Teun Remmers

96 | Chapter 6 mechanisms between weather elements and their association with PA. Such moderation was found for temperature and solar radiation. Apart from the potential influence of water-related activities, sun protection policies may also have influenced children’s PA particularly during the warmer months. At higher intensities of solar radiation (UV-levels  3), SunSmart protection measures in Australia encourage children to stay in the shade and wear protective clothing (e.g. hats and long sleeve tops) during periods of the day with extreme UV-levels. This may have hindered children’s participation in outdoor PA (18), and thus may have contributed to the moderation between temperature and day type in the relationship with VPA (see Figure SDC6 in Appendix 3). Similar school policies regarding promotion of indoor PA during wet weather conditions (14), may also be required during periods of the day with extreme UV-levels. While we found consistent associations between relative humidity and lower PA, Lewis et al. (21) found no such associations with MVPA (21). Humidity may influence PA in a way that the sensation of heat (at higher temperatures) depends on relative humidity (27, 28), which suggests a moderating mechanism between temperature and humidity in the relationship with PA. However no evidence for interaction was found in our study. Therefore regardless of temperature, PA may be more exhausting and uncomfortable for children in higher relative humidity, and therefore it may be best to encourage indoor activities on days of high humidity. Moreover, no associations were found between humidity and light PA. Similar to temperatures above 25 degrees Celsius, increments in humidity are especially related to declines in PA at high intensity. This may mean that these meteorological circumstances may discourage and/or limit children from performing MPA/VPA because of discomfort and/or increased fatigue. Except for positive associations in the Canadian sample examined by Lewis et al. (21), five studies reported negative associations between rainfall and PA (9, 12, 21). In the current study however, results regarding rainfall were mixed. Interestingly, only Duncan et al. (9) and Harrison et al. (14, 15) categorized of rainfall due to its skewed distribution (9). Unlike the study of Duncan et al. (9) but in line with the studies of Harrison (14, 15), categorization of rainfall in our study was based on days with no rainfall, and subsequently split into two categories of similar size. When we applied the same category-thresholds as Duncan et al. (9) however, results remained unchanged. A potential explanation for the mixed rainfall findings in our study may be that children performed PA between rain showers, or inside during rainfall. As we had no data on the rainfall characteristics (e.g. intensity, duration or timing of rainfall) or indoor/outdoor timing, we cannot speculate how much influence rainfall had on children’s daily PA levels. In addition, future studies that have access to wind speed data may also investigate whether average daily wind speed (or in a moderation mechanism with rainfall) influence children’s PA behavior. Interestingly, one study reported clear declines in lunchtime and after-school PA with increasing rainfall in 9-11 year-old children, but no associations were found for 13-14 year-old children (15). Unfortunately our daily rainfall data and limited age range do not allow direct comparison with these findings (15). Future studies are encouraged to include rainfall characteristics and indoor/outdoor time in their analyses,