American Journal of Respiratory and Critical Care Medicine

Rationale: Whereas associations between air pollution and respiratory morbidity for adults 65 years and older are well documented in the United States, the evidence for people under 65 is less extensive. To address this gap, the Centers for Disease Control and Prevention’s National Environmental Public Health Tracking Program collected respiratory emergency department (ED) data from 17 states.

Objectives: To estimate age-specific acute effects of ozone and fine particulate matter (particulate matter ≤2.5 mm in aerodynamic diameter [PM2.5]) on respiratory ED visits.

Methods: We conducted time-series analyses in 894 counties by linking daily respiratory ED visits with estimated ozone and PM2.5 concentrations during the week before the date of the visit. Overall effect estimates were obtained with a Bayesian hierarchical model to combine county estimates for each pollutant by age group (children, 0–18; adults, 19–64; adults ≥ 65, and all ages) and by outcome group (acute respiratory infection, asthma, chronic obstructive pulmonary disease, pneumonia, and all respiratory ED visits).

Measurements and Main Results: Rate ratios (95% credible interval) per 10-μg/m3 increase in PM2.5 and all respiratory ED visits were 1.024 (1.018–1.029) among children, 1.008 (1.004–1.012) among adults younger than 65 years, and 1.002 (0.996–1.007) among adults 65 and older. Per 20-ppb increase in ozone, rate ratios were 1.017 (1.011–1.023) among children, 1.051 (1.046–1.056) among adults younger than 65, and 1.033 (1.026–1.040) among adults 65 and older. Associations varied in magnitude by age group for each outcome group.

Conclusions: These results address a gap in the evidence used to ensure adequate public health protection under national air pollution policies.

Scientific Knowledge on the Subject

Short-term exposure to ambient ozone and fine particulate matter (particulate matter ≤2.5 mm in aerodynamic diameter [PM2.5]) is associated with increases in mortality and morbidity. In the United States, robust evidence from multicity studies exists for mortality and for hospitalizations among adults 65 years of age and older. These studies have contributed to the establishment of National Ambient Air Quality Standards. The evidence for morbidity outcomes in people under 65 years, however, is limited to single-city studies. Such studies have limitations with respect to nationwide generalizability because of between-city differences in air pollution composition and population characteristics as well as differences in study methodology.

What This Study Adds to the Field

Our analysis of 894 counties is the largest U.S. multicity study of air pollution and respiratory emergency department visits among all ages. We observed associations between ozone and respiratory emergency department visits among all age groups, with the highest magnitude association among adults younger than 65 years of age. Fine particulate matter was associated with respiratory emergency department visits among children younger than 19 and adults younger than 65 but not among adults 65 and older. The associations also varied by age group for specific respiratory outcomes. Reliance on associations from a single age group or outcome to evaluate air pollution standards may over- or underestimate the health benefits.

Substantial and consistent evidence supports the conclusion that short-term exposure to ambient ozone and fine particulate matter (particulate matter ≤2.5 mm in aerodynamic diameter [PM2.5]) is associated with increases in mortality and morbidity (1, 2). Robust evidence exists for mortality and for hospitalizations among adults 65 years and older, due to the availability of national vital statistics data and Medicare data. These data have enabled many multicity analyses, which have been instrumental in setting National Ambient Air Quality Standards for the Clean Air Act (39). Estimates from multicity studies provide strong evidence for determining causality between air pollution and health, for evaluating potential health benefits of proposed policies across the United States, and for establishing ambient air quality standards that provide adequate protection for the U.S. population.

Because national data sets are lacking, the evidence for morbidity outcomes for U.S. populations not covered by Medicare (i.e., individuals < 65 yr) often comes from single-city studies (1013). This is a particularly important issue for respiratory emergency department (ED) visits, because the vast majority of these visits occur in people under age 65 years (14). Collectively, the evidence from single-city studies indicates positive associations between air pollution and ED visits for all respiratory diseases combined, and for asthma, chronic obstructive pulmonary disease (COPD), and respiratory infections (7, 1521). Although informative, these studies have limitations with respect to nationwide generalizability due to between-city differences in air pollution composition and population characteristics as well as differences in study methodology. For the current ozone and particulate matter (PM) standards (22, 23), the U.S. Environmental Protection Agency (EPA) conducted national-scale mortality risk assessments using results from multicity studies. For morbidity outcomes, the EPA conducted similar risk assessments for hospital admissions among all ages in only 12 cities for ozone and 15 cities for PM where epidemiologic study results and necessary data were available. For respiratory-related ED visits among all ages, their assessment was conducted only in Atlanta and New York City, indicating a strong need for a multicity study.

In response to a Pew Commission report (24), the Centers for Disease Control and Prevention (CDC) National Environmental Public Health Tracking Program (Tracking Program) was launched to integrate health, exposure, and environmental hazard data to inform environmental health programs and policies. The National Environmental Public Health Tracking Network (Tracking Network) is a web-based system with components at national, state, and local levels (https://ephtracking.cdc.gov/). It is used to collect, integrate, analyze, and disseminate health and environmental data that drive actions to improve the health of communities. We invited 30 states known to have ED data centralized within their state to participate in the project. Using the Tracking Network, we collected daily, county respiratory ED data for all ages from 17 states representing 45% of the U.S. population (138.5 million individuals). We used these data to perform the first nationally relevant study to estimate associations between ozone and PM2.5 and respiratory ED visits among all ages in the United States. Partial results have been previously reported in the form of an abstract (25).

We requested daily number of ED visits with a primary diagnosis of respiratory disease (International Classification of Diseases, Ninth Revision [ICD-9] codes 460–519) aggregated by county, age group, and outcome group to minimize confidentiality concerns and increase state participation. Seventeen states (California, Colorado, Florida, Illinois, Iowa, Louisiana, Maine, Massachusetts, Minnesota, Missouri, New Hampshire, New Mexico, New York, North Carolina, South Carolina, Utah, and Vermont) provided data for 2000 through 2014 with 3 to 13 years per state (data availability varied by state) (see Table E1 in the online supplement). We created 19 age-specific outcomes by grouping the data into four age groups by five outcome groups, excluding COPD among children. The age groups included all ages combined, children 0 to less than 19 years (subsequently referred to as “children”), adults 19 to less than 65 years (“adults”), and adults 65 years and older (“older adults”). The outcome groups included all respiratory outcomes combined (ICD-9 codes 460–519); acute respiratory infection (ARI) including upper respiratory infections, bronchitis, and bronchiolitis (ICD-9 codes 460–466.0); asthma (ICD-9 code 493); COPD (ICD-9 code 491, 492, or 496); and pneumonia (ICD-9 codes 480–486).

For ambient air pollution concentrations, we used data from the Bayesian space–time downscaling (DS) fusion modeling approach, developed by the EPA and its partners (26). The DS modeling approach uses a Bayesian framework and spatially varying coefficients to regress monitoring measurements from Air Quality System sites onto the gridded output from the Community Multiscale Air Quality model. The process yields gridded, daily predictions of ozone and PM2.5 concentrations that are more accurate and precise than with previously used modeling approaches (2628). Predictions of daily maximum 8-hour average ozone concentrations in parts per billion ozone and daily 24-hour average PM2.5 from the DS model were generated at census-tract centroids for the contiguous United States as part of an interagency agreement between the CDC and EPA. For this study, we generated daily, population-weighted county-level estimates of ozone and PM2.5, for Years 2001 through 2012. Population-weighted, county-level estimates correlate well with county-specific air monitors, which are used in the regulatory process (29, 30). Daily, county-level estimates of maximum temperature and dew point temperature in degrees Fahrenheit were generated from the North American Land Data Assimilation System (NLDAS) model (31). We converted predictions from the NLDAS model from a grid resolution of 14 km × 14 km to the county level, using a previously cited geo-imputation approach (32). We used SAS version 9.3, Python version 3.3.2, and ArcGIS version 9.3 (provided by the CDC) to prepare the environmental data sets.

We used a two-stage model to obtain nationally relevant estimates of short-term associations between ozone and PM2.5 and respiratory ED visits. In the first stage, we fit time-series models (n = 16,986) for each combination of county (n = 894) and age-specific outcome (n = 19). To minimize issues with model convergence, we excluded those county- and age-specific outcome combinations where more than 90% of days had zero ED visits (n = 4,681). In these counties the population is low, resulting in too few ED visits for the time-series model to fit. We modeled an unconstrained distributed lag for Lag Days 0 through 6 to capture the cumulative association of exposures over the past week, using a Poisson log-linear model that accounted for overdispersion. We fit single-pollutant models for ozone and PM2.5 and two-pollutant models with both. Models included 1) nonlinear functions of same-day maximum temperature, same-day maximum dew point temperature, and previous 6-day average maximum temperature as natural cubic splines with 3 degrees of freedom; 2) a nonlinear function of calendar date as a natural cubic spline with 12 degrees of freedom per year of county data; and 3) indicator variables for day of week and for holidays. For the second stage, two-level Bayesian hierarchical models with noninformative priors were fit to combine county-specific effect estimates for each age-specific outcome to obtain nationally relevant effect estimates for ozone and PM2.5 (33). We report the rate ratio (RR) and 95% credible interval (CI), which we generated using the posterior probability distribution. Given our assumptions, the 95% CI has the property that there is a 95% probability that the true RR falls within the interval. We evaluated the sensitivity of our results by running the models with various degrees of freedom on calendar date (6, 8, 10, and 12 per year) and with three combinations of weather variables: 1) temperature, 2) temperature and dew point temperature, and 3) temperature, dew point temperature, and previous 6-day average temperature. We also ran the models with Lag Day –1 (pollution on the following day) as a negative control exposure to estimate the association with ozone and PM2.5 on the day after the ED visit (34). All models were implemented with R statistical software (version 3.3.2; R Foundation for Statistical Computing) on the CDC high-performance computing Linux cluster. Models were implemented using the glm function and purrr, broom, and TLNise packages.

Our analysis included 38.4 million respiratory ED visits from 869 counties representing 45% of the U.S. population (Table 1). Children and adults had approximately 16 million visits each whereas older adults had about 6 million visits. The mean daily rates of respiratory ED visits per 10,000 people were 1.20 for all ages combined, 1.94 for children, 0.91 for adults, and 1.37 for older adults. Rates of ARI and asthma were highest among children, whereas the rates for COPD and pneumonia were highest among older adults (Figure 1). We calculated the interquartile range (IQR) of each pollutant for each county for the specific years of data used in that county’s time-series analysis. The 869 county-specific IQR for daily 8-hour maximum ozone varied from 8.0 to 34.0 ppb with a mean IQR of 16.54 ppb (Figure 2). For 24-hour average PM2.5 levels, the mean county IQR was 5.3 μg/m3 and ranged from 1.9 to 9.8 μg/m3.

Table 1. Overall Mean and Variance of County-Level Mean Daily Counts and Rates per 10,000 Population of Respiratory Emergency Department Visits for Counties Analyzed by Outcome and Age Group

OutcomeAge Group (yr)ED VisitsCountiesDaily CountDaily Rate
MeanVarianceMeanVariance
AllAll39,975,41186917.262,383.551.200.313
0 to <1916,096,4438367.34523.191.940.914
19 to <6516,398,4388367.50400.900.910.206
65+5,884,3337892.7851.351.370.296
        
ARIAll18,169,8168368.13498.030.570.090
0 to <199,534,5468124.48189.911.220.400
19 to <657,176,1477743.4968.610.420.058
65+701,2124190.611.160.210.014
        
AsthmaAll5,761,7126913.10100.700.140.005
0 to <192,265,8105171.6422.150.230.014
19 to <652,717,7815941.7630.870.130.004
65+379,4232060.681.650.090.002
        
COPDAll2,385,1486781.327.190.100.003
0 to <19NANANANANANA
19 to <65896,8085060.691.430.080.002
65+1,342,4795710.882.960.380.027
        
PneumoniaAll4,659,8637492.3135.250.160.004
0 to <191,294,8444671.046.020.160.005
19 to <651,425,6215660.974.550.090.001
65+1,660,3535891.045.270.460.036

Definition of abbreviations: ARI = acute respiratory infection; COPD = chronic obstructive pulmonary disease; ED = emergency department; NA = not applicable.

We observed statistically significant positive associations between ozone and all respiratory ED visits combined in both single- and two-pollutant models for all age groups (Figure 3). For all ages combined, the RR per 20-ppb increase in ozone on Lag Days 0 through 6 for all respiratory ED visits was 1.039 (CI, 1.036–1.042). In the two-pollutant models, the association between all respiratory ED visits and ozone was strongest among adults followed by older adults. In contrast, PM2.5 was associated with all respiratory ED visits in both single- and two-pollutant models for all age groups, except in the two-pollutant model, where the association was consistent with the null among older adults (Figure 4). For all ages combined, the RR per 10-μg/m3 increase in PM2.5 on Lag Days 0 through 6 for all respiratory ED visits was 1.020 (CI, 1.017–1.023). The RRs between all respiratory ED visits combined and PM2.5 was higher among children than adults.

For both pollutants, the RR varied in magnitude by outcome and age group. For ozone, we observed significant and positive associations with asthma, ARI, COPD, and pneumonia among all ages combined in both single-pollutant and two-pollutant models (Figure 3). We also observed significant, positive associations between ozone and each outcome for all age groups except for asthma among older adults. The highest magnitude RR for ozone was asthma among adults (RR, 1.064; 95% CI, 1.053–1.076). RRs were generally highest among the adult group except for ARI, where the RRs for adults and older adults were similar. The RR for the association between ozone and ARI among children was elevated and significant but lower than the RRs for adults and older adults. The RRs for pneumonia exhibit less variation than the RRs for the other outcomes. For PM2.5 and all ages combined, we observed positive, significant associations for asthma and ARI in single- and two-pollutant models and pneumonia in single-pollutant models (Figure 4). We also observed positive, significant associations for asthma and ARI among each age group. In contrast to our ozone results, we observed less variation in the RR by age group for asthma and ARI, and the association for pneumonia was significant only among adults. We found little evidence of an association between PM2.5 and COPD among any age groups in either single- or two-pollutant models.

In our sensitivity analyses, the RR decreased slightly as the number of degrees of freedom decreased for ozone; association estimates for PM2.5 changed little (Figure E1). It is notable that, for children, many RRs for ozone decreased in magnitude from a positive association to a negative or null association as we decreased temporal control. The RRs for ozone and ARI among children were consistent with the null when using either 6 or 8 degrees of freedom, but were significant and positive with either 10 or 12 degrees of freedom (Figure E1). It is plausible that the most valid degree of temporal control may vary by outcome in relation to the degree of seasonality. For the three combinations of weather variables we evaluated, our results fluctuated slightly for ozone but were robust for PM2.5, with overlapping 95% credible intervals (Figure E2). The associations observed between ozone or PM2.5 concentrations on Lag Day −1 (pollution on the following day) and ED visits for asthma, COPD, and pneumonia were consistent with the null supporting our model specifications (Figure E3).

Our nationally relevant study, based on nearly 40 million respiratory ED visits representing all ages, addresses an important gap in air pollution epidemiology and is a valuable reference for future national air pollution standards. Our results support the EPA’s determination of a likely causal relationship between PM2.5 and respiratory effects and a causal relationship between ozone and respiratory effects (1, 2). However, our results highlight important variation in magnitude across age groups, outcomes, and pollutants. PM2.5 and all respiratory ED visits were strongly associated among children, moderately associated among adults, and not associated among older adults. Whereas the associations with PM2.5 were elevated and similar among all age groups for asthma, they varied by age group for ARI and pneumonia. Conversely, the association between ozone and all respiratory ED visits was higher in magnitude among both adult groups than the association among children. The association between ozone and asthma varied by age group, with the highest magnitude among adults. Associations varied by age group for ARI and COPD but were similar across age groups for pneumonia. These findings indicate that multicity studies of populations over 65 years of age may not be a good proxy of acute respiratory impacts on populations under 65, and that reliance on such studies could underestimate population respiratory health impacts of PM2.5 or ozone that were stronger in our study for younger age groups.

Few single-city studies have estimated associations of ozone or PM2.5 across different age groups for ED visits or hospitalizations. Those that have looked specifically at asthma and generally found a greater impact on children (10, 15, 35). We observed similar associations between PM2.5 and asthma across age groups and a stronger association for ozone on asthma among adults younger than 65 years, an age group previously not included in other multicity studies of morbidity. Differences in results could be due to differences in age group specification. For example, we included all children under 19 years in one age group whereas other studies separate children under 5 and children 5 to 19. The diagnosis of asthma in children under 5 years of age is difficult and could result in more outcome measurement error when included in the analysis (36).

For ozone, the EPA found consistent evidence of positive associations for asthma and COPD based on the few studies evaluating hospital admissions and ED visits for respiratory outcomes (2). Although the EPA found strong toxicologic evidence supporting an association between ozone and respiratory infections and pneumonia, the epidemiologic evidence was inconsistent. We observed positive associations between ozone and ARI, asthma, COPD, and pneumonia for all age groups, with the only exception asthma among older adults while adjusting for PM2.5. This supports the EPA’s conclusions and adds to epidemiologic evidence for ARI and pneumonia. For PM2.5, the EPA’s review of available studies produced stronger and more consistent evidence for COPD and respiratory infection effects than for asthma effects, including both ED visits and hospitalizations. In contrast, we observed significant, positive associations between PM2.5 and asthma and acute respiratory infections across all age groups. Pneumonia was associated with PM2.5 among adults only, whereas COPD was not associated with PM2.5 in any age group analyzed. These disparate results could be due to our evaluation of ED visits among all ages versus hospitalization among adults 65 years and older used in most previous studies.

The differences in magnitude across associations of outcomes, age groups, and pollutants are consistent with differences in disease pathology, respiratory tract physiology, pollutant chemistry, exposure patterns, or a combination thereof. The specific respiratory diseases leading to ED visits varies by age group, with ARI, asthma, and pneumonia prominent in children; asthma and COPD in adults; and COPD and pneumonia in older adults (14). The pathology of these diseases and their impact on the respiratory tract also vary by age. Small-airway obstruction related to COPD is often observed in the smallest airways, where gas exchange occurs, compared with a more proximal location among individuals with asthma (37). Similarly, among those with asthma, hyperresponsiveness tends to be more proximally located in younger individuals versus more peripherally located in older individuals (37). Age also influences respiratory tract morphology, breathing patterns, physiochemical properties of the extracellular lining fluid, immunologic responses, and mechanical properties of the lung (1, 2, 37). These differences can influence not only disease pathology but also pollutant uptake, dose, and effect on the respiratory tract. The impact of these differences on our results is further complicated by the differing physiochemical properties of ozone and PM2.5 (1, 2). For example, ozone is a gas that can penetrate deep into the lower respiratory tract, with more distal penetration in larger lungs. Conversely, evidence suggests that children may receive a larger dose of particulate matter in the lower respiratory tract compared with adults. Together, these differences lend biological plausibility to our results.

Limitations

Although our statistical model is well established and has been evaluated extensively (33, 3841), model misspecification is nevertheless a concern. Our county time-series analyses could be biased by an unmeasured or inadequately modeled predictor or confounder if that confounder fluctuates over time in a manner similar to ozone or PM2.5. Our sensitivity analyses, which include modeling time trend and meteorology in various ways and estimating associations with a negative control exposure, suggest that such a bias is unlikely (34). However, our results for ozone and ARI and asthma among children were sensitive to the number of degrees of freedom included in the time control. Our main results are presented using 12 degrees of freedom per year, which is more control than was used in many previous multicity studies. Respiratory ED visits exhibit strong seasonal patterns with sharp within-season spikes due to infectious disease, especially among children (42), which may differ from the outcomes of mortality and respiratory hospitalizations among populations over 65 years of age covered by previous multicity studies. As such, time-series analyses of respiratory ED visits may require more temporal control as evidenced by the amount of control used in many single-city studies of respiratory ED visits (11, 20, 43, 44). As can be seen from the sensitivity analyses, which use fewer degrees of freedom, most RRs are similar but some are substantially lower, highlighting the importance of selecting a smoother that adequately captures the spikes in outcome rates.

Defining respiratory diseases on the basis of primary diagnosis and ICD-9 codes could potentially lead to misclassification; however, such misclassification would not affect the results for all respiratory diseases combined. Although county air pollution metrics are not measures of personal exposure, they are relevant metrics for air pollution policy and can produce robust and unbiased results for ambient ozone and PM2.5 that tend to be more spatially homogeneous than other air pollutants (45). Modeled air pollution data may introduce measurement error into our analyses. Wintertime DS estimates of ozone may rely more on Community Multiscale Air Quality predictions and less on air pollutant measurements, as ozone monitoring occurs predominantly during the warmer months of the year, and year-round monitoring may not be available for colder areas of the United States. As a result, model predictions may have more error in colder months or areas with sparse ozone monitoring. Evaluation of the DS model has shown it to be a reasonable solution to the temporal and spatial gaps in available air pollution data (27, 28). Although our study provides estimates to inform national environmental health policy by combining the local estimates, we average across the local variation in the short-term associations between air pollution and respiratory ED visits that exists, masking any between-county heterogeneity in effect estimates that may exist. Generalizability to the most rural and sparsely populated areas may be limited, as counties with 90% of days without a respiratory ED visit were excluded. Last, although these 17 states analyzed represent 45% of the U.S. population, the northwest and mid-Atlantic states are underrepresented. The national generalizability of our results could be impacted by regional or state differences in factors that alter the association between air pollutants and respiratory ED visits such as air pollution composition, chronic morbidity prevalence, or population behaviors.

Conclusions

Both ozone and PM2.5 were associated with respiratory ED visits among all ages combined, and we observed variation in the magnitude of these associations across age, respiratory outcome, and pollutant. Our results provide the first nationally comprehensive risk estimates for ARI, asthma, COPD, pneumonia, and all respiratory outcomes combined for children, adults younger than 65 years, older adults, and all ages combined. Before our study, U.S. multicity estimates were available only for mortality or for morbidity among older adults, due to limitations in nationally standardized and accessible health data. Such gaps in available data, and consequently our gaps in understanding of the relationship between health and environmental hazards, were the primary motivation for the creation of the CDC Tracking Network. By examining associations with ozone and PM2.5 for people of all ages across hundreds of counties in the United States, we address a key gap in the evidence used to inform national ambient air pollution policy.

Madhusudan Chaganthi (Waltair Technical Services, Inc) for assistance in data management; Barbara Anderson (CDC/Agency of Toxic Substances and Disease Registry) for creation of manuscript figures; Judith Qualters (CDC) and Fuyuen Yip (CDC) for support and guidance; CDC staff that maintains the high-performance computing Linux cluster; Environmental Protection Agency and National Aeronautics and Space Administration staff that compiled the air pollution and meteorology data; and all the staff from the state health departments, tracking programs, and data organizations that provided the emergency department data.

1. U.S. Environmental Protection Agency (EPA). Integrated Science Assessment (ISA) for particulate matter. EPA/600/R-08/139F. Washington, D.C.: U.S. Environmental Protection Agency; 2009.
2. U.S. Environmental Protection Agency (EPA). Integrated Science Assessment (ISA) for ozone and related photochemical oxidants. EPA/600/R-10/076F. Washington, D.C.: U.S. Environmental Protection Agency; 2013.
3. Bell ML, Ebisu K, Peng RD, Walker J, Samet JM, Zeger SL, et al. Seasonal and regional short-term effects of fine particles on hospital admissions in 202 US counties, 1999–2005. Am J Epidemiol 2008;168:13011310.
4. Bell ML, McDermott A, Zeger SL, Samet JM, Dominici F. Ozone and short-term mortality in 95 US urban communities, 1987–2000. JAMA 2004;292:23722378.
5. Dominici F, Peng RD, Bell ML, Pham L, McDermott A, Zeger SL, et al. Fine particulate air pollution and hospital admission for cardiovascular and respiratory diseases. JAMA 2006;295:11271134.
6. Medina-Ramón M, Zanobetti A, Schwartz J. The effect of ozone and PM10 on hospital admissions for pneumonia and chronic obstructive pulmonary disease: a national multicity study. Am J Epidemiol 2006;163:579588.
7. Peng RD, Bell ML, Geyh AS, McDermott A, Zeger SL, Samet JM, et al. Emergency admissions for cardiovascular and respiratory diseases and the chemical composition of fine particle air pollution. Environ Health Perspect 2009;117:957963.
8. Samet JM, Dominici F, Curriero FC, Coursac I, Zeger SL. Fine particulate air pollution and mortality in 20 U.S. cities, 1987–1994. N Engl J Med 2000;343:17421749.
9. Zanobetti A, Schwartz J. Mortality displacement in the association of ozone with mortality: an analysis of 48 cities in the United States. Am J Respir Crit Care Med 2008;177:184189.
10. Mar TF, Koenig JQ. Relationship between visits to emergency departments for asthma and ozone exposure in greater Seattle, Washington. Ann Allergy Asthma Immunol 2009;103:474479.
11. Peel JL, Tolbert PE, Klein M, Metzger KB, Flanders WD, Todd K, et al. Ambient air pollution and respiratory emergency department visits. Epidemiology 2005;16:164174.
12. Slaughter JC, Kim E, Sheppard L, Sullivan JH, Larson TV, Claiborn C. Association between particulate matter and emergency room visits, hospital admissions and mortality in Spokane, Washington. J Expo Anal Environ Epidemiol 2005;15:153159.
13. Tolbert PE, Klein M, Peel JL, Sarnat SE, Sarnat JA. Multipollutant modeling issues in a study of ambient air quality and emergency department visits in Atlanta. J Expo Sci Environ Epidemiol 2007;17:S29S35.
14. Rui P, Kang K. National hospital ambulatory medical care survey: 2014 Emergency department summary tables [accessed 2017 Oct 28]. Available from: http://www.cdc.gov/nchs/data/ahcd/nhamcs_emergency/2014_ed_web_tables.pdf.
15. Alhanti BA, Chang HH, Winquist A, Mulholland JA, Darrow LA, Sarnat SE. Ambient air pollution and emergency department visits for asthma: a multi-city assessment of effect modification by age. J Expo Sci Environ Epidemiol 2016;26:180188.
16. Darrow LA, Hess J, Rogers CA, Tolbert PE, Klein M, Sarnat SE. Ambient pollen concentrations and emergency department visits for asthma and wheeze. J Allergy Clin Immunol 2012;130:630638.e4.
17. Glad JA, Brink LL, Talbott EO, Lee PC, Xu X, Saul M, et al. The relationship of ambient ozone and PM2.5 levels and asthma emergency department visits: possible influence of gender and ethnicity. Arch Environ Occup Health 2012;67:103108.
18. Gleason JA, Bielory L, Fagliano JA. Associations between ozone, PM2.5, and four pollen types on emergency department pediatric asthma events during the warm season in New Jersey: a case-crossover study. Environ Res 2014;132:421429.
19. Strickland MJ, Darrow LA, Klein M, Flanders WD, Sarnat JA, Waller LA, et al. Short-term associations between ambient air pollutants and pediatric asthma emergency department visits. Am J Respir Crit Care Med 2010;182:307316.
20. Winquist A, Klein M, Tolbert P, Flanders WD, Hess J, Sarnat SE. Comparison of emergency department and hospital admissions data for air pollution time-series studies. Environ Health 2012;11:70.
21. Horne BD, Joy EA, Hofmann MG, Gesteland PH, Cannon JB, Lefler JS, et al. Short-term elevation of fine particulate matter air pollution and acute lower respiratory infection. Am J Respir Crit Care Med 2018;198:759766.
22. U.S. Environmental Protection Agency (EPA). Quantitative health risk assessment for particulate matter. EPA-452/R-10-005. Washington, D.C.: U.S. Environmental Protection Agency; 2010.
23. U.S. Environmental Protection Agency (EPA). Health risk and exposure assessment for ozone. EPA-452/R-14-004a. Washington, D.C.: U.S. Environmental Protection Agency; 2014.
24. Pew Environmental Health Commission. America’s environmental health gap: why the country needs a nationwide health tracking network. Baltimore, MD: Johns Hopkins School of Hygiene and Public Health; 2000.
25. Strosnider HCH, Darrow L, Liu Y, Vaidyanathan R, Strickland M. Addressing gaps in age-specific evidence used for United States air pollution policy [abstract]. Presented at the 29th Annual Scientific Conference of the International Society of Environmental Epidemiology. September 24–28, 2017, Sydney, Australia. 2017.
26. Berrocal VJ, Gelfand AE, Holland DM. Space–time data fusion under error in computer model output: an application to modeling air quality. Biometrics 2012;68:837848.
27. Berrocal VJ, Gelfand AE, Holland DM. A spatio-temporal downscaler for output from numerical models. J Agric Biol Environ Stat 2010;15:176197.
28. Berrocal VJ, Gelfand AE, Holland DM. A bivariate space–time downscaler under space and time misalignment. Ann Appl Stat 2010;4:19421975.
29. Ivy D, Mulholland JA, Russell AG. Development of ambient air quality population-weighted metrics for use in time-series health studies. J Air Waste Manag Assoc 2008;58:711720.
30. Vaidyanathan A, Dimmick WF, Kegler SR, Qualters JR. Statistical air quality predictions for public health surveillance: evaluation and generation of county level metrics of PM2.5 for the environmental public health tracking network. Int J Health Geogr 2013;12:12.
31. Mitchell KE, Lohmann D, Houser PR, Wood EF, Schaake JC, Robock A, et al. The multi‐institution North American Land Data Assimilation System (NLDAS): utilizing multiple GCIP products and partners in a continental distributed hydrological modeling system. J Geophys Res 2004;109:D07S90.
32. Fechter-Leggett ED, Vaidyanathan A, Choudhary E. Heat stress illness emergency department visits in National Environmental Public Health Tracking states, 2005–2010. J Community Health 2016;41:5769.
33. Dominici F, Samet JM, Zeger SL. Combining evidence on air pollution and daily mortality from the 20 largest US cities: a hierarchical modeling strategy. J R Stat Soc Ser A Stat Soc 2000;163:263302.
34. Flanders WD, Klein M, Darrow LA, Strickland MJ, Sarnat SE, Sarnat JA, et al. A method to detect residual confounding in spatial and other observational studies. Epidemiology 2011;22:823826.
35. Silverman RA, Ito K. Age-related association of fine particles and ozone with severe acute asthma in New York City. J Allergy Clin Immunol 2010;125:367373.e5.
36. National Asthma Education and Prevention Program, National Heart, Lung, and Blood Institute. Expert Panel Report 3: guidelines for the diagnosis and management of asthma [2007 Aug 28; accessed 2018 Oct 31]. NIH Publication No. 07-4051. Bethesda, MD: National Heart, Lung, and Blood Institute; 2007.
37. Postma DS, Reddel HK, ten Hacken NH, van den Berge M. Asthma and chronic obstructive pulmonary disease: similarities and differences. Clin Chest Med 2014;35:143156.
38. Bell ML, Samet JM, Dominici F. Time-series studies of particulate matter. Annu Rev Public Health 2004;25:247280.
39. Welty LJ, Zeger SL. Are the acute effects of particulate matter on mortality in the National Morbidity, Mortality, and Air Pollution Study the result of inadequate control for weather and season? A sensitivity analysis using flexible distributed lag models. Am J Epidemiol 2005;162:8088.
40. Zeger SL, Irizarry R, Peng RD. On time series analysis of public health and biomedical data. Annu Rev Public Health 2006;27:5779.
41. Zeger SL, Thomas D, Dominici F, Samet JM, Schwartz J, Dockery D, et al. Exposure measurement error in time-series studies of air pollution: concepts and consequences. Environ Health Perspect 2000;108:419426.
42. Gerhardsson de Verdier M, Gustafson P, McCrae C, Edsbäcker S, Johnston N. Seasonal and geographic variations in the incidence of asthma exacerbations in the United States. J Asthma 2017;54:818824.
43. Darrow LA, Klein M, Flanders WD, Mulholland JA, Tolbert PE, Strickland MJ. Air pollution and acute respiratory infections among children 0–4 years of age: an 18-year time-series study. Am J Epidemiol 2014;180:968977.
44. Sarnat SE, Winquist A, Schauer JJ, Turner JR, Sarnat JA. Fine particulate matter components and emergency department visits for cardiovascular and respiratory diseases in the St. Louis, Missouri–Illinois, metropolitan area. Environ Health Perspect 2015;123:437444.
45. Strickland MJ, Gass KM, Goldman GT, Mulholland JA. Effects of ambient air pollution measurement error on health effect estimates in time-series studies: a simulation-based analysis. J Expo Sci Environ Epidemiol 2015;25:160166.
Correspondence and requests for reprints should be addressed to Heather M. Strosnider, Ph.D., M.P.H., 4770 Buford Highway, MS F60, Atlanta, GA 30341. E-mail: .

Supported by the Centers for Disease Control and Prevention (CDC), National Center for Environmental Health, Division of Environmental Hazards and Health Effects, Environmental Health Tracking Branch. The CDC supported the staff responsible for the design and conduct of the study; in the collection, analysis, and interpretation of the data; in the preparation, review, and approval of the manuscript; and in the decision to submit the manuscript for publication. This work of Y.L. was partially supported by the NASA Applied Sciences Program (grant NNX16AQ28G).

Disclaimer: The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the CDC.

Author Contributions: M.J.S. and H.M.S. conceived of and designed the study; H.M.S. and A.V. acquired and processed all data; H.M.S. conducted analysis and drafted the manuscript; H.H.C. provided statistical expertise; M.J.S., H.M.S., H.H.C., L.A.D., Y.L., and A.V. interpreted the results and provided critical input into the manuscript; all authors reviewed and approved the final manuscript.

This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.

Originally Published in Press as DOI: 10.1164/rccm.201806-1147OC on October 2, 2018

Author disclosures are available with the text of this article at www.atsjournals.org.

Comments Post a Comment




New User Registration

Not Yet Registered?
Benefits of Registration Include:
 •  A Unique User Profile that will allow you to manage your current subscriptions (including online access)
 •  The ability to create favorites lists down to the article level
 •  The ability to customize email alerts to receive specific notifications about the topics you care most about and special offers
American Journal of Respiratory and Critical Care Medicine
199
7

Click to see any corrections or updates and to confirm this is the authentic version of record