A risk assessment of atrazine use in California: human health and ecological aspects

Abstract

A risk assessment of the triazine herbicide atrazine has been conducted by first analyzing the toxicity database and subsequently estimating exposure. Margins of safety (MOS) were then calculated. Toxicity was assessed in animal studies and exposure was estimated from occupational and dietary sources. In acute toxicity studies, atrazine caused developmental toxicity in the rabbit [no observed effect level (NOEL) 5 mg kg⁻¹ day⁻¹] and cardiotoxicity in a dog chronic study (NOEL 0.5 mg kg⁻¹ day⁻¹); cancer (mammary glands) resulted from lifetime exposure. The mammary tumors, which occurred specifically in female Sprague–Dawley rats, were malignant, increased in a dose‐dependent manner and were also observed with other, related triazines. Evidence for a genotoxic basis for these tumors was either equivocal or negative. Triazines have been shown to be clastogenic in Chinese hamster ovary cells, in vitro, but without showing a convincing dose/response relationship. Atrazine can be converted into genotoxic N‐nitrosoatrazine in the environment or the digestive system, suggesting that N‐nitrosamines derived from triazines could be oncogenic. However, it was concluded that N‐nitrosotriazines are unlikely to play a significant role in triazine‐induced rat mammary gland tumors. An endocrine basis for the mammary tumors, involving premature aging of the female SD rat reproductive system, has been proposed. A suppression of the luteinizing hormone surge during the estrus cycle by atrazine leads to the maintenance of elevated blood levels of 17β‐estradiol (E₂) and prolactin. The mechanism for tumor development may include one or more of the following: the induction of aromatase (CYP19) and/or other P450 oxygenases, an antagonist action at the estrogen feedback receptor in the hypothalamus, an agonist action at the mammary gland estrogen receptor or an effect on adrenergic neurons in the hypothalamic–pituitary pathway. None of these has been excluded as a target because there has been a lack of a rigorous attempt to address the mechanism of action for mammary tumors at the molecular level. The potential occupational exposure to atrazine was assessed during mixing, loading and application. Absorbed daily dosage values were 1.8–6.1 µg kg⁻¹ day⁻¹. The MOS values (animal NOEL/human exposure) for short‐term (acute) exposure were 820–2800. Longer‐term occupational exposure and risk were also calculated. Detectable crop residues are generally absent at harvest. Theoretical calculations of acute dietary exposure used tolerance levels, along with secondary residues, and water, for which there is a maximum contamination level; atrazine plus the three main chlorotriazine metabolites were combined. MOS values were above 2000 for all population subgroups. Dietary exposure to atrazine is therefore extremely unlikely to result in human health hazard. Recent publications have reported a possible feminization of frogs, measured in laboratory and field studies. This is assumed to be due to the induction of aromatase, but no measurements of enzyme activity have been reported. In field studies, the water bodies with the greatest numbers of deformed frogs sometimes had the lowest concentrations of atrazine. Other studies have also cast doubt on the feminization theory, except perhaps at very high levels of atrazine. Epidemiology studies have investigated the possibility that atrazine may result in adverse effects in humans. Although some studies have claimed that atrazine exposure results in an elevated risk of prostate cancer, the published literature is inconclusive with respect to cancer incidence. Copyright © 2005 Society of Chemical Industry