To assess the association between vigorous physical activity, participation on sport teams and in exercise programs, television watching, family environment, and weight status in youth.
Cross-sectional data from the Third National Health and Nutrition Examination Survey.
A group of 2791 youth between the ages of 8 and 16 years who were enrolled in the Third National Health and Nutrition Examination Survey.
Main Outcome Measures
Overweight was defined using age- and sex-specific 85th percentile of body mass index using Center for Disease Control and Prevention growth charts.
Both males and females who had an overweight mother and or father were more likely to be overweight compared with youth who did not have an overweight parent. Females who watched 4 or more hours of television were more likely to be overweight than those who watched less than 4 hours. Males and 14- to 16-year-old females who participated in sport team and exercise programs were less likely to be overweight than their counterparts who did not participate. Also, females with larger families and males from families with higher family incomes were less likely to be overweight.
These results suggest that family environment is associated with overweight in youth and that sport and exercise program participants are less like to be overweight and that, for females, increased television watching is related to overweight.
FROM 1976 to 1990, there was a 40% increase in the prevalence of overweight among youth between ages 6 and 17 years.1,2 Using a 95th percentile cutoff for body mass index (BMI), prevalence of overweight had increased from 4% to 11%. Using an 85th percentile cutoff for BMI, it is estimated that 22% of youth aged 6 to 17 years were overweight in 1995 and 25% in 1998.2,3 Being overweight during adolescence is associated with risk factors for cardiovascular disease,4 being an overweight adult,5 and increased risk of all-cause mortality, an increased prevalence of coronary heart disease and other chronic diseases.6,7
The increase of overweight and obesity among youth may be associated with decreased physical activity. Traditional opportunities for physical activity, such as required participation in school physical education at all levels, have been reduced for many youth. The 1995 Youth Risk Behavior Survey (YRBS) reported that 59.6% of surveyed students in grades 9 through 12 participated in physical education classes, but only 48.8% participated in 1997.8,9 The relationship of moderate to vigorous exercise and weight status is open to interpretation. A body of work10- 12 supports the traditional view that physical activity is inversely associated with being overweight, while other studies have failed to support the idea that participation in physical activity is associated with a lower BMI.13,14
Overweight has been associated with an increase in sedentary activities, such as watching television. In the Third National Health and Nutrition Examination Survey (NHANES III) children, interviewed between 1988 and 1994, who reported watching 4 or more hours of television had on average greater BMI (P<.001) than children who reported watching less than 2 hours.15 Possible reasons for the association between television watching and body fatness include the following: television may reduce time spent in activities that require more energy expenditure,16 advertising may influence the selection of high-energy dietary food among youth,17 and a positive relationship exits between watching television and fat intake.18,19 However, the association between watching television and obesity is also inconsistent with some studies failing to find a relationship.10,20
Because of the increased prevalence of overweight among youth and the risk of subsequent chronic disease in adulthood, it is important to determine the correlates of overweight in youth. Environmental factors that may be related to overweight in youth include having overweight parents and other family factors.21- 24 The purpose of this study was to determine what associations exist between overweight in youth and vigorous physical activity, participation on sport teams and in exercise programs, and time spent in a sedentary activity in a triethnic sample of youth aged 8 to 16 years. Environmental factors, including overweight status of mother and father, family size, and poverty index, were also examined.
Study participants were enrolled in the Third National Health and Nutrition Examination Survey (NHANES III). The NHANES was designed to describe the health and nutritional status and estimate the prevalence of common chronic diseases and associated risk factors of the noninstitutionalized civilian population of the United States.25 Children 2 months to 5 years old, persons 60 years or older, non-Hispanic blacks, and Mexican Americans are oversampled to ensure reliable estimates for those groups.
A group of 2791 children, 8 to 16 years of age, interviewed between 1988 to 1994, for whom data were complete, were included in the study. From the original data set of 5052, youth were deleted who did not complete the physical activity questions (n = 951). Other deletions included youth categorized in the "other" race group (n = 184), those diagnosed as having physical conditions that precluded physical activity or its measurement (n = 44), missing BMI data for themselves or their biological parents (n = 917), missing poverty index (n = 161), and time spent watching television (n = 4).
The subject's age used in the study was that obtained during the physical examination held in a mobile examination center.25 Children were divided into 3 age groups: 8 to 10, 11 to 13, and 14 to 16 years. The race or ethnicity of the child was reported by an adult 18 years or older during the home interview, and categorized as non-Hispanic white, non-Hispanic black, and Mexican American. The respondent was the mother in 89% of the interviews and the father in 7% of the interviews. Other respondents included older siblings, grandparents, aunts, and uncles (4%). Family variables included family size (number of family members living at home) and poverty index an indication of socioeconomic status.25,26 Poverty index was expressed as a ratio calculated by the US Census Bureau, based on annual family income, family size, and other economic information. Scores above 1.00 indicate that family income was above the poverty line.
Physical measurements were performed during the physical examination. Standing height, body weight, and skinfolds were measured using standard methods.25 Equipment was calibrated between each measurement. Height and weight of the biologic parents were reported during the home interview.
Body mass index was calculated by dividing weight in kilograms by height in meters squared. Youth were dichotomized into overweight and nonoverweight, using age- and sex-specific 85th percentile of BMI from the Centers for Disease Control and Prevention growth charts.27 Weight status was used as the dependent variable in all analyses, with nonoverweight the referent group. Biological parents were categorized as overweight if their BMI was 25 or greater.28
Physical activity behavior was assessed by asking how many times per week the subjects "played or exercised enough to make them sweat and breathe hard." Youth who reported 3 or more sessions of physical activity were classified as active. Those reporting fewer than 3 sessions were in the low-active group. A question was asked concerning the number of sport teams and exercise programs participated in within the past year. Youth were instructed not to include physical education or gym class. Those who reported no participation on sport teams and in exercise programs were classified in the no-participation group, and those who reported 1 or more sport teams and exercise programs were in the participation group. Youth were also asked how many hours they had watched television the day preceding the examination. Respondents were grouped into 4 or more hours or fewer than 4 hours watched per day.15
We used χ2 analyses to determine differences, by sex, in physical activity behaviors, television watching, and percentage overweight, and t tests to determine differences in age, weight indexes, family size, and poverty index between the weight groups. Crude odds ratios (ORs) and 95% confidence intervals (CIs) were used to examine associations between weight status of the youth and demographic characteristics, vigorous physical activity, participation on sport teams and in exercise programs, television watching, and the mother's and father's weight status. Multiple logistic analysis models were calculated to examine the relationships of the independent variables with weight status. Two-way interaction terms were tested for age and race/ethnicity group by physical activity, sport teams and exercise programs, hours of television watching, and weight status of mother and father. Only interactions significant at P<.01 were considered to minimize the possibility of a type I error. All analyses were conducted separately for males and females.
The complex survey design that was used for NHANES III data collection was incorporated into data analyses by using SUDAAN.29 Sampling weights were incorporated into all analyses. The sampling weights had been calculated to take into account unequal selection probabilities resulting from the cluster design and oversampling of some groups.
Demographic characteristics, physical activity behaviors, and television watching are presented in Table 1. Weighted percentages were not statistically different for sex, age groups, or weight status between the subset (N = 2791) used in this study and those with missing information (N = 2261). There were statistical differences (P<.05) for race/ethnicity, participation in vigorous physical activity, sport teams and exercise programs, and watching television for the subset, compared with those youth who did not meet the inclusion criteria. About 51% of the study sample were male, 7.5% Mexican American, 14.2% non-Hispanic black, and 78.3% non-Hispanic white, and were equally distributed among the age groups. About 30% of the youth were overweight based on age- and sex-specific 85th percentile BMI from the Centers for Disease Control and Prevention growth charts.27 Nearly 81% of the youth reported vigorous physical activity 3 or more times per week and 24% watched 4 or more hours of television, similar to the results reported by Andersen et al.15 About 65% of the youth participated on 1 or more sport teams or in 1 or more exercise programs. There were no differences between males and females regarding weight status, but more males reported participation in 3 or more sessions of vigorous physical activity, participation on sport teams and in exercise programs, and watching 4 or more hours of television (Table 2).
The percentages of overweight youth were 7%, 39%, and 54%, with 0, 1, or 2 overweight parents, respectively (Figure 1). Youth classified as overweight had significantly higher values for BMI, the 3 skinfold measurements, and sum of skinfolds (P<.001) (Table 3). There were no significant differences in mean age between the weight groups.
Percentage of overweight youth (n = 2791) with 0, 1, or 2 overweight parents.
Associations of categorical variables with overweight are shown in Table 4. Among females, non-Hispanic black females were more likely to be overweight compared with non-Hispanic white females. Male and female sport team and exercise program participants during the past year were less likely to be in the overweight group. Females who watched 4 or more hours of television were more likely to be overweight compared with those who watched less. Females with an overweight father were more likely to be overweight, as were those with an overweight mother for both males and females. Family income (poverty index) was significantly higher among those in the nonoverweight group compared with the overweight group for males and females. No other significant associations were observed.
Table 5 shows the results of multiple logistic analyses by sex. Non-Hispanic black males were less likely to be in the overweight group when compared with non-Hispanic white males (OR = 0.68). For males, participants on sport teams and in exercise programs were less likely to be in the overweight group (OR = 0.63). Among the females, a significant interaction was found between age group and sport team and exercise program participation. Female sport team and exercise program participants in the group aged 14 to 16 years were less likely to be in the overweight group (OR = 0.23). Larger family size was associated with nonoverweight status among females (OR = 0.88). Higher family income was associated with nonoverweight for males (OR = 0.85). Females who watched 4 or more hours of television were more likely to be in the overweight group than those who watched less television (OR = 1.88). An overweight father increased the risk of being in the overweight group (OR = 1.74 for males, OR = 1.81 for females) compared with those who did not have an overweight father. Youth with an overweight mother were more likely (OR = 2.31 for males, OR = 2.73 for females) to be in the overweight group. No other significant associations were found.
This study evaluated the environmental influences and physical activity behaviors of overweight and nonoverweight 8- to 16-year-olds. Participation in vigorous physical activity was not related to being overweight, while being a sport team or exercise program participant was associated with not being overweight among all males and 14- to 16-year-old females. Females who watched 4 or more hours of television on the day preceding the survey were more likely to be overweight. Environmental factors associated with overweight were being from smaller families for females, being from a family with a smaller family income for males, and having at least 1 overweight parent for both.
In this study, 26% of youth who had participated on sport teams and in exercise classes were overweight compared with 36% of youth who did not participate. This association was significant for all males and for 14- to 16-year-old females. Ross et al30 reported that most children's physical activity takes place in organized programs outside of school. Sports participation is associated with better dietary habits, with youth involved in organized sports reporting more fruit and vegetable consumption than nonparticipants.31,32
As part of dietary guidelines it is recommended that all Americans be physically active each day.33 Among the suggested activities for children and teens was joining after-school or community physical activity programs. But data from the 1997 Youth Risk Behavior Survey indicate that there is a decline in participation in both school and organized sports unaffiliated with schools during high school, especially among females.9 This may be because of a decrease in accessibility or availability of structured activity.
The association between parental weight status and children's BMI was consistent with other studies. Youth with an overweight father or mother had an increased risk of being overweight. Familial patterns of obesity have been attributed to both genetics and family environment.1,34 Youth in families with 1 or 2 overweight parents consume a higher percentage of their energy intake as fat.18,35 No doubt this contributes to the development of overweight in youth. The relationship between overweight and other family variables is similar to previous findings. Troiano and Flegal2 using NHANES III found an inverse relationship between poverty index and non-Hispanic white males. Increases in overweight in youth with smaller family size has been noted.24
Using NHANES III data, Andersen et al15 found a significant positive relationship between television watching and BMI. This study, after controlling for demographic, physical activity behaviors, and environmental variables, found that females who reported watching 4 or more hours of television were more likely overweight than those who watched fewer than 4 hours of television per day. Reducing television watching may be a way to reduce and prevent overweight. Several recent school-based interventions that were successful in reducing the amount of time spent watching television reduced the BMI among females.36,37
This study found that there is an association between overweight status and lack of participation on sport teams and in exercise programs, but the use of a cross-sectional study design precludes conclusions about temporal sequence. Thus, it is not known if nonoverweight youth elected to participate on sport teams and in exercise programs or if participation in these activities helped to prevent youth from becoming overweight. It should also be noted that the duration of vigorous physical activity was not assessed in NHANES III and conclusions about a dose-response relation between the amount of physical activity and overweight cannot be assessed. The subset used in this study was composed of a slightly higher percentage of white youth; they were slightly more active and watched fewer hours of television than those from the NHANES III sample who were not included. Also, this study was limited to youth with reported BMIs for both biologic parents. Despite these limitations, the inclusion of data from a large national data set of youth with a wide range of ages and 3 race/ethnic groups increases the external validity of these results.
School-based interventions have shown that physical education classes can provide vigorous physical activity and promote skills necessary for developing good activity patterns.38- 40 Schools and communities need to provide physical activity programs that meet the needs and interests of all youth.41 Physicians can serve as advisors for schools and communities to ensure that such programs are safe and enjoyable and provide activities for youth of all ages and race/ethnic groups and for youth with different abilities and interests. Such programs may help to prevent more youth from becoming overweight.42
Parents are instrumental in bringing about weight loss in youth. Epstein and colleagues43- 46 presented a series of successful family-based interventions based on principles of behavior modification in 5- to 12-year-old obese children that incorporated physical activity, nutrition, and decreased television watching. Physicians need to educate parents about the complications of overweight and evaluate the diet and physical activity of their pediatric and adolescent overweight patients.47 The physician should encourage families to make changes that include increases in physical activity of the overweight youth, limiting television watching, and development of healthy eating habits.
No differences were observed in the number of sessions of reported vigorous physical activity between overweight and nonoverweight youth. Overweight youth were found to be less likely to have participated on sport teams and in exercise programs and were more likely to have overweight parents. Normal-weight females were more likely to have watched less television and come from larger families, while males were from families with higher incomes than their overweight counterparts.
Accepted for publication January 5, 2001.
Corresponding author: Marsha Dowda, DrPH, Department of Exercise Science, University of South Carolina, Columbia, SC 29208 (e-mail: email@example.com).
Dowda M, Ainsworth BE, Addy CL, Saunders R, Riner W. Environmental Influences, Physical Activity, and Weight Status in 8- to 16-Year-Olds. Arch Pediatr Adolesc Med. 2001;155(6):711-717. doi:10.1001/archpedi.155.6.711