Female Youth Soccer Participation and Continued Engagement: Associations With Community Size, Community Density, and...

Environmental context can impact youth engagement in sport and athlete development. Previous work has examined the population size of the birthplace of elite athletes; commonly known as the birthplace or community size effect. Community density has also been recognized as an important variable. Exact estimates for the ideal community characteristics and a thorough understanding of the underlying mechanisms has been somewhat elusive. Existing studies are cross-sectional in nature and there is evidence to suggest that significant variation exists within imposed categories. An athlete's birthdate position in a similar-age cohort can also impact development and has been associated with (dis)advantages resulting from subtle age differences (i.e., the relative age effect); it remains unknown if this variable is associated with population density. The objective of this study was to establish longitudinal participation trends among female youth soccer players in Ontario Canada, with consideration of community size, community density, and relative age. Within-category variation and associations between the variables were assessed. Registration entries at age 10 years (n = 9,826) and 16 years (n = 2,305) were isolated for analysis. Odds ratio analyses were conducted within each community size and density category for all 10 year old registrants; 95% confidence intervals were obtained. This procedure was repeated for all registrants at 16 years of age using the expected distribution at age 10 years to examine continued engagement. Findings suggest medium-sized communities (i.e., 10,000-249,999 inhabitants) provide the best odds of participation and continued engagement. Less densely populated communities (i.e., 50-<400 population/km²) appeared to be ideal for facilitating participation at age 10 years, but not for engagement at age 16 years. However, within-category variation was evident when each community was inspected individually. Consistent with previous attempts to find an association between community size and the relative age effect, there did not appear to be an association between community density and birth quartile distribution. Observations from this study show that community size and community density are truly unique and separate variables. Future studies should consider the underlying contributions to both low and high participation and continued engagement, while being mindful of within-category variation.