Extrapolating personal exposure measurements to air pollution from outdoor sources into annual mean estimates

A persistent disconnect exists between epidemiological studies emphasising long-term health effects and exposure assessments relying on short-term measurements, as well as between health responses linked to outdoor (ambient) concentrations versus translation of that to indoor or personal environments. To fill these gaps, we developed an extrapolation method to estimate annual personal exposure to air pollutants from outdoor sources, including fine particles, nitrogen dioxide, ozone and black carbon, using measurements from personal exposure campaigns. The methodology accounts for (a) hourly variations in background ambient concentrations, (b) magnitude and variation in infiltration efficiency, (c) adjustment for microenvironments, and (d) individual time activity patterns. Longer measurement periods appeared to be associated with greater agreement and lower variability between measurement-period and annually extrapolated estimates of personal exposure to PM_2.5 and BC from outdoor sources. We found that ambient measurements assigned by the nearest monitor approach were consistently higher than those estimated by the matched-scaled approach and higher than personal exposure to air pollution from outdoor sources for all pollutants, indicating potential bias in epidemiological associations, under specific conditions concerning variability and random error. Given this discrepancy, caution is warranted when applying outdoor air quality health guideline values to indoor environments.