University of Southern California Dissertations and Theses

The association of asthma and asthma-related medications on subclinical atherosclerosis: a cross-sectional analysis of four randomized clinical trials

(USC Thesis Other)

The association of asthma and asthma-related medications on subclinical atherosclerosis: a cross-sectional analysis of four randomized clinical trials

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
THE ASSOCIATION OF ASTHMA AND ASTHMA-RELATED
MEDICATIONS ON SUBCLINICAL ATHEROSCLEROSIS:
A CROSS-SECTIONAL ANALYSIS OF FOUR RANDOMIZED CLINICAL
TRIALS

by
Carlos Eliseo Carballo

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOSTATISTICS)

December 2012

Copyright 2012 Carlos Eliseo Carballo
ii

DEDICATION
Thank you to everyone that has helped me along the way towards the
completion of this Master’s thesis. It has been a long road but one with a fulfilling
end.
I especially want to thank Dr. Mack for her tremendous guidance, wisdom,
and most of all patience throughout this entire process.
Thank you Maria Angelica Cortes for your unconditional support throughout
this journey. Many sacrifices were made and I believe they were worth it. I hope you
feel the same way. You mean the world to me!
Thank you to my mom, Dora M. Garcia, to my brother, Jose Maldonado, and to
the rest of the Maldonado family. I hope the Maldonado kids can gather inspiration
from this accomplishment. I love you all!
Also, thank you to Gary Shields, Steve Moreno, Miguel Chable and John Wells.
Yes. It is finally over. Thank you to George Martinez for the technical help. Thank
you everyone. You know who you are.
iii

TABLE OF CONTENTS
Dedication: ii
List of Tables: iv
Abstract: vi
Chapter 1: Introduction 1

Chapter 2: Materials and Methods
Section 2.1. Study Design 5
Section 2.2. Measurement of Main Outcome – 8
Carotid IMT (CIMT)
Section 2.3. Asthma Status & 10
Respiratory Medication Use
Section 2.4. Measurement of Lipids 12
Section 2.5. Measurement of Vital Signs 13
and Anthropometry
Section 2.6. Assessment of Demographic Information 14
and the Use of Cardiovascular Disease-Related Risk Factors
Section 2.7. Statistical Analysis 15

Chapter 3: Results
Section 3.1. Participant Characteristics 18
Section 3.2. CIMT Comparisons by Asthma Status 27
Section 3.3. Association of Asthma with CIMT 32
Excluding WELLHART Trial Data
Section 3.4. CIMT Comparisons by Use of 34
Respiratory Medication Status
Section 3.5. Association of Respiratory Medication 42
Use with CIMT
Section 3.6. Association of Respiratory Medication Use 44
with CIMT Excluding WELLHART Trial Data

Chapter 4: Discussion 45

Bibliography 51
iv

LIST OF TABLES
Table 1: Demographic and Clinical 18-21
Characteristics by Trial

Table 2: Main Characteristics of 22-24
Asthmatics vs. Non-Asthmatics

Table 3: Main Characteristics of 25-26
Asthmatics vs. Non-Asthmatics
(Excluding WELLHART Trial)

Table 4: CIMT Comparisons by Asthma Status by 28
Cholesterol Risk Groups Among Total Sample (n=1284)

Table 5: CIMT Comparisons by Asthma Status by 28
Cholesterol Risk Groups
Excluding WELLHART Trial (n=1058)

Table 6: Association of Asthma Status on CIMT 29
by Trial (n=1284)

Table 7: Linear Regression Relating Asthma Status to CIMT 29

Table 8: Main Characteristics of the Users of 34-35
Respiratory Medications vs. the Non-Users
of Respiratory Medications

Table 9: Main Characteristics of the Users of 36-38
Respiratory Medications vs. the Non-Users
of Respiratory Medications
(Excluding WELLHART Trial)

v

Table 10: CIMT Comparisons by Respiratory Medications 40
Use Status by Cholesterol Risk Groups Among
Total Sample (n=1284)

Table 11: CIMT Comparisons by Respiratory Medications 40
Use Status by Cholesterol Risk Groups
Excluding WELLHART Trial (n=1058)

Table 12: Association of Respiratory Medications Use 41
Status on CIMT by Trial (n=1284)

Table 13: Linear Regression Relating the Use of 42
Respiratory Medications Status to CIMT
vi

ABSTRACT
Prior studies have evaluated the role of inflammation on atherosclerosis. The
inflammatory process observed in asthmatics plays a central role in the cascade of
events that may result in cardiac events such as plaque erosion and plaque fissuring that
can trigger Cardiovascular Disease (CVD) complications such as myocardial infarction
and stroke. The current research evaluates the association between carotid artery
intima-media thickness (CIMT), a measure of subclinical atherosclerosis, and asthma.
We also evaluated the association between CIMT and use of asthma respiratory
medications.
Baseline data from 1248 subjects in four randomized clinical trials were used to
investigate the association between CIMT and asthma and respiratory medication (anti-
inflammatory non-steroidal, anti-inflammatory steroidal, bronchial dilator and
leukotriene receptor antagonist) use. The primary outcome measure for all trials was
CIMT. Asthmatics were identified by self-report on a baseline health questionnaire.
Respiratory medication use was ascertained through collection of non-study medication
use (prescription and non-prescription) that used a standard medication-coding scheme
for all trials. Data were analyzed using multiple linear regression, with CIMT as the
dependent variable and self-reported asthma and use of respiratory medications as the
primary independent variables. Models were adjusted for age, gender, trial, diabetes,
race, weight, systolic blood pressure, high-density lipoprotein (HDL) cholesterol,
smoking status, education and income when evaluating the association with asthma.
vii

Models were adjusted for age, gender, trial, diabetes, race, weight, systolic blood
pressure, HDL cholesterol risk group, smoking status and education when evaluating the
association with respiratory medication use.
Mean CIMT did not differ between asthmatics and non-asthmatics in unadjusted
(p=0.95) and adjusted (p=0.92) models. Exclusion of the subjects with pre-existing
coronary artery disease at baseline (n=226) did not alter these results (p=0.81 for
comparison of CIMT between asthmatics and non-asthmatics).
Similar non-significant results were observed evaluating the association of
respiratory medication use on CIMT (p=0.66 in adjusted models in the total sample;
p=0.87 in adjusted models excluding subjects with coronary artery disease)
These data provide no evidence of an association between self-reported asthma,
use of respiratory medications, and CIMT measured cross-sectionally. Future data
analyses in these trials should utilize the longitudinally-collected data to evaluate the
association of asthma and respiratory medications with progression of atherosclerosis.

1
CHAPTER1: INTRODUCTION

Cardiovascular disease (CVD) accounts for more deaths than any other major
cause of death in the United States; CVD claims more lives each year than cancer,
chronic lower respiratory disease (CLRD), and accidents combined (Roger et al.,
2011). As the central underlying pathology for CVD, atherosclerosis is a systemic
process in which fatty deposits, inflammatory cells, and scar tissue aggregate within
the arterial wall. Atherosclerosis is a life-long process involving lipid peroxidation
and inflammation affecting the vascular wall (Künzli et al., 2010), and develops in
both large and small arteries feeding the heart, brain, kidneys, and extremities
(Roger et al., 2011).
Atherosclerosis is a complex disease initiated by the trapping and oxidation
of low-density lipoprotein in the sub-endothelial layer of the arterial wall, followed
by the generation of biologically active species that stimulate vascular cells to
produce inflammatory molecules (Cipollone et al., 2005). There is increasing
evidence that inflammation plays a central role in the cascade of events resulting in
plaque erosion and fissuring which trigger the key events that lead to myocardial
infarction and stroke (Cipollone et al., 2005).
An association between asthma and atherosclerosis (and ultimately, CVD)
has been suggested (Crosslin et al., 2009). Derivatives of arachidonic acid, such as
the leukotrienes that are produced by the 5-lipoxygenase (5-LO)-pathway, have long
been known for their inflammatory properties as well as their involvement with

2
asthma (Crosslin et al., 2009). Leukotriene-based inflammation has also been shown
to play an important role in atherosclerosis (Funk et al., 2005). ALOX5AP and
LTA4H, genes in the leukotriene biosynthesis pathway, have individually been
shown to be associated with cardiovascular disease (CVD) phenotypes such as
coronary artery disease and myocardial infarction (Helgadottir et al. 2004)).
Research involving the 5-lipoxygenase (5-LO) enzyme and arachidonic acid
has generally focused on asthma and other inflammatory disorders. More recently,
investigation related to the 5-LO enzyme has focused on CVD, including
atherosclerosis, myocardial infarction, stroke, and plaque stability (Zhao & Funk,
2004). The production of leukotrienes in inflammatory cells begins with the
cleavage of arachidonic acid from nuclear membrane glycerophospholipids. The 5-
LO enzyme then catalyses the conversion of arachidonic acid to 5-
hydroperoxyeicosatetraenoic acid (5-HPETE); the unstable intermediate
leukotriene A 4 (LTA 4) is subsequently produced with the aid of the accessory 5-LO
activating protein (FLAP). The LTA 4 hydrolase can metabolize LTA 4 to leukotriene
B 4 (LTB 4), which acts as a very potent neutrophil chemotaxin. LTB 4 also promotes
the adhesion of leukocytes to vascular endothelium (Funk, Colin D., 2005). For
example, higher levels of leukotriene LTB 4 were observed in symptomatic human
carotid artery lesions compared to asymptomatic lesions, with categorization based
on clinical evidence of plaque instability (Cipollone et al., 2005). The pathway in
which the 5-LO enzyme interacts has been proposed as a novel mechanism for
potential intervention in CVD; such interventions may be particularly relevant in

3
asthmatics as drugs that target the pathway are available (Allayee et al., 2007). LT
modifiers approved for the treatment of asthma include the 5-LO inhibitor, zileuton,
and the LT receptor antagonists montelukast and zafirlukast (Allayee et al., 2007).
Theophylline is also used to treat asthma and is a type of bronchodilator. Moderate
to severe asthmatics (at values of predicted FEV 1 <80%) receiving montelukast had
significantly lower levels of the CVD-associated inflammatory biomarker C-reactive
protein (CRP) in comparison to moderate to severe asthmatics (at values of
predicted FEV 1 <80%) receiving placebo at 1-month and 6-months; moderate to
severe asthmatics (at values of predicted FEV 1 <80%) receiving low-dose
theophylline had lower levels of CRP at 6-months (Allayee et al., 2007) in
comparison to moderate to severe asthmatics (at values of predicted FEV 1 <80%)
receiving placebo. In contrast, in asthmatics with predicted FEV 1 values ≥80%, no
significant effects of montelukast on the CVD-associated inflammatory biomarker
interleukin (IL)-6 were observed in comparison to placebo. By comparison, IL-6
levels were significantly lower in the theophylline group at 6-months for asthmatics
with predicted FEV 1 values <80% in comparison to asthmatics with predicted FEV 1
values <80% receiving placebo (Allayee et al., 2007).
Given the recent evidence suggesting that the 5-LO/LT pathway, known to be
related to asthma, also plays a role in CVD-related inflammatory properties, we
aimed to: (1) evaluate the association between asthma and subclinical
atherosclerosis, hypothesizing that asthmatics will have elevated levels of
atherosclerosis in comparison to non-asthmatics; (2) compare the level of

4
subclinical atherosclerosis in users of asthma-related respiratory medications to
non-users of such respiratory medication.
The analysis will utilize data from four randomized controlled trials
completed at the University of Southern California Atherosclerosis Research Unit
(ARU) that tested atherosclerosis interventions using longitudinally-measured
carotid artery intima-media thickness (CIMT) as a trial outcome. We use these CIMT
measurements in a cross-sectional analysis of baseline trial data.

5
CHAPTER 2: MATERIALS AND METHODS
Section 2.1. Study Design

This cross-sectional analysis incorporated existing data from four
randomized controlled trials (B-Vitamin Atherosclerosis Intervention Trial (BVAIT),
Estrogen in the Prevention of Atherosclerosis Trial (EPAT), Vitamin E
Atherosclerosis Prevention Study (VEAPS), and the Women’s Estrogen-Progestin
Lipid-Lowering Hormone Atherosclerosis Regression Trial (WELLHART))
conducted at the University of Southern California Atherosclerosis Research Unit
(ARU). These trials tested the anti-atherogenic efficacy of menopausal hormone
replacement therapy (EPAT and WELLHART) and vitamin supplementation (BVAIT
and VEAPS). The trial inclusion and exclusion criteria differed according to the
hypothesis that was being tested as well as the intervention of interest.
The BVAIT (n=506 randomized) subjects were men and postmenopausal
women ≥40 years old with fasting total homocysteine (tHcy) ≥8.5 µmol/L and no
clinical signs or symptoms of CVD. Exclusion criteria were fasting triglycerides >500
mg/dL, diabetes mellitus or fasting serum glucose >126 mg/dL, systolic blood
pressure ≥160 mm Hg and/or diastolic blood pressure ≥100 mm Hg (Hodis et al.,
2009).
EPAT (n=222 randomized) participants were postmenopausal women 45
years of age or older, without preexisting cardiovascular disease and with a low-
density lipoprotein (LDL) cholesterol level of 130 mg/dL or greater. Women with
diabetes mellitus were eligible provided that their fasting blood glucose level was

6
<200 mg/dL. Women with triglyceride levels greater than 400 mg/dL, high density
lipoprotein (HDL) cholesterol levels less than 30 mg/dL, and current smokers were
excluded (Hodis et al., 2001). Of the n=222 that were randomized, n=199 completed
the trial with at least one follow-up measurement of CIMT; in EPAT, CIMT
measurements were only obtained on the 199 subjects who had at least one post-
randomization ultrasound for measurement of CIMT.
The VEAPS (n=353 randomized) trial was comprised of men and women who
were ≥40 years old, with LDL cholesterol >130 mg/dL, and who had no clinical signs
or symptoms of CVD. Exclusion criteria included fasting triglycerides >500 mg/dL,
diabetes mellitus or fasting serum glucose ≥140 mg/dL, regular vitamin E
supplement intake > 1 year, and diastolic blood pressure (DBP) ≥100 mm Hg (Hodis
et al., 2002).
The WELLHART (n=226 randomized) cohort consisted of postmenopausal
women with pre-existing heart disease who had at least one coronary artery lesion
that occluded at least 30% of the lumen diameter. Other eligibility criteria included
age ≤75 years, low-density lipoprotein (LDL) cholesterol level of 100-250 mg/dL,
and a total triglyceride level <400 mg/dL. Women were excluded if they had a
diastolic blood pressure >110 mm Hg or a fasting serum glucose >200 mg/dL (Hodis
et al., 2003).
In this cross-sectional analysis, pooled data from these four trials contributed
a total of n=1307 subjects with n=1284 used for cross-sectional analysis (due to the
n=23 randomized EPAT subjects for whom CIMT was not measured in follow-up

7
visits. These subjects were excluded and cross-sectional data was not made
available. While eligibility requirements differed across trials, there were many
common eligibility criteria and data collection was standardized across all trials.
Most notably, the primary outcome for all trials was the rate of change in CIMT,
measured by ultrasound examinations conducted prior to randomization and at 6-
month intervals throughout trial follow-up.

8
Section 2.2. Measurement of Main Outcome – Carotid IMT
(CIMT)

The main outcome of interest in our analysis is the baseline measurement of
CIMT. CIMT measures the thickness of two layers (the intima and the media) of the
wall of the carotid artery, the largest conduit of blood supply to the brain.
Measurements of CIMT have been used in large observational studies as well as
randomized controlled trials as a measure of sub-clinical atherosclerosis (Kunzli et
al., 2010). CIMT is a well-established quantitative measure of generalized sub-
clinical atherosclerosis that correlates well with major cardiovascular risk factors,
with coronary artery atherosclerosis, and with clinical cardiovascular events (Künzli
et al., 2010). As a measure of subclinical atherosclerosis, CIMT may help define the
burden of atherosclerosis in individuals before they develop clinical events such as
heart attack or stroke (Künzli et al., 2010).
In all trials, high-resolution far-wall B-mode ultrasound images of the right
common carotid artery were obtained with a 7.5-MHz linear array transducer
attached to an ATL Ultramark-4 Plus Ultrasound System (Advanced Technology
Laboratory). Subjects were placed in a supine position with the head rotated to the
left using a 45°-degree head block. The jugular vein and carotid artery were located
in the transverse view, with the jugular vein stacked above the carotid artery. The
transducer was then rotated 90 degrees around the central line of the transverse
image of the stacked jugular vein and carotid artery to obtain a longitudinal image
while the stacked position of the vessels was maintained. All images contained

9
anatomic landmarks for reproducing probe angulation, and a hard copy of each
participant’s baseline image was used as a guide for follow-up examinations. Carotid
intima-media thickness was the average of approximately 70 to 100 individual
measurements between the intima-lumen and media-adventitia interfaces along a
1-cm length just distal to the carotid artery bulb (Hodis et al., 2001). Further details
on the ultrasound procedure and measurement of CIMT are published (Hodis et al.,
2001).

10
Section 2.3. Asthma Status & Respiratory Medication Use
On a baseline medical history questionnaire, participants indicated if they
had physician-diagnosed asthma by answering “yes” or “no”. These self-reports of
physician-diagnosed asthma were not verified with medical records.
Data on the current use of prescription and non-prescription medications
were collected at baseline. Data on all medications were recorded, including name,
start date, end date (if applicable), frequency and dosage. A standard medication-
coding scheme was applied to all trials; each medication was coded using a 2-digit
primary class category (e.g., respiratory drugs), a 2-digit secondary category (e.g.,
anti-inflammatory agents (non-steroidal), anti-inflammatory agents (steroidal),
bronchial dilators, or leukotriene receptor antagonists) and a final 3-digit drug code
specific to a drug within each class. Each medication was therefore uniquely
indentified by a 7-digit code. For the purpose of our analysis, the use of respiratory
medication included single use or combination use of respiratory medications at the
trial baseline. The main interest with regards to the use of respiratory medication
will focus on the use of anti-inflammatory non-steroidal medication (AINS), anti-
inflammatory steroidal (AIS) medication, the use of bronchodilators, and the use of
leukotriene anti-receptor antagonists.
For the 2-digit secondary code, non-steroidal anti-inflammatory medications
included cromolyn sodium, nedocromil sodium; steroidal anti-inflammatory
medications included beclomethasone dipropionate, triamcinolone acetonide,

11
flunisolide, budenoside, fluticasone propionate; bronchial dilators included albuterol,
salmeterol xinafoate, ipratropium bromide, theophylline, metaproterenol sulfate,
pirbuterol acetate, tarbutaline sulfate, ephedrine sulfate, levalbuterol; and leukotriene
receptor antagonists included montelukast and zafirlukast.

12
Section 2.4. Measurement of Lipids
Participants fasted for 8 hours before blood sample collection. Total plasma
cholesterol and triglycerides levels were measured using an enzymatic method of
the Standardization Program of the National Centers for Disease Control and
Prevention. High-density lipoprotein (HDL) cholesterol levels were measured after
lipoproteins containing apolipoprotein B were precipitated in whole plasma by
using heparin manganese chloride. Low-density lipoprotein (LDL) cholesterol levels
were estimated using the Friedewald equation (Hodis et al., 2001).

13
Section 2.5. Measurement of Vital Signs and Anthropometry
Vital signs (blood pressure, pulse rate), height and weight were ascertained
using standard procedures and were collected uniformly with the same instruments
in each trial and at each follow-up. Body mass index (BMI) was calculated as the
ratio of the value of weight (lbs.) multiplied by 703 and the squared value of height
(in.). The equation is as follows: ((weight (lbs.)) * 703) / ((height (in.))
2
).

14
Section 2.6. Assessment of Demographic Information and the Use
of Cardiovascular Disease-Related Risk Factors

Gender, age, race (non-Hispanic White, non-Hispanic Black, Hispanic, Asian,
and Native-American), education, marital status and income were obtained with a
structured questionnaire for all four trials. Other structured questionnaires included
an assessment of self-reported medical conditions (including diabetes), smoking
status (never smoker, previous smoker, current smoker), and the use of CVD-related
medications (lipid lowering medication, blood pressure medication, anticoagulant
medication, anticholinergic medication).

15
Section 2.7. Statistical Analysis
We tested the univariate and multivariate associations between CIMT and
self-reported asthma (asthmatics vs. non-asthmatics) using linear regression.
Subjects who did not know whether or not they were asthmatics (n=2, 1 subject in
BVAIT and 1 subject in EPAT) were placed in the non-asthmatic category, with the
rationale that a person would recall a physician diagnosis of asthma.
Baseline data for this analysis were obtained from screening visits and the
randomization visit. The distribution of CIMT was evaluated for normality with
plots. Baseline CIMT was first compared between asthmatics and non-asthmatics
and between users and non-users of respiratory medications using the independent
samples t-test in an unadjusted analysis.
Demographic characteristics including age, gender, marital status, and
income; baseline clinical variables including weight, height, BMI, systolic blood
pressure (SBP), diastolic blood pressure (DBP), pulse rate, and diabetes; and
baseline laboratory variables including total cholesterol, HDL cholesterol, LDL
cholesterol, and triglycerides were compared between asthmatics and non-
asthmatics to determine if any of these variables should be used in the multivariate
linear regression model-building process as possible correlates of CIMT and hence
as possible confounders of the association of asthma with CIMT. A variable was
found to be a confounder if the addition of the variable caused a 20% (or greater)

16
change in the asthma variable parameter estimate on the univariate association
with CIMT.
Then, all variables were added in a step-by-step process and kept if they
proved to be significant (p<0.05) after their addition to the previously constructed
model. Marginally significant (p<0.10) added variables were kept only if they were
known to be significantly (p<0.05) associated with CIMT and/or asthma from
previous univariate tests or from the literature related to atherosclerosis.
Interaction variables were also univariately evaluated to test if there were
significant differences between various groups (or categories) in the association of
asthma with CIMT. Interactions tested with the asthma variable included the
following indicator variables: gender, race, smoking status, diabetic status, marital
status, income group, the use of blood pressure medication, the use of hypolipidemic
medication, and the use of anticoagulation medication. After the formation of the
multivariate linear regression model, evaluation of the linear regression model
assumptions included checks for normality and homoscedasticity of model
residuals.
Since the combined dataset included the WELLHART (n=226) study, where
all participants had coronary artery disease and a high number had diabetes (Table
1), sensitivity analyses excluding WELLHART participants were performed (n=1058,
in this sensitivity analysis) to determine the influence of the WELLHART data on our
overall findings and interpretations.

17
The self-reported use, at baseline, of respiratory medication was categorized
into “yes” or “no” groups. These two groups were broken down into whether a
subject used (individually or in combination) any one of the four different types of
respiratory medications including anti-inflammatory non-steroidal (AINS), anti-
inflammatory steroidal (AIS), bronchial dilator, and leukotriene receptor antagonists.
Univariate and multivariate associations between CIMT and the self-reported
use of respiratory medication were tested using the same analytic approach as
detailed above for asthma-CIMT associations. In addition, analyses were completed
with the exclusion of the WELLHART trial subjects for the use of respiratory
medication.

18
CHAPTER 3: RESULTS
Section 3.1. Participant Characteristics
Table 1 summarizes the main characteristics of the four trials individually as
well as combined. Relatively few subjects self-reported having physician-diagnosed
asthma (n=108, 8.4%). The BVAIT trial reported the highest number of asthmatics
(8.9%), while the WELLHART trial reported the least (7.1%). The remaining
subjects were classified as non-asthmatics (n=1176, 91.6%). The pooled sample
showed a near two-to-one, female-to-male ratio (n=807, 62.9%, vs. n=477, 37.1%),
reflecting the trial designs.
Table 1 - Demographic and Clinical Characteristics by Trial
Characteristic BVAIT EPAT VEAPS WELLHART
Total
(Combined
Trials)
N 506 199 353 226 1284
CIMT (mm)
0.7535 ±
0.1492
0.7640 ±
0.1317
0.7551 ±
0.1329
0.8452 ±
0.2192
0.7717 ±
0.1609
Asthmatics at
Baseline
†

(n (%))
45 (8.9%) 20 (10.1%) 27 (7.6%) 16 (7.1%) 108 (8.4%)
Respiratory
Medication Use
at Baseline
(n (%))
‡

29
(5.7%)
4
(2.0%)
7
(2.0%)
14
(6.2%)
54
(4.2%)
On-Trial
Intervention
B vitamins
(folic acid,
vitamin B 6,
vitamin
B 12)
Unopposed
Estradiol
Vitamin E
(α-
Tocopherol)
Estradiol/Estradiol-
Progestin
N/A
Data are mean ± SD except where indicated
† = Self-reported
‡ = Includes anti-inflammatory non-steroidal; anti-inflammatory steroidal; bronchial dilator;
leukotriene antagonist medication (individually or combined)
* = Diabetes defined by fasting glucose (BVAIT > 140 mg/dL; EPAT > 200 mg/dL ; VEAPS >140
mg/dL; WELLHART > 200 mg/dL)
§ = White (Non-Hispanic); Black (Non-Hispanic)

19
Table 1 (Continued) - Demographic and Clinical Characteristics by Trial
Characteristic BVAIT EPAT VEAPS WELLHART
Total
(Combined
Trials)
Treatment
(n (%))
Vitamin B –
254(50.2%)
Placebo –
252(49.8%)
Estrogen –
97(48.7%)
Placebo –
102(51.3)
Vitamin E –
177(50.1%)
Placebo –
176(49.9%)

Estrogen –
76(33.6%)
Estrogen-Progestin
– 74(32.7%)
Placebo –
76(33.6%)
N/A
Age (yrs.)

61.3 ± 9.8

61.3 ± 6.9 56.1 ± 9.0 63.4 ± 6.5 60.2 ± 9.1
Total
Cholesterol
(mg/dL)

221.1 ±
39.4

250.5 ± 36.8

237.9± 26.4

234.0 ± 48.8

232.6 ± 39.2
HDL
Cholesterol
(mg/dL)

56.8 ± 15.2

53.9 ± 12.2

56.0 ± 13.0

49.2 ± 10.6

54.8 ± 13.7
LDL Cholesterol
(mg/dL)

138.3 ±
36.5

166.1 ± 32.2

155.2 ± 22.0

143.8 ± 42.4

148.2 ± 35.2
Triglyceride
Cholesterol
(mg/dL)

134.4 ±
131.6

155.3 ± 90.4

133.6 ± 55.4

207.9 ± 140.8

150.4 ±
114.7
Gender (n (%))
Males
Females

309(61.1%)
197(38.9%)

N/A
199(100.0%)

168 (47.6%)
185 (52.4%)

N/A
226 (100.0%)

477 (37.1%)
807 (62.9%)
Race (n (%))
White
§
Black
§
Hispanic
Asian
Native-American

328(64.8%)
75(14.8%)
55(10.9%)
45(8.9%)
3(0.6%)

118(59.3%)
22(11.1%)
40(20.1%)
18(9.0%)
1(0.5%)

260(73.7%)
35(9.9%)
38(10.8%)
19(5.4%)
1(0.3%)

69(30.5%)
38(16.8%)
100(44.3%)
18(8.0%)
1(0.4%)

775(60.4%)
170(13.2%)
233(18.1%)
100(7.8%)
6(0.5%)

Smoking Status
(n (%))
Current Smoker
Previous Smoker
Never Smoker

18 (3.6%)
178 (35.2%)
310 (61.3%)

0 (0.0%)
98 (49.2%)
101 (50.8%)

14 (4.0%)
113 (32.0%)
226 (64.0%)

26 (11.5%)
89 (39.3%)
111 (49.1%)

58 (4.5%)
478 (37.2%)
748 (58.3%)
Data are mean ± SD except where indicated
† = Self-reported
‡ = Includes anti-inflammatory non-steroidal; anti-inflammatory steroidal; bronchial dilator;
leukotriene antagonist medication (individually or combined)
* = Diabetes defined by fasting glucose (BVAIT > 140 mg/dL; dEPAT > 200 mg/dL ; VEAPS
>140 mg/dL; WELLHART > 200 mg/dL)
§ = White (Non-Hispanic); Black (Non-Hispanic)

20

Table 1 (Continued) - Demographic and Clinical Characteristics by Trial
Characteristic BVAIT EPAT VEAPS WELLHART
Total
(Combined
Trials)
Education (yrs.)

15.6 ± 2.1

14.7 ± 2.1

15.5 ± 2.2

11.7 ± 2.9

14.8 ± 2.7
DBP (mm Hg)

80.8 ± 10.4

77.2 ± 8.3

77.0 ± 10.3

76.5 ± 11.3

78.4 ± 10.4
SBP (mm Hg)

129.6 ±
16.8

126.1 ± 14.8

128.8 ± 18.2

142.1 ± 23.6

131.1 ± 19.0
Pulse Rate
(bpm)

64.5 ± 8.2

64.9 ± 6.4

64.6 ± 7.6

62.8 ± 8.1

64.3 ± 7.8
Weight (lbs.)

179.2 ±
36.3

164.0 ± 34.5

174.8 ± 36.9

162.1 ± 33.3

172.6 ± 36.3
BMI

28.1 ± 5.0

28.9 ± 6.7

27.7 ±

30.3 ± 5.5

28.5 ± 5.4
Diabetic
*

(n (%))
1 (0.2%) 6 (3.0%) 2 (0.6%) 103 (45.6%) 112 (8.7%)
Marital Status
(n (%))
Single
Married
Separated
Divorced
Widowed
Refused to Answer

45 (8.9%)
324 (64.0%)
5 (1.0%)
90 (17.8%)
41 (8.1%)
1 (0.2%)

15 (7.5%)
91 (45.7%)
5 (2.5%)
51 (25.6%)
37 (18.6%)
0 (0.0%)

50 (14.2%)
228 (64.6%)
6 (1.7%)
54 (15.3%)
15 (4.2%)
0 (0.0%)

15 (6.6%)
103 (45.6%)
21 (9.3%)
36 (15.9%)
51 (22.6%)
0 (0.0%)

125 (9.7%)
746 (58.1%)
37 (2.9%)
231 (18.0%)
144 (11.2%)
1 (0.1%)
Income (n (%))
<$10k
$10k - $19999
$20k - $29999
$30k - $39999
$40k - $49999
$50k - $59999
$60k+
Refused to Answer

9 (1.8%)
31 (6.1%)
41 (8.1%)
33 (6.5%)
48 (9.5%)
47 (9.3%)
265 (52.4%)
32 (6.3%)

17 (8.5%)
35 (17.6%)
34 (17.1%)
25 (12.6%)
16 (8.0%)
18 (9.0%)
35 (17.6%)
19 (9.5%)

6 (1.7%)
26 (7.4%)
37 (10.5%)
49 (13.9%)
35 (9.9%)
43 (12.2%)
150 (42.5%)
7 (2.0%)

68 (30.1%)
56 (24.8%)
25 (11.1%)
19 (8.4%)
14 (6.2%)
3 (1.3%)
12 (5.3%)
29 (12.8%)

100 (7.8%)
148 (11.5%)
137 (10.7%)
126 (9.8%)
113 (8.8%)
111 (8.6%)
462 (36.0%)
87 (6.8%)
Data are mean ± SD except where indicated
† = Self-reported
‡ = Includes anti-inflammatory non-steroidal; anti-inflammatory steroidal; bronchial dilator;
leukotriene antagonist medication (individually or combined)
* = Diabetes defined by fasting glucose (BVAIT > 140 mg/dL; EPAT > 200 mg/dL ; VEAPS >140
mg/dL; WELLHART > 200 mg/dL)
§ = White (Non-Hispanic); Black (Non-Hispanic)

21
Table 1 (Continued) - Demographic and Clinical Characteristics by Trial
Characteristic BVAIT EPAT VEAPS WELLHART
Total
(Combined
Trials)
Blood Pressure
Medication
(n (%))
Yes

172
(34.0%)

46 (23.1%)

86 (24.4%)

214 (94.7%)

518 (40.3%)
Hypolipidemics
Medication
(n (%))
Yes

178
(35.2%)

10 (5.0%)

70 (19.8%)

96 (42.5%)

354 (27.6%)
Anti-
Inflammatory
Non-Steroidal
(AINS) Use
(n (%))
Yes

1 (0.2%)

2 (1.0%)

1 (0.3%)

2 (0.9%)

6 (0.5%)
Anti-
Inflammatory
Steroidal (AIS)
Use (n (%))
Yes

23 (4.5%)

1 (0.5%)

5 (1.4%)

10 (4.4%)

39 (1.9%)
Bronchial
Dilator Use
(n (%))
Yes

30 (5.9%)

7 (3.5%)

9 (2.5%)

22 (9.7%)

68 (5.3%)
Leukotriene
Receptor
Antagonist
(LTRA) Use
(n (%))
Yes

3 (0.6%)

0 (0.0%)

2 (0.6%)

2 (0.9%)

7 (0.5%)
Data are mean ± SD except where indicated
† = Self-reported
‡ = Includes anti-inflammatory non-steroidal; anti-inflammatory steroidal; bronchial dilator;
leukotriene antagonist medication (individually or combined)
* = Diabetes defined by fasting glucose (BVAIT > 140 mg/dL; dEPAT > 200 mg/dL ; VEAPS
>140 mg/dL; WELLHART > 200 mg/dL)
§ = White (Non-Hispanic); Black (Non-Hispanic)

Table 2 compares the main characteristics of asthmatics versus non-
asthmatics. Asthmatics reported more use of bronchial dilators (32.4%) than non-
asthmatics (2.8%) (p<0.0001). The two groups did not differ on major variables

22
related to CIMT, including age (p=0.64), gender (p=0.86), race (p=0.97), smoking
status (p=0.41) and education (p=0.90). There are few asthmatics (n=108) and many
of the subjects were non-smokers. Asthmatics and non-asthmatics did not differ on
demographic and clinical characteristics.
Table 2 - Main Characteristics of Asthmatics vs. Non-Asthmatics
Characteristic Asthmatics (n=108) Non-Asthmatics (n=1176) p-value
CIMT (mm) 0.7726 ± 0.14 0.7716 ± 0.16 0.95
Age (yrs.) 60.6 ± 9.2 60.2 ± 9.1 0.64
Total Cholesterol
(mg/dL)
232.2 ± 39.4 232.6 ± 39.2 0.91
HDL Cholesterol
(mg/dL)
54.9 ± 13.2 54.8 ± 13.8 0.92
LDL Cholesterol
(mg/dL)
148.0 ± 35.1 148.3 ± 35.2 0.95
Triglycerides
(mg/dL)
147.6 ± 72.9 148.3 ± 35.2 0.70
Gender (n (%))
Males
Females

41 (38.0%)
67 (62.0%))

436 (37.1%)
740 (62.9%)
0.86
Race (n (%))

White
Black
Hispanic
Asian
Native American

65 (60.2%)
14 (13.0%)
20 (18.5%)
8 (7.4%)
1 (0.9%)

710 (60.4%)
156 (13.3%)
213 (18.1%)
92 (7.8%)
5 (0.4%)

0.97

Smoking Status
(n (%))
Current Smoker
Previous Smoker
Never Smoker

6 (5.6%)
34 (31.5%)
68 (63.0%)

52 (4.4%)
444 (37.8%)
680 (57.8%)

0.41
Education (yrs.) 14.8 ± 2.82 14.8 ± 2.7 0.90
DBP (mm Hg) 78.9 ± 11.7 78.4 ± 10.3 0.67
SBP (mm Hg) 130.1 ± 19.8 131.2 ± 18.9 0.58
Pulse Rate (bpm) 64.5 ± 8.2 64.3 ± 7.8 0.79
Weight (lbs.) 175.1 ± 34.9 172.4 ± 36.4 0.46
BMI 28.8 ± 5.5 28.5 ± 5.4 0.50
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
asthmatics and non-asthmatics.

23
Table 2 (Continued) – Main Characteristics of Asthmatics vs. Non-Asthmatics
Characteristic Asthmatics (n=108) Non-Asthmatics (n=1176) p-value
Diabetes
(n (%))

10 (9.3%) 102 (8.7%) 0.86
Marital Status
(n (%))
Single
Married
Separated
Divorced
Widowed
Refused to Answer

13 (12.0%)
59 (54.6%)
7 (6.5%)
19 (17.6%)
10 (9.3%)
0 (0.0%)

112 (9.5%)
687 (58.4%)
30 (2.6%)
212 (18.0%)
134 (11.4%)
1 (0.1%)

0.24
Income (n (%))
<$10k
$10k - $19999
$20k - $29999
$30k - $39999
$40k - $49999
$50k - $59999
$60k+
Refused to Answer

11 (10.2%)
13 (12.0%)
13 (12.0%)
12 (11.1%)
9 (8.3%)
11 (10.2%)
35 (32.4%)
4 (3.7%)

89 (7.6%)
135 (11.5%)
124 (10.5%)
114 (9.7%)
104 (8.8%)
100 (8.5%)
427 (36.3%)
83 (7.1%)

0.81
Blood Pressure
Medication Use at
Baseline
(n (%))

Yes

45 (41.7%)

473 (40.2%)

0.77
Hypolipidemics
Medication Use at
Baseline
(n (%))

Yes

29 (26.9%)

325 (27.6%)

0.86
Anti-Inflammatory
Non-Steroidal (AINS)
Medication Use at
Baseline
(n (%))
Yes

2 (1.9%)

4 (0.3%)

0.08
Anti-Inflammatory
Steroidal (AIS)
Medication Use at
Baseline
(n (%))

Yes

27 (25.0%)

12 (1.0%)

<0.0001
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
asthmatics and non-asthmatics.

24
Table 2 (Continued) – Main Characteristics of Asthmatics vs. Non-Asthmatics
Characteristic Asthmatics (n=108) Non-Asthmatics (n=1176) p-value
Bronchial Dilator
Medication Use at
Baseline (n (%))

Yes

35 (32.4%)

33 (2.8%)

<0.0001
Leukotriene
Receptor Antagonist
(LTRA) Medication
Use at Baseline
(n (%))

Yes

4 (3.7%)

3 (0.3%)

<0.01
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
asthmatics and non-asthmatics.

Table 3 summarizes the main characteristics in the comparison of asthmatics
versus non-asthmatics when the WELLHART trial participants are excluded. Except
for asthmatics taking more respiratory medications, the two groups do not differ on
any other characteristics.
We graphically evaluated the distribution of CIMT for normality. The
distribution displayed right-hand skewness (skewness = 2.53) with thick tails
(kurtosis = 14.02) when evaluated over the entire sample. The distribution of CIMT
improved when the WELLHART trial subjects (n=226) were excluded. A decrease in
right-hand skewness was observed (1.96 vs. 2.53) as well as a decrease in the
kurtosis (10.22 vs. 14.02).

25
Table 3 - Main Characteristics of Asthmatics vs. Non-Asthmatics (Excluding WELLHART Trial)
Characteristic Asthmatics (n=92)
Non-Asthmatics
(n=966)
p-value
CIMT (mm) 0.7594 ± 0.1372 0.7557 ± 0.1411 0.81
Age (yrs.) 60.4 ± 9.7 59.5 ± 9.4 0.40
Total Cholesterol (mg/dL) 233.9 ± 37.1 232.1 ± 36.9 0.66
HDL Cholesterol (mg/dL) 56.1 ± 13.4 56.0 ± 14.0 0.95
LDL Cholesterol (mg/dL) 150.1 ± 32.7 149.1 ± 33.5 0.78
Triglycerides (mg/dL) 138.5 ± 60.9 138.0 ± 107.6 0.95
Gender (n (%))
Males
Females

41 (44.6%)
51 (55.4%)
436 (45.1%)
530 (54.9%)
0.92
Race (n (%))
White (Non-Hispanic)
Black (Non-Hispanic)
Hispanic
Asian
Native American

61 (66.3%)
10 (10.9%)
12 (13.0%)
8 (8.7%)
1 (1.1%)

645 (66.8%)
122 (12.6%)
121 (12.5%)
74 (7.7%)
4 (0.4%)

0.89
Smoking Status
(n (%))
Current Smoker
Previous Smoker
Never Smoker

3 (3.3%)
27 (29.3%)
62 (67.4%)

29 (3.0%)
362 (37.5%)
575 (59.5%)

0.30
Education (yrs.) 15.4 ± 2.3 15.4 ± 2.1 0.88
DBP (mm Hg) 79.5 ± 12.0 78.8 ± 10.0 0.59
SBP (mm Hg) 129.5 ± 19.3 128.6 ± 16.7 0.64
Heart Pulse Rate (bpm) 64.2 ± 8.4 64.6 ± 7.6 0.63
Weight (lbs.) 174.6 ± 34.2 174.9 ± 36.8 0.94
BMI 28.3 ± 5.2 28.1 ± 5.3 0.79
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
asthmatics and non-asthmatics.

26
Table 3 (Continued) - Main Characteristics of Asthmatics vs. Non-Asthmatics (Excluding
WELLHART Trial)
Characteristic Asthmatics (n=92)
Non-Asthmatics
(n=966)
p-value
Diabetes (n (%))
Yes

0 (0.0%)

9 (0.9%)

1.00
Marital Status
(n (%))

Single
Married
Separated
Divorced
Widowed
Refused to Answer

12 (13.0%)
54 (58.7%)
4 (4.3%)
14 (15.2%)
8 (8.7%)
0 (0.0%)

98 (10.1%)
589 (61.0%)
12 (1.2%)
181 (18.7%)
85 (8.8%)
1 (0.1%)

0.24
Income (n (%))
<$10k
$10k - $19999
$20k - $29999
$30k - $39999
$40k - $49999
$50k - $59999
$60k+
Refused to Answer

4 (4.3%)
9 (9.8%)
11 (12.0%)
11 (12.0%)
8 (8.7%)
11 (12.0%)
35 (38.0%)
3 (3.3%)

28 (2.9%)
83 (8.6%)
101 (10.5%)
96 (9.9%)
91 (9.4%)
97 (10.0%)
415 (43.0%)
55 (5.7%)

0.89
Blood Pressure
Medication Use at
Baseline (n (%))

Yes

30 (32.6%)

274 (28.4%)

0.39
Hypolipidemics
Medication Use at
Baseline (n (%))

Yes

23 (25.0%)

235 (24.3%)

0.89
Anti-Inflammatory Non-
Steroidal (AINS)
Medication Use at
Baseline
(n (%))
Yes

1 (1.1%)

3 (0.3%)

0.31
Anti-Inflammatory
Steroidal (AIS) Medication
Use at Baseline (n (%))

Yes

22 (23.9%)

7 (0.7%)

<0.0001
Bronchial Dilator
Medication Use at
Baseline (n (%))
Yes

29 (31.5%)

17 (1.8%)

<0.0001
Leukotriene Receptor
Antagonist (LTRA)
Medication Use at
Baseline (n (%))
Yes

3 (3.3%)

2 (0.2%)

<0.01
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
asthmatics and non-asthmatics.

27
Section 3.2. CIMT Comparisons by Asthma Status
An independent samples t-test comparing the mean CIMT values between
asthmatics and non-asthmatics was performed for the entire sample (n=1284). The
findings were not statistically significant (p=0.95), with asthmatics (n=108) having a
mean CIMT (SD) of 0.7726 mm (0.14) compared to non-asthmatics (n=1176) having
a mean of 0.7716 mm (0.16). Similar results were obtained with the exclusion of the
WELLHART trial subjects (p=0.81 between groups). The mean CIMT (SD) value for
asthmatics (n=92) was 0.7594 mm (0.14) compared to non-asthmatics (n=966;
0.7557 mm (0.14)).
Independent samples t-tests comparing asthmatics and non-asthmatics were
performed for all demographic and clinical variables (Tables 2 and 3) to determine
if any differences might highlight possible confounders to be modeled in subsequent
multivariate linear regression models. No significant differences (p <0.05) (other
than respiratory medications) were found for any variable, either with all four trials
combined (Table 2) or with the exclusion of the WELLHART trial data (Table 3).
In other preliminary analyses, mean CIMT values were compared between
asthmatics and non-asthmatics with stratification on high and low risk categories for
lipid variables (e.g. total cholesterol, high-density lipoprotein [HDL] cholesterol,
low-density lipoprotein [LDL] cholesterol, and total triglycerides), using the
American Heart Association guidelines (American Heart Association, www.aha.org)

28
for optimal cholesterol levels. These analyses were intended to explore possible
modification of the CIMT-asthma association by lipid values.
Tables 4 and 5 summarize the findings for the various cholesterol risk groups
for the entire baseline sample (Table 4) and for the baseline sample excluding the
WELLHART trial data (Table 5).
Table 4 – CIMT Comparisons by Asthma Status by Cholesterol Risk Groups Among Total
Sample (n=1284)

TOTAL
CHOLESTEROL
HDL
CHOLESTEROL
LDL
CHOLESTEROL
TOTAL
TRIGLYCERIDES
CHARACTERISTIC Low High Low High Low High Low High
Non-Asthmatics
(n)
219 957 914 262 93 1083 730 446
Asthmatics (n) 23 85 82 26 9 99 71 37
Non-Asthmatics :
CIMT (mm) ± SD
0.7833
± 0.19
0.7689
± 0.15
0.7620
± 0.15
0.8052
± 0.21
0.7659
± 0.14
0.7721
± 0.16
0.7655
± 0.16
0.7816
± 0.16
Asthmatics : CIMT
(mm) ± SD
0.7506
± 0.17
0.7786
± 0.14
0.7705
± 0.15
0.7793
± 0.13
0.7956
± 0.21
0.7705
± 0.14
0.7741
± 0.14
0.7698
± 0.15
p-value 0.44 0.58 0.61 0.38 0.57 0.93 0.67 0.67

Table 5 – CIMT Comparisons by Asthma Status by Cholesterol Risk Groups Excluding
WELLHART Trial (n=1058)

TOTAL
CHOLESTEROL
HDL
CHOLESTEROL
LDL
CHOLESTEROL
TOTAL
TRIGLYCERIDES
CHARACTERISTIC Low High Low High Low High Low High
Non-Asthmatics
(n)
171 795 773 193 64 902 648 318
Asthmatics (n) 16 76 74 18 5 87 64 28
Non-Asthmatics:
CIMT (mm) ± SD
0.7690
± 0.18
0.7528
± 0.13
0.7506
± 0.13
0.7758
± 0.17
0.7486
± 0.13
0.7562
± 0.14
0.7535
± 0.14
0.7600
± 0.14
Asthmatics: CIMT
(mm) ± SD
0.7149
± 0.13
0.7687
± 0.14
0.7530
± 0.13
0.7855
± 0.15
0.6877
± 0.08
0.7635
± 0.14
0.7690
± 0.15
0.7374
± 0.11
p-value 0.24 0.32 0.88 0.82 0.30 0.65 0.41 0.40

Asthmatics and non-asthmatics did not differ on mean CIMT values when
stratified by lipid-stratified risk groups, with and without exclusion of WELLHART
subjects (Tables 4 & 5).

29
Table 6 – Association of Asthma Status on CIMT by Trial (n=1284)

BVAIT EPAT VEAPS WELLHART
Non-Asthmatics
(n)
461 179 326 210
Asthmatics (n) 45 20 27 16
Non-Asthmatics :
CIMT (mm) ± SD
0.7536 ± 0.15 0.7663 ± 0.13 0.7528 ± 0.13 0.8449 ± 0.22
Asthmatics :
CIMT (mm) ± SD
0.7524 ± 0.13 0.7435 ± 0.13 0.7826 ± 0.16 0.8489 ± 0.16
p-value 0.96 0.46 0.26 0.94
NOTE: Independent sample t-tests (p-value) were used to test difference between asthma groups on
CIMT.

Independent samples t-test comparisons on all the four individual trials
comparing the mean CIMT values between asthmatics and non-asthmatics showed
no statistically significant findings within trial (all p>0.05) (Table 6).
In linear regression, we first analyzed the unadjusted associations of asthma
with CIMT (Table 7). No statistically significant associations of mean CIMT were
observed between self-reported asthmatics in comparison to non-asthmatics
(p=0.95) for the entire sample. In an analysis among female asthmatics, there also
were no significant differences in mean CIMT in comparison to female non-
asthmatics (p=0.54). Asthmatic males also showed no significant differences in mean
CIMT in comparison to non-asthmatic males (p=0.37).
Table 7 Linear Regression Relating Asthma Status to CIMT
Model
Total Sample
(n=1284)
Excluding
WELLHART Trial
(n=1058)
Females
(n=807)
Males (n=477)
Unadjusted
Estimate (SE)
0.00102 (0.016) 0.00369 (0.015) -0.0127 (0.021) 0.0238 (0.026)
p-value 0.95 0.81 0.54 0.37
Adjusted
Estimate (SE)
-0.00140 (0.014) -0.000750 (0.014) -0.00465 (0.018) 0.00403 (0.024)
p-value 0.92 0.96 0.80 0.87
NOTE: Adjusted for age, gender, trial, diabetes, race, weight, systolic blood pressure, HDL
cholesterol, smoking status, education, and income.

30
As these unadjusted associations may be confounded, a multivariate linear
regression model was constructed to adjust for other variables that may affect the
association between asthma and CIMT. Independent variables that were initially
considered confounders included age, gender, and race. In addition, we included the
following independent variables: indicator variables for trial, self-reported
diagnosis of diabetes, weight, systolic blood pressure, and high-density lipoprotein
(HDL) cholesterol. All of these variables were significantly associated with CIMT
(p<0.05), with the exception of weight (p=0.10), when they were added to the crude
model in a step-by-step fashion. In addition to the statistical significance on the
association with CIMT, we evaluated variables as confounders with regards to their
effect on the primary association of asthma with CIMT.
The “trial” variable is a marker for the different trial populations as well as
the different time periods over which the four trials occurred. To the extent that
these populations differ both on CIMT and asthma, confounding will occur.
Similarly, the self report of diabetes was used to adjust for some diabetics
participating in the WELLHART trial; weight, systolic blood pressure and HDL
cholesterol are known to be key correlates of atherosclerosis risk. Adjustment for
these variables further reduced the difference in mean CIMT between asthmatics
and non-asthmatics (Table 7). Although not significantly associated with CIMT, we
added other variables that might act as confounders. Independent variables for
smoking status, years of education, and income were incorporated into our final
model.

31
With the final multivariate model, asthmatics did not have a statistically
significant difference in mean CIMT in comparison to non-asthmatics (p=0.92). An
analysis on gender in the final model showed that female asthmatics did not have a
statistically significant difference in mean CIMT in comparison to female non-
asthmatics (p=0.80); male asthmatics, also did not have a statistically significant
difference in mean CIMT in comparison to male non-asthmatics (p=0.87).

32
Section 3.3. Association of Asthma with CIMT
Excluding WELLHART Trial Data

The same linear modeling approach was used excluding the WELLHART trial
data because this trial included post-menopausal women who had established
coronary artery disease. We performed this analysis to determine if the asthma
association with CIMT differed when limiting the analysis to healthy individuals.
The sample was reduced from n=1284 to n=1058, with fewer females since
the WELLHART trial consisted of only women (n=226); the number of asthmatics
was reduced from n=108 to n=92. While there remained no association of asthma
with CIMT, the difference in the mean CIMT estimate for the unadjusted models
comparing asthmatics and non-asthmatics was nearly three times larger excluding
WELLHART subjects (β No WELLHART = 0.00369) than that which was observed in the
total sample including WELLHART subjects (β with WELLHART = 0.00102). Female
asthmatics did not have a statistically significant different mean CIMT value in
comparison to female non-asthmatics (p=0.49) (Table 7); male asthmatics did not
have a statistically significant different mean CIMT than male non-asthmatics
(p=0.37).

33
The same variables were used to adjust for variables that might impact the
association between asthma and mean CIMT. Even with the exclusion of the
WELLHART trial data, all of the previously named variables (age, gender, race, trial,
diabetes, weight, systolic blood pressure, and HDL cholesterol) were significantly
associated with CIMT when added to the crude model in a stepwise approach. When
the remaining variables (smoking status and education) were added to the model,
the asthma regression parameter estimate was β = -0.000750 (p=0.96).

34
Section 3.4. CIMT Comparisons by Use of
Respiratory Medication Status

We first compared users of respiratory medication and non-users of
respiratory medication on several factors that may confound the association of
respiratory medication use with CIMT.
Table 8 - Main Characteristics of the Users of Respiratory Medications vs. the Non-Users of
Respiratory Medications
Characteristic
Users of Respiratory
Medication (n=54)
Non-Users of
Respiratory
Medication (n=1230)
p-value
CIMT (mm) 0.7745 ± 0.15 0.7716 ± 0.16 0.90
Age (yrs.) 61.9 ± 8.7 60.2 ± 9.1 0.17
Total Cholesterol
(mg/dL)
230.6 ± 41.6 232.7 ± 39.2 0.70
HDL Cholesterol
(mg/dL)
57.5 ± 14.9 54.7 ± 13.6 0.14
LDL Cholesterol
(mg/dL)
143.9 ± 33.8 148.4 ± 35.2 0.36
Triglycerides (mg/dL) 148.3 ± 81.4 150.4 ± 116.0 0.85
Gender (n (%))
Males
Females

17 (31.5%)
37 (68.5%)

460 (37.4%)
770 (62.6%)

0.38
Race (n (%))

White (Non-Hispanic)
Black (Non-Hispanic)
Hispanic
Asian
Native American

30 (55.6%)
9 (16.7%)
10 (18.5%)
5 (9.3%)
0 (0.0%)

745 (60.6%)
161 (13.1%)
223 (18.1%)
95 (7.7%)
6 (0.5%)

0.89
Smoking Status
(n (%))
Current Smoker
Previous Smoker
Never Smoker

4 (7.4%)
21 (38.9%)
29 (53.7%)

54 (4.4%)
457 (37.2%)
719 (58.5%)

0.53
Education (yrs.) 14.2 ± 2.61 14.8 ± 2.7 0.15
DBP (mm Hg) 79.9 ± 13.9 78.4 ± 10.2 0.42
SBP (mm Hg) 130.6 ± 18.0 131.2 ± 19.0 0.86
Pulse Rate (bpm) 65.8 ± 9.2 64.2 ± 7.7 0.14
Weight (lbs.) 177.9 ± 37.8 172.4 ± 36.2 0.28
BMI 29.3 ± 5.1 28.5 ± 5.4 0.29
Diabetes
(n (%))

9 (16.7%)

103 (8.4%)

0.05
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
the users of respiratory medication and the non-users of respiratory medication.

35

Table 8 (Continued) - Main Characteristics of the Users of Respiratory Medications vs. the
Non-Users of Respiratory Medications
Characteristic
Users of Respiratory
Medication (n=54)
Non-Users of Respiratory
Medication (n=1230)
p-value
Marital Status (n (%))
Single
Married
Separated
Divorced
Widowed
Refused to Answer

6 (11.1%)
27 (50.0%)
3 (5.6%)
13 (24.1%)
5 (9.3%)
0 (0.0%)

119 (9.7%)
719 (58.5%)
34 (2.8%)
218 (17.7%)
139 (11.3%)
1 (0.1%)

0.62
Income (n (%))
<$10k
$10k - $19999
$20k - $29999
$30k - $39999
$40k - $49999
$50k - $59999
$60k+
Refused to Answer

7 (13.0%)
8 (14.8%)
5 (9.3%)
7 (13.0%)
5 (9.3%)
5 (9.3%)
15 (27.8%)
2 (3.7%)

93 (7.6%)
140 (11.4%)
132 (10.7%)
119 (9.7%)
108 (8.8%)
106 (8.6%)
447 (36.3%)
85 (6.9%)

0.66
Blood Pressure Medication
(n (%))

Yes

28 (51.9%)

490 (39.8%)

0.08
Hypolipidemics
Medication
(n (%))
Yes

23 (42.6)

331 (26.9%)

0.01
Anti-Inflammatory Non-
Steroidal (AINS) Use
(n (%))
Yes

3 (5.6%)

3 (0.2%)

<0.01
Anti-Inflammatory
Steroidal (AIS) Use (n (%))

Yes

29 (53.7%)

10 (0.8%)

<0.0001
Bronchial Dilator Use
(n (%))
Yes

46 (85.2%)

22 (1.8%)

<0.0001
Leukotriene Receptor
Antagonist (LTRA) (n (%))

Yes

4 (7.4%)

3 (0.2%)

<0.0001
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
the users of respiratory medication and the non-users of respiratory medication.

A relatively small number of subjects reported the use of respiratory
medication (n=54, 4.2%). The WELLHART trial reported the highest number of

36
users of respiratory medication (6.2%), while the VEAPS trial reported the least
(2.0%) when the tallies are counted in each individual trial.
Table 8 compares the main characteristics of the users of respiratory
medication versus the non-users of respiratory medication. Of all the users of
respiratory medication, 85.2% used bronchial dilators (p<0.0001) (Table 8). The two
respiratory medications groups did not differ on major variables related to CIMT,
including age (p=0.17), gender (p=0.38), race (p=0.89), smoking status (p=0.53) and
education (p=0.15). Users of respiratory medication and non-users of respiratory
medication did differ on the self-report diagnosis of diabetes (p=0.05),
hypolipidemic medications (p=0.08), and blood pressure medications (p=0.10).
Table 9 - Main Characteristics of the Users of Respiratory Medications vs. the Non-Users of
Respiratory Medications (Excluding WELLHART Trial)
Characteristic
Users of Respiratory
Medication (n=40)
Non-Users of
Respiratory
Medication (n=1018)
p-value
CIMT (mm) 0.7560 ± 0.11 0.7560 ± 0.14 1.00
Age (yrs.) 60.9 ± 9.2 59.5 ± 9.4 0.36
Total Cholesterol
(mg/dL)
235.5 ± 40.2 232.1 ± 36.8 0.58
HDL Cholesterol
(mg/dL)
58.7 ± 15.7 55.9 ± 13.9 0.21
LDL Cholesterol
(mg/dL)
149.0 ± 32.9 149.2 ± 33.4 0.98
Triglycerides
(mg/dL)
138.5 ± 62.8 138.0 ± 105.7 0.96
Gender (n (%))
Males
Females
17 (42.5%)
23 (57.5%)
460 (45.2%)
558 (54.8%)
0.74
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
the users of respiratory medication and the non-users of respiratory medication.

37
Table 9 (Continued) - Main Characteristics of the Users of Respiratory Medications vs. the
Non-Users of Respiratory Medications (Excluding WELLHART Trial)
Characteristic
Users of Respiratory
Medication (n=40)
Non-Users of
Respiratory
Medication (n=1018)
p-value
Race (n (%))

White
Black
Hispanic
Asian
Native American

25 (62.5%)
7 (17.5%)
4 (10.0%)
4 (10.0%)
0 (0.0%)

681 (66.9%)
125 (12.8%)
129 (12.7%)
78 (7.7%)
5 (49.1%)

0.80

Smoking Status
(n (%))

Current Smoker
Previous Smoker
Never Smoker

2 (5.0%)
14 (35.0%)
24 (60.0%)

30 (2.9%)
375 (36.8%)
613 (60.2%)

0.75
Education (yrs.) 15.0 ± 2.1 15.4 ± 2.1 0.23
DBP (mm Hg) 82.8 ± 14.1 78.7 ± 10.0 0.08
SBP (mm Hg) 129.9 ± 17.8 128.7 ± 16.9 0.65
Pulse Rate (bpm) 66.2 ± 10.0 64.5 ± 7.6 0.32
Weight (lbs.) 177.8 ± 41.8 174.7 ± 36.4 0.60
BMI 28.3 ± 5.1 28.1 ± 5.3 0.80
Diabetes (n (%))
Yes

0 (0.0%)

9 (0.9%)

1.00
Marital Status
(n (%))
Single
Married
Separated
Divorced
Widowed
Refused to Answer

4 (10.0%)
24 (60.0%)
2 (5.0%)
7 (17.5%)
3 (7.5%)
0 (0.0%)

106 (10.4%)
619 (60.8%)
14 (1.4%)
188 (18.5%)
90 (8.8%)
1 (0.1%)

0.62
Income (n (%))

<$10k
$10k - $19999
$20k - $29999
$30k - $39999
$40k - $49999
$50k - $59999
$60k+
Refused to Answer

1 (2.5%)
5 (12.5%)
3 (7.5%)
6 (15.0%)
4 (10.0%)
5 (12.5%)
15 (37.5%)
1 (2.5%)

31 (3.0%)
87 (8.5%)
109 (10.7%)
101 (9.9%)
95 (9.3%)
103 (10.1%)
435 (42.7%)
57 (5.6%)

0.86
Blood Pressure
Medication
(n (%))

Yes

15 (37.5%)

289 (28.4%)

0.21
Hypolipidemics
Medication
(n (%))

Yes

16 (40.0%)

242 (23.8%)

0.02
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
the users of respiratory medication and the non-users of respiratory medication.

38
Table 9 (Continued) - Main Characteristics of the Users of Respiratory Medications vs. the
Non-Users of Respiratory Medications (Excluding WELLHART Trial)
Characteristic
Users of Respiratory
Medication (n=40)
Non-Users of Respiratory
Medication (n=1018)
p-value
Anti-Inflammatory Non-
Steroidal (AINS) Use (n (%))

Yes

2 (5.0%)

2 (0.2%)

<0.01
Anti-Inflammatory Steroidal
(AIS) Use (n (%))

Yes

23 (57.5%)

6 (0.6%)

<0.0001
Bronchodilator Use (n (%))

Yes

33 (30.6%)

13 (1.3%)

<0.0001
Leukotriene Receptor
Antagonist (LRTA) (n (%))

Yes

3 (7.5%)

2 (0.2%)

<0.0001
NOTE: Data are mean ± SD except where indicated. p-values obtained by Chi-Squared (χ
2
) test
for categorical variables and t-test for continuous variables testing for differences between
the users of respiratory medication and the non-users of respiratory medication.

Table 9 summarizes the main characteristics in the comparison of the users
of respiratory medication versus the non-users of respiratory medication when the
WELLHART trial participants are excluded.
An independent samples t-test comparing the mean CIMT values between the
users of respiratory medication and the non-users of respiratory medication,
performed for the entire sample (n=1284), showed no differences (p=0.90), with the
users of respiratory medication (n=54) having a mean CIMT (SD) of 0.7745mm (0.15)
compared to non-users of respiratory medication (n=1230) having a mean of
0.7716mm (0.16)). Similar results were obtained for mean CIMT with the exclusion
of the WELLHART trial subjects (p=1.00, between groups). The mean CIMT (SD)
value for the users of respiratory medication (n=40) was 0.7560mm (0.11) compared
to the non-users of respiratory medication (n=1018; 0.7560mm (0.14)).

39
In other preliminary analyses, mean CIMT values were compared between
the users of respiratory medication and the non-users of respiratory medication with
stratification on high and low risk categories for lipid variables (e.g. total
cholesterol, high-density lipoprotein [HDL] cholesterol, low-density lipoprotein
[LDL] cholesterol, and total triglycerides), using the American Heart Association
guidelines (American Heart Association, www.aha.org) for optimal cholesterol
levels. These analyses were intended to explore possible modification of the CIMT-
respiratory medication use association by lipid values.
Table 10 and Table 11 summarize the findings for the various cholesterol
risk groups for the entire baseline sample (Table 10) and for the baseline sample
excluding the WELLHART trial data (Table 11).

40
Table 10 – CIMT Comparisons by Respiratory Medications Use Status by Cholesterol Risk
Groups Among Total Sample (n=1284)

TOTAL
CHOLESTEROL
HDL
CHOLESTEROL
LDL
CHOLESTEROL
TOTAL
TRIGLYCERIDES
CHARACTERISTIC Low High Low High Low High Low High
Non-Users of
Respiratory
Medication
227 1003 952 278 96 1134 767 463
Users of
Respiratory
Medication
15 39 44 10 6 48 34 20
Non-Users of
Respiratory
Medication: CIMT
(mm) ± SD
0.780
±
0.192
0.770
±
0.153
0.762
±
0.146
0.806
±
0.203
0.759 ±
0.140
0.773 ±
0.163
0.766
±
0.163
0.781 ±
0.158
Users of
Respiratory
Medication: CIMT
(mm) ± SD
0.815
±
0.180
0.759
±
0.134
0.789
±
0.158
0.711
±
0.072
0.927 ±
0.218
0.755 ±
0.129
0.773
±
0.115
0.778 ±
0.197
p-value 0.47 0.66 0.22 0.002 0.11 0.37 0.75 0.93

Table 11 – CIMT Comparisons by Respiratory Medications Use Status by Cholesterol Risk
Groups Excluding WELLHART Trial (n=1058)

TOTAL
CHOLESTEROL
HDL
CHOLESTEROL
LDL
CHOLESTEROL
TOTAL
TRIGLYCERIDES
CHARACTERISTIC Low High Low High Low High Low High
Non-Users of
Respiratory
Medication
180 838 815 203 67 951 685 333
Users of
Respiratory
Medication
7 33 32 8 2 38 27 13
Non-Users of
Respiratory
Medication: CIMT
(mm) ± SD
0.763 ±
0.177
0.754 ±
0.133
0.750 ±
0.132
0.779 ±
0.175
0.742
±
0.122
0.757 ±
0.143
0.754 ±
0.144
0.760 ±
0.137
Users of
Respiratory
Medication: CIMT
(mm) ± SD
0.795 ±
0.142
0.748 ±
0.109
0.766 ±
0.121
0.716 ±
0.080
0.825
±
0.266
0.752 ±
0.108
0.772 ±
0.113
0.723 ±
0.115
p-value 0.64 0.77 0.50 0.07 0.73 0.80 0.52 0.34

Users of respiratory medication and non-users of respiratory medication
differed in mean CIMT in the high risk HDL cholesterol category (high risk: HDL <45
mg/dL; p=0.002) when the entire baseline sample was used (Table 10). Non-users

41
of respiratory medication had a higher mean CIMT (SD) of 0.806 mm (0.203) in
comparison to users of respiratory medication who had a mean CIMT (SD) of 0.711
mm (0.072). When WELLHART data were excluded users of respiratory medication
and non-users of respiratory medication differed in mean CIMT in the high risk HDL
cholesterol category (high risk: HDL <45 mg/dL; p=0.07). Non-users of respiratory
medication had a higher mean CIMT (SD) of 0.779 mm (0.175) in comparison to
users of respiratory medication who had a mean CIMT (SD) of 0.716 mm (0.080).
A t-test comparison for each of the four individual trials for mean CIMT
values between the users of respiratory medication and the non-users of respiratory
medication showed no statistically significant differences (Table 12).
Table 12 – Association of Respiratory Medications Use Status on CIMT by Trial (n=1284)

BVAIT EPAT VEAPS WELLHART
Non-Users of
Respiratory
Medication
477 195 346 212
Users of
Respiratory
Medication
29 4 7 14
Non-Users of
Respiratory
Medication :
CIMT (mm) ± SD
0.7533 ± 0.15 0.7654 ± 0.13 0.7544 ± 0.13 0.8464 ± 0.22
Users of
Respiratory
Medication :
CIMT (mm) ± SD
0.7565 ± 0.12 0.6976 ± 0.13 0.7870 ± 0.08 0.8274 ± 0.22
p-value 0.91 0.31 0.52 0.75
NOTE: Independent sample t-tests (p-value) were used to test difference between respiratory
medication use groups on CIMT.

42
Section 3.5. Association of Respiratory Medication Use
with CIMT

We analyzed the unadjusted associations of respiratory medication use with
CIMT in linear regression models (Table 13). In the total sample, CIMT was not
associated with respiratory medication use (p=0.90). CIMT was also not associated
with respiratory medication use when stratified by gender in unadjusted models.
Table 13 Linear Regression Relating the Use of Respiratory Medications Status to CIMT
Model Total Sample
(n=1284)
Excluding
WELLHART Trial
(n=1058)
Females (n=807) Males (n=477)
Unadjusted
Estimate (SE)
0.00293 (0.022) -0.0000148
(0.023)
-0.00114 (0.027) 0.00933 (0.040)
p-value 0.90 1.00 0.97 0.81
Adjusted Estimate
(SE)
-0.0090 (0.02) -0.00336 (0.020) -0.0171 (0.024) 0.0096 (0.036)
p-value 0.66 0.87 0.48 0.79
NOTE: ‡ = p<0.05; Adjusted for age, gender, trial, diabetes, race, weight, systolic blood pressure, HDL
cholesterol risk, smoking status, and education.

As these unadjusted associations may be confounded, a multivariate linear
regression model was constructed to adjust for other variables that may affect the
association between respiratory medication use and CIMT. Independent variables
that were initially considered confounders included age, gender, and race. In
addition, we included the following independent variables: indicator variables for
trial, self-reported diagnosis of diabetes, weight, systolic blood pressure, high-
density lipoprotein (HDL) cholesterol risk categories, smoking status (e.g. non-
smoker, previous smoker, current smoker), and total years of education were
included in the final model.

43
With adjustments in the final model, mean CIMT did not differ (p=0.66)
between the users of respiratory medication and the non-users of respiratory
medication (Table 13). Subgroup analysis on gender showed no significant
differences between female users of respiratory medication and female non-users of
respiratory medication (p=0.48) in this final model. No significant differences were
present between male users of respiratory medication and male non-users of
respiratory medication (p=0.79) in this final model.

44
Section 3.6. Association of Respiratory Medication Use with CIMT
Excluding WELLHART Trial Data

When the WELLHART trial data were excluded (n=226), there are n=1018
non-users of respiratory medication and there are n=40 users of respiratory
medication. The total number of users of respiratory medications decreased by
25.9% (n=54 including WELLHART data vs. n=40 excluding WELLHART data). With
the exclusion of the WELLHART data, the unadjusted model estimate for the
respiratory medication parameter was non-significant (p=1.00).
Adjustment for the same variables was done to construct a linear regression
model to evaluate the association between the use of respiratory medication status
and mean CIMT. The users of respiratory medication did not have a statistically
significant difference in mean CIMT in comparison to the non-users of respiratory
medication (p=0.87) according to the final model in which the WELLHART trial data
is excluded.

45
CHAPTER 4: DISCUSSION
Our current analysis extends the evaluation of the association between
asthma (as well as a general collection of key respiratory medications) with
atherosclerosis, specifically carotid artery intima-media thickness (CIMT). Many
studies have hypothesized that respiratory diseases characterized by inflammation,
such as asthma, may correlate directly with atherosclerosis (and higher levels of
CIMT)(Allayee et al., 2007). Plaque rupture correlates with increased inflammation
within the plaque. It is this build-up in the artery walls, together with inflammation,
that lead to atherosclerosis and ultimately CVD (Roger et al., 2011). In this combined
cross-sectional analysis of four trials, we found no evidence of an association
between asthma and CIMT, either in unadjusted or covariate-adjusted analyses
(Table 7). Similar results in mean CIMT occurred when the WELLHART subjects
were excluded (Table 7).
The 5-LO enzyme pathway involves an important cascade of events in
atherosclerosis because the 5-LO enzyme that initiates the pathway occurs not only
during the development of atherosclerotic plaques but also during the progression
of atherosclerotic plaques toward instability (Cipollone et al., 2005).
Resulting products in the cascade of events due the 5-LO enzyme pathway
include leukotrienes. It has been noted that increased leukotriene production, and
the genes that regulate them, are linked to both asthma and atherosclerosis (Dwyer
et al., 2004). Leukotriene (LTB 4) is a pro-inflammatory mediator in the pathogenesis

46
of several inflammatory diseases such as asthma (Cipollone et al., 2005). While no
individual analyses were done with respect to any type of leukotriene, leukotriene
receptor antagonist (LTRA) medications were included in the category of
respiratory medications.
When evaluating the association of respiratory medication use on CIMT, we
did not observe significant differences (p=0.66; Table 13) in our final model after
adjustment, and the same was the case when the WELLHART subjects were
excluded (p=0.87; Table 13). Previous studies have shown that montelukast use was
associated with a significant lower risk for recurrent myocardial infarction in male
subjects (Ingelsson et. al, 2012). However, the sample size for montelukast use was
too small to allow for such an investigation. Also, a subgroup analysis on gender did
not reveal significant results among male users of respiratory medication in
comparison to male non-users of respiratory medication (p=0.79) nor did it reveal
significant results among female users of respiratory medication in comparison to
female non-users of respiratory medication (p=0.48). Our evaluation of the
association between users of respiratory medication (i.e. anti-inflammatory non-
steroidal, anti-inflammatory steroidal, bronchial dilators, leukotriene receptor
antagonists) with CIMT was appropriate because it allows for the comparison on the
use of leukotriene receptor antagonists since previous studies have noted that there
may be a link with the over-expression of leukotrienes produced and
atherosclerosis (Dwyer et al., 2004). The reason why any one of these individual

47
classes of respiratory medications was not analyzed individually is because the
sample size within each class was very low.
Significant differences were found in the univariate analyses when the users
of respiratory medication were compared with the non-users of respiratory
medication on the indicator variables for the self-report diagnosis of diabetes
(p=0.05), the indicator variable for the use of blood pressure medication (p=0.08)
and the indicator variable for the use of hypolipidemics medications (p=0.01) for
the entire sample. Exclusion of the WELLHART trial data only produced univariate,
significant differences in the comparison of the users of respiratory medication with
the non-users of respiratory medication for the variables diastolic blood pressure
(p=0.08) and the use of hypolipidemics medication (p=0.02). The association of
medication use with CIMT in an adjusted model was not found to be statistically
significant (p=0.66) when all trials were combined. This was also the case for the
adjusted model when the WELLHART trial subjects were excluded (p=0.87).
Although not considered in our analysis, the age of asthma onset for
asthmatics and other respiratory diseases requiring the use of respiratory
medication may be an important modifier in investigating a link with
atherosclerosis. As we did not collect data on age at onset, it is likely that our
analysis included subjects with child onset and adult onset asthma; distinguishing
between the two entities could be useful, as the association with CIMT may vary.
Childhood-onset asthma versus adult-onset asthma are two distinct entities;
both in relation to inflammatory pathophysiology and patient susceptibility. Adult-

48
onset asthma is more common in women, while child-onset asthma is more common
in men (Onufrak et al., 2006). Other studies have also shown an increased hazard of
Coronary Heart Disease (CHD) among women and an increased hazard of CHD in
both younger and older women (Iribarren et. al, 2004). Our analyses were limited
by the lack of information on the age of onset, categorization by the current use of
respiratory medications (as opposed to the historical use of respiratory
medications), and lack of medical record documentation of the self-reported
physician-diagnosed asthma. We therefore could not differentiate between child-
onset and adult-onset to determine if prolonged exposure to inflammation-related
asthma may promote higher levels of CIMT . This information may provide greater
insight for future analyses.
The use of leukotriene receptor antagonists (LTRAs) was very limited for
asthmatics and users of respiratory medication (Tables 2-3 & Tables 8-9). Only four
asthmatics and only four users of respiratory medications reported the use of LTRAs.
With regards to the other class of respiratory medications, 2 asthmatics took anti-
inflammatory non-steroidal (AINS) medication, 27 took anti-inflammatory steroidal
(AIS) medication and 35 took bronchial dilators. For those subjects that were
classified as users of respiratory medication only 3 took anti-inflammatory non-
steroidal (AINS) medication while 29 users of respiratory medication took anti-
inflammatory steroidal (AIS) medication and 46 took bronchial dilators. With these
tallies, we were unable to evaluate the association of CIMT with the use of this

49
specific class (LTRAs) of respiratory medications that impacts leukotriene
production. These numbers are very low compared to the overall size of our sample.
Our subjects were compiled from four independent trials, each with separate
inclusion criteria. Using these rich trial databases, various adjustments were made
while evaluating the primary associations of interest. In relation to asthma, the
sample size of asthmatics was relatively small (n=108) and the mean difference
between asthmatics and non-asthmatics was small, with an extremely small effect
size (mean difference divided by SD of IMT) of -0.01. The sample size for users of
respiratory medication (n=54) was even smaller, however the effect size for this
association was larger (effect size = -0.06). The relatively small number of
asthmatics (and respiratory medication users) reflects the fact that these trials in
general recruited healthy volunteers who were not pre-selected for respiratory
conditions.
Another limitation of this analysis was the limited use of LTRAs in these trial
populations, as this is a relatively new class of respiratory medication. Bronchial
dilators were the most frequent method of treatment (Tables 2-3 & Table 8-9).
Thickening of arterial walls may be the result of standard CVD risk factors as
well as inflammatory-related diseases. Our cross-sectional data show no such
association with asthma. While montelukast and low-dose theophylline may
potentially reduce inflammatory-related CVD risk in asthmatics (Allayee et al.,
2007), our data showed in aggregate that use of respiratory medications are
associated with non-significant, higher rather than lower levels of CIMT. Future

50
analyses should consider specific classes of respiratory medications; we did not
have sufficient numbers of users of respiratory medication in this analysis to
perform such class-specific analyses. Finally, our analyses were limited to cross-
sectional associations with CIMT and did not incorporate asthma history (duration,
severity), history of respiratory medication use, or longitudinal measures of CIMT to
assess associations with atherosclerosis progression.
In summary, our analyses provide no evidence on the association of asthma
with subclinically-measured atherosclerosis. These trial data provide limited
evidence of association of respiratory medications with elevated atherosclerosis.
Future such studies will benefit from consideration of class-specific associations of
respiratory medications and asthma history on CIMT, as well as the evaluation of
longitudinal associations of these respiratory variables with atherosclerosis
progression.

51
BIBLIOGRAPHY
Allayee, Hooman, Hartiala, Jaana et al. The Effect of Montelukast and Low-Dose
Theophylline on Cardiovascular Disease Risk Factors in Asthmatics. Chest September
2007; Volume 132, Number 3: 868 – 874.

American Heart Association (AHA). www.aha.org.

Cipollone, Francesco, Mezzetti, Andrea et al. Association Between 5-Lipoxygenase
Expression and Plaque Instability in Humans. Arterioscler Thromb Vasc Biol. 2005;
25:1665-1670.

Crosslin, David R., Shah, Svati H. et al. Genetic Effects in the Leukotriene Biosynthesis
Pathway and Asociation with Atherosclerosis. Hum Genet 2009; 125: 217-229.

Dwyer JH,, Allayee H., Dwyer KM, et al. Arachidonate 5-lipoxygenase Promoter
Genotype, Dietary Arachidonic Acid, and Atheroclerosis. New England Journal of
Medicine 2004; Volume 350: 29-37.

Funk, Colin D.. Leukotriene Modifiers as Potential Therapeutics for Cardiovascular
Disease. Nature Reviews: Drug Discovery; Volume 4, August 2005: 664-672.

Hodis, Howard N., Mack, Wendy J. et al. Estrogen in the Prevention of Atherosclerosis:
A Randomized, Double-Blind, Placebo-Controlled Trial. American College of
Physicians-American Society of Internal Medicine/Annals of Internal Medicine
December 4, 2001; Volume 135, Number 11: 939 – 953.

Hodis, Howard N., Mack, Wendy J. et al. Alpha-Tocopherol Supplementation in
Healthy Individuals Reduces Low-Density Lipoprotein Oxidation but Not
Atherosclerosis: The Vitamin E Atherosclerosis Prevention Study (VEAPS). American
Heart Association/Circulation 2002; Volume 106: 1453 – 1459.

52
Hodis, Howard N., Mack, Wendy J. et al. Hormone Thearpy and the Progression of
Coronary-Artery Atherosclerosis in Postmenopausal Women. New England Journal of
Medicine August 7 2003; Volume 349, Number 3; 535 – 545.
Hodis, Howard N., Mack, Wendy J. et al. High-Dose B Vitamin Supplementation and
Progression of Subclinical Atherosclerosis. Stroke: Journal of the American Heart
Association 2009; 40:00-00.

Ingelsson, Erik, Yin, Li, Back, Magnus. Nationwide Cohort Study of the Leukotriene
Receptor Antagonist Montelukast and Incident or Recurrent Cardiovascular Disease.
The Journal of Allergy and Clinical Immunology March 2012; Volume 129, Issue 3;
702-707.

Iribarren, Carlos, Tolstykh, Irina V., Eisner, Mark D. Are Patients with Asthma at
Increased Risk of Coronary Heart Disease? International Journal of Epidemiology
2004; 33; 743-748.

Künzli, Nino, Jerett, Michael, Mack, Wendy J. et al. Ambient Air Pollution and the
Progression of Atherosclerosis in Adults. PLoS ONE February 2010; Volume 5, Issue 2;
1-10.

Kunzli, Nino, Jerret Michael, Mack, Wendy J. et al. Ambien Air Pollution and
Atherosclerosis in Los Angeles. Environmental Health Perspectives; Volume 113,
Number 2. February 2005.

Onufrak, Stephen, Abramson, Jerome, Vaccarino, Viola. Adult-Onset Asthma is
Associated with Increased Carotid Atherosclerosis Among Women in the
Atherosclerosis Risk in Communities (ARIC) Study. Atherosclerosis 2006.

Roger, Véronique L., Go Alan S., et al. Heart Disease and Stroke Statistics 2011 Update:
A Report From the American Heart Association. Circulation: Journal of the American
Heart Association. 2011; 123; e18-e209.

Zhao, Lei, Funk, Colin D. Lipoxygenase Pathways in Atherogenesis. Trends Cardiovasc
Med 2004; 14:191-195.

Abstract (if available)

Abstract Prior studies have evaluated the role of inflammation on atherosclerosis. The inflammatory process observed in asthmatics plays a central role in the cascade of events that may result in cardiac events such as plaque erosion and plaque fissuring that can trigger Cardiovascular Disease (CVD) complications such as myocardial infarction and stroke. The current research evaluates the association between carotid artery intima-media thickness (CIMT), a measure of subclinical atherosclerosis, and asthma. We also evaluated the association between CIMT and use of asthma respiratory medications. ❧ Baseline data from 1248 subjects in four randomized clinical trials were used to investigate the association between CIMT and asthma and respiratory medication (anti- inflammatory non-steroidal, anti-inflammatory steroidal, bronchial dilator and leukotriene receptor antagonist) use. The primary outcome measure for all trials was CIMT. Asthmatics were identified by self-report on a baseline health questionnaire. Respiratory medication use was ascertained through collection of non-study medication use (prescription and non-prescription) that used a standard medication-coding scheme for all trials. Data were analyzed using multiple linear regression, with CIMT as the dependent variable and self-reported asthma and use of respiratory medications as the primary independent variables. Models were adjusted for age, gender, trial, diabetes, race, weight, systolic blood pressure, high-density lipoprotein (HDL) cholesterol, smoking status, education and income when evaluating the association with asthma. Models were adjusted for age, gender, trial, diabetes, race, weight, systolic blood pressure, HDL cholesterol risk group, smoking status and education when evaluating the association with respiratory medication use. ❧ Mean CIMT did not differ between asthmatics and non-asthmatics in unadjusted (p=0.95) and adjusted (p=0.92) models. Exclusion of the subjects with pre-existing coronary artery disease at baseline (n=226) did not alter these results (p=0.81 for comparison of CIMT between asthmatics and non-asthmatics). ❧ Similar non-significant results were observed evaluating the association of respiratory medication use on CIMT (p=0.66 in adjusted models in the total sample

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The association of apolipoprotein E genotype with cholesterol levels and atherosclerosis: three randomized clinical trials

PDF

Relationship between progression of atherosclerosis and coagulation measures in a randomized-controlled trial

PDF

Relationship of blood pressure and antihypertensive medications to cognitive change in the BVAIT, WISH, and ELITE clinical trials

PDF

The associations of inflammatory markers with progression of subclinical atherosclerosis in early and late postmenopausal women

PDF

Association of arterial stiffness progression with subclinical atherosclerosis measurements in postmenopausal women

PDF

The effects of hormone therapy on carotid artery intima-media thickness and serum lipids by ApoE4 genotype

PDF

Association between endogenous sex hormone levels and cognition: the Women’s Isoflavone Soy Health (WISH) trial

PDF

Bone mineral density is associated with carotid atherosclerosis in healthy postmenopausal women: a longitudinal analysis of randomized clinical trial data

PDF

The effect of vitamin D supplementation on the progression of carotid intima-media thickness and arterial stiffness in elderly African American women: Results of a randomized placebo-controlled trial

PDF

Association of subclinical atherosclerosis with plasma B-vitamin, cysteine, homocysteine, and cysteinyl glycine in a cardiovascular disease-free population

PDF

Associations between inflammatory markers and change in cognitive endpoints

PDF

Associations between physical activities with bone mineral density in postmenopausal women

PDF

Sex hormones and atherosclerosis in postmenopausal women

PDF

Associations between isoflavone soy protein (ISP) supplementation and breast cancer in postmenopausal women in the Women’s Isoflavone Soy Health (WISH) clinical trial

PDF

Association of bone density and bone metabolism markers: the Women’s Isoflavone Soy Health (WISH) trial

$Effect of menopausal hormone replacement therapy on lipoprotein particle fractions and association with carotid artery intima-media thickness and grey-scale median, independent measurements of su...$

PDF

Effect of menopausal hormone replacement therapy on lipoprotein particle fractions and association with carotid artery intima-media thickness and grey-scale median, independent measurements of su...

PDF

The relationship of resistin, leptin, adiponectin, and ghrelin with bone mineral density in healthy postmenopausal women: longitudinal analysis

PDF

Effect of estradiol on circulating levels of inflammatory cytokines in postmenopausal women

PDF

Exploring the relationship between menopausal hot flushes and Alzheimer's disease biomarkers: a cross-sectional analysis in postmenopausal women

PDF

Subclinical atherosclerosis in overweight Latino youth: influence of cardiometabolic risk factors

Asset Metadata

Creator Carballo, Carlos Eliseo (author)

Core Title The association of asthma and asthma-related medications on subclinical atherosclerosis: a cross-sectional analysis of four randomized clinical trials

School Keck School of Medicine

Degree Master of Science

Degree Program Biostatistics

Publication Date 09/24/2012

Defense Date 09/07/2012

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

Tag asthma,asthma-related medications,atherosclerosis,OAI-PMH Harvest

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Mack, Wendy Jean (committee chair), Allayee, Hooman (committee member), Hodis, Howard N. (committee member)

Creator Email carlosecarballo@gmail.com,ccarball@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-97973

Unique identifier UC11288323

Identifier usctheses-c3-97973 (legacy record id)

Legacy Identifier etd-CarballoCa-1213.pdf

Dmrecord 97973

Document Type Thesis

Rights Carballo, Carlos Eliseo

Type texts

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

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

Repository Name University of Southern California Digital Library

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