Investigation of biomarkers related to manganese exposure in metalworkers

Abstract

Manganese (Mn) is found in both environmental and occupational settings, and can cause neurological conditions affecting cognition and mobility. Existing Mn exposure biomarker studies have not reached consensus on a valid and reproducible biomarker when looking cross-sectionally at biological specimens from Mn-exposed individuals. Here, we investigate biomarkers of Mn in three metalworking cohorts using a variety of novel techniques. A mixed model was fit to longitudinal air and blood Mn (MnB measurements to obtain estimates of airborne exposure. Estimated Mn exposures were summed over exposure windows and related to measured MnB. We found that after 30 days of exposure, an increase in air Mn was associated with an increase in MnB. Using the same exposure model, the relationship between modeled air Mn and Mn at the target organ (as determined by magnetic resonance imaging) was investigated. T1-weighted indices of basal ganglia exhibited increases in signal intensity in relation to increased Mn exposure, demonstrating that T1-weighted changes can be detected in the brain even at low levels of exposure. Global and targeted urine metabolomics data, generated using mass spectrometry (MS) and nuclear magnetic resonance spectroscopy (NMR), were utilized to investigate biochemical differences between persons exposed and unexposed to Mn. We used univariate analysis to investigate differences between groups defined by exposure. The reproducibility of the global (MS, NMR) and targeted (NMR) metabolomics profiles were investigated by building elastic net models in the primary data set, and testing the models in a second sample collected from the study participants, and in samples from a different cohort. From univariate global analyses, nine ions were found to be related to Mn exposure using the MS data, but no NMR bins were found to be associated with exposure. Mn exposure could be classified fairly well in a second set of samples from the study participants using the elastic net models, however the models were less robust when using samples from a different cohort. This discrepant results could be due to differences in Mn exposure. Work presented here underscores the challenges of using and identifying putative biomarkers for use in occupational settings.