This article was downloaded by: [North Dakota State University]On: 19 November 2014, At: 16:38Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Toxicology and Environmental Health, PartA: Current IssuesPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/uteh20

Fine Particulate Air Pollution and Hospital Admissionsfor Pneumonia in a Subtropical City: Taipei, TaiwanShang-Shyue Tsai a & Chun-Yuh Yang b ca Department of Healthcare Administration , I-Shou University , Kaohsiung , Taiwanb Department of Public Health, College of Health Sciences , Kaohsiung Medical University ,Kaohsiung , Taiwanc Division of Environmental Health and Occupational Medicine , National Health ResearchInstitute , Miaoli , TaiwanPublished online: 20 Feb 2014.

To cite this article: Shang-Shyue Tsai & Chun-Yuh Yang (2014) Fine Particulate Air Pollution and Hospital Admissions forPneumonia in a Subtropical City: Taipei, Taiwan, Journal of Toxicology and Environmental Health, Part A: Current Issues, 77:4,192-201, DOI: 10.1080/15287394.2013.853337

To link to this article: http://dx.doi.org/10.1080/15287394.2013.853337

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Journal of Toxicology and Environmental Health, Part A, 77:192–201, 2014Copyright © Taylor & Francis Group, LLCISSN: 1528-7394 print / 1087-2620 onlineDOI: 10.1080/15287394.2013.853337

FINE PARTICULATE AIR POLLUTION AND HOSPITAL ADMISSIONS FORPNEUMONIA IN A SUBTROPICAL CITY: TAIPEI, TAIWAN

Shang-Shyue Tsai1, Chun-Yuh Yang2,3

1Department of Healthcare Administration, I-Shou University, Kaohsiung, Taiwan2Department of Public Health, College of Health Sciences, Kaohsiung Medical University,Kaohsiung, Taiwan3Division of Environmental Health and Occupational Medicine, National Health ResearchInstitute, Miaoli, Taiwan

This study was undertaken to determine whether there was a correlation between fineparticles (PM2.5) levels and hospital admissions for pneumonia in Taipei, Taiwan. Hospitaladmissions for pneumonia and ambient air pollution data for Taipei were obtained for theperiod from 2006 to 2010. The relative risk of hospital admissions for pneumonia was esti-mated using a case-crossover approach, controlling for weather variables, day of the week,seasonality, and long-term time trends. For the single-pollutant model (without adjustment forother pollutants), increased numbers of admissions for pneumonia were significantly associ-ated with higher PM2.5 levels both on warm days (>23◦C) and on cool days (<23◦C). This wasaccompanied by an interquartile range elevation correlated with a 12% (95% CI = 16%–13%)and 4% (95% CI = 3%–6%) rise in number of admissions for pneumonia, respectively. In thetwo-pollutant models, PM2.5 remained significant after inclusion of sulfur dioxide (SO2) orozone (O3) both on warm and on cool days. This study provides evidence that higher levels ofPM2.5 increase the risk of hospital admissions for pneumonia.

Over the past decade, many epidemio-logic studies demonstrated positive associationsbetween ambient levels of airborne particulatematter (PM) (generally measured as PM withan aerodynamic diameter ≤10 µm [PM10]) andrate of daily mortality (Levy et al., 2000; Popeet al., 2004; Goodman et al., 2004; Schwartz,2004; Analitis et al., 2006; World HealthOrganization [WHO], 2006; Cohen et al.,2005) and hospital admissions or emergencyroom (ER) visits for cardiovascular and respi-ratory morbidity (Samet and Krewski, 2007;Zanobetti et al., 2000; Le Tertre et al., 2002;Bedeschi et al., 2007; Krewski et al., 2005).Evidence that PM exerts adverse effects andconsequently negatively affects public healthhas led to more stringent regulations and stan-dards for levels of PM in outdoor air in the

Received 26 July 2013; accepted 27 September 2013.Address correspondence to Chun-Yuh Yang, PhD, MPH, Department of Public Health, College of Health Sciences, Kaohsiung

Medical University, 100 Shih Chuan 1st RD, Kaohsiung, Taiwan. E-mail: chunyuh@kmu.edu.tw

United States and other countries (Dominiciet al., 2006; Craig et al., 2008).

While previous studies primarily used PM10as an exposure indicator, fine particles (definedas PM with an aerodynamic diameter less than2.5 µm; PM2.5) have become a greater healthand regulatory concern due to data based uponepidemiologic studies suggesting that PM2.5might exert greater toxicity than larger parti-cles (Cifuentes et al., 2000; Schwartz et al.,1996; Zanobetti et al., 2009; Liao et al., 2011).It is now generally accepted that fine particlesare more harmful to health than larger parti-cles (PM10) because fine particles offer a greatersurface area and hence potentially larger con-centrations of adsorbed or condensed toxicair pollutants per unit mass (Wilson and Suh,1997; Pope and Dockery, 2006). Indeed, this

192

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14

FINE PARTICULATE AIR POLLUTION AND PNEUMONIA 193

is the underlying basis for the World HealthOrganization (WHO) recommendations usingPM2.5 rather than PM10 concentrations as airquality indicators (WHO, 2006).

Relatively few epidemiologic studies havebeen undertaken that address specifically theadverse health effects of PM2.5, as few citieshave monitored PM2.5 (Host et al., 2008).Considerable attention has focused on the cor-relation between exposure to PM and hospitaladmission for all respiratory-related symptoms(Vigotti et al., 2007; Peng et al., 2009; Hostet al., 2008; Bell et al., 2008; Zanobetti et al.,2009; Halonen et al., 2009; Linares and Diaz,2010a), possibly combining outcomes with dif-ferent sensitivities to air pollution and differ-ent lags between exposure and hospitalization(Medina-Ramon et al., 2006). Several studiesinvestigated the relationship between fine PMand hospital admissions (or ER visits) for pneu-monia (Dominici et al., 2006; Zanobetti andSchwartz, 2006; Host et al., 2008; Belleudiet al., 2010; Halonen et al., 2009). Becausethese studies were conducted primarily inAmerica and some European cities, the findingsmay not be applicable to Taiwan, which maypossess different air pollutant mixtures.

While many epidemiological investigationsin Taiwan reported associations of increasedrate of mortality and morbidity with higherlevels of ambient PM10 (Yang et al., 2004,2007; Chang et al., 2005; Chiu et al., 2008;Yang, 2008; Hsieh et al., 2010), fewer studiesassessed correlation with PM2.5 levels, whichmay be due to the lack of monitoring data(Hung et al., 2012a, 2012b; Hsieh et al., 2013;Chiu et al., 2013; Chang et al., 2013).

This study was undertaken to examine theshort-term impact of PM2.5 on daily hospitaladmissions for pneumonia among individualsresiding in Taipei city, the largest metropoli-tan city in Taiwan, over a 5-year period,2006–2010, using a case-crossover design.

MATERIALS AND METHODS

Taipei CityThis study examined daily variations in hos-

pital admissions for pneumonia in relation to

PM2.5 levels in Taipei for the 6-year periodfrom 2006 through 2010. Taipei is the largestmetropolitan city in Taiwan with a populationof approximately 2.64 million located in north-ern Taiwan. The major air pollution sourceis automobile exhaust emissions. Taipei has asubtropical climate, with an annual averagetemperature of 23◦C.

Hospital Admission DataThe National Health Insurance (NHI)

Program, which provides compulsory universalhealth insurance, was implemented in Taiwanon March 1, 1995. Under the NHI, 98% ofthe island’s population receives all forms ofhealth care services including outpatient ser-vices, inpatient care, Chinese medicine, dentalcare, childbirth services, physical therapy, pre-ventive health care, home care, and rehabili-tation for chronic mental illness. Most medicalinstitutions (93%) are contracted to the Bureauof NHI (BNHI), and those not contracted pro-vide fewer health services. More than 96%of the population who are covered by NHIused health services at least one time throughcontracted medical institutions.

Computerized records of daily clinic vis-its or hospital admissions are available foreach contracted medical institution. All med-ical institutions must submit standard claimdocuments for medical expenses on a com-puterized form, which includes the date ofadmission and discharge, identification num-ber, gender, date of birth, and diagnostic codeof each admission. Therefore, the informationfrom the NHI database appears to be suffi-ciently complete, reliable, and accurate for usein epidemiological studies. Daily counts of hos-pital admissions for pneumonia (InternationalClassification of Diseases, 9th revision [ICD-9],codes 480–486) were extracted from the med-ical insurance files for the period 2006–2010.

PM2.5 and Meteorological DataSix air quality monitoring stations were

established in Taipei city by the TaiwaneseEnvironmental Protection Administration (EPA),a central governmental agency, in 1994. The

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14

194 S.-S. TSAI AND C.-Y. YANG

monitoring stations were fully automated androutinely monitored five “criteria” pollutants,namely, sulfur dioxide (SO2) (by ultravioletfluorescence), particulate matter (PM10) (bybeta-ray absorption), nitrogen dioxide (NO2)(by ultraviolet fluorescence), carbon monoxide(CO) (by nondispersive infrared photometry),and ozone (O3) (by ultraviolet photometry)levels. However, PM2.5 was not regularly mon-itored. PM2.5 concentrations in Taiwan weremeasured continuously since 2006. The avail-ability of the monitoring network for PM2.5 pro-vided an opportunity to investigate the impactof PM2.5 on hospital admissions for pneumo-nia. For each day, hourly air pollution datawere obtained for all of the monitoring sta-tions. After calculating the hourly mean of eachpollutant from the 6 stations, the 24-h aver-age levels of these pollutants were computed.Daily information on mean temperature andmean humidity was provided by the TaipeiObservatory of the Central Weather Bureau.

StatisticsData were analyzed using the case-

crossover technique (Maclure, 1991; Marshalland Jackson, 1993; Mittleman et al., 1995).This design is an alternative to Poisson time-series regression models for studying the short-term effects attributed to air pollutants (Levyet al., 2001). In general, the case-crossoverdesign and the time-series approach yieldedalmost identical results (Neas et al., 1999; Leeand Schwartz, 1999; Lu and Zeger, 2007).

The time-stratified approach was used forthe case-crossover analysis (Levy et al., 2001).A stratification of time into separate monthswas made to select referent days as the daysfalling on the same day of the week withinthe same month as the index day. Air pollutionlevels during the case period were comparedwith exposures occurring on all referent days.This time-stratified referent selection schememinimizes bias due to nonstationarity of air pol-lution time-series data (Lumley and Levy, 2000;Janes et al., 2005; Mittleman, 2005). Results ofprevious investigations indicated that increasednumber of hospital admissions were associated

with higher ambient air pollutant levels on thesame day or the previous two days (Katsouyanniet al., 1997). Longer lag times have rarely beendescribed. Thus, the cumulative lag period upto 2 previous days (i.e., the average air pollu-tant levels of the same and previous 2 days) wasused. Because pollutants vary considerably byseason, especially O3 and PM, seasonal inter-actions between PM2.5 and hospital admissionshave often been reported. However, previousstudies were conducted mostly in countrieswhere the climates are substantially differentfrom that in Taipei (Yang et al., 2004; Changet al., 2005), which has a subtropical climatewith no apparent four-season cycle. Hence inthis study the potential interactions of season-ality on the adverse effects of PM2.5 were notconsidered, but temperature was used instead.The adverse health effects of each air pollutantwere examined for the “warm” days (days witha mean temperature above 23◦C) and “cool”days (days with a mean temperature below23◦C) separately.

The associations between hospital admis-sions for pneumonia and levels of PM2.5 wereestimated using the odds ratio (OR) and the95% confidence intervals (CI), which weregenerated using conditional logistic regressionwith weights equal to the number of hospitaladmissions on that day. All statistical analy-ses were performed using the SAS package(version 9.2; SAS Institute, Inc., Cary, NC).Both single-pollutant models and multipollu-tant models were fitted with a different com-bination of pollutants (up to two pollutants permodel) to assess the stability of the effects ofPM2.5. Exposure levels to air pollutants wereentered into the models as continuous vari-ables. Meteorologic variables such as daily aver-age temperature and humidity on the sameday, which might play a confounding role, wereincluded in the model. Inclusion of baromet-ric pressure did not markedly change the effectestimates and therefore this was not consid-ered in the final model. OR were calculatedfor the interquartile difference (IQR, betweenthe 25th and the 75th percentile) for PM2.5, asnoted during the study period. The criterion forsignificance was set at p < .05.

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14

FINE PARTICULATE AIR POLLUTION AND PNEUMONIA 195

RESULTS

During the 5 years of the study, there werea total of 62,478 hospital admissions for pneu-monia for the 47 hospitals in Taipei city. Thedescriptive statistics for admissions and cor-responding environmental data are shown inTable 1. There was an average of 62.98 dailyhospital admissions for pneumonia in the cityover the study period.

Pearson’s correlation coefficients amongthe air pollutants are presented in Table 2.There was a degree of correlation among thepollutants, especially between PM10 and PM2.5(r = 0.78), PM2.5 and SO2 (r = 0.61), PM2.5 andNO2 (r = 0.54), PM2.5 and CO (r = 0.54), SO2and NO2 (r = 0.52), SO2 and CO (r = 0.5), andNO2 and CO (r = 0.89).

Table 3 shows the effect estimates ofPM2.5 on hospital admissions for pneumoniain single-pollutant models and two-pollutantmodels. For the single pollutant model (with-out adjustment for other pollutants), increasednumbers of admissions for pneumonia weresignificantly associated with levels of PM2.5both on warm (>23◦C) and on cool (<23◦C)days. There was an IQR rise associated witha 12% (95% CI = 10%–13%) and 4% (95%CI = 3%–6%) elevation in number of pneumo-nia admissions, respectively. In two-pollutantmodels, PM2.5 remained significant after theinclusion of SO2 or O3 both on warm and cooldays.

TABLE 2. Correlation Coefficients Among Air Pollutants

Variable PM10 PM2.5 SO2 NO2 CO O3

PM10 1.0 0.78 0.43 0.35 0.35 0.26PM2.5 – 1.0 0.61 0.54 0.54 0.31SO2 – – 1.0 0.52 0.50 0.06NO2 – – – 1.0 0.89 −0.07CO – – – – 1.0 −0.23O3 – – – – – 1.0

TABLE 3. Association Between PM2.5 and Admissions forPneumonia in Taipei, Taiwan, 2006–2010

Temperature PM2.5 OR(95% CI)a

≥23◦C0C Without adjustmentb 1.12 (1.10–1.13) (1021 d)Adjusted for SO2 1.14 (1.11–1.16)Adjusted for NO2 1.01 (0.99–1.03)Adjusted for CO 1.01 (0.99–1.03)Adjusted for O3 1.11 (1.09–1.13)

<23◦C Without adjustmentb 1.04 (1.03–1.06) (805 d)Adjusted for SO2 1.15 (1.13–1.17)Adjusted for NO2 0.98 (0.97–0.99)Adjusted for CO 1.03 (1.01–1.05)Adjusted for O3 1.05 (1.04–1.07)

aCalculated for an interquartile range increases of PM2.5(17.46 µg/m3) and adjusted for temperature and humidity.

bSingle-pollutant model.

DISCUSSION

This study is one of the few that inves-tigated the association between exposure toPM2.5 and hospital admissions for pneumonia,as well as being the first investigation in Asia.Data demonstrated that the levels of PM2.5

TABLE 1. Distribution of Daily Pneumonia Admissions, Weather Conditions, and Air Pollution Variables in Taipei, Taiwan, 2006–2010

Percentile

Variablea Min Max Mean 25% 50% 75%

PM10 (µg/m3) 14.26 34.89 46.83 62.37 888.02 51.71PM2.5 (µg/m3) 8.35 19.46 27.06 36.92 140.54 29.99SO2 (ppb) 1.00 2.73 3.65 4.91 11.14 3.94NO2 (ppb) 3.22 19.97 23.86 28.81 55.59 24.67CO (ppm) 0.13 0.50 0.63 0.80 1.76 0.68O3 (ppb) 4.00 17.95 23.95 30.23 70.89 24.65Temperature (◦C) 9.35 19.50 24.11 28.42 33.18 23.69Humidity (%) 31.37 66.54 73.11 79.57 94.19 72.82Pneumonia admissions 0 51 62 73 139 62.98

Note. Min, minimum value; Max, maximum value.a24-h average.

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14

196 S.-S. TSAI AND C.-Y. YANG

were positively associated with elevation in thedaily number of hospitalizations for pneumo-nia after inclusion of SO2 or O3 both on warmand on cool days. The observed adverse effectsof PM2.5 were not maintained in the presenceof NO2 or CO. This might be attributed to thecollinearity between levels of PM2.5 and NO2or CO, which is a common problem in this typeof study.

Investigations on the effects of PM2.5 onadmissions for pneumonia are rare. Host et al.(2008) found a 2.5% (95 % CI = 0.1%–4.8%) excess risk of hospitalization for lowerrespiratory tract infections per 10 µg/m3 in6 French cities. Dominici et al. (2006) reporteda 1.39% (95% CI = 0.6%–2.09%) rise in hospi-talizations for respiratory-tract infection includ-ing acute bronchitis and pneumonia per 10-µg/m3 elevation in PM2.5 levels. Zanobetti andSchwartz (2006) demonstrated an increase of6.48% (95% CI = 1.13%–11.43%) in pneumo-nia admissions for a 17.14-µg/m3 rise in PM2.5levels (difference between 90th and 10th per-centile of PM2.5). A study in Rome by Belleudiet al. (2010) reported a 10-µg/m3 increment inthe level of PM2.5 was correlated with a 2.82%(95% CI = 0.52%–5.19%) increase in hospi-talizations for lower-respiratory-tract infectionsincluding acute bronchitis and pneumonia.A study in Helsinki by Halonen et al. (2009)found an IQR increment (6.2 µg/m3) in the lev-els of PM2.5 was associated with a 2.41% (95%CI = 0.64%–4.21%) increase in admissions forpneumonia. Data in this study demonstrateda 6.87% (which corresponds to 12% eleva-tion per IQR increment) and 2.29% (whichcorresponds to 4% rise per IQR increment)increase in hospitalizations for pneumonia per10-µg/m3 increment in the 3-d moving aver-age (lag 2) concentrations of PM2.5 for warmdays and cool days, respectively.

In our study, adverse effects were observedon both warm and cool days, but were greateron warm days (effect modification). It waspossible to confirm that noted PM effects var-ied by season (Zanobetti et al., 2009; Bellet al., 2008). The observed seasonal varia-tion in effect estimates might be explained byvariation in exposure patterns. Individuals in

Taipei are more likely to go outdoors and openthe windows in the warm season than in thecool season (higher exposure); thus, monitoredPM2.5 concentrations may be closer to per-sonal exposure to PM2.5 in the warm than coolseason (better exposure assessment). This con-dition may attenuate the PM2.5-induced effectin the cool season. On the other hand, seasonaldifferences in air pollution mixture may alsoaffect the effect estimates. Further, comparedwith other studies in developed countries, ourstudy found larger effect estimates per unitincrease of PM2.5. One potential explanationfor this discrepancy is that most published stud-ies only demonstrated straightforward pooledeffect estimates (lack of effect estimates strat-ified by season). This may mask inherent dif-ferences between different climates and airpollution mixtures. Variations in seasonal andregional effect estimates may in part resultfrom differences in the chemical compositionof PM2.5 (Bell et al., 2008). Nevertheless, theseasonal pattern of air pollution adverse healtheffects needs to be further investigated.

The most common and consistent asso-ciations between air pollutants and hospitaladmissions for respiratory disease were foundwith PM (Brunekreef and Holgate, 2002). A sig-nificant association was found between PM2.5exposure and admissions for pneumonia in thisstudy. This finding is in agreement with previ-ous studies (Dominici et al., 2006; Zanobettiand Schwartz, 2006; Host et al., 2008; Belleudiet al., 2010; Halonen et al., 2009). Somepathophysiological hypotheses may be inferredto explain the correlation between short-termadverse effects of PM2.5 and pneumonia occur-rence. Fine particles were suggested as theeffective toxic fraction of PM, because PM pro-mote and maintain oxidative stress both atthe respiratory level (the entry system) and atthe systemic level where oxidative stress pro-duces inflammation (MacNee and Donaldson,2003; Ghio et al., 2012). The mechanismsunderlying the adverse respiratory effects ofexposure to PM2.5 are unclear (Medina-Ramonet al., 2006). Animal studies demonstratedan increased vulnerability to PM resulting incardiopulmonary disease (Costa and Dreher,

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14

FINE PARTICULATE AIR POLLUTION AND PNEUMONIA 197

1997), and exacerbations of ongoing pneumo-coccal infection after exposure to concentratedambient PM2.5 (Zelikoff et al., 2003).

Air pollution has consistently been asso-ciated with increased hospital admissions forrespiratory and cardiovascular diseases in citiesthroughout the world. Recent studies suggestthat the rise in hospital admissions is due pri-marily to exposure to PM2.5 (Schwartz et al.,1996; Suh et al., 2011). Major PM2.5 com-ponents vary by region and by season, buttypically include ammonium sulfate and nitrate,elemental carbon, carbonaceous species, car-bonates, metals, and water (Peng et al.,2009; Suh et al., 2011). Despite considerableresearch, the relative toxicity of different con-stituents of PM2.5 remains unclear but likelyvaries, depending upon the components (Suhet al., 2011).

The origin of chemical pollutants in anurban atmosphere is known to be predomi-nantly attributed to road traffic (Linares andDiaz, 2010a). PM2.5 concentrations possess aless important natural component than PM10.This smaller natural component makes PM2.5 amore reliable indicator than PM10 for measur-ing anthropic activity in a large city (Linares andDiaz, 2010b)

The case-crossover study design was pro-posed by Maclure (1991) to examine theinfluence of transient, intermittent exposureson the subsequent risk of rare acute-onsetevents in close temporal proximity to expo-sure. This design offers the ability to controlmany confounders by design rather than by sta-tistical modeling. This design is an adaptationof the case-control study in which each caseserves as his or her own referent. Thereforetime-invariant subject-specific variables suchas gender, age, underlying chronic disease, orother individual-level characteristics do not actas confounders. In addition, a time-stratifiedapproach (Levy et al., 2001) was found tobe effective in controlling for seasonality, timetrends, and chronic and slowly varying poten-tial confounders (Lumley and Levy, 2000; Janeset al., 2005; Mittleman, 2005). In general, thecase-crossover design and the general additivemodel (GAM) approach, which has been the

analytic method of choice for studying short-term adverse effects of air pollution since 1990(Schwartz and Marcus, 1990), produced almostidentical results (Neas et al., 1999; Lee andSchwartz, 1999; Lu and Zeger, 2007).

For a factor to confound the relationshipbetween PM2.5 levels and admissions for pneu-monia it needs to be correlated with bothvariables. It is unlikely that smoking and otherindoor pollutants confounded the present asso-ciation, since day-to-day variations in indooremissions, including smoking, may not be cor-related with fine PM air pollution.

Exposure measurement error is a com-mon concern in environmental epidemiology.PM2.5 levels were assigned from fixed, outdoormonitoring stations to individuals to estimateexposure (assuming that exposure was homo-geneous, encompassing all the studied area).Exposure measurement errors resulting fromdifferences between the population-averageexposure and ambient PM2.5 levels are notavoidable. However, the potential for misclassi-fication of exposure due to the lack of personalmeasurements of PM2.5 exposure in this studyis of the Berkson type, known to produce abias toward the null and an underestimate ofthe association (Katsouyanni et al., 1997; Zegeret al., 2000).

Our study population is homogeneous interms of race compared with populations inother cities. This study was conducted in asubtropical city. These facts may restrict some-what the generalizability of these findings toother locations with different meteorologicaland racial characteristics. Further, behaviorsuch as air conditioning usage or time spentoutdoors may affect personal exposures. Thismight affect the magnitude of the observedassociations compared with other geographicallocations.

In summary, this study provided evidenceof associations between short-term exposureto fine PM and increased number of hospi-tal admissions for pneumonia. The ecologicaldesign of this investigation precludes the infer-ence of cause and effect. However, these find-ings reinforce the possible role of PM2.5 onhospital admissions for pneumonia.

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14

198 S.-S. TSAI AND C.-Y. YANG

FUNDING

This study is based in part on data fromthe National Insurance Research Databaseprovided by the Bureau of National HealthInsurance, Department of Health and man-aged by National Health Research Institutes.The interpretation and conclusions containedherein do not represent those of Bureauof National Health Insurance, Department ofHealth or National Health Research Institutes.This study was supported by a grant fromNational Health Research Institutes, Taiwan(EO-101-PP-08).

REFERENCES

Analitis, A., Katsouyanni, K., Dimakopoulou,K., Samoli, E., Nikoloulopoulos, A. K.,Petasakis, Y., Touloumi, G., Schwartz, J.,Anderson, H. R., Cambra, K., Forastiere,F., Zmirou, D., Vonk, J. M., Clancy, L.,Kriz, B., Bobvos, J., and Pekkanen, J.2006. Short-term effects of ambient particleson cardiovascular and respiratory mortality.Epidemiology 17: 230–233.

Bedeschi, E., Campari, C., Candela, S., Collini,G., Caranci, N., Frasca, G., Galassi, C.,Francesca, G., and Vigotti, M. A. 2007.Urban air pollution and respiratory emer-gency visits at pediatric unit, Reggio Emilia,Italy. J. Toxicol. Environ. Health A 70:261–265.

Bell, M. L., Ebisu, K., Peng, R. D., Walker, J.,Samet, J. M., Zeger, S. L., and Dominici,F. 2008. Seasonal and regional short-termeffects of fine particles on hospital admis-sions in 202 US counties, 1999–2005. Am.J. Epidemiol. 168: 1301–1310.

Belleudi, V., Faustini, A., Stafoggia, M., Cattani,G., Marconi, A., Perucci, C. A., andForastiere, F. 2010. Impact of fine and ultra-fine particles on emergency hospital admis-sions for cardiac and respiratory diseaes.Epidemiology 21: 414–423.

Brunekreef, B., and Holgate, S. T. 2002.Air pollution and health. Lancet 360:1233–1242.

Chang, C. C., Tsai, S. S., Ho, S. C., and Yang,C. Y. 2005. Air pollution and hospital admis-sions for cardiovascular disease in Taipei,Taiwan. Environ. Res. 98: 114–119.

Chang, C. C., Kuo, C. C., Liou, S. H., and Yang,C. Y. 2013. Fine particulate air pollution andhospital admissions for myocardial infarc-tion in a subtropical city: Taipei, Taiwan. J.Toxicol. Environ. Health Part A 76: 440–448.

Chiu, H. F., Cheng, M. H., and Yang, C. Y.2008. Air pollution and hospital admissionsfor pneumonia in a subtropical city: Taipei,Taiwan. Inhal. Toxicol. 21: 32–37.

Chiu, H. F., Tsai, S. S., Weng, H. H., and Yang,C. Y. 2013. Short-term effects of fine partic-ulate air pollution on emergency room vis-its for cardiac arrhythmias: A case-crossoverstudy in Taipei. J. Toxicol. Environ. Health A76: 614–623.

Costa, D. L., and Dreher, K. L. 1997.Bioavailable transition metals in particu-late matter mediate cardiopulmonary injuryin healthy and compromised animal mod-els. Environ. Health Perspect. 105(suppl. 5):1053–1060.

Cohen, A. J., Anderson, H. R., Ostro, B.,Pandey, K. D., Krzyzanowsli, M., Kunzli,N., Gutschmidt, K., Pope, A., Romieu, I.,Samet, J. M., and Smith, K. 2005. Theglobal burden of disease due to outdoor airpollution. J. Toxicol. Environ. Health A 68:1301–1307.

Cifuentes, L. A., Vega, J., Kopfer, K., and Lave,L. B. 2000. Effect of the fine fraction of partic-ulate matter versus the coarse mass and otherpollutants on daily nortality in Santiago Chile.J. Air Waste Manage. Assoc. 50: 1287–1298.

Craig, L., Brook, J. R., Chiotti, Q., Croes, B.,Gower, S., Hedley, A., Krewski, D., Krupnick,A., Krzyzanowski, M., Moran, M. D., Pennell,W., Samet, J. M., Schneider, J., Schortreed,J., and Williams, M. 2008. Air pollution andpublic health: A guide document for riskmanagers. J. Toxicol. Environ. Health A 71:588–698.

Dominici, F., Peng, R. D., Bell, M. L., Pham, L.,McDermott, A., Zeger, S. L., and Samet, J.M. 2006. Fine particulate air pollution andhospital admission for cardiovascular and

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14

FINE PARTICULATE AIR POLLUTION AND PNEUMONIA 199

respiratory diseases. J. Am. Med. Assoc. 295:1127–1134.

Ghio, A. J., Carraway, M. S., and Madden, M.C. 2012. Composition of air pollution parti-cles and oxidative stress in cells, tissues, andliving systems. J. Toxicol. Environ. Health B15: 1–21.

Goodman, P. G., Dockery, D. W., and Clancy,L. 2004. Cause-specific mortality and theextended effects of particulate pollutionand temperature exposure. Environ. HealthPerspect. 112: 179–185.

Halonen, J. I., Lanki, T., Yli-Tuomi, T., Tiittanen,P., Kulmala, M., and Pekkanen, J. 2009.Particulate air pollution and acute cardiores-piratory hospital admissions and mortal-ity among the elderly. Epidemiology 20:143–153.

Host, S., Larrieu, S., Pascal, L., Blanchard, M.,Declercq, C., Fabre, P., Jusot, J. F., Chardon,B., Le Tertre, A., Wagner, V., Prouvost, H.,and Lefrance, A. 2008. Short-term associa-tions between fine and coarse particles andhospital admissions for cardiorespiratory dis-eases in six French cities. Occup. Environ.Med. 65: 544–551.

Hsieh, Y. L., Yang, Y. H., Wu, T. N., andYang, C. Y. 2010. Air pollution and hospi-tal admissions for myocardial infarction ina subtropical city: Taipei, Taiwan. J. Toxicol.Environ. Health A 73: 757–765.

Hsieh, Y. L., Tsai, S. S., and Yang, C. Y. 2013.Fine particulate air pollution and hospitaladmissions for congestive heart failure: Acase-crossover study in Taipei. Inhal. Toxicol.25: 455–460.

Hung, L. J., Tsai, S. S., Chen, P. S., Yang, Y. H.,Liou, S. H., Wu, T. N., and Yang, C. Y. 2012a.Traffic air pollution and risk of death frombreast cancer in Taiwan: Fine particulate mat-ter (PM2.5) as a proxy marker. Aerosol AirQual. Res. 12: 275–282.

Hung, L. J., Chan, T. F., Wu, C. H., Chiu, H.F., and Yang, C. Y. 2012b. Traffic air pollu-tion and risk of death from ovarian cancerin Taiwan: Fine particulate matter (PM2.5) asa proxy marker. J. Toxicol. Environ. Health A75: 174–182.

Janes, H., Sheppard, L., and Lumley, T. 2005.Case-crossover analyses of air pollutionexposure data: Referent selection strategiesand their implications for bias. Epidemiology16: 717–726.

Katsouyanni, K., Touloumi, G., Spix, C.,Schwartz, J., Balducci, F., Medina, S., Rossi,G., Wojtyniak, B., Sunyer, J., Bacharova, L.,Schouten, J. P., Ponka, A., and Anderson,H. R. 1997. Short term effects of ambientsulphur dioxide and particulate matter onmortality in 12 European cities: Results fromtime series data from the APHEA project. Br.Med. J. 314: 1658–1663.

Krewski, D., Burnett, R., Jerrett, M., Pope,C.A., Rainham, D., Calle, E., Thurston, G.,and Thun, M. 2005. Mortality and long-termexposure to ambient air pollution: Ongoinganalyses based on the American cancer soci-ety cohort. J. Toxicol. Environ. Health A 68:1093–1109.

Le Tertre, A., Medina, S., Samoli, E., Forsberg,B., Michelozzi, P., Boumghar, A., Boumghar,A., Vonk, J.M., Bellini, A., Atkinson, R.,Ayres, J.G., Sunyer, J., Schwartz, J.,and Katsouyanni, K. 2002. Short-termeffects of particulate air pollution oncardiovascular diseases in eight Europeancities. J. Epidemiol. Commun. Health 56:773–779.

Lee, J. T., and Schwartz, J. 1999. Reanalysis ofthe effects of air pollution on daily mortal-ity in Seoul, Korea: A case-crossover design.Environ. Health Perspect. 107: 633–636.

Levy, J. I., Hammitt, J. K., and Spengler, J.D. 2000. Estimating the mortality impacts ofparticulate matter: What can be learned frombetween-study variability? Environ. HealthPerspect. 108: 109–117.

Levy, D., Lumley, T., Sheppard, L., Kaufman, J.,and Checkoway, H. 2001. Referent selectionin case-crossover analyses of acute healtheffects of air pollution. Epidemiology 12:186–192.

Liao, D., Shaffer, M. L., He, F., Rodriguez-Colon, S., Eu, R., Whitsel, E. A., Bixler,E. O., and Cascio, W. E. 2011. Fine par-ticulate air pollution is associated with

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14

200 S.-S. TSAI AND C.-Y. YANG

higher vulnerability to atrial fibrillation—TheAPACR study. J. Toxicol. Environ. Health A74:693–705.

Linares, C., and Diaz, J. 2010a. Short-termeffect of concentrations of fine particu-late matter on hospital admissions due tocardiovascular and respiratory causes amongthe over-75 age group in Madrid, Spain.Public Health 124: 28–36.

Linares, C., and Diaz, J. 2010b. Short-termeffect of PM2.5 on daily hospital admis-sions in Madrid (2003–2005). Int. J. Environ.Health Res. 20: 129–140.

Lu, Y., and Zeger, S. 2007. On the equivalenceof case-crossover and time series methods inenvironmental epidemiology. Biostatistics 8:337–344.

Lumley, T., and Levy, D. 2000. Bias in the case-crossover design: Implications for studies ofair pollution. Environmetrics 11: 689–704.

Maclure, M. 1991. The case-crossover design:A method for studying transient effects on therisk of acute events. Am. J. Epidemiol. 133:144–153.

MacNee, W., and Donaldson, K. 2003.Mechanisms of lung injury caused by PM10and ultrafine particles with special refer-ences to COPD. Eur. Respir. J. 21(suppl. 40):47s–51s.

Marshall, R. J., and Jackson, R. T. 1993. Analysisof case-crossover designs. Stat. Med. 12:23332341.

Medina-Ramon, M., Zanobetti, A., andSchwartz, J. 2006.The effect of ozone andPM10 on hospital admissions for pneumoniaand chronic obstructive pulmonary disease:A national multicity study. Am. J. Epidemiol.163: 579–588.

Mittleman, M. A. 2005. Optimal referent selec-tion strategies in case-crossover studies: Asettled issue. Epidemiology 16: 15–16.

Mittleman, M. A., Maclure, M., and Robins, J.M. 1995. Control sampling strategies for case-crossover studies: An assessment of relativeefficiency. Am. J. Epidemiol. 142:91–98.

Neas, L. N., Schwartz, J., and Dockery, D.1999. A case-crossover analysis of air pol-lution and mortality in Philadelphia. Environ.Health Perspect. 107: 629–631.

Peng, R. D., Bell, M. L., Geyh, A. S.,McDermott, A., Zeger, S. L., Samet, J. M.,and Dominici, F. 2009. Emergency admis-sions for cardiovascular and respiratory dis-eases and the chemical composition offine particle air pollution. Environ. HealthPerspect. 117: 957–963.

Pope, C. A. 3rd, Burnett, R. T., Thurston, G. D.,Thun, M. J., Calle, E. E., Krewski, D. P., andGodleski, J. J. 2004. Cardiovascular mortal-ity and long-term exposure to particulate airpollution: Epidemiological evidence of gen-eral pathophysiological pathways of disease.Circulation 109: 71–77.

Pope, C. A., and Dockery, D. W. 2006. Healtheffects of fine particulate air pollution: Linesthat connect. J. Air Waste Manage. Assoc. 36:709–742.

Samet, J., and Krewski, D. 2007. Health effectsassociated with exposure to ambient airpollution. J. Toxicol. Environ. Health A 70:227–242.

Schwartz, J. 2004. The effects of particulate airpollution on daily deaths: A multi-city case-crossover analysis. Occup Environ Med 61:956–961.

Schwartz, J., Dockery, D. W., and Neas, L. M.1996. Is daily mortality associated specifi-cally with fine particles? J. Air Waste Manage.Assoc. 46: 927–939.

Schwartz, J., and Marcus, A. 1990. Mortalityand air pollution in London: A time seriesanalysis. Am. J. Epidemiol. 131: 185–194.

Suh, H. H., Zanobetti, A., Schwartz, J., andCoull, B. A. 2011. Chemical properties of airpollutants and cause-specific hospital admis-sions among the elderly in Atlanta, Georgia.Environ. Health Perspect. 119: 1421–1428.

Vigotti, M. A., Chiaverini, F., Biagiola, P., andRossi, G. 2007. Urban air pollution andemergency visits for respiratory complaints inPisa, Italy. J. Toxicol. Environ. Health A 70:266–269.

World Health Organization. 2006. WHOair quality guidelines. Global update2005. Geneva, Switzerland: World HealthOrganization.

Wilson, W. E., and Suh, H. H. 1997. Fineparticles and coarse particles: Concentration

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14

FINE PARTICULATE AIR POLLUTION AND PNEUMONIA 201

relationship relevant to epidemiologic stud-ies. J. Air Waste Manage. Assoc. 47:1238–1249.

Yang, C. Y. 2008. Air pollution and hospitaladmissions for congestive heart failure in asubtropical city: Taipei, Taiwan. J. Toxicol.Environ. Health A 71: 1085–1090.

Yang, C. Y., Chang, C. C., Chuang, H. Y.,Tsai, S. S., Wu, T. N., and Ho, C. K. 2004.Relationship between air pollution and dailymortality in a subtropical city: Taipei, Taiwan.Environ. Int. 30: 519–523.

Yang, C. Y., Chen, C. C., Chen, C. Y., andKuo, H. W. 2007. Air pollution and hospitaladmissions for asthma in a subtropical city:Taipei, Taiwan. J. Toxicol. Environ. Health A70: 111–117.

Zanobetti, A., Schwartz, J., and Dockery, D.W. 2000. Airborne particles are a risk fac-tor for hospital admissions for heart andlung disease. Environ. Health Perspect. 108:1071–1077.

Zanobetti, A., Franklin, M., Koutrakis, P., andSchwartz, J. 2009. Fine particulate air pol-lution and its components in associationwith cause-specific emergency admissions.Environ. Health 8: 58.

Zanobetti, A., and Schwartz, J. 2006. Air pol-lution and emergency admissions in Boston,MA. J. Epidemiol. Commun. Health 60:890–895.

Zeger, S. L., Thomas, D., Dominici, F., Samet,J. M., Schwartz, J., Dockery, D., and Cohen,A. 2000. Exposure measurement error intime-series studies of air pollution: Conceptsand consequences. Environ. Health Perspect.108: 419–426.

Zelikoff, J. T., Chen, L. C., Cohen, M. D.,Fang, K., Gordon, T., Li, Y., Nadziejko,C., and Schlesinger, R. B. 2003. Effectsof inhaled ambient particulate matter onpulmonary antimicrobial immune defense.Inhal. Toxicol. 15: 131–150.

Dow

nloa

ded

by [

Nor

th D

akot

a St

ate

Uni

vers

ity]

at 1

6:38

19

Nov

embe

r 20

14