Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Jun 26.
Published in final edited form as: Respir Med. 2013 Aug 28;107(11):1763–1772. doi: 10.1016/j.rmed.2013.08.010

Nitrogen Dioxide and Allergic Sensitization in the 2005–2006 National Health and Nutrition Examination Survey

Charles H Weir 1, Karin B Yeatts 2, Jeremy A Sarnat 3, William Vizuete 1, Päivi M Salo 4, Renee Jaramillo 5, Richard D Cohn 5, Haitao Chu 6, Darryl C Zeldin 4, Stephanie J London 7
PMCID: PMC4071349  NIHMSID: NIHMS585225  PMID: 24045117

Abstract

Background

Allergic sensitization is a risk factor for asthma and allergic diseases. The relationship between ambient air pollution and allergic sensitization is unclear.

Objective

To investigate the relationship between ambient air pollution and allergic sensitization in a nationally representative sample of the US population.

Methods

We linked annual average concentrations of nitrogen dioxide (NO2), particulate matter ≤ 10 µm (PM10), particulate matter ≤ 2.5 µm (PM25), and summer concentrations of ozone (O3), to allergen-specific immunoglobulin E (IgE) data for participants in the 2005–2006 National Health and Nutrition Examination Survey (NHANES). In addition to the monitor-based air pollution estimates, we used the Community Multiscale Air Quality (CMAQ) model to increase the representation of rural participants in our sample. Logistic regression with population-based sampling weights was used to calculate adjusted prevalence odds ratios per 10 ppb increase in O3 and NO2, per 10 µg/m3 increase in PM10, and per 5 µg/m3 increase in PM2.5 adjusting for race, gender, age, socioeconomic status, smoking, and urban/rural status.

Results

Using CMAQ data, increased levels of NO2 were associated with positive IgE to any (OR 1.15, 95% CI 1.04, 1.27), inhalant (OR 1.17, 95% CI 1.02, 1.33), and outdoor (OR 1.16, 95% CI 1.03, 1.31) allergens. Higher PM2.5 levels were associated with positivity to indoor allergen-specific IgE (OR 1.24, 95% CI 1.13, 1.36). Effect estimates were similar using monitored data.

Conclusions

Increased ambient NO2 was consistently associated with increased prevalence of allergic sensitization.

Keywords: air pollution, allergic, sensitization, epidemiology, NHANES, IgE

INTRODUCTION

Both particulate and gaseous air pollutants have been hypothesized to play a role in the development and exacerbation of allergic diseases 1. Allergic or atopic sensitization is a strong risk factor for childhood and adult asthma and is characterized by increased immunoglobulin E (IgE) production to specific antigens that can be detected by measurements in blood 2, 3.

The evidence for a link between air pollution and allergic sensitization is inconsistent. Experimental studies provide a biologic basis for gaseous and particulate air pollutants as risk factors for allergic sensitization by showing enhanced IgE production after exposure to NO2, O3, and particulates 4, 5, 6, 1c. However, results from epidemiologic studies are equivocal. Positive associations between traffic-related air pollution and allergic sensitization were reported in two birth cohort studies in Germany and Sweden 7, 8. Nine cross-sectional studies also found positive associations between ambient air pollution and allergic sensitization.9, 10, 11, 12, 13, 14, 15, 16

In contrast, four prospective birth cohort studies conducted in Europe did not find associations between air pollution and allergic sensitization 17, 18,19, 20. Positive associations in the study by Brauer17 were limited to sensitization to food allergens and not inhalant allergens. Several cross sectional studies also did not find associations between ambient air pollution and allergic sensitization 21, 22, 23, 24. To date, most epidemiologic studies of air pollution and allergic sensitization have been conducted in Europe and have focused on air pollution from traffic sources. Diesel emissions represent the largest source of particulate matter from motor vehicles and have been hypothesized to be an adjuvant for allergic sensitization 25. Diesel vehicles are a much larger percentage of the vehicle fleet in Europe than the US 26. Recent studies of air pollution and asthma or allergies using nationally representative samples of the US population did not assess allergic sensitization. In addition, these studies relied on monitoring data alone, and as a result, have focused on study subjects mostly in major metropolitan areas 27,28. No population-based studies of air pollution and allergic sensitization representative of the US population have been conducted.

The National Health and Nutrition Examination Survey (NHANES) is a nationally representative survey of adults and children in the United States. The 2005–2006 NHANES survey included measurements of allergen-specific IgE. We linked monitored and modeled air pollution concentrations to the NHANES 2005–2006 dataset to investigate the relationship between ambient air pollution and allergic sensitization. By using an air quality model to assign exposures, we were able to increase the sample size for the investigation by including participants that did not live near air pollution monitors resulting in a sample more representative of the US population.

METHODS

We analyzed data from the NHANES 2005–2006 database. The 2005–2006 survey oversampled Mexican Americans, African Americans, ages 60 and older, adolescents 12– 19, and persons with low income to increase the reliability and precision of health status indicator estimates for these groups 29,30, 31. Our analysis was reviewed and approved by the University of North Carolina Chapel Hill Institutional Review Board. The study participants gave informed consent when they agreed to participate in the NHANES study.

Population and Study Sample

The 2005–2006 NHANES included 10,348 participants. We limited our analysis to participants ages 6 and older that were examined in the mobile exam center (MEC) (n=8086). Among the 8086 participants, 7268 had complete data for all 19 specific IgEs, 686 had no IgE data, and 132 were missing 1 or more specific IgEs. We further limited eligibility to 6917 persons with no missing values for any of the covariates used in our analysis.

Air Pollution Exposure Assignment

At the request of the National Center for Health Statistics (NCHS), the US Department of Housing and Urban Development geocoded the 2005–2006 NHANES (CDC, 2009). The NCHS linked US Environmental Protection Agency Air Quality System (AQS) monitored data and Community Multiscale Air Quality (CMAQ) model data to the 2005–2006 NHANES data by geocoded participant address 29,32. Because the data contains identifiable geographic information, it is not available for public use. We submitted a proposal to NCHS that specified our analysis plan and the variables we required from the public NHANES data file. NCHS approved our proposal and created a data set with AQS and CMAQ data linked to the NHANES data file.

We used AQS monitored data for particulate matter with aerodynamic diameter ≤ 2.5 µm (PM25) and ≤ 10 µm (PM10), ozone (O3), and nitrogen dioxide (NO2) to assign exposure estimates to participants within 20 miles (32.2 km) of a monitor. We selected annual calendar year estimates for NO2, PM25, and PM10. Annual calendar year estimates were the only long term exposure option for monitored data available in NCHS data files. For example, if a participant came into the MEC on June 1, 2005, then the participant received an estimate based on concentrations averaged from 1 Jan 2005 - 31 Dec 2005.

O3 is monitored at different times throughout the year in different locations. Average 8 hour daily maximum concentrations were calculated from 1 May through 30 September since O3 is monitored in most locations during this period. For monitored pollutants, inverse distance weighted estimates were calculated using the inverse of the squared distance between the participant residence and monitors within 20 miles (32.2 km) of the residence. Since we included only participants within 20 miles (32.2 km) of a monitor, the sample size differs by pollutant because the location of the monitors varies by pollutant based on regulatory requirements. Of the 6917 participants with a complete panel of allergen-specific IgE and covariates, the number of participants with monitored estimates was 4331 for NO2, 4492 for PM10, 5201 for O3, and 5298 for PM2.5.

In addition to monitored data, we obtained CMAQ model estimates available from the EPA National Exposure Research Lab Atmospheric Modeling and Analysis Division. Estimates were available for PM2.5, O3 and NO2, but not for PM10. CMAQ is often used by state air pollution control agencies to assess how proposed air quality management changes might impact air pollution concentrations 33, 34. CMAQ generates pollutant estimates by simulating the chemistry and physics of the atmosphere using air pollution emissions and meteorological data as inputs.

CMAQ output consisted of hourly surface concentrations for each day of calendar years 2004–2006 for the continental US at a resolution of 36×36 kilometers. Using CMAQ output, NCHS calculated averages for one year prior to the participant medical exam date. Participants received exams throughout the calendar year. By using CMAQ, we increased the number of participants with air pollution estimates in our study sample to 6227 for PM2.5, NO2, and O3. Participants had missing air pollution concentration data because they did not live within 20 miles (32.2 km) of a monitor, lived outside the domain of the model, or they did not have sufficient address information for data linkage.

Allergic Sensitization

Survey participants ages 6 and older were tested for each of 19 allergen-specific IgE antibodies using the Pharmacia Diagnostics ImmunoCAP 1000 System. The panel included IgE to 15 aeroallergens (Alternaria alternata, Aspergillus fumigates, Bermuda grass, birch, cat dander, cockroach, dog dander, dust mite [Dermatophagoides farinae and Dermato-phagoides pteronyssinus], mouse urine proteins, oak, ragweed, rat urine proteins, Russian thistle, rye grass, and 4 food allergens (egg white, cow’s milk, peanut, and shrimp). The lower limit of detection was 0.35 kU/L for each specific IgE. For samples below the detection limit, NHANES reported values equal to the lower limit of detection divided by the square root of 2. The upper limit of detection was 1000 kU/L. Samples that exceeded the upper limit of detection were assigned a value of 1000 kU/L 35.

Variable Definitions

Sensitization was defined as detectable specific IgE (≥0.35 kU/L). We investigated five allergic sensitization outcome variables 36. These included: 1) any of the IgE antibodies; 2) outdoor allergen-specific IgEs (Alternaria alternata, Aspergillus fumigatus, Bermuda grass, birch, oak, ragweed, Russian thistle, rye grass); 3) indoor allergen-specific IgEs [cat dander, cockroach, dog dander, dust mite (Dermatophagoides farinae and Dermatophagoides pteronyssinus), mouse proteins, rat urine proteins]; 4) inhalant (indoor or outdoor allergen-specific IgEs); and 5) food allergen-specific IgEs (egg white, cow’s milk, peanut, shrimp). These five outcomes are not mutually exclusive.

We considered several covariates in our analyses. We obtained data for age, race/ethnicity, gender, and poverty income ratio based on participant responses in the survey questionnaire. Cotinine, a biomarker for smoking and secondhand smoke exposure, was obtained from the medical exam 30. We used a dichotomous cotinine variable with a cut point of 10 ng/ml to distinguish smokers from non-smokers 37. We used the NCHS 2005–2006 urban-rural classification scheme to characterize the degree of urbanization where a participant resided 38. The scheme consists of four metropolitan categories and two non-metropolitan categories. We recoded the six-category NCHS urban-rural variable into five categories to preserve participant confidentiality and eliminate small cell sizes by combining the small and medium metropolitan categories into one category (Table 1). We used poverty income ratio as a surrogate for socioeconomic status. Race/ethnicity was categorized into non-Hispanic white, non-Hispanic black, Hispanic, and other.

Table 1.

Characteristics of the total sample and subsamples with pollutant dataa

Total Monitored Datab CMAQ Datac
n=6917 NO2
n=4331		 O3
n=5201		 PM10
n=4492		 PM2.5
n=5298		 n=6227
Race/Ethnicity (%)
Non Hispanic White 70.4 60.9 65.9 60.9 66.7 68.9
Non Hispanic Black 11.7 14.4 12.9 14.7 12.5 13.0
Mexican American 8.7 13.4 11.0 12.8 10.2 9.1
Other 9.2 11.3 10.2 11.6 10.5 9.1
Age (%)
6 – 17 17.2 17.4 17.4 17.3 17.5 17.0
≥18 82.8 82.6 82.6 82.7 82.5 83.0
Cotinine (%)
< 10 ng/ml 75.8 76.8 76.8 77.4 77.3 75.7
≥ 10 ng/ml 24.2 23.2 23.2 22.6 22.7 24.3
Gender
% Female 51.4 52.3 51.8 52.3 51.8 51.5
Urbanicity (%)
Large Metropolitan 31.3 53.8 42.9 52.7 41.0 34.9
Large Fringe
Metropolitan 18.6 28.5 24.5 25.8 22.2 19.0
Small and
Medium Metro 28.7 17.7 21.2 21.7 26.0 23.8
Micropolitan 15.5 0 10.4 0 9.8 15.8
Noncore 6.0 0 0.9 0 0.9 6.5
PIR (mean,
Std deviation) 2.5(0.02) 2.5(0.02) 2.5(0.02) 2.5(0.02) 2.5(0.02 ) 2.5(0.02)
Open in a new tab

Abbreviation: PIR=poverty income ratio, CMAQ=Community Multiscale Air Quality Model

a

All percentages were weighted using NHANES survey weights

b

Subsamples for monitored NO2, O3, PM2.5, and PM10 linked to NHANES participants

c

Subsamples for monitored NO2, O3, PM2.5, and PM10 from the CMAQ model linked to NHANES participants

Statistical Analyses

We calculated descriptive statistics for NHANES participants both with and without air pollution estimates (Table 2). Descriptive statistics were generated using SAS version 9.2. We used NHANES analytic and reporting guidelines to select the appropriate sub-sample weights (wtmec2yr) and design variables for our analysis all analyses except for descriptive statistics of pollutant concentrations since all participants received the medical exam 30.

Table 2.

Weighted prevalence of sensitization for participants with and without air pollution dataa

Monitored Datab CMAQc
Total
n=6917		 NO2
With
n=4331		 Without
n=2586		 O3
With
n=5201		 Without
n=1716		 PM2.5
With
n=5298		 Without
n=1619		 PM10
With
n=4492		 Without
n=2425		 With
n=6227		 Without
n=690
Sensitization (%)
Anyd 44.8 48.0 41.1 46.2 41.6 45.6 42.7 47.6 41.2 45.3 41.6
Inhalante 42.7 45.9 38.9 43.9 39.8 43.5 40.6 45.5 38.9* 43.1 39.6
Outdoorf 30.1 34.4 24.9 32.0 25.7 31.8 25.7 34.4* 24.5* 30.7 26.2
Indoorg 30.4 31.1 29.6 30.6 30.0 30.3 30.7 30.7 30.0 30.6 29.3
Foodh 6.5 5.8 7.3 6.1 7.4 5.9 8.2 5.5 7.8* 6.4 7.4
Open in a new tab

Abbreviations: CMAQ=Community Multiscale Air Quality Model

*

Prevalence differences between with and without are significant at p<.05

a

All percentages were weighted using NHANES survey weights

b

Subsamples for monitored PM2.5, O3, NO2, and PM10 linked to NHANES participants

c

Subsamples for PM2.5, O3, and NO2 linked to NHANES participants based on CMAQ data

d

Any=Detectable specific IgE to indoor, outdoor, or food allergens

e

Inhalant= Detectable specific IgE to indoor or outdoor allergens

f

Outdoor= Detectable specific IgE to Alternaria alternata, Asperigillus fumigatus, Bermuda grass, birch, oak, ragweed, Russian thistle, or rye grass

g

Indoor= Detectable specific IgE to cat dander, cockroach, dog dander, dust mite (Dermataphagoides farinae and Dermatophagoides pteronyssinus), mouse proteins, rat urine proteins

h

Food= Detectable specific IgE to egg white, milk, peanut or shrimp

We calculated crude and adjusted prevalence odds ratios using the SUDAAN R Logistic procedure release 10.0.1 to account for the clustering and stratification in the sample design. We used logistic regression to produce separate odds ratios for 1) monitored air pollution concentrations, 2) CMAQ estimates for participants with monitored and CMAQ estimates, and 3) all participants with CMAQ estimates. Odds ratios are scaled per 10 parts per billion for NO2 and O3, per 10 µg/m3 for PM10 and per 5 µg/m3 for PM2.5. We chose scaling factors to have consistency between our modeled air pollution data and our monitored air pollution data. Based on the existing literature we selected age, gender, race/ethnicity, poverty income ratio, cotinine, and level of urbanization as covariates. We included poverty income ratio as a continuous variable and gender, race/ethnicity, age, cotinine, and level of urbanization were included as categorical variables. We also adjusted for indoor air exposures of mold, housing type, and pets; they made no difference in effect estimates, so they were not included in the final models. We conducted interaction testing for age and gender using p value < 0.10 as a criterion for positive interaction.

RESULTS

The total study sample (n=6917) was more white and less urban than the sub-samples created from linking AQS air pollution estimates to NHANES participants (Table 1). As expected, given that CMAQ data are available for subjects living in rural areas far from monitors, the subsample created from linking CMAQ estimates was more similar to the overall sample than the subsamples created from linking monitored data, thus more representative of the overall US population.

Table 2 shows the frequency of sensitization among the total sample and the sub-samples. With the exception of food allergen sensitization, the percentage of sensitization was lower in the subset of participants without linked air pollution estimates compared with the subsample of participants with air pollution estimates. Most of the participants without air pollution estimates live in rural areas. In our sample, the percentage of sensitization is lower in rural areas compared to urban areas (prevalence of sensitization to any allergen = 44.7% for rural subjects and 49.4% for urban subjects). Sensitization was also lower for all subtypes of allergens except food allergens for rural versus urban subjects (data not shown).

Descriptive air pollution statistics are shown in Tables 3 and 4. For participants that had both modeled and monitored estimates, modeled estimates of O3 and PM2.5 based on the year prior to the participant medical exam date were higher than the inverse distance weighted monitored calendar year estimates on average. Model estimates for NO2 were lower than inverse distance weighted monitored estimates. NO2 and PM10 were most strongly correlated (r=0.48) among monitored pollutants. In contrast, NO2 was most strongly correlated with PM2.5 (r=0.60) among modeled pollutants.

Table 3.

Monitored and CMAQ Pollutant Concentrations

Monitored CMAQ
NO2
(ppb)		 O3
(ppb)		 PM2.5
(ug/m3)		 PM10
(ug/m3)		 NO2
(ppb)		 O3
(ppb)		 PM2.5
(ug/m3)
Measure
5th Percentile 11.4 37.5 8.7 19.4 2.0 45.6 6.9
10th Percentile 11.7 40.4 9.5 20.6 2.5 47.6 7.8
25th Percentile 13.2 48.1 11.4 23.6 3.7 52.4 9.5
Median 17.6 51.5 12.7 27.1 10.6 57.0 13.4
Mean 18.6 52.0 12.7 28 11.6 57.2 12.6
75th Percentile 24.3 55.3 13.9 30.9 15.3 61.2 15.1
95th Percentile 27.0 60.3 16.5 44.1 27.6 70.8 20.0
Open in a new tab

CMAQ PM10 air pollution concentrations were not available in our data set

Abbreviations: CMAQ=Community Multiscale Air Quality Model

Table 4.

Pearson Correlations of Monitored and CMAQ Pollutant Concentrations

Monitored CMAQ
NO2
(ppb)		 O3
(ppb)		 PM2.5
(ug/m3)		 PM10
(ug/m3)		 NO2
(ppb)		 O3
(ppb)		 PM2.5
(ug/m3)
Correlation
NO2 1 −0.25 −0.01 0.48 0.57 −0.38 −0.1
O3 −0.25 1 0.08 0.4 −0.22 0.66 −0.29
PM2.5 −0.01 0.08 1 0.11 0.45 −0.21 0.57
PM10 0.48 0.4 0.11 1 0.32 0.09 −0.2
CMAQ NO2 0.57 −0.22 0.45 0.32 1 −0.42 0.48
CMAQ O3 −0.01 0.66 −0.09 0.09 −0.42 1 −0.21
CMAQ PM2.5 −0.1 −0.29 0.57 −0.2 0.48 −0.21 1
Open in a new tab

Abbreviations: CMAQ=Community Multiscale Air Quality Model

Table 5 displays adjusted prevalence odds ratios and 95% confidence intervals for each air pollutant in relation to each category of allergen-specific IgE based on the following: 1) monitored data, 2) CMAQ data among participants with monitored data and CMAQ data, and 3) the larger sample of all participants with CMAQ estimates. A similar table including crude and adjusted odds ratios is provided as supplemental material (Table S1). The largest percent change in crude odds ratios with adjustment for potential confounders was from the addition of urbanicity and ethnicity. Using an alternative categorization of the current smoking (with 3 categories of cotinine and cut points (< 0.015 ng/ml and < 0.050 ng/ml)) produced similar results.

Table 5.

Adjusted odds ratios and 95% Cls between air pollution concentrations and allergen-specific IgE for ages ≥ 6a,b

Results Based
on Monitored Data		 CMAQ Results for
Subjects with Monitored Data		 CMAQ Results Including
Subjects without Monitored Data
Sensitization
Pollutant n OR (95% CI) n OR (95% CI) n OR (95% CI)
Anyc
NO2 4331 1.24 (1.07, 1.44) 4331 1.13 (1.02, 1.26) 6277 1.15 (1.04, 1.27)
O3 5201 1.07 (0.94, 1.21) 5151 1.09 (0.91, 1.31) 6277 1.10 (0.93, 1.29)
PM2.5 5298 1.13 (0.94, 1.36) 5208 1.02 (0.90, 1.17) 6277 1.04 (0.93, 1.17)
PM10 4492 1.08 (0.95, 1.25)
Inhalantd
NO2 4331 1.23 (1.04, 1.46) 4331 1.15 (0.99, 1.34) 6277 1.17 (1.02, 1.33)
O3 5201 1.06 (0.93, 1.20) 5151 1.09 (0.90, 1.32) 6277 1.11 (0.93, 1.32)
PM2.5 5298 1.11 (0.93, 1.32) 5208 1.05 (0.88, 1.24) 6277 1.07 (0.93, 1.24)
PM10 4492 1.07 (0.93, 1.23)
Outdoore
NO2 4331 1.24 (1.00, 1.55) 4331 1.03 (0.83, 1.29) 6277 1.10 (0.89, 1.35)
O3 5201 1.17 (0.99, 1.38) 5151 1.12 (0.88, 1.42) 6277 1.14 (0.90, 1.43)
PM2.5 5298 0.84 (0.62, 1.15) 5208 0.88 (0.68, 1.13) 6277 0.93 (0.74, 1.17)
PM10 4492 1.18 (0.98, 1.42)
Indoorf
NO2 4331 1.14 (0.97, 1.35) 4331 1.20 (1.09, 1.33) 6277 1.16 (1.03, 1.31)
O3 5201 0.91 (0.78, 1.06) 5151 1.03 (0.86, 1.23) 6277 1.02 (0.86, 1.22)
PM2.5 5298 1.27 (1.12, 1.45) 5208 1.26 (1.16, 1.38) 6277 1.24 (1.13, 1.36)
PM10 4492 0.93 (0.85, 1.03)
Foodg
NO2 4331 1.10 (0.77, 1.55) 4331 1.18 (0.91, 1.55) 6277 1.08 (0.82, 1.44)
O3 5201 0.80 (0.54, 1.19) 5151 1.07 (0.76, 1.51) 6277 1.01 (0.77, 1.32)
PM2.5 5298 1.27 (0.78, 2.08) 5208 1.22 (0.89, 1.69) 6277 1.09 (0.83, 1.44)
PM10 4492 0.78 (0.53, 1.14)
Open in a new tab

Abbreviations: OR – odds ratio, CI – confidence interval, CMAQ – Community Multiscale Air Quality Model

a

Odds ratios are per 10 ppb for NO2 and O3, per 5 µg/m for PM2.5 and 10 µg/m for PM10

b

Adjusted for age, ethnicity, poverty income ratio, gender, cotinine, urban/rural status

c

Any= Detectable specific IgE to indoor, outdoor, or food allergens

d

Inhalant= Detectable specific IgE to indoor or outdoor allergens

e

Outdoor= Detectable specific IgE to Alternaria alternata, Aspergillus fumigatus, Bermuda grass, birch, oak, ragweed, Russian thistle, or rye grass

f

Indoor= Detectable specific IgE to cat dander, cockroach, dog dander, dust mite (Dermatophagoides farina and Dematophagoides pteronyssinus), mouse proteins, rat urine proteins

g

Food = Detectable specific IgE to egg white, milk, peanut, or shrimp

The results were similar among the three analyses, but a greater number of significant associations were detected using modeled estimates than monitored estimates which might reflect the larger sample size for this analysis. The most frequent associations were observed for NO2 with most adjusted odds ratios near 1.2. After adjustment for confounders, the only significant association identified using modeled data that was not identified using monitored data was for NO2 in relation to indoor allergen-specific IgE. Similar effect estimates from CMAQ and monitored data of the same participants provide some confidence that odds ratios produced from our larger CMAQ sample (n=6277) that includes subjects without monitoring data provide reasonable effect estimates in the absence of monitored data. Testing for interaction for age or gender indicated very little evidence of effect modification. Age-stratified analyses with interaction P values are provided as supplemental material (Table S2 and Figure S1. Forest Plot of Age-stratified Analyses). For gender, the relationship between PM2.5 and outdoor air pollution was the only relationship with an interaction p value < 0.10.

DISCUSSION

We found associations between increased NO2 and PM2.5 concentrations and allergic sensitization in the US population. NO2 exposure was significantly associated with three allergic sensitization categories using CMAQ data. Overall, we found similar results using monitored data but with fewer statistically significant results in this smaller subset of the data. PM2.5 was consistently associated with sensitization to indoor allergens. This is the first population-based study of air pollution and allergic sensitization that used a nationally representative sample of the US population.

Most previous studies that have identified associations between ambient air pollution and allergic sensitization were studies of traffic-related air pollution in Europe. In contrast, our study was not designed to specifically assess traffic-related air pollution since our exposure metrics do not differentiate near roadway exposures. Additionally, our sample contains both children and adults, whereas most other studies have assessed either children or adults.

Our findings are plausible based on several recent mechanistic studies in mice that provide support for NO2 exposure as a contributor to the development of allergic sensitization. Ckless 39 provided evidence that NO2 contributes to allergic sensitization as an exogenous reactive nitrative species and contributes to the production of endogenous reactive oxidative and reactive nitrative species 40,39.

Consistent with several recent studies, our most frequent associations involve NO2. In a Swedish birth cohort, Nordling8 found an association between traffic-related NO2 and sensitization to pollens (OR =1.67 95% CI 1.10, 2.53 per 44 µg/m3, n=2543) at age 4 years. Similarly Kramer 11 identified an association (OR=4.96 95% CI 1.56, 15.74 per 10 µg/m3) between ambient NO2 and sensitization to pollens for children 9 years of age residing in urban areas. This cross-sectional study of 317 German children lost significance (OR=1.05 95% CI 0.70, 1.56) when urban and suburban children were analyzed together. In addition, a cross sectional study by Janssen 10 also found a positive association between NO2 and sensitization to inhalant allergens (OR=1.70 95% CI 1.03, 2.81 per 17.6 µg/m3) among 1114 Dutch children 7–12 years of age. Overall our effect estimates are generally smaller than reported in these studies. This may be in part because our exposure assessment approach could not resolve within city exposure contrasts or near roadway exposures. Possible reasons for the differences in association are that our sample included children and adults, was larger than the samples of the studies that reported positive associations, and used a different scaling factor. Also, the allergens included in the definition of sensitization are not consistent across studies. In contrast to our findings, several epidemiologic studies did not find positive associations between NO2 and sensitization to inhalant allergens. Three of these were birth cohort studies 17,18,20 while six were cross sectional studies 9, 22, 23, 19, 41, 24.

Across studies, there are differences in methods of exposure assessment, differences between the interpretation of skin tests and laboratory variability in assays of specific IgE to assess allergic sensitization, as well as differences in ambient pollutant levels that may all contribute to variation in the associations observed 42, 43. The combination of these factors makes comparisons difficult. For studies where an association was detected, no one pollutant appeared to be most frequently associated with allergic sensitization. This observation raises a question regarding whether our findings for NO2 represents a pollutant specific finding or if NO2 is a surrogate for traffic-related pollutants.

Two previous studies reported positive associations between PM2.5 and allergic sensitization. A cohort study by Morgenstern 7 found an association between PM2.5 and sensitization to inhalant allergens (OR=1.45 95% CI 1.21, 1.74 per 1.5 µg/m3) but was largely driven by sensitization to outdoor allergens. A cross sectional study by Annesi-Maesano 9 found a positive association between PM2.5 and sensitization to indoor allergens (OR=1.29 95% CI 1.11, 1.50 for high versus low pollutant exposure. Low pollutant exposure ranged from 1.6–12.2 µg/m3. Our finding of an association of PM2.5 with IgE of indoor allergens (OR=1.24 95% CI 1.13, 1.36 per 5 µg/m3) was similar in magnitude to Annesi-Maesano 9. Our findings were driven largely by dust mite (data not shown), which is the most common antigen to which subjects in this category are sensitized 44. Other studies identified included four studies that were not consistent with our results for PM2.5: two birth cohort studies 17,18 and two cross sectional studies 21,10.

Our study has several strengths. The NHANES study population is representative of the entire US population. We believe this is particularly important since most studies of air pollution and allergic sensitization have been conducted in Europe, which may have different pollutant mixtures and allergen species. In addition, the study is relatively large and the assessment of sensitization is comprehensive, based on 19 specific allergen IgEs. We also included data on a number of potential confounders, including cotinine to objectively assess smoking and exposure to environmental tobacco smoke.

Another important strength of the current analysis is that we used an air quality model as an alternate method of assigning air pollution exposures to increase inclusion of participants living outside of major metropolitan areas. We found consistent associations using both monitored and modeled air pollution estimates. To our knowledge, this is the first time that both monitoring and air quality modeling exposure assignment methods have been used in an epidemiologic study to assess the US population. Using an air quality model provides air pollution estimates that capture the nonlinear atmospheric chemistry and physics of the atmosphere that linear interpolation methods cannot 33. Finally, the general concordance of the results using both exposure assignment approaches adds strength to the validity of our findings.

The study has limitations. We adjusted for a number of potential confounders, but we cannot completely rule out unmeasured factors that might be spatially associated with air pollution that biased our effect estimates. Arbes 45 found that the prevalence of atopy differed by census region within the US. We were not able to conduct geographic level stratified analyses with our data. Our primary method of estimating air pollution concentrations relied on US EPA criteria pollutant monitoring. We did not have information on distance to roadway or traffic density to estimate near roadway exposures. The monitoring network has limited coverage of rural areas. However, we used CMAQ to increase the spatial coverage of air pollutant estimates for rural participants. Although we increased the number of rural participants CMAQ has limitations inherent to simulating air pollution concentrations. Meteorological data, emissions data, and the chemical and physical processes that CMAQ is simulating all introduce uncertainty into estimates of air pollution concentration 33.

Both of our exposure metrics are relatively coarse. We limited our investigation to participants within 20 miles (32.2 km) of a monitor and used a 36 kilometer grid to generate modeled estimates of ambient concentration. Because NHANES is a national sample, we were primarily concerned with exposure contrasts between areas and not within an area. Despite our exposure assignment approach being limited by not being able to capture within area exposure contrasts, we still detected positive associations between air pollution and allergic sensitization. Since our study was aimed at looking at differences between areas, and not within an area, we believe that our estimates of air pollution concentration are suitable for estimating associations under these conditions. Two key factors in how well ambient monitors estimate personal exposure are how close participants are to monitors and how homogeneous the pollutant concentrations are in space. The degree of pollutant spatial homogeneity varies across the study areas selected by NHANES based on the inventory of sources, topography, type of pollutant, atmospheric conditions, locations of monitors relative to study participants, model performance, and size of the area 46. Ambient concentrations of O3 and PM2.5 are relatively homogeneous over short distances compared to NO2. NO2 concentrations vary more over short distances as a result of traffic sources. Fourteen of the seventeen studies epidemiologic studies of air pollution and allergic sensitization we referenced estimated exposures from traffic or captured variability in air pollution concentration within an urban area. Since our study cannot, we may miss areas of highest concentration within an urban area that may have attenuated our effect estimates.

We chose to base our estimate of monitored pollutant levels on monitors within 20 miles (32.2 km), based on the work of Parker 47 who linked air pollution estimates for NHIS participants based on an average of 1) all monitors within the county, 2) monitors within a 5 mile radius of the participant census block group, and 3) monitors within 20 miles (32.2 km) of the participant census block group. Parker 47 suggested that these methods gave similar association results but have tradeoffs. Linking air pollution estimates to national survey data sets with finer spatial resolution reduces measurement error but also reduces sample size. On the other hand using air pollution estimates with coarser spatial resolution increases the likelihood of measurement error, increases sample size, and reduces the potential for selection bias.

In summary, our study suggests that ambient air pollution is associated with allergic sensitization. Our main finding of an association with NO2 and allergic sensitization is seen for both monitored and modeled data and across several categories of allergen-specific IgE. Our study is the first to assess the relationship between air pollution and allergic sensitization in a nationally representative sample of the US population.

Supplementary Material

supp1

Acknowledgements

We especially thank Stephanie Robinson, Jennifer Parker, Nataliya Kravets, Ajay Yesupriya, and Carolyn Neal at the National Center for Health Statistics for their help throughout the project. We also thank Ambarish Vaidyanathan for his assistance with preliminary exploration of Community Multiscale Air Quality Modeling data.

Data files were created by Nataliya Kravets, at the National Center for Health Statistics Research Data Center. Community Multiscale Air Quality Model data was provided by Alfreida Torian, US EPA Office of Research and Development Atmospheric Modeling and Analysis Division.

This research was supported in part by Division of Health Studies Agency for Toxic Substances and Disease Registry, and the Division of Intramural Research, National Institute of Environmental Health Sciences, NIH, DHHS. The findings and conclusions in this article are those of the authors and do not represent the views of the National Center for Environmental Health/Agency for Toxic Substances and Disease Registry or National Institute of Environmental Health Sciences.

Abbreviations

AQS

Air Quality System

CI

confidence interval

CMAQ

Community Multiscale Air Quality

IgE

Immunogloblin E

MEC

mobile examination center

NCHS

National Center for Health Statistics

NO2

nitrogen dioxide

NHANES

National Health and Nutrition Examination Survey

PM2.5

Particulate matter with aerodynamic diameter ≤ 2.5 µm

PM10

Particulate matter with aerodynamic diameter ≤ 10 µm

Footnotes

The authors declare they have no actual or potential competing financial interests. Renee Jaramillo and Richard Cohn are employed by SRA International, INC, Durham, NC.

REFERENCES

  • 1.(a) Bjorksten B. Environmental risk factors for atopy. Clin Rev Allergy Immunol. 1997;15(2):125–143. doi: 10.1007/BF02826583. [DOI] [PubMed] [Google Scholar]; (b) Parnia S, Brown JL, Frew AJ. The role of pollutants in allergic sensitization and the development of asthma. Allergy. 2002;57(12):1111–1117. doi: 10.1034/j.1398-9995.2002.02167.x. [DOI] [PubMed] [Google Scholar]; (c) Takafuji S, Nakagawa T. Air pollution and allergy. J Investig Allergol Clin Immunol. 2000;10(1):5–10. [PubMed] [Google Scholar]
  • 2.Becker A, Chan-Yeung M. Primary asthma prevention: is it possible? Curr Allergy Asthma Rep. 2008;8(3):255–261. doi: 10.1007/s11882-008-0042-6. [DOI] [PubMed] [Google Scholar]
  • 3.Ring J. Allergy in Practice. New York: Springer-Verlag; 2005. Pathophysiology of Allergic Reactions In; pp. 8–27. [Google Scholar]
  • 4.Gilmour MI. Interaction of air pollutants and pulmonary allergic responses in experimental animals. Toxicology. 1995;105(2–3):335–342. doi: 10.1016/0300-483x(95)03230-d. [DOI] [PubMed] [Google Scholar]
  • 5.Gilmour MI, Park P, Selgrade MK. Increased immune and inflammatory responses to dust mite antigen in rats exposed to 5 ppm NO2. Fund Appl Toxicol. 1996;31(1):65–70. doi: 10.1006/faat.1996.0076. [DOI] [PubMed] [Google Scholar]
  • 6.Osebold JW, Zee YC, Gershwin LJ. Enhancement of allergic lung sensitization in mice by ozone inhalation. Proc Soc Exp Biol Med. 1988;188(3):259–264. doi: 10.3181/00379727-188-42733. [DOI] [PubMed] [Google Scholar]
  • 7.Morgenstern V, Zutavern A, Cyrys J, Brockow I, Koletzko S, Kramer U, Behrendt H, Herbarth O, von Berg A, Bauer CP, Wichmann HE, Heinrich J, Grp GS, Grp LS. Atopic diseases, allergic sensitization, and exposure to traffic-related air pollution in children. Am J Resp Crit Care. 2008;177(12):1331–1337. doi: 10.1164/rccm.200701-036OC. [DOI] [PubMed] [Google Scholar]
  • 8.Nordling E, Berglind N, Melen E, Emenius G, Hallberg J, Nyberg F, Pershagen G, Svartengren M, Wickman M, Bellander T. Traffic-related air pollution and childhood respiratory symptoms, function and allergies. Epidemiology. 2008;19(3):401–408. doi: 10.1097/EDE.0b013e31816a1ce3. [DOI] [PubMed] [Google Scholar]
  • 9.Annesi-Maesano I, Moreau D, Caillaud D, Lavaud F, Le Moullec Y, Taytard A, Pauli G, Charpin D. Residential proximity fine particles related to allergic sensitisation and asthma in primary school children. Respir Med. 2007;101(8):1721–1729. doi: 10.1016/j.rmed.2007.02.022. [DOI] [PubMed] [Google Scholar]
  • 10.Janssen NA, Brunekreef B, van Vliet P, Aarts F, Meliefste K, Harssema H, Fischer P. The relationship between air pollution from heavy traffic and allergic sensitization, bronchial hyperresponsiveness, and respiratory symptoms in Dutch schoolchildren. Environ Health Perspect. 2003;111(12):1512–1518. doi: 10.1289/ehp.6243. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Kramer U, Koch T, Ranft U, Ring J, Behrendt H. Traffic-related air pollution is associated with atopy in children living in urban areas. Epidemiology. 2000;11(1):64–70. doi: 10.1097/00001648-200001000-00014. [DOI] [PubMed] [Google Scholar]
  • 12.Mortimer K, Neugebauer R, Lurmann F, Alcorn S, Balmes J, Tager I. Early-lifetime exposure to air pollution and allergic sensitization in children with asthma. J Asthma. 2008;45(10):874–881. doi: 10.1080/02770900802195722. [DOI] [PubMed] [Google Scholar]
  • 13.Nicolai T, Carr D, Weiland SK, Duhme H, von Ehrenstein O, Wagner C, von Mutius E. Urban traffic and pollutant exposure related to respiratory outcomes and atopy in a large sample of children. Eur Respir J. 2003;21(6):956–963. doi: 10.1183/09031936.03.00041103a. [DOI] [PubMed] [Google Scholar]
  • 14.Penard-Morand C, Raherison C, Charpin D, Kopferschmitt C, Lavaud F, Caillaud D, Annesi-Maesano I. Long-term exposure to close-proximity air pollution and asthma and allergies in urban children. Eur Respir J. 2010;36(1):33–40. doi: 10.1183/09031936.00116109. [DOI] [PubMed] [Google Scholar]
  • 15.Penard-Morand C, Charpin D, Raherison C, Kopferschmitt C, Caillaud D, Lavaud F, Annesi-Maesano I. Long-term exposure to background air pollution related to respiratory and allergic health in schoolchildren. Clin Exp Allergy. 2005;35(10):1279–1287. doi: 10.1111/j.1365-2222.2005.02336.x. [DOI] [PubMed] [Google Scholar]
  • 16.Wyler C, Braun-Fahrlander C, Kunzli N, Schindler C, Ackermann-Liebrich U, Perruchoud AP, Leuenberger P, Wuthrich B. Exposure to motor vehicle traffic and allergic sensitization. The Swiss Study on Air Pollution and Lung Diseases in Adults (SAPALDIA) Team. Epidemiology. 2000;11(4):450–456. doi: 10.1097/00001648-200007000-00015. [DOI] [PubMed] [Google Scholar]
  • 17.Brauer M, Hoek G, Smit HA, de Jongste JC, Gerritsen J, Postma DS, Kerkhof M, Brunekreef B. Air pollution and development of asthma, allergy and infections in a birth cohort. Eur Respir J. 2007;29(5):879–888. doi: 10.1183/09031936.00083406. [DOI] [PubMed] [Google Scholar]
  • 18.Gehring U, Wijga AH, Brauer M, Fischer P, de Jongste JC, Kerkhof M, Oldenwening M, Smit HA, Brunekreef B. Traffic-related air pollution and the development of asthma and allergies during the first 8 years of life. Am J Respir Crit Care Med. 2010;181(6):596–603. doi: 10.1164/rccm.200906-0858OC. [DOI] [PubMed] [Google Scholar]
  • 19.Kramer U, Sugiri D, Ranft U, Krutmann J, von Berg A, Berdel D, Behrendt H, Kuhlbusch T, Hochadel M, Wichmann HE, Heinrich J. Eczema, respiratory allergies, and traffic-related air pollution in birth cohorts from small-town areas. J Dermatol Sci. 2009;56(2):99–105. doi: 10.1016/j.jdermsci.2009.07.014. [DOI] [PubMed] [Google Scholar]
  • 20.Gruzieva O, Bellander T, Eneroth K, Kull I, Melen E, Nordling E, van Hage M, Wickman M, Moskalenko V, Hulchiy O, Pershagen G. Traffic-related air pollution and development of allergic sensitization in children during the first 8 years of life. J Allergy Clin Immunol. 2012;129(1):240–246. doi: 10.1016/j.jaci.2011.11.001. [DOI] [PubMed] [Google Scholar]
  • 21.Bedada GB, Heinrich J, Gotschi T, Downs SH, Forsberg B, Jarvis D, Luczynska C, Soon A, Sunyer J, Toren K, Kunzli N. Urban background particulate matter and allergic sensitization in adults of ECRHS II. Int J Hyg Environ Health. 2007;210(6):691–700. doi: 10.1016/j.ijheh.2006.11.005. [DOI] [PubMed] [Google Scholar]
  • 22.Charpin D, Pascal L, Birnbaum J, Armengaud A, Sambuc R, Lanteaume A, Vervloet D. Gaseous air pollution and atopy. Clin Exp Allergy. 1999;29(11):1474–1480. doi: 10.1046/j.1365-2222.1999.00685.x. [DOI] [PubMed] [Google Scholar]
  • 23.Hirsch T, Weiland SK, von Mutius E, Safeca AF, Grafe H, Csaplovics E, Duhme H, Keil U, Leupold W. Inner city air pollution and respiratory health and atopy in children. Eur Respir J. 1999;14(3):669–677. doi: 10.1034/j.1399-3003.1999.14c29.x. [DOI] [PubMed] [Google Scholar]
  • 24.Rosenlund M, Forastiere F, Porta D, De Sario M, Badaloni C, Perucci CA. Traffic-related air pollution in relation to respiratory symptoms, allergic sensitisation and lung function in schoolchildren. Thorax. 2009;64(7):573–580. doi: 10.1136/thx.2007.094953. [DOI] [PubMed] [Google Scholar]
  • 25.Riedl M, Diaz-Sanchez D. Biology of diesel exhaust effects on respiratory function. J Allergy Clin Immunol. 2005;115(2):221–228. doi: 10.1016/j.jaci.2004.11.047. quiz 229. [DOI] [PubMed] [Google Scholar]
  • 26.Bauner D, Laestadius S, Iida N. Evolving technological systems for diesel engine emission control: balancing GHG and local emissions. Clean Technol Envir. 2009;11(3):339–365. [Google Scholar]
  • 27.Akinbami LJ, Lynch CD, Parker JD, Woodruff TJ. The association between childhood asthma prevalence and monitored air pollutants in metropolitan areas, United States, 2001–2004. Environ Res. 2010;110(3):294–301. doi: 10.1016/j.envres.2010.01.001. [DOI] [PubMed] [Google Scholar]
  • 28.Parker JD, Akinbami LJ, Woodruff TJ. Air pollution and childhood respiratory allergies in the United States. Environ Health Perspect. 2009;117(1):140–147. doi: 10.1289/ehp.11497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.NCHS Documentation, Codebook, and Frequencies Geocoding Variables Survey Years 2005–2006. [accessed May 18]; http://www.cdc.gov/nchs/data/nhanes/limited_access/N0506_GE.pdf.
  • 30.NCHS Analytic and Reporting Guidelines: The National Health Nutrition and Examination Survey. [accessed May 18]; http://www.cdc.gov/nchs/data/nhanes/nhanes_03_04/nhanes_analytic_guidelines_dec_2005.pdf.
  • 31.Heeringa S WT, Berglund P. Applied Survey Data Analysis. Vol. 7–8. Chapman Hall/CRC: Boca Raton; 2010. Applied survey data analysis overview. [Google Scholar]
  • 32.NCHS National Center Health Statistics Data Linked to Air Quality Data. [accessed May 18]; http://www.cdc.gov/nchs/data_access/data_linkage/air_quality.htm.
  • 33.EPA, U. S. E. P. A. [accessed May 18];Community Multi-scale Air Quality Model. http://www.epa.gov/amad/Research/RIA/cmaq.html.
  • 34.EPA, U. S. E. P. A. Hierarchical Bayesian Model (HBM) - Derived estimates of air quality for 2006. 2010 EPA/600/R-10/021. [Google Scholar]
  • 35.NCHS Laboratory procedure manual for NHANES 2005–2006 data – Specific IgE/total IgE allergens in serum. http://www.cdc.gov/nchs/data/.nhanes/nhanes_05_06/general_data_release_doc_05_06.pdf.
  • 36.Salo PM, Calatroni A, Gergen PJ, Hoppin JA, Sever ML, Jaramillo R, Arbes SJ, Jr., Zeldin DC. Allergy-related outcomes in relation to serum IgE: results from the National Health and Nutrition Examination Survey 2005–2006. J Allergy Clin Immunol. 2011;127(5):1226–1235. doi: 10.1016/j.jaci.2010.12.1106. e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Pirkle JL, Flegal KM, Bernert JT, Brody DJ, Etzel RA, Maurer KR. Exposure of the US population to environmental tobacco smoke: the Third National Health and Nutrition Examination Survey, 1988 to 1991. JAMA. 1996;275(16):1233–1240. [PubMed] [Google Scholar]
  • 38.NCHS NCHS Urban-Rural Classification Scheme for Counties. [accessed May 18 2013]; http://www.cdc.gov/nchs/data_access/urban_rural.htm.
  • 39.Ckless K, Hodgkins SR, Ather JL, Martin R, Poynter ME. Epithelial, dendritic, and CD4(+) T cell regulation of and by reactive oxygen and nitrogen species in allergic sensitization. Biochim Biophys Acta. 2011;1810(11):1025–1034. doi: 10.1016/j.bbagen.2011.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Bevelander M, Mayette J, Whittaker LA, Paveglio SA, Jones CC, Robbins J, Hemenway D, Akira S, Uematsu S, Poynter ME. Nitrogen dioxide promotes allergic sensitization to inhaled antigen. J Immunol. 2007;179(6):3680–3688. doi: 10.4049/jimmunol.179.6.3680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Oftedal B, Brunekreef B, Nystad W, Nafstad P. Residential outdoor air pollution and allergen sensitization in schoolchildren in Oslo, Norway. Clin Exp Allergy. 2007;37(11):1632–1640. doi: 10.1111/j.1365-2222.2007.02823.x. [DOI] [PubMed] [Google Scholar]
  • 42.Bousquet PJ, Castelli C, Daures JP, Heinrich J, Hooper R, Sunyer J, Wjst M, Jarvis D, Burney P. Assessment of allergen sensitization in a general population-based survey (European Community Respiratory Health Survey I) Ann Epidemiol. 2010;20(11):797–803. doi: 10.1016/j.annepidem.2010.05.012. [DOI] [PubMed] [Google Scholar]
  • 43.Cox L, Williams B, Sicherer S, Oppenheimer J, Sher L, Hamilton R, Golden D. Pearls and pitfalls of allergy diagnostic testing: report from the American College of Allergy, Asthma and Immunology/American Academy of Allergy, Asthma and Immunology Specific IgE Test Task Force. Ann Allergy Asthma Immunol. 2008;101(6):580–592. [PubMed] [Google Scholar]
  • 44.Gergen PJ, Arbes SJ, Jr., Calatroni A, Mitchell HE, Zeldin DC. Total IgE levels and asthma prevalence in the US population: results from the National Health and Nutrition Examination Survey 2005–2006. J Allergy Clin Immunol. 2009;124(3):447–453. doi: 10.1016/j.jaci.2009.06.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Arbes SJ, Jr., Gergen PJ, Vaughn B, Zeldin DC. Asthma cases attributable to atopy: results from the Third National Health and Nutrition Examination Survey. J Allergy Clin Immunol. 2007;120(5):1139–1145. doi: 10.1016/j.jaci.2007.07.056. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Sarnat SE, Klein M, Sarnat JA, Flanders WD, Waller LA, Mulholland JA, Russell AG, Tolbert PE. An examination of exposure measurement error from air pollutant spatial variability in time-series studies. J Expo Sci Environ Epidemiol. 2010;20(2):135–146. doi: 10.1038/jes.2009.10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Parker JD, Woodruff TJ, Akinbami LJ, Kravets N. Linkage of the US National Health Interview Survey to air monitoring data: an evaluation of different strategies. Environ Res. 2008;106(3):384–392. doi: 10.1016/j.envres.2007.11.001. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supp1
NIHMS585225-supplement-supp1.doc (343KB, doc)

RESOURCES