Klish, William J.*; Karavias, Kellie E.‡; White, Katie S.‡; Balch, Angela J.‡; Fraley, J. Kennard*; Mikhail, Carmen*; Abrams, Stephanie H.*; Terrazas, Norma L.†; Smith, E. O’Brian*; Wong, William W.*
*Department of Pediatrics, Baylor College of Medicine
†Texas Children's Hospital
‡Sylvan Rodriguez Elementary School, Houston, TX.
Address correspondence and reprint requests to William W. Wong, PhD, USDA/ARS Children's Nutrition Research Center, 1100 Bates St, Houston, TX 77030 (e-mail: firstname.lastname@example.org).
Received 15 June, 2011
Accepted 5 August, 2011
The project was entirely supported by the Houston Independent School District.
The authors report no conflicts of interest.
Childhood obesity has increased dramatically in the United States during the last 2 decades (1). School-based initiatives directed at either increasing exercise or mandating a healthy diet have been the most popular approach toward a solution to the problem. These approaches are logical given that most children in America attend school and are in essence a captive audience. The meta-analysis done by Katz et al (2) implies that weight reduction can be achieved through a school-based childhood obesity intervention initiative; however, studies done by Harris et al (3) and Dobbins et al (4) contend that school-based initiatives have no effect on weight or body composition. We report the results of a school-initiated comprehensive initiative at a large elementary school in Houston, Texas, with a huge Hispanic population.
PATIENTS AND METHODS
The schools involved in the present study are located in a primarily Hispanic, low-income area of Houston. The intervention school had an enrollment of 1022 students, all of whom lived in densely populated apartment complexes with little or no access to safe play areas. The control school had an enrollment of 692 children with similar demographics and was in the same general area of the intervention school. No interventions related to healthy lifestyle were initiated in the control school during the study period. The intervention school staff recruited the control school for the study and collected all of the outcome data. The faculty at Baylor College of Medicine (BCM) provided the stadiometers and digital scales for the measurements and the training on proper measurement techniques. None of the faculty and staff at BCM and Texas Children's Hospital was involved in the recruitment, the implementation, and the measurements. The study was approved by the institutional review board for human studies at BCM.
All of the students in both schools had their height and weight measured in duplicate at the beginning and the end of the school year. Children were classified into the following categories: underweight (body mass index [BMI] <5th percentile), normal weight (5th percentile ≤ BMI ≤85th percentile), overweight (85th percentile < BMI <95th percentile), and obese (BMI ≥95th percentile).
At the intervention school, the program was initiated with a lecture to all of the teaching personnel in the intervention school by one of our team (W.J.K.) about the increasing prevalence of childhood obesity and the medical consequences. This was followed by a presentation of the initial BMI data collected at the intervention school. The data were distributed to the appropriate teachers and shared with the parents. During an open house at the beginning of school, flyers were distributed that discussed, in English and Spanish, healthy snack choices, healthy food suggestions for meals, and physical fitness activities for parents and their children. Obese children were invited with their parents to participate in a 1-hour after-school behavior modification program that offered 30 minutes of dietary instruction and behavioral therapy and 30 minutes of physical activity. The Child and Adolescent Trial for Cardiovascular Health program (5) was introduced through student-generated billboards of large-size foods displayed in each category, as well as classroom sorting games and activities to teach children about healthy eating habits by categorizing foods into 3 groups: GO, SLOW, and WHOA foods. The 5-A-Day Friday, Recapture Recess, and GO-SLOW-WHOA foods were incorporated in an ancillary classroom. Five-A-Day Friday is a schoolwide weekly event in which staff, students, and faculty are encouraged to drop everything at the principal's announcement and enjoy a serving of their favorite fruit or vegetable. Additionally, live announcements and Dole 5-A-Day commercial clips were shown weekly on our in-house broadcasting system. For the Recapture Recess program, Child and Adolescent Trial for Cardiovascular Health fitness equipment was provided at the beginning of the school year for each grade level. The equipment included jump ropes, various outdoor balls, team vests, Frisbees, and cones to promote active recess time. The 4th- and 5th-grade students also had the opportunity to participate in an after-school program, Culinary Classroom and Gourmet Garden for the 21st Century. The goal of this program was to provide students with an understanding of the role of food for life and how the natural world sustains us. It also attempts to help students improve their own nutrition, develop healthy eating habits, and become culinary risk takers by trying and tasting new foods. The 4th-grade students also participated in a chef-in-school program provided by a nonprofit organization, Recipe for Success. In this program, a professional chef comes to the school once per month to teach students during their regular class time how to create healthy meals. This group also has field experiences such as visiting a farm, shopping at a large supermarket, and visiting a restaurant to view the kitchen facilities and preparing a healthy meal. The in-house school broadcast system also promoted schoolwide nutrition messages weekly. All of the third- to fifth-grade students viewed videos that discussed healthy eating and participated in hands-on nutrition label reading exercises. Aramark, the commercial provider of the school lunch program, was petitioned to modify a-la-cart lunchroom items to make them healthier and provide fruits and vegetables at least 2 to 3 times per week in the after-school snack program. The after-school program also provided soccer, cheerleading, dance, and karate as a way to increase physical activity not available during regular school hours.
An independent samples t test was used to compare the means of the continuous variables and χ2 or Fisher exact test to compare proportions of the categorical variables for the baseline characteristics of the participants within the control and intervention schools. Binary logistic regression analysis was used to compare control and intervention groups with respect to proportion (prevalence, incidence, and remission) of overweight children and the proportion of obese children while adjusting for ethnicity and sex. The adjusted odds ratio was calculated to quantify the effect of the intervention relative to the control school. Because the elementary school students represented a heterogeneous group of children from a behavioral and developmental perspective, a hierarchical statistical analysis of BMI outcome by classroom teachers was performed while adjusting for potential differences in grade, sex, race, age, and baseline BMI.
Of the 1022 children in the intervention school, 779 had measurements at the beginning and the end of the school year and were included in the statistical analyses. The control school had a total enrollment of 692, and 510 children had both measurements. Transfers, parents moved, or absenteeism were the reasons for the reduced participants at the end of the school year. The incomplete measurement rates were similar between the intervention school (23.8%) and control school (26.3%). As shown in Table 1, the baseline characteristics of participants within the 2 schools were essentially equivalent. Hispanics, mostly of Mexican descent, were the predominant group at both schools. Weight distribution, BMI, and BMI z scores were also identical between the 2 schools. The difference in the prevalence of underweight between the schools could be ignored because of the small sample size. The prevalence of obesity in both schools at the beginning of the school year was >30%, which is much higher than the national average (6).
Table 2 shows the prevalence, incidence, and remission of overweight and obesity at the beginning and the end of the 9-month-long school year for those students who had paired observations. There was no difference in the prevalence of overweight at the end of the study period in both schools. The same was true with the prevalence of obesity. The total number of overweight and obese children did not change during the course of the school year within the control school starting. The intervention school started at 384 obese and overweight students and ended with 378. The percentage of students who became overweight during the course of the school year was less (P = 0.02) in the intervention group (4.1%) than in the control group (8.9%). Because the sample (n) used to calculate this number only represents those children who were normal at the beginning of the study, however, the children who moved from the obese to the overweight group were not included. This is important because the prevalence of overweight at follow-up was not statistically different. More students became obese in the intervention group than in the control group (7.7% vs 4.2%, respectively, P = 0.04) during the course of the school year. The remission data showed that the intervention school had a slight increase in the number of students going into remission than the control school, but the differences are not statistically significant.
As multiple classroom teachers (caregivers) and various programs were implemented across a varied student body, a hierarchical statistical analysis of BMI outcome by classroom teachers was performed while adjusting for potential differences in grade, sex, race, age, and baseline BMI. The analysis again showed no significant difference in BMI outcome (P = 0.22) between the 2 schools. With the fourth graders receiving an additional chef-in-school program, their BMI outcomes may have distorted the results. A hierarchical statistical analysis of BMI outcome by classroom teachers among the fourth graders while adjusting for potential differences in sex, race, age, and baseline BMI also showed no significant difference in BMI outcome (P = 0.14) between the 2 schools.
In spite of the fact that the teaching staff at the intervention school devoted significant time and effort to design a program to change the environment in the school and offer students every chance possible to develop a healthy lifestyle, there was no change in the average weight at the intervention school. Foster et al (7) described a successful school-based intervention to prevent overweight and obesity. Their program was similar to ours with several exceptions. Participants in their program were selected based on the parent's willingness to sign consent for their child to participate, and the largest segment of their student population was black. The program presented here was initiated by the school, so all of the students participated and most of the students were Hispanic. Perhaps signing consent signals a parent's willingness to participate and help their child make lifestyle changes. If this is true, then mandated school-based initiatives based on political legislation may not be successful unless the parents themselves lobby for the change. It is also possible that the Hispanic population that participated in our study was more resistant to change or was uncomfortable with participation because of language or cultural differences. In any case, the definition of success in the Foster et al study was weak in that the authors only saw a significant difference in the incidence of overweight just as we did. They reported an increase in the number of children who went from normal to overweight in the control schools but no difference in the numbers that went from overweight to normal. Their observations more likely were because they analyzed their data like us, and left out the children who moved between the obese and overweight categories.
There are limitations to the present study. We were able to carry on the study for a period of only 1 school year. It is possible that we could have seen changes in the second or third year; however, doing a longitudinal study that studies individual children in a school such as the 2 studied is extremely difficult. First, the dropout rate in the population studied is high. The fifth-grade students will graduate and move on to other schools, making them difficult to study. All of these changes diminish the power of the study and probably explain why there are few long-term longitudinal school-based obesity intervention studies.
The population studied was primarily of Hispanic origin and underprivileged. Perhaps data from this group cannot be generalized into other racial, ethnic, or socioeconomic groups. We studied this group because these children have some of the highest rates of obesity in the nation (8). Cognitive and behavior changes were not studied. There could well have been positive changes in these areas. The ultimate goal of any obesity prevention or treatment program is to achieve weight loss. For this reason, we believed monitoring weight should be the principal aim of the study.
As elementary school students represent a heterogeneous group of children from a behavioral and developmental perspective and the additional intervention program such as the chef-in-school program being offered to fourth graders may have distorted the outcomes, hierarchical statistical analyses on BMI outcome by classroom teachers while adjusting for potential differences in grade, sex, race, age, and baseline BMI were performed on all of the students and on the fourth graders alone. The results again showed no significant differences in BMI outcomes between the control and intervention schools.
The after-school program being offered to the obese children also attracted few participants (10 morbidly obese children from the third, fourth, and fifth grades) with no parental participants. At the end of the school year, only 5 of these children completed the program. The lack of participation further illustrates the difficulty of implementing after-school programs targeting obese children because they do not want to be singled out. The lack of parental participation is not surprising because both parents may be working to make ends meet or must stay home to take care of the younger siblings.
The possibility that students from the control school may have adopted some of the intervention programs could have distorted the outcomes. Although the 2 schools are approximately 1 mi apart, they are separated by 1 busy major street and another smaller street. Because the families who sent their children to these schools lived in apartment complexes within walking distance to the schools, the chance of cross-contamination was not likely. Furthermore, the teachers who performed the assessments at both schools did not observe any similar intervention programs at the control school.
We concluded from this exploratory study that in this population, a multicomponent school-based behavioral intervention program that incorporated environmental change, curriculum change, and engagement of parents for the prevention and treatment of childhood obesity may require further evaluation.
1. Ogden CL, Carroll MD, Curtin LR, et al. Prevalence of high body mass index in US children and adolescents, 2007–2008. JAMA 2010; 303:242–249.
2. Katz DL, O’Connell M, Njike VY, et al. Strategies for the prevention and control of obesity in the school setting: systematic review and meta-analysis. Int J Obes (Lond) 2008; 32:1780–1789.
3. Harris KC, Kuramoto LK, Schulzer M, et al. Effect of school-based physical activity interventions on body mass index in children: a meta-analysis. CMAJ 2009; 180:719–726.
4. Dobbins M, De CK, Robeson P, et al. School-based physical activity programs for promoting physical activity and fitness in children and adolescents aged 6-18. Cochrane Database Syst Rev 2009; 1:CD007651.
5. Luepker RV, Perry CL, McKinlay SM, et al. Outcomes of a field trial to improve children's dietary patterns and physical activity. The Child and Adolescent Trial for Cardiovascular Health. JAMA 1996;768–76.
6. Ogden CL, Carroll MD, Flegal KM. High body mass index for age among US children and adolescents, 2003-2006. JAMA 2008; 299:2401–2405.
7. Foster GD, Sherman S, Borradaile KE, et al. A policy-based school intervention to prevent overweight and obesity. Pediatrics 2008; 121:e794–e802.
8. Preventing Childhood Obesity: Health in the Balance. Washington, DC: Institute of Medicine; 2004.