 

 

 

¢
; ' 9'
- mum;

" I I} ”7‘: -'
. “’i‘yéﬂs: "1,1
g: fx‘vf :3 A

“a
‘Ile -
m
u
u.

a.»
a

.

.m
:r'
'r

 

- ”a
u.
“my“.

3" ‘

.,
...
“r.

- -.::~

x:

0-»
4-1.

5::
f:
'3

v P: ’5? . A t J
.‘fingéggak 52,02}.
‘l ,"jiﬂﬂgts‘ $5,: .

4353 {Mic ’

c1?
..
:H- .
. <
“-
4‘1"“

 

:y ,
, .
..

.:,
M. K
.L-

.

W55“?

1/

(DB)

 

.LlBRAFiY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

The Air in There — Should We Care?
An Investigation into the Relationship between Indoor Air
Quality and Tracheal Mucus in Thoroughbred Racehorses

presented by

Melissa Millerick-May

has been accepted towards fulﬁllment
of the requirements for the

PhD degree in Comparative Medicine and
lgtegrative Bioﬁgy

 

. >'
i 1/ I ' .
k .A ‘ {‘4’ L

Major Professor’s Sign ture

Wars HQ; 1002

Date

MSU is an aﬁnnative-action, equal-opportunity employer

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K /Pro,lAcc&Pres/ClRC/DateDue indd

 

THE AIR IN THERE — SHOULD WE CARE?
AN INVESTIGATION INTO THE RELATIONSHIP BETWEEN INDOOR
AIR QUALITY AND TRACHEAL MUCUS IN THOROUGHBRED
RACEHORSES

By

Melissa Millerick-May

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Comparative Medicine and Integrative Biology

2008

ABSTRACT
THE AIR IN THERE — SHOULD WE CARE?
AN INVESTIGATION INTO THE RELATIONSHIP BETWEEN INDOOR AIR
QUALITY AND TRACHEAL MUCUS IN THOROUGHBRED RACEHORSES
By

Melissa Millerick-May

Tracheal mucus accumulation is a Signiﬁcant problem in racehorses, as its
presence in sufﬁcient quantities negatively affects racing performance. The cause is
likely multifactorial in nature, but the stable environment has been implicated as one risk
factor, as it has been shown that some healthy horses brought into the stable from pasture
experience airway inﬂammation. The purpose of my work was to describe the range of
concentrations of airborne particulate matter within the stable, and factors that inﬂuence
its presence. I hypothesized that tracheal mucus accumulations result, at least in part,
from exposure to particulate matter (PM) in size ranges (less than 10 and less than 2.5
microns in diameter) determined to be important in human respiratory disease. I
hypothesized that area particle concentration and personal particulate exposure varies with
season, stable, stall location within a stable, and activity level, and correlates in an exposure-
dependent manner with established measures of inﬂammatory airway disease (IAD). Using
direct-reading instruments that measure particulate concentrations, I demonstrated that
dusts present in the breathing zone of horses during routine stable activities (cleaning,
feeding, and riding) contained concentrations of PMIO and PM2.5 that would be considered
unacceptable in industrial environments, but that consistent use of easy to implement dust-

suppression techniques more than adequately reduced exposures. Investigations in stables at

a Midwestern Thoroughbred racetrack, again using direct-reading instruments, revealed
signiﬁcant variations in area PMlO and PM2.5 concentrations by season (month), time of
day, stable, and certain stalls within stables. PM was lowest during months following
periods of wet weather and highest aﬁer periods of hot dry weather. Early morning PM
concentrations, when feeding, cleaning, and training activities occur, were highest in
comparison to mid-day and aﬁemoon concentrations when activity was minimal. Enclosed
stables with few sources of natural dilution ventilation experienced the highest
concentrations of PM as compared to the open-aired stables designed for maximum
ventilation. Stalls located in either high-trafﬁc areas or areas devoid of natural dilution
ventilation contained PM concentrations signiﬁcantly higher than adjacent stalls. The
presence of tracheal mucus was signiﬁcantly associated with area PM concentrations, and
overall prevalence was highest during months and in stables and stalls where area PM
concentrations were the greatest. Personal particulate exposure monitoring (PMlO) in
horses both with and without tracheal mucus suggests that personal idiosyncratic activities
did not inﬂuence personal exposures to the same degree as area PM; furthermore, overall
prevalence of tracheal mucus shows a striking association to personal PM exposures that
likely reﬂects ambient concentrations (e. g. monthly variations in PM). If this is true, then
implementation of dust suppression techniques within stables should signiﬁcantly reduce the
overall prevalence of Si gniﬁcant quantities of tracheal mucus. Percent neutrophils present in
tracheal wash samples did not correlate with the presence of tracheal mucus, and I therefore
suggest that its use in diagnosing IAD be reconsidered, as endoscopic examination scoring
of tracheal mucus is easy to perform, and likely the most accurate predictor of airway

disease sufﬁcient to affect performance.

For Star - The inspiration behind these investigations

iv

ACKNOWLEDGMENTS

First and foremost, I would like to thank my mentor Dr. Ed Robinson for his
invaluable support and guidance over the past 4.5 years, allowing me to combine my
previous educational and work experience with a personal interest. I would also like to
thank the members of my guidance committee, Drs. Frederik Derksen, Jack Harkema,
Susan Holcombe, and Vilma Yuzbasiyan-Gurkan.

The success of this project would not be possible without the help and support of
Heather DeFeijter-Rupp, Cathy Bemey, Sue Eberhardt, Mike Casemeir, Dr. Derksen, and
Dr. Holcombe. There were countless hours Spent by these individuals at the racetrack
working with the horses and in the laboratory both at the racetrack and at MSU, and
without their knowledge, dedication, and good humor, we would not have been able to
conduct a ﬁeld investigation of this magnitude. I would also like to thank Dr. Wilﬁied
Karmaus for his assistance with the data analysis, and his patience while I was attempting
to learn how to use SAS. Particular thanks go to Victoria Hoelzer-Maddox for her
patience and support while trying to navigate this degree program, and with the many
(oﬂen last-minute) drafts of this dissertation.

I would like to thank the management, trainers, and veterinarians at Thistledown
racetrack for their dedication to this research project. Without continued access to their
stables and horses, this project would not have been possible. I would also like to thank
the Grayson-Jockey Club Research Foundation and the Matilda Wilson Fund for

supporting this project.

 

I would like to extend a special thank you to my Sleepy Hollow family for
allowing me to ‘practice’ my sampling skills, and listening to countless hours of babble
regarding ventilation status and airborne dust. Nikki, Nancy, Debbie, Sarah M, Lauren S,
and Alex — thank you so much for taking such good care of my two 4-legged children
(a.k.a. my sanity).

Finally I would like to thank my family. Jamie, you have been so supportive of
my educational endeavors over the years that I do not know how to begin to thank you.
Justin, I want to thank you for your patience and understanding during the times when I
have not been the most ‘fun’ mommy to be around. I want to thank my mother and father
for their continued encouragement and support of my dreams. I also would like to thank
all of the grandparents and aunts for helping Jamie and me to juggle the day-to-day
school/weekend schedules that made this effort possible. Last, but not least, Jamie and I
would like to thank Jim and Lois Paradise for being such an integral part of Justin’s daily

life —— our appreciation cannot be put into words.

vi

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................................... x
LIST OF FIGURES ..................................................................................................... xi
LIST OF TERMS ......................................................................................................... xiv
INTRODUCTION ......................................................................................................... 1
Literature Review ................................................................................................ 4
Deﬁnition of IAD ........................................................................................... 4

Risk Factors for IAD ...................................................................................... 4

Mucus, Airway Inﬂammation, and Particulates ............................................ 5

Particle Composition ..................................................................................... 6
Co-exposures ................................................................................................. 6

Particle Size ................................................................................................... 7
Consequence of Particulate Exposures ......................................................... 8
Particulate Monitoring in Stables ................................................................. 9
Justiﬁcation of Our Studies in Relation to the Current Literature ................ 9
References ............................................................................................................. 12
CHAPTER 1: Indoor Air Quality in a Boarding Stable ................................................. 18
Summary .............................................................................................................. 1 8
Reasons for Performing Study ...................................................................... 18

Objectives ..................................................................................................... 1 8

Methods ........................................................................................................ l 8

Results ........................................................................................................... 1 9
Conclusions .................................................................................................. 19

Potential Relevance ...................................................................................... 20
Introduction ......................................................................................................... 20
Materials and Methods ....................................................................................... 22

The Boarding Stable ..................................................................................... 22
Particulate Measurements ............................................................................ 22

Feed .............................................................................................................. 23

Stable Activities ............................................................................................ 23

Riding ............................................................................................................ 24

Bulk Samples ................................................................................................. 24

Gas Sampling ................................................................................................ 25

Statistical Analysis ........................................................................................ 26

Results ................................................................................................................... 26
Feeds ............................................................................................................. 26

Stable Activities ............................................................................................ 27

Riding ............................................................................................................ 27

Bulk Samples ................................................................................................. 28

Gases ............................................................................................................ 28

vii

Discussion ............................................................................................................. 29
References .............................................................................................................. 42

CHAPTER 2: Evaluation of Factors Affecting the Concentration of Particulate

Matter (PMlO and PM2.5) in Stables at an American Thoroughbred Racetrack ........... 44
Summary .............................................................................................................. 44
Reasons for Performing Study ...................................................................... 44
Objectives ..................................................................................................... 44
Methods ........................................................................................................ 44
Results ........................................................................................................... 45
Conclusions .................................................................................................. 45
Potential Relevance ...................................................................................... 46
Introduction ......................................................................................................... 46
Materials and Methods ....................................................................................... 49
Stables ........................................................................................................... 49
Measurement of Particle Concentration and Numbers ................................ 50
Experimental Design .................................................................................... 5 1

Data Analysis ................................................................................................ 51
Results ................................................................................................................... 52
Effect of Month and Time of Day .................................................................. 53
Eﬂect of Stable .............................................................................................. 53
Effect of Stall Location ................................................................................. 54
Discussion ............................................................................................................. 54
References .............................................................................................................. 68

CHAPTER 3: Horse-Environment Interaction: The Relationship between
Environmental Particulate Matter and Airway Inﬂammation in Thoroughbred

Racehorses ...................................................................................................................... 72
Summary ............................................................................................................... 72
Reason for Performing Study ........................................................................ 72
Objectives ...................................................................................................... 72
Methods ......................................................................................................... 72
Conclusions ................................................................................................... 73
Potential Relevance ....................................................................................... 73
Introduction ......................................................................................................... 74
Materials and Methods ....................................................................................... 76
Experimental Design .................................................................................... 77
Endoscopic Examination .............................................................................. 77
Tracheal Lavage ........................................................................................... 78
Stables ........................................................................................................... 79
Measurement of Particle Concentration ...................................................... 80
Data Analysis ................................................................................................ 81
Bivariate Analysis ................................................................................ 81
Multivariate Analysis .......................................................................... 82

Results ................................................................................................................... 83
Study Population ........................................................................................... 83

viii

Tracheal Mucus ............................................................................................ 84

Multivariate Analysis .................................................................................... 84
M821 ................................................................................................... 84

M822 ................................................................................................... 85
Inﬂammatory Cells ....................................................................................... 85
Discussion ............................................................................................................. 86
References ........................................................................................................... l 01

CHAPTER 4: Airborne Particulates (PMlO) and Tracheal Mucus: A Case-Control

Study at an American Thoroughbred Racetrack .......................................................... 104
Summary ............................................................................................................ 104
Reasons for Performing Study ................................................................... 104
Objectives ................................................................................................... 1 04
Methods ...................................................................................................... 104
Results ......................................................................................................... 1 05
Conclusions ................................................................................................ 105
Potential Relevance .................................................................................... 106
Introduction ....................................................................................................... 107
Materials and Methods ..................................................................................... 109
Experimental Design .................................................................................. 1 O9
Endoscopic Examination ............................................................................ 1 10
Tracheal Lavage ......................................................................................... 1 10
Selection of Cases and Controls ................................................................. 111
Measurement of Breathing Zone Concentrations ....................................... l 12
Data Analysis .............................................................................................. 1 13
Results ................................................................................................................. 115
Validation of Particulate Measurements .................................................... 116
Mucus Scores and Particle Concentrations ............................................... 116
Inﬂammatory Cell Counts .......................................................................... 117
Variation in PM Exposures between Matched Cases (M522) and
Controls (MS =0) ........................................................................................ 1 17
Discussion .......................................................................................................... 1 18
References .......................................................................................................... 133
CONCLUSIONS ......................................................................................................... 136
Air Quality in a Boarding Stable ...................................................................... 136
Indoor Air Quality and Airway Inﬂammation in Thoroughbreds ............... 138
Case-Control Study ........................................................................................... 141
Practical Implications from the Investigation ................................................. 142
Future Investigations ......................................................................................... 142

ix

Table 1-1.

Table 1-2.

Table 2-1.

Table 2-2.

Table 3-1.

Table 3-2

Table 3-3.

Table 4-1.

Table 4-2.

Table 4-3.

Table 4-4.

LIST OF TABLES

Metal (mg/kg) and crystalline silica (% of total) concentration in bulk

samples taken from the arena footing and wall ledges. .............................

Summer and winter ammonia concentrations (ppm) in each of seven

stalls. ..........................................................................................................

Summary of number of stalls and total number of observations within
each stable, including the median, 5th, and 95th percentile values for

PM10 and PM2.5 (average, minimum, and maximum). ...........................

Number of stalls within each stable that exceed the 50th and 75th

percentile for PMlO and PM2.5 (Average, Minimum, and Maximum)...

Summary of number of examinations performed per stable, age,
gender, and assigned mucus score, as well as the prevalence of M821

and M822. .................................................................................................

Result of bivariate analysis of factors affecting mucus scores and

inﬂammatory cells. P-values <0.05 have been listed. ..............................

Concentrations of particulate matter (PM) within each quartile utilized
for GEE and mixed model analysis. Average (avg), maximum (max),
and minimum (min) values are presented, with numeric notations
indicating time of day sampling occurred (1 - morning, 2 - midday, 3 —

evening). ....................................................................................................

Summary of number of examinations performed per stable, age,
gender, and assigned mucus score, as well as the prevalence of MS_>_2 ..

Summary of age and gender of cases and controls included in the
study .........................................................................................................

Summary of numbers of cases and controls selected from each stable...

Description of the average, minimum, maximum, and time-weighted
average concentrations of PM] 0 (mg/m3) during each of three
sampling periods, as well as the number of 1-second samples
exceeding 0.569, 1, and 10 mg/m3 as experienced by cases and
controls ....................................................................................................

.34

.35

.60

.61

.92

.93

.95

123

124

125

126

LIST OF FIGURES

Figure 1-1. Diagram of the boarding stable showing the location of the stalls in
which ammonia concentrations were measured. ....................................... 36

Figure 1-2. Breathing zone concentration (mg/m3) of PM10 and PM2.5 during
consumption of various dry feedstuffs. Median values and range are
shown for both average and peak particle concentration .......................... 37

Figure 1-3. Breathing zone concentration (mg/m3) of PM10 and PM2.5 during
routine stable activities. Median values and range are shown for both
average and peak particle concentration. .................................................. 38

Figure 1-4. PM10 (top) and PM2.5 (bottom) concentrations (mg/m3) in the riding
arena during a single riding event. Aerosol photometer samples were
recorded every second for the duration of the activity .............................. 39

Figure 1-5. Effect of moisture content on breathing zone concentration (mg/m3)
of PM10 (left) and PM2.5 (right) during arena grading (upper) and
riding (lower). Values are maximum, minimum and mean. ..................... 40

Figure 1-6. Distribution of particle diameters (micron) in bulk samples taken from
the arena footing (top) and wall ledges (bottom). ..................................... 41

Figure 2-1. Effect of month on mean particle concentrations (PM10 and PM2.5)
in mg/m3. Light gray bars denote July, dark gray bars denote
September, and black-striped bars indicate November measurement
values. Error bars indicate the 25th and 75th percentile. An * denotes
signiﬁcant difference from referent month (July). .................................. 62

Figure 2-2. Effect of time of day on mean particle concentrations (PM10 and
PM2.5) in mg/m3. Light gray bars denote early morning, dark gray
bars denote mid-day, and black-striped bars indicate late aﬁemoon
measurement values. Late aﬂemoon measurements were set as the
referent group. Error bars indicate the 25th and 75th percentile. An
* denotes signiﬁcant difference from referent time of day (late
afternoon). . .............................................................................................. 63

Figure 2-3. Effect of stable on mean particle concentrations (PM10 and PM2.5) in
mg/m3. Light gray bars denote Stable 1, dark gray bars denote Stable
2, and black-striped bars indicate Stable 3. Error bars indicate the
25th and 75th percentile. An * denotes Signiﬁcant difference from
referent stable (stable 3). ........................................................................... 64

xi

Figure 2-4.

Figure 2-5.

Figure 2-6.

Figure 3-1.

Figure 3-2.

Figure 3-3.

Figure 3-4.

Effect of stable/month interaction on mean particle concentrations

(PM10 and PM2.5) in mg/m3. Light gray bars denote July, dark gray

bars denote September, and black-striped bars indicate November
measurement periods. Stable 3 and the month of July were used as

the referent group. Error bars indicate the 25th and 75th percentile.

An * denotes signiﬁcant difference from referent groups. ..................... 65

Map showing the locations of stalls within each stable that exceed the

50th and 75th percentiles for PM10 Average values. Note that

almost all of the stalls in Stable 2 exceed the 50th percentile, while

less than half of all stalls in both Stables 1&3 exceed the 50th

percentile. Also, Stable 2 contains the majority of the stalls that

exceed the 75th percentile while Stable 1 does not contain any. The
‘North’ notation located adjacent to each diagram denotes stable
orientation. .............................................................................................. 66

Map showing the locations of stalls within each stable that exceed the

50th and 75th percentiles for PM2.5 Average values. Note that

almost all of the stalls in Stable 2 exceed the 50th percentile, while

less than half of all stalls in both Stables 1&3 exceed the 50th

percentile. Also, Stable 2 contains the majority of the stalls that

exceed the 75th percentile while Stable 1 does not contain any. The
‘North’ notation located adjacent to each diagram denotes stable
orientation. .............................................................................................. 67

Effect of particle concentrations on overall prevalence of M821

(black line) by stable. Gray bars indicate average PM10

concentrations (mg/m3), and black bars indicate average PM2.5
concentrations (mg/m3). .......................................................................... 96

Effect of particle concentrations on overall prevalence of M821

(black line) by sampling month. Gray bars indicate average PM10
concentrations (mg/m3), and black bars indicate average PM2.5
concentrations (mg/m3). .......................................................................... 97

Effect of particle concentrations on overall prevalence of M832

(black line) by stable. Light gray bars indicate average PM10
concentrations (mg/m3), and dark gray bars indicate average PM2.5
concentrations (mg/m3). ......................................................................... 98

Effect of particle concentrations on percent neutrophils (solid line)

and lymphocytes (dashed line) by month. Dark gray bars indicate

average PM10 concentrations (mg/m3), and light gray bars indicate

average PM2.5 concentrations (mg/m3). ................................................ 99

xii

Figure 3-5.

Figure 4-1.

Figure 4-2.

Figure 4-3.

Figure 4-4.

Figure 4-5.

Figure 4-6.

Effect of particle concentrations on numbers of neutrophils (solid
line) and lymphocytes (dashed line) by month. Dark gray bars
indicate average PM10 concentrations (mg/m3), and light gray bars

indicate average PM2.5 concentrations (mg/m3). .................................

Example of 1 second tracings of PM10 concentrations of a selected

C386. ........................................................................................................

Picture of personal monitor attached to a surcingle, with sample
tubing along the length of the horse’s neck, and ending near the

nostril. ....................................................................................................

Location of sampling tube close to the nostril. .....................................

Percent of PM10 concentrations (grey bars) that fall into the ‘high’
calculated time-weighted average (TWA) category by sampling
month, and percent neutrophils (black line) by month plotted along

the same axis. .........................................................................................

Percent of overnight PM10 concentrations (grey bars) that fall into
the ‘high’ category by sampling month, and percent of horses with

MS_>_2 (black line) by month plotted along the same axis ......................

Percent of evening PM10 concentrations (grey bars) that fall into the
‘high’ category by sampling month, and percent of horses with

MS_>_2 (black line) by month plotted along the same axis. ....................

xiii

100

127

128

129

130

131

132

ACGIH

ANOVA

BAL

CAPS

COPD

GEE

LIST OF TERMS

American Conference of Governmental Industrial Hygienists
Analysis of variation

Bronchoalveolar lavage

Concentrated ambient particles

Chronic Obstructive Pulmonary Disease
General estimating equation

Inﬂammatory airway disease

Mucus score

Mucus score equal to zero

Mucus score greater than or equal to 1

Mucus score greater than or equal to 2
Milligram per kilogram

Milligram per cubic meter of air

National Ambient Air Quality Standard
Occupational exposure limits

Occupational Safety and Health Administration
Particulate matter

Particles not otherwise classiﬁed

Recurrent airway obstruction

Time-weighted average

xiv

INTRODUCTION

Airway inﬂammation and associated tracheobronchial mucus accumulation in
racehorses has been shown to reduce racing performance (Holcombe et al. 2006;
MacNamara et al. 1990b), a signiﬁcant problem for trainers and owners, as the overall
prevalence has been reported to be 13% (Wood et al. 2005a). Mucus accumulations are
generally a consequence of airway inﬂammation, the cause of which is most likely
multifactorial in nature. Risk factors of interest in racehorses are infection and exposure
to particulate matter (dust). Many of the horses in racing stables found to have mild to
moderate airway disease do not culture positive for bacteria (Christley et al. 2001a;
Wood et al. 1993; Wood et al. 2005a, c), therefore our focus has turned to the evaluation

of exposures to particulate matter.

It is likely that multiple factors acting alone or synergistically cause the severity of
inﬂammation that is responsible for mucus production and accumulation. Of particular
relevance to the development of airway inﬂammation in racehorses are the observations in
humans that increases in ambient ‘coarse’ (<10 micron in diameter) and ‘ﬁne’ (<2.5 micron
in diameter) particulates are associated with worsening of disease (mucus production,
cough, wheeze) in patients with asthma and COPD, and increased upper respiratory
infection visits to hospitals in individuals without pre-existing disease (Pope 1991a; Pope III

2000 ; Pope III et al. 2002; Schwartz 2004).

Sources of particulate matter in the racehorse environment include feed and
bedding, ﬂooring materials, track footing, road dust, vehicle exhaust, and emissions from
adjacent industrial operations. Particle sizes of interest in this study are classiﬁed as those
less than 10 (PM10) and less than 2.5 (PM2.5) microns in diameter. A small percentage of
PM10 have the potential to reach the lower airways but the majority of these particles are
deposited in the central airways such as the trachea (ACGIH 2005a). The smaller particles

represented by PM2.5 penetrate deep within the airways.

Our long-range goal is to understand the causes of inﬂammatory airway disease
(IAD) in racehorses. My investigations focused on the role of airborne particulates as a
cause of airway inﬂammation, and resultant mucus production. In order to test the
hypothesis that area particle concentration and personal particulate exposure varies with
stable, stall location within a stable, and activity level, and correlates in an exposure-
dependent manner with established measures of airway inﬂammation (mucus score, and

inﬂammatory cell counts) in racehorses with and without IAD, I did the following:

— Evaluated the range of concentrations of PM10 and PM2.5 in the breathing zone of
horses during routine stable activities using real-time particulate monitors (chapter 1)
— Investigated factors that inﬂuence area (background) concentrations of PM10 and

PM2.5 in 3 stables of differing design and management at a midwestem Thoroughbred

racetrack (chapter 2)

— Determined the relationship between area concentrations of PM10 and PM2.5 by stable,
stall, month, and time of day, and the overall prevalence of tracheal mucus in a
population of Thoroughbred racehorses (chapter 3)

— Used a case-control study design to identify significant differences in PM10 exposures

between cases and their matched controls (chapter 4)

The individual studies (chapters) have been written such that they can be read
independently, with references following each chapter. The studies are written in ﬁrst
person plural form (e. g. ‘we investigated), as a team of individuals was necessary to

perform these investigations.

Literature Review
Definition of [AD

Inﬂammatory airway disease (IAD) of racehorses is a syndrome characterized by
cough, exercise intolerance, excess tracheobronchial mucus that is visible upon
endoscopic exam, and inﬂammatory cells in tracheal aspirates and bronchoalveolar
lavage (BAL) (Anon 2003). Up to 45% of racehorses in a stable may be affected at any
one time (Newton et al. 2003a; Wood et al. 2005b), leading to a signiﬁcant ﬁnancial
impact as relatively small amounts of mucus in the airways result in horses ﬁnishing
further back in a race (Holcombe et al. 2006; MacNamara et al. 1990b). Due to the
potential impact of IAD on the racing industry, investigators have focused on elucidating

its cause, which is most likely multifactorial in nature.

Risk Factors for [AD

IAD is common in young racehorses, with incidence declining as the horse ages
and length of time in training increases (Christley et al. 2001a; Newton et al. 2003b;
Wood et al. 2005a, c), suggesting an infectious component, as increasing time spent in a
stable and exposure to bacterial and viral pathogens would confer immunity, Similar in
nature to that of a child entering daycare. As a result, several investigations have focused
on the role of infectious agents in the etiology of IAD. Bacterial, and not viral infections
are associated with neutrophilic inﬂammation and mucus accumulation in racehorse
populations (Chapman et al. 2000; Christley et al. 2001b; Wood et al. 1993; Wood et al.
2005b, c), and in one study as age of the population increased, the prevalence and

incidence of Streptococcus zooepidemicus and Streptococcus pneumoniae infection and

IAD decreased at the same rate (Wood et al. 2005a). Bacterial infections, however, are
not the only cause of IAD in racehorses, as many racehorses (over 20%) with airway
inﬂammation culture negative (Chapman et al. 2000; Christley et al. 2001a; Wood et al.
2005a). The stable environment has been identiﬁed as a second risk factor for airway
inﬂammation (Malikides and Hodgson 2005), as young racehorses are kept primarily
outdoors until, when at around the age of two, they are brought into a training facility or

racetrack to begin their careers.

Mucus, Airway Inﬂammation, and Particulates

Mucus secretion is a response to local irritation or inﬂammation through either
neural reﬂexes or secretion/recruitment of inﬂammatory cells (Wanner et al. 1996).
Particulates comprised of organic compounds, bacterial components (endotoxin), and
transition metals have been demonstrated to catalyze reactive oxygen species production
(Gonzalez-Flecha 2004; Shukla et al. 2000) potentially stimulating the release of stored
mucosubstance via signaling or cytotoxic events (Gonzalez-Flecha 2004). Inhalation of
particulate matter initiates an inﬂammatory response in both humans and animal models,
with a common component of this response being neutrophil recruitment (Ghio and
Devlin 2001; Ghio et al. 2006; Li et al. 1997). Exposure to organic dusts commonly
found in stables, dispersed from feed and bedding, have been shown to be associated with
airway neutrophilia and mucus accumulation in healthy horses (Gerber V 2004;

Holcombe et al. 2001; Tremblay et al. 1993).

Particle Composition

Particle composition in a stable is a complex mixture. Airborne exposure to
organic dusts (mold and fungi) and endotoxin resulting from agitation of feed and
bedding materials result in airway neutrophilia and an increase in mucus production in
both healthy horses (Gerber V 2004; Holcombe et al. 2001; Tremblay et al. 1993), and
those diagnosed with recurrent airway obstruction (RAO) (Bartner et al. 2005; Bemdt et
al. 2006, 2007; Derksen et al. 1985; Gerber et al. 2004a; McGorum et al. 1993a, b;
McGorum BC 1998 Sep; Robinson et al. 2006). Inorganic components of particulate
matter including metals (iron) and crustal materials (crystalline silica) are present in the
dusts to which horses are exposed (unpublished data), and have been associated with the
development and worsening of airway disease in humans and other animal models
(ACGIH 20053). As yet, these materials have not been evaluated in terms of their impact

on equine airway inﬂammation/disease (Ghio et al. 2006).

Co—exposures

As particles in a dust-cloud are not of uniform size or composition, exposures
over a period of time are varied. For example, as a racehorse eats its hay in the morning,
it may be exposed to a myriad of particles of differing composition. Airborne mold
spores and endotoxin may be elevated during morning cleanout and feeding activities,
and diesel exhaust persists throughout the day as a result of vehicle and tractor
movement. Evaluations of co-exposures in horses are limited, however, one such study
involving challenge with a hay dust suspension both with and without endotoxin

(endotoxin depleted hay dust suspension), demonstrated a ‘Synergistic proinﬂammatory

response’ (increased airway neutrophilia and dysfunction) in horses with RAO (Pirie et
al. 2003b). Low dose co-exposure in rats to particles harvested from rural and urban
locations revealed that regardless of sampling location and metal composition, the coarse
fraction (PM10) induced neutrophilic inﬂammation in the lung, presumably due to their
endotoxin content (Schins et al. 2004). Human studies reveal that relatively small
increases in ﬁne particles (PM2.5) are associated with increased hospital admissions due
to increased upper airway infection (Peel et al. 2005), and exposure studies in rats have
demonstrated a compromised ability of the lung to handle streptococcal infections

(Zelikoff et al. 2003; Zelikoff et al. 2002).

Particle Size

Airborne PM (dust) in racehorse stables originates from feed and bedding,
ﬂooring materials, racetrack footing materials, and emissions from neighboring roadways
and industrial operations. As there are many sources of PM in the environment of the
racehorse, size and composition of the particulate matter is not uniform. In regards to
human disease, particulate matter is generally classiﬁed in size ranges that include
inhalable (< 125 micron in diameter), thoracic (< 25 micron in diameter), and respirable
(< 10 micron in diameter) (ACGIH 2005a). Approximately 50% of particles 10 micron
in diameter have the ability to be deposited in the thoracic region (anywhere in the lung
airways), while that number goes up to approximately 90% for particles 2.5 micron in
diameter (ACGIH 2005a). On the same note, approximately 1% of particles 10 micron in
diameter have the ability to be deposited in the respirable region (gas exchange region),

increasing to 95% of particles 2.5 micron in diameter (ACGIH 2005a). The

Environmental Protection Agency (EPA) has classiﬁed these particles as ‘coarse’ or
PM10 (2.5-10 micron in diameter) and ‘ﬁne’ or PM2.5 (<2.5 micron in

diameter)(USEPA).

Consequence of Particulate Exposures

Human epidemiology studies consistently show that small increases in ambient
(background) concentrations of PM 10 and PM 2.5 of mixed composition results in a
worsening of symptoms in patients with pre-existing airway disease (Brunekreef and
Forsberg 2005; Oberdorster 1996; Pope 1991b; Pope et al. 2002), an increase in hospital
admissions (Peel et al. 2005; Pope 1991a), days away from school (Morgenstern et al.
2007; Pierse et al. 2006; Schwartz 2004), and lost work days (Pope 1991a; Pope III 2000
; Pope III et al. 2002; Schwartz 2004; Sunyer 2001). Overloading of the lungs by
particulate matter that is considered to be relatively inert in nature has the potential to
impair macrophage function resulting in a prolonged inﬂammatory response (Morrow
1992). Mechanisms that have been suggested in the ‘overload’ model include neutrophil
and alveolar macrophage activation and release of inﬂammatory cytokines (Morrow
1992; Oberdorster 1995). In-vitro exposure studies in mice with ﬁne (PM2.5)
concentrated air particles (CAPS) resulted in impaired macrophage interactions (binding,
internalization, and killing) with Streptococcus pneumoniae (Zhou and Kobzik 2007). A
similar mechanism could be responsible for the high prevalence of streptococcal

infections when racehorses ﬁrst enter training.

Particulate Monitoring in Stables

With regard to particulate effects on the airways in horses, organic dust
challenges have been used to induce airway inﬂammation (Derksen et al. 1985; Pirie et
al. 2003a), area particulate concentrations have been measured in a riding stable
(Crichlow et al. 1980), the dust-generating potential of various feeds and bedding have
been reported (Vandenput et al. 1997), and particulate load has been measured with
personal samplers under various management systems (McGorum et al. 1998; Webster et
al. 1987; Woods et al. 1993). Previous studies conducted in controlled settings as well as
private boarding stables utilizing traditional ﬁlter/pump methods (Clarke and Madelin
1987; Clarke et al. 1987; Crichlow et al. 1980; Kirschvink et al. 2002; McGorum et al.
1998; Woods et al. 1993) and direct reading instruments (Clements and Pirie 2007a, b)
have shown that dusts present in the breathing zone of horses are composed of particles
that include those in size ranges (PM10 and PM2.5) consistent with those described in the

human literature.

Justiﬁcation of Our Studies in Relation to the Current Literature

Our studies are the ﬁrst to investigate the association between airborne exposure
to particulates and the presence of tracheal mucus at an American Thoroughbred
Racetrack. The ﬁrst step in characterizing the environment of the racetrack is to survey
the stables (particulate mapping) to better understand the impact of stable design,
ventilation status, and management practices on area concentrations and numbers of

particles in size ranges (PM10 and PM2.5) known to inﬂuence disease in humans and

animal models. Because overexposure to particulate matter is recognized as a
contributing factor in airway inﬂammation (Morrow 1992; Oberdorster 1995), and is the
basis for setting exposure limits for ‘particulates not otherwise classiﬁed’ (PNOC) in
occupational settings (ACGIH 2005a), we chose to utilize real-time particulate monitors
which determine average and peak PM10 and PM2.5 particulate concentrations.
‘Mapping’ particulates, the act of determining PM concentrations at points equally
distributed throughout a pre-determined space, is also an important step in determining
the contribution of area particulates dispersed by sources not actively inﬂuenced by the
horse (e. g. stall cleaning, the act of feeding, ventilation status, exhaust and dust from
roadways). Careful tracking of activities and ventilation status (doors/windows open or
closed, air moving fans, etc.) gives insight into the relative contribution of these external
sources on area particulate concentrations, and is crucial for developing the rationale for

cost-effective and comprehensive sampling protocols focusing on composition.

Area particulate concentrations are important to understand in the racehorse
environment as the majority of the epidemiological investigations of the effects of air
quality on respiratory health, as measured by excess morbidity and mortality, focus on
concentrations of PM10 and PM2.5, and these values would likely be the minimum
concentrations to which a horse is exposed. However, depending on the activity in which
an individual is involved, area concentrations may or may not reﬂect personal exposure.
In the case of horses, individual behavior may determine daily exposure to particulates.
Some horses bury their faces in hay when they eat, others sniff and dig at the bedding,

and some lie down more than others. All of these idiosyncrasies have the potential to

10

affect both average and peak exposures. For this reason, we used small-scale personal
monitors, with the sampling point located at the nostril of the horse. Our sampling
protocol, for the ﬁrst time, allowed us to characterize average, minimum, and maximum
particulate concentrations, and the length of time these exposures occur. Previous
particulate sampling technology used in IAD and RAO studies, consisting of pump and
ﬁlter apparatus, measured only average concentration values and did not allow for the
evaluation of ‘peak’ exposures. Use of real-time monitors allowed us to determine
whether or not there were signiﬁcant differences (both average and peak values) in
exposures of horses both with and without tracheal mucus, and the factors that lead to

those exposures.

11

References

ACGIH (2005) American Conference of Governmental Industrial Hygienists TLVS and
BEIS. Threshold Limit Values for Chemical Substances and Physical Agents and
Biological Exposure Indices. Cincinnatti, OH.

Anon (2003) Workshop report: Inﬂammatory airway disease: deﬁning the syndrome.
Equine Veterinary Education 5, 81-82.

Bartner, L.R., Tesfaigzi, Y. and Robinson, NE. (2005) Persistent mucus accumulation: A
consequence of delayed mucous cell death in heaves-affected horses. American Journal
of Respiratory Critical Care Medicine 171, In press.

Bemdt, A., Derksen, F .J ., Venta, P., Ewart, 8., Yuzbasiyan-Gurkan, V. and Robinson,
NE. (2007) Elevated amount of toll-like receptor 4 mRNA in bronchial epithelial cells is
associated with airway inﬂammation in horses with recurrent airway obstruction. Am J
Physiol Lung Cell Mol Physiol 292, 936-943

Brunekreef, B. and Forsberg, B. (2005) Epidemiological evidence of effects of coarse
airborne particles on health. European Respiratory Journal 26, 309-318.

Chapman, P.S., Green, C., Main, J .P., Taylor, P.M., Cunningham, F.M., Cook, AI. and
Marr, CM. (2000) Retrospective study of the relationships between age, inﬂammation
and the isolation of bacteria from the lower respiratory tract of thoroughbred horses.
Veterinary Record 146, 91-95.

Christley, R.M., Hodgson, D.R., Rose, R.J., Hodgson, J .L., Wood, J .L. and Reid, 8.W.
(2001a) Coughing in thoroughbred racehorses: risk factors and tracheal endoscopic and
cytological ﬁndings. Veterinary Record 148, 99-104.

Christley, R.M., Hodgson, D.R., Rose, R.J., Wood, J .L., Reids, S.W., Whitear, KC. and
Hodgson, J .L. (2001b) A case-control study of respiratory disease in Thoroughbred
racehorses in Sydney, Australia. Equine Veterinary Journal 33, 256-264.

Clarke, AF. and Madelin, T. (1987) Technique for assessing respiratory health hazards
from hay and other source materials. Equine Veterinary Journal 19, 442-447.

Clarke, A.F., Madelin, T. and Alpress, KG. (1987) The relationship of air hygiene in
stables to lower airway disease and pharyngeal lymphoid hyperplasia in two groups of
Thorough-bred horses. Equine Veterinary Journal 19, 524—530.

Clements, J .M. and Pirie, R.S. (2007a) Respirable dust concentrations in equine stables.
Part 1: Validation of equipment and effect of various management systems. Research in
Veterinary Science 83, 256-262.

12

Clements, J .M. and Pirie, R.S. (2007b) Respirable dust concentrations in equine stables.
Part 2: The beneﬁts of soaking hay and optimising the environment in a neighbouring
stable. Research in Veterinary Science 83, 263-268.

Crichlow, E.C., Yoshida, K. and Wallace, K. (1980) Dust levels in a riding stable. Equine
Veterinary Journal 12, 185-188.

Derksen, F .J ., Robinson, N.E., Armstrong, P.J., A., SJ. and Slocombe, RF. (1985)
Airway reactivity in ponies with recurrent airway obstruction (heaves). Journal of
Applied Physiology 58, 598-604.

Gerber V, L.A., Bemey C, Robinson NE (2004) Airway mucus in recurrent airway
obstruction--short-term response to environmental challenge. J Vet Internal Med. 18, 92-
97.

Gerber, V., Lindberg, A., Bemey, C. and Robinson, NE. (2004) Airway mucus in
recurrent airway obstruction--Short-term response to environmental challenge. J Vet
Intern Med 18, 92-97.

Ghio, A]. and Devlin, RB. (2001) Inﬂammatory Lung Injury after Bronchial Instillation
of Air Pollution Particles. pp 704-708.

Ghio, A.J., Mazan, M.R., Hoffman, AM. and Robinson, NE. (2006) Correlates between
human lung injury after particle exposure and recurrent airway obstruction in the horse.
Equine Vet J 38, 362-367.

Gonzalez-Flecha, B. (2004) Oxidant mechanisms in response to ambient air particles.
Molecular Aspects of Medicine 25, 169-182.

Holcombe, 8.J., Jackson, C., Gerber, V., Jefcoat, A., Bemey, C., Eberhardt, S. and
Robinson, NE. (2001) Stabling is associated with airway inﬂammation in young Arabian
horses. Equine Veterinary Journal 33, 244-249.

Holcombe, 8.J., Robinson, N.E., Derksen, F.J., Berthold, B., Genovese, R., Miller, R.,
DeFeijter-Rupp, H., Carr, E.A., Eberhard, S.W., Boruta, D. and Kaneene, J .B. (2006)
Effect of tracheal mucus and tracheal cytology on racing performance in Thoroughbred
racehorses. Equine Vet J 38(4), 300-304.

Kirschvink, N., Di silvestro, F., Sbai, I., Vandenput, 8., Art, T., Roberts, C. and Lekeux,
P. (2002) The Use of Cardboard Bedding Material as Part of an Environmental Control
Regime for Heaves-affected Horses: In Vitro Assessment of Airborne Dust and

Aeroallergen Concentration and In Vivo Effects on Lung Function. The Veterinary
Journal 163, 319-325.

13

Li, X.Y., Gilrnour, P.S., Donaldson, K. and MacNee, W. (1997) In vivo and in vitro
proinﬂammatory effects of particulate air pollution (PM10). Environ Health Perspect 105
Suppl 5, 1279-1283.

MacNamara, 8., Bauer, 8. and Iafe, J. (1990) Endosc0pic evaluation of exercise-induced
pulmonary hemorrhage and chronic obstructive pulmonary disease in association with
poor performance in racing Standardbreds. Journal of the American Veterinary Medical
Association 196, 443-445.

Malikides, N. and Hodgson, J .L. (2005) Endotoxin is an Early Component of the Causal
Pathway in Inﬂammatory Airway Disease (IAD) in Yount TB Racehorses. In: 2005
World Equine Airways Symposium, Ithica, New York.

McGorum, B.C., Dixon, PM. and Halliwell, R.E.W. (1993a) Quantiﬁcation of histamine
in plasma and pulmonary ﬂuids from horses with chronic obstructive pulmonary disease,
before and after "natural (hay and straw) challenges". Veterinary Immunology and
Immunopathology 36, 223-237.

McGorum, B.C., Dixon, RM. and Halliwell, R.E.W. (1993b) Responses of horses
affected with chronic obstructive pulmonary disease to inhalation challenges with mould
antigens. Equine Veterinary Journal 25, 261-267.

McGorum BC, E.J., Cullen RT (1998 Sep) Total and respirable airborne dust endotoxin
concentrations in three equine management systems. Equine Vet J. 30, 430-434.

McGorum, B.C., Ellison, J. and Cullen, RT. (1998) Total and respirable airborne dust
endotoxin concentrations in three equine management systems. Equine Veterinary
Journal 30, 430-434.

Morgenstem, V., Zutavem, A., Cyrys, J ., Brockow, I., Gehring, U., Koletzko, 8., Bauer,
C.P., Reinhardt, D., Wichmann, HE. and Heinrich, J. (2007) Respiratory health and
individual estimated exposure to trafﬁc-related air pollutants in a cohort of young
children. pp 8-16.

Morrow, PE. (1992) Dust overloading of the lungs: update and appraisal. T oxicol Appl
Pharmacol 113, 1-12.

Newton, J .R., Wood, J .L. and Chanter, N. (2003a) A case control study of factors and
infections associated with clinically apparent respiratory disease in UK Thoroughbred
racehorses. Preventive Veterinary Medicine 60, 107-132.

Newton, J .R., Wood, J .L.N. and Chanter, N. (2003b) A case control study of factors and

infections associated with clinically apparent respiratory disease in UK Thoroughbred
racehorses. Preventive Veterinary Medicine 60, 107-132.

14

Oberdorster, G. (1995) Lung Particle Overload: Implications for Occupational Exposures
to Particles. Regulatory Toxicology and Pharmacology 21, 123-135.

Oberdorster, G. (1996) Signiﬁcance of particle parameters in the evaluation of exposure-
dose-response relationships of inhaled particles. Inhalation Toxicology 8 Suppl, 73-89.

Peel, J .L., Tolbert, P.E., Klein, M., Metzger, K.B., Flanders, W.D., Todd, K.,
Mulholland, J .A., Ryan, PB. and Frumkin, H. (2005) Ambient air pollution and
respiratory emergency department visits. Epidemiology 16, 164-174.

Pierse, N., Rushton, L., Harris, R.S., Kuehni, C.E., Silverrnan, M. and Grigg, J. (2006)
Locally generated particulate pollution and respiratory symptoms in young children. pp
216-220.

Pirie, R.S., Collie, D.D., Dixon, PM. and McGorum, B.C. (2003a) Inhaled endotoxin and
organic dust particulates have synergistic proinﬂammatory effects in equine heaves
(organic dust-induced asthma). Clinical and Experimental Allergy 33, 676-683.

Pirie, R.S., Collie, D.D.S., Dixon, PM. and McGorum, B.C. (2003b) Inhaled endotoxin
and organic dust particulates have synergistic proinﬂammatory effects in equine heaves
(organic dust-induced asthma). Clinical & Experimental Allergy 33, 676-683.

Pope, C. (19913) Respiratory hospital admissions associated with PM10 pollution in
Utah, Salt Lake, and Cache valleys. Arch Env Health 46, 90-97.

Pope, C.A., 3rd (1991b) Respiratory hospital admissions associated with PM10 pollution
in Utah, Salt Lake, and Cache Valleys. Archives of Environmental Health 46, 90-97.

Pope, C.A., 3rd, Burnett, R.T., Thun, M.J., Calle, E.E., Krewski, D., Ito, K. and Thurston,
GD. (2002) Lung cancer, cardiopulmonary mortality, and long-term exposure to ﬁne
particulate air pollution. Journal of the American Medical Association 287, 1132-1141.

Pope 111, CA. (2000 ) What do epidemiologic ﬁndings tell us about health effects of
environmental aerosols? J Aerosol Med. 13, 335-354.

Pope HI, C.A., Burnett, R.T., Thun, M.J., Calle, E.E., Krewski, D., Ito, K. and Thurston,
GD. (2002) Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine
Particulate Air Pollution. JAMA 287, 1132-1141.

Robinson, N.E., Karmaus, W., Holcombe, 8.J., Carr, EA. and Derksen, F .J . (2006)
Airway inﬂammation in Michigan pleasure horses: prevalence and risk factors. Equine
Veterinary Journal 38, 293-299.

Schins, R.P., Lightbody, J .H., Borm, P.J., Shi, T., Donaldson, K. and Stone, V. (2004)

Inﬂammatory effects of coarse and ﬁne particulate matter in relation to chemical and
biological constituents. T oxicol Appl Pharmacol 195, 1-11.

15

Schwartz, J. (2004) Air Pollution and Children's Health. Pediatrics 113, 1037-1043.

Shukla, A., Timblin, C., BeruBe, K., Gordon, T., McKinney, W., Driscoll, K., Vacek, P.
and Mossman, ET (2000) Inhaled particulate matter causes expression of nuclear factor
(NF )-kappaB-related genes and oxidant-dependent NF-kappaB activation in vitro.
American Journal of Respiratory Cell and Molecular Biology 23, 182-187.

Sunyer, J. (2001) Urban air pollution and chronic obstructive pulmonary disease: a
review. Eur Respir J 17, 1024-1033.

Tremblay, G.M., F erland, C., Lapointe, J .-M., Vrins, A., Lavoie, J .P. and Corrnier, Y.
(1993) Effect of stabling on bronchoalveolar cells obtained from normal and COPD
horses. Equine Veterinary Journal 25, 194-197.

USEPA National Ambient Air Quality Standards. http://www. epagov/ttn/naaqs/
Referenced April 1, 2006.

Vandenput, 8., Istasse, L., Nicks, B. and Lekeux, P. (1997) Airborne dust and
aeroallergen concentrations in different sources of feed and bedding for horses.
Veterinary Quarterly 19, 154-158.

Wanner, A., Salathe, M. and O'Riordan, T.G. (1996) Mucociliary clearance in the
airways. Am J Respir Crit Care Med 154, 1868-1902.

Webster, A.J.F., Clarke, A.F., Madelin, TM. and Wathes, CM. (1987) Air hygiene in
stables 1: Effects of stable design, ventilation and management on the concentration of
respirable dust. Equine Veterinary Journal 19, 448-453.

Wood, J .L., Burrell, M.H., Roberts, C.A., Chanter, N. and Shaw, Y. (1993) Streptococci
and Pasteurella spp. associated with disease of the equine lower respiratory tract. Equine
Veterinary Journal 25, 314-31 8.

Wood, J .L., Newton, J .R., Chanter, N. and Mumford, J .A. (2005a) Association between
respiratory disease and bacterial and viral infections in British racehorses. J Clin
Microbiol. 43, 120-126.

Wood, J .L., Newton, J .R., Chanter, N. and Mumford, J .A. (20050) Inﬂammatory airway
disease, nasal discharge, and respiratory infections in young British racehorses. Equine
Vet J 37, 236-242.

Woods, P.S., Robinson, N.E., Swanson, M.C., Reed, C.E., Broadstone, RV. and

Derksen, F .J . (1993) Airborne dust and aeroallergen concentration in a horse stable under
two different management systems. Equine Veterinary Journal 25, 208-213.

16

Zelikoff, J .T., Chen, L.C., Cohen, M.D., Fang, K., Gordon, T., Li, Y., Nadziejko, C. and
Schlesinger, RB. (2003) Effects of inhaled ambient particulate matter on pulmonary
antimicrobial immune defense. Inhalation Toxicology 15, 131-150.

Zelikoff, J .T., Schermerhom, K.F., Cohen, MD. and Schlesinger, RB. (2002) A Role for
Associated Transition Metals in the Irnmunotixicity of Inhaled Ambient Particulate
Matter. Environ Health Perspect 110, 871-875.

Zhou, H. and Kobzik, L. (2007) Effect of Concentrated Ambient Particles on

Macrophage Phagocytosis and Killing of Streptococcus pneumoniae. Am J Respir Cell
Mol Biol. 36, 460-465.

17

CHAPTER 1

Indoor Air Quality in a Boarding Stable

Summary
Reasons for Performing Study

Recurrent airway obstruction (RAO) and inﬂammatory airway disease (IAD) are
equine respiratory syndromes that are characteristically associated with inhalation of
organic dust. F eedstuffs, bedding, and arena footing are all sources of particulate matter

that, upon agitation, have the potential for dispersion into the horse's breathing zone.

Objectives

Our study had four objectives: 1) to measure particulate concentrations in the
breathing zone of horses eating a variety of feeds and during a variety of routine stable
activities, including riding in an arena; 2) to determine the amount of some crustaceous
metals and crystalline silica in arena footing and dust on ledges; 3) to investigate the
efﬁcacy of adding water to reduce particulate generation concentrations arising from
feeds and activities; and 4) to determine stall ventilation and ammonia concentrations in

stalls during winter and summer.

Methods
Two aerosol photometers measured coarse and ﬁne particulate concentrations in
the breathing zone of the horses eating hay, alfalfa cubes, complete pelleted feed, sweet

feed, and a “senior” diet. Particle concentrations also were measured during stall

18

 

cleanout, raking of aisles, removal of cobwebs, and during l-hour riding lessons in the
indoor arena with footing of varying water content. Particle size distribution and the
content of several metals and crystalline silica were measured in arena footing and dust
on ledges. Stall ventilation was evaluated by measurement of ammonia concentrations in

the air.

Results
In all situations, concentrations of coarse particles always were greater than those

of ﬁne. Eating dry hay and alfalfa cubes generated the highest concentrations of

 

particulates, with lower concentrations originating from complete feed and sweet feeds.
Stall cleaning generated fewer particles than aisle sweeping or removal of cobwebs.
Adding moisture reduced particle concentrations dramatically. Riding in the arena
generated large excessive concentrations of dust unless moisture content was greater than

10 percent. Stall ammonia concentrations were greater in summer than in winter.

Conclusions

Horses may be routinely exposed to levels of particulate matter and ammonia that
exceed occupational exposure limits for people. Inorganic components of dust sampled
included iron and crystalline silica, both known to induce respiratory disease in
occupational settings. Wetting was very effective in controlling dusts in the breathing
zone of horses. Location of stall, season, and proximity to source of natural ventilation

affects ammonia concentrations.

19

Potential Relevance

Understanding sources of dust and the importance of breathing zone
concentrations will help in managing horses with respiratory disease. Horse owners need
to know that dust can be suppressed by use of water, and both dust and ammonia can be
removed by improved ventilation. Real-time monitors are an effective tool in the
education of stable owners about inhalation hazards present in their stables. Particulate

exposures in horses may be greatly underestimated.

Introduction

Recurrent airway obstruction (RAO) and inﬂammatory airway disease (IAD) are
two equine respiratory syndromes that are characteristically associated with poor air
quality in stables (Anon 2001, 2003). Feedstuffs and bedding materials are common
sources of particulate matter in the equine environment. If sufﬁciently agitated, as occurs
during eating and movement of the horse within the stall, these materials have the
potential to disperse particulate into the horse's breathing zone in signiﬁcant
concentrations (Vandenput et al. 1997; Woods et al. 1993).

Human epidemiology studies suggest a strong association between exposure to
coarse and ﬁne airborne particulate matter and worsening of disease in patients with
asthma and COPD (Oberdorster 1996; Pope 1991b; Pope et al. 2002). Coarse
particulates have an aerodynamic diameter between 2.5 and 10 p. (PM10), and ﬁne
particulates are those with a mean aerodynamic diameter of less than 2.5 u (PM2,5). In
the US, the Environmental Protection Agency (EPA) has established National Ambient

Air Quality Standards (NAAQS) for particles in these two size ranges. In addition, the

20

American Conference of Governmental Industrial Hygiene (ACGIH) has established
chemical and size-Speciﬁc occupational exposure limits (OELs) for coarse and ﬁne
particulates, but their size ranges differ from those referred to in the NAAQS (ACGIH
2005b; Anon 2006). They are inhalable (<100 u diameter), thoracic (1-25 It in diameter),
and respirable (<10 u in diameter) particulates.

With regard to particulate effects on the airways in horses, organic dust
challenges have been used to induce airway inﬂammation (Derksen et al. 1985; Pirie et
al. 2003a), area particulate concentrations have been measured in a riding stable
(Crichlow et al. 1980), the dust-generating potential of various feeds and bedding have
been reported (Vandenput et al. 1997), and particulate load has been measured with
personal samplers under various management systems (McGorum et al. 1998; Webster et
al. 1987; Woods et al. 1993). These investigations have been conducted in relatively
controlled academic environments, and none have measured particulate load in the
breathing zone of horses in stables during a variety of routine activities. This paper
begins to correct this deﬁcit by measuring PM2_5 and PM10 in a large boarding stable that
a) uses various forms of feedstuffs as speciﬁed by each horse owner, b) is managed as a
cooperative, allowing for routine tasks to be performed by multiple individuals, and c)
has an active riding/lesson schedule allowing for sampling to occur in an indoor riding
arena. Through the use of real-time particle samplers, not only the average values but
peak exposures were measured. The bulk dust samples were analyzed for inorganic
components known to cause respiratory disease in humans. Measurements of

concentrations of ammonia allowed estimation of regional ventilation in the stable.

21

Materials and Methods
The Boarding Stable

The boarding stable consists of two buildings connected by an aisle (Figure 1).
The larger building measures 66 x 27 m, and consists of a 40 x 20 m indoor arena with 13
stalls along the east wall, each measuring approximately 3 x 3.5 m. The stalls are
separated from the arena by an aisle, and a wall measuring 1.2 m in height. Flooring in
the stalls consists of packed clay on top of which sawdust is partially replaced daily.
Flooring in the aisle consists of packed dirt/clay with rubber matting down the center.
Arena footing consists of a solid clay base underneath a mixture of sand (approximately
20 cm in depth). The footing has not been completely changed for over 10 years, but has
been added to approximately every other year. There is a large sliding door (4.5 x 6 m) at
one end of the arena, and a Sliding door (2.4 x 3 m) at each end of the aisle. The older
and smaller of the two buildings is wooden with a hayloﬁ that spans the entire length and

width of the structure. It contains ﬁve stalls (3 x 3.5 m) and measures 10 x 14 m overall.

Particulate Measurements

All particulate air sampling was conducted by use of two aerosol photometers
(DustTrak, TSI, Shoreview, Minnesota). Each photometer was ﬁtted with either a 10 u
or a 2.5 u particle size conditioner, and was calibrated before and after measurements

according to manufacturer instructions (Anon 2005).

22

Feed

Feedstuffs utilized in this study were baled alfalfa/timothy hay, commercially
prepared sweet feed, a complete feed, and a senior feed. All feed, with the exception of
the alfalfa/timothy hay, was consumed from a feed tub suspended in the front comer of
the stall. The hay was fed from the ﬂoor. To determine the effect of wetting on
particulate concentrations, feedstuffs were watered down immediately prior to feeding
via a spray nozzle attached to a garden hose. Five sets of measurements were taken for
each type of feed under both wet and dry conditions. The ﬂexible tube (0.64 mm IAD)
was attached to the particle size conditioners on each of the two aerosol photometers and
its end was held adjacent to the nostril of the horse for the duration of feeding in order to
make Simultaneous measurements of PM10 and PM2,5 in the breathing zone of the horse.
For each measurement, minimum, maximum, and average values were recorded in

mg/m3 .

Stable Activities

Five sets of measurements were taken during aisle sweeping, stall cleaning and
removal of cobwebs, and grading the arena. Arena grading was performed by pulling a
metal harrow (drag) over the arena for 13 minutes with a four-wheeled utility vehicle. To
determine the effect of wetting, a garden hose with a spray nozzle was utilized to dampen
the surface of the arena footing, aisle footing, and stall sawdust immediately before each
activity began. It was not possible to study removal of cobwebs in the wet conditions

because the aerosol photometer views water aerosol as particulate.

23

A horse was placed in crossties in the center of the aisle during sweeping and
arena grading, and measurements were taken adjacent to the nostril. The horse was
placed back in its stall during stall cleaning and removal of cobwebs, with all
measurements taken adjacent to the nostril. Measurements were taken for the duration of

. . . . . . 3
the actrvrty, and the mrnrmum, maxrmum, and average values In mg/m were recorded.

Riding

Five students rode at a walk, trot, and canter for 60 minutes in each of six trials.
Five trials were conducted with varying moisture content in the footing. The sixth was
performed after application of a commercial dust suppressant containing calcium
chloride. Moisture content was measured by weighing a sample of footing, and drying it
to constant weight.

Both aerosol photometers were placed side-by-side in the center of the arena
approximately 1 m above the ﬂoor in order to approximate the breathing zone of the
horse. The monitors ran continuously throughout each 60-minute trial. Data were
recorded every second, and minimum, maximum, and average values were computed for

each trial.

Bulk Samples

Three bulk dust samples were taken from the arena footing, and four samples
from ledges on the wall of the stable at least 3 m above the ground. The height of
sampling was selected so that particulates represented those that would have passed

through the breathing zone and settled on a ﬂat surface. For both footing and wall

24

samples, the individual samples were combined and mixed and a representative sample
was taken from each. Each sample was split and sent to the laboratories for measurement
of particle size distribution and analysis of composition. Particle size analysis was
performed using an L813 320 particle size analyzer with the Tornado dry powder sample
module, and a speciﬁed size range of 0.4 u to 2000 u in diameter (Beckman-Coulter
(2006) http://www. beckmanco ulter. com/products/instrumentLpartChar/pcjsI 3320. asp).
Composition analysis of the arena footing included crystalline silica (quartz, crystobalite,
tridymite) using the NIOSH 7500 method (Galson 2006) and x-ray diffraction, and ten
metals (silver, cadmium, cobalt, chromium, copper, iron, manganese, nickel, lead, and
zinc) using a "modiﬁed" NIOSH 7300 method and ICAP/ICP-MS (Galson 2006). Dust
samples taken from the ledges were analyzed for crystalline silica and iron, as well as
particle size distribution via methods described above. Metal content was reported in

mg/kg and crystalline silica was reported as a percentage of total dust.

Gas Sampling

Gas sampling for ammonia was performed utilizing pumps and sampling tubes
(Dragar, Accuro hand-held pump, Luebeck, Germany). Seven stalls were identiﬁed for
this portion of the sampling scheme (Figure l) and each stall was sampled once in winter
when all stable doors were closed, and once in summer while all stable doors were open.
Samples were taken before cleaning out, in the center of each stall, 1 m above the ground,

approximating the breathing zone on a horse.

25

Statistical Analysis

Effects of feeding and stable activities on particle loads in the PM10 and PM2.5
ranges were compared by AN OVA. When signiﬁcant defects were identiﬁed,
differences between means were compared by Student Neuman Keuls procedure.
Ammonia concentrations in summer and winter were compared by a paired t-test. In all

analyses, signiﬁcance was set at p<0 .05.

Results
Feeds

In all feeds, the average and maximum load of PM10 particles was signiﬁcantly
greater than those of PM2.5. Dry hay provided the greatest average and maximum PM 10
and PM2.5 particulate load in the horse's breathing zone (Figure 2). The maximal PM10
value was underestimated, because in two samples, the particle load was greater than the
readable range of the aerosol photometer, i.e. greater than 150 mg/m3 . These high values
originated from hay that was taken from the bottom of the stack. The average but not the
maximum particulate load from dry alfalfa cubes was signiﬁcantly less than that from dry
hay. Pouring the cubes into the feed tub resulted in a visible cloud of dust that remained
suspended while the horse ate. Complete pelleted feed generated a signiﬁcantly smaller
particle load than hay or alfalfa cubes, but the load was greater than that from sweet feed

or the commercially available feed for older horses.

26

Stable Activities

In the case of each stable activity, PM10 particle average and maximum load was
signiﬁcantly greater than PM”. Cobweb removal resulted in the greatest particle load in
the horse's breathing zone for both Size ranges (Figure 3). Aisle sweeping produced
signiﬁcantly less PM10 load than cobweb removal but not signiﬁcantly less PM2.5. Stall
cleaning and bedding with sawdust generated signiﬁcantly less PM10 and PM2.5 than
either of the former procedures. Wetting the aisles and the stalls decreased the average
and maximum load of both PM10 and PM2.5 in the horse's breathing zone to less than 2

mgma

Riding

In the absence of riding, the area particulate load averaged 0.045 mg/m3 for PM10
and 0.021 mg/m3 for PM; 5. During a one-hour riding lesson with ﬁve horses under dry
footing conditions, average and maximum load were 9.1 and 55.3 mg/m3 and 2.3 and
60.6 mg/m3 for PM10 and PM“, respectively. Figure 4 shows how the particle load
varied during the hour, increasing during warm-up and decreasing during cooling down.
As moisture content of footing increased, particulate load decreased. When moisture
content was ten percent or greater, particulate load was almost one hundredth that of dry
footing (Figure 5). PM10 particulates were maximally reduced when water content was
6.1 percent or greater. In the case of PM2.5, maximal reduction occurred at 9.33 percent
water content. After application of the commercial dust suppressant, average and

maximum load were 0.03 and 2.2 mg/m3 and 0.02 and 0.17 mg/m3 for PM10 and PM2.5,

27

respectively, similar values to those resulting from a water content of 9.33 percent or

greater.

Bulk Samples
The distribution of particle sizes in footing and wall samples is shown in Figure 6.

In the footing samples, particles greater than 10 p in diameter constituted virtually the

entire sample. While larger particles also constituted most of the wall sample, there was

a considerable percentage below 10 u and a secondary peak centered around 1 u.

Crystalline silica (quartz) comprised the bulk of both footing and wall samples.
Iron was present in both samples, with the concentration being greater in wall sample.
Manganese and copper were also present in the footing samples. Other metals were

below the laboratory limit of detection (Table 1-1).

Gases

In both summer and winter, concentrations of ammonia were below the limits of
detection in the stable aisle. The mean concentration of ammonia in stalls was
signiﬁcantly higher in Stunmer (31 :l:l9 ppm, mean :t SD) than in winter. In summer, the
concentration of ammonia was highest in the center of the stable and decreased toward
the doors at each end with the lowest value being in stall number 7 (Figure 1) where the

prevailing wind entered the stable.

28

Discussion

In this ﬁrst investigation of particulate concentrations in the breathing zone of
horses in a large boarding stable, average particulate concentrations only occasionally
exceeded recommended occupational exposure limits for humans (ACGIH 2005b), but
peak exposures regularly exceeded the excursion limits for both PM 10 and PM2.5, i.e.,
inhalable and respirable Sizes. Eating certain feeds generated the highest exposures
followed by cleaning and riding. Results from gas sampling indicated that ventilation
within the stable was not uniform, so that suspended particulates and irritant gases such

as ammonia were not removed.

The process of eating generates dust concentrations in the breathing zone of the
horse that are generally not accurately perceived by the observer who is breathing the air
several meters away from the horse's nose (Woods et al. 1993). When the horse
consumes hay, the act of biting and pulling the captured wisp of hay from the ﬂake or
bale results in the immediate dispersion of high concentrations of dusts at the point of the
nostril. The results of our study indicate that the peak concentrations in the breathing
zone regularly exceeded 87.5 mg/m3 PM10 and 20 mg/m3 PM2_5 while eating hay. These
values are more than two times greater than the excursion limits of 30 mg/m3 and 9
mg/m3 PM") and PM2.5 allowed in an occupational setting (ACGIH 2005b). The hay
utilized in our study was all grown in the same ﬁeld, and was baled and hauled the same
day. Despite this common source, dust levels varied considerably between the bales of
hay over the course of the study. Hay that was fed close to the time of bailing resulted in

less dust generation during feeding than the hay that was stored and fed several months

29

later. Interestingly, none of the hay that was fed during the study was considered to be
dusty or moldy by observers, even the hay that resulted in breathing zone concentrations

beyond our sampler's maximal range, which was 150 mg/m3.

Frequently, horses that are diagnosed with heaves are fed cubed hay or complete
pelleted feed to reduce particulate exposure. Our results suggest that this should be done
with great caution, especially in the case of cubed hay, because peak dust levels in the
breathing zone are similar to those found during consumption of baled hay. In this stable,
cubes are consumed from feed tubs suspended in the comer of the stall. When cubes
were poured into the feed tubs, particle clouds formed immediately adjacent to the horse's
nostrils, and did not disperse for several minutes. This phenomenon also occurred when
complete pelleted feed was taken from the bottom of the bag and poured into the feed
tubs.

Sweet feeds and senior feeds resulted in very low dust exposures. Similar levels
were measured during feeding of complete pelleted feeds especially when the pellets
were taken from the top of the bag. Molasses used in sweet feeds and senior feeds are a
ﬂavor enhancer but also act as a binding agent, thus dramatically reducing particle
liberation. For this reason, these types of feeds should be optimal in the management of

horses with heaves.
As might be expected, routine stable-cleaning activities also resulted in high

exposures to particulates. Removal of cobwebs resulted in very high exposures to small

particles, while raking and sweeping of the aisles dispersed clouds of particles

30

predominantly in the PM10 range. Stall cleaning resulted in moderate increases in dust
levels of predominantly larger particles (PM10). Exposure concentrations during stall
cleaning may have been underestimated as the wood shavings available during this study
appeared to be particularly clean in relation to other bulk loads received. Straw was not
available at this boarding stable, but breathing zone particulate concentrations during
cleanout at another stable were similar to those found in the breathing zone of horses

eating hay (unpublished data).

Under dry conditions, riding in the arena or grading its surface generated high
concentrations of both small and large size particles. This is a common observation by
stable owners and riders who see the dust, and experience upper respiratory responses
such as itchy and watery eyes, sneezing, and cough. For this reason, dust management
practices are often focused primarily on the arena. Dust concentrations in both size
ranges increased progressively during the one-hour riding lesson, and then decreased
rapidly to background concentrations when riding ceased. In this particular stable, riding
in the arena does not continue throughout the day, thus allowing for dust levels to return
to background concentrations. This may underestimate exposure potentials in other
stables, where, in many cases, riding occurs continually throughout the day, so the dust
does not have time to settle out. Unfortunately, we did not have sufﬁcient samplers to
concurrently measure particulate exposure in the stalled horses while the arena was in
use. However, in one quick measurement made during arena use under dry conditions,
particle load in the breathing zone of a stalled horse was only slightly less than that

recorded in the arena.

31

Crystalline silica and iron were the two components of the dust identiﬁed in bulk
samples ﬁ'om the stable that are known to result in occupational respiratory morbidity in
overexposure situations. Particulates in general and silica in particular can induce
generation of reactive oxygen species and production of inﬂammatory cytokines (e. g. IL-
6, IL-8, TNF-alpha) and metals such as iron contribute to the toxic effects of such
particulates (Dye et al. 1999; F ubini and Hubbard 2003; Ghio et al. 1992). Their role in
equine pulmonary disease is presently unknown. The actual exposures to silica and iron
could not be determined in the present study because our sampling equipment did not
distinguish between the various components of dust. Occupational exposure limits for
iron and crystalline silica are signiﬁcantly lower than the established guidelines for
nuisance dust, which is what was measured in this study and is our reference point.
Future studies should include personal sampling speciﬁcally for components of inorganic
dusts such as those mentioned above in order to begin to estimate exposures over 3

represent period of time as dictated by established sampling methods.

Ammonia is a well-known respiratory irritant (Wang et al. 1996). Before stall
cleaning, concentrations of ammonia in the stall hovered near occupational exposure
levels with some sample concentrations exceeding the 15-minute short-term exposure
limits for humans (ACGIH 2005c). Ammonia concentrations in summer were
signiﬁcantly higher than those observed in the winter, even though ventilation in summer
was superior to that in winter when all stable doors were closed. This can be explained

by the increased volatilization of ammonia during the warmer summer months. During

32

summer when stable doors were opened, ammonia concentrations increased in stalls

towards the center of the stable.

Fortunately, implementing wet methods to reduce particulate exposures is easy
and inexpensive. In all cases, exposures during feeding, cleaning, and riding were
dramatically reduced aﬂer wetting. In arena footing, water content of 10 percent or
higher provides maximal dust suppression. The commercial dust suppressant was as
effective as water in reducing particulate concentrations. Wetting (not soaking) of
feedstuffs immediately prior to initiation of feeding reduces exposures to that of

background levels.

In conclusion, our investigation has reinforced the fact that particulate concentrations can
be very high in the breathing zone of horses eating hay and other dry feeds (Vandenput et
al. 1997; Woods et al. 1993). To decrease exposures to concentrations of particles and
gases in the microenvironment of the stall, it is important to improve ventilation within
the stable as much as possible and to use clean feeds and dust suppression techniques in
order to reduce the potential for particle generation and dispersion. A little water goes a
long way in suppressing dust. Unfortunately, stable and horse owners tend to rely on
their own senses to evaluate whether or not a stall is well ventilated and dust free.
Decisions made from the aisle do not reﬂect what is going on in the breathing zone of the
horse. Use of real-time particulate monitors is useﬁrl to educate stable owners about the

inhaled hazards present in a stable.

33

 

Table 1-1. Metal (mg/kg) and crystalline Silica (% of total) concentration in bulk samples

taken from the arena footing and wall ledges.

 

 

 

 

34

Footing
concentration Wall ledge concentration
Metals (mgﬁg) (mglkg)
Silver <99.4
Cadmium <5
Cobalt <15
Chromium <50
Copper 11
Iron 3800 11000
Manganese 160
Nickel <9.9
Lead <25
Zinc <248
Footing
concentration (% of Wall ledge concentration
Crystalline Silica total) (% of total)
Quartz 22 22
Cristobalite <1 <1
Tridymite <1 <1

 

Table 1-2. Summer and winter ammonia concentrations (ppm) in each of seven stalls.

 

 

 

 

 

 

 

 

 

 

 

Stall number Ammonia
Winter Summer

1 13 41

2 20 47

3 4 59

4 10 30

5 2 30

6 9 10

7 11 20
MeaniSD 9.1i5.3 31:1:19

 

 

 

 

35

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Storage Bulk Observation Room
Sawdust
Storage II 7
6
5
4
Riding ,
Arsle
Arena
_
L—
3
2
1

 

 

 

 

 

 

 

Figure 1-1. Diagram of the boarding stable Showing the location of the stalls in which

ammonia concentrations were measured.

36

mg/m3
on
O

160 -
140 r
120 -
100 -

mg/m3
on
O

60’
40-
20-

 

120 [

PM2.5 ave.
0 _

1..

PM2.5 max

 

Dry hay Alfalfa Complete Senior Sweet

cubes pellets

ll

feed

J;

feed

mg/m3

mg/ma

160

160
140

120 '
100 -
80 -
60 -
40 -
20 -

PM10 ave.

PM10 max
,

L

 

i

Dry hay Alfalfa Complete Senior Sweet
cubes pellets feed feed

Figure 1-2. Breathing zone concentration (mg/m3) of PM10 and PM2_5 during consumption

of various dry feedstuffs. Median values and range are shown for both average and peak

particle concentration. Values with similar superscripts do not differ signiﬁcantly.

Removing cobwebs
100 -

80-

mg/m3

20

l
—I-

 

 

Aisle sweeping
100 P

80-
60- l
40+

20-
o L

mg/m3

 

Cleaning and bedding stall with sawdust
100 -
80 -
6O -
40
20 -
0 r i ' I I 1 a '
PM10 PM10 PM2.5 PM2.5
ave. max. ave. max.

I

mg/m3

 

 

Figure 1-3. Breathing zone concentration (mg/m3) of PM10 and PM2.5 during routine
stable activities. Median values and range are shown for both average and peak particle

concentration.

38

 

60—

mg/m3

 

 

60—

mg/m3

 

 

0 L .... 1W_-ll

l l I
0 30 60

mins.

 

Figure 1-4. PM10 (top) and PM2.5 (bottom) concentrations (mg/m3) in the riding arena
during a single riding event. Aerosol photometer samples were recorded every second for

the duration of the activity.

39

 

 

 

 

p PM10
100 -rag ‘00
10 P 10 -
E 1 l E 1
a O)
E 0.1 - I I I I ' E
0.01 . 0.01 r
0.001 0.001
Riding PM10
100 r 100
m 10 - m 10 -
E 1 - l E 1 -
or , or .
E 0.1 E 0.1
0.01 - 0.01 -
0.001 1 ‘ 1 J L ‘ 0.001
Dry 6.1 9.3 11.2 12.8 14.2
Percent moisture

0.1 r

Drag PM2.5

 

 

 

 

,
U I I I T
Riding PM2.5
- l l l 1
Dry 6.1 9.3 11.2 12.8 14.2
Percent moisture

Figure 1-5. Effect of moisture content on breathing zone concentration (mg/m3) of PM 10

(left) and PM2.5 (right) during arena grading (upper) and riding (lower). Values are

maximum, minimum and mean.

40

  

 

 

 

;\3
Q)
E
2
O
> will
0 r 'l I l .quTﬂPIIIITIT'l'Ii'TEIlllll,III
1 1 1o
Particle diameter (pm)
5-
g .
a) _
E
2
g -
0 r W?“ 1 . I f
1 1000

 

Particle diameter (pm)

Figure 1-6. Distribution of particle diameters (micron) in bulk samples taken ﬁom the

arena footing (top) and wall ledges (bottom).

41

References

ACGIH (2005a) American Conference of Governmental Industrial Hygienists TLVS and
BEIS. Threshold limits values for chemical substances and physical agents and biological
exposure indices., Cincinnati, Ohio.

ACGIH (2005b) Documentation of threshold limits values: Ammonia. In: Supplement to
the 7th edition. ACGIH Worldwide Signature Publications.

Anon (2001) Chairperson's introduction: International Workshop on Equine Chronic
Airway Disease, Michigan State University, 16-18 June 2000. Equine Vet. J. 33, 5-19.

Anon (2003) Workshop report: Inﬂammatory airway disease: deﬁning the syndrome.
Equine Vet. Educ. 5, 81-82.

Anon (2005) T81 Model 8520 DustTrak Aerosol Monitor. Operation and service manual.
1980198 revision Q.

Anon (2006) National ambient air quality standards, US Environmental Protection
Agency.

Crichlow, E.C., Yoshida, K. and Wallace, K. (1980) Dust levels in a riding stable. Equine
Vet. J. 12, 185-188.

Derksen, F.J., Robinson, N.E., Armstrong, P.J., A., SJ and Slocombe, RF. (1985)
Airway reactivity in ponies with recurrent airway obstruction (heaves). J. Appl. Physiol.
58, 598-604.

Dye, J.A., Adler, K.B., Richards, J .H. and Dreher, KL. (1999) Role of soluble metals in
oil ﬂy ash-induced airway epithelial injury and cytokine gene expression. Am. J.
Physiol.: Lung Cell. Mol. Physiol. 277, L498-510.

Fubini, B. and Hubbard, A. (2003) Reactive oxygen species (ROS) and reactive nitrogen
species (RN 8) generation by silica in inﬂammation and ﬁbrosis. Free Radic Biol Med 34,
1507-1 5 16.

Galson (2006) Sampling and analysis guide.

Ghio, A.J., Kennedy, T.P., Whorton, A.R., Crumbliss, A.L., Hatch, GE. and Hoidal, JR.

(1992) Role of surface complexed iron in oxidant generation and lung inﬂammation
induced by silicates. Am. J. Physiol.: Lung Cell. Mol. Physiol. 263, L51 1-518.

42

 

McGorum, B.C., Ellison, J. and Cullen, RT. (1998) Total and respirable airborne dust
endotoxin concentrations in three equine management systems. Equine Vet. J. 30, 430-
434.

Oberdorster, G. (1996) Signiﬁcance of particle parameters in the evaluation of exposure-
dose-response relationships of inhaled particles. Inhal Tox 8 Suppl, 73-89.

Pirie, R.S., Collie, D.D., Dixon, PM. and McGorum, BC. (2003) Inhaled endotoxin and
organic dust particulates have synergistic proinﬂammatory effects in equine heaves
(organic dust-induced asthma). Clin. Exp. Allergy 33, 676-683.

Pope, C.A., 3rd (1991) Respiratory hospital admissions associated with PM10 pollution
in Utah, Salt Lake, and Cache Valleys. Arch Environ Health 46, 90-97.

Pope, C.A., 3rd, Burnett, R.T., Thun, M.J., Calle, E.E., Krewski, D., Ito, K. and Thurston,
GD. (2002) Lung cancer, cardiopulmonary mortality, and long-term exposure to ﬁne
particulate air pollution. JAMA 287, 1132-1141.

Vandenput, 8., Istasse, L., Nicks, B. and Lekeux, P. (1997) Airborne dust and
aeroallergen concentrations in different sources of feed and bedding for horses. Vet.
Quarterly 19, 154-158.

Wang, A.L., Blackford, TL. and Lee, L.Y. (1996) Vagal bronchopulmonary C-ﬁbers and
acute ventilatory response to inhaled irritants. Respir. Physiol. 104, 231-239.

Webster, A.J.F., Clarke, A.F., Madelin, TM. and Wathes, CM. (1987) Air hygiene in
stables 1: Effects of stable design, ventilation and management on the concentration of
respirable dust. Equine Vet. J. 19, 448-453.

Woods, P.S., Robinson, N.E., Swanson, M.C., Reed, C.E., Broadstone, RV. and

Derksen, F.J. (1993) Airborne dust and aeroallergen concentration in a horse stable under
two different management systems. Equine Vet. J. 25, 208-213.

43

 

CHAPTER 2
Evaluation of Factors Affecting the Concentration of Particulate Matter (PM10 and

PM2.5) in Stables at an American Thoroughbred Racetrack

Summary
Reasons for Performing Study

Inﬂammatory Airway Disease (IAD) is common in Thoroughbred Racehorses
with an estimated overall prevalence of 13%, and is most likely multifactorial in nature.
Particles present in the environment of the racehorse are in size ranges known to result in
worsening of pre-existing respiratory disease in humans, and compromised ability to

handle streptococcal infections in laboratory animals.

Objectives
The purpose of this study was to determine the size ranges and concentrations of
airborne small particles in the stable environment of the racehorse, and the factors that

inﬂuence their presence.

Methods

Direct reading instruments were utilized to determine the concentration and
factors affecting the concentration of particulate matter (PM10 and PM2.5) in three
stables at an American Thoroughbred Racetrack. Particulate mapping took place three

times daily (early morning, midday, late afternoon) during each of three months (early,

44

mid, and late racing season) corresponding with times of varying activity within the

stable, and changes in racing activity and season.

Results

There was a signiﬁcant effect of month, time of day, and stable on concentrations
of both PM10 and PM2.5. Average PM10 and PM2.5 concentrations were highest in
September following a period of dry weather, and lowest in July following a period of
rainy weather. November concentrations of PM10 and PM2.5 were slightly lower than
those of September even though monitoring took place during a period of snowy/wet
weather. Early morning concentrations of PM were signiﬁcantly higher (P<0.0001) than
those measured throughout the rest of the day. The stable that was completely enclosed
and relatively devoid of natural sources of ventilation, consistently had signiﬁcantly
higher average concentrations of both PM10 and PM2.5 than the open-Sided stable.
There was a signiﬁcant difference in number of stalls within a stable that exceeded the

75th percentile for both PM10 and PM2.5.

Conclusions

We report signiﬁcant effects of month, time of day, stable, and stalls within a
stable on particle concentrations. PM10 and PM2.5 concentrations were lowest overall in
the month following a period of rain, and highest in the morning hours during peak
activity. The stable which was most enclosed, fed from hay nets, and had the most

activity had signiﬁcantly higher average PM10 and PM2.5 concentrations than the stable

45

with open sides and less activity. Speciﬁc stalls within stables have consistently elevated

concentrations of PM compared to adjacent stalls.

Potential relevance
Understanding sources and factors inﬂuencing concentrations of particulate
matter in the stables of racehorses will allow for the development of strategies aimed at

reduction of exposures and thus potentially a reduction in the prevalence of IAD.

Introduction

Inﬂammatory airway disease (IAD) of racehorses is a syndrome characterized by
cough, exercise intolerance, excess tracheobronchial mucus that is visible upon
endoscopic examination, and inﬂammatory cells in tracheal aspirates and
bronchoalveolar lavage (BAL) (Anon 2003). There is an overall prevalence of IAD of
approximately 13% of Thoroughbreds in training (Wood et al. 2005c), with up to 45% of
racehorses in a stable affected at any one time (unpublished data) leading to a signiﬁcant
ﬁnancial impact as relatively small amounts of mucus in the airways result in horses
ﬁnishing further back in a race (Holcombe et al. 2006; MacNamara et al. 1990b). Due to
the potential impact of IAD on the racing industry, investigators have focused on

elucidating its cause, which is most likely multifactorial in nature.

IAD is common in young racehorses, with incidence declining as the horse ages

and length of time in training increases, suggesting an infectious component, as

increasing time spent in a stable and exposure to bacterial and viral pathogens would be

46

expected to confer immunity (Wood et al. 2005c). As a result, several investigations
have focused on the role of infectious agents in the etiology of IAD. Bacterial, but not
viral infections have been associated with neutrophilic inﬂammation and mucus
accumulation in asymptomatic racehorse populations (Chapman et al. 2000; Christley et
al. 2001b; Wood et al. 1993; Wood et al. 2005a, c). Infections, however, most likely are
not the only cause of IAD in racehorses, as many racehorses (over 20%) with airway
inﬂammation culture negative (Chapman et al. 2000; Christley et al. 2001 a; Wood et al.
2005a). The stable environment has been identiﬁed as a second risk factor for airway
inﬂammation in thoroughbred racehorses (Malikides and Hodgson 2005), as they are kept
primarily outdoors until, when at around the age of two, they are brought into a training
facility or racetrack to begin their careers. Exposure to organic dusts commonly found in
stables, dispersed from feed and bedding, have been shown to be associated with airway
neutrophilia and mucus accumulation in healthy horses (Gerber V 2004; Holcombe et al.
2001; Tremblay et al. 1993). For these reasons, our attention has turned to the evaluation
Of exposures to particulate matter (dust) in the etiology of IAD in Thoroughbred

racehorses.

Airborne particulate matter (dust) in racehorse stables can originate from feed and
bedding, ﬂooring materials, and racetrack footing materials. As there are many sources
of particulate matter in the environment of the racehorse, size and composition of the
particulate matter would not be expected to be uniform. Particle Sizes of interest in this
study are classiﬁed as those less than 10 (PM10) and less than 2.5 (PM2.5) microns in

diameter. In htunans, a small percentage of PM10 have the potential to reach the lower

47

airways but the majority of these particles are deposited in the central airways such as the

trachea. The smaller particles represented by PM2.5 penetrate deep within the airways.

Previous studies conducted in controlled settings as well as private boarding
stables utilizing traditional ﬁlter/pump methods (Clarke and Madelin 1987; Clarke et al.
1987; Crichlow et al. 1980; Kirschvink et al. 2002; McGorum et al. 1998; Woods et al.
1993) and direct reading instruments (Clements and Pirie 2007 a, b) have shown that dusts
present in the breathing zone of horses are composed of particles that include those in
size ranges (PM10 and PM2.5) consistent with those described in the human literature.
When airborne concentrations of particles in these size ranges increase, there is
worsening of symptoms in people with pre-existing airway disease, increased lost
school/work days, increased hospital admissions, and increased mortality (Brunekreef
and Forsberg 2005; Oberdorster G 1996; Peel et al. 2005; Pope 1991a; Pope HI et al.

2002; Schwartz 2004).

The current study is part of a larger study that aims to describe the relationship
between exposure to particulate matter and the presence of IAD in Thoroughbred racehorses
at an American Thoroughbred Racetrack. The purpose of this ﬁrst study is to characterize

area particulate concentrations and the factors that inﬂuence their presence.

48

Materials and Methods
Stables

This study was conducted at a Thoroughbred racetrack in the Midwestern United
States, where there are stables of various ages and styles. The three stables used in this
study were selected because of their different designs and management styles, as well as

the cooperation of the trainers that use them.

Stable l was of newer construction and design Orientation of the stable was north-
south with roll-up shutters along the entire length (east and west sides) of the stable in
combination with high vaulted ceilings, large sliding doors on both ends (front and back),
and open fronted stalls (bars and gates) that face the outdoors. This stable is located

immediately adjacent to the main roadway and a common parking area west of the stable.

Stable 2 is of brick construction and the long axis of the stable is oriented in the
north-south direction. This stable has 3 meter ceilings, closed-front stalls, and small,
high windows (only in stalls that are on outer walls) that are kept closed. A single row of
stalls is located along each of the outer walls of the stable, with a double row (back-to-
back) of stalls down the center. The main source of ventilation is from two sets of large
doorways both at the north and south ends of the stable, and one set of smaller doorways
in the center of the west side of the stable. A large manure handling facility is located 6
meters from the north end of the stable, with another stable of similar design located a
similar distance from the south end. Stable management attempts to improve ventilation

in the summer months by placing large air-moving fans in each end doorway. This is the

49

only stable in the study that feeds from hay nets. This stable is also located immediately
adjacent to a busy racetrack road along the length of the western wall, and a busy city

road along the east side of the stable.

Stable 3 is of identical construction to Stable 2, however some differences do
exist. The stable is oriented with its long axis east to west, and has open windows during
the warm months. This stable is located in a secluded area (away from roads), but is

adjacent to a common parking area along its western end.

Measurement of particle concentration and numbers

Concentration and number of area airborne particles were evaluated using two TSI
DustTrakl monitors and one MetOne HHPC-6 Airborne Particle Counterl, respectively.
These machines are small and portable and use optical methods to quantify particle
concentrations and numbers. One DustTrak was set to measure the concentration (mg/m3 of
air) of particles less than 10 microns and the other to measure particles less than 2.5
microns. The particle counter provided the number of particles within a given size range:
0.5, 0.7, 1.0, 2.0, 5.0 and 10.0 microns. Equipment calibration and cleaning was performed
prior to the start of each day of sampling per manufacturers instructions, and periodic ﬂow

evaluations were made throughout each sampling day.

All measurements were one minute in duration, taken at the front-center of each stall

1 meter above the ground, approximating the height of the nostril of the horse. In an attempt

 

I TSI Incorporated, 500 Cardigan Road, Shoreview, MN 55126-3996

2 Hach Ultra, 481 California Ave., Grants Pass, OR 97526

50

to minimize our inﬂuence in the stable, horses remained in the stalls during monitoring, and

all activities (cleaning, grooming, feeding, etc.) took place as scheduled.

Experimental Design

Three visits were made to the racetrack over the 2005 racing season (June,
September, and November), staying 3 days at each visit. The visits were selected to be a)
early in the racing season, b) at the height of the season when conditions were usually
driest, and c) when the weather was cooler and damp and the new stable (Stable 1) had
closed its outer Shutters. The weather (rain/snow), temperature, and humidity were
recorded during our visit. Information on total rainfall for one week prior to each visit

was purchased from Weather Source, LLC.

We measured particulate concentrations in one stable each day during the visit.
Concentrations were measured for one minute in every stall three times of day - a) early
morning during clean out, feeding and grooming, b) mid-day when there was less

activity, and c) late afternoon which coincided with feeding and racing.

Data Analysis

Particle concentrations measured in each stall within each stable were analyzed
using Wilcoxon rank sums and the Kruskal-Wallis test of signiﬁcance to determine if
there were regional pockets of high particulate concentrations. Maps were prepared to
show the stalls within each stable that exceeded the 50th and 75th percentile of

particulate concentrations overall. PROC CROSSTABS was used to determine

51

signiﬁcant differences in the number of stalls in each stable with PM concentrations
greater than the 75th percentile overall. Mixed models were employed in SAS v.9.l.
(PROC MIXED). This procedure is designed to handle hierarchical data that is collected
repeatedly on stalls over time. In addition, it is able to handle missing data and has the
power to analyze a number of variance covariance structures while accounting for within-
stall correlations. Then, the possible covariance structure including unstructured (UN),
spatial Gaussian (SP(GAU)), ﬁrst order autoregressive (AR(1)), TOEPLITZ (TOEP), and
compound symmetry (C8) were tested. The best covariance structure was chosen based
on the Akaike Information Criterion, which was TOEPLITZ for all models. Using PROC
MD(ED, the effects of month, stable, and time of day on particle concentrations (PM10
and PM2.5) within each stable, and between stables were determined. Level of

signiﬁcance was set at P5005.

Results

A total of 717 measurements were taken throughout the sampling season (Table
2-1), with each consisting of average, minimum, and maximum values recorded for
PM10 and PM2.5 during the one minute measurement. The average PM10 and PM2.5
concentrations ranged between 0.078 and 4.090 mg/m3 and 0.030 and 1.170 mg/m3,
respectively. Peak PM10 and PM2.5 concentrations ranged from 0.124 to 12.800 and
0.052 to 4.16 mg/m3, respectively. The median PM10 and PM2.5 average, minimum, and
maximum concentrations for all samples are given in Table 2-1. While the larger

particles (PM5.0 _<_PM10) comprised the majority of the mass of the measurements that

52

were taken, those particles smaller than 5 microns in diameter (PM 5.0) made up the

largest percentage of particles (84-99%) (data not shown).

Effect of Month and Time of Day

Average particle concentrations (PM10 and PM 2.5) were signiﬁcantly (P <
0.0001) affected by month (Figure 2-1) with highest concentrations occurring in
September and the lowest in July. PM2.5 minimum concentrations were signiﬁcantly

lower in November than in July.

Average concentrations of PM10 and PM2.5 changed signiﬁcantly (P < 0.0001)
throughout the day (Figure 2-2). Concentrations were highest in the morning during
cleanout and decreased by the midday and late aﬂemoon sampling periods. Average
concentrations of PM10 and PM 2.5 were signiﬁcantly different between the morning
and late afternoon (reference measurement) (P<0.0001), but were not signiﬁcantly
different between the midday and late afternoon concentrations. This is in contrast to the
minimum PM2.5 concentrations in which both the morning and midday measurements

were signiﬁcantly different than those taken during the late afternoon (P<0.0001).

Effect of Stable

There was a highly signiﬁcant (P < 0.0001) effect of stable on PM10 and PM2.5
concentrations (Figure 2-3) with Stable 2 having the highest concentrations of both PM10
and PM 2.5. However, stable 1 had signiﬁcantly higher concentrations of average and

minimum PM2.5 than Stable 3 (referent stable). Figure 2-4 shows the interaction

53

between stable and month. While PM10 and PM2.5 average concentrations are always
higher in Stable 2 than Stable 1, overall trends are identical with average concentrations
increasing from July to September, and decreasing slightly in November. Stable 3 did
not follow a similar trend. By contrast, trends through the time of day were similar in all

stables (data not shown).

Effect of Stall Location

There was a signiﬁcant (P<0.001) effect of location of a stall within a stable on
PM10 average concentrations. Signiﬁcant effects of stall location also existed for PM10
(P<0.0005) and PM2.5 maxima (P<0.0008) and PM2.5 minimum (P<0.005)
concentrations. With regard to PM10 and PM2.5 average measurements, Stable 2
contained Signiﬁcantly more stalls in the 75th percentile than Stables 1 and 3 (Table 2-2,
Figures 2-5 and 2-6). In general, Stable 2 consistently had the highest number of stalls
which exceeded both the 50th and 75th percentiles for both PM10 and PM2.5; however, in
the case of PM 2.5 min, all the stalls of Stable 1 exceeded the 50th percentile (Table 2-2).
Stalls exceeding the 50th percentile for both PM10 and PM2.5 average concentrations

were consistently located along the west aisle of Stable 1 (Figures 2-5 and 2-6).

Discussion

These are the ﬁrst reported measurements of indoor air quality (real-time particle
concentration) in American racing stables. Our measurements indicate that there are
signiﬁcant effects of time of day, month, stable, and location of stalls within a stable on

particulate matter in sizes and concentrations of interest in terms of airway inﬂammation

54

in both humans and laboratory animals (Becker et al. 1996; Ghio et al. 2006; Gilmour et

al. 1996; Kappos et al. 2004; Li et al. 1997; Peel et al. 2005; Zelikoff et al. 2003).

Particle concentrations and numbers increased from July to September, and then
decreased slightly during November. These observations are consistent with weather
patterns, ventilation status, and numbers of horses in the stables. While Stable 2
consistently had higher monthly PM10 and PM2.5 average concentrations than Stable l,
particulate concentrations both followed the same increasing pattern from July to
September, and then experienced a slight decrease in November, suggesting a component
aside from stable construction/orientation and management style that affects particle
concentrations (e. g. rainfall, humidity, etc.). While it rained just prior to our visit in July,
and snowed during our visit in November, it was very dry just prior to and during our
visit in September. Wet/humid weather attenuated particle concentrations in July, but not
to the same degree in November, as the stable doors and windows were closed tight due
to the cold weather, thus eliminating all sources of natural dilution ventilation. Particle
concentrations in Stable 3 did not follow a Similar pattern, presumably because the trainer
regularly shipped horses in and out of the stable, altering which stalls were occupied.
within. Furthermore, this stable was the only one of the three oriented east-west and
away from major roadways, so external inﬂuences may have played a smaller role than in

the other two stables.

As might be expected, the new construction stable (Stable l) with roll-up (open-

air) sides and vaulted ceilings had signiﬁcantly lower concentrations of large particles

55

(PM10) throughout the day than the older brick style building with low ceilings, closed
windows, and hay fed from hay nets rather than the ﬂoor (Stable 2). Not surprisingly, all
particles were at their highest concentration during the morning measurement period
when cleaning, feeding, and preparation for exercise were occurring. The average
concentration of PM10 and PM 2.5 fell signiﬁcantly by mid-day, but only PM 2.5
minimum values continued to decrease signiﬁcantly throughout the remainder of the day.
The rapid decrease of PM10 can be explained by the higher rate of settling of the larger

particles. In contrast, the smaller and lighter PM 2.5 settles out more slowly.

In addition to particles (PM10 and PM2.5) generated and dispersed within the
stable due to activity (feeding, cleaning, etc.), factors external to the stable appear to
inﬂuence particle concentrations within the stable environment. Stable 1 can illustrate
this point with its open sides and high vaulted ceilings that allow for complete mixing
with outside air. In this stable all 30 stalls had median PM2.5 minimum values meeting or
exceeding the 50th percentile (Table 2-2). If all measured PM2.5 were generated indoors,
Stable 2 rather than Stable 1 would have had the greatest number of such stalls as it does
for all other PM values. This is not the case, and therefore, other factors must be affecting
the small particles. The consistently elevated levels of minimum PM 2.5 in Stable 1 are
consistent with its location in proximity to a busy roadway and the main parking area, as

small particles can originate from road dust (Gillies et al. 2001; Singh et al. 2006).

We observed a signiﬁcant variation in management style between stables that

probably explained particle concentrations throughout the day with Stable 2 consistently

56

having high particle concentrations. Stables l and 3 were busy in the morning hours
during feeding/cleanout activities and preparations for training, and then quiet throughout
the rest of the day. In these two stables, activity began daily between 5 and 5:30 am,
and was usually at a minimal level by around 9 am. Both stables fed grain from feeders
suspended in the corner of the stall, and hay was fed from the ﬂoor. In each of these
stables, no more than 3 individuals took part in the morning activities (cleaning,
grooming). In contrast, Stable 2 assigned one groom per 4-5 horses. This allowed for
simultaneous cleaning/ grooming in a minimum of six stalls at any one time — at least
double that of the other two stables, resulting in greater dispersion of particles. Stable 2
also fed grain from a comer feeder, but fed hay from hay nets that were ﬁlled each
morning by the grooms. Hay for each net was taken from the bale and shaken in the aisle
to loosen it before it was packed into hay nets and hung near the doorway of each stall.

In order to eat from the hay net, horses pulled the hay-ﬁlled net into the doorway, which
both released particles and obstructed ventilation of the stall. Aisle raking took place
continuously throughout the morning until the training session ended and all horses were
returned to their stalls. Activity in Stable 3 decreased by 11 am. daily, approximately 2

hours later than in Stables l& 2.

Our data indicate a signiﬁcant effect of stall location within the stable on PM
concentrations. Stable 2 consistently had stalls that exceeded the 75th percentile in terms
of PM10 and PM2.5 average and maximum concentrations. These stalls were located
along the center partition, an area furthest from doorways and windows, and likely

devoid of natural mixing/dilution.

57

In humans and laboratory animals, inhalation of PM10 typically results in a
neutrophil inﬂux into the lung (Li et al. 1997) and upregulation of pro-inﬂammatory
cytokines (Becker et al. 1996). The pathogenesis of PMI 0 damage has been related to their
metal content, which leads to oxidative stress (Gihnour et al. 1996). PM2.5, because of their
large surface area per unit mass, provide a large capacity to generate free radicals and are
inherently more damaging than coarse particles (Oberdorster et al. 2000). Exposing rats to
PM 2.5, in concentrations observed at this racetrack, has been shown to compromise the
ability of the lung to handle streptococcal infections (Zelikoff et al. 2003). Compounds such
as endotoxin attached to the particle surface also may be responsible for particle-induced
injury or may prime the airway for particle-induced inﬂammation (O'Grady et al. 2001;
Oberdorster et al. 2000). Increases in ambient PM2.5 are consistently associated with
increased morbidity and mortality from respiratory disease (Kappos et al. 2004), while
increases in ambient PM10 tend to be linked to short-term transient worsening of disease

(Brunekreef and F orsberg 2005).

A required ﬁrst step in any thorough Indoor Air Quality (IAQ) assessment of
facilities not previously evaluated is a survey of the building environment. This ‘survey’
typically consists of identiﬁcation of building construction, type(s) of ventilation present
(forced or dilution), location(s) and characteristics of ventilation sources, identiﬁcation of
structural components (walls, beams, etc.) that inﬂuence airﬂow within the building, and
identiﬁcation of potential indoor air contaminants and factors inﬂuencing their presence.

This information is then used to develop process controls to reduce exposures, exposure

58

risk assessments, and targeted air sampling plans for future contaminant speciﬁc personal

exposure monitoring.

We have conducted a thorough survey of particle concentrations (PM10 and
PM2.5) and factors that affect them in 3 stables at an American Thoroughbred racetrack
which follows the rationale mentioned above for conducting IAQ assessments. We have
identiﬁed multiple factors that inﬂuence particulate concentrations within these stables,
and have shown that particle concentrations consistently vary between stables, by season,
time of day, and location of the stable with regard to neighboring roads. Additionally, we
have identiﬁed regions (stalls) within stables that consistently experience higher
concentrations of PM, thus allowing for the identiﬁcation of horses potentially at

increased risk for airway inﬂammation.

59

see: n B:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

38 48.0 wood 33 Rod 08.0 :2 :2. ~86 wad Sod Sod Geneva: 22m
See 0:3 83 S2 god SS :3 Sod $3 82 89o wtd AneuaEESEA
ONE Sod :3 cm: mood $2 one.“ sec 34.0 83 was $3 Geneva: 22a
ow: $3 $2 33 23 $2 3: ﬁg 3.3 22 33 22 ABBEVRESE
83 $3 2.3 33 ~36 Sod ~28 Sod SS 83 ~86 5d mania? 2 SE
a: an; :2 cm: SS 83 83 $3 33 a: 23 Ed Anamavwéeza

Ba 53 2% _E as: Em $2 2% 5m ES. Ba .23 in in 32 Eu 5: Eu .3 BE 38. EVER
Q: .30 Eat 2: So as: 38 .30 39: SR .80 as:

in: 5 835 =< 2.5” as n «saw 2:: 5 N Beam Ens c9 _ 23m

 

 

AEEEORE Ea .SEEEE .omﬁozwv WNSE 98 SSE com moi? 2:583 5mm

98 .em .5608 08 @6203 .0735 :08 ESE meoumtomno mo BEBE H88 Ba 23% Co Conga mo Suﬁsm .TN 2an

6O

Table 2-2. Number of stalls within each stable that exceed the 50‘h and 75th percentile for

PM10 and PM2.5 (Average, Minimum, and Maximum).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Stable 1 (30 Stable 2 (31 Stable 3 (20 All Stables (81
stalls) stalls) stalls) stalls)

PM Values >5 0th >75th >5 0th >75th >5 0th >75th >5 0th >751}:
LCM“ M all M 2211 EL“ M M! Bell
percentiles

{mg/m3)

PM 10 Avg

(0331/0556) 10 0 26 7 11 2 47 9
PM 2.5 Avg
Q.122/0.229L l3 0 23 3 9 0 45 3
PM 10 Max

(0582/0977) 8 l 23 6 l4 2 45 9
PM 2.5 Max

0351/0636) 6 0 26 8 12 2 44 10
PM 10 Min

0196/0298) 6 l 27 6 8 1 41 8
PM 2.5 Min

0068/0158) 30 1 5 0 5 O 40 1

 

 

 

 

 

 

 

61

N
or

       

E * *
B 2‘
E
C l *
31.5‘ *
E l *
a l -
8 l
s 1'
U l * *
2 l * * *
€0.5-‘
I: l 'l |*
0+3 3'! A!” rune ,_ H

PM10 Ave PM2.5 Ave PM 10 Min PM2.5 Min PM 10 Max PM 2.5.Max

Figure 2-1. Effect of month on mean particle concentrations (PM10 and PM2.5) in
mg/m3 . Light gray bars denote July, dark gray bars denote September, and black-striped
bars indicate November measurement values. Error bars indicate the 25th and 75‘h

percentile. An * denotes signiﬁcant difference from referent month (July).

62

.N
o:

_x
01
l _,l—,_L—4A_l

(mo/m ’)
N

Partigle Concentration
in _- '

il-

*

in. II

t

PM10 Ave

PM2.5 Ave PM10 Min PM2.5 Min PM10 Max

 

   

SJ
PM2.5 Max

Figure 2-2. Effect of time of day on mean particle concentrations (PM10 and PM2.5) in

mg/m3. Light gray bars denote early morning, dark gray bars denote mid-day, and black-

striped bars indicate late afternoon measurement values. Late aftemoon measurements

were set as the referent group. Error bars indicate the 25th and 75th percentile. An *

denotes signiﬁcant difference from referent time of day (late afternoon).

Particle Concentration (mg/m 3)

 

PM10 Ave PM2.5 Ave PM10 Min PM2.5 Min PM10 Max PM2.5 Max

Figure 2-3. Effect of stable on mean particle concentrations (PM10 and PM2.5) in
mg/m3. Light gray bars denote Stable 1, dark gray bars denote Stable 2, and black-striped
bars indicate Stable 3. Error bars indicate the 25th and 75th percentile. An * denotes

signiﬁcant difference from referent stable (stable 3).

64

—L
on

ma)
_l
'o>
#4

_x _5
N -§
LLl 4 L 4144

9.0.0.0
ON-bODCD—I

   

Particle Concentration (mgl

       

as l i 7 .. ,- ..
PM10Ave l PM2.5Ave } PM10 Ave PM2.5 Ave ‘ PM10Ave I PM2.5Ave 1

Stable 1 1 Stable 2 Stable 3 ‘

Figure 2-4. Effect of stable/month interaction on mean particle concentrations (PM10
and PM2.5) in mg/m3 . Light gray bars denote July, dark gray bars denote September, and
black-striped bars indicate November measurement periods. Stable 3 and the month of
July were used as the referent group. Error bars indicate the 25th and 75‘h percentile. An

* denotes signiﬁcant difference from referent groups.

65

Stable 1 North Stable 2

 

orlloe BR TR GR

0.392 0.236
0.436 0.332
0.339 0.213
0.335 0.246
0.333 0.25
0.533 0.21

0.223 0.193
0.253 0.137
0.293 0.204
0.271 0.32

05331 0.313
0.255 0.279
0.217 0.292
0.174 0.352
0.339 0.275

 

 

 

' ’ ' ' 50th Percentile
75m Percentile
GR Grain Room
TR Tack Room
Ofﬁce Ofﬁce
WR Wash Rack
%ZZ§ZZ Not Sampled

 

 

 

Figure 2-5. Map Showing the locations of stalls within each stable that exceed the 50th
and 75th percentiles for PM10 Average values. Note that almost all of the stalls in Stable
2 exceed the 50th percentile, while less than half of all stalls in both Stables 1&3 exceed
the 50th percentile. Also, Stable 2 contains the majority of the stalls that exceed the 75th
percentile while Stable 1 does not contain any. The ‘North’ notation located adjacent to

each diagram denotes stable orientation.

66

Stable 1 North Stable 2
R 0 089

 

 

 

 

 

 

Office B TR GR
0.102 0.1
0.19 0.121
0.204 0.096
0.185 0.103
0.149 0.101
0.143 0.098
0.135 03:13
0.132 0.099
0.12 0.1
0.145 0.108
0.148 0.107
0.128 0.105
0.102 0.104 Stable 3
0.12 0.115
0.141 0.105
' Ste 5001 Percentile
751h Percentile
GR Grain Room
TR Tack Room North
Office Ofﬁce
WR Wash Rack
221%??? Not Sampled
BR Bathroom

 

 

 

Figure 2-6. Map Showing the locations of stalls within each stable that exceed the 50th
and 75th percentiles for PM2.5 Average values. Note that almost all of the stalls in Stable
2 exceed the 50th percentile, while less than half of all stalls in both Stables 1&3 exceed
the 50th percentile. Also, Stable 2 contains the majority of the stalls that exceed the 75th
percentile while Stable 1 does not contain any. The ‘North’ notation located adjacent to

each diagram denotes stable orientation.

67

References

Anon (2003) Workshop report: Inﬂammatory airway disease: deﬁning the syndrome.
Equine veterinary education 5, 81-82.

Becker, 8., Soukup, J .M., Gilmour, M.I. and Devlin, RB. (1996) Stimulation of human
and rat alveolar macrophages by urban air particulates: effects on oxidant radical
generation and cytokine production. Toxicology and Applied Pharmacology 141, 63 7-
648.

Brunekreef, B. and Forsberg, B. (2005) Epidemiological evidence of effects of coarse
airborne particles on health. European Respiratory Journal 26, 309-318.

Chapman, P.8., Green, C., Main, J .P., Taylor, P.M., Cunningham, F.M., Cook, A.J. and
Marr, CM. (2000) Retrospective study of the relationships between age, inﬂammation
and the isolation of bacteria from the lower respiratory tract of thoroughbred horses.
Veterinary Record 146, 91-95.

Christley, R.M., Hodgson, D.R., Rose, R.J., Hodgson, J .L., Wood, J .L. and Reid, S.W.
(2001a) Coughing in thoroughbred racehorses: risk factors and tracheal endoscopic and
cytological ﬁndings. Veterinary Record 148, 99-104.

Christley, R.M., Hodgson, D.R., Rose, R.J., Wood, J .L., Reids, S.W., Whitear, KC. and
Hodgson, J .L. (2001b) A case-control study of respiratory disease in Thoroughbred
racehorses in Sydney, Australia. Equine Veterinary Journal 33, 256-264.

Clarke, AF. and Madelin, T. (1987) Technique for assessing respiratory health hazards
from hay and other source materials. Equine Veterinary Journal 19, 442-447.

Clarke, A.F., Madelin, T. and Alpress, R.G. (1987) The relationship of air hygiene in
stables to lower airway disease and pharyngeal lymphoid hyperplasia in two groups of
Thorough-bred horses. Equine Veterinary Journal 19, 524-530.

Clements, J .M. and Pirie, R.S. (2007a) Respirable dust concentrations in equine stables.
Part 1: Validation of equipment and effect of various management systems. Research in
Veterinary Science 83, 256-262.

Clements, J.M. and Pirie, R.S. (2007b) Respirable dust concentrations in equine stables.
Part 2: The beneﬁts of soaking hay and optimising the environment in a neighbouring

stable. Research in Veterinary Science 83, 263-268.

Crichlow, E.C., Yoshida, K. and Wallace, K. (1980) Dust levels in a riding stable. Equine
Veterinary Journal 12, 185-188.

68

Gerber V, L.A., Bemey C, Robinson NE (2004) Airway mucus in recurrent airway
obstruction--Short-term response to environmental challenge. J Vet Internal Med. 18, 92-
97 .

Ghio, A.J., Mazan, M.R., Hoffman, A.M. and Robinson, NE. (2006) Correlates between
human lung injury after particle exposure and recurrent airway obstruction in the horse.
Equine Vet J 38, 362-367.

Gillies, J.A., Gertler, A.W., Sagebiel, J.C. and Dippel, WA. (2001) On-road particulate
matter (PM2.5 and PM10) emissions in the Sepulveda Tunnel, Los Angeles, California.
Environ Sci Technol. 15; 35, 1054-1063.

Gilmour, P.S., Brown, D.M., Lindsay, T.G., Beswick, P.H., MacNee, W. and Donaldson,
K. (1996) Adverse health effects of PM10 particles: involvement of iron in generation of
hydroxyl radical. Occupational and Environonmental Medicine 53, 817-822.

Holcombe, 8.J., Jackson, C., Gerber, V., Jefcoat, A., Bemey, C., Eberhardt, S. and
Robinson, NE. (2001) Stabling is associated with airway inﬂammation in young Arabian
horses. Equine Veterinary Journal 33, 244-249.

Holcombe, 8.J., Robinson, N.E., Derksen, F.J., Berthold, B., Genovese, R., Miller, R.,
DeFeijter-Rupp, H., Carr, E.A., Eberhard, S.W., Boruta, D. and Kaneene, J .B. (2006)
Effect of tracheal mucus and tracheal cytology on racing performance in Thoroughbred
racehorses. Equine Vet J 38(4), 300-304.

Kappos, A.D., Bruckmann, P., Eikmann, T., Englert, N., Heinrich, U., Hoppe, P., Koch,
E., Krause, G.H., Kreyling, W.G., Rauchfuss, K., Rombout, P., Schulz-Klemp, V., Thiel,
W.R. and Wichmann, HE. (2004) Health effects of particles in ambient air. Int J Hyg
Environ Health 207, 399-407.

Kirschvink, N., Di silvestro, F, Sbai, I., Vandenput, 8., Art, T., Roberts, C. and Lekeux,
P. (2002) The Use of Cardboard Bedding Material as Part of an Environmental Control
Regime for HeaveS-affected Horses: In Vitro Assessment of Airborne Dust and

Aeroallergen Concentration and In Vivo Effects on Lung Function. The Veterinary
Journal 163, 319-325.

Li, X.Y., Gilmour, P.S., Donaldson, K. and MacNee, W. (1997) In vivo and in vitro
proinﬂammatory effects of particulate air pollution (PM10). Environ Health Perspect 105
Suppl 5, 1279-1283.

MacNamara, B., Bauer, 8. and Iafe, J. (1990) Endoscopic evaluation of exercise-induced
pulmonary hemorrhage and chronic obstructive pulmonary disease in association with
poor performance in racing Standardbreds. Journal of the American Veterinary Medical
Association 196, 443-445.

69

Malikides, N. and Hodgson, J .L. (2005) Endotoxin is an Early Component of the Causal
Pathway in Inﬂammatory Airway Disease (IAD) in Yount TB Racehorses. In: 2005
World Equine Airways Symposium, Ithica, New York.

McGorum, B.C., Ellison, J. and Cullen, RT. (1998) Total and respirable airborne dust
endotoxin concentrations in three equine management systems. Equine Veterinary
Journal 30, 430-434.

O'Grady, N.P., Preas, H.L., Pugin, J ., Fiuza, C., Tropea, M., Reda, D., Banks, SM. and
Suffredini, AF. (2001) Local inﬂammatory responses following bronchial endotoxin
instillation in humans. American Journal of Respiratory Critical Care Medicine 163,
1591-1598.

Oberdorster G (1996) Signiﬁcance of particle parameters in the evaluation of exposure-
dose-response relationships of inhaled particles. Inhal Toxicol. 8 Suppl, 73-89.

Oberdorster, G., F inkelstein, J.N., Johnston, C., Gelein, R., Cox, C., Baggs, R. and Elder,
AC. (2000) Acute pulmonary effects of ultraﬁne particles in rats and mice. Res Rep
Health Eff Inst, 5-74; disc 75-86.

Peel, J .L., Tolbert, P.E., Klein, M., Metzger, K.B., Flanders, W.D., Todd, K.,
Mulholland, J .A., Ryan, PB. and Frumkin, H. (2005) Ambient air pollution and
respiratory emergency department visits. Epidemiology 16, 164-174.

Pope, C. (1991) Respiratory hospital admissions associated with PM10 pollution in Utah,
Salt Lake, and Cache valleys. Arch Env Health 46, 90-97.

Pope III, C.A., Burnett, R.T., Thun, M.J., Calle, E.E., Krewski, D., Ito, K. and Thurston,
GD. (2002) Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine
Particulate Air Pollution. J AMA 287, 1132-1141.

Schwartz, J. (2004) Air Pollution and Children's Health. Pediatrics 113, 1037-1043.

Singh, R.B., Desloges, C. and Sloan, J .J . (2006) Application of a microscale emission
factor model for particulate matter to calculate vehicle-generated contributions to ﬁne
particle emissions. J Air Waste Manag Assoc. 56, 37-47.

Tremblay, G.M., Ferland, C., Lapointe, J .-M., Vrins, A., Lavoie, J .P. and Corrnier, Y.
(1993) Effect of stabling on bronchoalveolar cells obtained from normal and COPD
horses. Equine Veterinary Journal 25, 194-197.

Wood, J .L., Burrell, M.H., Roberts, C.A., Chanter, N. and Shaw, Y. (1993) Streptococci

and Pasteurella spp. associated with disease of the equine lower respiratory tract. Equine
Veterinary Journal 25, 314-318.

70

Wood, J .L., Newton, J.R., Chanter, N. and Mumford, J.A. (2005a) Association between
respiratory disease and bacterial and viral infections in British racehorses. J Clin
Microbiol. 43, 120-126.

Wood, J .L., Newton, J .R., Chanter, N. and Mumford, J.A. (2005b) Inﬂammatory airway
disease, nasal discharge, and respiratory infections in young British racehorses. Equine
Veterinary Journal 37, 236-242.

Woods, P.S., Robinson, N.E., Swanson, M.C., Reed, C.E., Broadstone, RV. and
Derksen, F .J . (1993) Airborne dust and aeroallergen concentration in a horse stable under
two different management systems. Equine Veterinary Journal 25, 208-213.

Zelikoff, J .T., Chen, L.C., Cohen, M.D., Fang, K., Gordon, T., Li, Y., Nadziejko, C. and
Schlesinger, RB. (2003) Effects of inhaled ambient particulate matter on pulmonary
antimicrobial immune defense. Inhalation Toxicology 15, 131-150.

71

CHAPTER 3
Horse-Environment Interaction: The Relationship between Environmental Particulate

Matter and Airway Inﬂammation in Thoroughbred Racehorses

Summary
Reason for Performing Study

Inﬂammatory airway disease (IAD) is characterized by excess tracheobronchial
mucus, which negatively affects performance. The stable environment has been

implicated as a risk factor for IAD in Thoroughbred racehorses.

Objectives

The purpose of the present study was to determine the relationship between area
concentrations of PM10 and PM2.5 as measured in three racing stables and indices of
airway inﬂammation (i.e. tracheal mucus and inﬂammatory cells in tracheal wash) in the

horses that inhabit them.

Methods

We studied 107 racehorses from 3 stables, in 3 different months. Horses were
endoscopically examined from the nares to the trachea, assigned a mucus score, and
tracheal lavage was performed. Bivariate procedures, general estimating equations
(GEE), and mixed models were applied to estimate the association between repeated

measurements of PM each stall and stable, mucus presence, and inﬂammatory cells.

72

Conclusions

Month, stable, and PM were signiﬁcantly associated with the presence of tracheal
mucus, which had an overall prevalence of 67%. GEE models show statistically
signiﬁcant association between PM10 [OR 5.8, 95% CI (164-2056), p<0.0064] and
PM2.5 [OR 4.5, CI (1.35-14.90), p<0.0151]. The prevalence of tracheal mucus was
highest in the stables and months with highest PM and least in the open-sided stable with
low PM and after a period of wet weather when the PM concentration was low. Of all
tracheal wash samples, 68% contained > 20% neutrophils, and there was a signiﬁcant

association between PM2.5 and inﬂammatory cells.

Potential Relevance

Airborne particulates affect the presence of tracheal mucus. Further studies are
warranted to investigate associations of PM and severity of mucus, allowing for
recommendations that would reduce overall prevalence of tracheal mucus. More than
20% neutrophils in 80% of horses without visible mucus underscores the need to

reevaluate the use of inﬂammatory cells in the deﬁnition of IAD.

73

Introduction

Inﬂammatory airway disease (IAD) is a syndrome with an overall prevalence of
13% (Wood et al. 2005a), and affects up to 45% of racehorses in a training stable at any
one time. Racehorses with IAD may not Show clinical signs of respiratory illness (i.e.
cough, nasal discharge, malaise), however, the presence of excess tracheobroncheal
mucus visible upon endoscopic examination, and inﬂammatory cells in tracheal aspirates
and/or bronchoalveolar lavage ﬂuid are used to deﬁne this syndrome (Anon 2003).
Previous studies have shown that the presence of visible mucus in the trachea of
racehorses may result in a Signiﬁcant negative ﬁnancial impact to owners and trainers, as
these horses tend to ﬁnish ﬁlrther back in a race (Holcombe et al. 2006; MacNamara et

al. 1990a).

Bacterial infection (Newton et al. 2003b; Wood et al. 1993; Wood et al. 2005a, c)
and stable environment (Malikides and Hodgson 2005) are two factors that have been
implicated as causative agents of IAD in racehorses. AS more than 20% of racehorses
have been diagnosed with IAD culture negative from the trachea (Chapman et al. 2000;
Christley et al. 2001a; Wood et al. 2005a), the stable environment may be responsible for
airway inﬂammation in healthy horses (Gerber et al. 2004a; Holcombe et al. 2001;
Tremblay et al. 1993). Because particulate exposure is associated with airway disease in
people and experimental animals, we focused on the exposure of racehorses to dusts
(particulate matter) in the stable environment, and its relationship to measures of airway

inﬂammation (i.e. tracheal mucus and inﬂammatory cells in tracheal aspirates).

74

Sources of particulate matter in the stable environment are many, and may include
feed and bedding, ﬂooring materials, and particulates that while originating outside of the
stable environment (e. g. racetrack footing, road dust), have the potential to disperse and
be carried indoors. These sources of dust all contribute to area particulate concentrations
in the stalls. Multiple studies investigating particulate matter both in controlled settings as
well as in private stables have demonstrated that the aerodynamic diameter of particulates
found in the stable environment (Clarke and Madelin 1987; Clarke et al. 1987; Crichlow
et al. 1980; Kirschvink et al. 2002; McGorum et al. 1998; Woods et al. 1993) are
consistent with those known to result in worsening of pre-existing respiratory disease in
humans, increased lost work/school days, and increased mortality (Brunekreef and

Forsberg 2005; Pope 1991a; Pope ID 2000 ; Pope HI et al. 2002; Schwartz 2004).

We have previously demonstrated that particles less than 10 and 2.5 p in diameter
(PM10 and PM2.5, respectively) consistent with worsening of disease in humans are
present in racing stables, and that their concentrations are inﬂuenced by month, activity
level (time of day), stable, and stalls within a stable (Millerick-May et al. 2008a).
Concentrations of PM10 and PM2.5 are high in the morning hours during peak activity
(e. g. feeding, cleaning), and in sampling months following a period of dry weather. Large
particles (PM10) settle out quickly, while the smaller/lighter particles (PM2.5) remain
suspended. Stable design may affect the rate of decline, with the more enclosed stable
experiencing higher PM10 and PM2.5 peak and average concentrations throughout the

day as compared to the stable with a more open design. Low-level concentrations of

75

small particles (PM 2.5) potentially representing chronic low-level exposures are affected
by stable design and orientation to track roadways, with the open-air stable experiencing
the highest concentrations. The location of stalls within each stable also is signiﬁcantly
associated with PM concentration, with stalls located in hi gh-trafﬁc areas and/or away
from doors and windows experiencing signiﬁcantly higher concentrations than adjacent

stalls.

The purpose of the present study was to determine the relationship between area
concentrations of PM10 and PM2.5 as measured in three racing stables (Millerick-May et
al. 2008a) and indices of airway inﬂammation (i.e. tracheal mucus and inﬂammatory
cells in tracheal wash) in the horses that inhabit them. We hypothesized that indices of
airway inﬂammation would have a high prevalence during the months in which
particulate concentrations are increased, in stables with the highest overall PM

concentrations, and in horses whose stalls had the highest concentrations of PM.

Materials and Methods

This study was conducted as repeated measurement in July, September, and November
2005 at a Thoroughbred racetrack in the Midwestern United States, where, Holcombe et al.,
(2006) demonstrated that the presence of mucus in the airways negatively affected racing
performance (Holcombe et al. 2006), and where we investigated factors affecting area
particulate concentrations in stables (Millerick-May et al. 2008a). Three trainers with
differing management styles and inhabiting stables of differing construction were chosen for

inclusion into this study.

76

Experimental Design

Three visits were made to the racetrack during the 2005 racing season (June,
September, and November), staying 3 days at each visit. The visits were selected to be a)
early in the racing season, b) at the height of the season when conditions were usually
driest, and c) when the weather was cooler and damp and the new stable (Stable 1) had

closed its outer shutters.

The upper airway and trachea of each horse was endoscopically evaluated, and a
tracheal aspirate obtained one day during each of the three visits. The name, age, sex,
stable, and stall location of each horse was recorded along with mucus score. AS reported
in Millerick-May et al. 2008, particulate concentrations were measured in one stable each
day during the visit. Concentrations of PM10 and PM2.5 were measured for one minute
in every stall three times of day - a) early morning during clean out, feeding and
grooming, b) mid-day when there was less activity, and 0) late afternoon, which

coincided with feeding and racing.

Endoscopic Examination

Endoscopic examination was performed with the horse in its own stall. A member
of the trainer's staff restrained the horse; no chemical restraint was used. The
nasopharynx, larynx, and trachea were examined. The amount of mucus visible in the
trachea was graded on a scale of 0-5: 0 = no visible mucus; l = singular small drops; 2 =

multiple partially conﬂuent drops; 3 = a ventral stream of mucus; 4 = a ventral pool; and

77

5 = profuse amounts of mucus occupying more than 25% of tracheal lumen (Gerber et al.

2004b). None of the horses in this study scored grade 5.

Tracheal Lavage

After the mucus accumulation was scored, 10 mL of saline was infused into the
upper trachea through a sterile polyethylene catheter placed in the biopsy channel of the
endoscope. The saline ﬂowed down the trachea and formed a pool at the thoracic inlet.
The endoscope was then advanced caudally until the saline pool was visible and a sample
was aspirated through the catheter. After collection, samples were placed in storage
tubes on ice until they were analyzed (within 20 minutes). The outside of the endoscope

was thoroughly cleaned with betadine solution and distilled water between horses.

Measurements of total cell count and preparation of slides for differential cell
counts were conducted at a temporary laboratory that was set up in one of the stables.
The tracheal lavage samples were diluted 1:4 in sterile buffer. In the case of extremely
viscous samples, further dilutions were made. For total cell count, 10 microliters of
diluted sample was loaded into one side of a hemacytometer. Leukocytes were counted
in 4 chambers and total cell count was calculated by using the formula: cell
concentration/mL = total cell count in 4 squares x 2500 x dilution factor. In addition, 2
Slides were prepared by use of a cytocentrifuge. Each slide was loaded with 175
microliters of sample and the positively charged slides Spun at 600 rpm for 8 minutes,
removed from the cassette, air dried and sprayed with histological ﬁxative for transport

back to Michigan State University. Differential cell counts were conducted at the

78

University on the cytospin Slides stained with Wright Geimsa and examined under a
microscope by use of a 40X objective. A total of 200 leukocytes were counted in order to
determine the percentage of neutrophils, macrophages, lymphocytes, eosinophils, and
mast cells. Absolute cell counts were calculated ﬁom the total and differential cell

counts.

Stables

Stable 1 was of newer construction and design. Orientation of the stable was north-
south with roll-up shutters along the entire length (east and west sides) of the stable in
combination with high vaulted ceilings, large sliding doors on both ends (front and back),
and open fronted stalls (bars and gates) that face the outdoors. This stable, while
consistently experiencing the lowest PM10 and PM2.5 average and maximum
concentrations, has the highest PM2.5 minimum concentrations, presumably from road

dust dispersed from adjacent roadways and parking lots.

Stable 2 was of brick construction and the long axis of the stable was oriented in
the north-south direction with busy track and city roadways ﬂanking the lengths of the
stable. This stable had 3-meter ceilings, closed-front stalls, and small, high windows
(only in stalls that are on outer walls) that were kept closed. A single row of stalls was
located along each of the outer walls of the stable, with a double row (back-to-back) of
stalls down the center. The main source of ventilation was from two sets of large
doorways both at the north and south ends of the stable, and one set of smaller doorways

in the center of the west Side of the stable. Large buildings located close to the main

79

doorways served to impede natural airﬂow. This was the only stable in the study to feed
from hay nets, and had the greatest number of individuals simultaneously working within
the stable. This stable consistently experienced the highest PM10 and PM2.5 average

and peak concentrations.

Stable 3 was Of identical construction to Stable 2, however some differences
existed. The stable was oriented with its long axis east to west, and had open windows
during the warm months. This stable was located in a secluded area (away from roads),

but was adjacent to a common parking area along its western end.

Measurement of Particle Concentration

The stables and protocol for determining particle concentrations has been
described previously (Millerick-May et al. 2008a), and is therefore only brieﬂy reported
here. Brieﬂy, concentration of particles were assessed with two TSI DustTrak3 monitors.
Using particle size conditioning inlets, one was set to measure concentrations of PM10, and
the other PM2.5. Equipment calibration and cleaning was performed prior to the start of
each day of sampling per manufacturers instructions, and periodic ﬂow evaluations were

made throughout each sampling day.

All measurements were one minute in duration, taken at the front-center of each stall

1 meter above the ground, approximating the height of the nostril of the horse. In an attempt

 

3 TS] Incorporated, 500 Cardigan Road, Shoreview, MN 55126-3996

80

to minimize our inﬂuence in the stable, horses remained in the stalls during monitoring, and

all activities (cleaning, grooming, feeding, etc.) took place as scheduled.

Data Analysis
Bivariate Analysis

Pearson’s chi-square tests were performed to determine which factors (e. g. age,
month, stable, and particulate concentrations (PM10 and PM2.5)) were associated with
tracheal mucus scores. For purposes of this analysis, we grouped tracheal mucus scores
into two categories. A tracheal mucus score of (MS) 21 includes the presence of any
amount of mucus in the trachea observed during endoscopic examination. M822
corresponds with a mucus score of 2 or above, that which has previously been determined
to negatively impact racing performance (Holcombe et al. 2006). Average, minimum,
and maximum PM10 and PM2.5 concentrations were also grouped into categories which
1) represented concentrations across all day times (morning, noon, afternoon) and
sampling months (e. g. PM10Ave), 2) by day time across all months (e. g. PM10Ave
morning). To facilitate analysis, PM concentrations in each measurement category were

grouped into quartiles.

In order to determine the association between the above exposures and continuous
variables (i.e. cell count data from tracheal lavage), Speannan correlation coefﬁcients
were estimated. Inﬂammatory cell counts by cell type (e. g. neutrophils), as well as

percent of each type of inﬂammatory cell were analyzed to determine whether or not

81

there was a Signiﬁcant effect of particulate exposure. PM10 and 2.5 concentrations were

grouped as stated above.

Multivariate Analysis

Generalized estimating equation (GEE) models were conducted (GENMOD
procedure) to determine the association between the categorical variables of M821 and
M822 and factors found to be signiﬁcant in bivariate analyses (e.g. age, month, PM
concentrations). The outcome variables were measured three times (July, September, and
November) with horses being potentially evaluated endoscopically up to three times.
GEE analysis allowed for the determination of the effect of PM exposures on the
aforementioned dichotomous outcome variables (MS21 and MS22), while controlling for
within-horse variability (Zeger and Liang 1986). The potential effect of stable on the
presence of tracheal mucus was considered, however, not included in the models as
horses were not randomly assigned to stable. Because concentrations of particulates were
strongly associated with stable, the latter was a possible surrogate for exposure level, and

therefore could not be evaluated independently.

Linear mixed models (MIXED procedure) were conducted to determine the effect
of the above exposures on continuous outcome variables (inﬂammatory cell counts).
This procedure is designed to handle hierarchical data that is collected repeatedly on
subjects (horses) over time. In addition, it is able to handle missing data and has the
power to analyze a munber of variance covariance structures while accounting for within-

horse variability (Littel et al. 1996). The regular maximum likelihood (REML) method

82

was applied, assuming the random effects and error vector were normally distributed.
This was achieved by log-transforming the cell count data, which were then transformed
in geometric means. As possible covariance structures, unstructured (UN), spatial
Gaussian (SP [GAU]), ﬁrst order autoregressive (AR[1]), TOEPLITZ (TOEP), and
compound symmetry (C8) were tested. The best covariance structure was chosen based

on the Akaike Information Criterion, which was TOEPLITZ for all models.

All data analysis was conducted using SAS v.9.1. Level of Signiﬁcance was set at

P3005.

Results
Study Population

The population consisted of a total of 107 horses from 3 stables. Each horse that
was present during each of our 3 visits was examined (July, September, and November of
2005). Thirty-eight percent of horses were examined once, with thirty-one percent of
horses each examined a second and third time, resulting in a total of 206 individual
examinations. Of the 206 attempted examinations, endoscopy was not possible 18 times,
resulting in a total of 188 mucus score/tracheal lavage combinations (Table 1). Fourteen
tracheal lavage samples were not of sufﬁcient quality to perform analysis, leaving 174

samples of sufﬁcient quality for total and differential cell count analysis.

83

Tracheal Mucus
Of the 188 endoscopic evaluations performed, 67% (n=126) had a mucus score Of
greater than or equal to l (M821), and 19% (n=35) had a mucus score of greater than or

equal to 2 (M822) (Table 1).

Bivariate analysis (Table 2) showed that the following had a signiﬁcant effect on
M821: age, month, stable, PM10 and PM2.5 average and maximum values, as well as
PM10 and PM2.5 average and maximum values during the early morning sampling
periods . PM10 minimum concentrations, as well early morning PM10 minimum
concentrations were also signiﬁcantly associated with M821 (Table 2). Age and stable

were the only two factors signiﬁcantly associated with M822 (Table 2).

Multivariate Analysis
M821

There was a statistically signiﬁcant positive association between PM10 average
concentrations (4th quartile - OR 5.8, 95% CI (1.64-2056), p<0.0064) and a mucus score
of 21. There was also a statistically signiﬁcant association between PM2.5 average
concentrations (1St quartile — OR 4.5, CI (1.35-14.90), p<0.0151) and the presence of
tracheal mucus. The overall trend between PM10 and PM2.5 average concentrations by
stable, and the association with M821 can be seen in Figure 1. In a model that included
age, gender, sample month, and PM2.5 average concentrations (quartiles), there was a
signiﬁcant effect of month and of PM2.5 average on M821 (Figure 2), with the July

measurement period being the lowest in terms of M821 prevalence (OR 0.10, CI (0.03-

84

0.38), p<0.0007) and of PM2.5 concentration. There were no signiﬁcant effects of age or

gender on MUC21.

M822

Bivariate analysis shows a signiﬁcant relationship between age and MS22 (Table
2). This association remained signiﬁcant in the models (GEE), which were deveIOped to
evaluate the association between average, minimum, and maximum particle
concentrations, sampling month, and age, and the presence of tracheal mucus in
quantities that affect racing performance (M822). Two-year oldS had a signiﬁcantly
greater risk (OR 3.36-3.96, 95% CI (1.08-13.46), p<0.0275 — 0.0371) of M822 as

compared to horses 4 years Old and older (reference group).

While the GEE models used to determine the effect of particle concentrations on
MUC2 or greater did not show a signiﬁcant association, the overall trend remained
consistent with that of M821 , with a greater prevalence of M822 in Stable 2 as compared
to Stables 1&3 (Figure 3), and an increasing trend between the months of July and

September, with a slight decline in November (data not shown).

Inﬂammatory Cells

In this study, 68% of tracheal wash samples contained 20% or more neutrophils.
Chi-square analysis revealed no signiﬁcant association between inﬂammatory cell counts
(grouped in tertiles) and MS21. However, there was a Signiﬁcant association of M822

with numbers of neutrophils (p<0.02), macrophages (p<0.02), and lymphocytes (p<0.01).

85

The construction of GEE models using M821 and MS22 as outcome variables, and
inﬂammatory cell count data as predictor variables conﬁrmed that there was no
association between inﬂammatory cell counts and MS21, but a Signiﬁcant negative

association between number of lymphocytes and MS22 remained (<00009).

Chi-square analysis showed a signiﬁcant association between percent neutrophils
and lymphocytes, and PM 2.5 average concentrations, PM 2.5 average concentrations
during morning and evening sampling periods, and PM2.5 minimum concentrations (all)
and sampling time points (morning, midday, and evening) (Table 2). Signiﬁcant
associations between percent neutrophils are Similar to those of numbers of neutrophils
and particle concentrations, in that associations remain with PM 2.5 average
concentrations and the morning time point, as well as a PM2.5 maximum evening time

point (Table 2).

Mixed model analysis did not reveal any signiﬁcant relationship between PM10
and PM2.5, age, gender, or month, and numbers or percentage of inﬂammatory cells in
tracheal wash samples. However, the trend in numbers and percentage of neutrophils is
similar to that of PM10 and PM2.5 average concentrations by month (Figures 4 & 5),

with increasing values from July to September, and then a decline in November.

Discussion
These are the ﬁrst reported measurements of the association between indices of

airway inﬂammation in Thoroughbred racehorses and area concentrations of particulate

86

matter (PM10 and PM2.5) in American racing stables. Previously, a study conducted by
our laboratory (Holcombe et al. 2006) revealed that horses with M822 ﬁnished ﬁrrther
back in a race. This ﬁnding prompted racetrack management and trainers to inquire as to
the etiology of mucus in the airways, and supported our efforts to determine whether or
not there was an association between area (background) particle concentrations (PM10

and PM2.5), and measures of airway inﬂammation (tracheal mucus, inﬂammatory cells).

Brieﬂy, we found a Signiﬁcant effect of month, time of day, stable, and stall on
area concentrations of PM10 and PM2.5 (Millerick-May et al. 2008a). Measurements of
PM in the month following a period of hot/dry weather revealed the highest
concentrations, with those following a period of wet/rainy weather the lowest. AS to be
expected, times of day had a signiﬁcant effect on particle concentrations, with the early
morning hours during activities such as feeding and stall cleaning giving rise to the
highest concentrations of PM, and evening measurements the lowest. We found a
signiﬁcant effect of stable, with the busy, completely enclosed brick-style stable with
low ceilings (Stable 2) having the highest concentrations, and the open-sided stable with

high vaulted ceilings, the lowest.

Based on the aforementioned information, we hypothesized that if exposures to
area particle concentrations were to affect measures of airway inﬂammation, we would
expect the prevalence of airway inﬂammation to be greatest during the month of highest

PM concentrations (September), lowest during the month of lowest concentrations (July),

87

and highest in the stable which consistently had the highest concentrations of PM (Stable

2).

In support of this hypothesis, bivariate analysis revealed signiﬁcant effects of
month, stable, and concentrations of PM10 & PM2.5, on the presence of tracheal mucus
(MS21). PM10 and PM2.5 average and maximum concentrations as measured during the
early morning period, when peak activity was taking place, was also determined to have a
signiﬁcant effect on M821. Furthermore, multivariate analysis also revealed a Signiﬁcant
effect of PM10 & PM2.5 average concentrations on the presence of tracheal mucus
(MS21), and also a signiﬁcant effect of month (PM2.5 average) on M821, with the
month of July (following a period of wet weather) having the lowest concentrations of
PM2.5 average measurements, and the lowest prevalence of M821 (Figure 2). While
there was not a signiﬁcant effect of PM on the prevalence of M822 as there was with

M821, there was a Similar trend between the prevalence of M821 or M822 and stable

(Figure 3).

We were not able to include ‘stable’ as a variable in the model as it is a surrogate
for PM concentrations, however, further investigation into which stable contributed the
highest percentage of PM measurements to quartiles deemed to be signiﬁcant, indeed
revealed that Stable 2 had the greatest inﬂuence (Figure 1), as it was the most enclosed
and busiest of the three, contributed the greatest percent of PM1 0 and PM2.5 average
measurements to each of the 4th quartiles (highest concentrations), and was found

previously to contain the greatest number of stalls whose PM concentrations were in the

88

4th quartile (Millerick-May et al. 2008a), with the 4th quartile of PM10 (average) being

signiﬁcantly associated to M821.

There was a signiﬁcant effect of age on M822, as was determined by both
bivariate and multivariate analysis. 2-year olds, in comparison to those aged 4 and older,
had a higher prevalence of M822. Others investigating factors affecting the presence Of
IAD in racehorses have also found an association between age and the prevalence of
airway inﬂammation (Christley et al. 2001a; Wood et al. 1993; Wood et al. 20050),
however, more recent reports suggest an association between length of time in training,
rather than age, and indices of airway inﬂammation (Wood et al. 20050), with those
horses that have been in training longer, having a lower prevalence. While we were not
able to collect data regarding length of time in training, we did include length of time in
stable (data not shown) in each of our multivariate models. Length of time in stable was
estimated by number of months included in our study and reports of ‘shipped in’ dates by
our trainers. We found no signiﬁcant association between length of time in stable and
either M821 or M822. Others have hypothesized that young horses recently brought into
a stable environment would be more susceptible to respiratory infection, and over time
would become more resistant (Wood et al. 2005a). If this were the case, one would
expect the highest prevalence of M821 or M822 in our population to occur during our
initial sampling period (July), and decline throughout the rest of the year. This was not
the case in our population, as our highest prevalence of both MUC21 and MUC22

occurred in September.

89

Our data suggest that small amounts of mucus (MUC=1) are a result of a local
irritant response to particulate exposure. We reach this conclusion for the following
reasons: 1) the prevalence of M821 tracks particulate concentrations by month and by
stable, and is strongly associated with PM10 and PM2.5, 2) the prevalence of M821 is
not related to either the percent or number of inﬂammatory cells, and 3) the prevalence of
M822 is associated with changes in inﬂammatory cell populations. The fact that two-
thirds of the observations of MS>1 are comprised of MUC=1 (Table 1), suggests
therefore that MUC=1 is not associated with the presence of inﬂammatory cells, but

rather is a local response to particulates perhaps mediated through the epithelium.

Our overall study was designed to determine the effects of month, time of day,
and stable (design features and management) on area particle concentrations, and
resulting measures of airway inﬂammation (tracheal mucus and inﬂammatory cell
counts). As such, our study design required repeated measures in the same stables over
time. This study was not designed or sufﬁciently powered to distinguish between
concentrations of particulate matter, and degrees tracheal mucus (i.e. M822) previously
found to be associated with racing performance (Holcombe et al. 2006)). Based on the
trends observed in the present study, a more extensive investigation of this association is
warranted. As we have shown that the prevalence of M821 is Signiﬁcantly associated
with the 4th quartile PM10 Ave concentrations, and that prevalence of M822 follows a
Similar pattern, if data from a larger study provided more clarity as to the area
concentrations of PM that are associated with degrees of tracheal mucus, it should be

possible to set area PM10/PM2.5 exposure limits to reduce the occurrence of tracheal

90

mucus in quantities shown to be associated with racing performance (Holcombe et al.

2004).

The usefulness of inﬂammatory cell counts is an unresolved issue with regard to
diagnosis of IAD (Anon 2003) as there are other newer and better references. The
phenotype for IAD has been deﬁned as the presence of tracheal mucus observed during
endoscopic examination, and >20% neutrophils in tracheal wash/BAL samples.

However, in the present study, 68% of tracheal wash samples contained greater than 20%
neutrophils. Furthermore, 82% of horses with no visible tracheal mucus (MUC=0), had
greater than 20% neutrophils. Almost exactly the same prevalence was observed by
(Robinson et al. 2006), in a study of pleasure horses. These observations indicate that the
deﬁnition of IAD may need to be revisited, with tracheal mucus score rather than

inﬂammatory cells used as the gold standard for diagnosis.

91

Table 3-1. Summary of number of examinations performed per stable, age, gender, and

assigned mucus score, as well as the prevalence of M821 and MS22.

 

 

 

 

 

 

 

Mucus
Mucus Affecting
Present Racing
Age Gender Mucus Score“ (M821) (M822)
Stable # Examinations 2 3 24 Unk. M F 0 l 2 23 # % # %
l 67 1 3 7 46 1 54 1 3 29 25 l 1 27 48 2 4
2 84 44 25 12 3 43 41 20 41 14 7 62 76 21 26
3 55 16 18 17 4 38 17 13 25 12 0 37 74 12 24
Total 206 73 50 75 8 135 71 62 91 27 8 126 67 35 19

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

92

 

 

vmved emmed deed wged Need memed $38.sz

 

0222a

 

$85 0222a

 

_emed Hm><e~2m

 

Neeed wveed EEON—ZQ

 

Hmmed Need emded £2233

 

meeed 082m . gm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

meed
ﬂood xuzeﬁzm
tamed pied mmeed vowed Sled RNed 9556.333
286 9222.2
28d mmeed Deed 035m
vNeed vied 5:22
Seed ewmed :ed deﬁed :lved ow<
# % # W % # % # 1 % # % W W
W W W n ma 3 m. m w N _6 _«s
m m w m s. w w d n m .0 l
D D d u. .m o u. 0 .w 0
al.. a u. cw u. d w ,M u. d
H H aw e ".1. N ,M e H mm
5 9 S S N 9 S S q
s s
deem:

Son. 033 medv negate .m=oo >88ng 93 888 25:5 wﬁuooam macaque eo $0.398 08535 eo 23.0% .N-m 2an

93

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

885 :36 88.0 $86 3235

886 ﬂood 2222;

823 good €223:

So 9:225

356 88.0 3225

good BEBE

$86 £86 835 3%.sz

$23 :23 $3222;

28d ~83 £23 3222;

322E

$8.0 BEBE

:35 3225

”Rod £85 083 E3 @222;

# EMS # w /o # /o # 1 /o # /o w w Nw><n~2m
5 ﬂ am 9. ﬂ «4. W. m y. “1..

 

 

 

 

 

 

 

 

 

.5283 3 29¢

94

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ohm.“
eemd emwd ewmd - £6 QNN mod need 5 ._ eeé - mde mug. one. mvod Ned -
- C _d - wad - vemd mnvd - Nemd - mend - Em.— - vemd - memd mend - cad - oehd - SA - emmd eemd v
emvd emnd Need wmmd
m: "d eN—d memd - eemd dike vm; - mend eemd - mew—d Sud - ommd -
- Vmed - coed - _eed _emd - memd - vad :wd - need .. «Ed ede - i.e.e - 03d Need - 12d Rmd m
eemd mmwd :md weed ecmd
mmed weed owed - vemd Suvd - weed 02d - mned wgd - emﬂd -
- mved - hmed - Edd mad - :_.e - vfd cmmd - need - need _mmd - owed - wEd eevd - mwed meme N
dad wﬁmd vmmd cevd thd
Nwed owed deed - eeﬁe Ngd - wood wood - mmed 53d - mwed -
- ﬂed - ﬂed - weed wmed - vmed . Nmed wgd - Nved - emed need - fed - owed mwﬁd - Vmed emﬁe _
mEE NEE _EE :3: GEE _me ﬁxmﬁ mo>~ ~26 ~o>a 38 x5: x9: 02“ o>a M
WNSE WNSE WNSE 33E “~33 WNSE 32m 9sz WNSE 32m 0sz 092m 253 932m 32m W

 

 

 

95

.Awﬁﬁgo l m SweEE - N .wEEoE - C @2580 wEEEwm
b6 we 08: wcumoweﬁ £8390: oceans 2:3 68:38.“ 98 mos—S, EEG SEEMS 98 .9225 83::me Amid owﬁu><

29:28 688 @228 23 mmO 8e BEE: 05.83 some E53, 22% 83$: BEESEQ mo mcosmbzoocou .m-m 2an

.o 9
01 O)

53
.1;

Concentration (mglma)
_e e
N w

_e
-L

 

Stable 1 Stable 2 Stable 3

Figure 3-1. Effect of particle concentrations on overall prevalence of M831 (black line)
by stable. Gray bars indicate average PM10 concentrations (mg/m3), and black bars

indicate average PM2.5 concentrations (mg/m3).

96

.o .0
U1 0)
—#—l

_o
A

9
N
1 .

Concentration (mg/m3)
e
O.)

O
_s

O .
+4444
l

 

June September November

Figure 3-2. Effect of particle concentrations on overall prevalence of MS>1 (black line)
by sampling month. Gray bars indicate average PM10 concentrations (mg/m3), and

black bars indicate average PM2.5 concentrations (mg/m3).

97

.o .0
01 O)

.o
.3:

Concentration (mg/m3)
e e
N w
Twei+A~4+~——4-—nev~Ae~——a

9
—L

 

0

Stable 1 Stable 2 Stable 3

Figure 3-3. Effect of particle concentrations on overall prevalence of MS>2 (black line)
by stable. Light gray bars indicate average PM10 concentrations (mgm3), and dark gray

bars indicate average PM2.5 concentrations (mg/m3).

98

35

a 0.3

E

U)

E

C

3 0.2 .-

i!

H

C

0

0

C

O

U 0.1 1*

0 ~ —l~—* ' 0

 

July September November

Figure 3-4. Effect of particle concentrations on percent neutrophils (solid line) and
lymphocytes (dashed line) by month. Dark gray bars indicate average PM10
concentrations (mg/m3), and light gray bars indicate average PM2.5 concentrations

(mg/m3).

99

   

Percent Cell Type

v 180

~ 160
A «l 140 E
00 =
0
g 1 120 .c_'3
g E
c -, 100 g
.g a)
9 + 80 2
’E o
g —— 60 g
o E
o r 40 2

20

e o

 

 

July September November

Figure 3-5. Effect of particle concentrations on numbers of neutrophils (solid line) and
lymphocytes (dashed line) by month. Dark gray bars indicate average PM10
concentrations (mg/m3), and light gray bars indicate average PM2.5 concentrations

(mg/m3).

100

References

Anon (2003) Workshop report: Inﬂammatory airway disease: deﬁning the syndrome.
Equine veterinary education 5, 81-82.

Brunekreef, B. and Forsberg, B. (2005) Epidemiological evidence of effects of coarse
airborne particles on health. European Respiratory Journal 26, 309-318.

Chapman, P.S., Green, C., Main, J .P., Taylor, P.M., Cunningham, F.M., Cook, A.J. and
Marr, C .M. (2000) Retrospective study of the relationships between age, inﬂammation
and the isolation of bacteria from the lower respiratory tract of thoroughbred horses.
Veterinary Record 146, 91-95.

Christley, R.M., Hodgson, D.R., Rose, R.J., Hodgson, J .L., Wood, J.L. and Reid, SW.
(2001) Coughing in thoroughbred racehorses: risk factors and tracheal endoscopic and
cytological ﬁndings. Veterinary Record 148, 99-104.

Clarke, AF and Madelin, T. (1987) Technique for assessing respiratory health hazards
from hay and other source materials. Equine Veterinary Journal 19, 442-447.

Clarke, A.F., Madelin, T. and Alpress, R.G. (1987) The relationship of air hygiene in
stables to lower airway disease and pharyngeal lymphoid hyperplasia in two groups of
Thorough-bred horses. Equine Veterinary Journal 19, 524-530.

Crichlow, E.C., Yoshida, K. and Wallace, K. (1980) Dust levels in a riding stable. Equine
Veterinary Journal 12, 185-188.

Gerber, V., Lindberg, A., Bemey, C. and Robinson, N.E. (2004a) Airway mucus in
recurrent airway obstruction--short-tenn response to environmental challenge. J Vet
Intern Med 18, 92-97.

Gerber, V., Straub, R., Marti, E., Hauptman, J ., Herholz, C., King, M., Imhof, A., Tahon,
L. and Robinson, N.E. (2004b) Endoscopic scoring of mucus quantity and quality:
observer and horse variance and relationship to inﬂammation, mucus viscoelasticity and
volume. Equine Veterinary Journal 36, 576 - 582.

Holcombe, S.J., Jackson, C., Gerber, V., J efcoat, A., Bemey, C., Eberhardt, S. and
Robinson, NE. (2001) Stabling is associated with airway inﬂammation in young Arabian
horses. Equine Veterinary Journal 33, 244-249.

Holcombe, S.J., Robinson, N.E., Derksen, F.J., Berthold, B., Genovese, R., Miller, R.,
DeFeijter-Rupp, H., Carr, E.A., Eberhard, S.W., Boruta, D. and Kaneene, J .B. (2006)
Effect of tracheal mucus and tracheal cytology on racing performance in Thoroughbred
racehorses. Equine Vet J 38(4), 300-304.

101

Holcombe, S.J., Robinson, N.E., Derksen, F .J . and Kaneene, J. (2004) Tracheal mucus is
associated with poor racing perfprmance in thoroughbred horses. In: 50th Annual
Convention of the American Association of Equine Practitioners, AAEP, Denver. p In
press.

Kirschvink, N., Di silvestro, F ., Sbai, I., Vandenput, S., Art, T., Roberts, C. and Lekeux,
P. (2002) The Use of Cardboard Bedding Material as Part of an Environmental Control
Regime for Heaves-affected Horses: In Vitro Assessment of Airborne Dust and
Aeroallergen Concentration and In Vivo Effects on Lung Function. The Veterinary
Journal 163, 319-325.

Littel, R.C., Milliken, G., Stroup, W. and Wolﬁnger, R. ( 1996) SAS System for mixed
models, SAS Institute Inc., Cary, NC. p 633.

MacNamara, B., Bauer, S. and Iafe, J. (1990) Endoscopic evaluation of exercise-induced
pulmonary hemorrhage and chronic obstructive pulmonary disease in association with
poor performance in racing Standardbreds. Journal of the American Veterinary Medical
Association 196, 443-445.

Malikides, N. and Hodgson, J .L. (2005) Endotoxin is an Early Component of the Causal
Pathway in Inﬂammatory Airway Disease (IAD) in Yount TB Racehorses. In: 2005
World Equine Airways Symposium, Ithica, New York.

McGorum, B.C., Ellison, J. and Cullen, RT. (1998) Total and respirable airborne dust
endotoxin concentrations in three equine management systems. Equine Veterinary
Journal 30, 430-434.

Millerick-May, M., Karmaus, W., Derksen, F.J., Berthold, B., Holcombe, SJ. and
Robinson, NE. (2008) Evaluation of Factors Affecting the Concentration of Particulate
Matter (PM10 and PM2.5) in Stables at an American Thoroughbred Racetrack. Equine
Vet J Submitted.

Newton, J .R., Wood, J .L.N. and Chanter, N. (2003) A case control study of factors and
infections associated with clinically apparent respiratory disease in UK Thoroughbred
racehorses. Preventive Veterinary Medicine 60, 107-132.

Pope, C. (1991) Respiratory hospital admissions associated with PM10 pollution in Utah,
Salt Lake, and Cache valleys. Arch Env Health 46, 90-97.

Pope HI, CA. (2000 ) What do epidemiologic ﬁndings tell us about health effects of
environmental aerosols? J Aerosol Med. 13, 335-354.

Pope III, C.A., Burnett, R.T., Thun, M.J., Calle, E.E., Krewski, D., Ito, K. and Thurston,

GD. (2002) Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine
Particulate Air Pollution. J AMA 287, 1132-1141.

102

Robinson, N.E., Karmaus, W., Holcombe, S.J., Carr, EA. and Derksen, F.J. (2006)
Airway inﬂammation in Michigan pleasure horses: prevalence and risk factors. Equine
Veterinary Journal 38, 293-299.

Schwartz, J. (2004) Air Pollution and Children's Health. Pediatrics 113, 1037-1043.

Tremblay, G.M., F erland, C., Lapointe, J .-M., Vrins, A., Lavoie, J .P. and Cormier, Y.
(1993) Effect of stabling on bronchoalveolar cells obtained from normal and COPD
horses. Equine Veterinary Journal 25, 194-197.

Wood, J .L., Burrell, M.H., Roberts, C.A., Chanter, N. and Shaw, Y. (1993) Streptococci
and Pasteurella spp. associated with disease of the equine lower respiratory tract. Equine
Veterinary Journal 25, 314-318.

Wood, J .L., Newton, J .R., Chanter, N. and Mumford, J .A. (2005a) Association between
respiratory disease and bacterial and viral infections in British racehorses. J Clin
Microbiol. 43, 120-126.

Wood, J .L., Newton, J .R., Chanter, N. and Mumford, J .A. (2005b) Inﬂammatory airway
disease, nasal discharge, and respiratory infections in young British racehorses. Equine
Vet J 37, 236—242.

Woods, P.S., Robinson, N.E., Swanson, M.C., Reed, C.E., Broadstone, RV. and
Derksen, F.J. (1993) Airborne dust and aeroallergen concentration in a horse stable under

two different management systems. Equine Veterinary Journal 25, 208-213.

Zeger, S.L. and Liang, KY. (1986) Longitudinal data analysis for discrete and
continuous outcomes. Biometrics 42, 121-130.

103

CHAPTER 4
Airborne Particulates (PM10) and Tracheal Mucus: A Case-Control Study at an

American Thoroughbred Racetrack

Summary
Reasons for Performing Study

The presence of sufﬁcient quantities of tracheal mucus (MS32) negatively affects
racing performance. We have previously demonstrated a signiﬁcant association between
concentrations of particulate matter (PM10 and PM2.5) and the presence of tracheal

mucus in a population of Thoroughbred racehorses.

Objectives

Previous measurements of particulates (PM10 and PM2.5) by our laboratory were
of background concentrations rather than at the horse’s nostril, which would account for
not only area concentrations of PM, but also that which is generated as a result of
personal idiosyncrasies of the horse (manner of eating, in-stall activity, rolling, etc.). The
objective of this study was to determine the differences in personal particulate exposures
between those with (M822) and without (MS=0) tracheal mucus visible upon endoscopic

exam in quantities previously associated with poor racing performance.

Methods
We studied 653 racehorses from 6 stables over seven months (April — October

2006). Horses were endoscopically examined, and assigned a mucus score. Nine cases

104

and controls were selected and age matched each month based on the presence (M822) or
absence (MS=O) of tracheal mucus. Tracheal wash was performed on all cases and
controls. Each month, the selected case-control pairs wore direct reading personal
particulate monitors for on average 17.5 hours. The personal monitors were attached to
vaulting surcingles and a sufﬁcient length of Tygon tubing to span the crest of the horse,
with the end of the sampling tube secured approximately 2 inches above the nostril. The
monitors were set to take samples every second during 3 time periods (mid-day, evening,

and overnight). Particulate concentrations (PM10) were recorded after each time period.

Results

Overall prevalence of M822 was 23% during the 7-month duration of the study.
Bivariate analysis revealed a signiﬁcant effect of trainer and month on prevalence of
MS32, and a signiﬁcant effect of month on personal PM concentrations and percent
neutrophils. There was a signiﬁcant difference between exposures to PM10
concentrations in cases and controls during the evening (maximum) and overnight
(minimum) time periods, as well as signiﬁcant differences in absolute, but not percent
neutrophils between cases and controls, with both groups having greater than 20%

neutrophils in their tracheal wash ﬂuid.

Conclusions
The fact that overall prevalence of MSZZ tracked closely with monthly personal
PM exposure data, that personal particulate concentrations between cases and controls

were not signiﬁcantly different in most instances, and were similar to area concentrations

105

measured in a prior study, suggests that area PM is what predominantly inﬂuences
personal exposures. We report a signiﬁcant difference in neutrophil counts between
cases and controls, but there was no difference between percent neutrophils, a value

which is frequently used in the diagnosis of IAD.

Potential Relevance

If indeed it is the area concentrations of PM that predominantly inﬂuence personal
exposures as our data suggests, and it is these concentrations that are associated with
tracheal mucus (M832), following low-dust stable management practices may greatly

reduce the prevalence of tracheal mucus.

106

Introduction

The presence of tracheal mucus in sufﬁcient quantities negatively affects racing
performance, potentially resulting in a signiﬁcant ﬁnancial impact to the racing industry
(Holcombe et al. 2006; MacNamara et al. 1990b). Tracheal mucus is one feature of a
syndrome known as Inﬂammatory Airway Disease (IAD), which has an overall
prevalence in racing Thoroughbreds of 13% (Wood et al. 20050), with up to 45% of

horses in a single racing stable affected at any one time (unpublished data).

The cause of IAD is likely multifactorial in nature, however, bacterial infection
(Newton et al. 2003a; Wood et al. 2005a, c) and the environment (Malikides and
Hodgson 2005) are two factors that have been implicated in otherwise young and healthy
racehorses. Airway inﬂammation, as determined by the presence of tracheal mucus and
increased inﬂammatory cells in tracheal wash or bronchoalveolar lavage ﬂuid (BALF),
occurs in healthy horses upon exposure to the stable environment (Gerber V 2004;

Holcombe et al. 2001; Tremblay et al. 1993).

We have previously demonstrated a signiﬁcant association between
concentrations of particles less than 10 and 2.5 u in diameter (PM10 and PM2.5), and the
presence of tracheal mucus in Thoroughbred racehorses (Millerick-May et al. 2008b), as
well as signiﬁcant associations between month, time of day, stable, and stall on area
concentrations of these particles (Millerick-May et al. 2008a). Area concentrations of

particulate matter, however, do not reﬂect personal exposures as they may underestimate

107

actual exposures. Area measurements in stables reﬂect particles in the outside air that
permeate throughout the building, as well as those dispersed as a result of stable activity
(i.e. cleaning, feeding, and raking), and are useful in determining ventilation patterns and
‘hot spots’/dead zones within the stable. These measurements are invaluable in

formulating/recommending control strategies to reduce overall concentrations.

Personal exposure concentrations, as would be measured at the level of the nostril
(breathing zone), reﬂect not only the area concentrations mentioned above, but the
additional particulates generated as a result of ‘personal’ activities of the horse (e. g.
eating, rolling, and stall walking). These concentrations would be expected to be
different for each individual as personal behavior differs. Examples of this would be the
vigor with which an animal pulls hay from the ﬂake/hay net, the number of times a horse

rolls or lies down in a day, and movement (walking) behaviors in the stall, etc.

Dusts (particulates) measured in the breathing zone of horses have been recorded
in concentrations known to induce airway inﬂammation in humans and laboratory
animals (Millerick-May et al. 2007 ; Woods PS 1993), and originate from bedding,
feedstuffs, and footing materials. Components of these dusts are both organic (e. g. molds
and endotoxin) (Vandenput et al. 1997; Woods PS 1993) and inorganic (e.g. iron and
crystalline silica) (Millerick-May et al. 2007) in nature, many of which have occupational
exposure limits (ACGIH 2005a) aimed at preventing occupationally induced lung disease

in people.

108

The purpose of the present study, which used a case-control approach, was to
determine the association between breathing zone concentrations of particulate matter in
a population of Thoroughbred racehorses, and the presence of tracheal mucus in
quantities previously determined to negatively affect racing performance (Holcombe et
al. 2006). We hypothesized that horses with MSZZ would experience greater personal

PM exposures as compared to horses with MS=0.

Materials and Methods

This study was conducted at the same Thoroughbred racetrack in the Midwestern
United States, where, Holcombe et al., (2006) demonstrated that the presence of mucus in
the airways negatively affected racing performance (Holcombe et al. 2006), where we
investigated factors affecting area particulate concentrations in stables (Millerick-May et
al. 2008a), and where we demonstrated an association between particulate matter in

racing stables and the presence of tracheal mucus in resident horses (Millerick-May et al.

2008b)

Experimental Design

Seven visits were made to the racetrack during the 2006 racing season (April
through October), staying 4 days at each visit. Horses stabled in 6 racing stables were
chosen for inclusion into this study. Cases and controls were selected from these racing

stables, and their personal PM10 exposures recorded over a period of 24 hours.

109

In each of the 7 months, the upper airway and trachea of approximately 100
horses was endoscopically evaluated. The name, age, gender, stable, and stall location of
each horse was recorded along with mucus score. From these horses, 9 cases and 9
controls were identiﬁed based on the presence or absence of tracheal mucus. Each horse
was ﬁtted with a personal particulate monitor, and particulate concentrations (PM10)

were measured over most of a 24-hour period of time.

Endoscopic Examination

Endoscopic examination was performed with the horse in its own stall. A member
of the trainer's staff restrained the horse; no chemical restraint was used. The
nasopharynx, larynx, and trachea were examined. The amount of mucus visible in the
trachea was graded on a scale of 0-5: 0 = no visible mucus; 1 = singular small drops; 2 =
multiple partially conﬂuent drops; 3 = a ventral stream of mucus; 4 = a ventral pool; and
5 = profuse amounts of mucus occupying more than 25% of tracheal lumen (Gerber et al.

2004b). None of the horses in this study scored grade 5.

Tracheal Lavage

Tracheal lavage was performed only in horses identiﬁed as either a case or
control. Ten milliliters of saline was infused into the upper trachea through a sterile
polyethylene catheter placed in the biopsy channel of the endoscope. The saline ﬂowed
down the trachea and formed a pool at the thoracic inlet. The endoscope was then
advanced caudally until the saline pool was visible and a sample was aspirated through

the catheter. After collection, samples were placed in storage tubes on ice until they were

110

analyzed (within 20 minutes). The outside of the endoscope was thoroughly cleaned with

betadine solution and distilled water between horses.

Measurements of total cell count and preparation of slides for differential cell
counts were conducted at a temporary laboratory that was set up in one of the stables.
The tracheal lavage samples were diluted 1:4 in sterile buffer. In the case of extremely
viscous samples, ﬁrrther dilutions were made as necessary. For total cell count, 10
microliters of diluted sample was loaded into one side of a hemacytometer. Leukocytes
were counted in 4 chambers and total cell count was calculated by using the formula: cell
concentration/ml = total cell count in 4 squares x 2500 x dilution factor. In addition, 2
slides were prepared by use of a cytocentrifuge. Each slide was loaded with 175
microliters of sample and the positively charged slides spun at 600 rpm for 8 minutes,
removed from the cassette, air dried and sprayed with histological ﬁxative for transport
back to Michigan State University. Differential cell counts were conducted at the
University on the cytospin slides stained with Wright Geimsa and examined under a
microscope by use of a 40X objective. A total of 200 leukocytes were counted in order to
determine the percentage of neutrophils, macrophages, lymphocytes, eosinophils, and
mast cells. Absolute cell counts were calculated from the total and differential cell

counts.

Selection of Cases and Controls

Nine cases and nine controls were selected each month from the approximately

100 horses that were endoscopically examined. Pools of potential cases were identiﬁed

111

as horses with a tracheal mucus score of 2 or greater (MS_>_2), and controls as those
without tracheal mucus (MS-=0). The 9 cases and 9 controls were randomly chosen from
the pool and were matched by age. Individuals that had previously been identiﬁed as a
case were no longer eligible to be a case in subsequent months, and the same selection
rule applied to controls. That said - an individual that had been previously selected as a
case was eligible to be a control during a subsequent month, and vice versa. Once horses

had been used both as a case and as a control, they were excluded from further study.

Measurement of Breathing Zone Concentrations

Concentrations of breathing zone particulates were evaluated using TSI SidePak
AM5104 personal monitors. These monitors are small, lightweight, and utilize light-
scattering technology to determine real-time particle concentrations. Using particle size
conditioning inlets, each monitor was set to measure concentrations of PM10. Equipment
calibration (ﬂow and zero ﬁlter) was performed prior to the start of each sampling segment,

with cleaning occurring once daily per manufacturers’ instructions.

Samplers were afﬁxed to the horse by the use of vaulting surcingles (Figure 1)
secured just behind the shoulder. Each sampler was secured within (encircled by) a handle
located on the surcingle, the intent of which was to offer additional protection to the monitor
during lying down/rolling. Six feet of TygonTM tubing (1/ ” ID) was attached at one end

of the sampler to the inlet, run along the crest of the horses’ neck (secured to the mane),

 

‘ TSI Incorporated, 500 Cardigan Road, Shoreview, MN 55126-3996

112

down the right side of the halter, with the other end of the tubing located approximately two

inches above the nostril (Figure 2).

The use of four feet of tubing or less was recommended by the manufacturer
(personal communication) in order to minimize sampling losses due to impaction/adhesion
onto the inner walls of the tubing. However, to allow the horse to have full range of motion
of its head and neck and be assured that the sampling tube remain connected to the sampler,
it was necessary to extend the length of the tube to six feet. In order to determine potential
losses due to impaction/sedimentation in the sampling tube, we took 6 matched one minute
samples (six different horse/tubing set-ups) at the nostril of the horse using a TSI DustTrak1
monitor and compared 1-minute average concentrations to those measured by the personal

monitor/tubing conﬁguration.

Personal monitors were programmed to record PM10 concentrations every second
(Figure 3) for the duration of each of three sampling periods, which were midday (1 1a.m. to
4 pm), evening (5 pm. to 11 pm), and overnight (11:30 pm. until 6 am.) It was
necessary to remove the monitors between sampling periods to download data,
clean/calibrate/check ﬂow rates, and recharge/replace battery packs. We were unable to

take measurements between 6 am. and 11 am. due to stable and training activities.

113

Data Analysis

For each horse and each sampling time segment (mid-day, evening, and
overnight), average, maximum, and minimum PM10 concentrations were recorded. In
addition, the number of 1 second concentrations exceeding 0.569 mg/m3 (a concentration
equal to or above which is associated with the presence of tracheal mucus) (Millerick-
May et al. 2008b), 1 mg/m3, and 10 mg/m3 was calculated. Time weighted average
(TWA) concentrations for each horse were calculated from average concentrations/length
of time sampled (in .5 hour segments) recorded during each of the three sampling
segments. In the event of equipment failure, any concentrations measured during that

sampling period were not used.

Pearson’s chi-square analysis was used to determine which factors (age, stable,
month) were associated with M832 (Holcombe et al. 2006). As personal monitoring data
reﬂects not only concentrations of particulates liberated in close proximity to the nostril,
but also area concentrations of dust previously determined to be inﬂuenced by other
factors (i.e. month, stable, and time of day) (Millerick-May et al. 2008a), we utilized the
personal monitoring data collected during the case/control portion of our study to
determine whether or not these samples were inﬂuenced by such factors (month). For
these analyses, it was necessary to use Fisher’s exact test in lieu of Pearson’s chi-square,
as the expected frequency in certain cells was determined to be less than 5. Therefore,
the particulate data were grouped into ‘high’ and ‘low’ categories for each of the

sampling time points (mid-day, evening, and overnight) and were analyzed by month.

114

Effect of month on inﬂammatory cell counts and percentages were also
determined using Fisher’s exact test. In order to determine the association between the
particle concentrations and cell count data ﬁ'om tracheal lavage, Spearman correlation
coefﬁcients were employed. Inﬂammatory cell counts by cell type (c. g. neutrophils), as
well as percent of each type of inﬂammatory cell were analyzed to determine whether or

not there was a signiﬁcant effect of particulate exposure.

Conditional logistic regression procedures (PROC PHREG) were employed to
determine if there were signiﬁcant differences in personal exposures to particulates
during mid-day, evening, and overnight sampling periods within matched cases and
controls. Time-weighted average exposures, as well as numbers of 1 second samples
above 0.569, 1, and 10 mg/m3 were also considered. These procedures were also utilized
to determine weather there was a difference in total neutrophils between cases and their

matched controls. Odds ratios and 95% conﬁdence limits were calculated.

All data analysis was conducted using SAS v.9.l. Level of signiﬁcance was set at

P5005.

Results

The population consisted of a total of 653 horses from 6 racing stables. All but 4
horses (n=649) were endoscopically examined and a mucus score of 0-5 assigned (Table 1).
Characteristics (age, gender, and stable) of horses selected for inclusion into the study are

listed in Tables 2&3. At the conclusion of the 7-month sampling period, we had obtained

115

personal monitoring data on a total of 113 cases and controls, with a total of 104 matched
pairs. We obtained an average of 17.5 hours of personal particulate monitoring data per
horse, giving rise to a total of 6,557,000 l-second data points for inclusion in our analysis.
Median PM concentrations, and numbers of 1-second samples above 3 thresholds (0.569, 1,

and 10 mg/m3) for cases and controls are presented in Table 4.

Validation of Particulate Measurements
Comparison of simultaneousl-minute measurements of particle concentrations at the
end of the tubing by the nostril with those recorded by the monitor located on the surcingle,

revealed losses due to the tubing attachment ranged from 9-11% (data not shown).

Mucus Scores and Particle Concentrations
In the 649 endoscopic examinations, the overall prevalence of M822 was 23.1
percent (n=150) (Table 1). Bivariate analysis revealed a signiﬁcant effect of trainer

(p<0.0153) and month (p<0.0430) on MSZZ.

There was a signiﬁcant effect of month on the following measurements of particle
concentrations: midday average (p<0.0077) and minimum (p<0.0280) concentrations,
evening average (p<0.001) (Figure 5) and minimum (p<0.0395) concentrations, overnight
average (p<0.0005) (Figure 4), minimum (p<0.0028), and maximum (p<0.0008)
concentrations, time-weighted average concentrations (0.0001) and number of data points

(seconds) (p<0.0187) exceeding 1 mg/m3.

116

Differences in mean concentrations of PM10 throughout the three sampling periods
were negligible, however, signiﬁcant differences did exist between evening and overnight
minimum concentrations (p<0.0066), as well as between midday and evening minimum

concentrations (p<0.02) with evening concentrations being the greatest in both instances.

Inﬂammatory Cell Counts
There was a signiﬁcant effect of month on percent of neutrophils (p<0.0049),
macrophages (p<0.0013), and eosinophils (p<0.023 7) in tracheal wash and also on number

of eosinophils (p<0.0201).

Evening average (p<0.0002), mid-day average (p<0.0280), overnight average
(p<0.0234), and time-weighted average (p<0.0005) PM10 concentrations were signiﬁcantly
associated with absolute neutrophil counts, and evening average concentrations were
signiﬁcantly associated with numbers of macrophages (p<0.0066) and lymphocytes
(p<0.0324). Number of eosinophils was associated with mid-day average (p<0.0164) and
overnight maximum (p<0.0017) PM concentrations. There was a signiﬁcant difference
between cases and their matched controls in numbers of neutrophils (p<0.0486, OR 1.863,

95%CI (1.004 — 3.456), with cases having greater numbers in relation to controls.

Variation in PM Exposures between Matched Cases (M532) and Controls (MS =0)
Analysis of differences between matched case/control exposures to PM during three
sampling periods (evening, overnight, and mid-day) revealed signiﬁcant differences only in

exposure to evening maximum PM concentrations (p<0.0265, OR 2.083, 95% CI (1.090-

117

3.092), and overnight minimum PM concentrations (p<0.0309, OR 2.560, 95% CI (1 .090-
6.013). Signiﬁcant differences in exposure between cases and controls were not identiﬁed

with the remaining measures of PM1 0.

Discussion

This is the ﬁrst reported large-scale case-control study of the association between
personal exposures to airborne particulates (PM10) in horses with (M822) and without
(MS=O) visible accumulations of tracheal mucus. In a previous study, we reported
signiﬁcant associations of time of day, month, stable, and stall on concentrations of PM, and
the presence (M521) of tracheal mucus, but no signiﬁcant association between area
concentrations and MS22 even though the data showed similar trends to that of M831. As a
result, we hypothesized that greater personal rather than area PM exposure was responsible

for MS_>_2.

Particulate concentrations in the breathing zone of horses have previously been
measured by both the traditional pump and ﬁlter methods (Woods PS 1993), allowing for
gravimetric analysis and speciation of dusts over a set period of time, as well as the use of
real-time monitors (Clements and Pirie 2007a, b; Millerick-May et al. 2007), which
provide data in terms of average, minimum, and maximum particle concentrations
recorded during the sampling period. There are beneﬁts to using both methods, however,
the real-time monitoring used in the present study allowed for measurement of the
previously uncharacterized range of PM concentrations (peak and minimum) experienced

over a long sampling period. The anatomy of the horse necessitated use of longer

118

sampling tubes than recommended by the manufacturer and indeed, about 10 percent of
particulates were lost. However, because all personal sampler/tubing conﬁgurations were
standardized, matched cases and controls were monitored during the same sampling period,
and our goal was to determine relative differences in exposures rather than absolute values,

these losses were deemed acceptable for the purposes of this study.

Despite any losses due to impaction/sedimentation in the tubing, the PM
concentrations measured with the personal samplers were similar to those previously
measured with the area monitors during the midday and evening time periods (we could not
compare morning measurements because personal samplers had to be removed during that
time). The similarities between personal and area concentrations suggest that PM exposures
are primarily dependent on area concentrations rather than additional contributions by
personal idiosyncratic activities of the horse, such as nostril placement while eating (leaving

nose inside hay pile), rolling, or sleeping while lying down rather than standing.

The overall prevalence of MS_>_2 was 19% in our previous study as compared to
23% in our present study. The latter may be a more accurate reﬂection of true prevalence
because horses were examined for 7 consecutive months during this study as compared to
only 3 months (July, September, and November) in the previous study. Overall trends
between the two studies were consistent; the lowest prevalence occurred during the cooler

early summer months, and the highest during hot dry weather.

119

While there was a signiﬁcant difference between cases and controls and numbers of
neutrophils in tracheal wash ﬂuids, percent neutrophils were not found to be signiﬁcantly
different. Median values for percent neutrophils in both cases and controls exceeded 27%,
with 20% being recognized as the value above which a diagnosis of airway inﬂammation is
generally made. In fact, 59% of horses with mucus scores of 0 (MS=O) had neutrophil

percentages that exceeded the 20% threshold for IAD.

In support of our hypothesis that cases had higher PM exposures than controls, we
found signiﬁcant differences in their exposures to evening maximum, and ovemight
minimum concentrations of PM1 O. In our previous investigation of the relationship between
area PM and the presence of visible mucus, MUC_>_1 also was signiﬁcantly associated with
both maximum and minimum PM10 concentrations. Our failure to ﬁnd more case/control
differences in PM concentrations may be a function of study design. Although the number
of cases and controls were similar in each stable, they were not evenly distributed among the
stables (Table 3). This was because it was sometimes challenging to ﬁnd an adequate
numbers of controls (MS=0) that ﬁrlﬁlled our matching criteria. As a consequence, more
than half of the horses recruited (both cases and controls) came from 2 stables (Table 3).
Because stable design and management determine area PM concentrations (Chapter 3), and
the latter rather than individual factors is the most important determinant of personal
exposure, having large numbers of horses from the same two stables likely reduced our

chances in ﬁnding a difference in PM exposures.

120

We have previously reported that both stable and month inﬂuence area
concentrations of PM and the presence of tracheal mucus and inﬂammatory cells (Millerick-
May et al. 2008b). If, as we suspect that personal exposures are largely the result of area
concentrations, then the data ﬁ'om personal samplers should reﬂect monthly changes in area
PM. For this reason, we examined temporal trends in personal PM concentrations and
prevalence of M832 (Figure 4 and 5). Indeed, as personal PM concentrations ﬂuctuated
month-by-month, they were tracked by M832 prevalence. There was a similar tracking
between personal PM10 and percent neutrophils (Figure 6). These observations strongly
suggest that area particulate exposure is an important determinant of the presence of mucus

in the airways.

Elevations in area concentrations of particulate matter (PM10 & PM2.5) are known
to result in worsening of pre-existing respiratory disease in humans, increased lost
work/school days, and increased mortality (Brunekreef and Forsberg 2005; Pope 1991 a;
Pope III 2000 ; Pope III et al. 2002; Schwartz 2004). Control of area particulate
concentrations within the stable environment is much easier than attempting to control
individual behaviors (e. g. eating, rolling, pacing, etc.) that inﬂuence personal dust
exposures. If indeed it is the area concentrations of PM that predominantly inﬂuence
personal exposures as our data suggests, and it is these concentrations that, in turn, are
associated with tracheal mucus (M832), efforts to improve ventilation within stables, and
to follow dust management practices aimed at minimizing dispersion of dust within the

stable environment will reduce the prevalence of tracheal mucus in horses within those

121

stables, and during months where factors external to the stable environment (e. g. weather

conditions) also inﬂuence PM concentrations.

122

Table 4-1. Summary of number of examinations performed per stable, age, gender, and
assigned mucus score, as well as the prevalence of M832. *4 horses unable to be

endoscopically examined

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mucus Affecting
Age Gender Mucus Score“ Racin (M832)
Stable 2 3 34 Unk. M F 0 l 2 33 # %
1 5 27 53 61 24 22 49 12 1 13 15.5
2 116 96 26 112 126 75 95 51 16 67 28.3
3 22 31 34 44 44 23 51 10 3 13 14.9
4 22 57 82 88 73 66 59 25 10 35 21.9
5 4 17 21 1 23 19 12 15 ll 4 15 35.7
6 0 21 18 21 18 5 27 6 1 7 18
Total 169 249 234 1 349 304 203 296 1 15 35 150 23.1

 

 

 

 

 

 

 

 

 

123

Table 4-2. Summary of age and gender of cases and controls included in the

study

 

 

 

 

 

Age Gender
2 3 34 M F
Case (M832) 15 24 14 26 27
Control (MS=O) 16 25 19 29 31
Total 31 49 33 55 58

 

 

 

 

 

 

 

 

124

 

Table 4-3. Summary of numbers of cases and controls selected from each stable

Controls
7
24
5

17
6
1

0

 

125

Table 4-4. Description of the average, minimum, maximum, and time-weighted average
concentrations of PM1 O (mg/m3) during each of three sampling periods, as well as the

nmnber of 1-second samples exceeding 0.569, 1, and 10 mg/m3 as experienced by cases

and controls.

 

Case

Median (5th-95th Dctl)

Control
Median (5th-95th

pctl)

 

Evening Average (mg/m3)

0.075 (0.023 - 0.254)

0.059 (0.022 - 0.235)

 

Evening Maximum (mg/m3)

12.346 (1.940 - 19.140)

9.402 (2.758 - 19.392)

 

EveningMinimum Eng/m3)

0.003 (0 - 0.056)

0.002 (0 - 0.056)

 

 

 

Midday Average (mg/m3) 0.073 (0.024 - 0.197) 0.072 (0.017 - 0.149)
10.933 ( 3.868 - 11.311 (3.189 —

Midday Maximum (mg/m3) 19.071) 19.496)

Midday Minimum (mg/m3) 0 (0 - 0.046) 0 (0 - 0.031)

 

Overnight Average (mm?)

0.059 (0.017 - 0.178)

0.054 (0.018 - 0.179)

 

11.463 (3.001 -

 

 

 

Overnight Maximum (mg/m3) 9.733 (1.659 - 19.100) 19.406)
Overnight Minimum (EM) 0 (0 - 0.055) 0 (0 — 0.037)
Time-Weighted Average (mg/m3) 0.070 (0.030 - 0.178) 0.075 (0.022 - 0.167)
# ofpoints >0.569 mg/m3 353 (46 - 892) 231 (32 - 1020)

 

# of points > 1 mg/m3

3(0- 18)

4 (0-14)

 

 

# of points > 10mg/m3

 

726 (122 - 1977)

 

522 (100 - 1778)

 

126

 

 

 

 

 

 

6
2° 5
E
014
E
6’ 3
(I)
o
6 2
<
1
O .
26 Sat 3AM 6AM
Aug 2006 Date & Time

Figure 4-1. Example of 1 second tracings of PM1 0 concentrations of a selected case.

127

 

Figure 4-2. Picture of personal monitor attached to a surcingle, with sample tubing along

the length of the horse’s neck, and ending near the nostril.

128

 

Figure 4-3. Location of sampling tube close to the nostril

129

 

$6 of Concentrations In 'ngh' Category for TWA

 

April May June July August September October

Figure 4-4. Percent of PM10 concentrations (grey bars) that fall into the ‘high’
calculated time-weighted average (TWA) category by sampling month, and percent

neutrophils (black line) by month plotted along the same axis.

130

Percent Neutrophlle

56 01 Horses with MS>2

 

x of Overnight PM Concentnﬂone In ‘ngh' Category

 

April May June July August September October

Figure 4-5. Percent of overnight PM10 concentrations (grey bars) that fall into the ‘high’
category by sampling month, and percent of horses with M832 (black line) by month

plotted along the same axis.

100 y

‘1
O
. #4 4 A- 44’ 14 4,A4

% of Horeee with us>2

 

56 of Evening PM Samples In the 'ngh' Category
01
O

  

oi». , . a? 1+ 1

April May June July August September October

 

Figure 4-6. Percent of evening PM10 concentrations (grey bars) that fall into the ‘high’
category by sampling month, and percent of horses with MS_>_2 (black line) by month

plotted along the same axis.

132

References

ACGIH (2005a) American Conference of Governmental Industrial Hygienists TLVS and
BEIs. Threshold Limit Values for Chemical Substances and Physical Agents and
Biological Exposure Indices. Cincinnatti, OH.

ACGIH (2005b) American Conference of Governmental Industrial Hygienists TLVs and
BEIs. Threshold limits values for chemical substances and physical agents and biological
exposure indices., Cincinnati, Ohio.

ACGIH (2005c) Documentation of threshold limits values: Ammonia. In: Supplement to
the 7th edition. ACGIH Worldwide Signature Publications.

Anon (2001) Chairperson's introduction: International Workshop on Equine Chronic
Airway Disease, Michigan State University, 16-18 June 2000. Equine Veterinary Journal
33, 5-19.

Anon (2003) Workshop report: Inﬂammatory airway disease: deﬁning the syndrome.
Equine veterinary education 5, 81-82.

Anon (2005) T81 Model 8520 DustTrak Aerosol Monitor. Operation and service manual.
1980198 revision Q.

Anon (2006) National ambient air quality standards, US Environmental Protection
Agency.

Bartner, L.R., Tesfaigzi, Y. and Robinson, NE. (2005) Persistent mucus accumulation: A
consequence of delayed mucous cell death in heaves-affected horses? American Journal
of Respiratory Critical Care Medicine 171, In press.

Becker, S., Soukup, J .M., Gilmour, MI. and Devlin, RB. (1996) Stimulation of human
and rat alveolar macrophages by urban air particulates: effects on oxidant radical

generation and cytokine production. Toxicology and Applied Pharmacology 141, 637-
648.

Bemdt, A., Derksen, F .J ., Venta, P., Ewart, S., Yuzbasiyan-Gurkan, V. and Robinson,
NE. (2007) Elevated amount of toll-like receptor 4 mRNA in bronchial epithelial cells is
associated with airway inﬂammation in horses with recurrent airway obstruction. Am J
Physiol Lung Cell Mol Physiol 292, 936-943.

Brunekreef, B. and F orsberg, B. (2005) Epidemiological evidence of effects of coarse
airborne particles on health. European Respiratory Journal 26, 309-318.

Chapman, P.S., Green, C., Main, J .P., Taylor, P.M., Cunningham, F.M., Cook, A.J. and
Marr, CM. (2000) Retrospective study of the relationships between age, inﬂammation

133

and the isolation of bacteria from the lower respiratory tract of thoroughbred horses.
Veterinary Record 146, 91 -95.

Christley, R.M., Hodgson, D.R., Rose, R.J., Hodgson, J .L., Wood, J .L. and Reid, S.W.
(2001a) Coughing in thoroughbred racehorses: risk factors and tracheal endoscopic and
cytological ﬁndings. Veterinary Record 148, 99-104.

Christley, R.M., Hodgson, D.R., Rose, R.J., Wood, J .L., Reids, S.W., Whitear, K.G. and
Hodgson, J .L. (2001b) A case-control study of respiratory disease in Thoroughbred
racehorses in Sydney, Australia. Equine Veterinary Journal 33, 256-264.

Clarke, AF. and Madelin, T. (1987) Technique for assessing respiratory health hazards
from hay and other source materials. Equine Veterinary Journal 19, 442-447.

Clarke, A.F., Madelin, T. and Alpress, R.G. (1987) The relationship of air hygiene in
stables to lower airway disease and pharyngeal lymphoid hyperplasia in two groups of
Thorough-bred horses. Equine Veterinary Journal 19, 524-530.

Clements, J .M. and Pirie, R.S. (2007a) Respirable dust concentrations in equine stables.
Part 1: Validation of equipment and effect of various management systems. Research in
Veterinary Science 83, 256-262.

Clements, J .M. and Pirie, R.S. (2007b) Respirable dust concentrations in equine stables.
Part 2: The beneﬁts of soaking hay and optimising the environment in a neighbouring
stable. Research in Veterinary Science 83, 263-268.

Crichlow, E.C., Yoshida, K. and Wallace, K. (1980) Dust levels in a riding stable. Equine
Veterinary Journal 12, 185-188.

Derksen, F .J ., Robinson, N.E., Armstrong, P.J., A., 8.]. and Slocombe, RF. (1985)
Airway reactivity in ponies with recurrent airway obstruction (heaves). Journal of
Applied Physiology 58, 598-604.

Dye, J .A., Adler, K.B., Richards, J .H. and Dreher, KL. (1999) Role of soluble metals in
oil ﬂy ash-induced airway epithelial injury and cytokine gene expression. American
Journal of Physiology: Lung Cellular and Molecular Physiology 277 , L498-510.

Fubini, B. and Hubbard, A. (2003) Reactive oxygen species (ROS) and reactive nitrogen
species (RN S) generation by silica in inﬂammation and ﬁbrosis. Free radical biology and
medicine 34, 1507-1516.

Galson (2006) Sampling and analysis guide.

Gerber V, L.A., Berney C, Robinson NE (2004) Airway mucus in recurrent airway

obstruction--short-term response to environmental challenge. J Vet Internal Med. 18, 92-
97 .

134

Gerber, V., Lindberg, A., Bemey, C. and Robinson, NE. (2004a) Airway mucus in
recurrent airway obstruction--short-tenn response to environmental challenge. J Vet
Intern Med 18, 92—97.

Gerber, V., Straub, R., Marti, E., Hauptman, J ., Herholz, C., King, M., Imhof, A., Tahon,
L. and Robinson, N.E. (2004b) Endoscopic scoring of mucus quantity and quality:
observer and horse variance and relationship to inﬂammation, mucus viscoelasticity and
volume. Equine Veterinary Journal 36, 576 - 582.

Ghio, A.J. and Devlin, RB. (2001) Inﬂammatory Lung Injury aﬁer Bronchial Instillation
of Air Pollution Particles. pp 704-708.

Ghio, A.J., Kennedy, T.P., Whorton, A.R., Crumbliss, A.L., Hatch, GE. and Hoidal, J .R.
(1992) Role of surface complexed iron in oxidant generation and lung inﬂammation

induced by silicates. American Journal of Physiology: Lung Cellular and Molecular
Physiology 263, L51 1-518.

Ghio, A.J., Mazan, M.R., Hoffman, A.M. and Robinson, NE. (2006) Correlates between
human lung injury after particle exposure and recurrent airway obstruction in the horse.
Equine Vet J38, 362-367.

Gillies, J .A., Gertler, A.W., Sagebiel, J .C. and Dippel, W.A. (2001) On-road particulate
matter (PM2.5 and PM10) emissions in the Sepulveda Tunnel, Los Angeles, California.
Environ Sci T echnol. 15; 35, 1054-1063.

Gilmour, P.S., Brown, D.M., Lindsay, T.G., Beswick, P.H., MacNee, W. and Donaldson,
. K. (1996) Adverse health effects of PM10 particles: involvement of iron in generation of
hydroxyl radical. Occupational and Environonmental Medicine 53, 817-822.

Gonzalez-Flecha, B. (2004) Oxidant mechanisms in response to ambient air particles.
Molecular Aspects of Medicine 25, 169-182.

Holcombe, S.J., Jackson, C., Gerber, V., J efcoat, A., Bemey, C., Eberhardt, S. and
Robinson, NE. (2001) Stabling is associated with airway inﬂammation in young Arabian
horses. Equine Veterinary Journal 33, 244-249.

Holcombe, S.J., Robinson, N.E., Derksen, F .J ., Berthold, B., Genovese, R., Miller, R.,
DeFeijter-Rupp, H., Carr, E.A., Eberhard, S.W., Boruta, D. and Kaneene, J .B. (2006)
Effect of tracheal mucus and tracheal cytology on racing performance in Thoroughbred
racehorses. Equine Vet J 38(4), 300-304.

Holcombe, S.J., Robinson, N.E., Derksen, F.J. and Kaneene, J. (2004) Tracheal mucus is
associated with poor racing perfprrnance in thoroughbred horses. In: 5 0th Annual
Convention of the American Association of Equine Practitioners, AAEP, Denver. p In
press.

135

Kappos, A.D., Bruckmann, P., Eikmann, T., Englert, N., Heinrich, U., Hoppe, P., Koch,
E., Krause, G.H., Kreyling, W.G., Rauchfuss, K., Rombout, P., Schulz-Klemp, V., Thiel,
W.R. and Wichmann, HE. (2004) Health effects of particles in ambient air. Int J Hyg
Environ Health 207, 399-407.

Kirschvink, N., Di silvestro, F ., Sbai, 1., Vandenput, S., Art, T., Roberts, C. and Lekeux,
P. (2002) The Use of Cardboard Bedding Material as Part of an Environmental Control
Regime for Heaves-affected Horses: In Vitro Assessment of Airborne Dust and

Aeroallergen Concentration and In Vivo Effects on Lung Function. The Veterinary
Journal 163, 319-325.

Li, X.Y., Gilmour, P.S., Donaldson, K. and MacNee, W. (1997) In vivo and in vitro
proinﬂammatory effects of particulate air pollution (PM10). Environ Health Perspect 105
Suppl 5, 1279-1283.

Littel, R.C., Milliken, G., Stroup, W. and Wolﬁnger, R. (1996) SAS System for mixed
models, SAS Institute Inc., Cary, NC. p 633.

MacNamara, 3., Bauer, S. and Iafe, J. (1990a) Endoscopic evaluation of exercise-induced
pulmonary hemorrhage and chronic obstructive pulmonary disease in association with
poor performance in racing Standardbreds. Journal of the American Veterinary Medical
Association 196, 443-445.

MacNamara, 3., Bauer, S. and Iafe, J. (1990b) Endoscopic evaluation of exercise-
induced pulmonary hemorrhage and chronic obstructive pulmonary disease in association
with poor performance in racing Standardbreds. Journal of the American Veterinary
Medical Association 196, 443-445.

Malikides, N. and Hodgson, J .L. (2005) Endotoxin is an Early Component of the Causal
Pathway in Inﬂammatory Airway Disease (IAD) in Yount TB Racehorses. In: 2005
World Equine Airways Symposium, Ithica, New York.

McGorum, B.C., Dixon, RM. and Halliwell, R.E.W. (1993a) Quantiﬁcation of histamine
in plasma and pulmonary ﬂuids from horses with chronic obstructive pulmonary disease,

before and after "natural (hay and straw) challenges". Veterinary Immunology and
Immunopathology 36, 223-237.

McGorum, B.C., Dixon, PM. and Halliwell, R.E.W. (1993b) Responses of horses
affected with chronic obstructive puhnonary disease to inhalation challenges with mould

antigens. Equine Veterinary Journal 25, 261-267.

McGorum BC, E.J., Cullen RT (1998 Sep) Total and respirable airborne dust endotoxin
concentrations in three equine management systems. Equine Vet J. 30, 430-434.

136

McGorum, B.C., Ellison, J. and Cullen, RT (1998) Total and respirable airborne dust
endotoxin concentrations in three equine management systems. Equine Veterinary
Journal 30, 430-434.

Millerick-May, M., Derksen, F .J . and Robinson, NE. (2007) Indoor Air Quality in a
Boarding Stable. Equine Vet J Under Revision.

Millerick-May, M., Karmaus, W., Derksen, F.J., Berthold, B., Holcombe, SJ. and
Robinson, N.E. (2008a) Evaluation of Factors Affecting the Concentration of Particulate
Matter (PM10 and PM2.5) in Stables at an American Thoroughbred Racetrack. Equine
Vet J Submitted.

Millerick-May, M., Karmaus, W., Derksen, F.J., Berthold, B., Holcombe, SJ. and
Robinson, N.E. (2008b) Horse-Environment Interaction: the relationship between
environmental particulate matter and airway inﬂammation in Thoroughbred racehorses.
Equine Vet J Submitted.

Morgenstem, V., Zutavem, A., Cyrys, J ., Brockow, I., Gehring, U., Koletzko, S., Bauer,
C.P., Reinhardt, D., Wichmann, HE. and Heinrich, J. (2007) Respiratory health and
individual estimated exposure to trafﬁc-related air pollutants in a cohort of young
children. pp 8-16.

Morrow, RE. (1992) Dust overloading of the lungs: update and appraisal. T oxicol Appl
Pharmacol 113, 1-12.

Newton, J .R., Wood, J .L. and Chanter, N. (20033) A case control study of factors and
infections associated with clinically apparent respiratory disease in UK Thoroughbred
racehorses. Preventive Veterinary Medicine 60, 107-132.

Newton, J .R., Wood, J.L.N. and Chanter, N. (2003b) A case control study of factors and
infections associated with clinically apparent respiratory disease in UK Thoroughbred
racehorses. Preventive Veterinary Medicine 60, 107-132.

O'Grady, N.P., Preas, H.L., Pugin, J ., Fiuza, C., Tropea, M., Reda, D., Banks, SM. and
Suffredini, A.F. (2001) Local inﬂammatory responses following bronchial endotoxin

instillation in humans. American Journal of Respiratory Critical Care Medicine 163,
1 591-1 598.

Oberdorster, G. (1995) Lung Particle Overload: Implications for Occupational Exposures
to Particles. Regulatory Toxicology and Pharmacology 21, 123-135.

Oberdorster, G. (1996) Signiﬁcance of particle parameters in the evaluation of exposure-
dose-response relationships of inhaled particles. Inhalation Toxicology 8 Suppl, 73-89.

Oberdorster G (1996) Signiﬁcance of particle parameters in the evaluation of exposure-
dose-response relationships of inhaled particles. Inhal T oxicol. 8 Suppl, 73-89.

137

Oberdorster, G., F inkelstein, J .N., Johnston, C., Gelein, R., Cox, C., Baggs, R. and Elder,
AC. (2000) Acute pulmonary effects of ultraﬁne particles in rats and mice. Res Rep
Health Eﬂ Inst, 5-74; disc 75-86.

Peel, J .L., Tolbert, P.E., Klein, M., Metzger, K.B., Flanders, W.D., Todd, K.,
Mulholland, J .A., Ryan, PB. and Frumkin, H. (2005) Ambient air pollution and
respiratory emergency department visits. Epidemiology 16, 164-174.

Pierse, N., Rushton, L., Harris, R.S., Kuehni, C.E., Silverman, M. and Grigg, J. (2006)
Locally generated particulate pollution and respiratory symptoms in young children. pp
216-220.

Pirie, R.S., Collie, D.D., Dixon, PM. and McGorum, B.C. (2003a) Inhaled endotoxin and
organic dust particulates have synergistic proinﬂammatory effects in equine heaves
(organic dust-induced asthma). Clinical and Experimental Allergy 33, 676-683.

Pirie, R.S., Collie, D.D.S., Dixon, RM. and McGorum, B.C. (2003b) Inhaled endotoxin
and organic dust particulates have synergistic proinﬂammatory effects in equine heaves
(organic dust-induced asthma). Clinical & Experimental Allergy 33, 676-683.

Pope, C. (1991a) Respiratory hospital admissions associated with PM10 pollution in
Utah, Salt Lake, and Cache valleys. Arch Env Health 46, 90-97.

Pope, C.A., 3rd (1991b) Respiratory hospital admissions associated with PM10 pollution
in Utah, Salt Lake, and Cache Valleys. Archives of Environmental Health 46, 90-97.

Pope, C.A., 3rd, Burnett, R.T., Thun, M.J., Calle, E.E., Krewski, D., Ito, K. and Thurston,
GD. (2002) Lung cancer, cardiopulmonary mortality, and long-term exposure to ﬁne
particulate air pollution. Journal of the American Medical Association 287, 1132-1141.

Pope 111, CA. (2000 ) What do epidemiologic ﬁndings tell us about health effects of
environmental aerosols? J Aerosol Med. 13, 335-354.

Pope III, C.A., Burnett, R.T., Thun, M.J., Calle, E.E., Krewski, D., Ito, K. and Thurston,
GD. (2002) Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine
Particulate Air Pollution. JAMA 287, 1132-1141.

Robinson, N.E., Karmaus, W., Holcombe, S.J., Carr, EA. and Derksen, F.J. (2006)
Airway inﬂammation in Michigan pleasure horses: prevalence and risk factors. Equine
Veterinary Journal 38, 293-299.

Schins, R.P., Lightbody, J .H., Borm, F.J., Shi, T., Donaldson, K. and Stone, V. (2004)

Inﬂammatory effects of coarse and ﬁne particulate matter in relation to chemical and
biological constituents. T oxicol Appl Pharmacol 195, 1-11.

138

Schwartz, J. (2004) Air Pollution and Children's Health. Pediatrics 113, 103 7-1043.

Shukla, A., Timblin, C., BeruBe, K., Gordon, T., McKinney, W., Driscoll, K., Vacek, P.
and Mossman, B.T. (2000) Inhaled particulate matter causes expression of nuclear factor
(NF )-kappaB-related genes and oxidant-dependent NF-kappaB activation in vitro.
American Journal of Respiratory Cell and Molecular Biology 23, 182-187.

Singh, R.B., Desloges, C. and Sloan, J .J . (2006) Application of a microscale emission
factor model for particulate matter to calculate vehicle-generated contributions to ﬁne
particle emissions. J Air Waste Manag Assoc. 56, 37—47.

Sunyer, J. (2001) Urban air pollution and chronic obstructive pulmonary disease: a
review. Eur Respir J 17, 1024-1033.

Tremblay, G.M., Ferland, C., Lapointe, J .-M., Vrins, A., Lavoie, J .P. and Corrnier, Y.
(1993) Effect of stabling on bronchoalveolar cells obtained from normal and COPD
horses. Equine Veterinary Journal 25, 194-197.

USEPA National Ambient Air Quality Standards. http://www. epagov/ttn/naaqs/
Referenced April I, 2006.

 

Vandenput, S., Istasse, L., Nicks, B. and Lekeux, P. (1997) Airborne dust and
aeroallergen concentrations in different sources of feed and bedding for horses.
Veterinary Quarterly 19, 154-158.

Wang, A.L., Blackford, TL. and Lee, L.Y. (1996) Vagal bronchopulmonary C-ﬁbers and
acute ventilatory response to inhaled irritants. Respiration Physiology 104, 231-239.

Wanner, A., Salathe, M. and O'Riordan, T.G. (1996) Mucociliary clearance in the
airways. Am J Respir Crit Care Med 154, 1868-1902.

Webster, A.J.F., Clarke, A.F., Madelin, T.M. and Wathes, OM. (1987) Air hygiene in
stables 1: Effects of stable design, ventilation and management on the concentration of
respirable dust. Equine Veterinary Journal 19, 448-453.

Wood, J .L., Burrell, M.H., Roberts, C.A., Chanter, N. and Shaw, Y. (1993) Streptococci
and Pasteurella spp. associated with disease of the equine lower respiratory tract. Equine
Veterinary Journal 25, 314-318.

Wood, J .L., Newton, J .R., Chanter, N. and Mumford, J .A. (2005a) Association between
respiratory disease and bacterial and viral infections in British racehorses. J Clin
Microbiol. 43, 120-126.

Wood, J .L., Newton, J .R., Chanter, N. and Mumford, J .A. (2005b) Inﬂammatory airway

disease, nasal discharge and respiratory infections in young British racehorses. Equine
Veterinary Journal 37, 236-242.

139

Wood, J .L., Newton, J .R., Chanter, N. and Mumford, J .A. (20050) Inﬂammatory airway
disease, nasal discharge, and respiratory infections in young British racehorses. Equine
Vet J37, 236-242.

Woods PS, R.N., Swanson MC, Reed CE, Broadstone RV, Derksen FJ (1993) Airborne
dust and aeroallergen concentration in a horse stable under two different management
systems. Equine Vet J. 25, 208-213.

Woods, P.S., Robinson, N.E., Swanson, M.C., Reed, C.E., Broadstone, RV. and
Derksen, F .J . (1993) Airborne dust and aeroallergen concentration in a horse stable under
two different management systems. Equine Veterinary Journal 25, 208-213.

Zeger, S.L. and Liang, KY. (1986) Longitudinal data analysis for discrete and
continuous outcomes. Biometrics 42, 121-130.

Zelikoff, J .T., Chen, L.C., Cohen, M.D., Fang, K., Gordon, T., Li, Y., Nadziejko, C. and
Schlesinger, RB. (2003) Effects of inhaled ambient particulate matter on pulmonary
antimicrobial immune defense. Inhalation Toxicology 15, 131-150.

Zelikoff, J .T., Schermerhom, K.F., Cohen, MD. and Schlesinger, RB. (2002) A Role for
Associated Transition Metals in the Immunotixicity of Inhaled Ambient Particulate
Matter. Environ Health Perspect 110, 871-875.

Zhou, H. and Kobzik, L. (2007) Effect of Concentrated Ambient Particles on

Macrophage Phagocytosis and Killing of Streptococcus pneumoniae. Am J Respir Cell
Mol Biol. 36, 460-465.

140

CONCLUSIONS

This investigation focused on the exposure of horses to airborne particulate matter
(PM) within the stable and its effects on measures of airway inﬂammation (tracheal
mucus and inﬂammatory cells) in Thoroughbred racehorses. The purpose of my work
was to quantify these exposures, better understand the factors that inﬂuence them, and
determine whether or not there was a signiﬁcant effect on airway inﬂammation. My
overall hypothesis was that ambient particle load and personal particulate exposure varies
with stable, stall location within a stable, and activity level, and correlates in an exposure-
dependent manner with established measures of airway inﬂammation (mucus score, and
inﬂammatory cell counts) in racehorses.

Determining concentrations of PM that accurately reﬂect area concentrations or
personal exposures is difﬁcult in indoor environments when sources of PM generation,
dispersion, and ventilation are relatively constant, as these measurements are simply a
‘snapshot in time’. Add to this the complexity of monitoring an environment such as a
stable in which there is a potential for substantial variation in PM generation and dispersion
due to factors that include: weather patterns, ﬁxed (doors/windows) and non-ﬁxed (fans)
sources of ventilation, inconsistencies in human work habits, variability in feed/bedding
quality, and adaptation of equipment originally developed for human use to be worn by
horses, and one has a very inconsistent environment in which to attempt to characterize PM
concentrations. Attempts were made to overcome these variations in the present indoor air
quality/ airway inﬂammation investigation by utilizing a repeated measures study design in a
limited number of stables, allowing me to determine whether or not there were relatively

‘constant’ differences in PM concentrations over time and between stables. While this study

141

design allowed me to investigate variations in PM concentrations and measures of airway
inﬂammation by month (PM and inﬂamm.), time of day (PM and inﬂamm.), stable (PM
only), and stall (PM and inﬂamm.), I was unable to perform an analysis that would
incorporate the human element (effect of management style/ stable) on airway inﬂammation
as it was likely a surrogate for PM concentration. As horses were endoscopically examined
only once per month and PM monitoring took place during the same time frame, I was only
able to make statements regarding PM concentrations and their association to airway
inﬂammation (mucus score and inﬂammatory cells) at that particular point in time. This
study design did not allow me to track PM concentrations and airway inﬂammation
throughout each month and over several months, which would allow for further clarity in
terms of duration of time horses experienced particular levels of airway inﬂammation (e. g.
moving through a range of mucus scores) and their association to PM concentrations
measured over prolonged periods of time (days/months).

For the case-control study, it would have been ideal to be able to monitor each horse
over several days. However, in order to achieve the power required to detect a difference in
exposure between cases and controls, it was necessary to study each horse for no more than
24 hours. Therefore, it is possible that I was not able to truly capture the variability in
personal exposures due to intra—animal variation in behavior over time (e. g. reduced
movement due to exhaustion after a race). I was unable to monitor the horses during the
early morning period, a time at which 1 determined there to be signiﬁcant elevations in area
PM concentrations, due to the training schedule of the horses, which also resulted in an
underestimation of personal PM exposures. While I made every attempt to select

cases/controls from a ‘pool’ of horses that were evenly distributed through multiple stables,

142

I was unable to do so, ultimately resulting in a skewed distribution of cases/controls coming
from each stable. This was unavoidable as there were several months in which it was
difﬁcult to ﬁnd horses that would meet the criteria to become a ‘control’ and I had to take
them where I could ﬁnd them. In addition, I quickly found that there were horses from
certain stables that would not tolerate the sampling equipment for one reason or another
(nervousness, etc.), thus reducing the total ‘pool’ of horses from which to choose. I also had
to inevitably drop one stable from our study as the type of bedding used was routinely
plugging the personal monitors resulting in lost/missing data and the inability to utilize those
monitors for the rest of the monthly sampling period.

Despite these inherent limitations to this study, I was able to determine with
certainty that there was an association between airborne PM concentrations as measured by
personal and area monitors, and airway inﬂammation in racehorses. Based on what I have
learned from conducting this study, and the results that are summarized in the following

sections, future studies will be designed to take these limitations into account.

Air Quality in a Boarding Stable

Prior to initiating activities at the racetrack, I wanted to better understand the range
of PM concentrations in the breathing zone of the horse as a result of routine activities and
management practices which commonly occur in stables. As a result, I conducted a small-
scale ﬁeld study in a boarding stable where I was conﬁdant that I would have complete co-

operation by horses and people alike.

143

I chose to use direct—reading instruments in my investigations as they have the ability
to provide not only average concentrations of PM, but also peak and minimum
concentrations, which we hypothesized may be important in the etiology of IAD, as short-
tem1/peak exposures, or chronic low-level exposures had not been previously quantiﬁed in
this particular environment. These monitors are ideal for ‘mapping’ particle concentrations,
allowing the investigator to determine regional ‘hot spots’ of PM, which indicate areas of

poor ventilation, and I had utilized them in the past in industrial settings.

I found that even when horses ate what would be considered by horse owners to be
good quality hay, particle concentrations at the nostril of the horse exceeded short-term
excursion limits set for industrial workers, and that eating ‘slightly dusty hay’ exceeded the
capacity of the monitor (150 mg/m3). Veterinarians often recommend feeding horses with
recurrent airway obstruction (RAO) a diet that includes pellets or hay cubes in order to
minimize dust exposure, but surprisingly, breathing zone concentrations of PM while
consuming these types of feeds were high as well, as they are fed out of deep bins that
essentially trap the particles in the immediate vicinity of the horses nostril. Activities such
as de-cobwebbing, sweeping/raking of the aisles, dragging and riding in the indoor arenas,
also dispersed very high concentrations of PM, which remained suspended in the horses
breathing zone. Fortunately, however, a small trial using simple dust suppression
techniques commonly utilized in industry (application of water) virtually eliminated the dust

when water was applied just prior to initiating the activity (feeding, cleaning, etc.).

144

I used portable pump/colorimetric tube methods to determine ammonia
concentrations within stalls, and found seasonal variations (high in summer, low in winter),
consistent with what I’d expect due to the volatile nature of gasses in the heat/cold. What
was interesting was that ammonia concentrations increased in stalls located furthest from the
doorways (natural ventilation). I also discovered that even though the entire front of the
stalls were ﬁtted with bars to allow for movement of air, the air exchange within these stalls
was minimal, as any air movement/exchange predominantly occurred in the aisleways,
leading me to believe that the majority of decision making in terms of ventilation status (6. g.
doors open vs. doors closed), temperature regulation, and sense of ‘air quality’ (e. g.
dustiness) occurs ‘from the aisleway’ where the stable management, owners, and riders

spend the majority of their time (rather than in the stall with the horse).

Bulk samples of dust were collected from ledges greater than six feet off the ground
to estimate the composition of dust that would pass up through the breathing zone of the
horse and then settle-out. These dusts were analyzed for both particle composition and size,
and I was interested to ﬁnd that two components known to result in airway
inﬂammation/chronic disease in occupational lung disease were present in the stable
environment, namely crystalline silica (sand footing/ﬂooring materials) and iron (crustal

materials).

These initial ﬁndings, which veriﬁed that 1) particulates common in the breathing

zone of horses were of the same size (PM10 and PM2.5) and in concentrations that were of

interest in the current human and animal literature, 2) there were regional variations in

145

ventilation within the stable resulting in ‘hot spots’ of PM, 3) there were certain activities
that generated signiﬁcant concentrations of dust in the breathing zone of horses, and 4) that
components of these dusts were of biological importance, armed me with the knowledge

necessary to go forward and attempt a similar study on a much larger scale.

Indoor Air Quality and Airway Inﬂammation in Thoroughbreds

I was very fortunate to be able to continue a relationship with racetrack veterinarians
and trainers that had been built previously during a study conducted by our laboratory which
ultimately determined that there was a signiﬁcant effect of tracheal mucus (M822) on racing
performance. As a result of this study, these stakeholders were very interested in the
cause(s) of tracheal mucus, and allowed us to perform our study in the hopes that we would
be able to provide recommendations back to them in order to reduce the prevalence of

airway mucus.

I was able to demonstrate that indeed there were signiﬁcant effects of season
(month), time of day, stable, and location within a stable on area concentrations of PM10
and PM2.5. The sampling month that followed a period of hot dry weather resulted in the

highest concentrations of PM, with the lowest occurring after a period of rainy weather.
Morning PM measurements were always the highest because this is when all of the

activity at a racetrack occurs (feeding, cleaning, grooming, training, etc.). Large particles

(PM10) tended to settle out rather quickly, and our mid-day measurements were not

146

signiﬁcantly different from the late afternoon. The smaller particles (PM2.5) settled out

more slowly, particularly in the stable that was the most enclosed.

There was a signiﬁcant effect of stable on area PM concentrations; however, it was
difﬁcult to separate out the effect of stable design vs. management style. The newer open-
air design stable consistently had the lowest concentrations of PM, but also had the least
amount of human activity throughout the day. The two older-style enclosed brick buildings
of the same design differed from one another in terms of PM concentrations most likely due
to orientation of buildings along roadways as well as management/activity level. The stable
that consistently experienced the highest concentrations of PM was ﬂanked on both sides by
busy roadways, but also was the only stable to feed from hay nets, and had the greatest
number of employees, which performed all chores (cleaning, grooming, raking)
simultaneously. Interestingly, during the month of cold weather when the doors and
windows/shutters were closed tight on all three stables, the magnitude of variation in PM

between the stables was reduced.

There was a signiﬁcant effect of stall location on PM concentrations. Stalls that
were located either in hi gh-trafﬁc areas or the ﬁrrthest from natural sources of dilution

ventilation had the highest concentrations of PM as compared to adjacent stalls.

Taking all of this into consideration, I expected to see the highest prevalence of

tracheal mucus during the months with highest PM concentrations, and in the stables and

stalls with the highest PM concentrations. Indeed, the highest prevalence of the presence of

147

tracheal mucus occurred during our September sampling month when PM concentrations
were the greatest, in the stable that consistently had the highest concentrations of PM, and in
those stalls with the highest PM concentrations. I was unable to take stable and stall into
consideration in the hierarchical statistical models, as they were most likely a surrogate for

PM exposure, but bivariate analysis did determine signiﬁcance.

As the study was designed to determine the differences in PM concentrations
between months, stables, and stalls, repeated measures design was necessary. As a result, I
did not have sufﬁcient power to distinguish the relationship between ambient concentrations

of PM and differing degrees of mucus accumulation.

Interpretation of total and percent inﬂammatory cells in regards to airway
inﬂammation was not straight-forward, as numbers and percent of inﬂammatory cells
tracked nicely with mucus prevalence by month, however, they were not signiﬁcantly

associated with the presence (MS_>_1) of tracheal mucus.

Case-Control Study

Based on the particle concentrations that were measured at the level of the nostril in
the boarding stable, and my hypothesis that horses with signiﬁcant (MS32) quantities of
tracheal mucus were exposed to higher concentrations of PM than their counterparts without
visible mucus (MS=0), I decided to use personal real-time monitors to track particle

concentrations over as much of a 24 hour period as possible.

148

Incredibly, we were able, on average, to monitor the horses over a 17.5 hour period
of time, with the monitors taking samples every second so as to ‘catch’ peak exposures that
would result from the horse’s personal activities (eating, rolling, pacing, etc.). We were
unable to monitor the horses during the early morning hours, the time period in our last

study in which PM concentrations were the greatest.

PM concentrations measured with personal samplers in cases and their matched
controls did not differ in most instances; however, the data also suggested that they did not
differ greatly from ambient PM concentrations. Overall prevalence of M822 in all horses
endoscopically examined tracked convincingly with personal monitoring results. As in the
previous study, there was a signiﬁcant effect of month, with the lowest PM concentrations
occurring in the early summer months during periods of wet weather, and peaking late

summer during hot/dry periods.

While there was a signiﬁcant difference between cases and controls in numbers of
neutrophils in tracheal wash ﬂuids, there was no difference in percent neutrophils. As with
the previous study, percent neutrophils tracked nicely with PM concentrations stratiﬁed by
month. Based on our data, I question the useﬁrlness of utilizing percent neutrophils in
tracheal wash samples as a diagnostic criterion for IAD. Our data consistently show no
signiﬁcant differences in percent neutrophils between horses with and without tracheal
mucus, and in fact, in our second study, 82% of horses with no visible tracheal mucus had

greater than 20% neutrophils, similar to what has been found in other studies. As

149

endoscopic examination is readily available, perhaps the criteria for diagnosis of IAD should

be limited to tracheal mucus score.

Practical Implications from the Investigation

The fact that my data suggest that personal monitoring reﬂects primarily area PM
exposures, and that it is area PM that is associated with tracheal mucus, is very exciting.
This means that individuals armed with the right equipment can enter stables and measure
the area PM concentrations during times of peak activity, determine where ‘pockets’ of high
PM/low ventilation exist, and make appropriate recommendations as to how to maximize
ventilation and implement dust management practices within that facility in order to reduce
the prevalence of IAD. Furthermore, because I have demonstrated that simple changes in
management can dramatically reduce particulate load, it should be easy to implement

control measures.

Future Investigations

As I have determined an effect of season and time of day on PM concentrations, it is
no longer necessary to perform repeated monitoring in stables and repeated examinations of
horses. To be able to obtain signiﬁcant power to determine the association between area
particulate concentrations and degree of tracheal mucus, any future study should encompass
additional stables and horses, thus providing additional clarity as to an area
concentration/threshold above which one would expect to ﬁnd horses with signiﬁcant
quantities of tracheal mucus. A similar study in non-racing stables would be beneﬁcial in

order to understand the variation in areat PM concentrations as a result of stable

150

construction/design and management practices, as currently insufﬁcient stables have studied
to deﬁne a ‘normal’ and a ‘dusty’ environment. Limited monitoring of speciﬁc PM
components ‘of concern’ (endotoxin, mold, iron, crystalline silica, etc.) would be useful to
begin to understand what is commonly found in stables, and in what concentrations/ratios.
Finally, an ‘intervention’ study should be performed at a racing stable to determine the

effect of a dust control strategy on the prevalence of mucus within that stable.

151

lllllllllllllljlllllllﬂllllll15116111711