Background: Incidence of hypospadias varies considerably across countries, ranging from 4 to 43 cases per 10,000 births. Environmental factors might explain these differences. The classical approach is to use case-control studies to identify these factors. However, this approach suffers from the unavoidable arbitrariness of the definition of controls, and of recall bias.
Objective and hypotheses: To identify environmental markers of the place of birth of children with hypospadias.
Method: 8766 patients were recruited by 15 surgical centers. Addresses at birth were geolocalized using ArcGIS 9.3.1. This was used to map the addresses of the patients i) with socioeconomic variables available from INSEE and anonymous public databases (spatial resolution:1 km.) ii) with the land cover characteristics assessed with satellite imaging (resolution: 250 m, database: Corine Land Cover). Then 100 sets of 8766 matched virtual controls were chosen randomly on the map thanks to the pps software and provided the reference environmental values. Bonferroni correction for multiple comparison was used to assess test significance.
Results: Overall, a significant positive association was found with urbanization. When restricted to the 2279 patients whose residence was >10 km from their recruiting surgical center, more hypospadias were found in places close to non-irrigated arable lands, complex cultivation patterns, and broad-leaved forests. A residence close from vineyards was not found predictive of hypospadias.
Conclusion: Few environmental correlations, none of which being associated with pesticides, were found with our approach, which relies on an objective and unbiased assessment of environmental exposures, but may still lack spatial and temporal resolution.
10 - 12 Sep 2016
European Society for Paediatric Endocrinology