Strengths and Limitations: Integration of Exposure Sciences and Epidemiology in Health Risk Assessment of Environmental Contaminants

Risk assessment of environmental contaminants is based on an integrated evaluation of in-vitro, in-vivo and human data. More than a 100.000 new chemicals have been synthesized in the last 5 decades. A few thousand of these have become wide-spread in the environment. Of these many have not undergone testing for safety or toxicity as thorough pre-market testing has only been institutionalized in the last decade in some regions of the world. Additionally, toxicity tests are not a 100% predictive for adverse human health effects in itself. Therefore, environmental epidemiology plays a key role in post-market research and can play a role in both hazard and risk characterization. As more epidemiological studies have become available in more recent years their use in risk assessment has increased. For example, the Dutch Health Council committee on cancer risk values has increasingly used epidemiological studies to set such limits (increasing from 0% in 1995 to 25% in 2013 of all evaluations in that time period). However, the utility of environmental epidemiology in risk assessment depends strongly on the quality of the exposure assessment. For example, in a hazard evaluation we conducted for electromagnetic low field (ELF-MF) exposure and risk of Amyotrophic Lateral Sclerosis (ALS), studies that had limitations in the exposure assessment did not show a clear association while studies that had better exposure assessment showed a clear increased risk between ELF-MF exposure and ALS. Similarly, in a risk evaluation of asbestos exposure and lung cancer, we observed a difference in risk per unit of exposure of a factor of 4 between the best to all studies. These two examples are indicative of the importance of good quality exposure assessment in epidemiological studies. In order to facilitate the transparent use of epidemiological studies in risk assessment evaluation tools for study design (e.g. ROBINS-E) with a strong focus on exposure assessment quality should be used.