Can one use ambient air concentration data to estimate personal and population exposures to particles? An approach...

The objective of this paper is to devise a way to facilitate the use of fixed air monitors data in order to assess population exposure. A weighting scheme that uses the data from different monitoring sites and takes into account the time-activity patterns of the study population is proposed. PM2.5 personal monitoring data were obtained within the European EXPOLIS study, in Grenoble, France (40 adult non-smoking volunteers, winter 1997). Volunteers carried PM2.5 personal monitors during 48 h and filled in time-activity diaries. Workplaces and places of residence were classified into two categories using a Geographic Information System (GIS): some volunteers' life environments are seen as best represented by PM10 ambient air monitors located in urban background sites; others by monitors situated close to high traffic density sites (proximity sites). Measurements from the Grenoble fixed monitoring network using a TEOM PM10 sampler were available across the same period for these two types of sites (PM10block and PM10prox). These data were used to compute a translator parameter deltai that forces the measured PM2.5 personal exposures (PM2.5persoi) to equate the average PM10 urban ambient air concentrations ([PM10back + PM10prox]/2) measured the same days. Average deltai was 4.2 microg/m3 (CI95%[-3.4; 11.9]), with true average PM2.5 personal exposure being 36.2 microg/m3 (28.2; 44.1). PM10 ambient levels at the proximity site and at the background site were respectively PM10prox = 43.8 microg/m3 (37.1; 50.6) and PM10back = 37.0 microg/m3 (31.8; 42.3). In order to assess the consistency of this approach, six scenarios of 'proximity' and 'background' environments were accommodated, according to traffic intensity and road distance. Deltai was estimated for the entire EXPOLIS population and for subgroups, using terciles based on the percentage of time spent in proximity by each subject. Other similar studies need to be conducted in different urban settings, and with other pollutants, in order to assess the generalizability of this simple approach to estimate population exposures from air quality surveillance data.