Phase 1

Mapping Health Risks:

The first phase, setting the foundation for the work to follow, will involve a detailed mapping of the health risks of PM across the United States in order to identify those cities and regions where risks have been estimated to be the highest and the lowest.  This approach addresses a critical obstacle in approaching PM characteristics and risks to health—the many potential combinations of characteristics that may be relevant to the diverse health outcomes linked to PM.  Research on PM has been complicated by the numerous characteristics of PM potentially relevant to their toxicity and the possibility that different characteristics may be relevant to different health outcomes.  Our epidemiologically-based approach provides that guiding principle, using the actually-occurring variation in risks to health as a basis for exploring the matrix defined by PM characteristics and sources and health outcomes.

The mapping of health risks in this first phase will build on the methods and data bases already assembled for the National Medicare Cohort.  These health data bases have been linked to the monitoring data for PM₁₀ and PM_2.5, as well as other pollutants, along with information on demographic characteristics from the Census Bureau and meteorology from the National Weather Center. In addition, these health data bases will be linked by geographical location to the data from the Speciation Trends Network and to the EPA’s Supersite Program.

State-of-the-art statistical methods:

Novel methods will be developed to analyze these data bases for a variety of health outcomes. Methods will build upon the previously developed Bayesian hierarchical regression models for NMMAPS, which will provide regression coefficients that estimate the increment in risk per unit increase in PM concentration for each of these outcomes.  To guard against random variation in these coefficients and to gain evidence across locations, the coefficients will be smoothed using Bayesian methods.   We anticipate generating maps for key outcomes: total mortality and cardiovascular mortality, and hospitalizations for cardiovascular outcomes (myocardial infarction and other ischemic events, congestive heart failure, and arrhythmias), for pneumonia and influenza, and for chronic obstructive pulmonary disease (COPD) and asthma.  These maps will guide the work of subsequent phases.

Analyzing Data:

During this first phase, we will also develop methods and carry out analyses of the accumulating data from Speciation Trends Network and also of the already collected data from the Supersites Program.  These analyses will have the purpose of better understanding the correlations among pollutants, the extent of intra-site and cross-site variation in pollution patterns, and their association with health outcomes.  The findings will be useful in guiding monitoring strategies for the second and third phases of the Center’s activities.

Assessing particle toxicity:

The first phase will also involve the development and testing of laboratory approaches for assessing particle toxicity.  A battery of in vitro and in vivo bioassays will be developed, building from the work of Dr. Garcia and colleagues in lung injury and in pulmonary/cardiovascular interactions.  The range of assays will include not only widely-used indicators of injury, e.g., inflammatory cytokines and markers of oxidant activity, but novel approaches based on microarrays for gene expression.  Mouse models will be used that include not only wild type C57B6 mice, but murine models of asthma and congestive heart failure (CHF).  In this first phase, these assays will be assessed using a set of real-world PM collected from multiple, distinct sites. The selected battery of genes will reflect candidates identified in previous work on lung injury.  For the second and third phases, particles will be collected in sufficient quantity for characterization and use in the bioassays.  The emphasis in the second phase will be on comparative toxicity of PM from the selected locations, while the activities of the third phase will be more hypothesis-based and directed at mechanisms of injury.

Protocol Development:

In the first phase, we also extend and further evaluate methods for collection of sufficient mass of particles; a key goal will be to evaluate the extent to which the collection method alters PM characteristics and to measure the stability of the materials collected.  The core set of measurements of PM characteristics will be developed, along with the protocol for detailed on-site monitoring by the Johns Hopkins team.