The U.S. Household Food Security Scale, developed with federal support for use in national surveys, is an effective research tool. This study uses these new measures to examine associations between food insecurity and health outcomes in young children. The purpose of this study was to determine whether household food insecurity is associated with adverse health outcomes in a sentinel population ages ≤ 36 mo. We conducted a multisite retrospective cohort study with cross-sectional surveys at urban medical centers in 5 states and Washington DC, August 1998–December 2001. Caregivers of 11,539 children ages ≤ 36 mo were interviewed at hospital clinics and emergency departments (ED) in central cities. Outcome measures included child's health status, hospitalization history, whether child was admitted to hospital on day of ED visit (for subsample interviewed in EDs), and a composite growth-risk variable. In this sample, 21.4% of households were food insecure (6.8% with hunger). In a logistic regression, after adjusting for confounders, food-insecure children had odds of “fair or poor” health nearly twice as great [adjusted odds ratio (AOR) = 1.90, 95% CI = 1.66–2.18], and odds of being hospitalized since birth almost a third larger (AOR = 1.31, 95% CI = 1.16–1.48) than food-secure children. A dose-response relation appeared between fair/poor health status and severity of food insecurity. Effect modification occurred between Food Stamps and food insecurity; Food Stamps attenuated (but did not eliminate) associations between food insecurity and fair/poor health. Food insecurity is associated with health problems for young, low-income children. Ensuring food security may reduce health problems, including the need for hospitalizations.
food security, food insecurity, hunger, child health, children
Food security is defined as access at all times to enough nutritious food for an active and healthy life, whereas food insecurity is defined as limited or uncertain access to enough nutritious food (1–4). Although lack of access to enough nutritious food can occur for a variety of reasons, the U.S. Household Food Security Scale (U.S. HFSS)4 was designed to identify food insecurity arising specifically from the lack of adequate financial resources to purchase enough food. This kind of food insecurity is sometimes called “resource-constrained” or “poverty-linked” food insecurity, although some households with incomes above the poverty threshold experience it (1–3).5
As defined, food insecurity at its least severe levels does not necessarily involve reductions in the quantity of food intake below normal levels, but is evident in adult respondents' concerns about the sufficiency of their household food supply and adjustments to household food management, including reductions in diet quality and variety. At moderately severe levels of food insecurity, food intake for adults in the household is reduced below normal levels by reducing meal or serving sizes or skipping meals, sometimes leading to hunger. At more severe levels, households with children also reduce the children's food intake to an extent that the children experience hunger as a result of inadequate household resources, whereas adults in households with or without children experience even more extensive reductions in food intake, possibly going whole days without food (3). As these definitions imply, hunger and undernutrition may occur as a result of food insecurity, depending on its severity and duration (2). Moreover, recent research suggests that food insecurity may exacerbate the onset or persistence of other adverse health conditions, including overweight and obesity among some subpopulations (5–11).
Household food insecurity is a concern to pediatricians because it has implications for child health in several ways. Earlier versions of food security measures similar to the 18-item U.S. Food Security Scale were associated with inadequate intakes of several important nutrients (10–15), cognitive developmental deficits (16–23), behavioral and psychosocial dysfunction in children and adults (16,24–27), and poor health in children and adults (11,12,28–30). The association of micronutrient and protein-energy deficits with impaired immunity and wound healing and thus with increased risk of serious illness is also well established (29–37). Recent research also suggests that affective or psychologic stresses such as those accompanying resource-constrained food insecurity can influence child health and well-being adversely, independent of associated nutritional deficits (23–25,29). Not being able to purchase enough nutritious food, and the resultant emotional or psychologic stresses arising in the household, can contribute to adverse health effects or exacerbate poor health caused by other factors, including malnutrition (38–43).
Young, low-income children in households utilizing urban medical centers represent a sentinel population at high risk of adverse health outcomes, and may exhibit health effects of food insecurity at levels of clinical severity or at prevalence rates that are not noted among children in the general population (44–51). This study evaluates whether in inner-city settings, young children in households exposed to food insecurity have significantly different odds of experiencing negative health outcomes than similar children in food-secure families.
SUBJECTS AND METHODS
Setting and instruments.
The Children's Sentinel Nutrition Assessment Project (C-SNAP) conducted household-level surveys and medical record audits between August 1998 and December 2001 at central-city medical centers in Baltimore, Boston, Little Rock, Los Angeles, Minneapolis, and Washington, D.C. A convenience sample comprising adult caregivers accompanying 11,539 children age ≤ 36 mo at acute- and primary-care clinics and hospital emergency departments (ED) was interviewed in private settings by trained interviewers scheduled during peak patient-flow times. At 3 sites (Boston, Little Rock, and Los Angeles, n = 6502), interviews were conducted in the hospital ED. Caregivers of critically ill or injured children at any site were not approached. Potential respondents were excluded if they did not speak English, Spanish, or Somali (Minneapolis only), were not knowledgeable about the child's household, the child's caregiver had been interviewed within the previous 6 mo, or they refused consent for any reason.
The survey instrument included questions on household characteristics, food security, federal assistance program participation, changes in benefits, child's health status, and child's hospitalization history. Household food security status was derived from responses to the U.S. Food Security Scale in accordance with established procedures (3,4). The questionnaire contains a combination of items drawn from other validated survey instruments (developed by C-SNAP researchers or others) and items specific to C-SNAP study goals and objectives. Where possible, we used wording from existing surveys that had been validated. The core set of 18 food-security items were taken from the U.S. HFSS, and scored and categorized in accordance with established procedures (4). The survey instrument and both the surveillance and interview protocols were pilot-tested at Boston Medical Center on several hundred subjects over 1996–1997. The instrument was revised as necessary before being distributed to all 6 C-SNAP sites for implementation in 1998. Slight modifications were made since 1998, but these have been mainly to improve skip patterns or to clarify aspects of a few questions.
Additional information was obtained from medical record audits of all children whose caregivers were interviewed. These data include anthropometric measures (height and weight) and, for the subsample of children interviewed in the ED, whether the child was admitted to the hospital on the day of the visit. Institutional Review Board approval was obtained at each of the 6 C-SNAP sites through application to the parent institution's IRB.
Sample characteristics.
The analytic cohort (Tables 1 and 2) comprised 11,539 children whose adult caregivers were interviewed at the 6 C-SNAP sites. These children comprise 78% of a larger pool of potential participants approached at the 6 study sites. Of the total approached, 22% did not respond, 7% refused the interview, and an additional 15% were ineligible due to language, lack of knowledge of the child's household, or having been interviewed previously.
TABLE 1
Characteristics of caregivers in the analytic cohort by exposure to variation in household food security status, 1998–20021, 2, 3
Caregiver characteristics n % Food insecure
Study site*
Baltimore 1017 14.8
Boston 3102 19.8
D.C. 725 35.0
Little Rock 1556 8.6
Los Angeles 1844 20.1
Minneapolis 3295 28.6
Subtotal 11,539 21.4
Race/Ethnicity*
African American 5886 17.4
Hispanic 4052 31.2
Caucasian 1272 10.2
Other 326 13.5
Born in the United States
Yes* 6801 13.7
No 4713 32.4
Marital status
Single* 6082 21.1
Married/Partner 5420 21.6
Age
<21 y* 2177 15.0
≥21 y 9259 22.8
Schooling
<Grade 12* 4474 28.3
≥Grade 12 7015 16.9
Employed
Yes* 4710 16.6
No 6692 24.6
Receive SSI
Yes* 777 25.5
No 10,685 21.0
Subsidized housing
Yes 2442 21.4
No 8910 21.4
Live in temporary housing
Yes* 3126 26.3
No 8413 19.5
Receive WIC
Yes* 9085 22.6
No 2389 16.5
Receive Food Stamps*
Currently 3718 24.9
Previously 1632 23.8
Never 6089 18.5
Receive TANF*
Currently 3136 24.9
Previously 1820 23.6
Never 6528 19.0
1
Group comparisons used χ2 tests.
2
Asterisks indicate statistical significance, * P ≤ 0.05.
3
Totals may differ due to missing data on some variables.
View Large
TABLE 2
Characteristics of children in the analytic cohort by exposure to variation in HFSS, 1998–20021, 2
Child characteristics n % Food insecure
Age
≤1 y 6595 21.8
>1 to ≤2 y 3051 20.6
>2 to ≤3 y 1890 20.9
Birth weight
<2500 g 1434 22.9
≥2500 g 9763 21.0
Child's insurance status**
Public 8693 23.1
Private 1202 9.7
None 1547 20.8
In daycare**
Yes 3757 17.0
No 7696 23.3
Weight-for-age Z-score
(Mean, 95% CI) −0.006 (−0.033, 0.021) 0.031 (−0.023, 0.085)
(SD, 95% CI) 1.29 (1.27, 1.31) 1.34 (1.31, 1.39)