Estimating exposure to pollutants generated from indoor and outdoor sources within vulnerable populations using personal air quality monitors: A London case study

Personal exposure to air pollution can originate from indoor or outdoor sources, depending on location and activity. This study aimed to quantify personal exposure from each source separately, allowing comparison of the associated epidemiological estimates from each source type. We utilised 12,901 participant-day personal measurements of exposure to multiple pollutants collected from 344 London dwelling participants of four panel studies conducted between 2015 and 2019. A four-step process was applied to personal measurements incorporating 1) GPS spatial analysis including address identification and location tagging; 2) estimating outdoor home pollutant levels from matched fixed ambient monitors; 3) calculation of infiltration efficiency when participants were at home; and 4) indoor and outdoor source separation for personal exposure measurements. From the results, our participants with Chronic Obstructive Pulmonary Disease (COPD) dataset had an average (SD) personal exposure from outdoor sources of 4.0 (1.3) μg/m³ for NO₂ and 5.1 (3.0) μg/m³ for PM_2.5, the school children's average (SD) personal exposure to PM_2.5 from outdoor sources was 5.5 (4.3) μg/m³, the professional drivers' average (SD) personal exposure to black carbon from outdoor sources was 1.7 (1.0) μg/m³, and the healthy young adults' average (SD) personal exposure to black carbon from outdoor sources was 1.2 (0.5) μg/m³. Compared to the average total personal exposures, outdoor sources accounted for 49 % of NO₂ exposure, 41 % to 55 % of PM_2.5, and 60 % to 85 % of black carbon, dependent on the panel study - demonstrating a strong influence from outdoor sources for personal exposures to air pollution in London. Our findings highlighted that endeavours should continue to be made towards reducing pollution from both outdoor and indoor sources. The between-panel and within-panel exposure differences, derived from our novel partitioning methodology, can contribute to the estimation of health effects from indoor and outdoor sources and inform targeted interventions for vulnerable groups.