PM2.5 metal exposures and nocturnal heart rate variability:: a panel study of boilermaker construction workers

Background: To better understand the mechanism(s) of particulate matter (PM) associated cardiovascular effects, research priorities include identifying the responsible PM characteristics. Evidence suggests that metals play a role in the cardiotoxicity of fine PM (PM2.5) and in exposure-related decreases in heart rate variability (HRV). We examined the association between daytime exposure to the metal content of PM2.5 and night HRV in a panel study of boilermaker construction workers exposed to metal-rich welding fumes. Methods: Twenty-six male workers were monitored by ambulatory electrocardiogram (ECG) on a workday while exposed to welding fume and a non-workday (baseline). From the ECG, rMSSD (square root of the mean squared differences of successive intervals) was summarized over the night (0: 00-7:00). Workday, gravimetric PM2.5 samples were analyzed by x-ray fluorescence to determine metal content. We used linear mixed effects models to assess the associations between night rMSSD and PM2.5 metal exposures both with and without adjustment for total PM2.5. Matched ECG measurements from the non-workday were used to control for individual cardiac risk factors and models were also adjusted for smoking status. To address collinearity between PM2.5 and metal content, we used a two-step approach that treated the residuals from linear regression models of each metal on PM2.5 as surrogates for the differential effects of metal exposures in models for night rMSSD. Results: The median PM2.5 exposure was 650 mu g/m(3); median metal exposures for iron, manganese, aluminum, copper, zinc, chromium, lead, and nickel ranged from 226 mu g/m(3) to non-detectable. We found inverse linear associations in exposure- response models with increased metal exposures associated with decreased night rMSSD. A statistically significant association for manganese was observed, with a decline of 0.130 msec (95% CI: -0.162, -0.098) in night rMSSD for every 1 mu g/m(3) increase in manganese. However, even after adjusting for individual metals, increases in total PM2.5 exposures were associated with declines in night rMSSD. Conclusion: These results support the cardiotoxicity of PM2.5 metal exposures, specifically manganese. However the metal component alone did not account for the observed declines in night HRV. Therefore, results suggest the importance of other PM elemental components.