Anyone walking along a highway during morning rush hour is likely to get a lungful of engine exhaust, but actually they won’t be as bad off as the people in the passing cars.

That’s perhaps a surprising result of a study of the exposure of drivers to pollution done by scientists at Duke University, Emory University and the Georgia Institute of Technology. In their study published this summer, researchers say they found that the levels of soot and chemicals detected inside cars during rush hour in Atlanta were twice as high as those measured by roadside detectors.

Although exposure to traffic emissions is frequently associated with negative health impacts, few studies have measured air pollution directly in-vehicle, and limited measurements of daily commuter exposure exist. This research, part of the Atlanta Commuter Exposures (ACE) Study, assessed on-roadway in-cabin particulate pollution (PM2.5) collected from scripted rush hour commutes on highways and on non-highway side streets. Water-soluble extracts from PM2.5 filters were analyzed for oxidative potential of water-soluble species using the dithiothreitol (DTT) assay, and results suggest that there may be substantial gas-phase DTT activity in fresh emissions.

Researchers measured DTTv activities (i.e., DTT activity normalized to the sampled air volume) that were on average two times higher than comparable measurements collected by stationary roadside monitoring sites. Although some of this difference may be attributable to positive artifacts due to relatively brief (2-h) quartz filter sampling durations, the current findings provide some indication that commuters encounter notably higher exposure to redox-active PM2.5 in the on-road environment.

Strong correlations were observed between water-soluble DTT activity and water-soluble organic carbon (WSOC), specifically for the “semivolatile” WSOC component (measured as the difference between denuded and non-denuded filters). Although potential for artifacts when measuring DTT activity of fresh emissions using filter-based methods is considerable, these results suggest that semivolatile organic species are important contributors to DTT activity, at least in environments where ambient PM2.5 is dominated by vehicular sources.