Teun Remmers

General discussion | 181 accessibility to PA facilities and social environmental determinants (e.g. parenting influences, having active siblings) were important in explaining parent-reported outside play. These results underline the role of parents as gatekeepers in providing children the freedom to perform unstructured outside play. Fourth, chapters 6, 7, 8 and 9 reported on associations between the physical environment and children's PA, using methodological advances. As these methodological advances are still developing, these studies have addressed both empirical and methodological aims. Namely, results from chapter 6 discuss the influence of humidity, solar radiation, humidity, day length and rainfall on children's PA, and furthermore demonstrate the non-linear relationship between temperature and PA. The latter is in line with one previous study (7), and the study presented in chapter 6 was unique in investigating the influence of weather elements across all four seasons. The study presented in chapter 9 builds upon results from earlier studies showing declines of PA in the transition towards adolescence (8-14), and provides unique in-depth insight of where and in which time-segments changes in PA occur during the transition from primary to secondary school. Studies presented in chapters 7 and 8 focus on environmental determinants of afterschool PA. Previous studies highlight the importance of the afterschool period and its potential in increasing PA (15-17), but results from chapter 7 showed that the influence of playgrounds in the school environment is limited to the time-segment directly afterschool (until 6 PM), and for children living within 800-m from their school. This study provides an example of how important insights can be obtained from time- and spatial-filtering approaches. Chapter 8 focused the relationships between attributes of the physical (multi-place) environment and children's PA. Previous studies using accelerometers and GPS methodologies suggest that children are more active in greenspace environments such as parks (18-21), but results in chapter 8 add that children with higher densities of greenspace performed more afterschool leisure time PA compared to children with lower densities, irrespective of whether PA was actually performed at these greenspaces. These results may be an important step forward in acknowledging the importance of greenspace environments in objectively measured PA. Hence, urban planners and health promotors are encouraged to design natural experiments that strategically position greenspace in the shared publicly accessible home- school environment. In addition, future studies are encouraged to replicate these analyses in order to open opportunities for transcultural comparisons. In line with the previously described EnRG framework (3), this thesis both demonstrated that a direct pathway exists between environment and PA behaviour, and also that the pathway between (especially perceived) environment and PA is indeed moderated by parenting influences. Methodological and Theoretical Considerations Study Designs In this thesis, two observational study designs were applied to investigate environmental determinants of PA; cross-sectional and longitudinal designs. The studies in chapters 3, 7 and 8 used cross-sectional designs. Although using these designs limits the ability to draw conclusions on causal inference, in this thesis they were useful in exploring alternative