ensured a sufficient statistical power to highlight a moderate
excess risk.
Our methodology also presented some limitations. The
validity or appropriateness of the dispersion models have
not been assessed (and had to be assumed over time in
this study of long-term effects), because we lacked the necessary
household and soil measurements in the contaminated
areas. Residence location as a surrogate of exposure
cannot distinguish contributions from the direct and the
indirect exposure pathways (e.g., from air to soil and