PMCC PMCC

Search tips
Search criteria

Advanced
Results 1-25 (33)
 

Clipboard (0)
None

Select a Filter Below

Year of Publication
Document Types
1.  A new technique for evaluating land use regression models and their impact on health effect estimates 
BACKGROUND
Leave-one-out cross-validation that fails to account for variable selection does not properly reflect prediction accuracy when the number of training sites is small. The impact on health effect estimates has rarely been studied.
OBJECTIVES
Develop an improved validation procedure for land-use regression models with variable selection and investigate health effect estimates in relation to land-use regression model performance.
METHODS
We randomly generated ten training and test sets for nitrogen dioxide and particulate matter. For each training set we developed models and evaluated them using a cross-holdout validation approach. Cross-holdout validation develops new models for each evaluation compared to refitting the model without variable selection, as in standard leave-one-out cross-validation. We also implemented holdout validation, which evaluates model predictions using independent test sets. We evaluated the relationship between cross-holdout validation and holdout validation R2 and estimates of the association between air pollution and forced vital capacity in the Dutch birth cohort.
RESULTS
Cross-holdout validation R2s were generally identical to holdout validation R2s, but were notably smaller than leave-one-out cross-validation R2s. Decreases in forced vital capacity in relation to air pollution exposure were larger for land-use regression models that had larger holdout validation and cross-holdout validation R2s rather than leave-one-out cross-validation R2.
Conclusion
Cross-holdout validation accurately reflects predictive ability of land-use regression models and is a useful validation approach for small datasets. Land-use regression predictive ability in terms of hold-out validation and cross-holdout validation rather than leave-one-out cross-validation was associated with the magnitude of health effect estimates in a case study.
doi:10.1097/EDE.0000000000000404
PMCID: PMC5221608  PMID: 26426941
2.  Development of Long-term Spatiotemporal Models for Ambient Ozone in Six Metropolitan regions of the United States: The MESA Air Study 
Background
Current epidemiologic studies rely on simple ozone metrics which may not appropriately capture population ozone exposure. For understanding health effects of long-term ozone exposure in population studies, it is advantageous for exposure estimation to incorporate the complex spatiotemporal pattern of ozone concentrations at fine scales.
Objective
To develop a geo-statistical exposure prediction model that predicts fine scale spatiotemporal variations of ambient ozone in six United States metropolitan regions.
Methods
We developed a modeling framework that estimates temporal trends from regulatory agency and cohort-specific monitoring data from MESA Air measurement campaigns and incorporates land use regression with universal kriging using predictor variables from a large geographic database. The cohort-specific data were measured at home and community locations. The framework was applied in estimating two-week average ozone concentrations from 1999 to 2013 in models of each of the six MESA Air metropolitan regions.
Results
Ozone models perform well in both spatial and temporal dimensions at the agency monitoring sites in terms of prediction accuracy. City-specific leave-one (site)-out cross-validation R2 accounting for temporal and spatial variability ranged from 0.65 to 0.88 in the six regions. For predictions at the home sites, the R2 is between 0.60 and 0.91 for cross-validation that left out 10% of home sites in turn. The predicted ozone concentrations vary substantially over space and time in all the metropolitan regions.
Conclusion
Using the available data, our spatiotemporal models are able to accurately predict long-term ozone concentrations at fine spatial scales in multiple regions. The model predictions will allow for investigation of the long-term health effects of ambient ozone concentrations in future epidemiological studies.
doi:10.1016/j.atmosenv.2015.10.042
PMCID: PMC5021184  PMID: 27642250
Ozone; spatio-temporal; geo-statistical model; multi-city; MESA Air
3.  Prediction of fine particulate matter chemical components with a spatio-temporal model for the Multi-Ethnic Study of Atherosclerosis cohort 
Although cohort studies of the health effects of PM2.5 have developed exposure prediction models to represent spatial variability across participant residences, few models exist for PM2.5 components. We aimed to develop a city-specific spatio-temporal prediction approach to estimate long-term average concentrations of four PM2.5 components including sulfur, silicon, and elemental and organic carbon for the Multi-Ethnic Study of Atherosclerosis cohort, and to compare predictions to those from a national spatial model. Using 2-week average measurements from a cohort-focused monitoring campaign, the spatio-temporal model employed selected geographic covariates in a universal kriging framework with the data-driven temporal trend. Relying on long-term means of daily measurements from regulatory monitoring networks, the national spatial model employed dimension-reduced predictors using universal kriging. For the spatio-temporal model, the cross-validated and temporally-adjusted R2 was relatively higher for EC and OC, and in the Los Angeles and Baltimore areas. The cross-validated R2s for both models across the six areas were reasonably high for all components except silicon. Predicted long-term concentrations at participant homes from the two models were generally highly correlated across cities but poorly correlated within cities. The spatio-temporal model may be preferred for city-specific health analyses, whereas both models could be used for multi-city studies.
doi:10.1038/jes.2016.29
PMCID: PMC5104659  PMID: 27189258
empirical/statistical models; epidemiology; exposure modeling; particulate matter
4.  Exposure to Traffic-Related Air Pollution in Relation to Progression in Physical Disability among Older Adults 
Environmental Health Perspectives  2016;124(7):1000-1008.
Background:
Physical disability is common though not inevitable in older age and has direct bearing on a person’s ability to perform activities essential for self-care and independent living. Air pollution appears to increase the risk of several chronic diseases that contribute to the progression of disability.
Objective:
We evaluated long-term exposure to traffic-related air pollution (TRAP) in relation to progression in physical disability.
Methods:
We conducted our investigation within the Chicago Health and Aging Project. We measured participants’ exposures to TRAP using two surrogates: residential proximity to major roads (1993 onwards) and ambient concentrations of oxides of nitrogen (NOX; 1999 onwards), predicted via a geographic information systems-based spatiotemporal smoothing model (cross-validation R2 = 0.87) that incorporated community-based monitoring and resolved intraurban exposure gradients at a spatial scale of tens of meters. Participants’ lower-extremity physical ability was assessed every 3 years (1993–2012) via tandem stand, chair stand, and timed walking speed.
Results:
In multivariable-adjusted analyses (n = 5,708), higher long-term NOX exposure was associated with significantly faster progression in disability. Compared with the 5-year decline in physical ability score among participants in the lowest quartile of NOX exposure, decline among those in the highest exposure quartile was 1.14 units greater (95% confidence interval [CI]: –1.86, –0.42), equivalent to 3 additional years of decline among those in the lowest exposure quartile. The association was linear across the continuum of NOX exposure: per 10-ppb increment in exposure, the 5-year decline in physical ability score was 0.87 unit greater (95% CI: –1.35, –0.39). Proximity to a major road was not associated with disability progression (n = 9,994).
Conclusions:
These data join a growing body of evidence suggesting that TRAP exposures may accelerate aging-related declines in health.
Citation:
Weuve J, Kaufman JD, Szpiro AA, Curl C, Puett RC, Beck T, Evans DA, Mendes de Leon CF. 2016. Exposure to traffic-related air pollution in relation to progression in physical disability among older adults. Environ Health Perspect 124:1000–1008; http://dx.doi.org/10.1289/ehp.1510089
doi:10.1289/ehp.1510089
PMCID: PMC4937863  PMID: 27022889
5.  Long-term Exposure to Air Pollution and Markers of Inflammation, Coagulation, and Endothelial Activation 
Epidemiology (Cambridge, Mass.)  2015;26(3):310-320.
Background
Air pollution is associated with cardiovascular disease, and systemic inflammation may mediate this effect. We assessed associations between long- and short-term concentrations of air pollution and markers of inflammation, coagulation, and endothelial activation.
Methods
We studied participants from the Multi-Ethnic Study of Atherosclerosis from 2000 to 2012 with repeat measures of serum C-reactive protein (CRP), interleukin-6 (IL-6), fibrinogen, D-dimer, soluble E-selectin, and soluble Intercellular Adhesion Molecule-1. Annual average concentrations of ambient fine particulate matter (PM2.5), individual-level ambient PM2.5 (integrating indoor concentrations and time–location data), oxides of nitrogen (NOx), nitrogen dioxide (NO2), and black carbon were evaluated. Short-term concentrations of PM2.5 reflected the day of blood draw, day prior, and averages of prior 2-, 3-, 4-, and 5-day periods. Random-effects models were used for long-term exposures and fixed effects for short-term exposures. The sample size was between 9,000 and 10,000 observations for CRP, IL-6, fibrinogen, and D-dimer; approximately 2,100 for E-selectin; and 3,300 for soluble Intercellular Adhesion Molecule-1.
Results
After controlling for confounders, 5 µg/m3 increase in long-term ambient PM2.5 was associated with 6% higher IL-6 (95% confidence interval = 2%, 9%), and 40 parts per billion increase in long-term NOx was associated with 7% (95% confidence interval = 2%, 13%) higher level of D-dimer. PM2.5 measured at day of blood draw was associated with CRP, fibrinogen, and E-selectin. There were no other positive associations between blood markers and short- or long-term air pollution.
Conclusions
These data are consistent with the hypothesis that long-term exposure to air pollution is related to some markers of inflammation and fibrinolysis.
doi:10.1097/EDE.0000000000000267
PMCID: PMC4455899  PMID: 25710246
6.  Reduced-Rank Spatio-Temporal Modeling of Air Pollution Concentrations in the Multi-Ethnic Study of Atherosclerosis and Air Pollution1 
The annals of applied statistics  2014;8(4):2509-2537.
There is growing evidence in the epidemiologic literature of the relationship between air pollution and adverse health outcomes. Prediction of individual air pollution exposure in the Environmental Protection Agency (EPA) funded Multi-Ethnic Study of Atheroscelerosis and Air Pollution (MESA Air) study relies on a flexible spatio-temporal prediction model that integrates land-use regression with kriging to account for spatial dependence in pollutant concentrations. Temporal variability is captured using temporal trends estimated via modified singular value decomposition and temporally varying spatial residuals. This model utilizes monitoring data from existing regulatory networks and supplementary MESA Air monitoring data to predict concentrations for individual cohort members.
In general, spatio-temporal models are limited in their efficacy for large data sets due to computational intractability. We develop reduced-rank versions of the MESA Air spatio-temporal model. To do so, we apply low-rank kriging to account for spatial variation in the mean process and discuss the limitations of this approach. As an alternative, we represent spatial variation using thin plate regression splines. We compare the performance of the outlined models using EPA and MESA Air monitoring data for predicting concentrations of oxides of nitrogen (NOx)—a pollutant of primary interest in MESA Air—in the Los Angeles metropolitan area via cross-validated R2.
Our findings suggest that use of reduced-rank models can improve computational efficiency in certain cases. Low-rank kriging and thin plate regression splines were competitive across the formulations considered, although TPRS appeared to be more robust in some settings.
doi:10.1214/14-AOAS786
PMCID: PMC4803447  PMID: 27014398
Spatiotemporal modeling; reduced-rank; air pollution; kriging; thin plate splines
7.  Long-Term Exposure to Air Pollution and Type 2 Diabetes Mellitus in a Multiethnic Cohort 
American Journal of Epidemiology  2015;181(5):327-336.
Although air pollution has been suggested as a possible risk factor for type 2 diabetes mellitus (DM), results from existing epidemiologic studies have been inconsistent. We investigated the associations of prevalence and incidence of DM with long-term exposure to air pollution as estimated using annual average concentrations of particulate matter with an aerodynamic diameter of 2.5 μm or less (PM2.5) and nitrogen oxides at baseline (2000) in the Multi-Ethnic Study of Atherosclerosis. All participants were aged 45–84 years at baseline and were recruited from 6 US sites. There were 5,839 participants included in the study of prevalent DM and 5,135 participants without DM at baseline in whom we studied incident DM. After adjustment for potential confounders, we found significant associations of prevalent DM with PM2.5 (odds ratio (OR) = 1.09, 95% confidence interval (CI): 1.00, 1.17) and nitrogen oxides (OR = 1.18, 95% CI: 1.01, 1.38) per each interquartile-range increase (2.43 µg/m3 and 47.1 ppb, respectively). Larger but nonsignificant associations were observed after further adjustment for study site (for PM2.5, OR = 1.16, 95% CI: 0.94, 1.42; for nitrogen oxides, OR = 1.29, 95% CI: 0.94, 1.76). No air pollution measures were significantly associated with incident DM over the course of the 9-year follow-up period. Results were partly consistent with a link between long-term exposure to air pollution and the risk of type 2 DM. Additional studies with a longer follow-up time and a greater range of air pollution exposures, including high levels, are warranted to evaluate the hypothesized association.
doi:10.1093/aje/kwu280
PMCID: PMC4339386  PMID: 25693777
air pollution; diabetes; nitrogen oxides; particulate matter; prospective cohort study
8.  Breast Cancer Risk in Relation to Ambient Air Pollution Exposure at Residences in the Sister Study Cohort 
Background
Some but not all past studies reported associations between components of air pollution and breast cancer, namely fine particulate matter ≤ 2.5 μm (PM2.5) and nitrogen dioxide (NO2). It is yet unclear whether risks differ according to estrogen receptor (ER) and progesterone receptor (PR) status.
Methods
This analysis includes 47,591 women from the Sister Study cohort enrolled from August 2003-July 2009, in whom 1,749 invasive breast cancer cases arose from enrollment to January 2013. Using Cox proportional hazards and polytomous logistic regression, we estimated breast cancer risk associated with residential exposure to NO2, PM2.5, and PM10.
Results
While breast cancer risk overall was not associated with PM2.5 (Hazards ratio [HR] = 1.03; 95% CI: 0.96–1.11), PM10 (HR = 0.99; 95% CI: 0.98–1.00), or NO2 (HR = 1.02; 95% CI: 0.97–1.07), the association with NO2 differed according to ER/PR subtype (p = 0.04). For an interquartile range (IQR) difference of 5.8 parts per billion (ppb) in NO2, the relative risk (RR) of ER+/PR+ breast cancer was 1.10 (95% CI: 1.02–1.19), while there was no evidence of association with ER−/PR− (RR=0.92; 95% CI: 0.77–1.09; pinteraction=0.04).
Conclusions
Within the Sister Study cohort, we found no significant associations between air pollution and breast cancer risk overall. But we observed an increased risk of ER+/PR+ breast cancer associated with NO2.
Impact
Though these results suggest there is no substantial increased risk for breast cancer overall in relation to air pollution, NO2, a marker of traffic related air pollution, may differentially affect ER+/PR+ breast cancer.
doi:10.1158/1055-9965.EPI-15-0787
PMCID: PMC4686338  PMID: 26464427
Air pollution; Breast cancer risk; Particulate matter; Nitrogen dioxide; Cancer survival
9.  Individual-Level Concentrations of Fine Particulate Matter Chemical Components and Subclinical Atherosclerosis: A Cross-Sectional Analysis Based on 2 Advanced Exposure Prediction Models in the Multi-Ethnic Study of Atherosclerosis 
American Journal of Epidemiology  2014;180(7):718-728.
Long-term exposure to outdoor particulate matter with an aerodynamic diameter less than or equal to 2.5 µm (PM2.5) has been associated with cardiovascular morbidity and mortality. The chemical composition of PM2.5 that may be most responsible for producing these associations has not been identified. We assessed cross-sectional associations between long-term concentrations of PM2.5 and 4 of its chemical components (sulfur, silicon, elemental carbon, and organic carbon (OC)) and subclinical atherosclerosis, measured as carotid intima-media thickness (CIMT) and coronary artery calcium, between 2000 and 2002 among 5,488 Multi-Ethnic Study of Atherosclerosis participants residing in 6 US metropolitan areas. Long-term concentrations of PM2.5 components at participants' homes were predicted using both city-specific spatiotemporal models and a national spatial model. The estimated differences in CIMT associated with interquartile-range increases in sulfur, silicon, and OC predictions from the spatiotemporal model were 0.022 mm (95% confidence interval (CI): 0.014, 0.031), 0.006 mm (95% CI: 0.000, 0.012), and 0.026 mm (95% CI: 0.019, 0.034), respectively. Findings were generally similar using the national spatial model predictions but were often sensitive to adjustment for city. We did not find strong evidence of associations with coronary artery calcium. Long-term concentrations of sulfur and OC, and possibly silicon, were associated with CIMT using 2 distinct exposure prediction modeling approaches.
doi:10.1093/aje/kwu186
PMCID: PMC4172166  PMID: 25164422
atherosclerosis; cardiovascular diseases; carotid intima-media thickness; cohort studies; particulate matter
10.  Long-Term Air Pollution Exposure and Blood Pressure in the Sister Study 
Environmental Health Perspectives  2015;123(10):951-958.
Background
Exposure to air pollution has been consistently associated with cardiovascular morbidity and mortality, but mechanisms remain uncertain. Associations with blood pressure (BP) may help to explain the cardiovascular effects of air pollution.
Objective
We examined the cross-sectional relationship between long-term (annual average) residential air pollution exposure and BP in the National Institute of Environmental Health Sciences’ Sister Study, a large U.S. cohort study investigating risk factors for breast cancer and other outcomes.
Methods
This analysis included 43,629 women 35–76 years of age, enrolled 2003–2009, who had a sister with breast cancer. Geographic information systems contributed to satellite-based nitrogen dioxide (NO2) and fine particulate matter (≤ 2.5 μm; PM2.5) predictions at participant residences at study entry. Generalized additive models were used to examine the relationship between pollutants and measured BP at study entry, adjusting for cardiovascular disease risk factors and including thin plate splines for potential spatial confounding.
Results
A 10-μg/m3 increase in PM2.5 was associated with 1.4-mmHg higher systolic BP (95% CI: 0.6, 2.3; p < 0.001), 1.0-mmHg higher pulse pressure (95% CI: 0.4, 1.7; p = 0.001), 0.8-mmHg higher mean arterial pressure (95% CI: 0.2, 1.4; p = 0.01), and no significant association with diastolic BP. A 10-ppb increase in NO2 was associated with a 0.4-mmHg (95% CI: 0.2, 0.6; p < 0.001) higher pulse pressure.
Conclusions
Long-term PM2.5 and NO2 exposures were associated with higher blood pressure. On a population scale, such air pollution–related increases in blood pressure could, in part, account for the increases in cardiovascular disease morbidity and mortality seen in prior studies.
Citation
Chan SH, Van Hee VC, Bergen S, Szpiro AA, DeRoo LA, London SJ, Marshall JD, Kaufman JD, Sandler DP. 2015. Long-term air pollution exposure and blood pressure in the Sister Study. Environ Health Perspect 123:951–958; http://dx.doi.org/10.1289/ehp.1408125
doi:10.1289/ehp.1408125
PMCID: PMC4590742  PMID: 25748169
11.  A Flexible Spatio-Temporal Model for Air Pollution with Spatial and Spatio-Temporal Covariates 
The development of models that provide accurate spatio-temporal predictions of ambient air pollution at small spatial scales is of great importance for the assessment of potential health effects of air pollution. Here we present a spatio-temporal framework that predicts ambient air pollution by combining data from several different monitoring networks and deterministic air pollution model(s) with geographic information system (GIS) covariates. The model presented in this paper has been implemented in an R package, SpatioTemporal, available on CRAN.
The model is used by the EPA funded Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air) to produce estimates of ambient air pollution; MESA Air uses the estimates to investigate the relationship between chronic exposure to air pollution and cardiovascular disease. In this paper we use the model to predict long-term average concentrations of NOx in the Los Angeles area during a ten year period. Predictions are based on measurements from the EPA Air Quality System, MESA Air specific monitoring, and output from a source dispersion model for traffic related air pollution (Caline3QHCR). Accuracy in predicting long-term average concentrations is evaluated using an elaborate cross-validation setup that accounts for a sparse spatio-temporal sampling pattern in the data, and adjusts for temporal effects. The predictive ability of the model is good with cross-validated R2 of approximately 0.7 at subject sites.
Replacing four geographic covariate indicators of traffic density with the Caline3QHCR dispersion model output resulted in very similar prediction accuracy from a more parsimonious and more interpretable model. Adding traffic-related geographic covariates to the model that included Caline3QHCR did not further improve the prediction accuracy.
doi:10.1007/s10651-013-0261-4
PMCID: PMC4174563  PMID: 25264424
12.  Markers of Inflammation and Coagulation after Long-Term Exposure to Coarse Particulate Matter: A Cross-Sectional Analysis from the Multi-Ethnic Study of Atherosclerosis 
Environmental Health Perspectives  2015;123(6):541-548.
Background
Toxicological research suggests that coarse particles (PM10–2.5) are inflammatory, but responses are complex and may be best summarized by multiple inflammatory markers. Few human studies have investigated associations with PM10–2.5 and, of those, none have explored long-term exposures. Here we examine long-term associations with inflammation and coagulation in the Multi-Ethnic Study of Atherosclerosis.
Methods
Participants included 3,295 adults (45–84 years of age) from three metropolitan areas. Site-specific spatial models were used to estimate 5-year concentrations of PM10–2.5 mass and copper, zinc, phosphorus, silicon, and endotoxin found in PM10–2.5. Outcomes included interleukin-6, C-reactive protein, fibrinogen, total homocysteine, D-dimer, factor VIII, plasmin–antiplasmin complex, and inflammation and coagulation scores. We used multivariable regression with multiply imputed data to estimate associations while controlling for potential confounders, including co-pollutants such as fine particulate matter.
Results
Some limited evidence was found of relationships between inflammation and coagulation and PM10–2.5. Endotoxin was the PM10–2.5 component most strongly associated with inflammation, with an interquartile range (IQR) increase (0.08 EU/m3) associated with 0.15 (95% CI: 0.01, 0.28; p = 0.03) and 0.08 (95% CI: –0.07, 0.23; p = 0.28) higher inflammation scores before and after control for city, respectively. Copper was the component with the strongest association with coagulation, with a 4-ng/m3 increase associated with 0.19 (95% CI: 0.08, 0.30; p = 0.0008) and 0.12 (95% CI: –0.05, 0.30; p = 0.16) unit higher coagulation scores before and after city adjustment, respectively.
Conclusions
Our cross-sectional analysis provided some evidence that long-term PM10–2.5 exposure was associated with inflammation and coagulation, but associations were modest and depended on particle composition.
Citation
Adar SD, D’Souza J, Mendelsohn-Victor K, Jacobs DR Jr, Cushman M, Sheppard L, Thorne PS, Burke GL, Daviglus ML, Szpiro AA, Diez Roux AV, Kaufman JD, Larson TV. 2015. Markers of inflammation and coagulation after long-term exposure to coarse particulate matter: a cross-sectional analysis from the Multi-Ethnic Study of Atherosclerosis. Environ Health Perspect 123:541–548; http://dx.doi.org/10.1289/ehp.1308069
doi:10.1289/ehp.1308069
PMCID: PMC4455582  PMID: 25616153
13.  Traffic-related Air Pollution and the Right Ventricle. The Multi-ethnic Study of Atherosclerosis 
Rationale: Right heart failure is a cause of morbidity and mortality in common and rare heart and lung diseases. Exposure to traffic-related air pollution is linked to left ventricular hypertrophy, heart failure, and death. Relationships between traffic-related air pollution and right ventricular (RV) structure and function have not been studied.
Objectives: To characterize the relationship between traffic-related air pollutants and RV structure and function.
Methods: We included men and women with magnetic resonance imaging assessment of RV structure and function and estimated residential outdoor nitrogen dioxide (NO2) concentrations from the Multi-ethnic Study of Atherosclerosis, a study of individuals free of clinical cardiovascular disease at baseline. Multivariable linear regression estimated associations between NO2 exposure (averaged over the year prior to magnetic resonance imaging) and measures of RV structure and function after adjusting for demographics, anthropometrics, smoking status, diabetes mellitus, and hypertension. Adjustment for corresponding left ventricular parameters, traffic-related noise, markers of inflammation, and lung disease were considered in separate models. Secondary analyses considered oxides of nitrogen (NOx) as the exposure.
Measurements and Main Results: The study sample included 3,896 participants. In fully adjusted models, higher NO2 was associated with greater RV mass and larger RV end-diastolic volume with or without further adjustment for corresponding left ventricular parameters, traffic-related noise, inflammatory markers, or lung disease (all P < 0.05). There was no association between NO2 and RV ejection fraction. Relationships between NOx and RV morphology were similar.
Conclusions: Higher levels of NO2 exposure were associated with greater RV mass and larger RV end-diastolic volume.
doi:10.1164/rccm.201312-2298OC
PMCID: PMC4098110  PMID: 24593877
air pollutants; pulmonary circulation; heart ventricles; pulmonary hypertension
14.  A Unified Spatiotemporal Modeling Approach for Predicting Concentrations of Multiple Air Pollutants in the Multi-Ethnic Study of Atherosclerosis and Air Pollution 
Environmental Health Perspectives  2014;123(4):301-309.
Background:
Cohort studies of the relationship between air pollution exposure and chronic health effects require predictions of exposure over long periods of time.
Objectives:
We developed a unified modeling approach for predicting fine particulate matter, nitrogen dioxide, oxides of nitrogen, and black carbon (as measured by light absorption coefficient) in six U.S. metropolitan regions from 1999 through early 2012 as part of the Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air).
Methods:
We obtained monitoring data from regulatory networks and supplemented those data with study-specific measurements collected from MESA Air community locations and participants’ homes. In each region, we applied a spatiotemporal model that included a long-term spatial mean, time trends with spatially varying coefficients, and a spatiotemporal residual. The mean structure was derived from a large set of geographic covariates that was reduced using partial least-squares regression. We estimated time trends from observed time series and used spatial smoothing methods to borrow strength between observations.
Results:
Prediction accuracy was high for most models, with cross-validation R2 (R2CV) > 0.80 at regulatory and fixed sites for most regions and pollutants. At home sites, overall R2CV ranged from 0.45 to 0.92, and temporally adjusted R2CV ranged from 0.23 to 0.92.
Conclusions:
This novel spatiotemporal modeling approach provides accurate fine-scale predictions in multiple regions for four pollutants. We have generated participant-specific predictions for MESA Air to investigate health effects of long-term air pollution exposures. These successes highlight modeling advances that can be adopted more widely in modern cohort studies.
Citation:
Keller JP, Olives C, Kim SY, Sheppard L, Sampson PD, Szpiro AA, Oron AP, Lindström J, Vedal S, Kaufman JD. 2015. A unified spatiotemporal modeling approach for predicting concentrations of multiple air pollutants in the Multi-Ethnic Study of Atherosclerosis and Air Pollution. Environ Health Perspect 123:301–309; http://dx.doi.org/10.1289/ehp.1408145
doi:10.1289/ehp.1408145
PMCID: PMC4384200  PMID: 25398188
15.  Estimating Acute Air Pollution Health EFFects from Cohort Study Data 
Biometrics  2013;70(1):164-174.
Summary
Traditional studies of short-term air pollution health effects use time series data, while cohort studies generally focus on long-term effects. There is increasing interest in exploiting individual level cohort data to assess short-term health effects in order to understand the mechanisms and time scales of action. We extend semiparametric regression methods used to adjust for unmeasured confounding in time series studies to the cohort setting. Time series methods are not directly applicable since cohort data are typically collected over a prespecified time period and include exposure measurements on days without health observations. Therefore, long-time asymptotics are not appropriate, and it is possible to improve efficiency by exploiting the additional exposure data. We show that flexibility of the semiparametric adjustment model should match the complexity of the trend in the health outcome, in contrast to the time series setting where it suffices to match temporal structure in the exposure. We also demonstrate that pre-adjusting exposures concurrent with the health endpoints using trends in the complete exposure time series results in unbiased health effect estimation and can improve efficiency without additional confounding adjustment. A recently published article found evidence of an association between short-term exposure to ambient fine particulate matter (PM2.5) and retinal arteriolar diameter as measured by retinal photography in the Multi-Ethnic Study of Atherosclerosis (MESA). We reanalyze the data from this article in order to compare the methods described here, and we evaluate our methods in a simulation study based on the MESA data.
doi:10.1111/biom.12125
PMCID: PMC4080094  PMID: 24571570
Air pollution; Environmental epidemiology; Generalized least squares; Mixed models; Semiparametric regression; Time series; Unmeasured confounding.
16.  Measurement error in two-stage analyses, with application to air pollution epidemiology 
Environmetrics  2014;24(8):501-517.
Summary
Public health researchers often estimate health effects of exposures (e.g., pollution, diet, lifestyle) that cannot be directly measured for study subjects. A common strategy in environmental epidemiology is to use a first-stage (exposure) model to estimate the exposure based on covariates and/or spatio-temporal proximity and to use predictions from the exposure model as the covariate of interest in the second-stage (health) model. This induces a complex form of measurement error. We propose an analytical framework and methodology that is robust to misspecification of the first-stage model and provides valid inference for the second-stage model parameter of interest.
We decompose the measurement error into components analogous to classical and Berkson error and characterize properties of the estimator in the second-stage model if the first-stage model predictions are plugged in without correction. Specifically, we derive conditions for compatibility between the first- and second-stage models that guarantee consistency (and have direct and important real-world design implications), and we derive an asymptotic estimate of finite-sample bias when the compatibility conditions are satisfied. We propose a methodology that (1) corrects for finite-sample bias and (2) correctly estimates standard errors. We demonstrate the utility of our methodology in simulations and an example from air pollution epidemiology.
doi:10.1002/env.2233
PMCID: PMC3994141  PMID: 24764691
measurement error; spatial statistics; two-stage estimation; air pollution; environmental epidemiology
17.  A regionalized national universal kriging model using Partial Least Squares regression for estimating annual PM2.5 concentrations in epidemiology 
Many cohort studies in environmental epidemiology require accurate modeling and prediction of fine scale spatial variation in ambient air quality across the U.S. This modeling requires the use of small spatial scale geographic or “land use” regression covariates and some degree of spatial smoothing. Furthermore, the details of the prediction of air quality by land use regression and the spatial variation in ambient air quality not explained by this regression should be allowed to vary across the continent due to the large scale heterogeneity in topography, climate, and sources of air pollution. This paper introduces a regionalized national universal kriging model for annual average fine particulate matter (PM2.5) monitoring data across the U.S. To take full advantage of an extensive database of land use covariates we chose to use the method of Partial Least Squares, rather than variable selection, for the regression component of the model (the “universal” in “universal kriging”) with regression coefficients and residual variogram models allowed to vary across three regions defined as West Coast, Mountain West, and East. We demonstrate a very high level of cross-validated accuracy of prediction with an overall R2 of 0.88 and well-calibrated predictive intervals. In accord with the spatially varying characteristics of PM2.5 on a national scale and differing kriging smoothness parameters, the accuracy of the prediction varies by region with predictive intervals being notably wider in the West Coast and Mountain West in contrast to the East.
doi:10.1016/j.atmosenv.2013.04.015
PMCID: PMC3763950  PMID: 24015108
Ambient air quality; Land use regression; National air quality model; Partial Least Squares; Particulate matter; Universal kriging
18.  Characterizing Spatial Patterns of Airborne Coarse Particulate (PM10–2.5) Mass and Chemical Components in Three Cities: The Multi-Ethnic Study of Atherosclerosis 
Environmental Health Perspectives  2014;122(8):823-830.
Background: The long-term health effects of coarse particular matter (PM10–2.5) are challenging to assess because of a limited understanding of the spatial variation in PM10–2.5 mass and its chemical components.
Objectives: We conducted a spatially intensive field study and developed spatial prediction models for PM10–2.5 mass and four selected species (copper, zinc, phosphorus, and silicon) in three American cities.
Methods: PM10–2.5 snapshot campaigns were conducted in Chicago, Illinois; St. Paul, Minnesota; and Winston-Salem, North Carolina, in 2009 for the Multi-Ethnic Study of Atherosclerosis and Coarse Airborne Particulate Matter (MESA Coarse). In each city, samples were collected simultaneously outside the homes of approximately 40 participants over 2 weeks in the winter and/or summer. City-specific and combined prediction models were developed using land use regression (LUR) and universal kriging (UK). Model performance was evaluated by cross-validation (CV).
Results: PM10–2.5 mass and species varied within and between cities in a manner that was predictable by geographic covariates. City-specific LUR models generally performed well for total mass (CV R2, 0.41–0.68), copper (CV R2, 0.51–0.86), phosphorus (CV R2, 0.50–0.76), silicon (CV R2, 0.48–0.93), and zinc (CV R2, 0.36–0.73). Models pooled across all cities inconsistently captured within-city variability. Little difference was observed between the performance of LUR and UK models in predicting concentrations.
Conclusions: Characterization of fine-scale spatial variability of these often heterogeneous pollutants using geographic covariates should reduce exposure misclassification and increase the power of epidemiological studies investigating the long-term health impacts of PM10–2.5.
Citation: Zhang K, Larson TV, Gassett A, Szpiro AA, Daviglus M, Burke GL, Kaufman JD, Adar SD. 2014. Characterizing spatial patterns of airborne coarse particulate (PM10–2.5) mass and chemical components in three cities: the Multi-Ethnic Study of Atherosclerosis. Environ Health Perspect 122:823–830; http://dx.doi.org/10.1289/ehp.1307287
doi:10.1289/ehp.1307287
PMCID: PMC4123025  PMID: 24642481
19.  Predicting Intra-Urban Variation in Air Pollution Concentrations with Complex Spatio-Temporal Dependencies 
Environmetrics  2009;21(6):606-631.
We describe a methodology for assigning individual estimates of long-term average air pollution concentrations that accounts for a complex spatio-temporal correlation structure and can accommodate spatio-temporally misaligned observations. This methodology has been developed as part of the Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air), a prospective cohort study funded by the U.S. EPA to investigate the relationship between chronic exposure to air pollution and cardiovascular disease. Our hierarchical model decomposes the space-time field into a “mean” that includes dependence on covariates and spatially varying seasonal and long-term trends and a “residual” that accounts for spatially correlated deviations from the mean model. The model accommodates complex spatio-temporal patterns by characterizing the temporal trend at each location as a linear combination of empirically derived temporal basis functions, and embedding the spatial fields of coefficients for the basis functions in separate linear regression models with spatially correlated residuals (universal kriging). This approach allows us to implement a scalable single-stage estimation procedure that easily accommodates a significant number of missing observations at some monitoring locations. We apply the model to predict long-term average concentrations of oxides of nitrogen (NOx) from 2005–2007 in the Los Angeles area, based on data from 18 EPA Air Quality System regulatory monitors. The cross-validated R2 is 0.67. The MESA Air study is also collecting additional concentration data as part of a supplementary monitoring campaign. We describe the sampling plan and demonstrate in a simulation study that the additional data will contribute to improved predictions of long-term average concentrations.
doi:10.1002/env.1014
PMCID: PMC4029437  PMID: 24860253
Air Pollution; Exposure Assessment; Hierarchical Modeling; Spatio-Temporal Modeling; Maximum Likelihood; Universal Kriging
20.  Exposure measurement error in PM2.5 health effects studies: A pooled analysis of eight personal exposure validation studies 
Environmental Health  2014;13:2.
Background
Exposure measurement error is a concern in long-term PM2.5 health studies using ambient concentrations as exposures. We assessed error magnitude by estimating calibration coefficients as the association between personal PM2.5 exposures from validation studies and typically available surrogate exposures.
Methods
Daily personal and ambient PM2.5, and when available sulfate, measurements were compiled from nine cities, over 2 to 12 days. True exposure was defined as personal exposure to PM2.5 of ambient origin. Since PM2.5 of ambient origin could only be determined for five cities, personal exposure to total PM2.5 was also considered. Surrogate exposures were estimated as ambient PM2.5 at the nearest monitor or predicted outside subjects’ homes. We estimated calibration coefficients by regressing true on surrogate exposures in random effects models.
Results
When monthly-averaged personal PM2.5 of ambient origin was used as the true exposure, calibration coefficients equaled 0.31 (95% CI:0.14, 0.47) for nearest monitor and 0.54 (95% CI:0.42, 0.65) for outdoor home predictions. Between-city heterogeneity was not found for outdoor home PM2.5 for either true exposure. Heterogeneity was significant for nearest monitor PM2.5, for both true exposures, but not after adjusting for city-average motor vehicle number for total personal PM2.5.
Conclusions
Calibration coefficients were <1, consistent with previously reported chronic health risks using nearest monitor exposures being under-estimated when ambient concentrations are the exposure of interest. Calibration coefficients were closer to 1 for outdoor home predictions, likely reflecting less spatial error. Further research is needed to determine how our findings can be incorporated in future health studies.
doi:10.1186/1476-069X-13-2
PMCID: PMC3922798  PMID: 24410940
Exposure measurement error; Fine particles; Fine particles of ambient origin; Monitoring data; Spatio-temporal models
21.  Vascular Responses to Long- and Short-Term Exposure to Fine Particulate Matter 
Objectives
This study evaluated the association of long- and short-term air pollutant exposures with flow-mediated dilation (FMD) and baseline arterial diameter (BAD) of the brachial artery using ultrasound in a large multicity cohort.
Background
Exposures to ambient air pollution, especially long-term exposure to particulate matter <2.5 μm in aerodynamic diameter (PM2.5), are linked with cardiovascular mortality. Short-term exposure to PM2.5 has been associated with decreased FMD and vasoconstriction, suggesting that adverse effects of PM2.5 may involve endothelial dysfunction. However, long-term effects of PM2.5 on endothelial dysfunction have not been investigated.
Methods
FMD and BAD were measured by brachial artery ultrasound at the initial examination of the Multi-Ethnic Study of Atherosclerosis. Long-term PM2.5 concentrations were estimated for the year 2000 at each participant’s residence (n = 3,040) using a spatio-temporal model informed by cohort-specific monitoring. Short-term PM2.5 concentrations were based on daily central-site monitoring in each of the 6 cities.
Results
An interquartile increase in long-term PM2.5 concentration (3 μg/m3) was associated with a 0.3% decrease in FMD (95% confidence interval [CI] of difference: −0.6 to −0.03; p = 0.03), adjusting for demographic characteristics, traditional risk factors, sonographers, and 1/BAD. Women, nonsmokers, younger participants, and those with hypertension seemed to show a greater association of PM2.5 with FMD. FMD was not significantly associated with short-term variation in PM2.5 (−0.1% per 12 μg/m3 daily increase [95% CI: −0.2 to 0.04] on the day before examination).
Conclusions
Long-term PM2.5 exposure was significantly associated with decreased endothelial function according to brachial ultrasound results. These findings may elucidate an important pathway linking air pollution and cardiovascular mortality.
doi:10.1016/j.jacc.2012.08.973
PMCID: PMC3665082  PMID: 23103035
air pollution; atherosclerosis; cardiovascular mortality; endothelial function; flow-mediated dilation; traffic
22.  Prospective Study of Particulate Air Pollution Exposures, Subclinical Atherosclerosis, and Clinical Cardiovascular Disease 
American Journal of Epidemiology  2012;176(9):825-837.
The Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air) was initiated in 2004 to investigate the relation between individual-level estimates of long-term air pollution exposure and the progression of subclinical atherosclerosis and the incidence of cardiovascular disease (CVD). MESA Air builds on a multicenter, community-based US study of CVD, supplementing that study with additional participants, outcome measurements, and state-of-the-art air pollution exposure assessments of fine particulate matter, oxides of nitrogen, and black carbon. More than 7,000 participants aged 45–84 years are being followed for over 10 years for the identification and characterization of CVD events, including acute myocardial infarction and other coronary artery disease, stroke, peripheral artery disease, and congestive heart failure; cardiac procedures; and mortality. Subcohorts undergo baseline and follow-up measurements of coronary artery calcium using computed tomography and carotid artery intima-medial wall thickness using ultrasonography. This cohort provides vast exposure heterogeneity in ranges currently experienced and permitted in most developed nations, and the air monitoring and modeling methods employed will provide individual estimates of exposure that incorporate residence-specific infiltration characteristics and participant-specific time-activity patterns. The overarching study aim is to understand and reduce uncertainty in health effect estimation regarding long-term exposure to air pollution and CVD.
doi:10.1093/aje/kws169
PMCID: PMC3571256  PMID: 23043127
air pollution; atherosclerosis; cardiovascular diseases; environmental exposure; epidemiologic methods; particulate matter
23.  A National Prediction Model for PM2.5 Component Exposures and Measurement Error–Corrected Health Effect Inference 
Environmental Health Perspectives  2013;121(9):1017-1025.
Background: Studies estimating health effects of long-term air pollution exposure often use a two-stage approach: building exposure models to assign individual-level exposures, which are then used in regression analyses. This requires accurate exposure modeling and careful treatment of exposure measurement error.
Objective: To illustrate the importance of accounting for exposure model characteristics in two-stage air pollution studies, we considered a case study based on data from the Multi-Ethnic Study of Atherosclerosis (MESA).
Methods: We built national spatial exposure models that used partial least squares and universal kriging to estimate annual average concentrations of four PM2.5 components: elemental carbon (EC), organic carbon (OC), silicon (Si), and sulfur (S). We predicted PM2.5 component exposures for the MESA cohort and estimated cross-sectional associations with carotid intima-media thickness (CIMT), adjusting for subject-specific covariates. We corrected for measurement error using recently developed methods that account for the spatial structure of predicted exposures.
Results: Our models performed well, with cross-validated R2 values ranging from 0.62 to 0.95. Naïve analyses that did not account for measurement error indicated statistically significant associations between CIMT and exposure to OC, Si, and S. EC and OC exhibited little spatial correlation, and the corrected inference was unchanged from the naïve analysis. The Si and S exposure surfaces displayed notable spatial correlation, resulting in corrected confidence intervals (CIs) that were 50% wider than the naïve CIs, but that were still statistically significant.
Conclusion: The impact of correcting for measurement error on health effect inference is concordant with the degree of spatial correlation in the exposure surfaces. Exposure model characteristics must be considered when performing two-stage air pollution epidemiologic analyses because naïve health effect inference may be inappropriate.
Citation: Bergen S, Sheppard L, Sampson PD, Kim SY, Richards M, Vedal S, Kaufman JD, Szpiro AA. 2013. A national prediction model for PM2.5 component exposures and measurement error–corrected health effect inference. Environ Health Perspect 121:1017–1025; http://dx.doi.org/10.1289/ehp.1206010
doi:10.1289/ehp.1206010
PMCID: PMC3764074  PMID: 23757600
24.  Association of Long-Term Air Pollution with Ventricular Conduction and Repolarization Abnormalities 
Epidemiology (Cambridge, Mass.)  2011;22(6):773-780.
Background
Short-term exposure to air pollution may affect ventricular repolarization, but there is limited information on how long-term exposures might affect the surface ventricular electrocardiographic (ECG) abnormalities associated with cardiovascular events. We carried out a study to determine whether long-term air pollution exposure is associated with abnormalities of ventricular repolarization and conduction in adults without known cardiovascular disease.
Methods
A total of 4783 participants free of clinical cardiovascular disease in the Multi-Ethnic Study of Atherosclerosis underwent 12-lead ECG examinations, cardiac-computed tomography and calcium scoring, as well as estimation of air pollution exposure using a finely resolved spatio-temporal model to determine long-term average individual exposure to fine particulate matter (PM2.5) and proximity to major roadways. We assessed ventricular electrical abnormalities including presence of QT prolongation (Rautaharju QTrr criteria) and intraventricular conduction delay (QRS duration > 120 msec). We used logistic regression to determine the adjusted relationship between air pollution exposures and ECG abnormalities.
Results
A 10 µg/m3-increase in estimated residential PM2.5 was associated with an increased odds of prevalent QT prolongation (adjusted odds ratio [OR]= 1.6 [95% confidence interval (CI)= 1.2 to 2.2]) and intraventricular conduction delay (OR 1.7, 95% CI: 1.0 to 2.6, independent of coronary-artery calcium score. Living near major roadways was not associated with ventricular electrical abnormalities. No significant evidence of effect modification by traditional risk factors or study site was observed.
Conclusions
This study demonstrates an association between long-term exposure to air pollution and ventricular repolarization and conduction abnormalities in adults without clinical cardiovascular disease, independent of subclinical coronary arterial calcification.
doi:10.1097/EDE.0b013e31823061a9
PMCID: PMC3197855  PMID: 21918454
25.  Efficient measurement error correction with spatially misaligned data 
Biostatistics (Oxford, England)  2011;12(4):610-623.
Association studies in environmental statistics often involve exposure and outcome data that are misaligned in space. A common strategy is to employ a spatial model such as universal kriging to predict exposures at locations with outcome data and then estimate a regression parameter of interest using the predicted exposures. This results in measurement error because the predicted exposures do not correspond exactly to the true values. We characterize the measurement error by decomposing it into Berkson-like and classical-like components. One correction approach is the parametric bootstrap, which is effective but computationally intensive since it requires solving a nonlinear optimization problem for the exposure model parameters in each bootstrap sample. We propose a less computationally intensive alternative termed the “parameter bootstrap” that only requires solving one nonlinear optimization problem, and we also compare bootstrap methods to other recently proposed methods. We illustrate our methodology in simulations and with publicly available data from the Environmental Protection Agency.
doi:10.1093/biostatistics/kxq083
PMCID: PMC3169665  PMID: 21252080
Environmental epidemiology; Environmental statistics; Exposure modeling; Kriging; Measurement error

Results 1-25 (33)