Ka Hung Chan1, Xi Xia2, Kin-Fai Ho2, Yu Guo3, Om P Kurmi4, Huaidong Du5, Derrick A Bennett1, Zheng Bian3, Haidong Kan6, John McDonnell1, Dan Schmidt1, Rene Kerosi1, Liming Li7, Kin Bong Hubert Lam8, Zhengming Chen5. 1. Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, University of Oxford, UK. 2. Jockey Club School of Public Health and Primary Care, The Chinese University of Hong Kong, Hong Kong Special Administrative Region. 3. Chinese Academy of Medical Sciences, China. 4. Faculty Research Centre for Intelligent Healthcare, Faculty of Health and Life Sciences, Coventry University, UK. 5. Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, University of Oxford, UK; MRC Population Health Research Unit, Nuffield Department of Population Health, University of Oxford, UK. 6. School of Public Health, Fudan University, China. 7. Department of Epidemiology and Biostatistics, Peking University, China. 8. Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, University of Oxford, UK. Electronic address: hubert.lam@ndph.ox.ac.uk.
Abstract
BACKGROUND: Previous studies of the health impact of ambient and household air pollution (AAP/HAP) have chiefly relied on self-reported and/or address-based exposure modelling data. We assessed the feasibility of collecting and integrating detailed personal exposure data in different settings and seasons. METHODS/ DESIGN: We recruited 477 participants (mean age 58 years, 72% women) from three (two rural [Gansu/Henan] and one urban [Suzhou]) study areas in the China Kadoorie Biobank, based on their previously reported fuel use patterns. A time-resolved monitor (PATS+CO) was used to measure continuously for 120-hour the concentration of fine particulate matter (PM2.5) at personal and household (kitchen and living room) levels in warm (May-September 2017) and cool (November 2017-January 2018) seasons, along with questionnaires on participants' characteristics (e.g. socio-demographic, and fuel use) and time-activity (48-hour). Parallel local ambient monitoring of particulate matter (PM1, PM2.5 and PM10) and gaseous pollutants (CO, ozone, nitrogen oxides) was conducted using regularly-calibrated devices. The air pollution exposure data were compared by study sites and seasons. FINDINGS: Overall 76% reported cooking at least weekly (regular-cooks), and 48% (urban 1%, rural 65%) used solid fuels (wood/coal) for cooking. Winter heating was more common in rural sites than in urban site (74-91% vs 17% daily), and mainly involved solid fuels. Mixed use of clean and solid fuels was common for cooking in rural areas (38%) but not for heating (0%). Overall, the measured mean PM2.5 levels were 2-3 fold higher in the cool than warm season, and in rural (e.g. kitchen: Gansuwarm_season = 142.3 µg/m3; Gansucool_season = 508.1 µg/m3; Henanwarm_season = 77.5 µg/m3; Henancool_season = 222.3 µg/m3) than urban sites (Suzhouwarm_season = 41.6 µg/m3; Suzhoucool_season = 81.6 µg/m3). The levels recorded tended to be the highest in kitchens, followed by personal, living room and outdoor. Time-resolved data show prominent peaks consistently recorded in the kitchen at typical cooking times, and sustained elevated PM2.5 levels (> 100 µg/m3) were observed in rural areas where use of solid fuels for heating was common. DISCUSSION: Personal air pollution exposure can be readily assessed using a low-cost time-resolved monitor in different settings, which, in combination with other personal and health outcome data, will enable reliable assessment of the long-term health effects of HAP/AAP exposures in general populations.
BACKGROUND: Previous studies of the health impact of ambient and household air pollution (AAP/HAP) have chiefly relied on self-reported and/or address-based exposure modelling data. We assessed the feasibility of collecting and integrating detailed personal exposure data in different settings and seasons. METHODS/ DESIGN: We recruited 477 participants (mean age 58 years, 72% women) from three (two rural [Gansu/Henan] and one urban [Suzhou]) study areas in the China Kadoorie Biobank, based on their previously reported fuel use patterns. A time-resolved monitor (PATS+CO) was used to measure continuously for 120-hour the concentration of fine particulate matter (PM2.5) at personal and household (kitchen and living room) levels in warm (May-September 2017) and cool (November 2017-January 2018) seasons, along with questionnaires on participants' characteristics (e.g. socio-demographic, and fuel use) and time-activity (48-hour). Parallel local ambient monitoring of particulate matter (PM1, PM2.5 and PM10) and gaseous pollutants (CO, ozone, nitrogen oxides) was conducted using regularly-calibrated devices. The air pollution exposure data were compared by study sites and seasons. FINDINGS: Overall 76% reported cooking at least weekly (regular-cooks), and 48% (urban 1%, rural 65%) used solid fuels (wood/coal) for cooking. Winter heating was more common in rural sites than in urban site (74-91% vs 17% daily), and mainly involved solid fuels. Mixed use of clean and solid fuels was common for cooking in rural areas (38%) but not for heating (0%). Overall, the measured mean PM2.5 levels were 2-3 fold higher in the cool than warm season, and in rural (e.g. kitchen: Gansuwarm_season = 142.3 µg/m3; Gansucool_season = 508.1 µg/m3; Henanwarm_season = 77.5 µg/m3; Henancool_season = 222.3 µg/m3) than urban sites (Suzhouwarm_season = 41.6 µg/m3; Suzhoucool_season = 81.6 µg/m3). The levels recorded tended to be the highest in kitchens, followed by personal, living room and outdoor. Time-resolved data show prominent peaks consistently recorded in the kitchen at typical cooking times, and sustained elevated PM2.5 levels (> 100 µg/m3) were observed in rural areas where use of solid fuels for heating was common. DISCUSSION: Personal air pollution exposure can be readily assessed using a low-cost time-resolved monitor in different settings, which, in combination with other personal and health outcome data, will enable reliable assessment of the long-term health effects of HAP/AAP exposures in general populations.