Epidemiological studies frequently use central site concentrations as surrogates of exposure to
air pollutants. Variability in
air pollutant infiltration due to differential air exchange rates (AERs) is potentially a major factor affecting the relationship between central site concentrations and actual exposure, and may thus influence observed health risk estimates. In this analysis, we examined AER as an effect modifier of associations between several urban
air pollutants and corresponding emergency department (ED) visits for
asthma and wheeze during a 4-year study period (January 1999-December 2002) for a 186 ZIP code area in metro Atlanta. We found positive associations for the interaction between AER and pollution on
asthma ED visits for both
carbon monoxide (CO) and
nitrogen oxides (NO(x)), indicating significant or near-significant effect modification by AER on the
pollutant risk-ratio estimates. In contrast, the interaction term between
particulate matter (PM)(2.5) and AER on
asthma ED visits was negative and significant. However, alternative distributional tertile analyses showed PM(2.5) and AER epidemiological model results to be similar to those found for NOx and CO (namely, increasing risk ratios (RRs) with increasing AERs when ambient PM(2.5) concentrations were below the highest tertile of their distribution). Despite the fact that
ozone (O(3)) was a strong independent predictor of
asthma ED visits in our main analysis, we found no O(3)-AER effect modification. To our knowledge, our findings for CO, NOx, and PM(2.5) are the first to provide an indication of short-term (i.e., daily) effect modification of multiple air pollution-related risk associations with daily changes in AER. Although limited to one outcome category in a single large urban locale, the findings suggest that the use of relatively simple and easy-to-derive AER surrogates may reflect intraurban differences in short-term exposures to
pollutants of ambient origin.