Adverse effects of exposure to air pollutants during fetal development and early life with focus on pre-eclampsia, preterm delivery, and childhood asthma

David Olsson

Department of Public Health and Clinical Medicine

Occupational and Environmental Medicine

Umeå University 2014

Responsible publisher under Swedish law: the Dean of the Medical Faculty

This work is protected by the Swedish Copyright Legislation (Act 1960:729)

ISBN: 978-91-7601-139-3

ISSN: 0346-6612

Elektronisk version tillgänglig på http://umu.diva-portal.org/

Tryck/Printed by: Print & Media, Umeå University

Umeå, Sweden 2014

He få vål va he vejl

Sankte Per – Skapelseberättelsen, burträskarens uppkomst

i

Table of Contents

Table of Contents i Abstract ii Abbreviations iv Enkel sammanfattning på svenska v Original papers vi Introduction 1

Air pollution exposure 2 Air pollution and birth outcomes 2 Air pollution and asthma 3

Aims of the thesis 5 Overall aims 5 Specific aims 5

Materials and methods 6 Study populations 6 Outcome definitions 6 Covariate definitions 7 Exposure assessment 8 Statistical analysis 8

Results 13 Paper I 13 Paper II 14 Paper III 14 Paper IV 15

Discussion 17 Exposure assessment 17 Statistical modeling 17 Mechanism 18 Comparison of findings 18 Policy implications 20 Summary of findings 20 Future research 21 Conclusions 22

Acknowledgements 23 References 25

ii

Abstract

Background Air pollution exposure has been shown to have adverse effects

on several health outcomes, and numerous studies have reported

associations with cardiovascular morbidity, respiratory disease, and

mortality. Over the last decade, an increasing number of studies have

investigated possible associations with pregnancy outcomes, including

preterm delivery. High levels of vehicle exhaust in residential neighborhoods

have been associated with respiratory effects, including childhood asthma,

and preterm birth is also associated with childhood asthma.

The first aim of this thesis was to investigate possible associations between

air pollution exposure and pregnancy outcomes – primarily preterm delivery

but also small for gestational age (SGA) and pre-eclampsia – in a large

Swedish population (Papers I–III). The second aim was to study any

association between exposure to high levels of vehicle exhaust during

pregnancy and infancy and prescribed asthma medication in childhood

(Paper IV).

Methods The study cohorts were constructed by matching other individual

data to the Swedish Medical Birth Register. In the first two studies, air

pollution data from monitoring stations were used, and in the third and

fourth studies traffic intensity and dispersion model data were used.

Preterm delivery was defined as giving birth before 37 weeks of gestation.

SGA was defined as having a birth weight below the 10th percentile for a

given duration of gestation. Pre-eclampsia was defined as having any of the

ICD-10 diagnosis codes O11 (pre-existing hypertension with pre-eclampsia),

O13 (gestational hypertension without significant proteinuria), O14

(gestational hypertension with significant proteinuria), or O15 (eclampsia).

Childhood asthma medication was defined as having been prescribed asthma

medication between the ages of five and six years.

Results We observed an association between ozone exposure during the

first trimester and preterm delivery. First trimester ozone exposure was also

associated with pre-eclampsia. The modeled concentration of nitrogen

oxides at the home address was associated with pre-eclampsia, but critical

time windows were not possible to investigate due to high correlations

between time windows. We did not observe any association between air

pollution exposure and SGA. High levels of vehicle exhaust at the home

address, estimated by nitrogen oxides and traffic intensity, were associated

with a lower risk of asthma medication.

iii

Conclusion Air pollution exposure during pregnancy was associated with

preterm delivery and pre-eclampsia. We did not observe any association

between air pollution levels and intrauterine growth measured as SGA. No

harmful effect of air pollution exposure during pregnancy or infancy on the

risk of being prescribed asthma medication between five and six years of age

was observed.

iv

Abbreviations

ATC – Anatomical therapeutic chemical

BMI – Body mass index

CO – Carbon monoxide

ICD-10 – International Statistical Classification of Diseases, version 10

LBW – Low birth weight NO2 – Nitrogen dioxide NOx – Nitrogen oxides O3 – Ozone

OR – Odds ratio

PM10 – Particulate matter with a diameter

v

Enkel sammanfattning på svenska

Ogynnsamma födelseutfall, exempelvis förtida födsel, har visats öka risken

för sämre studieresultat och mer sjuklighet genom hela livet. Astma under

barndomen kan leda till en lägre fysisk aktivitet eller ett lägre psykosocialt

välbefinnande.

Denna avhandling har två huvudspår: (1) Att studera samband mellan

luftföroreningshalter under graviditeten och ogynnsamma utfall, främst

havandeskapsförgiftning, för tidigt födda barn samt tillväxthämning. (2) Att

studera samband mellan luftföroreningshalter under graviditeten eller

spädbarnstiden och astmamedicinering under barndomen.

De viktigaste fynden är att förhöjda ozonhalter i ett tidigt skede av

graviditeten ledde till en ökad risk för barnet att bli för tidigt fött och för

mamman att drabbas av havandeskapsförgiftning. Bodde mamman i ett

område med högre avgashalter hade hon en högre risk att drabbas av

havandeskapsförgiftning, och mammor som bodde i områden med lägst

avgashalter hade en mindre risk att föda ett tillväxthämmat barn.

Det fanns inga tecken på att bo i områden med högre avgashalter eller mer

trafikerade områden under graviditeten eller spädbarnstiden ökade risken

för att barnet skulle behöva medicineras för astma före skolåldern.

Alla fyra delarbeten i avhandlingen har omfattat Storstockholm. Den första

studien baserades på alla kvinnor i området som var gravida mellan 1988

och 1995. Det andra och tredje delarbetet utgår från alla kvinnor som var

gravida mellan augusti 1997 och februari 2006. Det fjärde delarbetet baseras

på alla födslar mellan juli 2000 och oktober 2005.

De luftföroreningar som studerades var ozon och kväveoxider, där

kväveoxider användes som ett mått på trafikavgaser. Som ett ytterligare mått

på trafik studerades det genomsnittliga antalet fordon som passerade

hemadressen.

vi

Original papers

I: Olsson D, Ekström M, Forsberg B. Temporal variation in air pollution

concentrations and preterm birth – a population based epidemiological

study. Int. J. Environ. Res. Public Health, 2012, 9:272-285.

II: Olsson D, Mogren I, Forsberg B. Air pollution exposure in early

pregnancy and adverse pregnancy outcomes: a register-based cohort study.

BMJ Open, 2013, 3:e001955.

III. Olsson D, Mogren I, Eneroth K, Forsberg B. Traffic pollution at home

address and pregnancy outcomes. Submitted.

IV. Olsson D, Bråbäck L, Forsberg B. Traffic pollution exposure at home

during pregnancy and infancy and childhood asthma medication.

Manuscript.

vii

1

Introduction

Adverse pregnancy outcomes such as a shorter period of gestation and

intrauterine growth restriction have been shown to have an effect on

morbidity throughout childhood, adolescence, and adult life1-2. A shorter

period of gestation is also associated with decreased school performance and

is an important predictor of lower socioeconomic status in adulthood2-4.

Common proxies for intrauterine growth restriction in the scientific

literature over the years have been low birth weight (LBW) at term and small

for gestational age (SGA)5-6. LBW is defined as having a birth weight lower

than 2,500 g, and SGA is commonly defined as having a birth weight lower

than either the 10th percentile for a given gestational age or lower than 2

standard deviations below the average birth weight for a given gestational

age. LBW was the more common metric in earlier studies, but as estimates of

gestational age have been increasingly reliable and more readily available

SGA has become more common. Similarly, preterm delivery has also been

more commonly studied as the reliability of gestational age estimates have

improved, and preterm delivery is defined as being born before the 37th week

of gestation7-9. Known risk factors for preterm delivery and LBW are

smoking and pre-eclampsia.

Approximately 3%–8% of all pregnancies in Western countries are

complicated by pre-eclampsia10-11. The main symptoms of pre-eclampsia are

hypertension and proteinuria12. Pre-eclampsia is more common among

nulliparous women and women with previous chronic hypertension11.

A shorter period of gestation in particular is associated with a higher rate of

asthma, possibly because the lung tissue might not be fully developed at the

time of delivery13.

Childhood asthma and wheezing are heterogeneous conditions with multiple

causes14, including environmental tobacco smoke and dampness in the

home15. Wheezing in infants is often a transient condition that is closely

associated with respiratory infections, but the causes of persistent wheezing

remain unclear. Asthma is the most common chronic disease in childhood

and can affect both physical fitness and psychosocial wellbeing16.

2

Air pollution exposure

Nitrogen oxides (NOx) consist of nitrogen dioxide (NO2, an oxidant with

inflammatory effects17) and nitrogen oxide (NO). NOx are formed at high

temperature in combustion engines and are, therefore, a marker of vehicle

exhaust and traffic pollution.

Ground level ozone (O3, an oxidant with short-term effects on lung function

and airway inflammation17) is formed along with NO through a chemical

reaction between oxygen and NO2 under the influence of sunlight. In areas

with high NO emissions, more O3 is consumed to oxidize NO to NO2.

Long-term exposure to air pollution has been associated with increased

mortality and incidence of chronic diseases, and there is stronger evidence

for the negative effects of particulate matter than for NO2 and O317-18.

Primary combustion particles, including soot particles and elemental carbon,

are likely to be more harmful than particle mass in general. However, black

carbon and elemental carbon are seldom measured by typical monitoring

stations.

Exposure misclassification is always a potential problem in epidemiological

studies using measured or modeled outdoor concentrations in the area or at

home. This is because study subjects have different time-activity patterns

and might have a true exposure that is lower or higher than what the

exposure variable suggests. This type of exposure misclassification will likely

dilute any association between exposure and outcome.

The association between air pollution exposure and health outcome might be

confounded by risk factors that are correlated with both exposure and

outcome. Potential confounders must vary together in the same

dimension(s) as the exposure, for example, if the exposure varies over time,

then potential confounders must also vary over time.

Air pollution and birth outcomes

Early studies on the effects of air pollution on birth outcomes focused mainly

on long-term average exposures in different districts within the study area.

The main outcome of interest was LBW5-6,19-22, but preterm delivery and SGA

were also studied23. The air pollutants investigated in those studies were

mainly carbon monoxide (CO), sulfur dioxide (SO2), NOx, and total

suspended particles. Most studies reported that elevated levels of SO2 were

associated with increased odds of LBW5,20-23.

3

Later studies focused on preterm delivery as the outcome of interest to a

much larger extent than the earlier studies, although LBW was still

frequently studied7-9,24-33. Long-term air pollution averages were still the

most common exposure metric used, but the exposure assessment shifted

from area averages to using the nearest monitoring station and including

only subjects living within a certain distance from the monitoring station.

There were some studies that used modeling approaches in the exposure

assessment, including kriging32, dispersion modelling29 and land-use

regression25. Particulate matter with an average diameter less than 10 µm

(PM10) or 2.5 µm (PM2.5) were the most commonly studied air pollutants,

although others such as NO2, NOx, SO2, O3, and CO were also studied. Most

studies showed positive associations between air pollution exposure and the

studied birth outcome. However, the critical timing of the exposure varied

across the different study populations.

In more recent studies, preterm delivery, SGA, and LBW have been the most

commonly studied outcomes, although a few studies have used pre-

eclampsia as the outcome34-44. The exposure assessments were similar to

those of the earlier studies, i.e., measuring exposure levels from the nearest

monitoring station within a certain radius or using a dispersion model to

estimate the level of exposure at the home address. As in earlier studies,

higher levels of air pollution were positively associated with preterm

delivery, LBW, and SGA. Higher levels of PM10 and NO2 have been

associated with an increased risk of pre-eclampsia37,40, but higher levels of

CO were associated with a decreased risk of pre-eclampsia44.

A potential pathway through which ambient air pollution could lead to

adverse pregnancy outcomes is through systemic inflammation45.

Proinflammatory cytokines could disrupt trophoblastic invasion during

placentation leading to suboptimal function of the placenta46-47. This could in

turn lead to pre-eclampsia, preterm delivery, or intrauterine growth

restriction48.

Air pollution and asthma

Air pollution exposure and proximity to traffic have been linked to asthma or

asthma symptoms in several studies49-55, and air pollution exposure has also

been linked to hospital admissions and emergency room visits for asthma56-

57. Associations between early life exposure to air pollution and childhood

asthma have been reported in several studies58-63, and there is some evidence

that exposure to air pollution during gestation might be associated with an

increased risk of developing asthma during childhood64. These findings are

4

inconsistent52, however, and an association between air pollution exposure

during childhood and asthma has not always been found65-66.

Air pollution exposure might lead to epigenetic reprogramming that could

potentially lead to an elevated risk of developing respiratory diseases in

general67. Alternatively, air pollution exposure during periods of rapid

development of lung tissue, such as during the last stage of pregnancy or

during infancy, might cause disruption in the development of the respiratory

organs that could lead to a higher risk for asthma59,67.

5

Aims of the thesis

Overall aims

The overall aims of this thesis were (1) to assess any possible association

between air pollution exposure during pregnancy and important adverse

pregnancy outcomes and (2) to assess any possible associations between air

pollution exposure during pregnancy or infancy and asthma medication

during childhood.

Specific aims

In Paper I, the aim was to study any association between O3 and NO2

exposure during the first and second trimesters and the last week of

gestation and preterm delivery and duration of gestation.

Paper II aimed to assess the association between first trimester levels of O3

and NOx and preterm delivery in a different cohort as well as to study pre-

eclampsia and SGA as outcomes.

The aim of Paper III was to study any association between traffic pollution at

the home address during pregnancy, as indicated by NOx and traffic

intensity, and preterm delivery, pre-eclampsia, and SGA.

Paper IV studied the association between traffic pollution during pregnancy

and from birth to the first birthday, as indicated by NOx and traffic intensity,

and being prescribed asthma medication between five and six years of age.

6

Materials and methods

Study populations

All four studies were performed on study populations in the greater

Stockholm area. Paper I was based on all singleton births conceived between

1988 and 1995. Papers II and III were based on all singleton births conceived

between August 1997 and February 2006 and delivered at one of the five

hospitals in the greater Stockholm area (Danderyd, Huddinge, Karolinska,

Södersjukhuset, and Södertälje). The third study was further restricted to

include only women who did not change addresses during pregnancy and

who had a delivery that started spontaneously. The fourth study was based

on all singleton births between July 1, 2000, and October 30, 2005.

Depending on the timing of the exposure of interest in the fourth study, the

inclusion criteria were altered. When studying exposure during pregnancy,

only children born to women who did not change addresses during

pregnancy were included. When studying exposure during infancy, those

who did not change addresses between birth and their first birthday were

included in one datset and those who did not change address during infancy

but whose mothers changed addresses during pregnancy were included in a

separate dataset.

Outcome definitions

Duration of gestation was based on ultrasound examination or the date of

the last menstrual period (Paper I). Preterm delivery was defined as having a

gestational period shorter than 37 weeks (Papers I–III). SGA was defined as

having a birth weight below the 10th percentile for the given duration of

gestation in days (Papers II and III) and according to sex (Paper III). Pre-

eclampsia was defined as having any of the ICD-10 (International Statistical

Classification of Diseases, version 10) diagnoses of O11 (pre-existing

hypertension with pre-eclampsia), O13 (gestational hypertension without

significant proteinuria), O14 (gestational hypertension with significant

proteinuria), or O15 (eclampsia) (Papers II and III). Two different

definitions of childhood asthma medication were used. The first included

any prescribed asthma medication between the ages 5 and 6 years

(Anatomical Therapeutical Chemical (ATC) codes R03AC, R03AK, R03BA,

R03BC, R03CC, or R03DC), and the second included being prescribed anti-

inflammatory medicines or leukotriene antagonists on two or more

occasions between the ages of 5 and 6 years (ATC codes R03AK, R03BA or

R03DC) (Paper IV).

7

Covariate definitions

The Swedish Medical Birth Register contains data on a number of different

covariates such as smoking habits, family situation, and parity.

In Paper I we included maternal smoking habits (non-smoker, moderate

smoker (1–9 cigarettes/day) and heavy smoker (10 or more cigarettes/day)),

maternal parity (0, 1, 2, 3, or more para), and infant sex (male, female). The

date of conception was used to adjust for possible confounding from

temporal and seasonal trends, and data on temperature (°C) and relative

humidity (%) were used in the modeling.

Paper II used the same covariates as Paper I, but it also included data on

maternal age, maternal asthma status, highest level of maternal education,

BMI (kg/m2) at antenatal care registration, family situation, and maternal

region of origin in the statistical modeling. Maternal age was a continuous

variable. Maternal asthma status was a dichotomous factor defined as a case

if the mother was prescribed any asthma medication between July 2005 and

December 2011 (ATC codes R03AC, R03AK, R03BA, R03BC, R03CC, or

R03DC) or had ever received hospital care for asthma bronchiale since 1997.

Highest level of education had five category levels (pre-upper secondary

school (

8

of times the child changed addresses up to the age of four was followed

yearly resulting in 5 possible discrete states of 0–4 address changes.

Exposure assessment

The City of Stockholm Environment and Health Administration provided

time-series of air pollution measurements for all studies. The modeled

annual averages for the 100 m × 100 m squares where the home addresses of

the study population were located were provided for Papers III and IV. NO2

(1-hour maxima, Paper I) and NOx (daily mean, Paper II) were measured

daily at rooftop level at three monitoring stations. O3 (8-hour maxima,

Papers I and II) was measured at rooftop level at two monitoring stations,

Figure 1. In the case where one station had a missing value, the data were

imputed from the other stations by linear regression. The time-series from

the monitoring stations were used to create citywide averages that were used

to calculate time-dependent averages over weeks and trimesters (12 weeks).

If there were more than 19 days of missing data, no trimester average was

calculated.

Annual average NOx levels were estimated at the location of the home

address using the SMHI-Airviro Gauss dispersion model68, Figure 2. The

input to the model was provided by the emission inventory of the City of

Stockholm Environment and Health Administration and the Uppsala Air

Quality Management System, which consist of road traffic NOx emissions

(i.e. traffic flow, vehicle fleet composition, emission factors and

meteorology). The ratio between the estimated annual averages at the

location of the home address and the annual city average from the

monitoring stations were used to estimate trimester, pregnancy, and first-

year exposures for each study subject in the third and fourth studies.

Annual daily average traffic flow at the location of the home address was

used as an alternative vehicle exhaust exposure metric in Papers III and IV.

These data were also provided by the City of Stockholm Environment and

Health Administration.

Statistical analysis

Logistic regression was used for the main analyses in Papers I, II, and IV.

Linear regression was used for the continuous outcomes in Paper I. Mixed-

model logistic regression was used in Paper III. All analyses were performed

in R programming software69.

9

Figure 1. Monthly average NO2 and O3 levels.

In Paper I, each time window was studied separately (first trimester, second

trimester, last week of gestation, and last week at risk for preterm delivery).

For each outcome, the estimated air pollution effect was modeled in three

steps. First, unadjusted models were fitted. Next, the models were adjusted

for maternal smoking, parity, sex of the infant, temperature, relative

humidity, season, and long-term trend. The last model estimated the effects

of the air pollutants simultaneously. The seasonal patterns were accounted

for by fitting the day of the year of conception as a cyclic spline function. A

cubic regression spline was fitted to the conception year to adjust for long-

term trends.

Given the results in Paper I, only exposures to NOx and O3 during the first

trimester were studied in Paper II. The modeling was performed in six steps.

First, unadjusted models were fitted for each pollutant separately. Then

maternal age (as a cubic regression spline), parity, highest level of education,

10

region of origin, maternal asthma, and seasonal trend (as a cyclic spline)

were added to the model. In the third step, a cubic regression spline to

account for any long-term trend was added. The fourth step was to study the

exposures simultaneously. First-trimester temperature and relative humidity

were added in the fifth step. Finally, maternal smoking habits, family

situation, and BMI at the first antenatal visit were added. Additionally, we

explored if maternal asthma acted as an effect modifier for O3 by adding an

interaction term in the modeling procedure.

Figure 2. Modeled annual NOx means over the study area

The modeling approach in Paper III was similar to that of the previous

papers. The initial model consisted of home address NOx levels or traffic

flow and a random intercept for home municipality. In the second modeling

step, a set of maternal variables was added (asthma status, level of

education, region of origin, age and parity), along with conception date, first

trimester O3, and temperature. In the third and final step, maternal smoking

status, family situation, and BMI at the first antenatal visit were included in

11

the model. Additional analysis was performed on the subset of the study

population that had complete information for all explanatory variables in

order to investigate if changes in estimates were due to adjustment for

additional factors or due to restriction of the study population.

Similarly to the other studies, the statistical analyses in Paper IV were

performed in a forward stepwise fashion. Initially, a crude model including

only outcome and exposure was fitted. In the second step, older siblings,

parental asthma, maternal region of origin, maternal age and level of

education at the time of delivery, number of address changes, birth month,

and date of birth were added to the model. Pregnancy outcomes (duration of

gestation (cubic regression spline), pre-eclampsia, and SGA) were added in

the third step. In the last step, maternal BMI at the first antenatal visit was

added. The modeling approach for those who changed address differed in

step 2 in that older siblings, level of education, and number of address

changes were excluded.

12

Table 1. Outcome frequencies and selected mean exposure levels across

studies.

Paper Ia Number (%) Mean first trimester NO2 (sd), µg/m3

Mean first trimester O3 (sd), µg/m3

Preterm delivery, Yes

6,154 (5.3) 38.4 (5.4) 57.6 (13.4)

No 109,434 38.5 (5.4) 57.1 (13.1) Paper IIa Number (%) Mean first

trimester NOx (sd), µg/m3

Mean first trimester O3 (sd), µg/m3

Pre-eclampsia delivery, Yes

3,239 (2.7) 96.6 (16.3) 68.6 (13.3)

No 117,516 96.9 (16.3) 67.8 (13.3) Preterm delivery, Yes

5,341 (4.4) 96.7 (16.3) 68.4 (13.2)

No 115,404 96.9 (16.3) 67.8 (13.3) SGAb, Yes 11,334 (9.4) 97.0 (16.2) 67.6 (13.2) No 109,216 96.9 (16.3) 67.9 (13.3) Paper IIIc Number (%) Mean first

trimester NOx (sd), µg/m3

Pre-eclampsia delivery, Yes

2,774 (2.7) 15.4 (7.6)

No 99,994 15.1 (7.4) Preterm delivery, Yes

2,823 (2.8) 15.3 (7.5)

No 98,220 15.1 (7.4) SGAb, Yes 9,977 (9.8) 14.9 (7.2) No 91,960 15.1 (7.4) Paper IVc Number (%) Mean first

year NOx (sd), µg/m3

Two prescriptions of anti-inflammatory medicine or leukotriene antagonists

2,354 (3.0) 13.2 (6.2)

Any prescribed asthma medication

7,614 (9.7) 13.1 (6.1)

No prescribed asthma medication

70,526 13.4 (6.4)

aAir pollution data from monitoring stations. bSmall for gestational age. cAir

pollution data from dispersion model.

13

Results

The proportion of preterm deliveries was higher in the earlier cohort (all

singleton children who were conceived between 1988 and 1995, Paper I)

than in the later cohort (all singleton children conceived between August

1997 and February 2006, Paper II) (Table 1). Almost 3% of the pregnant

women suffered from pre-eclampsia and pregnancy-induced hypertension.

The average O3 levels increased from the first study cohort to the second, but

the average difference in first trimester O3 levels between preterm and term

deliveries were consistently 0.5–0.6 µg/m3.

Figure 3. Annual average O3 and proportion of preterm delivery between

1988 and 1995.

Paper I

Higher first trimester O3 levels were associated with slightly increased odds

of preterm delivery and a shorter period of gestation in both the unadjusted

and adjusted models. The second trimester and last week of gestation O3

levels were not associated with an altered risk of preterm delivery, but were

14

weakly associated with a shortened period of gestation. Figure 3 indicates a

co-variation between O3 levels and preterm delivery during the study period,

in this case without any adjustment.

Elevated NO2 levels during the first and second trimester were not associated

with an increased risk of preterm delivery, but they were associated with a

slightly longer period of gestation. Elevated NO2 levels during the late stage

of pregnancy were associated with an increased risk of preterm delivery.

Paper II

Higher levels of first trimester O3 were again associated with higher odds of

preterm delivery in a new cohort. The association was stronger when elective

Caesarean sections were excluded from the analysis. There was some

evidence of effect modification by maternal asthma status where the

association between O3 and preterm delivery was stronger among asthmatic

mothers than non-asthmatic mothers. Higher levels of NOx did not appear

to have an effect on the risk of having a preterm delivery.

There was no evidence of any increased risk of SGA by having elevated first

trimester levels of O3 or NOx, in fact, nearly all estimated associations

indicated a slightly decreased risk of SGA.

Increased first trimester O3 was associated with higher odds of pre-

eclampsia. The association was slightly weaker when excluding elective

Caesarean sections; however, having an elective Caesarean section is a likely

outcome of having pre-eclampsia. Increased levels of first trimester NOx did

not exhibit any association with pre-eclampsia.

Paper III

Due to high correlation between the NOx levels at the home location through

all periods of pregnancy, it was difficult to identify periods of increased

susceptibility. The NOx level at the home address was positively correlated

with a slight, but not statistically significant, increased risk of preterm

delivery. Higher levels of NOx were associated with an increased risk of SGA

when comparing any of quartiles 2, 3, or 4 with the first quartile. There was,

however, no difference in estimated effect among quartiles 2, 3, and 4. There

was a marked association between home address NOx and pre-eclampsia

with an OR of 1.17 per 10 µg/m3 increase in NOx in the fully adjusted model.

15

There were no associations between average traffic intensity and preterm

delivery or SGA. Traffic intensity was associated with slightly increased odds

of pre-eclampsia, but the association was close to the null.

Figure 4. Main findings of the studies in terms of ORs per 10 µg/m3 increase in air pollution level. Exposure periods were the first trimester in Papers I, IIand III and the pregnancy period in Paper IV. The results for Paper I are from the multi-pollutant model in Table 4 of Paper I. For Paper II, the results were collected from Model 4 in Paper II as shown in Tables 1, 3, and 4 of Paper II for the non-restricted study population. The results from Papers III and IV were collected from Model 2 in Table 2 of Papers III and IV.

Paper IV

Higher levels of NOx during pregnancy or during infancy were associated

with a lower risk of using asthma medication between the ages five and six

16

years. When restricting the analysis to only those who changed addresses

prior to delivery, the protective association remained but were not as marked

as for those who remained at the same address throughout their pregnancy.

Traffic intensity showed a similar pattern of association with asthma

medication as NOx, although for those who changed address there was

neither a protective nor a harmful effect.

17

Discussion

Exposure assessment

Papers I and II used city-wide fluctuations in urban background air pollution

values over time to estimate individual exposure. In other words, individual

exposure was a function of conception date only, and two women who

conceived on the same day would have been assigned the same levels of

exposure regardless of where they lived. This can lead to exposure

misclassification, particularly for pollutants with important local sources,

such as NO2 and NOx, which vary both in time and strongly in space because

they are formed primarily through combustion in vehicle engines. This is less

of an issue for O3 that, although it is consumed by locally emitted NO,

depends mainly on incoming air masses and varies seasonally over time70.

This means that there was a greater chance to detect an association between

O3 and pregnancy outcomes than between NO2 or NOx and pregnancy

outcomes in Papers I and II. Papers III and IV were better designed to

estimate associations between spatially varying traffic air pollution and the

studied outcome. However, due to the exposure assessment it was possible to

separate between effects of exposure during different time windows.

Statistical modeling

The statistical models in the four papers include several covariates, some of

which might be considered confounders and some of which might be

important determinants but unlikely confounders. For example, parity

would not be considered a confounder in the papers that focus on the

correlation between the temporal variation of exposure and pregnancy

outcomes unless one suspects that families with different numbers of

children are differentially prone to plan for a child at a certain time during

the year and that parity is associated with the outcome. In the papers

focusing on spatial variation, parity could be a confounder if families of

different sizes have different probabilities of living in environments with

certain air pollution concentrations.

In the first two papers, the potential confounders of interest were time-

varying variables – such as season of conception and the meteorological

variables – because they could affect the outcome and because O3 varies with

season. In Papers III and IV, the confounders that were identified were the

proxies for socioeconomic status, i.e. level of education and maternal region

of origin. The reason why they could be confounders is that people with a

higher level of education tend to live in the central areas in Stockholm where

traffic air pollution levels are higher; therefore, neglecting to adjust for

socioeconomic status might lead to the conclusion that living in more

18

polluted areas might have a beneficial effect on pregnancy outcomes and

childhood asthma medication.

Mechanisms

In the studies of preterm delivery, pre-eclampsia was thought of as a possible

intermediate step along the causal pathway between air pollution exposure

and preterm delivery. In Paper II this was explored by adjusting the models

in the main analysis for pre-eclampsia. This proposed that the previously

observed association would be masked by adding pre-eclampsia as a factor in

the model. This did not change the estimated association indicating that pre-

eclampsia is not an important intermediate step between air pollution

exposure and preterm delivery.

In a similar manner, the adverse pregnancy outcomes studied in the first

three papers could be possible pathways through which air pollution

exposure during pregnancy could lead to childhood asthma. Being born after

a shorter gestational period or being growth restricted were associated with

an increased risk of asthma medication (Paper IV, Table 1). If air pollution

levels during pregnancy were associated with both pregnancy outcomes and

the need to use asthma medication, it is possible that the relationship

between air pollution exposure and childhood asthma medication would be

masked by adjusting for pregnancy outcomes. However, no adverse

association between traffic pollution during pregnancy (or infancy) and

asthma medication was apparent, and no changes in the estimated

association were observed after adjusting for birth outcomes.

Comparison of findings

There was an association between elevated O3 levels during the first

trimester of pregnancy and preterm birth in Paper I and Paper II. This

association has been reported previously in a few other studies9,36-37,47,71, but

other studies have found no such association24,33,72. In one of the studies,

however, the association did not remain after adjustment for possible

confounders71. The differing results might be explained in part by different

timing and intensity of the seasonal O3 peaks, for example, in Stockholm and

Shanghai9 the highest levels of O3 are found during spring73, while the

highest levels of O3 in the US and London are found during summer36,74-75.

Vitamin D status varies with the season – with a trough in the spring76-80 –

and a lower level of vitamin D status is associated with increased sensitivity

to inflammation81. Higher O3 levels have been reported to be associated with

inflammation in early pregnancy45, and inflammation during early

19

pregnancy has been associated with a higher prevalence of preterm

delivery82.

High NO2 levels late in pregnancy were associated with preterm delivery in

Paper I, but no association was observed in the other stages of pregnancy.

The association between elevated levels of late-pregnancy NO2 and preterm

delivery is in agreement with previous studies33,47,71-72,83. Higher levels of

NOx early in pregnancy were not associated with preterm delivery in Paper

II. In Paper III, only the results for the average levels of NOx during the first

trimester were reported because the NOx levels were highly correlated

between different pregnancy time windows. No statistically significant

association between elevated NOx levels and preterm delivery was observed.

The problem with correlated exposure estimates over different time windows

have been reported elsewhere26,29. Although no statistically significant

associations between early pregnancy or pregnancy NOx levels and preterm

delivery were observed in Papers II and III, the observed tendencies are in

the same direction as previously published results25,29,84.

Similarly to Paper II, a study from Pennsylvania reported a tendency that

higher levels of first trimester O3 led to an increased risk of pre-eclampsia37.

A study from California reported similar associations as in Paper II for O3

and pre-eclampsia in Orange County but not in Los Angeles85. The differing

results for Los Angeles might be because they reported entire pregnancy

exposure associations only. A few studies have reported associations between

vehicle exhaust and pre-eclampsia, and these associations were of similar

magnitude as reported in Paper III37,85-87.

The association between late pregnancy levels of NO2 (Paper I) and preterm

delivery and between pregnancy average levels of NOx and pre-eclampsia

(Paper III) suggest that vehicle exhaust exposure in Stockholm could be high

enough to cause adverse effects on pregnancy outcomes.

The observed association between NOx and SGA in Paper III – where

mothers who resided in areas with NOx levels exceeding the 1st quartile were

more likely to give birth to an SGA infant – supports the conclusion of small

or no effect in a review by Glinianaia et al., although that review focused on

particulate matter88. No association between O3 or NOx and SGA was

observed in Paper II, which might be because only the exposure in the first

trimester was studied and effects on fetal growth are likely to occur later in

pregnancy89.

No positive association between NOx levels at home and being prescribed

asthma medication was found, and similar studies have reported

20

inconsistent results59,90. Several smaller studies have, however, reported an

association between traffic air pollution and incident or new-onset asthma61-

63. The differing results might be because a more accurate exposure

assessment is possible in a smaller study population62 or because the studied

population was likely to be more sensitive to airway insults than the general

population61. A possible explanation for the negative associations observed

in Paper IV and in the other Swedish study by Lindgren et al.91 is that in

populations living in a relatively clean environment there is some beneficial

adaptation among children living in neighborhoods with moderately higher

levels of air pollution.

Policy implications

In the recent WHO REVIHAAP (2013) report, O3 was “upgraded” as a health

problem18. Climate change will increase O3 exposure assuming that

everything else stays the same. The increased proportion of diesel cars in

countries like Sweden is changing the ratio between NO and NO2 in local

emissions and is increasing O3 levels in the cities92. The local regulations for

O3 could be stricter, but large-scale measures would have to be undertaken

in order to effectively reduce concentrations.

In Paper I it was estimated that 129 cases (7.8%) of preterm delivery could be

attributed to being exposed to the highest quartile of O3 instead of the lowest

quartile. Similarly, 138 cases (5.5%) of pre-eclampsia and 211 cases (5.2%) of

preterm delivery could be attributed to being exposed to O3 levels exceeding

the 25th percentile in Paper II. Being exposed to the highest quartile of NO2

was associated with 85 cases (5.2%) of preterm delivery in Paper I. Stronger

regulations of air pollution might lead to substantial monetary savings from

a public point of view partly because preterm infants are more likely to spend

a prolonged time in the hospital before being discharged and women

suffering from pre-eclampsia need more medical attention prior to delivery,

and partly because preterm children have increased morbidity. Supporting

this it was shown that preterm children were prescribed more asthma

medication than term children (Table 1 in Paper IV).

Summary of findings

Papers I and II showed a consistent association between first trimester O3

exposure and preterm delivery. These results support previous findings9,47,71

and have been repeated in later publications36-37. Paper II was one of the first

papers reporting an association between first trimester O3 and pre-

eclampsia37,85,87. The strong observed association between traffic-based air

pollution and pre-eclampsia in Paper III adds additional strength to the few

previously published findings37,85-87.

21

Future research

Future research should strive to explore if the spatial pattern of O3 is

associated with birth outcomes or childhood morbidity. Spatiotemporal

models should be implemented in order to improve the exposure assessment

of traffic-related air pollution and O3. Instead of using proxies for traffic

pollution, such as NO2 and NOx, more biologically active substances, such as

soot and elemental carbon, should be used. In order to further improve the

exposure assessment, occupational exposure and exposure levels during

commuting should be accounted for.

22

Conclusions

First trimester O3 was associated with an increased likelihood of

preterm delivery. There was some evidence of an association

between higher early pregnancy O3 levels and pre-eclampsia.

Traffic pollution at the home address during gestation was strongly

associated with pre-eclampsia.

There was an indication that traffic pollution at the home address

during gestation could lead to an increased risk of preterm delivery.

Living in a neighborhood with levels of traffic pollution above the 1st

quartile was associated with SGA.

There was no increased likelihood of being prescribed asthma

medication between five and six years of age among children living

in more polluted areas during their first year or among children

having a mother who lived in a more polluted area during

pregnancy.

23

Acknowledgements

I would like to thank:

My main supervisor Bertil Forsberg for guiding me through the process of

conducting scientific research, for your never-ending enthusiasm and

constant flow of ideas. A special thanks for every time I have been stuck on a

problem and you told me “But if you tried this…”, solving the issue.

My co-supervisors Magnus Ekström and Anders Bucht for all your help with

the conception and birth of the studies.

My co-authors Ingrid Mogren and Lennart Bråbäck for your brilliance and

contagious enthusiasm.

All the people at the Environment and Health Administration in

Stockholm who provided me with exposure data, in particular Christer

Johansson, Boel Löwenheim and my co-author Kristina Eneroth.

Lars Rylander, Eva Rönmark and Joacim Rocklöv for reviewing my

manuscripts and your helpful comments at the mid-seminar.

The administrators at the unit, Thomas Kling, Jenny Bagglund and Kristina

Lindblom for all your help.

My fellow PhD students over the years for listening to my complaints about

my lack of progress, allowing me to win a set or two when playing squash,

grabbing a sparkling beverage after work, and helping with my work.

My colleagues for your fantastic tirades about subjects ranging from

fireplaces to statistics and for telling me what’s what.

My friends from the crown of creation, Burträsk, whom I would name if I

wasn’t certain someone would be unintentionally omitted.

24

My parents, Bosse and Barbro, and sisters, Sofia and Johanna, with families

for all your love and support, additional thanks to my sister for teaching me

to read and write, so I wouldn’t have trouble keeping up with the curriculum

once school started, and for teaching me to chew with my mouth closed.

The Norlunds for babysitting and whatnot.

Edith and Imre for all the joy you give me, there is no feeling in the world

like when I come home and am met with a big smile and a delighted ‘Da’ or

‘Pappa’.

Sofia for letting me love you. Or in the words of Dylan: “There’s beauty in the

silver, singin’ river, There’s beauty in the sunrise in the sky; But none of

these and nothing else, can match the beauty That I remember in my true

love’s eyes”.

This thesis was funded by the Centre for Environmental Research in Umeå

(CMF) and is a part of the SIMSAM project.

25

References

1. Crump C, Winkleby MA, Sundquist J, Sundquist K. Risk of asthma in young adults who were born preterm: a Swedish national cohort study. Pediatrics 2011;127(4):e913-20.

2. Vogt H, Lindstrom K, Braback L, Hjern A. Preterm birth and inhaled

corticosteroid use in 6- to 19-year-olds: a Swedish national cohort study. Pediatrics 2011;127(6):1052-9.

3. Ekeus C, Lindstrom K, Lindblad F, Rasmussen F, Hjern A. Preterm birth, social disadvantage, and cognitive competence in Swedish 18- to 19-year-old men. Pediatrics 2010;125(1):e67-73.

4. Lindstrom K, Winbladh B, Haglund B, Hjern A. Preterm infants as young adults: a Swedish national cohort study. Pediatrics 2007;120(1):70-7.

5. Bobak M, Leon DA. Pregnancy outcomes and outdoor air pollution:

an ecological study in districts of the Czech Republic 1986-8. Occupational & Environmental Medicine 1999;56(8):539-43.

6. Chen L, Yang W, Jennison BL, Goodrich A, Omaye ST. Air pollution and birth weight in northern Nevada, 1991-1999. Inhal Toxicol 2002;14(2):141-57.

7. Bell ML, Ebisu K, Belanger K. Ambient air pollution and low birth weight in Connecticut and Massachusetts. Environ Health Perspect 2007;115(7):1118-24.

8. Huynh M, Woodruff TJ, Parker JD, Schoendorf KC. Relationships

between air pollution and preterm birth in California. Paediatr Perinat Epidemiol 2006;20(6):454-61.

9. Jiang LL, Zhang YH, Song GX, et al. A time series analysis of outdoor air pollution and preterm birth in Shanghai, China. Biomed Environ Sci 2007;20(5):426-31.

10. Myatt L, Webster RP. Vascular biology of preeclampsia. J Thromb Haemost 2009;7(3):375-84.

26

11. Uzan J, Carbonnel M, Piconne O, Asmar R, Ayoubi JM. Pre-eclampsia: pathophysiology, diagnosis, and management. Vasc Health Risk Manag 2011;7:467-74.

12. Kanasaki K, Kalluri R. The biology of preeclampsia. Kidney Int

2009;76(8):831-7.

13. Liu X, Olsen J, Agerbo E, et al. Birth weight, gestational age, fetal growth and childhood asthma hospitalization. Allergy Asthma Clin Immunol 2014;10(1):13.

14. Herzog R, Cunningham-Rundles S. Pediatric asthma: natural history, assessment, and treatment. Mt Sinai J Med 2011;78(5):645-60.

15. Ronmark E, Jonsson E, Platts-Mills T, Lundback B. Different

pattern of risk factors for atopic and nonatopic asthma among children--report from the Obstructive Lung Disease in Northern Sweden Study. Allergy 1999;54(9):926-35.

16. WHO. Global surveillance, prevention and control of Chronic Respiratory Disaeses. Geneva: World Health Organization Europe, 2007.

17. WHO. Air Quality Guidelines Global Update 2005: Particulate matter, ozone, nitrogen dioxide and sulfur dioxide. Geneva: World Helath Organization, 2006.

18. WHO. Review of evidence on health aspects of air pollution -

REVIHAAP Project. Bonn: World Health Organization Europe, 2013.

19. Bobak M, Richards M, Wadsworth M. Air pollution and birth weight in Britain in 1946. Epidemiology 2001;12(3):358-9.

20. Ha EH, Hong YC, Lee BE, et al. Is air pollution a risk factor for low birth weight in Seoul? Epidemiology 2001;12(6):643-8.

21. Maisonet M, Bush TJ, Correa A, Jaakkola JJ. Relation between

ambient air pollution and low birth weight in the Northeastern United States. Environ Health Perspect 2001;109 Suppl 3:351-6.

22. Wang X, Ding H, Ryan L, Xu X. Association between air pollution and low birth weight: a community-based study. Environ Health Perspect 1997;105(5):514-20.

27

23. Bobak M. Outdoor air pollution, low birth weight, and prematurity. Environ Health Perspect 2000;108(2):173-6.

24. Lee SJ, Hajat S, Steer PJ, Filippi V. A time-series analysis of any

short-term effects of meteorological and air pollution factors on preterm births in London, UK. Environ Res 2008;106(2):185-94.

25. Llop S, Ballester F, Estarlich M, et al. Preterm birth and exposure to air pollutants during pregnancy. Environ Res 2010;110(8):778-85.

26. Malmqvist E, Rignell-Hydbom A, Tinnerberg H, et al. Maternal exposure to air pollution and birth outcomes. Environ Health Perspect 2011;119(4):553-8.

27. Maroziene L, Grazuleviciene R. Maternal exposure to low-level air

pollution and pregnancy outcomes: a population-based study. Environ Health 2002;1(1):6.

28. Parker JD, Mendola P, Woodruff TJ. Preterm birth after the Utah Valley Steel Mill closure: a natural experiment. Epidemiology 2008;19(6):820-3.

29. Wu J, Ren C, Delfino RJ, et al. Association between local traffic-generated air pollution and preeclampsia and preterm delivery in the south coast air basin of California. Environ Health Perspect 2009;117(11):1773-9.

30. Xu X, Sharma RK, Talbott EO, et al. PM10 air pollution exposure

during pregnancy and term low birth weight in Allegheny County, PA, 1994-2000. Int Arch Occup Environ Health 2011;84(3):251-7.

31. Zhao Q, Liang Z, Tao S, Zhu J, Du Y. Effects of air pollution on neonatal prematurity in Guangzhou of China: a time-series study. Environ Health 2011;10:2.

32. Seo JH, Leem JH, Ha EH, et al. Population-attributable risk of low birthweight related to PM10 pollution in seven Korean cities. Paediatr Perinat Epidemiol 2010;24(2):140-8.

33. Ritz B, Wilhelm M, Hoggatt KJ, Ghosh JK. Ambient air pollution

and preterm birth in the environment and pregnancy outcomes study at the University of California, Los Angeles. Am J Epidemiol 2007;166(9):1045-52.

28

34. Chang HH, Reich BJ, Miranda ML. Time-to-event analysis of fine particle air pollution and preterm birth: results from North Carolina, 2001-2005. Am J Epidemiol 2012;175(2):91-8.

35. Ghosh JK, Wilhelm M, Su J, et al. Assessing the influence of traffic-

related air pollution on risk of term low birth weight on the basis of land-use-based regression models and measures of air toxics. Am J Epidemiol 2012;175(12):1262-74.

36. Le HQ, Batterman SA, Wirth JJ, et al. Air pollutant exposure and preterm and term small-for-gestational-age births in Detroit, Michigan: long-term trends and associations. Environ Int 2012;44:7-17.

37. Lee PC, Roberts JM, Catov JM, Talbott EO, Ritz B. First trimester exposure to ambient air pollution, pregnancy complications and adverse birth outcomes in Allegheny County, PA. Matern Child Health J 2013;17(3):545-55.

38. Padula AM, Mortimer K, Hubbard A, et al. Exposure to traffic-

related air pollution during pregnancy and term low birth weight: estimation of causal associations in a semiparametric model. Am J Epidemiol 2012;176(9):815-24.

39. Pereira G, Belanger K, Ebisu K, Bell ML. Fine particulate matter and risk of preterm birth in Connecticut in 2000-2006: a longitudinal study. Am J Epidemiol 2014;179(1):67-74.

40. Pereira G, Haggar F, Shand AW, et al. Association between pre-eclampsia and locally derived traffic-related air pollution: a retrospective cohort study. J Epidemiol Community Health 2013;67(2):147-52.

41. Pereira G, Nassar N, Cook A, Bower C. Traffic emissions are

associated with reduced fetal growth in areas of Perth, Western Australia: an application of the AusRoads dispersion model. Aust N Z J Public Health 2011;35(5):451-8.

42. Romao R, Pereira LA, Saldiva PH, et al. The relationship between low birth weight and exposure to inhalable particulate matter. Cad Saude Publica 2013;29(6):1101-8.

43. Wilhelm M, Ghosh JK, Su J, et al. Traffic-related air toxics and term low birth weight in Los Angeles County, California. Environ Health Perspect 2012;120(1):132-8.

29

44. Zhai D, Guo Y, Smith G, et al. Maternal exposure to moderate ambient carbon monoxide is associated with decreased risk of preeclampsia. American Journal of Obstetrics and Gynecology 2012;207(1):57 e1-9.

45. Lee PC, Talbott EO, Roberts JM, et al. Particulate Air Pollution

Exposure and C-reactive Protein During Early Pregnancy. Epidemiology 2011;22(4):524-31.

46. Gitto E, Reiter RJ, Karbownik M, et al. Causes of oxidative stress in the pre- and perinatal period. Biology of the Neonate 2002;81(3):146-157.

47. Hansen C, Neller A, Williams G, Simpson R. Maternal exposure to low levels of ambient air pollution and preterm birth in Brisbane, Australia. BJOG: An International Journal of Obstetrics and Gynaecology 2006;113(8):935-941.

48. Kannan S, Misra DP, Dvonch JT, Krishnakumar A. Exposures to

airborne particulate matter and adverse perinatal outcomes: a biologically plausible mechanistic framework for exploring potential effect modification by nutrition. Environ Health Perspect 2006;114(11):1636-42.

49. McConnell R, Berhane K, Yao L, et al. Traffic, susceptibility, and childhood asthma. Environ Health Perspect 2006;114(5):766-72.

50. Waldron G, Pottle B, Dod J. Asthma and the motorways--one District's experience. J Public Health Med 1995;17(1):85-9.

51. Andersson M, Modig L, Hedman L, Forsberg B, Ronmark E. Heavy

vehicle traffic is related to wheeze among schoolchildren: a population-based study in an area with low traffic flows. Environ Health 2011;10:91.

52. Gowers AM, Cullinan P, Ayres JG, et al. Does outdoor air pollution induce new cases of asthma? Biological plausibility and evidence; a review. Respirology 2012;17(6):887-98.

53. Miyake Y, Yura A, Iki M. Relationship between distance from major roads and adolescent health in Japan. J Epidemiol 2002;12(6):418-23.

30

54. van Vliet P, Knape M, de Hartog J, et al. Motor vehicle exhaust and chronic respiratory symptoms in children living near freeways. Environ Res 1997;74(2):122-32.

55. Venn A, Yemaneberhan H, Lewis S, Parry E, Britton J. Proximity of

the home to roads and the risk of wheeze in an Ethiopian population. Occupational & Environmental Medicine 2005;62(6):376-80.

56. Atkinson RW, Kang S, Anderson HR, Mills IC, Walton HA. Epidemiological time series studies of PM2.5 and daily mortality and hospital admissions: a systematic review and meta-analysis. Thorax 2014.

57. Strickland MJ, Darrow LA, Klein M, et al. Short-term associations between ambient air pollutants and pediatric asthma emergency department visits. American Journal of Respiratory & Critical Care Medicine 2010;182(3):307-16.

58. Braback L, Forsberg B. Does traffic exhaust contribute to the

development of asthma and allergic sensitization in children: findings from recent cohort studies. Environ Health 2009;8:17.

59. Guarnieri M, Balmes JR. Outdoor air pollution and asthma. Lancet 2014;383(9928):1581-92.

60. Tzivian L. Outdoor air pollution and asthma in children. J Asthma 2011;48(5):470-81.

61. Carlsten C, Dybuncio A, Becker A, Chan-Yeung M, Brauer M. Traffic-

related air pollution and incident asthma in a high-risk birth cohort. Occupational & Environmental Medicine 2011;68(4):291-5.

62. McConnell R, Islam T, Shankardass K, et al. Childhood incident asthma and traffic-related air pollution at home and school. Environ Health Perspect 2010;118(7):1021-6.

63. Zhou C, Baiz N, Zhang T, Banerjee S, Annesi-Maesano I. Modifiable exposures to air pollutants related to asthma phenotypes in the first year of life in children of the EDEN mother-child cohort study. BMC Public Health 2013;13:506.

64. Clark NA, Demers PA, Karr CJ, et al. Effect of early life exposure to

air pollution on development of childhood asthma. Environ Health Perspect 2010;118(2):284-90.

31

65. Fuertes E, Standl M, Cyrys J, et al. A longitudinal analysis of associations between traffic-related air pollution with asthma, allergies and sensitization in the GINIplus and LISAplus birth cohorts. PeerJ 2013;1:e193.

66. Ranzi A, Porta D, Badaloni C, et al. Exposure to air pollution and

respiratory symptoms during the first 7 years of life in an Italian birth cohort. Occupational & Environmental Medicine 2014;71(6):430-6.

67. Ho SM. Environmental epigenetics of asthma: an update. Journal of Allergy & Clinical Immunology 2010;126(3):453-65.

68. SMHI. Modules. http://www.smhi.se/airviro/modules/ Accessed March 31, 2014.

69. R Core Team. R: A Language and Environment for Statistical

Computing. Vienna, Austria: R Foundation for Statistical Computing, 2013.

70. Andersson-Sköld Y, Janhäll S. Simulations of ozone formation from different emission sources in Sweden. Water, Air, & Soil Pollution 1995;85(4):2045-2050.

71. Liu S, Krewski D, Shi Y, Chen Y, Burnett RT. Association between gaseous ambient air pollutants and adverse pregnancy outcomes in Vancouver, Canada. Environmental Health Perspectives 2003;111(14):1773-1778.

72. Ritz B, Yu F, Chapa G, Fruin S. Effect of air pollution on preterm

birth among children born in Southern California between 1989 and 1993. Epidemiology (Cambridge, Mass.) 2000;11(5):502-511.

73. Xu JL, Zhu YX, Li JL. Seasonal cycles of surface ozone and NOx in Shanghai. Journal of Applied Meteorology 1997;36(10):1424-1429.

74. Vingarzan R, Taylor B. Trend analysis of ground level ozone in the greater Vancouver/Fraser Valley area of British Columbia. Atmospheric Environment 2003;37(16):2159-2171.

75. Air Quality Expert Group. Ozone in the United Kingdom. London:

Department for the Environment, Food and Rural Affairs, 2009.

76. van der Mei IA, Ponsonby AL, Engelsen O, et al. The high prevalence of vitamin D insufficiency across Australian populations is only

http://www.smhi.se/airviro/modules/

32

partly explained by season and latitude. Environ Health Perspect 2007;115(8):1132-9.

77. Rucker D, Allan JA, Fick GH, Hanley DA. Vitamin D insufficiency in

a population of healthy western Canadians. CMAJ 2002;166(12):1517-24.

78. Moan J, Dahlback A, Ma L, Juzeniene A. Influenza, solar radiation and vitamin D. Dermatoendocrinol 2009;1(6):307-9.

79. McCullough ML, Weinstein SJ, Freedman DM, et al. Correlates of circulating 25-hydroxyvitamin D: Cohort Consortium Vitamin D Pooling Project of Rarer Cancers. Am J Epidemiol 2010;172(1):21-35.

80. Craig SM, Yu F, Curtis JR, et al. Vitamin D status and its

associations with disease activity and severity in African Americans with recent-onset rheumatoid arthritis. J Rheumatol 2010;37(2):275-81.

81. Ngo DT, Sverdlov AL, McNeil JJ, Horowitz JD. Does vitamin D modulate asymmetric dimethylarginine and C-reactive protein concentrations? Am J Med 2010;123(4):335-41.

82. Catov JM, Bodnar LM, Ness RB, Barron SJ, Roberts JM. Inflammation and dyslipidemia related to risk of spontaneous preterm birth. Am J Epidemiol 2007;166(11):1312-9.

83. Leem J-H, Kaplan BM, Shim YK, et al. Exposures to Air Pollutants

during Pregnancy and Preterm Delivery. Environmental Health Perspectives 2006;114(6):905-910.

84. Yorifuji T, Naruse H, Kashima S, et al. Residential proximity to major roads and adverse birth outcomes: a hospital-based study. Environ Health 2013;12(1):34.

85. Wu J, Wilhelm M, Chung J, Ritz B. Comparing exposure assessment methods for traffic-related air pollution in an adverse pregnancy outcome study. Environ Res 2011;111(5):685-92.

86. Malmqvist E, Jakobsson K, Tinnerberg H, Rignell-Hydbom A,

Rylander L. Gestational diabetes and preeclampsia in association with air pollution at levels below current air quality guidelines. Environ Health Perspect 2013;121(4):488-93.

33

87. van den Hooven EH, de Kluizenaar Y, Pierik FH, et al. Air pollution, blood pressure, and the risk of hypertensive complications during pregnancy: the generation R study. Hypertension 2011;57(3):406-12.

88. Glinianaia SV, Rankin J, Bell R, Pless-Mulloli T, Howel D.

Particulate air pollution and fetal health: a systematic review of the epidemiologic evidence. Epidemiology 2004;15(1):36-45.

89. Shah PS, Balkhair T. Air pollution and birth outcomes: a systematic review. Environ Int 2011;37(2):498-516.

90. Anderson HR, Favarato G, Atkinson RW. Long-term exposure to air pollution and the incidence of asthma: meta-analysis of cohort studies. Air Quality Atmosphere and Health 2013;6(1):47-56.

91. Lindgren A, Stroh E, Bjork J, Jakobsson K. Asthma incidence in

children growing up close to traffic: a registry-based birth cohort. Environ Health 2013;12:91.

92. Orru H, Andersson C, Ebi KL, et al. Impact of climate change on ozone-related mortality and morbidity in Europe. Eur Respir J 2013;41(2):285-94.