Citation:

Breen, M. Exposure Modeling of Individuals in ?Air Pollution Epidemiology Stu. Emory University Seminar, Atlanta, GA, March 21, 2019.

Impact/Purpose:

Dr. Breen's research focuses on development of computational models to improve exposure assessments for air pollution epidemiology studies. Environmental decision-making used to direct EPA programs and policies concerning air pollution (e.g., setting the National Ambient Air Quality Standards, NAAQS) must be based on sound science, which often include air pollution health studies. Dr. Breen's models can reduce exposure misclassifications, which can lead to uncertainty to risk estimates from air pollution health studies, and can address the needs of public health to reduce exposures for susceptible individuals.

Description:

Air pollution health studies often use outdoor concentrations from a central-site monitor as exposure surrogates. To improve exposure assessments, we previously developed and evaluated an exposure model for individuals (EMI), which predicts five tiers of individual-level exposure metrics for ambient air pollutants using outdoor concentrations, questionnaires, weather, and time-location information. We linked a mechanistic air exchange rate (AER) model to a mass-balance air pollutant infiltration model to predict residential AER (Tier 1), infiltration factors (Finf, Tier 2), indoor concentrations (Cin, Tier 3), personal exposure factors (Fpex, Tier 4), and personal exposures (E, Tier 5) for ambient air pollutants. In this study, we extended EMI by including: (1) an air quality model to predict hourly census-block outdoor concentrations for four pollutants (PM, NOX, CO, EC), (2) a GPS-based microenvironment tracker (MicroTrac) model to predict time spent by individuals in various microenvironments. We predicted daily exposure metrics (Tiers 1-5) for the 15 participants across 10 consecutive weeks in a cohort health study in central North Carolina called Coronary Artery Disease and Environmental Exposure (CADEE). Our modeled predictions for a total of 708 participant-days showed substantial house-to-house and temporal variability of AER, Finf, and Cin (Tiers 1-3); and subject-to-subject variability of Fpex and E (Tiers 4-5) for the four pollutants. The capability of EMI could help reduce uncertainty of ambient pollutant exposure metrics used in health studies, such as CADEE, in support of improving health risk estimates.

Record Details: