews differed significantly by baseline housing tenure (rented, 69% non-participants vs. 65% participants [P=0.0062]), baseline education level (less than high school education, 45% non-participants vs. 36% participants [P<0.001]), and race/ethnicity (Black, 44% non-participants vs. 52% participants [P<0.001]; Latino, 33% non-participants vs. 25% participants [P<0.001]) from those who participated.
Outcomes
Behaviors
Mothers reported on child behaviors within the last 6 months based on eight behavior problem scale items from the validated Child Behavior Checklist/6–18 years (CBCL). Responses were scored as not true (0), somewhat true (1), or often or very true (2), and raw scale scores were calculated by summing the (1) and (2) scores. Sex- and age-normed standardized T-scores, with lower limit truncated at T=50 (with standard deviation=10) in the Achenbach System of Empirically Based Assessment (ASEBA) scoring package,23 were dichotomized into normal range and borderline/clinical range based on problem scale scores ≥ 93rd percentile.23 T-scores ≥65 were designated the borderline/clinical range and reflect symptom severity.
Internalizing behaviors included the Anxious/Depressed, Withdrawn/Depressed, and Somatic Complaints problem scales. Scores in the borderline/clinical range suggest a concern for maladaptive functioning in inward-focused behaviors. Externalizing behaviors included the Rule Breaking Behaviors and Aggressive Behavior problem scales. Scores in the borderline/clinical range suggest a concern for maladaptive functioning in behaviors toward the external environment. Other behaviors were encompassed by the Attention Problems, Social Problems, and Thought Problems scales. Scores in the borderline/clinical range suggest a concern for maladaptive functioning in social-emotional communication and interaction.
Predictors
Food Hardship
Household food hardship was measured as a mother’s yes/no response to either of the following FFCWS survey questions: “In the past 12 months, were you ever hungry, but you just couldn’t afford more food?” and “In the past 12 months, did you receive free food or meals?” Although earlier years of the FFCWS included the validated U.S. Household Food Security Survey for a subset of family interviews, neither this screen nor the USDA 6-item Short Form screen12 were used by the FFCWS for 9-year interviews.
Our food hardship measure generally depicts both low food intake and high food need, and the first part of the measure is similar to other brief screening questions for food insecurity validated in low-income populations and counts households with very low food insecurity.24,25 We assumed that maternal food hardship is a proxy for family food hardship. The second question in the measure reflects use of emergency food resources by families who receive Supplemental Nutrition Assistance Program (SNAP) benefits as well as families who are not eligible to receive SNAP benefits. Only 19% of mothers who reported receiving income from SNAP also reported receiving free food or meals (Pearson’s r=0.19) in the last 12 months. This suggests that most FFCWS households did not consider the SNAP benefits to be free food.
Energy Hardship
Household energy hardship was measured as a mother’s yes/no response to the following FFCWS survey questions: “In the past 12 months, did you not pay full amount of gas/oil/electric bill because there wasn’t enough money?” and “In the past 12 months, gas/oil/electric utility shut off because there wasn’t enough money?” Neither additional components of energy insecurity4, nor the validated Home Energy Insecurity Scale13 were asked by FFCWS in the 9-year interviews. Our energy hardship measure is similar to questions included in the Home Energy Insecurity Scale and captured households struggling to meet energy needs.
Covariates
We identified characteristics detected a priori as significant predictors of children’s social-emotional development and behavior problems as covariates in the regression analyses.6,10,14,16,17,19,21,26 These included child sex and health insurance status and maternal factors: race and age; Latino ethnicity; education level; employment status; housing tenure (dichotomized to renting or living with renters vs. owning or living with owners); household income-to-poverty ratio, using year-specific poverty thresholds established by the U.S. Census Bureau at the time of the 9-year interview; self-reported health status (dichotomized to good vs. fair/poor); depressive symptoms; attention deficit hyperactivity disorder (ADHD) history; smoking; illicit and prescribed substance abuse; and alcohol abuse (dichotomized to “3 drinks or less” vs. “4 drinks or more” in a single day).
The FFCWS based the presence of maternal depressive symptoms on questions derived from the Composite International Diagnostic Interview - Short Form (CIDI-SF), Section A. The CIDI-SF is a validated and standardized instrument meeting criteria from the Diagnostic and Statistical Manual of Mental Disorders – Fourth Edition (DSM-IV).27 The FFCWS classified mothers in the 9-year cohort with depressive symptoms via a liberal scale of two weeks of depressive symptoms experienced in the past 12 months and designated mothers as probable cases if the major depressive score was greater than or equal to 3. Maternal substance abuse was dichotomized as a yes/no variable based on pooled response to FFCWS questions about ever using amphetamines or other stimulants, sedatives, tranquilizers, analgesics/prescription painkillers without healthcare provider recommendation, heroin, cocaine or crack, LSD, inhalants, and/or marijuana in the last 12 months.
Statistical Analyses
Descriptive statistics were computed for our total sample. Unpaired t-tests compared differences in mean raw scores and T-scores for behavior problem scales between children with dual food and energy hardship and children without hardship. Multiple logistic regressions then estimated associations between T-scores in the borderline/clinical range for child internalizing behaviors (anxious/depressed, withdrawn/depressed, somatic complaints), externalizing behaviors (rule-breaking and aggression), and other problems (attention, thought, and social problems) and a four-category variable for “dual food and energy hardship,” “only food hardship,” “only energy hardship,” and “no hardship.” Primary regression analysis used “no hardship” as the comparison group, and secondary analyses used “only energy hardship” and “only food hardship” as comparison groups. Regression models controlled for child sex and health insurance status and maternal factors of race and age, Latino ethnicity, education level, employment status, housing tenure, self-reported health status, depressive symptoms, ADHD history, smoking, substance misuse, alcohol abuse, and household income-to-poverty ratio.
Analyses were conducted using the SAS® statistical program (version 9.1.4) and significant levels of difference were based on a Bonferroni correction to control for the number of comparisons, which yielded an alpha value of 0.006. The Columbia University Medical Center Institutional Review Board approval was obtained for this study.
RESULTS
Sample Characteristics
Almost 10% of households reported dual food and energy hardship (n=334), 7% reported only food hardship (n=233), and 24% reported only energy hardship (n=827). Sociodemographic characteristics of the study population are summarized in Table 1. Roughly half of the mothers were Black and unmarried and about two-thirds reported an annual household income less than 200% of the federal poverty level. Seventeen percent met criteria for depressive symptoms and 32% reported substance abuse. The effect of dual food and energy hardship could not be entirely explained by poverty threshold level after cross-tabulation showed the variables were not highly correlated (Pearson’s r<0.22, p<0.001).
Table 1:
Sociodemographic data and food and energy hardship characteristics, Fragile Families and Child Wellbeing Study sample (n=3,508)
| Variable | Value |
|---|---|
| Child’s sex, n (%) | |
| Female | 1,843 (52.5) |
| Male | 1,665 (47.5) |
| Maternal race/ethnicity, n (%) | |
| White | 1,059 (30.7) |
| Black | 1,790 (51.9) |
| Asian | 79 (2.3) |
| Amerindian | 146 (4.2) |
| Other | 374 (10.9) |
| Latino | 876 (25.2) |
| Maternal age, mean ± SD, years | 34.4 ± 6.0 |
| Maternal marital status married, n (%) | 1,321 (37.8) |
| Maternal employment, n (%) | 2,176 (62.1) |
| Mean household gross annual income, mean ± SD | $44,849 ± $53,603 |
| Mean household income-to-poverty ratio, mean ± SD | 1.1 ± 5.3 |
| Household annual income < 200% federal poverty level, n (%) | 2,305 (66.3) |
| Received income from SNAP in last 12 months, n (%) | 1,603 (45.7) |
| Maternal highest educational attainment less than high school, n (%) | 764 (21.8) |
| Maternal housing tenure, n (%) | |
| Rents/lives with renters | 2,381 (69.6) |
| Owns/lives with owners | 1,012 (29.6) |
| Other housing arrangement | 28 (0.8) |
| Maternal tobacco use, n (%) | 1,031 (29.3) |
| Maternal alcohol abuse, n (%) | 554 (15.8) |
| Maternal substance abuse, n (%) | 318 (32.2) |
| Maternal ADHD historya, n (%) | 20 (2.6) |
| Maternal self-reported health fair/poor, n (%) | 584 (16.6) |
| Maternal depressive symptomsb, n (%) | 612 (17.5) |
| Reported no child health insurance coverage, n (%) | 171 (4.9) |
| Food hardship measure, n (%) | |
| Only hungry but couldn’t afford food | 134 (3.8) |
| Only received free food | 314 (9.0) |
| Hungry and received free food | 119 (3.4) |
| Energy hardship measure, n (%) | |
| Only couldn’t afford to pay full utility bill | 792 (22.6) |
| Only had utility shut-off | 83 (2.4) |
| Couldn’t pay full utility bill and had utility shut-off | 286 (8.2) |
| Dual food and energy hardship, n (%) | 334 (9.5) |
Total n=769 for sample that furnished responses to the two FFCWS questions on taking medication for attention deficit and receiving counseling/therapy for attention problems; the remaining 2,734 mothers skipped the questions due to the 9-year survey skip pattern sequence.
From the Composite International Diagnostic Interview-Short Form. Indicates whether respondent meets the liberal criteria for depressive symptoms.
Children experiencing dual food and energy hardship had highest T-scores on the Withdrawn/Depressed scale (T-score=57.1; standard deviation [SD]=7.4) and Attention Problems scale (T-score=57.1; SD=8.2) (Table 2). Mean raw scores and T-scores for behavior problems scales were both higher for children experiencing dual food and energy hardship and significantly different from the scores of children without dual food and energy hardship.
Table 2:
Mean Child Behavioral Problems Scale Scores at 9 years-old
| Behavior Problem Scalea | All Children | Children with no hardship | Children with dual food and energy hardship | ||||||
|---|---|---|---|---|---|---|---|---|---|
|
|
|||||||||
| Mean Raw score (SD) | Mean T-score (SD) | % with T-scores in Borderline/Clinical Range (T-score ≥ 65) | Mean Raw score (SD) | Mean T-score (SD) | % with T-scores in Borderline/Clinical Range (T-score ≥ 65) | Mean Raw score (SD) | Mean T-score (SD) | % with T-scores in Borderline/Clinical Range (T-score ≥ 65) | |
| Anxious/Depressed†‡ | 2.4 (2.8) | 52.8 (5.4) | 5.0 | 2.2 (2.8) | 52.7 (5.4) | 4.2 | 3.2 (2.9) | 54.6 (5.8) | 8.0 |
| Withdrawn/Depressed†‡ | 1.4 (1.8) | 54.5 (6.0) | 9.1 | 1.2 (1.7) | 54.1 (5.9) | 6.8 | 2.1 (2.2) | 57.1 (7.4) | 20.1 |
| Somatic Complaints†‡ | 1.3 (2.1) | 53.7 (6.0) | 6.4 | 1.2 (2.2) | 53.8 (6.1) | 5.4 | 1.7 (2.1) | 55.4 (6.4) | 10.0 |
| Attention Problems†‡ | 3.6 (3.6) | 54.6 (6.6) | 8.6 | 3.3 (3.5) | 54.3 (6.4) | 7.3 | 4.9 (4.2) | 57.1 (8.2) | 18.2 |
| Social Problems†‡ | 2.1 (2.3) | 53.8 (4.9) | 4.2 | 1.8 (2.3) | 53.1 (4.8) | 3.0 | 2.9 (2.6) | 55.3 (5.7) | 7.8 |
| Thought Problems†‡ | 1.9 (2.6) | 54.2 (6.3) | 9.0 | 1.7 (2.7) | 53.9 (6.3) | 7.5 | 2.5 (2.7) | 56.3 (7.2) | 16.1 |
| Rule Breaking†‡ | 1.8 (2.3) | 53.9 (5.3) | 5.4 | 1.6 (2.3) | 53.5 (5.2) | 3.8 | 2.4 (2.3) | 55.9 (6.4) | 12.9 |
| Aggressive†‡ | 4.5 (5.1) | 54.0 (6.5) | 9.5 | 3.9 (4.7) | 53.5 (6.1) | 6.7 | 6.2 (5.8) | 56.4 (7.9) | 17.2 |
Independent groups t-tests compared significant differences in mean raw score and mean T-score for behavior problem scales between children with dual food and energy hardship and children with no hardship at 9 years-old.
Significant at Bonferroni-corrected p< 0.003 for mean raw scores t-tests
Significant at Bonferroni-corrected p<0.003 for mean T-scores t-tests
After adjustment for covariates, children experiencing dual food and energy hardship had almost 3 times greater odds of withdrawn/depressed behavior (adjusted odds ratio [AOR]= 2.8, 95% CI:1.4–5.5), 3 times greater odds of somatic complaints in the borderline/clinical range (AOR=3.2, 95% CI:1.5–6.9), and almost 4 times greater odds of rule breaking behavior (AOR=3.7, 95% CI:1.5–9.2) compared to children with no hardship (Table 3).
Table 3.
The Adjusted Odds of Exhibiting Behavioral Problems among Children with Dual Food and Energy Hardshipa
| Behavior Problem Scale | Dual Food and Energy Hardship | ||
|---|---|---|---|
| Reference: No hardship | Reference: Energy hardship | Reference: Food Hardship | |
| Adjusted OR (95% CI) | Adjusted OR (95% CI) | Adjusted OR (95% CI) | |
| Anxious/Depressed | 0.7 (0.3–1.8) | 0.5 (0.2–1.4) | 0.5 (0.1–1.6) |
| Withdrawn/Depressed | 2.8 (1.4–5.5) † | 2.0 (0.9–4.2) | 2.9 (1.0–0.8) |
| Somatic Complaints | 3.2 (1.5–6.9) † | 4.2 (1.7–10.0) † | 2.0 (0.7–5.5) |
| Attention Problems | 2.2 (1.1–4.3) | 2.0 (0.9–4.4) | 1.6 (0.6–4.2) |
| Social Problems | 1.6 (0.6–3.8) | 1.7 (0.6–4.6) | 1.0 (0.3–3.0) |
| Thought Problems | 2.0 (1.0–3.9) | 1.7 (0.8–3.6) | 2.2 (0.8–5.9) |
| Rule Breaking | 3.7 (1.5–9.2) † | 2.0 (0.8–5.0) | 2.7 (0.8–9.5) |
| Aggressive | 1.8 (0.9–3.8) | 1.1 (0.5–2.3) | 1.0 (0.8–2.0) |
Models adjusted for covariates of maternal age, race, Latino ethnicity, education level, marital status, employment status, housing tenure, household income-to-poverty ratio, self-reported health status, depressive symptoms, history of ADHD, substance abuse, alcohol abuse, smoking, child sex, and child health insurance status.
Significant at Bonferroni-corrected p<0.006
When compared to children experiencing only energy hardship, children with dual food and energy hardship had 4 times greater odds of somatic complaints in the borderline/clinical range (AOR=4.2, 95% CI:1.7–10.3). There were no significant differences in the odds of behavior problems when children with dual hardship were compared to children with only food hardship or when children with each individual hardship were compared to either those with no hardship or those with only the other individual hardship.
DISCUSSION
In this nationally representative sample, household dual food and energy hardship was associated with internalizing and externalizing behaviors in the elementary school-aged children. Compared to children with no hardship, children experiencing dual food and energy hardship had significantly increased odds of borderline/clinical withdrawn/depressed problems, somatic complaints, and rule-breaking behaviors after controlling for potential confounders. Children with dual food and energy hardship also showed increased odds of somatic complaints compared to those with only energy hardship. Dual hardship is significantly associated with specific problem behaviors compared to either hardship alone or none at all. The increased odds of both internalizing and externalizing behaviors exhibited by children experiencing dual food and energy hardship may lead to poor academic motivation.8,28
Our findings suggest a concurrent association between household food and energy hardship and child behavior problems and builds upon the work of Zilanawala and Pilsaukas,21 who established a link between material hardship at 3 years-old and 5 years-old in the FFCWS cohort and increases in child externalizing and internalizing behaviors at those ages. Our results are also consistent with findings by Frank et al.,10 who described an ordinal association between a cumulative material hardship index and developmental risk for children age 4–36 months, and by Whitaker et al.,6 who described greater odds of child behavior problems with increased levels of maternal food insecurity after controlling for maternal health problems. Our study expands upon this work by a) focusing on the specific associations of food and energy hardship, two common and impactful forms of material hardship in meeting basic household needs and b) assessing behavior in older children from a national cohort and identifying middle childhood as a time for intervention on material hardship to clear a path to better illuminate the root causes of the behavior problems.
Our results also show that children experiencing dual food and energy hardship have a greater odds of borderline/clinical range somatic complaints than children experiencing only energy hardship. Somatic complaints in childhood, apart from anxiety, has been associated with poorer academic performance29 as well as increased risk of depression, anxiety disorders, and panic disorders in adulthood.30
These findings cannot be entirely explained by poverty. More than half of the entire sample reported lower incomes below 200% of the federal poverty level yet only 10% reported dual food and energy hardship. Of those with dual food and energy hardship, almost 10% of mothers reported incomes greater than 200% of the federal poverty level, suggesting that living in or near poverty was not the sole driver of hardship. This suggests that government measures of poverty that use narrow income standards do not account for variations in material hardships and family resources.31 Variations in material hardship may be due to diminished resources related to adult relationship discord, substance and alcohol abuse, smoking, unemployment challenges from the economic downturn, and lower wages from decreased work productivity and disruption secondary to ADHD. Even after controlling for these factors that can fluctuate material hardship and family resources, we still determined that dual food and energy hardship was significantly associated with specific child behavioral problems.
There are several potential mechanisms for poor behavioral adjustment in late elementary school-aged children who experience dual food and energy hardship. Parental stress from juggling multiple economic competing priorities has been associated with parental mental health issues, such as ADHD and depression26, smoking, substance and alcohol abuse, and parenting practice.5,17,32 Similarly, child behavior appears to be correlated with hardship experiences around daily household needs of food and utilities. Children have been found to experience a cognitive and emotional awareness of food insecurity33 and this may extrapolate to an awareness of multiple household material hardship experiences.1 Emotional dysregulation and social dysfunction that may coexist with food and energy hardship experiences can impact school performance as these children can get more easily frustrated, given up on tasks more quickly, and struggle to work collaboratively.2
Our study was limited by several factors. A temporal association between dual food and energy hardship and onset of child behavior problems could not be inferred by our cross-sectional study design. Although we controlled for sociodemographic characteristics and maternal depressive symptoms, smoking, substance and alcohol abuse, and ADHD history in the analyses, there may be unmeasured socio-environmental confounders. Alternatively, the inclusion of several covariates with a priori clinical significance may have caused overadjustment and resulted in bias of some associations toward the null. Food and/or energy hardship may serve as measureable proxies for subtle elements in parent-child and parent-parent relationships and caregiver health, such as maternal negativity and perceived lack of co-parent support, which cause household upheaval and also correlate with child behavior problems.
Food hardship and energy hardship were used to approximate food insecurity and energy insecurity in our sample in the absence of validated food insecurity and energy insecurity screening questions in the FFCWS 9-year interview survey. We were limited by the FFCWS questions to account for food hardship based on potentially more severe cases involving hunger, meal affordability issues, and the receipt of free food. In a national survey of SNAP new-entrant households, 76% of SNAP households that were using food pantries upon enrolling in SNAP were still using food pantries about 6 months later,34 and we were unable to explore this question of adequacy of the SNAP benefit for families with the secondary data analysis. A recent systematic review demonstrated that even marginal levels of household food insecurity were associated with child behavior problems20, and the absence of an association between only food hardship and behavior problems in our study was likely due to limitations of our food hardship measure and small cell size across behavioral domains as the fewest number of mothers experienced only food hardship.
We were also limited to account for energy hardship from questions about inability to afford complete payment of utility bill, which is only one component of energy hardship4 and may not have been applicable to low-income families who do not have to pay for energy as a separate cost. Neither these potentially subtler cases of food hardship and energy hardship, nor a determination of the extent and frequency of hardship into acute versus chronic cases, could be established from the pre-determined FFCWS 9-year survey questions in this secondary data analysis and our cross-sectional design.
Approximately 9% of mothers were unable to be located for 9-year interviews and they may have been experiencing such significant hardship that they did not have a working telephone or stable address. It is also possible that the data as nationally representative of families in large urban cities would not generalize to the experiences of children in more rural parts of the country.
Additionally, the child behavior problem scale scores and food hardship and energy hardship measures were self-reported by the biological mothers and were not confirmed by objective measures, making them subject to potential recall bias and reporter bias. There may have been social desirability bias from maternal perception of their child’s behavior problems as a reflection on their parenting style and practice. This may explain in part the smallest cell sizes for T-scores in the borderline/clinical range for the anxiety/depression scale and a non-significant apparent reduced risk for anxiety/depression with dual food and energy hardship, and additional studies are needed to explore this further.
Notwithstanding these limitations, our study was strengthened by its robust sample size from a national longitudinal birth cohort. We also controlled for maternal depressive symptoms concurrent to the report of hardship to minimize variance from common method bias and reduce measurement error.21 In addition, controlling for maternal substance and alcohol abuse, smoking, and ADHD history in the logistic regression models helped minimize the potential confounding effects of these known factors associated with child behavior problems. The focus on two important dimensions of material hardship, food and energy, as opposed to using one composite or constructed material hardship variable enables us to hone in on specific issues that can be measured by health professionals and addressed. Given the complexity in identifying the root causes of child behavior problems in families with food and energy hardship, the tackling of these issues through referrals and social program supports clears a path to better expose those factors causing the behavior problems.
CONCLUSION
Our study demonstrates a significant association between dual food and energy hardship and concurrent internalizing and externalizing behavior problems in elementary school-aged children. Dual hardship rather than individual hardship is associated with withdrawn/depressed, somatic complaints, and rule breaking behaviors even after controlling for sociodemographic and maternal health and wellness factors. Additional studies to longitudinally examine food and energy hardship across childhood and the onset of behavior problems are needed to determine the extent to which hardships may be causing and aggravating problem behaviors. Further research can also relate the degree of hardship with degree of behavior problems as well as how food and energy hardship, as tangible measurable needs, may serve as proxies for less tangible elements of household upheaval and family dysfunction also associated with child behavior. Future work raises implications for pediatricians to integrate energy hardship screening into social needs screening, many of which already include food insecurity. Screening and intervention opportunities in the outpatient setting can clear a path to identify causes of behavior problems. Pediatricians are uniquely positioned to refer to community organizations for food and energy assistance and to advocate for actionable child health policies around food and energy safety net programs.
What’s New:
Children experiencing dual food and energy hardship have greater odds of internalizing (withdrawn/depressed behavior and somatic complaints) and externalizing behaviors (rule-breaking) compared to those with no hardship and greater odds of somatic complaints compared to those with only energy hardship.
ACKNOWLEDGEMENTS
We would like to thank Dr. Melissa Stockwell and Dr. Julie Lumeng for their valuable comments on previous drafts of this manuscript. Research reported in this publication was supported by the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health (NIH) under Award Number R25DK096944, and the Academic Pediatric Association (APA). This work was also supported in part by the Health Resources and Services Administration Ruth L. Kirchstein National Research Service Award Institutional Research Training Grant in the Department of Medicine at the Columbia University Irving Medical Center under Training Grant T32HP10260. The funding sources had no involvement in study design, data collection, analyses, interpretation, or report construction and submission decisions. The content is solely the responsibility of the authors, and does not necessarily represent the official views of NIH or the APA. The Fragile Families and Child Wellbeing Study was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) of the National Institutes of Health under award numbers R01HD36916, R01HD39135, and R01HD40421, as well as a consortium of private foundations.
Footnotes
Competing Interests Statement: The above authors have indicated that they have no potential conflicts of interest to disclose.
REFERENCES
- 1.Halfon N, Larson K, Son J, et al. Income Inequality and the Differential Effect of Adverse Childhood Experiences in US Children. Acad Pediatr. 2017;17(7S):S70–S78. [DOI] [PubMed] [Google Scholar]
- 2.Eamon MK. The Effects of Poverty on Children’s Socioemotional Development: An Ecological Systems Analysis. Soc Work. 2001;46(3):256–266. [DOI] [PubMed] [Google Scholar]
- 3.Weinreb L, Wehler C, Perloff J, et al. Hunger: its impact on children’s health and mental health. Pediatrics. 2002;110(4):e41. [DOI] [PubMed] [Google Scholar]
- 4.Hernández D Understanding “energy insecurity” and why it matters to health. Soc Sci Med 1982. 2016;167:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Neece CL, Green SA, Baker BL. Parenting Stress and Child Behavior Problems: A Transactional Relationship Across Time. Am J Intellect Dev Disabil. 2012;117(1):48–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Whitaker RC, Phillips SM, Orzol SM. Food Insecurity and the Risks of Depression and Anxiety in Mothers and Behavior Problems in their Preschool-Aged Children. Pediatrics. 2006;118(3):e859–e868. [DOI] [PubMed] [Google Scholar]
- 7.Newland RP, Crnic KA. Developmental Risk and Goodness of Fit in the Mother-Child Relationship: Links to Parenting Stress and Children’s Behaviour Problems. Infant Child Dev. 2017;26(2). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Miles SB, Stipek D. Contemporaneous and longitudinal associations between social behavior and literacy achievement in a sample of low-income elementary school children. Child Dev. 2006;77(1):103–117. [DOI] [PubMed] [Google Scholar]
- 9.Aber JL, Grannis K, Owen S, Sawhill I. Middle Childhood Success and Economic Mobility. Washingon, D.C.: Center on Children and Families - Brookings Institute; 2013. https://www.brookings.edu/research/middle-childhood-success-and-economic-mobility/. Accessed July 8, 2017. [Google Scholar]
- 10.Frank DA, Casey PH, Black MM, et al. Cumulative Hardship and Wellness of Low-Income, Young Children: Multisite Surveillance Study. Pediatrics. 2010;125(5):e1115–e1123. [DOI] [PubMed] [Google Scholar]
- 11.Research Food and Center Action. Food Hardship in America: Households with Children Especially. Washington, D.C.; 2016:15. http://www.frac.org/wp-content/uploads/food-hardship-report-households-with-children-sep-2016.pdf. Accessed July 7, 2017. [Google Scholar]
- 12.<au>Economic Research Service, U.S. Department of Agriculture. Food Security in the U.S. Food & Nutrition Assistance. https://www.ers.usda.gov/topics/food-nutrition-assistance/food-security-in-the-us/. Published September 6, 2017. Accessed November 8, 2017.
- 13.Cook JT, Frank DA, Casey PH, et al. A Brief Indicator of Household Energy Security: Associations With Food Security, Child Health, and Child Development in US Infants and Toddlers. Pediatrics. 2008;122(4):e867–e875. [DOI] [PubMed] [Google Scholar]
- 14.Frank DA, Neault NB, Skalicky A, et al. Heat or Eat: The Low Income Home Energy Assistance Program and Nutritional and Health Risks Among Children Less Than 3 Years of Age. Pediatrics. 2006;118(5):e1293-e1302. [DOI] [PubMed] [Google Scholar]
- 15.Rose-Jacobs R, Black MM, Casey PH, et al. Household Food Insecurity: Associations with At-Risk Infant and Toddler Development. Pediatrics. 2008;121(1):65–72. [DOI] [PubMed] [Google Scholar]
- 16.Jyoti DF, Frongillo EA, Jones SJ. Food insecurity affects school children’s academic performance, weight gain, and social skills. J Nutr. 2005;135(12):2831–2839. [DOI] [PubMed] [Google Scholar]
- 17.Alaimo K, Olson CM, Frongillo EA. Food insufficiency and American school-aged children’s cognitive, academic, and psychosocial development. Pediatrics. 2001;108(1):44–53. [PubMed] [Google Scholar]
- 18.Huang J, Vaughn MG. Household Food Insecurity and Children’s Behaviour Problems: New Evidence from a Trajectories-Based Study. Br J Soc Work. 2016;46(4):993–1008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Frongillo EA, Jyoti DF, Jones SJ. Food Stamp Program participation is associated with better academic learning among school children. J Nutr. 2006;136(4):1077–1080. [DOI] [PubMed] [Google Scholar]
- 20.Shankar P, Chung R, Frank DA. Association of Food Insecurity with Children’s Behavioral, Emotional, and Academic Outcomes: A Systematic Review. J Dev Behav Pediatr. 2017;38(2):135–150. [DOI] [PubMed] [Google Scholar]
- 21.Zilanawala A, Pilkauskas NV. Material hardship and child socioemotional behaviors: Differences by types of hardship, timing, and duration. Child Youth Serv Rev. 2012;34(4):814–825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Reichman NE, Teitler JO, Garfinkel I, McLanahan SS. Fragile Families: sample and design. Child Youth Serv Rev. 2001;23(4):303–326. [Google Scholar]
- 23.Achenbach TM, Rescorla LA. Manual for the ASEBA School-Age Forms & Profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, and Families; 2001. [Google Scholar]
- 24.Hager ER, Quigg AM, Black MM, 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] [PubMed] [Google Scholar]
- 25.Urke HB, Cao ZR, Egeland GM. Validity of a single item food security questionnaire in Arctic Canada. Pediatrics. 2014;133(6):e1616–1623. [DOI] [PubMed] [Google Scholar]
- 26.Williams DT, Cheadle JE. Economic Hardship, Parents’ Depression, and Relationship Distress among Couples With Young Children. Soc Ment Health. 2016;6(2):73–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kessler RC, Andrews G, Mroczek D, Ustun B, Wittchen H-U. The World Health Organization Composite International Diagnostic Interview short-form (CIDI-SF). Int J Methods Psychiatr Res. 1998;7(4):171–185. [Google Scholar]
- 28.Breslau J, Miller E, Breslau N, Bohnert K, Lucia V, Schweitzer J. The impact of early behavior disturbances on academic achievement in high school. Pediatrics. 2009;123(6):1472–1476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hughes AA, Lourea-Waddell B, Kendall PC. Somatic Complaints in Children with Anxiety Disorders and their Unique Prediction of Poorer Academic Performance. Child Psychiatry Hum Dev. 2008;39(2):211–220. [DOI] [PubMed] [Google Scholar]
- 30.Shanahan L, Zucker N, Copeland WE, Bondy C, Egger HL, Costello EJ. Childhood Somatic Complaints Predict Generalized Anxiety and Depressive Disorders During Adulthood in a Community Sample. Psychol Med. 2015;45(8):1721–1730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Cauthen NK, Fass S. Measuring poverty in the United States. June 2008. http://www.nccp.org/publications/pub_825.html. Accessed May 7, 2016.
- 32.Mistry RS, Vandewater EA, Huston AC, McLoyd VC. Economic well-being and children’s social adjustment: The role of family process in an ethnically diverse low-income sample. Child Dev. 2002;73(3):935–951. [DOI] [PubMed] [Google Scholar]
- 33.Fram MS, Frongillo EA, Jones SJ, et al. Children Are Aware of Food Insecurity and Take Responsibility for Managing Food Resources. J Nutr. 2011;141(6):1114–1119. [DOI] [PubMed] [Google Scholar]
- 34.Mabli J, Worthington J. Supplemental Nutrition Assistance Program Participation and Emergency Food Pantry Use. J Nutr Educ Behav. 2017;49(8):647–656.e1. [DOI] [PubMed] [Google Scholar]