Richa Adhikari, Jennifer D’Souza, Elsayed Z. Soliman, Gregory L. Burke, Martha L. Daviglus, David R. Jacobs, Sung Kyun Park, Lianne Sheppard, Peter S. Thorne, Joel D. Kaufman, Timothy V. Larson, Sara Dubowsky Adar

Reduced heart rate variability, a marker of impaired cardiac autonomic function, has been linked to short-term exposure to airborne particles. This research adds to the literature by examining associations with long-term exposures to coarse particles (PM10-2.5). Using electrocardiogram recordings from 2,780 participants (45-84 years) from three Multi-Ethnic Study of Atherosclerosis sites, we assessed the standard deviation of normal-tonormal intervals (SDNN) and root-mean square differences of successive normal-to-normal intervals (rMSSD) at a baseline (2000-2002) and follow-up (2010-2012) examination (mean visits/person=1.5). Annual average concentrations of PM10-2.5 mass, copper, zinc, phosphorus, silicon, and endotoxin were estimated using site-specific spatial prediction models. We assessed associations for baseline heart rate variability and rate of change in heart rate variability over time using multivariable mixed models adjusted for time, sociodemographic, lifestyle, health, and neighborhood confounders, including co-pollutants. In our primary models adjusted for demographic and lifestyle factors and site, PM10-2.5 mass was associated with 1.0% (95% CI: -4.1, 2.1%) lower SDNN levels per interquartile range of 2 µg/m3. Stronger associations, however, were observed prior to site adjustment and with increasing residential stablity. Similar patterns were found for rMSSD. We found little evidence for associations with other chemical species and with with the rate of change in heart rate variability, though endotoxin was associated with increasing heart rate variability over time. We found only weak evidence that long-term PM10-2.5 exposures are associated with lowered heart rate variability. Stronger associations among residentially stable individuals suggest that confirmatory studies are needed.