Predictors of Dermal Exposures to Polycyclic Aromatic Compounds Among Hot-Mix Asphalt Paving Workers

Objectives: The primary objective of this study was to identify the source and work practices that affect dermal exposure to polycyclic aromatic compounds (PACs) among hot-mix asphalt (HMA) paving workers. Methods: Four workers were recruited from each of three asphalt paving crews (12 workers) and were monitored for three consecutive days over 4 weeks for a total of 12 sampling days per worker (144 worker days). Two sampling weeks were conducted under standard conditions for dermal exposures. The third week included the substitution of biodiesel for diesel oil used to clean tools and equipment and the fourth week included dermal protection through the use of gloves, hat and neck cloth, clean pants, and long-sleeved shirts. Dermal exposure to PACs was quantified using two methods: a passive organic dermal (POD) sampler specifically developed for this study and a sunflower oil hand wash technique. Linear mixed-effects models were used to evaluate predictors of PAC exposures. Results: Dermal exposures measured under all conditions via POD and hand wash were low with most samples for each analyte being below the limit of the detection with the exception of phenanthrene and pyrene. The geometric mean (GM) concentrations of phenanthrene were 0.69 ng cm−2 on the polypropylene layer of the POD sampler and 1.37 ng cm−2 in the hand wash sample. The GM concentrations of pyrene were 0.30 ng cm−2 on the polypropylene layer of the POD sampler and 0.29 ng cm−2 in the hand wash sample. Both the biodiesel substitution and dermal protection scenarios were effective in reducing dermal exposures. Based on the results of multivariate linear mixed-effects models, increasing frequency of glove use was associated with significant (P < 0.0001) reductions for hand wash and POD phenanthrene and pyrene concentrations; percent reductions ranged from 40 to 90%. Similar reductions in hand wash concentrations of phenanthrene (P = 0.01) and pyrene (P = 0.003) were observed when biodiesel was substituted for diesel oil as a cleaning agent, although reductions were not significant for the POD sampler data. Although task was not a predictor of dermal exposure, job site characteristics such as HMA application temperature, asphalt grade, and asphalt application rate (tons per hour) were found to significantly affect exposure. Predictive models suggest that the combined effect of substituting biodiesel for diesel oil as a cleaning agent, frequent glove use, and reducing the HMA application tem