CHEST
Original Research
S
hort-term changes in the occurrence of bronchitic
symptoms are more likely to be infl uenced by
changes in the environment, diet, or life-style risks
than by changes in the genetic pool. From a preventive
perspective, information on environmental, dietary,
and behavioral factors is crucial.
1In a systematic
Medline search, we identifi ed seven studies
concern-ing exposure to ambient air pollution on development
of bronchitic symptoms, but only two studies
2,3focused on children with asthma, both of which
provided suggestive but inconclusive results.
In 2007, we conducted a nationwide cross-sectional
study in Taiwan, where we collected information
about bronchitic symptoms during the past 12 months
and also those important potential determinants of
allergic disease in children. In the present study, we
elaborated the associations between exposure to
urban air pollution and the prevalence of bronchitic
symptoms in schoolchildren, focusing on nitrogen
oxides (NO
2), carbon monoxide (CO), ozone (O
3),
sulfur dioxide (SO
2), and particles with an
aerody-namic diameter
ⱕ 2.5 m m (PM
2.5). Further, we
applied the two-stage hierarchical model to adjust for
confounding, to elaborate effect modifi cation on an
Background: There were limited studies concerning ambient air pollution exposure on
develop-ment of bronchitic symptoms among children. These studies provided suggestive but
inconclu-sive results.
Objectives: To assess the association between air pollutants and the prevalence of bronchitic
symptoms in the Taiwan Children Health Study.
Methods: We conducted a nationwide cross-sectional study of 5,049 Taiwanese children in 2007.
Routine air pollution monitoring data were used for sulfur dioxide (SO
2), nitrogen dioxides (NO
2),
ozone (O
3), carbon monoxide (CO), and particles with an aerodynamic diameter
ⱕ 2.5 m m (PM
2.5).
The exposure parameters were calculated using the between-community 3-year average
concen-tration. The effect estimates were presented as odds ratios (ORs) per interquartile changes for
SO
2, NO
2, O
3, CO, and PM
2.5.
Results: In the two-stage hierarchical model adjusting for confounding, the prevalence of
bron-chitic symptoms with asthma was positively associated with the between-community 3-year
aver-age concentrations of NO
2(adjusted OR, 1.81 per 8.79 ppb; 95% CI, 1.14-2.86), and CO (OR, 1.31
per 105 ppb; 95% CI, 1.04-1.64). The prevalence of phlegm with no asthma was related to O
3(OR,
1.32 per 8.77 ppb; 95% CI, 1.06-1.63).
Conclusions: The results suggest that long-term exposure to outdoor air pollutants, such as NO
2,
CO, and O
3, may increase the prevalence of bronchitic symptoms among children.
CHEST 2010; 137(
䊏):1 –9
Abbreviations: CO 5 carbon monoxide; NO 2 5 nitrogen dioxide; O 3 5 ozone; OC 5 organic carbon; OR 5 odds ratio;PM 2.5 5 particles with an aerodynamic diameter ⱕ 2.5 m m; SO 2 5 sulfur dioxide
Air Pollution and Prevalence of Bronchitic
Symptoms Among Children in Taiwan
Bing-Fang Hwang, MS, PhD ; and Yungling Leo Lee, MD, PhD
Manuscript received November 2, 2009; revision accepted February 16, 2010.
Affi liations: From the Department of Occupational Safety and Health and Graduate Program (Dr Hwang), College of Public Health, China Medical University, Taichung; and the Institute of Preventive Medicine and Research Center for Genes (Dr Lee), Environment and Human Health, College of Public Health, National Taiwan University, Taipei, Taiwan .
Funding/Support: This study was supported by the National Science Council in Taiwan [Grants NSC 96-2314-B-039-019, 96-2314-B-006-053].
Correspondence to: Yungling Leo Lee, MD, PhD, Institute of Preventive Medicine, College of Public Health, National Taiwan University, No.17 Xu-Zhou Rd, 516R, Taipei 100, Taiwan; e-mail: leolee@ntu.edu.tw
© 2010 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( http://www.chestpubs.org/ site/misc/reprints.xhtml ).
DOI: 10.1378/chest.09-2600
[AQ1]
fl uorescence, and PM 2.5 by b -gauge. A daily (24-h) averaged
con-centration was calculated when at least 13 valid hourly values were available with not more than six successive hourly values missing and an 8-h averaged concentration was calculated when at least six valid hourly values were available. We hypothesized that long-term exposure to outdoor air pollutants based on historic routine air monitoring data from 2005 through 2007 would increase the prevalence of bronchitic symptoms among children. Exposure parameters in the present study were 3-year average (2005-2007) and the yearly deviations from the 3-year average concentrations in each municipality, calculated from the 24-h NO 2 , CO, SO 2 , PM 2.5 , and 10:00 am to 6:00 pm 8-h O 3 .
Covariates
Information on potential confounders was obtained from the parental-administered questionnaire. The covariates in the pres-ent analyses included age, gender, parpres-ental education, family annul income, duration of breastfeeding, maternal smoking his-tory during pregnancy, environmental tobacco smoke, cock-roaches note, carpet used, home dampness and mold, and parental atopy ( Table 1 ). Parental atopy was a measure of genetic predis-position to asthma and it was defi ned as the father or mother of the index child ever having been diagnosed as having asthma, allergic rhinitis, or atopic eczema.
Statistical Methods
The association between air pollution and the prevalence of bronchitic symptoms was examined in two groups (children with
individual level, and to assess the community-level
effects of air pollution.
4,5Materials and Methods
Data Collection and Study Population
Taiwan Children Health Study was based on a multipurpose nationwide design that focused on outdoor air pollutants as pri-mary interest. Communities of Taiwan were selected with the aim of maximizing the variability and minimizing the correlations in criteria outdoor pollutants based on historic routine air monitor-ing data. In communities with pollution patterns of interest, neighborhoods with stable, largely middle-income populations, ethnically representative of Taiwan as a whole, were identifi ed from 2004 census data.
A total of 5,804 seventh- to eighth-grade children were recruited from public schools in 14 communities covering diverse parts of Taiwan, which was representative of Taiwanese middle-school children. Written consent was obtained from parents or guardians. The questionnaire was distributed in all communities simultaneously in September, and subjects were given the forms by project staff following their pulmonary function tests and asked to complete and return them the fol-lowing day. Questionnaire responses by parents or guardians were used to categorize children’s basic information, medical history, family history, personal habits, housing characteristics, and environmental conditions. The fi nal study population was 5,052 schoolchildren (response rate, 87.0%). The study protocol was approved by the Respiratory Health Screening Steering Committee of the Taiwan Department of Health and the Insti-tutional Review Board of National Taiwan University, and it complied with the principles outlined in the Helsinki Declaration. 6
Health Outcomes
Questions on respiratory symptoms and illnesses were modi-fi ed after those used in the Children’s Health Study in Southern California. 2,7 The outcome of interest was bronchitic symptoms.
The bronchitic symptoms were defi ned on the basis of having any one of the following: (1) one or more episodes of bronchitis (defi ned by the question: “How many times in the past 12 months did your child have bronchitis?”), without the following symp-toms: (2) chronic cough (defi ned by a yes answer to the question “During the past 12 months, has this child had a cough fi rst thing in the morning that lasted for as long as 3 months in a row?” or to the question “During the past 12 months, has this child had a cough at other times of the day that lasted for as much as 3 months in a row?”), or (3) chronic phlegm (defi ned by a yes answer to the question: “Other than with colds, does this child usually seem congested in the chest or bring up phlegm?”). A history of asthma for each child was assessed on the basis of the answer to the question, “Has a physician ever diagnosed your child as having asthma?”
Exposure Assessment
Complete monitoring data for the air pollutants, including SO 2 ,
NO 2 , O 3 , CO, and PM 2.5 , as well as daily temperature and relative
humidity, are available for 14 EPA monitoring stations in Taiwan Children Health Study communities since 2005 ( Fig 1 ). Concen-trations of each pollutant are measured continuously and reported hourly—CO by nondispersive infrared absorption, NO 2 by
chemi-luminescence, O 3 by ultraviolet absorption, SO 2 by ultraviolet
[AQ3]
Figure 1. Locations of 14 communities in the Taiwan Children Health Study.
the fi rst stage for the risk of bronchitic symptoms as a function of site-specifi c intercepts, j, where a j 5 1, ..., 14, and personal covari-ates. The adjusted site-specifi c intercepts and prevalence rates are related by Pj 5 e a j /(1 1 e a j ). In the second stage, these intercept
terms representing the logit of the site-specifi c prevalence rates (Pj; j 5 1, ..., 14), adjusted for personal covariates, were regressed on each site-specifi c ambient pollutant level by using a linear “eco-logic” regression, that is, logit a j 5 a 1 Uj 1 b Zj, where Uj denotes the random departure from the general prevalence a j on the logit scale for site j and Zj denotes the ambient pollution level for site j. Thus, b can be interpreted as the log odds ratio (per interquartile changes) for each pollutant, adjusted for personal characteristics. and without asthma) to determine whether children with asthma
are more likely to develop bronchitic symptoms. We estimated adjusted odds ratios (ORs) in a two-stage hierarchical model using logistic and ecologic model analyses. The models assume two sources of variation: the variation among subjects in the fi rst stage, part of which could be explained by the individual confounders, and the variation of air pollution between communities in the sec-ond stage, part of which could be explained by variables measured at the municipal level. In the analyses we assumed that (1) the outcome variable follows Bernoulli distribution, (2) intercept terms are random at the municipal level, and (3) all the explanatory variables are fi xed effects. A logistic regression model was fi tted in
Table 1— Distribution of Bronchitic Symptoms, Demographics, and Other Characteristics in Patients With/Without History of Asthma Characteristic Asthma (N 5 376 ) No Asthma (N 5 4,676) x 2 No. % No. % Bronchitis 81 21.5 205 4.4 181.1 ( P , .001) Chronic phlegm 53 14.1 164 3.5 96.6 ( P , .001) Chronic cough 52 13.8 119 2.5 134.5 ( P , .001) Bronchitic symptoms 137 36.4 429 9.2 253.1 ( P , .001) Age, y 2.51 ( P 5 .29) 12 255 67.8 2,971 63.5 13 98 26.1 1,361 29.1 14 23 6.1 344 7.4 Sex 5.14 ( P , .05) Male 202 53.7 2,245 48.0 Female 174 46.3 2,431 52.0 Parental education, y a 1.92 ( P 5 .38) , 8 65 17.4 841 18.1 8-11 244 65.4 3,111 67.1 ⱖ 12 64 17.2 686 14.8
Family annual income a 5.16 ( P 5 .08)
Low 111 32.0 1,651 38.1
Medium 187 53.9 2,174 50.2
High 49 14.1 505 11.7
Environmental tobacco smoke a 0.41 ( P 5 .52)
Yes 188 50.4 2,267 48.8
No 185 49.6 2,375 51.2
Maternal smoking during pregnancy a 4.58 ( P 5 .03)
Yes 22 5.9 174 3.7
No 353 94.1 4,472 96.3
Cockroaches 0.02 ( P 5 .88)
Yes 331 89.5 4,137 89.4
No 39 10.5 493 10.6
Any home dampness and mold 5.15 ( P = .02)
Yes 147 39.1 2,569 54.9 No 229 60.9 2,107 45.1 Duration of breastfeeding, mo a 0.67 ( P 5 .88) 0 188 50.7 2,430 52.8 1-2 143 38.5 1,714 37.3 3-5 22 5.9 246 5.3 ⱖ 6 18 4.9 209 4.5 Carpet used a 1.12 ( P 5 .29) Yes 33 8.8 503 10.8 No 340 91.2 4,151 89.2 Pet 0.03 ( P 5 .86) Yes 222 59.0 2,743 58.7 No 154 41.0 1,933 41.3 Parental atopy 53.4 ( P , .001) Yes 160 42.6 1,170 25.0 No 216 57.4 3,506 75.0
a Number of subjects does not add up to total number because data were missing.
[AQ10]
[AQ11]
Air Pollution and Bronchitic Symptoms
In the one-pollutant model, the prevalence of
bron-chitic symptoms with asthma was related to NO
2lev-els (adjusted OR 1.81 per 8.79 ppb change; 95% CI,
1.14-2.86). With the addition of SO
2(adjusted OR
1.76; 95% CI, 0.99-3.14), PM
2.5(adjusted OR, 2.01;
95% CI, 1.20-3.36), or O
3(adjusted OR, 1.79; 95%
CI, 1.12-2.85), the effect estimate for NO
2remained
signifi cant ( Table 6 ). The adjusted OR for 105 ppb
change in CO was 1.31 (95% CI, 1.04-1.64) and the
estimates changed little when a second pollutant was
added. The adjusted odds ratio for 1.31 ppb change in
SO
2alone was 1.17 (95% CI, 0.97-1.40), but inclusion
of O
3increased the effect estimate substantially
(adjusted OR, 1.26; 95% CI, 1.03-1.53), whereas
addi-tion of CO (adjusted OR, 1.11; 95% CI, 0.92-1.34)
and NO
2(adjusted OR, 1.02; 95% CI, 0.81-1.28) had
little infl uence. The prevalence of bronchitic
symp-toms with asthma was weak or not related to PM
2.5concentrations in any combination of air pollutants.
The risk of bronchitic symptoms with asthma was not
related to O
3in the one-pollutant model (adjusted
OR, 0.80 per 8.77 ppb change; 95% CI, 0.59-1.09),
but addition of PM
2.5reduced the effect estimate
sub-stantially (adjusted OR, 0.64; 95% CI, 0.41-1.00), and
inclusion of other pollutants changes the effect
esti-mates a little. Furthermore, the prevalence of chronic
phlegm without asthma in the one-pollutant model
was related to O
3(adjusted OR, 1.32 per 8.77 ppb
change; 95% CI, 1.06-1.63). With the addition of
either NO
2(adjusted OR, 1.35; 95% CI, 1.07-1.69),
CO (adjusted OR, 1.37; 95% CI, 1.08-1.75), SO
2(adjusted OR, 1.28; 95% CI, 1.03-1.59), or PM
2.5(adjusted OR, 1.34; 95% CI, 1.00-1.78), the effect
estimate for O
3remained signifi cant.
Discussion
The prevalence of bronchitic symptoms increased
according to increased 3-year average concentrations
of two pollutants, NO
2and CO, among children with
asthma. The prevalence of bronchitic symptoms with
asthma was weak or not related to the levels of PM
2.5,
SO
2, and O
3. Our results provide evidence that
chil-dren with asthma are more likely to develop
bron-chitic symptoms when exposed to the air pollutants
NO
2and CO. Furthermore, the prevalence of chronic
phlegm was also associated with O
3among children
without asthma.
Validity of Results
We used routine air pollution monitoring data as
the basis for exposure assessment. These data
reason-ably represented exposures both in the school and in
[AQ5]
[AQ6]
The results from the models are presented as ORs, along with their95% CIs. The goodness of fi t was assessed with likelihood ratio tests to determine whether a variable contributed signifi cantly to the model. First, we fi tted a full model with a complete set of cova-riates. To elaborate sources of confounding, we fi tted models with different combinations of covariates and compared the effect from models with and without the covariate of interest. If the adjusted OR differed from the crude OR by . 10%, that covariate was included in the fi nal model.
We considered the effect of multiple pollutants on the preva-lence of bronchitic symptoms. The correlation between NO 2 and
CO concentrations was high (0.86), and the concentrations of PM 2.5 and SO 2 were also highly correlated (0.68). We fi rst fi tted
one-pollutant models ( Table 2 ) and then considered two-pollutant models by fi tting one (NO 2 or CO) and the other (SO 2 or PM 2.5 )
pollutant . Finally, we fi tted two-pollutant models with O 3 and
another pollutant. The two-pollutant models provide estimates of the independent effects of CO, NO 2 , SO 2 , PM 2.5 , and O 3 on the
bronchitic symptoms controlling for the second pollutant in the model ( Table 3 ). The effect of each pollutant on the risk of bron-chitic symptoms was presented as ORs per interquartile changes for SO 2 , NO 2 , O 3 , and PM 2.5 .
Results
Study Population and Prevalence
of Bronchitic Symptoms
The characteristics of the study population and the
prevalence of bronchitic symptoms with and without
asthma according to the baseline covariates are shown
in Table 1 . The prevalence of bronchitic symptoms
with and without asthma during the past 12 months
was 36.4% and 9.2%, respectively. A larger
propor-tion of subjects with asthma than subjects without
asthma were male ( x
25 5.14, P , .05) and had
mater-nal smoking during pregnancy ( x
25 4.58, P 5 .03),
parental atopy ( x
25 53.4, P , .001), and less
pres-ence of any home dampness and mold ( x
25 5.15,
P 5 .02) ( Table 1 ).
Air Pollution
The distributions of the annual mean air pollutant
concentrations in the 14 monitoring stations in the
years 2005 through 2007 are presented in Table 4 , and
the correlations between different pollutants between
communities are presented in Table 5 . The correlation
between NO
2and CO concentrations was high (0.86),
which refl ects the common source of gasoline-power
vehicles or natural gas power plant combustion. The
concentrations of PM
2.5and SO
2were also highly
cor-related (0.68), indicating a common source of
station-ary fuel combustion or diesel-power vehicles, although
SO
2concentrations were also correlated with NO
2(0.56). The concentration of O
3was negatively
corre-lated with CO (0.33), but positively correcorre-lated with
PM
2.5(0.73) and SO
2(0.37), and it was only weakly
cor-related with that of one (NO
2or CO) and another
(PM
2.5or SO
2) air pollutants.
T
able 2
—
Adjusted Odds Ratios and 95% CIs of Bronchitic Symptoms Among Children W
ith Asthma in T wo Pollutant Models Symptom Two-Pollutant Model 1 (NO 2 1 SO 2 ) T wo-Pollutant Model 2 (NO 2 1 PM 2.5 ) T wo-Pollutant Model 3 (NO 2 1 O 3 ) T wo-Pollutant Model 4 (CO 1 SO 2 ) T wo-Pollutant Model 5 (CO 1 PM 2.5 ) T wo-Pollutant Model 6 (CO 1 O 3 ) T wo-Pollutant Model 7 (SO 2 1 O 3 ) T wo-Pollutant Model 8 (PM 2.5 1 O 3 ) NO 2 (8.79 ppb) Bronchitis 1.99 (1.02-3.87) 2.04 (1.15-3.63) 1.81 (1.06-3.10) … … … … … Chronic phlegm 1.27 (0.59-2.74) 1.51 (0.79-2.89) 1.49 (0.81-2.73) … … … … … Chronic cough 1.25 (0.49-3.23) 1.26 (0.57-2.80) 1.10 (0.58-2.07) … … … … … Bronchitic symptoms 1.76 (0.99-3.14) 2.01 (1.20-3.36) 1.79 (1.12-2.85) … … … … … CO (105 ppb) Bronchitis … … … 1.26 (0.97-1.64) 1.28 (0.99-1.65) 1.26 (0.97-1.64) … … Chronic phlegm … … … 1.17 (0.87-1.59) 1.22 (0.91-1.62) 1.20 (0.89-1.63) … … Chronic cough … … … 1.15 (0.80-1.67) 1.15 (0.81-1.65) 1.05 (0.75-1.45) … … Bronchitic symptoms … … … 1.27 (1.00-1.60) 1.31 (1.04-1.66) 1.27 (1.00-1.61) … … SO 2 (1.31 ppb) Bronchitis 0.95 (0.73-1.23) … … 1.06 (0.85-1.32) … … 1.18 (0.94-1.48) … Chronic phlegm 1.13 (0.83-1.52) … … 1.15 (0.90-1.46) … … 1.26 (0.98-1.61) … Chronic cough 0.93 (0.64-1.35) … … 0.96 (0.71-1.29) … … 1.09 (0.83-1.43) … Bronchitic symptoms 1.02 (0.81-1.28) … … 1.11 (0.92-1.34) … … 1.26 (1.03-1.53) … PM 2.5 (16.84 µg/m 3 ) Bronchitis … 0.75 (0.42-1.33) … … 0.92 (0.54-1.56) … … 1.39 (0.64-2.99) Chronic phlegm … 1.02 (0.53-1.93) … … 1.12 (0.62-2.04) … … 1.98 (0.82-4.76) Chronic cough … 0.75 (0.34-1.66) … … 0.79 (0.39-1.62) … … 1.68 (0.67-4.22) Bronchitic symptoms … 0.75 (0.46-1.24) … … 0.92 (0.58-1.44) … … 1.61 (0.83-3.10) O3 (8.77 ppb) Bronchitis … … 0.85 (0.59-1.23) … … 0.93 (0.64-1.33) 0.76 (0.52-1.12) 0.72 (0.43-1.20) Chronic phlegm … … 0.87 (0.57-1.32) … … 0.92 (0.60-1.42) 0.75 (0.48-1.17) 0.61 (0.33-1.13) Chronic cough … … 0.64 (0.42-0.99) … … 0.65 (0.42-1.02) 0.61 (0.38-0.98) 0.49 (0.26-0.93) Bronchitic symptoms … … 0.82 (0.60-1.11) … … 0.88 (0.64-1.21) 0.71 (0.51-0.98) 0.64 (0.41-1.00) T
wo-stage hierarchical analysis adjusting for age, sex, parental education, yearly income, during of breastfeeding, maternal sm
oking during pregnancy
, environmental tobacco smoke, cockroaches note
monthly
, carpet, pets, home dampness and mold, parental atopy
. CO 5 carbon monoxide; NO 2 5 nitrogen dioxide; O 3 5 ozone; OR 5 odds ratio; PM 2.5 5
particles with aerodynamic diameter 2.5
m m or less; SO 2 5 sulfur dioxide.
T
able 3—
Adjusted ORs and 95% CIs of Bronchitic Symptoms Among Children W
ithout Asthma in T wo-Pollutant Models Symptom T wo-Pollutant Model 1 (NO 2 1 SO 2 ) T wo-Pollutant Model 2 (NO 2 1 PM 2.5 ) T wo-Pollutant Model 3 (NO 2 1 O 3 ) T wo-Pollutant Model 4 (CO 1 SO 2 ) T wo-Pollutant Model 5 (CO 1 PM 2.5 ) T wo-Pollutant Model 6 (CO 1 O 3 ) T wo-Pollutant Model 7 (SO 2 1 O 3 ) T wo-Pollutant Model 8 (PM 2.5 1 O 3 ) NO 2 (8.79 ppb) Bronchitis 1.11 (0.76-1.63) 1.05 (0.78-1.42) 0.98 (0.71-1.35) … … … … … Chronic phlegm 0.83 (0.54-1.28) 0.91 (0.62-1.32) 1.12 (0.80-1.58) … … … … … Chronic cough 1.49 (0.94-2.35) 1.29 (0.86-1.94) 1.27 (0.87-1.86) … … … … … Bronchitic symptoms 1.06 (0.88-1.10) 1.06 (0.82-1.39) 1.06 (0.83-1.34) … … … … … CO (105 ppb) Bronchitis … … … 1.05 (0.90-1.22) 1.04 (0.90-1.21) 1.02 (0.87-1.20) … … Chronic phlegm … … … 0.93 (0.78-1.11) 0.94 (0.79-1.12) 1.06 (0.89-1.26) … … Chronic cough … … … 1.16 (0.97-1.39) 1.16 (0.97-1.38) 1.18 (0.98-1.42) … … Bronchitic symptoms … … … 1.02 (0.90-1.15) 1.02 (0.90-1.15) 1.04 (0.91-1.17) … … SO 2 (1.31 ppb) Bronchitis 0.93 (0.80-1.07) … … 0.94 (0.83-1.06) … … 0.95 (0.83-1.09) … Chronic phlegm 1.14 (0.98-1.33) … … 1.11 (0.98-1.26) … … 1.04 (0.93-1.18) … Chronic cough 0.90 (0.76-1.07) … … 0.96 (0.83-1.10) … … 0.98 (0.84-1.14) … Bronchitic symptoms 0.99 (0.88-1.10) … … 1.00 (0.91-1.10) … … 0.99 (0.90-1.27) … PM 2.5 (16.84 µg/m 3 ) Bronchitis … 0.83 (0.63-1.10) … … 0.82 (0.62-1.10) … … 0.73 (0.49-1.10) Chronic phlegm … 1.35 (0.95-1.90) … … 1.32 (0.96-1.81) … … 0.97 (0.64-1.47) Chronic cough … 0.95 (0.64-1.39) … … 1.01 (0.71-1.44) … … 1.05 (0.63-1.76) Bronchitic symptoms … 0.95 (0.74-1.22) … … 0.97 (0.77-1.23) … … 0.85 (0.62-1.15) O 3 (8.77 ppb) Bronchitis … … 0.96 (0.79-1.18) … … 0.97 (0.79-1.21) 0.99 (0.80-1.23) 1.13 (0.86-1.47) Chronic phlegm … … 1.35 (1.07-1.69) … … 1.37 (1.08-1.75) 1.28 (1.03-1.59) 1.34 (1.00-1.78) Chronic cough … … 1.02 (0.80-1.32) … … 1.07 (0.82-1.39) 1.02 (0.79-1.32) 0.98 (0.69-1.39) Bronchitic symptoms … … 1.07 (0.91-1.25) … … 1.08 (0.91-1.28) 1.07 (0.90-1.27) 1.15 (0.93-1.43) T
wo-stage hierarchical analysis adjusting for age, sex, parental education, yearly income, during of breastfeeding, maternal sm
oking during pregnancy
, environmental tobacco smoke, cockroaches note
monthly
, carpet, pets, home dampness and mold, parental atopy
. See T
diesel-powered motor. NO
2, CO, and PM
2.5are
com-monly from both types of emissions. In the present
study, NO
2and CO concentrations were highly
cor-related and SO
2and PM
2.5concentrations were also
correlated. In the modeling, we were able to control
for one pollutant (SO
2or PM
2.5) at a time as a
poten-tial confounder when assessing the effect of the other
pollutant (NO
2or CO) and vice versa .
Synthesis With Previous Knowledge
In the present study, we found an 80% increase in
the prevalence of bronchitic symptoms per 8.79 ppb
increase in NO
2and a 30% increased prevalence of
bronchitic symptoms per 105 ppb increase in CO
exposure among children with asthma. In addition,
we showed a 30% increased prevalence of chronic
phlegm per 8.77 ppb in O
3among children without
asthma.
Seven previous studies from southern California,
2,7the eastern United States,
10,11Germany,
3Switzer-land,
12and Poland
13have elaborated the relationships
between exposure to outdoor air pollutants and the
risk of bronchitic symptoms, but only two studies
2,3focused on children with asthma. Our present study
and a southern California study
2reported an increased
risk for NO
2, but the fi ndings are inconsistent for SO
2,
PM
2.5, and O
3.
2,3Two surveys conducted in six and 24 cities in the
eastern United States showed positive associations
between exposure to NO
2, SO
2, and PM
2.5and the
prevalence of chronic cough and bronchitis.
10,11In a
cross-sectional study carried out in southern
Califor-nia, increased bronchitic symptoms were associated
the home for two reasons. The schools were chosen
to be in the vicinity of monitoring stations. Because
the density of middle schools in Taiwan is very high,
almost all children attended schools within 1 km of
their homes. In addition, the two-stage hierarchical
models took both individual-level and
community-level information into account, which would make
our results more valid.
From previous literature, we know that a high
pro-portion of outdoor air pollutants (NO
2, CO) penetrate
indoors.
8,9Most of the Taiwanese schoolchildren
spend at least 8 h/d in the school. Air conditioning is
rare in Taiwanese classrooms. In addition,
mechani-cal fi ltration is practimechani-cally the only type of fi ltration in
Taiwanese homes during the summer, even if the
home is air conditioned. Any known or unknown
fac-tors, such as time outdoors, level of exercise, exchange,
and deposition as well as penetration of air pollutants
into the indoor microenvironments could be
attributed to the potential problem of exposure misclassifi
-cation using community-level exposure to represent
personal exposure. This is also a limitation in all
pre-vious studies assessing the effects of ambient air
pol-lution on risk of respiratory illnesses/symptoms
among children.
Assessment of the independent effects of different
pollutants is diffi cult, because urban air pollution
constitutes a complex mixture of several compounds.
Although all the measured pollutants have several
sources, NO
2and CO are predominantly from
natu-ral gas power plant combustion, whereas the main
sources of SO
2and PM
2.5are stationary fossil
com-bustion processes. In addition, busy roads typically
have two types of vehicles, gasoline-powered and
Table 4— Mean and Distribution of 3-y Average Air Pollutant Concentrations Between Communities, Taiwan 2005-2007
Pollutant Mean 6 SD Minimum Maximum Interquartile Range
NO 2 , ppb 17.68 6 0.25 10.06 26.83 8.79 CO, 100 ppb 5.24 6 0.06 3.04 7.78 1.05 SO 2 , ppb 4.33 6 0.10 2.16 10.09 1.31 PM 2.5 , m g/m 3 33.38 6 0.50 19.83 51.34 16.84 O 3 , ppb 44.64 6 0.39 30.34 59.12 8.77 Temperature, °C 24 6 1.04 22.46 25.91 1.23 Relative humidity 74.0% 6 3.0% 69.0% 80.0% 4.3%
See Table 2 for expansion of abbreviations.
Table 5— Correlations of Air Pollutants Across 14 Communities
Pollutant NO 2 CO SO 2 PM 2.5 O 3 N O 2 1.00 0.86 a 0.55 a 0.37 2 0.07 CO … 1.00 0.16 0.09 2 0.33 SO 2 … … 1.00 0.68 a 0.37 PM 2.5 … … … 1.00 0.73 a O 3 … … … … 1.00
See Table 2 for expansion of abbreviations.
symptoms with asthma was not related to the levels of
PM
2.5, it was likely that there was an association with
OC in PM typically present in motor vehicle exhausts
and in particular in diesel exhausts. Further studies
should assess these relationships.
Our fi ndings also suggested that the prevalence of
chronic phlegm without asthma was related to O
3exposure. O
3is a secondary pollutant in the
atmo-sphere produced from traffi c exhausts but scavenged
by direct motor vehicle emissions. O
3is a known
respiratory irritant and has been shown to increase
the synthesis of the allergic antibody IgE in human
beings.
16It could increase sensitization to common
allergens and infl uence the development of phlegm.
Conclusion
The present study provides additional evidence
that exposure to outdoor air pollutants increases the
prevalence of bronchitic symptoms with asthma and
without asthma in schoolchildren. The prevalence
of bronchitic symptoms with asthma was related to
NO
2and CO exposure. The present fi ndings also
suggest that exposure to O
3may increase the
preva-lence of chronic phlegm among children without
asthma.
with the levels of NO
2and PM
2.5.
7Later, a
prospec-tive southern California study found posiprospec-tive
associa-tions between bronchitic symptoms with asthma and
NO
2, organic carbon (OC), and PM
2.5.
2The risks of
bronchitic symptoms with asthma were associated
with NO
2, OC, and PM
2.5. In a cohort study
con-ducted in Germany, the risk of bronchitic symptoms
with asthma was elevated for PM
2.5.
3In a study
of Swiss schoolchildren, the risk of bronchitis was
associated with NO
2, SO
2, and PM
10for the most-
compared with the least-polluted community.
12A survey in Poland found that outdoor air pollution
level was associated with an increased risk of chronic
phlegm.
13The possible mechanisms of NO
2are through
interaction with the immune system or impairment
of respiratory response to infection, which could
result in increased risk of bronchitic symptoms.
14,15There are no plausible mechanisms through which
CO exposure would infl uence the airways and
increase the risk of bronchitic symptoms. In the
pres-ent study, it was not possible to elaborate to what
extent NO
2would have direct effects on children’s
airways. CO is unlikely to have any direct effects on
the airways. Our results did not show any association
between the prevalence of bronchitic symptoms with
asthma and PM
2.5. Although the risk of bronchitic
Table 6— Adjusted ORs and 95% CIs of Bronchitic Symptoms Among Patients With and Without Asthma in Single Pollutant Models Symptoms Asthma No Asthma OR 95% CI OR 95% CI Bronchitis (n 5 81/373) (n 5 205/4,666) NO 2 (8.79 ppb) 1.83 1.07-3.14 0.99 0.72-1.35 CO (105 ppb) 1.28 0.99-1.65 1.03 0.88-1.20 SO 2 (1.31 ppb) 1.11 0.90-1.38 0.95 0.84-1.07 PM 2.5 (16.84 m g/m 3 ) 0.96 0.57-1.62 0.83 0.62-1.11 O 3 (8.77 ppb) 0.84 0.60-1.19 0.96 0.79-1.18 Chronic phlegm (n 5 53/373) (n 5 164/4,636) NO 2 (8.79 ppb) 1.52 0.83-2.78 1.04 0.72-1.51 CO (105 ppb) 1.22 0.92-1.63 0.97 0.81-1.16 SO 2 (1.31 ppb) 1.20 0.95-1.51 1.10 0.97-1.24 PM 2.5 (16.84 m g/m 3 ) 1.17 0.65-2.11 1.30 0.94-1.79 O 3 (8.77 ppb) 0.86 0.58-1.28 1.32 1.06-1.63 Chronic cough (n 5 52/376) (n 5 119/4,676) NO 2 (8.79 ppb) 1.12 0.53-2.40 1.28 0.87-1.85 CO (105 ppb) 1.14 0.79-1.64 1.16 0.97-1.38 SO 2 (1.31 ppb) 0.98 0.73-1.33 0.98 0.85-1.13 PM 2.5 (16.84 m g/m 3 ) 0.83 0.40-1.71 1.03 0.72-1.47 O 3 (8.77 ppb) 0.65 0.42-0.99 1.01 0.79-1.28 Bronchitic symptoms (n 5 137/376) (n 5 429/4,676) NO 2 (8.79 ppb) 1.81 1.14-2.86 1.04 0.82-1.33 CO (105 ppb) 1.31 1.04-1.64 1.02 0.90-1.15 SO 2 (1.31 ppb) 1.17 0.97-1.40 1.00 0.91-1.10 PM 2.5 (16.84 m g/m 3 ) 0.99 0.63-1.57 0.98 0.78-1.23 O 3 (8.77 ppb) 0.80 0.59-1.09 1.06 0.91-1.25
Two-stage hierarchical analysis adjusting for age, sex, parental education, yearly income, during of breastfeeding, maternal smoking during pregnancy, environmental tobacco smoke, cockroaches note monthly, carpet, pets, home dampness and mold, parental atopy. See Table 2 for expansion of defi nitions.
Acknowledgments
Author contributions: Dr Hwang: contributed to coordinating the data analysis, data interpretation, and writing the draft of this paper.
Dr Lee: contributed as the coordinator of Taiwan Children Health Study and was involved with the critical revision of the manuscript.
Financial/nonfi nancial disclosures: The authors have reported to CHEST that no potential confl icts of interest exist with any companies/organizations whose products or services may be dis-cussed in this article .
Other contributions: We thank all the fi eld workers who sup-ported data collection, the school administrators and teachers, and especially the parents and children who participated in this study.
References
1 . Bråbäck L , Forsberg B . Does traffi c exhaust contribute to the development of asthma and allergic sensitization in chil-dren: fi ndings from recent cohort studies . Environ Health . 2009 ; 8 : 17 .
2 . McConnell R , Berhane K , Gilliland F , et al . Prospective study of air pollution and bronchitic symptoms in children with asthma . Am J Respir Crit Care Med . 2003 ; 168 ( 7 ): 790 - 797 . 3 . Morgenstern V , Zutavern A , Cyrys J , et al ; GINI Study Group ;
LISA Study Group . Atopic diseases, allergic sensitization, and exposure to traffi c-related air pollution in children . Am J
Respir Crit Care Med . 2008 ; 177 ( 12 ): 1331 - 1337 .
4 . Dominici F . Invited commentary: air pollution and health-what can we learn from a hierarchical approach? Am J Epidemiol . 2002 ; 155 ( 1 ): 11 - 15, discussion 16 .
5 . Greenland S . Principles of multilevel modelling . Int J
Epidemiol . 2000 ; 29 ( 1 ): 158 - 167 .
6 . 41st World Medical Assembly. Declaration of Helsinki: rec-ommendations guiding physicians in biomedical research
involving human subjects . Bull Pan Am Health Organ .
1990 ; 24 ( 䊏 ): 606 - 609 .
7. McConnell R , Berhane K , Gilliland F , et al . Air pollution and bronchitic symptoms in Southern California children with asthma . Environ Health Perspect . 1999 ; 107 ( 9 ): 757 - 760 . 8 . Partti-Pellinen K , Marttila O , Ahonen A , Suoninen O , Haahtela
T . Penetration of Nitrogen Oxides and Particles from outdoor into indoor air and removal of the pollutants through fi ltration of incoming air . Indoor Air . 2000 ; 10 (2): 126 - 132 .
9 . Rojas-Bracho L , Suh HH , Oyola P , Koutrakis P . Measurement of children’s exposures to particles and nitrogen dioxide in Santiago, Chile . Sci Total Environ . 2002 ; 287 (3): 249 - 264 . 10 . Dockery DW , Speizer FE , Stram DO , Ware JH , Spengler
JD , Ferris BG Jr . Effects of inhalable particles on respiratory health of children . Am Rev Respir Dis . 1989 ; 139 ( 3 ): 587 - 594 . 11 . Dockery DW , Pope CA III . Acute respiratory effects of
particulate air pollution . Annu Rev Public Health . 1994 ; 15 : 107 - 132 .
12 . Braun-Fahrländer C , Vuille JC , Sennhauser FH , et al . Respiratory health and long-term exposure to air pollutants in Swiss schoolchildren. SCARPOL Team. Swiss Study on Childhood Allergy and Respiratory Symptoms with Respect to Air Pollution, Climate and Pollen . Am J Respir Crit Care
Med . 1997 ;1 55 (3): 1042 - 1049 .
13 . Jedrychowski W , Flak E . Effects of air quality on chronic respiratory symptoms adjusted for allergy among preadolescent children . Eur Respir J . 1998 ; 11 ( 6 ): 1312 - 1318 .
14 . Chen TM , Gokhale J , Shofer S , Kuschner WG . Outdoor air pollution: nitrogen dioxide, sulfur dioxide, and carbon monox-ide health effects . Am J Med Sci . 2007 ; 333 ( 4 ): 249 - 256 . 15 . Anonymous ; Committee of the Environmental and
Occupational Health Assembly of the American Thoracic Society . Health effects of outdoor air pollution. Part 2 . Am J
Respir Crit Care Med . 1996 ; 153 ( 2 ): 477 - 498 .
16 . Brunekreef B , Holgate ST . Air pollution and health . Lancet . 2002 ; 360 ( 9341 ): 1233 - 1242 .