Weather and children's physical activity; How and why do relationships vary between countries?

Flo Harrison, Anna Goodman, E. M.F. van Sluijs, Lars Bo Andersen, G Cardon, Rachel Davey, Kathleen F. Janz, Susi Kriemler, Lynn Molloy, Angie S. Page, Russ Pate, Jardena Puder, Luis B. Sardinha, Anna Timperio, Niels Wedderkopp, Andy P. Jones, Sigmund Anderssen, G Cardon, Alan Cooper, Ulf EkelundD. W. Esliger, L. B. Sherar, Karsten Froberg, Pedro Hallal, Katarzyna Kordas, J Reilly, Jo Salmon, E. M.F. van Sluijs, on behalf the ICAD collaborators

Research output: Contribution to journalArticlepeer-review

74 Citations (Scopus)
11 Downloads (Pure)


Background: Globally most children do not engage in enough physical activity. Day length and weather conditions have been identified as determinants of physical activity, although how they may be overcome as barriers is not clear. We aim to examine if and how relationships between children's physical activity and weather and day length vary between countries and identify settings in which children were better able to maintain activity levels given the weather conditions they experienced. Methods: In this repeated measures study, we used data from 23,451 participants in the International Children's Accelerometry Database (ICAD). Daily accelerometer-measured physical activity (counts per minute; cpm) was matched to local weather conditions and the relationships assessed using multilevel regression models. Multilevel models accounted for clustering of days within occasions within children within study-cities, and allowed us to explore if and how the relationships between weather variables and physical activity differ by setting. Results: Increased precipitation and wind speed were associated with decreased cpm while better visibility and more hours of daylight were associated with increased cpm. Models indicated that increases in these variables resulted in average changes in mean cpm of 7.6/h of day length, -13.2/cm precipitation, 10.3/10 km visibility and -10.3/10kph wind speed (all p < 0.01). Temperature showed a cubic relationship with cpm, although between 0 and 20 degrees C the relationship was broadly linear. Age showed interactions with temperature and precipitation, with the associations larger among younger children. In terms of geographic trends, participants from Northern European countries and Melbourne, Australia were the most active, and also better maintained their activity levels given the weather conditions they experienced compared to those in the US and Western Europe. Conclusions: We found variation in the relationship between weather conditions and physical activity between ICAD studies and settings. Children in Northern Europe and Melbourne, Australia were not only more active on average, but also more active given the weather conditions they experienced. Future work should consider strategies to mitigate the impacts of weather conditions, especially among young children, and interventions involving changes to the physical environment should consider how they will operate in different weather conditions.

Original languageEnglish
Article number74
Pages (from-to)1-13
Number of pages13
JournalInternational Journal of Behavioral Nutrition and Physical Activity
Issue number1
Publication statusPublished - 30 May 2017


Dive into the research topics of 'Weather and children's physical activity; How and why do relationships vary between countries?'. Together they form a unique fingerprint.

Cite this