Background/Aim: Global increases in sedentary behavior are a major public health concern requiring research into its causes. We investigate environmental factors such as temperature and green space that may predict physical activity among children 4-6 years of age, since active children are more likely to exercise regularly over the course of their life. Methods: We use children's data collected in the years 2013-15 in a subset of the Mexico City PROGRESS cohort (n=383). Children wore accelerometers for a week recording their activity levels. For physical activity, we calculated mean, maximum, and minimum hourly vertical counts by day and by week. We estimated children's exposure to green space using 50m resolution seasonal maximum NDVI (Normalized Difference Vegetation Index), a satellite-based measure of greenness; and exposure to air temperature with a daily 1km resolution model developed by our group using data from satellite-based land surface temperature and weather stations (CV R2=0.89). For our analysis, we use multiple linear regression models adjusted for child and mother covariates, including child BMI z-scores. Results: We find maximum daily temperature to be positively associated with maximum hourly vertical counts. Children had on average 0.06 SD units higher exercise for each 1°C higher ambient temperature (p=0.002, temperature range 11-31°C). When stratified by gender, the significant and positive association holds for girls but not for boys. We do not find a significant relationship between green space and physical activity. Discussion: Our findings suggest that temperature, but not green space, is an important environmental influence associated with children's physical activity, particularly for early school age girls in Mexico City. Our findings may not generalize to other climates or non-urban areas that have more green space. We recommend that future epidemiological projects consider weather variations when analyzing physical activity data.
展开▼