Personal exposures to PM(2.5) and polycyclic aromatic hydrocarbons and their relationship to environmental tobacco smoke at two locations in Greece

Abstract

In the context of a large-scale molecular epidemiology study of biomarkers of genotoxicity of air pollution, 24-h mean personal exposures to airborne PM(2.5) (particulate matter <2.5 microm) and associated polycyclic aromatic hydrocarbon (PAHs) were measured in 194 non-smoking technical institute students living in the city of Athens, Greece (an area with moderately high levels of air pollution) and the nearby small town of Halkida anticipated to have lower pollution levels. Extensive information relevant to the assessment of long-term and recent exposure to PAH was obtained from questionnaires as well as a time-location-activity diary (TLAD) which was kept by all subjects during a 4-day observation period. During the last 24 h of this period, subjects underwent personal exposure monitoring for PM(2.5) and PAH, while a sample of blood was donated at the end of this period. All subjects were monitored in this way twice; once during a winter season (October-February) and once during the following summer season (June-September). Nine subjects with plasma cotinine levels above 20 ng/ml were considered as unreported smokers and excluded from the study. Winter PM(2.5) exposures were lower in Athens (geometric mean 39.7 microg/m(3)) than Halkida (geometric mean 56.2 microg/m(3)) (P<0.001), while there was no significant location difference during the summer (Athens: geometric mean 32.3 microg/m(3), Halkida: geometric mean 32.9 microg/m(3); P=0.79). On the other hand, PAH exposures (sum of the eight carcinogenic PAHs) were significantly higher in Athens than in Halkida during the winter (Athens: geometric mean 8.26 ng/m(3), Halkida: geometric mean 5.80 ng/m(3); P<0.001) as well as during the summer (Athens: geometric mean 4.44 ng/m(3), Halkida: geometric mean 1.48 ng/m(3); P<0.001). There was a significant difference in the profile of the PAH exposures at the two locations, the proportion of lighter PAH (benzo[a]anthracene, chrysene [CHRYS], benzo[k]fluoranthene, and benzo[b]fluoranthene) being higher, and that of heavier PAH (benzo[ghi]perylene [BPer] and indeno[1,2,3,cd]pyrene) lower, in Halkida than in Athens, regardless of season. This difference appeared to be related to individual exposure to environmental tobacco smoke (ETS), as indicated by (a) the correlation at the individual level between the CHRYS/BPer ratio and declared time of recent exposure to ETS as well as plasma cotinine levels, especially during the winter; (b) the parallel variation of the mean levels of all three markers (declared ETS exposure, cotinine levels, CHRYS/BPer ratio) among three subgroups of subjects (Athens subjects who had lowest levels of all three markers; Halkida subjects other than those living in the institute campus area; and Halkida subjects living in the institute campus area who had the highest levels of all three markers). This demonstrates that ETS can have a distinctive effect on the PAH exposure profile of subjects exposed to relatively low levels of urban air pollution.