Calculating human exposure to endocrine disrupting pesticides via agricultural and non-agricultural exposure routes.

Endocrine Disrupting Chemicals (EDCs) are of increasing concern because of their potential impacts on the environment, wildlife and human health. Pesticides and some pesticide metabolites are an important group of EDC, and exposure to them is a poorly quantified source of human and environmental exposure to such chemicals generally. Models for estimating human exposure to Endocrine Disrupting (ED) pesticides are an important risk management tool. Probabilistic models are now being used in addition to deterministic ones in all areas of risk assessment. These can provide more realistic exposure estimates, because they are better able to deal with variation and uncertainty more effectively and better inform risk management decisions. Deterministic models are still used and are of great value where exposure data are scarce. Models or groups of models that provide holistic human ED pesticide exposure estimates are required if the risk posed to humans by ED pesticides is to be better assessed. Much more research is needed to quantify different exposure routes such as exposure from agricultural spray drift and the medical use of pesticides to develop such models. Most available probabilistic models of human exposure were developed in the USA and require modification for use elsewhere. In particular, datasets equivalent to those used to create and apply the American models are required. This paper examines the known routes of human pesticide exposure with particular reference to ED pesticides and their quantification as unlike pesticides generally, many ED pesticides are harmful at very low doses, especially if exposure occurs during sensitive stages of development, producing effects that may not manifest for many years or that affect descendants via epigenetic changes. It also summarises available deterministic and probabilistic models commonly used to calculate human exposure. The main requirement if such models are to be used in the UK is more quantitative data on the sources and pathways of human ED pesticide exposure.