A randomised cross-over cohort studyof exposure to emissions from a roadtunnel ventilation stack

Christine T Cowie12 Wafaa Ezz12 Wei Xuan12 William Lilley3 Nectarios Rose14

Michael Rae3 Guy B Marks1

To cite Cowie CT Ezz WXuan W et al A randomisedcross-over cohort studyof exposure to emissionsfrom a road tunnel ventilationstack BMJ Open 201200e001201 doi101136bmjopen-2012-001201

Prepublication history andadditional material for thispaper are available online Toview these files please visitthe journal online (httpdxdoiorg101136bmjopen-2012-001201)

Received 3 April 2012Accepted 12 July 2012

This final article is availablefor use under the terms ofthe Creative CommonsAttribution Non-Commercial20 Licence seehttpbmjopenbmjcom

For numbered affiliations seeend of article

Correspondence toProfessor Guy B MarksRespiratory andEnvironmental EpidemiologyWoolcock Institute of MedicalResearch PO Box M77Missenden Road PO SydneyNSW 2050 Australia guymarkssydneyeduau

ABSTRACTBackground and objective Road tunnels areincreasingly important components of urbaninfrastructure However knowledge of their healthimpact on surrounding communities is limited Ourobjective was to estimate the short-term respiratoryhealth effects of exposure to emissions from a roadtunnel ventilation stackMethods We conducted a randomised cross-overcohort study in 36 volunteers who underwent threeexposure scenarios in 2006 before the road tunnelopened and in 2007 (n=27) and 2008 (n=20) after thetunnel opened Exposure downwind of the stack wascompared to upwind of the stack and to a distant heavilytrafficked location adjacent to a main road Spirometryexhaled nitric oxide (eNO) and symptom scores weremeasured repeatedly during each 2 h exposure sessionResults Downwind locations were associated withincreased reports of lsquodry nosersquo (score difference 03695 CI 009 to 063) compared with the control location(2006 vs 20072008) but not with impaired lungfunction increased airway inflammation or othersymptoms The heavily trafficked location was associatedwith significantly increased eNO (ratio=109 95 CI 104to 114) eye (score difference 005 95 CI 001 to 010)and chest (score difference 021 95 CI 009 to 033)symptoms compared to the stack locationsConclusions There was no consistent evidence ofadverse respiratory effects from short-term exposuresdownwind of the tunnel ventilation stack except for drynose symptoms However the findings of increasedairway inflammation and symptoms in subjects after only2 h exposure at the heavily trafficked location aresuggestive of detrimental effects of short-term exposuresto traffic-related air pollution

INTRODUCTIONRoad tunnels are increasingly important com-ponents of urban infrastructure However fewstudies have investigated the adverse impacts ofexposure to emissions from road tunnel portalsor their ventilation stacks or the beneficialeffects of diverting traffic from surface roadsinto tunnels This may be due to the difficulty

in designing studies that distinguish betweenemissions from road tunnels and emissionsfrom general surrounding traffic1 2 Only onestudy has investigated the association betweenexposure to emissions from a road tunnel venti-lation stack and health effects That cross-sectional study found no evidence of increasedrespiratory or irritant symptoms in high expos-ure zones compared to low exposure zones3

ARTICLE SUMMARY

Article focus Road tunnels are being increasingly commis-

sioned in major cities around the world to allevi-ate traffic on surface roads

There are little existing data on the healthimpacts of exposure to tunnel emissions onsurrounding communities

Our objective was to estimate the short-termrespiratory health effects of exposure to emis-sions from a road tunnel ventilation stack

Key messages There was no consistent evidence of adverse

respiratory effects of being located downwind ofthe tunnel ventilation stack except for anincrease in dry nose symptoms

Short-term (2 h) exposures to a heavily traffickedcontrol site were associated with increasedairway inflammation eye and chest symptoms

The findings suggest detrimental effects ofshort-term exposure to traffic-related air pollu-tion but are not supportive of consistent adverseeffects from short-term exposures downwind ofthe tunnel ventilation stacks

Strengths and limitations of this study Strengths of the study include the randomised

cross-over cohort design and blinding of partici-pants use of a heavily trafficked site as a posi-tive control the beforeafter nature of our studythe repeated measures within and across yearsand the sensitivity analysis

A limitation of our study is the differential lossfrom the groups over time

Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201 1

Open Access Research

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

Several recent studies have reported designs that may behelpful for assessing this type of exposure Each has investi-gated the respiratory4ndash8 or cardiovascular9ndash13 effects of veryshort-term exposures (typically 2 h or less) to traffic-relatedair pollution (TRAP) and diesel emissions in real-worldsettings All are characterised by their quasi-experimentalstudy design in which participants were taken to locations(busy traffic inner city with diesel only quiet park like) orunderwent realistic commutes (in buses cars or bicycling)where the investigators expected elevated exposures orcontrasts in exposures to TRAP to occur However thesedesigns have not yet been applied to the assessment of newpoint sources such as emissions from road tunnels andtheir ventilation stacksA proposal to build a new road tunnel in Sydney with

emissions vented to the surface via ventilation stacks pre-sented us with an opportunity to prospectively test thehypothesis that short-term exposure to emissions from thetunnel ventilation stacks has an adverse effect on respira-tory health

METHODSThe study was conducted in Sydney Australia between2006 and 2008 and approved by the University of

Sydney Human Research Ethics Committee (HREC) on17 August 2006

Study design and subjectsThe study was a randomised controlled cross-over trialThirty-six volunteers were recruited through a universitystudent and workplace networks and the WoolcockInstitute of Medical Researchrsquos (WIMR) VolunteerDatabase Eligibility criteria included nonsmokers abilityto walk easily for 2 h able to participate in 2006 and2007 sense of smell not compromised and not pregnantStudy participants were randomised into three equal-

size groups which defined the order of exposure in thecross-over study Three exposure conditions were testedlsquodownwindrsquo of the LCTwestern ventilation stack (putativeadverse exposure) lsquoupwindrsquo of the stack (negativecontrol) and a lsquoheavily trafficked sitersquo (positive control)The three exposures took place over 6 days in OctoberndashNovember in 2006 before the tunnel opened andduring the same months in 2007 and 2008 after thetunnel opened to account for seasonal influences Datacollection extended to 2008 the second follow-up year asit was expected that traffic volumes in the tunnel wouldbe increased between 2007 and 2008 Traffic volumes in

Figure 1 Schematic of fieldwork

2 Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201

A randomised cross-over cohort study of exposure

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

November 2007 and 2008 were 49 218 vehicles per day(vpd) and 58 218 vpd respectively Each exposure com-prised two morning sessions during peak traffic periods(0700ndash0900) on two consecutive weekdays (Tuesdaysand Wednesdays) (figure 1) There was a 6 day washoutperiod between exposure conditions The order of groupallocation to exposure type was the same each yearexcept in 2008 where order of testing for two groups waschanged twice due to changing wind conditions

Exposure locationsThe tunnel ventilation stacks are located at its easternand western ends respectively As the eastern stack islocated close to tunnel feeder roads attribution of aneffect solely due to the stack would be difficult Hencewe decided to conduct the study at sites located aroundthe western ventilation stack using predetermined

walking routes (figure 2) The heavily trafficked (posi-tive control) site was located alongside a major urbancommuter road close to Sydneyrsquos central business districtand remote from the study area The annual averagedaily traffic volume at this site was 64 418 vehicles perday during 200514

We selected six sites in a 500ndash1000 m circumferencearound the stack and two sites on either side of themajor commuter road The stack sites were chosenbased on the feasibility of accommodating testing proce-dures and identification of 20 min walking routes whichwere all on quiet residential streets with low trafficvolumes and were not undulating Participants wereasked to walk at a comfortable even paceFor each study day two of the six stack sites were selected

and defined as lsquoupwindrsquo or lsquodownwindrsquo sites based on thewind direction recorded at 0600 that morning

Figure 2 Study area and field sites

Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201 3

A randomised cross-over cohort study of exposure

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

Baseline assessmentPrior to fieldwork we conducted spirometry measurednitric oxide concentration in exhaled breath (eNO) andperformed allergen skin prick tests (SPTs) on partici-pants in the WIMR laboratory Participants completed aquestionnaire recording asthma diagnosis wheezemedication use respiratory symptoms and odour percep-tion15ndash17 as experienced in the preceding 3 monthsParticipants also completed a series of 14 visual ana-logue scales (VAS) to rate their perception of respiratoryand chest nose throat and eye symptoms They ratedodour perception (strength and annoyance) on a VAS ifthey responded affirmatively to detecting odourSymptoms were ranked on VAS scales from lsquono symp-tomsrsquo (0) to lsquothe worst I can imaginersquo (100) These datawere used to determine pre-existing disease and to char-acterise the study population

FieldworkParticipants were either transported in a mini-van ormade their own way to the field site They were asked towear N95 mask (Alpha Protech N95 particulate facemask certified to NIOSH standards) from when they lefthome each morning until measurements were made tominimise exposure on route They were asked to with-hold long-acting bronchodilators for 12 h and short-acting bronchodilators for 6 h before fieldworkEach exposure period was started at 0700 and lasted

approximately 2 h incorporating two 20 min periods ofwalking and two rest periods (figure 1) During eachexposure subjects underwent three rounds of testingincluding completion of VAS forms eNO measurementand spirometry in that order and at timepoints as indi-cated in figure 1 Participants were allocated to pairs fortesting and walking with each pairing maintained (asmuch as possible) throughout the 3 years of testing

Clinical testsWe used a hand-held spirometer (QRS DiagnosticZ-5000-2668) linked to Office Medic software (V 45i) torecord forced expiratory volume in 1 s (FEV1) andforced vital capacity (FVC) Spirometry was performedaccording to the American Thoracic SocietyEuropeanRespiratory Society (ATSERS) criteria with technicallyunsatisfactory curves identified at the time of testing andagain after fieldwork These technically unsatisfactorycurves were excluded from analysis To improve reliabil-ity of the measures we chose the highest of two valuesreproducible within 150 ml for FEV1 and FVC Wherenot reproducible curves were rechecked for quality andif acceptable the highest measures for FEV1 and FVCwere selectedWe used an off-line technique to measure eNO

according to ATSERS criteria18 at an expiratory flowrate of 200 mls19 Exhaled breath was collected into a3-litre chemically inert bag through a rotameter incorp-orating a filter and an NO scrubber on the inlet valve toexclude the influence of ambient NO levels on personal

NO concentrations The NO concentration was mea-sured within 24 h of collection using a chemilumines-cence analyser (ThermoEnvironmental 42degC)SPTs were performed using airborne allergens for

house dust mites (Dermatophagoides pteronyssinusD farinae) mould (Alternaria Aspergillus) cat pelt dogcockroach ryegrass and grass mix (Hollister-Stier)20

Wheal sizes were measured as the average of the largestdiameter and its perpendicular at 15 min Wheals largerthan 3 mm in diameter and larger than the negativecontrol were considered positive Subjects with any posi-tive SPTwere classified as atopic

Air quality monitoringAir quality monitoring took place on each field testingday at each visited site constituting a total of 18 meas-urement days We collected samples for particulatematter (PM10 PM25 and PM1) volatile organic com-pounds (VOCSs) total oxides of nitrogen (NOx) nitro-gen dioxide (NO2) and nitrogen oxide (NO)We used Osiris instruments to collect PM data in 2006

and 2007 and Topas units for collection of PM data in2008 due to the unavailability of Osiris instrumentsConcentrations of PM data were recorded as 2 minaveragesVOC samples were collected using a stainless-steel canis-

ter and inlet designed to collect an evenly spaced sampleover the 2 h of fieldwork providing an average readingThe samples were analysed at the CSIRO Laboratory in2006 and 2008 for 13-butadiene benzene toluene ethyl-benzene m-+p-xylene o-xylene Samples were analysed byan alternative laboratory in 2007 but the limit of detec-tion (DL) was higher and given that many of the readingswere below this DL we excluded the 2007 dataSamples for ambient NOxNO2 were collected via

grab samples every 15 min at each site using identicalcollection bags as those used for eNO collection Thesewere analysed (within 24 h) as for eNO analysisWe obtained continuously recorded data for 2006ndash2007

for NO2 PM10 and PM25 from two air quality monitorssited in elevated positions near the ventilation stacks andestablished by the tunnel proponents for statutorypurposes21

Wind directionWind direction for each field day was determined usingmeteorological forecasts from the Bureau of Meteorologyand by releasing slightly buoyant balloons filled with amixture of air and helium from an elevated location nearthe western stackTo validate our choice of testing sites wind speed and

direction data were obtained from three air quality mon-itoring stations located near the western ventilationstack including the elevated site mentioned above Theother two were ground-level stations operated for allthree study years We constructed wind roses for thethree monitoring sites using Lakes EnvironmentalWRPlot View (V651)

4 Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201

A randomised cross-over cohort study of exposure

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

Post hoc validation of assignment of downwind andupwind study sites for each testing day was checkedusing wind rose data Where the chosen locations werewithin plusmn45deg of the measured downwind or upwind direc-tion they were deemed to be accurate downwind orupwind locations respectively This validation was usedin the sensitivity analysis Data for the heavily traffickedsite (north or south side of the road) were included inthe sensitivity analysis in the same way Data were alsoexcluded where wind speed was lt1 ms

AnalysisHealth outcomesThe distributions of eNO were skewed and so thesevalues were log-transformed All other dependent vari-ables were analysed without transformationPrincipal components analysis (PCA) was conducted to

reduce the 14 symptom scores to five factors which werefound to explain 77 of the overall variance The screeplot suggested that this was the optimum number offactors Hence a five-factor structure was fitted withvarimax rotation using data for all 3 years The five-factorscores were output for each observation and these wereused as continuous measures of symptom intensity Thefive symptom factors were (1) itchydrysore eyes andblocked nose (EYE) (2) soredry throat sore chest andrunny nose (THROAT) (3) chest tightness difficultybreathing (CHEST) (4) sneezing cough watery eyes(SNEEZE) and (5) dry-nose symptoms (DRY NOSE) Thedistributions of VAS scores for lsquostrong smellrsquo and lsquoannoyingsmellrsquo were highly skewed and included a large proportionof zero values and so were treated as binary outcomeszero (lsquonorsquo) and greater than zero (lsquoyesrsquo)To estimate the air pollutant means we first calculated

the daily means for each pollutant and then calculatedthe mean pollutant level by location and yearWe analysed the data using an lsquointention to treatrsquo

approach based on the locations chosen for each groupfor each field-testing session Data for timepoint_0 wereexcluded from the main analyses based on the assump-tion that exposure had not started when that measure-ment was made We used a mixed effects linearregression model in which the fixed effects were loca-tion (upwind downwind control) and year (2006 20072008) and used contrast statements to test specific inter-actions between them We also used contrast statementsto estimate the overall difference (averaged over allyears) in effect between the heavily trafficked locationand the upwind and downwind locations combinedSince we expected that health outcomes would be worseat the heavily trafficked site this served as a validation ofthe design We included a random intercept term forsubjects with a variance components covariance struc-ture to account for repeated measures within subjectsSpirometric outcomes eNO and the five symptomfactors were separately included as outcome variablesSignificance of the fixed effects was tested using F testsLeast squares means were estimated for each

combination of year and location Analyses were con-ducted using SAS V92 PROC MIXED proceduresThe binary variables lsquostrong smellrsquo and lsquoannoying smellrsquo

were treated identically except that a generalised linearmixed model was fitted with Proc GLIMMIX using theadaptive quadrature estimation method a logistic link anda binomial error distribution in SAS V92 Examination ofthe distribution of residuals indicated that there were nomajor departures from normality and transformation ofthe data was not required

Air pollutant concentrationsWe used a similar mixed effects linear regression modelwith contrast statements to test the effects of location andyear and their interaction on air pollutant measurements(NOx NO2 PM10 PM25 and PM1) excluding VOCswhere there were too few readings A random interceptterm for days was used to account for repeated measuresWe conducted a sensitivity analysis for the air quality databy re-fitting the models using data that were validatedusing the wind-rose method as described earlierFor the elevated monitor data we used an autoregres-

sive error model with Yule-Walker estimates and stepwiseregression to determine whether there was a differencein air quality between 2006 (pretunnel) and 2007 (post-tunnel opening) by comparing daily averages for the2 years We adjusted this analysis for changes in regionalair quality measured at three regulatory monitoring sta-tions operated by the Department of EnvironmentClimate Change and Water (DECCW) including as a cov-ariate the change in average daily readings from theDECCW sites combined We tested up to 20 lags residualswere plotted to check for normality and the autocorrel-ation function and partial autocorrelation function werechecked for autocorrelation These models were imple-mented in SAS 92 using the AUTOREG procedure

Sample size and power calculationThe sample size for this study (36 subjects) was selectedby reference to a study conducted in London that wasable to detect significant within-subject differences inlung function in 60 subjects exposed to a high pollutionarea and a low pollution area5 noting that our studyhad a greater number of repeated measures

RESULTSIn 2006 25 of the 36 participants reported having hada diagnosis of asthma and 42 reported having hadwheeze including 25 who had wheeze in the last3 months (table 1) The majority of participants wereatopic Of the 36 subjects who started the study in 200626 (72) participated in 2007 and 20 (56) in 2008While the cohort characteristics remained relativelystable between 2006 and 2007 in 2008 the remainingcohort was less symptomatic in the 3 months prior to2008 field testing (table 1 supplementary material)

Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201 5

A randomised cross-over cohort study of exposure

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

Health outcomesFEV1 and FVC were similar across locations but were gen-erally higher in 2006 compared to 2007 or 2008 (table 2)Mean values for eNO were highest at the heavily traf-ficked (positive control) location for all years except for2008 and did not differ between the downwind andupwind locations EYE THROAT and CHEST symptomscores were highest at the heavily trafficked location com-pared to the downwind or upwind locations during allyears with one exception (table 2)There was little evidence that the downwind locations

when compared with the heavily trafficked or upwind loca-tions were associated with adverse health outcomes in2007 or 2008 compared to 2006 (table 3) The only excep-tion was that changes in DRY NOSE scores were higher fordownwind versus heavily trafficked locations Downwindlocations were associated with a decrease in reporting ofchest symptoms for almost all year comparisonsOver the 3-year period as a whole exposure at the

heavily trafficked locations compared with both thedownwind and upwind stack locations was associatedwith more airway inflammation (eNO) (ratio=109) and

Table

2Mean(SD)valuesforlungfunctionandsymptom

scoresbylocationandyear

Location

Year

Noof

obs

FEV1(L)

FVC

(L)

eNO

(ppb)

(median)

EYEsymptoms

scoredagger

THROAT

symptoms

scoredagger

CHEST

symptoms

scoredagger

SNEEZE

symptoms

scoredagger

DRYNOSE

symptoms

scoredagger

Heavily

trafficked

Average

463

272(080)364(100)

176

(130)

minus0048(0714)

0052(1187)

0120(1098)

0351(1590)

minus0158(1076)

2006

208

284(081)370(100)

152

(121)

minus0022(0881)

0120(1146)

0095(1120)

0059(1153)

0086(1146)

2007

156

267(083)362(101)2434(144)

minus0054(0587)

0108(1305)

0137(1084)

0547(1929)

minus0409(1003)

2008

99

258(069)353(089)

122

(66)

minus0092(0467)

minus0182(1049)

0143(1082)

0651(1692)

minus0268(0923)

Downwind

Average

461

268(079)362(099)

152

(109)

minus0033(0840)

minus0012(0962)

minus0071(0874)

0365(1867)

minus0248(1025)

2006

207

281(084)373(105)

128

(95)

0033(1073)

minus0031(0873)

minus0054(0907)

minus0005(0867)

minus0139(0933)

2007

143

257(073)353(095)

213

(129)

minus0139(0414)

0010(1043)

minus0023(0720)

0663(2422)

minus0264(1155)

2008

111

259(074)355(090)

125

(75)

minus0024(0728)

minus0002(1024)

minus0163(0984)

0688(2289)

minus0432(0997)

Upwind

Average

447

268(081)361(095)

157

(114)

minus0042(0801)

minus0065(0980)

minus0015(1045)

0159(1264)

minus0113(1046)

2006

204

279(083)369(097)

135

(103)

minus0013(1039)

minus0090(0958)

minus0040(0961)

minus0056(0964)

0060(0891)

2007

138

263(082)357(096)

209

(132)

minus0048(0545)

0022(1141)

0020(1317)

0409(1529)

minus0161(1313)

2008

105

254(073)350(088)

127

(76)

minus0087(0544)

minus0135(0757)

minus0015(0745)

0220(1305)

minus0374(0834)

NumberofobservationsforFEV1providedasobservationsvariedslightlywithhealthoutcomesmeasured

daggerEYEsymptomsitchydrysore

eyesandblockednoseTHROATsymptomssoredry

throatsore

chestamprunnynoseCHESTsymptomschesttightnessdifficultybreathingSNEEZE

symptomssneezingcoughwatery

eyesDRYNOSEsymptomsdry

nose

eNOexhalednitricoxideFEV1forcedexpiratory

volumein

1sFVCforcedvitalcapacity

Table 1 Baseline characteristics as measured at

recruitment in 2006

Symptomcharacteristic N ()

Total number of subjects 36 (100)

Asthma (ever diagnosis) 9 (25)

Ever wheezed 15 (42)

Wheeze (in last 3 months) 9 (25)

Breathless (in last 3 months) 4 (11)

Cough (in last 3 months) 20 (56)

Activities limited (in last 3 months) 3 (8)

Ever taken asthma medication 15 (42)

Asthma medication (in last 3 months) 8 (22)

Inhaled steroids (in last 3 months) 6 (17)

Sensitive eyesdagger 12 (33)

Eye symptoms (in the last 3 months without

a coldflu)Dagger

14 (39)

Nose symptoms (in the last 3 months

without a coldflu)

18 (50)

Throat symptoms (in the last 3 months

without a coldflu)

10 (28)

Mouth symptoms (in the last 3 months

(soreness of gums or teeth))

13 (36)

Atopicsect 26 (72)

Female 25 (69)

Spirometry and chemical sensitivity score Mean (SD)

FEV1 (L) 299 (094)

FVC (L) 378 (111)

eNO (ppb) 111 (121)

Chemical sensitivity scale score (CSS)17 363 (87)

Inhaled steroids=having taken either seretide or pulmicortdaggerQuestion Do you usually regard your eyes as being sensitive tocigarette smoke smog air conditioning or central heatingDaggerQuestion Have you had any of the following eye symptoms inthe last 3 months when you did not have a cold or flu (eyeredness burning feeling scratchiness grittiness dryness orwatery eyes)sectAny allergen SPTge3 mm

6 Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201

A randomised cross-over cohort study of exposure

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

Table 3 Contrasts in lung function eNO and symptom scores between the downwind site and the other two sites (upwind and heavily trafficked) adjusted for change from baseline (2006)

to subsequent years

Contrast FEV1 (diff in litres)

(95 CI)

FVC (diff in litres)

(95 CI) eNO ratio (95 CI)

EYE symptoms

(diff in score) (95 CI)

THROAT symptoms

(diff in score) (95 CI)Location Year

Downwind versus

(heavily trafficked and

upwind)dagger

20072008 vs 2006Dagger 001 (minus004 to 006) 003 (minus003 to 009) 103 (094 to 112) minus0006(minus0097 to 0086)

0130 (minus0084ndash0344)

2007 vs 2006 003 (minus003 to 009) 005 (minus003 to 013) 101 (091 to 111) minus0055(minus0159 to 0049)

0041 (minus0220ndash0301)

2008 vs 2006 minus0004 (minus007 to 006) 002 (minus007 to 010) 105 (094 to 117) 0043 (minus0074 to 0159) 0207 (minus0027 to 0441)

Downwind versus

upwind

20072008 vs 2006Dagger 002 (minus004 to 008) 006 (minus002 to 013) 103 (093 to 114) minus0044(minus0151 to 0063)

0080 (minus0169 to 0329)

2007 vs 2006 004 (minus003 to 011) 006 (minus003 to 015) 105 (094 to 117) minus0108(minus0229 to 0012)

minus0006 (minus0308 to 0297)

2008 vs 2006 0004 (minus007 to 008) 006 (minus004 to 015) 101 (089 to 115) 0015 (minus0120 to 0149) 0179 (minus0092 to 0449)

Downwind versus

heavily trafficked

20072008 vs 2006Dagger minus000 (minus006 to 006) 0004 (minus007 to 008) 103 (093 to 114) 0033 (minus0073 to 0140) 0180 (minus0068 to 0428)

2007 vs 2006 002 (minus005 to 009) 004 (minus005 to 012) 097 (087 to 108) minus0002(minus0121 to 0117)

0087 (minus0212 to 0385)

2008 vs 2006 minus001 (minus009 to 006) minus003 (minus012 to 007) 109 (096 to 123) 0071 (minus0065 to 0206) 0236 (minus0038 to 0509)

Heavily trafficked

versus (upwind and

downwind)sect

Overall minus001 (minus004 to 001) minus003 (minus006 to 0002) 109 (104 to 114) 0054 (0006 to 0102) 0107 (minus0005 to 0218)

Downwind versus

(heavily trafficked and

upwind)dagger

20072008 vs 2006Dagger minus0080(minus0302 to 0142)

0288

(minus0139 to 0716)

0276 (0040 to 0511) 345 (157 to 760) 169 (072 to 396)

2007 vs 2006 minus0031(minus0277 to 0214)

0375

(minus0093 to 0844)

0320 (0053ndash0587) 572 (235 to 1391) 195 (077 to 494)

2008 vs 2006 minus0133(minus0409 to 0143)

0182

(minus0254 to 0618)

0229

(minus0057 to 0516)

214 (077 to 597) 133 (044 to 400)

Downwind versus

upwind

20072008 vs 2006Dagger minus0017(minus0275 to 0242)

0430

(minus0067 to 0927)

0193

(minus0081 to 0466)

351 (140 to 880) 123 (045 to 340)

2007 vs 2006 0033

(minus0252 to 0318)

0422

(minus0123 to 0966)

0187

(minus0124 to 0498)

411 (148 to 1143) 112 (038 to 337)

2008 vs 2006 minus0042(minus0362 to 0277)

0424

(ndash0080 to 0929)

0178

(minus0154 to 0510)

331 (098 to 1118) 126 (034 to 471)

Downwind versus

heavily trafficked

20072008 vs 2006Dagger minus0143(minus0400 to 0114)

0147

(minus0349 to 0642)

0358 (0086 to 0631) 339 (139 to 829) 231 (090 to 594)

2007 vs 2006 minus0096(minus0377 to 0185)

0329

(minus0208 to 0866)

0453 (0146 to 0759) 796 (289 to 2193) 338 (120 to 954)

2008 vs 2006 minus0224(minus0547 to 0099)

minus0060(minus0570ndash0450)

0281

(minus0054 to 0617)

138 (044 to 440) 140 (041 to 475)

Heavily trafficked

versus (upwind and

downwind)sect

Overall 0209

(0094 to 0325)

0150

(minus0072 to 0372)

0022 (minus0101 to 0144) 234 (158 to 349) 408 (267 to 625)

Contrasts are derived from mixed effects regression model in which location and year and the interaction between them are fixed effects and individual subject intercepts are a randomeffect Main contrasts of interest are those included in the table and consist of downwind versus other exposures as the downwind exposure is the primary exposure of interest in this studydaggerDownwind versus (heavily trafficked and upwind) is the comparison of downwind versus heavily trafficked and upwind exposures combinedDagger20072008 vs 2006 represents the comparison of both post-tunnel years combined (20072008) versus the pre-tunnel year (2006)sectRepresents the overall comparison of the heavily trafficked site compared to the upwind and downwind sites combined (data from all years included) as we anticipated the heavilytrafficked site to have higher air pollution levels and that health effects would be more prominenteNO exhaled nitric oxide FEV1 forced expiratory volume in 1 s FVC forced vital capacity

CowieCTEzz

WXuan

WetalBM

JOpen

201200e001201doi101136bmjopen-2012-001201

7

Ara

ndomisedcro

ss-o

vercohort

stu

dyofexposure

on May 27 2020 by guest Protected by copyright httpbmjopenbmjcom BMJ Open first published as 101136bmjopen-2012-001201 on 17 August 2012 Downloaded from

higher scores for EYE CHEST and THROAT symptoms(last row table 3) These findings support the validity ofthe study design and analysisThe change in lsquosmell_strongrsquo from 2006 to subsequent

years was higher for downwind locations than the othertwo locations (table 3) However in general this patternwas not observed for lsquosmell-annoyingrsquo Overall bothlsquosmell-strongrsquo and lsquosmell-annoyingrsquo were higher at theheavily trafficked locations than the two stack locations(last row table 3)

Sensitivity analysis for health outcomesIn the sensitivity analysis we excluded data that mighthave led to exposure misclassification due to unexpected

wind direction as described earlier Data for five lsquodown-windrsquo sites in 2006 and two in 2008 for two lsquoupwindrsquosites in 2006 and two in 2008 and for one heavily traf-ficked site in 2006 were excludedIn this analysis FVC was increased for downwind loca-

tions versus the other two locations for most year compari-sons (table 2 supplementary material) FEV1 was alsoincreased for downwind versus heavily trafficked locationsConversely eNO was higher at downwind compared toupwind locations for 2007 vs 2006 The increasedodds of lsquosmell-strongrsquo at downwind locations did notdiffer to the other locations in the sensitivity analysisOverall eNO levels CHEST symptoms lsquosmell-strongrsquo andlsquosmell-annoyingrsquo were higher at the heavily trafficked

Table 4 Mean pollutant levels by location (2006ndash2008 overall)

Pollutant Location Sample days Samples Mean Median IQR Min Max

Nitrogen oxidesdagger

NOx Heavily trafficked 18 157 1262 952 281ndash1331 79 6507

Downwind 18 159 239 162 80ndash385 35 663

Upwind 18 150 262 158 86ndash328 36 844

NO2 Heavily trafficked 18 160 61 38 12ndash57 minus23 278

Downwind 18 159 31 23 12ndash42 08 94

Upwind 18 151 31 18 14ndash45 06 104

Particulate matterDagger

PM10 Heavily trafficked 17 1034 526 451 354ndash642 283 1085

Downwind 17 1099 419 335 297ndash499 171 904

Upwind 17 1087 459 352 254ndash495 167 1328

PM25 Heavily trafficked 17 1034 130 139 96ndash165 42 200

Downwind 17 1099 110 104 79ndash135 15 197

Upwind 17 1087 117 106 92ndash134 17 285

PM1 Heavily trafficked 17 1034 24 25 19ndash30 11 34

Downwind 17 1099 19 19 15ndash23 03 32

Upwind 17 1087 20 20 15ndash24 05 51

VOCssect

Benzene Heavily trafficked 12 12 086 086 044ndash107 035 17

Downwind 12 12 039 037 018ndash055 016 075

Upwind 12 12 037 027 023ndash049 013 087

Butadiene Heavily trafficked 12 12 025 025 011ndash033 004 059

Downwind 12 12 011 011 004ndash018 004 024

Upwind 12 12 011 009 007ndash016 004 030

Toluene Heavily trafficked 10 10 525 460 260ndash680 150 1080

Downwind 10 10 261 255 220ndash300 140 390

Upwind 10 10 358 355 230ndash400 120 700

Ethyl-benzene Heavily trafficked 12 12 066 059 029ndash102 015 170

Downwind 12 12 045 038 021ndash060 009 150

Upwind 12 12 051 036 022ndash046 009 190

mp-xylene Heavily trafficked 12 12 243 170 081ndash331 055 770

Downwind 12 12 125 135 049ndash170 028 280

Upwind 12 12 122 088 046ndash160 030 400

o-xylene Heavily trafficked 12 12 090 070 034ndash123 022 270

Downwind 12 12 050 052 019ndash077 013 110

Upwind 12 12 049 037 021ndash064 010 160

Means and medians calculated by taking the mean of daily averages at each site IQR minimum and maximum values using all (notaveraged) datadaggerGrab sample for NO2 and NOx collected at each site every 15 min and averaged per 2 h exposure periodDaggerPM data collected every 2 minsectOne sample collected at site providing an average VOC reading per 2 h exposure period 2007 data not used in analysis due to differentdetection limits used by different laboratory 95th percentile not reported for VOCs as sample size too small 95th percentile equivalent tomaximum readingPM particulate matter NOx oxides of nitrogen

8 Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201

A randomised cross-over cohort study of exposure

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

locations compared to upwind or downwind locationsconfirming the findings of the intention-to-treat analysis(table 2 supplementary material)

Air pollutant levelsAs expected the concentrations of all pollutants werehigher at the heavily trafficked sites than the upwind ordownwind sites However there was little difference inmean pollutant levels between the downwind andupwind locations around the ventilation stack (tables 3and 4 supplementary material) NOx and NO2 ratioswere lower for the downwind sites than the other twosites for 2007 vs 2006 In contrast the NO2 ratio for2008 vs 2006 was higher at the downwind sites than atthe heavily trafficked sites (table 5) The change in PM(for all size fractions) from 2006 to subsequent years didnot vary by site Overall the change in pollutants wasgreatest at the heavily trafficked sites compared to theupwind and downwind sites combinedIn the sensitivity analysis the pollutant ratios for down-

wind locations compared to upwind locations increasedfor most pollutants supporting the validity of the exclu-sion criteria (table 4 supplementary material)NO2 and PM10 decreased at the western and eastern

elevated monitoring sites between 2006 and 2007 afteradjusting for changes in regional air quality (table 6)

DISCUSSIONWe did not detect any consistent evidence of detrimen-tal effects on lung function airway inflammation orrespiratory symptoms attributable to short-term exposureto emissions from the tunnel ventilation stack with theexception of increased reporting of dry nose symptomsThis absence of any adverse health effect might beexplained by the fact that we did not measure higherpollutant concentrations at the downwind stack sitesafter the tunnel opened The finding of increased pollu-tant levels and increased symptoms and airway inflam-mation during periods of exposure at the heavilytrafficked site supports previous evidence of the detri-mental effects of short-term exposures to TRAPThis study relied on the proposition that the downwind

sites we selected were in fact exposed to emissions fromthe tunnel ventilation stack during peak traffic periodsFor a downwind site to be impacted by the emissionplume from the stack the plume needed to be mixed atground level prior to reaching the site and subjected toconsistent wind direction during the sampling periodWe used weather balloons early on each study morning inorder to select the most appropriate downwind siteCompared to the upwind locations pollutant concentra-tions at the downwind locations were similar or some-times slightly lower This may have reflected actualconditions in the field suggesting that there was littlemeasurable impact of the plume on the pollutants mea-sured However this finding may have also arisen due toinaccurate prediction of the ground-level location of the

Table

5Ratiosofpollutantlevelscontrastingthedownwindsiteandtheothertwosites(upwindandheavily

trafficked)adjustedforchangefrom

baseline(2006)to

subsequentyears

Contrast

NOxratio(95

CI)

NO

2ratio(95

CI)

PM

10ratio(95

CI)

PM

25ratio95

CI)

PM

1ratio(95

CI)

Location

Year

Downwindversus(heavily

trafficked

andupwind)dagger

20072008vs2006Dagger

078(061to

099)

077(059to

100)

110(066to

181)

109(055to

216)

106(058to

193)

2007vs2006

059(044to

080)

049(038to

063)

118(061to

228)

104(042to

258)

098(044to

217)

2008vs2006

102(077to

135)

120(088to

164)

102(056to

185)

114(067to

194)

114(070to

187)

Downwindversusupwind

20072008vs2006Dagger

084(064to

112)

063(046to

086)

116(065to

207)

110(050to

241)

104(052to

207)

2007vs2006

072(051to

102)

052(039to

071)

116(054to

248)

109(038to

309)

102(041to

255)

2008vs2006

099(071to

137)

076(053to

109)

116(058to

230)

111(060to

204)

106(060to

187)

Downwindversusheavily

trafficked

20072008vs2006Dagger

072(055to

094)

093(068to

127)

104(058to

186)

109(050to

239)

108(054to

216)

2007vs2006

049(035to

069)

046(034to

061)

120(056to

257)

100(035to

285)

094(038to

236)

2008vs2006

105(076to

145)

190(133to

272)

089(045to

178)

118(064to

218)

124(070to

219)

Heavily

traffickedversus(upwind

anddownwind)sect

Overall

408(364to

457)

145(128to

165)

127(100to

162)

121(087to

168)

131(099to

175)

Ratiosare

derivedfrom

amixedeffects

regressionmodelin

whichlocationandyearandtheinteractionbetweenthem

are

fixedeffectsandindividualintercepts

are

arandom

effectMain

contrasts

ofinterestare

thoseincludedin

thetable

andconsistofdownwindversusotherexposuresasthedownwindexposure

istheprimary

exposure

ofinterestin

this

study

daggerDownwindversus(heavily

traffickedandupwind)isthecomparisonofdownwindversusheavily

traffickedandupwindexposurescombined

Dagger20072008vs2006represents

thecomparisonofboth

post-tunnelyears

combined(20072008)versusthepre-tunnelyear(2006)

sectRepresents

theoverallcomparisonoftheheavily

traffickedsitecomparedto

thetwostacksites(downwindandupwind)combined(data

from

allyears

included)

NOxoxidesofnitrogenPMparticulate

matter

Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201 9

A randomised cross-over cohort study of exposure

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

plume Nevertheless the sensitivity analysis we conductedwas designed to address this potential weakness The find-ings of that analysis support the original analysisOur finding that eNO an indicator of airway inflam-

mation was increased at the heavily trafficked site is con-sistent with evidence of increased airway inflammationafter exposure to TRAP (or diesel) in previous studiesThese studies have varied in their measured end-pointsbut include lsquoreal-worldrsquo exposure studies investigatingneutrophilic airway inflammation4 5 associations betweeneNO and PM25 black carbon and fine particle counts6

and chamber studies reporting increased neutrophilicairway inflammation22ndash25 increases in pro-inflammatorycytokines interleukin (IL)-626 and IL-825 increasedexhaled carbon monoxide (CO)22 and increased airwayresistance26 27

The absence of effects on spirometric function evenat the heavily trafficked site is consistent with findings ofchamber studies of diesel exhaust22 24ndash27 high particleconcentrations28 and NO2

29 an in-tunnel study30 and acommuter study8 However our findings are inconsistentwith the London study of diesel exhaust exposureswhich demonstrated small reductions in FEV1 after 1ndash2 h exposures5

Strengths of the study include the randomised cross-over design and blinding of participants by usingupwind and downwind stack sites It is unlikely that theparticipants were aware of the upwinddownwind statusof their location on any given day The study was furtherstrengthened by including the heavily trafficked sites asa positive control The finding that pollutant levels werehigher and that subjects had greater airway inflamma-tion eye and chest symptom scores at this site supportthe validity of the study methodology The beforeafternature of our study and repeated measures within andacross years also strengthened the study designA limitation of our study is the differential loss from

the groups over time Common reasons for nonpartici-pation in follow-up years were travel overseas residence

relocation and changing work practices all factors unre-lated to the experience of study participationIn conclusion while we found that this ventilation stack

under current operating conditions had little measurableshort-term health impact we did demonstrate adversehealth effects attributable to short-term exposures to TRAPat heavily trafficked locations These findings support evi-dence from existing studies of the adverse health effectsattributable to short-term exposures to TRAP

Author affiliations1Woolcock Institute of Medical Research Sydney New South Wales Australia2Cooperative Research Centre for Asthma and Airways Sydney Australia3Commonwealth Scientific and Industrial Research Organisation NewcastleAustralia4NSW Health Department Sydney New South Wales Australia

Acknowledgements We are grateful to the volunteers for their participation inthis study Professor Bert Brunekreef Vicky Sheppeard and Michael Staff gaveinitial advice on the study design We acknowledge the invaluable assistance ofthe following research staff Brett Toelle Elena Belousova AdrianaCortes-Waterman Kitty Ng Angelica Ordonez Paula Garay Adrian ForeroJennifer Manning Renate Thielbeer The study was overseen by a ScientificAdvisory Group established by the NSW Health Department which providedadvice on study design methodology analysis and interpretation It consistedof Professor Bruce Armstrong Professor Judy Simpson Associate ProfessorBin Jalaludin Dr Peter Lewis and Dr Gavin Fisher

Contributors CC organised and participated in all aspects of the studyincluding study design ethics clearances recruiting subjects field workstatistical analysis and led manuscript preparation WE was involved in all fieldwork WX contributed to study design and data analysis WL conducted airquality exposure assessment measurements in the field and provided advice onair quality measurement NR contributed biostatistical expertise MR conductedair quality measurements in the field and wind rose analyses GM participatedin the study design statistical analysis and provided general oversight of thestudy All authors contributed to the preparation of the manuscript

Funding The study was funded by the CRC for Asthma amp Airways and theNSW Health Department

Competing interests Nectarios Rose was employed with the NSW HealthDepartment on the Biostatistics Training Scheme at the time the study wasconducted

Patient consent The presented data are anonymised and risk of identificationis low

Table 6 Mean pollutant concentrations for elevated monitors by year and difference in pollutant concentration from

post-tunnel year compared to pre-tunnel year

Site and pollutant

Mean pollutant levels (SD) Difference in pollutant levels 2007 versus 2006

2006 2007 Unadjusted estimate (95 CI) Adjusted estimate (95 CI)dagger

Elevated site near western ventilation stack

NO2 (ppb) 134 (56) 121 (45) minus193 (minus306 minus080) minus106 (minus167 minus044)NOx (ppb) 261 (199) 206 (146) minus597 (minus1120 minus074) minus274 (minus442 to minus106)PM10 (μgm

3) 187 (74) 162 (70) minus240 (minus423 minus057) minus045 (minus195 104)PM25 (μgm3) 62 (35) 53 (33) minus101 (minus227 025) minus020 (minus066 027)

Elevated site near eastern ventilation stack

NO2 (ppb) 167 (70) 157 (53) minus250 (minus441 minus060) minus099 (minus221 023)NOx (ppb) 307 (178) 271 (141) minus628 (minus1093 minus162) minus205 (minus291 minus120)PM10 (μgm

3) 202 (81) 161 (68) minus402 (minus624 minus180) minus175 (minus305 minus044)PM25 (μgm3) 64 (36) 53 (32) minus114 (minus240 011) minus026 (minus065 012)

Represents the change in pollutant concentration between pre-tunnel (25 March 2006ndash24 March 2007) and post-tunnel(25 March 2007ndash24 March 2008)daggerAdjusted for change in air quality at the three regional sitesNOx oxides of nitrogen PM particulate matter

10 Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201

A randomised cross-over cohort study of exposure

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

Ethics approval University of Sydney Human Research Ethics Committee

Provenance and peer review Not commissioned externally peer reviewed

Data sharing statement No additional data are available

REFERENCES1 NHMRC Air quality in and around traffic tunnels Final Report 2008

Commonwealth of Australia ISBN online 1864963573 httpwwwnhmrcgovauguidelinespublicationseh42 (accessed 25 Feb 2011)

2 Kuykendall JR Shaw SL Paustenbach D et al Chemicals presentin automobile traffic tunnels and the possible community healthhazards a review of the literature Inhal Toxicol 200921747ndash92

3 Capon A Sheppeard V Irvine K et al Investigating health effects ina community surrounding a road tunnel stackmdasha cross sectionalstudy Environ Health 2008710

4 Jacobs L Nawrot T de Geus B et al Subclinical responses inhealthy cyclists briefly exposed to traffic-related air pollution anintervention study Environ Health 2010964

5 McCreanor J Cullinan P Nieuwenhuijsen MJ et al Respiratoryeffects of exposure to diesel traffic in persons with asthma N EnglJ Med 20073572348ndash58

6 Adar SD Adamkiewicz G Gold DR et al Ambient andmicroenvironmental particles and exhaled nitric oxide before andafter a group bus trip Environ Health Perspect 2007115507ndash12

7 Strak M Boogaard H Meliefste K et al Respiratory health effects ofultrafine and fine particle exposure in cyclists Occup Environ Med201067118ndash24

8 Zuurbier M Hoek G Oldenwening M et al Respiratory effects ofcommutersrsquo exposure to air pollution in traffic Epidemiology201122219ndash27

9 Adar SD Gold DR Coull BA et al Focused exposures to airbornetraffic particles and heart rate variability in the elderly Epidemiology20071895ndash103

10 Dales R Liu L Szyszkowicz M et al Particulate air pollution andvascular reactivity the bus stop study Int Arch Occup EnvironHealth 200781159ndash64

11 Laumbach RJ Rich DQ Gandhi S et al acute changes in heartrate variability in subjects with diabetes following a highway trafficexposure J Occup Environ Med 201052324ndash31

12 Zuurbier M Hoek G Oldenwening M et al In-traffic air pollutionexposure and CC16 blood coagulation and inflammation markers inhealthy adults Environ Health Perspect 20111191384ndash9

13 Weichenthal S Kulka R Dubeau A et al Traffic-related airpollution and acute changes in heart rate variability and respiratoryfunction in urban cyclists Environ Health Perspect 20111191373ndash8

14 RTA Annual Average Daily Traffic (AADT) data Sydney Region2005 httpwwwrtanswgovautrafficinformationdownloadsaadtdata_dl1html (accessed 16 Dec 2010)

15 ISAAC Worldwide variation in the prevalence of symptoms ofasthma allergic rhinoconjunctivitis and atopic eczema ISAACLancet 19983511225ndash32

16 Toelle B Ng K Belousova E et al The prevalence of asthma andallergy in schoolchildren in Belmont Australia three cross sectionalsurveys over 20 years BMJ 2004328386ndash7

17 Nordin S Millqvist E Lowhagen O et al A short chemical sensitivityScale for assessment of airway sensory hyperreactivity Int ArchOccup Environ Health 200477249ndash54

18 American Thoracic Society ATSERS recommendations for thestandardization procedures for the online and offline measurementof exhaled lower respiratory nitric oxide and nasal nitric oxide 2005Am J Respir Crit Care Med 2005171912ndash30

19 Salome C Roberts A Brown N et al Exhaled nitric oxide in apopulation sample of young adults Am J Respir Crit Care Med1999159911ndash16

20 Peat J Mihrshahi S Kemp A et al Three-year outcomes of dietaryfatty acid modification and house dust mite reduction in theChildhood Asthma Prevention Study J Allergy Clin Immunol2004114807ndash13

21 Team Ferrari Environmental Audit of ambient air monitoringLane cove tunnel 2006 November 2006 NSW Australia

22 Nightingale JA Maggs R Cullinan P et al Airway inflammation aftercontrolled exposure to diesel exhaust particulates Am J Respir CritCare Med 2000162161ndash6

23 Nordenhall C Pourazar J Blomberg A et al Airway inflammationfollowing exposure to diesel exhaust a study of time kinetics usinginduced sputum Eur Respir J 2000151046ndash51

24 Salvi S Blomberg A Rudell B et al Acute inflammatory responsesin the airways and peripheral blood after short-term exposure todiesel exhaust in healthy human volunteers Am J Respir Crit CareMed 1999159702ndash9

25 Stenfors N Nordenhall C Salvi SS et al Different airwayinflammatory responses in asthmatic and healthy humans exposedto diesel Eur Respir J 20042382ndash86

26 Nordenhall C Pourazar J Ledin M-C et al Diesel exhaustenhances airway responsiveness in asthmatic subjects Eur Respir J200117909ndash15

27 Rudell B Ledin MC Hammarstrom U et al Effects on symptomsand lung function in humans experimentally exposed to dieselexhaust Occup Environ Med 199653658ndash62

28 Gong H Linn WS Sioutas C et al Controlled exposures of healthyand asthmatic volunteers to concentrated ambient fine particles inLos Angeles Inhal Toxicol 200315305ndash25

29 Hesterberg TW Bunn WB McClellan RO et al Critical reviewof the human data on short-term nitrogen dioxide (NO2)exposures evidence for NO2 no-effect levels Crit Rev Toxicol200939743ndash81

30 Larsson B-M Grunewald J Skoumlld CM et al Limited airway effectsin mild asthmatics after exposure to air pollution in a road tunnelResp Med 20101041912ndash18

Cowie CT Ezz W Xuan W et al BMJ Open 201200e001201 doi101136bmjopen-2012-001201 11

A randomised cross-over cohort study of exposure

on May 27 2020 by guest P

rotected by copyrighthttpbm

jopenbmjcom

B

MJ O

pen first published as 101136bmjopen-2012-001201 on 17 A

ugust 2012 Dow

nloaded from

Correction

Cowie CT Ezz W Xuan W et al A randomised cross-over cohort study of exposure to emis-sions from a road tunnel ventilation stack BMJ Open 20122e001201 Spacing in table 3 wasinaccurate and missed some header rows The corrected table is below Also in the thirdparagraph of the section entitled ldquoHealth outcomesrdquo the sentence ldquo higher scores for EYECHEST and THROAT symptoms (last row table 3)rdquo should actually be ldquohigher scores for EYEand CHEST symptoms (table 3)rdquo In addition the volume number should be listed as ldquo2rdquo inthis article We apologise for these errors

BMJ Open 20122e001201corr1 doi101136bmjopen-2012-001201corr1

BMJ Open 20122 doi101136bmjopen-2012-001201corr1 1

Miscellaneous

Table 3 Contrasts in lung function eNO and symptom scores between the downwind site and the other two sites (upwind and heavily trafficked) adjusted for change from

baseline (2006) to subsequent years

ContrastFEV1 (diff in litres)(95 CI)

FVC (diff in litres)(95 CI) eNO Ratio (95 CI)

EYE symptoms(diff in score)(95 CI)

THROAT symptoms(diff in score)(95 CI)Location Year

Downwind vs (Heavily

trafficked and Upwind)dagger200708 vs 2006Dagger 001 (minus004ndash006) 003 (minus003ndash009) 103 (094ndash112) minus0006 (minus0097ndash0086) 0130 (minus0084ndash0344)2007 vs 2006 003 (minus003ndash009) 005 (minus003ndash013) 101 (091ndash111) minus0055 (minus0159ndash0049) 0041 (minus0220ndash0301)2008 vs 2006 minus0004 (minus007ndash006) 002 (minus007ndash010) 105 (094ndash117) 0043 (minus0074ndash0159) 0207 (minus0027ndash0441)

Downwind vs Upwind 200708 vs 2006Dagger 002 (minus004ndash008) 006 (minus002ndash013) 103 (093ndash114) minus0044 (minus0151ndash0063) 0080 (minus0169ndash0329)2007 vs 2006 004 (minus003ndash011) 006 (minus003ndash015) 105 (094ndash117) minus0108 (minus0229ndash0012) minus0006 (minus0308ndash0297)2008 vs 2006 0004 (minus007ndash008) 006 (minus004ndash015) 101 (089ndash115) 0015 (minus0120ndash0149) 0179 (minus0092ndash0449)

Downwind vs Heavily

trafficked

200708 vs 2006Dagger minus000 (minus006ndash006) 0004 (minus007ndash008) 103 (093ndash114) 0033 (minus0073ndash0140) 0180 (minus0068ndash0428)2007 vs 2006 002 (minus005ndash009) 004 (minus005ndash012) 097 (087ndash108) minus0002 (minus0121ndash0117) 0087 (minus0212ndash0385)2008 vs 2006 minus001 (minus009ndash006) minus003 (minus012ndash007) 109 (096ndash123) 0071 (minus0065ndash0206) 0236 (minus0038ndash0509)

Heavily trafficked vs

(Upwind and

Downwind)sect

Overall minus001 (minus004ndash001) minus003 (minus006ndash0002) 109 (104ndash114) 0054 (0006ndash0102) 0107 (minus0005ndash0218)

Contrast CHEST symptoms(diff in score) (95 CI)

SNEEZE symptoms(diff in score) (95 CI)

DRY NOSE symptoms(diff in score) (95 CI)

SMELL-STRONGOdds ratio (95 CI)

SMELL-ANNOYINGOdds ratio (95 CI)Location Year

Downwind vs (Heavily

trafficked and Upwind)dagger200708 vs 2006Dagger minus0080 (minus0302ndash0142) 0288 (minus0139ndash0716) 0276 (0040ndash0511) 345 (157ndash760) 169 (072ndash396)

2007 vs 2006 minus0031 (minus0277ndash0214) 0375 (minus0093ndash0844) 0320 (0053ndash0587) 572 (235ndash1391) 195 (077ndash494)

2008 vs 2006 minus0133 (minus0409ndash0143) 0182 (minus0254ndash0618) 0229 (minus0057ndash0516) 214 (077ndash597) 133 (044ndash400)

Downwind vs Upwind 200708 vs 2006Dagger minus0017 (minus0275ndash0242) 0430 (minus0067ndash0927) 0193 (minus0081ndash0466) 351 (140ndash880) 123 (045ndash340)

2007 vs 2006 0033 (minus0252ndash0318) 0422 (minus0123ndash0966) 0187 (minus0124ndash0498) 411 (148ndash1143) 112 (038ndash337)

2008 vs 2006 minus0042 (minus0362ndash0277) 0424 (minus0080ndash0929) 0178 (minus0154ndash0510) 331 (098ndash1118) 126 (034ndash471)

Downwind vs Heavily

trafficked

200708 vs 2006Dagger minus0143 (minus0400ndash0114) 0147 (minus0349ndash0642) 0358 (0086ndash0631) 339 (139ndash829) 231 (090ndash594)

2007 vs 2006 minus0096 (minus0377ndash0185) 0329 (minus0208ndash0866) 0453 (0146ndash0759) 796 (289ndash2193) 338 (120ndash954)

2008 vs 2006 minus0224 (minus0547ndash0099) minus0060 (minus0570ndash0450) 0281 (minus0054ndash0617) 138 (044ndash440) 140 (041ndash475)

Heavily trafficked vs

(Upwind and

Downwind)sect

Overall 0209 (0094ndash0325) 0150 (minus0072ndash0372) 0022 (minus0101ndash0144) 234 (158ndash349) 408 (267ndash625)

Contrasts are derived from mixed effects regression model in which location and year and the interaction between them are fixed effects and individual subject intercepts are a random effectMain contrasts of interest are those included in the table and consist of downwind versus other exposures as the downwind exposure is the primary exposure of interest in this studydaggerDownwind vs (heavily trafficked and upwind) is the comparison of downwind versus heavily trafficked and upwind exposures combinedDagger200708 vs 2006 represents the comparison of both post-tunnel years combined (200708) versus the pre-tunnel year (2006)sectRepresents the overall comparison of the heavily trafficked site compared to the upwind and downwind sites combined (data from all years included) as we anticipated the heavily trafficked siteto have higher air pollution levels and that health effects would be more prominent

2BM

JOpen

20122doi101136bmjopen-2012-001201corr1

Table 3 Contrasts in lung function eNO and symptom scores between the downwind site and the other two sites (upwind and heavily trafficked) adjusted for change from

baseline (2006) to subsequent years

ContrastFEV1 (diff in litres)(95 CI)

FVC (diff in litres)(95 CI) eNO Ratio (95 CI)

EYE symptoms(diff in score)(95 CI)

THROAT symptoms(diff in score)(95 CI)Location Year

Downwind vs (Heavily

trafficked and Upwind)dagger200708 vs 2006Dagger 001 (minus004ndash006) 003 (minus003ndash009) 103 (094ndash112) minus0006 (minus0097ndash0086) 0130 (minus0084ndash0344)2007 vs 2006 003 (minus003ndash009) 005 (minus003ndash013) 101 (091ndash111) minus0055 (minus0159ndash0049) 0041 (minus0220ndash0301)2008 vs 2006 minus0004 (minus007ndash006) 002 (minus007ndash010) 105 (094ndash117) 0043 (minus0074ndash0159) 0207 (minus0027ndash0441)

Downwind vs Upwind 200708 vs 2006Dagger 002 (minus004ndash008) 006 (minus002ndash013) 103 (093ndash114) minus0044 (minus0151ndash0063) 0080 (minus0169ndash0329)2007 vs 2006 004 (minus003ndash011) 006 (minus003ndash015) 105 (094ndash117) minus0108 (minus0229ndash0012) minus0006 (minus0308ndash0297)2008 vs 2006 0004 (minus007ndash008) 006 (minus004ndash015) 101 (089ndash115) 0015 (minus0120ndash0149) 0179 (minus0092ndash0449)

Downwind vs Heavily

trafficked

200708 vs 2006Dagger minus000 (minus006ndash006) 0004 (minus007ndash008) 103 (093ndash114) 0033 (minus0073ndash0140) 0180 (minus0068ndash0428)2007 vs 2006 002 (minus005ndash009) 004 (minus005ndash012) 097 (087ndash108) minus0002 (minus0121ndash0117) 0087 (minus0212ndash0385)2008 vs 2006 minus001 (minus009ndash006) minus003 (minus012ndash007) 109 (096ndash123) 0071 (minus0065ndash0206) 0236 (minus0038ndash0509)

Heavily trafficked vs

(Upwind and

Downwind)sect

Overall minus001 (minus004ndash001) minus003 (minus006ndash0002) 109 (104ndash114) 0054 (0006ndash0102) 0107 (minus0005ndash0218)

Contrast CHEST symptoms(diff in score) (95 CI)

SNEEZE symptoms(diff in score) (95 CI)

DRY NOSE symptoms(diff in score) (95 CI)

SMELL-STRONGOdds ratio (95 CI)

SMELL-ANNOYINGOdds ratio (95 CI)Location Year

Downwind vs (Heavily

trafficked and Upwind)dagger200708 vs 2006Dagger minus0080 (minus0302ndash0142) 0288 (minus0139ndash0716) 0276 (0040ndash0511) 345 (157ndash760) 169 (072ndash396)

2007 vs 2006 minus0031 (minus0277ndash0214) 0375 (minus0093ndash0844) 0320 (0053ndash0587) 572 (235ndash1391) 195 (077ndash494)

2008 vs 2006 minus0133 (minus0409ndash0143) 0182 (minus0254ndash0618) 0229 (minus0057ndash0516) 214 (077ndash597) 133 (044ndash400)

Downwind vs Upwind 200708 vs 2006Dagger minus0017 (minus0275ndash0242) 0430 (minus0067ndash0927) 0193 (minus0081ndash0466) 351 (140ndash880) 123 (045ndash340)

2007 vs 2006 0033 (minus0252ndash0318) 0422 (minus0123ndash0966) 0187 (minus0124ndash0498) 411 (148ndash1143) 112 (038ndash337)

2008 vs 2006 minus0042 (minus0362ndash0277) 0424 (minus0080ndash0929) 0178 (minus0154ndash0510) 331 (098ndash1118) 126 (034ndash471)

Downwind vs Heavily

trafficked

200708 vs 2006Dagger minus0143 (minus0400ndash0114) 0147 (minus0349ndash0642) 0358 (0086ndash0631) 339 (139ndash829) 231 (090ndash594)

2007 vs 2006 minus0096 (minus0377ndash0185) 0329 (minus0208ndash0866) 0453 (0146ndash0759) 796 (289ndash2193) 338 (120ndash954)

2008 vs 2006 minus0224 (minus0547ndash0099) minus0060 (minus0570ndash0450) 0281 (minus0054ndash0617) 138 (044ndash440) 140 (041ndash475)

Heavily trafficked vs

(Upwind and

Downwind)sect

Overall 0209 (0094ndash0325) 0150 (minus0072ndash0372) 0022 (minus0101ndash0144) 234 (158ndash349) 408 (267ndash625)

Contrasts are derived from mixed effects regression model in which location and year and the interaction between them are fixed effects and individual subject intercepts are a random effectMain contrasts of interest are those included in the table and consist of downwind versus other exposures as the downwind exposure is the primary exposure of interest in this studydaggerDownwind vs (heavily trafficked and upwind) is the comparison of downwind versus heavily trafficked and upwind exposures combinedDagger200708 vs 2006 represents the comparison of both post-tunnel years combined (200708) versus the pre-tunnel year (2006)sectRepresents the overall comparison of the heavily trafficked site compared to the upwind and downwind sites combined (data from all years included) as we anticipated the heavily trafficked siteto have higher air pollution levels and that health effects would be more prominent

2BM

JOpen

20122doi101136bmjopen-2012-001201corr1