Ambient concentrations are commonly used as proxies for personal PM2.5 exposure in epidemiological studies, despite indoor settings being the places where people spend most of their time. In a panel study of 110 nonsmoking, healthy college students in Lhasa, Tibet, indoor PM2.5 was monitored using calibrated low-cost sensors for two multiweek periods, in over 40 dormitories where participants resided. We also repeatedly measured fractional exhaled nitric oxide (FeNO), an acute respiratory inflammation biomarker, for each participant. Time-averaged indoor PM2.5 concentrations in individual dormitories ranged from 3.2 to 30 μg/m3 in the summer and from 3.6 to 57 μg/m3 in the fall, in most cases exceeding the outdoor level (4.3 and 4.9 μg/m3, respectively). The hourly mean indoor PM2.5 concentrations displayed a clear trimodal diel pattern, with peaks coincident with periods of increased activities. Further questionnaire-based analysis suggests that incense burning and smoking contributed to elevated levels of indoor PM2.5. Overnight PM2.5 levels in the dormitories were significantly associated with increased FeNO the following morning, with the effects attenuated as the hourly lag increased. In contrast, inconclusive associations were observed for ambient PM2.5. The results demonstrate that disregarding indoor exposure can result in biased estimates of acute health effects of PM2.5 in low PM2.5 areas.
Keywords: FeNO; dormitory; incense burning; low-cost sensors; personal exposure.