[Exposure assessment to air pollutants: dispersion models versus human biomonitoring]

the assessment of individual exposure to toxicants in industrially contaminated areas is difficult when multiple productions are actively close to residential areas. Two thermoelectric power plants (one recently converted to coal) and a large harbour have been operating since the Sixties in the area of Civitavecchia (North of Rome, Lazio Region, Central Italy). Detailed exposure assessment of pollutants from industrial sources, heating (biomass combustion), and urban traffic were not available.

to assess the individual exposure of residents by using both dispersion models and human biomonitoring.

cross-sectional study.

residents in the area of Civitavecchia since 2001 aged 35-69 years were enrolled and their addresses were geocoded. Exposure assessment to power plants emissions (PM10), traffic (NOx), harbour (PM10), and biomass burning (PM10) was performed at individual residential address using Lagrangian dispersion models. The "Ambiente e Biomarcatori a Civitavecchia" study (Environment and biomarkers in Civitavecchia - ABC study) made available urinary concentrations of metals (i.e., Cd, Pt, W, Pd, Hg, Tl, Cr, Pd, Ni, As) in a sample of residents, while information on personal characteristics, lifestyles, work history, clinical history, and use of drugs was collected by interview.

for each metal, a linear regression analysis was implemented to study the association between its log-transformed values (adjusted for urinary creatinine) and the linear terms of the estimated pollutants concentrations, adjusted for age, gender, period, and other risk factors (i.e., educational level, smoking habit, alcohol, BMI). Geometric Mean Ratios (GMR) and the corresponding confidence intervals (95%CI) were computed.

in the ABC sample of 1,141 residents (42% male, mean age 53.5 years, SD 9.7), there was a positive association between PM10 from power plants and urinary cadmium (Cd) (GMR: 1.12; 95%CI 1.00-1.25), traffic pollution (NOx) and platinum (Pt) (GMR: 1.17; 95%CI 1.00-1.38), PM10 from biomass burning and Cd (GMR: 1.14; 95%CI 1.05-1.24), and tungsten (W) (GMR: 1.19; 95%CI 1.03-1.37) and palladium (Pd) (GMR: 1.11; 95%CI 1.00-1.23). Other associations observed did not reach statistical significance.

although it may not be said that dispersion models can replace biomonitoring studies, they certainly represent an excellent tool for exposure assessment and, therefore, they may be used for the estimation of the individual exposure of populations living in industrially contaminated areas.