Glyphosate exposure and urinary oxidative stress biomarkers in the Agricultural Health Study

Abstract

Background: Glyphosate is the most widely applied herbicide worldwide, and its use has been associated with increased risks of certain hematopoietic cancers in epidemiologic studies. Animal and in vitro experiments suggest that glyphosate may induce oxidative stress, a key characteristic of carcinogens; however, evidence in human populations remains scarce. We investigated associations between glyphosate exposure and urinary oxidative stress biomarkers in the Biomarkers of Exposure and Effect in Agriculture study, a molecular epidemiologic subcohort in the Agricultural Health Study.

Methods: This analysis included 268 male farmers selected based on self-reported recent and lifetime occupational glyphosate use and 100 age- and geography-matched male nonfarmers. Concentrations of glyphosate and oxidative stress biomarkers (8-hydroxy-2'-deoxyguanosine [8-OHdG], 8-iso-prostaglandin-F2α, and malondialdehyde [MDA]) were quantified in first-morning-void urine. We performed multivariable linear regression to evaluate associations of urinary glyphosate and self-reported glyphosate use with each oxidative stress biomarker.

Results: Urinary glyphosate concentrations were positively associated with levels of 8-OHdG (highest vs lowest glyphosate quartile; geometric mean ratio = 1.15, 95% confidence interval = 1.03 to 1.28; Ptrend = .02) and MDA (geometric mean ratio = 1.20, 95% confidence interval = 1.03 to 1.40; Ptrend = .06) overall. Among farmers reporting recent glyphosate use (last 7 days), use in the previous day was also associated with statistically significantly increased 8-OHdG and MDA levels. Compared with nonfarmers, we observed elevated 8-iso-prostaglandin-F2α levels among farmers with recent, high past 12-month, or high lifetime glyphosate use.

Conclusions: Our findings contribute to the weight of evidence supporting an association between glyphosate exposure and oxidative stress in humans and may inform evaluations of the carcinogenic potential of this herbicide.

Figure 1.

Box and whisker plots of urinary (A) 8-OHdG, (B) 8-isoprostane, and (C) MDA concentrations across quartiles of urinary glyphosate concentration among all 368 participants (quartile 1 = <LOD-0.289 µg/L; quartile 2 = 0.290-0.506 µg/L; quartile 3 = 0.507-0.933 µg/L; quartile 4 = 0.934-35.2 µg/L). The top and bottom edges of the boxes represent the upper (75th percentile) and lower (25th percentile) quartiles of the oxidative stress biomarker, respectively, and the whiskers above and below the boxes indicate the range of the data that lie within 1.5 times the IQR (ie, 1.5 IQR above the 75th percentile for the upper whisker and the minimum observed value for the lower whisker). The thick horizontal lines represent the median, and the solid circles represent the geometric mean concentration of the oxidative stress biomarker. For ease of visualization, the y-axis was truncated at a value of 3 IQRs above the 75th percentile for each oxidative stress biomarker. 8-isoprostane = 8-iso-prostaglandin-F2α; 8-OHdG = 8-hydroxy-2′-deoxyguanosine; IQR = interquartile range; LOD = limit of detection; MDA = malondialdehyde.