Teun Remmers

Playability of school environments and afterschool PA | 109 Statistical analyses Our analyses were performed using SPSS 20.0 for Windows (IBM SPSS Inc., Armonk, NY), and p <0.05 indicated statistical significance. Our dependent variable was the ASPA performed in light and moderate to vigorous intensity for each hourly time-interval of measurement. Our primary independent variable was the combined playability index of the different school-environments. We first described the percentages of light PA (LPA) and moderate-to-vigorous PA (MVPA) across four two-hour time-periods after school, and across four distance-buffers from schools using univariate analyses of variance (Table 1). We performed multilevel linear mixed models in order to account for the time-dependent structure of the data. We specified a random intercept and slope for the hourly time- intervals, nested within the specific dates at which a respondent’s accelerometry commenced. Analyses were also adjusted for hourly weartime, average daily temperature, daily duration of rainfall, and daily duration of sunshine per day. We evaluated whether age and gender moderated the association between playability and ASPA, but as we did not find moderation, we only adjusted for age and gender. To investigate the influence of the four time-periods and distance-categories on the relationship between ASPA and playability, we entered the appropriate interaction terms in our linear mixed models. Using dummy-coded interaction-terms between distance- categories and the playability index-score, we were able to estimate main effects of playability for each of the two-hour time periods, while still acknowledging the time- dependent structure with the random intercept and slope. Distance-categories were both conceptualized as crow-fly and network-distances. Finally, we also repeated our analyses now stratifying for the time-periods and distance-categories simultaneously to investigate their interactive influence (Table 4) Results In total, 587 children (74.2%) provided valid ASPA measurements, for two (44.8%) and three (55.2%) valid weekdays, respectively. The 280 participating boys and 307 girls were aged 10.2 years on average (range 8 to 11 years) (Table 1). Across all time-periods, 27.9% of the time after-school was spent in LPA, which accumulated to 103.4 minutes per day (SD=23.9). MVPA accounted for 7.7% of the after-school time, accumulating to 28.4 minutes per day (SD=14.6). Daily percentages of LPA declined across the four time- periods, while MVPA slightly declined after 16:00, but increased again after 18:00. Daily percentages of LPA and MVPA were comparable across the distance-categories (no statistically significant differences in analysis of variance, data not shown).