Childhood obesity has been rising since the 1970s; recent estimates indicate that 17.1% of US children and adolescents are overweight (1). Obesity and its associated comorbidities are of particular concern for residents of rural areas in the United States, where obesity rates are higher in adults (2) and children (3-5). Children living in rural communities in the United States may be at heightened risk for obesity because of broader obstacles to health and well-being such high rates of poverty (6) and limited access to medical care (6,7).

Children's participation in regular physical activity (PA) is important for the prevention of obesity and is associated with several other health benefits, including the prevention of cardiovascular disease and type 2 diabetes, improved musculoskeletal health, and better mental and emotional health outcomes (8). Several large national health surveys in the United States have found leisure-time PA to be lower among adults who reside in rural areas compared with those in urban areas (2,9-11). In contrast to findings in adult studies, a study that used national data from the 2003 Youth Risk Behavior Survey found that urban adolescents generally report lower engagement in PA and higher engagement in sedentary behaviors than do their suburban and rural counterparts (12). Similarly, a 2008 study of 3,416 elementary school students in Iowa found urban children were the least active overall (5). Further research on urban and rural differences in PA in children is warranted to assess potential disparities in PA by metropolitan status.

The South is an important region to investigate urban and rural obesity prevention behaviors. Southern states have the largest proportion of rural children in the United States (5) and some of the highest prevalence rates of obesity. A national study of rural US counties found that counties in Louisiana, Mississippi, and Texas had the highest prevalence of obesity (13). Adolescent overweight has also been found to be higher in southern states (14). This study compares the prevalence of PA and sedentary behaviors among children who resided in urban, suburban, and rural metropolitan areas by using state data representative of public school students living in Texas. Recognizing that PA behaviors may vary by developmental age, this study examines PA and sedentary behaviors by metropolitan status in 4th-, 8th-, and 11th-grade students.

The sex and age composition was similar across metropolitan status classifications for each grade (Table 3). Across grades, the proportion of African American students was highest in the urban stratum and lowest in the rural stratum, as expected from the sampling frame. Hispanic students were the largest ethnic group in the urban stratum, and white/other students were the largest ethnic group in the rural stratum across grades. The composition of Hispanic and white/other students was similar for the suburban stratum of 4th and 8th grades, and white/other students (50.3%) represented the largest ethnic group for the suburban stratum in 11th grade. The proportion of economically disadvantaged students was highest in urban students across grade levels, ranging from 59.9% of urban 11th-grade students to 69.7% of urban 4th-grade students (Table 3).

Prevalence estimates of engagement in PA and sedentary behavior by grade and metropolitan status were stratified by sex (Table 4 and Table 5). Urban students reported the largest decreases in PA from 4th to 11th grades, with significant decreases in PA in girls based on nonoverlapping confidence intervals found for sports team participation, vigorous PA, and participation in other PA/lessons. Suburban and rural boys' VPA increased from 4th to 11th grade, and urban boys' VPA decreased (Table 4).

With the exception of PE class attendance, PA prevalence estimates were similar across metropolitan status classifications for 4th-grade students. Prevalence of daily PE class attendance (≥4 days per week) in suburban and rural 4th-grade students was roughly double that of urban students (Table 5). Exploring mean PE days attended in an average week by metropolitan status, we found the mean daily PE attendance was lower for urban students in 4th (2.31; 95% confidence interval [CI], 2.09-2.54) and 11th grades (0.88; 95% CI, 0.12, 1.64) compared with same-grade suburban (4th-grade mean, 2.91; 95% CI, 2.62-3.19; 11th-grade mean, 2.17; 95% CI, 1.66-2.69) and rural students (4th-grade mean, 3.23; 95% CI, 2.84-3.62; 11th-grade mean, 1.78, 95% CI, 1.33-2.24).

For 8th- and 11th-grade students, the percentage of students who reported engaging in any of the 6 PA indicators examined was generally lower among urban students (Tables 4 and 5). Vigorous PA in 8th-grade urban boys was the only PA indicator that ranked higher compared with the suburban and rural status classifications. For all students, a higher percentage of 8th- and 11th-grade students reported playing on a sports team at school than on a sports team outside of school. This difference in sports team participation at school versus outside of school was highest in rural school students in 8th grade girls and boys (Table 4). Finally, generally similar prevalence estimates within each grade were found for sedentary behaviors of television watching and video game playing across the metropolitan status classifications (Table 6).

Differences in PA among metropolitan status classifications varied by PA indicator, grade level, and sex. Despite adjustments made for age, ethnicity, and percentage of economically disadvantaged students, suburban and rural students were more likely to report PA behavior than were urban students (Table 7 and Table 8). Suburban and rural 4th- and 11th-grade students were 2 to 3 times more likely than urban students to report attending PE on 4 or more days of the week (Table 8). In 8th grade, rural girls and boys were more than twice as likely to report participation on a sports team at school compared with urban counterparts from the same grade (Table 7). Suburban and rural 11th-grade boys were roughly twice as likely to report participating in vigorous PA on 3 or more days of the week compared with urban boys from the same grade level (Table 7), and 11th-grade suburban boys were 1.71 times more likely to report participation in strengthening exercise. No significant differences by metropolitan status were found for participation in other structured PA in 8th- and 11th-grade students (Table 8). Urban students generally reported lower prevalence of PA, but urban 4th-grade girls reported a higher likelihood of participating in sports teams and in other structured physical activities than did their nonurban counterparts (Tables 7 and 8), and urban 8th-grade boys were more likely to report engaging in vigorous PA than were their rural counterparts (Table 7). No significant differences among the metropolitan status classifications were found for 8th-grade students' participation in sports teams outside of school or engagement in strengthening exercise (Table 7).

Few significant differences were found by metropolitan status for sedentary behaviors, and the direction of the significant associations was mixed (Table 9). In 4th grade, suburban and rural girls were 1.54 times more likely to report watching at least 3 hours of television per day compared with their urban counterparts. In 8th grade, rural boys were more than twice as likely to play computer/video games compared to urban boys of the same age. In 11th grade, suburban boys were less likely to watch at least 3 hours of television per day compared with urban boys. No significant differences in television watching by metropolitan status were found for 8th grade boys, and no significant differences in computer/video game playing were found for 4th- or 11th-grade girls and boys (Table 9).

We aimed to assess potential disparities in child and adolescent PA behaviors by residence in urban, suburban, or rural areas. Contrary to findings of lower prevalence of PA among adults in rural areas (2,9-11), we found that urban 8th- and 11th-grade students generally reported lower prevalence of PA and urban 4th-grade students reported lower daily PE participation than did their rural or suburban counterparts. This finding is supported in part by previous research of elementary (5) and high school students (12) that found urban children reported the lowest prevalence of PA behaviors.

The lower prevalence of PA in 8th- and 11th-grade urban students may result from greater concerns about the safety of children who live in urban environments. Several studies have documented an inverse association between parent and child perceptions of neighborhood safety and children's engagement in PA (21-23). Urban residents may have more concerns about neighborhood safety because violent and property crime rates are higher in urban than in suburban and rural areas (24). A national report of school crime and safety in the United States found that urban students were more likely than suburban and rural students to report fear of attack at school, on the way to or from school, and away from school (25). Parental anxiety about neighborhood safety is negatively correlated with children's PA levels, and urban parents express more anxiety about neighborhood safety than do suburban parents (22).

Our finding of a higher prevalence of sports team participation and other structured PA participation in 4th-grade students and lower prevalence of these PA behaviors in 11th-grade students across metropolitan status corroborates previous research findings that engagement in PA declines as children mature through adolescence (8). We found that between 4th and 11th grade, urban students reported the largest decline in sports team participation and vigorous PA, and among the largest declines in other structured PA participation. A longitudinal study of adolescent PA found the decline in the total amount of PA during adolescence to be primarily a function of a decrease in the number of activities rather than a decrease in the time spent on specific activities (26). Our findings of a lower prevalence of participation in team sports and other organized PA among urban 8th- and 11th-grade students suggest that older children in urban environments may face greater barriers to participation in structured PA. Barriers that merit further investigation to understand lower PA in urban students include the possibility of greater competition for participation on sports teams, higher costs for sports and other PA participation, and decreasing availability of structured PA opportunities for the higher grade levels.

Given that our results indicated higher prevalence of participation in school-based sports compared with sports by outside organizations across metropolitan status classifications, schools appear to be an important context for providing sports participation opportunities across metropolitan status. Our finding that rural 8th-grade boys and girls are more than twice as likely as their urban counterparts to participate in school sports may indicate that more opportunities for school-based sports are available in rural areas than in urban areas. A stronger social norm for sports team participation among rural students may also be a factor. Further research is needed to assess the availability of and access to organized school- and community-based physical activity opportunities by metropolitan status to better understand our finding of less participation in organized PA among urban 8th- and 11th-grade students.

The lower prevalence of some PA behaviors in urban students may also result from less access to recreational areas and facilities. Closer proximity and higher density of exercise facilities have been associated with increased frequency of exercise (27). In a study of commercial PA-related outlets across the United States, suburban areas were significantly more likely than urban and rural areas to have physical fitness facilities, membership sports and recreation clubs, dance studios, schools and halls, and public golf courses (28). Lower access to areas of recreation such as parks is another factor that may account for lower prevalence of PA behaviors in urban students observed in this study. Access to parks has been found to be associated with higher engagement in PA behaviors in adolescent girls (29). A study based on the 2003 California Health Interview Survey found that teens living in neighborhoods with a greater concentration of crowded households and a higher proportion of residents living in poverty had less access to parks and were less physically active (30). Research is needed to assess differences in access to recreational facilities by metropolitan status as a potential explanation for lower PA in urban students.

Our data indicate that urban 4th- and 11th-grade students reported significantly less PE attendance than their nonurban counterparts, with a mean difference of approximately 1 day per week compared with rural students. A study of 277 elementary schools in Montreal found that suburban location was associated with greater opportunity for student PA in schools (31). In a national study of PA in US high school students (12), point prevalence estimates of PE attendance on at least 1 day per week were the lowest for urban students. With the passage of Texas Senate Bill 530, elementary and middle schools in Texas are now required to implement 30 minutes of PA per day or 125 minutes of PA per week. As such, this requirement may reduce the disparity in PE minutes by metropolitan status over time, provided that schools consider PE an opportunity for addressing PA-mandated minutes. In high schools, however, where no recent PA legislation has been passed in Texas, this disparity in PE merits further attention to identify factors such as local district policies, competing academic interests, and budget constraints that may account for the lower PE participation in urban students.

Sedentary behaviors such as television watching have been associated with child overweight (32) and have been hypothesized to displace time spent in PA (8). Our prevalence estimates of television watching for at least 3 hours per day in this sample of 11th-grade students from Texas are similar to national estimates for US high school students (33). In observing sedentary behavior by metropolitan status, we found few significant differences among the grade levels examined. The similar prevalence estimates of sedentary behavior in urban, suburban, and rural students is supported in part by our previous study of US high school students in which we reported no significant differences by metropolitan status in television watching for either male or female students or for computer/video game playing in male students (12). In contrast to these findings, adolescents living in inner city neighborhoods of low SES and urban locations with diverse race/ethnicity were significantly more likely to report at least 14 hours of television/video viewing and video/computer gaming per week than those living in newer suburban developments (34). These results are most likely due to differences in the way in which geographic location was defined; Nelson et al used a cluster analysis approach to identify specific, nonoverlapping patterns of neighborhoods that depart from the traditional classification of urban, rural, and suburban (34). We recognize the potential for better differentiation of location based on an approach that takes into consideration other neighborhood characteristics, but our findings nonetheless suggest that a large proportion of 4th-, 8th-, and 11th-grade students across metropolitan status classifications (range, 28.0%-62.2%) engage in long hours of media-related sedentary behavior every day.