University of Southern California Dissertations and Theses

Evaluation of new methods for estimating exposure to traffic-related pollution and early health effects for large population epidemiological studies

(USC Thesis Other)

Evaluation of new methods for estimating exposure to traffic-related pollution and early health effects for large population epidemiological studies

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

EVALUATION OF NEW METHODS FOR ESTIMATING EXPOSURE TO TRAFFIC-
RELATED POLLUTION AND EARLY HEALTH EFFECTS FOR LARGE
POPULATION EPIDEMIOLOGICAL STUDIES

by

Donna Carmel Dueker

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(EPIDEMIOLOGY)

May 2012

Copyright 2012 Donna Carmel Dueker

ii
ACKNOWLEDGEMENTS

I would like to acknowledge the contributions of my collaborators and co-authors on
manuscripts in preparation. For the third chapter of this dissertation, Rob McConnell
assisted in study design, implementation and serves as a co-author on the manuscript in
preparation. Jim Gauderman provided statistical analysis advice and also is a co-author.
For the fourth chapter of this dissertation, my co-authors of the manuscript in preparation
Maryam Taher and John Wilson provided valuable GPS data analysis and advice on
interpreting my findings. Rob McConnell is also a co-author. For the fifth chapter, Scott
Fruin, Arantza Eiguren-Fernandez provided

on-road traffic pollution assessment and
laboratory analyses and guidance on interpreting the results of the pollutant
measurements. Lars Barregard, Andreas Sjodin, Samar Basu, Zheng Li, Junfeng Zhang,
Lena Samuelsson conducted the urinary biomarker laboratory analyses and gave
feedback on a manuscript in preparation for submission.W. James Gauderman provided
statistical advice. Rob McConnell assisted in the study design and collaborated on
interpreting the study’s findings.

iii
TABLE OF CONTENTS

Acknowledgements ii

List of Tables v

List of Figures vi

Abstract

Chapter 1: Introduction 1
Chapter 1 References 5

Chapter 2: Background and Significance 6
Chapter 2 References 12

Chapter 3: Accuracy of a Time-Resolved Step Counter 19
in Children
Abstract 19
Introduction 20
Methods 22
Statistical Analysis 24
Results 28
Discussion 34
Conclusion 38
Chapter 3 References 39

Chapter 4: Evaluation of a personal GPS logging instrument: 41
Limitations for assessing children’s locations
Abstract 41
Introduction 42
Methods 43
Data Analysis 48
Results 48
Discussion 63
Chapter 4 References 69

iv
Chapter 5: Urinary metabolites of polycyclic aromatic 72
hydrocarbon exposure and biomarkers
of early effect after travel on a busy road
Abstract 72
Introduction 73
Methods 76
Statistical Analysis 80
Results 81
Discussion 89
Chapter 5 References 100

Chapter 6 Conclusion 108
Chapter 6 References 116

Comprehensive References 118

v

LIST OF TABLES

Table 3.1: Participant characteristics 28

Table 3.2: Device and observer recorded steps 29

Table 3.3: Mean percent error of SportBrain and Digiwalker 29
counts

Table 3.4: Mean percent error of SportBrain and Digiwalker 32
by normal and overweight categories

Table 3.5: Distribution of daily average SportBrain wear 33
time and steps recorded

Table 5.1 Measurements of PAHs and BTEX on the 83
bus and during field trip

Table 5.2 Urinary biomarker measurements 85

Table 5.3 Intraclass correlation of pre-trip measurements 86

Table 5.4 Pre-trip and post-trip biomarker median values 87

Table 5.5 Sample size required to detect 10%, 20%, 96
and 30% change in mean urinary measurements

vi
LIST OF FIGURES

Figure 1.1: Overview 2

Figure 3.1a: Accelerometer chart used to determine wear 26
time of pedometer

Figure 3.1b: Example of accelerometer chart in which 27
non-wear time is not easily determined

Figure 3.2: Step counter recorded steps compared to observer 31
recorded steps during self-paced walking for
SportBrain (2a) and Digiwalker (2b)

Figure 4.1a: Wintec WBT 201, a wearable GPS data logger 45

Figure 4.1b Wintec WBT 201 covered with piece of Styrofoam 45

Figure 4.2 Representation of GPS raw data, showing example 46
of data with travel between locations on land and
well into the ocean between 15 second recordings

Figure 4.3 Percent possible waypoints recorded by day 49

Figure 4.4 Percent possible five minute intervals with at least 50
one location recorded, by day

Figure 4.5a Percent possible five minute time intervals 51
recorded with at least one location, weekdays

Figure 4.5b Percent possible five minute time intervals 51
recorded with at least one location, weekend

Figure 4.6a Percent possible waypoints recorded 12am-6am 53

Figure 4.6b Percent possible five minute time intervals 53
recorded with at least one location 12am to 6am

Figure 4.7a Nighttime (12am-6am) 15-second location 54
recordings with a large amount of scatter
outside a residential parcel (shaded)

vii
Figure 4.7b Location recording when restricted to 5-minute 55
rolling average during overnight hours

Figure 4.7c 5- minute rolling average with 50% data threshold 56

Figure 4.8a: Nighttime (12am-6am) 15-second interval 57
locations with few points outside the home
parcel (shaded)

Figure 4.8b Improved location recording when restricted 58
to 5-minute rolling average during overnight hours

Figure 4.8c Precision and accuracy was further improved 59
when restricted to 5-minute rolling average
with 50% data threshold

Figure 4.9 A study participant’s locations recorded while 60
at school

Figure 4.10a Typical freeway travel record 61

Figure 4.10b Travel path approaching school (in blue) 62
on small roadways for one participant

Figure 4.11 Percent of possible five-minute time intervals 63
with at least one location recorded by subject,
across days

Figure 5.1 Timing of data collection 78

Figure 5.2 Real time measurements of 82
black carbon (BC), C) and
NO
2
from the lead vehicle

Figure 5.3a Urinary 1-naphthol concentration by collection 88

Figure 5.3b: Urinary 1-naphthol averages of pre- 88
and post-trip collections

Figure 5.4 1-Aminopyrene by pooled pre- 88
and post-trip collections

viii

Figure 6.1: A way forward in epidemiological studies 108
of traffic-related air pollution and respiratory
health effects

Figure 6.2: Diurnal pattern of NO at community monitors 110
during weekdays in Long Beach and
Santa Barbara

ix
ABSTRACT

Objective: There is emerging evidence that local traffic-related pollution (TRP) has
adverse health effects that are independent of regional pollution effects. Current methods
to assess TRP exposure have limitations that may account for uncertainty and
inconsistency in the observed traffic-related health effects. New methods are needed to
assess TRP exposure in different microenvironments and to assess early biological effects
of these exposures in population based studies of air pollution. I evaluated new tools that
can be used to assess time and activity and also evaluated the impact on novel biological
markers of exposure to TRP and biomarkers of early biological effects.
Specifically the objectives of my research were: 1) to evaluate the accuracy of a
time-resolved step counter in children and the duration of consecutive zero step count
minutes that indicated non-wear time periods 2) evaluate the ability of a GPS data logger
to assess location of children during usual activity and 3) to assess feasibility of
collecting urine samples at school and to evaluate urinary biomarkers of exposure and
early effect of traffic-related air pollution. The overall goal of my research was to identify
methods that could greatly improve exposure assessments of TRP by providing an
integrated metric of dose that could be used to strengthen studies investigating the
relationship between TRP exposure and health affects.
Methods: To evaluate these tools, I conducted two studies. In the first study, a new time-
resolved step counter, the SportBrain, was evaluated for accuracy. Seventeen children
walked or ran on a treadmill at 2, 3, 4 and 5 miles/hour and walked around a track while
wearing the SportBrain and Digiwalker SW-701 pedometers. We compared percent error
in step counts for the two pedometers relative to observer counts. A sub-sample wore an

x
accelerometer and SportBrain pedometer during up to 5 days of usual activity. In the
second study, up to five urine samples per child were collected before and after a school
field trip with bus travel on a busy highway from fifteen 9-10 year olds recruited from
two classrooms in a low pollution region of Los Angeles. Samples were analyzed for
biomarkers of exposure (ten polycyclic aromatic hydrocarbon metabolites and 1-
aminopyrene) and of effect (Clara cell protein 16 and 8-iso-PGF
2α
, a major F
2
-
isoprostane). Four days of data from 17 children wearing GPS loggers recording every 15
seconds were evaluated for completeness by time of day during weekend and weekdays
and for accuracy during nighttime at home. Percentage of possible GPS recorded points
and of 5-minute intervals with at least one recorded location were examined.
Results: The SportBrain pedometer performed with acceptable accuracy at all evaluated
treadmill speeds and during self-paced walking, recording steps within an average of 4%
of observed step counts. During normal wear only 1% of zero count periods were less
than 60 minutes. 60% of participants collected all five urine samples. There was no
statistically significant difference between the pre- and post-trip measurements of any
biomarker. There was a high proportion of missing GPS location data. Mean percentage
of total possible 15-second interval locations recorded daily was less than 30%. Across
participants, the GPS loggers recorded 1% to 47% of total possible location points on
weekends and 1% to 55% on weekdays. More complete data were measured during travel
to school (average 90%).. During nighttime (12am-6am), on average, location was
recorded for less than 25% of 5-minute intervals and accuracy was poor. Urine collection
was feasible in a school setting. 60% of participants collected all urine five samples.

xi
However, there was no statistically significant difference between the pre- and post-trip
measurements of any biomarker.
Conclusion: The SportBrain iStep X1 pedometer provides a valid measure of step counts
in short averaging times useful for assessing patterns of physical activity in population
studies and periods of non-wear. Collecting urine samples from children at school was
feasible, but traffic effects were not detected after exposure in this small population
sample. The large proportion of missing data, which varied by location, limits the
usefulness of GPS logging instruments for population studies. They have potential utility
for assessing on-road travel time and route.

1

CHAPTER 1: INTRODUCTION

Goals of My Research: The broad goal of my dissertation research was to evaluate new
methods to assess exposure to traffic-related air pollution (TRP) based on locations in
key pollution microenvironments (e.g. in-transit, at home and school), methods that will
be applicable to large population epidemiological studies. I evaluated new tools that can
be used to assess time and physical activity, which can then potentially be used to
estimate inhaled dose in these locations. I also evaluated the impact on novel biological
markers of exposure to TRP and biomarkers of early biological effect. To evaluate these
new tools, I conducted two pilot studies. I designed the first study to evaluate the
accuracy of the SportBrain
TM
pedometer and its potential for estimating time in moderate
and vigorous physical activity (MVPA) (which can be linked to lung ventilation). The
second study was a quasi-experimental study evaluating these new methods using data I
collected. Traditional analysis of traffic and adverse health effects has examined the
effect of residential exposure (Gauderman, et al., 2007; McConnell, et al., 2006;
McConnell, et al., 2010). My research expands on this approach by including time,
location, and physical activity (Figure 1.1).

2

Figure 1.1: A way forward in epidemiological studies of traffic-related air pollution

My research was a methods-development project with a small sample size. It was
a pilot study within the Children’s Health Study (CHS). Part of the goal of my research
was to evaluate and demonstrate feasibility of these new methods. With demonstrated
feasibility, the methods can potentially be expanded for use in large population
epidemiological studies (such as the CHS).

STUDY 1: Accuracy of a Time-Resolved Step Counter in Children

Time spent in moderate and vigorous physical activity outdoors may be an important
determinant of exposure to traffic-related air pollutants (Long, et al., 2001 ). With the
increased ventilation of physical activity, the potential dose may be higher.
Accelerometers can be used to assess physical activity, but the accelerometer’s high cost
(>$200) limits their use in large population studies. NHANES has used accelerometers
with many subjects, however, it involved great effort and cost to equip the subjects and to
ensure minimal loss of the expensive equipment (Troiano, et al., 2008 ). For many
researchers, the cost of accelerometers would limit the practicality of use in large studies.
Time
Location
Physical
Activity
Traffic
Proximity (freeway
and major roads)
“Personal”
TRP
Exposure
Urinary
biomarkers:
Exposure (PAHs)

Effect (F2
Isoprostanes,
Clara Cell 16)

“Traditional” Analysis

“Potential”
Dose
Modeled TRP

3
The SportBrain
TM
is a low-cost alternative to accelerometers for physical activity
assessment. However, it has

not yet been validated for accuracy and reliability. Therefore,
I designed a study to evaluate accuracy of a new Sportbrain
TM
model (iStep X1),
comparing it to the gold standard of directly observed steps using methods previously
reported in the literature (Beets, Patton and Edwards, 2005; Le Masurier, Lee and Tudor-
Locke, 2004; Rowlands and Eston, 2005 ; Schneider, et al., 2003).
There is no consensus on the minimum interval with no activity needed to identify
periods of non-wear of motion sensors. For accelerometers between 10–60 minutes have
been used to define periods not worn (Kristensen, et al., 2008; Troiano, Berrigan, Dodd,
Masse, Tilert and McDowell, 2008 ). Therefore, because the pedometer may be less
sensitive to motion than accelerometers, I examined periods of zero counts while the
Sportbrain
TM
was worn to determine thresholds for non-compliance with wear for
potential use in larger studies.
The hypotheses for my study were:
1. The SportBrain
TM
pedometer is accurate in a controlled setting on a treadmill
and in a free walk around a track
2. The distribution of zero counts can be used to identify time periods of non-wear

Specific aims included:

1. To measure step counts with the Sportbrain
TM
and direct observation at selected
speeds on a treadmill and at normal paced walking around a track
2. To examine the distribution of periods of zero counts when the Sportbrain
TM
is
being worn by children (to identify the threshold of zero count time likely to
indicate non-wear).

4

STUDY 2: A Pilot Study of Urinary Metabolites of Exposure and Effect and Their
Relationship to On-Road Exposure and Evaluation of GPS to Assess Location

Specific Aims

A sample of 4
th
grade children was studied before, during, and after a field trip with
travel on a major traffic corridor in Los Angeles. Urine samples were collected pre- and
post- field trip in order to evaluate exposure. In addition, children were equipped with a
global positioning system (GPS) logger, and the children provided daily compliance and
activity diary data.
I tested the following hypotheses:
1. Urinary polyaromatic hydrocarbon (PAH) metabolites, Clara Cell secretory
protein (CC16) and F2-Isoprostane will increase significantly after exposure to
high levels of on-road traffic pollutants during the school field trip.
2. An increase in urinary PAH metabolites will be associated with increased
biological markers of effect (CC16 and F2-Isoprostane).
3. The GPS logger can be used to identify the location of study participants.
Specific aims included:

1. To measure urinary PAHs, F2-isoprostane, and CC16 pre- and post-field trip.
2. To evaluate assigning children to specific microenvironments (home, school, and
in transit) based on GPS data during the two days preceding and concurrent with
the urine collection.

5
CHAPTER 1 REFERENCES

Beets M.W., Patton M.M., and Edwards S. The accuracy of pedometer steps and time
during walking in children. Med Sci Sports Exerc 2005: 37(3): 513-520.

Gauderman W.J., Vora H., McConnell R., Berhane K., Gilliland F., Thomas D., et al.
Effect of exposure to traffic on lung development from 10 to 18 years of age: a cohort
study. Lancet 2007: 369(9561): 571-577.

Kristensen P.L., Moller N.C., Korsholm L., Wedderkopp N., Andersen L.B., and Froberg
K. Tracking of objectively measured physical activity from childhood to adolescence: the
European youth heart study. Scand J Med Sci Sports 2008: 18(2): 171-178.

Le Masurier G.C., Lee S.M., and Tudor-Locke C. Motion sensor accuracy under
controlled and free-living conditions. Med Sci Sports Exerc 2004: 36(5): 905-910.

Long C.M., Suh H.H., Catalano P.J., and Koutrakis P. Using time- and size-resolved
particulate data to quantify indoor penetration and deposition behavior. Environ Sci
Technol 2001: 35(10): 2089-2099.

McConnell R., Berhane K., Yao L., Jerrett M., Lurmann F., Gilliland F., et al. Traffic,
susceptibility, and childhood asthma. Environ Health Perspect 2006: 114(5): 766-772.

McConnell R., Islam T., Shankardass K., Jerrett M., Lurmann F., Gilliland F., et al.
Childhood incident asthma and traffic-related air pollution at home and school. Environ
Health Perspect 2010: 118(7): 1021-1026.

Rowlands A.V., and Eston R.G. Comparison of accelerometer and pedometer measures
of physical activity in boys and girls, ages 8-10 years. Res Q Exerc Sport 2005: 76(3):
251-257.

Schneider P.L., Crouter S.E., Lukajic O., and Bassett D.R., Jr. Accuracy and reliability of
10 pedometers for measuring steps over a 400-m walk. Med Sci Sports Exerc 2003:
35(10): 1779-1784.

Troiano R.P., Berrigan D., Dodd K.W., Masse L.C., Tilert T., and McDowell M. Physical
activity in the United States measured by accelerometer. Med Sci Sports Exerc 2008:
40(1): 181-188.

6

CHAPTER 2: BACKGROUND AND SIGNIFICANCE

Health Effects of Air Pollution: Regional outdoor air pollution is associated with a wide
range of adverse health effects including lung function deficits, asthma exacerbation,
respiratory symptoms such as cough and wheeze, cardiovascular disease, perhaps
developmental effects such as low birth weight, and possible neurobehavioral and
neurocognitive effects (Gauderman, et al., 2005; Gauderman, et al., 2007; Gotschi, et al.,
2008; Islam, et al., 2007; Kim, et al., 2008; Li, et al., 2003; Salam, Islam and Gilliland,
2008)
,
(Chen and Schwartz, 2009; Gauderman, et al., 2007; Sin, Wu and Man,
2005)
,
(Hole, et al., 1996; Knuiman, et al., 1999 ). Regulation of regional ambient air
pollution has reduced levels of criteria pollutants, but local traffic related sources remain
a public health concern, and ambient concentration measurements likely underestimate
actual exposures to traffic emissions. Nearly 10 million Americans live in areas
exceeding the regional NO
2
standard, and more than 20 million live in areas exceeding
the standard for particulate matter with aerodynamic diameter less than 10 microns
(PM
10
) (Anonymous, 1996 )
.
In addition, there have been observed effects of regional
pollution extending to levels below current regulatory standards (Kunzli, et al., 2003;
Schwartz, et al., 2008 ). Children are particularly susceptible to air pollution exposure
because their organs are still developing (Heinrich and Slama, 2007; Kajekar, 2007 ; Sly
and Flack, 2008; Soto-Martinez and Sly, 2009). Regional ambient pollutant exposures,
for example, have been strongly associated with childhood lung function deficits and
other respiratory health outcomes (Avol, et al., 2001; Frischer, et al., 1999; Gauderman,
et al., 2000; Gauderman, et al., 2002; Horak, et al., 2002; Jedrychowski, Flak and Mroz,

7
1999)
,
(Anonymous, 1996; Anonymous, 1996; Dockery, et al., 1989; Parker, Akinbami
and Woodruff, 2009 ; Peters, et al., 1999; Raizenne, et al., 1996; Schwartz, 1989; Ware,
et al., 1986).
There is emerging evidence that local traffic related pollution (TRP) has adverse
health effects that are independent of regional pollution effects (English, et al., 1999;
Holguin, et al., 2007; McConnell, et al., 2006; Salam, Islam and Gilliland, 2008 ). TRP
that vary locally with traffic proximity near homes (exposures which are currently largely
unregulated) have been associated with lung function deficits and with asthma, childhood
wheeze, and other asthma symptoms and morbidity in studies in Europe, Japan and the
US (Brauer, et al., 2002; English, Neutra, Scalf, Sullivan, Waller and Zhu, 1999;
Gauderman, et al., 2005; Kim, et al., 2004; Nicolai, et al., 2003; van Vliet, et al., 1997;
Venn, et al., 2000; Waldron, Pottle and Dod, 1995; Wjst, et al., 1993; Zmirou, et al.,
2004) (Gauderman, et al., 2005; McConnell, et al., 2006; Molitor, et al., 2006; Salam,
Islam and Gilliland, 2008 ). Traffic related air pollution has also been associated with
preterm delivery and preeclampsia (Wu, et al., 2009 ). Proximity to traffic is important in
assessing exposure. Studies show a sharp spatial gradient of TRP (Gilbert, et al., 2003 ;
Zhu, et al., 2002) and there is also diurnal variation in pollution levels (Zhu, et al., 2006 ).
Research Needs: There is also increasing interest in on-road exposures. Commuting is a
key route of exposure for TRP, but commuting exposure is challenging to evaluate. The
pollutants in the TRP mixture that are responsible for observed health associations are not
known. However, toxicologically relevant ultrafine particulate matter has been found to
be present in very high concentrations in the vicinity of major roadways (Zhu, Hinds,
Kim and Sioutas, 2002) and exceedingly high ultrafine exposures have been found in

8
vehicles and school buses (Behrentz, et al., 2005; Sabin, et al., 2005 ; Westerdahl, et al.,
2005). In Los Angeles 33-45% of daily ultrafine particulate exposure can occur from on-
road commuting exposure (Fruin, et al., 2008 ). Health effects of these on-road
exposures are not well-studied.
Most large scale epidemiological studies of air pollution have assessed exposure
with central site monitoring measurements, residential proximity to roads, or have used
modeled residential exposure to nearby fresh traffic emissions (Brauer, et al., 2002;
Brunekreef, et al., 1997 ; Gauderman, et al., 2007; Gotschi, et al., 2008; Kunzli, et al.,
2009; Nordling, et al., 2008). A major limitation of these exposure assessment methods
is that they do not account for exposures at other locations that may make substantial
contributions to overall exposure (for example, indoors and outdoors, at school or work,
or in transit) and the size of effects may be underestimated because these crude metrics
misclassify exposure (Jerrett, et al., 2005; Salam, Islam and Gilliland, 2008 ). In
addition, there are spatial variations in air pollution that may not be accounted for with
these methods (Zhu, Hinds, Kim and Sioutas, 2002; Zhu, Kuhn, Mayo and Hinds, 2006 ).
These limitations to exposure assessment may account for uncertainty and inconsistency
in the observed traffic-related health effects (English, Neutra, Scalf, Sullivan, Waller and
Zhu, 1999; Jerrett, 2007 ; Samet, 2007).
Recent studies suggest that time of day and locations of children’s activities are
important determinants of air pollution exposure and health effects, and there is evidence
that effects of air pollution exposure are modified by markers for potential lung dose
based on outdoor exercise and time spent outside (Gauderman, et al., 2002; McConnell,
et al., 2002; McConnell, et al., 2006 ). Increased rates of new onset asthma at schools

9
with higher TRP exposure have been observed (McConnell, et al., 2010). These
associations were independent of large adverse respiratory effects among children in
homes in “hot zones” within 75m of major roadways or 500m of freeways (McConnell,
et al., 2006 ). In communities with high TRP exposure, spending more time outdoors and
playing team sports that increase inhaled exposure resulted in more asthmatic bronchitic
symptoms and larger deficits in lung function growth (Gauderman, et al., 2002 ;
McConnell, et al., 2002). A limitation to these studies has been the crude assessment of
exercise and the inability to link exercise with location or time of day that may markedly
influence dose. Better characterization of time-activity patterns in key pollution
microenvironments is needed to identify high risk children and to develop prevention
strategies for reducing exposure at schools and in other key pollution microenvironments.
New Tools for Exposure Assessment: New methods clearly are needed to assess TRP
exposure in other locations, and to assess early biological effects of such exposures in
population based studies of air pollution. Fortunately, new tools have the potential to
improve exposure assessment at costs feasible in large population studies and thereby
improve the precision and accuracy of the estimates of health effects associated with
these exposures. These tools include personal global positioning system (GPS) loggers
that track location, data which can be used to assign time spent in different
microenvironments such as home, school and on-road (Elgethun, et al., 2003; Phillips, et
al., 2001; Rodriguez, Brown and Troped, 2005; Wiehe, et al., 2008 ). Research quality
accelerometers that can identify bursts of moderate and vigorous physical activity
(MVPA) associated with increased ventilation (and therefore increased potential pollutant
dose to the lung) are expensive to use in large studies, but a new pedometer,

10
SportBrain
TM
(based on a low-cost piezoelectric accelerometer), records step counts in
one minute epochs and has promise to identify MVPA at low cost.
Urinary biomarkers, including polyaromatic hydrocarbon (PAH) metabolites have
been used as indicators of exposure. Measurements of heavy occupational traffic-related
exposure to PAH’s have been found to correlate with these metabolites

and studies have
also found associations between traffic related pollutant exposure and urinary PAH’s
(Kang, et al., 2002 ). New biological markers of oxidative stress and inflammatory
damage to lungs for use in epidemiologic studies would enhance our understanding of the
effects of air pollution and might be used to assess early health effects of traffic exposure.
8-iso-prostaglandin-F2-alpha (F2-Isoprostane) can be measured in urine and is a marker
of systemic oxidative stress (Basu, 2008; Morrow, 2005 ). Oxidative stress has been
postulated to be an important mechanism for the pulmonary effects of oxidant ambient air
pollutants such as ozone and particulate matter (Gilliland, et al., 1999; Li, et al., 2003).
There is a developing literature examining the association of F2-IsoProstane with
exposure to another prototype air pollutant, tobacco smoke, and the results suggest that
urinary F2-IsoProstane might be a useful intermediate marker of the effect of oxidant air
pollutants in epidemiological studies (Bachi, et al., 1996; Chiabrando, et al., 1999;
Morrow, et al., 1995; Obata, et al., 2000 ). Urinary clara cell secretory protein 16 (CC16)
is a promising peripheral marker of damage to the lung (Broeckaert and Bernard, 2000 ).
CC16 can be measured in urine and was observed to be increased after experimental
exposure to wood smoke (Barregard, et al., 2006; Barregard, et al., 2008) and in
association with daily variation in ambient fine particulate exposure (Timonen, et al.,

11
2004 ). However, there has been little study of the feasibility of collection of urine
samples or in the utility of these markers for large population studies of children.

12

CHAPTER 2 REFERENCES

Anonymous. Health effects of outdoor air pollution. Part 2. Committee of the
Environmental and Occupational Health Assembly of the American Thoracic Society.
American Journal of Respiratory and Critical Care Medicine 1996: 153(2): 477-498.

Anonymous. Health effects of outdoor air pollution. Committee of the Environmental and
Occupational Health Assembly of the American Thoracic Society. American Journal of
Respiratory and Critical Care Medicine 1996: 153(1): 3-50.

Avol E.L., Gauderman W.J., Tan S.M., London S.J., and Peters J.M. Respiratory effects
of relocating to areas of differing air pollution levels. Am J Respir Crit Care Med 2001:
164(11): 2067-2072.

Bachi A., Zuccato E., Baraldi M., Fanelli R., and Chiabrando C. Measurement of urinary
8-Epi-prostaglandin F2alpha, a novel index of lipid peroxidation in vivo, by
immunoaffinity extraction/gas chromatography-mass spectrometry. Basal levels in
smokers and nonsmokers. Free Radic Biol Med 1996: 20(4): 619-624.

Barregard L., Sällsten G., Gustafson P., Andersson L., Johansson L., Basu S., et al.
Experimental Exposure to Wood-Smoke Particles in Healthy Humans: Effects on
Markers of Inflammation, Coagulation, and Lipid Peroxidation. Inhalation Toxicology
2006: 18(11): 845-853.

Barregard L., Sallsten G., Andersson L., Almstrand A.C., Gustafson P., Andersson M., et
al. Experimental exposure to wood smoke: effects on airway inflammation and oxidative
stress. Occup Environ Med 2008: 65(5): 319-324.

Basu S. F2-isoprostanes in human health and diseases: from molecular mechanisms to
clinical implications. Antioxid Redox Signal 2008: 10(8): 1405-1434.

Behrentz E., Sabin L.D., Winer A.M., Fitz D.R., Pankratz D.V., Colome S.D., et al.
Relative importance of school bus-related microenvironments to children's pollutant
exposure. J Air Waste Manag Assoc 2005: 55(10): 1418-1430.

Brauer M., Hoek G., Van Vliet P., Meliefste K., Fischer P.H., Wijga A., et al. Air
pollution from traffic and the development of respiratory infections and asthmatic and
allergic symptoms in children. Am J Respir Crit Care Med 2002: 166(8): 1092-1098.

Broeckaert F., and Bernard A. Clara cell secretory protein (CC16): characteristics and
perspectives as lung peripheral biomarker. Clin Exp Allergy 2000: 30(4): 469-475.

Brunekreef B., Janssen N.A., de Hartog J., Harssema H., Knape M., and van Vliet P. Air
pollution from truck traffic and lung function in children living near motorways.
Epidemiology 1997: 8(3): 298-303.

13

Chen J.C., and Schwartz J. Neurobehavioral effects of ambient air pollution on cognitive
performance in US adults. Neurotoxicology 2009: 30(2): 231-239.

Chiabrando C., Valagussa A., Rivalta C., Durand T., Guy A., Zuccato E., et al.
Identification and measurement of endogenous beta-oxidation metabolites of 8-epi-
Prostaglandin F2alpha. J Biol Chem 1999: 274(3): 1313-1319.

Dockery D.W., Speizer F.E., Stram D.O., Ware J.H., Spengler J.D., and Ferris B.G., Jr.
Effects of inhalable particles on respiratory health of children. Am Rev Respir Dis 1989:
139(3): 587-594.

Elgethun K., Fenske R.A., Yost M.G., and Palcisko G.J. Time-location analysis for
exposure assessment studies of children using a novel global positioning system
instrument. Environ Health Perspect 2003: 111(1): 115-122.

English P., Neutra R., Scalf R., Sullivan M., Waller L., and Zhu L. Examining
associations between childhood asthma and traffic flow using a geographic information
system. Environ Health Perspect 1999: 107(9): 761-767.

Frischer T., Studnicka M., Gartner C., Tauber E., Horak F., Veiter A., et al. Lung
function growth and ambient ozone: a three-year population study in school children. Am
J Respir Crit Care Med 1999: 160(2): 390-396.

Fruin S., Westerdahl D., Sax T., Sioutas C., and Fine P. Measurements and predictors of
on-road ultrafine particle concentrations and associated pollutants in Los Angeles.
Atmospheric Environment 2008: 42: 207-219.

Gauderman W.J., McConnell R., Gilliland F., London S., Thomas D., Avol E., et al.
Association between air pollution and lung function growth in southern California
children. Am J Respir Crit Care Med 2000: 162(4 Pt 1): 1383-1390.

Gauderman W.J., Gilliland G.F., Vora H., Avol E., Stram D., McConnell R., et al.
Association between air pollution and lung function growth in southern California
children: results from a second cohort. Am J Respir Crit Care Med 2002: 166(1): 76-84.

Gauderman W.J., Avol E., Lurmann F., Kuenzli N., Gilliland F., Peters J., et al.
Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology 2005:
16(6): 737-743.

Gauderman W.J., Vora H., McConnell R., Berhane K., Gilliland F., Thomas D., et al.
Effect of exposure to traffic on lung development from 10 to 18 years of age: a cohort
study. Lancet 2007: 369(9561): 571-577.

Gilbert N.L., Woodhouse S., Stieb D.M., and Brook J.R. Ambient nitrogen dioxide and
distance from a major highway. Sci Total Environ 2003: 312(1-3): 43-46.

14

Gilliland F.D., McConnell R., Peters J., and Gong H., Jr. A theoretical basis for
investigating ambient air pollution and children's respiratory health. Environ Health
Perspect 1999: 107 Suppl 3: 403-407.

Gotschi T., Heinrich J., Sunyer J., and Kunzli N. Long-term effects of ambient air
pollution on lung function: a review. Epidemiology 2008: 19(5): 690-701.

Gotschi T., Sunyer J., Chinn S., de Marco R., Forsberg B., Gauderman J.W., et al. Air
pollution and lung function in the European Community Respiratory Health Survey. Int J
Epidemiol 2008: 37(6): 1349-1358.

Heinrich J., and Slama R. Fine particles, a major threat to children. Int J Hyg Environ
Health 2007: 210(5): 617-622.

Hole D.J., Watt G.C., Davey-Smith G., Hart C.L., Gillis C.R., and Hawthorne V.M.
Impaired lung function and mortality risk in men and women: findings from the Renfrew
and Paisley prospective population study. BMJ 1996: 313(7059): 711-715; discussion
715-716.

Holguin F., Flores S., Ross Z., Cortez M., Molina M., Molina L., et al. Traffic-related
exposures, airway function, inflammation, and respiratory symptoms in children. Am J
Respir Crit Care Med 2007: 176(12): 1236-1242.

Horak F., Jr., Studnicka M., Gartner C., Spengler J.D., Tauber E., Urbanek R., et al.
Particulate matter and lung function growth in children: a 3-yr follow-up study in
Austrian schoolchildren. Eur Respir J 2002: 19(5): 838-845.

Islam T., Gauderman W.J., Berhane K., McConnell R., Avol E., Peters J.M., et al.
Relationship between air pollution, lung function and asthma in adolescents. Thorax
2007: 62(11): 957-963.

Jedrychowski W., Flak E., and Mroz E. The adverse effect of low levels of ambient air
pollutants on lung function growth in preadolescent children. Environ Health Perspect
1999: 107(8): 669-674.

Jerrett M., Arain A., Kanaroglou P., Beckerman B., Potoglou D., Sahsuvaroglu T., et al.
A review and evaluation of intraurban air pollution exposure models. J Expo Anal
Environ Epidemiol 2005: 15(2): 185-204.

Jerrett M. Does traffic-related air pollution contribute to respiratory disease formation in
children? Eur Respir J 2007: 29(5): 825-826.

Kajekar R. Environmental factors and developmental outcomes in the lung. Pharmacol
Ther 2007: 114(2): 129-145.

15
Kang J.W., Cho S.H., Kim H., and Lee C.H. Correlation of urinary 1-hydroxypyrene and
2-naphthol with total suspended particulates in ambient air in municipal middle-school
students in Korea. Arch Environ Health 2002: 57(4): 377-382.

Kim J.J., Smorodinsky S., Lipsett M., Singer B.C., Hodgson A.T., and Ostro B. Traffic-
related air pollution near busy roads: the East Bay Children's Respiratory Health Study.
Am J Respir Crit Care Med 2004: 170(5): 520-526.

Kim J.J., Huen K., Adams S., Smorodinsky S., Hoats A., Malig B., et al. Residential
traffic and children's respiratory health. Environ Health Perspect 2008: 116(9): 1274-
1279.

Knuiman M.W., James A.L., Divitini M.L., Ryan G., Bartholomew H.C., and Musk A.W.
Lung function, respiratory symptoms, and mortality: results from the Busselton Health
Study. Ann Epidemiol 1999: 9(5): 297-306.

Kunzli N., McConnell R., Bates D., Bastain T., Hricko A., Lurmann F., et al. Breathless
in Los Angeles: the exhausting search for clean air. Am J Public Health 2003: 93(9):
1494-1499.

Kunzli N., Bridevaux P.O., Liu L.J., Garcia-Esteban R., Schindler C., Gerbase M.W., et
al. Traffic-related air pollution correlates with adult-onset asthma among never-smokers.
Thorax 2009: 64(8): 664-670.

Li N., Sioutas C., Cho A., Schmitz D., Misra C., Sempf J., et al. Ultrafine particulate
pollutants induce oxidative stress and mitochondrial damage. Environ Health Perspect
2003: 111(4): 455-460.

McConnell R., Berhane K., Gilliland F., London S.J., Islam T., Gauderman W.J., et al.
Asthma in exercising children exposed to ozone: a cohort study. Lancet 2002: 359(9304):
386-391.

McConnell R., Berhane K., Yao L., Jerrett M., Lurmann F., Gilliland F., et al. Traffic,
susceptibility, and childhood asthma. Environ Health Perspect 2006: 114(5): 766-772.

McConnell R., Islam T., Shankardass K., Jerrett M., Lurmann F., Gilliland F., et al.
Childhood incident asthma and traffic-related air pollution at home and school. Environ
Health Perspect 2010: 118(7): 1021-1026.

Molitor J., Molitor N.T., Jerrett M., McConnell R., Gauderman J., Berhane K., et al.
Bayesian modeling of air pollution health effects with missing exposure data. Am J
Epidemiol 2006: 164(1): 69-76.

Morrow J.D., Frei B., Longmire A.W., Gaziano J.M., Lynch S.M., Shyr Y., et al.
Increase in circulating products of lipid peroxidation (F2-isoprostanes) in smokers.
Smoking as a cause of oxidative damage. N Engl J Med 1995: 332(18): 1198-1203.

16

Morrow J.D. Quantification of isoprostanes as indices of oxidant stress and the risk of
atherosclerosis in humans. Arterioscler Thromb Vasc Biol 2005: 25(2): 279-286.

Nicolai T., Carr D., Weiland S.K., Duhme H., von Ehrenstein O., Wagner C., et al. Urban
traffic and pollutant exposure related to respiratory outcomes and atopy in a large sample
of children. Eur Respir J 2003: 21(6): 956-963.

Nordling E., Berglind N., Melen E., Emenius G., Hallberg J., Nyberg F., et al. Traffic-
related air pollution and childhood respiratory symptoms, function and allergies.
Epidemiology 2008: 19(3): 401-408.

Obata T., Tomaru K., Nagakura T., Izumi Y., and Kawamoto T. Smoking and oxidant
stress: assay of isoprostane in human urine by gas chromatography-mass spectrometry. J
Chromatogr B Biomed Sci Appl 2000: 746(1): 11-15.

Parker J.D., Akinbami L.J., and Woodruff T.J. Air pollution and childhood respiratory
allergies in the United States. Environ Health Perspect 2009: 117(1): 140-147.

Peters J.M., Avol E., Gauderman W.J., Linn W.S., Navidi W., London S.J., et al. A study
of twelve Southern California communities with differing levels and types of air
pollution. II. Effects on pulmonary function. Am J Respir Crit Care Med 1999: 159(3):
768-775.

Phillips M.L., Hall T.A., Esmen N.A., Lynch R., and Johnson D.L. Use of global
positioning system technology to track subject's location during environmental exposure
sampling. J Expo Anal Environ Epidemiol 2001: 11(3): 207-215.

Raizenne M., Neas L.M., Damokosh A.I., Dockery D.W., Spengler J.D., Koutrakis P., et
al. Health effects of acid aerosols on North American children: pulmonary function.
Environ Health Perspect 1996: 104(5): 506-514.

Rodriguez D.A., Brown A.L., and Troped P.J. Portable global positioning units to
complement accelerometry-based physical activity monitors. Med Sci Sports Exerc 2005:
37(11 Suppl): S572-581.

Sabin L.D., Behrentz E., Winer A.M., Jeong S., Fitz D.R., Pankratz D.V., et al.
Characterizing the range of children's air pollutant exposure during school bus commutes.
J Expo Anal Environ Epidemiol 2005: 15(5): 377-387.

Salam M.T., Islam T., and Gilliland F.D. Recent evidence for adverse effects of
residential proximity to traffic sources on asthma. Curr Opin Pulm Med 2008: 14(1): 3-8.

Samet J.M. Traffic, air pollution, and health. Inhal Toxicol 2007: 19(12): 1021-1027.

17
Schwartz J. Lung function and chronic exposure to air pollution: a cross-sectional
analysis of NHANES II. Environ Res 1989: 50(2): 309-321.

Schwartz J., Coull B., Laden F., and Ryan L. The effect of dose and timing of dose on the
association between airborne particles and survival. Environ Health Perspect 2008:
116(1): 64-69.

Sin D.D., Wu L., and Man S.F. The relationship between reduced lung function and
cardiovascular mortality: a population-based study and a systematic review of the
literature. Chest 2005: 127(6): 1952-1959.

Sly P.D., and Flack F. Susceptibility of children to environmental pollutants. Ann N Y
Acad Sci 2008: 1140: 163-183.

Soto-Martinez M., and Sly P. Relationship between environmental exposures in children
and adult lung disease: The case for outdoor exposures. Chron Respir Dis 2009.

Timonen K.L., Hoek G., Heinrich J., Bernard A., Brunekreef B., de Hartog J., et al. Daily
variation in fine and ultrafine particulate air pollution and urinary concentrations of lung
Clara cell protein CC16. Occup Environ Med 2004: 61(11): 908-914.

van Vliet P., Knape M., de Hartog J., Janssen N., Harssema H., and Brunekreef B. Motor
vehicle exhaust and chronic respiratory symptoms in children living near freeways.
Environ Res 1997: 74(2): 122-132.

Venn A., Lewis S., Cooper M., Hubbard R., Hill I., Boddy R., et al. Local road traffic
activity and the prevalence, severity, and persistence of wheeze in school children:
combined cross sectional and longitudinal study. Occup Environ Med 2000: 57(3): 152-
158.

Waldron G., Pottle B., and Dod J. Asthma and the motorways--one District's experience.
J Public Health Med 1995: 17(1): 85-89.

Ware J.H., Ferris B.G., Jr., Dockery D.W., Spengler J.D., Stram D.O., and Speizer F.E.
Effects of ambient sulfur oxides and suspended particles on respiratory health of
preadolescent children. Am Rev Respir Dis 1986: 133(5): 834-842.

Westerdahl D., Fruin S., Sax T., Fine P.M., and Sioutas C. Mobile platform
measurements of ultrafine particles and associated pollutant concentrations on freeways
and residential streets in Los Angeles. Atmospheric Environment 2005: 39(20): 3597-
3610.

Wiehe S.E., Carroll A.E., Liu G.C., Haberkorn K.L., Hoch S.C., Wilson J.S., et al. Using
GPS-enabled cell phones to track the travel patterns of adolescents. Int J Health Geogr
2008: 7: 22.

18
Wjst M., Reitmeir P., Dold S., Wulff A., Nicolai T., von Loeffelholz-Colberg E.F., et al.
Road traffic and adverse effects on respiratory health in children. BMJ 1993: 307(6904):
596-600.

Wu J., Ren C., Delfino R.J., Chung J., Wilhelm M., and Ritz B. 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): 1773-1779.

Zhu Y., Hinds W.C., Kim S., and Sioutas C. Concentration and size distribution of
ultrafine particles near a major highway. J Air Waste Manag Assoc 2002: 52(9): 1032-
1042.

Zhu Y., Kuhn T., Mayo P., and Hinds W.C. Comparison of daytime and nighttime
concentration profiles and size distributions of ultrafine particles near a major highway.
Environ Sci Technol 2006: 40(8): 2531-2536.

Zmirou D., Gauvin S., Pin I., Momas I., Sahraoui F., Just J., et al. Traffic related air
pollution and incidence of childhood asthma: results of the Vesta case-control study. J
Epidemiol Community Health 2004: 58(1): 18-23.

19

CHAPTER 3: ACCURACY OF A TIME-RESOLVED STEP
COUNTER IN CHILDREN

CHAPTER 3 ABSTRACT

Most pedometers record cumulative steps, limiting ability to assess level of physical
activity or non-wear periods. The SportBrain iStep X1 has potential to overcome this
limitation by recording and storing step count data in 60-second epochs. We evaluated
the accuracy of this instrument in children and the duration of consecutive zero step count
minutes that indicated non-wear time periods. Seventeen children walked or ran on a
treadmill at 2, 3, 4 and 5 miles/hour and walked around a track while wearing the
SportBrain and Digiwalker SW-701 pedometers. We compared percent error in step
counts for the two pedometers relative to observer counts. A sub-sample wore an
accelerometer and SportBrain pedometer during up to 5 days of usual activity. The
SportBrain pedometer performed with acceptable accuracy at all evaluated treadmill
speeds and during self-paced walking, recording steps within an average of 4% of
observed step counts. During normal wear only 1% of zero count periods were less than
60 minutes. We conclude that the SportBrain iStep X1 pedometer provides a valid
measure of step counts in short averaging times useful for assessing patterns of physical
activity in population studies and periods of non-wear.

20
CHAPTER 3 INTRODUCTION
Pedometers and accelerometers are motion sensors that measure physical activity.
Motion sensors provide objective direct methods of physical activity measurement and
have advantages over self-reported activity, especially in children, for whom self-report
may be less reliable than in adults (Scruggs, et al., 2003; Tudor-Locke, et al., 2006 ).
Research accelerometers accurately measure physical activity and record movement in
intervals from 1 to 60 seconds. Although research accelerometers have been used in large
population studies of physical activity (Troiano, Berrigan, Dodd, Masse, Tilert and
McDowell, 2008), their high cost (>$200) is a limitation, especially in studies of children
who may be more likely to lose the instruments.
Pedometers are a low cost method to measure physical activity objectively. Like
accelerometers, they have little subject burden and can be unobtrusively worn on a belt or
waist band. Pedometer counts are strongly correlated with accelerometer output. In a
review of 25 studies the median reported correlation was 0.86 between the two
instruments (Tudor-Locke, et al., 2002 ). Pedometers, like accelerometers, have been
correlated with energy expenditure in laboratory and field studies(Eston, Rowlands and
Ingledew, 1998; Stone, Esliger and Tremblay, 2007), and new hybrid models of
accelerometer-based step counters are more accurate than earlier step counters at
comparably low cost (Crouter, Schneider and Bassett, 2005 ).
Pedometers have limitations for use in studies of physical activity because most
record cumulative steps taken. Although some models record and store daily total step
counts for several days, this temporal resolution is not adequate to identify specific

21
periods of physical activity or to classify physical activity by intensity level. For
example, a subject taking 10,000 steps daily might accumulate the steps over the course
of the day through walking or by taking a single five mile run.
A new model of pedometer, the SportBrain
TM
iStep X1, uses a piezoelectric
accelerometer-based mechanism to record step counts and overcomes some of the
limitations of other pedometers. For example, the SportBrain records and stores time-
resolved steps in 60 second epochs in addition to displaying cumulative step counts. Step
count data can be downloaded into a spreadsheet for further analysis. (The SportBrain
step counter and assess to time-resolved data are available from Device-
Research@humana.com). From these data, patterns of physical activity can be identified.
The data could potentially be used to identify periods of non-compliance, based on long
periods with zero steps, and periods of moderate and vigorous physical activity, based on
a threshold of steps per minute.
In this study we evaluated the accuracy of the SportBrain iStep X1 pedometer in
children in both paced treadmill and free walking conditions. We compared the accuracy
of the SportBrain

in normal and overweight children to a gold standard of directly
observed steps and to the Digiwalker SW-701, a brand which has demonstrated accuracy
in several previous validation studies (Crouter, et al., 2003; Schneider, Crouter and
Bassett, 2004 ). There is no consensus on the minimum interval with no activity needed
to identify periods of non-wear of motion sensors and this assessment has not been made
for pedometers because they have not recorded in short intervals. For accelerometers
between 10–60 minutes have been used to define periods not worn (Kristensen, Moller,

22
Korsholm, Wedderkopp, Andersen and Froberg, 2008; Troiano, Berrigan, Dodd, Masse,
Tilert and McDowell, 2008 ). Pedometers may be less sensitive to motion than
accelerometers, especially in children; therefore, we examined periods of zero counts
while the Sportbrain
TM
was worn to define thresholds of zero counts which might be used
to indicate periods of non-wear.

CHAPTER 3 METHODS
Participants
A convenience sample of 17 children (10 females, 7 males) aged 10-17
participated in the Sportbrain validation study. A sub-sample of 9 children (3 females, 6
males) participated in an evaluation of thresholds of continuous minutes of zero count
suggestive of non-wear time. The study was approved by the Institutional Review Board
at the University of Southern California. Parental informed consent and child assent were
obtained before testing.
Procedures
Height and weight of each participant, in light clothing without shoes, were
measured by study staff. We calculated BMI percentiles using the BMI SAS program
from the Centers for Disease Control and Prevention website (National, ). Children with
BMI less than the 85
th
percentile were classified as normal weight (6 participants).
Children with BMI greater than the 85
th
percentile were classified as overweight or obese
(11 participants).

23
Treadmill evaluation: Each child walked and ran on a Trackmaster TMX425 treadmill
while wearing SportBrain and Digiwalker pedometers attached to pants/shorts on the
right side of the waist. The SportBrain was worn midway between the midline and
anterior superior iliac crest, and the Digiwalker was placed laterally to the SportBrain.
Prior to use, each device was shaken manually 100 times and the step count display
checked to verify that the shakes were recorded.
Participants walked or ran on a treadmill at 2, 3, 4, and 5 mph for 3 minutes at
each speed. These speeds were selected to elicit slow to fast walking and running. At 5
mph every participant ran. Participants straddled the treadmill belt prior to each 3 minute
bout while the pedometers were reset to zero. Two staff members counted steps with a
click counter. Every foot strike was counted at 2mph and 3mph and every other foot
strike at 4mph and 5mph (and the count doubled).
Usual paced walk evaluation: To test the pedometers during normal walking, 13 of the 17
participants wore the devices while walking around a school track. Children were
instructed to walk at their usual walking pace. Each child walked one lap around the track
while two observers followed and recorded every other foot strike. After recording
pedometer and directly observed step counts, the devices were reset to zero and the
children walked a second lap. The pedometer- and observer-recorded step count values
for lap 1 and for lap 2 were then summed.
Usual activity non-wear evaluation: Participants wore the SportBrain
TM
and an Actigraph
7164 accelerometer for up to five days during waking hours. Participants were contacted
via telephone or text message every evening to remind them to wear the devices and to

24
complete a very brief structured checklist identifying periods of non-wear (and the
reason).
CHAPTER 3 STATISTICAL ANALYSIS
Validation study: We calculated the intraclass coefficients (ICC) to assess the agreement
between observers 1 and 2. We had high agreement at all speeds and during self paced
walking (ICC > 0.96, after excluding the counts at 5mph for a single subject with a
discrepancy of 42 steps between the two observers). Therefore, we used the average of
observer 1 and observer 2 as a “gold standard” for the evaluation of the pedometers.
At each speed on the treadmill and during self-paced walking around the track we
calculated the percent error for each pedometer count relative to the observers’ average
count, i.e. Percent error= [(Pedometer steps-Observed steps)/observed steps] x 100. We
also calculated the mean of the absolute value of percent errors since the averaging of
over- and under-counting can overestimate the device’s accuracy (Le Masurier, Lee and
Tudor-Locke, 2004) . We used paired t-tests to examine the difference in mean percent
errors between the SportBrain and the Digiwalker in self-paced walking and an F test to
assess the difference in the variability of these errors. We also used t tests to examine the
difference in mean percent errors in self-paced walking between overweight and normal
weight participants within each device.
We used analysis of variance (ANOVA) to investigate the association of percent
error with device, speed, and BMI percentile during treadmill walking/running.
Specifically, we used split plot ANOVA to account for between-subject (weight) and
within-subject variables (speed and device). We tested the three way interaction, then

25
each two-way interaction, and independent effects. An alpha of 0.05 was considered
statistically significant. The effect of BMI percentile on percent error was also examined
as a dichotomized (normal weight and overweight) variable with normal and overweight
categorized as described above. We also examined the interaction between weight (as
categorical or continuous BMI percent) and device during self-paced walking using these
analysis of variance methods.
Usual activity non-wear analysis: The daily recall of non-wear times was not useful as
there were multiple reported non-wear periods during which activity was recorded on the
accelerometer and pedometer. However, with rare exceptions, non-wear periods in these
children were apparent on visual inspection of the tracing of the accelerometer time series
(see Figure 3.1a for a typical tracing). Sedentary activity registered with low counts that
were frequent, but periods not wearing the accelerometer demonstrated no activity.
There was one child for whom there were isolated, low accelerometer counts during
periods with no step counts on a single day (Figure 3.1b), perhaps reflecting movement of
an instrument belt not being worn. This time period between 9 and 11 AM was classified
as non-wear time.

26
Figure 3.1a: Accelerometer chart used to determine wear time of pedometer

0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
ACTIVITY
TIME
Period of non-wear

27

Figure 3.1b: Example of accelerometer chart in which non-wear time is not easily
determined

The accelerometer time series data were inspected for each participant to identify
periods of time of non-wear. The accelerometer-determined times of non-wear were
excluded from the analysis of the distribution of duration of zero pedometer step counts.
The distribution of consecutive zero step counts was examined for each participant for
each day, for each participant with all days combined, and for all participants combined.
We also examined the total number of minutes with greater than 130 steps, a value that
has been shown to be an approximate threshold for moderate and vigorous physical
activity in children (Lubans, et al., 2009).
All analyses were performed using SAS version 9.2.

0
500
1000
1500
2000
2500
3000
ACTIVITY
TIME
Period of
indeterminate wear

28
CHAPTER 3 RESULTS
Validation study:
The sample included 10 girls and 7 boys, with an average age of 12.8 years (Table
3.1). Most participants were overweight (11 of 17; 65%). The mean Digiwalker step
count was lower at all speeds and during the walk around the track than the directly
observed count (Table 3.2). The variability (standard deviation) was greater for the
Digiwalker than for directly observed counts. The large variability was only partially
explained by the impact of zero step counts recorded for one obese participant at 2 and 3
mph. The mean step count, variability and range for the SportBrain were generally
similar to the measurements from direct observation.
Table 3.1: Participant characteristics (N=17)

Mean (SD) Range
Age (years) 12.8 (1.7) 10-17
Height (cm) 157 (7.4) 145-170
Weight (kg) 65 (16.1) 42-104
BMI (%) 87.3 (14.8) 45.3-99.4

BMI %
Normal wt:
Overweight/obese:
N (%)
6 (35)
11 (65)

29

Speed SportBrain Counts
*
Mean (SD) Range
Digiwalker Counts*

Mean (SD) Range
Observed
*†
Mean (SD) Range
2mph 314 (31.5) 254-360 257 (95.3) 0-360 318 (29.4) 264-360
3mph 364 (23.5) 323-409 334 (92.9) 0-407 366 (23.8) 323-410
4mph 404 (33.6) 251-454 368 (124.8) 3-454 418 (21.2) 389-459
5mph 484 (43.9) 368-540 484 (90.6) 168-575 506 (34.9) 445-588
Track 1177 (76.1) 1071-1327 1141(137.6) 774-1328 1176 (73.4) 1070-
1327
*
N=17 except at 5 mph N=16 and on track N=13
†
Average of counter 1 and counter 2

For the treadmill evaluation of percent error, the three way interactions between
device, speed, and weight category (p=0.89) or continuous BMI percentile (p=0.54) were
not significant. There was a statistically significant interaction between device and speed
(p=0.003), adjusted for the effect of weight category. At each treadmill speed the
SportBrain recorded steps, on average, within 4% of observer recorded steps (Table 3.3).

Table 3.3: Mean percent error of SportBrain and Digiwalker counts

Speed SportBrain
Percent Error
Mean(SD)

Digiwalker
Percent Error
Mean (SD)

2mph -1.3 (3.5) -19.3 (28.4)
3mph -0.5 (1.4) -8.3 (35.5)
4mph -3.2 (10.1) -11.8 (29.9)
5mph
*
-3.9 (9.6) -4.3 (16.5)
Track
*
0.06 (0.6) -2.8 (11.0)
*
N=17 except at 5 mph N=16 and on track N=13

The Digiwalker undercounted steps by almost 20% at 2mph, decreasing to less than 5%
at 5 mph (p=0.008 for the trend in Digiwalker percent error across speed). The variability
in error at each speed was also considerably larger for the Digiwalker than for the
Table 3.2: Device and observer recorded steps

30
SportBrain. At 5 mph, the Digiwalker and SportBrain had similar percent errors and the
difference was not significant. However, the standard deviation in percent error at 5 mph
was significantly larger (p=0.04) in the Digiwalker (SD=16.5) than in the SportBrain
(SD=9.6). During self-paced walking on the track both the SportBrain and the Digiwalker
had low mean percent error. However, the variability was larger for the Digiwalker
(SD=11.0) than for the SportBrain (SD=0.6), a difference that was highly significant
(p<0.001). This larger variability and the pattern of Digiwalker undercount compared
with the observed counts is apparent in Figure 3.2b and was not seen for the SportBrain
(Figure 3.2a). We also calculated the mean of the absolute value of the percent errors.
These values were almost identical to the mean percent errors because the error was
largely due to undercounting.

31
Figure 3.2: Step counter recorded steps compared to observer recorded
steps during self-paced walking for SportBrain (3.2a) and Digiwalker (3.2b)

After stratifying by device, neither the interaction of speed with weight category
nor with BMI percentile was significantly associated with percent error. The main effect
of BMI percentile or weight category also was not significant for either instrument.

32
However, the undercounts were generally greater among overweight children for both
instruments and statistical power to identify relatively modest differences in percent error
in this small sample of children was limited (Table 3.4). There was more variability in the
percent error of both instruments among overweight than among normal weight
participants, especially for the Digiwalker. For example, the standard deviation of the
percent error for the Digiwalker during the self-paced walk on a track (13.2) was
significantly greater in the overweight participants than the SD (0.5) of the normal weight
participants. Although the SD of the percent error for the SportBrain did not vary by
weight category during the self-paced walk (p=0.33), the variance was significantly
greater among overweight compared with normal weight children at 4 and 5mph
treadmill speeds (p=0.007 and p<0.001 respectively).

Table 3.4: Mean percent error of SportBrain and Digiwalker by normal
and overweight categories
Speed SportBrain Percent Error Digiwalker Percent Error
Normal Wt
(N=6)
*
Mean(SD)
Overweight
(N=11)

Mean(SD)
Normal Wt
(N=6)

Mean(SD)
Overweight
(N=11)

Mean(SD)
2mph -0.88 (1.5) -1.5 (4.3) -15.1 (20.2) -21.6 (32.8)
3mph -1.03 (2.1) -0.19 (0.7) -7.9 (15.2) -8.6 (30.4)
4mph -1.7 (3.10) -4.06 (12.5) -3.8 (8.3) -16.1 (36.6)
5mph -0.3 (0.5) -5.6 (11.4) 0.03 (0.4) -6.2 (19.9)
Track -0.06 (0.4) 0.12 (0.7) -0.07 (0.5) 4.1 (13.2)
*
N=5 at 5mph; on track N=4 for normal weight and N=9 for overweight

33
Non-wear study:
Nine children aged 13-16 participated in the study. All children were instructed to
wear the SportBrain

pedometer and accelerometer for five days. However, two
participants had unusable data because of instrument malfunction (the accelerometer for
one and pedometer for the other), so useable data were available for analysis for seven
children (1 girl, 6 boys). Based on inspection of the accelerometer tracing, 3 children
provided data for all five days. Of the other four, three participants wore the devices for
four days and one for three days. A total of 30 days had useable data for analysis.
The pedometers were worn for an average 501 minutes daily (55% of waking
hours based on self-reported hours awake, information which was available for 29 of the
30 days for which data were collected), but this varied from an hour (9%) on average for
one child to essentially the entire waking 5-day period (97% assuming 980 minutes (16.3
hours) awake daily; Table 3.5).
Table 3.5 Distribution of daily average SportBrain wear time and steps recorded (N=30
days*)

Mean(SD) Min 25% Median 75% Max
Minutes Worn Per Day 501 (224)
60 381 514 647 950
Percent of Time Awake Worn** 55 (24)
9 43 54 73 97
Total Steps
7839 (4434)
733 5016 7553 11050
16584
Percent Zero Steps** 61 (13)
14 52 65 69 80
Minutes >= 130 steps 0.7 (1.5)
0 0 0 1 7
Minutes >= 100 steps 25 (24)
0 6 17 44 85
Consecutive minutes with zero steps 7.6 (13)
1 1 3 8 158
____________________________________________________________________________
*N= Number of person days with collected pedometer data
**(Minutes of zero steps/minutes of pedometer wear time) x 100; N=29 days because one
participant did not report time arising and going to bed

34
The total number of daily steps across days and individuals combined ranged from 733 to
16584. The mean steps per day was 7839 (SD=4434). There were more than 10,000 steps
per day for 37% of the days. However, based on the number of minutes with more than
130 steps, there were few periods of MVPA (no day with more than 7 minutes for any
participant). There were zero steps on average for 61% of the minutes worn during which
the Sportbrain was worn (after removing minutes of non-wear based on the Actigraph),
but the average number of minutes of SportBrain recorded consecutive zero steps was
7.6. Seventy-five percent of the consecutive zero count step intervals was 8 minutes or
less. The maximum zero count interval was 158 minutes. Ninety-five percent of the
consecutive zero steps were less than 30 minutes during normal wear, and 99% of zero
steps intervals were less than 58 minutes (results not shown).

CHAPTER 3 DISCUSSION
The SportBrain iStep X1 demonstrated accurate and precise step counting at all
tested speeds and in self-paced walking in both normal and overweight children, and
outperformed the Digiwalker, especially at slower speeds and in overweight children. In
previous studies, Digiwalker pedometers have demonstrated more accurate counts at
higher speeds (Beets, Patton and Edwards, 2005 ). Previous studies have also found
pedometers to be less accurate in obese adults and children (Beets, Patton and Edwards,
2005 ; Crouter, Schneider and Bassett, 2005; Mitre, et al., 2009).
The reduced accuracy of pedometers in overweight and obese participants and at
slow walking speeds has been associated with the pedometer mechanism (Crouter,
Schneider and Bassett, 2005; Melanson, et al., 2004 ). The Digiwalker has a spring-

35
loaded horizontal lever that moves up and down with the hip’s vertical movements. The
movement of the lever opens and closes an electrical circuit and the lever arm makes an
electrical contact and a step is counted (Crouter, Schneider and Bassett, 2005 ).
Accelerometer-based devices like the SportBrain iStep X1 count steps based on
movement detected by the piezo-electric accelerometer. When a horizontal cantilevered
beam compresses a piezo-electric crystal during movement, the resulting voltage
oscillations are used to record steps (Crouter, Schneider and Bassett, 2005 ). Our results
suggest that piezo-electric pedometers are more accurate than spring-loaded lever
pedometers during slow walking and in overweight populations. However, among
overweight participants the SportBrain also had significantly greater variability in the
percent error at 4 and 5mph (Table 3.4). Crouter et al. studied the effects of BMI, waist
circumference, and pedometer tilt on the accuracy of spring loaded and piezo-electric
pedometers (Crouter, Schneider and Bassett, 2005 ). They found pedometer tilt
influenced accuracy of the spring-loaded pedometer more than waist circumference or
BMI. The piezo-electric pedometer’s accuracy was not affected by waist circumference,
BMI, or pedometer tilt. SportBrain also makes a spring loaded model, the iStep X. We
have evaluated this model and found it to be considerably less accurate than the newer
SportBrain iStep X1 (results not shown).
Because they are inexpensive, pedometers have been recommended for use in
large scale epidemiological research on physical activity (Tudor-Locke, Williams, Reis
and Pluto, 2002 ). In this setting the SportBrain is a major improvement on existing
pedometers because it provides time-resolved data, which can potentially be used to

36
classify individual physical activity into broad categories of time spent in physical
activity of different intensities. For example, in adolescents 130 steps per minute has
been reported to correspond approximately to 70% maximum heart rate (Lubans,
Morgan, Collins, Boreham and Callister, 2009), a threshold for moderate to vigorous
physical activity (MVPA). (The exact step count corresponding to MVPA varied
modestly depending on fitness level). A high correlation of cumulative pedometer-
recorded step counts with energy expenditure has also been reported (Eston, Rowlands
and Ingledew, 1998; Stone, Esliger and Tremblay, 2007 ). This suggests that an
approximate estimate of time-resolved energy expenditure might be possible based on the
time resolved step counts provided by the SportBrain, using prediction equations like
those developed for the interpretation of accelerometer counts (Trost, Way and Okely,
2006 ). However, other studies suggest that the prediction of energy expenditure based on
heart rate, a more physiologically relevant marker of energy expenditure than step counts,
varies markedly between individuals based on fitness and other characteristics (Zuurbier,
et al., 2009 ). Further research is needed to address the utility of time-resolved step
counts in predicting energy expenditure.
A limitation to most pedometers in the research setting is the inability to identify
periods of non-compliance that can produce biased estimates of physical activity.
Because the SportBrain provides time-resolved step count data, periods of non-wear can
potentially be identified based on long periods with zero counts. Such data reduction
algorithms have been developed to identify non-compliance of individuals wearing
accelerometers (National, ). Based on our small sample size, we propose a cut-off to

37
identify periods of non-wear of 30-60 minutes. Choice of a cut-off might depend on the
goal of the study, although there are some limitations to a choice anywhere in this range.
For a study evaluating sedentary behavior, a 30 minute threshold to classify a period of
non-wear might result in underestimating time spent in sedentary behavior, because some
sedentary periods (about 5% based on our results) would be classified as non-wear time.
With a more conservative threshold of 60 minutes (essentially capturing all wear time),
some non-wear time would be classified as sedentary behavior, resulting in an
overestimate of time spent in sedentary activity. If the goal of a study were to assess time
spent physically active, a 60 minute threshold would overestimate sedentary time. Further
study to assess the distribution of duration of zero step counts, for example by examining
the variation by age and gender, might result in better algorithms for distinguishing
periods of non-wear from periods of sedentary activity. However, some of our study
participants also reported removing the pedometer because of sports practice. In this
setting, either a 30 or a 60 minute period of vigorous physical activity might be
misclassified as sedentary based on the consecutive string of zero counts. This issue of
informative missingness is a generic limitation to physical activity monitors that require
participant compliance, and some data imputation methods have been suggested
(Catellier, et al., 2005 ).
Although the SportBrain overcomes several limitations of currently available low-
cost pedometers, other generic limitations of pedometers include the limited ability to
assess non-ambulatory physical activity, such as bicycling, or the additional work
associated with steps climbing stairs. The ability to record step count data in 60 second

38
epochs has advantages over recording cumulative steps taken; however, this epoch is too
long to assess accurately very short periods of physical activity that occur in children
during periods less than 60 seconds (Bailey, et al., 1995 ). In addition, we did not
examine the effect of motion not reflective of physical activity, for example, steps that
might be recorded from the motion of riding in a car.
CHAPTER 3 CONCLUSION
We conclude that the SportBrain iStep X1 pedometer accurately measures step
counts on a treadmill and during self-paced walking, in both normal-weight and
overweight children. Although further investigation is warranted, time resolved step
counts have potential to identify non-wear periods, to classify level of physical activity
and perhaps to estimate time spent in moderate and vigorous physical activity. Because
of its accuracy and low cost, the SportBrain merits consideration for use in large
epidemiological studies aimed at measuring physical activity in children.

39

CHAPTER 3 REFERENCES

Bailey R.C., Olson J., Pepper S.L., Porszasz J., Barstow T.J., and Cooper D.M. The level
and tempo of children's physical activities: an observational study. Med Sci Sports Exerc
1995: 27(7): 1033-1041.

Beets M.W., Patton M.M., and Edwards S. The accuracy of pedometer steps and time
during walking in children. Med Sci Sports Exerc 2005: 37(3): 513-520.

Catellier D.J., Hannan P.J., Murray D.M., Addy C.L., Conway T.L., Yang S., et al.
Imputation of missing data when measuring physical activity by accelerometry. Med Sci
Sports Exerc 2005: 37(11 Suppl): S555-562.

Crouter S.E., Schneider P.L., Karabulut M., and Bassett D.R., Jr. Validity of 10
electronic pedometers for measuring steps, distance, and energy cost. Med Sci Sports
Exerc 2003: 35(8): 1455-1460.

Crouter S.E., Schneider P.L., and Bassett D.R., Jr. Spring-levered versus piezo-electric
pedometer accuracy in overweight and obese adults. Med Sci Sports Exerc 2005: 37(10):
1673-1679.

Eston R.G., Rowlands A.V., and Ingledew D.K. Validity of heart rate, pedometry, and
accelerometry for predicting the energy cost of children's activities. J Appl Physiol 1998:
84(1): 362-371.

Kristensen P.L., Moller N.C., Korsholm L., Wedderkopp N., Andersen L.B., and Froberg
K. Tracking of objectively measured physical activity from childhood to adolescence: the
European youth heart study. Scand J Med Sci Sports 2008: 18(2): 171-178.

Le Masurier G.C., Lee S.M., and Tudor-Locke C. Motion sensor accuracy under
controlled and free-living conditions. Med Sci Sports Exerc 2004: 36(5): 905-910.

Lubans D.R., Morgan P.J., Collins C.E., Boreham C.A., and Callister R. The relationship
between heart rate intensity and pedometer step counts in adolescents. J Sports Sci 2009:
27(6): 591-597.

Melanson E.L., Knoll J.R., Bell M.L., Donahoo W.T., Hill J.O., Nysse L.J., et al.
Commercially available pedometers: considerations for accurate step counting. Prev Med
2004: 39(2): 361-368.

40
Mitre N., Lanningham-Foster L., Foster R., and Levine J.A. Pedometer accuracy for
children: can we recommend them for our obese population? Pediatrics 2009: 123(1):
e127-131.

National. Cancer Institute [Internet]. Risk factor monitoring and methods: SAS programs
for analyzing NHANES 2003-2004 accelerometer data.[Cited 2009 August 1] Available
from: http://riskfactor.cancer.gov/tools/nhanes_pam. . Accessed.

Schneider P.L., Crouter S.E., and Bassett D.R. Pedometer measures of free-living
physical activity: comparison of 13 models. Med Sci Sports Exerc 2004: 36(2): 331-335.

Scruggs P.W., Beveridge S.K., Eisenman P.A., Watson D.L., Shultz B.B., and Ransdell
L.B. Quantifying physical activity via pedometry in elementary physical education. Med
Sci Sports Exerc 2003: 35(6): 1065-1071.

Stone M.R., Esliger D.W., and Tremblay M.S. Comparative validity assessment of five
activity monitors: does being a child matter? Pediatr Exerc Sci 2007: 19(3): 291-309.

Troiano R.P., Berrigan D., Dodd K.W., Masse L.C., Tilert T., and McDowell M. Physical
activity in the United States measured by accelerometer. Med Sci Sports Exerc 2008:
40(1): 181-188.

Trost S.G., Way R., and Okely A.D. Predictive validity of three ActiGraph energy
expenditure equations for children. Med Sci Sports Exerc 2006: 38(2): 380-387.

Tudor-Locke C., Williams J.E., Reis J.P., and Pluto D. Utility of pedometers for
assessing physical activity: convergent validity. Sports Med 2002: 32(12): 795-808.

Tudor-Locke C., Lee S.M., Morgan C.F., Beighle A., and Pangrazi R.P. Children's
pedometer-determined physical activity during the segmented school day. Med Sci Sports
Exerc 2006: 38(10): 1732-1738.

Zuurbier M., Hoek G., van den Hazel P., and Brunekreef B. Minute ventilation of
cyclists, car and bus passengers: an experimental study. Environ Health 2009: 8: 48.

41

CHAPTER 4: EVALUATION OF A PERSONAL GPS LOGGING
INSTRUMENT: LIMITATIONS FOR ASSESSING CHILDREN’S
LOCATIONS

CHAPTER 4 ABSTRACT
Global positioning system (GPS) technology is increasingly used to assess geographically
varying exposure in population studies. However, there has been limited evaluation of
accuracy and completeness of personal GPS data. The ability of a GPS data logger to
assess location of children during usual activity was evaluated. Four days of data from 17
children wearing GPS loggers recording every 15 seconds were evaluated for
completeness by time of day during weekend and weekdays and for accuracy during
nighttime at home. Percentage of possible GPS recorded points and of 5-minute intervals
with at least one recorded location were examined. Mean percentage of total possible 15-
second interval locations recorded daily was less than 30%. Across participants, the GPS
loggers recorded 1% to 47% of total possible location points on weekends and 1% to
55% on weekdays. More complete data were measured during travel to school (average
90%). The percentage of daily 5-minute intervals with recorded data was as high as 53%.
At least one location was recorded during 69% of 5-minute intervals before school
(6:30am-8am), 62% during school (8am-2pm), and 56% after school (2pm-5pm). On
weekends 5-minute interval data were less complete than during weekdays. During
nighttime (12am-6am), on average, location was recorded for less than 25% of 5-minute
intervals and accuracy was poor. The large proportion of missing data, which varied by
location, limits the usefulness of GPS logging instruments for population studies. They
have potential utility for assessing on-road travel time and route.

42

CHAPTER 4 INTRODUCTION

Epidemiological studies often need to define the places in which participants spend time
and the amount of time spent in a specific location. Characterizing time and location is
important for assessing determinants of physical activity, effects of the built environment
and air pollution exposure, among other outcomes (Grow, et al., 2008; Maddison and Ni
Mhurchu, 2009; McKone, Ryan and Ozkaynak, 2009 ; Phillips, Hall, Esmen, Lynch and
Johnson, 2001). Time-location information can be assessed with self-reporting methods
such as diaries and questionnaires (Elgethun, Fenske, Yost and Palcisko, 2003; Freeman,
et al., 1999 ). However, relying on self-report has well-recognized limitations such as
inaccurate recall and limited compliance, and these limitations may be greater in children
(Elgethun, et al., 2007 ; Stone and Shiffman, 2002). With improvements in technology,
the use of global positioning system (GPS) provides an objective method for time-
location assessment. Portable GPS devices with memory to store substantial amounts of
data over time are now available at relatively low-cost, suitable for large population
studies. Compared to self-report, GPS devices have reduced participant burden and the
potential for improved accuracy of data.
GPS has been shown to be useful for research in fields ranging from navigation to
farming (Maddison and Ni Mhurchu, 2009 ). Increasingly, GPS has been used in studies
of transportation, physical activity, and time-location tracking to assess various
environmental exposures such as air pollution (Cooper, et al., 2010; Duncan, Badland and
Mummery, 2009 ; Maddison, et al., 2010; Nuckols, Ward and Jarup, 2004). In physical
activity research, GPS has potential for determining the spatial context of activity, factors
such as distance traveled and speed, and there is potential for GPS data to augment

43
information from accelerometer measured physical activity (Cooper, et al., 2010 ;
Maddison and Ni Mhurchu, 2009; Quigg, et al., 2010). Personal GPS logging instruments
have also been used in the assessment of air pollution exposure (Adams, Riggs and
Volckens, 2009; Gerharz, Kruger and Klemm, 2009 ). However, utility of GPS requires
accuracy and complete data, or a sample of location data that is randomly sampled across
different activities commonly conducted on weekends, weekdays and different times of
day.
We evaluated the ability of a commercially available GPS logger to accurately
assign time-location to a sample of children during usual activity over 4 days.

CHAPTER 4 METHODS
Study population: Participants were recruited from two 4
th
grade classrooms at a school
in Palos Verdes, CA, to wear GPS loggers for 5 days in order to measure weekend and
weekday location. Forty students were eligible to participate, 18 (45%) agreed and
returned signed parental informed consent. However, one participant joined on the final
day after observing classmates participating (and this child was not included in the
analysis). Of the remaining 17 contributing to analyses, three joined on the

third day of
the study.
Sampling period and protocol: The GPS loggers were set to record location every 15
seconds. The devices were turned on before giving them to the study participants and the
participants were asked not to turn the loggers off or on. They were instructed to charge
the instrument every night. Participants received a reminder flyer to put on their

44
refrigerators and parents were called twice during the study period to remind them to
encourage the child to charge the instrument.
Participants were asked to wear the GPS devices everyday during waking hours
except while bathing or swimming during the study period. Devices were distributed on
Friday, May 2, 2008 and returned the following Wednesday morning. Data were
analyzed starting at 8pm Friday.
Specifications of instrument: We used Wintec WBT-201 devices in our study (Figure
4.1a), a relatively low-cost commercially available GPS logger. (The price was $75 per
unit). The instrument has a convenient size of 6.0 x 3.8 x 1.6 cm and weighs 48 grams.
Pre-testing demonstrated comparable reception and accuracy when carried in a small
pouch on the belt, in a pocket, purse or backpack. A piece of Styrofoam was taped over
the device’s power button to prevent participants from turning it off, and the covered
device was placed in a small pouch and attached to a belt (Figure 4.1b). The
manufacturer reports that the horizontal position accuracy is 1-5 meters when it can
perform a wide area augmentation system correction and otherwise is approximately 10
meters. The device stores up to 131,072 waypoints (recorded locations), allowing
approximately 22 days of data archiving with 15-second recording epochs. The data
output includes latitude, longitude, date, time, speed, and elevation. A limitation to the
instrument was a battery life of 15 hours, so daily recharging was necessary to obtain
multi-day information.

Figure 4.1a. The Wintec WBT 201, a wearable GPS data logger
Figure 4.1b: Wintec WBT 201 covered with a piece of Styrofoam

Data Processing: GPS data were downloaded using Time machine X Programming
software provided by the Wintec manufacturer. The output format of
TK files, and these files can be converted to KMZ (for viewing in Google Earth
4.2) and GPX (GPS eXchange
software using SAFE’s FME
attribute tables in ArcGIS we identified errors b
The Wintec WBT 201, a wearable GPS data logger

Wintec WBT 201 covered with a piece of Styrofoam

GPS data were downloaded using Time machine X Programming
software provided by the Wintec manufacturer. The output format of the
TK files, and these files can be converted to KMZ (for viewing in Google Earth
) and GPX (GPS eXchange file) formats. The data were exported to Esri’s
’s FME software to convert the GPX data to shapefiles. Viewing the
attribute tables in ArcGIS we identified errors based on location (i.e. x,y coordinates
45
GPS data were downloaded using Time machine X Programming
the GPS logger is
TK files, and these files can be converted to KMZ (for viewing in Google Earth Figure
Esri’s ArcGIS
GPX data to shapefiles. Viewing the
ased on location (i.e. x,y coordinates

46
placing the child and accompanying GPS unit in the Pacific Ocean) as well as improbable
changes in elevation or speed. The elevation errors were identified from individual
waypoint elevations (e.g. a child and GPS unit at 10,735 m above sea level) or rapid
changes in elevation (> 300 m) between consecutive waypoints spanning short distances
(< 500 m) or unrealistically rapid movement between consecutive waypoints (e.g. 300
km/hr). These erroneous waypoints were then deleted from the dataset.

Figure 4.2. Representation of GPS raw data, showing example of data with travel
between locations on land and well into the ocean between 15 second recordings

For each subject we calculated the number of 15-second interval waypoints
recorded each day and the total number of waypoints recorded during the study period.
We also calculated the proportion of all 5-minute intervals with at least one GPS
location reading. Time and location point data were assessed by day (over each 24 hour

47
period from midnight to midnight), and by periods of interest during each day (overnight
8pm-6:30am, travel to school 6:30am-8am, at school 8am-2pm, after school 2pm-5pm,
and evening 5pm-8pm). In addition, we examined overnight data between 12am and
6am. Review of these nighttime data, for which we assumed the instrument would have
been at each participant’s home in a single location, demonstrated high variability in
recorded location, so locations were determined by averaging location data over a five
minute rolling average (eg. 1:00am to 1:05am, 1:01am to 1:06am, etc.) to determine
whether more accurate location data could be obtained. Finally, we required a 5-minute
period-specific 50% data completeness threshold for inclusion in this analysis. For each
five minute rolling average there should have been 20 waypoints logged, so if less than
10 logged points were recorded, that interval was not assigned a location. The rationale
for this analysis was to evaluate whether the locational precision and accuracy was
improved by excluding intervals with less complete data.
Geocoding addresses: Participants’ home addresses and the address of the school were
geocoded using the USC Geocoder (Goldberg, 2011 ). This research platform
incorporates the standard components found within typical geocoding system
architectures, including the representation and storage of reference data layers, a feature
matching algorithm based on a deterministic candidate scoring scheme, and a set of
feature interpolation algorithms (Boscoe, 2008 ). GPS data were overlaid with census
parcel data to view locations recorded within home and school parcels.

48
CHAPTER 4 DATA ANALYSIS
To summarize data, percent possible GPS waypoint recordings for each
participant was calculated by dividing the number of 15-second waypoint readings
recorded by the total number possible and multiplying by 100 ((waypoints
recorded/possible recordings) x 100)). Each 24 hour period had a total possible of 5760
(1440 minutes x 4 records per minute). Percent possible waypoints were calculated for
each study participant by day, by time of day, and for the total study period. We
performed similar calculations for each study period (e.g. 6:30am to 8am-specific data)
and calculated the percent possible time monitored ((time monitored in minutes/possible
time in minutes) x100)).

CHAPTER 4 RESULTS
Fourteen of the 17 participants included in the analysis had GPS units for all study
days. However, of the 14 one lost the instrument and one reported a malfunctioning
device, and these children were provided new units on day 3 (Monday morning). Three
additional children enrolled on day 3.Therefore, there were twelve participants with data
logged for the entire study period and five participants with data for two days beginning
Monday.
There was substantial missing data (Figure 4.3). Of the 5760 possible daily
recordings, the mean percent waypoints per day across study participants was less than
30% for each study day. Across subjects, on weekends the GPS loggers recorded from
1% to 47% of the total. A similar range was recorded on weekdays, 1% to 55%. Sunday
had the lowest percent waypoints recorded, with a mean of 16% compared to 26-28% for
the other days.

49

Figure 4.3: Percent possible waypoints recorded by day*

The proportion of the day for which at least one location was recorded in each 5-
minute interval was higher (Figure 4.4). Saturday and Sunday’s mean percents of 40%
and 31%, respectively, were lower than Monday and Tuesday’s means of 53% and 41%.
There was considerable variability between participants on any given day, from 0% (a
participant with a day of missing data) to 93%.

50
Figure 4.4: Percent possible five minute intervals with at least one location recorded, by
day*

*24 hour day: 12am to 12am; +: mean

Completeness of at least 1 recording in each 5-minute interval varied markedly by
time of day. On weekdays (Fig 4.5a) the data were most complete between 6:30am and
8am (69%), during school hours from 8am-2pm (62%) and during the afternoon from
2pm-5pm (57%), and were least complete from 8pm-6:30am (28%). On the weekend
(Figure 4.5b), data were most complete for 5pm-8pm (58%) and least complete from
6:30am-8am (20%), although the period from 8pm-6:30am was almost as incomplete
(21%). There was substantial between-participant variability in completeness of data
regardless of the time of day (generally ranging from 0 to 100%).

51
Figure 4.5a: Percent possible five minute intervals recorded with at least one location,
weekdays*
*8pm Sunday to 8am Wednesday; + : mean

Figure 4.5b: Percent possible five-minute intervals with at least one location recorded,
weekend*

*Friday 8pm to Sunday 8pm; + : mean

52
Data were least complete between 12am-6 am both for all possible 15-second
waypoints (Figure 4.6a), on average less than 12% (range 0-62%), and for at least one
recording in each 5-minute interval (Figure 4.6b), on average less than 25% (range 0-
74%). Some participants had a large number of waypoints outside the home parcel and
poor precision and accuracy to the location during the 12am-6am period when the GPS
loggers should have been stationary and charging (see for example Figure 4.7a showing
scatter outside of the residential parcel and into the next block). The accuracy and
precision for this instrument’s location was not improved by using the 5-minute rolling
average location (Figure 4.7b at the same residence). Using the 5-minute rolling average
with a 50% data completion threshold resulted in no waypoints located in the parcel
(Figure 4.7c). There was, however, considerable variability between children, with
waypoints for some children located primarily within the home parcel during the night
(Figure 4.8a), and this precision was generally improved by taking the 5-minute rolling
average (Figure 4.8b) and further improved using the 5-minute rolling average with 50%
data completion threshold (Figure 4.8c). However, there were no children for whom
accuracy and precision were sufficient to locate the child within the home (not just within
the parcel).

53
Figure 4.6a: Percent possible waypoints recorded 12am-6am

+ : mean
Figure 4.6b: Percent possible five minute time intervals recorded with at least one
location, 12am to 6am*

+ : mean

54
Figure 4.7a: Nighttime (12am-6am) 15-second location recordings with a large amount
of scatter outside a residential parcel (shaded)

55
Figure 4.7b: Location recording when restricted to 5-minute rolling average during
overnight hours

56
Figure 4.7c: 5-minute rolling average with 50% data threshold

57

Figure 4.8a: Nighttime (12am-6am) 15-second interval locations with few
points outside the home parcel (shaded)

58
Figure 4.8b: Improved location recording when restricted to the 5-minute rolling average
during overnight hours

59
Figure 4.8c: Precision and accuracy was further improved when restricted to 5-minute
rolling average with 50% data threshold

During the school day, recorded locations showed considerable scatter, but these
were almost exclusively within the school boundaries, which were relatively large. (See,
for example, Figure 4.9).

60
Figure 4.9: A study participant’s locations recorded while at school

61
The 6:30am-8am weekday period with the most complete data collection (Figure
645a) included the morning commute to school for each child. In order to understand
whether time spent in transit was more complete than other activities, we examined each
participant’s data individually for the trip from home to school. We calculated GPS
recorded location completeness during travel to school by dividing the number of 15-
second interval locations recorded during travel by the total number of intervals spent in
transit. On average across participants, the GPS recorded location for 91% of the transit
time (SD=16, range 31-100%). On-road travel route was readily identified even if there
were some waypoints that were not recorded (Figure 4.10a and 4.10b).
Figure 4.10a: Typical freeway travel record

62
Figure 4.10b: Travel path approaching school (in blue) on small roadways
for one participant*

We investigated whether the variability in missing data was consistently
accounted for by specific individuals. However, across complete study days (N=4 for this
analysis), there was considerable variability in the percent of time recorded for each
participant (Figure 4.11). For example, participant 1 had a range of 23% to 80% percent
time recorded across the four study days. Participants 2 and 3 had less variability across
the study days (20% to 39% and 65 to 78%, respectively).

63
Figure 4.11: Percent of possible five-minute time intervals with at least one location
recorded by subject, across days*
*Restricted to participants with 4 days of data collection: Saturday, Sunday, Monday, and
Tuesday; + is mean

CHAPTER 4 DISCUSSION
In this evaluation of a commercially available low-cost GPS logger, we observed a large
proportion of time with missing data among children engaged in their usual pattern of
activities. A notable exception was time spent in on-road travel, for which data were, on
average, greater than 90% complete. This made it possible to identify an individual route
travelled by each participant (Figure 4.10a and 4.10b), even where data were missing for
short periods. For other locations (at school and after school, for example), it was
possible to identify a larger proportion of time during which a child’s location could be
assigned by requiring only one data point in each 5-minute interval. However, for school
time relatively imprecise GPS data provide little additional information beyond that
provided by school attendance records. In addition, there was large variability in record
completeness both between-child (within day) and within-child (across day). Nighttime,

64
when children were at home with the instrument charging, had the worst data
completeness. Precision and accuracy of measurement was poor during this period, could
be improved in some cases by using a rolling 5-minute average of location, but was still
not sufficiently accurate to distinguish indoor from outdoor location within the home
parcel.
Valid location information about time spent in different locations and activities
could be obtained from GPS loggers if missing data were randomly distributed over the
day, but could bias inference about where children spend time if data were not missing at
random (Donders, et al., 2006 ). For example, accelerometers for assessing physical
activity have been used with GPS loggers to evaluate the location and context of physical
activity, requiring accelerometry recorded physical activity data to be matched to GPS
location, data that will be incomplete (based on our findings). If physical activity occurs
indoors at a gymnasium where there is poor signal capture, the physical activity data will
not have matched GPS data and exclusion of these data could result in an underestimation
of the role of indoor gymnasium time in physical activity. The reported proportion of
missing matched accelerometer and GPS data in selected physical activity studies has
been variable, from 28% to 76% (Cooper, et al., 2010 ; Maddison, Jiang, Vander Hoorn,
Exeter, Mhurchu and Dorey, 2010). Another scenario in which the use of GPS loggers
could result in biased results is in the assessment of exposure to air pollution, which is
known to vary markedly by proximity to traffic corridors and time of day (Behrentz, et
al., 2005) (Hu, et al., 2009; Ozkaynak, et al., 2007; Zhou and Levy, 2007 ; Zhu, Hinds,
Kim and Sioutas, 2002). Our results indicate that any exposure assignment made based
simply on GPS recorded location would over-estimate the contribution of heavy on-road

65
exposure (with little missing information) to the total time sampled and under-estimate
the contribution of exposure at home at night. As GPS technology evolves and becomes
more commonly used in larger scale studies of time, location, and activity, it will be
important to assess the potential bias associated with missing data. One useful step would
be for investigators to report information on sampling frequency (e.g. every 15 seconds in
our study), averaging time and other restrictions used to impute location more precisely;
the proportion of missing or inaccurate data by microenvironment, (e.g. signal lost at
home or work), time of day. This would facilitate comparison of results with other studies
and the possibility that informatively missing data might explain different results.
There are several possible causes for missing data. Equipment malfunction or
dead batteries could result in no location data being recorded. Instrument problems in one
study using GPS to track location resulted in location being recorded during only 30% of
the total monitoring time (Phillips, Hall, Esmen, Lynch and Johnson, 2001 ). Although
we provided participants with reminders to charge the GPS loggers nightly, we do not
know the level of compliance and if some missing data might be a result of loss of battery
life. We do know that if there was any logged data, even if sparse, that the logger was
powered on. But for the participants who had no recordings during a day or an entire time
interval, we do not know if the device was powered and not capturing or not functioning.
It is likely that problems with signal capture were an important cause of missing data or
erroneous assessed locations in our study rather than an uncharged battery, because using
any data in each 5-minute interval as a criterion for non-missingness fairly markedly
improved the percentage of time for which data were available. (This would not have
occurred if continuous periods with an uncharged battery were accounting for these

66
missing data periods). We also observed a high percentage of missing data during the
overnight hours when we assumed the participants were at home. If the devices were
plugged in for charging, loss of battery life should not have been the cause of nighttime
missing data. GPS signals can be blocked by materials such as concrete and steel which
reduce or eliminate reception within buildings. In addition, poor satellite geometry can
affect GPS accuracy. Interference in the satellite signals from multi-path errors that arise
from the reflection of satellite signals from other surfaces, including buildings, cars, trees,
the ground or water may have contributed to missing data (Rainham, et al., 2008 ). When
the available satellites are in one part of the sky or nearly lined up (relative to GPS
antenna), the precision of the GPS is reduced. Embankments or buildings which partially
block the line-of-sight of the GPS unit can reduce accuracy of the location measurement.
Therefore, time spent outside is likely to have more complete information than time spent
indoors. When the GPS has open sky and receives signals from satellites that are
dispersed in elevation and angular distance, the positional coordinates of the GPS have
the greatest accuracy (Phillips, Hall, Esmen, Lynch and Johnson, 2001 ). GPS may
perform well during on-road travel because there is a view of the open sky and less signal
interference. Even if an obstruction results in loss of signal, as the vehicle moves it will
be likely to pass through locations with open sky where a waypoint will be recorded
accurately. Tall buildings in central cities may also interfere with signals, and this has
been called the “urban canyon” effect (Morabia, et al., 2009 ).
Other investigators have interpolated prolonged periods with missing signal to
indoor locations based on the location of the last previously recorded location (Cooper,
Page, Wheeler, Hillsdon, Griew and Jago, 2010; Elgethun, Fenske, Yost and Palcisko,

67
2003). In a refinement of this approach, time with no signal was classified as time
indoors at home if lost signal occurred at home, unless the subsequent first recorded
location was greater than 1km from the residence (suggesting that signal was lost for
some other reason than having gone indoors) (Elgethun, Yost, Fitzpatrick, Nyerges and
Fenske, 2007 ). However, in our data it was common for the next location recorded after
a prolonged loss of signal to be many kilometers from the previously recorded location,
perhaps because when GPS devices lose their signal, it takes time to re-initialize (in order
to find the orbit and clock data for the relevant satellites). This loss of signal can be brief
(for example, when driving through a tunnel) but after a prolonged loss of signal can take
up to 13 minutes to reinitialize even with a clear satellite view, according to the Wintec
WBT-201 user’s manual. We conclude that interpolating missing GPS data requires
judgment and assumptions that may result in location misclassification and that to
automate this procedure in a way that could be applied in large population studies would
entail considerable uncertainty to location assessment.
Although there were limitations to the instrument we evaluated, ongoing
technological advances in GPS instrumentation might increase the potential for inferring
useful time-location information in large population studies. Increased battery life to
allow continuous data recording for several days without recharging would reduce data
loss. Battery life varies by instrument, but until recently has been relatively short (10-48
hours) (Wu, et al., 2010 ). Indicators in the data file to distinguish lack of power from
loss of signal would improve the usefulness of the data. Newer commercially available
GPS data loggers report the number of satellites in view and contributing to each
waypoint recorded (Wu, Jiang, Liu, Houston, Jaimes and McConnell, 2010), which may

68
make it possible to weight the precision of different measurements in order to estimate
more accurately the true location of a child. There has been increased interest in using
GPS installed within cell phones to evaluate location, but location assessed with phones
has been shown to be considerably less accurate than stand-alone GPS devices
(Zandbergen, 2009; Zandbergen and Barbeau, 2011 ).
Even with the current limitations to data collection, GPS data might complement
diary data with other known limitations for large population studies of children, for
example to identify locations in specific time intervals of interest, such as during travel or
after school. GPS has been used in this way, for example to examine the walk to school
and to evaluate time outdoors after school (Cooper, et al., 2010 ). Although overall
participation rate by children eligible for this study was only 45%, the late enrollment by
some children after seeing their classmates’ enthusiasm with the study suggests that
participation rate could be improved by including a run-in period prior to the final
enrollment and start of data collection.
Position accuracy and instrument precision under static and dynamic conditions in
a variety of environments is critical for time-location analysis. We conclude that personal
GPS data logging instruments have promise for identifying time spent in transit. For other
locations, these data may best be used as a complement to other data sources.
Epidemiological studies of effects of environmental exposures using these devices to
identify exposure based on location require consideration of the potential of
informatively missing data to bias the interpretation of results.

69
CHAPTER 4 REFERENCES

Adams C., Riggs P., and Volckens J. Development of a method for personal,
spatiotemporal exposure assessment. J Environ Monit 2009: 11(7): 1331-1339.

Behrentz E., Sabin L.D., Winer A.M., Fitz D.R., Pankratz D.V., Colome S.D., et al.
Relative importance of school bus-related microenvironments to children's pollutant
exposure. J Air Waste Manag Assoc 2005: 55(10): 1418-1430.

Boscoe F. The science and art of geocoding. In: Rushton G., Armstrong, M.P., Gittler, J.,
Greene, B.R., Pavlik, C.E., West, M.M., and Zimmerman, D.L. (ed). Geocoding Health
Data: The Use of Geographic Codes in Cancer Prevention and Control, Research and
Practice. CRC Press: Boca Raton, 2008, pp 95-109.

Cooper A.R., Page A.S., Wheeler B.W., Griew P., Davis L., Hillsdon M., et al. Mapping
the walk to school using accelerometry combined with a global positioning system. Am J
Prev Med 2010: 38(2): 178-183.

Cooper A.R., Page A.S., Wheeler B.W., Hillsdon M., Griew P., and Jago R. Patterns of
GPS measured time outdoors after school and objective physical activity in English
children: the PEACH project. Int J Behav Nutr Phys Act 2010: 7: 31.

Donders A.R., van der Heijden G.J., Stijnen T., and Moons K.G. Review: a gentle
introduction to imputation of missing values. J Clin Epidemiol 2006: 59(10): 1087-1091.

Duncan M.J., Badland H.M., and Mummery W.K. Applying GPS to enhance
understanding of transport-related physical activity. J Sci Med Sport 2009: 12(5): 549-
556.

Elgethun K., Fenske R.A., Yost M.G., and Palcisko G.J. Time-location analysis for
exposure assessment studies of children using a novel global positioning system
instrument. Environ Health Perspect 2003: 111(1): 115-122.

Elgethun K., Yost M.G., Fitzpatrick C.T., Nyerges T.L., and Fenske R.A. Comparison of
global positioning system (GPS) tracking and parent-report diaries to characterize
children's time-location patterns. J Expo Sci Environ Epidemiol 2007: 17(2): 196-206.

Freeman N.C.G., Lioy P.J., Pellizzari E.D.O., Zelon H., Thomas K., Clayton A., et al.
Responses to the Region 5 NHEXAS time/activity diary. 1999: 9(5): 414-426.

Gerharz L.E., Kruger A., and Klemm O. Applying indoor and outdoor modeling
techniques to estimate individual exposure to PM2.5 from personal GPS profiles and
diaries: a pilot study. Sci Total Environ 2009: 407(18): 5184-5193.

70
Goldberg D. The USC WebGIS Geocoding Platform. GIS Research Laboratory Research
Report 2011: 11.

Grow H.M., Saelens B.E., Kerr J., Durant N.H., Norman G.J., and Sallis J.F. Where are
youth active? Roles of proximity, active transport, and built environment. Med Sci Sports
Exerc 2008: 40(12): 2071-2079.

Hu S., Fruin S., Kozawa K., Mara S., Paulson S.E., and Winer A.M. A wide area of air
pollutant impact downwind of a freeway during pre-sunrise hours. Atmospheric
Environment 2009: 43(16): 2541-2549.

Maddison R., and Ni Mhurchu C. Global positioning system: a new opportunity in
physical activity measurement. Int J Behav Nutr Phys Act 2009: 6: 73.

Maddison R., Jiang Y., Vander Hoorn S., Exeter D., Mhurchu C.N., and Dorey E.
Describing patterns of physical activity in adolescents using global positioning systems
and accelerometry. Pediatr Exerc Sci 2010: 22(3): 392-407.

McKone T.E., Ryan P.B., and Ozkaynak H. Exposure information in environmental
health research: current opportunities and future directions for particulate matter, ozone,
and toxic air pollutants. J Expo Sci Environ Epidemiol 2009: 19(1): 30-44.

Morabia A., Amstislavski P.N., Mirer F.E., Amstislavski T.M., Eisl H., Wolff M.S., et al.
Air pollution and activity during transportation by car, subway, and walking. Am J Prev
Med 2009: 37(1): 72-77.

Nuckols J.R., Ward M.H., and Jarup L. Using geographic information systems for
exposure assessment in environmental epidemiology studies. Environ Health Perspect
2004: 112(9): 1007-1015.

Ozkaynak H., Palma T., Touma J.S., and Thurman J. Modeling population exposures to
outdoor sources of hazardous air pollutants. J Expos Sci Environ Epidemiol 2007: 18(1):
45-58.

Phillips M.L., Hall T.A., Esmen N.A., Lynch R., and Johnson D.L. Use of global
positioning system technology to track subject's location during environmental exposure
sampling. J Expo Anal Environ Epidemiol 2001: 11(3): 207-215.

Quigg R., Gray A., Reeder A.I., Holt A., and Waters D.L. Using accelerometers and GPS
units to identify the proportion of daily physical activity located in parks with
playgrounds in New Zealand children. Prev Med 2010: 50(5-6): 235-240.

Rainham D., Krewski D., McDowell I., Sawada M., and Liekens B. Development of a
wearable global positioning system for place and health research. Int J Health Geogr
2008: 7: 59.

71
Stone A.A., and Shiffman S. Capturing momentary, self-report data: a proposal for
reporting guidelines. Ann Behav Med 2002: 24(3): 236-243.

Wu J., Jiang C., Liu Z., Houston D., Jaimes G., and McConnell R. Performances of
different global positioning system devices for time-location tracking in air pollution
epidemiological studies. Environ Health Insights 2010: 4: 93-108.

Zandbergen P.A. Accuracy of iPhone Locations: A Comparison of Assisted GPS, WiFi
and Cellular Positioning. Transactions in GIS 2009: 13: 5-25.

Zandbergen P.A., and Barbeau S.J. Positional Accuracy of Assisted GPS Data from
High-Sensitivity GPS-enabled Mobile Phones. The Journal of Navigation 2011: 64(03):
381-399.

Zhou Y., and Levy J.I. Factors influencing the spatial extent of mobile source air
pollution impacts: a meta-analysis. BMC Public Health 2007: 7: 89.

Zhu Y., Hinds W.C., Kim S., and Sioutas C. Concentration and size distribution of
ultrafine particles near a major highway. J Air Waste Manag Assoc 2002: 52(9): 1032-
1042.

72

CHAPTER 5: URINARY METABOLITES OF POLYCYCLIC
AROMATIC HYDROCARBON EXPOSURE AND BIOMARKERS
OF EARLY EFFECT AFTER TRAVEL ON A BUSY ROAD

CHAPTER 5 ABSTRACT

The objective of the study was to assess feasibility of collecting urine samples at school
and to evaluate urinary biomarkers of exposure and early effect of traffic-related air
pollution. Up to five samples per child were collected before and after a school field
trip with bus travel on a busy highway from fifteen 9-10 year olds recruited from two
classrooms in a low pollution region of Los Angeles. Samples were analyzed for
biomarkers of exposure (ten polycyclic aromatic hydrocarbon metabolites and a nitro-
PAH,1-aminopyrene) and of effect (Clara cell protein 16 and 8-iso-PGF
2α
, a major F
2
-
isoprostane). 60% of participants collected all five samples. There was no statistically
significant difference between the pre- and post-trip measurements of any biomarker.
Collecting urine samples from children at school was feasible, but traffic effects were
not detected after exposure in this small population sample.

73

CHAPTER 5 INTRODUCTION

Regional ambient air pollution has well studied adverse effects on cardiovascular and
respiratory health (Chen and Schwartz, 2009; Gauderman, et al., 2005; Gauderman, et al.,
2007; Gotschi, et al., 2008; Hole, Watt, Davey-Smith, Hart, Gillis and Hawthorne, 1996;
Islam, et al., 2007; Kim, et al., 2008; Knuiman, James, Divitini, Ryan, Bartholomew and
Musk, 1999 ; Li, et al., 2003; Salam, Islam and Gilliland, 2008; Sin, Wu and Man, 2005).
Emerging evidence suggests that local near-roadway traffic-related pollution (TRP) has
adverse effects that are independent of those of regional pollutants (English, Neutra,
Scalf, Sullivan, Waller and Zhu, 1999; Holguin, et al., 2007; McConnell, et al., 2006;
Salam, Islam and Gilliland, 2008 ). However, most large scale epidemiological studies of
air pollution have assessed exposure based on central site monitoring measurements,
residential proximity to roads, or modeled residential exposure to nearby fresh traffic
emissions (Brauer, et al., 2002; Brunekreef, Janssen, de Hartog, Harssema, Knape and
van Vliet, 1997; Gauderman, et al., 2007; Gotschi, et al., 2008; Jerrett, et al., 2005 ;
Kunzli, et al., 2009; Nordling, et al., 2008). A limitation of these exposure assessment
methods is that they do not account for exposures at other locations that may make
substantial contributions to overall exposure (for example, indoors and outdoors, at
school or work, or in transit). Ambient concentration measurements likely underestimate
actual exposures to traffic emissions, and this underestimation may account for
uncertainty and inconsistency in the observed traffic-related health effects (Braback and
Forsberg, 2009 ; Jerrett, et al., 2005; Salam, Islam and Gilliland, 2008). Better exposure
assessment will result in better estimates of the associated health effects.

74
On-road microenvironmental exposure may be especially important to assess
because commuting is a key source of exposure to some components of the TRP
mixture. For example, in Los Angeles 33-45% of daily ultrafine particulate exposure can
occur from 90 minutes of on-road commuting exposure, an average amount of on-road
time in California (Fruin, Westerdahl, Sax, Sioutas and Fine, 2008 ; Klepeis, et al.,
2001). Ultrafine particles (less than 0.1 µm in aerodynamic diameter) which have
adverse biological effects have been found to be present in very high concentrations in
the vicinity of major roadways (Zhu, Hinds, Kim and Sioutas, 2002) and exceedingly
high ultrafine exposures have been found in vehicles and school buses (Behrentz, et al.,
2005; Sabin, et al., 2005 ; Westerdahl, Fruin, Sax, Fine and Sioutas, 2005). However,
on-road exposure is challenging to evaluate due to the chemical complexity of the TRP
mixture, widely-varying traffic emission rates, wind and other meteorological factors
affecting pollutant levels, and effects of timing of exposure measurements (Behrentz, et
al., 2005; Laumbach and Kipen, 2010 ; Sabin, et al., 2005; Westerdahl, Fruin, Sax, Fine
and Sioutas, 2005; Zhou and Levy, 2007). Health effects of these on-road exposures are
not well-studied. New methods that are feasible for large population studies are needed
to assess exposure to TRP and to assess early biological effects of these exposures.
Urinary biomarkers have potential for assessing exposure and biological effects
of TRP for large population epidemiologic studies. Urinary polycyclic aromatic
hydrocarbons (PAH) metabolites have been found to correlate with exposure among
workers occupationally exposed (Jongeneelen, et al., 1990 ; Merlo, et al., 1998). Vehicle
emissions are a major source of PAH which are found in ambient air both in vapor- and

75
particle-phase, depending largely on the molecular weight of the PAH (Larsen and
Baker, 2003; Marr, et al., 1999 ; Panther, Hooper and Tapper, 1999; Sorensen, et al.,
2003). Associations have been observed between non-occupational TRP exposure and
urinary PAHs in some studies, although the findings are inconsistent (Bae, et al., 2010 ;
Kang, Cho, Kim and Lee, 2002; Tuntawiroon, et al., 2006). Polycyclic mononitroarenes
(nitro-PAHs) are nitro-substituted derivatives of PAHs that are specific in ambient air to
combustion of diesel exhaust (Zwirner-Baier and Neumann, 1999) 1-nitropyrene is
metabolized to 1-aminopyrene and this metabolite can be measured in urine (Laumbach,
et al., 2009) 8-iso-prostaglandin-F
2α
is a major F
2
-isoprostane that can be measured in
urine and is a marker of systemic lipid oxidation and oxidative stress (Basu, 2008;
Morrow, 2005 ). Oxidative stress has been postulated to be an important mechanism for
pulmonary effects of inhaled ambient particulate matter including asthma exacerbation,
pulmonary inflammation, and reduced lung function (Gilliland, McConnell, Peters and
Gong, 1999; Li, et al., 2003; Nordling, et al., 2008 ). There is a developing literature
examining the association of F
2
-isoprostanes with exposure to tobacco smoke, and the
results suggest that urinary F
2
-isoprostanes might be a useful intermediate marker of the
effect of air pollutants in epidemiological studies (Bachi, Zuccato, Baraldi, Fanelli and
Chiabrando, 1996; Chiabrando, et al., 1999; Helmersson, et al., 2005; Morrow, et al.,
1995 ; Obata, Tomaru, Nagakura, Izumi and Kawamoto, 2000). Urinary F2-isoprostanes
were also increased after an experimental exposure to wood smoke (Barregard, et al.,
2006 ). Urinary Clara cell secretory protein 16 (CC16) is a promising peripheral marker
of damage to the lung (Broeckaert and Bernard, 2000) CC16 can be measured in urine

76
and was observed to be increased after experimental exposure to wood smoke
(Barregard, et al., 2008) and in association with daily variation in ambient fine
particulate exposure (Timonen, et al., 2004 ). The time course of these urinary
biomarkers of exposure and early effect is short, on the order of hours, so they are
appropriate for assessment of recent exposures (Basu, 2008; Broeckaert, et al., 2000 ; Li,
et al., 2008).
The aims of our study were to evaluate the feasibility of collecting urine from
children in a school setting and to examine whether these biomarkers of exposure and
acute pre-clinical systemic and pulmonary effects are associated with on-road traffic
pollution. We conducted a study on a sample of 4
th
grade children before, during, and
after a school field trip with travel on a major freeway traffic corridor in Los Angeles.
Urine samples were collected before and after the field trip in order to evaluate exposure
and biological effects. We assessed participation and compliance. Urine samples were
analyzed to determine whether PAH and nitro-PAH metabolites, CC16 and F
2
-
isoprostane increased after exposure to high levels of on-road traffic pollutants during
the school field trip and to evaluate whether an increase in urinary PAH metabolites was
associated with increased CC16 and F
2
-isoprostane.

CHAPTER 5 METHODS

Children were recruited from a school in Palos Verdes, CA, a location with low
levels of regional pollution. The school is almost 3 km from the nearest highway
(coastal Highway 1) and 1 km from the nearest major road. TRP exposures occurred
during a field trip to downtown Los Angeles along the 110 freeway, a major traffic

77
corridor with 300,000 vehicles per day, including heavy truck traffic from the largest
port complex in the United States (CalTrans 2008 ). A chartered Coach America Van
Hool diesel bus transported children during the field trip travel. Travel occurred during
morning rush hour, leaving the school at 8:41am and arriving at a museum in the
downtown area at 9:29am, traveling a distance of 42 km. The return trip to school took
place between 1:41pm and 2:18pm, for a total travel time of 1 hour and 25 minutes,
Study Participants: Participants were recruited from two 4
th
grade classrooms. Forty
students were eligible to participate, fifteen (38%) agreed and returned signed parental
informed consent. All fifteen (ten girls, five boys, 100%) collected at least one urine
sample and nine of the fifteen ( 6 girls, 3 boys,60%) collected all five samples. The
mean age of participants was 9.9 years (SD=0.35)
Sample Collection: Subjects collected up to five urine samples before and after the field
trip, which occurred on a Tuesday (Figure 5.1). Two samples were collected prior to the
field trip (Monday evening and Tuesday morning) and three samples were collected
after the field trip (Tuesday afternoon immediately after the trip, Tuesday evening, and
Wednesday morning). Participants were instructed to collect the first morning void urine
and the last void before bed. Except for the Tuesday afternoon sample collected at
school upon returning from the field trip, all samples were collected at the child’s home
by the child. Children brought the samples to school in coolers with ice packs provided
daily to them, and the coolers were delivered to study staff. Samples were aliquoted into
vials and then samples were stored at -80
O
C .

78
Figure 5.1: Timing of data collection
Friday Saturday Sunday Monday Tuesday Wednesday
Urine Collection

Evening
Morning
Post-trip
Evening
Morning
Diet diary
Traffic
questionnaire

Traffic-Related Pollutant Assessment: Integrated air pollution measurements were made
on the bus and continuous, real-time measurements of on-road TRP concentrations were
made in a lead vehicle driven in front of the bus. The lead vehicle was an electric-
powered, zero-emission Toyota RAV4 sub-SUV carrying sampling instruments and
equipment, as described previously (Kozawa, Fruin and Winer, 2009; Westerdahl,
Fruin, Sax, Fine and Sioutas, 2005). Black carbon concentrations were measured by
light absorption with a Magee Scientific Aethalometer; NO and NO2 were measured by
chemiluminescence with a Teledyne API model 200E; and CO2 was measured by non-
dispersive infrared gas analyzer with a Licor LI-820. Particle-bound PAHs were
measured with an EcoChem PAS2000 detecting UV-induced photo-emission of particle-
bound PAHs three rings and larger.
On the bus, integrated ambient volatile PAH samples during the field trip were
collected using XAD-4 resin tubes (SKC, model 226-170). Volatile organic compounds
(VOCs) were collected with glass tubes of activated carbon, and later speciated with
standard gas chromatography with mass spectrometry (GC-MS) techniques (analytical
portion of EPA Method TO-14). These measurements were made in the front and the
back of the bus during on-road time only. An additional measurement was made by each

79
of 2 research staff accompanying the children, combining on-road time and time during
the field trip activities in a downtown museum.
Laboratory Analysis of Urinary PAHs: The samples were shipped on dry ice to the
Centers for Disease Control, National Center for Environmental Health. Ten PAH
metabolites which are present in relatively high concentrations in urine(Huang, et al.,
2006) were measured, using high resolution GC-MS with isotope dilution, a method
developed for analysis of samples in the National Exposure and Nutrition Examination
Survey (Li, et al., 2006). The metabolites measured were 1-hydroxypyrene, 1- and 2-
hydroxynaphthalene, 1-, 2-, 3-, and 4-hydroxyphenanthrene, and 2-, 3-, 9-
hydroxyfluorene. All measurements were adjusted for urinary creatinine.
Laboratory Analysis of Urinary 1-aminopyrene: 1-aminopyrene was analyzed at the
University of Southern California with a HPLC-fluorescence technique (Laumbach, et
al., 2009 ). Prior to the analysis, the pre-field trip samples and the post-field trip samples
for each participant were pooled, resulting in a total of 18 samples for analysis (9 pre-
and 9 post-field trip samples). The pre- and post-field trip samples were weighted by
creatinine concentration to determine the volume each sample contributed to the pooled
sample.
Laboratory Analysis of Biological Markers of Effect: F
2
-isoprostanes were analyzed
using radioimmunoassay developed by Basu (Basu, 1998 ). CC16 was measured with
an enzyme-linked immunosorbent assay (ELISA), using previously described methods
(Andersson, Lundberg and Barregard, 2007 ). Urine creatinine was measured using the

80
Jaffe´ method and used as an adjustment for urinary dilution (Andersson, Lundberg and
Barregard, 2007 ).
Covariate Assessment: Children completed a diet diary on Monday and Tuesday (Figure
1), with help from parents. They were asked to record any meats eaten and to note the
cooking method of meats (e.g. grilled). They were also to note the cooking method of
other foods and whether any foods were browned or blackened. A short questionnaire
was completed during each day of the study reporting on time spent in on-road travel,
the type of road on which most travel occurred, and the level of traffic on the roads.
Study participants reported no environmental tobacco smoke exposure.

CHAPTER 5 STATISTICAL ANALYSIS

The distribution of each creatinine-adjusted metabolite was examined by timing
of urine sample collection. We created a pre-field trip variable by averaging the
measurements from samples collected prior to the field trip ((Monday evening +
Tuesday morning)/2) from Figure 1 and a post-field trip variable from samples collected
after the field trip ((Tuesday afternoon + Tuesday evening + Wednesday morning)/3).
We examined the difference between the pre-field trip and the post-field trip
measurements. Because the measurements were skewed and the sample size not large
enough to assume normality we conducted Wilcoxon signed rank tests on the difference
variable for each of the 10 PAH metabolites, CC16, and F
2
-isoprostane. In sensitivity
analyses to account for diurnal variation in biomarker concentrations unrelated to on-
road exposure, we compared evening post-trip with evening pre-trip measurements,
morning post-trip with morning pre-trip measurements, and afternoon post-trip to

81
evening pre-trip measurements. We calculated intraclass correlations for the pre-trip
measurements and evaluated correlations between PAH metabolite measurements, F2-
Isoprostane, and CC16 using the Spearman correlation coefficient at each collection. We
also examined the correlations between the pre- and post-trip measurements within each
individual and between the pre- and post-trip difference of each biological marker across
individuals. All analyses were conducted with SAS version 9.2 (SAS Institute, Inc.
Cary, North Carolina).

CHAPTER 5 RESULTS

Field Trip Exposure: There were elevated levels of black carbon, CO
2
, NO, and during
travel as measured by the lead vehicle (Figure 5.2). Levels increased markedly when the
bus departed the school. While the bus was parked on a side street in downtown Los
Angeles, the pollutant concentrations measured in the accompanying lead vehicle
decreased markedly. The concentrations again increased during the return trip. A similar
pattern of increased real-time levels of particle-bound PAHs during on-road travel was
also observed (results not shown).

82
Figure 5.2: Real time measurements of black carbon (BC), CO
2
and NO from the lead
vehicle
0
200
400
600
800
1000
1200
0
10
20
30
40
50
7:31
7:47
8:03
8:19
8:36
8:52
9:08
9:24
9:41
9:57
10:13
10:29
10:46
11:02
11:18
11:34
11:51
12:07
12:23
12:39
12:56
1:12
1:28
1:44
2:01
2:17
ppm CO2 or ppb NO
ug/m3 BC
Black Carbon (ug/m3)
CO2 (ppm)
NO (ppb)

Gaseous PAH concentrations measured on the bus during on-road travel were
generally similar to those measured from the lead vehicle (Table 5.1). The speciated
PAHs during travel were generally higher in the rear of the bus than either the front of
the bus or from the lead vehicle (with the exception of benz[a]anthracene and chrysene,
Table 5.1). The higher values measured in the rear of the bus may have been due to
exhaust from the bus itself (Behrentz, et al., 2004 ; Behrentz, et al., 2005). Samples
collected by staff accompanying the children on the bus and in the museum (reflecting
the average exposure on and off the bus) had consistently lower concentrations than the
corresponding samples taken during just on-road travel. There were detectable levels of
benzene throughout the bus and in the lead vehicle. Toluene, ethylbenzene and xylene
were detected only in the rear of the bus.
Departure from school
Arrival at museum
in downtown
Depart
from
downtown
Return to
school

83

Table 5.1: Measurements of PAHs* and BTEX** on the bus and during field trip
Lead
Vehicle
BuBus front.
Travel time
only

Bus, front.
Travel plus
children’s
time off
bus
Bus, back.
Travel time
only
Bus, back
Travel plus
children’s
time off
bus

Naphthalene
Acenaphthene
Fluorene
†

Phenanthrene
Anthracene
Fluoranthrene
Pyrene
Benz[a]anthracene
Chrysene
597
5
-9
156
1
0
14
18
3.9
645
23
41
128
1.6
0
16
13
3.5
416
10
-4
63
0.8
0
4.1
8
0.7
940
33
61
186
1.4
3.5
33
14
3.4
390
13
23
7
0.9
0.7
11
6.1
1.6

Toluene Ethylbenzene Xylene Benzene 0.34
*ng/m
3
; **ppb
†
Negative values: Concentrations close to zero, therefore blank-subtracted data resulted
in negative concentration values.

***SRL (sample reporting limit) for toluene, ethylbenzene, xylene: 0.03 ppb; benzene:
0.06 ppb

Urinary biomarkers: Collection compliance varied across collections. Twelve
participants collected samples Monday evening, 13 Tuesday morning, 15 Tuesday
afternoon following the field trip, 11 Tuesday evening, and 13 Wednesday morning. Of
the 15 participants who collected any urine sample, we report results for the 9 (6 female,
3 male) who collected all five samples. However, a similar pattern of results (not shown)
was observed in sensitivity analyses comparing evening pre- to evening post-trip,

84
morning pre- to morning post-trip, and morning pre- to afternoon post-trip using
measurements from all participants contributing to those comparisons.
The distribution of illustrative metabolites by sample collection is shown in
Table 5.2. In addition to CC16 and F2-Isoprostane, the markers of early effect, we show
results for the naphthols, the metabolites of naphthalene. Naphthalene is the ambient
PAH present in greatest concentration. The 1-hydroxypyrene metabolite of the pyrene
PAH has been examined widely in other studies. There was a wide between-child range
of measurements at each wave of collection. For example, the first evening sample of 1-
naphthol varied almost 20-fold from 699 to 13078 ng/g creatinine. Although the
naphthalene metabolites 1- and 2-naphthol were present in greater concentrations than
other metabolites, the range of concentrations for all PAH metabolites varied by at least
10-fold between individuals, and some, for example 3-hydroxyfluorene, varied at some
collections by more than 100-fold between children (results not shown). CC16 had the
widest range (greater than 2000-fold variation between children at the Wednesday
morning collection). Only F
2
-Isoprostane had a comparatively narrow 2-3 fold range
between children. Compared to this large between-child variability the within-child
variability was relatively smaller. Table 5.3 shows the between-and within-child
variation for measurements of each PAH metabolite, F
2
-Isoprostane and CC-16, based
on the first two measurements (prior to the field trip). The intraclass correlations were
relatively high with 2-naphthol and CC16 having high ICCs of 0.95 and 0.99,
respectively. F2-isoprostane, which had less between-child variability than other
metabolites, had relatively more within-child variability and a low ICC of 0.33.

85
Table 5.2: Distribution of selected urinary biomarker measurements by collection
Biomarker Median Interquartile Range Min-Max

1-naphthol*
Monday evening 1291 1242 699-13078
Tuesday morning 1437 953 508-6471
Tuesday afternoon 1062 505 446-3947
Tuesday evening 1814 1354 612-4833
Wednesday morning 1510 827 673-5600
2-naphthol*
Monday evening 1442 2391 852-10744
Tuesday morning 1304 2903 805-11089
Tuesday afternoon 1283 2387 619-12062
Tuesday evening 2282 2137 989-11718
Wednesday morning 1269 1980 881-12184
1-hydroxypyrene*
Monday evening 114 103 38-215
Tuesday morning 112 113 40-214
Tuesday afternoon 82 73 31-141
Tuesday evening 114 86 51-485
Wednesday morning 110 71 53-265
Clara Cell 16**
Monday evening
†
0.47 0.67 0.07-25.4
Tuesday morning 0.31 0.14 0.05-29.6
Tuesday afternoon 0.26 0.59 0.03-34.9
Tuesday evening 0.69 2.1 0.08-21.6
Wednesday morning 0.34 1.1 0.02-47.4
F2-isoprostane***
Monday evening
†
0.86 0.23 0.49-0.95
Tuesday morning 0.66 0.13 0.28-1.07
Tuesday afternoon 0.73 0.15 0.36-0.88
Tuesday evening 0.77 0.26 0.61-1.16
Wednesday morning 0.61 0.23 0.45-0.91

*ng/g creatinine ; ** µg/g creatinine; ***nmol/mmol creatinine
†N=8, insufficient sample volume for analysis

86
Table 5.3: Standard deviation of between- and within-subject variability and intraclass
correlation coefficients (ICC) of pre-trip measurements*

Between** Within** ICC
1-naphthol 4053.58 1605.66 0.73
2-naphthol 460.73 237.08 0.95
1-hydroxypyrene 267.88 28.07 0.8
2-hydroxyfluorene 1120.40 84.61 0.99
3-hydroxyfluorene 522.73 27.92 0.99
9-hydroxyfluorene 786.98 287.99 0.76
1-hydroxyphenanthrene 392.55 19.68 0.99
2-hydroxyphenanthrene 91.97 11.61 0.97
3-hydroxyphenanthrene 317.99 20.30 0.99
4-hydroxyphenanthrene 160.12 157.49 0.02
Clara Cell 16 12.73 1.06 0.99
F2-isoprostane 0.24 0.17 0.33

*Calculated based on the two pre-trip measurements
**Between and within-subject variability calculated based on the square root of the [mean square
between] and [mean square within] variance, respectively. PAHs in ng/g creatinine; 1-
aminopyrene in pg/mg creatinine; CC16 in µg/g creatinine; F2-isoprostane in nmol/mmol
creatinine

Table 5.4 shows the pre- and post-field trip median concentrations for each
metabolite and the median of the difference between the pre- and post-field trip
measurements for each child. (A positive difference denotes an increase from pre-trip to
post-trip). There was no clear pattern of increased metabolite concentration after the
field trip and the differences between pre- and post-trip measurements were not
statistically significant. CC16 decreased slightly and F
2
-isoprostane increased after the
field trip; however, the changes were not statistically significant.

87

Table 5.4: Pre-trip and post-trip biomarker median values

Biomarker* Pre-trip median Post-trip median Median difference P-value
(IQR
†
) (IQR
†
) (post-pre) (IQR
†
)_______
1-naphthol 1229 (950) 1337 (660) -74 (1036) 0.73
2-naphthol 1285(2654) 1576 (1926) 17 (372) 0.73
2-hydroxyfluorene 147 (89) 114 (163) -5 (72) 0.65
3-hydroxyfluorene 47 (60) 61 (135) 22 (39) 0.25
9-hydroxyfluorene 179 (100) 239 (308) -2 (261) 0.64
1-hydroxyphenanthrene 90 (51) 123 (157) 17 (36) 0.13
2-hydroxyphenanthrene 30 (14) 37 (61) 4 (10) 0.49
3-hydroxyphenanthrene 64 (27) 76 (98) 11 (14) 0.13
4-hydroxyphenanthrene 12 (7) 13 (50) 1 (16) 0.82
1-hydroxypyrene 116 (105) 94 (180) 4 (39) 0.91
1-aminopyrene 33 (20) 53 (28) 14 (28) 0.57
Clara cell 16 0.37 (0.43) 0.37 (1.7) -0.02 (1.5) 0.94
F
2
-isoprostane 0.76 (0.21) 0.7 (0.15) 0.0007 (0.21) 0.84

*PAHs in ng/g creatinine; 1-aminopyrene in pg/mg creatinine; CC16 in µg/g creatinine; F2-isoprostane in
nmol/mmol creatinine
†Interquartile range
N=9 except F2-isoprostane N=8, CC16 and 1-Hydroxyfluorene N=8 due to insufficient sample for
analysis

Figures 5.3a and 5.3b demonstrate the distribution of between-participant variability at
each urine collection and for the pre- and post-trip average for an illustrative metabolite,
1-naphthol. The plot on the left (5.3a) shows 1-naphthol by collection and the plot on
the right (5.3b) shows the pre- and post-field trip values. There was no consistent
pattern of change either between the median pre- and post-trip or between Monday
evening before the field trip and Tuesday evening after the trip. The highest value for
each collection was from the same study subject for whom the post-trip measurement
was lower than the pre-trip measurement. Figure 5.4 illustrates the between-participant
variability for 1-aminopyrene before and after the on-road exposure.

88

Figure 5.3: Urinary 1-naphthol concentration (a) by collection and (b) by averages of
pre- and post-trip collections

Figure 5.4: 1-Aminopyrene by pooled pre- and post-trip collections

89

We also examined the correlation of the difference between each PAH (post-pre)
with the pre-clinical biomarker of physiological effect, CC16, and F
2
-isoprostane and
there were no correlations that were statistically significant. There also was no
consistent pattern of correlation of each PAH with CC16 and F
2
-isoprostane assessed
across individuals (cross-sectionally) at each of the 5 sampling times.
CHAPTER 5 DISCUSSION

We demonstrated the feasibility of recruiting children in a classroom setting to
collect repeated urine samples. However, we did not find a statistically significant
difference between the pre- and post- field trip measurements for any of the ten PAH
metabolites examined also in NHANES, or for 1-aminopyrene, CC16, or F2-
isoprostane. Possible reasons for the lack of association include insufficiently high or
duration of exposure to traffic-related pollutants during the field trip, the small sample
size, that other exposures or physiological variability influenced levels of these
biomarkers, or that the sampling intervals missed an effect.
Traditional American yellow school buses without air conditioning are not well
sealed and outside air can easily penetrate (Behrentz, et al., 2005 ; Borak and Sirianni,
2007). The bus used in this field trip was better sealed and had air conditioning, so
pollutant concentrations in the bus may have differed from ambient levels. However,
naphthalene and other gaseous PAH levels in the bus were comparable to those
measured in the lead vehicle (Table 5.1), which also recorded marked increases in the

90
real-time markers of TRP exposure (in Figure 5.1) corresponding to the on-road periods.
The naphthalene concentrations ranged from 390-940 ng/m
3
. For comparison, these are
higher than levels measured in large surveys of ambient pollutants in southern California
as part of the Multiple Air Toxics Exposure Study (MATES III)(2008) (although the
measurements were not exactly comparable as the MATES samples were averaged over
longer intervals than we have measured). Lead vehicle black carbon measurements were
comparable to on-bus measurements in previous studies (Fruin, Winer and Rodes, 2004
; Sabin, et al., 2005). In a study of bus commuting in southern California, a range of 3-
19 µg/m
3
(Behrentz, et al., 2005) was reported, consistent with the black carbon range in
our study of 2-20µg/m
3
(Figure 5.2). Median in-vehicle NO measurements in another
southern California study were reported to be 390 ppb with an interquartile range of
330-470 ppb (Fruin, Westerdahl, Sax, Sioutas and Fine, 2008 ). Our lead vehicle NO
levels were similar, with an on-road high of 527 ppb (Figure 2). Benzene concentrations
in our study ranged from 0.3 to 1.9 µg/m
3
(Table 5.1), lower than the range of 0.1 to 11
µg/m
3
measured in an earlier southern California school bus study,(Sabin, et al., 2005)
when it is possible that primary tailpipe emissions were higher.
Although we can conclude that the children were likely exposed to high
concentrations of gaseous traffic-related air pollutants typical of freeway travel in Los
Angeles, the level of exposure may not have been enough to elicit a significant change
in urinary biomarker measurements,. In fact, the range of PAH metabolite
concentrations in our study was generally consistent with US population ranges
(Grainger, et al., 2006; Huang, Caudill, Grainger, Needham and Patterson, 2006; Li, et

91
al., 2008 ). For example, in the National Health and Nutrition Examination Survey
(NHANES) the 1999-2000 U.S. geometric mean of 1-hydroxypyrene for children 6-11
years old was 94.1 ng/g creatinine (Grainger, et al., 2006 ), compared with a geometric
mean of 89.3 ng/g creatinine for Monday evening’s collection in our study. A
subsequent NHANES study found a U.S. population geometric mean of 1-
hydroxypyrene for children 6-11 years old of 67 ng/g creatinine (Li, et al., 2008 ). The
U.S. geometric means of 1-naphthol and 2-naphthol for children 6-11 were 1600 and
1890 ng/g creatinine, respectively (Li, et al., 2008 ). We measured geometric means
1597 and 2008 ng/g creatinine for the first measurement of 1- and 2-naphthol. We
measured geometric means 1597 and 2008 ng/g creatinine for the first measurement of
1- and 2-naphthol. Geometric mean urinary CC16 in our study ranged between 0.36 and
0.7 µg/g creatinine across urine collections, with the consistently high outlier across
collections from a boy. Much higher urinary CC16 levels were reported in one previous
study of adults (Timonen, et al., 2004 ). CC16 is, however, also secreted by the prostate
in adult men, and a study restricting urine sampling to mid-portions showed median
urinary CC16 well below 1 µg/g creatinine, in agreement with the present study
(Andersson, Lundberg and Barregard, 2007 ). It is possible that the high level of CC16
we measured in a single boy reflected prostatic sources in a child nearing puberty or
some other reason for a high lung source. There has been little study of urinary CC16 in
children.
Average F2-isoprostane levels of 0.54 (SD=0.49) nmol/mmol creatinine and of
0.27 (SD=0.11) nmol/mmol creatinine have been reported in adults in two studies by the

92
same investigators (Helmersson and Basu, 1999 ; Helmersson and Basu, 2001). In a
larger study, these investigators reported a median level of F2-isoprostane of 0.18
nmol/mmol creatinine in males and 0.2 nmol/mmol creatinine in females (Helmersson
and Basu, 1999 ; Helmersson and Basu, 2001). The mean and median F
2
-isoprostane
measurements in the current study were slightly higher, ranging between a mean .63 to
0.85 nmol/mmol creatinine and median 0.61 to 0.81 nmol/mmol creatinine across the
five sampling points, but there are few data available on normal levels in children.
Previous studies examining associations between traffic-related pollutant
exposures and urinary PAHs and CC16 have not been consistent (Fiala, et al., 2001 ;
Kang, Cho, Kim and Lee, 2002). For example, a Korean study of 137 children in four
schools with differing levels of ambient exposure PAH exposure found an association
with 2-naphthol but not with 1-hydroxypyrene (Kang, Cho, Kim and Lee, 2002 ). A
Czech study found higher urinary 1-hydroxypyrene levels in children at kindergartens in
high traffic areas than in low pollution areas, but the investigators concluded that dietary
intake was the primary source of total PAH dose (Fiala, et al., 2001 ).
Studies have also been conducted of workers with generally higher or longer
exposure. A study of PAH exposure and urinary 1-hydroxypyrene demonstrated 20-fold
greater ambient exposure to PAHs among traffic police than office duty police, which
was reflected in statistically significant higher levels of 1-hydoxypyrene (Ruchirawat, et
al., 2002 ). In another study evaluating the association between atmospheric PAHs and
urinary 1-hydroxypyrene among 99 policemen, those working in patrol cars had 30%
greater exposure than control subjects working in a station, based on personal exposure

93
samples of pyrene. However, the difference of urinary 1-hydroxypyrene between the
control and exposed subjects was not significant, and the investigators speculated that a
higher level of PAH exposure was required to detect a significant difference in urinary
metabolite measurements (Zhang, et al., 2007 ). A study of 34 adult commuters
repeatedly measured serum CC16 before and after commuting but found no consistent
associations with exposure (Zuurbier, et al., 2011 ). A quasi-experimental study in
which 34 healthy adults were exercised above a busy street with heavy particulate
exposure and with filtered air also found no effects on CC16 levels in serum or urine
(Brauner, et al., 2009 ).
Several studies have investigated elimination times for PAH metabolites,
CC16, and F2-isoprostane. Half-lives from PAH metabolite studies vary widely
(Brzeznicki, Jakubowski and Czerski, 1997 ). For example, half-life of 1-hydroxypyrene
ranged from 6 to 35 hours in highly exposed coke workers (Jongeneelen, et al., 1990 ).
The half-life of F
2
-isoprostane is 4 min in rabbits and 16 min in humans, (Basu, 1998;
Basu, 2008; Kaviarasan, et al., 2009) so it is likely that a transient increase associated
with exposure would not have been detected in urine that may have reflected several
hours of collection. For F
2
-isoprostane there is no diurnal variation (Helmersson and
Basu, 1999), but there is for urinary CC16 (Andersson, Lundberg and Barregard, 2007 ).
To account for the half-life and diurnal variation, we collected samples immediately
after the field trip as well as the evening after the trip and the next morning. Although it
is possible that the sampling interval missed transient excursions in some markers or
that complete timed urine collections would have detected a difference between pre- and

94
post-trip measurements, it seems unlikely that a big impact of exposure would not have
been reflected in at least one of the biological markers of interest if these were sensitive
markers of the exposures of interest.
Other exposures might have explained why significant increases in metabolite
measurements were not observed after the field trip. Possible PAH exposure sources
include diet, tobacco smoke, and wood combustion (Fiala, et al., 2001; Grainger, et al.,
2006), and other studies have shown elevated 1-hydroxypyrene levels, for example, in
subjects who consumed charred meat and in smokers (Sorensen, et al., 2003 ). Both F
2
-
isoprostane and CC16 levels are influenced by tobacco smoke exposure (Basu, et al.,
2009 ; Timonen, et al., 2004). We examined the diet diary collected by each child with
help from a parent, and no participant reported eating charred or grilled meat or other
charred food consumption that might explain the higher pre-trip metabolite levels seen
in some subjects. Because half-life of PAH metabolites can be up to 35 hours
(Jongeneelen, et al., 1990) it is possible that diet before the two days of diet diaries
could have influenced the pre-trip measurements. However, diet does not affect F
2
-
isoprostane levels(Helmersson and Basu, 2001), and is thought not to influence the
metabolites of (gaseous) naphthalene or 1-aminopyrene, so poorly measured dietary
exposures are unlikely to have explained the lack of an effect of the on-road exposure.
Questionnaire data on secondhand smoke exposure and other potential sources of
exposure such as heating sources also did not provide an explanation for our results or
for the high biomarker measurements of some individuals. The participant with the
second highest evening pre-field trip measurement of 1-naphthol and the highest pre-

95
field trip level of 1-aminopyrene reported several hours of freeway travel during the
weekend prior to the field trip (Figure 5.3a, 5.3b, and 5.4). On-road travel time did not
provide an explanation for the participant with the highest levels of 1-naphthol or 1-
hydroxyprene.
A few studies have examined between- and within-individual variability. For
example, a study of 200 Dutch children found relatively low correlation of 0.40 between
two measurements of urinary 1-hydroxypyrene in samples collected 3 weeks apart (van
Wijnen, et al., 1996 ). However, in our study there was a correlation of 0.87 between
the pre-trip and post-trip measurements of 1-hydroxypyrene. In another study an ICC of
0.55 was found for spot urine samples of 1-hydroxypyrene collected across seven
days(Li, et al., 2009), lower than the ICC of 0.8 in our study (Table 5.3), however, our
calculation was based on only the two pre-trip sampling time periods, A study of
variability of repeated measurements of urinary F
2
-isoprostanes during 10 days in adults
demonstrated a coefficient of variation of approximately 40%, compared with an
intraclass correlation coefficient of 33% across the two pre-trip collections in our study
(Table 5.3) (Helmersson and Basu, 2001 ). The small sample size in our study limited
the potential to identify effects of ambient exposure using our study design. However,
the data allow estimation of appropriate sample size needed to detect different effects
based on levels observed in Table 5.4. For example, to have 80% power to detect a 20%
change (a change of 478 ng/g creatinine) in the mean pre-exposure 1-naphthol
concentration would require 125 participants (two-sided test, alpha=0.05). To detect a

96
20% change (431 ng/g creatinine) of 2-naphthol would require 40 participants, and a
20% change in 1-hydroxypyrene (23 ng/g creatinine), 43 participants.
Table 5.5: Sample size (N) required to detect 10%, 20% and 30%
changes in mean urinary metabolite levels*
Biomarker Difference % Difference (absolute)** N
1-naphthol
10% 239.1 497
20% 478.2 125
30% 717.3 56
2-naphthol
10% 215.5 159
20% 431 40
30% 646.5 18
2-hydroxyfluorene
10% 42.1 552
20% 84.1 138
30% 126.2 63
3-hydroxyfluorene
10% 17.3 807
20% 34.5 204
30% 51.8 92
9-hydroxyfluorene
10% 37.1 1363
20% 74.1 342
30% 111.2 154
1-hydroxyphenanthrene
10% 18.9 577
20% 37.8 146
30% 56.7 66

97
Table 5.5 continued

2-hydroxyphenanthrene
10% 5.1 727
20% 10.2 184
30% 15.3 83
3-hydroxyphenanthrene
10% 13 793
20% 26 200
30% 39 90
4-hydroxyphenanthrene
10% 6.1 2978
20% 12.2 746
30% 18.4 333
1-hydroxypyrene
10% 11.4 169
20% 22.8 43
30% 34.2 19
1-aminopyrene
10% 5.6 942
20% 11.2 236
30% 16.7 106
Clara cell 16
10% 0.03 8379
20% 0.06 2095
30% 0.09 931
F2-isoprostane
10% 0.07 34
20% 0.14 9
30% 0.21 4

*For 80% power, alpha 0.05; based on change from the pre-trip mean from Table 5.4
**PAHs in ng/g creatinine; 1-aminopyrene in pg/mg creatinine;
CC16 in µg/g creatinine; F2-isoprostane in nmol/mmol creatinine

98
There has been little study of the feasibility of collection of urine samples or in
the utility of these markers in large population studies of children. We demonstrated the
feasibility of urine sample collection from younger children and demonstrated that
collection is possible in a school setting. The procedures for recruitment required a short
meeting with the classes to explain the study and distribute information for children to
take home to parents, and the subsequent logistics of providing materials for home
collection took only 2-3 minutes of class time at the end of the day. Collecting the early
morning samples from children upon return to school required no class time. The overall
rate of agreement to participate in the study by students and parents (15/40; 38%)
suggests that repeated urine collection would be feasible in informative sub-studies of a
substantial proportion of participants in a large cohort, but that special care would be
required to avoid selection bias. There was good compliance with at least one urine
collection (100% once children agreed to participate). The 100% compliance occurred
at the school immediately following the field trip, including one child who completed
none of the collections at home. Although multiple collections are desirable to obtain
more stable estimates of exposure, it is possible that a single spot urine collection at
school that required less commitment from busy parents would have resulted in greater
participation. Such sampling would also ensure exact sampling times, which may be
important for biomarkers for which there is diurnal variability. It is unknown whether
our experience with this private school can be generalized to other settings. The school
principal endorsed the study and encouraged teachers to allow students to participate,

99
but teacher enthusiasm was conditional on the project not interfering with the many
demands of a difficult job.
We conclude that collecting urine samples from children in a school setting is
feasible, but challenging. Additional research is required to determine whether urinary
biomarkers will be useful to assess traffic-related pollutant exposure and early biological
effects of exposure in epidemiological studies of children. Our results suggest that the
markers we studied may be useful only in the setting of relatively high exposures and
large sample sizes.

100

CHAPTER 5 REFERENCES

Andersson L., Lundberg P.A., and Barregard L. Methodological aspects on measurement
of Clara cell protein in urine as a biomarker for airway toxicity, compared with serum
levels. J Appl Toxicol 2007: 27(1): 60-66.

AQMD. MATES III Final Report. http://www.aqmd.gov/prdas/matesIII/matesIII.html,
2008. Accessed.

Bachi A., Zuccato E., Baraldi M., Fanelli R., and Chiabrando C. Measurement of urinary
8-Epi-prostaglandin F2alpha, a novel index of lipid peroxidation in vivo, by
immunoaffinity extraction/gas chromatography-mass spectrometry. Basal levels in
smokers and nonsmokers. Free Radic Biol Med 1996: 20(4): 619-624.

Bae S., Pan X.C., Kim S.Y., Park K., Kim Y.H., Kim H., et al. Exposures to particulate
matter and polycyclic aromatic hydrocarbons and oxidative stress in schoolchildren.
Environ Health Perspect 2010: 118(4): 579-583.

Barregard L., Sällsten G., Gustafson P., Andersson L., Johansson L., Basu S., et al.
Experimental Exposure to Wood-Smoke Particles in Healthy Humans: Effects on
Markers of Inflammation, Coagulation, and Lipid Peroxidation. Inhalation Toxicology
2006: 18(11): 845-853.

Barregard L., Sallsten G., Andersson L., Almstrand A.C., Gustafson P., Andersson M., et
al. Experimental exposure to wood smoke: effects on airway inflammation and oxidative
stress. Occup Environ Med 2008: 65(5): 319-324.

Basu S. Radioimmunoassay of 15-keto-13,14-dihydro-prostaglandin F2alpha: an index
for inflammation via cyclooxygenase catalysed lipid peroxidation. Prostaglandins Leukot
Essent Fatty Acids 1998: 58(5): 347-352.

Basu S. Metabolism of 8-iso-prostaglandin F2α. FEBS Letters 1998: 428(1-2): 32-36.

Basu S. F2-isoprostanes in human health and diseases: from molecular mechanisms to
clinical implications. Antioxid Redox Signal 2008: 10(8): 1405-1434.

Basu S., Helmersson J., Jarosinska D., Sällsten G., Mazzolai B., and Barregård L.
Regulatory factors of basal F2-isoprostane formation: Population, age, gender and
smoking habits in humans. Free Radical Research 2009: 43(1): 85-91.

101
Behrentz E., Fitz D.R., Pankratz D.V., Sabin L.D., Colome S.D., Fruin S.A., et al.
Measuring self-pollution in school buses using a tracer gas technique. Atmospheric
Environment 2004: 38(23): 3735-3746.

Behrentz E., Sabin L.D., Winer A.M., Fitz D.R., Pankratz D.V., Colome S.D., et al.
Relative importance of school bus-related microenvironments to children's pollutant
exposure. J Air Waste Manag Assoc 2005: 55(10): 1418-1430.

Borak J., and Sirianni G. Studies of self-pollution in diesel school buses: methodological
issues. J Occup Environ Hyg 2007: 4(9): 660-668.

Braback L., and 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.

Brauer M., Hoek G., Van Vliet P., Meliefste K., Fischer P.H., Wijga A., et al. Air
pollution from traffic and the development of respiratory infections and asthmatic and
allergic symptoms in children. Am J Respir Crit Care Med 2002: 166(8): 1092-1098.

Brauner E.V., Mortensen J., Moller P., Bernard A., Vinzents P., Wahlin P., et al. Effects
of ambient air particulate exposure on blood-gas barrier permeability and lung function.
Inhal Toxicol 2009: 21(1): 38-47.

Broeckaert F., and Bernard A. Clara cell secretory protein (CC16): characteristics and
perspectives as lung peripheral biomarker. Clin Exp Allergy 2000: 30(4): 469-475.

Broeckaert F., Clippe A., Knoops B., Hermans C., and Bernard A. Clara cell secretory
protein (CC16): features as a peripheral lung biomarker. Ann N Y Acad Sci 2000: 923: 68-
77.

Brunekreef B., Janssen N.A., de Hartog J., Harssema H., Knape M., and van Vliet P. Air
pollution from truck traffic and lung function in children living near motorways.
Epidemiology 1997: 8(3): 298-303.

Brzeznicki S., Jakubowski M., and Czerski B. Elimination of 1-hydroxypyrene after
human volunteer exposure to polycyclic aromatic hydrocarbons. Int Arch Occup Environ
Health 1997: 70(4): 257-260.

CalTrans Traffic and Vehicle Data Systems Unit, 2008.

Chen J.C., and Schwartz J. Neurobehavioral effects of ambient air pollution on cognitive
performance in US adults. Neurotoxicology 2009: 30(2): 231-239.

102
Chiabrando C., Valagussa A., Rivalta C., Durand T., Guy A., Zuccato E., et al.
Identification and measurement of endogenous beta-oxidation metabolites of 8-epi-
Prostaglandin F2alpha. J Biol Chem 1999: 274(3): 1313-1319.

English P., Neutra R., Scalf R., Sullivan M., Waller L., and Zhu L. Examining
associations between childhood asthma and traffic flow using a geographic information
system. Environ Health Perspect 1999: 107(9): 761-767.

Fiala Z., Vyskocil A., Krajak V., Viau C., Ettlerova E., Bukac J., et al. Environmental
exposure of small children to polycyclic aromatic hydrocarbons. Int Arch Occup Environ
Health 2001: 74(6): 411-420.

Fruin S., Westerdahl D., Sax T., Sioutas C., and Fine P. Measurements and predictors of
on-road ultrafine particle concentrations and associated pollutants in Los Angeles.
Atmospheric Environment 2008: 42: 207-219.

Fruin S.A., Winer A.M., and Rodes C.E. Black carbon concentrations in California
vehicles and estimation of in-vehicle diesel exhaust particulate matter exposures.
Atmospheric Environment 2004: 38(25): 4123-4133.

Gauderman W.J., Avol E., Lurmann F., Kuenzli N., Gilliland F., Peters J., et al.
Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology 2005:
16(6): 737-743.

Gauderman W.J., Vora H., McConnell R., Berhane K., Gilliland F., Thomas D., et al.
Effect of exposure to traffic on lung development from 10 to 18 years of age: a cohort
study. Lancet 2007: 369(9561): 571-577.

Gilliland F.D., McConnell R., Peters J., and Gong H., Jr. A theoretical basis for
investigating ambient air pollution and children's respiratory health. Environ Health
Perspect 1999: 107 Suppl 3: 403-407.

Gotschi T., Sunyer J., Chinn S., de Marco R., Forsberg B., Gauderman J.W., et al. Air
pollution and lung function in the European Community Respiratory Health Survey. Int J
Epidemiol 2008: 37(6): 1349-1358.

Grainger J., Huang W., Patterson D.G., Jr., Turner W.E., Pirkle J., Caudill S.P., et al.
Reference range levels of polycyclic aromatic hydrocarbons in the US population by
measurement of urinary monohydroxy metabolites. Environ Res 2006: 100(3): 394-423.

Helmersson J., and Basu S. F2-isoprostane excretion rate and diurnal variation in human
urine. Prostaglandins Leukot Essent Fatty Acids 1999: 61(3): 203-205.

103
Helmersson J., and Basu S. F(2)-isoprostane and prostaglandin F(2 alpha)metabolite
excretion rate and day to day variation in healthy humans. Prostaglandins Leukot Essent
Fatty Acids 2001: 65(2): 99-102.

Helmersson J., Larsson A., Vessby B., and Basu S. Active smoking and a history of
smoking are associated with enhanced prostaglandin F(2alpha), interleukin-6 and F2-
isoprostane formation in elderly men. Atherosclerosis 2005: 181(1): 201-207.

Hole D.J., Watt G.C., Davey-Smith G., Hart C.L., Gillis C.R., and Hawthorne V.M.
Impaired lung function and mortality risk in men and women: findings from the Renfrew
and Paisley prospective population study. BMJ 1996: 313(7059): 711-715; discussion
715-716.

Holguin F., Flores S., Ross Z., Cortez M., Molina M., Molina L., et al. Traffic-related
exposures, airway function, inflammation, and respiratory symptoms in children. Am J
Respir Crit Care Med 2007: 176(12): 1236-1242.

Huang W., Caudill S.P., Grainger J., Needham L.L., and Patterson D.G., Jr. Levels of 1-
hydroxypyrene and other monohydroxy polycyclic aromatic hydrocarbons in children: a
study based on U.S. reference range values. Toxicol Lett 2006: 163(1): 10-19.

Islam T., Gauderman W.J., Berhane K., McConnell R., Avol E., Peters J.M., et al.
Relationship between air pollution, lung function and asthma in adolescents. Thorax
2007: 62(11): 957-963.

Jerrett M., Arain A., Kanaroglou P., Beckerman B., Potoglou D., Sahsuvaroglu T., et al.
A review and evaluation of intraurban air pollution exposure models. J Expo Anal
Environ Epidemiol 2005: 15(2): 185-204.

Jongeneelen F.J., van Leeuwen F.E., Oosterink S., Anzion R.B., van der Loop F., Bos
R.P., et al. Ambient and biological monitoring of cokeoven workers: determinants of the
internal dose of polycyclic aromatic hydrocarbons. Br J Ind Med 1990: 47(7): 454-461.

Kang J.W., Cho S.H., Kim H., and Lee C.H. Correlation of urinary 1-hydroxypyrene and
2-naphthol with total suspended particulates in ambient air in municipal middle-school
students in Korea. Arch Environ Health 2002: 57(4): 377-382.

Kaviarasan S., Muniandy S., Qvist R., and Ismail I.S. F(2)-isoprostanes as novel
biomarkers for type 2 diabetes: a review. J Clin Biochem Nutr 2009: 45(1): 1-8.

Kim J.J., Huen K., Adams S., Smorodinsky S., Hoats A., Malig B., et al. Residential
traffic and children's respiratory health. Environ Health Perspect 2008: 116(9): 1274-
1279.

104
Klepeis N.E., Nelson W.C., Ott W.R., Robinson J.P., Tsang A.M., Switzer P., et al. The
National Human Activity Pattern Survey (NHAPS): a resource for assessing exposure to
environmental pollutants. J Expo Anal Environ Epidemiol 2001: 11(3): 231-252.

Knuiman M.W., James A.L., Divitini M.L., Ryan G., Bartholomew H.C., and Musk A.W.
Lung function, respiratory symptoms, and mortality: results from the Busselton Health
Study. Ann Epidemiol 1999: 9(5): 297-306.

Kozawa K.H., Fruin S.A., and Winer A.M. Near-road air pollution impacts of goods
movement in communities adjacent to the Ports of Los Angeles and Long Beach.
Atmospheric Environment 2009: 43(18): 2960-2970.

Kunzli N., Bridevaux P.O., Liu L.J., Garcia-Esteban R., Schindler C., Gerbase M.W., et
al. Traffic-related air pollution correlates with adult-onset asthma among never-smokers.
Thorax 2009: 64(8): 664-670.

Larsen R.K., 3rd, and Baker J.E. Source apportionment of polycyclic aromatic
hydrocarbons in the urban atmosphere: a comparison of three methods. Environ Sci
Technol 2003: 37(9): 1873-1881.

Laumbach R., Tong J., Zhang L., Ohman-Strickland P., Stern A., Fiedler N., et al.
Quantification of 1-aminopyrene in human urine after a controlled exposure to diesel
exhaust. J Environ Monit 2009: 11(1): 153-159.

Laumbach R.J., and Kipen H.M. Acute effects of motor vehicle traffic-related air
pollution exposures on measures of oxidative stress in human airways. Ann N Y Acad Sci
2010: 1203: 107-112.

Li N., Sioutas C., Cho A., Schmitz D., Misra C., Sempf J., et al. Ultrafine particulate
pollutants induce oxidative stress and mitochondrial damage. Environ Health Perspect
2003: 111(4): 455-460.

Li Z., Romanoff L.C., Trinidad D.A., Hussain N., Jones R.S., Porter E.N., et al.
Measurement of urinary monohydroxy polycyclic aromatic hydrocarbons using
automated liquid-liquid extraction and gas chromatography/isotope dilution high-
resolution mass spectrometry. Anal Chem 2006: 78(16): 5744-5751.

Li Z., Sandau C.D., Romanoff L.C., Caudill S.P., Sjodin A., Needham L.L., et al.
Concentration and profile of 22 urinary polycyclic aromatic hydrocarbon metabolites in
the US population. Environ Res 2008: 107(3): 320-331.

Li Z., Romanoff L.C., Lewin M.D., Porter E.N., Trinidad D.A., Needham L.L., et al.
Variability of urinary concentrations of polycyclic aromatic hydrocarbon metabolite in

105
general population and comparison of spot, first-morning, and 24-h void sampling. J
Expo Sci Environ Epidemiol 2009.

Marr L.C., Kirchstetter T.W., Harley R.A., Miguel A.H., Hering S.V., and Hammond
S.K. Characterization of Polycyclic Aromatic Hydrocarbons in Motor Vehicle Fuels and
Exhaust Emissions. Environmental Science & Technology 1999: 33(18): 3091-3099.

McConnell R., Berhane K., Yao L., Jerrett M., Lurmann F., Gilliland F., et al. Traffic,
susceptibility, and childhood asthma. Environ Health Perspect 2006: 114(5): 766-772.

Merlo F., Andreassen A., Weston A., Pan C.F., Haugen A., Valerio F., et al. Urinary
excretion of 1-hydroxypyrene as a marker for exposure to urban air levels of polycyclic
aromatic hydrocarbons. Cancer Epidemiol Biomarkers Prev 1998: 7(2): 147-155.

Morrow J.D., Frei B., Longmire A.W., Gaziano J.M., Lynch S.M., Shyr Y., et al.
Increase in circulating products of lipid peroxidation (F2-isoprostanes) in smokers.
Smoking as a cause of oxidative damage. N Engl J Med 1995: 332(18): 1198-1203.

Morrow J.D. Quantification of isoprostanes as indices of oxidant stress and the risk of
atherosclerosis in humans. Arterioscler Thromb Vasc Biol 2005: 25(2): 279-286.

Nordling E., Berglind N., Melen E., Emenius G., Hallberg J., Nyberg F., et al. Traffic-
related air pollution and childhood respiratory symptoms, function and allergies.
Epidemiology 2008: 19(3): 401-408.

Obata T., Tomaru K., Nagakura T., Izumi Y., and Kawamoto T. Smoking and oxidant
stress: assay of isoprostane in human urine by gas chromatography-mass spectrometry. J
Chromatogr B Biomed Sci Appl 2000: 746(1): 11-15.

Panther B.C., Hooper M.A., and Tapper N.J. A comparison of air particulate matter and
associated polycyclic aromatic hydrocarbons in some tropical and temperate urban
environments. Atmospheric Environment 1999: 33(24-25): 4087-4099.

Ruchirawat M., Mahidol C., Tangjarukij C., Pui-ock S., Jensen O., Kampeerawipakorn
O., et al. Exposure to genotoxins present in ambient air in Bangkok, Thailand — particle
associated polycyclic aromatic hydrocarbons and biomarkers. Science of The Total
Environment 2002: 287(1-2): 121-132.

Sabin L.D., Behrentz E., Winer A.M., Jeong S., Fitz D.R., Pankratz D.V., et al.
Characterizing the range of children's air pollutant exposure during school bus commutes.
J Expo Anal Environ Epidemiol 2005: 15(5): 377-387.

Salam M.T., Islam T., and Gilliland F.D. Recent evidence for adverse effects of
residential proximity to traffic sources on asthma. Curr Opin Pulm Med 2008: 14(1): 3-8.

106

Sin D.D., Wu L., and Man S.F. The relationship between reduced lung function and
cardiovascular mortality: a population-based study and a systematic review of the
literature. Chest 2005: 127(6): 1952-1959.

Sorensen M., Autrup H., Moller P., Hertel O., Jensen S.S., Vinzents P., et al. Linking
exposure to environmental pollutants with biological effects. Mutat Res 2003: 544(2-3):
255-271.

Timonen K.L., Hoek G., Heinrich J., Bernard A., Brunekreef B., de Hartog J., et al. Daily
variation in fine and ultrafine particulate air pollution and urinary concentrations of lung
Clara cell protein CC16. Occup Environ Med 2004: 61(11): 908-914.

Tuntawiroon J., Mahidol C., Navasumrit P., Autrup H., and Ruchirawat M. Increased
health risk in Bangkok children exposed to polycyclic aromatic hydrocarbons from
traffic-related sources. Carcinogenesis 2006: 28(4): 816-822.

van Wijnen J.H., Slob R., Jongmans-Liedekerken G., van de Weerdt R.H., and
Woudenberg F. Exposure to polycyclic aromatic hydrocarbons among Dutch children.
Environ Health Perspect 1996: 104(5): 530-534.

Westerdahl D., Fruin S., Sax T., Fine P.M., and Sioutas C. Mobile platform
measurements of ultrafine particles and associated pollutant concentrations on freeways
and residential streets in Los Angeles. Atmospheric Environment 2005: 39(20): 3597-
3610.

Zhang W., Xu D., Zhuang G., Ding C., Wang G., Chang J., et al. A pilot study on using
urinary 1-hydroxypyrene biomarker for exposure to PAHs in Beijing. Environ Monit
Assess 2007: 131(1-3): 387-394.

Zhou Y., and Levy J.I. Factors influencing the spatial extent of mobile source air
pollution impacts: a meta-analysis. BMC Public Health 2007: 7: 89.

Zhu Y., Hinds W.C., Kim S., and Sioutas C. Concentration and size distribution of
ultrafine particles near a major highway. J Air Waste Manag Assoc 2002: 52(9): 1032-
1042.

Zuurbier M., Hoek G., Oldenwening M., Meliefste K., Krop E., van den Hazel P., et al.
In-Traffic Air Pollution Exposure and CC16, Blood Coagulation, and Inflammation
Markers in Healthy Adults. Environ Health Perspect 2011.

Zwirner-Baier I., and Neumann H.-G. Polycyclic nitroarenes (nitro-PAHs) as biomarkers
of exposure to diesel exhaust. Mutation Research/Genetic Toxicology and Environmental
Mutagenesis 1999: 441(1): 135-144.

107

108
CHAPTER 6: CONCLUSION

New tools I have evaluated have promise to provide better estimates of the true
health effects of air pollution and reduce their uncertainty. They have potential to assess
the contribution of ambient exposures at locations away from home (based on GPS
logging), the effect of exercise on dose (based on step counts), and to provide an
integrated assessment of dose (based on biological markers of PAH) and an assessment
of the biological mechanisms of health effects based on oxidative stress (F2-isoprostane)
and inflammation (CC16). (See Figure 6.1).
Figure 6.1: A way forward in epidemiological studies of traffic-related air pollution and
respiratory health effects

Considerable effort has been made to improve traffic-related pollutant exposure
assessment at home, but there has been much less attention paid to the contribution of
time-activity patterns to actual dose. For example, many studies have examined the
association between modeled residential near-roadway TRP exposure and asthma.
However, provocative new research suggests that time spent at other locations, such as
school and during on road travel, are associated as strongly with childhood asthma
incidence and exacerbation as are residential exposures, even though children spent much

109
less time in these locations compared to time at home (McConnell, et al., 2010;
McConnell, et al., 2010). Studies conducted at USC have suggested that exercise based
on self-reported participation in team sports can increase the risk of asthma by 3-4 fold in
high pollution communities, but not in low pollution communities (McConnell, et al.,
2002). The investigators suggested that this crude proxy for physical activity was a
marker of increased dose of air pollution associated with the increased ventilation
associated with physical activity. Children reporting greater time spent outdoors in
communities and therefore more exposed to ambient pollution had stronger associations
of particulate matter with lung function growth deficits, compared to children reporting
less time spent outdoors(Gauderman, et al., 2002). These are very crude indicators of the
influence of physical activity and time outdoors on dose. If the observed associations are
causal then the use of these misclassified indicators of time-activity almost certainly
underestimates the true effects.
If we were to combine better exercise and time outdoors information we could
dramatically improve our estimates of health effects. Consider, for example, a
hypothetical situation of a child in Long Beach who is physically active on a school
playground within 50m of a freeway with heavy truck traffic compared with the TRP
exposure of a child in Santa Barbara sitting in a classroom 500m from a freeway with low
truck traffic. Studies of the dispersion of TRP such as ultrafine particulates suggest that
the more TRP exposed child would experience, on average, about a fivefold increase
exposure based on distance to a freeway (Zhu, Hinds, Kim and Sioutas, 2002). Moderate
and vigorous physical activity (MVPA) is associated with almost a six-fold increase in
minute ventilation (and therefore of potential dose of pollution to the lungs)(Cabrera, et

110
al., 2002). A further two-fold gradient in exposure to small particles is associated with
being outdoors(Long, Suh, Catalano and Koutrakis, 2001). The local spatial gradient in
the school-morning traffic peak is unknown but is at least three-fold between Long Beach
and Santa Barbara (Figure 6.2, courtesy of Rob McConnell). Thus, a potential difference
between these two children is on the order of an astounding 180-fold. Although this is an
extreme example of dose gradients between children, if these differences are reflected at
all in health effect estimates then we are leaving unmeasured large sources of variability
in the inter-individual differences in potential dose in our studies that estimate only
yearly average residential TRP effects on asthma prevalence and incidence.

Figure 6.2: Diurnal pattern of NO at community monitors during weekdays in Long Beach (left)
and Santa Barbara (right)

Integrated data from the GPS logger and SportBrain pedometer could potentially
be used to model personal exposure (based on information on location) and inhaled
exposure (based on location and physical activity) to spatial and temporal variation in
TRP at school, home and other locations. This better characterization of these time-
activity patterns in key pollution microenvironments would improve of our understanding
Long Beach

111
of the relationship of exposure, time-activity and susceptibility to asthma-related
outcomes and lung function growth.
The exposure biomarkers could play an important role in refining the estimates of
TRP epidemiological effect to the extent they integrate time-activity and exposure in a
dose metric. In addition there is potential to identify specific toxicological groups of
compounds in the case of PAHs or sources in the case of 1-aminopyrene which is specific
to diesel. 1-aminopyrene is an especially interesting marker, because there is little way
using current exposure assessment methods to distinguish diesel exhaust pollution health
effects from gasoline health effects. Thus epidemiological studies using these metabolites
have profound implications for policy because they can provide a target for more specific
source regulation. The ability to demonstrate a pathway of effects from an integrated
exposure biomarker to early oxidative stress or inflammation effects known to be
involved in pathogenesis of asthma, lung function damages and other air pollution effects
could markedly strengthen etiological causal inferences that are a prerequisite for
effective regulatory action.
Now I will consider the possibility (based on my studies) of achieving the
promise of new approaches to air pollution epidemiology that I have outlined.
The SportBrain iStep X1 shows great promise for use in epidemiological studies
to refine estimates of air pollution dose. The low cost per unit makes it feasible for large
population studies and the SportBrain provides a clear advantage over other step counters
because it provides time resolved step counts. As described in Chapter 3 an important
challenge is to relate the time resolved step counts to minute ventilation (and dose).
Although there has been some evaluation of the number of steps per minute associated

112
with thresholds for MVPA and minute ventilation (Lubans, et al., 2008; Lubans, Morgan,
Collins, Boreham and Callister, 2009; Scruggs, Beveridge, Eisenman, Watson, Shultz and
Ransdell, 2003), further study is needed to better understand the association between the
continuum of steps per minute and minute ventilation, or at least to be able to assign
duration of low, moderate, or vigorous PA to children of different ages and fitness levels.
One way forward would be develop regression equations such as those used with
accelerometers to translate count data into estimates of energy expenditure which is a
good marker for minute ventilation (Freedson, Melanson and Sirard, 1998; Strath, Bassett
and Swartz, 2003). The role of height, weight, and stride length measurements as
moderators of the relationship between steps and energy expenditure warrants study as
these can be measured in the field. Better understanding of the relationship of step counts,
fitness, and energy expenditure would be helpful, as relatively simple tests of fitness
such the Queen’s College Step Test (McArdle, et al., 1972) can be conducted in the field.
Additional laboratory studies examining these research questions require equipment not
available to me to measure energy expenditure (for example by measuring oxygen
consumption), but a partnership between epidemiologists and exercise physiologists
extending beyond my collaboration with the Occupational Therapy department would
make these studies feasible. The collection of SportBrain data was not logistically
difficult but appropriate real-time protocols and procedures for quality control, data
collection, data reduction, and analysis would be required for large studies.
My assessment of the potential of GPS loggers to link physical activity with
location in proximity to roadways and associated TRP exposures is much less promising.
The GPS loggers I studied have limited utility in identifying time spent in TRP

113
microenvironments relevant to air pollution research because there was a large amount of
missing data. Only a small percentage of time can be assigned to specific
microenvironments. An exception was on-road time which can be highly influential in
overall TRP exposure estimates. Primary traffic emissions, including ultrafine
particulates have been found to be present in very high concentrations in the vicinity of
major roadways (Zhu, Hinds, Kim and Sioutas, 2002), and exceedingly high ultrafine
exposures have been found in vehicles and school buses (Behrentz, et al., 2005; Sabin, et
al., 2005; Westerdahl, Fruin, Sax, Fine and Sioutas, 2005). In Los Angeles 33-45% of
daily ultrafine particulate exposure can occur from on-road commuting exposure (Fruin,
Westerdahl, Sax, Sioutas and Fine, 2008). It is difficult to speculate on the impact on
health effect estimates of better measurements of on-road time but a USC study showing
an association of travel time to school (assessed using a Mapquest algorithm) with
suggests that the impact may be substantial (McConnell, Liu, Wu, Lurmann, Peters and
Berhane, 2010).
As GPS technology advances, the utility of GPS loggers to assess time in different
locations may improve. GPS devices which provide data on the number of satellites in
view could at least help refine uncertainty as to whether the device in on or off or out of
view of satellites.
My study leaves many questions unanswered regarding the utility of urinary
biomarkers of TRP exposure for assessing respiratory (or other) health effects in large
epidemiological studies. The relationship of exposure to urine biomarkers in relevant
epidemiologic settings, for example in determining on-road exposure for childhood
asthma epidemiology, must be demonstrated before these markers would warrant large-

114
scale use. We did answer an important logistical question by demonstrating the feasibility
of collecting urine samples from children and found that high compliance with little staff
time for a single sample is likely among children who agree to participate. Considerably
more effort was required to obtain multiple collections from non-school hours with lower
collection compliance. For large population studies spot urine collections at school are a
logistically feasible option. Although urine biomarkers of PAH exposure and oxidative
stress response were not associated with a heavy TRP exposure challenge in the school
field trip, I was able to estimate sample size requirement for detecting moderate effects of
exposure. For a future study, I could increase the sample size to examine association of
acute exposures on bus travel with the biomarkers. However, the Children’s Health Study
investigators at USC are interested in chronic exposure, and these markers may have
better applications to more steady state exposures relevant to chronic effects. For
example, if they are associated with distance to a busy roadway over the constant period
of exposure this entails, the marker could be useful as an integrative dose metric. I would
like to evaluate the association of urinary PAHs and 1-aminopyrene in single spot urine
collected from a large population of children who reside at differing distances from a
major roadway with heavy truck traffic and examine associations with distance and with
asthma prevalence and incidence.
In conclusion, better assessment of time-activity and integrated biomarkers of
exposure and effect are needed to improve epidemiological estimates of air pollution
health effects. There are important implications for public health policy and regulations.
My studies have provided a rationale for the use of the SportBrain in large
epidemiological studies and for GPS in assessing effects of on-road exposure. I have

115
identified limitations of GPS not commonly addressed by the many individuals using this
tool for time-activity assessment. I have identified uncertainties in the potential of urinary
PAH metabolites, biomarkers of oxidative stress and lung inflammation for large
population epidemiological studies.

116

CHAPTER 6 REFERENCES

Behrentz E., Sabin L.D., Winer A.M., Fitz D.R., Pankratz D.V., Colome S.D., et al.
Relative importance of school bus-related microenvironments to children's pollutant
exposure. J Air Waste Manag Assoc 2005: 55(10): 1418-1430.

Cabrera M.E., Portzline G., Aach S., Condie C., Dorostkar P., and Mianulli M. Can
current minute ventilation rate adaptive pacemakers provide appropriate chronotropic
response in pediatric patients? Pacing Clin Electrophysiol 2002: 25(6): 907-914.

Freedson P.S., Melanson E., and Sirard J. Calibration of the Computer Science and
Applications, Inc. accelerometer. Med Sci Sports Exerc 1998: 30(5): 777-781.

Fruin S., Westerdahl D., Sax T., Sioutas C., and Fine P. Measurements and predictors of
on-road ultrafine particle concentrations and associated pollutants in Los Angeles.
Atmospheric Environment 2008: 42: 207-219.

Gauderman W.J., Gilliland G.F., Vora H., Avol E., Stram D., McConnell R., et al.
Association between air pollution and lung function growth in southern California
children: results from a second cohort. Am J Respir Crit Care Med 2002: 166(1): 76-84.

Long C.M., Suh H.H., Catalano P.J., and Koutrakis P. Using time- and size-resolved
particulate data to quantify indoor penetration and deposition behavior. Environ Sci
Technol 2001: 35(10): 2089-2099.

Lubans D.R., Morgan P.J., Callister R., and Collins C.E. The Relationship Between
Pedometer Step Counts and Estimated VO₂Max as Determined by a
Submaximal Fitness Test in Adolescents. Pediatric Exercise Science 2008: 20(3): 273-
284.

Lubans D.R., Morgan P.J., Collins C.E., Boreham C.A., and Callister R. The relationship
between heart rate intensity and pedometer step counts in adolescents. J Sports Sci 2009:
27(6): 591-597.

McArdle W.D., Katch F.I., Pechar G.S., Jacobson L., and Ruck S. Reliability and
interrelationships between maximal oxygen intake, physical work capacity and step-test
scores in college women. Med Sci Sports 1972: 4(4): 182-186.

McConnell R., Berhane K., Gilliland F., London S.J., Islam T., Gauderman W.J., et al.
Asthma in exercising children exposed to ozone: a cohort study. Lancet 2002: 359(9304):
386-391.

117
McConnell R., Islam T., Shankardass K., Jerrett M., Lurmann F., Gilliland F., et al.
Childhood incident asthma and traffic-related air pollution at home and school. Environ
Health Perspect 2010: 118(7): 1021-1026.

McConnell R., Liu F., Wu J., Lurmann F., Peters J., and Berhane K. Asthma and school
commuting time. J Occup Environ Med 2010: 52(8): 827-828.

Sabin L.D., Behrentz E., Winer A.M., Jeong S., Fitz D.R., Pankratz D.V., et al.
Characterizing the range of children's air pollutant exposure during school bus commutes.
J Expo Anal Environ Epidemiol 2005: 15(5): 377-387.

Scruggs P.W., Beveridge S.K., Eisenman P.A., Watson D.L., Shultz B.B., and Ransdell
L.B. Quantifying physical activity via pedometry in elementary physical education. Med
Sci Sports Exerc 2003: 35(6): 1065-1071.

Strath S.J., Bassett D.R., Jr., and Swartz A.M. Comparison of MTI accelerometer cut-
points for predicting time spent in physical activity. Int J Sports Med 2003: 24(4): 298-
303.

Westerdahl D., Fruin S., Sax T., Fine P.M., and Sioutas C. Mobile platform
measurements of ultrafine particles and associated pollutant concentrations on freeways
and residential streets in Los Angeles. Atmospheric Environment 2005: 39(20): 3597-
3610.

Zhu Y., Hinds W.C., Kim S., and Sioutas C. Concentration and size distribution of
ultrafine particles near a major highway. J Air Waste Manag Assoc 2002: 52(9): 1032-
1042.

118

COMPREHENSIVE REFERENCES

Adams C., Riggs P., and Volckens J. Development of a method for personal,
spatiotemporal exposure assessment. J Environ Monit 2009: 11(7): 1331-1339.

Andersson L., Lundberg P.A., and Barregard L. Methodological aspects on measurement
of Clara cell protein in urine as a biomarker for airway toxicity, compared with serum
levels. J Appl Toxicol 2007: 27(1): 60-66.

Anonymous. Health effects of outdoor air pollution. Part 2. Committee of the
Environmental and Occupational Health Assembly of the American Thoracic Society.
American Journal of Respiratory and Critical Care Medicine 1996: 153(2): 477-498.

Anonymous. Health effects of outdoor air pollution. Committee of the Environmental and
Occupational Health Assembly of the American Thoracic Society. American Journal of
Respiratory and Critical Care Medicine 1996: 153(1): 3-50.

AQMD. MATES III Final Report. http://www.aqmd.gov/prdas/matesIII/matesIII.html,
2008. Accessed.

Avol E.L., Gauderman W.J., Tan S.M., London S.J., and Peters J.M. Respiratory effects
of relocating to areas of differing air pollution levels. Am J Respir Crit Care Med 2001:
164(11): 2067-2072.

Bachi A., Zuccato E., Baraldi M., Fanelli R., and Chiabrando C. Measurement of urinary
8-Epi-prostaglandin F2alpha, a novel index of lipid peroxidation in vivo, by
immunoaffinity extraction/gas chromatography-mass spectrometry. Basal levels in
smokers and nonsmokers. Free Radic Biol Med 1996: 20(4): 619-624.

Bae S., Pan X.C., Kim S.Y., Park K., Kim Y.H., Kim H., et al. Exposures to particulate
matter and polycyclic aromatic hydrocarbons and oxidative stress in schoolchildren.
Environ Health Perspect 2010: 118(4): 579-583.

Bailey R.C., Olson J., Pepper S.L., Porszasz J., Barstow T.J., and Cooper D.M. The level
and tempo of children's physical activities: an observational study. Med Sci Sports Exerc
1995: 27(7): 1033-1041.

Barregard L., Sällsten G., Gustafson P., Andersson L., Johansson L., Basu S., et al.
Experimental Exposure to Wood-Smoke Particles in Healthy Humans: Effects on
Markers of Inflammation, Coagulation, and Lipid Peroxidation. Inhalation Toxicology
2006: 18(11): 845-853.

119
Barregard L., Sallsten G., Andersson L., Almstrand A.C., Gustafson P., Andersson M., et
al. Experimental exposure to wood smoke: effects on airway inflammation and oxidative
stress. Occup Environ Med 2008: 65(5): 319-324.

Basu S. Radioimmunoassay of 15-keto-13,14-dihydro-prostaglandin F2alpha: an index
for inflammation via cyclooxygenase catalysed lipid peroxidation. Prostaglandins Leukot
Essent Fatty Acids 1998: 58(5): 347-352.

Basu S. Metabolism of 8-iso-prostaglandin F2α. FEBS Letters 1998: 428(1-2): 32-36.

Basu S. F2-isoprostanes in human health and diseases: from molecular mechanisms to
clinical implications. Antioxid Redox Signal 2008: 10(8): 1405-1434.

Basu S., Helmersson J., Jarosinska D., Sällsten G., Mazzolai B., and Barregård L.
Regulatory factors of basal F2-isoprostane formation: Population, age, gender and
smoking habits in humans. Free Radical Research 2009: 43(1): 85-91.

Beets M.W., Patton M.M., and Edwards S. The accuracy of pedometer steps and time
during walking in children. Med Sci Sports Exerc 2005: 37(3): 513-520.

Behrentz E., Fitz D.R., Pankratz D.V., Sabin L.D., Colome S.D., Fruin S.A., et al.
Measuring self-pollution in school buses using a tracer gas technique. Atmospheric
Environment 2004: 38(23): 3735-3746.

Behrentz E., Sabin L.D., Winer A.M., Fitz D.R., Pankratz D.V., Colome S.D., et al.
Relative importance of school bus-related microenvironments to children's pollutant
exposure. J Air Waste Manag Assoc 2005: 55(10): 1418-1430.

Borak J., and Sirianni G. Studies of self-pollution in diesel school buses: methodological
issues. J Occup Environ Hyg 2007: 4(9): 660-668.

Boscoe F. The science and art of geocoding. In: Rushton G., Armstrong, M.P., Gittler, J.,
Greene, B.R., Pavlik, C.E., West, M.M., and Zimmerman, D.L. (ed). Geocoding Health
Data: The Use of Geographic Codes in Cancer Prevention and Control, Research and
Practice. CRC Press: Boca Raton, 2008, pp 95-109.

Braback L., and 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.

Brauer M., Hoek G., Van Vliet P., Meliefste K., Fischer P.H., Wijga A., et al. Air
pollution from traffic and the development of respiratory infections and asthmatic and
allergic symptoms in children. Am J Respir Crit Care Med 2002: 166(8): 1092-1098.

120
Brauner E.V., Mortensen J., Moller P., Bernard A., Vinzents P., Wahlin P., et al. Effects
of ambient air particulate exposure on blood-gas barrier permeability and lung function.
Inhal Toxicol 2009: 21(1): 38-47.

Broeckaert F., and Bernard A. Clara cell secretory protein (CC16): characteristics and
perspectives as lung peripheral biomarker. Clin Exp Allergy 2000: 30(4): 469-475.

Broeckaert F., Clippe A., Knoops B., Hermans C., and Bernard A. Clara cell secretory
protein (CC16): features as a peripheral lung biomarker. Ann N Y Acad Sci 2000: 923: 68-
77.

Brunekreef B., Janssen N.A., de Hartog J., Harssema H., Knape M., and van Vliet P. Air
pollution from truck traffic and lung function in children living near motorways.
Epidemiology 1997: 8(3): 298-303.

Brzeznicki S., Jakubowski M., and Czerski B. Elimination of 1-hydroxypyrene after
human volunteer exposure to polycyclic aromatic hydrocarbons. Int Arch Occup Environ
Health 1997: 70(4): 257-260.

Cabrera M.E., Portzline G., Aach S., Condie C., Dorostkar P., and Mianulli M. Can
current minute ventilation rate adaptive pacemakers provide appropriate chronotropic
response in pediatric patients? Pacing Clin Electrophysiol 2002: 25(6): 907-914.

CalTrans Traffic and Vehicle Data Systems Unit, 2008.

Catellier D.J., Hannan P.J., Murray D.M., Addy C.L., Conway T.L., Yang S., et al.
Imputation of missing data when measuring physical activity by accelerometry. Med Sci
Sports Exerc 2005: 37(11 Suppl): S555-562.

Chen J.C., and Schwartz J. Neurobehavioral effects of ambient air pollution on cognitive
performance in US adults. Neurotoxicology 2009: 30(2): 231-239.

Chiabrando C., Valagussa A., Rivalta C., Durand T., Guy A., Zuccato E., et al.
Identification and measurement of endogenous beta-oxidation metabolites of 8-epi-
Prostaglandin F2alpha. J Biol Chem 1999: 274(3): 1313-1319.

Cooper A.R., Page A.S., Wheeler B.W., Griew P., Davis L., Hillsdon M., et al. Mapping
the walk to school using accelerometry combined with a global positioning system. Am J
Prev Med 2010: 38(2): 178-183.

Cooper A.R., Page A.S., Wheeler B.W., Hillsdon M., Griew P., and Jago R. Patterns of
GPS measured time outdoors after school and objective physical activity in English
children: the PEACH project. Int J Behav Nutr Phys Act 2010: 7: 31.

121
Crouter S.E., Schneider P.L., Karabulut M., and Bassett D.R., Jr. Validity of 10
electronic pedometers for measuring steps, distance, and energy cost. Med Sci Sports
Exerc 2003: 35(8): 1455-1460.

Crouter S.E., Schneider P.L., and Bassett D.R., Jr. Spring-levered versus piezo-electric
pedometer accuracy in overweight and obese adults. Med Sci Sports Exerc 2005: 37(10):
1673-1679.

Dockery D.W., Speizer F.E., Stram D.O., Ware J.H., Spengler J.D., and Ferris B.G., Jr.
Effects of inhalable particles on respiratory health of children. Am Rev Respir Dis 1989:
139(3): 587-594.

Donders A.R., van der Heijden G.J., Stijnen T., and Moons K.G. Review: a gentle
introduction to imputation of missing values. J Clin Epidemiol 2006: 59(10): 1087-1091.

Duncan M.J., Badland H.M., and Mummery W.K. Applying GPS to enhance
understanding of transport-related physical activity. J Sci Med Sport 2009: 12(5): 549-
556.

Elgethun K., Fenske R.A., Yost M.G., and Palcisko G.J. Time-location analysis for
exposure assessment studies of children using a novel global positioning system
instrument. Environ Health Perspect 2003: 111(1): 115-122.

Elgethun K., Yost M.G., Fitzpatrick C.T., Nyerges T.L., and Fenske R.A. Comparison of
global positioning system (GPS) tracking and parent-report diaries to characterize
children's time-location patterns. J Expo Sci Environ Epidemiol 2007: 17(2): 196-206.

English P., Neutra R., Scalf R., Sullivan M., Waller L., and Zhu L. Examining
associations between childhood asthma and traffic flow using a geographic information
system. Environ Health Perspect 1999: 107(9): 761-767.

Eston R.G., Rowlands A.V., and Ingledew D.K. Validity of heart rate, pedometry, and
accelerometry for predicting the energy cost of children's activities. J Appl Physiol 1998:
84(1): 362-371.

Fiala Z., Vyskocil A., Krajak V., Viau C., Ettlerova E., Bukac J., et al. Environmental
exposure of small children to polycyclic aromatic hydrocarbons. Int Arch Occup Environ
Health 2001: 74(6): 411-420.

Freedson P.S., Melanson E., and Sirard J. Calibration of the Computer Science and
Applications, Inc. accelerometer. Med Sci Sports Exerc 1998: 30(5): 777-781.

Freeman N.C.G., Lioy P.J., Pellizzari E.D.O., Zelon H., Thomas K., Clayton A., et al.
Responses to the Region 5 NHEXAS time/activity diary. 1999: 9(5): 414-426.

122
Frischer T., Studnicka M., Gartner C., Tauber E., Horak F., Veiter A., et al. Lung
function growth and ambient ozone: a three-year population study in school children. Am
J Respir Crit Care Med 1999: 160(2): 390-396.

Fruin S., Westerdahl D., Sax T., Sioutas C., and Fine P. Measurements and predictors of
on-road ultrafine particle concentrations and associated pollutants in Los Angeles.
Atmospheric Environment 2008: 42: 207-219.

Fruin S.A., Winer A.M., and Rodes C.E. Black carbon concentrations in California
vehicles and estimation of in-vehicle diesel exhaust particulate matter exposures.
Atmospheric Environment 2004: 38(25): 4123-4133.

Gauderman W.J., McConnell R., Gilliland F., London S., Thomas D., Avol E., et al.
Association between air pollution and lung function growth in southern California
children. Am J Respir Crit Care Med 2000: 162(4 Pt 1): 1383-1390.

Gauderman W.J., Gilliland G.F., Vora H., Avol E., Stram D., McConnell R., et al.
Association between air pollution and lung function growth in southern California
children: results from a second cohort. Am J Respir Crit Care Med 2002: 166(1): 76-84.

Gauderman W.J., Avol E., Lurmann F., Kuenzli N., Gilliland F., Peters J., et al.
Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology 2005:
16(6): 737-743.

Gauderman W.J., Vora H., McConnell R., Berhane K., Gilliland F., Thomas D., et al.
Effect of exposure to traffic on lung development from 10 to 18 years of age: a cohort
study. Lancet 2007: 369(9561): 571-577.

Gerharz L.E., Kruger A., and Klemm O. Applying indoor and outdoor modeling
techniques to estimate individual exposure to PM2.5 from personal GPS profiles and
diaries: a pilot study. Sci Total Environ 2009: 407(18): 5184-5193.

Gilbert N.L., Woodhouse S., Stieb D.M., and Brook J.R. Ambient nitrogen dioxide and
distance from a major highway. Sci Total Environ 2003: 312(1-3): 43-46.

Gilliland F.D., McConnell R., Peters J., and Gong H., Jr. A theoretical basis for
investigating ambient air pollution and children's respiratory health. Environ Health
Perspect 1999: 107 Suppl 3: 403-407.

Goldberg D. The USC WebGIS Geocoding Platform. GIS Research Laboratory Research
Report 2011: 11.

Gotschi T., Heinrich J., Sunyer J., and Kunzli N. Long-term effects of ambient air
pollution on lung function: a review. Epidemiology 2008: 19(5): 690-701.

123
Gotschi T., Sunyer J., Chinn S., de Marco R., Forsberg B., Gauderman J.W., et al. Air
pollution and lung function in the European Community Respiratory Health Survey. Int J
Epidemiol 2008: 37(6): 1349-1358.

Grainger J., Huang W., Patterson D.G., Jr., Turner W.E., Pirkle J., Caudill S.P., et al.
Reference range levels of polycyclic aromatic hydrocarbons in the US population by
measurement of urinary monohydroxy metabolites. Environ Res 2006: 100(3): 394-423.

Grow H.M., Saelens B.E., Kerr J., Durant N.H., Norman G.J., and Sallis J.F. Where are
youth active? Roles of proximity, active transport, and built environment. Med Sci Sports
Exerc 2008: 40(12): 2071-2079.

Heinrich J., and Slama R. Fine particles, a major threat to children. Int J Hyg Environ
Health 2007: 210(5): 617-622.

Helmersson J., and Basu S. F2-isoprostane excretion rate and diurnal variation in human
urine. Prostaglandins Leukot Essent Fatty Acids 1999: 61(3): 203-205.

Helmersson J., and Basu S. F(2)-isoprostane and prostaglandin F(2 alpha)metabolite
excretion rate and day to day variation in healthy humans. Prostaglandins Leukot Essent
Fatty Acids 2001: 65(2): 99-102.

Helmersson J., Larsson A., Vessby B., and Basu S. Active smoking and a history of
smoking are associated with enhanced prostaglandin F(2alpha), interleukin-6 and F2-
isoprostane formation in elderly men. Atherosclerosis 2005: 181(1): 201-207.

Hole D.J., Watt G.C., Davey-Smith G., Hart C.L., Gillis C.R., and Hawthorne V.M.
Impaired lung function and mortality risk in men and women: findings from the Renfrew
and Paisley prospective population study. BMJ 1996: 313(7059): 711-715; discussion
715-716.

Holguin F., Flores S., Ross Z., Cortez M., Molina M., Molina L., et al. Traffic-related
exposures, airway function, inflammation, and respiratory symptoms in children. Am J
Respir Crit Care Med 2007: 176(12): 1236-1242.

Horak F., Jr., Studnicka M., Gartner C., Spengler J.D., Tauber E., Urbanek R., et al.
Particulate matter and lung function growth in children: a 3-yr follow-up study in
Austrian schoolchildren. Eur Respir J 2002: 19(5): 838-845.

Hu S., Fruin S., Kozawa K., Mara S., Paulson S.E., and Winer A.M. A wide area of air
pollutant impact downwind of a freeway during pre-sunrise hours. Atmospheric
Environment 2009: 43(16): 2541-2549.

Huang W., Caudill S.P., Grainger J., Needham L.L., and Patterson D.G., Jr. Levels of 1-
hydroxypyrene and other monohydroxy polycyclic aromatic hydrocarbons in children: a
study based on U.S. reference range values. Toxicol Lett 2006: 163(1): 10-19.

124

Islam T., Gauderman W.J., Berhane K., McConnell R., Avol E., Peters J.M., et al.
Relationship between air pollution, lung function and asthma in adolescents. Thorax
2007: 62(11): 957-963.

Jedrychowski W., Flak E., and Mroz E. The adverse effect of low levels of ambient air
pollutants on lung function growth in preadolescent children. Environ Health Perspect
1999: 107(8): 669-674.

Jerrett M., Arain A., Kanaroglou P., Beckerman B., Potoglou D., Sahsuvaroglu T., et al.
A review and evaluation of intraurban air pollution exposure models. J Expo Anal
Environ Epidemiol 2005: 15(2): 185-204.

Jerrett M. Does traffic-related air pollution contribute to respiratory disease formation in
children? Eur Respir J 2007: 29(5): 825-826.

Jongeneelen F.J., van Leeuwen F.E., Oosterink S., Anzion R.B., van der Loop F., Bos
R.P., et al. Ambient and biological monitoring of cokeoven workers: determinants of the
internal dose of polycyclic aromatic hydrocarbons. Br J Ind Med 1990: 47(7): 454-461.

Kajekar R. Environmental factors and developmental outcomes in the lung. Pharmacol
Ther 2007: 114(2): 129-145.

Kang J.W., Cho S.H., Kim H., and Lee C.H. Correlation of urinary 1-hydroxypyrene and
2-naphthol with total suspended particulates in ambient air in municipal middle-school
students in Korea. Arch Environ Health 2002: 57(4): 377-382.

Kaviarasan S., Muniandy S., Qvist R., and Ismail I.S. F(2)-isoprostanes as novel
biomarkers for type 2 diabetes: a review. J Clin Biochem Nutr 2009: 45(1): 1-8.

Kim J.J., Smorodinsky S., Lipsett M., Singer B.C., Hodgson A.T., and Ostro B. Traffic-
related air pollution near busy roads: the East Bay Children's Respiratory Health Study.
Am J Respir Crit Care Med 2004: 170(5): 520-526.

Kim J.J., Huen K., Adams S., Smorodinsky S., Hoats A., Malig B., et al. Residential
traffic and children's respiratory health. Environ Health Perspect 2008: 116(9): 1274-
1279.

Klepeis N.E., Nelson W.C., Ott W.R., Robinson J.P., Tsang A.M., Switzer P., et al. The
National Human Activity Pattern Survey (NHAPS): a resource for assessing exposure to
environmental pollutants. J Expo Anal Environ Epidemiol 2001: 11(3): 231-252.

Knuiman M.W., James A.L., Divitini M.L., Ryan G., Bartholomew H.C., and Musk A.W.
Lung function, respiratory symptoms, and mortality: results from the Busselton Health
Study. Ann Epidemiol 1999: 9(5): 297-306.

125
Kozawa K.H., Fruin S.A., and Winer A.M. Near-road air pollution impacts of goods
movement in communities adjacent to the Ports of Los Angeles and Long Beach.
Atmospheric Environment 2009: 43(18): 2960-2970.

Kristensen P.L., Moller N.C., Korsholm L., Wedderkopp N., Andersen L.B., and Froberg
K. Tracking of objectively measured physical activity from childhood to adolescence: the
European youth heart study. Scand J Med Sci Sports 2008: 18(2): 171-178.

Kunzli N., McConnell R., Bates D., Bastain T., Hricko A., Lurmann F., et al. Breathless
in Los Angeles: the exhausting search for clean air. Am J Public Health 2003: 93(9):
1494-1499.

Kunzli N., Bridevaux P.O., Liu L.J., Garcia-Esteban R., Schindler C., Gerbase M.W., et
al. Traffic-related air pollution correlates with adult-onset asthma among never-smokers.
Thorax 2009: 64(8): 664-670.

Larsen R.K., 3rd, and Baker J.E. Source apportionment of polycyclic aromatic
hydrocarbons in the urban atmosphere: a comparison of three methods. Environ Sci
Technol 2003: 37(9): 1873-1881.

Laumbach R., Tong J., Zhang L., Ohman-Strickland P., Stern A., Fiedler N., et al.
Quantification of 1-aminopyrene in human urine after a controlled exposure to diesel
exhaust. J Environ Monit 2009: 11(1): 153-159.

Laumbach R.J., and Kipen H.M. Acute effects of motor vehicle traffic-related air
pollution exposures on measures of oxidative stress in human airways. Ann N Y Acad Sci
2010: 1203: 107-112.

Le Masurier G.C., Lee S.M., and Tudor-Locke C. Motion sensor accuracy under
controlled and free-living conditions. Med Sci Sports Exerc 2004: 36(5): 905-910.

Li N., Sioutas C., Cho A., Schmitz D., Misra C., Sempf J., et al. Ultrafine particulate
pollutants induce oxidative stress and mitochondrial damage. Environ Health Perspect
2003: 111(4): 455-460.

Li Z., Romanoff L.C., Trinidad D.A., Hussain N., Jones R.S., Porter E.N., et al.
Measurement of urinary monohydroxy polycyclic aromatic hydrocarbons using
automated liquid-liquid extraction and gas chromatography/isotope dilution high-
resolution mass spectrometry. Anal Chem 2006: 78(16): 5744-5751.

Li Z., Sandau C.D., Romanoff L.C., Caudill S.P., Sjodin A., Needham L.L., et al.
Concentration and profile of 22 urinary polycyclic aromatic hydrocarbon metabolites in
the US population. Environ Res 2008: 107(3): 320-331.

Li Z., Romanoff L.C., Lewin M.D., Porter E.N., Trinidad D.A., Needham L.L., et al.
Variability of urinary concentrations of polycyclic aromatic hydrocarbon metabolite in

126
general population and comparison of spot, first-morning, and 24-h void sampling. J
Expo Sci Environ Epidemiol 2009.

Long C.M., Suh H.H., Catalano P.J., and Koutrakis P. Using time- and size-resolved
particulate data to quantify indoor penetration and deposition behavior. Environ Sci
Technol 2001: 35(10): 2089-2099.

Lubans D.R., Morgan P.J., Callister R., and Collins C.E. The Relationship Between
Pedometer Step Counts and Estimated VO₂Max as Determined by a
Submaximal Fitness Test in Adolescents. Pediatric Exercise Science 2008: 20(3): 273-
284.

Lubans D.R., Morgan P.J., Collins C.E., Boreham C.A., and Callister R. The relationship
between heart rate intensity and pedometer step counts in adolescents. J Sports Sci 2009:
27(6): 591-597.

Maddison R., and Ni Mhurchu C. Global positioning system: a new opportunity in
physical activity measurement. Int J Behav Nutr Phys Act 2009: 6: 73.

Maddison R., Jiang Y., Vander Hoorn S., Exeter D., Mhurchu C.N., and Dorey E.
Describing patterns of physical activity in adolescents using global positioning systems
and accelerometry. Pediatr Exerc Sci 2010: 22(3): 392-407.

Marr L.C., Kirchstetter T.W., Harley R.A., Miguel A.H., Hering S.V., and Hammond
S.K. Characterization of Polycyclic Aromatic Hydrocarbons in Motor Vehicle Fuels and
Exhaust Emissions. Environmental Science & Technology 1999: 33(18): 3091-3099.

McArdle W.D., Katch F.I., Pechar G.S., Jacobson L., and Ruck S. Reliability and
interrelationships between maximal oxygen intake, physical work capacity and step-test
scores in college women. Med Sci Sports 1972: 4(4): 182-186.

McConnell R., Berhane K., Gilliland F., London S.J., Islam T., Gauderman W.J., et al.
Asthma in exercising children exposed to ozone: a cohort study. Lancet 2002: 359(9304):
386-391.

McConnell R., Berhane K., Yao L., Jerrett M., Lurmann F., Gilliland F., et al. Traffic,
susceptibility, and childhood asthma. Environ Health Perspect 2006: 114(5): 766-772.

McConnell R., Islam T., Shankardass K., Jerrett M., Lurmann F., Gilliland F., et al.
Childhood incident asthma and traffic-related air pollution at home and school. Environ
Health Perspect 2010: 118(7): 1021-1026.

McConnell R., Liu F., Wu J., Lurmann F., Peters J., and Berhane K. Asthma and school
commuting time. J Occup Environ Med 2010: 52(8): 827-828.

127
McKone T.E., Ryan P.B., and Ozkaynak H. Exposure information in environmental
health research: current opportunities and future directions for particulate matter, ozone,
and toxic air pollutants. J Expo Sci Environ Epidemiol 2009: 19(1): 30-44.

Melanson E.L., Knoll J.R., Bell M.L., Donahoo W.T., Hill J.O., Nysse L.J., et al.
Commercially available pedometers: considerations for accurate step counting. Prev Med
2004: 39(2): 361-368.

Merlo F., Andreassen A., Weston A., Pan C.F., Haugen A., Valerio F., et al. Urinary
excretion of 1-hydroxypyrene as a marker for exposure to urban air levels of polycyclic
aromatic hydrocarbons. Cancer Epidemiol Biomarkers Prev 1998: 7(2): 147-155.

Mitre N., Lanningham-Foster L., Foster R., and Levine J.A. Pedometer accuracy for
children: can we recommend them for our obese population? Pediatrics 2009: 123(1):
e127-131.

Molitor J., Molitor N.T., Jerrett M., McConnell R., Gauderman J., Berhane K., et al.
Bayesian modeling of air pollution health effects with missing exposure data. Am J
Epidemiol 2006: 164(1): 69-76.

Morabia A., Amstislavski P.N., Mirer F.E., Amstislavski T.M., Eisl H., Wolff M.S., et al.
Air pollution and activity during transportation by car, subway, and walking. Am J Prev
Med 2009: 37(1): 72-77.

Morrow J.D., Frei B., Longmire A.W., Gaziano J.M., Lynch S.M., Shyr Y., et al.
Increase in circulating products of lipid peroxidation (F2-isoprostanes) in smokers.
Smoking as a cause of oxidative damage. N Engl J Med 1995: 332(18): 1198-1203.

Morrow J.D. Quantification of isoprostanes as indices of oxidant stress and the risk of
atherosclerosis in humans. Arterioscler Thromb Vasc Biol 2005: 25(2): 279-286.

National. Cancer Institute [Internet]. Risk factor monitoring and methods: SAS programs
for analyzing NHANES 2003-2004 accelerometer data.[Cited 2009 August 1] Available
from: http://riskfactor.cancer.gov/tools/nhanes_pam. . Accessed.

Nicolai T., Carr D., Weiland S.K., Duhme H., von Ehrenstein O., Wagner C., et al. Urban
traffic and pollutant exposure related to respiratory outcomes and atopy in a large sample
of children. Eur Respir J 2003: 21(6): 956-963.

Nordling E., Berglind N., Melen E., Emenius G., Hallberg J., Nyberg F., et al. Traffic-
related air pollution and childhood respiratory symptoms, function and allergies.
Epidemiology 2008: 19(3): 401-408.

Nuckols J.R., Ward M.H., and Jarup L. Using geographic information systems for
exposure assessment in environmental epidemiology studies. Environ Health Perspect
2004: 112(9): 1007-1015.

128

Obata T., Tomaru K., Nagakura T., Izumi Y., and Kawamoto T. Smoking and oxidant
stress: assay of isoprostane in human urine by gas chromatography-mass spectrometry. J
Chromatogr B Biomed Sci Appl 2000: 746(1): 11-15.

Ozkaynak H., Palma T., Touma J.S., and Thurman J. Modeling population exposures to
outdoor sources of hazardous air pollutants. J Expos Sci Environ Epidemiol 2007: 18(1):
45-58.

Panther B.C., Hooper M.A., and Tapper N.J. A comparison of air particulate matter and
associated polycyclic aromatic hydrocarbons in some tropical and temperate urban
environments. Atmospheric Environment 1999: 33(24-25): 4087-4099.

Parker J.D., Akinbami L.J., and Woodruff T.J. Air pollution and childhood respiratory
allergies in the United States. Environ Health Perspect 2009: 117(1): 140-147.

Peters J.M., Avol E., Gauderman W.J., Linn W.S., Navidi W., London S.J., et al. A study
of twelve Southern California communities with differing levels and types of air
pollution. II. Effects on pulmonary function. Am J Respir Crit Care Med 1999: 159(3):
768-775.

Phillips M.L., Hall T.A., Esmen N.A., Lynch R., and Johnson D.L. Use of global
positioning system technology to track subject's location during environmental exposure
sampling. J Expo Anal Environ Epidemiol 2001: 11(3): 207-215.

Quigg R., Gray A., Reeder A.I., Holt A., and Waters D.L. Using accelerometers and GPS
units to identify the proportion of daily physical activity located in parks with
playgrounds in New Zealand children. Prev Med 2010: 50(5-6): 235-240.

Rainham D., Krewski D., McDowell I., Sawada M., and Liekens B. Development of a
wearable global positioning system for place and health research. Int J Health Geogr
2008: 7: 59.

Raizenne M., Neas L.M., Damokosh A.I., Dockery D.W., Spengler J.D., Koutrakis P., et
al. Health effects of acid aerosols on North American children: pulmonary function.
Environ Health Perspect 1996: 104(5): 506-514.

Rodriguez D.A., Brown A.L., and Troped P.J. Portable global positioning units to
complement accelerometry-based physical activity monitors. Med Sci Sports Exerc 2005:
37(11 Suppl): S572-581.

Rowlands A.V., and Eston R.G. Comparison of accelerometer and pedometer measures
of physical activity in boys and girls, ages 8-10 years. Res Q Exerc Sport 2005: 76(3):
251-257.

129
Ruchirawat M., Mahidol C., Tangjarukij C., Pui-ock S., Jensen O., Kampeerawipakorn
O., et al. Exposure to genotoxins present in ambient air in Bangkok, Thailand — particle
associated polycyclic aromatic hydrocarbons and biomarkers. Science of The Total
Environment 2002: 287(1-2): 121-132.

Sabin L.D., Behrentz E., Winer A.M., Jeong S., Fitz D.R., Pankratz D.V., et al.
Characterizing the range of children's air pollutant exposure during school bus commutes.
J Expo Anal Environ Epidemiol 2005: 15(5): 377-387.

Salam M.T., Islam T., and Gilliland F.D. Recent evidence for adverse effects of
residential proximity to traffic sources on asthma. Curr Opin Pulm Med 2008: 14(1): 3-8.

Samet J.M. Traffic, air pollution, and health. Inhal Toxicol 2007: 19(12): 1021-1027.

Schneider P.L., Crouter S.E., Lukajic O., and Bassett D.R., Jr. Accuracy and reliability of
10 pedometers for measuring steps over a 400-m walk. Med Sci Sports Exerc 2003:
35(10): 1779-1784.

Schneider P.L., Crouter S.E., and Bassett D.R. Pedometer measures of free-living
physical activity: comparison of 13 models. Med Sci Sports Exerc 2004: 36(2): 331-335.

Schwartz J. Lung function and chronic exposure to air pollution: a cross-sectional
analysis of NHANES II. Environ Res 1989: 50(2): 309-321.

Schwartz J., Coull B., Laden F., and Ryan L. The effect of dose and timing of dose on the
association between airborne particles and survival. Environ Health Perspect 2008:
116(1): 64-69.

Scruggs P.W., Beveridge S.K., Eisenman P.A., Watson D.L., Shultz B.B., and Ransdell
L.B. Quantifying physical activity via pedometry in elementary physical education. Med
Sci Sports Exerc 2003: 35(6): 1065-1071.

Sin D.D., Wu L., and Man S.F. The relationship between reduced lung function and
cardiovascular mortality: a population-based study and a systematic review of the
literature. Chest 2005: 127(6): 1952-1959.

Sly P.D., and Flack F. Susceptibility of children to environmental pollutants. Ann N Y
Acad Sci 2008: 1140: 163-183.

Sorensen M., Autrup H., Moller P., Hertel O., Jensen S.S., Vinzents P., et al. Linking
exposure to environmental pollutants with biological effects. Mutat Res 2003: 544(2-3):
255-271.

Soto-Martinez M., and Sly P. Relationship between environmental exposures in children
and adult lung disease: The case for outdoor exposures. Chron Respir Dis 2009.

130
Stone A.A., and Shiffman S. Capturing momentary, self-report data: a proposal for
reporting guidelines. Ann Behav Med 2002: 24(3): 236-243.

Stone M.R., Esliger D.W., and Tremblay M.S. Comparative validity assessment of five
activity monitors: does being a child matter? Pediatr Exerc Sci 2007: 19(3): 291-309.

Strath S.J., Bassett D.R., Jr., and Swartz A.M. Comparison of MTI accelerometer cut-
points for predicting time spent in physical activity. Int J Sports Med 2003: 24(4): 298-
303.

Timonen K.L., Hoek G., Heinrich J., Bernard A., Brunekreef B., de Hartog J., et al. Daily
variation in fine and ultrafine particulate air pollution and urinary concentrations of lung
Clara cell protein CC16. Occup Environ Med 2004: 61(11): 908-914.

Troiano R.P., Berrigan D., Dodd K.W., Masse L.C., Tilert T., and McDowell M. Physical
activity in the United States measured by accelerometer. Med Sci Sports Exerc 2008:
40(1): 181-188.

Trost S.G., Way R., and Okely A.D. Predictive validity of three ActiGraph energy
expenditure equations for children. Med Sci Sports Exerc 2006: 38(2): 380-387.

Tudor-Locke C., Williams J.E., Reis J.P., and Pluto D. Utility of pedometers for
assessing physical activity: convergent validity. Sports Med 2002: 32(12): 795-808.

Tudor-Locke C., Lee S.M., Morgan C.F., Beighle A., and Pangrazi R.P. Children's
pedometer-determined physical activity during the segmented school day. Med Sci Sports
Exerc 2006: 38(10): 1732-1738.

Tuntawiroon J., Mahidol C., Navasumrit P., Autrup H., and Ruchirawat M. Increased
health risk in Bangkok children exposed to polycyclic aromatic hydrocarbons from
traffic-related sources. Carcinogenesis 2006: 28(4): 816-822.

van Vliet P., Knape M., de Hartog J., Janssen N., Harssema H., and Brunekreef B. Motor
vehicle exhaust and chronic respiratory symptoms in children living near freeways.
Environ Res 1997: 74(2): 122-132.

van Wijnen J.H., Slob R., Jongmans-Liedekerken G., van de Weerdt R.H., and
Woudenberg F. Exposure to polycyclic aromatic hydrocarbons among Dutch children.
Environ Health Perspect 1996: 104(5): 530-534.

Venn A., Lewis S., Cooper M., Hubbard R., Hill I., Boddy R., et al. Local road traffic
activity and the prevalence, severity, and persistence of wheeze in school children:
combined cross sectional and longitudinal study. Occup Environ Med 2000: 57(3): 152-
158.

131
Waldron G., Pottle B., and Dod J. Asthma and the motorways--one District's experience.
J Public Health Med 1995: 17(1): 85-89.

Ware J.H., Ferris B.G., Jr., Dockery D.W., Spengler J.D., Stram D.O., and Speizer F.E.
Effects of ambient sulfur oxides and suspended particles on respiratory health of
preadolescent children. Am Rev Respir Dis 1986: 133(5): 834-842.

Westerdahl D., Fruin S., Sax T., Fine P.M., and Sioutas C. Mobile platform
measurements of ultrafine particles and associated pollutant concentrations on freeways
and residential streets in Los Angeles. Atmospheric Environment 2005: 39(20): 3597-
3610.

Wiehe S.E., Carroll A.E., Liu G.C., Haberkorn K.L., Hoch S.C., Wilson J.S., et al. Using
GPS-enabled cell phones to track the travel patterns of adolescents. Int J Health Geogr
2008: 7: 22.

Wjst M., Reitmeir P., Dold S., Wulff A., Nicolai T., von Loeffelholz-Colberg E.F., et al.
Road traffic and adverse effects on respiratory health in children. BMJ 1993: 307(6904):
596-600.

Wu J., Ren C., Delfino R.J., Chung J., Wilhelm M., and Ritz B. 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): 1773-1779.

Wu J., Jiang C., Liu Z., Houston D., Jaimes G., and McConnell R. Performances of
different global positioning system devices for time-location tracking in air pollution
epidemiological studies. Environ Health Insights 2010: 4: 93-108.

Zandbergen P.A. Accuracy of iPhone Locations: A Comparison of Assisted GPS, WiFi
and Cellular Positioning. Transactions in GIS 2009: 13: 5-25.

Zandbergen P.A., and Barbeau S.J. Positional Accuracy of Assisted GPS Data from
High-Sensitivity GPS-enabled Mobile Phones. The Journal of Navigation 2011: 64(03):
381-399.

Zhang W., Xu D., Zhuang G., Ding C., Wang G., Chang J., et al. A pilot study on using
urinary 1-hydroxypyrene biomarker for exposure to PAHs in Beijing. Environ Monit
Assess 2007: 131(1-3): 387-394.

Zhou Y., and Levy J.I. Factors influencing the spatial extent of mobile source air
pollution impacts: a meta-analysis. BMC Public Health 2007: 7: 89.

Zhu Y., Hinds W.C., Kim S., and Sioutas C. Concentration and size distribution of
ultrafine particles near a major highway. J Air Waste Manag Assoc 2002: 52(9): 1032-
1042.

132
Zhu Y., Kuhn T., Mayo P., and Hinds W.C. Comparison of daytime and nighttime
concentration profiles and size distributions of ultrafine particles near a major highway.
Environ Sci Technol 2006: 40(8): 2531-2536.

Zmirou D., Gauvin S., Pin I., Momas I., Sahraoui F., Just J., et al. Traffic related air
pollution and incidence of childhood asthma: results of the Vesta case-control study. J
Epidemiol Community Health 2004: 58(1): 18-23.

Zuurbier M., Hoek G., van den Hazel P., and Brunekreef B. Minute ventilation of
cyclists, car and bus passengers: an experimental study. Environ Health 2009: 8: 48.

Zuurbier M., Hoek G., Oldenwening M., Meliefste K., Krop E., van den Hazel P., et al.
In-Traffic Air Pollution Exposure and CC16, Blood Coagulation, and Inflammation
Markers in Healthy Adults. Environ Health Perspect 2011.

Zwirner-Baier I., and Neumann H.-G. Polycyclic nitroarenes (nitro-PAHs) as biomarkers
of exposure to diesel exhaust. Mutation Research/Genetic Toxicology and Environmental
Mutagenesis 1999: 441(1): 135-144.

Abstract (if available)

Abstract Objective: There is emerging evidence that local traffic-related pollution (TRP) has adverse health effects that are independent of regional pollution effects. Current methods to assess TRP exposure have limitations that may account for uncertainty and inconsistency in the observed traffic-related health effects. New methods are needed to assess TRP exposure in different microenvironments and to assess early biological effects of these exposures in population based studies of air pollution. I evaluated new tools that can be used to assess time and activity and also evaluated the impact on novel biological markers of exposure to TRP and biomarkers of early biological effects. ❧ Specifically the objectives of my research were: 1) to evaluate the accuracy of a time-resolved step counter in children and the duration of consecutive zero step count minutes that indicated non-wear time periods 2) evaluate the ability of a GPS data logger to assess location of children during usual activity and 3) to assess feasibility of collecting urine samples at school and to evaluate urinary biomarkers of exposure and early effect of traffic-related air pollution. The overall goal of my research was to identify methods that could greatly improve exposure assessments of TRP by providing an integrated metric of dose that could be used to strengthen studies investigating the relationship between TRP exposure and health affects. ❧ Methods: To evaluate these tools, I conducted two studies. In the first study, a new time-resolved step counter, the SportBrain, was evaluated for accuracy. Seventeen children walked or ran on a treadmill at 2, 3, 4 and 5 miles/hour and walked around a track while wearing the SportBrain and Digiwalker SW-701 pedometers. We compared percent error in step counts for the two pedometers relative to observer counts. A sub-sample wore an accelerometer and SportBrain pedometer during up to 5 days of usual activity. In the second study, up to five urine samples per child were collected before and after a school field trip with bus travel on a busy highway from fifteen 9-10 year olds recruited from two classrooms in a low pollution region of Los Angeles. Samples were analyzed for biomarkers of exposure (ten polycyclic aromatic hydrocarbon metabolites and 1-aminopyrene) and of effect (Clara cell protein 16 and 8-iso-PGF2

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Ambient air pollution and lung function in children

PDF

Early life air pollution exposure, gestational diabetes mellitus, and autism spectrum disorder

PDF

Disparities in exposure to traffic-related pollution sources by self-identified and ancestral Hispanic descent in participants of the USC Children’s Health Study

PDF

Examining exposure to extreme heat and air pollution and its effects on all-cause, cardiovascular, and respiratory mortality in California: effect modification by the social deprivation index

PDF

Airway inflammation and respiratory health in the Southern California children's health study

PDF

Early life risk factors for childhood asthma

PDF

Uncertainty quantification in extreme gradient boosting with application to environmental epidemiology

PDF

Personal exposure to particulate matter PM2.5 sources during pregnancy and birthweight

PDF

The role of pesticide exposure in breast cancer

PDF

Effects of stress and the social environment on childhood asthma in the children' s health study

PDF

A cohort study of air-pollution and childhood obesity incidence

PDF

Effects of near-roadway air pollution exposure on obesity, obesity-related behavior, and neurobehavioral deficits during peripuberty

PDF

Sensor-based mobile health approaches for personal air pollution and pediatric asthma studies

PDF

Associations between ambient air pollution and hypertensive disorders of pregnancy

PDF

Genetic epidemiological approaches in the study of risk factors for hematologic malignancies

PDF

A pilot study of a global approach to assessing air pollution exposure in port communities: passive air monitoring of nitrogen dioxide concentrations

PDF

Prenatal exposure to organophosphate esters and early neurodevelopment

PDF

Prenatal air pollution exposure, newborn DNA methylation, and childhood respiratory health

PDF

The relationship between per- and polyfluoroalkyl substances, microRNA, and non-alcoholic fatty liver disease

PDF

Air pollution, smoking, and multigenerational DNA methylation Signatures: a study of two southern California cohorts

Asset Metadata

Creator Dueker, Donna Carmel (author)

Core Title Evaluation of new methods for estimating exposure to traffic-related pollution and early health effects for large population epidemiological studies

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Epidemiology

Publication Date 04/30/2012

Defense Date 01/24/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Air pollution,exposure assessment,GPS,OAI-PMH Harvest,pedometer,physical activity,time-activity,urinary biomarker

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor McConnell, Rob (committee chair), Cockburn, Myles (committee member), Fruin, Scott (committee member), Gauderman, William James (committee member), Wilson, John P. (committee member)

Creator Email donna.dueker@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-17036

Unique identifier UC11290099

Identifier usctheses-c3-17036 (legacy record id)

Legacy Identifier etd-DuekerDonn-676.pdf

Dmrecord 17036

Document Type Dissertation

Rights Dueker, Donna Carmel

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA