University of Southern California Dissertations and Theses

Dietary fats, atherosclerosis, and cardiovascular disease

(USC Thesis Other)

Dietary fats, atherosclerosis, and cardiovascular disease

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
DIETARY FATS, ATHEROSCLEROSIS, AND CARDIOVASCULAR DISEASE

by

Nitzan C. Roth

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

December 2009

Copyright 2009 Nitzan C. Roth
ii
ACKNOWLEDGMENTS

First and foremost, I want to thank my advisor, Wendy Mack. I stumbled upon your
research group due to the excellent advice of Stan Azen and stayed past the first summer
due to your excellent mentorship. For the past four years, you served as an inspiring role
model for me as both a researcher and teacher. You nurtured my passion for research in
epidemiology by providing wonderful opportunities and insight and through unwavering
support and encouragement, even in the face of unexpected results and other difficulties.
I appreciate and enjoy all of our conversations, whether about epidemiology or opera.
Thank you.
My dissertation committee members, Hooman Allayee, Eileen Crimmins, Brian
Henderson, Howard Hodis, and Kristine Monroe, provided me with invaluable advice
and guidance. I particularly want to thank Dr. Henderson for offering me the opportunity
to explore cardiovascular risk factors in the Multiethnic Cohort Study, Dr. Hodis for
granting me access to the rich data from the Atherosclerosis Research Unit trials, Dr.
Crimmins for providing me with a broader education through the Multidisciplinary
Research Training in Gerontology fellowship, Dr. Monroe for her friendship and insights
in data management and analysis, and Dr. Allayee for allowing me to gain bench research
experience in his laboratory. During the latter experience, I benefited from the kindness
of Susanna Vikman and other members of the Allayee laboratory, who were incredibly
patient even after realizing that I didn't know how to pipette yet. My dissertation also
would not have been possible without the support of the Keck School of Medicine
M.D./Ph.D. Program and its tireless staff. Thank you.
iii
I want to thank the ladies of the Mack research group. I was fortunate to be able to
follow the exemplary examples of Nicole Gatto, Roksana Karim, Cheryl Vigen, and Ling
Zheng, and to share the journey through graduate school with Shawna Christensen,
Farzana Choudhury, Melissa Frasco, and Yani Lu. Graduate school can be a lonely
place, but you helped to make it fun and to keep me on track. Likewise, I want to thank
all of my wonderful friends for their love and support, particularly: Henry, Beth, Ian,
Jason (who I have complete faith will graduate soon too), Rogier, Joel, and Toi. My life
is filled with joy because of you. Thank you.
Finally, saving the best for last, I want to thank my family. It would be impossible to
list all of the reasons to thank you. When we talk, I get my daily dose of laughter,
support, and love. You encourage me to succeed, believe in me when I struggle, and
rejoice in my accomplishments. Aba, you inspired me to become a scientist and continue
to inspire me with your boundless scientific curiosity. Ema, thank you for keeping me
sane and reminding me that I should neither "work to live" nor "live to work". Yoel, I
appreciate all of your technical advice and have tremendous faith that, regardless of
whether you become the fourth Dr. Roth in the family, there are amazing things in store
for you. Maayan, I valued your insight that graduate school is supposed to be hard, and
have been proud to follow in your footsteps since childhood. I love you all very much.
Thank you.

iv
TABLE OF CONTENTS

ACKNOWLEDGMENTS ii

LIST OF TABLES vi

LIST OF FIGURES viii

ABBREVIATIONS ix

ABSTRACT xi

CHAPTER 1: Literature Review 1
1.1 Introduction to fatty acids 1
1.2 Definitions of cardiovascular disease and atherosclerosis 6
1.3 Biologic mechanisms for the effects of dietary fatty acids on
risk of cardiovascular disease 8
1.3.1 Effects of dietary fatty acids on serum lipids 8
1.3.2 Effects of dietary fatty acids on vascular function 10
1.3.3 Effects of dietary fatty acids on inflammation and
thrombosis 11
1.3.4 Other effects of dietary fatty acids 12
1.4 Chapter 1 References 14

CHAPTER 2: Dietary Intake of Monounsaturated Fat and Marine-Derived
Long-chain N-3 Polyunsaturated Fatty Acids Reduces the Risk of
Myocardial Infarction: the Multiethnic Cohort Study 17
2.1 Introduction 17
2.2 Materials and Methods 18
2.2.1 Population 18
2.2.2 Ascertainment of diet 19
2.2.3 Follow-up and ascertainment of MI 20
2.2.4 Statistical analysis 21
2.3 Results 23
2.4 Discussion 40
2.5 Chapter 2 References 45

CHAPTER 3: Cross-sectional Associations Between 5-Lipoxygenase Promoter
Polymorphisms, Dietary Fats, and Carotid Artery Atherosclerosis 49
3.1 Introduction 49
3.2 Methods 51
3.2.1 Study populations 51
3.2.2 Clinical and laboratory measurements 52
3.2.3 Carotid intima-media thickness 53
v
3.2.4 Dietary assessment 54
3.2.5 Genotyping 54
3.2.6 Statistical analysis 55
3.3 Results 56
3.3.1 Characteristics of the study populations 56
3.3.2 Associations of 5-LO promoter polymorphisms with
carotid artery atherosclerosis 58
3.3.3 Gene-diet interactions between 5-LO promoter genotype
and dietary fats on carotid artery atherosclerosis 60
3.3.4 Interactions between 5-LO promoter genotype and plasma
lipids on carotid artery atherosclerosis 63
3.4 Discussion 64
3.5 Chapter 3 References 69

CHAPTER 4: Validation of Multiethnic Cohort Study Self-Reports of
Physician-Diagnosed Myocardial Infarction and Other Conditions
Through Comparison with California Hospital Discharge Data 74
4.1 Introduction 74
4.2 Materials and Methods 75
4.2.1 Population 75
4.2.2 Assessment of medical history via self-reports 76
4.2.3 Assessment of medical history via hospital records 76
4.2.4 Statistical analysis 77
4.3 Results 79
4.4 Discussion 86
4.5 Chapter 4 References 91

BIBLIOGRAPHY 93

APPENDICES
Appendix A Dietary Intake of Monounsaturated Fat and Marine-
Derived Long-Chain N-3 Polyunsaturated Fatty Acids
Reduces the Risk of Myocardial Infarction: the
Multiethnic Cohort Study 104
Appendix B Cross-sectional Associations Between 5-Lipoxygenase
Promoter Polymorphisms, Dietary Fats, and Carotid
Artery Atherosclerosis 118

vi
LIST OF TABLES

Table 1.1: Fatty acid composition of common edible fats and oils 3

Table 1.2: Fatty acid composition of common fat- and protein-containing
foods 4

Table 2.1: Cardiovascular risk factors and risk of MI among MEC men 24

Table 2.2: Cardiovascular risk factors and risk of MI among MEC women 28

Table 2.3: Baseline diet of the MEC, stratified by sex and ethnicity 32

Table 2.4: Dietary fats and risk of MI in the MEC using carbohydrate and
protein replacement models 33

Table 2.5: EPA plus DHA and fish and risk of MI in the MEC using
carbohydrate and protein replacement models 35

Table 2.6: Dietary fats and fish and risk of MI among men in the MEC
using carbohydrate and protein replacement models, stratified by
ethnicity 36

Table 2.7: Dietary fats and fish and risk of MI among women in the MEC
using carbohydrate and protein replacement models, stratified by
ethnicity 38

Table 3.1: Baseline characteristics of subjects with known 5-LO promoter
genotypes from four randomized controlled atheroclerosis trials 56

Table 3.2: Baseline diet of subjects with known 5-LO promoter genotypes
from four randomized controlled atherosclerosis trials 57

Table 3.3: 5-LO promoter genotype and cardiovascular risk factors 59

Table 3.4: CIMT according to 5-LO promoter genotype 61

Table 4.1: Sensitivity of MEC self-reports compared to OSHPD records,
by diagnosis 81

Table 4.2: Odds ratios for confirmation of OSHPD diagnoses of MI by
MEC self-reports according to patient and admission
characteristics 83

vii
Table A.1: Dietary fats and risk of MI in the MEC using a saturated fat
replacement model 104

Table A.2: EPA plus DHA and fish and risk of MI in the MEC using
saturated fat replacement models 105

Table A.3: Dietary fats and fish and risk of MI in the MEC using
carbohydrate and protein replacement models, stratified by
ethnicity 106

Table A.4: Dietary fats and fish and risk of MI in the MEC using
carbohydrate and protein replacement models, stratified by sex 108

Table A.5: Dietary fats and fish and risk of MI in the MEC using saturated
fat replacement models, stratified by ethnicity 110

Table A.6: Dietary fats and fish and risk of MI in the MEC using saturated
fat replacement models, stratified by sex 112

Table A.7: Dietary fats and fish and risk of MI in the MEC using saturated
fat replacement models, stratified by sex and ethnicity 114

Table B.1: Design and eligibility criteria of four randomized controlled
atherosclerosis trials 118

Table B.2: Baseline characteristics of subjects who were or were not
genotyped 119

Table B.3: Baseline characteristics of genotyped subjects with or without
known 5-LO genotypes 120

Table B.4: 5-LO promoter allele frequencies by ethnicity 121

Table B.5: LDL cholesterol according to 5-LO promoter genotype 122

Table B.6: CIMT according to dietary fat intakes 123

Table B.7: CIMT according to 5-LO promoter genotype and dietary fat
intakes 124

Table B.8: CIMT according to 5-LO promoter genotype and plasma lipids
among non-users and users of lipid-lowering medication 126

viii
LIST OF FIGURES

Figure 3.1: Interaction between dietary intake of linoleic acid and 5-LO
promoter genotype on CIMT 62

Figure 3.2: Interaction between use of lipid-lowering medication and 5-LO
promoter genotype on CIMT 64

ix
ABBREVIATIONS

5-LO 5-lipoxygenase

ALA (alpha-)linolenic acid [18:3n-3]
Ara arachidonic acid [20:4n-6]
BMI body mass index
BVAIT B-Vitamin Atherosclerosis Intervention Trial
CABG coronary artery bypass grafting
CCA common carotid artery
CIMT carotid intima-media thickness
CHD coronary heart disease
CVD cardiovascular disease
DBP diastolic blood pressure
DHA docosahexaenoic acid [22:6n-3]
ELITE Early versus Late Intervention Trial with Estradiol
EPA eicosapentaenoic acid [20:5n-3]
EPAT Estrogen in the Prevention of Atherosclerosis Trial
HDL high-density lipoprotein
hs-CRP high-sensitivity C-reactive protein
LA linoleic acid [18:2n-6]
LDL low-density lipoprotein
MEC Multiethnic Cohort
MI myocardial infarction
x
MUFA monounsaturated fatty acid
NSAID non-steroidal anti-inflammatory drug
PCI percutaneous coronary intervention
PUFA polyunsaturated fatty acid
SBP systolic blood pressure
SFA saturated fatty acid
sICAM-1 soluble intercellular adhesion molecule-1
VEAPS Vitamin E Atherosclerosis Prevention Study
VLDL very low-density lipoprotein
WELL-HART Women's Estrogen-Progestin Lipid-Lowering Hormone
Atherosclerosis Regression Trial

WISH Women's Isoflavone Soy Health trial

xi
ABSTRACT

Atherosclerosis is the underlying cause of myocardial infarction (MI) and numerous
environmental and genetic factors contribute to its development. Despite advances in our
understanding of atherogenesis, many uncertainties on the relationship between diet and
cardiovascular disease (CVD) risk remain. Chapter 1 reviews the biologic mechanisms
for the effects of dietary fats on CVD risk, including effects on serum lipids, vascular
function, chronic inflammation, and thrombosis.
Chapters 2 and 3 examine the effects of sex, ethnicity, and genetic variation on the
associations between specific types of dietary fat and CVD risk and atherosclerosis.
Among 53,120 Latino and African-American men and women in the Multiethnic Cohort
Study, greater intakes of monounsaturated fat and marine-derived long-chain n-3
polyunsaturated fatty acids were associated with lower risk of MI. Greater intakes of
saturated fat and n-6 polyunsaturated fat were associated with higher risk of MI among
Latinos. Among 1,827 healthy subjects who participated in one of four randomized
controlled atherosclerosis trials, atherosclerosis measured using carotid intima-media
thickness (CIMT) was not associated with promoter polymorphisms for 5-lipoxygenase,
an inflammatory gene in the overall sample. However, there were significant gene-
environment interactions such that homozygosity for shorter 5-lipoxygenase promoter
alleles was atherogenic among users of lipid-lowering medication or individuals with low
dietary intake of linoleic acid, the intermediate-chain n-6 polyunsaturated fatty acid that
comprised the majority of energy intake from polyunsaturated fats. Taken together, these
findings suggest that nutritional recommendations for prevention of cardiovascular
xii
disease may need to account for differences in the relationship between diet and
cardiovascular disease across population subgroups. Further investigation is needed to
understand the interactions between diet, inflammation, and other factors underlying
atherosclerosis and cardiovascular disease.
Chapter 4 validates Multiethnic Cohort Study self-reports of physician-diagnosed MI
and other conditions through comparison with a statewide database of California hospital
discharge diagnoses. The sensitivity of self-reports ranged from 59 to 93% and was
higher for chronic conditions (such as hypertension, diabetes, and cancers) than for acute
events (such as MI, stroke, hip fracture, and cholecystectomy). Sensitivity of self-reports
also varied by sex, ethnicity, age, education, length of the hospital stay, and illness
severity.

1
Chapter 1:
Literature Review

1.1 Introduction to fatty acids
Fatty acids are a class of lipid compounds that consist of a hydrocarbon skeleton with
a carboxyl group (-COOH) at one end (1). They are classified as saturated fatty acids
(SFAs), monounsaturated fatty acids (MUFAs), or polyunsaturated fatty acids (PUFAs)
based on the number of double bonds between atoms in their carbon chain: SFAs have no
double bonds and are therefore fully saturated with hydrogen atoms, whereas MUFAs
and PUFAs have one double bond and more than one double bond per molecule,
respectively (1). Individual fatty acids can be identified by formal chemical names that
indicate the length of the carbon chain and the locations of any double bonds. In medical
literature, however, they are usually identified using common names and a notation
[C:N], where C is the length of the carbon chain and N is the number of double bonds.
Because double bonds prevent rotation of the participating carbon atoms along the bond
axis, unsaturated fatty acids can also be classified as cis or trans isomers based on their
three-dimensional configurations. MUFAs and PUFAs are further classified as omega-3
(n-3), omega-6 (n-6), or omega-9 (n-9) based on whether the carbon furthest away from
the carboxyl group is three, six, or nine carbons away from a double bond (1).
Classification of fatty acids is important because the chemical and physical properties
of the molecules are reflected in their differing dietary sources and physiologic effects,
discussed below. Fatty acids also serve as the "building blocks" for other lipid

2
compounds -- including monoglycerides (a combination of glycerol plus one fatty acid),
diglycerides (a combination of glycerol plus two fatty acids), triglycerides (a combination
of glycerol plus three fatty acids), phospholipids (a combination of glycerol plus one or
two fatty acids plus a phosphoric acid compound), and sterol esters (a combination of
sterol plus one fatty acid) (1) -- that have a wide range of physiologic actions including
supplying energy and serving as structural components of cell membranes throughout the
body.
Edible fats and oils consist of monoglycerides, diglycerides, and triglycerides in solid
or liquid form (1). Their fatty acid compositions depend on their sources (Table 1.1) (2).
Fats from animal sources (e.g., butter and lard) and certain plant oils (e.g., coconut oil)
contain high amounts of SFAs and very low amounts of PUFAs. In contrast, other plant
oils contain relatively low amounts of SFAs and higher amounts of MUFAs (e.g., almond
oil, canola oil, and olive oil) or PUFAs (e.g., corn oil, flaxseed oil, safflower oil, sesame
oil, soybean oil, sunflower oil, and walnut oil). With the exceptions of flaxseed oil and
canola oil, the ratio of n-6 to n-3 PUFAs in commonly used oils is relatively high. The
fatty acid composition of fat- and protein-containing food items varies similarly (Table
1.2) (2). Foods from land animals (e.g., beef, pork, and cheese) contain higher amounts
of SFAs than those from poultry (e.g., chicken, turkey, and eggs), fish, or plant sources
(e.g., beans and nuts). Plant-derived n-6 and n-3 PUFAs such as linoleic acid (LA)
[18:2n-6], α-linolenic acid (ALA) [18:3n-3], and arachidonic acid (Ara) [20:4n-6] are
found in large amounts not only in beans and nuts but also in eggs and poultry. However,
the marine-derived long-chain n-3 PUFAs -- namely, eicosapentaenoic acid (EPA)

3

4

5
[20:5n-3] and docosahexaenoic acid (DHA) [22:6n-3] -- are found predominantly in fatty
fish and other seafood, with only small amounts found in eggs and poultry.
In the United States, fat accounts for an average of 34% of total energy intake (3). As
percents by weight, SFAs, MUFAs, and PUFAs account for roughly 36-37%, 41-42%,
and 22% of total fat, respectively, although within- and between-person variation is high
(3). LA, the main n-6 PUFA, comprises 84-89% of energy intake from PUFAs while
ALA, the n-3 homologue of LA, comprises only 9-11% of energy intake from PUFAs
(4). Longer-chain n-6 and n-3 PUFAs (e.g., Ara, EPA, and DHA) together provide less
than 5% of total PUFA energy (4). As a result, the ratio of n-6 to n-3 PUFAs consumed
in the United States is quite high, estimated at 15-20:1, probably due to increased
consumption of LA over several decades (4, 5). In comparison, the ratio of n-6 to n-3
PUFAs consumed in European countries has increased more moderately to 7-16:1 (5) and
in Japan, remained relatively stable over the past two decades at approximately 4:1 (6)
due to more frequent consumption of fatty fish.
LA and ALA are unique among fatty acids in that they are "essential" in the diet
because they cannot be synthesized de novo in mammals and must be consumed in
modest amounts in order to ensure normal physiologic functions (7-9). Long-chain n-6
and n-3 PUFAs are not strictly "essential" because they can be synthesized from LA and
ALA, respectively (9). LA and ALA compete as substrates for the Δ6- and Δ5-desaturase
enzymes that convert them into Ara and EPA, respectively (9). Ara and EPA can then be
converted into other long-chain PUFAs (e.g., docosapentaenoic acid, DPA [22:5n-6] and

6
DHA) or a variety of metabolites (7-9). Importantly, n-6 PUFAs cannot be converted
into n-3 PUFAs or vice versa in mammalian cells (9).

1.2 Definitions of cardiovascular disease and atherosclerosis
Cardiovascular disease (CVD), a generic term for disorders of the heart or blood
vessels, has been the leading cause of death in the United States since 1900 (10).
Although CVD encompasses a wide range of conditions including congenital heart
diseases, valvular heart diseases, cardiomyopathies, and peripheral vascular diseases, by
far the single largest contributors to CVD mortality are myocardial infarction (MI),
known colloquially as a "heart attack", and stroke (10). While an MI or stroke is a
sudden event, the underlying cause of both conditions is the progressive narrowing of
large- and medium-sized arteries that develops over many years and is known as
atherosclerosis (11), a term that stems from the Greek words "athere", meaning gruel or
grit, and "skleros", meaning hard (12). Today, either directly or indirectly,
atherosclerosis causes over one half of the total number of deaths in the United States --
more than the next six leading causes of death combined (10).
Atherosclerosis was once considered to be a degenerative disease that inevitably
progressed with age due to the accumulation of lipids in arterial walls. The classic "diet-
heart" hypothesis, developed largely based on evidence from animal studies in the 1930s
and ecologic studies in the 1960s and 1970s, ascribed to that view and suggested that
high dietary consumption of fat and cholesterol leads to accelerated progression of
atherosclerosis due to increases in total serum cholesterol concentrations that promote

7
lipid accumulation (3). During the past few decades, advances in understanding the
biochemical mechanisms of atherosclerosis and landmark discoveries of atherogenic risk
factors, have led researchers to adopt a new paradigm that recognizes atherosclerosis as a
complex, highly-regulated and active process involving resident cells of the arterial wall
and infiltrating lipoproteins and chronic inflammatory cells (8). Diet has far-reaching
effects on many, if not all, of the disease's components. Nutritional epidemiologist
Dariush Mozaffarian wrote:
...the impact of diet on disease risk cannot be judged solely by effects on total and
LDL cholesterol levels, as dietary habits also affect numerous other intermediary
risk factors, including other circulating lipoproteins, vascular hemodynamics,
inflammation, endothelial function, insulin sensitivity, satiety and weight gain,
coagulation and thrombosis, and arrhythmic risk. (13)

Moreover, it is clear that a large number of dietary factors other than total fat composition
of foods may influence risk of CVD, including: consumption of specific fats (e.g., SFAs,
MUFAs, n-6 PUFAs, n-3 PUFAs, and trans fatty acids); consumption of carbohydrates
and their sources (including their quality as judged by their fiber content or glycemic
index); consumption of protein and its sources; alcohol consumption; consumption of
fruits and vegetables; intake of micronutrients (e.g., vitamins, minerals, and
antioxidants); and food processing and preparation methods (3, 10). Despite this
extensive list of potential dietary risk factors for CVD, there remains considerable
interest in the roles of dietary fats due to their wide range of physiologic effects.

8
1.3 Biologic mechanisms for the effects of dietary fatty acids on risk of
cardiovascular disease
1.3.1 Effects of dietary fatty acids on serum lipids
Recent studies suggest that total serum cholesterol is not as strong of a predictor of
CVD risk as serum concentrations of triglycerides, low-density lipoprotein (LDL)
cholesterol, and high-density lipoprotein (HDL) cholesterol, and their respective ratios
(3). LDL cholesterol, known colloquially as the "bad" cholesterol, accumulates in the
intima of arteries by binding to proteoglycans in the extracellular matrix and then
undergoes biochemical modification. Modified LDL cholesterol promotes further
atherogenesis because, unlike normal LDL cholesterol, it is not recognized by surface
LDL receptors regulated by negative feedback inhibition (11). Instead, modified LDL
cholesterol is ingested by macrophages and other cells in unlimited quantities through
scavenger receptors, forming engorged "foam cells" found in atherosclerotic lesions (11).
Modified LDL cholesterol and foam cells also induce endothelial expression of cellular
adhesion molecules, chemokines, and inflammatory factors that promote the recruitment
and entry of circulating leukocytes and smooth muscle cells from the media into the
intimal layer (11). Small, dense LDL particles are particularly harmful. In contrast, HDL
cholesterol, known colloquially as the "good" cholesterol, appears to protect against
atherosclerosis, possibly due to its role in facilitating transport of lipids from peripheral
tissues to the liver for disposal (11).
In a pooled analysis of experimental studies on dietary fatty acids and serum lipids,
Hegsted et al. found that serum total cholesterol and LDL cholesterol concentrations

9
were elevated by increased intake of SFAs and dietary cholesterol and decreased intake
of PUFAs, but were not significantly altered by changes in intake of MUFAs (14). In the
same study, predicted changes in serum HDL cholesterol concentrations were positively
associated with intakes of SFAs and PUFAs. A more recent pooled analysis of 60 studies
(15) confirmed the earlier findings that serum total cholesterol was positively associated
with intake of SFAs, inversely associated with intake of PUFAs, and not significantly
associated with intake of MUFAs. However, Mensink et al. found that all three major
classes of fatty acids (i.e., SFAs, cis MUFAs, and cis PUFAs) had significant effects on
serum LDL cholesterol and HDL cholesterol concentrations. In particular, isocaloric
replacement of carbohydrates with SFAs was associated with increases in LDL and HDL
cholesterol but not in the ratio of total cholesterol to HDL cholesterol, a factor that the
authors argued is more predictive of CVD risk. In comparison, both cis MUFAs and cis
PUFAs were associated with decreases in LDL cholesterol and increases in HDL
cholesterol when replacing carbohydrates, leading to decreases in the ratio of total
cholesterol to HDL cholesterol. Increased intake of trans MUFAs, on the other hand,
substantially increased LDL cholesterol without changing HDL cholesterol
concentrations and thereby increased the ratio of total cholesterol to HDL cholesterol.
All three major classes of fatty acids, but not trans MUFAs, were also associated with
decreases in serum triglycerides when replacing carbohydrates in the diet. In addition,
serum concentrations of apolipoprotein B (Apo B), a marker of the number of LDL
particles, were significantly increased by isocaloric replacement of carbohydrates with
trans MUFAs and decreased by replacement of carbohydrates with cis MUFAs or cis

10
PUFAs, whereas serum concentrations of apolipoprotein A-I (Apo A-I), a marker of the
number of HDL particles, were significantly increased by replacement of carbohydrates
with SFAs or cis MUFAs.
The effects of dietary fatty acids on serum lipids are not homogeneous even within the
three major classes of fatty acids. In particular, the triglyceride-lowering effect of PUFA
intake appears to be stronger for marine-derived n-3 PUFAs -- namely, EPA and DHA
from fish oil -- than for n-6 PUFAs such as LA or plant-derived n-3 PUFAs such as ALA
(16-18). Fish oil is also unique in that unlike LA and ALA, it significantly reduces serum
very low-density lipoprotein (VLDL) cholesterol but may increase LDL cholesterol (16,
19), an effect that may be due to increases in LDL particle size (17). Studies also suggest
that intake of PUFAs, especially EPA and DHA, may increase the susceptibility of LDL
and VLDL cholesterol to oxidative modification when compared to MUFAs (18-21).
1.3.2 Effects of dietary fatty acids on vascular function
According to the "response-to-injury" hypothesis, atherogenesis is initiated by injuries
to the arterial endothelium due to hemodynamic forces (i.e., shear stress due to turbulent
blood flow in regions of arterial branching or curvature) or reactive oxygen species (e.g.,
superoxide O
2
•
-
) that alter its normal functions (11). In particular, endothelial
dysfunction is characterized by increased permeability of the endothelium and altered
expression of endothelium-derived substances (11). Although it is unclear whether
dietary fatty acids directly increase or decrease production of reactive oxygen species in
vivo (22, 23), intake of MUFAs and PUFAs counteracts some injury-induced changes in
endothelial cell expression. Dietary MUFAs and PUFAs increase endothelial secretion of

11
nitric oxide, an effect that allows vasodilation and reductions in blood pressure and
promotes an antiinflammatory and antithrombotic state. Intake of n-3 PUFAs,
particularly EPA and DHA, also enhances production of endothelium-derived vascular
relaxing factor (which further improves vasodilation) (23, 24) and inhibits endothelial
cell expression of inflammatory factors such as E-selectin, vascular cell adhesion
molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1), interleukin-4 (IL-
4), tumor necrosis factor-α (TNF-α), bacterial endotoxin, and interleukin-6 (IL-6) and
interleukin-1 (IL-1) in response to interleukin-1 (IL-1) (9, 22-25). Similar but more
modest effects on endothelial cell expression are seen with intake of MUFAs or n-6
PUFAs, but not SFAs (25).
Vascular smooth muscle cell proliferation is another characteristic of atherosclerosis
progression, particularly in later stages (11). Dietary PUFAs, and to a lesser extent
dietary MUFAs, inhibit smooth muscle cell proliferation by interfering with DNA
synthesis or progression of the cell cycle (22, 24). DHA may also act to induce apoptosis
of smooth muscle cells (23). SFA-induced changes in vascular smooth muscle cell
proliferation have not been reported.
1.3.3 Effects of dietary fatty acids on inflammation and thrombosis
In addition to their effects on endothelial cell expression, dietary PUFAs reduce
recruitment of circulating leukocytes into atheromatous plaques by modifying gene
expression in leukocytes themselves (22, 24). For example, dietary n-3 PUFAs reduce
expression of monocyte chemoattractant protein-1 (MCP-1) in undifferentiated

12
monocytes and intake of EPA and DHA reduces TNF-α and IL-1 expression in
macrophages (24).
The effects of dietary PUFAs on inflammation and thrombosis are also driven in large
part by competition between Ara and EPA as substrates for enzymes that catalyze release
of the fatty acids from cell membranes (22) or conversion of the fatty acids into a variety
of metabolites (8, 23). Ara can be converted by cyclooxygenase into prostaglandin (PG)
H
2
(PGH
2
) or by 5-lipoxygenase into leukotriene (LT) A
4
(LTA
4
) (8, 26). PGH
2
can be
further metabolized to produce PGE
2
, PGI
2
, or thromboxane (TX) A
2
(TXA
2
), whereas
LTA
4
can be further metabolized to produce LTB
4
or the cysteinyl LTs (LTC
4
, LTD
4
, and
LTE
4
) (8, 26, 27). Alternatively, the same enzymes can metabolize EPA to produce
PGE
3
, PGI
3
, TXA
3
, LTB
5
, LTC
5
, LTD
5
, and LTE
5
(8, 23, 26). The Ara-derived LTs are
potent proinflammatory lipid mediators that have been associated with a variety of acute
and chronic inflammatory diseases including asthma (8, 27). Similarly, TXA
2
stimulates
vasoconstriction and platelet aggregation (23, 26). In contrast, the EPA-derived LTs and
TXA
3
are associated with less inflammatory and thrombotic activity (8, 23). Ara-derived
PGI
2
and EPA-derived PGI
3
have roughly equivalent biologic activity (23). Increasing n-
3 PUFA intake, particularly of EPA and DHA, and decreasing n-6 PUFA intake may
therefore lead to production of less potent inflammatory and thrombotic factors and a
shift towards an antiinflammatory and antithrombitic state (8, 23, 24).
1.3.4 Other effects of dietary fatty acids
Other physiologic effects of dietary fatty acids that have been reported include their
effects on insulin sensitivity, coagulation, and arrythmogenesis (3, 22, 24). Both n-6 and

13
n-3 PUFAs have been shown to promote insulin secretion, an effect that could lead to
insulin resistance as seen in type 2 diabetes and the metabolic syndrome (7). High
dietary intake of MUFAs may reduce coagulation by decreasing plasma levels of von
Willebrand factor, plasminogen activator inhibitor type 1, α2-antiplasmin, and
coagulation factors VII and XIIa (22). Dietary EPA and DHA consumed in fish oil may
also reduce coagulation by decreasing plasma levels of fibrinogen and coagulation factors
VII and X (22), while dietary LA may enhance coagulation by increasing plasma levels
of coagulation factor VII (7). Finally, although it is unclear whether n-6 PUFAs are pro-
or anti-arrhythmogenic (7), there is growing evidence that dietary intake of EPA and
DHA protects against potentially fatal arrhythmias including ventricular fibrillation and
sudden cardiac death (13, 24, 28, 29).

14
1.4 Chapter 1 References

1. Marks DB. Board Review Series. Biochemistry, 3rd ed. Lippincott Williams &
Williams: Baltimore, ML; 1999.

2. United States Department of Agriculture (USDA), Agricultural Research Service.
What's In The Food You Eat Search Tool, 3.0. USDA Home Page,
http://www.ars.usda.gov/Services/docs.htm?docid=17032 .

3. Willett W. Nutritional Epidemiology, 2nd ed. Oxford University Press: New York,
NY: 1998.

4. Kris-Etherton PM, Taylor DS, Yu-Poth S, et al. Polyunsaturated fatty acids in the
food chain in the United States. Am J Clin Nutr 2000; 71(suppl): 179S-188S.

5. Sanders TAB. Polyunsaturated fatty acids in the food chain in Europe. Am J Clin
Nutr 2000; 71(suppl): 176S-178S.

6. Sugano M, Hirahara F. Polyunsaturated fatty acids in the food chain in Japan. Am J
Clin Nutr 2000; 71(suppl): 189S-196S.

7. Dubnov G, Berry EM. Omega-6 fatty acids and coronary artery disease: the pros and
cons. Curr Atheroscler Rep 2004; 6: 441-446.

8. James MJ, Gibson RA, Cleland LG. Dietary polyunsaturated fatty acids and
inflammatory mediator production. Am J Clin Nutr 2000; 71(suppl): 343S-348S.

9. Simopoulos AP. The importance of the omega-6/omega-3 fatty acid ratio in
cardiovascular disease and other chronic diseases. Exp Biol Med 2008; 233: 674-688.

10. Rosamond W, Flegal K, Furie K, et al. Heart disease and stroke statistics 2008
update: a report from the American Heart Association Statistics Committee and Stroke
Statistics Subcommittee. Circulation 2008; 117: e25-e146.

11. Lusis AJ. Atherosclerosis. Nature 2000; 407: 233-241.

12. Bing R. Cardiology: The Evolution of the Science and the Art. Rutgers University
Press: Piscataway, NJ: 1999.

13. Mozaffarian D. Fish and n-3 fatty acids for the prevention of fatal coronary heart
disease and sudden cardiac death. Am J Clin Nutr 2008; 87(suppl): 1991S-1996S.

14. Hegsted DM, Ausman LM, Johnson JA, et al. Dietary fat and serum lipids: an
evaluation of the experimental data. Am J Clin Nutr 1993; 57: 875-883.

15

15. Mensink RP, Zock PL, Kester ADM, et al. Effects of dietary fatty acids and
carbohydrates on the ratio of serum total to HDL cholesterol and on serum lipids and
apolipoproteins: a meta-analysis of 60 controlled trials. Am J Clin Nutr 2003; 77: 1146-
1155.

16. Kestin M, Clifton P, Belling B, et al. n-3 fatty acids of marine origin lower systolic
blood pressure and triglycerides but raise LDL cholesterol compared with n-3 and n-6
fatty acids from plants. Am J Clin Nutr 1990; 51: 1028-1034.

17. Griffin MD, Sanders TAB, Davies IG, et al. Effects of altering the ratio of dietary n-
6 to n-3 fatty acids on insulin sensitivity, lipoprotein size, and postprandial lipemia in
men and postmenopausal women aged 45-70 y: the OPTILIP Study. Am J Clin Nutr
2006; 84: 1290-1298.

18. Finnegan YE, Minihane AM, Leigh-Firbank EC, et al. Plant- and marine-derived n-
3 polyunsaturated fatty acids have differential effects on fasting and postprandial blood
lipid concentrations and on the susceptibility of LDL to oxidative modification in
moderately hyperlipidemic subjects. Am J Clin Nutr 2003; 77: 783-795.

19. Hau MF, Smelt AHM, Bindels AJGH, et al. Effects of fish oil on oxidative
resistance of VLDL in hypertriglyceridemic patients. Arterioscler Thromb Vasc Biol
1996; 16: 1197-1202.

20. Mata P, Varela O, Alonso R, et al. Monounsaturated and polyunsaturated n-6 fatty
acid-enriched diets modify LDL oxidation and decrease human coronary smooth muscle
cell DNA synthesis. Arterioscler Thromb Vasc Biol 1997; 17: 2088-2095.

21. Tsimikas S, Philis-Tsimikas A, Alexopoulos S, et al. LDL isolated from Greek
subjects on a typical diet or from American subjects on an oleate-supplemented diet
induces less monocyte chemotaxis and adhesion when exposed to oxidative stress.
Arterioscler Thromb Vasc Biol 1999; 19: 122-130.

22. Moreno JJ, Mitjavila MT. The degree of unsaturation of dietary fatty acids and the
development of atherosclerosis (review). J Nutr Biochem 2003; 14: 182-195.

23. Mori TA, Beilin LJ. Omega-3 fatty acids and inflammation. Curr Atheroscler Rep
2004; 4: 461-467.

24. Jung UJ, Torrejon C, Tighe AP, et al. n-3 fatty acids and cardiovascular disease:
mechanisms underlying beneficial effects. Am J Clin Nutr 2008; 87(suppl): 2003S-
2009S.

16
25. De Caterina R, Liao JK, Libby P. Fatty acid modulation of endothelial activation.
Am J Clin Nutr 2000; 71(suppl): 213S-223S.

26. Vanden Heuvel JP. Diet, fatty acids, and regulation of genes important for heart
disease. Curr Atheroscler Rep 2004; 6: 432-440.

27. Radmark O, Werz O, Steinhilber D, et al. 5-lipoxygenase: regulation of expression
and enzyme activity. Trends Biochem Sci 2007; 32(7): 332-341.

28. Kang JX, Leaf A. Prevention of fatal cardiac arrhythmias by polyunsaturated fatty
acids. Am J Clin Nutr 2000; 71(suppl): 202S-207S.

29. Chrysohoou C, Panagiotakos DB, Pitsavos C, et al. Long-term fish consumption is
associated with protection against arrhythmia in healthy persons in a Mediterranean
region -- the ATTICA study. Am J Clin Nutr 2007; 85: 1385-1391.

17
Chapter 2:
Dietary Intake of Monounsaturated Fat and Marine-Derived Long-chain N-3
Polyunsaturated Fatty Acids Reduces the Risk of Myocardial Infarction:
the Multiethnic Cohort Study

2.1 Introduction
Current dietary recommendations for reducing the risk of myocardial infarction (MI)
and other cardiovascular diseases (CVD) include limiting intake of saturated fat and
increasing intake of polyunsaturated fat, particularly from fish sources (1). These
recommendations are largely based on evidence from animal studies and human clinical
trials that greater intake of saturated fat adversely affects serum lipids (2), whereas
greater intake of unsaturated fats, particularly the marine long-chain n-3 polyunsaturated
fatty acids eicosapentaenoic acid (EPA) [20:5n-3] and docosahexaenoic acid (DHA)
[22:6n-3], has beneficial effects on serum lipids (2, 3), vascular function (4-7),
inflammation and thrombosis (4-8), coagulation (4), insulin secretion (4), and arrhythmic
risk (9). However, prospective epidemiologic studies on the relations between specific
types of dietary fat and risk of CVD have yielded inconsistent results (10-14), even for
consumption of fish and EPA and DHA (14-20). Previous studies were also limited to
cohorts primarily comprised of individuals from a single ethnic group, and therefore
unable to assess potential interethnic differences in the relation between diet and CVD
risk. In this study, we examined the associations between specific types of dietary fat and

18
risk of MI among 53,120 Latino and African-American men and women in the
Multiethnic Cohort Study.
2.2 Materials and Methods
2.2.1 Population
The Multiethnic Cohort (MEC) is a prospective cohort of 215,251 adult men and
women, 45 to 75 years of age at the time of recruitment, residing in Hawaii or California
(primarily Los Angeles County). The cohort was established between 1993 and 1996 to
examine associations between diet, lifestyle, and genetic factors and incidence of cancers
and other chronic diseases in a representative sample of five major ethnic groups:
African-Americans, Japanese-Americans, Latinos, Native Hawaiians, and Whites. The
details of the study design and baseline characteristics of the cohort have been published
(21). Briefly, potential participants were identified through drivers' license files, voters'
registration files, and Health Care Financing Administration files. Individuals entered the
cohort by returning a 26-page mailed and self-administered questionnaire that asked
about demographic factors (including self-assigned ethnicity), anthropometrics, health-
related behaviors, medical history, use of medications, and for women, reproductive
history and use of exogenous hormones. Participants of mixed ethnicity were assigned to
one category according to the following priority ranking: African-American, Native
Hawaiian, Latino, Japanese-American, White, and other.
For this study, we restricted our population to individuals from the two main ethnic
groups recruited in California: Latinos (n = 44,092) and African-Americans (n = 33,864).
We used previously published MEC exclusion criteria (22) to exclude n = 17,304

19
individuals with missing or implausible questionnaire information on education, body
mass index (BMI), smoking, physical activity, diet, or for women, menopausal status or
use of postmenopausal hormone therapy. We also excluded n = 7,532 individuals with a
prior history of cardiovascular disease defined as either a self-report of angina, MI, or
stroke on the baseline questionnaire or a hospitalization for MI, stroke, or to undergo a
coronary revascularization procedure anytime between January 1, 1991 and the date of
cohort entry. After these exclusions, n = 53,120 individuals were available for analysis.
2.2.2 Ascertainment of diet
Dietary intake was assessed at cohort entry using an 18-page, self-administered
quantitative food frequency questionnaire designed for use in this multiethnic population.
Individuals were asked to report how frequently, on average, they ate each food item
during the past year and their usual portion size. Usual frequency for each food item was
reported by marking one of eight categories, ranging from "never or hardly ever" to "two
or more times a day". For beverages (including alcohol), the highest category was
separated into "two or three times a day" and "four or more times a day", yielding nine
possible responses. Usual portion size for each food item was reported by marking one
of three categories indicated using standard measurements and photographs showing
different portions of representative foods. Average daily intake of nutrients or food
groups were calculated by summing contributions from multiple food items using an
extensive food composition and recipe database developed and maintained at the Cancer
Research Center of Hawaii. The development and validation of the MEC food frequency
questionnaire has been published (21, 23). The questionnaire provided a reasonable

20
measure of most nutrients when compared with multiple 24-hr recalls, particularly after
energy adjustment through the use of nutrient densities (i.e., as g/1000 kcal) or energy-
adjusted absolute nutrient intakes. Average correlation coefficients over all nutrient
densities ranged from 0.57 to 0.74 across the sex- and ethnicity-specific subgroups (23).
2.2.3 Follow-up and ascertainment of MI
The vital status of cohort members is ascertained by annual linkage of the cohort with
state death files and periodic linkage with National Death Index files. The underlying
causes of death are established from the death certificate and coded according to the
Ninth Revision of the International Classification of Disease (ICD-9) through 1998 and
according to the Tenth Revision of the International Classification of Disease (ICD-10)
starting in 1999. In addition, cohort members residing in California are linked each year
to the California Office of Statewide Health Planning and Development's hospitalization
discharge database. The database consists of mandatory records of all in-patient
hospitalizations in acute-care facilities in California except those at federal facilities or at
City of Hope, a comprehensive cancer center. Records include data on patient
demographics, the date of admission, the principal diagnosis and up to 24 other diagnoses
(coded using ICD-9), and the principal procedure and up to 20 other procedures (also
coded using ICD-9). Discharge diagnoses were used as our primary source of
information on non-fatal cardiovascular events and as a secondary source of information
on comorbid conditions including hypertension and diabetes. Linkage of the cohort with
death files and hospitalization records was based solely on Social Security numbers

21
obtained from the self-administered questionnaires and was complete from January 1,
1991 through December 31, 2005 as of the time of this analysis.
Our primary outcome was MI, identified among any of the discharge diagnoses for a
hospitalization or as an underlying cause of death and coded as 410 (acute or subsequent
MI) or 412 (old or silent MI) in ICD-9 and as I21 (acute MI), I22 (subsequent MI), or
I25.2 (old or silent MI) in ICD-10. Our secondary outcome included MI and coronary
revascularization procedures (coded as ICD-9 P36) such as coronary artery bypass graft
surgery and percutaneous coronary intervention. For each individual, censoring occurred
at the time of the first event of interest, death, or the last date of cohort linkage on
December 31, 2005 (i.e., assuming no loss to follow-up due to individuals moving out of
California). Because associations with dietary factors and other potential cardiovascular
risk factors were similar for both outcomes, we present results for our primary outcome
only.
2.2.4 Statistical analysis
We used Cox proportional hazards models stratified on age at recruitment (in 1-yr
intervals) and year of recruitment (in 1-yr intervals) and using days since cohort entry
(defined as completion of the baseline questionnaire) as the time metric. First, we
estimated the sex-specific hazard ratio (HR) with 95% confidence interval (CI) of MI by
the following potential cardiovascular risk factors: ethnicity; education; smoking; alcohol
use; BMI; physical activity; history of hypertension or diabetes; use of aspirin or
antihypertensive medication; intakes of energy (in kcal per day), dietary fiber (in g/1000
kcal per day), and dietary cholesterol (in mg/1000 kcal per day); and for women,

22
menopausal status and use of postmenopausal hormone therapy. Next, we estimated the
HR of MI in the overall study population and in sex- and ethnicity-specific subgroups for
dietary intakes (in g/1000 kcal per day) of fish and specific types of fats: saturated fat,
monounsaturated fat, n-6 polyunsaturated fat, n-3 polyunsaturated fat, and EPA plus
DHA. Dietary factors were categorized using cohort-wide quintiles to allow comparison
across sex- and ethnicity-specific subgroups. In general, cutpoints chosen for non-dietary
risk factors were those used in prior MEC analyses of cardiovascular disease mortality
(24).
To control for potential confounding, we used basic models further stratified on sex
and ethnicity (for associations in the overall study population) and adjusted for smoking;
history of diabetes; and intakes of energy (25), dietary fiber, and dietary cholesterol.
Other cardiovascular risk factors did not substantially alter the associations between
dietary fats and risk of MI and were therefore excluded from the models. Our final
multivariate models took an "isocaloric" replacement approach to examine the effect of
replacing a certain percentage of energy from one source for the same amount from
another source (26). In our carbohydrate and protein replacement models, we estimated
the effects of isocaloric replacement of carbohydrate or protein by each specific type of
fat. We simultaneously adjusted for intakes of saturated fat, monounsaturated fat, and
polyunsaturated fats that are mutually exclusive, while excluding carbohydrate and
protein intakes from the models. Because protein was not a confounder, carbohydrate
replacement models yielded identical results and are not presented. In our saturated fat
replacement models, we estimated the effects of isocaloric replacement of saturated fat

23
by each specific type of unsaturated fat. We simultaneously adjusted for intakes of
carbohydrate, protein, monounsaturated fat, and polyunsaturated fats that are mutually
exclusive, while excluding saturated fat intake from the models. Tests for trend by
dietary factors were calculated with likelihood ratio tests (degrees of freedom, df, = 1) by
assigning the sex- and ethnicity-specific median values to each cohort-wide quintile and
modeling these values as a continuous variable. Tests for interactions of dietary
associations by sex and ethnicity were calculated using likelihood ratio tests (df = 4). We
used a significance level of 0.05 and present two-sided P values for all hypotheses. Data
management used SAS software version 9.1 (SAS Institute Inc., Cary, North Carolina).
Statistical analyses used STATA software version 10.0 (StataCorp, College Station,
Texas).

2.3 Results
Among the 53,120 individuals followed for up to 12.7 years (median: 11.8 years), we
observed 3,229 incident cases of MI; this included 1,090 cases among Latino men (n
=15,877), 760 cases among African-American men (n = 8,529), 564 cases among Latina
women (n = 14,930), and 815 cases among African-American women (n = 13,784). In
both men (Table 2.1) and women (Table 2.2), smoking, history of hypertension, history
of diabetes, and use of aspirin, and use of antihypertensive medication were
independently associated with higher risk of MI whereas alcohol use, physical activity,
and dietary fiber intake were independently associated with lower risk of MI. Intake of
dietary cholesterol was independently associated with higher risk of MI in men only. In

24

25

26

27
women only, lower risk of MI was significantly associated with education, younger age
at natural menopause (relative to natural menopause at age 50-54 years), and current use
of postmenopausal hormone therapy. Compared to Latinos, African-American ethnicity
was independently associated with higher risk of MI in women (HR = 1.21, 95% CI:
1.04, 1.40; P = 0.01) but not men (HR = 0.96, 95% CI: 0.85, 1.08; P = 0.51) after
controlling for other cardiovascular risk factors.
Across all sex- and ethnicity-specific subgroups, median total fat intake ranged from
34.4 to 37.5 g/1000 kcal. Saturated fat, monounsaturated fat, and polyunsaturated fat
accounted for roughly 33%, 40%, and 27% of dietary fat intake, respectively (Table 2.3).
N-6 polyunsaturated fat comprised approximately 89% of all polyunsaturated fat
consumed (median ratio of n-6 to n-3 polyunsaturated fat: 8.8). Higher intakes of
saturated, monounsaturated, and polyunsaturated fats were positively associated with
African-American ethnicity, smoking, BMI, history of diabetes, and use of medications
and inversely associated with use of alcohol and, for women, use of postmenopausal
hormone therapy (data not shown). Physical activity was significantly associated with
lower intakes of saturated and monounsaturated fats and higher intake of polyunsaturated
fat (data not shown). Higher intakes of EPA plus DHA and fish were positively
associated with African-American ethnicity, education, physical activity, history of
diabetes, and use of medications (data not shown).

28

29

30

31

32
Table 2.3. Median daily intake of nutrients and fish according to sex and ethnicity at baseline in 1993-
1996, Multiethnic Cohort, California, United States

Men Women

Latinos
(n = 15,877)
African-
Americans
(n = 8,529)
Latinas
(n = 14,930)
African-
Americans
(n = 13,784)
Energy intake (kcal) 2328 1960 1905 1669
Fat (g/1000 kcal):
Saturated fat 11.1 11.1 10.7 10.5
Monounsaturated fat 13.1 14.0 12.3 13.1
n-6 polyunsaturated fat 7.4 8.1 7.3 8.0
n-3 polyunsaturated fat 0.8 0.9 0.8 0.9
EPA + DHA 0.03 0.05 0.03 0.06
Carbohydrate (g/1000 kcal) 122.4 118.6 131.1 126.1
Protein (g/1000 kcal) 38.3 37.2 38.8 38.6
Fiber (g/1000 kcal) 12.1 10.3 13.6 12.2
Cholesterol (mg/1000 kcal) 110.8 122.1 102.1 112.7
Fish + shellfish (g/1000 kcal) 4.9 7.4 4.3 7.6

DHA = docosahexaenoic acid [22:6n-3]; EPA = eicosapentaenoic acid [20:5n-3].

In analyses stratified on sex, ethnicity, age at recruitment, and year of recruitment,
intakes of saturated fat, monounsaturated fat, n-6 polyunsaturated fat, and n-3
polyunsaturated fat were each significantly associated with higher risk of MI (data not
shown). These associations became attenuated and lost significance after adjustment for
other cardiovascular risk factors (Table 2.4). When we incorporated all four types of fat
in the same model, such that the HRs represent isocaloric substitution of carbohydrate or
protein with each type of fat, and controlled for other confounders, risk of MI was
inversely associated with intake of monounsaturated fat (HR for the highest quintile
relative to the lowest quintile = 0.81, 95% CI: 0.65, 1.02; P
trend
= 0.04) but not
significantly associated with intakes of saturated fat (HR = 1.18, 95% CI: 0.97, 1.44;
P
trend
= 0.11), n-6 polyunsaturated fat (HR = 1.21, 95% CI: 0.98, 1.50; P
trend
= 0.09), or n-
3 polyunsaturated fat (HR = 0.85, 95% CI: 0.71, 1.03; P
trend
= 0.13). In contrast, when
we incorporated monounsaturated fat, n-6 polyunsaturated fat, carbohydrate, and protein

33

34
in the same model, such that the HRs represent isocaloric substitution of saturated fat
with carbohydrate, protein, or each type of unsaturated fat, the association between
greater intake of monounsaturated fat and lower risk of MI was attenuated and not
significant (HRs for each quintile relative to the lowest quintile = 1.03, 1.00, 1.00, and
1.05; 95% CI for the highest quintile relative to the lowest quintile: 0.87, 1.28; P
trend
=
0.71); data shown in Appendix A.1. Similarly, in the saturated fat replacement model,
risk of MI was not significantly associated with intakes of n-6 polyunsaturated fat (HRs =
1.03, 1.14, 1.13, and 1.18; 95% CI: 0.96, 1.44; P
trend
= 0.13) or n-3 polyunsaturated fat
(HRs = 0.94, 0.94, 0.97, and 0.90; 95% CI: 0.75, 1.09; P
trend
= 0.34); data shown in
Appendix A.1.
Intakes of EPA plus DHA and fish were not significantly associated with risk of MI in
basic models or in carbohydrate/protein replacement models adjusted for intakes of
saturated fat, monounsaturated fat, n-6 polyunsaturated fat, and linolenic acid [18:3n-3]
(Table 2.5). In saturated fat replacement models (i.e., adjusted for cardiovascular risk
factors and intakes of monounsaturated fat, n-6 polyunsaturated fat, linolenic acid,
carbohydrate, and protein), greater intake of EPA plus DHA was associated with lower
risk of MI (HRs = 0.98, 1.04, 0.90, and 0.89; 95% CI: 0.77, 1.03; P
trend
= 0.04) but
greater intake of fish was not significantly associated with risk of MI (HRs = 0.91, 0.94,
0.90, and 0.90; 95% CI: 0.80, 1.02; P
trend
= 0.17); data shown in Appendix A.2.
The relations between specific types of dietary fat and risk of MI across sex- and
ethnicity-specific subgroups are presented in Table 2.6 and Table 2.7 using
carbohydrate/protein replacement models. There was a significant ethnicity-diet

35

36

37

38

39

40
interaction (P = 0.04) between Latino ethnicity and intake of saturated fat on risk of MI,
such that greater saturated fat intake was associated with higher risk of MI among Latinos
(HR for the highest quintile relative to the lowest quintile = 1.48, 95% CI: 1.13, 1.94;
P
trend
= 0.006) but not among African-Americans (HR = 0.97, 95% CI: 0.73, 1.29; P
trend
=
0.84); data shown in Appendix A.3. Among Latinos, the effect of saturated fat intake on
risk of MI was consistent across the sexes but the sex-specific estimates and interactions
did not achieve statistical significance, possibly due to the small numbers of cases in each
sex-specific subgroup of Latinos. Likewise, risk of MI was positively associated with
intake of n-6 polyunsaturated fat among Latinos (HR = 1.40, 95% CI: 1.04, 1.88; P
trend
=
0.02) but not among African-Americans (HR = 1.09, 95% CI: 0.80, 1.47; P
trend
= 0.59),
but the ethnicity-diet interaction did not achieve statistical significance (P = 0.18); data
shown in Appendix A.3. Intakes of monounsaturated fat, n-3 polyunsaturated fat, EPA
plus DHA, and fish had no clear effect modification by ethnicity on risk of MI (Appendix
A.3). There also did not appear to be any effect modification of dietary associations by
sex in carbohydrate/protein replacement models (Appendix A.4), or any effect
modification of dietary associations by sex or ethnicity in saturated fat replacement
models (Appendices A.5, A.6, and A.7).

2.4 Discussion
In this multiethnic cohort, we found that greater intake of monounsaturated fat was
associated with lower risk of MI when replacing carbohydrate or protein in the diet, but
not when replacing saturated fat. Greater intake of EPA and DHA (the long-chain n-3

41
polyunsaturated fatty acids found in marine sources), on the other hand, was associated
with lower risk of MI when replacing saturated fat in the diet, but not when replacing
carbohydrate or protein. In addition, when replacing carbohydrate or protein in the diet,
greater intakes of saturated fat and n-6 polyunsaturated fats were associated with higher
risk of MI among Latinos but not among African-Americans.
Our finding of a significant positive association between saturated fat intake and risk
of MI among Latinos is consistent with results from controlled trials on the effects of
dietary fats on serum lipids (2) and early ecological studies (27). Similar positive
associations have also been found in six prospective cohort studies: among Japanese-
American men in the Honolulu Heart Study (28), among men in the Ireland-Boston Diet-
Heart Study (10), among younger men in the Framingham Study (29), among male Israeli
civil servants (30), among men in the Health Professionals Follow-up Study (11), and
among younger American Indian men and women in the Strong Heart Study (12). In
other studies, such as the Nurses' Health Study (13) and the Alpha-Tocopherol, Beta-
Carotene Cancer Prevention Study (14), the positive association between saturated fat
intake and risk of CVD became attenuated and was not significant after controlling for
intakes of fiber and other types of fat. In our study, adjustment for dietary factors and
other cardiovascular risk factors attenuated the relation between saturated fat intake and
risk of MI among African-Americans but not among Latinos.
Our results with regard to the protective effect of monounsaturated fat are consistent
with evidence from experimental studies suggesting that monounsaturated fat acts like
polyunsaturated fat to decrease serum levels of low-density lipoprotein cholesterol and

42
triglycerides (2), increase serum levels of high-density lipoprotein cholesterol (2), and
counteract injury-induced changes in endothelial cell expression, e.g., by enhancing
secretion of nitric oxide and inhibiting secretion of various inflammatory factors (4-5).
Greater intake of monounsaturated fat may also reduce coagulation by decreasing plasma
levels of von Willebrand factor, plasminogen activator inhibitor type 1, α2-antiplasmin,
and coagulation factors VII and XIIa (4). Despite these myriad potential mechanisms for
a protective effect of monounsaturated fat intake, previous epidemiologic studies have in
general failed to find a significant association between monounsaturated fat intake and
risk of cardiovascular disease (14-15, 28-30). A high ratio of monounsaturated to
saturated fat was one component of the Mediterranean diet that was associated with
reduced all-cause and coronary heart disease mortality in the Healthy Ageing: a
Longitudinal study in Europe (HALE) cohort (31); however, the investigators did not
distinguish between the effects of the ratio and other Mediterranean diet components. In
the Strong Heart Study (12), mortality from coronary heart disease was inversely
associated with monounsaturated fat intake among older American Indians (aged 60-79
years) but positively associated with monounsaturated fat intake among younger
American Indians (aged 47-59 years). The authors suggested that their findings may
have been confounded by intake of saturated fat, which was highly correlated with intake
of monounsaturated fat and therefore not included in the same models (12). In our study,
adjusting for saturated fat intake may have created a spurious inverse association between
monounsaturated fat intake and risk of MI; this bias, if present, would also explain the

43
observed differences in the effects of monounsaturated fat intake in the
carbohydrate/protein and saturated fat replacement models.
In contrast, our finding of a protective effect of EPA and DHA from fish and other
marine sources on risk of MI is consistent with results from numerous epidemiologic
studies (15-19). Like other polyunsaturated fatty acids, EPA and DHA have beneficial
effects on serum lipids (2) and may have a particularly strong triglyceride-lowering effect
(32-34). There is a growing body of evidence that they may also lower cardiovascular
risk by increasing vasodilation through enhanced production of nitric oxide and
endothelium-derived vascular relaxing factor (6, 7), reducing inflammation by modifying
gene expression in endothelial cells and circulating leukocytes (4-7), inhibiting smooth
muscle proliferation (4, 6, 7), reducing coagulation by decreasing plasma levels of
fibrinogen and coagulation factors VII and X (4), and protecting against potentially-fatal
arrhythmias including ventricular fibrillation and sudden cardiac death (6, 9, 35).
Dietary intake of EPA may also directly impact inflammation and thrombosis via
competition with arachidonic acid [20:4n-6], its n-6 homologue, as substrates for 5-
lipoxygenase and other enzymes that catalyze the release of the fatty acids from cell
membranes (4) or conversion of the fatty acids into a variety of metabolites including
prostaglandins, leukotrienes (LTs), and thromboxanes (TXs) (6-8, 36). Because the
EPA-derived LTs and TXA
3
are associated with less inflammatory and thrombotic
activity than the arachidonic acid-derived LTs and TXA
2
(7, 36), increasing dietary
intake of EPA and DHA and decreasing intake of n-6 polyunsaturated fat may lead to a
shift towards an antiinflammatory and antithrombotic state. In a cross-sectional analysis

44
of men from the Health Professionals Follow-up Study and women from the Nurses'
Health Study, the highest concentrations of inflammatory markers were observed in
individuals with a high intake of linoleic acid [18:2n-6] and low intake of EPA and DHA
(37). Moreover, in two recent studies, individuals with polymorphisms in the promoter
of the gene encoding 5-lipoxygenase had increased atherosclerosis and an increased risk
of MI when dietary intake of arachidonic acid was high and dietary intake of EPA and
DHA was low (38-39). Our finding of a positive association between n-6
polyunsaturated fat intake and risk of MI among Latinos is consistent with this
mechanism. Interestingly, however, we did not find any association between n-6
polyunsaturated fat intake and risk of MI among African-Americans, an ethnic group
with a higher prevalence of 5-lipoxygenase promoter polymorphisms. It should also be
noted that the American Heart Association recently concluded that there is insufficient
evidence for a role of the balance of n-6 and n-3 polyunsaturated fats in the development
of cardiovascular diseases, and that greater intake of both classes of polyunsaturated fat
continues to be recommended (40).
In summary, our results provide evidence that greater intake of monounsaturated fat
and marine long-chain n-3 polyunsaturated fatty acids reduces the risk of MI in a
multiethnic cohort of men and women. To our knowledge, this is the first study to
directly address differences in dietary associations with cardiovascular risk between
Latinos and African-Americans.

45
2.5 Chapter 2 References
1. Lichtenstein AH, Appel LJ, Brands, M, et al. Diet and lifestyle recommendations
revision 2006: a scientific statement from the American Heart Association Nutrition
Committee. Circulation 2006; 114: 82-96.

2. Mensink RP, Zock PL, Kester ADM, et al. Effects of dietary fatty acids and
carbohydrates on the ratio of serum total to HDL cholesterol and on serum lipids and
apolipoproteins: a meta-analysis of 60 controlled trials. Am J Clin Nutr 2003; 77: 1146-
1155.

3. Nestel PJ. Fish oil and cardiovascular disease: lipids and arterial function. Am J Clin
Nutr 2000; 71(suppl): 228S-231S.

4. Moreno JJ, Mitjavila MT. The degree of unsaturation of dietary fatty acids and the
development of atherosclerosis (review). J Nutr Biochem 2003; 14: 182-195.

5. De Caterina R, Liao JK, Libby P. Fatty acid modulation of endothelial activation. Am
J Clin Nutr 2000; 71(suppl): 213S-223S.

6. Jung UJ, Torrejon C, Tighe AP, et al. n-3 fatty acids and cardiovascular disease:
mechanisms underlying beneficial effects. Am J Clin Nutr 2008; 87(suppl): 2003S-
2009S.

7. Mori TA, Beilin LJ. Omega-3 fatty acids and inflammation. Curr Atheroscler Rep
2004; 6: 461-467.

8. Vanden Heuvel JP. Diet, fatty acids, and regulation of genes important for heart
disease. Curr Atheroscler Rep 2004; 6: 432-440.

9. Kang JX, Leaf A. Prevention of fatal cardiac arrhythmias by polyunsaturated fatty
acids. Am J Clin Nutr 2000; 71(suppl): 202S-207S.

10. Kushi LH, Lew RA, Stare FJ, et al. Diet and 20-year mortality from coronary heart
disease: the Ireland-Boston Diet-Heart Study. N Engl J Med 1985; 312(13): 811-818.

11. Ascherio A, Rimm EB, Giovannucci EL, et al. Dietary fat and risk of coronary heart
disease in men: cohort follow up study in the United States. BMJ 1996; 313: 84-90.

12. Xu J, Eilat-Adar S, Loria C, et al. Dietary fat intake and risk of coronary heart
disease: the Strong Heart Study. Am J Clin Nutr 2006; 84: 894-902.

46
13. Oh K, Hu FB, Manson JE, et al. Dietary fat intake and risk of coronary heart disease
in women: 20 years of follow-up of the Nurses' Health Study. Am J Epidemiol 2005;
161(7): 672-679.

14. Pietinen P, Ascherio A, Korhonen P, et al. Intake of fatty acids and risk of coronary
heart disease in a cohort of Finnish men: the Alpha-Tocopherol, Beta-Carotene Cancer
Prevention Study. Am J Epidemiol 1997; 145(10): 876-887.

15. Kromhout D, Bosschieter EB, de Lezenne Coulander C. The inverse relation
between fish consumption and 20-year mortality from coronary heart disease. N Engl J
Med 1985; 312(19): 1205-1209.

16. Daviglus ML, Stamler J, Orencia AJ, et al. Fish consumption and the 30-year risk of
fatal myocardial infarction. N Engl J Med 1997; 336(15): 1046-1053.

17. Albert CM, Hennekens CH, O'Donnell CJ, et al. Fish consumption and risk of
sudden cardiac death. JAMA 1998; 279(1): 23-28.

18. Yuan JM, Ross RK, Gao YT, et al. Fish and shellfish consumption in relation to
death from myocardial infarction among men in Shanghai, China. Am J Epidemiol 2001;
154(9): 809-816.

19. Hu FB, Bronner L, Willett WC, et al. Fish and omega-3 fatty acid intake and risk of
coronary heart disease in women. JAMA 2002; 287(14): 1815-1821.

20. Ascherio A, Rimm EB, Stampfer MJ, et al. Dietary intake of marine n-3 fatty acids,
fish intake, and the risk of coronary disease among men. N Engl J Med 1995; 332(15):
977-982.

21. Kolonel LN, Henderson BE, Hankin JH, et al. A multiethnic cohort in Hawaii and
Los Angeles: baseline characteristics. Am J Epi 2000; 151(4): 346-357.

22. Nöthlings U, Wilkens LR, Murphy SP, et al. Meat and fat intake as risk factors for
pancreatic cancer: the Multiethnic Cohort Study. J Natl Cancer Inst 2005; 97(19): 1458-
1465.

23. Stram DO, Hankin JH, Wilkens LR, et al. Calibration of the dietary questionnaire
for a multiethnic cohort in Hawaii and Los Angeles. Am J Epidemiol 2000; 151(4): 358-
370.

24. Henderson SO, Haiman CA, Wilkens LR, et al. Established risk factors account for
most of the racial differences in cardiovascular disease mortality. PLoS ONE 2007; 2(4):
e377.

47
25. Willett W, Stampfer M. Implications of total energy intake for epidemiologic
analysis. In: Willett W. Nutritional Epidemiology, 2nd ed. New York, NY: Oxford
University Press, 1998: 273-301.

26. Willett W. Issues in analysis and presentation of dietary data. In: Willett W.
Nutritional Epidemiology, 2nd ed. New York, NY: Oxford University Press, 1998: 321-
346.

27. Willett W. Diet and coronary heart disease. In: Willett W. Nutritional
Epidemiology. 2nd ed. New York, NY: Oxford University Press; 1998: 414-466.

28. McGee DL, Reed DM, Yano K, et al. Ten-year incidence of coronary heart disease
in the Honolulu Heart Program: relationship to nutrient intake. Am J Epidemiol 1984;
119: 667-676.

29. Posner BM, Cobb JL, Belanger AJ, et al. Dietary lipid predictors of coronary heart
disease in men: the Framingham Study. Arch Intern Med 1991; 151: 1181-1187.

30. Goldbourt U, Yaari S, Medalie JH. Factors predictive of long-term coronary heart
disease mortality among 10,059 male Israeli civil servants and municipal employees: a
23-year mortality follow-up in the Israeli ischemic heart disease study. Cardiology 1993;
82: 100-121.

31. Knoops KT, De Groot LC, Kromhout D, et al. Mediterranean diet, lifestyle factors,
and 10-year mortality in elderly European men and women: the HALE project. JAMA
2004; 292(12): 1433-1439.

32. Kestin M, Clifton P, Belling B, et al. n-3 fatty acids of marine origin lower systolic
blood pressure and triglycerides but raise LDL cholesterol compared with n-3 and n-6
fatty acids from plants. Am J Clin Nutr 1990; 51: 1028-1034.

33. Griffin MD, Sanders TA, Davies IG, et al. Effects of altering the ratio of dietary n-6
to n-3 fatty acids on insulin sensitivity, lipoprotein size, and postprandial lipemia in men
and postmenopausal women aged 45-70 y: the OPTILIP Study. Am J Clin Nutr 2006; 84:
1290-1298.

34. Finnegan YE, Minihane AM, Leigh-Firbank EC, et al. Plant- and marine-derived n-
3 polyunsaturated fatty acids have differential effects on fasting and postprandial blood
lipid concentrations and on the susceptibility of LDL to oxidative modification in
moderately hyperlipidemic subjects. Am J Clin Nutr 2003; 77: 783-795.

35. Mozaffarian D. Fish and n-3 fatty acids for the prevention of fatal coronary heart
disease and sudden cardiac death. Am J Clin Nutr 2008; 87(suppl): 1991S-1996S.

48
36. James MJ, Gibson RA, Cleland LG. Dietary polyunsaturated fatty acids and
inflammatory mediator production. Am J Clin Nutr 2000; 71(suppl): 343S-348S.

37. Pischon T, Hankinson SE, Hotamisligil GS, et al. Habitual dietary intake of n-3 and
n-6 fatty acids in relation to inflammatory markers among US men and women.
Circulation 2003; 108: 155-160.

38. Dwyer JH, Allayee H, Dwyer KM, et al. Arachidonate 5-lipoxygenase promoter
genotype, dietary arachidonic acid, and atherosclerosis. N Engl J Med 2004; 350(1): 29-
37.

39. Allayee H, Baylin A, Hartiala J, et al. Nutrigenetic association of the 5-lipoxygenase
gene with myocardial infarction. Am J Clin Nutr 2008; 88: 934-940.

40. Harris WS, Mozaffarian D, Rimm E, et al. Omega-6 fatty acids and risk for
cardiovascular disease: a science advisory from the American Heart Association
Nutrition Subcommittee of the Council on Nutrition, Physical Activity, and Metabolism;
Council on Cardiovascular Nursing; and Council on Epidemiology and Prevention.
Circulation 2009; 119. doi: 10.1161/circulationaha.108.191627

49
Chapter 3:
Cross-sectional Associations Between 5-Lipoxygenase Promoter Polymorphisms,
Dietary Fats, and Carotid Artery Atherosclerosis

3.1 Introduction
Atherosclerosis is characterized by narrowing of the lumens of large- and medium-
sized arteries due to the infiltration of low-density lipoprotein (LDL) cholesterol, chronic
inflammatory cells, and smooth muscle cells into the intimal layer of arterial walls (1). It
is the underlying cause of myocardial infarction and stroke and contributes to more than
half of all deaths in the United States (2). Atherosclerosis has myriad environmental and
genetic factors that have been identified as contributing to its development. Although the
majority of genes that have been implicated in atherogenesis relate to lipid metabolism, a
number of "nonconventional" genetic risk factors have been reported that relate to
inflammation (3). One such gene codes for 5-lipoxygenase (5-LO), an enzyme expressed
in numerous inflammatory cells including monocytes and macrophages (4) and that is the
rate-limiting enzyme in the biosynthesis of leukotrienes from long-chain polyunsaturated
fatty acids (4, 5). In differentiated myeloid cells and other leukocytes, 5-LO catalyzes the
conversion of arachidonic acid [20:4n-6], a long-chain n-6 polyunsaturated fatty acid,
into leukotriene A
4
(LTA
4
) with the assistance of 5-LO activating protein (FLAP) (4).
LTA
4
is further metabolized by LTA
4
hydrolase to produce leukotriene B
4
(LTB
4
) or by
LTA
4
synthase to produce the cysteinyl leukotrienes LTC
4
, LTD
4
, and LTE
4
(4).
Eicosapentaenoic acid (EPA) [20:5n-3], the n-3 homologue of arachidonic acid, functions

50
as an alternative substrate in the 5-LO pathway leading to the formation of leukotriene B
5

(LTB
5
) and the cysteinyl leukotrienes LTC
5
, LTD
5
, and LTE
5
(4). Arachidonic acid-
derived leukotrienes are proinflammatory lipid mediators that have been associated with
a variety of acute and chronic inflammatory diseases including asthma (5). In contrast,
LTB
5
and the other EPA-derived leukotrienes are associated with less inflammatory
activity (5). The competition between arachidonic acid and EPA as substrates for 5-LO
suggests that there may be a direct link between nutrition and inflammation via
consumption of dietary fats.
5-LO and its associated enzymes have been localized to macrophages in
atherosclerotic lesions in mice and humans (6-8) and there is strong evidence that the 5-
LO pathway may affect atherosclerosis through its effects on monocyte adhesion,
chemotaxis, and migration; monocyte differentiation and proliferation; smooth muscle
cell proliferation; regulation of vascular tone; and plaque erosion (9). In mouse models,
5-LO deficiency appeared to protect against the formation of aortic lesions (6, 10). In
humans, polymorphisms in the genes encoding FLAP and LTA
4
hydrolase have been
linked to an increased risk of myocardial infarction and stroke (11-14) and a
polymorphism in the gene encoding LTC
4
synthase has been linked with elevated carotid
artery atherosclerosis and coronary artery calcium (15).
The human 5-LO gene spans over 82 kilobases on chromosome 10q11 and contains 14
exons and 13 introns (16). Although the human and mouse 5-LO genes share a high
degree of homology, the human core promoter region is unique in that it contains tandem
hexanucleotide repeats (5'GGGCGG3') that function as overlapping binding sites for the

51
transcription factors Sp1 and early growth-response protein 1 (16, 17). In previously-
studied populations, the number of tandem repeats ranged from three to eight and the
most common promoter allele contained five tandem repeats (16-22). Although early in
vitro studies were inconsistent with regard to the relative 5-LO expression of variant
alleles (17, 18), allele-specific expression of the shorter ("3" and "4" repeats) alleles was
recently shown to be twice that of the "5" allele (22) and asthmatics with the shorter
alleles exhibited an enhanced response to 5-LO inhibitors (19). The hypothesis that 5-LO
promoter polymorphisms may affect risk of atherosclerosis and its complications was
previously tested in three human cohorts yielding inconsistent results (20-22).
Interestingly, the positive associations between 5-LO promoter polymorphisms and
atherosclerosis reported in two of these studies appeared to be modified by dietary
consumption of long-chain polyunsaturated fatty acids (20, 22). Our aim was to replicate
the findings of Dwyer et al. (20) on carotid artery atherosclerosis in a larger cohort of
adults and to investigate possible interactions of 5-LO promoter polymorphisms with
cardiovascular risk factors and consumption of dietary fats.

3.2 Methods
3.2.1 Study populations
Our study population consisted of 1,852 adults, 40 to 92 years old, who participated in
one of four randomized controlled atherosclerosis trials conducted at the Atherosclerosis
Research Unit at the University of Southern California: the Vitamin E Atherosclerosis
Prevention Study (VEAPS, n = 353) (23), the B-Vitamin Atherosclerosis Intervention

52
Trial (BVAIT, n = 506) (24), the Women's Isoflavone Soy Health Trial (WISH, n = 350,
in progress), and the Early versus Late Intervention Trial with Estradiol (ELITE, n = 643,
in progress). The design and eligibility criteria for each trial are summarized in
Appendix B.1. Subjects were non-diabetic and free of symptomatic cardiovascular
disease at study entry. Two trials (WISH and ELITE) were restricted to postmenopausal
women (with serum estradiol <20 pg/mL). All subjects provided written informed
consent and the study protocol of each trial was approved by the University of Southern
California Institutional Review Board.
3.2.2 Clinical and laboratory measurements
Baseline data on subjects' characteristics and clinic and laboratory measurements were
collected using the same methods in all four trials. Information on subjects' age,
ethnicity, years of education, current and past smoking habits, and past medical history
were assessed using structured questionnaires. Vital signs (including systolic and
diastolic blood pressure), anthropometrics (including height, weight, and waist and hip
circumferences), and use of non-study medications (including aspirin, antiinflammatory
medication, antihypertensive medication, and lipid-lowering medication) and
supplements were recorded by clinic staff. Body mass index (BMI) was calculated by
dividing weight (in kg) by the square of height (in m) and waist-to-hip ratio was
calculated by dividing waist circumference (in cm) by hip circumference (in cm).
Minimum 8-hr fasting blood samples were drawn to measure plasma concentrations of
glucose, total cholesterol, high-density lipoprotein (HDL) cholesterol, and triglycerides

53
using standard methods. Plasma LDL cholesterol was calculated using the Friedewald
formula (25) for subjects with a fasting triglyceride concentration less than 400 mg/dL.
3.2.3 Carotid intima-media thickness
Rate of change in carotid intima-media thickness (CIMT) was the primary endpoint in
each trial. In this study, we focused on CIMT at study entry (baseline) as the primary
measure of atherosclerosis. CIMT was measured using high-resolution B-mode
ultrasonograms of the right common carotid artery obtained with a 7.5-MHz linear-array
transducer attached to an ultrasound system. Subjects were placed in a supine position
with the head rotated to the left using a 45-degree block. First, ultrasonographers located
a transverse image with the jugular vein stacked above the common carotid artery. Then
they rotated the transducer by 90 degrees around the central line of the transverse image
to obtain a longitudinal view while maintaining the stacked position of the blood vessels.
Automated computerized edge detection was used to measure the distance between the
intima-lumen and media-adventitia interfaces in 70-100 segments within a 1-cm length of
the common carotid artery immediately distal to the carotid artery bulb. CIMT was
defined as the average of the measurements. These methods have been shown to yield
high reliability for CIMT measurements (27, 28). Ultrasonographic measurement of
CIMT has been directly validated by microscopy (29) and indirectly validated by the
close correlations reported between CIMT and many risk factors for atherosclerosis (29-
33). CIMT is strongly associated with both angiographically-documented coronary
atherosclerosis (34-39) and prevalence and incidence of cardiovascular events including
myocardial infarction and stroke (40-47).

54
3.2.4 Dietary assessment
Baseline diet was assessed using 3-day dietary booklets in which each subject would
record his or her diet during the week before an initial clinic visit for two weekdays and
either Saturday or Sunday. Food intakes from the booklets were analyzed using the
computerized Nutrient Analysis System (Nutrition Scientific, Inc., Pasadena, CA), which
converted food intakes into average daily nutrient intakes using the University of
Minnesota Nutrition Coordinating Center nutrient database. We then converted absolute
nutrient intakes (in g or mg) into nutrient densities (in g or mg per 1000 kcal) by dividing
each nutrient intake by the participant's average daily energy intake (in kcal). Diet data
was not available for 65 subjects (n = 2 from VEAPS, n = 5 from BVAIT, n = 4 from
WISH, and n = 54 from ELITE). Subjects with missing diet were included in the main
analysis if they had known genotypes but excluded from the analyses of gene-diet
interactions.
3.2.5 Genotyping
Genotyping of the number of tandem repeats in the 5-LO promoter region was
performed using previously described methods (20) in isolated DNA from 1,839 subjects
(n = 349 from VEAPS, n = 499 from BVAIT, n = 350 from WISH, and n = 641 from
ELITE), such that sampling was nearly complete (99.3%). After stratification by trial,
genotyped subjects were similar to subjects who were not genotyped (n = 13) with
respect to baseline characteristics (Appendix B.2). Ninety duplicate specimens were
included in the genotyping and yielded identical genotypes, indicating a high reliability.
The genotype of 12 subjects (0.7%) could not be determined, leaving 1,827 subjects with

55
known genotypes (n = 343 from VEAPS, n = 497 from BVAIT, n = 349 from WISH, and
n = 638 from ELITE). Included subjects were generally similar to subjects with missing
genotypes (Appendix B.3).
3.2.6 Statistical analysis
Our main analysis compared mean CIMT among the 5-LO promoter genotypes. We
used multivariate linear regression to adjust for trial, ethnicity, and age (centered at the
sample mean of 59.9 years of age). Additional cardiovascular risk factors (e.g., sex,
education, smoking, body mass index, waist circumference, blood pressure, plasma
lipids, and use of non-study medications) were not independently associated with 5-LO
promoter genotype and did not substantially affect results when included in the regression
models. In a secondary analysis, we compared mean concentrations of plasma lipids
among the 5-LO promoter genotypes in non-users and users of lipid-lowering medication.
We then tested for effect modification of the gene-CIMT association by consumption of
dietary fats, use of lipid-lowering medication, and plasma lipid concentrations. Dietary
fat intakes and plasma lipids concentrations were categorized as above or below the
sample medians. Likelihood ratio tests were used to test for interactions after adjustment
for confounders including trial, ethnicity, age, college education (yes or no), ever
smoking (yes or no), systolic blood pressure (in tertiles), waist circumference (in tertiles),
and total energy intake (in tertiles). We used a significance level of 0.05 and present two-
sided P values for all hypotheses. Data management used SAS software version 9.1
(SAS Institute Inc., Cary, North Carolina). Statistical analyses used STATA software
version 10.0 (StataCorp, College Station, Texas).

56
3.3 Results
3.3.1 Characteristics of the study populations
Table 3.1 presents the baseline characteristics of the subjects used for this study,
stratified by trial. The distribution of sex, ethnicity, age, blood pressure, abdominal
Table 3.1. Baseline characteristics of subjects with known 5-lipoxygenase promoter genotypes from four
randomized controlled atherosclerosis trials

VEAPS
(n = 343)
BVAIT
(n = 497)
WISH
(n = 349)
ELITE
(n = 638)
Female sex (%) *** 51.3 39.0 100 100
Ethnicity ***
Non-hispanic white (%) 74.3 65.8 63.9 68.8
Non-hispanic black (%) 10.2 14.9 6.3 8.9
Hispanic (%) 10.8 10.7 16.3 13.6
Asian or Pacific Islander (%) 4.7 8.7 13.5 8.6

No college education (%) 12.8 12.1 10.3 9.4
Current or past smoking (%) 36.4 39.2 40.7 40.3
Overweight (%) *** 69.7 74.7 57.6 62.9
Use of aspirin (%) *** 12.2 24.1 23.2 23.2
Use of NSAID (%) 7.0 14.9 17.2 19.8
Use of non-NSAID antiinflammatory
medication (%) ***

1.5

2.0

1.2

1.1
Use of antihypertensive medication (%) *** 19.0 30.2 22.4 23.5
Use of lipid-lowering medication (%) *** 7.0 19.5 22.1 19.9

Age (years) *** 56.1 ± 0.5 61.6 ± 0.4 61.4 ± 0.4 59.9 ± 0.3
Systolic blood pressure (mmHg) *** 128.8 ± 1.0 129.8 ± 0.8 122.1 ± 0.8 116.5 ± 0.5
Diastolic blood pressure (mmHg) *** 77.0 ± 0.6 80.8 ± 0.5 77.0 ± 0.5 74.5 ± 0.3
Body mass index (kg/m
2
) *** 27.8 ± 0.3 28.2 ± 0.2 26.7 ± 0.3 27.1 ± 0.2
Waist circumference (cm) *** 91.5 ± 0.7 90.8 ± 0.6 82.6 ± 0.6 84.4 ± 0.5
Waist-hip ratio *** 0.86 ± 0.004 0.86 ± 0.004 0.81 ±0.003 0.81 ± 0.002
Total cholesterol (mg/dL) *** 238.0 ± 1.4 220.8 ± 1.7 220.5 ± 1.8 222.1 ± 1.4
LDL cholesterol *** 155.4 ± 1.2 138.0 ± 1.6 136.2 ± 1.6 134.4 ± 1.3
HDL cholesterol *** 55.8 ± 0.7 56.9 ± 0.7 62.2 ± 0.9 66.2 ± 0.7
Triglycerides (mg/dL) *** 133.7 ± 3.0 129.5 ± 3.0 110.8 ± 3.3 107.2 ± 2.4
Carotid intima-media thickness (um) *** 754.7 ± 7.1 754.1 ± 6.7 810.2 ± 5.4 770.8 ± 4.2

BVAIT = B-Vitamin Atherosclerosis Intervention Trial; ELITE = Early versus Late Intervention Trial with
Estradiol; HDL = high-density lipoprotein; LDL = low-density lipoprotein; NSAID = nonsteroidal
antiinflammatory drug; VEAPS = Vitamin E Atherosclerosis Prevention Study; WISH = Women's
Isoflavone Soy Health Trial.
Data given as percentages or means ± standard errors.
Significant heterogeneity: * P <0.05, ** P <0.01, *** P <0.001 by Pearson chi-square test or analysis of
variance.

57
adiposity, plasma lipids, use of non-study medications, and CIMT differed across the
trials (P <0.001). Baseline dietary consumption of fats, carbohydrate, and protein also
varied slightly across the trials (Table 3.2); P <0.001.

Table 3.2. Baseline mean (± standard error) daily diets of subjects with known 5-lipoxygenase promoter
genotypes from four randomized controlled atherosclerosis trials

VEAPS
(n = 341)
BVAIT
(n = 492)
WISH
(n = 345)
ELITE
(n = 584)
Energy intake (kcal) *** 1966 ± 38 2011 ± 33 1775 ± 29 1667 ± 24
Fat (g/1000 kcal) *** 35.0 ± 0.4 37.0 ± 0.4 38.0 ± 0.5 38.5 ± 0.4
Saturated fat ** 11.3 ± 0.2 11.9 ± 0.2 11.2 ± 0.2 11.7 ± 0.1
Monounsaturated fat *** 13.3 ± 0.2 14.1 ± 0.2 15.2 ± 0.3 15.2 ± 0.2
n-6 polyunsaturated fat *** 6.7 ± 0.1 6.9 ± 0.1 7.4 ± 0.1 7.4 ± 0.1
Linoleic acid [18:2n-6] *** 6.6 ± 0.1 6.9 ± 0.1 7.3 ± 0.1 7.3 ± 0.1
Ara {20:4n-6] *** 0.08 + 0.003 0.08 ± 0.002 0.09 ±0.003 0.09 ± 0.002
n-3 polyunsaturated fat * 0.8 ± 0.02 0.9 ± 0.02 0.9 ± 0.02 0.9 ± 0.02
Linolenic acid [18:3n-3] 0.7 ± 0.02 0.7 ± 0.02 0.7 ± 0.02 0.7 ± 0.02
EPA [20:5n-3] *** 0.03 ± 0.003 0.05 ± 0.003 0.06 ±0.004 0.06 ± 0.004
DHA [22:6n-3] *** 0.06 ± 0.005 0.09 ± 0.006 0.10 ±0.007 0.10 ± 0.006
Ratio of n-6 to n-3 polyunsaturated fats 10.0 ± 0.4 10.2 ± 0.4 10.5 ± 0.4 10.1 ± 0.3
Ratio of Ara to EPA + DHA * 4.8 ± 0.4 3.7 ± 0.2 3.6 ± 0.4 3.5 ± 0.3
Carbohydrate (g/1000 kcal) *** 127.7 ± 1.3 124.2 ± 1.0 120.6 ± 1.3 117.3 ± 0.9
Starch *** 57.7 ± 0.8 53.0 ± 0.7 47.8 ± 1.0 46.5 ± 0.6
Soluble fiber *** 3.4 ± 0.1 3.5 ± 0.1 4.0 ± 0.1 4.0 ± 0.1
Insoluble fiber *** 5.9 ± 0.1 6.3 ± 0.1 7.3 ± 0.2 7.2 ± 0.1
Protein (g/1000 kcal) *** 40.7 ± 0.5 40.4 ± 0.4 44.6 ± 0.6 45.9 ± 0.4
Animal protein *** 27.0 ± 0.6 26.4 ± 0.5 29.3 ± 0.6 30.9 ± 0.5
Vegetable protein *** 13.2 ± 0.2 13.6 ± 0.2 14.7 ± 0.2 14.6 ± 0.2
Alcohol (g/1000 kcal) 4.0 ± 0.3 4.1 ± 0.3 3.1 ± 0.3 3.7 ± 0.2
Cholesterol (mg/1000 kcal) * 119 ± 3 125 ± 3 125 ± 4 133 ± 3
Sodium (mg/1000 kcal) 1419 ± 21 1470 ± 18 1453 ± 23 1449 ± 19
Caffeine (mg/1000 kcal) *** 90 ± 5 79 ± 4 91 ± 5 104 ± 4

Ara = arachidonic acid; BVAIT = B-Vitamin Atherosclerosis Intervention Trial; DHA = docosahexaenoic
acid; ELITE = Early versus Late Intervention Trial with Estradiol; EPA = eicosapentaenoic acid; VEAPS =
Vitamin E Atherosclerosis Prevention Study; WISH = Women's Isoflavone Soy Health Trial.
Significant heterogeneity: * P <0.05, ** P <0.01, *** P <0.001 by analysis of variance.

58
3.3.2 Associations of 5-LO promoter polymorphisms with carotid artery
atherosclerosis
The number of tandem repeats in the 5-LO promoter region ranged from two to nine
in our study, with the "5" allele being the most common (76.8%). The sample
frequencies of the other alleles were 15.5% for the "4" allele, 3.4% each for the "3" and
"6" alleles, 0.5% for the "7" allele, and 0.1% each for the "2", "8", and "9" alleles. The
distribution of alleles varied substantially by ethnicity (Appendices B.4), with the "3"
allele being more frequent among non-Hispanic blacks (27.7%) and the "4" and "6"
alleles being more frequent among Asians and Pacific Islanders (24.8% and 16.5%,
respectively). Prevalence of use of antihypertensive medication and mean body mass
index and waist circumference varied by 5-LO genotype (Table 3.3). However, these
differences were attenuated and lost significance after adjustment for ethnicity (data not
shown). Compared to individuals homozygous for the "5" allele, mean (±SE) plasma
LDL cholesterol was elevated by 8±4 mg/dL in individuals homozygous for shorter
alleles ("33", "34, or "44"; P = 0.04) and reduced by 4±2 mg/dL in individuals carrying
one shorter allele (P = 0.03), after adjustment for ethnicity (Appendix B.5). Sex, age,
education, smoking habits, blood pressure, plasma HDL cholesterol or triglycerides, and
use of antiinflammatory or lipid-lowering medication were not significantly associated
with 5-LO genotype.
Means of CIMT according to 5-LO promoter genotype are presented in Table 3.4.
Compared to individuals homozygous for the "5" allele, mean (±SE) CIMT was elevated
by 137±31 µm in individuals homozygous for longer alleles ("66" or "67"; P = 0.004)

59
Table 3.3. Baseline characteristics of subjects from four randomized controlled atherosclerosis trials,
according to 5-lipoxygenase promoter genotype

55
(n = 1,110)

33, 34, 44
(n = 81)
25, 35, 45,
36, 37, 46,
47, 48, 49
(n = 533)
56, 57, 58,
59,
66, 67
(n = 103)
Female sex (%) 74.1 70.4 75.2 73.8
Ethnicity ***
Non-hispanic white (%) 76.9 27.2 60.6 44.7
Non-hispanic black (%) 5.3 50.6 16.1 1.9
Hispanic (%) 13.2 9.9 11.8 15.5
Asian or Pacific Islander (%) 4.6 12.4 11.4 37.9

No college education (%) 11.2 8.6 10.3 13.6
Current or past smoking (%) 40.6 37.0 38.7 31.1
Overweight (%) * 67.2 76.5 65.2 56.3
Use of aspirin (%) 21.5 25.9 20.3 22.3
Use of NSAID (%) 16.2 14.8 15.2 10.7
Use of non-NSAID antiinflammatory
medication (%)

1.8

1.2

0.8

1.0
Use of antihypertensive medication (%) * 22.2 30.9 27.8 23.3
Use of lipid-lowering medication (%) 18.6 16.1 17.1 14.6

Age (years) 60.2 ± 0.3 58.8 ± 1.0 59.6 ± 0.4 59.6 ± 0.9
Systolic blood pressure (mmHg) 123.4 ± 0.5 126.8 ± 2.2 123.3 ± 0.7 124.0 ± 1.5
Diastolic blood pressure (mmHg) 77.1 ± 0.3 78.5 ± 1.0 77.2 ± 0.4 76.9 ± 0.8
Body mass index (kg/m
2
) ** 27.4 ± 0.2 28.5 ± 0.6 27.7 ± 0.2 26.1 ± 0.5
Waist circumference (cm) ** 87.1 ± 0.4 90.3 ± 1.5 87.4 ± 0.6 83.3 ± 1.2
Waist-hip ratio 0.83 ± 0.002 0.85 ±0.009 0.83 ±0.003 0.82 ±0.007
Total cholesterol (mg/dL) ** 225.0 ± 1.1 235.1 ± 5.0 220.7 ± 1.5 228.5 ± 2.8
LDL cholesterol ** 140.2 ± 1.0 149.7 ± 4.5 136.5 ± 1.4 141.9 ± 2.6
HDL cholesterol 61.2 ± 0.5 60.8 ± 2.1 60.1 ± 0.7 62.3 ± 1.6
Triglycerides (mg/dL) 117.8 ± 1.8 122.8 ± 8.6 120.3 ± 2.9 121.2 ± 6.0

HDL = high-density lipoprotein; LDL = low-density lipoprotein; NSAID = nonsteroidal antiinflammatory
drug.
Data given as percentages or means ± standard errors.
Significant heterogeneity: * P <0.05, ** P <0.01, *** P <0.001 by analysis of variance.

60
and by 27±15 µm in individuals homozygous for shorter alleles (P = 0.06). These
elevations were attenuated and lost significance after adjustment for trial, ethnicity, and
age. Ethnicity was the main confounder of the association between CIMT and shorter 5-
LO alleles, whereas trial and age were the main confounders of the association between
CIMT and longer 5-LO alleles. Under a dominant model with individuals classified as (i)
homozygous for the "5" allele, (ii) carrying at least one shorter allele, or (iii) homozygous
for longer alleles or carrying the "5" allele and one longer allele, 5-LO genotype was not
significantly associated with CIMT (P for likelihood ratio test of an overall gene effect =
0.35). Similarly, under a recessive model with individuals classified as (i) homozygous
for the "5" allele, (ii) carrying one shorter allele, (iii) homozygous for shorter alleles, or
(iv) homozygous for longer alleles or carrying the "5" allele and one longer allele, 5-LO
genotype was not significantly associated with CIMT (P for likelihood ratio test of an
overall gene effect = 0.18).
3.3.3 Gene-diet interactions between 5-LO promoter genotype and dietary
fats on carotid artery atherosclerosis
After adjustment for other cardiovascular risk factors, CIMT was inversely associated
with dietary intake of linoleic acid [18:2n-6], the main polyunsaturated fatty acid in
Western diets (mean±SE difference in CIMT for low versus high intake = 19±5 µm; P =
0.001), and positively associated with dietary intake of arachidonic acid [20:4n-6], the
long-chain derivative of linoleic acid (mean ± SE difference in CIMT for high versus low
intake = 12±5 µm; P = 0.03). CIMT was not significantly associated with dietary intakes
of saturated fat, monounsaturated fat, or n-3 polyunsaturated fats (Appendix B.6).

61

62
Figure 3.1. Interaction between dietary intake of linoleic acid [18:2n-6] and 5-lipoxygenase promoter
genotype on carotid intima-media thickness (CIMT) among subjects from four randomized controlled
atherosclerosis trials

Means (± standard errors) are shown.

Under a recessive model with the same genotype classifications as described above,
there was a significant gene-diet interaction (P = 0.048) between homozygosity for
shorter alleles and dietary intake of linoleic acid on CIMT (Figure 3.1). Among
individuals with low intake of linoleic acid, mean (±SE) CIMT was elevated by 43±19
µm in individuals homozygous for shorter alleles compared to individuals homozygous
for the "5" allele (P = 0.02). In contrast, mean (±SE) CIMT was reduced by 24±19 µm in
individuals homozygous for shorter alleles and with high intake of linoleic acid compared

63
to individuals homozygous for the "5" allele with low intake of linoleic acid (P = 0.21).
There were no significant gene-diet interactions on CIMT between 5-LO promoter
polymorphisms and dietary intakes of saturated fat, monounsaturated fat, n-3
polyunsaturated fats, or arachidonic acid (Appendix B.7).

3.3.4 Interactions between 5-LO promoter genotype and plasma lipids on
carotid artery atherosclerosis
We examined whether the associations between 5-LO promoter polymorphisms and
CIMT varied by concentrations of plasma lipids or use of lipid-lowering medication, a
surrogate for past elevations in plasma lipids. Under a recessive model with the same
genotype classifications as described above, there was a significant interaction (P = 0.03)
between homozygosity for shorter alleles and use of lipid-lowering medication on CIMT
(Figure 3.2). Mean (±SE) CIMT was elevated by 77±32 µm in users of lipid-lowering
medication homozygous for shorter alleles compared to non-users of lipid-lowering
medication homozygous for the "5" allele (P = 0.02). In contrast, among non-users of
lipid-lowering medication, CIMT was not significantly elevated in individuals
homozygous for shorter alleles compared to individuals homozygous for the "5" allele (P
= 0.63). Among users of lipid-lowering medication, the atherogenic effect of
homozygosity for shorter alleles was further enhanced by high concentrations of total
cholesterol, LDL cholesterol, and HDL cholesterol (Appendix B.8), although these
interactions did not achieve statistical significance (P > 0.05 for each interaction) due to
the low numbers of lipid-lowering medication users in each genotype-cholesterol group.

64
Figure 3.2. Interaction between baseline use of lipid-lowering medication and 5-lipoxygenase promoter
genotype on carotid intima-media thickness (CIMT) among subjects from four randomized controlled
atherosclerosis trials

Means (± standard errors) are shown.

3.4 Discussion
In this study, we did not find a significant association between 5-LO promoter
polymorphisms and carotid artery atherosclerosis as measured with CIMT. Although
CIMT was slightly elevated in both individuals homozygous for shorter alleles and
individuals homozygous for longer alleles compared to individuals homozygous for the
"5" allele, any atherogenic effect of the variant 5-LO alleles may have been due to chance
alone and was of a lesser magnitude than that reported by Dwyer et al. (20) in a smaller
cohort of healthy adults. To our knowledge, this is the first study aiming to replicate the

65
Los Angeles Atherosclerosis Study findings in another population free of symptomatic
cardiovascular disease. Our finding is consistent with results from two studies on 5-LO
promoter polymorphisms and risk of cardiovascular disease (21, 22) that did not find an
overall gene effect. Both of those studies compared 5-LO promoter genotypes between
myocardial infarction survivors and healthy controls. Other case-control studies have
reported an increased risk of myocardial infarction and stroke among individuals with
polymorphisms in the genes encoding FLAP and LTA
4
hydrolase (11-14), and variants in
the gene encoding LTC
4
synthase have been linked to carotid artery atherosclerosis and
coronary artery calcium (15).
In previous studies, the atherogenic effect of 5-LO promoter polymorphisms was
modified by dietary intakes of long-chain polyunsaturated fatty acids (20, 22). In
particular, it was enhanced by high dietary intake of arachidonic acid (20, 22) and blunted
by low intake of arachidonic acid (20, 22) or high intakes of EPA and docosahexaenoic
acid (DHA) [22:6n-3] (20), the long-chain n-3 polyunsaturated fatty acids found in fish
and other marine sources. Our study provides additional evidence for a nutrigenetic
association between 5-LO promoter polymorphisms and atherosclerosis. However, we
found a gene-diet interaction between 5-LO promoter polymorphisms and dietary intake
of linoleic acid, the intermediate-chain n-6 polyunsaturated fatty acid that comprises 84-
89% of energy intake from polyunsaturated fats among Americans (48). Moreover, the
atherogenic effect of homozygosity for shorter 5-LO promoter alleles was enhanced by
low intake of linoleic acid and blunted by high intake of linoleic acid. This observation is

66
therefore in the opposite direction as has been previously reported for arachidonic acid,
the long-chain derivative of linoleic acid.
Notably, the main dietary effect of linoleic acid on carotid artery atherosclerosis was
also in the opposite direction (i.e., an inverse association with CIMT) as that of
arachidonic acid among carriers of two variant 5-LO promoter alleles in the Los Angeles
Atherosclerosis Study (20). Although linoleic acid is converted in low amounts (~0.2%)
into arachidonic acid in vivo (49), increased dietary intake of linoleic acid does not seem
to translate into increased tissue concentrations of arachidonic acid (50). Inverse
associations between cardiovascular risk and blood, tissue, or dietary concentrations of
linoleic acid have been reported in a number of epidemiologic studies (51-55), and the
American Heart Association recently concluded that current evidence strongly supports a
protective role for consumption of n-6 polyunsaturated fats (56). Taken together with our
findings, this information suggests that the role of dietary fats in the 5-LO inflammatory
pathway as relates to atherogenesis may be more complex than direct competition
between arachidonic acid and EPA as substrates for 5-LO, and warrants further
investigation.
Unlike in the Los Angeles Atherosclerosis Study (20), we found significant
differences in plasma total and LDL cholesterol concentrations across 5-LO promoter
genotypes, such that homozygosity for shorter alleles was a risk factor for higher
cholesterol concentrations. We also found a significant interaction between use of lipid-
lowering medication and homozygosity for shorter alleles on CIMT, such that the
atherogenic effect of the shorter alleles was enhanced by use of lipid-lowering

67
medication, a surrogate for past elevations in plasma lipids. Among users of lipid-
lowering medication, the atherogenic effect of the shorter alleles was further enhanced by
high plasma concentrations of total, LDL, and HDL cholesterol at baseline. Equivalently,
total and LDL cholesterol were more atherogenic among individuals homozygous for
shorter 5-LO promoter alleles. Interestingly, a stronger atherogenic effect of other
cardiovascular risk factors, such as air pollution, has been reported for users of lipid-
lowering medications (57). Taken together with the finding of a significant gene-LDL
cholesterol interaction on CIMT in the Los Angeles Atherosclerosis Study (20), these
results suggest that the 5-LO inflammatory pathway may impact other pathophysiologic
mechanisms underlying atherogenesis.
A limitation of this study is its cross-sectional design, only using baseline assessment
of diet and other clinical and laboratory measurements. Lack of repeat measurements
could lead to measurement error that may have biased the observed effects of dietary
factors and other cardiovascular risk factors on CIMT. However, because the 5-LO
promoter polymorphisms were not independently associated with most cardiovascular
risk factors (except cholesterol concentrations) after controlling for ethnicity, we would
not expect any differential bias across the 5-LO promoter genotypes. Similarly, we
cannot rule out confounding of the association between 5-LO promoter polymorphisms
and CIMT by unmeasured environmental or genetic factors.
In summary, although we were unable to replicate previous findings of an overall
association between 5-LO promoter polymorphisms and carotid artery atherosclerosis,
our results suggest that genetic variation in the 5-LO promoter region may promote

68
atherogenesis in population subgroups defined by dietary factors or other cardiovascular
risk factors. Further investigation is needed in order to understand the interactions
between the 5-LO inflammatory pathway and other biologic mechanisms underlying
atherosclerosis and cardiovascular diseases.

69
3.5 Chapter 3 References
1. Lusis AJ. Atherosclerosis. Nature 2000; 407: 233-241.

2. Rosamond W, Flegal K, Furie K, et al. Heart disease and stroke statistics 2008
update: a report from the American Heart Association Statistics Committee and Stroke
Statistics Subcommittee. Circulation 2008; 117: e25-e146.

3. Tymchuk CN, Hartiala J, Patel PI, et al. Nonconventional genetic risk factors for
cardiovascular disease. Curr Atheroscler Rep 2006; 8: 184-192.

4. Radmark O, Werz O, Steinhilber D, et al. 5-lipoxygenase: regulation of expression
and enzyme activity. Trends Biochem Sci 2007; 32(7): 332-341.

5. James MJ, Gibson RA, Cleland LG. Dietary polyunsaturated fatty acids and
inflammatory mediator production. Am J Clin Nutr 2000; 71(suppl): 343S-348S.

6. Mehrabian M, Allayee H, Wong J, et al. Identification of 5-lipoxygenase as a major
gene contributing to atherosclerosis susceptibility in mice. Circ Res 2002; 91: 120-126.

7. Spanbroek R, Grabner R, Lotzer K, et al. Expanding expression of the 5-lipoxygenase
pathway within the arterial wall during human atherosclerosis. Proc Natl Acad Sci 2003;
100(3): 1238-1243.

8. Qiu H, Gabrielsen A, Agardh HE, et al. Expression of 5-lipoxygenase and leukotriene
A
4
hydrolase in human atherosclerotic lesions correlates with symptoms of plaque
instability. Proc Natl Acad Sci 2006; 103(21): 8161-8166.

9. Mehrabian M, Allayee H. 5-lipoxygenase and atherosclerosis. Curr Opin Lipidol
2003; 14: 447-457.

10. Ghazalpour A, Wang X, Lusis AJ, et al. Complex inheritance of the 5-lipoxygenase
locus influencing atherosclerosis in mice. Genetics 2006; 173: 943-951.

11. Helgadottir A, Manolescu A, Thorleifsson G, et al. The gene encoding 5-
lipoxygenase activating protein confers risk of myocardial infarction and stroke. Nat
Genet 2004; 36(3): 233-239.

12. Lohmussaar E, Gschwendtner A, Mueller JC, et al. ALOX5AP gene and the PDE4D
gene in a Central European population of stroke patients. Stroke 2005; 36: 731-736.

13. Helgadottir A, Gretarsdottir S, St Clair D, et al. Association between the gene
encoding 5-lipoxygenase-activating protein and stroke replicated in a Scottish population.
Am J Hum Genet 2005; 76: 505-509.

70
14. Helgadottir A, Manolescu A, Helgason A, et al. A variant of the gene encoding
leukotriene A4 hydrolase confers ethnicity-specific risk of myocardial infarction. Nat
Genet 2006; 38(1): 68-74.

15. Iovannisci DM, Lammer EJ, Steiner L, et al. Association between a leukotriene C4
synthase gene promoter polymorphism and coronary artery calcium in young women: the
Muscatine Study. Arterioscler Thromb Vasc Biol 2007; 27: 394-399.

16. Mehrabian M, Allayee H. 5-lipoxygenase and atherosclerosis. Curr Opin Lipidol
2003; 14: 447-457.

17. Silverman ES, Du J, De Sanctis GT, et al. Egr-1 and Sp1 interact functionally with
the 5-lipoxygenase promoter and its naturally occurring mutants. Am J Respir Cell Mol
Biol 1998; 19: 316-323.

18. In KH, Asano K, Beier D, et al. Naturally occurring mutations in the human 5-
lipoxygenase gene promoter that modify transcription factor binding and reporter gene
transcription. J Clin Invest 1997; 99: 1130-1137.

19. Drazen JM, Yandava CN, Dube L, et al. Pharmacogenetic association between
ALOX5 promoter genotype and the response to anti-asthma treatment. Nat Genet 1999;
22: 168-170.

20. Dwyer JH, Allayee H, Dwyer KM, et al. Arachidonate 5-lipoxygenase promoter
genotype, dietary arachidonic acid, and atherosclerosis. N Engl J Med 2004; 350: 29-37.

21. Gonzalez P, Reguero JR. A functional Sp1/Egr1-tandem repeat polymorphism in the
5-lipoxygenase gene is not associated with myocardial infarction. Int J Immunogenetics
2007; 34(2): 127-130.

22. Allayee H, Baylin A, Hartiala J, et al. Nutrigenetic association of the 5-lipoxygenase
gene with myocardial infarction. Am J Clin Nutr 2008; 88: 934-940.

23. Hodis HN, Mack WJ, LaBree L, et al. Alpha-tocopherol supplementation in healthy
individuals reduces low-density lipoprotein oxidation but not atherosclerosis: the Vitamin
E Atherosclerosis Prevention Study (VEAPS). Circulation 2002; 106(12): 1453-1459.

24. Hodis HN, Mack WJ, Dustin L, et al. High dose B-vitamin supplementation reduces
plasma homocysteine levels but not the progression of subclinical carotid, coronary or
aortic atherosclerosis: the B-vitamin Atherosclerosis Intervention Trial (BVAIT). Stroke,
in press.

71
25. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-
density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge.
Clin Chem 1972; 18: 499-502.

26. Hodis HN, St John JA, Xiang M, et al. Inflammatory markers and progression of
subclinical atherosclerosis in healthy postmenopausal women (from the Estrogen in the
Prevention of Atherosclerosis Trial). Am J Cardiol 2008; 101: 1131-1133.

27. Selzer RH, Hodis HN, Kwong-Fu H, et al. Evaluation of computerized edge tracking
for quantifying intima-media thickness of the common carotid artey from B-mode
ultrasound images. Atherosclerosis 1994; 111: 1-11.

38. Selzer RH, Mack WJ, Lee PL, et al. Improved common carotid elasticity and intima-
media thickness measurements from computer analysis of sequential ultrasound frames.
Atherosclerosis 2001; 154: 185-193.

29. Persson J, Formgren J, Israelsson B, et al. Ultrasound-determined intima-media
thickness and atherosclerosis: direct and indirect validation. Arterioscler Thromb 1994;
14: 261-264.

30. Salonen R, Salonen JT. Determinants of carotid intima-media thickness: a
population-based ultrasonography study in eastern Finnish men. J Intern Med 1991; 229:
225-231.

31. O'Leary DH, Polak JF, Kronmal RA, et al. Distribution and correlates of
sonographically detected carotid artery disease in the Cardiovascular Health Study.
Stroke 1992; 23: 1752-1760.

32. Geroulakos G, O'Gorman D, Nicolaides A, et al. Carotid intima-media thickness:
correlation with the British Regional Heart Study risk score. J Intern Med 1994; 234:
431-433.

33. El-Barghouti N, Elkeles R, Nicolaides A, et al. The ultrasonic evaluation of the
carotid intima-media thickness and its relation to risk factors of atherosclerosis in normal
and diabetic populations. Int Angiol 1997; 16(I): 50-54.

34. Craven TE, Ryu JE, Espeland MA, et al. Evaluation of the associations between
carotid artery atherosclerosis and coronary artery stenosis: a case-control study.
Circulation 1990; 82: 1230-1242.

35. Wofford JL, Kahl FR, Howard GR, et al. Relation of extent of extracranial carotid
artery atherosclerosis as measured by B-mode ultrasound to the extent of coronary
atherosclerosis. Arterioscler Thromb 1991; 11: 1786-1794.

72
36. Geroulakos G, O'Gorman DJ, Kalodiki E, et al. The carotid intima-media thickness
as a marker of the presence of severe symptomatic coronary artery disease. Eur Heart J
1994; 15: 781-785.

37. Crouse JR, Craven TE, Hagaman AP, et al. Association of coronary disease with
segment-specific intimal-medial thickening of the extracranial carotid artery. Circulation
1995; 92: 1141-1147.

38. Visona A, Pesavento R, Lusiani L, et al. Intimal medial thickening of common
carotid artery as indicator of coronary artery disease. Angiology 1996; 47: 61-66.

39. Mack WJ, LaBree L, Liu CR, et al. Correlations between measures of
atherosclerosis change using carotid ultrasonography and coronary angiography.
Atherosclerosis 2000; 150: 371-379.

40. Chambless LE, Heiss G, Folsom AR, et al. Association of coronary heart disease
incidence with carotid arterial wall thickness and major risk factors: the Atherosclerosis
Risk in Communities (ARIC) Study, 1987-1993. Am J Epidemiol 1997; 146: 483-494.

41. Chambless LE, Folsom AR, Clegg LX, et al. Carotid wall thickness is predictive of
incident clinical stroke: the Atherosclerosis Risk in Communities (ARIC) Study. Am J
Epidemiol 2000; 151: 478-487.

42. O'Leary DH, Polak JF, Kronmal RA, et al. Carotid-artery intima and media
thickness as a risk factor for myocardial infarction and stroke in older adults. N Engl J
Med 1999; 340: 14-22.

43. Iglesias del Sol A, Moons KG, Hollander M, et al. Is carotid intima-media thickness
useful in cardiovascular disease risk assessment? The Rotterdam Study. Stroke 2001;
32: 1532-1538.

44. Lorenz MW, Von Kegler S, Steinmetz H, et al. Carotid intima-media thickening
indicates a higher vascular risk across a wide age range: prospective data from the
Carotid Atherosclerosis Progression Study (CAPS). Stroke 2006; 37: 87-92.

45. Rosvall M, Janzon L, Berglund G, et al. Incident coronary events and case fatality in
relation to common carotid intima-media thickness. J Intern Med 2005; 257: 430-437.

46. Tsivgoulis G, Vemmos K, Papamichael C, et al. Common carotid artery intima-
media thickness and the risk of stroke recurrence. Stroke 2006; 37: 1913-1916.

47. Hodis HN, Mack WJ, LaBree L, et al. The role of carotid arterial intima-media
thickness in predicting clinical coronary events. Ann Intern Med 1998; 128: 262-269.

73
48. Kris-Etherton PM, Taylor DS, Yu-Poth S, et al. Polyunsaturated fatty acids in the
food chain in the United States. Am J Clin Nutr 2000; 71(suppl): 179S-188S.

49. Hussein N, Ah-Sing E, Wilkinson P, et al. Long-chain conversion of [
13
C]linoleic
acid and alpha-linolenic acid in response to marked changes in their dietary intake in
men. J Lipid Res 2005; 46: 269-280.

50. Sarkkinen ES, Agren JJ, Ahola I, et al. Fatty acid composition of serum cholesterol
esters, and erythrocyte and platelet membranes as indicators of long-term adherence to
fat-modified diets. Am J Clin Nutr 1994; 59: 364-370.

51. Oh K, Hu FB, Manson JE, et al. Dietary fat intake and risk of coronary heart disease
in women: 20 years of follow-up of the Nurses' Health Study. Am J Epidemiol 2005;
161: 672-679.

52. Shekelle RB, Shryock AM, Paul O, et al. Diet, serum cholesterol and death from
coronary heart disease: the Western Electric Study. N Engl J Med 1981; 304: 65-70.

53. Laaksonen DE, Nyyssonen K, Niskanen L, et al. Prediction of cardiovascular
mortality in middle-aged men by dietary and serum linoleic and polyunsaturated fatty
acids. Arch Intern Med 2005; 165: 193-199.

54. Harris WS, Poston WC, Haddock CK. Tissue n-3 and n-6 fatty acids and risk for
coronary heart disease events. Atherosclerosis 2007; 193: 1-10.

55. Kark JD, Kaufmann NA, Binka F, et al. Adipose tissue n-6 fatty acids and acute
myocardial infarction in a population consuming a diet high in polyunsaturated fatty
acids. Am J Clin Nutr 2003; 77: 796-802.

56. Harris WS, Mozaffarian D, Rimm E, et al. Omega-6 fatty acids and risk for
cardiovascular disease: a science advisory from the American Heart Association
Nutrition Subcommittee of the Council on Nutrition, Physical Activity, and Metabolism;
Council on Cardiovascular Nursing; and Council on Epidemiology and Prevention.
Circulation 2009; 119. doi: 10.1161/circulationaha.108.191627

57. Kunzli N, Jerrett M, Mack WJ, et al. Ambient air pollution and atherosclerosis in
Los Angeles. Environ Health Perspect 2005; 113: 201-206.

74
Chapter 4:
Validation of Multiethnic Cohort Study Self-Reports of Physician-Diagnosed
Myocardial Infarction and Other Conditions Through Comparison with California
Hospital Discharge Data

4.1 Introduction
Large prospective studies involving tens or hundreds of thousands of participants are
essential for the sensitive exploration of environmental and genetic influences on the
development of human diseases. One challenge inherent to these studies is ascertainment
of disease outcomes, particularly when non-fatal events are of interest. Existing medical
databases, when available, are a potential source of morbidity data for ascertainment of
conditions such as myocardial infarction (MI) that are relatively severe and require
hospitalization or frequent physician contact (1). However, many of these cohort studies
ask participants directly about their illnesses, hospitalizations, and surgeries and rely on
these self-reports either as outcome measures or as the basis for requesting medical
records from which adjudicated events are identified. Because accurate self-reporting
depends both on a participant's ability to remember a condition and correctly distinguish
it from related or similarly-named conditions as well as willingness to report the
condition (2), the validity of self-reporting may vary greatly depending on the population
and conditions of interest (2-5). We determined the sensitivity of self-reports of MI and
other physician-diagnosed conditions by comparing them against statewide hospital
discharge diagnoses in a multiethnic cohort recruited from the general population in

75
California. For each condition, we examined the accuracy of reporting of the year of
diagnosis and investigated whether patient and admission characteristics influenced
reporting in this population.

4.2 Materials and Methods
4.2.1 Population
The Multiethnic Cohort (MEC) is a prospective cohort of 215,251 adult men and
women, 45 to 75 years of age at the time of recruitment, residing in Hawaii or California
(mainly Los Angeles County). The cohort was established between 1993 and 1996 to
examine associations between diet, lifestyle, and genetic factors and incidence of cancer
and other chronic diseases in a representative sample of five major ethnic groups:
African-Americans, Japanese-Americans, Latinos, Native Hawaiians, and Whites. The
details of the study design and baseline characteristics of the cohort have been published
(6). Briefly, potential participants were identified through drivers' license files, voters'
registration files, and Health Care Financing Administration files. Individuals entered the
cohort by returning a 26-page mailed and self-administered questionnaire that asked
about demographic factors (including self-assigned ethnicity), anthropometrics, health-
related behaviors, past medical history, family history, use of medications, and for
women, reproductive history and use of exogeneous hormones. Participants of mixed
ethnicity were assigned to one category according to the following priority ranking:
African-American, Native Hawaiian, Latino, Japanese-American, White, and other.
Between 1999 and 2003, a second self-administered questionnaire was mailed to cohort

76
members to update their information on past medical history, family history, and certain
other factors. For the current study, we restricted our population to individuals recruited
in California and who returned both questionnaires (n = 80,635).
4.2.2 Assessment of medical history via self-reports
In the baseline questionnaire distributed in 1993-1996, participants were asked, "Has
your doctor ever told you that you had any of the following?" Participants then filled in
circles to self-report 28 conditions. In the second questionnaire distributed in 1999-2003,
participants were again asked to self-report a physician diagnosis of 24 conditions.
Participants were also asked to report the year in which each condition, if present, was
first diagnosed. Response options were before 1994, 1995, 1996, 1997, and 1998.
Importantly, neither questionnaire specifically asked about hospitalizations. Our primary
interest for this study was in reporting of MI, which was distinguished from angina
(described as "chest pain on exertion that is relieved by medication") in the second
questionnaire. For comparison, we analyzed eight other conditions that represent a range
of acute and chronic illnesses that may necessitate hospitalization. These conditions were
stroke, hypertension, diabetes, hip fracture, cholecystectomy, female breast cancer,
prostate cancer, and colorectal cancer.
4.2.3 Assessment of medical history via hospital records
Cohort members residing in California are linked each year to the California Office of
Statewide Health Planning and Development (OSHPD) hospitalization discharge
database. The database consists of mandatory records of all in-patient hospitalizations in
acute-care facilities in California except those at federal facilities or at City of Hope (a

77
comprehensive cancer center). Records include Social Security numbers and data on
patient demographics, the date and length of admission, type of admission (unscheduled
emergency or scheduled urgent or elective admission), illness severity (minor, moderate,
major, or extreme loss of function), the principal diagnosis and up to 24 other diagnoses,
and the principal procedure and up to 20 other procedures. Discharge diagnoses and
procedures are coded according to the Ninth Revision of the International Classification
of Disease (ICD-9) (7). The hospital records are checked by the OSHPD and are required
to meet data error tolerance levels of 0.1% for sex and date of birth (8). Linkage of the
cohort with hospital records is based solely on Social Security numbers obtained from the
self-administered questionnaires and was complete from January 1, 1991 through
December 31, 2005 at the time of this analysis. For comparability with the self-reports,
hospital records were included in this analysis only if the patient had been admitted prior
to January 1, 1999, i.e., prior to the distribution of the second questionnaire. If a
participant had multiple hospital records for the same diagnosis (as either the principal
diagnosis or any of the other discharge diagnoses) or procedure (as either the principal
procedure or any of the other procedures), only the first record was included.
4.2.4 Statistical analysis
We used discharge diagnoses from the OSHPD as the "gold standard" against which
self-reports from the second questionnaire were compared. For MI, our primary
outcome, we considered both "exact" and "nonspecific" matches. An "exact" match was
defined as ICD-9 codes 410 or 412 to a self-report of MI on the questionnaire. Three
"nonspecific" match definitions were used: (i) ischemic heart disease (ICD-9 codes 410-

78
414), (ii) heart disease (ICD-9 codes 390-398, 402, or 404-428) (4), and (iii) any disease
of the circulatory system (ICD-9 codes 390-459). Our secondary outcomes were stroke
(ICD-9 codes 433 or 434), hypertension (ICD-9 codes 401-405), diabetes (ICD-9 code
250), hip fracture (ICD-9 codes 820, 821.0, and 821.1), cholecystectomy (ICD-9 code
P51.2), female breast cancer (ICD-9 code 174), prostate cancer (ICD-9 code 185), and
colorectal cancer (ICD-9 codes 153 or 154). Notably, the ICD-9 definition of stroke was
restricted to ischemic stroke only; hemorrhagic stroke does not have unique ICD-9 codes,
and ICD-9 codes for cerebral hemorrhage would likely include conditions that we would
not expect participants to report as stroke.
For each outcome, we computed the percentage of "exact" or "nonspecific" OSHPD
diagnoses that were confirmed by self-reports on the second questionnaire. This
approach estimates the sensitivity of self-reporting to identify conditions for which
participants were hospitalized. OSHPD records that were not matched to self-reports
were considered "false negatives" since the participant had failed to self-report a
condition that, by virtue of having a discharge diagnosis, must have been physician-
diagnosed. Because the questionnaires did not specifically ask about hospitalizations, we
were not able to assess the specificity of self-reporting; self-reports that were not matched
to hospital records are not strictly "false positives" because they could reflect conditions
that were diagnosed out-of-hospital and which did not necessitate hospitalization during
the follow-up period. In order to assess the accuracy of reporting of the years in which
conditions were diagnosed, we conducted separate analyses with matches restricted to
conditions reported as being first diagnosed (i) in the same year as the first relevant

79
OSHPD record, or (ii) within three years before or after the year of the first relevant
OSHPD record, through 1998.
We used unconditional logistic regression to calculate the odds ratios of self-reporting
a condition for which the patient was hospitalized according to patient and admission
characteristics. Patient characteristics considered as potential predictors of reporting
were sex, ethnicity, age (<60, 60-64, 65-69, or >=70 years), maximum education
obtained (<=10th grade, 11th-12th grade, some further education, or college graduate),
and marital status (married, separated/divorced/widowed, or never married). Admission
characteristics considered as potential predictors of reporting were type of admission
(unscheduled emergency or scheduled urgent/elective admission), severity of the illness
(minor, moderate, major, or extreme loss of function), length of the hospital stay, and
time between the admission and the self-report (defined as the return of the second
questionnaire). Length of the hospital stay and time between the admission and the self-
report were grouped in quartiles of days for each condition. Data management used SAS
software version 9.1 (SAS Institute Inc., Cary, North Carolina). Statistical analyses used
STATA software version 10.0 (StataCorp, College Station, Texas).

4.3 Results
Of the 110,836 MEC participants recruited in California, 80,635 (73%) returned the
second questionnaire and were included in this study. Participants who returned the
second questionnaire were significantly more likely than non-respondents to be young
(i.e., less than 60 years old at cohort recruitment), female, married, and with at least some

80
further education beyond high school (P <0.001 for each characteristic); data not shown.
Latinos born in the United States, Japanese-Americans, and Whites were more likely to
return the second questionnaire than African-Americans or Latinos born outside the
United States (P <0.001); data not shown.
MI was reported by 5.6% of participants in the second questionnaire and was listed
among any of the discharge diagnoses in OSHPD records for 3.2% of participants.
Hypertension and diabetes were the conditions reported most frequently (44.8% and
16.9% of participants, respectively) and were the conditions for which the largest
proportions of the cohort had discharge diagnoses (16.3% and 7.2% of participants,
respectively). The prevalence of self-reporting of the other investigated conditions
ranged from 1.3% (hip fracture) to 9.1%. The prevalence rates of discharge diagnoses for
these conditions were even smaller, ranging from 0.3% (hip fracture) to 3.7% (prostate
cancer). In general, participants who reported MI or another health condition or who
were hospitalized for the condition were more likely to be older, male, African-
American, less educated, and unmarried (data not shown). However, a higher proportion
of women than men reported or were hospitalized for cholecystectomy or hip fractures
(data not shown). Married individuals were also more likely to undergo cholecystectomy
or be diagnosed with prostate cancer, and the proportion of participants reporting female
breast cancer or having a discharge diagnosis of prostate cancer increased with higher
education.
The sensitivity of self-reports to identify conditions for which participants were
hospitalized varied among the conditions investigated (Table 4.1). Of the 2,543 OSHPD

81

82
diagnoses of MI, 71.9% were reported. Sensitivity decreased considerably when
"nonspecific" OSHPD diagnoses for MI were considered. The rates for confirmation of
OSHPD diagnoses of hip fracture and cholecystectomy by self-reports were similar to
those of MI in magnitude, whereas the rates for confirmation of OSHPD diagnoses of
hypertension, diabetes, and cancers by self-reports were all greater than 80% and nearly
complete for female breast cancer and prostate cancer. For each condition, the sensitivity
of self-reports was greatly reduced if matches were restricted to a report in the same year
as the discharge diagnosis (i.e., such that OSHPD diagnoses with self-reports from non-
matching years were considered "false negatives") but only slightly reduced if matches
were restricted to a report within three years of the discharge diagnosis.
The mutually-adjusted odds ratios (ORs) of confirming an OSHPD diagnosis of MI
with a self-report within three years of the admission were calculated according to patient
and admission characteristics (Table 4.2). Hospitalizations for MI among women were
68% less likely to be reported (95% confidence interval [CI]: 0.56, 0.84) than those
among men. Compared to African-Americans, the sensitivity of MI reporting was 60%
higher among Whites (95% CI: 1.19, 2.16) and 37% higher among Latinos born in the
United States (95% CI: 1.06, 1.77). Participants who accurately reported MI were also
significantly younger (OR for age 70 years or older relative to age less than 60 years =
0.76, 95% CI: 0.59, 0.99; P
trend
= 0.03) and more highly educated (OR for college
graduate relative to less than an 11th grade education = 1.75, 95% CI: 1.29, 2.36; P
trend

<0.001) than those hospitalized for MI but without self-reports. A scheduled urgent or
elective admission with a discharge diagnosis of MI was 76% less likely to be reported as

83

84

85
physician-diagnosed MI than an unscheduled emergency admission for MI (95% CI:
0.61, 0.95), whereas longer length of hospital stay was significantly associated with
reporting of MI (OR for at least 7 days relative to 2 days or less = 1.45, 95% CI: 1.09,
1.94; P
trend
= 0.02).
Similar patterns were observed for the other conditions investigated, with a few
exceptions. For example, individuals hospitalized with a diagnosis of hypertension were
significantly more likely to report physician-diagnosed hypertension if they were women
than if they were men (OR = 1.22, 95% CI: 1.11, 1.34) and individuals hospitalized for a
hip fracture were significantly more likely to self-report hip fracture if they were older
(OR for age 70 years or older relative to age less than 60 years = 2.57, 95% CI: 1.02,
6.44; P
trend
= 0.04). Compared to African-Americans, Japanese-Americans had
significantly higher sensitivity of reporting for female breast cancer (OR = 4.24, 95% CI:
1.50, 11.98), whereas Latinos born outside of the United States had significantly lower
sensitivity of reporting for stroke (OR = 0.59, 95% CI: 0.37, 0.91), hypertension (OR =
0.56, 95% CI: 0.48, 0.65), and diabetes (OR = 0.76, 95% CI: 0.61, 0.95). Marital status,
illness severity, and time between the hospital admission and the self-report were not
significantly associated with reporting of MI. However, participants who were separated,
divorced, or widowed at the time of cohort recruitment were significantly less likely than
married participants to report hospitalization for diabetes (OR = 0.74, 95% CI: 0.63,
0.87) or prostate cancer (OR = 0.57, 95% CI: 0.36, 0.89) as a physician diagnosis of the
respective condition within three years of the admission. Illness severity had a significant
inverse association with the sensitivity of reporting for hypertension (OR for extreme loss

86
of function relative to minor loss of function = 0.64, 95% CI: 0.48, 0.85; P
trend
= 0.004),
diabetes (OR = 0.48, 95% CI: 0.32, 0.72; P
trend
<0.001), hip fracture (OR = 0.30, 95% CI:
0.04, 2.33; P
trend
= 0.04), and cholecystectomy (OR = 0.69, 95% CI: 0.36, 1.32; P
trend
=
0.03), whereas longer time between the hospital admission and the self-report had a
significant positive association with the sensitivity of reporting for stroke (OR for the
highest quartile relative to the lowest quartile = 1.69, 95% CI: 1.15, 2.51; P
trend
= 0.001),
hypertension (OR = 1.39, 95% CI: 1.19, 1.62; P
trend
<0.001), and diabetes (OR = 2.15,
95% CI: 1.64, 2.80; P
trend
<0.001).

4.4 Discussion
Medical conditions requiring hospitalization were accurately reported as physician-
diagnosed conditions by 59-93% of MEC participants. This range of sensitivity for self-
reports is comparable to that observed in the California Teachers Study (CTS), which
compared self-reported hospitalizations against OSHPD discharge diagnoses (4). As we
had expected, reporting varied greatly by condition. Accurate reporting of long-lasting
conditions such as hypertension, diabetes, and cancers (81-93%) was higher than
reporting of acute events such as MI, stroke, hip fracture, and cholecystectomy (59-72%).
One possible explanation for this finding is that the frequent physician contact often
necessitated by long-lasting medical problems may help participants correctly recall those
conditions, reducing under-reporting (2, 9). Hospitalizations with a principal diagnosis of
cancers also had the highest self-reporting in the CTS, but investigators did not
distinguish between cancers by site nor between acute and long-lasting conditions within

87
other diagnostic groups (4). In the MEC, sensitivity of reporting for cancers of the breast
and prostate (90-93%) was higher than for colorectal cancer (87%). These results are
consistent with the pattern of reporting by cancer site in the Cancer Prevention Study II
Nutrition Survey (10), which found that sensitivity of reporting for cancers was highest
for breast, prostate, and lung cancers (90-91%) and lowest for rectal cancer and
melanoma (16% and 53%, respectively). In contrast, the sensitivity of reporting for colon
cancer (58%) was lower than the sensitivity of reporting for breast cancer (79%) but
equivalent to that of prostate cancer (58%) in the 1980 New Haven Epidemiologic
Catchment Area Study (11). Notably, however, the latter study had much higher "false
negative" rates for all non-skin cancers than seen in the Cancer Prevention Study II
Nutrition Survey and in the MEC.
A unique feature of this study was its detailed assessment of the year of diagnosis
across a range of conditions. Sensitivity was reduced to 28-70% when matches were
restricted to self-reported diagnoses in the same years as their respective hospital
admissions, whereas sensitivity remained relatively high (55-89%) when matches were
restricted to self-reported diagnoses within three years of their respective hospital
admissions. These findings indicate that participants may have difficulty accurately
recalling the exact year in which a condition was first diagnosed, but are usually able to
recall the year of diagnosis within a much broader interval. For long-lasting conditions,
these findings also suggest that many participants diagnosed with conditions through
physician visits or outpatient services required hospitalization within three years of their
diagnosis.

88
Our results confirm that the accuracy of self-reporting depends on the population of
interest. Younger age and higher education were associated with higher sensitivity of
reporting for most conditions investigated; these findings are consistent with previous
studies of cancers and other diseases (3-5, 10, 11). In general, reporting was also
positively associated with the length of the hospital stay and inversely associated with
severity of the illness in both the MEC and CTS (4). Although scheduled hospital
admissions were more likely to be reported in the CTS for all diagnostic groups
combined (4), we found that scheduled admissions were more likely to be reported for
long-lasting conditions (e.g., hypertension, diabetes, and cancers) but less likely to be
reported for acute events such as MI and stroke. One possible explanation for the
apparent discrepancy between our results with regard to type of hospital admission and
those in the CTS is that in the latter study, participants were specifically asked about
hospitalizations and may not have known to also unscheduled admissions, such as those
occurring through the emergency department. The effects of sex and ethnicity on self-
reporting also varied somewhat by condition but was typically more accurate for men,
Latinos born in the United States, Japanese-Americans, and Whites and less accurate for
women, Latinos born outside the United States, and African-Americans. To our
knowledge, this is the first study to present data on the variability in the sensitivity of
self-reporting for a range of medical conditions across several non-White ethnic groups.
A limitation of this study is that we could not address potential over-reporting of
conditions by participants since the MEC questionnaires did not specifically ask about
hospitalizations. For example, MI may have been over-reported if participants confused

89
the condition with other forms of ischemic heart disease or non-cardiovascular conditions
that can cause chest pain (e.g., heartburn). In previous studies, the specificity of self-
reported MI ranged from 67-83% (3, 12, 13), meaning that 17-33% of self-reported cases
were not confirmed by review of medical records. In additional analyses using MEC
data, we found that the specificity of self-reporting (i.e., the confirmation rates of self-
reported conditions by OSHPD diagnoses) ranged from 22% (hypertension) to 46% (MI).
However, those rates are misleadingly low because we cannot distinguish between "false
positive" self-reports and accurate reporting of physician-diagnosed conditions that were
not severe enough to necessitate hospitalization.
Errors in the OSHPD data may explain some of the disparities with self-reports.
There may have been errors in the linkage of participants with the OSHPD records, since
linkage was based on Social Security numbers that may have contained errors. Whereas
ICD-9 coding of causes of death in mortality databases has been shown to have high
validity (14), coding of hospital discharge diagnoses and procedures may be more
variable. For example, estimates for the positive predictive value of an ICD-9 code of
410 to define MI in medical databases range from 55-95% compared to medical records
(13, 15-19) and may vary by sex and ethnicity (16). OSHPD records are monitored to
meet data error tolerance levels for sex and date of birth, but are not checked for accuracy
of reporting of discharge diagnoses, procedures, or other variables (8). In an audit of a
small number of OSHPD records, almost 65% of conditions identified in hospital records
were not included among the OSHPD discharge diagnoses (20). Coding practices also
differed across hospitals and appeared to be influenced by financial incentives to include

90
conditions that defined a patient's diagnosis-related group (DRG) or claims (20).
However, omission of discharge diagnoses from OSHPD records would not be expected
to effect our estimates for sensitivity of self-reporting.
Our findings suggest that the use of self-reports in epidemiologic studies as events or
as the basis for requesting medical records from which to identify adjudicated events may
result in substantial misclassification of outcomes due to under-reporting, particularly in
cohorts recruited from the general population. Because self-reporting depends on both
patient and admission characteristics, the misclassification is likely to be differential
across groups defined by sex, ethnicity, education, and other risk-related characteristics.
Medical databases, when available, may provide a more promising source of morbidity
data either alone or in combination with repeated self-reports.

91
4.5 Chapter 4 References
1. Harlow SD, Linet MS. Agreement between questionnaire data and medical records:
the evidence for accuracy of recall. Am J Epidemiol 1989; 129(2): 233-248.

2. Edwards WS, Winn DM, Kurlantzick V, et al. Evaluation of National Health
Interview Survey diagnostic reporting. (Vital and health statistics, series 2, no. 120).
Hyattsville, MD: National Center for Health Statistics, 1994.

3. Bergmann MM, Byers T, Freedman DS, et al. Validity of self-reported diagnoses
leading to hospitalization: a comparison of self-reports with hospital records in a
prospective study of American adults. Am J Epidemiol 1998; 147(10): 969-977.

4. Marshall SF, Deapen D, Allen M, et al. Validating California Teachers Study self-
reports of recent hospitalization: comparison with California hospital discharge data. Am
J Epidemiol 2003; 158(10): 1012-1020.

5. Mackenbach JP, Looman CWN, Van der Meer JBW. Differences in the misreporting
of chronic conditions, by level of education: the effect on inequalities in prevalence rates.
Am J Public Health 1996; 86(5): 706-711.

6. Kolonel LN, Henderson BE, Hankin JH, et al. A multiethnic cohort in Hawaii and
Los Angeles: baseline characteristics. Am J Epidemiol 2000; 151(4): 346-357.

7. Practice Management Information Corporation. International classification of
diseases, ninth revision, clinical modification. 6th ed. Los Angeles, CA: Practice
Management Information Corporation, 2000.

8. California Office of Statewide Health Planning and Development, Health Information
Division, Patient Discharge Section. California patient discharge data reporting manual,
third edition, for discharge data for the years 1999 and 2000. Sacramento, CA: Office of
Statewide Health Planning and Development, 2000.

9. Madow WG. Net differences in interview data on chronic conditions and information
derived from medical records. (Vital and health statistics, series 2, no. 57). Rockville,
MD: National Center for Health Statistics, 1973.

10. Bergmann MM, Calle EE, Mervis CA, et al. Validity of self-reported cancers in a
prospective cohort study in comparison with data from state cancer registries. Am J
Epidemiol 1998; 147(6): 556-562.

11. Desai MM, Bruce ML, Desai RA, et al. Validity of self-reported cancer history: a
comparison of health interview data and cancer registry records. Am J Epidemiol 2001;
153(3): 299-306.

92

12. Walker MK, Whincup PH, Shaper AG, et al. Validation of patient recall of doctor-
diagnosed heart attack and stroke: a postal questionnaire and record review comparison.
Am J Epidemiol 1998; 148(4): 355-361.

13. Heckbert SR, Kooperberg C, Safford MM, et al. Comparison of self-report, hospital
discharge codes, and adjudication of cardiovascular events in the Women's Health
Initiative. Am J Epidemiol 2004; 160(12): 1152-1158.

14. Sathiakumar N, Delzell E, Osman A. Using the National Death Index to obtain
underlying cause of death codes. J Occup Environ Med 1998; 40(9): 808-813.

15. Pladevall M, Goff DC, Nichaman MZ, et al. An assessment of the validity of ICD
code 410 to identify hospital admissions for myocardial infarction: the Corpus Christi
Heart Project. Int J Epidemiol 1996; 25(5): 948-952.

16. Rosamond WD, Chambless LE, Sorlie PD, et al. Trends in the sensitivity, positive
predictive value, false-positive rate, and comparability ratio of hospital discharge
diagnosis codes for acute myocardial infarction in four U.S. communities, 1987-2000.
Am J Epidemiol 2004; 160(12): 1137-1146.

17. Hammar N, Alfredsson L, Rosen M, et al. A national record linkage to study acute
myocardial infarction incidence and case fatality in Sweden. Int J Epidemiol 2001; 30:
S30-S34.

18. Peterson LA, Wright S, Normand SLT, et al. Positive predictive value of the
diagnosis of acute myocardial infarction in an administrative database. J Gen Intern Med
1999; 14: 555-558.

19. Kiyota Y, Schneeweiss S, Glynn RJ, et al. Accuracy of Medicare claims-based
diagnosis of acute myocardial infarction: estimating positive predictive value on the basis
of review of hospital records. Am Heart J 2004; 148(1): 99-104.

20. Green J, Wintfeld N. How accurate are hospital discharge data for evaluating
effectiveness of care? Med Care 1993; 31(8): 719-731.

93
Bibliography

Albert CM, Hennekens CH, O'Donnell CJ, et al. Fish consumption and risk of sudden
cardiac death. JAMA 1998; 279(1): 23-28.

Allayee H, Baylin A, Hartiala J, et al. Nutrigenetic association of the 5-lipoxygenase
gene with myocardial infarction. Am J Clin Nutr 2008; 88: 934-940.

Ascherio A, Rimm EB, Giovannucci EL, et al. Dietary fat and risk of coronary heart
disease in men: cohort follow up study in the United States. BMJ 1996; 313: 84-90.

Ascherio A, Rimm EB, Stampfer MJ, et al. Dietary intake of marine n-3 fatty acids, fish
intake, and the risk of coronary disease among men. N Engl J Med 1995; 332(15): 977-
982.

Bergmann MM, Byers T, Freedman DS, et al. Validity of self-reported diagnoses leading
to hospitalization: a comparison of self-reports with hospital records in a prospective
study of American adults. Am J Epidemiol 1998; 147(10): 969-977.

Bergmann MM, Calle EE, Mervis CA, et al. Validity of self-reported cancers in a
prospective cohort study in comparison with data from state cancer registries. Am J
Epidemiol 1998; 147(6): 556-562.

Bing R. Cardiology: The Evolution of the Science and the Art. Rutgers University Press:
Piscataway, NJ: 1999.

California Office of Statewide Health Planning and Development, Health Information
Division, Patient Discharge Section. California patient discharge data reporting manual,
third edition, for discharge data for the years 1999 and 2000. Sacramento, CA: Office of
Statewide Health Planning and Development, 2000.

Chambless LE, Folsom AR, Clegg LX, et al. Carotid wall thickness is predictive of
incident clinical stroke: the Atherosclerosis Risk in Communities (ARIC) Study. Am J
Epidemiol 2000; 151: 478-487.

Chambless LE, Heiss G, Folsom AR, et al. Association of coronary heart disease
incidence with carotid arterial wall thickness and major risk factors: the Atherosclerosis
Risk in Communities (ARIC) Study, 1987-1993. Am J Epidemiol 1997; 146: 483-494.

Chrysohoou C, Panagiotakos DB, Pitsavos C, et al. Long-term fish consumption is
associated with protection against arrhythmia in healthy persons in a Mediterranean
region -- the ATTICA study. Am J Clin Nutr 2007; 85: 1385-1391.

94
Craven TE, Ryu JE, Espeland MA, et al. Evaluation of the associations between carotid
artery atherosclerosis and coronary artery stenosis: a case-control study. Circulation
1990; 82: 1230-1242.

Crouse JR, Craven TE, Hagaman AP, et al. Association of coronary disease with
segment-specific intimal-medial thickening of the extracranial carotid artery. Circulation
1995; 92: 1141-1147.

Daviglus ML, Stamler J, Orencia AJ, et al. Fish consumption and the 30-year risk of
fatal myocardial infarction. N Engl J Med 1997; 336(15): 1046-1053.

De Caterina R, Liao JK, Libby P. Fatty acid modulation of endothelial activation. Am J
Clin Nutr 2000; 71(suppl): 213S-223S.

Desai MM, Bruce ML, Desai RA, et al. Validity of self-reported cancer history: a
comparison of health interview data and cancer registry records. Am J Epidemiol 2001;
153(3): 299-306.

Drazen JM, Yandava CN, Dube L, et al. Pharmacogenetic association between ALOX5
promoter genotype and the response to anti-asthma treatment. Nat Genet 1999; 22: 168-
170.

Dubnov G, Berry EM. Omega-6 fatty acids and coronary artery disease: the pros and
cons. Curr Atheroscler Rep 2004; 6: 441-446.

Dwyer JH, Allayee H, Dwyer KM, et al. Arachidonate 5-lipoxygenase promoter
genotype, dietary arachidonic acid, and atherosclerosis. N Engl J Med 2004; 350(1): 29-
37.

Edwards WS, Winn DM, Kurlantzick V, et al. Evaluation of National Health Interview
Survey diagnostic reporting. (Vital and health statistics, series 2, no. 120). Hyattsville,
MD: National Center for Health Statistics, 1994.

El-Barghouti N, Elkeles R, Nicolaides A, et al. The ultrasonic evaluation of the carotid
intima-media thickness and its relation to risk factors of atherosclerosis in normal and
diabetic populations. Int Angiol 1997; 16(I): 50-54.

Finnegan YE, Minihane AM, Leigh-Firbank EC, et al. Plant- and marine-derived n-3
polyunsaturated fatty acids have differential effects on fasting and postprandial blood
lipid concentrations and on the susceptibility of LDL to oxidative modification in
moderately hyperlipidemic subjects. Am J Clin Nutr 2003; 77: 783-795.

95
Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-
density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge.
Clin Chem 1972; 18: 499-502.

Geroulakos G, O'Gorman D, Nicolaides A, et al. Carotid intima-media thickness:
correlation with the British Regional Heart Study risk score. J Intern Med 1994; 234:
431-433.

Geroulakos G, O'Gorman DJ, Kalodiki E, et al. The carotid intima-media thickness as a
marker of the presence of severe symptomatic coronary artery disease. Eur Heart J
1994; 15: 781-785.

Ghazalpour A, Wang X, Lusis AJ, et al. Complex inheritance of the 5-lipoxygenase
locus influencing atherosclerosis in mice. Genetics 2006; 173: 943-951.

Goldbourt U, Yaari S, Medalie JH. Factors predictive of long-term coronary heart
disease mortality among 10,059 male Israeli civil servants and municipal employees: a
23-year mortality follow-up in the Israeli ischemic heart disease study. Cardiology 1993;
82: 100-121.

Gonzalez P, Reguero JR. A functional Sp1/Egr1-tandem repeat polymorphism in the 5-
lipoxygenase gene is not associated with myocardial infarction. Int J Immunogenetics
2007; 34(2): 127-130.

Green J, Wintfeld N. How accurate are hospital discharge data for evaluating
effectiveness of care? Med Care 1993; 31(8): 719-731.

Griffin MD, Sanders TA, Davies IG, et al. Effects of altering the ratio of dietary n-6 to n-
3 fatty acids on insulin sensitivity, lipoprotein size, and postprandial lipemia in men and
postmenopausal women aged 45-70 y: the OPTILIP Study. Am J Clin Nutr 2006; 84:
1290-1298.

Hammar N, Alfredsson L, Rosen M, et al. A national record linkage to study acute
myocardial infarction incidence and case fatality in Sweden. Int J Epidemiol 2001; 30:
S30-S34.

Harlow SD, Linet MS. Agreement between questionnaire data and medical records: the
evidence for accuracy of recall. Am J Epidemiol 1989; 129(2): 233-248.

Harris WS, Mozaffarian D, Rimm E, et al. Omega-6 fatty acids and risk for
cardiovascular disease: a science advisory from the American Heart Association
Nutrition Subcommittee of the Council on Nutrition, Physical Activity, and Metabolism;
Council on Cardiovascular Nursing; and Council on Epidemiology and Prevention.
Circulation 2009; 119. doi: 10.1161/circulationaha.108.191627

96

Harris WS, Poston WC, Haddock CK. Tissue n-3 and n-6 fatty acids and risk for
coronary heart disease events. Atherosclerosis 2007; 193: 1-10.

Hau MF, Smelt AHM, Bindels AJGH, et al. Effects of fish oil on oxidative resistance of
VLDL in hypertriglyceridemic patients. Arterioscler Thromb Vasc Biol 1996; 16: 1197-
1202.

Heckbert SR, Kooperberg C, Safford MM, et al. Comparison of self-report, hospital
discharge codes, and adjudication of cardiovascular events in the Women's Health
Initiative. Am J Epidemiol 2004; 160(12): 1152-1158.

Hegsted DM, Ausman LM, Johnson JA, et al. Dietary fat and serum lipids: an evaluation
of the experimental data. Am J Clin Nutr 1993; 57: 875-883.

Helgadottir A, Gretarsdottir S, St Clair D, et al. Association between the gene encoding
5-lipoxygenase-activating protein and stroke replicated in a Scottish population. Am J
Hum Genet 2005; 76: 505-509.

Helgadottir A, Manolescu A, Helgason A, et al. A variant of the gene encoding
leukotriene A4 hydrolase confers ethnicity-specific risk of myocardial infarction. Nat
Genet 2006; 38(1): 68-74.

Helgadottir A, Manolescu A, Thorleifsson G, et al. The gene encoding 5-lipoxygenase
activating protein confers risk of myocardial infarction and stroke. Nat Genet 2004;
36(3): 233-239.

Henderson SO, Haiman CA, Wilkens LR, et al. Established risk factors account for most
of the racial differences in cardiovascular disease mortality. PLoS ONE 2007; 2(4): e377.

Hodis HN, Mack WJ, Dustin L, et al. High dose B-vitamin supplementation reduces
plasma homocysteine levels but not the progression of subclinical carotid, coronary or
aortic atherosclerosis: the B-vitamin Atherosclerosis Intervention Trial (BVAIT). Stroke,
in press.

Hodis HN, Mack WJ, LaBree L, et al. Alpha-tocopherol supplementation in healthy
individuals reduces low-density lipoprotein oxidation but not atherosclerosis: the Vitamin
E Atherosclerosis Prevention Study (VEAPS). Circulation 2002; 106(12): 1453-1459.

Hodis HN, Mack WJ, LaBree L, et al. The role of carotid arterial intima-media thickness
in predicting clinical coronary events. Ann Intern Med 1998; 128: 262-269.

97
Hodis HN, St John JA, Xiang M, et al. Inflammatory markers and progression of
subclinical atherosclerosis in healthy postmenopausal women (from the Estrogen in the
Prevention of Atherosclerosis Trial). Am J Cardiol 2008; 101: 1131-1133.

Hu FB, Bronner L, Willett WC, et al. Fish and omega-3 fatty acid intake and risk of
coronary heart disease in women. JAMA 2002; 287(14): 1815-1821.

Hussein N, Ah-Sing E, Wilkinson P, et al. Long-chain conversion of [
13
C]linoleic acid
and alpha-linolenic acid in response to marked changes in their dietary intake in men. J
Lipid Res 2005; 46: 269-280.

Iglesias del Sol A, Moons KG, Hollander M, et al. Is carotid intima-media thickness
useful in cardiovascular disease risk assessment? The Rotterdam Study. Stroke 2001;
32: 1532-1538.

In KH, Asano K, Beier D, et al. Naturally occurring mutations in the human 5-
lipoxygenase gene promoter that modify transcription factor binding and reporter gene
transcription. J Clin Invest 1997; 99: 1130-1137.

Iovannisci DM, Lammer EJ, Steiner L, et al. Association between a leukotriene C4
synthase gene promoter polymorphism and coronary artery calcium in young women: the
Muscatine Study. Arterioscler Thromb Vasc Biol 2007; 27: 394-399.

James MJ, Gibson RA, Cleland LG. Dietary polyunsaturated fatty acids and
inflammatory mediator production. Am J Clin Nutr 2000; 71(suppl): 343S-348S.

Jung UJ, Torrejon C, Tighe AP, et al. n-3 fatty acids and cardiovascular disease:
mechanisms underlying beneficial effects. Am J Clin Nutr 2008; 87(suppl): 2003S-
2009S.

Kang JX, Leaf A. Prevention of fatal cardiac arrhythmias by polyunsaturated fatty acids.
Am J Clin Nutr 2000; 71(suppl): 202S-207S.

Kark JD, Kaufmann NA, Binka F, et al. Adipose tissue n-6 fatty acids and acute
myocardial infarction in a population consuming a diet high in polyunsaturated fatty
acids. Am J Clin Nutr 2003; 77: 796-802.

Kestin M, Clifton P, Belling B, et al. n-3 fatty acids of marine origin lower systolic
blood pressure and triglycerides but raise LDL cholesterol compared with n-3 and n-6
fatty acids from plants. Am J Clin Nutr 1990; 51: 1028-1034.

Kiyota Y, Schneeweiss S, Glynn RJ, et al. Accuracy of Medicare claims-based diagnosis
of acute myocardial infarction: estimating positive predictive value on the basis of review
of hospital records. Am Heart J 2004; 148(1): 99-104.

98

Knoops KT, De Groot LC, Kromhout D, et al. Mediterranean diet, lifestyle factors, and
10-year mortality in elderly European men and women: the HALE project. JAMA 2004;
292(12): 1433-1439.

Kolonel LN, Henderson BE, Hankin JH, et al. A multiethnic cohort in Hawaii and Los
Angeles: baseline characteristics. Am J Epi 2000; 151(4): 346-357.

Kris-Etherton PM, Taylor DS, Yu-Poth S, et al. Polyunsaturated fatty acids in the food
chain in the United States. Am J Clin Nutr 2000; 71(suppl): 179S-188S.

Kromhout D, Bosschieter EB, de Lezenne Coulander C. The inverse relation between
fish consumption and 20-year mortality from coronary heart disease. N Engl J Med
1985; 312(19): 1205-1209.

Kunzli N, Jerrett M, Mack WJ, et al. Ambient air pollution and atherosclerosis in Los
Angeles. Environ Health Perspect 2005; 113: 201-206.

Kushi LH, Lew RA, Stare FJ, et al. Diet and 20-year mortality from coronary heart
disease: the Ireland-Boston Diet-Heart Study. N Engl J Med 1985; 312(13): 811-818.

Laaksonen DE, Nyyssonen K, Niskanen L, et al. Prediction of cardiovascular mortality
in middle-aged men by dietary and serum linoleic and polyunsaturated fatty acids. Arch
Intern Med 2005; 165: 193-199.

Lichtenstein AH, Appel LJ, Brands, M, et al. Diet and lifestyle recommendations
revision 2006: a scientific statement from the American Heart Association Nutrition
Committee. Circulation 2006; 114: 82-96.

Lohmussaar E, Gschwendtner A, Mueller JC, et al. ALOX5AP gene and the PDE4D
gene in a Central European population of stroke patients. Stroke 2005; 36: 731-736.

Lorenz MW, Von Kegler S, Steinmetz H, et al. Carotid intima-media thickening
indicates a higher vascular risk across a wide age range: prospective data from the
Carotid Atherosclerosis Progression Study (CAPS). Stroke 2006; 37: 87-92.

Lusis AJ. Atherosclerosis. Nature 2000; 407: 233-241.

Mack WJ, LaBree L, Liu CR, et al. Correlations between measures of atherosclerosis
change using carotid ultrasonography and coronary angiography. Atherosclerosis 2000;
150: 371-379.

99
Mackenbach JP, Looman CWN, Van der Meer JBW. Differences in the misreporting of
chronic conditions, by level of education: the effect on inequalities in prevalence rates.
Am J Public Health 1996; 86(5): 706-711.

Madow WG. Net differences in interview data on chronic conditions and information
derived from medical records. (Vital and health statistics, series 2, no. 57). Rockville,
MD: National Center for Health Statistics, 1973.
Marks DB. Board Review Series. Biochemistry, 3rd ed. Lippincott Williams &
Williams: Baltimore, ML; 1999.
Marshall SF, Deapen D, Allen M, et al. Validating California Teachers Study self-
reports of recent hospitalization: comparison with California hospital discharge data. Am
J Epidemiol 2003; 158(10): 1012-1020.

Mata P, Varela O, Alonso R, et al. Monounsaturated and polyunsaturated n-6 fatty acid-
enriched diets modify LDL oxidation and decrease human coronary smooth muscle cell
DNA synthesis. Arterioscler Thromb Vasc Biol 1997; 17: 2088-2095.

McGee DL, Reed DM, Yano K, et al. Ten-year incidence of coronary heart disease in the
Honolulu Heart Program: relationship to nutrient intake. Am J Epidemiol 1984; 119:
667-676.

Mehrabian M, Allayee H, Wong J, et al. Identification of 5-lipoxygenase as a major gene
contributing to atherosclerosis susceptibility in mice. Circ Res 2002; 91: 120-126.

Mehrabian M, Allayee H. 5-lipoxygenase and atherosclerosis. Curr Opin Lipidol 2003;
14: 447-457.

Mensink RP, Zock PL, Kester ADM, et al. Effects of dietary fatty acids and
carbohydrates on the ratio of serum total to HDL cholesterol and on serum lipids and
apolipoproteins: a meta-analysis of 60 controlled trials. Am J Clin Nutr 2003; 77: 1146-
1155.

Moreno JJ, Mitjavila MT. The degree of unsaturation of dietary fatty acids and the
development of atherosclerosis (review). J Nutr Biochem 2003; 14: 182-195.

Mori TA, Beilin LJ. Omega-3 fatty acids and inflammation. Curr Atheroscler Rep 2004;
6: 461-467.

Mozaffarian D. Fish and n-3 fatty acids for the prevention of fatal coronary heart disease
and sudden cardiac death. Am J Clin Nutr 2008; 87(suppl): 1991S-1996S.

Nestel PJ. Fish oil and cardiovascular disease: lipids and arterial function. Am J Clin
Nutr 2000; 71(suppl): 228S-231S.

100
Nöthlings U, Wilkens LR, Murphy SP, et al. Meat and fat intake as risk factors for
pancreatic cancer: the Multiethnic Cohort Study. J Natl Cancer Inst 2005; 97(19): 1458-
1465.

Oh K, Hu FB, Manson JE, et al. Dietary fat intake and risk of coronary heart disease in
women: 20 years of follow-up of the Nurses' Health Study. Am J Epidemiol 2005; 161:
672-679.

O'Leary DH, Polak JF, Kronmal RA, et al. Carotid-artery intima and media thickness as
a risk factor for myocardial infarction and stroke in older adults. N Engl J Med 1999;
340: 14-22.

O'Leary DH, Polak JF, Kronmal RA, et al. Distribution and correlates of sonographically
detected carotid artery disease in the Cardiovascular Health Study. Stroke 1992; 23:
1752-1760.

Persson J, Formgren J, Israelsson B, et al. Ultrasound-determined intima-media
thickness and atherosclerosis: direct and indirect validation. Arterioscler Thromb 1994;
14: 261-264.

Peterson LA, Wright S, Normand SLT, et al. Positive predictive value of the diagnosis of
acute myocardial infarction in an administrative database. J Gen Intern Med 1999; 14:
555-558.

Pietinen P, Ascherio A, Korhonen P, et al. Intake of fatty acids and risk of coronary heart
disease in a cohort of Finnish men: the Alpha-Tocopherol, Beta-Carotene Cancer
Prevention Study. Am J Epidemiol 1997; 145(10): 876-887.

Pischon T, Hankinson SE, Hotamisligil GS, et al. Habitual dietary intake of n-3 and n-6
fatty acids in relation to inflammatory markers among US men and women. Circulation
2003; 108: 155-160.

Pladevall M, Goff DC, Nichaman MZ, et al. An assessment of the validity of ICD code
410 to identify hospital admissions for myocardial infarction: the Corpus Christi Heart
Project. Int J Epidemiol 1996; 25(5): 948-952.

Posner BM, Cobb JL, Belanger AJ, et al. Dietary lipid predictors of coronary heart
disease in men: the Framingham Study. Arch Intern Med 1991; 151: 1181-1187.

Practice Management Information Corporation. International classification of diseases,
ninth revision, clinical modification. 6th ed. Los Angeles, CA: Practice Management
Information Corporation, 2000.

101
Qiu H, Gabrielsen A, Agardh HE, et al. Expression of 5-lipoxygenase and leukotriene A
4

hydrolase in human atherosclerotic lesions correlates with symptoms of plaque
instability. Proc Natl Acad Sci 2006; 103(21): 8161-8166.

Radmark O, Werz O, Steinhilber D, et al. 5-lipoxygenase: regulation of expression and
enzyme activity. Trends Biochem Sci 2007; 32(7): 332-341.

Rosamond W, Flegal K, Furie K, et al. Heart disease and stroke statistics 2008 update: a
report from the American Heart Association Statistics Committee and Stroke Statistics
Subcommittee. Circulation 2008; 117: e25-e146.

Rosamond WD, Chambless LE, Sorlie PD, et al. Trends in the sensitivity, positive
predictive value, false-positive rate, and comparability ratio of hospital discharge
diagnosis codes for acute myocardial infarction in four U.S. communities, 1987-2000.
Am J Epidemiol 2004; 160(12): 1137-1146.

Rosvall M, Janzon L, Berglund G, et al. Incident coronary events and case fatality in
relation to common carotid intima-media thickness. J Intern Med 2005; 257: 430-437.

Salonen R, Salonen JT. Determinants of carotid intima-media thickness: a population-
based ultrasonography study in eastern Finnish men. J Intern Med 1991; 229: 225-231.

Sanders TA. Polyunsaturated fatty acids in the food chain in Europe. Am J Clin Nutr
2000; 71(suppl): 176S-178S.

Sarkkinen ES, Agren JJ, Ahola I, et al. Fatty acid composition of serum cholesterol
esters, and erythrocyte and platelet membranes as indicators of long-term adherence to
fat-modified diets. Am J Clin Nutr 1994; 59: 364-370.

Sathiakumar N, Delzell E, Osman A. Using the National Death Index to obtain
underlying cause of death codes. J Occup Environ Med 1998; 40(9): 808-813.

Selzer RH, Hodis HN, Kwong-Fu H, et al. Evaluation of computerized edge tracking for
quantifying intima-media thickness of the common carotid artey from B-mode ultrasound
images. Atherosclerosis 1994; 111: 1-11.

Selzer RH, Mack WJ, Lee PL, et al. Improved common carotid elasticity and intima-
media thickness measurements from computer analysis of sequential ultrasound frames.
Atherosclerosis 2001; 154: 185-193.

Shekelle RB, Shryock AM, Paul O, et al. Diet, serum cholesterol and death from
coronary heart disease: the Western Electric Study. N Engl J Med 1981; 304: 65-70.

102
Silverman ES, Du J, De Sanctis GT, et al. Egr-1 and Sp1 interact functionally with the 5-
lipoxygenase promoter and its naturally occurring mutants. Am J Respir Cell Mol Biol
1998; 19: 316-323.

Simopoulos AP. The importance of the omega-6/omega-3 fatty acid ratio in
cardiovascular disease and other chronic diseases. Exp Biol Med 2008; 233: 674-688.

Spanbroek R, Grabner R, Lotzer K, et al. Expanding expression of the 5-lipoxygenase
pathway within the arterial wall during human atherosclerosis. Proc Natl Acad Sci 2003;
100(3): 1238-1243.

Stram DO, Hankin JH, Wilkens LR, et al. Calibration of the dietary questionnaire for a
multiethnic cohort in Hawaii and Los Angeles. Am J Epidemiol 2000; 151(4): 358-370.

Sugano M, Hirahara F. Polyunsaturated fatty acids in the food chain in Japan. Am J Clin
Nutr 2000; 71(suppl): 189S-196S.

Tsimikas S, Philis-Tsimikas A, Alexopoulos S, et al. LDL isolated from Greek subjects
on a typical diet or from American subjects on an oleate-supplemented diet induces less
monocyte chemotaxis and adhesion when exposed to oxidative stress. Arterioscler
Thromb Vasc Biol 1999; 19: 122-130.

Tsivgoulis G, Vemmos K, Papamichael C, et al. Common carotid artery intima-media
thickness and the risk of stroke recurrence. Stroke 2006; 37: 1913-1916.

Tymchuk CN, Hartiala J, Patel PI, et al. Nonconventional genetic risk factors for
cardiovascular disease. Curr Atheroscler Rep 2006; 8: 184-192.

United States Department of Agriculture (USDA), Agricultural Research Service. What's
In The Food You Eat Search Tool, 3.0. USDA Home Page,
http://www.ars.usda.gov/Services/docs.htm?docid=17032 .

Vanden Heuvel JP. Diet, fatty acids, and regulation of genes important for heart disease.
Curr Atheroscler Rep 2004; 6: 432-440.

Visona A, Pesavento R, Lusiani L, et al. Intimal medial thickening of common carotid
artery as indicator of coronary artery disease. Angiology 1996; 47: 61-66.

Walker MK, Whincup PH, Shaper AG, et al. Validation of patient recall of doctor-
diagnosed heart attack and stroke: a postal questionnaire and record review comparison.
Am J Epidemiol 1998; 148(4): 355-361.

103
Willett W, Stampfer M. Implications of total energy intake for epidemiologic analysis.
In: Willett W. Nutritional Epidemiology, 2nd ed. New York, NY: Oxford University
Press, 1998: 273-301.

Willett W. Diet and coronary heart disease. In: Willett W. Nutritional Epidemiology.
2nd ed. New York, NY: Oxford University Press; 1998: 414-466.

Willett W. Issues in analysis and presentation of dietary data. In: Willett W. Nutritional
Epidemiology, 2nd ed. New York, NY: Oxford University Press, 1998: 321-346.

Wofford JL, Kahl FR, Howard GR, et al. Relation of extent of extracranial carotid artery
atherosclerosis as measured by B-mode ultrasound to the extent of coronary
atherosclerosis. Arterioscler Thromb 1991; 11: 1786-1794.

Xu J, Eilat-Adar S, Loria C, et al. Dietary fat intake and risk of coronary heart disease:
the Strong Heart Study. Am J Clin Nutr 2006; 84: 894-902.

Yuan JM, Ross RK, Gao YT, et al. Fish and shellfish consumption in relation to death
from myocardial infarction among men in Shanghai, China. Am J Epidemiol 2001;
154(9): 809-816.

104
Appendix A:
Dietary Intake of Monounsaturated Fat and Marine-Derived Long-chain N-3
Polyunsaturated Fatty Acids Reduces the Risk of Myocardial Infarction:
the Multiethnic Cohort Study

Table A.1. Hazard ratios of myocardial infarction according to quintiles of intake of specific types of fat
using a saturated fat replacement model
a
, Multiethnic Cohort, California, United States, 1993-2005

Saturated fat replacement model

Quintile of intake
(g/1000 kcal)
Median
intake
(g/1000 kcal)
No. cases
/
person-yrs HR 95% CI P trend
b
Monounsaturated fat
<10.25 8.78 586 / 115,996 1.00
10.25 to <12.19 11.32 631 / 115,707 1.03 0.90, 1.19
12.19 to <13.79 13.01 628 / 116,367 1.00 0.85, 1.17
13.79 to <15.54 14.60 645 / 115,285 1.00 0.83, 1.19
>=15.54 16.83 739 / 114,714 1.05 0.87, 1.28 0.71
n-6 polyunsaturated fat
<6.02 5.18 607 / 114,925 1.00
6.02 to <7.12 6.61 612 / 114,970 1.03 0.90, 1.18
7.12 to <8.08 7.60 669 / 115,911 1.14 0.97, 1.33
8.08 to <9.21 8.59 666 / 115,424 1.13 0.94, 1.35
>=9.21 10.17 675 / 116,839 1.18 0.96, 1.44 0.13
n-3 polyunsaturated fat
<0.69 0.60 616 / 111,338 1.00
0.69 to <0.81 0.76 643 / 120,325 0.94 0.82, 1.07
0.81 to <0.91 0.86 646 / 117,065 0.94 0.81, 1.10
0.91 to <1.03 0.96 662 / 112,210 0.97 0.82, 1.15
>=1.03 1.13 662 / 117,132 0.90 0.75, 1.09 0.34

CI = confidence interval; HR = hazard ratio.
a
Stratified on sex, ethnicity, age at recruitment, and year of recruitment and adjusted for smoking, history
of diabetes, total energy, and intakes of monounsaturated fat, n-6 polyunsaturated fat, n-3 polyunsaturated
fat, carbohydrate, protein, fiber, and cholesterol such that the HRs represent isocaloric substitution of
saturated fat with each type of unsaturated fat.
b
Two-sided P values for trend were calculated with likelihood ratio tests (df = 1) using the sex- and
ethnicity-specific median values for each quintile of intake as a continuous variable.

105

Table A.2. Hazard ratios of myocardial infarction according to quintiles of intake of EPA and DHA and
fish using saturated fat replacement models
a
, Multiethnic Cohort, California, United States, 1993-2005

Saturated fat replacement model

Quintile of intake
(g/1000 kcal)
Median
intake
(g/1000 kcal)
No. cases
/
person-yrs HR 95% CI P trend
b
EPA +DHA
<0.02 0.01 566 / 110,878 1.00
0.02 to <0.03 0.02 512 / 97,103 0.98 0.87, 1.11
0.03 to <0.05 0.04 915 / 156,900 1.04 0.92, 1.17
0.05 to <0.07 0.06 544 / 97,537 0.90 0.78, 1.03
>=0.07 0.09 692 / 115,650 0.89 0.77, 1.03 0.04
Fish
<2.10 0.90 651 / 115,876 1.00
2.10 to <4.42 3.24 616 / 115,994 0.91 0.81, 1.02
4.42 to <7.21 5.71 637 / 115,333 0.94 0.84, 1.05
7.21 to <11.72 9.07 643 / 115,268 0.90 0.80, 1.01
>=11.72 16.18 682 / 115,599 0.90 0.80, 1.02 0.17

CI = confidence interval; DHA = docosahexaenoic acid [22:6n-3]; EPA = eicosapentaenoic acid [20:5n-3];
HR = hazard ratio.
a
Stratified on sex, ethnicity, age at recruitment, and year of recruitment and adjusted for smoking, history
of diabetes, total energy, and intakes of monounsaturated fat, n-6 polyunsaturated fat, linolenic acid [18:3n-
3], carbohydrate, protein, fiber, and cholesterol such that the HRs represent isocaloric substitution of
saturated fat with EPA and DHA or fish.
b
Two-sided P values for trend were calculated with likelihood ratio tests (df = 1) using the sex- and
ethnicity-specific median values for each quintile of intake as a continuous variable.

106

107

108

109

110

111

112

113

114

115

116

117

118
Appendix B:
Cross-sectional Associations Between 5-Lipoxygenase Promoter Polymorphisms,
Dietary Fats, and Carotid Artery Atherosclerosis

119
Table B.2. Baseline characteristics of subjects from four randomized controlled atherosclerosis trials who
were or were not genotyped for 5-lipoxygenase promoter polymorphisms

Genotyped
(n = 1,839)
Not genotyped
(n = 13)
Trial *
VEAPS 19.0 15.4
BVAIT 27.1 0
WISH 19.0 53.9
ELITE 34.9 30.8

Female sex (%) * 74.4 46.2
Ethnicity
Non-hispanic white (%) 68.1 46.2
Non-hispanic black (%) 10.2 15.4
Hispanic (%) 12.8 23.1
Asian or Pacific Islander (%) 8.9 15.4

No college education (%) 10.9 15.4
Current or past smoking (%) 39.2 15.4
Use of aspirin (%) 21.5 15.4
Use of NSAID (%) 15.5 23.1
Use of non-NSAID antiinflammatory medication (%) 1.4 7.7
Use of antihypertensive medication (%) 24.3 15.4
Use of lipid-lowering medication (%) 17.8 23.1

Age (years) 59.9 ± 0.2 57.5 ± 2.7
Systolic blood pressure (mmHg) 123.6 ± 0.4 116.0 ± 4.3
Diastolic blood pressure (mmHg) 77.2 ± 0.2 73.9 ± 2.4
Body mass index (kg/m
2
) 27.5 ± 0.1 27.2 ± 1.2
Waist circumference (cm) 87.1 ± 0.3 87.9 ± 2.9
Waist-hip ratio 0.8 ± 0.002 0.8 ± 0.150
Total cholesterol (mg/dL) 224.5 ± 0.8 223.8 ± 9.1
LDL cholesterol 139.7 ± 0.8 148.2 ± 7.6
HDL cholesterol * 61.0 ± 0.4 52.9 ± 2.8
Triglycerides (mg/dL) 119.0 ± 1.5 113.2 ± 7.8
Carotid intima-media thickness (um) 770.5 ± 2.9 740.4 ± 33.8

BVAIT = B-Vitamin Atherosclerosis Intervention Trial; ELITE = Early versus Late Intervention Trial with
Estradiol; HDL = high-density lipoprotein; LDL = low-density lipoprotein; NSAID = nonsteroidal
antiinflammatory drug; VEAPS = Vitamin E Atherosclerosis Prevention Study; WISH = Women's
Isoflavone Soy Health Trial.
Data given as percentages or means ± standard errors.
Significant heterogeneity: * P <0.05, ** P <0.01, *** P <0.001 by Pearson chi-square test or t-test.

120
Table B.3. Baseline characteristics of subjects from four randomized controlled atherosclerosis trials who
were genotyped for 5-lipoxygenase promoter polymorphisms and did or did not have usable genotypes

Usable
genotype
(n = 1,827)
No usable
genotype
(n = 12)
Trial
VEAPS 18.8 50.0
BVAIT 27.2 16.7
WISH 19.1 8.3
ELITE 34.9 25.0

Female sex (%) * 74.3 100
Ethnicity
Non-hispanic white (%) 68.1 66.7
Non-hispanic black (%) 10.3 0
Hispanic (%) 12.8 8.3
Asian or Pacific Islander (%) 8.8 25.0

No college education (%) 11.0 8.3
Current or past smoking (%) 39.4 16.7
Use of aspirin (%) 21.4 41.7
Use of NSAID (%) 15.5 8.3
Use of non-NSAID antiinflammatory medication (%) 1.4 0
Use of antihypertensive medication (%) 24.3 33.3
Use of lipid-lowering medication (%) 17.8 16.7

Age (years) 59.9 ± 0.2 60.8 ± 2.3
Systolic blood pressure (mmHg) 123.5 ± 0.4 127.8 ± 5.4
Diastolic blood pressure (mmHg) 77.2 ± 0.2 79.3 ± 1.9
Body mass index (kg/m
2
) 27.5 ± 0.1 25.4 ± 1.2
Waist circumference (cm) 87.1 ± 0.3 81.0 ± 3.4
Waist-hip ratio 0.8 ± 0.002 0.8 ± 0.014
Total cholesterol (mg/dL) 224.4 ± 0.8 237.4 ± 16.4
LDL cholesterol 139.7 ± 0.8 146.3 ± 16.1
HDL cholesterol 61.0 ± 0.4 64.1 ± 3.9
Triglycerides (mg/dL) 118.9 ± 1.5 135.4 ± 17.4
Carotid intima-media thickness (um) 770.8 ± 2.9 732.1 ± 18.7

BVAIT = B-Vitamin Atherosclerosis Intervention Trial; ELITE = Early versus Late Intervention Trial with
Estradiol; HDL = high-density lipoprotein; LDL = low-density lipoprotein; NSAID = nonsteroidal
antiinflammatory drug; VEAPS = Vitamin E Atherosclerosis Prevention Study; WISH = Women's
Isoflavone Soy Health Trial.
Data given as percentages or means ± standard errors.
Significant heterogeneity: * P <0.05, ** P <0.01, *** P <0.001 by analysis of variance.

121

122

123
Table B.6. Carotid intima-media thickness (in µm) according to consumption of specific types of dietary
fats among subjects from four randomized controlled atherosclerosis trials

Model 1
a
Model 2
b
Fat (g/1000 kcal) n Mean ± SE P Mean ± SE P
Saturated fat
<=11.4 881 730 ± 8 732 ± 9
>11.4 881 732 ± 5 0.64 738 ± 6 0.41
Monounsaturated fat
<14.3 881 735 ± 8 732 ± 9
>=14.3 881 726 ± 5 0.12 727 ± 6 0.39
n-6 polyunsaturated fat
<=6.8 881 740 ± 8
>6.8 881 721 ± 5 <0.001
Linoleic acid [18:2n-6]
<=6.7 881 740 ± 8 732 ± 9
>6.7 881 721 ± 5 0.001 715 ± 6 0.003
Ara [20:4n-6]
<0.08 881 724 ± 8 732 ± 9
>=0.08 881 736 ± 5 0.03 745 ± 6 0.04
n-3 polyunsaturated fat
<=0.8 881 727 ± 8
>0.8 881 735 ± 5 0.18
Linolenic acid [18:3n-3]
<0.7 881 730 ± 8 732 ± 9
>=0.7 881 733 ± 5 0.60 733 ± 6 0.90
EPA [20:5n-3]
<0.008 881 729 ± 8 732 ± 9
>=0.008 881 732 ± 5 0.50 738 ± 9 0.50
DHA [22:6n-3]
<=0.03 881 729 ± 8 732 ± 9
>0.03 881 732 ± 5 0.64 726 ± 9 0.51
EPA + DHA
<0.04 881 728 ± 8
>=0.04 881 733 ± 5 0.45
Ratio of n-6 to n-3 polyunsaturated fats
<=8.3 881 737 ± 8
>8.3 881 726 ± 5 0.03
Ratio of Ara to EPA + DHA
<=2.1 880 730 ± 8
>2.1 882 732 ± 5 0.69

Ara = arachidonic acid; DHA = docosahexaenoic acid; EPA = eicosapentaenoic acid.
a
Adjusted for trial, ethnicity, age (continuous), college education (yes/no), ever smoking (yes/no), systolic
blood pressure (tertiles), waist circumference (tertiles), and energy intake (tertiles).
b
Mutually-adjusted and further adjusted for covariates as in model 1
a
.

124

125

126

Abstract (if available)

Abstract Atherosclerosis is the underlying cause of myocardial infarction (MI) and numerous environmental and genetic factors contribute to its development. Despite advances in our understanding of atherogenesis, many uncertainties on the relationship between diet and cardiovascular disease (CVD) risk remain. Chapter 1 reviews the biologic mechanisms for the effects of dietary fats on CVD risk, including effects on serum lipids, vascular function, chronic inflammation, and thrombosis.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Cardiovascular disease risk factors and cognitive function

PDF

Genetic studies of inflammation and cardiovascular disease

PDF

Genetic and dietary determinants of nonalcoholic fatty liver disease in Hispanic children

PDF

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

PDF

Hormone therapy timing hypothesis and atherosclerosis

PDF

Subclinical carotid atherosclerosis, psychosocial measures, and cognitive function in middle- to older-aged adults

PDF

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

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

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

PDF

The contribution of arachidonate 5-lipoxygenase to inflammatory response in coronary artery disease

PDF

Ectopic fat and adipose tissue inflammation in overweight and obese African Americans and Hispanics

PDF

Sex hormones and atherosclerosis in postmenopausal women

PDF

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

PDF

Quantity versus quality: how adipose tissue accumulation and immune cell profile associate with risk for type 2 diabetes in minority children and adults

PDF

Micelle nanoparticles for the detection and treatment of atherosclerosis

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

Longitudinal changes in physical activity and physical fitness: associations with blood pressure

PDF

Molecular mechanisms of young-onset type 2 diabetes: integration of diet and multi-omics biomarkers

PDF

Dietary and supplementary folate intake and prostate cancer risk

PDF

The role of inflammation in mediating effects of obesity on Alzheimer's disease

Asset Metadata

Creator Roth, Nitzan C. (author)

Core Title Dietary fats, atherosclerosis, and cardiovascular disease

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Epidemiology

Publication Date 11/11/2011

Defense Date 07/02/2009

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

Tag 5-lipoxygenase,atherosclerosis,cardiovascular disease,carotid intima-media thickness,dietary fats,Ethnicity,hospital records,OAI-PMH Harvest,prospective studies,self-reports

Place Name California (states)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Mack, Wendy J. (committee chair), Allayee, Hooman (committee member), Crimmins, Eileen M. (committee member), Henderson, Brian E. (committee member), Hodis, Howard Neil (committee member), Monroe, Kristine (committee member)

Creator Email nitzanroth@yahoo.com,nroth@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2728

Unique identifier UC1138093

Identifier etd-Roth-3324 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-271579 (legacy record id),usctheses-m2728 (legacy record id)

Legacy Identifier etd-Roth-3324.pdf

Dmrecord 271579

Document Type Dissertation

Rights Roth, Nitzan C.

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu