Predicting Chronic Personal Exposure to Airborne Polycyclic Aromatic Hydrocarbon ?Mixture in Krakow, Poland

Abstract

BACKGROUND: Polycyclic aromatic hydrocarbons (PAHs) are a globally distributed class of genotoxin and carcinogens. Inhalation PAHs, in particular, is an important route of exposure. Yet, in absence of directly monitoring of personal exposure (PE), estimating chronic PE to a mixture of airborne PAHs typically relies on proxy predictor variables. Here, we examined the performances of three prediction models of chronic PE. METHODS: In an on-going prospective birth cohort study, the pregnant women were directly monitored for their PE either once during the gestational period to eight carcinogenic PAHs and pyrene (n=266) or repeatedly, once during eachtrimester (n=78). Additional monitoring of the same PAHs in indoor and outdoor setting were also conducted. Three PE models were developed. Under the most resource-restricted scenario (model 1), we considered self-reported indoor and outdoor sources as proxy variables of PE. In the intermediate-resource scenario (model 2), we posited that the citywide demand for heating (unit, °C) is a proxy variable for PE to PAHs. Under the most resource abundant scenario (model 3), we posited that infiltrated indoor concentration of directly monitored outdoor mean concentrations represents a proxy of PE. RESULTS: Based on the Model 1, proxy variables for the pregnancy period as independent predictors explained ? 70% variability in PE to each of the nine PAHs (all Ps < 0.0001). Based on the Model 2, a single unit increase (°C) in citywide demand for heating was associated with 10% increase in PE to each of nine PAHs (all Ps < 0.0001). In addition, single unit increase in the wind speed (m/s) was associated with 30—40% reduction in PE to each of nine PAHs. Based on the Model 3, pooled 48-hour mean outdoor concentration for each of eight species explained 76—87% of total variability in PE to the same species. CONCLUSION: In Krakow, Poland, accurate and precise prediction of PE of nine PAHs is possible using inexpensive data sources.