Patterns and Correlates of Participation in a Weekly Mass Participation Physical Activity Event, parkrun, in Australia, 2011-2020.

Background: Studying effective interventions already operating at scale is critical to improving physical activity intervention research translation. The free, weekly, timed 5-km run or walk parkrun represents a unique opportunity to examine successful organic dissemination. We conducted an ecological analysis to identify patterns of growth in Australian parkrun participation and their correlates from 2011 to 2020.

Method: Outcome variables were (1) weekly counts of walkers/runners and (2) monthly number of new parkrun registrants. We used latent class analysis to characterize growth trajectories followed by logistic regression on class membership. Covariates included parkrun course characteristics (eg, surface type and route), site-level aggregate participant profile (eg, proportion women and mean age), and surrounding area characteristics (eg, population density and physical activity norm).

Results: Three hundred and sixty-three parkruns were included (n = 8,388,695 participation instances). Sixty-nine percent followed a low-growth and 31% a high-growth participation pattern. High growth was associated with greater participation by women, concrete/bitumen surface type, lower area socioeconomic status, and greater volunteer heterogeneity. Odds of being in the slow-growth class were higher if the course contained >1 km of repetition, higher average age of participants, better average parkrun performance, and higher running group membership. Two patterns of new registration were identified: high start followed by steep decline; and low start, slow decline with similar correlates to participation.

Conclusions: Parkruns with a less competitive social milieu may have more rapid dissemination. As a free and regular event, parkruns in low socioeconomic areas have the potential to improve the activity levels of those with fewer resources.