Airborne manganese, blood levels and neurobehavioral exposure response in smelter workers

Background and Objectives. Appropriate exposure metrics for manganese neurotoxicity need to be identified and the progression of impairment described. Methods. In a previously studied population of 115 silicon- and ferromanganese smelter workers and 145 community controls, Mn blood levels were modeled using multiple linear regression for dust (large respirable particulate) and fume (small respirable particulate). Neurobehavioral outcomes were analyzed using exposure metrics that included burdens and cumulative burdens. Results. Duration of employment was a strong predictor of Mn blood levels (t = 6.95, p <0.00001) and stronger with an 80-day half-life applied (t = 7.44, p <0.00001). Blood levels largely track the small respirable particulate fraction (6-day half-life). Using the square root of Mn concentrations (t = 7.87) produced a better model fit (R2 = 0.2334 vs. 0.2162). For three out of eight neurobehavioral outcomes, duration of Mn exposure was the best predictor with Luria Motor Scale having the strongest association (t = 5.0, p < 10-6). For three others the duration and fume metrics were comparable. When workers evaluated in both 1990 and 2004 were analyzed with a random-effects mixed model, seven of the eight outcomes had both reversible and progressive features. Conclusions. Small respirable Mn particulate (<< 0.1 µm diam.) best predicts blood Mn. Strong prediction by both duration and cumulative square root of Mn fume suggests homeostatic mechanisms are regulating systemic Mn levels. Observation of reversible or progressive manifestations of Mn exposure may help elucidate underlying mechanisms and better identify requirements for exposure control.