SEARCH

SEARCH BY CITATION

Keywords:

  • childhood obesity;
  • overweight;
  • body mass index;
  • nutritional education;
  • exercise;
  • BMI percentile;
  • BMI gain

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Design and Methods
  5. Results
  6. Discussion
  7. References

Objective

Study the effects of multiple exercise and nutritional educational interventions on adverse body mass index (BMI) gain of BMI sub-groups of predominately Hispanic six through eight year-old children at high risk for obesity.

Methods

BMI and demographic data were recorded at baseline and six months later in 749 first and second grade public school children at four elementary schools. Two schools (intervention group) received 150 min of extra physical education classes, weekly cooking classes, a structured nutritional curriculum, and parental counseling. BMI changes were calculated for each student and compared by BMI percentile subgroups using the two tailed T-test.

Results

No statistical BMI differences occurred between intervention and control group children below the 25th percentile. Significance differences in BMI gain were noted from the 25th to the 50th percentile (P = 0.027), 50th-75th percentile (P = 0.045), and 75th-95th percentile (P = 0.00007), but not for the 95th-98th percentile (P = 0.288), 98th and above (P = 0.223), or both obese groups combined (P = 0.085).

Conclusions

Nutritional education and exercise can prevent but not treat obesity in predominately Hispanic first and second grade children. BMI subgroups should be studied to avoid masking differing outcomes of obese and nonobese children.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Design and Methods
  5. Results
  6. Discussion
  7. References

Childhood Obesity now affects 19.6% of all children aged 6 through 11 [1]. Given the association of childhood obesity with adult obesity [2-4] and insulin resistance [5, 6], early intervention may decrease the considerable added disease burden associated with adult obesity [7-9].

Among children of Mexican-American descent, 22.3% of all girls and 27.1% of all boys now meet criteria for obesity [body mass index (BMI) at or exceeding the 95th percentile]. Multiple factors have been associated with higher rates of obesity. Hispanics may harbor a cultural perception that fat confers an element of health, thereby conflicting with medical advice [10]. More sedentary activities and fewer strenuous outdoor activities have contributed to markedly less energy expenditure of children as compared with 40 years ago [11]. Parental obesity doubles the risk of adult obesity among both obese and nonobese children under the age of 10 [12]. Excessive TV or screen time [13], inadequate neighborhood activities [14], or inadequate household environment [15] play a role in establishing long-term behaviors that elevate obesity rates. Low income, nonwhite racial status [16-18], and a higher rate of adult Hispanic obesity are three factors with public health implications for the changing demographics in the United States.

Single focus school attempts to influence childhood obesity rates have been modest in their ability to influence obesity rates. Exercise intervention programs at schools may fail in part because of compensatory inactivity of the intervention groups after leaving the daytime school environment [19]. The 2005 elimination of sugar-sweetened soft drinks, high fat foods, added sugar, and high per-item calorie foods in school meals by the State of California has shown no decrease in childhood obesity [20, 21]. Rates continued to climb until 2012, where they have at least temporarily plateaued [22]. An 18 month study of healthy weight children that gave sugar-sweetened drinks to one group every day, while denying them to another group, yielded an average weight difference of 1.02 Kg between the two groups [23]. Although 1.02 Kg was statistically significant, the similarity of weight gain between the two groups suggests a compensatory reduction of caloric intake in the sugar-sweetened drink group that moderated the clinical significance of this single focus intervention. In a separate randomized trial in adolescents that limited sugar-sweetened drinks over a one-year period, the effect of decreased BMI in the intervention group was present at the one-year mark, but absent at the two-year follow-up [24].

The primary objectives of this study were:

  1. To evaluate the efficacy of multiple interventions on children from six to eight years of age as compared to children without these interventions.
  2. To measure the responses to these interventions as a function of the degree of body habitus, as indicated by BMI sub-groups.

The primary outcome measured was the change in BMI over the six-month study period for each subgroup of BMI in the intervention group and the control group. It was hypothesized that multiple interventions would result in BMI reductions over the entire spectrum of overweight and obese children.

Palm Springs Unified School District (PSUSD) has a fairly homogeneous population of low income families, with most children receiving free or reduced fee school meals. Breakfast is served daily to all children in the classroom after the start of the school day, and lunch is served to all in the cafeteria. Participation in the PSUSD nutrition program is universal, with meal content and nutritional education closely managed by the program nutritionists.

Design and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Design and Methods
  5. Results
  6. Discussion
  7. References

First and second grade students in four elementary schools were studied. All four schools are participants in year-round education. All students of the four campuses received diets in accordance with 2005 California State mandates for reduced sugar, calories, and fat. Participation of students in the breakfast and lunch program was 100% at all schools. At two of the campuses (the intervention group), additional nutritional, motivational, educational, and physical interventions were initiated. These were not available to the first and second graders of the other two schools (control group).

The two intervention schools received the following additional interventions:

  1. Daily recess time (which was previously free time) was structured into organized play two days per week to increase exercise.
  2. One extra physical education class was given daily for a total of 150 extra minutes of exercise per week. Classes were conducted by physical education specialists and included activities such as relay races and obstacle course games.
  3. Children who stayed on-campus after school (∼50% of the students) received two 30-min cooking classes per week. Topics were taken from the 36 unit curriculum of the Harvest of the Month [25] program. A different produce item was selected and multiple recipes were made from that item each class.
  4. Assemblies were held every three months that exposed the children to nutritional and recreational topics with interactive productions based upon the Kaiser Permanente Amazing Food Detective [26] series.
  5. Nutritional education in the classroom consisted of weekly 45 min courses using the Dairy Council, [27] Team Nutrition, [28] and My Plate [29] curricula.
  6. Teachers received two seminars (one every three months) where health statistics and nutritional curricula [such as Rethink your Drink [30]] were introduced.
  7. Administrators received monthly 15 min nutritional presentations to highlight or update the curricular modifications that were occurring.
  8. A “Chef in the Classroom” program occurred bi-monthly in each class. Children learned to cook healthy foods while paying attention to the amount of fat, sodium, sugar, and salt on food labels.
  9. The one-month intersessions between quarters contained one or two week-long “Health Camps.” These consisted of 4 h of instruction from the Dairy Council, Harvest of the Month, or USDA Team Nutrition curricula, in addition to inclusion in games that required exercise.
  10. Parents of children who exceeded the 85th BMI percentile received one-on-one counseling twice during the six-month study. Parents were asked to recall the prior 24 h of food preparation. Other topics included learning about the importance of proper sleep time, selecting the right produce, the importance of nonsweetened beverages, reducing screen time to <2 h, and increasing daily activities. For parents without transportation, these were held at the family's home.
  11. All parents were requested to attend monthly group activities regarding nutrition, obesity, and physical activity. Family cooking night attendance was the most popular, with attendance of ∼200 persons in each of the schools. The audience was divided into four groups for instruction in cooking healthy with food provided by the school. Each group was assigned to prepare a dish, which was shared with all in potpourri style. Less popular, with attendance at 40–100, were the monthly “Bring a Parent to Breakfast” events. The Parent Cooking Club and Zumba dance classes (attendance 10–15) were less-well attended.

The BMI of each student was recorded in April, 2011, and at the end of the study six months later. Other data collected at the onset included racial background, gender, school site, and status for inclusion in the state of California free or reduced meal program. First and second graders at two of the schools received additional interventions, as detailed above. Of the initial 749 students measured, only the 561 students who were present at both weighing's were included in the study. The BMI was calculated using the standard formula of kilograms of weight divided by the square of the height, as measured in meters. The initial and follow-up BMIs were recorded for each student.

Student data was divided into quartile subgroups of increasing body habitus, based upon the student's entry BMI percentile. The highest quartile was further subdivided to allow further evaluation of nearly obese students (75th percentile up to but not including the 95th percentile), obese students (95th percentile up to but not including the 98th percentile), and highly obese students (equal to or greater than the 98th percentile). The selection of BMI subgroups was guided by the percentiles published by the National Center for Health Statistics [31, 32].

The changes in BMI over the six-month period were compiled for intervention and control students in each of the six BMI percentile subgroups. The BMI changes were compared using the two tailed T-test. The mean BMI changes for each subgroup were then graphed to allow a visual comparison of the changes for each BMI subgroup of the control group and of the intervention group.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Design and Methods
  5. Results
  6. Discussion
  7. References

With the exception of the percentage of children receiving subsidized meals, there were no baseline differences between the control and intervention groups (Table 1). The control group's meal subsidy rate was slightly higher (99 vs. 93%). Girls and boys had a similar incidence of obesity (Figure 1).

Table 1. Study group characteristics
 ControlInterventionP-Value
Number of Students289 (100%)272 (100%) 
Hispanic234 (83%)218 (80%)P =0.377
Meal subsidy269 (93%)269 (99%)P = 0.0004
Overweight (85th − <95%)49 (17%)46 (17%)P = 0.989
Obese (> = 95%)64 (22%)73 (27%)P = 0.198
image

Figure 1. BMI Percentile by Gender.

Download figure to PowerPoint

After six months of intervention, no statistical difference in BMI between the intervention and control groups was demonstrated (P = 0.64) for children below the 25th percentile of BMI (Table 2). From the 25th to the 95th percentiles, all subgroups of children in the intervention groups demonstrated significantly less BMI gain than those of the control groups. For the 75th-95th percentiles, a high risk subgroup that includes the overweight and nearly overweight children, the intervention six-month mean gain was 0.074 Kg/m2 as compared with the control group gain of 0.685 Kg/m2 (P = 0.00007).

Table 2. BMI subgroup changes over 6 months
Control subgroupsIntervention subgroups
BMI percentileNumber of studentsMean BMI change in 6 months (Kg/M2)Number of studentsMean BMI change in 6 months (Kg/M2)Two-tailed P value
<2536−0.02935−0.0900.64
25 to <5054+0.13331−0.1610.027
50 to <7559+0.10952−0.1410.045
75 to <9580+0.68582+0.0740.00007
95 to <9830+0.8127+0.5000.288
Over 9830+0.75745+0.4650.223
95-98 + over 9860+0.7872+0.480.085

The interventions on the obese and highly obese subgroups (95th-98th percentiles and over 98th percentiles, respectively) did not reach statistical superiority to those of the control groups for either the obese (P = 0.288), or highly obese (P = 0.233). Combining both obese subgroups together yielded a nonsignificant trend towards similarity (P = 0.085).

The BMI subgroup changes are displayed graphically for both control and intervention groups (Figure 2). Included for reference is the baseline plot for normal BMI sub-group changes derived from CDC growth charts. Between the 25th and 95th percentiles, the intervention yielded consistent reductions in BMI gain. Between the 75th and 95th percentiles, the separation between the studied groups is the greatest. Above the 95th percentile, there are no statistical differences between the groups, although a trend is noted.

image

Figure 2. Differences from the expected BMI growth by BMI percentile subgroups. Data Points represent the midline of the measured groups (the mean of the first through 25th percentile subjects are represented by the data point at 12.5, 25th-50th is represented by the data point at 37.5, etc.).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Design and Methods
  5. Results
  6. Discussion
  7. References

As children grow and naturally increase their BMI, some will ascend to unhealthy BMI percentiles in future years. The interventions demonstrated the greatest relative reduction in BMI growth in the children at greatest risk for obesity (75th-95th percentile). By contrast, those children already above the 95th percentile at entry to the study were relatively resistant to the same interventions.

For those children under the 25th percentile, it can be said that these interventions caused neither BMI gain nor loss. Importantly, there was no evidence that thin children became thinner (P = 0.64). For those between the 25th and 75th percentiles, there was a decrease in the rate of BMI growth (P = 0.027 and P = 0.045). While a decrease in BMI for children in the 50th through 75th percentiles can be construed as good prevention for possible future obesity, the significant drop seen in children of the 25th through 50th percentiles should be viewed as unintended and potentially unwanted. For children between the 75th and 95th percentiles, the suppression of BMI elevation was the most pronounced (P = 0.00007). Lastly, for children above the 95th and 98th percentiles, the interventions failed to significantly alter the expected rise in BMI and showed no significant differences from the control group (P = 0.288 and 0.233).

This inability to limit BMI gain in obese children aged six to eight shows similarities to the results noted with obese type 2 diabetic children aged 10–17, where intense education plus over 200 min of weekly exercise demonstrated only a nonsignificant improvement in glycemic control, but no improvement in obesity [33].

That this subgroup of obese first and second graders might represent a genotype or phenotype progressing through the natural history of childhood insulin resistance as it progresses to type 2 diabetes is worthy of further investigation. Given the considerable financial and human costs of adult obesity and diabetes, it follows that interventions to avoid impending insulin resistance associated with obesity [34] are a public health imperative. The durability of these school-imposed dietary, educational, behavioral, and motivational modifications upon the eventual BMI of nonobese children is unknown but deserving of further study.

Limitations of the study include the inability to obtain a six-month follow-up on 188 of the initial 749 eligible children due to a high transiency rate in the studied population. Whether the response to education is different in less transient populations is not known. The inability to control off-campus eating, measure differences in off-campus activities, account for different individual metabolic rates, or any combination of the above are unmeasured variables. Fifty percent of the children at all schools stayed after school and therefore received slightly more intervention than others. The characteristics of the after-school group were not recorded for comparison to those children who went home after school. Lastly, low numbers of underweight children (BMI under the fifth percentile) in the study preclude conclusions regarding possible effects on this subgroup.

Interventions that include augmented physical activity, nutritional education, and motivational interventions in the school setting using standardized guidelines and fixed diets are effective in decreasing BMI gain over a six-month period for first and second grade children for whom the presenting BMI is between the 25th and 95th percentiles. Children in the 75th-95th percentiles have the most robust reduction in BMI gain. These interventions, however, do not yield the same affirmative results in first and second grade school children who present with a BMI above the 95th percentile. The treatment of obesity in young children will require different interventions than will the prevention of obesity.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Design and Methods
  5. Results
  6. Discussion
  7. References