Background
Youth participation in sport programs is an essential part of their physical activity and a predictor of healthy adult lifestyles (Perkins et al., 2004). Co-curricular school sport programs have the potential to be important sources for youth sport, particularly among disadvantaged populations (Edwards, 2009). Despite adolescents wanting to play school sports, opportunities may be limited by policies restricting participation (Casper et al., 2011). To advance policy reforms, it may be beneficial to understand potential effects of proposed policies to increase participation.

Objectives
This analysis sought to simulate the impact of policy changes to increase co-curricular sport opportunities for adolescents in North Carolina middle schools. We estimated the potential effect of two policy aims on sport participation: 1) Repealing the state policy preventing 6th grade participation in interscholastic athletics, and 2) Implementation of co-curricular intramural programs.

Methods
Simulations were based on data from two ALR-funded grants in combination with nationally representative survey data. Individual-level data and estimations of participation rates were derived from a survey in four NC middle schools (ALR7) administered to 2,582 6-8th graders. From these data, statewide baseline school-level sport participation rates were estimated for stratified groups of students based on grade level (6th-8th), race (White/Minority), economically disadvantaged (ED) and school sport program type (interscholastic/intramural). Participation rates were estimated for schools with combined sport programs by applying rate adjustments based upon the NELS:88 national dataset. School-based sport program information was obtained from a questionnaire completed by personnel at 325 NC middle schools in January 2009 (ALR8). School enrollment data were obtained from NCES Common Core of Data and NC Department of Public Instruction. We assumed that stratified participation patterns in this sample would be representative of middle schools statewide and that levels of supply and demand were constant. Initial school-level sport participation rates were estimated for each school based on enrollment in the stratified groups and sport program type. These baseline rates were validated by comparison with aggregated results of the NCYRBS.

To compare potential changes in statewide rates of youth sport participation, regression models were initially estimated using the simulated school-level sample, controlling for percent minority and percent ED. Parameter estimates to calculate estimated variation were generated from existing school program conditions (with 53% percent of sampled schools offering no intramurals) and a simulated effect of allowing 6th grade participation in interscholastic sports in a random sub-sample of half the schools in the dataset.

Results
61.03% of middle school students statewide were estimated to participate in organized sports at or out of school. The estimate was similar to the aggregate results of the NCYRBS that indicated 61.5% of middle school students reported participating in sports in 2009. The current participation rate for ED students was estimated at 52.2% compared to 69.2% for non-ED students. The participation rate was 50.6% for ED minority students. Among all sport participants, 37% were estimated to be ED.

Regression analyses predicted offering intramural sports was associated with the greatest potential effect on overall participation rates (B = .204, SE = .004, p < .001) when controlling for school percent minority and percent ED. It was estimated that the universal replacement of interscholastic sports with intramural sports would increase overall sport participation by 0.83%. However, implementation of intramural sports statewide for all grades while maintaining current levels of support for interscholastic sports was estimated to increase overall sport participation by 13%. These simulations estimated that sport participation among 6th graders could increase by 17.5%.

Permitting 6th grade participation in interscholastic sports was estimated to reduce overall participation (B = -.037, SE = .004. p < .001), likely due to the displacement of current 7th and 8th grade participants in interscholastic sports. Based on simulations, removal of the statewide policy prohibiting 6th grade participation in interscholastic sports would increase sport participation among 6th graders by 8.9%, but decrease overall participation among middle school students by 0.17%.

Regression models indicated school sport policies had a more significant association with rates of sport participation among economically disadvantaged students. With the adoption of intramurals statewide, estimated participation rates among ED students could increase 18.8% compared to 9.7% for non-ED students. Further, sport participation among ED minority students could increase 28.27% from current levels.

Conclusions
Simulations suggested that policy changes to implement intramurals statewide could result in 26,000 new sport participants among North Carolina adolescents, of which 53% were estimated to be ED. Alternative policy proposals may not have the same impact on increasing participation rates. While simulations such as these have inherent limitations, the estimated effects suggested in these policy changes warrant further investigation. Overall, school policies designed to expand opportunities for participation in broader co-curricular sport have potential for increasing overall physical activity and reducing physical activity disparities.

Support/Funding
Funding provided by Active Living Research, a research program of the Robert Wood Johnson Foundation.