Abstract
BACKGROUND: Glyphosate is a commonly used herbicide worldwide and is purportedly associated with multiple health effects. Research assessing the association of glyphosate concentrations with glycosylated hemoglobin (HbA1c) levels and the prevalence of diabetes is scarce. We sought to evaluate the association between urinary glyphosate levels and HbA1c levels and the prevalence of diabetes. METHODS: A total of 2,745 adults in the National Health and Nutrition Examination Survey from 2013 to 2016 were included in this study. Generalized linear models (GLM) were applied to evaluate the associations of glyphosate concentrations with HbA1c levels and the prevalence of diabetes. The dose-response relationship was examined using restricted cubic splines (RCS). RESULTS: Significantly positive correlations of urinary glyphosate concentrations with HbA1c levels (percentage change: 1.45; 95% CI: 0.95, 1.96; P < 0.001) and the prevalence of diabetes (OR: 1.45; 95% CI: 1.24, 1.68; P < 0.001) were found after adjustment. Compared with the lowest quartile of glyphosate levels, the highest quartile was positively associated with HbA1c levels (percentage change: 4.19; 95% CI: 2.54, 5.85; P < 0.001) and the prevalence of diabetes (OR: 1.89; 95% CI: 1.37, 2.63; P < 0.001). The RCS curves demonstrated a monotonically increasing dose-response relationship between urinary glyphosate levels and the prevalence of diabetes and HbA1c levels. CONCLUSIONS: Urinary glyphosate concentrations are positively associated with HBA1c levels and the prevalence of diabetes. To verify our findings, additional large-scale prospective investigations are required.