ORIGINAL

ARTICLES
Predicting Early Emergence of Childhood Obesity in Underserved
Preschoolers
William J. Heerman, MD, MPH1, Evan C. Sommer, BA1, James C. Slaughter, DrPH2, Lauren R. Samuels, PhD2,
Nina C. Martin, EdD3, and Shari L. Barkin, MD, MSHS1

Objective To determine the magnitude of risk of factors that contribute to the emergence of childhood obesity
among low-income minority children.
Study design We conducted a prospective cohort analysis of parent child pairs with children aged 3-5 years
who were nonobese (n = 605 pairs) who participated in a 3-year randomized controlled trial of a healthy lifestyle
behavioral intervention. After baseline, height and weight were measured 5 times over 3 years to calculate body
mass index (BMI) percentiles and classify children as normal, overweight, or obese. Multivariable logistic regression
was used to estimate the odds of obesity after 36 months. Predictors included age, sex, birth weight, gestational
age, months of breastfeeding, ethnicity, baseline child BMI, energy intake, physical activity, food security, parent
baseline BMI, and parental depression.
Results Among this predominantly low-income minority population, 66% (398/605) of children were normal
weight at baseline and 34% (n = 207/605) were overweight. Among normal weight children at baseline, 24% (85/
359) were obese after 36 months; among overweight children at baseline, 55% (n = 103/186) were obese after
36 months. Age at enrollment (OR 2.11, 95% CI 1.64-2.72), child baseline BMI (OR 3.37, 95% CI 2.51-4.54), and
parent baseline BMI (OR for a 6-unit change 1.36, 95% CI 1.09-1.70) were significantly associated with the odds
of becoming obese for children.
Conclusions The combination of child age, parent BMI, and child overweight as predictors of child obesity sug-
gest a paradigm of family-centered obesity prevention beginning in early childhood, emphasizing the relevance of
child overweight as a phenotype highly predictive of child obesity. (J Pediatr 2019;-:1-6).
Trial registration Clinicaltrials.gov: NCT01316653.

A
lthough the overall prevalence of obesity in preschoolers appears to be plateauing,1 there is a continued rise in the prev-
alence of obesity for minority population subgroups.2 Estimates indicate that 25.8% of Latino children and 22.0% of
black children aged 2-17 years have childhood obesity, compared with 14.1% of white children.3 These racial disparities
in obesity prevalence develop before children enter kindergarten4 and are attrib-
utable to a range of risk factors, including genetic, family, and environmental fac-
tors; health behaviors like diet and physical activity; cultural norms; and
1 2
socioeconomic influences.5 Developing strategies to prevent disparities in child- From the Departments of Pediatrics and Biostatistics,
Vanderbilt University Medical Center; and Department 3

hood obesity requires careful examination of how these factors from multiple of Psychology and Human Development, Vanderbilt
University, Nashville, TN
levels of a child’s social ecology contribute to the emergence of childhood obesity Supported by grants (U01 HL103620, U01 HL103561,
during the early periods of child development. National Institutes of Health DK056350) with additional
support from the remaining members of the Childhood
Observational studies have identified a host of early life factors that are asso- Obesity Prevention and Treatment Research (COPTR)
Consortium (U01 HD068890, U01 HL103622, U01
ciated with developing childhood obesity, including birth weight,6,7 breastfeed- HL103629), from the National Heart, Lung, and Blood
8,9 10 Institute, the Eunice Kennedy Shriver National Institute of
ing, and maternal gestational weight gain. In addition, parental feeding Child Health and Development, and the Office of
behaviors and early-life dietary quality have been associated consistently with Behavioral and Social Sciences Research. The content
expressed in this paper is solely the responsibility of the
later childhood obesity. However, despite these well-recognized risk factors, authors and does not necessarily represent the official
views of the National Heart, Lung, and Blood Institute, the
childhood obesity-prevention interventions have demonstrated modest effect Eunice Kennedy Shriver National Institute of Child Health
11
sizes and often disappointing results. Furthermore, relatively little attention and Human Development, the National Institutes of
Health, or the US Department of Health and Human
has been paid to the risk of developing overweight or to the risk factors that pre- Services. The REDCap Database is supported by Na-
tional Center for Advancing Translational Sciences/Na-
dict which overweight children will develop obesity. tional Institutes of Health (UL1 TR000445). W.H. was
supported by the National Heart Lung and Blood Institute
This study used a prospective cohort analysis to determine the magnitude of (K23 HL127104). S.B. was supported by the National
risk among combinations of factors that contribute to the emergence of child- Institute of Diabetes and Digestive and Kidney Diseases
(P30DK092986). The authors declare no conflicts of in-
hood obesity among minority children from low-income families. Multiple terest.
Portions of this study were presented at the Pediatric
Academic Societies annual meeting, May 5-8, 2018,
Toronto, Canada.

BMI Body mass index 0022-3476/$ - see front matter. ª 2019 Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.jpeds.2019.06.031

1
THE JOURNAL OF PEDIATRICS  www.jpeds.com Volume -  - 2019

determinants of childhood obesity were considered, and 1 weekend day (using Nutrition Data System for
including individual, family, and social/community factors. Research software; Nutrition Coordinating Center, Minne-
apolis, Minnesota). Child physical activity was measured us-
Methods ing accelerometry, asking children to wear a GT3X+
accelerometer (ActiGraph, Pensacola, Florida) on their waist
Using a prospective cohort design, we conducted a secondary for 7 consecutive days. Previously validated cut points deter-
analysis of data collected as a part of a randomized clinical mined time spent in moderate and vigorous physical activ-
trial (Growing Right Onto Wellness, clinicaltrials.gov: ity.18 Parent/family characteristics included parent BMI
NCT01316653) implementing a behavioral intervention to calculated from measured height and weight, parent depres-
prevent childhood obesity among low-income families in sion from self-report using the Center for Epidemiologic
Nashville, Tennessee. The full methods of the intervention Studies Depression scale,19 parent race/ethnicity from
have been published previously, and the intervention did parent self-report, and food security from self-report using
not lead to a difference in early child body mass index the 6-item short form of the US Household Food Security
(BMI) trajectory between the intervention and control Survey Module.20
groups at 3 years of follow-up.12,13 The Vanderbilt University The primary outcome was child obesity at 36-month
Medical Center institutional review board approved the follow-up. Data collectors prospectively measured a child’s
study protocol. Bilingual data collectors obtained written height (without shoes, to nearest 0.1 cm, using wall-
informed consent in participants’ language of choice using mounted stadiometers) and weight (to nearest 0.1 kg, using
an enhanced, low-literacy approach.14 research-grade scales). These measurements were used to
Parent child pairs were recruited in Nashville, Tennessee, calculate a child’s BMI, which was then classified as normal
from community centers, doctor’s offices, and other commu- weight (BMI percentile ³50th and <85th); overweight (BMI
nity settings between December 2012 and June 2014. Chil- percentile ³85th and <95th); or obese (BMI ³95th) based
dren were included if their BMI percentile was ³50th and on the Centers for Disease Control and Prevention standard-
<95th based on standardized growth curves developed by ized growth curves.15 Data collectors responsible for
the Centers for Disease Control and Prevention (BMI percen- measuring height and weight were trained, certified, and
tile ³50th and <95th).15 By including children in the upper blinded to study group assignment.
range of normal weight and children with overweight, this in- Multivariable logistic regression models were used to
clusion criterion was chosen to identify children at greatest analyze the association between each of the predictors of in-
risk of developing obesity.16 To be eligible, participants had terest and childhood obesity at 36-month follow-up. Results
to qualify for 1 or more services for underserved populations are presented as ORs with corresponding 95% CIs.
(eg, Special Supplemental Nutrition Program for Women, We considered 4 sequential, nested multivariable models
Infants, and Children, Medicaid). Additional inclusion that contained (1) child baseline characteristics, (2) child base-
criteria consisted of (1) child age between 3 and 5 years, (2) line characteristics and child early life characteristics, (3) child
parent age ³18 years, (3) parents and children spoke English baseline and early life characteristics and child baseline health
or Spanish, (4) commitment to participate in the 36-month behaviors during preschool, or (4) child baseline and early life
study, and (5) consistent phone access. Exclusion criteria characteristics, child baseline health behaviors during pre-
consisted of medical condition precluding routine physical school, and parent/family characteristics. We define early life
activity or footpath (eg, lived or worked) outside a 5-mile characteristics as those factors in the first year of life that
radius of participating community centers. may affect the emergence of later childhood obesity. This
Data on parents and children were collected at baseline, 3, approach facilitates understanding of the unique contributions
9, 12, 24, and 36 months. Exposures for this analysis are of child demographics, child early life predictors, child health
grouped into 4 sets: (1) child baseline characteristics, (2) behaviors, and parent/household predictors. We present the
child early life characteristics (ie, in the first year of life),17 C statistic (area under the receiver operating characteristic
(3) child baseline health behaviors during preschool, and curve) for each of these models to summarize the predictive
(4) parent/family characteristics. Child baseline demo- ability of each model. The C statistic represents the ability of
graphics consisted of child baseline BMI (calculated from the model to discriminate between participants who experi-
measured height and weight), age (years), and sex. Child enced the outcome event (in this case, children who became
early life predictors consisted of birth weight (grams), gesta- obese at 36-month follow-up) and those who did not. Specif-
tional age (weeks), and breastfeeding length (0 months, any ically, it is the probability that the model-predicted risk will be
to 5 months, and 6 months or longer). Child baseline demo- greater for a random participant who experienced the event as
graphics (except BMI) and individual child early life predic- compared with the predicted risk for a random participant
tors were collected by parent report during baseline data who did not experience the event.21,22
collection. Child health behaviors during preschool included Likelihood ratio tests were used to compare model fit be-
mean daily total energy intake (kilocalories) and mean daily tween each pair of models. To test the assumption that allo-
minutes of moderate/vigorous physical activity (adjusting for cation to the intervention condition in the original
mean daily total wear time). Child energy intake was assessed randomized controlled trial had a negligible effect on the
by parent-report using 24-hour diet recalls on 2 weekdays emergence of obesity, we conducted an additional sensitivity
2 Heerman et al
- 2019 ORIGINAL ARTICLES

Table. Descriptive statistics for baseline characteristics for total baseline sample* and by child obesity status at 36-
month follow-up
Baseline sample Children without obesity Children with obesity at
Participant baseline characteristics (N = 605) at 36-month follow-up (n = 357) 36-month follow-up (n = 188)
Child baseline characteristics
Age at anthropometry collection, y 4.3 (0.9) 4.2 (0.9) 4.6 (0.9)
Sex
Male 290 (47.9%) 160 (44.8%) 95 (50.5%)
Female 315 (52.1%) 197 (55.2%) 93 (49.5%)
BMI, kg/m2 16.6 (0.8) 16.5 (0.7) 17.0 (0.7)
BMI category
Normal 398 (65.8%) 274 (76.8%) 85 (45.2%)
Overweight 207 (34.2%) 83 (23.2%) 103 (54.8%)
BMI z score 0.8 (0.5) 0.7 (0.4) 1.0 (0.4)
Child early life characteristics
Birth weight, g (N = 547) 3309.7 (574.6) 3313.0 (545.9) 3302.8 (621.1)
Gestational age, wk (N = 506) 38.9 (2.2) 39.0 (2.2) 38.9 (2.0)
Breastfeeding (N = 603)
None 99 (16.4%) 62 (17.4%) 29 (15.5%)
Any to <6 mo 214 (35.5%) 128 (35.9%) 60 (32.1%)
³6 mo 290 (48.1%) 167 (46.8%) 98 (52.4%)
Child baseline health behaviors during preschool
Mean daily total energy intake, kcal (N = 604) 1194.3 (384.6) 1179.9 (409.1) 1206.0 (340.0)
Mean daily total wear time, min (N = 599) 1010.2 (152.3) 1010.7 (148.2) 1013.8 (159.7)
Mean daily moderate/vigorous physical 85.2 (30.9) 81.7 (30.8) 90.3 (31.8)
activity, min (N = 599)
Parent/family characteristics
BMI, kg/m2 29.6 (5.8) 29.1 (5.8) 30.5 (5.6)
BMI category
Underweight 2 (0.3%) 1 (0.3%) 0 (0.0%)
Normal 119 (19.7%) 85 (23.8%) 26 (13.8%)
Overweight 237 (39.2%) 141 (39.5%) 70 (37.2%)
Obese 247 (40.8%) 130 (36.4%) 92 (48.9%)
Race/ethnicity
Hispanic Mexican 383 (63.3%) 226 (63.3%) 125 (66.5%)
Hispanic non-Mexican 168 (27.8%) 97 (27.2%) 49 (26.1%)
Non-Hispanic 54 (8.9%) 34 (9.5%) 14 (7.4%)
CES-D scale (N = 604) 9.8 (9.0) 9.7 (9.3) 10.2 (9.0)
Food security† (N = 601)
Food secure 346 (57.6%) 214 (60.5%) 96 (51.3%)
Food insecure without hunger 173 (28.8%) 95 (26.8%) 61 (32.6%)
Food insecure with hunger 82 (13.6%) 45 (12.7%) 30 (16.0%)
WIC and/or SNAP participation (N = 601) 525 (87.4%) 310 (87.3%) 161 (86.1%)

CES-D, Center for Epidemiologic Studies Depression; SNAP, Supplemental Nutrition Assistance Program; WIC, Women, Infants, and Children.
Total baseline sample N = 605, 36-month follow-up sample n = 545, and 34.5% (n = 188) were obese at 36-month follow-up.
*Total baseline sample is defined as all children who were either normal weight or overweight at baseline (n = 605), which excludes 5 children who were already obese at baseline.
†Standard 6-item indicator set for classifying households by food security status level. The scale ranges from 0 to 6, with lower scores indicating greater food security.

analysis by including the randomization condition (ie, inter- to account for correlation arising from taking repeated mea-
vention vs control) to the fully adjusted model. There was no surements on the same child over time. Model-based esti-
substantive difference in the model output (data not shown). mates for the probability of child obesity at 36 months
In addition, to understand how specific combinations of were calculated using representative covariate profiles, and
predictors identified in the primary models contributed to the results were plotted across the follow-up period. To illus-
the emergence of child obesity over time, we used generalized trate the effect of parent BMI on the probability of child
estimating equations to fit 2 marginal logistic regression obesity at 36 months, we show the model-based estimates
models with covariate main effects and covariate-by-time in- as a continuous function of parent BMI. Then, to illustrate
teractions. The first model included the main effects of child the combined effect of parent BMI with other predictors
baseline BMI category (overweight vs normal weight) and on the probability of child obesity at 36-month, we selected
parent baseline BMI (continuous) and each of their interac- a parent BMI of 24 kg/m2 (normal weight) and a parent
tions with time to assess whether their respective relation- BMI of 34 kg/m2 (obesity).
ships to child obesity differed over time. Similarly, the For all analyses, multiple imputation (10 imputed data-
second model included child baseline BMI category (over- sets) for missing covariates was performed using the “are-
weight vs normal weight) and child baseline age (continuous) gImpute” function available in the rms package for R. All
and each of their interactions with time. The Huber White analyses were conducted in R, version 3.4.4 (R Foundation
sandwich estimator, clustering on subject identifier, was used for Statistical Computing, Vienna, Austria).

Predicting Early Emergence of Childhood Obesity in Underserved Preschoolers 3

THE JOURNAL OF PEDIATRICS  www.jpeds.com Volume -

Figure 1. Predictors of childhood obesity emergence. The plot displays results from adjusted multivariable logistic regression,
showing ORs and 95% CIs of covariates predicting childhood obesity at 36 months among children who were overweight or
normal weight at baseline. For each covariate, ORs are shown for the comparison (e.g., “Insecure with hunger: Secure”) or
difference (e.g., 1kg/m2). The unit of measure for breastfeeding is reported in months.

with time, indicated that the greatest predicted probability of

Results
childhood obesity occurred among children who were over-
weight at baseline, and the probability was even greater
At baseline, 610 parent child pairs were included in the orig-
for children whose parents had a greater baseline BMI
inal study, 605 of which included a child who was not obese at
(Figure 2). The second model, with baseline child BMI
baseline. At 36-month follow-up, child BMI was available for
category, child baseline age, and each of their interactions
90% (n = 545) of the nonobese original sample. The Table
with time, indicated that the greatest predicted probability
displays baseline data on parent child pairs according to
of childhood obesity occurred, again, among children who
whether the children developed obesity at 36 months. At
were overweight at baseline, and the probability was greater
baseline, as per study design, 66% (n = 398) of children
for children who were older at baseline (Figure 2). Model-
were normal weight and 34% (n = 207) were overweight.
based estimates indicated that a child with normal weight
Just more than one-half of children (52%, n = 315) were
at baseline whose participating parent had obesity had a
female, and 91% (n = 550) were Hispanic. Mean child age
greater probability of having obesity by 36-month follow-
at baseline was 4.3 years, with all child participants between
up (27% probability) compared with a child whose
3 and 5 years. Mean parent BMI at baseline was 29.6 (SD
participating parent was normal weight (18% probability).
5.8), and 91% of parents (n = 551) were Hispanic. Among
This effect appeared to be even more pronounced among
children who were normal weight at baseline and had
children who were overweight at baseline; model-based
follow-up BMI, 24% (n = 85/359) became obese by the 36-
estimates indicated that an overweight child whose parent
month follow-up. In children who were overweight at
had obesity had a 61% probability of having obesity
baseline and had follow-up BMI, 55% (n = 103/186) were
after 36 months, compared with a 49% probability for a
obese by the 36-month follow-up.
child whose parent was normal weight. To illustrate
In the fully adjusted logistic regression model (with base-
the combined effect of parent BMI and child baseline
line, early life, preschool, and parent predictors), child age
obesity status on the predicted probability of child obesity
at enrollment (OR 2.11, 95% CI 1.64-2.72), child baseline
at 36 months, we show model-based estimates for a range
BMI (OR 3.37, 95% CI 2.51-4.54), and parent baseline BMI
of parent BMI values in Appendix 1 (available at www.
(OR 1.36 for a 6-unit change, 95% CI 1.09-1.70) all demon-
jpeds.com).
strated statistically significant increased odds of childhood
To evaluate the added contribution of early life, preschool,
obesity at 36-month follow-up (Figure 1).
and parent/family characteristics, each set of variables was
To evaluate the combined effect of child baseline weight
added sequentially to the base child demographics model
status and other predictors on the emergence of child obesity,
(Appendix 2; available at www.jpeds.com). Appendix 3
we conducted 2 additional models. Results from the first lon-
(available at www.jpeds.com) contains the C statistics for
gitudinal logistic regression model, with baseline child BMI
each of the 4 models. The predictive ability of each
category, parent baseline BMI, and each of their interactions
4 Heerman et al
- 2019 ORIGINAL ARTICLES

later childhood obesity indicated the predicted probability

of a child developing obesity at 36-month follow-up was
61% if the child was overweight and the child’s parent
had obesity (ie, parent BMI 34 kg/m2). By comparison,
the predicted probability of a child developing obesity at
36-month follow-up was 18% when the child was normal
weight at baseline and the child’s parent was normal
weight (ie, BMI 24 kg/m2) at baseline. Based on this abso-
lute difference of 43 percentage points in the predicted
probability of developing early childhood obesity,23 the
importance of treating parents for overweight/obesity
cannot be overstated. In addition, identifying child age
as a significant predictor of later child obesity supports
the growing body of literature that indicates child obesity
prevention efforts should begin before the preschool
period.
Previous literature has identified these specific risk fac-
tors,24 although many published analyses use large popula-
tion datasets with repeated cross-sectional analysis.23,25,26
Moreover, previous studies have not considered the simulta-
neous contribution of these covariates on the emergence of
child obesity.16,17,27-29 In addition, the current literature
has under-reported on data from traditionally under-
Figure 2. Predicted probability of childhood obesity across represented minority children. Our study adds to this litera-
36 months of follow-up. Graphs from 2 separate models using ture with a prospectively collected longitudinal dataset of
generalized estimating equations to fit marginal logistic re-
underserved minority preschool age children, examining
gressions are shown. Top, The first model included the main
the magnitude of combination of risk factors to examine
effects of child baseline BMI category and parent baseline
BMI and each of their interactions with time. Bottom, The emerging child obesity.
second model included child baseline BMI category and child With obesity-prevention efforts proving especially chal-
baseline age and each of their interactions with time. lenging in these specific population subgroups, using epide-
miologic methods to identify the most salient risk factors will
directly inform future clinical management strategies and
research agendas. The current study suggests that clinical
sequential model improves from 0.75 for the base model to practice should shift from the traditional paradigm of treat-
0.78 for the full model. Likelihood ratio test results ing childhood obesity and its comorbidities to include a pre-
comparing model fit for each pair of models also are vention strategy that focuses on preventing the development
presented in Appendix 3. Results indicate that none of the of overweight in early childhood. In addition, our findings
more complex models significantly improved model fit. direct us to a family-centered approach that encourages a
2-generation strategy to reduce childhood obesity, address-
Discussion ing both parental obesity and early childhood overweight,
as major risk factors for childhood obesity. Future research
In this cohort analysis of low-income minority preschoolers should focus on these multilevel influences across the lifespan
who were at risk for obesity but not yet obese, older age at to develop approaches for obesity prevention among low-
cohort enrollment, child overweight, and parent BMI at base- income minority populations.
line were the main predictors of the emergence of childhood This study has several limitations. As in any cohort study,
obesity at 36-month follow-up after we controlled for a there is the possibility of residual or unmeasured confound-
comprehensive set of individual child and family covariates. ing. Some of the measured confounders were based on
This specific combination of child age, parent BMI, and child parental report and therefore subject to recall or social desir-
overweight as predictors of emerging early child obesity sug- ability bias. We had incomplete data on household income,
gest a paradigm of family-centered obesity prevention that as 27% (n = 164) of participants did not know their annual
would need to begin in early childhood, emphasizing the household income. As such, the main measure of socioeco-
clinical relevance of child overweight as a phenotype highly nomic status was Women, Infants, and Children/Supple-
predictive of child obesity. mental Nutrition Assistance Program participation, which
The main contribution of this cohort analysis is to may not account for variability in socioeconomic condition
recognize the magnitude of the risk of childhood obesity among this low-income population.
for the combined predictors of interest. Models including In this cohort analysis of normal weight and overweight
interactions between statistically significant predictors of minority preschoolers living in poverty, the combination of
Predicting Early Emergence of Childhood Obesity in Underserved Preschoolers 5
THE JOURNAL OF PEDIATRICS  www.jpeds.com Volume -

child age at enrollment, child overweight, and parent BMI at 13. Po’e EK, Heerman WJ, Mistry RS, Barkin SL. Growing Right Onto Well-
baseline predicted the greatest probability of the emergence ness (GRoverweight): a family-centered, community-based obesity pre-
vention randomized controlled trial for preschool child-parent pairs.
of childhood obesity at 36-month follow-up. n
Contemp Clin Trials 2013;36:436-49.
14. Heerman WJ, White RO, Barkin SL. Advancing informed consent for
Submitted for publication Mar 6, 2019; last revision received Jun 7, 2019; vulnerable populations. Pediatrics 2015;135:e562-4.
accepted Jun 11, 2019. 15. Ogden CL, Kuczmarski RJ, Flegal KM, Mei Z, Guo S, Wei R, et al. Cen-
Reprint requests: William J. Heerman, MD, MPH, Assistant Professor of ters for Disease Control and Prevention 2000 growth charts for the
Pediatrics, 2146 Belcourt Ave, 2nd Floor, Nashville, TN 37209. E-mail: bill. United States: improvements to the 1977 National Center for Health Sta-
[email protected] tistics version. Pediatrics 2002;109:45-60.
16. Nader PR, O’Brien M, Houts R, Bradley R, Belsky J, Crosnoe R, et al.
Identifying risk for obesity in early childhood. Pediatrics 2006;118:
References e594-601.
1. Ogden CL, Carroll MD, Lawman HG, Fryar CD, Kruszon-Moran D, 17. Bichteler A, Gershoff ET. Identification of children’s BMI trajectories
Kit BK, et al. Trends in obesity prevalence among children and adoles- and prediction from weight gain in infancy. Obesity (Silver Spring)
cents in the United States, 1988-1994 through 2013-2014. JAMA 2018;26:1050-6.
2016;315:2292-9. 18. Butte NF, Wong WW, Lee JS, Adolph AL, Puyau MR, Zakeri IF. Predic-
2. Hales CM, Fryar CD, Carroll MD, Freedman DS, Ogden CL. Trends in tion of energy expenditure and physical activity in preschoolers. Med Sci
obesity and severe obesity prevalence in us youth and adults by sex and Sports Exerc 2014;46:1216-26.
age, 2007-2008 to 2015-2016. JAMA 2018;319:1723-5. 19. Radloff LS. The CES-D scale: a self-report depression scale for research in
3. Hales CM, Carroll MD, Fryar CD, Ogden CL. Prevalence of obesity the general population. Applied Psychol Meas 1997;3.
among adults and youth: United States, 2015-2016. NCHS Data Brief 20. Gulliford MC, Mahabir D, Rocke B. Reliability and validity of a short
2017;(288):1-8. form household food security scale in a Caribbean community. BMC
4. Isong IA, Richmond T, Avendano M, Kawachi I. Racial/ethnic dispar- Public Health 2004;4:22.
ities: a longitudinal study of growth trajectories among us kindergarten 21. Bamber D. The area above the ordinal dominance graph and the area
children. J Racial Ethn Health Disparities 2018;5:875-84. below the receiver operating characteristic graph. J Math Psychol
5. Huang TT, Drewnosksi A, Kumanyika S, Glass TA. A systems-oriented 1975;12:387-415.
multilevel framework for addressing obesity in the 21st century. Prev 22. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver
Chronic Dis 2009;6:A82. operating characteristic (ROC) curve. Radiology 1982;143:29-36.
6. Oken E, Gillman MW. Fetal origins of obesity. Obes Res 2003;11:496- 23. Rouse H, Goudie A, Rettiganti M, Leath K, Riser Q, Thompson J. Prev-
506. alence, patterns, and predictors: a statewide longitudinal study of child-
7. Johnsson IW, Haglund B, Ahlsson F, Gustafsson J. A high birth weight is hood obesity. J Sch Health 2019;89:237-45.
associated with increased risk of type 2 diabetes and obesity. Pediatr 24. Cunningham SA, Kramer MR, Narayan KM. Incidence of childhood
Obes 2015;10:77-83. obesity in the United States. N Engl J Med 2014;370:403-11.
8. Yan J, Liu L, Zhu Y, Huang G, Wang PP. The association between breast- 25. Pan L, McGuire LC, Blanck HM, May-Murriel AL, Grummer-
feeding and childhood obesity: a meta-analysis. BMC Public Health Strawn LM. Racial/ethnic differences in obesity trends among young
2014;14:1267. low-income children. Am J Prev Med 2015;48:570-4.
9. Gibbs BG, Forste R. Socioeconomic status, infant feeding practices and 26. Pan L, Park S, Slayton R, Goodman AB, Blanck HM. Trends in severe
early childhood obesity. Pediatr Obes 2014;9:135-46. obesity among children aged 2 to 4 years enrolled in special supplemental
10. Diesel JC, Eckhardt CL, Day NL, Brooks MM, Arslanian SA, Bodnar LM. nutrition program for women, infants, and children from 2000 to 2014.
Is gestational weight gain associated with offspring obesity at 36 months? JAMA Pediatr 2018;172:232-8.
Pediatr Obes 2015;10:305-10. 27. Pryor LE, Brendgen M, Tremblay RE, Pingault JB, Liu X, Dubois L, et al.
11. Waters E, de Silva-Sanigorski A, Hall BJ, Brown T, Campbell KJ, Gao Y, Early risk factors of overweight developmental trajectories during middle
et al. Interventions for preventing obesity in children. Cochrane Data- childhood. PLoS One 2015;10:e0131231.
base Syst Rev 2011:CD001871. 28. Dubois L, Girard M. Early determinants of overweight at 4.5 years in a
12. Barkin S, Heerman WJ, Sommer EC, Martin NC, Buchowski MS, population-based longitudinal study. Int J Obes (Lond) 2006;30:610-7.
Schlundt D, et al. Effect of a behavioral intervention for underserved 29. Woo Baidal JA, Locks LM, Cheng ER, Blake-Lamb TL, Perkins ME,
preschool-age children on change in body mass index: a randomized Taveras EM. Risk factors for childhood obesity in the first 1,000 days:
clinical trial. JAMA 2018;320:450-60. a systematic review. Am J Prev Med 2016;50:761-79.

6 Heerman et al