Statistical and epidemiological methods for studying the medium-short term association between industrial pollution and children's health

Abstract Background The present work is based on the “Viadana Study”, an important project aimed at assessing the impact of the wood industry on the health of the local population, which involved most part of the 3 years of my doctoral course The District of Viadana (Mantua, Northern Italy) is one of the largest industrial area for wood processing and chipboard production. Children aged 3-14 were chosen as target population for evaluating the effects of air pollution on public health. A first publication on the Viadana Study was published by de Marco et al. (de Marco et al. 2010) and the paper showed that living in the proximity of chipboard industries is a health hazard for children: children living close (<2 km of radius) to the chipboard industries had a statistically significant higher prevalence of eye irritations and respiratory symptoms. However, the use of the distance as a proxy of exposure implied the previously mentioned assumption of isotropic dispersion of air pollutants around the emission point sources. This assumption may be acceptable only in preliminary analyses and cannot be used to prove a clear link between symptoms and chipboard industry-related pollutants. For this reason, this pubblication was aimed at clarifying this association. Materials and Methods The investigation of the genotoxic damage was conducted on a sample selected from the cohort of children / adolescents (n = 3854) who had participated in the study of Viadana in December 2006 (de Marco et al. 2010). A follow-up questionnaire was administered to the children parent. Out of 656 selected children, 421 questionnaires was compiled. To each respondent child, a sample of cells of the oral mucosa of each participant was analyzed for the measurement of the genotoxic damage using Comet Assay and Micronucleus tests. The levels of formaldehyde and NO2 were monitored in the territory through ad hoc campaigns, conducted with a network of passive samplers, both in the warm season than in the cold one. The choice of monitoring sites was performed using a procedure that allowed us to take into account both the distribution of the target population, and pollution levels expected in the area. The pollution levels measured at monitoring sites were then used to estimate outdoor concentrations at homes of children, using techniques based on the Kriging spatial interpolation. Results The estimation of the pollution in the area, based on geo-statistical approach, confirmed the presence of a higher formaldehyde concentration in the south of the district, in particular around the chipboard placed in the south-west. The concentration of NO2 seems more intense around the towns and around main crossroads. Using these objective data of pollution, a positive and significant association between pollution and DNA damage measured by the comet and the micro-nucleus tests was confirmed. The use of objective markers of exposure and damage (comet assay and micronucleus test) highlighted the existence of a direct "early" cellular damage. Furthermore, preliminary analysis also showed a statistically significant association between exposure to formaldehyde and self-reported respiratory diseases in the follow-up questionnaire. Conclusions The use of statistical methods in epidemiological study answered the questions of this research. The follow-up study of Viadana has documented the existence of a significant increase in genotoxic damage in children exposed to higher levels of formaldehyde and NO2. In particular, for the spatial distribution highlighted in the present study, the formaldehyde can be considered a specific proxy of the activity of industries in the chipboard sector. Putting together the results of the two epidemiological investigations of Viadana, there is sufficient evidence supporting a causal relationship between exposure to pollutants emitted by chipboard industries and the health effects (genotoxic damage, symptoms and hospitalizations) of the population living near the plants.