John Libbey Eurotext

Environnement, Risques & Santé


Scientific results and public health decisions: Some paradoxes Volume 9, issue 4, Juillet-Août 2010

EHESP École des hautes études en santé publiqueDépartement d'épidémiologie et de recherche clinique Av du Pr Léon Bernard F-35043 Rennes cedex France

The relation between environment and health most often involves risks that are low at the individual level and diseases that are generally multifactorial and not specific for a given substance or agent. This situation makes it difficult to establish causal associations between exposures and their health effects. Moreover, to determine the priority of actions, it is necessary to estimate the extent of the risk avoided if exposure is reduced or stopped (attributable risk), which requires both knowledge of the distribution of exposure in the population and an estimate of the dose-response relation. These two conditions are rarely met simultaneously. We nonetheless observe, especially in France, that even in cases where an attributable risk can be estimated, the public health actions that should logically follow it are rarely if ever implemented. Inversely, some decisions in the field of environmental health are taken on scientifically fragile bases.