Exposure Model for Individuals (EMI)

Background

The Clean Air Act requires EPA to set National Ambient Air Quality Standards (NAAQS) for pollutants considered harmful to public health (primary standards), including the health of 'sensitive' populations such as asthmatics and children. According to the Clean Air Act, the primary standards for the six principal pollutants (criteria pollutants) must be based on sound science that often includes air pollution health studies used to understand the relationship between individual exposures and health outcomes. Due to cost and participant burden of indoor and personal air monitoring, health studies often estimate exposures based on outdoor ambient measurements from central‑site air monitors, and assume people are located in the same place, usually their residential address, throughout a 24 hour period. However, these ambient levels do not necessarily reflect personal exposures since pollutant concentrations observed indoors (e.g., homes, offices, factories, schools, motor vehicles) can differ from those observed at central‑site air monitors, and considerable time is spent by people in these indoor locations. According to the National Research Council report on epidemiology and air pollution, the impact of this discrepancy in the exposure of health study subjects leads to uncertainty and potential bias to risk estimates.

To address this critical need to reduce exposure error, the EPA is developing and evaluating an exposure model for individuals (EMI) in air pollution health studies (See Reference 1). The EMI is being developed to predict individual exposures for multiple air pollutants from ambient concentrations, meteorology, and questionnaire information such as building characteristics, occupant behavior related to building operation, indoor sources, and time-activity patterns. This research is being conducted in response to the needs of numerous air pollution cohort health studies, including the National Children's Study, and complements existing population‑based exposure models (e.g., SHEDS-PM, see References 2-4).

In brief, the EMI predicts personal exposure-dose metrics for individuals in air pollution health studies. The EMI inputs include outdoor air pollutant concentrations, meteorology, and questionnaire information such as building characteristics and operation, indoor sources, and time-location activity patterns. The EMI consists of three primary modules: (1) indoor air quality module, (2) time-location-activity module, and (3) respiratory tract dosimetry module. The EMI predicted exposure-dose metrics are then used for epidemiological analysis in health studies with individual health outcomes.

Expected Impact of EMI

This research supports the requirements under the Clean Air Act and recommendations by the National Research Council. The overall goal of this research is to provide a modeling tool to enhance the ability to conduct air pollution human exposure assessments with greater certainty. The EMI applications for client-specific needs will serve to strengthen the science behind regulatory decision-making on setting the National Ambient Air Quality Standards (NAAQS). This modeling effort will address significant challenges relating to:

better understanding of the linkages between ambient concentrations and individual exposures
identifying key exposure factors to help optimize the design of health studies
characterizing and reducing the uncertainty of individual exposure predictions for air pollutants associated with adverse cardiovascular and respiratory effects in "sensitive" individuals such as diabetics and asthmatic children
demonstrating the utility of individual exposure estimates for development and application of more accurate, cost-effective, and less burdensome exposure metrics for epidemiologic analysis in cohort health studies

This interdisciplinary research will leverage resources and develop research collaborations both within and outside EPA. The EMI is being developed, evaluated, and applied by EPA, the University of Michigan, Harvard University, Applied Research Associates, the German Research Center for Environmental Health, and the National Children's Study.

Objective and Results

Estimating human exposure to air pollutants and examining their influence on various health outcomes is a critical component of EPA’s human health research program. Within this context, this research focuses on the development and application of an individual‑based exposure model that will improve understanding of complex linkages between ambient concentrations and individual exposures, and apply model‑predicted exposure metrics for epidemiologic analysis in health studies. Initially, this modeling effort will provide exposure metrics for specific air pollutants associated with four cohort health studies that examine:

cardiovascular events for diabetics in central North Carolina,
exacerbation of asthma for children living near major roadways in Detroit, Michigan,
respiratory effects for asthmatics in central North Carolina, and
development of asthma for children across the country (National Children's Study).

The development and evaluation of EMI is ongoing. To support various health studies, initial development of EMI is focusing on estimating exposure to PM2.5, as summarized in an EMI presentation (PDF, 1pp, 478kb) ). To support future health studies, the EMI will be expanded for multiple pollutants.

Tiered Exposure Metrics for Individuals in Health Studies

The EMI is being developed to provide multiple tiers of individual-level exposure metrics. This tiered approach will support the development of exposure metrics appropriate for the information available from particular health studies, and enable the comparison of multiple exposure metrics with different levels of complexity.

In brief, the EMI is being developed to provide multiple tiers of individual-level exposure metrics with different levels of complexity and information needs. The four primary metrics listed from least to most complex are: (1) building air exchange rates, (2) indoor concentrations, (3) personal exposures, and (4) dose. Alternative metrics include time spent and activity level in specific microenvironments, and concentration during high activity levels. These exposure metrics are then used for epidemiological analysis in health studies with individual health outcomes.

Microenvironment Tracker (MicroTrac) for EMI

For EMI’s time-location-activity module, a microenvironment tracker (MicroTrac) is being developed. The MicroTrac utilizes personal GPS data loggers and novel algorithms to determine time spent in microenvironments as specified by EMI. The MicroTrac will provide time-location data with improved spatial and temporal resolution, and accuracy with reduced participant burden, as compared to time‑location diaries. This research supports the development and application of advanced technical capabilities for improved exposure assessments.

The development of MicroTrac is ongoing. Initially, the technical capabilities of the GPS data loggers are being characterized to make recommendations to health studies. In particular, the MicroTrac research effort initiated the incorporation of GPS data loggers into the design of a near roadway asthmatic health study in Detroit, Michigan. The algorithms needed to map GPS-derived locations into the specific microenvironments used by EMI are also being developed and evaluated with time-location diaries before applying MicroTrac in health studies.

Status of EMI

https://www.epa.gov/heasd/products/emi/emi_status.html

Questions and Answers for EMI

https://www.epa.gov/heasd/products/emi/emi_qa.html

References

1 - Breen M. S., Breen M., Williams, R. W., Schultz, B. D. (2010). Predicting residential air exchange rates from questionnaires and meteorology: model evaluation in central North Carolina. Environ. Sci. Technol., 44, 9349-9356. publication and supporting information.

2 - Burke, J., M. Zufall, H. Özkaynak (2001) A population exposure model for particulate matter: Case study results for PM2.5 in Philadelphia, PA. J Exp Anal Environ Epidemiol., 11, 470-489.

3 - Georgopoulos, P., Wang, S., Vyas, V., Sun, Q., Burke, J., Vedantham, R., McCurdy, T., Özkaynak, H. (2005) A source-to-dose assessment of population exposures to fine PM and ozone in Philadelphia, PA, during a summer 1999 episode. J Exp Anal Environ Epidemiol., 15, 439-457.

4 - Özkaynak, H., Frey, H.C., Burke, J., Pinder, R. (2009) Analysis of coupled model uncertainties in source-to-dose modeling of human exposures to ambient air pollution: A PM2.5 case study. Atmos. Environ., 43, 1641-1649.

Contacts

Michael Breen, Ph.D.
Principal Investigator
(919) 541-9409
breen.michael@epa.gov

Andrew Geller, Ph.D.
Branch Chief
(919) 541-4208
geller.andrew@epa.gov

Human Exposure and Atmospheric Sciences

Research & Development | National Exposure Research Laboratory