Teun Remmers

Investigating determinants of children's outside | 61 Discussion This study has shown that family environment was the strongest construct of variables predicting OP, compared to socio-demographic characteristics and perceived PE. More specifically, this study has demonstrated that habit strength and the presence of rules were positive, strong and stable predictors of OP over time. Positive modelling by siblings and child age at baseline were also positive predictors. In contrast, parent perceived difficulty in improving OP was a negative (i.e. related to less minutes of OP), relatively strong and stable predictor of OP over time. As this study focused on unstructured OP, results are not generalizable to organized sports, school PE and/or active transport PA. Therefore, one should be cautious in directly relating an increase in OP with more PA energy expenditure (PAEE) and the accompanied benefits regarding childhood obesity. However, considering the various benefits of OP (e.g. increased PA, motor abilities and social skills), it may contribute to an improved general health status of the child over time. This study has found no relationship between perceived PE and OP, which is in line with a cross-sectional study among 4- to 12-year-old children (13) and a longitudinal study among 5- to 6-year-old children (14). Hypothetically, an explanation for this may be that parents have other perceptions regarding PA opportunities in their neighborhood than their children, which may also explain the relatively weak associations regarding attractiveness, child friendliness and traffic busyness. Therefore, future studies are urged to assess both parental and children’s motives for OP. Another explanation may be that attributes of perceived PE are moderated or mediated by family environment. Future studies are encouraged to investigate these mechanisms. Our study has found that a positive parental attitude and family attitude were related to less time spent on OP. An explanation for this contra-intuitive finding may be that some parents in the present study perceived that their child needed more PA or OP and these parents had the intention to improve but were unable to achieve this yet compared to other children (23). Another explanation may be that reverse causation played a role here, as parents may consider their OP as sufficient and therefore did not think their child could/should improve on this. In addition, active parental encouragement towards improving OP was related to less OP over time, which may be explained by the possibility that active encouragement conflicts with the self-regulatory character of OP. It may be suggested that facilitation and providing autonomy may be more effective in promoting OP in children. Our positive association regarding parental rules and duration of OP were not in line with the findings of Sallis et al. (22). Discrepancy in the formulation of rules may explain this as Sallis et al. formulated rules merely related to the discouragement of OP (e.g. do not play rough games), while this study solely asked for the presence of rules regarding OP.