Abstract for the 2017 HEI Annual Conference
Metabolomic Indicators of Exposure to Primary Traffic for use in Air Pollution Epidemiologic Modeling
Donghai Liang1, Armistead G. Russell2, Rachel Golan3, Jennifer L. Moutinho2, Tianwei Yu1, Chandresh N. Ladva1, Roby Greenwald4, Stefanie Ebelt Sarnat1, Dean P. Jones1, Jeremy A. Sarnat1
1Emory University, GA; 2Georgia Institute of Technology, GA; 3Ben Gurion University, Israel; 4Georgia State University, GA
Background Traffic pollution health studies increasingly focus on the identification of sensitive, biologically-relevant indicators of exposure and response. Environmental metabolomics, where metabolites associated with endogenous and exogenous processes can be quantitated, holds promise as a powerful tool for improving internal exposure estimation to complex air pollution mixtures, including primary traffic emissions. To date, environmental metabolomics applications have either been conducted in cohorts of several thousand or in smaller panels of individuals exposed to extremely elevated concentrations of specific chemicals in occupational settings.
Methods We conducted the Dorm Room Inhalation to Vehicle Emissions (DRIVE) Study to measure traditional and novel primary traffic indicators along a complete emissions-to-dose pathway. Intensive field sampling was conducted on the campus of the Georgia Institute of Technology (GIT) at 8 monitoring sites (2 indoor and 6 outdoor) ranging from 0.01 to 2.3 km away from a congested highway artery in Atlanta In addition, 54 students living in GIT dormitories either near (20 m) or far (1.4 km) from the highway conducted personal sampling and contributed weekly biomonitoring (plasma and saliva). We used targeted and untargeted metabolomics-wide association analyses to examine associations between primary traffic and corresponding metabolomics profiles in the panel.
Results Exposures to traffic pollution differed between students living in the near and far dorms. A total of 20,766 metabolic features were reliably extracted from plasma and 29,013 from saliva samples. Linear random effects models were conducted to examine associations between feature intensity (relative concentration) and level of each single traffic pollutant indicator (BC, CO, NO, NO2, NOx, and PM2.5). In total, over 597 features were robustly identified and significantly associated with at least one or more single pollutant indicators (p < 0.05, Benjamini–Hochberg FDR correction). Of these features, 294 had matching mass to charge ratios (m/z) with metabolites in the Human Metabolome Database (HMDB), and 14 were identified with matching m/z ratios with features listed in EPA’s curated list of air toxic pollutants.
Conclusions This study is among the first to examine the metabolic response to complex traffic exposures. Collectively, the DRIVE metabolomics results demonstrate initial proof of concept for this approach in being able to resolve statistically robust metabolic differences in a small panel setting. Comprehensive pathway analysis and validation is currently being conducted to identify specific metabolite patterns and further develop biologically-relevant indicators to primary traffic exposures for use in panel-based exposure and epidemiologic studies.