Does improved exposure information for PM2.5 constituents explain differing results among epidemiological studies?

Abstract

Contrary findings are often found among epidemiological studies examining associations of different types of airborne particulates against the same health endpoints. Some studies of heart rate variability (HRV) in humans find associations with either regional particulate material 2.5 microns or smaller (PM_2.5) and/or with “sulfate” while some do not; some find associations with more local emissions such as black carbon (BC), while others do not. We explore if there might there be a consistent methodological explanation for inconsistent findings among HRV studies. To do this, we identify studies of HRV changes in humans examining associations with ambient PM_2.5 and sulfate, ambient PM_2.5 and BC, or all three; we briefly review findings and methodologies, including exposure issues; then we explore why studies may come to different conclusions.

We tentatively conclude that differences in accuracy of subject exposure information for health-relevant emissions such as BC, which vary spatially over short distances in urban areas, may explain conflicting study results. HRV studies with accurate exposure information for BC or urban/industrial PM_2.5 generally find large, significant associations with BC or urban/industrial PM_2.5, but rarely with secondary sulfate or regional emissions generally. However, absent accurate exposure information for BC, studies appear more likely to find associations with less spatially variable secondary sulfate or PM_2.5, and less likely to find strong associations with BC. However, research on this subject is limited, as are the number of studies evaluated here. Added research is necessary to confirm these findings (or otherwise), and to explore whether exposure misclassification might cause other health effects results to consistently vary.

Acknowledgments

Declaration of interest: The authors report no conflicts of interest.