Poster 188. Patterns of Multi-Sport Participation in High School Athletes: A Decade of Epidemiologic Data from More Than 80,000 Students

This is the authors' abstract. We don't add key points for this paper.

Objective: Sports specialization has been increasingly recognized as a modifiable risk factor for overuse injury in young athletes due to repetitive biomechanical loading in a single sport. Conversely, participation in multiple sports may distribute biomechanical forces across the body, potentially increasing force adaptation and reducing cumulative injury risk. Despite existing evidence, large-scale epidemiologic research on high school athletes remains limited, yet it is essential for developing effective injury prevention strategies and informing youth sports participation policies. The purpose of this study was to introduce a novel epidemiological approach to assess multi-sport participation using publicly available roster data from Michigan high schools between 2015 and 2025. Methods: In this retrospective, database study, participation and injury data were analyzed for n = 80,226 high school athletes across Michigan between 2015 and 2025. Roster data were obtained from MaxPreps, an online platform hosting high school sports rosters and statistics. Using a two-stage Python pipeline, rosters were automatically scraped across 10 academic years and exported into a combined Excel sheet. To identify single-versus multi-sport athletes, player names were matched across sports to detect repeated entries. The sports assessed included football, wrestling, baseball, boys' and girls' basketball, soccer, and girls' volleyball. For each academic year, rosters from all sports were merged into a single data frame and stored as a distinct worksheet, allowing consistent year-to-year organization. A secondary processing script standardized and cleaned the data. This included alphabetizing player records, identifying multi-sport athletes by counting unique pairs like name and grade, as well as generating annual summary statistics for single-versus multi-sport participation. Multi-sport athletes were flagged for manual verification, and a checkpointing system ensured that previously processed files were logged to avoid redundant analysis. Statistical analyses included calculation of the mean, standard deviation, and coefficient of variation, as well as a Chi-square test of independence to assess rates of multi-sport participation and their year-to-year fluctuations. Results: Over 10 years, data from 80,226 student-athletes were collected across 198 high schools in Michigan, encompassing all athletes who participated in fall, winter, and spring sports from 2015 to 2025. On average, there were 376 multi-sport athletes per year (SD 150) compared with 7,647 single-sport athletes (SD 2,300). The proportion of athletes participating in multiple sports was consistently low, averaging only 4.8% annually across the 10 years. Year-to-year variation was modest, with a mean deviation of 0.55 percentage points, an SD of 0.66, and a coefficient of variation of 1.19. However, a Chi-square test of independence demonstrated a significant difference in multi-sport versus single-sport athlete distribution across seasons (χ²(9) = 55.16, p < 0.001). Conclusions: In this large-scale epidemiologic analysis of Michigan high school athletes, multi-sport participation rates remained low, averaging less than 5% annually across a decade of data. Meaningful year-to-year fluctuations were observed, demonstrating that rates of sport diversification are not static at the high school level. This study introduces a novel methodology that enables large-scale assessment of multi-sport participation using publicly available data. The approach is reproducible, scalable, and may serve as a foundation for future research exploring how multi-sport participation relates to athletic injuries when linked with external datasets. Several limitations must be acknowledged. The study relies on MaxPreps, a voluntary self-reporting platform, which introduces variability in data quality across schools and seasons. Reporting practices are not standardized, and some sports were inconsistently represented. Furthermore, the dataset does not capture participation in club or travel teams outside of high school seasons, which may underestimate true multi-sport involvement. Despite these limitations, this work demonstrates the feasibility of constructing a large-scale, structured dataset of high school athlete participation. This methodology provides a new way to quantify and track patterns of sport diversification at scale, laying the groundwork for future studies that integrate roster data with other national datasets.