Journal of exposure science & environmental epidemiology · 2025
Comparison of residential and mobility-integrated air pollution exposures from tracking campaigns and agent-based modelling in Switzerland and the Netherlands.
СтатьяСравнительное исследованиеНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
de Hoogh Kees, Flückiger Benjamin, Probst-Hensch Nicole, Jeong Ayoung, Imboden Medea, Karsies Aletta, Schmitz Oliver, Vermeulen Roel, Kyriakou Kalliopi, Ndiaye Aisha, Shen Youchen, Karssenberg Derek, Vienneau Danielle, Hoek Gerard
Аннотация
BACKGROUND: Studies investigating the health effects of long-term exposure to air pollution generally rely on the outdoor air pollution exposure assigned at the residential address. By ignoring time activity, population exposure misclassification could potentially lead to loss of precision or bias in epidemiological studies.
OBJECTIVE: We aimed to assess how residential-based air pollution exposures compared with "real" tracking-based exposures.
METHODS: We conducted two tracking campaigns in Switzerland (CH) and the Netherlands (NL) with 686 participants followed for 2 weeks with GPS trackers whilst keeping time activity diaries. In addition, we simulated mobility and commuting tracks for the same subjects using agent-based modeling (ABM) with information from census and travel survey data to estimate mobility-integrated air pollution exposures. Exposures were calculated by overlaying residential address, measured (GPS) and modeled (ABM) tracks with annual average hourly NO2 and PM2.5 concentration surfaces.
RESULTS: We found strong agreements between residential and tracking-based exposures in CH for both pollutants (R2 > 0.76) and NL for NO2 (R2 = 0.79), and weaker agreement in NL for PM2.5 (R2 = 0.56). Similarly, the agreement between ABM and tracking-based exposures was strong for NO2 (R2 > 0.77 in CH and NL), while for PM2.5 it was stronger in CH (R2 = 0.80) than in NL (R2 = 0.54). The highest correlations were between residential and ABM exposures (R2 > 0.96 for both pollutants). Using information commonly available even in large administrative cohorts, we found that exposures derived from the tracking campaigns agreed well with ABM in our two study areas.
SIGNIFICANCE: Our study supports the use of residential exposures in epidemiological studies on long-term health effects of air pollution, whilst acknowledging that ABM, especially if the work location is known, can be a useful tool to estimate mobility-integrated exposures.
IMPACT STATEMENT: Our research supports the use of residential exposures in studies investigating the long-term health effects of air pollution, whilst acknowledging that agent-based modeling, especially if the work location is known, is valuable for estimating mobility-integrated exposures. Our findings are broadly applicable to air pollution epidemiology, in particular, studies of large populations that rely on exposure modeling.
OBJECTIVE: We aimed to assess how residential-based air pollution exposures compared with "real" tracking-based exposures.
METHODS: We conducted two tracking campaigns in Switzerland (CH) and the Netherlands (NL) with 686 participants followed for 2 weeks with GPS trackers whilst keeping time activity diaries. In addition, we simulated mobility and commuting tracks for the same subjects using agent-based modeling (ABM) with information from census and travel survey data to estimate mobility-integrated air pollution exposures. Exposures were calculated by overlaying residential address, measured (GPS) and modeled (ABM) tracks with annual average hourly NO2 and PM2.5 concentration surfaces.
RESULTS: We found strong agreements between residential and tracking-based exposures in CH for both pollutants (R2 > 0.76) and NL for NO2 (R2 = 0.79), and weaker agreement in NL for PM2.5 (R2 = 0.56). Similarly, the agreement between ABM and tracking-based exposures was strong for NO2 (R2 > 0.77 in CH and NL), while for PM2.5 it was stronger in CH (R2 = 0.80) than in NL (R2 = 0.54). The highest correlations were between residential and ABM exposures (R2 > 0.96 for both pollutants). Using information commonly available even in large administrative cohorts, we found that exposures derived from the tracking campaigns agreed well with ABM in our two study areas.
SIGNIFICANCE: Our study supports the use of residential exposures in epidemiological studies on long-term health effects of air pollution, whilst acknowledging that ABM, especially if the work location is known, can be a useful tool to estimate mobility-integrated exposures.
IMPACT STATEMENT: Our research supports the use of residential exposures in studies investigating the long-term health effects of air pollution, whilst acknowledging that agent-based modeling, especially if the work location is known, is valuable for estimating mobility-integrated exposures. Our findings are broadly applicable to air pollution epidemiology, in particular, studies of large populations that rely on exposure modeling.