Grantee Research Project Results
Final Report: Chronic Exposure to Particulate Matter and Cardiopulmonary Disease
EPA Grant Number: R830545Title: Chronic Exposure to Particulate Matter and Cardiopulmonary Disease
Investigators: Laden, Francine , Schwartz, Joel , Speizer, Frank E. , Suh, Helen H. , Camargo, Carlos , Puett, Robin C. , Yanosky, Jeff D.
Institution: Brigham and Women’s Hospital
EPA Project Officer: Chung, Serena
Project Period: January 20, 2003 through January 19, 2006 (Extended to January 19, 2008)
Project Amount: $933,602
RFA: Epidemiologic Research on Health Effects of Long-Term Exposure to Ambient Particulate Matter and Other Air Pollutants (2002) RFA Text | Recipients Lists
Research Category: Air Quality and Air Toxics , Human Health , Particulate Matter , Air
Objective:
We proposed (1) to develop a model estimating long-term exposure to air pollution in the continental U.S. using existing databases, including the EPA Air Quality System (AQS) and (2) to evaluate the association of chronic exposure to air pollution with incident coronary and respiratory disease and total mortality in the Nurses’ Health Study (NHS), an ongoing prospective cohort study of 121,700 women residing throughout the U.S. We hypothesize that the incidence of these diseases and total mortality are positively associated with air pollution and that exposure to air pollution exacerbates existing disease. We also hypothesize that the association with coronary heart disease will be greater among diabetics than nondiabetics and that consumption of antioxidants will modify the association.
Approach:
The basic approach is threefold: (1) to use existing data sources to create a model for exposure to air pollution throughout the U.S.; (2) to link the yearly average exposure to the residential addresses of the study participants; and (3) to evaluate the relative risk of the outcomes in the high compared with the low exposure areas. We will model long-term exposure to air pollution for the years 1986 through 2000 using data from the EPA's Aerometric Information Retrieval System (AIRS), the National Emissions Trends database, National Oceanic and Atmospheric Association, and commercially available traffic count data. Existing ambient monitoring data from specific sites will be used to supplement this information. Residential addresses are updated every two years and will be mapped using geographic information system (GIS) software and linked to the exposure model. Incident cases of cardiovascular disease and of chronic obstructive pulmonary disease, asthma, and lung cancer, diagnosed during the study period, are identified routinely on the NHS biennial self-administered questionnaire. Cases are confirmed by supplemental questionnaire and review of medical records. Mortality is reported by next-of-kin and also obtained by regular searches of the National Death Index. We will estimate the relative risks of these outcomes associated with air pollution using the proportional hazards model, including adjustment for smoking, and other confounders. We will also assess interactions with comorbid diabetes and consumption of antioxidants.
Summary/Accomplishments (Outputs/Outcomes):
This project consisted of two parts – (1) development of predictive models of particulate matter less than 10 microns in diameter (PM10) and less than 2.5 microns in diameter (PM2.5) for the northeastern United States for the years 1988 through 2002; and (2) evaluation of the association of chronic ambient exposure to PM10 and PM2.5, as defined by the models, with mortality and cardiopulmonary outcomes in a large cohort of US women.
Air pollution modeling:
Model description, key assumptions, version, source and intended use:
We used GIS-based spatial smoothing models to predict monthly outdoor PM concentrations.
We modeled spatial and temporal trends in PM monitoring data (US EPA's Air Quality System (AQS), the IMPROVE network and data from Harvard research studies) using generalized additive mixed models with bivariate penalized spline terms for space, and one-dimensional penalized spline terms for Geographic Information System (GIS) derived and meteorological predictors. These predictors included block group and county level population density; distance to nearest road by Census Feature Class Code road class; elevation from the United States Geological Survey (USGS) National Elevation Dataset; land use/land cover from the USGS National Land Cover Dataset; primary PM10 emissions information from the USEPA National Emissions Inventory (NEI); and meteorological variables, including wind speed and precipitation, from the National Climatic Data Center (NCDC). This model structure allows for highly spatially and temporally resolved predictions of chronic PM exposures, even for individuals living in areas with no nearby monitors (albeit with greater uncertainty for locations with distant monitors).
A summary of the distributions of the GIS-derived and meteorological covariates included in the final model, their units, and the number of estimated degrees of freedom in their smooth terms is presented in Table I.
PM2.5 monitoring data are not available on a national level until 1999. Therefore to predict PM2.5 levels for earlier time periods we are using observations of horizontal visual range made at Weather Bureau/Army/Navy (WBAN) stations (about 430 nationwide), after correcting for the truncated nature of the observations, as a predictor for PM2.5. Our models use the above mentioned corrected visibility observations, PM10 predictions, and the PM2.5 monitoring data collected from 1999 to 2002 to estimate monthly average PM2.5 levels back in time to 1988.
Our modeling approach makes several statistical assumptions: isotropy and stationarity of the spatial surfaces, normality and homoscedasticity of residuals, independence of the monthly spatial terms, uniform effects of GIS-derived and meteorological covariates across space, and no interactions of these covariates with one another. Diagnostic procedures have shown these assumptions to be reasonably well justified given our PM data. Our models are intended to be used for predicting monthly average PM concentrations at unmeasured locations.
Performance criteria for the model related to the intended use:
The strength of agreement between measured and observed PM levels, as described by the R2 from cross-validation, was used as the performance criteria for our PM models.
Test results to demonstrate that the model performance criteria were met:
We performed extensive validation of our models by comparing model predictions to measured PM levels using cross-validation. For our PM10 model, model validation results and the results of sensitivity analyses are presented in Yanosky et al. (2008) (Table III). The performance of our PM10 model was strong (cross-validation R2=0.62), with little bias (-0.4 μg/m3) and high precision (6.4 μg/m3). The PM10 model (with monthly spatial terms) performed better than a model with seasonal spatial terms (cross-validation R2=0.54). The addition of GIS-derived and meteorological predictors improved predictive performance over spatial smoothing (cross-validation R2=0.51) or inverse distance weighted interpolation (cross-validation R2=0.29) methods alone and increased the spatial resolution of predictions (Table IV). The model performed well in both rural and urban areas, across seasons, and across the entire time period.
Theory behind the model, expressed in non-mathematical terms:
The theory behind our models is that PM levels at a given location provide information about PM levels at nearby locations, after accounting for local factors which may influence measured PM levels. Thus, our model uses information on air quality measured at monitoring stations to determine how these levels vary across different locations and which local factors influence these levels. With this information, we then used the model to estimate PM levels at the residential addresses of Nurses’ Health Study participants (unmeasured locations) to estimate their chronic exposure to PM.
Mathematics used, including formulas and calculation methods:
The mathematical formula for our models, as presented in Yanosky et al. (2008), was:
1)
where yi,t is the natural-log transformed PM10 monthly site average (for i=1,…, I monitoring sites in the study region and t=1,…, T monthly time periods ), and gt(si) accounts for time-period-specific residual spatial variability whereas g(si) accounts for time-invariant spatial variability. Also, si is the projected spatial coordinate pair for the ith location, Zi,t,1 through Zi,t,P are time-varying covariates, and Xi,1 through Xi,Q are time-invariant GIS-derived covariates. We also note that bi is essentially a site-specific random effect in this model, hence our characterization of the model as a GAMM. Details regarding the fitting of these models in two-stages can be found in Yanosky et al. (2008).
Hardware requirements:
The hardware requirements of running the model code vary depending on the amount of input data, however any PC with a recent version of R installed can fit the model to a small (<1,000 observations) data set, though it may be slow to complete. Larger input data sets will likely require additional RAM, however 4 GB should be enough for all but the largest data sets.
Documentation (e.g., users’ guide, journal publications, model code):
As discussed in Yanosky et al. (2008), our GIS-based spatio-temporal PM models were fit using the statistical software application R using the mgcv library. Generic model code for fitting these models is included below. This code is not intended to run as-is, but rather is included as an example of the modeling code used to fit our GAMM models. In short, the approach involves estimating site-specific intercepts and smooth functions of time-varying covariate effects after adjusting these for residual spatial and temporal trends using an iterative process. Then, spatial variation in the site-specific intercepts is modeled by estimating smooth funtions of time-invariant covariates and a time-invariant smooth spatial function. See Appendix 1 for sample model code.
Epidemiologic Analyses:
Using the PM10 model described above, we predicted monthly average exposures at the addresses of the 66,250 Nurses’ Health Study participants residing in the northeastern US metropolitan areas. Residential address and potential risk factors for mortality and cardiovascular disease were updated every two years. We examined the relationship of chronic particulate exposures in the past 3, 12, 24, 36 and 48 months with all cause mortality and incident and fatal coronary heart disease (CHD). During follow-up 1992-2002, there were 3,785 deaths and 1,348 CHD events. In age- and calendar time-adjusted models, 10 μg/m3 increases in 12-month average PM10 were associated with increased all cause mortality (16%, 95%CI: 5, 28% ) and fatal CHD (43%, 95%CI:10, 86%) (Table VI). Adjustment for body mass index (BMI) and physical activity weakened these associations; however they remained significant for fatal CHD. BMI (Table VII) and smoking modified the association between PM10 exposure and fatal CHD. In this population, increases in PM10 were associated with increases in all cause and CHD mortality. Never smokers with larger BMIs were at greatest risk of fatal CHD (Table VIII). In models using PM2.5, as opposed to PM10, stronger associations with mortality outcomes were observed. These analyses are currently underway. Our findings add to the body of literature indicating an increased risk of all cause mortality and fatal CHD associated with chronic exposure to particulate matter.
Strengths and limitations
In this study we developed a detailed spatiotemporal model, allowing us to predict monthly levels of exposure to particulate matter at residential locations in the northeastern US over a 14 year period. Our exposure assessment is therefore more accurate then that used in previous epidemiologic studies of chronic exposure. However, the large amount of data and technological resources needed to build such a model precluded us from estimating exposures throughout the entire continental US as we had originally planned. Another strength of our study was our access to the Nurses’ Health Study cohort, a well characterized prospective study with updated residential and covariate information and validated health outcomes. In our proportional hazards models we created risk sets for every month of follow-up allowing us to evaluate past exposures specific to the date of death or MI. Although we confronted technological limitations, the contribution of more accurate exposure estimation for longer periods of time and the inclusion of time varying covariates provide valuable information that have been lacking in previous studies of long-term exposures to air pollution.
Table I: Summary of GIS-derived and meteorological covariates included in the final PM10 model.
Category | Description | Unitsb | Time-varying? | Summary of covariate distribution | Edf of smooth termc | ||||
---|---|---|---|---|---|---|---|---|---|
5% | 25% | 50% | 75% | 95% | |||||
Distance to road | Distance to nearest A1 roada | ln(m) | No | 4.7 | 6.7 | 7.8 | 9.2 | 10.9 | 2.5 |
Distance to nearest A2 roada | ln(m) | No | 3.9 | 5.9 | 7.2 | 8.2 | 9.3 | 1.0 | |
Distance to nearest A3 roada | ln(m) | No | 3.5 | 5.7 | 6.7 | 7.6 | 8.7 | 1.0 | |
Urban land use | Percentage of urban land use within 1 km | % | No | 0.1 | 21.0 | 59.7 | 81.7 | 95.4 | 2.4 |
Population density | Block group population density | persons mile-2 | No | 26 | 174 | 1 441 | 4 640 | 15 650 | 4.5 |
County population density | persons mile-2 | No | 24 | 140 | 375 | 1 428 | 11 038 | 4.9 | |
Point source emissions | PM10 emissions within 1 km | ln(tons year-1) | No | -6.9 | -6.9 | -6.9 | 0.7 | 5.1 | 1.0 |
PM10 emissions within 10 km | ln(tons year-1) | No | -6.5 | 4.2 | 5.8 | 7.0 | 8.5 | 2.3 | |
Elevation | Elevation above sea-level | m | No | 3.4 | 48.2 | 176.4 | 233.3 | 359.0 | 1.4 |
Area-source emissions | PM10 emissions by county and year | ln(tons year-1) | Yes | 7.4 | 8.7 | 9.3 | 10.0 | 10.7 | 7.2 |
Precipitation | Total monthly precipitation | inches (cm) | Yes | 1.2 (3.0) | 2.2 (5.6) | 3.2 (8.1) | 4.5 (11.4) | 6.9 (17.5) | 7.6 |
Wind speed | Monthly-average surface wind speed | m s-1 | Yes | 2.6 | 3.3 | 3.8 | 4.3 | 4.8 | 6.4 |
aDesignated by Census Feature Class Codes, from ESRI StreetMap road data. | |||||||||
b'ln' is natural log. | |||||||||
cEdf is estimated degrees of freedom of smooth term in the final model. |
Table II: Accuracy and precision of PM10 model predictions by state or state group, season, quartiles of urban land use and block group population density, monitoring network, and monitoring objective determined using cross-validation.
Grouping | Na | Biasb (μg/m3) | RMSPEc (μg/m3) | |
---|---|---|---|---|
By state or state groupd (1988-2002) | Northern New England (ME, NH, VT) | 6 362 | -1.1 | 6.6 |
Southern New England (CT,MA, RI) | 6 658 | -0.6 | 5.0 | |
NY | 5 307 | 0.1 | 5.8 | |
Western Mid-Atlantic (MD, NJ, PA) | 11 630 | -0.7 | 7.2 | |
Midwestern (OH, MI) | 12 919 | 0.2 | 6.4 | |
DE | 469 | -2.7 | 6.3 | |
By season (across study region) | Winter | 10 478 | -0.7 | 7.1 |
Spring | 10 896 | -0.5 | 6.5 | |
Summer | 10 967 | -0.04 | 6.0 | |
Fall | 11 004 | -0.4 | 5.9 | |
By quartiles of urban land use | 1 | 10 842 | -0.3 | 6.0 |
2 | 10 899 | 0.1 | 7.4 | |
3 | 10 805 | -0.8 | 6.1 | |
4 | 10 799 | -0.6 | 6.0 | |
By quartiles of block group population density | 1 | 10 827 | -0.3 | 6.2 |
2 | 10 819 | -0.9 | 7.3 | |
3 | 10 885 | 0.2 | 6.1 | |
4 | 10 814 | -0.6 | 5.9 | |
By monitoring network | AQSe | 42 271 | -0.5 | 6.4 |
IMPROVEf | 836 | 1.8 | 4.6 | |
Harvard 5 Cities | 172 | 4.7 | 7.7 | |
Harvard 24 Cities | 66 | -1.8 | 6.8 | |
By monitoring objectiveg | Unknown (Other networks) | 1 074 | 2.0 | 5.4 |
Unknown (AQS sites) | 5 286 | -1.8 | 8.0 | |
Population Exposure | 18 794 | 0.4 | 5.7 | |
Highest Concentration | 15 127 | -1.3 | 6.9 | |
General/Background | 1 173 | 0.8 | 5.0 | |
Other | 1 139 | -1.0 | 5.2 | |
Upwind Background | 287 | 2.1 | 5.2 | |
Source Oriented | 209 | 1.3 | 3.7 | |
Maximum Ozone Concentration | 145 | 1.8 | 4.3 | |
Maximum Precursor Emissions Impact | 111 | 4.8 | 5.4 | |
Across all observationsa | 43 345 | -0.4 | 6.4 | |
a"N" is number of pairs of measurements and cross-validation predictions; data from one site in Pennsylvania were removed. | ||||
bMean prediction error (cross-validation predictions minus measurements). | ||||
cRMSPE is root mean squared prediction error, describing precision of the model predictions. | ||||
dWhere results were similar, nearby states were grouped. | ||||
eFrom the USEPA's Air Quality System monitoring network. | ||||
fFrom the Interagency Monitoring of Protected Visual Environments (IMPROVE) monitoring network. | ||||
gDefined for sites in the USEPA Air Quality System (AQS) network only. |
Table III: Predictive performance of the final PM10 model and alternative models.
Model description | Number of spatial termsa | Covariates included | Model fit R2,b | Cross-validation results | ||
---|---|---|---|---|---|---|
Interceptc | Slopec | Cross-validation R2,d | ||||
Final model | 180 monthlye | All | 0.74 | 2.4 +/- 0.9 | 0.92 +/- 0.003 | 0.62 |
Seasonal spatial terms | 4 seasonale | All | 0.59 | 2.1 +/- 0.1 | 0.94 +/- 0.004 | 0.54 |
GAM spatial smoothing only | 180 monthly | None | 0.76 | 3.7 +/- 0.1 | 0.87 +/- 0.004 | 0.51 |
Inverse distance weighted interpolation | NA | None | NA | 8.3 +/- 0.1 | 0.65 +/- 0.005 | 0.29 |
Nearest neighbor interpolation | NA | None | NA | 15.5 +/- 0.2 | 0.42 +/- 0.006 | 0.22f |
aCorresponds to the extent of control for space-time interaction in the model. | ||||||
bDerived from fitting the model to all data from sets 1 through 9, including data from states adjacent to the study region. | ||||||
cPresented as (parameter estimate +/- standard error) from linear regression of observations on predictions. | ||||||
dDerived from cross-validation on sets 1 through 9, with each set held out in turn (one site in Pennsylvania was removed as an outlier); 43 345 observations total. | ||||||
eRefers to number of time-varying spatial terms fit in the first stage of the model in addition to one spatial term fit in the second stage. |
TABLE IV: Hazard ratios and 95%CIs for all cause and cause specific mortality associated with a 10 mg/m3 change in predicted PM10 exposure*
Cases | Person Months | Hazard Ratio (95% CI) | ||||
---|---|---|---|---|---|---|
month before | 3 month moving average | 12 month moving average | 48 month moving average | |||
All Cause Mortality | 3785 | 600752 | 1.04 (0.98,1.11) | 1.14 (1.05,1.23) | 1.16 (1.05,1.28) | 1.15 (1.04,1.28) |
MI | ||||||
First CHD Event | 1348 | 597456 | 1.08 (0.98,1.19) | 0.96 (0.84,1.09) | 1.10 (0.94,1.29) | 1.09 (0.92,1.29) |
Fatal CHD† | 494 | 597456 | 1.16 (0.98,1.36) | 1.21 (0.98,1.48) | 1.43 (1.10,1.86) | 1.43 (1.09,1.88) |
Nonfatal MI | 854 | 597458 | 1.03 (0.91,1.18) | 0.83 (0.71,0.98) | 0.94 (0.77,1.15) | 0.93 (0.75,1.15) |
*Modeled stratifying by age in months, adjusting for state of residence, year and season †Including sudden deaths, excluding prior nonfatal MI |
TABLE V: Hazard ratios and 95%CIs for the association between fatal CHD* and a 10 μg/m3 change in average predicted PM10 in the 12 months prior to death stratified by BMI
BMI< 30 | BMI=30.0+ | |
---|---|---|
Cases† | 310 | 149 |
Person Months | 432557 | 132250 |
Hazard Ratio (95%CI) | ||
Base Model ‡ | 1.20 (0.86,1.68) | 2.16 (1.35,3.45) |
Adjusted Model § | 1.08 (0.76,1.52) | 1.99 (1.23,3.22) |
* Including sudden deaths † Risks based on complete information for participants ‡ Modeled stratifying by age in months, adjusting for state of residence, year and season § Modeled stratifying by age in months, adjusting for state of residence, year, season, smoking status, family history of MI, high cholesterol, diabetes, hypertension, median family income in census tract of residence, physical activity, and median house value in census tract of residence |
TABLE VI: Hazard ratios and 95%CIs for the association between fatal CHD* and a 10 μg/m3 change in average predicted PM10 in the 12 months prior to death stratified by BMI and smoking status.†
ALL | Never Smoker | Former | Current |
---|---|---|---|
Cases* | 160 | 190 | 125 |
Person Months | 251153 | 244466 | 77153 |
Hazard Ratio (95%CI) | |||
Basic Model‡ | 1.87 (1.24,2.81) | 1.32 (0.89,1.96) | 0.90 (0.54,1.51) |
Adjusted Model§ | 1.83 (1.20,2.79) | 1.22 (0.82,1.83) | 0.88 (0.52,1.48) |
BMI < 30 | |||
Cases* | 97 | 120 | 100 |
Person Months | 181451 | 180120 | 64912 |
Hazard Ratio (95%CI) | |||
Basic Model‡ | 1.42 (0.83,2.41) | 1.01 (0.60,1.68) | 0.88 (0.50,1.58) |
Adjusted Model§ | 1.41 (0.82,2.42) | 0.98 (0.58,1.64) | 0.85 (0.47,1.53) |
BMI 30+ | |||
Cases* | 59 | 67 | 23 |
Person Months | 59425 | 59897 | 11614 |
Hazard Ratio (95%CI) | |||
Basic Model‡ | 2.85 (1.44,5.65) | 1.89 (0.99,3.58) | 0.98 (0.31,3.14) |
Adjusted Model§ | 2.82 (1.40,5.71) | 1.64 (0.86,3.13) | 1.03 (0.31,3.42) |
* Including sudden deaths † Risks based on complete information for participants ‡ Modeled stratifying by age in months, adjusting for state of residence, year and season § Modeled stratifying by age in months, adjusting for state of residence, year, season, smoking status, family history of MI, high cholesterol, diabetes, hypertension, median family income in census tract of residence, physical activity, and median house value in census tract of residence. |
Expected Results:
Although there is a substantial body of literature demonstrating the adverse health effects associated with air pollution, to date there have only been two large cohort studies of mortality. This proposed study will not only evaluate mortality but it will be the first study to prospectively evaluate cause-specific incident disease on a nationwide basis. Further, it will provide information on the extent of life shortening associated with the exposure by measuring survival and severity of disease after the first event.
Journal Articles on this Report : 8 Displayed | Download in RIS Format
Other project views: | All 12 publications | 8 publications in selected types | All 8 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Brochu PJ, Yanosky JD, Paciorek CJ, Schwartz J, Chen JT, Herrick RF, Suh HH. Particulate air pollution and socioeconomic position in rural and urban areas of the Northeastern United States. American Journal of Public Health 2011;101(Suppl 1):S224-S230. |
R830545 (Final) |
Exit Exit |
|
Paciorek CJ, Yanosky JD, Puett RC, Laden F, Suh HH. Practical large-scale spatio-temporal modeling of particulate matter concentrations. Annals of Applied Statistics 2009;3(1):370-397. |
R830545 (Final) |
Exit Exit |
|
Puett RC, Schwartz J, Hart JE, Yanosky JD, Speizer FE, Suh H, Paciorek CJ, Neas LM, Laden F. Chronic particulate exposure, mortality, and coronary heart disease in the Nurses' Health Study. American Journal of Epidemiology 2008;168(10):1161-1168. |
R830545 (Final) |
Exit Exit |
|
Puett RC, Hart JE, Yanosky JD, Paciorek C, Schwartz J, Suh H, Speizer FE, Laden F. Chronic fine and coarse particulate exposure, mortality, and coronary heart disease in the Nurses' Health Study. Environmental Health Perspectives 2009;117(11):1697-1701. |
R830545 (Final) |
|
|
Puett RC, Hart JE, Suh H, Mittleman M, Laden F. Particulate matter exposures, mortality and cardiovascular disease in the Health Professionals Follow-up Study. Environmental Health Perspectives 2011;119(8):1130-1135. |
R830545 (Final) |
|
|
Puett RC, Hart JE, Schwartz J, Hu FB, Liese AD, Laden F. Are particulate matter exposures associated with risk of type 2 diabetes? Environmental Health Perspectives 2011;119(3):384-389. |
R830545 (Final) |
|
|
Yanosky JD, Paciorek CJ, Schwartz J, Laden F, Puett R, Suh HH. Spatio-temporal modeling of chronic PM10 exposure for the Nurses' Health Study. Atmospheric Environment 2008;42(18):4047-4062. |
R830545 (Final) |
Exit Exit |
|
Yanosky J, Paciorek C, Suh HH. Predicting chronic fine and coarse particulate exposures using spatiotemporal models for the Northeastern and Midwestern United States. Environmental Health Perspectives 2009;117(4):522-529. |
R830545 (Final) |
|
Supplemental Keywords:
epidemiology, health effects, ambient air, particulates, Environmental Monitoring, PM2.5, PM10, air pollutants, cardiovascular disease, chronic effects, chronic exposure, human exposure, mortality, exposure modeling, GIS, geocoding, asthma, chronic obstructive pulmonary disease, myocardial infarction, antioxidants, lung cancer, diabetes, smoking, physical activity,, RFA, Health, Scientific Discipline, PHYSICAL ASPECTS, Air, Ecosystem Protection/Environmental Exposure & Risk, HUMAN HEALTH, particulate matter, Bioavailability, Health Risk Assessment, air toxics, Exposure, Epidemiology, Monitoring/Modeling, Risk Assessments, Disease & Cumulative Effects, Environmental Monitoring, Physical Processes, tropospheric ozone, particulates, health effects, ambient air quality, sensitive populations, urban air, atmospheric measurements, EMPACT, chronic exposure, monitoring, PM 2.5, air pollutants, effects assessment, particulate, stratospheric ozone, acute cardiovascular effects, airway disease, pulmonary disease, ozone, continuous monitoring, ambient air, air pollution, children, carbon black, particles, human exposue, clinical studies, human exposure, chronic effects, sensitive subgroups, ecological risk, ambient particulates, Acute health effects, PM2.5, allergic response, cardiotoxicity, mortality, measurement methods , atmospheric chemistry, long-term exposure, cardiopulmonery responses, cardiovascular diseaseProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.
Project Research Results
- 2006 Progress Report
- 2005 Progress Report
- 2004 Progress Report
- 2003 Progress Report
- Original Abstract
8 journal articles for this project