Locations of Adolescent Physical Activity in an Urban Environment and Their Associations with Air Pollution and Lung Function.

Rationale: Physical activity while being exposed to high concentrations of air pollution may lead to greater inhalation of pollutant particles and gases. Thus, owing to features of the built city environment, specific locations where physical activity take place may put individuals at increased risk for harmful inhaled exposures leading to decrements in lung function.
Objectives: The objectives were to determine locations throughout an urban landscape where children engage in moderate to vigorous activity (MVA). We hypothesized that outdoor activity would be associated with increased exposure to air pollution and reduced lung function.
Methods: Children aged 9-14 years living in New York City (NYC) (n = 151) wore global positioning system devices and wrist accelerometers for two 24-hour periods. Time-stamped global positioning system points and accelerometer data were aggregated and mapped using ArcGIS to determine locations where children engaged in MVA. Location-specific particulate matter <2.5 microns and nitrogen dioxide (NO2) was determined based on land use regression models of street-level pollution. Temporal air pollution exposure was determined based on daily concentrations collected at one central site in NYC. Forced expiratory volume in 1 second (FEV1), forced vital capacity (FVC), and forced expiratory flow, midexpiratory phase (FEF25-75) were collected following each 24-hour period. Data were analyzed using multivariable linear regression models to examine associations between MVA time and both lung function and air pollution in separate models. Additionally, a multiplicative interaction term (MVA time × season) was included to test whether the association between MVA time and lung function outcomes varied by warmer versus colder months.
Results: On average, children spent less MVA time outdoors (38.2 ± 39.6 min/d) compared with indoors (71.9 ± 74.7 min/d, P < 0.01), regardless of season. The majority of outdoor MVA occurred along sidewalks and roadbeds (30.2 ± 33.3 min/d, 76.9% of outdoor) where annual average concentrations of NO2 were relatively high. Interquartile range (IQR) increase in outdoor MVA time (44 min) was associated with higher levels of annual average NO2 (P < 0.01) but not particulate matter <2.5 microns. In warmer months, for IQR increase in outdoor MVA time, children had 1.41% lower FEV1/FVC (95% confidence interval [95% CI], -2.46 to -0.36) and 4.40% lower percent predicted FEF25-75 (95% CI, -8.02 to -0.78). These results persisted even after adjustment for location-specific annual average concentrations of NO2. No association was observed between MVA time and lung function in colder months (P > 0.05), and a formal test for interaction (MVA time × season) was significant (P value for interaction = 0.01 and 0.03 for FEV1/FVC and FEF25-75, respectively).Conclusions: Children in NYC spent less time active outdoors compared with indoors. Outdoor activity was greatest near traffic sources and associated with higher annual average concentrations of NO2. In warmer months, outdoor activity was associated with lower lung function, but this association did not appear to be mediated by higher exposure to outdoor pollution during exercise.