Faculty Bibliography

Central monitoring site (CMS) concentrations have been used to represent population-based personal exposures to particulate matter (PM) of ambient origin. We investigated the associations of the concentrations of PM(2.5) and PM(10) and their elemental components for elderly clinic patients with chronic obstructive pulmonary disease in two cities with different PM compositions, that is, New York City (NYC) and Seattle. Daily measurements of CMS, outdoor residential, and indoor PM(10) and PM(2.5) concentrations, as well as personal PM(10), were made concurrently for 12-consecutive winter days at 9 NYC and 15 Seattle residences, as well for 9 NYC residences in summer. Filters were analyzed for elemental components using X-ray fluorescence (XRF), and for black carbon (BC) by light reflectance, and outdoor-indoor-personal relationships of PM components were examined using mixed-effect models. Using sulfur (S) as a tracer of PM of ambient origin, the mean contributions of outdoor PM(2.5) was 55.2% of the indoor concentrations in NYC, and 80.0% in Seattle, and outdoor PM(2.5) in NYC and Seattle were 19.7 and 18.5% of personal PM(2.5) concentration. S was distributed homogeneously in both cities (R(2)=0.65), whereas nickel (R(2)=0.23) was much more spatially heterogeneous. Thus, CMS measurements can adequately reflect personal exposures for spatially uniform components, such as sulfate, but they are not adequate for components from more local sources.

Background: Asthma is a growing epidemic in the United States. Associations of acute air pollution with asthma hospital admissions have been shown to be modified by co-morbid conditions. However, no known previous study has investigated the modifying effects of concurrent respiratory infections, specifically bronchitis and upper respiratory infection (URI), on these associations. Methods: Data for 105,914 unscheduled hospital admissions in New York City with a primary diagnosis of asthma were obtained for years 2003-2006. A Poisson generalized linear model was applied to determine the association of inter-quartile increases in daily ambient air pollution (i.e., PM2.5, NO2, SO2, and O3) exposures with asthma hospital admissions, after controlling for meteorological and temporal variables. Stratified analyses were conducted for admissions with secondary diagnoses of bronchitis and URI, respectively. Results: Air pollution associations of asthma admissions with a secondary diagnosis of bronchitis were significantly higher vs. admissions without a secondary diagnosis of bronchitis (e.g., PM2.5 risk ratios are 1.19 (95% CI = 1.07-1.30) with bronchitis vs. 1.05 (95% CI = 1.02-1.07) without bronchitis). However, secondary diagnoses of URI did not modify the associations of air pollutants with asthma hospital admissions. These results were observed across multiple pollutants and age groups. Discussion: There is an increased risk of air pollution related asthma hospital admission among individuals with bronchitis, but not with URI. The identification of differential risks based on infection of the bronchioles, but not the upper respiratory tract, provides an opportunity for further investigation into the mechanisms by which air pollution affects respiratory function

Background: Multi-city time-series and long-term cohort studies have reported associations between PM2.5 and both mortality and hospitalizations. However, the specific PM components most responsible for the observed associations have not been established. Objective: To identify PM2.5 sources responsible for both short- and long-term health effects. Methods: We examined both acute and long-term epidemiological associations across the U.S. Using Poisson modeling, daily mortality (years 2001-2006), and elderly hospitalizations (years 2000-2008) were associated with PM2.5 and its chemical components using EPA's CSN data in 150 U.S. cities, and separately for a 64 cities sub-set where PM2.5, its chemical constituents, and gaseous pollutant data were all available. Using the nationwide American Cancer Society CPS-II Cohort, we similarly investigated U.S. mortality associations with components of long-term PM2.5 exposures.. Results: Nationwide factor analyses identified seven major sources: Traffic, Soil, Coal combustion, Salt, Metals, Biogenic Burning, and Residual Oil combustion. The Traffic factor showed most consistent associations with daily all-cause mortality, and with CVD and respiratory hospitalizations. Our extended ACS cohort follow-up analyses (through 2004) indicated that Coal Combustion PM2.5 was most consistently associated with increased risk of IHD mortality. Soil or Wood burning sources were consistently not associated with any causes of mortality. Conclusions: Associations of components of PM2.5 and health varied by outcome (hospitalization vs. mortality), by cause (cardiac vs. respiratory), and between acute and long-term exposures. In general, acute morbidity and mortality associations were most related to traffic-related components, while chronic mortality was most strongly associated with long-term exposure to Coal Combustion PM2.5

Background. Particulate matter (PM) is a complex mixture of chemical constituents affecting health risks. However, current health-based PM standards are limited to mass concentrations. Objective. To conduct an integrated series of toxicology and epidemiology studies to determine which PM components are most influential. Methods. Using EPA's CSN data enabled us to study the influence of PM2.5 components on: 1) shortterm human morbidity and mortality in 150 US cities; and 2) annual mortality rates in 100 US cities. We also conducted: 3) a series of 6-month subchronic inhalation studies (6 h/d, 5 d/wk) of concentrated (10X) ambient air PM2.5 (CAPs) in ApoE-/- mice in five U.S. airsheds; and 4) we administered aliquots of PM10-2.5, PM2.5-0.2, and PM<0.2 samples from the same 5 airsheds to cells in vitro, and to mouse lungs in vivo by aspiration to determine their comparative acute effects. Results. Across all four substudies, fossil-fuel combustion sources were most consistently associated with both short- and long-term cardiovascular disease (CVD) effects. Daily human CVD mortality and hospital admissions, and cardiac function in the mice, were most closely associated with constituents from residual oil combustion and traffic, while annual human CVD mortality and aortic plaque progression in the mice were more closely associated with coal combustion effluents. Conclusions. 1) PM toxicity is driven by a complex interaction of particle size, study site, and season; 2) the most influential constituents vary with endpoint. These results can: 1) guide future research; selection of emission controls; and 3) optimization of the public health benefits

Background: Chronic psychological stress has been linked to respiratory health, both independently and in combination with air pollution. Robust exposure assessment must, therefore, address geographic patterning among multiple stressors (e.g., exposures may be concentrated in communities of color and lower socioeconomic position) and interpretability of administrative data. Objective: To validate publicly-available indicators of stressors through a community-based qualitative process, and apply GIS-based methods to explore the potential for spatial confounding and effect modification. Methods: Approximately 25 spatially-distributed focus groups are being conducted to identify, describe and rank psychosocial stressors reported by residents of diverse communities. Stressors perceived as important by community-members are compared against publicly-available data (e.g., crime rates, access to resources). Spatial correlation and autocorrelation among stressors, and with community-level air pollution concentrations derived from the New York City Community Air Survey (NYCCAS), are assessed using Simultaneous Autoregressive (SAR) modelling. Results: Focus group discussions provide information on intra-urban variability in perceptions of key community social stressors, yielding a validated, city-wide GIS database (n=16). Preliminary analyses identified widespread spatial autocorrelation, and revealed inconsistent spatial correlations among community-level stressor and pollution exposures - rates of assault and noise disruption were positively correlated with poverty (r=.72 and .57, respectively), but not with air pollutant concentrations, while rates of serious housing violations were positively correlated with poverty (r=.80) and pollutant concentrations (PM2.5 r=.89, SO2 r=.81). Conclusions: Where community social stressors vary in both prevalence and perceived importance, challenges of construct validity of administrative data may be addressed through complementary, spatially-informed qualitative and!

Background: Which components of ambient air particulate matter (PM) are most associated with adverse health effects is less well known. In addition, current Chemical Speciation Network (CSN) data, with every 3rd/6th day sampling schedule, do not allow an examination of multi-day effects of chemical components. Objectives: To determine the associations between daily PM2.5 (PM < 2.5 mum in aerodynamic diameter) components and hospitalizations in Seattle and Detroit using time-series analyses. Methods: We obtained daily PM2.5 Teflon filters for the years of 2002-2004 and analyzed trace elements using X-ray fluorescence, and black carbon (BC) using light reflectance as a surrogate measure of elemental carbon (EC). We used Poisson regression and distributed-lag models to estimate cumulative effects (lags of 0 through 2 days) for cardiovascular and respiratory diseases, with adjustments for time-varying covariates. We computed the excess risks for interquartile range (IQR) increases of each pollutant component for both warm and cold seasons. Results: The PM2.5 components and gaseous pollutants most closely associated with cardiovascular and respiratory hospitalization in Detroit were secondary aerosols, traffic markers, and biomass combustion; while in Seattle, those for cold season traffic emissions and residual oil burning. Conclusions: The effects of PM2.5 on daily hospitalizations vary with source, season, and locale, consistent with the hypothesis that PM2.5 composition has an appreciable influence on the health effects attributable to PM2.5. The multi-day risk estimates were generally bigger than individual day effects, suggesting that risk assessments using a single day lag model are likely to underestimate the health impacts

BACKGROUND:Epidemiological studies have linked both noise and air pollution to common adverse health outcomes such as increased blood pressure and myocardial infarction. In urban settings, noise and air pollution share important sources, notably traffic, and several recent studies have shown spatial correlations between noise and air pollution. The temporal association between these exposures, however, has yet to be thoroughly investigated despite the importance of time series studies in air pollution epidemiology and the potential that correlations between these exposures could at least partly confound statistical associations identified in these studies. METHODS:An aethelometer, for continuous elemental carbon measurement, was co-located with a continuous noise monitor near a major urban highway in New York City for six days in August 2009. Hourly elemental carbon measurements and hourly data on overall noise levels and low, medium and high frequency noise levels were collected. Hourly average concentrations of fine particles and nitrogen oxides, wind speed and direction and car, truck and bus traffic were obtained from nearby regulatory monitors. Overall temporal patterns, as well as day-night and weekday-weekend patterns, were characterized and compared for all variables. RESULTS:Noise levels were correlated with car, truck, and bus traffic and with air pollutants. We observed strong day-night and weekday-weekend variation in noise and air pollutants and correlations between pollutants varied by noise frequency. Medium and high frequency noise were generally more strongly correlated with traffic and traffic-related pollutants than low frequency noise and the correlation with medium and high frequency noise was generally stronger at night. Correlations with nighttime high frequency noise were particularly high for car traffic (Spearman rho=0.84), nitric oxide (0.73) and nitrogen dioxide (0.83). Wind speed and direction mediated relationships between pollutants and noise. CONCLUSIONS:Noise levels are temporally correlated with traffic and combustion pollutants and correlations are modified by the time of day, noise frequency and wind. Our results underscore the potential importance of assessing temporal variation in co-exposures to noise and air pollution in studies of the health effects of these urban pollutants.

OBJECTIVE-Hemoglobin A(1c) (HbA(1c)) is recommended for identifying diabetes and prediabetes. Because HbAi, does not fluctuate with recent eating or acute illness, it can be measured in a variety of clinical settings. Although outpatient studies identified HbA(1c)-screening cutoff values for diabetes and prediabetes, HbA(1c)-screening thresholds have not been determined for acute-care settings. Using follow-up fasting blood glucose (FBG) and the 2-h oral glucose tolerance test (OGTT) as the criterion gold standard, we determined optimal HbA(1c)-screening cutoffs for undiagnosed dysglycemia in the emergency department setting. RESEARCH DESIGN AND METHODS This was a prospective observational study of adults aged >= 18 years with no known history of hyperglycemia presenting to an emergency department with acute illness. Outpatient FBS and 2-h OGTT were performed after recovery from the acute illness, resulting in diagnostic categorizations of prediabetes, diabetes, and dysglycemia (prediabetes or diabetes). Optimal cutoffs were determined and performance data identified for cut points. RESULTS-A total of 618 patients were included, with a mean age of 49.7 (+/- 14.9) years and mean HbA(1c) of 5.68% (+/- 0.86). On the basis of an OGTT, the prevalence of previously undiagnosed prediabetes and diabetes was 31.9 and 10.5%, respectively. The optimal HbA(1c)-screening cutoff for prediabetes was 5.7% (area under the curve [AUC] = 0.659, sensitivity = 55%, and specificity = 71%), for dysglycemia 5.8% (AUC = 0.717, sensitivity = 57%, and specificity = 79%), and for diabetes 6.0% (AUC = 0.868, sensitivity = 77%, and specificity = 87%). CONCLUSIONS-We identified HbA(1c) cut points to screen for prediabetes and diabetes in an emergency department adult population. The values coincide with published outpatient study findings and suggest that an emergency department visit provides an opportunity for HbA(1c)-based dysglycemia screening