Proximity to wood industries and respiratory symptoms in children: a sensitivity analysis.

Increased prevalence of respiratory and irritation symptoms was found in children who live near a large wood industrial park. Proximity to the wood industries was used as indicator of exposure. This study describes a sensitivity analysis for the results of the survey. All the children (3-14 years) living in the area were surveyed through a parental questionnaire (n=3854) and their addresses were geocoded. The distances from each child's home and school to the closest industry were combined, weighted and used as an indicator of exposure. A sensitivity analysis was performed to check 1) the robustness of the results to the choice of weights used for defining the exposure indicator, 2) the effect of outliers on risk estimates and 3) the sensitivity on the functional form used for modeling the dose-response function. The choice of the weights did not influence the association between proximity to the industries and respiratory symptoms. Excluding the subjects who lived far away from the industries showed that in a radius of 5km from the industries the study did not had enough power to estimate a gradient in the dose-response function. Besides, results were sensitive to the choice of the functional form used for modeling the minimum distance. The sensitivity analyses confirmed the overall increasing trend of respiratory symptoms with proximity to the industries and pointed out that all the assumptions made for defining a proxy of exposure need to be carefully checked.