Glyphosate internal dose estimation: comparing passive dosimetry and biomonitoring in simulated heavy residential herbicide application

Abstract

Glyphosate exposure is commonly assessed using either passive dosimetry or biomonitoring. This study tests the hypothesis that internal dose estimates derived from passive dosimetry are higher and more uncertain than those derived from biomonitoring by comparing concurrently collected passive dosimetry and biomonitoring data from a controlled simulation of heavy residential herbicide application. The two predominant exposure routes were evaluated separately: six applicators were protected from dermal exposures with hooded Tyvek^® coveralls and gloves and six were protected from inhalation exposure with a respirator. Urine was serially collected from all applicators between 30 min prior to and up to 36 h after application, while breathing-zone air samplers (left and right lapel) and four dermal patches were collected from dermally protected applicators. Internal doses were independently estimated from air and dermal patch measurements (passive dosimetry) and from urinary glyphosate residues (biomonitoring; dilution-adjusted). Average total internal daily doses of glyphosate estimated from passive dosimetry were approximately seven times greater than biomonitoring-based estimates. This overestimation appeared attributable to the dermal pathway and was consistent with the uncertainty in the assumed dermal absorption and clothing penetration factors as well as with the uncertainty introduced by extrapolating a limited number of dermal patch measurements to larger body regions. Even so, the highest individual internal daily dose of glyphosate derived from passive dosimetry remained below the internaldose benchmark derived from conservative regulatory health-based guidance values (60 μg/kg/day), suggesting a low likelihood of adverse health effects under the tested conditions. By directly comparing concurrently collected passive dosimetry and biomonitoring data in applicators with detectable urinary glyphosate, this study provides empirical support for using biomonitoring to evaluate and refine dermal passive-dosimetry assumptions in glyphosate exposure assessment.