Abstract:
Exposure to wood smoke has been associated with lung cancer and other adverse health outcomes, and therefore it is a great health concern for those in developing parts of the world as well as developed nations. Pyrene, naphthalene, fluorene, and phenanthrene are components of wood smoke and their metabolites have been frequently used as biomarkers of exposure to combustion emissions. In this study, we explored the use of polycyclic aromatic hydrocarbon (PAH) urinary metabolites to determine their effectiveness as biomarkers for wood smoke exposure. We recruited ten professional firefighters and two volunteers certified to work as firefighters on prescribed burns who performed controlled burns on the United States Department of Energy’s Savannah River Site (SRS) in North Carolina. We quantified metabolites in pre-shift and post-shift urine samples from the firefighters on 7 prescribed burn days and 3 non-burn days, which resulted in a total of 68 urine sample pairs (48 on burn days and 20 on non-burn days). We collected 69 personal air samples, including 48 samples on burn days and 21 samples on non-burn days; these were analyzed for PM2.5, levoglucosan, and light absorbing carbon (LAC). We also measured personal carbon monoxide (CO) exposures in conjunction with the personal air filter samples. We investigated associations between cross-shift PAH metabolite concentrations and the 2 main tasks performed, holding, and lighting, as well as the association between the air contaminants and PAH metabolite concentrations. The subjects’ breathing rate, derived from accelerometry data, was used to calculate the inhaled dose of particulate matter, levoglucosan, and LAC. We found a significant increase in geometric mean (GM) cross-shift metabolite concentration for all six metabolites on burn days and no significant change in GM cross-shift concentration for all six metabolites on non-burn days. 2-Napthol (2-Nap), 2+9-hydroxyfluorene (2+9-Flu), and 3-hydroxyphenanthrene (3-Phen) had the largest GM cross-shift ratios of 2.69, 2.61, and 3.19 respectively on burn days. When stratified by task, subjects that performed lighting had significant increases in the GM cross-shift ratio for all six metabolites (Range: 1.72-3.19), and for those who performed holding, all six metabolites exhibited an increase in GM cross-shift ratio, but none were significant. The burn day exposure data shows that those who performed lighting had a 4.2-fold higher GM LAC absorption coefficient, and a 2.6-fold lower levoglucosan/gravimetric PM2.5 ratio. This pattern of higher LAC with lower relative levoglucosan provides strong evidence that subjects who perform lighting are exposed to much lower levels of wood smoke than those who are holding. It also suggests that the large increase in GM cross-shift metabolite ratio in lighters is most likely due to exposure to the combustion emissions from the drip torch. Since those who performed holding had much higher levels of wood smoke exposure and substantially lower GM cross-shift metabolite ratios, it appears OH-PAH metabolites may not be effective biomarkers of wood smoke exposure in this setting. Strong associations were observed between LAC and 2-Nap, 2+9-Flu, and 3-Phen. URI http://hdl.handle.net/1773/40098