Partitioning personal measurements into personal exposure to pollution from indoor and outdoor sources in London inhabitants - Αn analysis within the MELONS* project

Abstract

Background and Aim. Air pollution, particularly particles, from indoor and outdoor sources originate from different activities, have different toxicities, may have different health effects and need to be addressed by separate interventions. For epidemiological studies, personal exposure to pollution from outdoor sources was proposed as true exposure often approximated by modelled estimates. This study aimed to derive partitioned personal exposure to pollution from indoor and outdoor sources from four diverse study populations gathering 13,871 participant-day personal measurements of exposure using wearables in London. Methods. A mixed approach, based on personal measurement frequency, fixed radius, clustering and reverse geocoding, was applied to tag one-minute personal measurements into personal microenvironments. Home addresses were paired with local ambient data to facilitate calculation of infiltration efficiencies. Then personal measurements were partitioned into personal exposure to pollution from indoor and outdoor sources at one-minute resolution for each participant. Results. COPD patients, compared to school children, had higher mean PM2.5 pollution from indoor sources (COPD: 6.9 ± 37.2, children: 4.1 ± 42.8 μg/m3), but similar pollution from outdoor sources (COPD: 5.1 ± 12.8, children: 5.3 ± 27.2 μg/m3). For black carbon, professional drivers, compared to healthy adults, had higher personal exposure to pollution from outdoor sources (drivers: 1.7 ± 4.3, healthy adults: 1.2 ± 2.5 μg/m3) and lower personal exposure to pollution from indoor sources (drivers: 0.3 ± 1.8, healthy adults: 0.8 ± 2.1 μg/m3). Conclusions. By utilising GPS data, personal measurements from wearable monitors can be partitioned into personal exposure to pollution from indoor and outdoor sources to aid epidemiological studies and policy development. Our results, highlight that sources of pollution can be very different even though personal measurements of total exposure levels are similar. Further work is required to extrapolate results over longer time periods and larger populations. Keywords. Pollution partition, Personal exposure; Wearables