Estimating pesticide exposures: areas of strength and weakness in epidemiologic studies

摘要

Comparing the merits of toxicology and epidemiology studies can become like parents extolling their child's prowess at a school function or sporting event. As an epidemiologist, I consider the real-life exposure experience in human populations to be highly relevant and informative for public health. However, just as I know (deep down) that my daughter will never play soccer in the Olympics, the practice of exposure assessment by epidemiologists is not at a world class level. We, epidemiologists, can learn from the robust nature of toxicology studies. The US EPA Office of Pesticide Programs (OPP) released its Framework for Incorporating Human Epidemiologic & Incident Data in Risk Assessments of Pesticides in 2016. Therein, the OPP presented quality considerations (low, medium and high) for exposure assessment using questionnaires and for biomonitoring data. These score sheets are a starting point for discussion. Knowing that we must rely upon observational methods and that the human dose levels are much lower than for laboratory animals, the demand for excellence is high. Epidemiologists need to ensure that biologically relevant exposure data are collected, that the data are sufficiently transparent, that measurement error has been quantified and that multiple exposures have been assessed.