Dynamic assessment of inhaled air pollution using GPS and accelerometer data

Abstract

Exposure to air pollution can have severe health impacts, especially for the elderly. To estimate the inhaled dose of air pollution, traditionally only the air pollution concentration at the home location is considered, without incorporating individual travel behavior and physical activity. This can lead to bias in health impact assessment and epidemiological studies, possibly underestimating exposure to air pollution and misinforming policy makers. Our paper addresses this issue using accurate 7-day GPS and accelerometer data on 180 participants aged between 58 and 65 living in Ghent (Belgium). NO2 concentration for Belgium is available from a land-use regression model. Three methods are used to calculate the inhaled dose of NO2. The first method is the traditional static method, using only the NO2 concentration at the home location. The second method incorporates travel behavior using GPS data, thus looking at the NO2 concentration at the exact location of the participant. The third method additionally incorporates accelerometer data and estimates the transport mode used and physical activity to calculate the ventilation rate. When incorporating geographical location, differences in inhaled dose of NO2 depend on the NO2 concentration at the home location and the individual travel behavior. When additionally incorporating ventilation rate, the inhaled dose of NO2 increases by more than 12%. In addition to comparing these three methods with each other, the influence of transport mode is tested. Cycling is associated with increased inhaled doses of NO2 relative to other modes. It is important for health impact assessment and epidemiological studies to incorporate individual travel behavior and physical activity to measure the inhaled dose of air pollution, and this can be done accurately using GPS and accelerometer data.