Promoting Weight Maintenance among Overweight and Obese Hispanic Children in a Rural Practice

Abstract

US Hispanic children experience a disproportionate burden of overweight and obesity. Comprehensive high-intensity behavioral programs have demonstrated effectiveness in improving weight status among obese children. However, there remains a need to develop more efficient interventions that are feasible in primary care and demonstrate effectiveness in Hispanic children.
The pilot study used a two-group randomized design. Eligible overweight (BMI between the 85th and 94th percentile for age and gender) or obese (BMI ≥95th percentile) Hispanic children and their parents (N=118 child/parent dyads) were recruited from a rural pediatric clinic and randomized to: standard care (SC; n=61 dyads) or behavioral intervention (INT; n=57 dyads). The primary outcomes-weight, waist circumference, and zBMI-were measured at baseline, 2, 6, and 18 weeks. Multivariate logistic regression was used to examine the effect of INT on the likelihood of weight maintenance adjusting for potential confounding variables.
Significantly fewer INT children (68.5%) experienced weight gain, compared to SC children (89.7%; p=0.009). The same pattern was observed for waist circumference, where fewer INT children (44%) experienced an increase in waist circumference, compared to SC children (68.6%; p=0.02). Although a trend of improvement in favor of the INT was observed for zBMI, it was not significant.
This study provides preliminary evidence for the feasibility of a primary-care-based approach to promoting weight maintenance among a high-risk population.

Full text

Promoting Weight Maintenance
among Overweight and Obese Hispanic
Children in a Rural Practice
Deborah Parra-Medina, MPH, PhD,
1
Cynthia Mojica, PhD,
1
Yuanyuan Liang, PhD,
2
Yongjian Ouyang, MSc,
2
Awilda I. Ramos, MD, FAAP,
3
and Ismaela Gomez, DPN, RN, CPNP
4
Abstract
Background: US Hispanic children experience a disproportionate burden of overweight and obesity. Comprehensive high-
intensity behavioral programs have demonstrated effectiveness in improving weight status among obese children. However, there
remains a need to develop more efficient interventions that are feasible in primary care and demonstrate effectiveness in Hispanic
children.
Methods: The pilot study used a two-group randomized design. Eligible overweight (BMI between the 85th and 94th percentile for
age and gender) or obese (BMI ‡95th percentile) Hispanic children and their parents (N=118 child/parent dyads) were recruited from
a rural pediatric clinic and randomized to: standard care (SC; n=61 dyads) or behavioral intervention (INT; n=57 dyads). The primary
outcomes—weight, waist circumference, and zBMI—were measured at baseline, 2, 6, and 18 weeks. Multivariate logistic regression
was used to examine the effect of INT on the likelihood of weight maintenance adjusting for potential confounding variables.
Results: Significantly fewer INT children (68.5%) experienced weight gain, compared to SC children (89.7%; p=0.009). The
same pattern was observed for waist circumference, where fewer INT children (44%) experienced an increase in waist circum-
ference, compared to SC children (68.6%; p=0.02). Although a trend of improvement in favor of the INT was observed for zBMI, it
was not significant.
Conclusions: This study provides preliminary evidence for the feasibility of a primary-care–based approach to promoting weight
maintenance among a high-risk population.
Introduction
Significant ethnic disparities in obesity prevalence
exist among US children. Hispanic children, specif-
ically Mexican Americans, are more likely to be
obese than other racial/ethnic groups.
1,2
A recent review of
childhood obesity prevention interventions targeting His-
panic children identified only two interventions with pos-
itive outcomes, both of which were school based.
3–5
Home/
family and healthcare settings have been increasingly
recognized as important settings in obesity prevention.
6
The primary care setting, in particular, provides the op-
portunity for incorporating health promotion and preven-
tion counseling into routine well-child visits.
7
Currently,
the US Preventive Services Task Force (USPSTF) recom-
mends that clinicians screen children ages 6 and older for
obesity and offer or refer them to comprehensive, intensive
behavioral interventions to promote improvement in
weight status.
8
However, pediatric obesity prevention
programs have not been routinely implemented in primary
care settings.
9,10
Primary care providers often face many
barriers, such as lack of time,
11,12
lack of reimbursement,
9
lack of awareness of community resources, and lack of
training or insufficient knowledge and skills on behavioral
management strategies.
11,12
Further, few obesity preven-
tion studies conducted in pediatric clinics have targeted
minorities or overweight versus obese populations.
In response to these evidence gaps, the Nutrition and
Exercise Start Today (NEST) pilot study evaluated an
evidence-based pediatric obesity management intervention
1
Institute for Health Promotion Research, University of Texas Health Science Center at San Antonio, San Antonio, TX.
2
Department of Epidemiology and Biostatistics, University of Texas Health Science Center at San Antonio, San Antonio, TX.
3
Central Texas Health Research, New Braunfels, TX.
4
New Braunfels Pediatrics Associates, New Braunfels, TX.
CHILDHOOD OBESITY
August 2015 jVolume 11, Number 4
ªMary Ann Liebert, Inc.
DOI: 10.1089/chi.2014.0120
355

for Hispanic children (ages 5–14) and their parent(s) that
could be efficiently delivered in a rural pediatric clinical
practice.
Methods
NEST is a two-group randomized study designed to assess
the effectiveness of a pediatric obesity management inter-
vention among Hispanic children in pediatric clinics. Pedia-
tric health care providers and their clinical staff were trained
to implement a standard care intervention—consistent with
the American Academy of Pediatrics (AAP) established
guidelines
9
—for each participant during a routine clinic visit.
After implementing the standard care intervention, the eli-
gible children were randomized to receive either additional
behavioral intervention components (intervention; INT) or no
additional intervention (standard care; SC). The primary
study outcomes were weight, waist circumference, and
zBMI. Secondary outcomes were sedentary behavior, sugar-
sweetened beverage (SSB) consumption, and fasting glucose
and insulin, which were not the focus of this article. The
study was funded by the Centers for Medicare and Medicaid
Services (CMS030457). The study protocol was approved by
The University of Texas Health Science Center at San An-
tonio (UTHSCSA; San Antonio, TX) Institutional Review
Board before participant recruitment. Participants were en-
rolled from March 2010 to April 2011.
Research Setting and Recruitment
Participants were recruited from a federally funded rural
health clinic in New Braunfels in Comal County, Texas.
New Braunfels has an estimated population of 51,000, with
85% white, 2% black, and 10% Asian, American Indian, or
some other race. Approximately 33% are Hispanic (of any
race)
13
and 13% of the population lives below the federal
poverty level.
Participant eligibility. Participants were eligible for the
trial if they met the following criteria: (1) were Hispanic
based on parent self-report; (2) 5–14 years of age; (3)
overweight (BMI between the 85th and 95th percentile for
age and gender) or obese (BMI ‡95th percentile for age and
gender); (4) one parent or adult caregiver who resided with
the participant had to agree to participate in intervention and
evaluation activities; and (5) parent had access to a tele-
phone. A participant was excluded if he or she had any of
the following: (1) a mental, emotional, or physical handicap
identified by a parent or provider that may interfere with
study participation; (2) a diagnosis of cardiovascular, pul-
monary, or digestive disease; or (3) planning to move from
the local area within the time span of the study.
Recruitment. A clinic-based licensed vocational nurse
trained on the protection of human subjects (e.g., Colla-
borative Institutional Training Initiative) facilitated re-
cruitment (e.g., managed the referral process and helped
schedule clinic appointments). Each week, the nurse used
the electronic medical record (EMR) system and/or
scheduling database to identify children that met the eli-
gibility criteria and had nonurgent medical appointments
scheduled. The nurse generated an EMR point-of-care alert
for clinic staff and healthcare providers to consider refer-
ring the child to the NEST study. At the nonurgent medical
appointment, the nurse explained the study to parents,
provided a recruitment brochure, and obtained a signed
referral from interested parents allowing UTHSCSA re-
search staff, a trained graduate research assistant, to con-
tact them by telephone to explain the study in more detail.
Research staff contacted referred parents by telephone to
obtain verbal consent to conduct eligibility screening,
complete a baseline survey, and schedule the child for their
first visit (visit 1; participants also have three additional
follow-up visits) with the study clinical provider (e.g., a
pediatrician or nurse practitioner trained on the standard
care protocol). During visit 1, the nurse obtained signed
informed consent from the parent and assent from the
child. Vitals (i.e., blood pressure, height, weight, and waist
circumference) were assessed by a trained nurse and lab
work requested (fasting glucose, insulin, and cholesterol). In
addition, the clinical provider delivered the standard care
intervention (brief behavioral counseling and goal setting to
the child and parent). Immediately after visit 1, children
were randomized to SC only or behavioral intervention
(INT). The SC group received brief behavioral counseling
and goal setting from the healthcare provider only. The INT
group received all elements of SC plus a face-to-face
counseling session with a health educator (immediately
after visit 1) and monthly telephone counseling calls and
newsletters through the end of the study. Participants were
randomized using a sequence of 200 random Bernoulli
values generated using Excel without any restriction by the
study statistician (Y.L.), and primary care providers (PCPs),
nurses, and research assistants responsible for data collec-
tion were blind to treatment assignment. The sequence was
concealed until interventions were assigned. Study partici-
pants were notified of their treatment assignment in person
by an intervention research staff member.
During recruitment, we screened 192 parent/child dyads
for eligibility. Of these, 20% were ineligible (n=38), 2%
refused participation (n=4), and 16% did not show up for
visit 1 (n=32); the remaining 61% were randomized
(N=118 child/adult dyads or 236 individual participants).
All study participants were scheduled to receive three
follow-up visits with their clinical provider: at 2 (visit 2), 6
(visit 3), and 18 weeks (visit 4). After completing visit 4,
parents from both groups were contacted by research
staff for a follow-up telephone interview. Six were lost to
follow-up resulting in a retention rate of 95% (6 of 118).
Research Methods
The NEST intervention was based on a Pediatric Obesity
Clinical Toolkit for Healthcare Providers from the Texas
Pediatric Society,
14
that is consistent with Stage 1: Prevention
Plus (healthy lifestyle change) of the AAP established
356 PARRA-MEDINA ET AL.

guidelines for the prevention and treatment of childhood
obesity.
9
The prevention plus stage is appropriate initial
treatment intervention for all overweight and obese children
2–18 years of age.
9
Specific behavioral strategies in the
prevention plus stage include: (1) consumption of five or
more servings of fruits and vegetables per day; (2) decreasing
or eliminating SSBs; (3) limiting screen time to 2 hours or
less a day; and (4) engaging in one or more hours of physical
activity (PA) a day. Additional strategies include: (1) eating
breakfast daily; (2) limiting meals outside the home; (3)
eating family meals at least five times a week; and (4) al-
lowing child to self-regulate his or her meals and (5) avoiding
overly restrictive behaviors. The prevention plus stage re-
quires close follow-up monitoring and can be implemented
by physicians or allied healthcare providers with training in
pediatric weight management or behavioral counseling.
Standard care. Participants in both INT and SC received
a Healthy Lifestyle Prescription (HLP) from their clinical
provider. The HLP lists 11 healthy lifestyle strategies (i.e.,
eat breakfast every day, play outside 1 hour a day) that are
recommended in the prevention plus stage for the pre-
vention and treatment of childhood obesity.
A computerized algorithm was applied to self-reported
behavioral data collected at baseline to identify the most
appropriate healthy lifestyle behavioral strategies for each
family. For example, if the child did not eat breakfast every
day or watched television more than 2 hours daily, then these
behaviors would be identified as behaviors the study par-
ticipant could work on. The recommended behavioral
strategies for each study participant were forwarded to their
clinical provider before visit 1. Research staff placed the
HLP with recommended behavioral strategies in the child’s
medical chart so that the clinical provider could review it
before or during visit 1. Visit 1 consisted of a clinical as-
sessment of vital signs (i.e., height, weight, waist circum-
ference, and blood pressure), a physical exam, ordering lab
tests, and provider counseling using the HLP. For this pro-
ject, checking fasting glucose and fasting insulin levels was
part of standard care for children with a new diagnosis of
excessive weight gain. Based on the health of the child, the
clinical provider ordered additional tests as needed. During
visit 1, the pediatrician reviewed the HLP with the parent/
child and offered suggestions on making prescribed changes.
Participants were prescribed two diet strategies and one PA
strategy. The clinical provider was able to modify the HLP in
consultation with the child and parent, as appropriate. Dur-
ing visit 2, the physician reassessed vital signs, discussed
laboratory test results and the BMI chart, and reviewed
progress on the HLP with the parent/child. During visits 3
and 4, the clinical provider reassessed vital signs, reviewed
the HLP with the parent/child, and (at visit 4 only) reordered
fasting serum glucose and fasting insulin lab tests.
Behavioral intervention. Parents and children assigned to
INT received all elements of SC, plus face-to-face coun-
seling, telephone counseling, and newsletters.
Face-to-face counseling. INT participants received one
30-minute, face-to-face counseling session at the clinic
with a masters-level health educator immediately after
visit 1. The face-to-face counseling session targeted the
family and included at least the child and parent. The
health educator discussed the healthy lifestyle prescription,
outlined the major intervention goals relative to PA and
diet, provided education about PA and diet and current
recommendations, and set short- and long-term PA and
diet goals with participants. Tips for exercising safely and
preventing injury (including warning signs and symptoms)
were provided and participants were referred to commu-
nity resources, as appropriate.
Telephone counseling. INT parents also received monthly
telephone counseling calls (approximately 15 minutes
each) from the health educator. Calls began after the face-
to-face counseling session and continued through visit 4.
The calls were designed to assess current PA levels and
dietary practices relative to the previous face-to-face
meeting or phone call. The health educator addressed
barriers to implementation of healthy lifestyle goals, pro-
vided encouragement and support, and discussed any
topics raised by the parents. Topic-specific tip sheets were
mailed after the call if the health educator deemed the
sheets useful.
Newsletters. INT parents and children also received four
monthly bilingual (English and Spanish) newsletters.
Newsletters, developed by the UTHSCA research team
(D.P.M. and C.M.) using evidence-based health informa-
tion sources (i.e., CDC website), were designed to provide
parents with age-appropriate tips on how to encourage
their child and family to continue following their healthy
lifestyle prescription, such as eating meals together and
engaging in more PA. The newsletters also featured age-
appropriate examples of fun, interactive family activities,
and healthy food and snack recipes.
Measures
Waist circumference, weight, and height were assessed
at baseline, two, six and 18 weeks, concurrent with stan-
dard clinic visits. Self-reported behavior (family/child
nutrition and physical activity) and labs (fasting glucose,
insulin and cholesterol) were assessed at baseline and
immediately post-intervention (18 weeks). To minimize
subject burden, demographic data was assessed at base-
line since it was unlikely to change during the brief study
period.
Child waist circumference (minimum waist girth) was
measured to the nearest 0.5 cm using a Seca
ª
tape measure
midway between the right iliac crests and the lower ribs
when the subject is standing erect with feet together.
15
Child weight was measured (to the nearest 0.1 kg) using a
Seca
ª
digital scale following standard protocol.
16
Child
height (measured to the nearest 0.1 cm) was obtained using
a stadiometer without shoes. We converted BMI (weight
CHILDHOOD OBESITY August 2015 357

Table 1. Participants’ Demographic Characteristics
Standard care Intervention Total
n561 n557 N5118 pvalue
Child
Age 0.30
a
Mean (SD) 9.92 (2.7) 9.4 (2.7) 9.67 (2.7)
Gender, N(%) 0.19
c
Male 28 (45.9) 19 (33.3) 47 (39.8)
Female 33 (54.1) 38 (66.7) 71 (60.2)
Health insurance, N(%) 0.65
c
Medicaid/CHIP 51 (83.6) 44 (78.6) 95 (81.2)
Private 9 (14.8) 10 (17.9) 19 (16.2)
Other 1 (1.6) 2 (3.6) 3 (2.6)
Adult
Age 0.76
b
Mean (SD) 36.5 (13.3) 35.8 (8.5) 36.2 (11.2)
Gender, N(%) 1.00
c
Male 5 (8.2) 4 (7.1) 9 (7.7)
Female 56 (91.8) 52 (92.9) 108 (92.3)
Marital status 0.71
c
Married/living as married 43 (72.9) 43 (76.8) 86 (74.8)
Single 8 (13.6) 9 (16.1) 17 (14.8)
Divorced/separated 7 (11.9) 4 (17.1) 11 (9.6)
Other/widowed 1 (1.7) 0 (0) 1 (0.9)
Education 0.32
d
Less than high school 29 (47.5) 26 (46.4) 55 (47.0)
High school graduate/GED 21 (34.4) 14 (25.0) 35 (29.9)
More than high school 11 (18.0) 16 (28.6) 27 (23.1)
Annual family income 0.42
c
<$10,000 8 (13.1) 8 (14.3) 16 (13.7)
$10,001–$20,000 12 (19.7) 17 (30.4) 29 (24.8)
$20,001–$30,000 16 (26.2) 7 (12.5) 23 (19.7)
$30,001–$40,000 7 (11.5) 9 (16.1) 16 (13.7)
>$40,000 10 (16.4) 7 (12.5) 17 (14.5)
Unknown 8 (13.1) 8 (14.3) 16 (13.7)
Country of origin 0.99
d
United States 31 (50.8) 28 (50.9) 59 (50.9)
Mexico 29 (47.5) 26 (47.3) 55 (47.4)
Guatemala 1 (1.6) 1 (1.8) 2 (1.7)
Years living in United States among
foreign born
Mean (SD) 15.9 (8.2) 15.37 (7.7) 15.6 (7.9) 0.82
a
a
t-test.
b
Mann-Whitney’s U test.
c
Fisher’s exact test.
d
Chi-square test.
SD, standard deviation; CHIP, Children’s Health Insurance Program; GED, General Educational Development.
358 PARRA-MEDINA ET AL.

[kg]/[height {m}
2
]) to standard deviation (SD) scores (or z
scores) and to percentile values using growth charts from
the CDC.
Demographic data were collected from the parent on
parental acculturation (i.e., language preference for read-
ing and speaking, country of birth, and years in the United
States) and socioeconomic status (i.e., parental education,
family income, and child’s health insurance).
Statistical Analysis
The primary hypothesis was that children assigned to the
INT group (compared to the SC group) would demonstrate
a greater proportion of weight maintenance (i.e., a smaller
proportion of weight gain). Descriptive statistics were used
to summarize demographic data collected at baseline. To
determine the effect of the INT on a child’s weight
maintenance, we used summary statistics to evaluate and
compare these outcomes (or changes) between the INT and
SC based on intention-to-treat analysis. For categorical
outcomes, Fisher’s exact test and/or the chi-square test
were used to examine the differences between the two
groups. For continuous outcomes, the t-test or Mann-
Whitney’s U test was used, depending on the distributions
of the outcomes.
For waist circumference, weight, zBMI, and BMI
percentile—measured repeatedly over four time points
(baseline, 2, 6, and 18 weeks)—the trend of change was
defined as a dichotomous variable and measured as fol-
lows. For each child, a linear regression line was fitted
using all available measurements at the four time points.
The trend of change was coded as 1 if the slope of the fitted
regression line was positive in sign and 0 otherwise. The
NEST research team’s clinical experience suggests a 10–
20% dropout for this 18-week study. In this pilot study, no
effect-size estimates were available; we conducted power
analysis assuming that children would continue to gain
weight without treatment. Growth data from the National
Institute of Child Health and Human Development Study
of Early Child Care and Youth Development
17
showed that
80% of children who are overweight at any time during the
elementary school period (ages 7, 9, and 11 years) will be
overweight at age 12 years. So, without any intervention,
we assumed that most overweight children would continue
to gain excess weight as they grow. We estimated that 80%
of children in the SC group would have a positive slope,
indicating weight gain or a failure of weight mainte-
nance.
17
With this assumption, a sample of 160 (80 chil-
dren per group) would achieve 80% power to detect a
minimum 20% difference with a significance level of 0.05
using a two-sided Z test with pooled variance. We would
need to increase the estimate, of weight gain to 90% of
children in the SC group, to detect the same difference with
a sample of 118 (59 children per group).
Multivariate logistic regression was used to examine the
effect of INT on the likelihood of weight maintenance after
adjusting for potential confounding variables (child age,
child gender, annual family income, parent education, and
parent country of origin). Sensitivity analyses were con-
ducted for children with less than two measures of waist
circumference, weight, or zBMI (i.e., missing slope). Two
different imputation approaches were applied: (1) assum-
ing that all missing slopes were positive (i.e., a failure of
Table 2. Participants’ Clinical Characteristics
Standard care Intervention Total
n561 n557 N5118 pvalue
Waist (cm) 0.77
a
Mean (SD) 85.85 (15.97) 84.99 (14.83) 85.44 (15.37)
Waist abnormal, N(%) 0.12
c
No 17 (29.3) 9 (16.7) 26 (23.2)
Yes 41 (70.7) 45 (83.3) 86 (76.8)
BMI 0.65
b
Mean (SD) 26.64 (5.53) 27.15 (5.6) 26.89 (5.55)
BMI percentile 0.1
b
Mean (SD) 97.51 (2.37) 97.89 (2.62) 97.7 (2.49)
BMI percentile abnormal, N(%) 0.40
c
Overweight 9 (14.8) 5 (18.9) 14 (12.0)
Obese 52 (85.2) 51 (91.1) 103 (88)
a
t-test.
b
Mann-Whitney’s U test.
c
Fisher’s exact test.
SD, standard deviation.
CHILDHOOD OBESITY August 2015 359

weight maintenance) and (2) assuming that all missing
slopes were negative or zero (i.e., a success of weight
maintenance). Analyses were performed by a doctoral-
level statistician (Y.L.) and masters-level data analyst
(Y.O.) using SAS software (version 9, 2008; SAS Institute
Inc., Cary, NC).
Results
Participant demographic characteristics, by study group,
are shown in Table 1. Most children were female (60%),
with a mean age of 9.67 years and on Medicaid/Children’s
Health Insurance Program (CHIP) (81.2%). Adult care-
givers were mostly female (92.3%), US born (50.9%), low
income (58.2% had £$30,000 annual family income), and
married (74.8%), with a mean age of 36.2 (SD =11.2).
Clinical characteristics of children at baseline (Table 2)
showed that 88% were obese, mean BMI percentile was
97.7 (SD =2.5), and 76.8% had abnormal waist circum-
ference. There were no significant differences in demo-
graphic or clinical characteristics observed at baseline by
study group
Table 3. Unadjusted Analysis for Waist Circumference, Weight, and zBMI
Standard care Intervention Total
Part I: before imputation N(%) N(%) N(%) pvalue
a
Waist circumference Negative or zero 16 (31.4) 28 (56) 44 (43.6) 0.02
Positive 35 (68.6) 22 (44) 57 (56.4)
Total 51 50 101
Weight Negative or zero 6 (10.3) 17 (31.5) 23 (20.5) 0.009
Positive 52 (89.7) 37 (68.5) 89 (79.5)
Total 58 54 112
zBMI Negative or zero 34 (58.6) 36 (67.9) 70 (63.1) 0.33
Positive 24 (41.4) 17 (32.1) 41 (36.9)
Total 58 53 111
Standard care Intervention Total
Part II: imputation method I
b
N(%) N(%) N(%) pvalue
a
Waist circumference Negative or zero 16 (26.2) 28 (49.1) 44 (37.3) 0.01
Positive 45 (73.8) 29 (50.9) 74 (62.7)
Total 61 57 118
Weight Negative or zero 6 (9.8) 17 (29.8) 23 (19.5) 0.01
Positive 55 (90.2) 40 (70.2) 95 (80.5)
Total 61 57 118
zBMI Negative or zero 34 (55.7) 36 (63.2) 70 (59.3) 0.46
Positive 27 (44.3) 21 (36.8) 48 (40.7)
Total 61 57 118
Standard care Intervention Total
Part III: imputation method II
c
N(%) N(%) N(%) pvalue
a
Waist circumference Negative or zero 26 (42.6) 35 (61.4) 61 (51.7) 0.045
Positive 35 (57.4) 22 (38.6) 57 (48.3)
Total 61 57 118
Weight Negative or zero 9 (14.8) 20 (35.1) 29 (24.6) 0.018
Positive 52 (85.2) 37 (64.9) 89 (75.4)
Total 61 57 118
zBMI Negative or zero 37 (60.7) 40 (70.2) 77 (65.3) 0.335
Positive 24 (39.3) 17 (29.8) 41 (34.7)
Total 61 57 118
a
Fisher’s exact test.
b
All children with missing data experienced weight gain.
c
All children with missing data did not experience weight gain.
360 PARRA-MEDINA ET AL.

Without adjusting for other confounding variables, there
was strong evidence of weight maintenance in favor of the
INT group (see Table 3). Particularly (Table 3: Part I), 44%
of children in the INT group experienced an increase in
waist circumference, compared to 68.6% in the SC group
(p=0.02), and 68.5% of children in the INT group expe-
rienced weight gain, compared to 89.7% in the SC group
(p=0.009). In addition, 32.1% of children in INT group
experienced increase in zBMI, compared to 41.4% in the
SC group ( p=0.33), indicating a trend of improvement in
favor of the INT group. After imputing missing data under
the assumption that all children experienced weight gain
(Table 3: Part II), 50.9% of children in the INT group
experienced an increase in waist circumference, compared
to 73.8% in the SC group ( p=0.01); 70.2% of children in
the INT group experienced weight gain, compared to
90.2% in the SC group ( p=0.01), and 36.8% of children in
the INT group experienced increase zBMI, compared to
44.3% in the SC group ( p=0.46). Similar trends were
observed after imputing missing data under the alternative
assumption that all children did not experience weight gain
(p=0.045, p=0.018, and p=0.335 for waist circumfer-
ence, weight, and zBMI, respectively; Table 3: Part III).
After adjusting for age, gender, income, education, and
country of origin (see Table 4), the odds of weight (or waist
circumference) gain was markedly reduced by 75% for
children in the INT group, compared to children in the SC
group (weight: odds ratio [OR] =0.25; 95% confidence in-
terval [CI] =0.08, 0.75; waist circumference: OR =0.25;
95% CI =0.09, 0.66). After adjusting for income, education,
and country of origin, the odds of BMI z-score increase
was reduced by 22% for children in the INT group,
compared to children in the SC group (OR =0.78; 95%
CI, 0.33–1.84), although it was not statistically significant
(p=0.573). After imputation, similar results were ob-
served hence not reported.
Discussion and Conclusion
Compared to children in SC, children in the INT group
were more likely to experience weight maintenance. Suc-
cessfully maintaining weight of children as they grow will
improve their weight status and can lower their risk for
obesity-related complications, such as high blood pressure
and metabolic syndrome.
The NEST behavioral intervention was designed to meet
current guidelines
9
and was offered to participants at no
cost. We were able to demonstrate improved weight
maintenance among those who received the program using
weight and waist circumference, but not zBMI. The in-
tervention period, however, was brief, lasting only 18
weeks. The behavioral change literature suggests that a
person must engage in a new behavior for 6 months before
it becomes habit. Thus, a longer intervention period and
follow-up from the provider and health educator may have
improved outcomes, maintaining child weight over a
Table 4. Summary of Logistic Regression Analysis
Waist
circumference Weight zBMI
Covariate OR (95% CI) pvalue OR (95% CI) pvalue OR (95% CI) pvalue
Group Standard care 1.00 0.007 1.00 0.013 1.00 0.573
Intervention 0.25 (0.09, 0.66) 0.25 (0.08, 0.75) 0.78 (0.33, 1.84)
Age 1.13 (0.95, 1.37) 0.205 0.98 (0.80, 1.19) 0.823 — —
Gender Female 1.00 0.025 1.00 0.909 — —
Male 0.30 (0.11, 0.86) 1.07 (0.34, 3.36) —
Income <$10k 0.08 (0.01, 0.60) 0.139 0.19 (0.02, 2.42) 0.415 0.60 (0.13, 2.68) 0.264
$10k*20k 0.12 (0.02, 0.71) 0.27 (0.03, 2.73) 0.56 (0.15, 2.11)
$20k*30k 0.08 (0.01, 0.54) 0.49 (0.04, 6.55) 0.82 (0.21, 3.18)
$30k*40k 0.12 (0.02, 0.91) 0.11 (0.01, 1.22) 0.12 (0.02, 0.80)
>$40k 0.06 (0.01, 0.49) 0.13 (0.01, 1.74) 0.25 (0.04, 1.48)
Unknown 1.00 1.00 1.00
Education <High school 1.00 0.479 1.00 0.752 1.00 0.855
HS/GED 0.47 (0.14, 1.60) 1.62 (0.42, 6.28) 1.32 (0.46, 3.78)
>High school 0.64 (0.16, 2.52) 1.04 (0.24, 4.44) 1.02 (0.26, 4.02)
Country
of origin
United States 1.00 0.378 1.00 0.742 1.00 0.469
Other 0.62 (0.21, 1.81) 1.21 (0.38, 3.85) 0.70 (0.26, 1.86)
OR, odds ratio; CI, confidence interval; HS, high school; GED, General Educational Development.
CHILDHOOD OBESITY August 2015 361

longer period and allowing time for growth and observed
shifts in BMI.
PCPs who have regular contact with children and their
parents are in an influential position to monitor and modify
factors that contribute to unhealthy weight gain. Major
barriers for implementation of guidelines, such as lack of
time,
11,12
lack of awareness of community resources, and
insufficient knowledge and skills on behavioral manage-
ment strategies,
11,12
persist. Health educators, as a group,
have not traditionally focused their efforts on the primary
care setting, yet, given their training and skills, are quali-
fied to address many of these barriers. Involvement of
health educators in primary care has the potential not only
to enhance how primary care is delivered, but also to im-
prove health outcomes.
18
Health educators can provide di-
rect delivery of patient education, such as health coaching,
serve as connectors to community resources, and facilitate
evidence-based practice and quality improvement.
19
If im-
plemented properly, involvement of health educators has
potential to enhance how primary care is delivered, improve
the health of people with regard to chronic conditions, and
reduce related healthcare costs.
18
To increase access to effective treatment for childhood
obesity, efforts are needed to accelerate implementation of
the USPSTF guidelines and translate evidence in practice.
New models for delivery of evidence-based preventive
services are needed. Our goal is to develop a replicable
model that could be adopted in primary care settings to
manage their overweight or obese pediatric patients. Re-
sults from this pilot study should be treated as preliminary
and interpreted with caution. That said, the study provides
preliminary evidence for the feasibility of the NEST in-
tervention, including logistical issues, potential effect sizes
for behavioral outcomes, and the cultural suitability and
reliability of measures and the intervention approach. It is
probable that 18 weeks is too short a time to establish the
desired behavior without additional support. For future
studies, we recommend extending the intervention period
to 6 months and adding a maintenance period. The inter-
vention can be strengthened by increasing the frequency
and duration and perhaps incorporating other social media
(Facebook, YouTube, Twitter, Instagram, and mobile
phone text messages) to enhance communication with
parents and older youth. Utilizing developmentally ap-
propriate strategies for children is key to success.
The USPSTF recommends that clinicians offer or refer
overweight or obese children to intensive counseling and
behavioral interventions to promote improvements in
weight status. The NEST program was designed to help
clinicians offer these clinical preventive services to their
patients in a low-cost, flexible approach. This study has
limited generalizability because we focus on one specific
ethnic group from one clinic. Future studies should include
multiple sites and a more diverse sample. Some children in
SC were able to maintain their weight. The pilot, as de-
signed, did not have a true control group. It may be that, for
some patients, brief provider counseling and goal setting is
sufficient to initiate the desired behavior change. Other
children and their families may require the additional
support provided by the NEST intervention. In addition,
there were a substantial proportion of children in INT that
did not respond positively. Further exploration of who
benefitted from the program and why is needed.
Acknowledgments
This project described was funded by the Centers for
Medicare and Medicaid Services (CMS030457) and also
supported by the National Institutes of Health’s National
Center for Research Resources and the National Center
for Advancing Translational Sciences, through Grant
8UL1TR000149. The statements contained in this article
are solely those of the authors and do not necessarily re-
flect the views and policies of the Centers for Medicare
and Medicaid Services or the NIH. The authors are
grateful to all of the families who participated in this
study. The authors also thank the providers and nurses at
New Braunfels Pediatric Associates, Inc., for their par-
ticipation as well as their time and feedback regarding
the study protocols. In addition, the authors acknowledge
Dr. Awilda Ramos for inspiring this work and cham-
pioning our efforts in the clinic and community. Special
thanks to Dr. Richard Bragg, Project Officer at the Cen-
ters for Medicare and Medicaid Services for his support
through the development of the authors’ proposal and
implementation of the study. Finally, the authors acknowl-
edge the substantial contributions of staff who participated
in the NEST program: Sandra Covarrubias, Dorothy Long
Parma, Sima Momin, Jerrie Ricarte, Laura Rubalcava, and
Laura Zepeda.
Author Disclosure Statement
No competing financial interests exist.
References
1. Centers for Disease Control and Prevention (CDC). CDC grand
rounds: Childhood obesity in the United States. MMWR Morb
Mortal Wkly Rep Vol 2011;60:42–46.
2. Krassas G, Tzotzas, T. Do obese children become obese adults:
Childhood predictors of adult disease. Pediatr Endocrinol Rev
2004;1(Suppl 3):455–459.
3. Branscum P, Sharma M. A systematic analysis of childhood obe-
sity prevention interventions targeting Hispanic children: Lessons
learned from the previous decade. Obes Rev 2011;12:e151–e158.
4. Johnston C, Tyler, C, Fullerton, et al. Effects of a school-based
weight maintenance program for Mexican-American children:
Results at 2 years. Obesity (Silver Spring) 2010;18:542–547.
5. Olvera N, Bush, J, Sharma SV, et al. BOUNCE: A community-
based mother-daughter healthy lifestyle intervention for low-in-
come Latino families. Obesity (Silver Spring) 2010;18(Suppl 1):
S102–S104.
362 PARRA-MEDINA ET AL.

6. Birch LL, Ventura AK. Preventing childhood obesity: What
works? Int J Obes (Lond) 2009;33(Suppl 1):S74–S81.
7. Tyler D. Family-centered collaborative negotiation: A model for
facilitating behavior change in primary care. J Am Acad Nurs
Pract 2008;20:194–203.
8. Barton M. Screening for obesity in children and adolescents: US
Preventive Services Task Force recommendation statement. Pe-
diatrics 2010;125:361–367.
9. Spear B. Recommendations for treatment of child and adolescent
overweight and obesity. Pediatrics 2007;120(Suppl 4):S254–S288.
10. Small L. Prevention and early treatment of overweight and obesity
in young children: A critical review and appraisal of the evidence.
Pediatr Nurs 2007;33:149–127.
11. Jelalian E. Survey of physician attitudes and practices related to
pediatric obesity. Clin Pediatr 2003;42:235–245.
12. Story M. Management of child and adolescent obesity: Attitudes,
barriers, skills, and training needs among health care professionals.
Pediatrics 2002;110(1 Pt 2):210–214.
13. US Census Bureau. 2005–2007 American Community Survey. 2000.
Available at http://factfinder.census.gov/servlet/ACSSAFFFacts?_
event=Search&geo_id=&_geoContext=&_street=&_county=new+
braunfels&_cityTown=new +braunfels&_state=04000US48&_zip=&_
lang=en&_sse=on&pctxt=fph&pgsl=010 Last accessed June 10,
2009.
14. Texas Pediatric Society. Pediatric Obesity: A Toolkit for Health-
care Providers. Texas Department of State Health Services: Aus-
tin, TX, 2005.
15. NHLBI Obesity Education Initiative Expert Panel on the Identifi-
cation, Evaluation, and Treatment of Obesity in Adults. Executive
summary of the clinical guidelines on the identification, evalua-
tion, and treatment of overweight and obesity in adults. Arch Intern
Med 1998;158:1855–1867.
16. Heyward VH, Stolarczyk LM. Applied Body Composition As-
sessment. Human Kinectics: Champaign, IL, 1996.
17. Nader PR, O’Brien M, Houts R, et al.; National Institute of Child
Health and Human Development Early Child Care Research Net-
work. Identifying risk for obesity in early childhood. Pediatrics
2006;118:e594–e601.
18. Society for Public Health Education. Affordable Care Act oppor-
tunities and challenges for health education specialists (issue
brief). Society for Public Health Education: Washington, DC, pp.
1–15, 2013.
19. Lorig K. Patient-centered care: Depends on the point of view.
Health Educ Behav 2012;39:523–525.
Address correspondence to:
Deborah Parra-Medina, MPH, PhD
Professor of Epidemiology and Biostatistics
Associate Director for Education
and Training Programs
Institute for Health Promotion Research
University of Texas Health Science Center at San Antonio
7411 John Smith Drive
Suite 1000
San Antonio, TX 78229
E-mail: parramedina@uthscsa.edu
CHILDHOOD OBESITY August 2015 363