1. Introduction
The U.S. Department of Agriculture (USDA)’s National School Lunch Program (NSLP) and School Breakfast Program (SBP) are integral parts of the U.S. federal government’s food assistance and nutrition safety net. Through these programs, participating schools serve lunches and breakfasts meeting federally established nutrition standards and receive reimbursements for meal costs from USDA that vary based on the household income of participating children. Children from families with incomes at or below 130% of federal poverty guidelines can receive meals for free; those whose family income is between 130% and 185% of poverty pay a reduced price [
1]. (In 2015, the federal poverty guideline was an annual income of
$24,250 for a family of four [
2].) Children are certified to receive FRP school meals by either submitting an application or being directly certified because their households participate in other federal assistance programs [
1]. USDA provides higher levels of reimbursements for free and reduced-price NSLP and SBP meals. Children from higher-income families pay what is called “full price”, although USDA still partially subsidizes their meals. Regulations require that all children receive the same meal and differences in payment status should not be identifiable.
In 2019, the NSLP served 29.4 million children on a typical school day, and the SBP served close to 15 million children [
3]. Almost all public schools participate in the programs, and meals are available to all attending students. However, 74% of lunches and 85% of breakfasts were served free or reduced-price (FRP) to low-income students in 2019.
In 2010, Congress passed the Healthy, Hunger-Free Kids Act (HHFKA) [
4]. It mandated updated nutrition standards in line with federal dietary guidance, including age-based calorie minimums and maximums; required inclusion of whole grain-rich foods; a wider range of vegetables, with dark green and red/orange vegetables and legumes required to be served weekly; more whole fruit; and low or nonfat milk as the standard milk offered. Implementation of menu changes began in the 2012–2013 school year (SY) [
5].
The HHFKA also established a new policy, the Community Eligibility Provision (CEP), under which schools serving primarily low-income children could offer meals to all students at no cost—a policy referred to as “universal free meals”. Previously, some schools offered universal free meals through other USDA provisions, but the CEP considerably expanded the number of schools offering them. The CEP required participating schools to offer the SBP, increasing availability of school breakfasts. The CEP option began as a pilot in selected states and was implemented in all states by SY 2014–2015 [
6]. Finally, the HHFKA established availability of afterschool suppers as part of afterschool programs in low-income areas.
These changes could be expected to have important impacts on the diets of children living in households that struggle with food insecurity. Food security is defined as access by all people at all times to enough food for an active, healthy life [
7]. Households unable to acquire adequate food for one or more household members because of insufficient money and other resources for food are considered food insecure [
8]. In 2019, 14% of households with children were classified as food insecure (adults, children, or both were food insecure) at some time during the year [
7]. Parents and other household adults typically try to shield children from experiencing food insecurity; nevertheless, children were reported to be food insecure in 7% of households with children [
7].
Food insecurity among children is associated with many adverse impacts on physical and mental health, as well as negative social and academic outcomes [
8,
9]. Children living in food-insecure households have been reported to have poorer overall health, have increased frequency of chronic health conditions and more hospitalizations, and be more likely to be iron-deficient, and male children have been found to have lower bone density, and they are more likely to experience psychological distress, including higher rates of depression and anxiety and, among adolescents, suicidal ideation [
8,
9]. Behavior problems and academic challenges, including slower progress in math and reading and higher likelihood of repeating a grade among elementary-age children, are also associated with food insecurity [
8,
9].
Previous research has shown school meals to be important resources for children in food-insecure households. In an analysis of data from the USDA’s School Nutrition Dietary Assessment III (SNDA-III), conducted in SY 2004–2005, Potamites and Gordon found that children from food-insecure households were more likely to participate in school meal programs and obtained a larger share of their food and nutrient intakes from school meals than other children [
10]. Using national food consumption survey data, Smith found that school meals provided the highest nutritional quality contribution to the diets of children from food-insecure households compared to food obtained at home or elsewhere, with quality measured by the USDA’s Healthy Eating Index (HEI) [
11]. Both studies were completed before the HHFKA-driven changes in school meals. Post HHFKA, substituting school meals for meals from home was associated with higher nutritional quality for students across the income distribution [
12]. The more widespread availability of school breakfasts and afterschool suppers may have increased the importance of school food in the diets of food-insecure children, whereas changes in the nutrition standards may have impacted its contribution to intakes of nutrients and important food groups for all students.
The objective of this paper is to examine the association between food security status, school meal participation, and students’ diet using data from the first School Nutrition and Meal Cost Study (SNMCS-I), a comprehensive, national data set collected following the implementation of HHFKA-mandated changes. SNMCS-I is unique in that it links student-reported dietary intakes at school with school-reported menus to identify the contributions of school meals to students’ diets. We hypothesized that, compared to food secure (FS) students, food insecure (FI) students and, to a lesser extent, marginally secure (MS) students make more use of the USDA school meal programs, and that these programs contribute more to their diet in terms of both quantity and quality, as assessed by calories and HEI scores, respectively.
4. Discussion
This study found that FI and MS students were more likely to attend schools that participated in SBP, offered universal free meals, and offered reimbursable afterschool snacks or suppers, which suggests that the efforts to encourage participation in healthy school meals appropriately focused on students with greater need. Although FI and MS students were more likely to participate in school meals, which is consistent with the first hypothesis, participation was not universal; only 38% of FI and 33% of MS students in SBP-participating schools participated in SBP. This is scarcely different from participation levels in SNDA-III (37% and 26% for FI and MS students in SBP-participating schools, respectively), despite a 10 percent increase in SBP availability in the decade between studies [
10]. The availability of SBP therefore does not seem to be sufficient to increase participation, although causality cannot be inferred. Instead, SBP participation was significantly higher in schools offering universal free meals or afterschool snacks or suppers. These policy options may decrease any stigma associated with participation or may be an indicator of an overall school culture that encourages participation. Strategies to make school breakfast more convenient or appealing to children, such as serving breakfast in the classroom at the beginning of the school day, rather than in the cafeteria before school starts, or offering “second-chance breakfasts” at mid-morning to late-arriving students have also been suggested to help increase participation [
22,
23]; further exploration of factors influencing SBP participation, particularly in schools with large proportions of students vulnerable to food insecurity, may provide more evidence of their effectiveness and identify other factors influencing participation.
School meals contributed significantly more to FI and MS students’ energy intakes, which reflects their higher rates of participation and is consistent with the hypothesis that school meal programs are an important contributor to diet quantity (that is, calories). Among participants, differences across food security groups were not statistically significant. In SNDA-III, school meals contributed significantly more energy to FI and MS participants’ diets than to FS participants’ diets; school meals made up 32% of the daily energy intake of FS participants, 30% of MS participants’ intake, and 27% of FS participants’ intake [
10]. Differences between the studies may reflect the calorie maximums in the updated nutrition standards. However, the two studies used different methods to identify school meal participants. SNDA-III identification was based on reported intakes rather than administrative data, although the methods were compared using SNMCS-I data and only small differences in classification were found [
14]. One might expect to see greater differences in the contributions of school meals to students’ energy intakes in SNMCS-I because participation was not based on whether any school meal foods were consumed, and FI and MS students may be more likely to consume those foods if their access to food outside of school is limited. It is unclear why differences were observed in SNDA-III but not SNMCS-I.
Analysis of the USDA’s National Food Acquisition and Purchasing Study (FoodAPS) found that FI households buy less food than other households of equivalent composition and, in particular, acquired less total fruit, whole fruit, total protein, and seafood and plant proteins compared to food-secure households, on a per-1000-calorie basis [
24]. When analyzing FoodAPS data, Kuhn found that school meal programs played an important role in reducing “consumption crunches” (monthly cycles of reduced energy intakes associated with food assistance benefits running out) in low-income families with children during periods of reduced food expenditures [
25]. This effect benefited both children and adults in the households, probably because the children’s ability to acquire food at school reduced strain on overall household food resources. Our finding that FI and MS children were more likely to participate in school meals and therefore acquire a larger share of their calories from these programs is consistent with Kuhn’s findings, and it indicates the important role these programs can play in supporting dietary adequacy for children in FI households.
Using data from the National Health and Nutrition Examination Surveys (NHANES) before implementation of the updated nutrition standards, Smith found that consumption of foods obtained at school improved the quality of diets of low-income and FI students as measured by the HEI, but it did not improve diet quality for higher-income children [
11]. However, a similar analysis conducted with NHANES data collected postimplementation found improvements in students’ diet quality across all income groups [
12].
Our findings are largely consistent with the post-HHFKA analysis, and because of our detailed information on school meal foods, they can provide some additional insights. Nonschool meal foods eaten by students in all food security groups scored well below the optimal score for most components, and the scores were similar across the groups with only two exceptions—dairy and fatty acids. The lower dairy scores for nonschool foods among FI students suggest school meal foods are particularly important for calcium and other essential nutrients contributed by dairy; this is consistent with findings from SNDA-III that school meal participants from FI households obtained 47% of their calcium intake from school meals, compared to 38% of intake for participants from FS households [
10]. Although fatty acid scores from non-school foods were higher among FI students, all groups scored less than 50% for this component, suggesting improvement is needed by all.
School foods scored at or close to 100% for five HEI components, including fruit, dairy, and whole grains. For some HEI components, the quality gap between school and nonschool foods was particularly large—for example, nonschool food scores averaged 31% of the HEI score for whole grains, whereas school foods scored 100%. For each student group, the average HEI scores for nonschool foods were between 55% and 57%, whereas school foods scored between 79% and 81%. These findings are consistent with the hypothesis that school meals contribute to students’ dietary quality, but differences across food security groups were not observed. Disappointingly, school food consumption scores for total vegetables, and especially dark green vegetables and beans, were below 50% for all student groups, despite SNMCS-I menu data indicating that meals provided to students scored much higher on these components [
26]. Low scores for these foods indicate a challenge for creating acceptance of these foods. Scores for sodium were also low, less than 50%; in this case, menu data indicate some schools found it challenging to meet sodium targets [
26]. The USDA has worked to assist SFAs with this challenge by developing low-sodium USDA Foods for schools; other strategies the USDA has identified as effective include using taste tests to identify student preferences, tailoring menu options to cultural and local preferences, and nutrition promotion [
27].
A few limitations must be noted. First, students were categorized using the adult measure of food security. Because students’ status may differ from that of the adults in the household, the adult measure may be less sensitive for detecting the contribution of school meals to students’ diets across food security groups. Second, children and adolescents, like adults, may misreport their dietary intakes, either intentionally or because they are still undergoing cognitive development [
28]. Several dietary recall validation studies have found children to accurately or slightly overreport their energy intakes, and younger children may report more accurately with parental assistance, but overweight or obese children and older children are more likely to underreport [
28,
29]. Third, some analyses were based on small sample sizes and may not be reliable; for example, few students attended schools that offered universal free meals. Finally, SNMCS-I was a descriptive study that was not designed to test causality of any hypotheses and the analysis in this paper does not attempt to account for any possible confounders or adjust for characteristics that may be associated with misreporting. The relationships among student, household, or school characteristics, school meal participation, and dietary intakes should not be interpreted as causal.
Implications for Practice and Future Research
U.S. federal dietary guidance emphasizes the importance of a healthy diet in childhood for growth, development, learning, and establishment of healthy eating habits that may reduce the risk of health problems later in life [
30]. For children from FI households, USDA school meals may be particularly important. Our findings support those of previous research indicating that USDA school meals can be important contributors to the diets of children in FI households, augmenting household food resources [
8,
10,
11]. Participation in school meals may also increase acceptance of healthier foods outside of school, similar to what has been found for another USDA-sponsored, school-based child nutrition program [
31]. Nevertheless, some children in these households do not participate in school meals, especially SBP. Providing universal free meals was associated with higher participation. The CEP provides a new mechanism for schools serving primarily low-income children to offer universal free meals, and its adoption has grown from 14,214 schools in SY 2014–2015, its first year of national availability, to 30,667 in SY 2019–2020 [
32]. Future research will provide more information on the association of CEP adoption with SBP participation among children from FI households. Research on other strategies to promote SBP participation, such as serving it in the classroom at the start of the school day rather than in the cafeteria before school, may also be helpful to school nutrition professionals interested in promoting SBP participation, and research may also shed light on strategies for motivating healthier eating outside of school.
It is notable that school foods promote diet quality among all students, whether food secure or food insecure. Emphasizing this finding may encourage participation among all students, creating a norm that would reduce any perceived stigma among FI students.