Teun Remmers

Daily weather and children's PA patterns | 87 Introduction The importance of regular physical activity (PA) and the range of associated short- and long-term benefits in children and adolescents is now globally established and accepted (16, 29). However, many children in Western Countries are insufficiently active (13), with inactivity tending to track from youth to adulthood (17). This makes the promotion of children’s PA a critical health promotion target. In order to develop effective interventions to increase PA, it is important to understand why, when, and how much PA is performed. The development and validation of accelerometers has been an important contribution to this, as accelerometers can provide rich and objective data over extended periods of time (4). An important complicating aspect for studies seeking to identify determinants of children’s PA patterns stems from variability in weather conditions. This may be particularly problematic for studies exploring determinants of PA that takes place outdoors. For example, relationships between aspects of walkability and active transport may be moderated by the weather as children may be more likely to walk to school instead of using motorized transport in favorable weather conditions (22). Lack of consideration of weather conditions may compromise the findings of observational studies and influence intervention outcomes. Hence, understanding how the weather affects PA may help future studies to analyze and interpret PA patterns, especially when measuring across diverse meteorological settings and with long follow-up periods. In addition, studies investigating the influence of the weather may help future interventions to target inactive periods to overcome weather-related PA declines (9, 25). The majority of studies that have examined weather and objectively-measured PA behavior in children have focused on seasonality (5, 25). Most European studies have generally found that PA is highest in summer and lowest in winter, but results from other continents are inconsistent (25). This inconsistency may be explained by scheduling of organized sports or school curricula (6, 26), or regional differences in specific weather elements (e.g. rainfall, temperature, wind, relative humidity, etc.) (11, 25). International comparisons are limited as average weather conditions within season vary considerably across the world (21). Consequently, although season is a proxy for general differences in weather conditions, seasonal variation does not fully capture the temporal influence of weather on PA behavior. In contrast to seasonal differences, consideration of the influence of actual daily weather elements (e.g. temperature, humidity, solar radiation) does enable international comparisons (21), and provides more detailed information about weather-related differences in children’s PA. The investigation of weather elements and their influence on activity levels is therefore warranted (5, 19). To the best of our knowledge, only the studies of Goodman et al. (12), Duncan et al. (9), Harrison et al. (14, 15), and Lewis et al. (21) have focused on weather elements by matching daily meteorological data to accelerometer measurements in children. However, in three studies data were only collected during a single measurement period (i.e. only