Skip to main content
NCBI home page
JAMA Network logoLink to JAMA Network
. 2022 May 5;5(5):e2210480. doi: 10.1001/jamanetworkopen.2022.10480

Association of the Healthy, Hunger-Free Kids Act of 2010 With Body Mass Trajectories of Children in Low-Income Families

Andrea S Richardson 1,✉, Margaret M Weden 2, Irineo Cabreros 3, Ashlesha Datar 4
PMCID: PMC9073566  PMID: 35511177

Key Points

Question

Has the Healthy, Hunger-Free Kids Act of 2010 (HHFKA) improved body mass trajectories of children in low-income families?

Findings

In this cohort study evaluating 3388 children before HHFKA implementation and 2570 children after HHFKA implementation, kindergarteners in low-income families who did not participate in the free or reduced-price National School Lunch Program before the HHFKA had steeper decreases in body mass index (BMI) difference from obesity threshold between grades 1 and 5 than their participating peers. After HHFKA implementation, this obesogenic association between the National School Lunch Program and BMI difference change was no longer observed.

Meaning

These findings suggest that implementation of the HHFKA may have helped curb obesogenic increase in BMI among low-income children.

Abstract

Importance

Implemented in 2012, the Healthy, Hunger-Free Kids Act of 2010 (HHFKA) increased nutritional requirements of the National School Lunch Program (NSLP) to reverse the potential role of the NSLP in childhood obesity.

Objective

To evaluate whether associations between the free or reduced-price NSLP and body mass growth differed after implementation of the HHFKA.

Design, Setting, and Participants

This cohort study used data from 2 nationally representative cohorts of US kindergarteners sampled in 1998 to 1999 and 2010 to 2011 and followed up for 6 years, through grade 5, in the Early Childhood Longitudinal Study Kindergarten Class of 1998-1999 (ECLS-K:1999, in 2003-2004) and Kindergarten Class of 2010-2011 (ECLS-K:2011, in 2015-2016). In total, 5958 children were selected for analysis from low-income families eligible for the free or reduced-price NSLP (household income <185% of the federal poverty level) who attended public schools and had no missing data on free or reduced-price NSLP participation or on body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) at kindergarten or grades 1 and 5. Data were analyzed from January 1 to September 7, 2021.

Exposures

Cross-cohort comparison of before vs after implementation of the HHFKA for free or reduced-price NSLP participation at kindergarten and grades 1 and 5.

Main Outcomes and Measures

Body mass index difference (BMID) from obesity threshold was the difference in BMI units from the age- and sex-specific obesity thresholds (95th percentile) and is sensitive to change at high BMI. Multigroup models by cohort included weights to balance the distribution of the 2 cohorts across a wide range of covariates. A Wald test was used to assess whether associations differed between the cohorts.

Results

In the final analysis, 3388 children in ECLS-K:1999 (1696 girls [50.1%]; mean [SD] age at baseline, 74.6 [4.3] months) and 2570 children in ECLS-K:2011 (1348 males [52.5%]; mean [SD] age at baseline, 73.6 [4.2] months) were included. The best fitting model for BMID change by free or reduced-price NSLP participation across the cohorts included fixed and time-varying associations. Before HHFKA implementation, grade 5 free or reduced-price NSLP participants had higher BMID, adjusted for their prior BMID trajectory, than nonparticipants (β = 0.54; 95% CI, 0.27-0.81). After HHFKA implementation, this association was attenuated (β = −0.07; 95% CI, −0.58 to 0.45), and grade 5 associations were different across cohorts (χ21 = 4.29, P = .04).

Conclusions and Relevance

In this cohort study using cross-cohort comparisons, children from low-income families who participated in the free or reduced-price NSLP had a higher likelihood of progression to high BMI that was no longer observed after HHFKA implementation. This finding suggests that the HHFKA may have attenuated the previous association of the NSLP with child obesity disparities.

This cohort study uses data from 2 nationally representative cohorts of US kindergarteners followed up through grade 5 to assess whether body mass growth differed after implementation of the Healthy, Hunger-Free Kids Act of 2010 among children in low-income families.

Introduction

Obesity remains high among children,1 and severe obesity is climbing,2 especially in low-income populations.3,4 Although school meals are a critical source of nutrients for low-income children,5 growing evidence suggests that prior to 2010, school meals may have contributed to childhood obesity.6,7,8,9,10

Congress passed the Healthy, Hunger-Free Kids Act of 2010 (HHFKA)11 to reduce childhood obesity by increasing the nutritional requirements in school meals. Since its implementation in 2012, growing evidence suggests that the HHFKA improved children’s dietary quality12,13,14,15,16,17,18; however, to our knowledge, only 2 observational studies have estimated associations with child growth, and those had mixed findings.19,20 An interrupted time series of National Survey of Children’s Health (2003-2018) observed decreased obesity trends after HHFKA implementation among impoverished children.20 In a multivariate regression analysis assessing data before and after HHFKA implementation in the Early Childhood Longitudinal Study (ECLS) of children in grades 1 and 3, the HHFKA was associated with slower body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) z score growth in grade 3 but only for boys.19 However, those studies’ findings may not be robust to problems of selection bias in observational data. Several studies before HHFKA implementation have shown that ignoring selection into the National School Lunch Program (NSLP) biases associations downward between the NSLP and child weight.7,8,9,21 The 2 studies that examined changes in child growth before and after HHFKA implementation used BMI z score19 and obesity20 outcomes, which raises concerns about their findings given BMI z score limitations for capturing extreme values22,23,24,25,26 and severe obesity increases,2,27,28 especially in low-income populations.29 We evaluated 2 national cohorts of children followed throughout elementary school to examine whether associations between free or reduced-price NSLP participation and body mass growth of children from low-income families changed following HHFKA implementation. We followed the Centers for Disease Control and Prevention guidelines30; thus, our analysis is sensitive to extreme BMI values that are due to severe obesity increases.2,27,28,29 We account for body mass tracking over time,31,32,33,34 and we model differences within child and between children to reduce selection bias.

Methods

Participants

We used data from cohorts of nationally representative samples of US kindergarteners in ECLS. The ECLS Kindergarten Class of 1998-1999 (ECLS-K:1999) recruited 21 409 kindergartens in fall of 1998 from 1000 public and private schools, and the ECLS-K:2011 recruited 18 174 kindergarteners in fall of 2010 from 970 public and private schools across the US. Both cohort studies followed up children through grade 5. The ECLS obtained parent consent before data collection. Our analyses are based on data collected in kindergarten and grades 1 and 5, when both cohort studies surveyed parents about their child’s NSLP participation. The present study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies. The RAND Human Subjects Protection Committee approved this study and waived the need for informed consent per the Common Rule.

Of 21 409 children (ECLS-K:1999) and 18 174 children (ECLS-K:2011) recruited for the studies, children were ineligible for the present study if they did not participate in any of the waves (kindergarten and grades 1 and 5) (9839 children in ECLS-K:1999 and 9962 children in ECLS-K:2011), if they ever attended private school (2467 in ECLS-K:1999 and 1089 children in ECLS-K:2011), if they never qualified for the free or reduced-price NSLP (ie, income never <185% of the federal poverty level at kindergarten, and grades 1 and 5)35 (3895 children in ECLS-K:1999 and 3393 children in ECLS-K:2011), or if they had missing household income at kindergarten or grades 1 and 5 (40 children in ECLS-K:1999 and 7 children in ECLS-K:2011) (Figure 1). Of 5168 eligible children in ELCSK:1999 and 3723 eligible children in ECLS-K:2011, we excluded those with missing NSLP data (1197 children [23.2%] in ECLS-K:1999 and 685 children [18.4%] in ECLS-K:2011) or BMI score (83 children in kindergarten [1.6%], 272 children in grade 1 [5.3%], and 228 children in grade 5 [4.4%] in ECLS-K:1999; and 102 children in kindergarten [2.7%], 73 children in grade 1 [2.0%], and 293 children in grade 5 [7.9%] in ECLS-K:2011). The study design and children’s changing schools reduced grade 1 and 5 samples, explaining most nonresponses.36,37 Compared with the analytic sample, eligible children excluded for missing BMI data differed significantly from children not excluded with respect to race and ethnicity, urbanicity, family dinners (ECLS-K:1999 only), and mother's educational level (ECLS-K:1999 only), but there were no significant differences in free or reduced-price NSLP participation, household income, child's birth weight, or mother's work status (eTable 1 in the Supplement). However, differences may be spurious owing to multiple testing. Compared with the analytic sample, BMI at kindergarten was similar for excluded children (500 children in ECLS-K:1999: difference in means, 0.1 (95% CI, −0.1 to 0.4); 366 children in ECLS-K:2011: difference in means, −0.1 (95% CI, −0.4 to 0.2).

Figure 1. Flow of Analytic Samples in Early Childhood Longitudinal Study Kindergarten Class of 1998-1999 (ECLS-K:1999) and Kindergarten Class of 2010-2011 (ECLS-K:2011) Cohorts.

Figure 1.

Open in a new tab

NSLP represents free or reduced-price National School Lunch Program.

Outcomes

During the spring of each wave, trained assessors measured children’s height (to the nearest quarter inch using a ShorrBoard; Weigh and Measure LLC) and weight (to the nearest half-pound using a Seca digital bathroom scale). The BMI percentiles and z scores have not matched empirical data for children with high BMI.22,23,24,25,26 Consequently, high BMIs are conflated to narrow distributions of BMI z scores and percentiles. Thus, BMI measures relative to the 95th percentile are better than BMI z scores or percentiles at identifying children with severe obesity.25 We used the SAS macro from the Centers for Disease Control and Prevention to calculate the difference between the child’s BMI and the 95th percentile BMI value (BMID) for sex and age.30 We describe BMI z scores and weight status using BMI percent of the 95th percentile (BMIP) for underweight (<5.3 BMIP), normal weight (5.3 to <89 BMIP), overweight (89 to <100 BMIP), obesity (100 to <120 BMIP), and severe obesity (≥120 BMIP).

Exposures

During fall of kindergarten and grades 1 and 5, the ECLS-K:1999 parent responded to 2 questions: “Does (CHILD) usually receive a complete lunch offered at school? By complete school lunch, I mean a complete meal such as a salad, soup, a sandwich, or a hot meal that is offered each day at a fixed price, not just milk, snacks, ice cream, or a lunch (he/she) brought from home”; and “Are these lunches free or reduced price?” Although ECLS-K:2011 parent surveys at kindergarten and grade 1 included the same questions, at grade 5 the parents were asked “Does (CHILD) receive complete school lunches for free or reduced price at school?” For consistency, we used free or reduced-price lunch participation to classify NSLP participation. The NSLP lunches were the same regardless of the price, yet school lunches may provide low-income children a greater proportion of their daily nutritional intake because low-income households may have lower nutritional quality than school lunches.38 We hypothesize that the association of the HHFKA with children’s dietary quality may be greater for children in low-income families than their more advantaged peers.

Covariates

Unless specified below, ECLS-K:1999 and ECLS-K:2011 used the same survey instruments.39,40 During the fall kindergarten wave, guardians reported time-invariant characteristics: child’s sex, race and ethnicity (Black non-Hispanic, Hispanic, White non-Hispanic, and other [American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, or more than 1 race and ethnicity]), birthweight, and mother’s highest level of education. Reported time-varying characteristics included child’s age, parent’s employment status in the last week (eg, “work at a job for pay”), whether “in a typical week” the family “eats the evening meal together,” number of hours a day the child “usually watch[es] TV or videos on school days” (not asked in ECLS-K:2011 for grade 5), and last year’s household income. We chose these covariates because they could be associated with unobserved factors that influence free or reduced-price NSLP participation and family norms that may increase obesity risk, such as watching television and eating habits.9,41,42,43 School urbanicity was linked to the cohort by the National Center for Education Statistics prior to releasing the data set.39

Statistical Analysis

Descriptive statistics were obtained with Stata, version 15.1 (StataCorp LLC). The HHFKA was first implemented in grade 2 of ECLS-K:2011, which enabled us to examine how the HHFKA may have changed associations between free or reduced-price NSLP and BMID trajectories. We made 3 comparisons to reduce bias: (1) within-child change in BMID; (2) differences in free or reduced-price NSLP associations with BMID change (grades 1 vs 5); and (3) differences between cohorts. The historical cohort enabled us to observe how changes in BMID over time between free or reduced-price NSLP and non-NSLP participants evolved before the HHFKA.

We used a general panel approach with structural equation modeling44 and tested which model best fit the data. We tested a fixed-effects model, a random-effects model, and a model with time varying and time-invariant coefficients, with or without lagged outcomes. The best fitting model for BMID change by free or reduced-price NSLP participation across the cohorts included fixed and time-varying associations (eTable 2 and eTable 3 in the Supplement). This allowed for associations between BMID and time-varying characteristics to change as children aged, adjusted for BMID in the previous waves and adjusted for time-invariant child characteristics (eg, birth weight). Because children’s heights and weights were measured in the spring, we posited that this timing provided time for participation to have a lagged effect on body mass. Furthermore, because BMID at time t is a function of BMID at time t−1, it is also a function of the association between lagged BMI at time t−1 and NSLP at time t−1. More details regarding the chosen model, commands, output, and mathematical specifications are given in eMethods in the Supplement.

We use entropy balancing45 to ensure that HHFKA-exposed and HHFKA-unexposed cohorts were comparable (eMethods in the Supplement). Weighted cohorts were similar across covariates, whereas unweighted cohorts had discrepancies (eTables 4 and 5 in the Supplement). We used Mplus, version 7.11,44 and missing covariate data were handled with multiple imputation using the mice package in R.46 Statistical significance was defined by a 2-sided P value <.05. We plotted estimated mean BMIDs by free or reduced-price NSLP participation and report standard errors.

We examined whether associations varied by sex but did not assess associations by race and ethnicity because of low statistical power. We conducted sensitivity analyses, including imputation of grade 5 full-price NSLP participation in ECLS-K:2011 (eMethods and eTable 6 in the Supplement). All data were analyzed from January 1 to September 7, 2021.

Results

In total, 3388 children in ECLS-K:1999 (1696 girls [50.1%] and 1692 boys [49.9%]; mean [SD] age at baseline, 74.6 [4.3] months) and 2570 children in ECLS-K:2011 (1222 girls [47.5%] and 1348 boys [52.5%]; mean [SD] age at baseline, 73.6 [4.2] months) were included in the analysis. The data given in the Table reflect severe obesity among US kindergarteners from low-income families, providing evidence that BMID is an appropriate measure to study body mass growth.30 The ECLS-K:1999 kindergarteners’ mean BMI was 2 BMI units below their obesity threshold. Between ECLS-K:1999 and ECLS-K:2011, the mean BMID of kindergarteners from low-income families moved closer to their obesity threshold. Free or reduced-price NSLP participation increased in ECLS-K:1999 from 1896 (56.0%) to 2309 (68.2%) and from 1767 (68.8%) to 1957 (76.7%) in ECLS-K:2011. The median household income decreased from $22 500 (IQR, $12 500-$32 500) in 1999 to $20 522 (IQR, $13 060-$27 985) in 2011, after adjusting for inflation. However, mother’s highest achieved educational level at kindergarten increased between the cohorts, with an approximately 10 percentage-point increase in at least some college attainment. The more recent cohort (ECLS-K:2011) included more Hispanic children (1026 [39.9%] vs 955 [28.2%]) and fewer non-Hispanic Black (605 [17.9%] vs 280 [10.9%]) or non-Hispanic White (1418 [41.9%] vs 979 [38.1%]) children than in ECLS-K:1999.

Table. Baseline Characteristics of ECLS Cohorts by Free or Reduced-Price National School Lunch Program Participation.

Characteristic No. (%)
ECLS-K:1999 ECLS-K:2011
All (n = 3388) NSLP All (n = 2570) NSLP
Not free or reduced pricea Free or reduced priceb Not free or reduced pricec Free or reduced priced
Age in kindergarten, mean (SD), mo 74.6 (4.3) 74.9 (4.3) 74.4 (4.2) 73.6 (4.5) 73.7 (4.5) 73.6 (4.6)
Race and ethnicity
Black non-Hispanic 605 (17.9) 129 (8.7) 476 (25.1) 280 (10.9) 29 (3.6) 251 (14.2)
Hispanic 955 (28.2) 277 (18.6) 678 (35.8) 1026 (39.9) 171 (21.3) 855 (48.4)
White non-Hispanic 1418 (41.9) 925 (62.0) 493 (26.0) 979 (38.1) 480 (59.8) 499 (28.2)
Othere 409 (12.1) 161 (10.8) 248 (13.1) 285 (11.1) 123 (15.3) 162 (9.2)
Missing 1 (0.03) 0 1 (0.1) 0 0 0
Sex
Female 1696 (50.1) 731 (49.0) 965 (50.9) 1222 (47.6) 375 (46.7) 847 (47.9)
Male 1692 (49.1) 761 (51.0) 931 (49.1) 1348 (52.5) 428 (53.3) 920 (52.1)
BMID, mean (SD)f
Kindergarten −2.1 (2.6) −2.3 (2.5) −2.0 (2.6) −1.8 (2.6) −2.1 (2.5) −1.7 (2.7)
Grade 1 −2.4 (3.2) −2.7 (2.9) −2.3 (3.4) −2.1 (3.2) −2.6 (2.9) −1.9 (3.2)
Grade 5 −2.6 (5.2) −3.1 (5.0) −2.4 (5.3) −2.6 (5.2) −3.7 (5.0) −2.3 (5.3)
BMI z score, mean (SD)
Kindergarten 0.5 (1.1) 0.4 (1.1) 0.5 (1.2) 0.6 (1.1) 0.5 (1.0) 0.7 (1.1)
Grade 1 0.4 (1.3) 0.4 (1.2) 0.5 (1.4) 0.6 (1.2) 0.4 (1.1) 0.6 (1.2)
Grade 5 0.8 (1.2) 0.7 (1.1) 0.8 (1.2) 0.7 (1.4) 0.5 (1.2) 0.8 (1.5)
Kindergarten weight status
Underweight 0 0 0 0 0 0
Normal weight 2144 (63.3) 980 (65.7) 1164 (61.4) 1456 (56.7) 502 (62.5) 954 (54.0)
Overweight 778 (23.0) 333 (22.3) 445 (23.5) 672 (26.2) 196 (24.4) 476 (26.9)
Obesity 323 (9.5) 123 (8.2) 200 (10.6) 311 (12.1) 70 (8.7) 241 (13.6)
Severe obesity 143 (4.2) 56 (3.8) 87 (4.6) 131 (5.1) 35 (4.4) 96 (5.4)
Grade 1 weight status
Underweight 0 0 0 0 0 0
Normal weight 2275 (67.2) 802 (69.9) 1473 (65.3) 1586 (61.7) 414 (67.5) 1172 (59.9)
Overweight 573 (16.9) 197 (17.2) 376 (16.8) 487 (19.0) 112 (18.3) 375 (19.2)
Obesity 339 (10.0) 95 (8.3) 244 (10.9) 345 (13.4) 64 (10.4) 281 (14.4)
Severe obesity 201 (5.9) 54 (4.7) 147 (6.6) 152 (5.9) 23 (3.8) 129 (6.6)
Grade 5 weight status
Underweight 0 0 0 0 0 0
Normal weight 2000 (59.0) 681 (63.1) 1319 (57.1) 1495 (58.2) 400 (66.8) 1095 (55.6)
Overweight 518 (15.3) 159 (14.7) 359 (15.6) 388 (15.1) 85 (14.2) 303 (15.4)
Obesity 544 (16.1) 153 (14.2) 391 (16.9) 435 (16.9) 84 (14.0) 351 (17.8)
Severe obesity 326 (9.6) 86 (8.0) 240 (10.4) 252 (9.8) 30 (5.0) 222 (11.3)
Birthweight, g
<2500 299 (8.8) 102 (6.8) 197 (10.4) 182 (7.1) 52 (6.5) 130 (7.4)
2500-3999 2601 (76.8) 1169 (78.4) 1432 (75.5) 1503 (58.5) 504 (62.8) 999 (56.5)
≥4000 355 (10.5) 174 (11.7) 181 (9.6) 193 (7.5) 60 (7.5) 133 (7.5)
Missing 133 (3.9) 47 (3.2) 86 (4.5) 692 (26.9) 187 (23.3) 505 (28.6)
Mother’s educational attainment at kindergarten
<9th grade 323 (9.5) 65 (4.4) 258 (13.6) 278 (10.8) 40 (5.0) 238 (13.5)
Grade 9-12 503 (14.9) 141 (9.5) 362 (19.1) 344 (13.4) 42 (5.2) 302 (17.1)
High school or GED 1333 (39.3) 588 (39.4) 745 (39.3) 764 (29.7) 197 (24.5) 567 (32.1)
Some college 944 (27.9) 510 (34.2) 434 (22.9) 853 (33.2) 333 (41.5) 520 (29.4)
Bachelor’s degree 160 (4.7) 109 (7.3) 51 (2.7) 240 (9.3) 137 (17.1) 103 (5.8)
≥Graduate school 61 (1.8) 48 (3.2) 13 (0.7) 91 (3.5) 54 (6.7) 37 (2.1)
Missing 64 (1.9) 31 (2.1) 33 (1.7) 0 0 0
Household income, median (IQR), $g
Kindergarten 22 500 (12 500-32 500) 27 500 (17 500-37 500) 17 500 (12 500-27 500) 20 522 (13 060-27 985) 27 985 (20 522-46 642) 16 791 (9328-24 253)
Grade 1 22 500 (12 500-32 500) 32 500 (27 500-45 000) 17 500 (12 500-27 500) 20 522 (13 060-27 985) 33 582 (24 253-46 641) 16 791 (9328-24 253)
Grade 5 27 500 (17 500-37 500) 37 500 (27 500-62 500) 22 500 (12 500-32 500) 24 254 (16 791-33 582) 46 642 (33 582-65 299) 20 522 (13 060-27 985)
Family dinners, mean (SD), No./wkh
Kindergarten 5.8 (1.8) 5.8 (1.8) 5.8 (1.8) 5.8 (1.7) 5.8 (1.7) 5.9 (1.8)
Grade 1 5.9 (1.7) 5.7 (1.7) 5.9 (1.7) 5.7 (1.8) 5.7 (1.7) 5.7 (1.8)
Grade 5 5.6 (1.8) 5.5 (1.8) 5.6 (1.8) 5.6 (1.8) 5.4 (1.7) 5.6 (1.8)
Television, mean (SD), h/di
Kindergarten 2.1 (1.4) 2.0 (1.3) 2.2 (1.6) 2.3 (1.5) 2.2 (1.5) 2.3 (1.5)
Grade 1 2.3 (1.5) 2.1 (1.3) 2.4 (1.5) 1.8 (1.4) 1.6 (1.2) 1.8 (1.4)
School in urban areaj
Kindergarten 1270 (37.5) 446 (29.9) 824 (43.4) 914 (35.6) 189 (23.5) 725 (41.0)
Grade 1 1261 (37.2) 337 (29.4) 924 (41.3) 920 (35.8) 149 (24.3) 771 (39.4)
Grade 5 1202 (35.5) 306 (28.4) 896 (38.8) 809 (31.5) 118 (19.7) 691 (35.1)
Mother employedk
Kindergarten 1258 (37.1) 594 (39.8) 664 (35.0) 644 (25.1) 218 (27.2) 426 (24.1)
Grade 1 1537 (46.8) 555 (48.3) 982 (43.8) 911 (35.5) 260 (42.4) 651 (33.3)
Grade 5 1630 (48.1) 583 (54.0) 1047 (45.3) 1117 (43.5) 330 (55.1) 787 (39.9)
Open in a new tab

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); BMID, difference from obesity threshold (95th percentile); ECLS, Early Childhood Longitudinal Study; ECLS-K:1999, ECLS Kindergarten Class of 1998-1999; ECLS-K:2011, ECLS Kindergarten Class of 2010-2011; NSLP, National School Lunch Program.

a

No. (%) in kindergarten, 1492 (44.0%); grade 1, 1148 (33.9%); and grade 5, 1079 (31.8%).

b

No. (%) in kindergarten, 1896 (56.0%); grade 1, 2240 (66.1%); and grade 5, 2309 (68.2%).

c

No. (%) in kindergarten, 803 (31.2%); grade 1, 613 (23.8%); and grade 5, 599 (23.3%).

d

No. (%) in kindergarten, 1767 (68.8%), grade 1, 1957 (76.1%); and grade 5, 1971 (76.7%).

e

American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, or more than 1 race and ethnicity.

f

BMID is child’s BMI minus the BMI at the age- and sex-specific 95th percentile.

g

Household incomes in ECLS-K:2011 were adjusted for inflation from 1999.

h

No. missing in ECLS-K:1999 kindergarten, 2; grade 1, 0; and grade 5, 2; and No. missing in ECLS-K:2011 kindergarten, 1; grade 1, 2; and grade 5, 0.

i

No. missing in ECLS-K:1999 kindergarten, 36; and grade 1, 0. No. missing in ECLS-K:2011 kindergarten, 0; and grade 1, 0.

j

No. missing in ECLS-K:1999 kindergarten, 0; grade 1, 17; and grade 5, 135. No. missing in ECLS-K:2011 kindergarten, 16; grade 1, 40; and grade 5, 84.

k

No. missing in ECLS-K:1999 kindergarten, 424; grade 1, 103; and grade 5, 117. No. missing in ECLS-K:2011 kindergarten, 573; grade 1, 0; and grade 5, 0.

At grade 1, there were no associations between BMID and the free or reduced-price NSLP after adjustment for kindergarten BMID and covariates in either ECLS-K:1999 (β = −0.25; 95% CI, −1.17 to 0.68) and ECLS-K:2011 (β = 0.08; 95% CI, −1.04 to 1.20) (eFigure 1 in the Supplement). By grade 5, free or reduced-price NSLP recipients had higher BMID, adjusted for grade 1 BMID, in ECLS-K:1999 (β = 0.54; 95% CI, 0.27-0.81), but there was no association in ECLS-K:2011 (β = −0.07; 95% CI, −0.58 to 0.45). Change in BMID for NSLP recipients between kindergarten and grade 1 was not different between cohorts (χ21 = 0.19, P = .66). However, the before vs after HHFKA change between cohorts (grades 1 and 5) was different (χ21 = 4.29, P = .04).

Figure 2 presents kindergarten and grades 1 and 5 mean BMID estimates for free or reduced price by NSLP participation and by cohort. Prior to the HHFKA, children participating in the free or reduced-price NSLP maintained their BMID levels from grade 1 through grade 5 compared with nonparticipating children who experienced a decrease in BMID. After the HHFKA, all children had the same declining BMID trajectory. The estimated BMIDs and SEs indicated a negative BMID trajectory in ECLS-K:2011, but BMID change was indistinguishable between the grade 1 and grade 5 years irrespective of the free or reduced-price NSLP. Notably, children in ECLS-K:2011 had higher BMIDs (mean [SE], −1.8 [0.6]) in kindergarten than children in ECLS-K:1999 (mean [SE], −2.2 [0.5]). Furthermore, nonparticipants did not have as steep a BMID decrease as their ECLS-K:1999 counterparts. Associations of before vs after HHFKA BMID change did not differ by sex (χ21, 1.84; P = .19) (Figure 3). Parents in ECLS-K:1999 reported that 2385 children in kindergarten (70.4%), 3021 children in grade 1 (89.2%), and 3060 children in grade 5 (90.3%) participated in either the full-price, free, or reduced-price NSLP. In ECLS-K:2011, 2043 children in kindergarten (79.5%) and 2257 children in grade 1 (87.8%) participated in the NSLP. After NSLP participation imputation at grade 5 in ECLS-K:2011, we classified 2424 children (94.3%) as participating in the NSLP. The NSLP participation associations were the same as for our main model (Figure 4). For grade 1, there were no associations between BMID and NSLP participation in ECLS-K:1999 (β = −0.21; 95% CI, −1.43 to 1.00) or in ECLS-K:2011 (β = −0.62; 95% CI, −2.87 to 1.63). By grade 5, NSLP recipients in ECLS-K:1999 had higher BMID, adjusted for BMID in grade 1, (β = 0.55; 95% CI, 0.19-0.91), but there was no association in ECLS-K:2011 (β = 0.36; 95% CI, −0.46 to 1.18). The change before vs after the HHFKA from grades 1 to 5 between the cohorts was similar (χ21 = 0.17, P = .68).

Figure 2. Body Mass Index Difference (BMID) From Obesity by Free or Reduced-Price National School Lunch Program (NSLP) Participation and Early Childhood Longitudinal Study (ECLS) Cohort.

Figure 2.

Open in a new tab

ECLS-K:1999 indicates ECLS Kindergarten Class of 1998-1999; ECLS-K:2011, ECLS Kindergarten Class of 2010-2011. Error bars represent SEs.

Figure 3. Body Mass Index Difference (BMID) From Obesity by Free or Reduced-Price National School Lunch Program (NSLP) Participation, Early Childhood Longitudinal Study (ECLS) Cohort, and Sex.

Figure 3.

Open in a new tab

ECLS-K:1999 indicates ECLS Kindergarten Class of 1998-1999; ECLS-K:2011, ECLS Kindergarten Class of 2010-2011. Error bars represent SEs.

Figure 4. Body Mass Index Difference (BMID) from Obesity by Imputed Full-Price, Free, or Reduced-Price National School Lunch Program (NSLP) Participation and Early Childhood Longitudinal Study (ECLS) Cohort.

Figure 4.

Open in a new tab

ECLS-K:1999 indicates ECLS Kindergarten Class of 1998-1999; ECLS-K:2011, ECLS Kindergarten Class of 2010-2011. Error bars represent SEs.

Using BMI z scores, the associations were similar with our main model. However, the obesogenic pre-HHFKA grade 5 association between free or reduced-price NSLP and change in BMI z score was attenuated (β = 0.08; 95% CI, 0.00-0.16) compared with models using BMID (eFigure 2 in the Supplement). Before vs after HHFKA grade 5 BMID change by free or reduced-price NSLP participation between cohorts was similar (χ21 = 1.57, P = .21). Associations across unweighted sensitivity models were similar (eTable 7 in the Supplement).

Discussion

In this cohort study, we estimated associations between the free or reduced-price NSLP and children’s change in BMID before and after the implementation of the HHFKA in 2 cohorts of school-aged children. Before the implementation of the HHFKA, children participating in the free or reduced-price NSLP had a more obesogenic BMI trajectory between grades 1 and 5 than children who did not participate. However, this association was not observed in the recent cohort who entered grade 5 after HHFKA implementation. Associations between free or reduced-price NSLP participation and BMID change between kindergarten to grade 1 (pre-HHFKA implementation period for both cohorts) were similar across cohorts, which provides evidence that implementation of the HHFKA, not cohort differences, explains our results.

Results of research evaluating school meals and child obesity preceding the HHFKA have been mixed, with several reports suggesting that school meals contribute to child obesity,6,7,8,9,43,47,48 whereas others do not.21,49,50 Among studies with null findings,49,50 data limitations51,52,53—that is, measurement bias from parent-reported height and weight, small numbers of children, and inability to distinguish body mass growth differences during development—may have biased findings toward the null. Vericker et al19 found that the HHFKA reversed the association between the NSLP and child BMI growth for only boys in the ECLS. Yet without quasi-experimental consideration, the authors estimated grade 1 to 3 changes in BMI z scores separately for the cohorts, and they did not test whether associations differed across cohorts.19

The implementation of the HHFKA increased fruit, vegetable, and whole grain amounts and limited saturated fats. In addition, the HHFKA introduced provisions to improve school environments, such as increasing nutritional requirements of all foods and beverages sold in schools (eg, Smart Snacks54). Although school-level provisions may have improved child body mass growth, on average, children in ECLS-K:2011 started school in kindergarten with BMID skewed more toward obesity than ECLS-K:1999 children. Despite the downward trajectory, all ECLS-K:2011 children had a mean BMI in grade 5 that was about as close to the obesity threshold as for fifth graders participating in the free or reduced-price NSLP in ECLS-K:1999. This finding suggests that school policy may need to overcome population-level drivers of child obesity when children may be more predisposed to obesity than children in previous decades.55 The gap between participants and nonparticipants may have closed in ECLS-K:2011 because nonparticipants did not experience the BMID decline that their historical counterparts did. This could reflect secular environmental differences (eg, competitive food or beverage availability) in which children in ECLS-K:2011 may have had greater exposure to obesogenic dietary options. Increasing access to school meals with more rigorous nutritional requirements that are culturally pleasing to children may be needed to achieve greater success in reducing child obesity.

Despite arguments against nutrition standards in the HHFKA,56 increasing evidence shows that the HHFKA has not reduced school meal participation57,58 or students’ acceptance of school lunches.59 Moreover, recent studies have found that the HHFKA was associated with improvements in the dietary intake of US school students,13,16,18 which is consistent with studies examining nutritional quality after HHFKA implementation.12,38 The present study ended in 2016, prior to the 2018 loosening of the nutritional requirements in the HHFKA.60 Whether future studies will find support for the HHFKA improving disparities in child body mass growth after 2018, or whether the proposed weakening of the nutrition standards becomes implemented, is unknown.61

Limitations and Strengths

Our study has limitations. To address selection bias of free or reduced-price NSLP participation we estimated BMID change within and between individuals across cohorts. We did not restrict our sample to children who changed free or reduced-price NSLP status, and we used weights to balance cohort differences. However, BMI trajectories may differ as a result of unobserved factors not adjusted for, as acknowledged previously.21 Within-cohort and cross-cohort comparisons of associations between NSLP participation and BMID should be interpreted cautiously. We addressed confounding by accounting for latent time-invariant child characteristics, BMID trajectory, and multiple imputation of missing covariate data. The lack of information on full-price NSLP participation after HHFKA implementation in ECLS-K:2011 is a limitation. However, we imputed missing participation information. Although finding statistically significant associations supports our findings, the lack of a cross-cohort difference between BMID change and free or reduced-price NSLP participation does not. However, most fifth graders (94%) in ECLS-K:2011 were categorized as participants, which limited our statistical power to detect a cross-cohort difference. We were unable to assess differential associations by race and ethnicity due to limited samples. In addition, the NSLP may have a longer lagged association with the BMID than we accounted for in our analyses. Our study was also limited by incomplete school breakfast participation information. We did not use survey weights because we selected a sample of students from low-income families; thus, our analytic sample is not nationally representative.

Despite these limitations, direct contrasts in body mass growth before vs after HHFKA implementation enabled us to identify a critical period for HHFKA-associated change in the free or reduce price-NSLP from grades 1 through 5. Our ability to include lagged BMID growth enabled us to estimate BMID change while accounting for children who may have been susceptible to early onset obesity,62 either a result of genetic factors or postnatal environments.63,64 This ability may have enabled us to detect body mass growth change associated with the NSLP using a measure of BMI that is sensitive to high values. Indeed, the attenuated obesogenic association between the NSLP and BMI z score change in grade 5 before HHFKA implementation highlights the limitations in the use of BMI z scores. In addition, our findings were robust to clustering by school, adjusting for time-varying mother’s educational level, and adjusting for time-varying school-level percent eligible for the free or reduced-price NSLP.

Conclusions

In this cohort study using cross-cohort comparisons of children in ECLS-K:1999 and ECLS-K:2011, we observed obesogenic growth of children from kindergarten through grade 5 among participants in the free or reduced-price NSLP before the implementation of the HHFKA that was absent after the implementation, suggesting that the nutritional standards in the HHFKA improved low-income children’s BMI trajectories.

Supplement.

eTable 1. Baseline Characteristics of Analytic Sample Compared to Children Excluded for Missing BMI

eTable 2. Tested Models and Fit Indices

eTable 3. Random and Fixed Effects Models by Cohort, Hybrid Model 2b by Cohort, and Hybrid Model 5 by Cohort With Lagged BMID

eMethods.

eTable 4. Child, Family, and School Characteristics of ECLS-K:2011 and ECLS-K:1999 Low-Income Child Samples

eTable 5. Deciles of Numerical Variables of ECLS-K:2011 and ECLS-K:1999 Low-Income Child Samples

eFigure 1. Path Estimates of Model

eTable 6. Performance Characteristics of Proposed Full and Simple Imputation Approaches for Predicting Grade 1 Full or Free or Reduced-Price NSLP Participation

eFigure 2. Path Estimates of Sensitivity Model Using BMI z Score

eTable 7. Free or Reduced-Price NSLP Estimates in Unweighted Sensitivity Models

eReferences

Click here for additional data file. (643.5KB, pdf)

References

  • 1.National Center for Health Statistics. Data brief No. 288: prevalence of obesity among adults and youth, United States, 2015-2016. October 2017. Accessed March 23, 2022. https://www.cdc.gov/nchs/data/databriefs/db288.pdf [Google Scholar]
  • 2.Hales CM, Fryar CD, Carroll MD, Freedman DS, Ogden CL. Trends in obesity and severe obesity prevalence in US youth and adults by sex and age, 2007-2008 to 2015-2016. JAMA. 2018;319(16):1723-1725. doi: 10.1001/jama.2018.3060 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Ogden CL, Carroll MD, Fakhouri TH, et al. Prevalence of obesity among youths by household income and education level of head of household—United States 2011-2014. MMWR Morb Mortal Wkly Rep. 2018;67(6):186-189. doi: 10.15585/mmwr.mm6706a3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Rogers R, Eagle TF, Sheetz A, et al. The relationship between childhood obesity, low socioeconomic status, and race/ethnicity: lessons from Massachusetts. Child Obes. 2015;11(6):691-695. doi: 10.1089/chi.2015.0029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.US Department of Agriculture Economic Research Service. Children’s food security and intakes from school meals. May 2010. Accessed March 23, 2022. https://www.ers.usda.gov/publications/pub-details/?pubid=84339
  • 6.National Bureau of Economic Research. US food and nutrition programs, working paper 21057. March 2015. Accessed March 23, 2022. https://www.nber.org/system/files/working_papers/w21057/w21057.pdf
  • 7.Capogrossi K, You W. The influence of school nutrition programs on the weight of low-income children: a treatment effect analysis. Health Econ. 2017;26(8):980-1000. doi: 10.1002/hec.3378 [DOI] [PubMed] [Google Scholar]
  • 8.Millimet DL, Tschernis R, Husain M. School nutrition programs and the incidence of childhood obesity. J Hum Resour. 2010;45:640-654. [Google Scholar]
  • 9.Schanzenbach DW. Do school lunches contribute to childhood obesity? J Hum Resour. 2009;44(3):684-709. [Google Scholar]
  • 10.National Bureau of Economic Research . Means-Tested Transfer Programs in the United States. University of Chicago Press; 2003. [Google Scholar]
  • 11.US Department of Agriculture . Healthy Hunger-Free Kids Act: school meals. Published November 20, 2013. Accessed May 29, 2018. https://www.fns.usda.gov/cn/healthy-hunger-free-kids-act
  • 12.Gearan EC, Fox MK. Updated nutrition standards have significantly improved the nutritional quality of school lunches and breakfasts. J Acad Nutr Diet. 2020;120(3):363-370. doi: 10.1016/j.jand.2019.10.022 [DOI] [PubMed] [Google Scholar]
  • 13.Berger AT, Widome R, Erickson DJ, Laska MN, Harnack LJ. Changes in association between school foods and child and adolescent dietary quality during implementation of the Healthy, Hunger-Free Kids Act of 2010. Ann Epidemiol. 2020;47:30-36. doi: 10.1016/j.annepidem.2020.05.013 [DOI] [PubMed] [Google Scholar]
  • 14.US Department of Agriculture. School nutrition and meal cost study. October 19, 2019. Accessed March 23, 2022. https://www.fns.usda.gov/school-nutrition-and-meal-cost-study
  • 15.Johnson DB, Podrabsky M, Rocha A, Otten JJ. Effect of the Healthy Hunger-Free Kids Act on the nutritional quality of meals selected by students and school lunch participation rates. JAMA Pediatr. 2016;170(1):e153918. doi: 10.1001/jamapediatrics.2015.3918 [DOI] [PubMed] [Google Scholar]
  • 16.Valizadeh P, Ng SW. The new school food standards and nutrition of school children: direct and indirect effect analysis. Econ Hum Biol. 2020;39:100918. doi: 10.1016/j.ehb.2020.100918 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Smith TA, Mojduszka EM, Chen S. Did the new school meal standards improve the overall quality of children’s diets? Paper presented at: Agricultural and Applied Economics Association Annual Meeting, July 21-23, 2019; Atlanta, GA. [Google Scholar]
  • 18.Kinderknecht K, Harris C, Jones-Smith J. Association of the Healthy, Hunger-Free Kids Act with dietary quality among children in the US National School Lunch Program. JAMA. 2020;324(4):359-368. doi: 10.1001/jama.2020.9517 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Vericker TC, Gearing ME, Kim SD. Updated nutrition standards for school meals associated with improved weight outcomes for boys in elementary school. J Sch Health. 2019;89(11):907-915. doi: 10.1111/josh.12828 [DOI] [PubMed] [Google Scholar]
  • 20.Kenney EL, Barrett JL, Bleich SN, Ward ZJ, Cradock AL, Gortmaker SL. Impact of the Healthy, Hunger-Free Kids Act on obesity trends. Health Aff (Millwood). 2020;39(7):1122-1129. doi: 10.1377/hlthaff.2020.00133 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Gundersen C, Kreider B, Pepper J. The impact of the National School Lunch Program on child health: a nonparametric bounds analysis. J Econom. 2012;166(1):79-91. doi: 10.1016/j.jeconom.2011.06.007 [DOI] [Google Scholar]
  • 22.Woo JG. Using body mass index z-score among severely obese adolescents: a cautionary note. Int J Pediatr Obes. 2009;4(4):405-410. doi: 10.3109/17477160902957133 [DOI] [PubMed] [Google Scholar]
  • 23.Kakinami L, Henderson M, Chiolero A, Cole TJ, Paradis G. Identifying the best body mass index metric to assess adiposity change in children. Arch Dis Child. 2014;99(11):1020-1024. doi: 10.1136/archdischild-2013-305163 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Paluch RA, Epstein LH, Roemmich JN. Comparison of methods to evaluate changes in relative body mass index in pediatric weight control. Am J Hum Biol. 2007;19(4):487-494. doi: 10.1002/ajhb.20608 [DOI] [PubMed] [Google Scholar]
  • 25.Flegal KM, Wei R, Ogden CL, Freedman DS, Johnson CL, Curtin LR. Characterizing extreme values of body mass index-for-age by using the 2000 Centers for Disease Control and Prevention growth charts. Am J Clin Nutr. 2009;90(5):1314-1320. doi: 10.3945/ajcn.2009.28335 [DOI] [PubMed] [Google Scholar]
  • 26.Cole TJ, Faith MS, Pietrobelli A, Heo M. What is the best measure of adiposity change in growing children: BMI, BMI %, BMI z-score or BMI centile? Eur J Clin Nutr. 2005;59(3):419-425. doi: 10.1038/sj.ejcn.1602090 [DOI] [PubMed] [Google Scholar]
  • 27.Hales CM, Fryar CD, Carroll MD, Freedman DS, Aoki Y, Ogden CL. Differences in obesity prevalence by demographic characteristics and urbanization level among adults in the United States, 2013-2016. JAMA. 2018;319(23):2419-2429. doi: 10.1001/jama.2018.7270 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Skinner AC, Skelton JA. Prevalence and trends in obesity and severe obesity among children in the United States, 1999-2012. JAMA Pediatr. 2014;168(6):561-566. doi: 10.1001/jamapediatrics.2014.21 [DOI] [PubMed] [Google Scholar]
  • 29.Skelton JA, Cook SR, Auinger P, Klein JD, Barlow SE. Prevalence and trends of severe obesity among US children and adolescents. Acad Pediatr. 2009;9(5):322-329. doi: 10.1016/j.acap.2009.04.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Centers for Disease Control and Prevention . Defining severe obesity and BMI metrics among children with very high BMIs. A SAS Program for the 2000 CDC Growth Charts (ages 0 to <20 years). 2019. Updated February 18, 2022. Accessed March 24, 2022. https://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/sas.htm
  • 31.Guo SS, Wu W, Chumlea WC, Roche AF. Predicting overweight and obesity in adulthood from body mass index values in childhood and adolescence. Am J Clin Nutr. 2002;76(3):653-658. doi: 10.1093/ajcn/76.3.653 [DOI] [PubMed] [Google Scholar]
  • 32.Serdula MK, Ivery D, Coates RJ, Freedman DS, Williamson DF, Byers T. Do obese children become obese adults? a review of the literature. Prev Med. 1993;22(2):167-177. doi: 10.1006/pmed.1993.1014 [DOI] [PubMed] [Google Scholar]
  • 33.Rolland-Cachera MF, Deheeger M, Avons P, Guilloud-Bataille M, Patois EMS. Tracking the development of adiposity from one month of age to adulthood. Ann Hum Biol. 1987;14:219-222. doi: 10.1080/03014468700008991 [DOI] [PubMed] [Google Scholar]
  • 34.Rolland-Cachera MF, Deheeger M, Maillot M, Bellisle F. Early adiposity rebound: causes and consequences for obesity in children and adults. Int J Obes (Lond). 2006;30(suppl 4):S11-S17. doi: 10.1038/sj.ijo.0803514 [DOI] [PubMed] [Google Scholar]
  • 35.United States Census Bureau . Poverty thresholds. 2016. Accessed January, 2017. https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-poverty-thresholds.html
  • 36.National Center for Education Statistics . Early Childhood Longitudinal Study, kindergarten class of 1998-99 (ECLS-K): eighth-grade methodology report. September 2009. Accessed March 24, 2022. https://nces.ed.gov/pubs2009/2009003.pdf
  • 37.National Center for Education Statistics . Early Childhood Longitudinal Study, Kindergarten Class of 2010–11 (ECLS-K:2011). 2019. Accessed March 24, 2022. https://nces.ed.gov/pubs2019/2019051.pdf
  • 38.Fox MK, Gearan E, Cabili C, et al. School Nutrition and Meal Cost Final Report Volume 4: Student Participation, Satisfaction, Plate Waste, and Dietary Intake. U.S. Department of Agriculture, Food and Nutrition Service, Office of Policy Support; 2019. [Google Scholar]
  • 39.National Center for Education Statistics . Instruments and assessments. 2020. Accessed September 2020. https://nces.ed.gov/ecls/kinderinstruments.asp
  • 40.National Center for Education Statistics . Data collection instruments. 2020. Accessed September 2020. https://nces.ed.gov/ecls/instruments2011.asp
  • 41.Gleason PM, Dodd AH. School breakfast program but not school lunch program participation is associated with lower body mass index. J Am Diet Assoc. 2009;109(2)(suppl):S118-S128. doi: 10.1016/j.jada.2008.10.058 [DOI] [PubMed] [Google Scholar]
  • 42.Lowry R, Wechsler H, Galuska DA, Fulton JE, Kann L. Television viewing and its associations with overweight, sedentary lifestyle, and insufficient consumption of fruits and vegetables among US high school students: differences by race, ethnicity, and gender. J Sch Health. 2002;72(10):413-421. doi: 10.1111/j.1746-1561.2002.tb03551.x [DOI] [PubMed] [Google Scholar]
  • 43.Li J, Hooker NH. Childhood obesity and schools: evidence from the national survey of children’s health. J Sch Health. 2010;80(2):96-103. doi: 10.1111/j.1746-1561.2009.00471.x [DOI] [PubMed] [Google Scholar]
  • 44.Muthén LK, Muthén BO. Mplus User's Guide. Sixth Edition. Muthén & Muthén; 1998-2010. [Google Scholar]
  • 45.Hainmueller J. Entropy balancing for causal effects: a multivariate reweighting method to produce balanced samples in observational studies. Polit Anal. 2012;20(1):25-46. doi: 10.1093/pan/mpr025 [DOI] [Google Scholar]
  • 46.van Buuren S, Groothuis-Oudshoorn K. mice: multivariate imputation by chained equations in R. J Stat Softw. 2011;45(3):1-67. doi: 10.18637/jss.v045.i03 [DOI] [Google Scholar]
  • 47.Taber DR, Chriqui JF, Powell L, Chaloupka FJ. Association between state laws governing school meal nutrition content and student weight status: implications for new USDA school meal standards. JAMA Pediatr. 2013;167(6):513-519. doi: 10.1001/jamapediatrics.2013.399 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Story M, Nanney MS, Schwartz MB. Schools and obesity prevention: creating school environments and policies to promote healthy eating and physical activity. Milbank Q. 2009;87(1):71-100. doi: 10.1111/j.1468-0009.2009.00548.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Hofferth SL, Curtin S. Poverty, food programs, and childhood obesity. J Policy Anal Manage. 2005;24(4):703-726. doi: 10.1002/pam.20134 [DOI] [PubMed] [Google Scholar]
  • 50.Mirtcheva DM, Powell LM. National School Lunch Program participation and child body weight. East Econ J. 2013;39:328-345. doi: 10.1057/eej.2012.14 [DOI] [Google Scholar]
  • 51.Weden MM, Brownell PB, Rendall MS, Lau C, Fernandes M, Nazarov Z. Parent-reported height and weight as sources of bias in survey estimates of childhood obesity. Am J Epidemiol. 2013;178(3):461-473. doi: 10.1093/aje/kws477 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Rendall MS, Weden MM, Lau C, Brownell P, Nazarov Z, Fernandes M. Evaluation of bias in estimates of early childhood obesity from parent-reported heights and weights. Am J Public Health. 2014;104(7):1255-1262. doi: 10.2105/AJPH.2014.302001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.The National Academies of Sciences Engineering Medicine. Assessing Prevalence and Trends in Obesity: Navigating the Evidence. National Academies Press; 2016. [Google Scholar]
  • 54.US Department of Agriculture . Tools for schools: focusing on smart snacks. 2019. Accessed February 1, 2021. https://www.fns.usda.gov/cn/tools-schools-focusing-smart-snacks
  • 55.Ogden CL, Carroll MD, Curtin LR, Lamb MM, Flegal KM. Prevalence of high body mass index in US children and adolescents, 2007-2008. JAMA. 2010;303(3):242-249. doi: 10.1001/jama.2009.2012 [DOI] [PubMed] [Google Scholar]
  • 56.School Nutrition Association. Releases SNA 2015 position paper calling for greater funding and flexibility for school meal programs. Published January 29, 2015. Accessed March 24, 2022. https://schoolnutrition.org/pressreleases/snareleases2015positionpaper/ [Google Scholar]
  • 57.Vaudrin N, Lloyd K, Yedidia MJ, Todd M, Ohri-Vachaspati P. Impact of the 2010 US Healthy, Hunger-Free Kids Act on school breakfast and lunch participation rates between 2008 and 2015. Am J Public Health. 2018;108(1):84-86. doi: 10.2105/AJPH.2017.304102 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Martinelli S, Acciai F, Au LE, Yedidia MJ, Ohri-Vachaspati P.Parental perceptions of the nutritional quality of school meals and student meal participation: before and after the Healthy Hunger-Free Kids Act. J Nutr Educ Behav. 2020;52(11):1018-1025. doi: 10.1016/j.jneb.2020.05.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Turner L, Chaloupka FJ. Perceived reactions of elementary school students to changes in school lunches after implementation of the United States Department of Agriculture’s new meals standards: minimal backlash, but rural and socioeconomic disparities exist. Child Obes. 2014;10(4):349-356. doi: 10.1089/chi.2014.0038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.US Department of Agriculture . Child nutrition programs: flexibilities for milk, whole grains, and sodium requirements. Fed Regist. 2018;83(238):63775-63784. 7 CFR §210, §215, §220, and §226. [PubMed] [Google Scholar]
  • 61.Federal Register. Simplifying meal service and monitoring requirements in the National School Lunch and School Breakfast Programs. 2020;85(15):4094-4134. 7 CFR §210, §215, §220, §226, and §235. [Google Scholar]
  • 62.Lakshman R, Elks CE, Ong KK. Childhood obesity. Circulation. 2012;126(14):1770-1779. doi: 10.1161/CIRCULATIONAHA.111.047738 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Catalano PM, Ehrenberg HM. The short- and long-term implications of maternal obesity on the mother and her offspring. BJOG. 2006;113(10):1126-1133. doi: 10.1111/j.1471-0528.2006.00989.x [DOI] [PubMed] [Google Scholar]
  • 64.Hillier TA, Pedula KL, Schmidt MM, Mullen JA, Charles MA, Pettitt DJ. Childhood obesity and metabolic imprinting: the ongoing effects of maternal hyperglycemia. Diabetes Care. 2007;30(9):2287-2292. doi: 10.2337/dc06-2361 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement.

eTable 1. Baseline Characteristics of Analytic Sample Compared to Children Excluded for Missing BMI

eTable 2. Tested Models and Fit Indices

eTable 3. Random and Fixed Effects Models by Cohort, Hybrid Model 2b by Cohort, and Hybrid Model 5 by Cohort With Lagged BMID

eMethods.

eTable 4. Child, Family, and School Characteristics of ECLS-K:2011 and ECLS-K:1999 Low-Income Child Samples

eTable 5. Deciles of Numerical Variables of ECLS-K:2011 and ECLS-K:1999 Low-Income Child Samples

eFigure 1. Path Estimates of Model

eTable 6. Performance Characteristics of Proposed Full and Simple Imputation Approaches for Predicting Grade 1 Full or Free or Reduced-Price NSLP Participation

eFigure 2. Path Estimates of Sensitivity Model Using BMI z Score

eTable 7. Free or Reduced-Price NSLP Estimates in Unweighted Sensitivity Models

eReferences

Click here for additional data file. (643.5KB, pdf)

Articles from JAMA Network Open are provided here courtesy of American Medical Association

RESOURCES