School day physical activity profiles and transitions across the school year: a latent analysis in 8–9 year-old children

Abstract

Abstract Background School day physical activity (PA) is important for children’s health and development, yet longitudinal evidence on within-year trajectory patterns is limited. Traditional approaches to studying PA change may obscure substantial heterogeneity in how individual children’s PA evolves over time. We used person-centered latent analyses and cut-point-free accelerometer metrics to identify distinct school day PA profiles and examine how profile membership changed across one school year. Methods Participants were 249 children aged 8–9 years from seven primary schools in northwest England. Wrist-worn accelerometers assessed school day PA at three timepoints during the 2023-24 school year (Autumn, Winter/Spring, Summer) using average acceleration (PA volume) and intensity gradient (PA intensity distribution) as population-independent metrics. Multiple imputation handled missing accelerometer data. Latent profile analysis was performed at each timepoint with latent transition analysis used to examine stability and change in profile membership across consecutive timepoints. Results Three distinct PA profiles were consistently identified across all timepoints: High PA (high volume, high intensity; 10.8–41.4% of sample), Average PA (below average volume, average intensity; 39.4–67.9%), and Low PA (low volume, low intensity; 13.7–48.6%). Profile membership shifted significantly across the school year ( p <.001) with predominantly downward transitions. The Autumn-to-Winter/Spring transition was particularly pronounced with 82.5% of High PA children transitioning to less active profiles and only 17.5% maintaining High PA membership. Despite modest Summer increases in aggregate mean PA, individual-level transitions remained predominantly downward, illustrating how group-level trajectories masked divergent subgroup patterns. Girls were consistently over-represented in the Low PA profile (67.9–73.5%) and boys in High PA (68.9–77.8%) at all timepoints, with these patterns replicated within each school. Conclusions School day PA in mid-childhood was characterised by three distinct, replicable profiles with predominantly downward within-year transitions that were most pronounced during winter months. The divergence between aggregate and individual-level PA trajectories underscores the value of person-centered approaches for understanding school day PA dynamics. School day PA promotion strategies should extend beyond targeting inactive children to also prioritise maintaining engagement among higher active children across the school year, with gender-responsive approaches needed to address the persistent and systemic disadvantage experienced by girls.