Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The multiethnic nature of chronic disease: studies in the multiethnic cohort
(USC Thesis Other)
The multiethnic nature of chronic disease: studies in the multiethnic cohort
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THE MULTIETHNIC NATURE OF CHRONIC DISEASE: STUDIES IN THE MULTIETHNIC
COHORT
by
Neal Atul Tambe
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)
August 2016
Copyright 2016 Neal Atul Tambe
ii
DEDICATION
To my half-Egyptian, half-German wife, and my multiethnic children.
iii
ACKNOWLEDGMENTS
Committee: Dr. Christopher Haiman, Dr. Veronica Wendy Setiawan, Dr. Wendy Cozen,
Dr. Dan Stram, Dr. Gerhard Coetzee
Lab members: Ying Han, Ye Feng, Peggy Wan, Kristen Rand, Jackie Porcel, Jane
Figueiredo, Alex Stram, Wangjing Ke, Grace Sheng, Lani Park, Loreall Pooler, Karen Park,
Enrique Gonzalez
Hawaii: Dr. Loic Le Marchand, Dr. Lynne Wilkens
Dr. Brian Henderson
Caroline Wesemuller, Sophia
Parents, sister, family and friends
iv
CONTENTS
Dedication ii
Acknowledgments iii
List of Tables vi
List of Figures viii
Abbreviations x
Abstract xii
Chapter 1: Introduction 1
The multiethnic nature and biology of disease 1
Epidemiological study design for multiethnic studies 2
The Multiethnic Cohort 3
Colorectal cancer and atopic conditions 7
Racial/ethnic differences in cardiovascular disease and stroke 13
Gallbladder disease: exploring genetic variation by ethnicity 16
Summary 21
References 22
Chapter 2: Atopic Allergic Conditions and Colorectal Cancer Risk in the Multiethnic
Cohort 34
Abstract 35
Introduction 36
Methods 37
Discussion 50
References 56
Chapter 3: Racial and Ethnic Differences in Cardiovascular Disease Mortality: The
Multiethnic Cohort 59
Abstract 60
Introduction 62
Methods 64
Results 68
Discussion 89
References 95
Chapter 4: Multiethnic Genome-wide Association Study of Gallbladder Disease 100
Abstract 100
Introduction 102
Methods 104
Results 109
v
Discussion 126
References 131
Chapter 5: Conclusions and Future Directions 137
AAC & CRC 137
Ethnic Differences in Cardiovascular Disease & Stroke risk 139
GBD and genetic susceptibility 142
Conclusion 144
References 144
Appendix A: AAC and CRC 149
Appendix B: CVD Risk factors 152
Appendix C: GBD GWAS 163
vi
LIST OF TABLES
Table 2.1 Baseline Characteristics of Analyzed Individuals by Sex and AAC
Status in the MEC (1993-2010)
42
Table 2.2 Estimated Relative Risk of CRC Incidence Associated with AAC Status
by Ethnicity and Subgroup in the Multiethnic Cohort (1993-2010)
45
Table 2.3 Estimated Relative Risk of CRC Mortality Associated with AAC Status
by Ethnicity and Subgroup in the Multiethnic Cohort (1993-2010)
47
Table 2.4 Estimated Relative Risk of CRC Incidence Associated with AAC Status
in Individuals with Self-Reported Endoscopy with Follow-up
Beginning with the Second Questionnaire (1999-2003)
49
Table 3.1 Mortality rates from AMI, OHD, and stroke by race and sex in the
MEC
70
Table 3.2 Age-standardized distributions of risk factors (%) with associated
RRs of mortality from AMI, OHD, and stroke amongst men in the
MEC
73
Table 3.3 Age-standardized distributions of risk factors and associated RRs of
mortality from AMI, OHD, and stroke amongst women in the MEC
75
Table 3.4 Observed and adjusted RR and 95% confidence interval of mortality
in men by race in the MEC
80
Table 3.5 Observed and adjusted RR and 95% confidence interval of mortality
in women by race in the MEC
81
Table 3.6 Least-squared lipid means by race/ethnicity and sex in the MEC 87
Table 4.1 Top single SNP associated with GBD by region in the multiethnic
meta-analysis and ethnic-specific analyses
110
Table 4.2 Top SNP in conditional analysis in the ABCG5/8 region of
chromosome 2
110
Table A.1 Multivariate Risks of Colorectal Cancer Associated with AAC Status
Stratified by Risk Factors in the MEC (1993-2010)
150
Table B.1 Prevalence of cholesterol medication use and number of events by
sex in the MEC
153
vii
Table B.2 Main effect RR of cholesterol medication usage on disease risk, by
sex in the MEC
154
Table B.3 Percent change in racial/ethnic specific risk estimates and percent
of crude RRs by sex and race/ethnicity
159
Table C.1 Characteristics of Studies included in GWA analysis of Gallbladder
Disease
164
Table C.2 Number of typed and imputed SNPs analyzed by substudy 165
Table C.3 Genome wide significant SNPs in the 2p21 region 166
Table C.4 Genome-wide significant SNPs in the 7p22.2 region in Japanese
Americans in the MEC
168
Table C.5 Allele frequencies of SNPs in LD with rs13226493, by race/ethnicity
in the MEC
169
Table C.6 Genome-wide significant SNPs in the 14q21.3 and 15q21.3 regions
in African Americans in the MEC
170
Table C.7 Known GBD risk SNPs and associations in MEC 171
Table C.8 Known lipid-related risk variants and associations with GBD in the
MEC
172
viii
LIST OF FIGURES
Figure 3.1a Crude and adjusted RRs of mortality from AMI, OHD, and stroke by
race in men in the MEC
82
Figure 3.2b Crude and adjusted RRs of mortality from AMI, OHD, and stroke by
race in women in the MEC
82
Figure 3.2 Calculated least-squared means of lipid levels by sex in the MEC 88
Figure 4.1a Manhattan plots of GWA inverse-weighted meta-analysis. (a)
Meta-analysis across all races/ethnicities.
115
Figure 4.1b Manhattan plots of GWA inverse-weighted meta-analysis. (b)
Meta-analysis across African Americans.
115
Figure 4.1c Manhattan plots of GWA inverse-weighted meta-analysis. (c)
Meta-analysis across Japanese Americans.
116
Figure 4.1d Manhattan plots of GWA inverse-weighted meta-analysis. (d)
Meta-analysis across Latinos.
116
Figure 4.2a Regional associations for findings P < 5.0 x 10
-8
.
(a) Region at 2p21
encompassing genes ABCG5/8
117
Figure 4.2b Regional associations for findings P < 5.0 x 10
-8
.
(b) Region at
7p22.2, present only in Japanese Americans in the MEC.
119
Figure 4.2c Regional associations for findings P < 5.0 x 10
-8
.
(c) Region at
14q21.3, present only in African Americans in the MEC.
120
Figure 4.2d Regional associations for findings P < 5.0 x 10
-8
.
(d) Region on
chromosome 15, present only in African Americans in the MEC.
121
Figure 4.3a Association plots of conditional analyses in the 2p21 region,
surrounding ABCG5/8. (a) Local association plot conditioning on
the top SNP in the region, rs11887534.
122
Figure 4.3b Association plots of conditional analyses in the 2p21 region,
surrounding ABCG5/8. (b) Local association plot, conditioning on
both rs11887534 and rs4953026.
124
ix
Figure 4.3c Association plots of conditional analyses in the 2p21 region,
surrounding ABCG5/8. (c) Third step of local association plot,
conditioning on rs11887534, rs4953026, and rs7599564.
125
Figure B.1 Sub-analysis of individuals reporting cholesterol medication usage,
and the effect on risk/ethnic specific risk differences in the MEC
using White individuals as the baseline.
155
Figure C.1 Quantile-quantile (QQ) plots of race/ethnic specific inverse
weighted meta-analyses
178
x
ABBREVIATIONS
AAC Atopic Allergic Conditions
AMI Acute myocardial infarction
CI Confidence Interval
CPS-II Cancer Prevention Study II
CRC Colorectal Cancer
CVD Cardiovascular Disease
DMS Diabetes in Mexico Study
FHS Framingham Health Study
GBD Gallbladder Disease
GWAS Genome-wide Association Study
GWAS Genome-wide Association Study
HDL High Density Lipoprotein
HR Hazard Ratio
HRT Hormone Replacement Therapy
ICD International Classification of Diseases
IgE Immunoglobulin E
IS Ischemic Stroke
LDL Low Density Lipoprotein
MCDS Mexico City Diabetes Study
MEC Multiethnic Cohort
xi
NSAID Non-Steroidal Anti-Inflammatory Drug
OHD Other heart diseases
PCA Principal Components Analysis
PCA Principal Components Analysis
RAF Risk Allele Frequency
RR Relative Risk
SD Standard Deviation
SEER Surveillance, Epidemiology, and End Results
SEER Surveillance, Epidemiology, and End Results
SIGMA Slim Initiative for Genomic Medicine in the Americas
SNP Single Nucleotide Polymorphism
T2D Type 2 Diabetes
TC Total Cholesterol
TH2 T-Helper cell type 2
xii
ABSTRACT
Racial/ethnic disparities exist in incidence and mortality of colorectal cancer (CRC),
cardiovascular diseases (CVD), stroke, and gallbladder disease (GBD). The prevalence of
risk factors for these diseases also differ by race/ethnicity. Here we examined common
chronic diseases that vary across racial/ethnic populations in the prospective
Multiethnic Cohort (MEC) in Los Angeles and Hawaii to define risk factors that are
important in each population which may contribute to differences in risk across Whites,
African Americans (AA), Native Hawaiians (NH), Japanese Americans (JA), and Latinos.
We used a prospective cohort analysis to examine the relationship between atopic
allergic conditions (AAC) and CRC and found that individuals with AAC had a 14%
decreased CRC risk across all races/ethnicities. In a mortality analysis of CVD and stroke
risk, we found that known risk factors accounted for much of the difference in risk of
mortality when compared to Whites, except for excess risk in AA women, NH men and
women, and decreased risk in JA men and women. A genome-wide association study
(GWAS) replicated a known risk locus for GBD in the gene ABCG5/8, and identified in a
JA-specific locus (7p22) and two AA-specific loci (14q21.3 & 15q21.3) associated with
increased risk of GBD. Both environmental and genetic risk factors contribute to varying
risks of incidence and mortality of common chronic diseases across race/ethnicity.
1
CHAPTER 1: INTRODUCTION
1.1 The multiethnic nature and biology of disease
Epidemiological methodology was initially developed in the 20
th
century, with the goal
of understanding the sufficient and component causes of disease in specific populations
(1,2). Chronic conditions, as defined by the National Center for Health Statistics, involve
diseases that persist for more than 3 months, are non-self-limiting in nature, and
produce recurrent health problems (3). These include diabetes, cancer, and
atherosclerotic diseases that could precipitate an acute mortal event such as stroke or
myocardial infarction. The incidence and prevalence of chronic conditions often differ
by race/ethnicity. For example, between 2007 and 2011, the incidence of cancer was
606.2/100,000 persons in Black males, 540.8/100,000 in White males, and
322.3/100,000 persons in Asian/Pacific Islander males. Rates of death due to cancer
were 275.5/100,000 males in Blacks, 214.0/100,000 persons in White males, and
131.0/100,000 persons in Asian/Pacific Islander males. While rates across all groups
were lower in women, the general trends by race/ethnicity remained the same. Similar
patterns are also seen in colorectal (CRC), lung & bronchial, prostate, and breast cancers
(4). Rates of non-cancer chronic conditions, such as cardiovascular disease (CVD) and
gallbladder disease (GBD) differ by ethnicity as well, and are explored in subsequent
chapters (5,6 In drafting). Over time, global and national demographics have shifted
with regards to age, race, and ethnicity. In 1940, the U.S. population consisted of 89.8%
2
White, 9.8% Black, and 0.4% “other” races and ethnicities. In 2010, population estimates
had shifted to 72.4% White, 12.6% Black, 4.8% Asian, and 7.3% other races alone (7).
Now, 16.3% of individuals report a Hispanic/Latino origin. It is now increasingly
important to study the causal factors and spread of chronic diseases from a multiethnic
context.
1.2 Epidemiological study design for multiethnic studies
There are multiple ways to study a multiethnic population, including cross-sectional
studies, case-control studies, and cohort studies. While case-control studies are an
excellent method of examining a large number of exposures given one outcome, a
cohort study is preferred as it allows for the direct comparison of risk among many
groups where the status of exposure onset is known to occur prior to the incidence of
disease. Many informative prospective cohort studies have been assembled over the
past few decades, though most of them primarily focus on White individuals. These
include: the Framingham Heart Study (FHS) and its various offshoot cohorts, the
Women’s Health Study (WHS), the Cancer Prevention Studies (CPS), the California
Teacher’s Study (CTS), the Harvard Grant and Glueck studies, the Physician Health
Studies I & II, and the Nurses Health Studies (8–14). Often studies will compare only one
ethnic minority with Whites, and those minorities apart from Blacks are frequently
underrepresented (15,16).
Multiethnic studies of chronic disease are important given changing national/global
demographics for a number of reasons. First, racial/ethnic differences in the incidence
3
and outcome of disease may be caused by a host of socioeconomic, cultural, genetic,
and non-genetic health factors. Including multiple races/ethnicities in a study allows for
the identification of causal factors in one population that may not be as common in
another, such as differences in migration patterns, diet, or other exposures. Diversifying
the population of a study can increase the variability and also inform biology,
uncovering hidden interactions between various risk factors and race/ethnicity (1). The
Southern Community Cohort Study (SCCS) and Black Women’s Health Study (BWHS)
were created to help identify causes of cancer and other diseases in a specific
racial/ethnic population (17,18). The Hispanic Community Health Study (HCHS) and
Study of Latinos (SOL) focus on the unique dietary and health factors in Hispanic/Latino
populations, and how they affect disease risk (19).
Finally, examining differences in risk by race/ethnicity can help advise public policy. The
differences in which exposures are prevalent in each population and how they affect
chronic disease can allow policy makers to finely tailor and target public health policy.
1.3 The Multiethnic Cohort
1.3.1 Construction of the MEC
The National Health and Nutritional Examination Survey (NHANES) Epidemiologic
Follow-up Study (NEHFS) examines a multiethnic population through longitudinal
surveys, while the Multiethnic Study of Atherosclerosis (MESA) is a cohort specifically
devoted to examining multiethnic exposures (20,21). The Multiethnic Cohort (MEC),
based out of Los Angeles, CA and Hawaii is one of the largest prospective multiethnic
4
cohorts in the United States, with 215,251 adult men and women enrolled at baseline. It
includes a multiethnic distribution of individuals: 26.4% Japanese-American, 22.9%
White, 22.0% Latino, 16.3% African-American, 6.5% Native-Hawaiian, and 5.8% "other"
ethnic individuals (22). This large prospective cohort was begun to examine how dietary
and lifestyle risk factors contribute to ethnic differences in cancer development. It has
since been used to study both cancer (i.e. prostate and breast) and non-cancer
outcomes (cardiovascular disease and allergies), as well as the effects of therapeutic
modalities on disease progression (mammography use and postmenopausal hormone
therapy on breast cancer outcomes) (23–27). The MEC was constructed between 1993-
1996 and includes individuals between 45 and 75 years of age who completed a 26-page
self-administered questionnaire containing information on basic demographic factors,
lifestyle factors (including diet, physical activity, and medication use), prior chronic
conditions, as well as reproductive history and hormone usage in women.
The comprehensive 26-page dietary questionnaire allows for the assessment of a
plethora of dietary factors, from macromolecules to minerals. The inclusion of
medication usage, physical activity and self-reported physician diagnosed conditions and
comorbidities allows for the analysis and stratification of various covariates and
confounders.
Multiple additional follow-up questionnaires have been sent to surviving participants of
the MEC which allows for the comparison of exposures across time. However, given the
large size of the cohort, it is very costly and time consuming to collect risk factor
information multiple times, especially when information cannot be gathered via a
5
mailed questionnaire. A nested case-control allows for the examination of smaller
subset of individuals targeting a specific outcome/disease, where it is easy to collect
data (such as blood samples). Multiple nested case-control studies have been
constructed within each racial/ethnic group with the purpose of examining specific
types of cancers (primarily prostate and breast). Investigators in the MEC have collected
biological specimens (blood and urine) to examine genetic variants and the interaction
of these variants with disease and environmental covariates (28). This biorepository was
constructed between 2001-2006, forming a sub-cohort of 67,594 individuals (29). Since
its inception in the early 1990’s, the MEC has generated hundreds of publications.
1.3.2 Advantages of the MEC
The primary purpose of the MEC was to examine the relationship of various
environmental factors (initially diet) with cancer risk through a long-term prospective
analysis and linkage with the National Cancer Institute’s (NCI) Surveillance,
Epidemiology, and End Results (SEER) program. The use of a cohort model additionally
allows for the examination of non-cancer related outcomes as well. Linkage with the
National Death Index (NDI) and state death indices provides information on all causes of
death expanding the utility of the MEC (30).
As a population-based cohort with an adequate response rate, the results from the MEC
are generalizable to a large segment of the population. Importantly, we can expect the
exposure profiles of the cohort to be representative of the population racial/ethnic
groups included in this study (31).
6
1.3.3 Issues of a cohort study
Since the initial data collection period, baseline covariates (diet, medication use, and
exercise) can change, resulting in negative and/or positive bias (2). Follow-up
questionnaires to track changes in covariates requires adequate response rates; in the
MEC, almost ¾ of the cohort completed the initial follow-up questionnaire.
While additional information can be collected on follow-up questionnaires, analysis of
the cohort must take into account a limited observational time. For example, in chapter
3, we examine the effect of cholesterol medication use on the risk of cardiovascular
disease. Not present in the baseline questionnaire, this information was fortunately
collected upon follow-up, necessitating the use of only the subset of individuals who
answered the survey and limiting the observational time that can be used in analysis,
leading to reduced power to detect risk of disease (2).
The MEC is a large cohort, and covariate information was collected via self-administered
questionnaires, leading to the potential for recall bias. Using self-reported data along
with changes in diet over time may result in exposure misclassification (31). The MEC
has performed calibration studies using 260 individuals to validate the questionnaire’s
dietary data, as dietary information is poorly reported (22,32). However, information
about history of physician-diagnosed conditions, prior medical history, and family
history may still retain recall bias. This recall bias is likely to be non-differential with
respect to most diseases as all covariates were collected prior to diagnosis.
7
1.4 Goals of this dissertation
In this dissertation we examine three common chronic diseases with differing rates of
incidence and mortality by race/ethnicity. We will take a look at how the prevalence of
various risk factors also differs by race/ethnicity, and examine the relationship between
disparities in risk of chronic disease and the association with the respective risk factors.
First we will examine the prevalence of atopic allergic conditions and their effect on CRC
risk across five racial/ethnic populations. Then we will look at racial/ethnic specific
differences in cardiovascular disease and stroke risk and how known risk factors
contribute to disparities in risk. Finally, we will examine gallbladder disease in a genetic
context, using the MEC to replicate known and identify novel risk alleles in multiple
ethnic populations.
1.5 Colorectal cancer and atopic conditions
1.5.1 Colorectal cancer epidemiology
Colorectal cancer (CRC) is an uncontrolled neoplasm of the epithelial lining of the gut
(33). Initial symptoms of CRC include abdominal pain, change in bowel habits, weight
loss, nausea and vomiting, and fatigue. Patients may present with any of those non-
specific symptoms, in addition to anemia, perianal symptoms, or bright red blood per
rectum (34). Only about 39% of CRCs are diagnosed when the cancer is limited to the
primary site (Stage I/IIb) (35). Early detection and prevention is difficult, as the criteria
for determining screening modality are numerous and the methods of screening are
unpleasant. Recommendations for screening are continually changing, and involve the
8
use of regular colonoscopies, sigmoidoscopies, CT scans, and fecal occult blood tests
(gFOBT) (36). The current guidelines advocate the use of gFOBT annually, sigmoidoscopy
every 5 years, or complete colonoscopy every 10 years in individuals 50-75. The task
force does not recommend CT colonoscopies at this time.
CRC mortality differs by ethnicity; as of 2012, African-Americans have a CRC rate of
21.4/100,000 persons compared to 15.0/100,000 persons in Whites (37 Table 6,38).
Hispanic and Asian/Pacific Islanders have reduced incidences of CRC, 12.2/100,000
persons and 11.0/100,000 persons, respectively. Thirty-five percent of African
Americans with CRC develop right-sided cancers, while 36% develop sigmoid cancers,
compared to 24% and 60% respectively in Whites. Migration patterns and lifestyle
factors can also affect ethnic differences in CRC incidence and mortality, as Japanese
Americans appear to have double the incidence rate of CRC as Japanese individuals born
and living abroad, though the rates of disease in Japanese living in Japan are rapidly
increasing (39,40).
CRC is a cancer that develops due to prolonged exposure to cytotoxic agents in the
environment. Family history, history of polyps, BMI, smoking status, alcohol
consumption, red meat, fiber and other dietary factors are all important risk factors of
CRC, which differ by race/ethnicity(41,42). Differences in screening and socioeconomic
status may also contribute to racial/ethnic differences (43,44). In an analysis in the MEC,
adjustment for these risk factors left increased risk in Japanese American men (RR 1.27)
and women (RR 1.49), and African American women (RR 1.48) (42).
9
1.5.2 Pathogenesis and Biology of CRC
The different types of CRC that can arise in the gut include adenocarcinomas, carcinoid
tumors, gastrointestinal stromal tumors, lymphomas, and squamous cancers. 95% of
CRCs are adenocarcinomas arising from intestinal glands (41). Pathologically, CRCs
appear ulcerated, infiltrating, and are poorly differentiated in late stages. Approximately
6-9% of CRCs are hereditary, while 73% are sporadic and 20% have a familial link (41).
Hereditary CRC originates from syndromes such as familial adenomatous polyposis (FAP)
where hundreds of adenomas form throughout the digestive system, or hereditary non-
polyposis colorectal cancer (HNPCC, or Lynch syndrome) (45,46). Sporadic colorectal
adenocarcinoma can arise from small serrated or villous adenomas, and can take
upwards of 10 years to develop into an infiltrating carcinoma.
The pathogenesis of CRC involves the accumulation of multiple somatic mutations
resulting in unchecked neoplastic growth. Mutations in oncogenes (Ki-ras), tumor-
suppressor genes (APC, DCC, p53, MCC), and DNA mismatch repair genes all contribute
to CRC development. The current model of CRC pathogenesis is thought to be the loss of
heterozygosity, in which normal mucosa turns in to hyperproliferative epithelium
(through APC or beta-catenin mutations), which can result in adenoma formation (via Ki-
ras and other mutations), with a final progression to adenocarcinoma (via p53
mutations) (35,47). An alternative pathway involves regions of microsatellite instability
that form due to defects in mismatch repair genes, which can account for approximately
20% of tumors (48).
10
Exposure to external antigens and pathogens introduced through the diet can result in
increased rates of mutation, which may slip by the DNA repair mechanisms resulting in
the defects noted above. To counter this, the gut itself contains its own distinct immune
system, known as gut-associated lymphoid tissue (GALT) (49). The physiological
inflammatory response within the gut is a highly regulated system that develops based
on the normal intestinal flora, and is activated by external antigens (50). As a
consequence, the epithelial layer of the gut is the site of heavy immune activity and
rapid tissue turnover. As increased and persistent inflammation can be associated with
cancer development, it is important to examine the relationship of inflammatory
processes with carcinogenesis (51).
1.5.3 Inflammation and Cancer
Inflammation has a very important role in cancer, and can have both carcinogenic and
anti-carcinogenic effects (52, chap.9). Hepatitis, an inflammatory condition of the liver
with many causes such as infection and alcoholism, is a risk factor for liver cancer.
Similarly, inflammatory bowel disease is a risk factor for CRC. One pathway of tumor
initiation due to chronic inflammation begins with irreversible DNA damage (which
could be epigenetic damage as well) initiated by reactive oxygen species (ROS) and
reactive nitrogen species (RNS) released by macrophages persistently activated at the
site of inflammation. Constitutive activation of NF-κB in cancer tissue can promote anti-
apoptotic and pro-neoplastic genes leading to cancer initiation and progression (53,
chap.19). However, there are different types of chronic inflammation which can play
different roles in carcinogenesis, including atopic inflammation.
11
1.5.4 Atopic allergic conditions
Atopic allergic conditions (AAC) are a complex set of diseases that involve the
hyperactive response of the immune system towards an environmental antigen. AAC
manifest in the body in different ways. In an acute asthmatic event, increased
inflammation in response to an environmental antigen leads to bronchial inflammation
and constriction. Edema, vasodilation, and increased mucous production can occur with
allergic rhinitis. The immune system reacts to three targets: autoimmune targets (a
pathological reaction against the self, misidentified as a foreign antigen), microbial
pathogens, and environmental allergens/triggers. Acute reactions to allergies are known
as a type I or type IV hypersensitivity response (54, chap.4). Exposure of environmental
antigens to antigen-presenting cells (APC) leads to the activation of T-helper cell type 2
(Th2), which leads to the release of specific cytokines. These cytokines promote
eosinophil production in the bone marrow, increased uptake of immune cells in the
peripheral tissue, and differentiation by B-cells to release the hallmark immunoglobulin
E (IgE). The IgE molecules bind to environmental antigens, and result in sensitization of
mast cells and eosinophils. Further exposure to allergens can cause degranulation of the
mast cells, leading to the release of histamine, TNF-a, and other cytokines that cause
edema and constriction. Additional excessive activation of eosinophils results in
increased mucosal secretions and airway hyper-responsiveness, a characteristic of
asthma (55). Hyperreactivity and extraneous recruitment of eosinophils and neutrophils
to the site of antigenic attachment can also lead to a late phase reaction, which can
cause additional damage to the tissues (56, chap.11).
12
1.5.5 Prevalence of AAC
The prevalence of AAC also differs amongst various racial/ethnic populations. According
to the National Health Information Survey (NHIS) from 2009-2011, Black children had a
higher prevalence of skin allergies (17.4%) and a lower prevalence of respiratory
allergies (15.6%) in comparison to white children (12.0% and 19.1% respectively).
Hispanic children had lower prevalence of both type of allergies (10.1% for skin allergies
and 13.0% for respiratory allergies) (57). A meta-analysis examining racial/ethnic
differences in food allergies using IgE measurement found that Black children were
more likely to have food allergies than White children (58,59). Genome wide association
studies (GWAS) are currently underway to understand the genetic basis for these
racial/ethnic differences, though very few validated markers have been demonstrated
that account for these differences in AAC prevalence (60,61).
In a review of atopic inflammatory diseases and the top three cancers (prostate, breast,
and colorectal), Vojtechova argues that the association of allergies with breast and CRC
is negligible, and that a positive association exists with prostate cancer. They ultimately
conclude that “there is little epidemiological support for the immune surveillance theory
or antigenic stimulation theory” (62). However, some analyses were improperly
conducted, combining incidence and mortality estimates for inappropriate pooled
estimates. In contrast, in the CPS-II cohort, Turner et al. reported that individuals with
atopic allergic inflammatory conditions experienced a 14% reduction in risk of CRC [95%
CI: 0.64-0.91] (63). These could ultimately account for the disparities in CRC risk as
discussed above. As such, a recent review by Turner et al. argues that asthma, allergy,
13
and/or increased IgE levels were inversely associated with meningiomas, gliomas, non-
Hodgkin’s lymphomas (also shown in the MEC), and pancreatic cancer (64,65).
1.5.6 AAC and CRC in the MEC
In chapter 2, we examine the association between AAC and CRC in five racial/ethnic
populations. We hypothesize that AAC is associated with an inverse association with
CRC incidence and mortality. This is the first multiethnic study to examine this
association, as previous studies (such as the in the CPS-II cohort) were performed in
primarily White populations (66). This study will provide additional insight on the
association between hyperactive inflammation on cancer development.
1.6 Racial/ethnic differences in cardiovascular disease and stroke
1.6.1 Cardiovascular disease and stroke
Cardiovascular disease and stroke are the leading causes of morbidity and mortality
both nationally and globally. In 2013, cardiovascular diseases (CVD) claimed 611,105
lives, while cerebrovascular diseases claimed 128,978 lives, per the National Center for
Health Statistics (NCHS) (33). As of 2006 (per CDC 2011), Blacks had an age-adjusted
coronary heart disease rate of 161.6/100,000 persons per year, compared to Whites
with 134.2/100,000 persons per year, and Asian/Pacific Islanders with 77.1/100,000
persons per year. Hispanics had a much lower rate (106.4/100,000 persons per year)
compared to non-Hispanics (136.8/100,000 persons per year) (67). In chapter 3, we
attempt to determine if these disparities in disease can be attributed to known risk
factors.
14
Atherosclerotic disease is a chronic condition that builds up over a long period of time,
precipitating in an acute myocardial infarction (AMI) or stroke. Over decades, plaques
build up in the intima of vascular walls, beginning with endothelial dysfunction and
infiltration of LDL and other lipogenic particles. Oxidization and consumption of lipid
particles by leukocytes results in the formation of foam cells. These foam cells release
matrix metalloproteases and other cytokines that damage the matrix membrane,
allowing infiltration of smooth muscle from the media to the intima. The smooth muscle
cells develop a fibrous cap through the calcification and generation of additional matrix
proteins, and a plaque is formed. Rupture or damage of this cap leads to a cascade of
events resulting in a thrombus, which can block coronary or cerebrovascular arteries,
resulting in a heart attack or stroke (54,68).
The final acute outcome is the result of a long-term process, the pathogenesis of which
differs by race/ethnicity. Here we will examine the acute outcomes of AMI, other heart
diseases, and stroke and how they differ by race/ethnicity.
1.6.2 Risk factors for CVD
There are numerous established risk factors for atherosclerotic disease, including
smoking, diabetes mellitus, family history and genetics, and other metabolic disorders
such as hyperlipidemia and hypertension (69). The prevalence of these risk factors
differs by race/ethnicity, and contributes to pathogenesis by exacerbating the process
outlined above (20,70). For example, smoking not only increases the risk of endothelial
damage, but also enhances the inflammatory response in a vessel wall upon interaction
15
with mast cells (71). Protective factors include use of hormone replacement therapy and
increased physical activity (68).
Blood serum cholesterol levels (carried by low density lipoproteins (LDL), high density
lipoproteins (HDL), and other lipoproteins in triglycerides) are associated with dietary
intake and physical activity, and can in part determine cardiovascular disease risk (72–
74). Individual lipid levels can therefore be analyzed as an additional biomarker of
exposure. Elevated lipid levels have been shown to have detrimental effects on coronary
artery hardening and thickening, exacerbating atherosclerotic disease (20,75–79).
Cholesterol medication (fibrates, statins, and others) reduces serum lipid levels, but
prescription rates differ by ethnicity (80,81). Unfortunately, the benefits of the use of
statins in primary prevention to reduce morbidity and mortality is still being debated
(74,82).
These modifiable risk factors are responsible for a large portion of disease (83). Many
patients are resistant to change, and more evidence based research is necessary to help
physicians better inform patients (70). Adherence and response to lifestyle risk factor
modifications differs by race/ethnicity. The AHA continues to examine the differences in
prevalence of these risk factors, and more research is necessary to understand the
effect on racial/ethnic differences in risk of disease (70,84,85).
1.6.3 CVD and known risk factors in the MEC
In chapter 3 we examine racial/ethnic disparities in mortality of AMI, other heart
disease (including acute/sub-acute CVD, healed MI or subsequent MI, angina,
16
atherosclerosis or aneurysm, cardiomyopathy, signal block, flutters and arrhythmias,
heart failure) and stroke (ischemic and hemorrhagic), and how known risk factors
account for some of the differences. Known risk factors here include: body mass index
(BMI), hypertension, diabetes, smoking (in pack-years and never/past/current), alcohol
consumption, physical activity, education (as a proxy for socioeconomic status), dietary
intake factors, and reproductive health factors in women (69,82,86–96). Initially, a crude
examination of cause-specific mortality rates within the MEC revealed great differences
by race/ethnicity (97). A prior study in the MEC showed that established risk factors (as
discussed above) account for most of the racial/ethnic specific differences in CVD-
related mortality, though there were a small number of cases (5). However, some
differences in risk did remain, for example, a 51% reduced risk of AMI in Japanese
American males, and a 26% reduced risk of AMI in Latino males, compared to White
males. Here we reevaluate the effect of these known risk factors on racial/ethnic
differences in CVD risk, with >14 years of follow-up. In addition, we also examine the
effect of known risk factors on racial/ethnic differences in stroke-related mortality
which has yet to be examined in the MEC.
1.7 Gallbladder disease: exploring genetic variation by ethnicity
1.7.1 Gallbladder disease epidemiology
Gallbladder disease (GBD) is an increasingly prevalent set of diseases including
cholelithiasis and cholecystitis. Though often asymptomatic, the burden of GBD has
increased to over 20% in the past few decades, is responsible for almost 2 million
17
ambulatory care visits, and has resulted in over 750,000 elective cholecystectomy
operations in the United States (98). The prevalence of GBD by race/ethnicity ranges
between 10%-40%, with North American Indian, Chilean, and Latino women having
increased rates of GBD compared to Asians and African Americans (99,100).
1.7.2 Gallbladder biology and pathology
The two types of gallstones include cholesterol and pigment stones. Gallstones form due
to abnormal biliary composition, secretion, or reabsorption. Supersaturation of
cholesterol compared to bile acids or phospholipids results in nucleation and crystal
precipitation. Any risk factors which increase cholesterol secretion (such as high-caloric
diets), or change the consistency of bile into a sludge (pregnancy and abnormal
gallbladder motor function) can increase the risk of cholelithiasis (101). Pigment stones,
while less common, are exacerbated by the secretion of unconjugated bilirubin stuck in
the biliary tree (54, chap.15)
Cholecystitis usually occurs due to cystic duct obstruction by a gallstone. Pain and
inflammation occur due to mechanical pressure on the duct, the release of various
tissue factors, or bacterial infection during cholestasis. In some cases, GBD can progress
to gallbladder carcinoma, which can be lethal due to its late presentation (102,103).
Despite this, prophylactic cholecystectomy in the presence of asymptomatic gallstones
is not recommended.
18
1.7.3 GBD risk factors
Risk factors for GBD include increased age, female sex, obesity, high caloric intake,
dyslipidemia, metabolic syndrome, cholesterol medication use, and
conditions/medications that decrease bile reabsorption (98,104). These risk factors have
also been confirmed within the MEC, smoking, red meat and cholesterol consumption
were significant risk factors while alcohol use, fruit and fiber intake were protective
factors (6). Twin and family-based studies have shown that the additive genetic
heritability of gallstones ranges from 15%-25% (99,105). Genetic variants associated
with GBD have been reported in a few racial/ethnic populations. There is one major
locus in the ABCG5/8 cholesterol transporter gene, and the most widely replicated
single nucleotide polymorphism (SNP) rs11887534 has been reported in German,
Chilean, and Chinese populations(106–108). The Epidemiologic Architecture for Genes
Linked to Environment (EAGLE) study examined 49 lipid trait-associated GWAS markers
in 852 cases and 4980 controls, discovering 13 significantly associated SNPs distributed
amongst Whites, Blacks, and Mexican Americans (109). Two additional loci were
reported in a study of ~15,000 women of European ancestry in ABCG5/8 (110). A meta-
analysis using a priori candidates for bilirubin secretion highlighted the UGT1A1 locus
(associated with Gilbert syndrome) as associated with GBD (106). Additional analyses
are necessary in a multiethnic population in order to replicate these known SNPs,
elucidate novel variants associated with GBD susceptibility and development, and
account for disparities in GBD risk across racial/ethnic populations.
19
1.7.4 Genetic variation
A single nucleotide polymorphism (SNP) is a mutation in the genetic code distributed
approximately every 1 kilobases, usually with a frequency of >1% in the general
population. SNPs can be synonymous (resulting in no change in the final translated
protein), or nonsynonymous. Nonsynonymous SNPs may or may not alter the final
protein structure, or even halt the translation of a transcripted mRNA strand (missense
mutation) (111, chap.4 – Human Genetic Diversity). The resultant mutations can affect
many biological functions, by altering protein transcription, folding, and function, and
are the sources of inherited susceptibility to disease. Most variants do not cause actual
disease, but may be associated markers for functional mutations.
There are many methods of identifying genomic variants that result in increased
susceptibility to disease. Candidate gene studies examine specific genetic loci that are a
priori hypothesized to cause disease (112,113). Linkage studies use related individuals to
discover highly penetrant alleles as opposed to association studies (111, chap.10-
Identifying the Genetic Basis for Human Disease).
Genome wide association studies scan the entire genome for low-penetrant
polymorphisms that are responsible for disease. Unfortunately, it is very costly to
genotype the entire human genome and every SNP of interest. Instead, marker SNPs are
selected for genotyping that are in linkage disequilibrium with many other SNPs. During
meiotic cellular divisions, blocks of genetic code recombine between chromosomes,
resulting in SNPs that “travel together” in blocks. Variants that are non-randomly
recombined together are said to be in linkage disequilibrium (LD) with each other. After
20
genotyping, the remaining SNPs are imputed via IMPUTE2, which uses a reference panel
such as the multiethnic 1000 Genomes Project or the HapMap project (114–116).
SNP allele frequencies can differ widely by race/ethnicity. A multiethnic population will
allow for the discovery of SNPs that are more common in a particular racial/ethnic
group (117–121). In addition, within a particular racial/ethnic group, cryptic relatedness
may exist which can result in hidden population structures that may mask SNPs with
true associations, or result in associations that are not real.
1.7.5 Genetic analysis in the MEC
The MEC has been used numerous times before to identify ethnic-specific risk variants
and the effect of racial/ethnic differences in allele frequencies on disparities in risk.
Examination of prostate cancer cases revealed African American specific risk variants in
the 8q24 susceptibility region (122–124). Racial/ethnic differences in prevalence of T2D
risk alleles were found to not explain differences in prevalence of diabetes (120). Latino
individuals in the MEC were used in the SIGMA consortium to identify variants in
SLC16A11 as risk factors for T2D in Mexico (125).
In chapter 4, we attempt to perform an agnostic genome-wide association study to
identify genetic risk variants for GBD in a multiethnic population (112). As discussed
previously racial/ethnic differences in risk of GBD exist, and some of the risk of GBD are
associated with family history and genetic effects. Here we use three races/ethnicities in
the MEC to replicate known GBD risk loci, and attempt to identify both ethnic-specific
and rare risk alleles. By using a large population, we will have additional power with
21
which to detect risk alleles with small effects. Using a multiethnic population allows us
to identify risk alleles that may be common in one racial/ethnic population, and rare in
other populations. We hope to identify risk alleles that not only contributes to genetic
GBD risk, but accounts for racial/ethnic differences in risk as well. We use samples from
the MEC as well as the Slim Initiative in Genomic Medicine for the Americas (SIGMA)
Type 2 Diabetes consortium to cover three different racial/ethnic groups.
1.8 Summary
The risk of incidence and mortality due to CRC, CVD, stroke, and GBD all differ by
race/ethnicity. In addition, the prevalence of risk factors for those diseases also differ by
race/ethnicity. The primary aim of this dissertation is to examine common chronic
conditions that vary in frequency across racial and ethnic populations in the MEC, and
attempt to reveal novel risk factors or the degree to which known risk factor might
explain disparities in risk across Whites, African Americans, Native Hawaiians, Japanese
Americans, and Latinos in the MEC. Here we use a wide range of analytic techniques
including a prospective cohort analysis of incidence when analyzing AAC and CRC, a
mortality analysis when examining CVD, and a GWAS to discover risk variants for GBD.
In chapter 2, we look at allergic rhinitis, atopic asthma, and allergies, collectively known
as atopic allergic conditions (AAC). Specifically, we examine the association between
AAC and CRC incidence and mortality in each of the races/ethnicities in the MEC.
Chapter 3 examines the risk of CVD and stroke, and the contribution of known risk
factors towards differences in risk of disease by race/ethnicity. These covariates include
22
lifestyle risk factors, comorbidities, medication usage, and dietary factors. We also
examine lipid biomarkers within each of the five racial/ethnic groups in the MEC, and
whether or not the patterns of lipids in each population match differences in CVD and
stroke risk.
In chapter 4 we take an agnostic approach to search for genetic risk loci of gallbladder
disease (GBD). Here we study multiple racial/ethnic populations to examine common
and ethnic-specific risk alleles for the development of GBD, to both replicate known risk
variants and discover novel signals, and understand why GBD differs across racial/ethnic
populations.
Finally, Chapter 5 will summarize the differences in risk by ethnicity we have examined
so far, and the potential for future work.
1.9 References
1. Rothman KJ. Epidemiology: An introduction. 2nd ed. New York, NY: Oxford
University Press; 2012 268 p.
2. Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. 2008.
3. Goodman R a, Posner SF, Huang ES, et al. Defining and measuring chronic
conditions: imperatives for research, policy, program, and practice. Prev. Chronic
Dis. [electronic article]. 2013;10(Mcc):E66.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3652713&tool=pmc
entrez&rendertype=abstract)
4. Siegel RL, Miller KD, Jemal A. Cancer Statistics, 2015. CA Cancer J Clin.
2015;65(1):5–29.
5. Henderson SO, Haiman CA, Wilkens LR, et al. Established Risk Factors Account for
Most of the Racial Differences in Cardiovascular Disease Mortality. PLoS One
[electronic article]. 2007;2(4):e377.
(http://dx.plos.org/10.1371/journal.pone.0000377)
23
6. Figueiredo J, Setiawan VW, Porcel J, et al. Gallbladder Disease risk factors. Curr.
Draft.
7. US Census Bureau. 1940-2010: How Has America Changed?
2012;(http://www.census.gov/library/infographics/1940_census_change.html)
8. Dawber TR, Meadors GF, Moore FE. Epidemiological approaches to heart disease:
the Framingham Study. Am. J. Public Health. 1951;41(3):279–281.
9. Bernstein L, Allen M, Anton-Culver H, et al. High breast cancer incidence rates
among California teachers: Results from the California Teachers Study (United
States). Cancer Causes Control. 2002;13(7):625–635.
10. Calle EE, Rodriguez C, Jacobs EJ, et al. The American Cancer Society Cancer
Prevention Study II Nutrition Cohort: rationale, study design, and baseline
characteristics. Cancer [electronic article]. 2002;94(9):2490–501.
(http://www.ncbi.nlm.nih.gov/pubmed/12015775). (Accessed January 10, 2014)
11. Vaillant GE, Mukamal K. Successful Aging. Am. J. Psychiatry [electronic article].
2001;158(6):839–847.
(http://dx.doi.org/10.1176/appi.ajp.158.6.839\nhttp://psychiatryonline.org/data
/Journals/AJP/3725/839.pdf)
12. Winter AC, Rexrode KM, Lee I, et al. Migraine and subsequent risk of breast
cancer: a prospective cohort study. Cancer Causes Control [electronic article].
2013;24(1):81–89. (http://link.springer.com/10.1007/s10552-012-0092-x)
13. Christen WG, Gaziano JM, Hennekens CH. Design of physicians’ health study II—A
randomized trial of beta-carotene, vitamins E and C, and multivitamins, in
prevention of cancer, cardiovascular disease, and eye disease, and review of
results of completed trials. Ann. Epidemiol. 2000;10(2):125–134.
14. Belanger CF, Hennekens CH, Rosner B, et al. The Nurses’ Health Study. Am. J.
Nurs. [electronic article]. 1978;78(6):1039–1040.
(http://www.jstor.org.libproxy1.usc.edu/stable/3462013)
15. Chaturvedi N. Ethnic differences in cardiovascular disease. Heart.
2003;89(6):681–686.
16. Kabat GC, Kim M, Hunt JR, et al. Body mass index and waist circumference in
relation to lung cancer risk in the Women’s Health Initiative. Am. J. Epidemiol.
[electronic article]. 2008;168(2):158–69.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2878097&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
17. Signorello LB, Cohen SS, Williams DR, et al. Socioeconomic status, race, and
mortality: A prospective cohort study. Am. J. Public Health. 2014;104(12):e98–
e107.
24
18. Cohen SS, Park Y, Signorello LB, et al. A pooled analysis of body mass index and
mortality among African Americans. PLoS One. 2014;9(11).
19. Bandiera FC, Arguelles W, Gellman M, et al. Cigarette smoking and depressive
symptoms among Hispanic/Latino adults: Results from the Hispanic Community
Health Study/Study of Latinos (HCHS/SOL). Nicotine Tob. Res. 2015;17(6):727–
734.
20. Bild DE, Detrano R, Peterson D, et al. Ethnic Differences in Coronary Calcification:
The Multi-Ethnic Study of Atherosclerosis (MESA). Circulation [electronic article].
2005;111(10):1313–1320.
(http://circ.ahajournals.org/content/111/10/1313.abstract)
21. Patel SA, Winkel M, Ali MK, et al. Cardiovascular Mortality Associated With 5
Leading Risk Factors: National and State Preventable Fractions Estimated From
Survey Data. Ann. Intern. Med. [electronic article].
2015;(http://annals.org/article.aspx?articleid=2362308). (Accessed July 1, 2015)
22. Kolonel LN, Henderson BE, Hankin JH, et al. A multiethnic cohort in Hawaii and
Los Angeles: baseline characteristics. Am. J. Epidemiol. [electronic article].
2000;151(4):346–57. (http://www.ncbi.nlm.nih.gov/pubmed/10695593)
23. Waters KM, Le Marchand L, Kolonel LN, et al. Generalizability of associations from
prostate cancer genome-wide association studies in multiple populations. Cancer
Epidemiol. Biomarkers Prev. [electronic article]. 2009;18(4):1285–9.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2917607&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
24. Henderson SO, Bretsky P, Henderson BE, et al. Risk factors for cardiovascular and
cerebrovascular death among African Americans and Hispanics in Los Angeles,
California. Acad. Emerg. Med. [electronic article]. 2001;8(12):1163–72.
(http://www.ncbi.nlm.nih.gov/pubmed/11733295)
25. Conroy SM, Maskarinec G, Wilkens LR, et al. Obesity and breast cancer survival in
ethnically diverse postmenopausal women: the Multiethnic Cohort Study. Breast
Cancer Res. Treat. [electronic article]. 2011;129(2):565–574.
(http://link.springer.com/10.1007/s10549-011-1468-4)
26. Edwards QT, Li A, Pike MC, et al. Patterns of regular use of mammography--body
weight and ethnicity: the Multiethnic Cohort. J Am Acad Nurse Pr. [electronic
article]. 2010;22(3):162–169. (http://www.ncbi.nlm.nih.gov/pubmed/20236401)
27. Lee S, Kolonel L, Wilkens L, et al. Postmenopausal hormone therapy and breast
cancer risk: the Multiethnic Cohort. Int J Cancer [electronic article].
2006;118(5):1285–1291. (http://www.ncbi.nlm.nih.gov/pubmed/16170777)
28. Kolonel LN, Altshuler D, Henderson BE. The multiethnic cohort study: exploring
genes, lifestyle and cancer risk. Nat Rev Cancer [electronic article].
25
2004;4(7):519–527.
(http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&do
pt=Citation&list_uids=15229477)
29. Epplein M, Franke A a, Cooney R V, et al. Association of plasma micronutrient
levels and urinary isoprostane with risk of lung cancer: the multiethnic cohort
study. Cancer Epidemiol. Biomarkers Prev. 2009;18(7):1962–1970.
30. Setiawan VW, Virnig BA, Porcel J, et al. LINKING DATA FROM THE MULTIETHNIC
COHORT STUDY TO MEDICARE DATA: LINKAGE RESULTS AND APPLICATION TO
CHRONIC DISEASE RESEARCH. Am. J. Epidemiol. [electronic article]. 2015;kwv055–
.
(http://aje.oxfordjournals.org.libproxy.usc.edu/content/early/2015/04/04/aje.kw
v055). (Accessed April 29, 2015)
31. Kelsey JL, Whittemore AS, Evans AS, et al. Methods in Observational
Epidemiology. 2nd ed. New York, NY: Oxford University Press; 1996.
32. Stram D, 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.
33. CDC/NCHS. LCWK11. Deaths, percent of total deaths and rank order for 113
selected causes of death and Enterocolitis due to Clostridium difficile, by Hispanic
origin, race for non-Hispanic origin and sex, United States, 2013. 2014 1-15 p.
34. Tsai S, Gearhart SL. 2 - Presentation and Initial Evaluation of Colorectal Cancer.
Elsevier Inc.; 2011 13-19
p.(http://www.sciencedirect.com/science/article/pii/B9781416046868500075)
35. Chu KM. Epidemiology and Risk Factors of Colorectal Cancer. Elsevier Inc.; 2011 1-
11 p.(http://dx.doi.org/10.1016/B978-1-4160-4686-8.50006-3)
36. USPSTF. Final Research Plan: Colorectal Cancer: Screening.
2014;(http://www.uspreventiveservicestaskforce.org/Page/Document/final-
research-plan54/colorectal-cancer-screening2). (Accessed January 8, 2015)
37. Howlader N, Noone AM, Krapcho M, et al. National Cancer Institute SEER Cancer
Statistics Review 1975-2012. Natl. Cancer Inst. [electronic article]. 2015;1975–
2012. (http://seer.cancer.gov/csr/1975_2012/)
38. Haggar FA, Boushey RP. Colorectal cancer epidemiology: Incidence, mortality,
survival, and risk factors. Clin. Colon Rectal Surg. 2009;22(4):191–197.
39. Flood DM, Weiss NS, Cook LS, et al. Colorectal cancer incidence in Asian migrants
to the United States and their descendants. Cancer Causes Control [electronic
article]. 2000;11(5):403–411.
(http://link.springer.com/10.1023/A:1008955722425)
26
40. Fung KYC, Fung K, Ooi CC, et al. Colorectal Carcinogenesis: A Cellular Response to
Sustained Risk Environment. Int. J. Mol. Sci. [electronic article].
2013;14(7):13525–13541. (http://www.mdpi.com/1422-0067/14/7/13525/)
41. Chu KM. Epidemiology and Risk Factors of Colorectal Cancer. Elsevier Inc.; 2011 1-
11 p.
42. Ollberding NJ, Nomura AM, Wilkens LR, et al. Racial/ethnic differences in
colorectal cancer risk: The multiethnic cohort study. Int. J. Cancer [electronic
article]. 2011;129(8):1899–1906. (http://doi.wiley.com/10.1002/ijc.25822)
43. DeLancey JOL, Thun MJ, Jemal A, et al. Recent trends in Black-White disparities in
cancer mortality. Cancer Epidemiol. Biomarkers Prev. [electronic article].
2008;17(11):2908–12. (http://www.ncbi.nlm.nih.gov/pubmed/18990730).
(Accessed January 10, 2014)
44. Laiyemo AO, Doubeni C, Pinsky PF, et al. Race and Colorectal Cancer Disparities:
Health-Care Utilization vs Different Cancer Susceptibilities. JNCI J. Natl. Cancer
Inst. [electronic article]. 2010;102(8):538–546.
(http://jnci.oxfordjournals.org/cgi/doi/10.1093/jnci/djq068)
45. Iacobuzio-Donahue CA, Montgomery E. Gastrointestinal and Liver Pathology. 2nd
ed. Elsevier Saunders; 2012.
46. Giardiello FM. Hereditary Colorectal Cancer and Polyp Syndromes. Elsevier Inc.;
2011 21-30 p.(http://dx.doi.org/10.1016/B978-1-4160-4686-8.50008-7)
47. Bateman a. C. Pathology of serrated colorectal lesions. J. Clin. Pathol. [electronic
article]. 2014;67(10):865–874. (http://jcp.bmj.com/cgi/doi/10.1136/jclinpath-
2014-202175)
48. Padussis JC, Beasley GM, McMahon NS, et al. Neoplasms of the Small Intestine,
Vermiform Appendix, and Peritoneum, and Carcinoma of the Colon and Rectum.
In: Kufe DW, III EF, Holland JF, et al., eds. HOLLAND-FREI CANCER MEDICINE.
Shelton, CT: People’s Medical Publishing House-USA;
2010(http://online.statref.com/Document.aspx?fxId=72&docId=935)
49. Freud AG, Caligiuri MA. The Organization and Structure of Lymphoid Tissues. In:
Kaushansky K, Lichtman MA, Prchal JT, et al., eds. Williams Hematology, 9e. New
York, NY: McGraw-Hill Education;
2015(http://mhmedical.com/content.aspx?aid=1121088799)
50. Dotan I, Mayer L. Mucosal Immunology and Vaccine Development. In: Sleisenger
and Fordtran’s Gastrointestinal and Liver Disease. Elsevier Inc.; 2016:16–
27.e8.(http://dx.doi.org/10.1016/B978-1-4557-4692-7.00002-8)
51. Laskin D, Laskin J. Inflammation and Cancer. In: Kufe DW, III EF, Holland JF, et al.,
eds. Holland-Frei Cancer Medicine. Shelton, CT: People’s Medical Publishing
House-USA; 2010(http://online.statref.com/Document.aspx?fxId=72&docId=935)
27
52. Kumar V, Abbas AK, Aster JC. Environmental and Nutritional Diseases. In: Robbins
Basic Pathology. 2013:269–307.
53. Mendelsohn J, Gray J, Howley P, et al. The Molecular Basis of Cancer. 4th ed.
Elsevier Saunders; 2015.
54. Kumar V, Abbas AK, Aster JC. Robbins Basic Pathology. 9th ed. Philadelphia:
Elsevier Saunders; 2013.
55. Abbas AK, Lichtman AH, Pillai S. Allergy. In: Cellular and Molecular Immunology.
Elsevier Inc.; 2015:417–435.
56. Abbas AK, Lichtman AH. Basic Immunology. 3rd ed. Elsevier Saunders; 2009.
57. Jackson K, Howie L, Akinbami L. Trends in allergic conditions among children:
United States, 1997-2011. NCHS Data Brief. 2013;121.
58. Greenhawt M, Weiss C, Conte ML, et al. Racial and ethnic disparity in food allergy
in the united states: A systematic review. J. Allergy Clin. Immunol. Pract.
[electronic article]. 2013;1(4):378–386.
(http://dx.doi.org/10.1016/j.jaip.2013.04.009)
59. Keet C a., McCormack MC, Pollack CE, et al. Neighborhood poverty, urban
residence, race/ethnicity, and asthma: Rethinking the inner-city asthma epidemic.
J. Allergy Clin. Immunol. [electronic article]. 2015;135(3):655–662.
(http://linkinghub.elsevier.com/retrieve/pii/S0091674914016765)
60. Barnes K. Genomewide association studies in allergy and the influence of
ethnicity. Curr Opin Allergy Clin Immunol [electronic article]. 2010;10(5):427–433.
(http://www.ncbi.nlm.nih.gov/pubmed/20724922)
61. Barnes KC. Genetic epidemiology of health disparities in allergy and clinical
immunology. J Allergy Clin Immunol [electronic article]. 2006;117(2):243–246.
(http://www.ncbi.nlm.nih.gov/pubmed/16461122)
62. Vojtechova P, Martin RM. The association of atopic diseases with breast,
prostate, and colorectal cancers: a meta-analysis. Cancer Causes Control
[electronic article]. 2009;20(7):1091–1105.
(http://link.springer.com/10.1007/s10552-009-9334-y)
63. Turner MC, Chen Y, Krewski D, et al. Cancer Mortality among US Men and Women
with Asthma and Hay Fever. Am. J. Epidemiol. [electronic article].
2005;162(3):212–221.
(http://aje.oxfordjournals.org/cgi/doi/10.1093/aje/kwi193). (Accessed January
10, 2014)
64. Erber E, Lim U, Maskarinec G, et al. Common immune-related risk factors and
incident non-Hodgkin lymphoma: The multiethnic cohort. Int. J. Cancer [electronic
article]. 2009;125(6):1440–1445. (http://doi.wiley.com/10.1002/ijc.24456)
28
65. Turner MC. Epidemiology: allergy history, IgE, and cancer. Cancer Immunol
Immunother [electronic article]. 2012;61(9):1493–1510.
(http://www.ncbi.nlm.nih.gov/pubmed/22183126)
66. Jacobs EJ, Gapstur SM, Newton CC, et al. Hay Fever and Asthma as Markers of
Atopic Immune Response and Risk of Colorectal Cancer in Three Large Cohort
Studies. Cancer Epidemiol. Biomarkers Prev. [electronic article]. 2013;22(4):661–
669. (http://cebp.aacrjournals.org/cgi/doi/10.1158/1055-9965.EPI-12-1229)
67. Keenan NL, Shaw KM, CDC. CDC Health Disparities and Inequalities Report -
United States, 2011. MMWR Suppl. [electronic article]. 2011;60(1):1–2.
(http://www.ncbi.nlm.nih.gov/pubmed/21430612)
68. Strom JB, Libby P. Atherosclerosis. In: Lilly LS, ed. Pathophysiology of Heart
Disease. Lippincott, Williams & Wilkins; 2011
69. Kim ESH, Gornik HL. Atherosclerotic Risk Factors. Seventh Ed. Elsevier Inc.; 2010
416-428.e3 p.(http://dx.doi.org/10.1016/B978-1-4557-5304-8.00027-3)
70. Mozaffarian D, Benjamin EJ, Go AS, et al. Heart Disease and Stroke Statistics—
2015 Update. Circulation [electronic article]. 2015;131(4):e29–e322.
(http://circ.ahajournals.org/cgi/doi/10.1161/CIR.0000000000000152)
71. Barua RS, Sharma M, Dileepan KN. Cigarette smoke amplifies inflammatory
response and atherosclerosis progression through activation of the H1R-TLR2/4-
COX2 axis. Front. Immunol. 2015;6(NOV).
72. 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(5):1146–1155.
73. Balk EM, Lichtenstein AH, Chung M, et al. Effects of omega-3 fatty acids on serum
markers of cardiovascular disease risk: A systematic review. Atherosclerosis
[electronic article]. 2006;189(1):19–30.
(http://www.sciencedirect.com/science/article/pii/S0021915006000694)
74. Baigent C, Keech A, Kearney PM, et al. Efficacy and safety of cholesterol-lowering
treatment: prospective meta-analysis of data from 90,056 participants in 14
randomised trials of statins. Lancet [electronic article]. 2005;366(9493):1267–78.
(http://www.ncbi.nlm.nih.gov/pubmed/16214597\nhttp://www.thelancet.com/j
ournals/lancet/article/PIIS0140-6736(05)67394-1/fulltext)
75. Arsenault BJ, Boekholdt SM, Kastelein JJP. Lipid parameters for measuring risk of
cardiovascular disease. Nat. Rev. Cardiol. [electronic article]. 2011;8(4):197–206.
(http://dx.doi.org/10.1038/nrcardio.2010.223)
76. Angelantonio E Di, Gao P, Pennells L, et al. Lipid-Related Markers and
Cardiovascular Disease Prediction. Jama. 2014;307(23):2499–2506.
29
77. Collaboration* TERF. Major lipids, apolipoproteins, and risk of vascular disease.
Jama [electronic article]. 2009;302(18):1993–2000.
(http://dx.doi.org/10.1001/jama.2009.1619)
78. Howard B V, Robbins DC, Sievers ML, et al. LDL cholesterol as a strong predictor
of coronary heart disease in diabetic individuals with insulin resistance and low
LDL: The Strong Heart Study. Arterioscler. Thromb. Vasc. Biol. 2000;20:830–835.
79. Gordon T, Castelli WP, Hjortland MC, et al. High density lipoprotein as a
protective factor against coronary heart disease. Am. J. Med. 1977;62(5):707–
714.
80. Mochari-Greenberger H, Liao M, Mosca L. Racial and ethnic differences in statin
prescription and clinical outcomes among hospitalized patients with coronary
heart disease. Am. J. Cardiol. [electronic article]. 2014;113(3):413–7.
(http://www.sciencedirect.com/science/article/pii/S0002914913021395).
(Accessed June 30, 2015)
81. Lewey J, Shrank WH, Bowry ADK, et al. Gender and racial disparities in adherence
to statin therapy: a meta-analysis. Am. Heart J. [electronic article].
2013;165(5):665–78, 678.e1.
(http://www.sciencedirect.com/science/article/pii/S0002870313001385).
(Accessed June 30, 2015)
82. Taylor F, Huffman MD, Macedo AF, et al. Statins for the primary prevention of
cardiovascular disease. Cochrane database Syst. Rev. [electronic article].
2013;1:CD004816. (http://www.ncbi.nlm.nih.gov/pubmed/23440795). (Accessed
January 31, 2015)
83. Yusuf S, Hawken S, Ôunpuu S, et al. Effect of potentially modifiable risk factors
associated with myocardial infarction in 52 countries (the INTERHEART study):
case-control study. Lancet [electronic article]. 2004;364(9438):937–952.
(http://linkinghub.elsevier.com/retrieve/pii/S0140673604170189)
84. Kurian AK, Cardarelli KM. Racial and ethnic differences in cardiovascular disease
risk factors: a systematic review. Ethn. Dis. 2007;17(1):143–152.
85. Lloyd-Jones D, Adams RJ, Brown TM, et al. Executive summary: Heart disease and
stroke statistics-2010 update: A report from the american heart association.
Circulation. 2010;121(7):46–215.
86. Lewis CE, McTigue KM, Burke LE, et al. Mortality, health outcomes, and body
mass index in the overweight range: A science advisory from the american heart
association. Circulation. 2009;119(25):3263–3271.
87. Davidson MH. Cardiovascular Risk Factors in a Patient with Diabetes Mellitus and
Coronary Artery Disease: Therapeutic Approaches to Improve Outcomes:
Perspectives of a Preventive Cardiologist. Am. J. Cardiol. [electronic article].
30
2012;110(9):43B–49B.
(http://linkinghub.elsevier.com/retrieve/pii/S0002914912019984)
88. Daviglus ML, Pirzada A, Talavera GA. Cardiovascular disease risk factors in the
Hispanic/Latino population: lessons from the Hispanic Community Health
Study/Study of Latinos (HCHS/SOL). Prog. Cardiovasc. Dis. [electronic article].
2014;57(3):230–6.
(http://www.sciencedirect.com/science/article/pii/S0033062014001042).
(Accessed April 29, 2015)
89. DeFina LF, Haskell WL, Willis BL, et al. Physical Activity Versus Cardiorespiratory
Fitness: Two (Partly) Distinct Components of Cardiovascular Health? Prog.
Cardiovasc. Dis. [electronic article]. 2015;57(4):324–329.
(http://linkinghub.elsevier.com/retrieve/pii/S0033062014001406)
90. Dong JY, Dong J, Zhang YH, et al. Meta-Analysis of Dietary Glycemic Load and
Glycemic Index in Relation to Risk of Coronary Heart Disease. Am. J. Cardiol.
[electronic article]. 2012;109(11):1608–1613.
(http://linkinghub.elsevier.com/retrieve/pii/S0002914912006017)
91. Wolk a, Manson JE, Stampfer MJ, et al. Long-term intake of dietary fiber and
decreased risk of coronary heart disease among women. JAMA [electronic
article]. 1999;281(21):1998–2004.
(http://www.ncbi.nlm.nih.gov/pubmed/10359388)
92. Stramba-Badiale M. Postmenopausal hormone therapy and the risk of
cardiovascular disease. J. Cardiovasc. Med. (Hagerstown). 2009;10(4):303–309.
93. Smyth A, Teo KK, Rangarajan S, et al. Alcohol consumption and cardiovascular
disease, cancer, injury, admission to hospital, and mortality: a prospective cohort
study. Lancet [electronic article]. 2015;386(10007):1945–1954.
(http://linkinghub.elsevier.com/retrieve/pii/S0140673615002354)
94. Schulte H, Cullen P, Assmann G. Obesity, mortality and cardiovascular disease in
the Münster Heart Study (PROCAM). Atherosclerosis [electronic article].
1999;144(1):199–209. (http://www.ncbi.nlm.nih.gov/pubmed/10381293)
95. Rosano G, Vitale C, Spoletini I, et al. Cardiovascular health in the menopausal
woman: impact of the timing of hormone replacement therapy. Climacteric
[electronic article]. 2012;15(4):299–305.
(http://www.ncbi.nlm.nih.gov/pubmed/22424090)
96. Rimm EB, Ascherio a, Giovannucci E, et al. Vegetable, fruit, and cereal fiber
intake and risk of coronary heart disease among men. JAMA [electronic article].
1996;275(6):447–51. (http://www.ncbi.nlm.nih.gov/pubmed/8627965)
97. Wan P. Crude mortality rates in the MEC. 2013.
98. Portincasa P, Moschetta A, Palasciano G. Cholesterol gallstone disease. Lancet.
31
2006;368(9531):230–239.
99. Shaffer E a. Epidemiology of gallbladder stone disease. Best Pract. Res. Clin.
Gastroenterol. 2006;20(6):981–996.
100. Stinton LM, Myers RP, Shaffer E a. Epidemiology of gallstones. Gastroenterol. Clin.
North Am. [electronic article]. 2010;39(2):157–169.
(http://dx.doi.org/10.1016/j.gtc.2010.02.003)
101. Greenberger NJ, Paumgartner G. Diseases of the Gallbladder and Bile Ducts. In:
Kasper D, Fauci A, Hauser S, et al., eds. Harrison’s Principles of Internal Medicine,
19e. New York, NY: McGraw-Hill Education;
2015(http://mhmedical.com/content.aspx?aid=1120812268)
102. Azodo IA, Parks RW, Garden OJ. Epidemiology of Cholangiocarcinoma and
Gallbladder Carcinoma BT - Biliary Tract and Gallbladder Cancer: A
Multidisciplinary Approach. In: Herman MJ, Pawlik MT, Thomas R. Charles J, eds.
. Berlin, Heidelberg: Springer Berlin Heidelberg; 2014:1–
31.(http://dx.doi.org/10.1007/978-3-642-40558-7_1)
103. Mack TM, Menck HR. Epidemiology of Cancer of the Digestive Tract. Boston:
Martinus Nijhoff Publishers; 1982.
104. Stinton LM, Shaffer EA. Epidemiology of gallbladder disease: Cholelithiasis and
cancer. Gut Liver. 2012;6(2):172–187. (http://pdf.medrang.co.kr/ekjg/ekjg006-02-
03.pdf). (Accessed May 6, 2015)
105. Katsika D, Grjibovski A, Einarsson C, et al. Genetic and environmental influences
on symptomatic gallstone disease: A Swedish study of 43,141 twin pairs.
Hepatology. 2005;41(5):1138–1143.
106. Buch S, Schafmayer C, Vlzke H, et al. Loci from a genome-wide analysis of bilirubin
levels are associated with gallstone risk and composition. Gastroenterology
[electronic article]. 2010;139(6):1942–1951.e2.
(http://dx.doi.org/10.1053/j.gastro.2010.09.003)
107. Goodloe R, Brown-Gentry K, Gillani NB, et al. Lipid trait-associated genetic
variation is associated with gallstone disease in the diverse Third National Health
and Nutrition Examination Survey (NHANES III). BMC Med. Genet. [electronic
article]. 2013;14:120.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3870971&tool=pmc
entrez&rendertype=abstract)
108. Katsika D, Magnusson P, Krawczyk M, et al. Gallstone disease in Swedish twins:
Risk is associated with ABCG8 D19H genotype. J. Intern. Med. 2010;268(3):279–
285.
109. Goodloe R, Brown-Gentry K, Gillani NB, et al. Lipid trait-associated genetic
variation is associated with gallstone disease in the diverse Third National Health
32
and Nutrition Examination Survey (NHANES III). BMC Med. Genet. [electronic
article]. 2013;14(1):120. (http://www.biomedcentral.com/1471-2350/14/120)
110. Rodriguez S, Gaunt TR, Guo Y, et al. Lipids, obesity and gallbladder disease in
women: insights from genetic studies using the cardiovascular gene-centric 50K
SNP array. Eur. J. Hum. Genet. [electronic article]. 2015;(April 2014):1–7.
(http://www.nature.com/doifinder/10.1038/ejhg.2015.63)
111. Nussbaum RL, McInnes RR, Willard HF. Thompson & Thompson Genetics in
Medicine. 8th ed. Elsevier Inc.; 2016.
112. Stram DO. Topics in Quantitative Genetics. 2014;31–77.
113. Pulst SM. Genetic linkage analysis. Arch. Neurol. 1999;56(6):667–672.
114. Howie BN, Donnelly P, Marchini J. A flexible and accurate genotype imputation
method for the next generation of genome-wide association studies. PLoS Genet.
2009;5(6).
115. Thorisson G a., Smith A V., Krishnan L, et al. The International HapMap Project
Web site. Genome Res. 2005;15(11):1592–1593.
116. 1KGP. 1000 Genomes Project. (http://www.1000genomes.org/)
117. Haiman CA, Patterson N, Freedman ML, et al. Multiple regions within 8q24
independently affect risk for prostate cancer. Nat. Genet. [electronic article].
2007;39(5):638–644. (http://www.nature.com/doifinder/10.1038/ng2015)
118. Haiman CA, Han Y, Feng Y, et al. Genome-Wide Testing of Putative Functional
Exonic Variants in Relationship with Breast and Prostate Cancer Risk in a
Multiethnic Population. PLoS Genet. [electronic article]. 2013;9(3):e1003419.
(http://dx.plos.org/10.1371/journal.pgen.1003419)
119. Park SL, Caberto CP, Lin Y, et al. Association of Cancer Susceptibility Variants with
Risk of Multiple Primary Cancers: The Population Architecture using Genomics
and Epidemiology Study. Cancer Epidemiol. Biomarkers Prev. [electronic article].
2014;23(11):2568–2578. (http://cebp.aacrjournals.org/cgi/doi/10.1158/1055-
9965.EPI-14-0129)
120. Waters KM, Stram DO, Hassanein MT, et al. Consistent association of type 2
diabetes risk variants found in europeans in diverse racial and ethnic groups. PLoS
Genet. [electronic article]. 2010;6(8).
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2928808&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
121. Watanabe R. Genetic Epidemiology. In: USC - PM599. 2014
122. Al Olama AA, Kote-Jarai Z, Berndt SI, et al. A meta-analysis of 87,040 individuals
identifies 23 new susceptibility loci for prostate cancer. Nat. Genet. [electronic
article]. 2014;46(10):1103–1109.
33
(http://www.nature.com/doifinder/10.1038/ng.3094)
123. Haiman CA, Chen GK, Blot WJ, et al. Genome-wide association study of prostate
cancer in men of African ancestry identifies a susceptibility locus at 17q21. Nat.
Genet. [electronic article]. 2011;43(6):570–573.
(http://www.nature.com/doifinder/10.1038/ng.839)
124. Han Y, Hazelett DJ, Wiklund F, et al. Integration of Multiethnic Fine-mapping and
Genomic Annotation to Prioritize Candidate Functional SNPs at Prostate Cancer
Susceptibility Regions. Hum. Mol. Genet. [electronic article]. 2015;ddv269.
(http://www.hmg.oxfordjournals.org/lookup/doi/10.1093/hmg/ddv269)
125. Williams AL, Jacobs SBR, Moreno-Macías H, et al. Sequence variants in SLC16A11
are a common risk factor for type 2 diabetes in Mexico. Nature [electronic
article]. 2013;506(7486):97–101.
(http://www.ncbi.nlm.nih.gov/pubmed/24390345\nhttp://www.nature.com/doif
inder/10.1038/nature12828)
34
CHAPTER 2: ATOPIC ALLERGIC CONDITIONS AND COLORECTAL
CANCER RISK IN THE MULTIETHNIC COHORT
Authors & Affiliations
Neal A. Tambe, Lynne R. Wilkens, Peggy Wan, Daniel O. Stram, Frank Gilliland, S. Lani
Park, Wendy Cozen, Otoniel Martínez-Maza, Loic Le Marchand, Brian E. Henderson,
Christopher A. Haiman
AUTHOR AFFILIATIONS: Department of Preventive Medicine, Norris Cancer Center,
University of Southern California, Los Angeles, California (Neal A. Tambe, Peggy Wan,
Daniel O. Stram, Frank Gilliland, S. Lani Park, Wendy Cozen, Brian E. Henderson,
Christopher A. Haiman); Department of Epidemiology, University of Hawaii Cancer
Center, Honolulu, Hawaii (Lynne R. Wilkens, Loic Le Marchand); Departments of
Obstetrics & Gynecology and Microbiology, Immunology & Molecular Genetics, David
Geffen School of Medicine at UCLA, Department of Epidemiology, UCLA Fielding School
of Public Health, and Jonsson Comprehensive Cancer Center, University of California, Los
Angeles, CA 90095 (Otoniel Martínez-Maza).
Grant acknowledgement: This work was supported by the National Cancer Institute at
the National Institutes of Health (UM1 CA164973).
35
CORRESPONDENCE: Neal A. Tambe, B.S., ntambe@usc.edu, phone: (408) 489-0129, fax:
(323) 442-7749, address: USC Keck School of Medicine, Norris Medical Hospital, 1441
Eastlake Avenue #4405, Los Angeles, CA 90089
RUNNING HEAD: ATOPIC ALLERGIC CONDITIONS AND COLORECTAL CANCER
2.1 Abstract
Studies have provided evidence of an inverse association between atopic allergic
conditions (AAC) and invasive colorectal cancer (CRC) incidence and mortality in
predominantly White populations. We examined the association between AAC (asthma,
hay fever, or allergy) and CRC among White, African American, Native Hawaiian,
Japanese American, and Latino men and women in the Multiethnic Cohort (MEC), within
Los Angeles and Hawaii. The prospective analysis included 4834 incident CRC cases and
1363 CRC-related deaths ascertained between 1993 and 2010. We examined
associations by ethnicity, location, stage, and potential effect modification by CRC risk
factors. AAC was associated with a reduced risk of CRC incidence among both men and
women (RR = 0.86, 95% CI: 0.80, 0.92). The reduction in risk was noted in all populations
except Latinos, and was significant in Whites (RR = 0.85, 95% CI: 0.73, 0.98), African
Americans (RR = 0.81, 95% CI: 0.70, 0.95), Native Hawaiians (RR = 0.72, 95% CI: 0.54,
0.96), and Japanese Americans (RR = 0.87, 95% CI: 0.78, 0.98). Individuals with AAC also
had a 20% reduction in CRC-related mortality (P = 0.001). These findings provide
evidence for the potential protective role of the reactive immune system in colorectal
cancer.
36
2.2 Introduction
Colorectal cancer (CRC) is one of the most common cancers with respect to
incidence and mortality both globally and nationally (1,2). There are a number of
established environmental and lifestyle risk factors for CRC including diet, body mass
index (BMI), smoking status, and type 2 diabetes, and protective factors including
hormone replacement therapy (HRT) usage, statin and non-steroidal anti-inflammatory
drug (NSAID) usage, and physical activity (3–8). Previous studies have reported
differences in CRC risk amongst racial/ethnic groups after accounting for these various
risk factors, including higher risks for Japanese American men and women, and for
African American women compared to Whites (9).
Atopic allergic conditions (AAC) such as asthma, hay fever, and food allergies
have been examined in connection with cancer risk, as these conditions may indicate a
heightened immune response. A more reactive immune system may contribute to lower
cancer risk. To date, studies have provided support for a protective association of AAC
with pancreatic cancer, glioma, as well as a number of hematologic malignancies, while
a positive association has been noted with lung cancer risk (10–12). Studies investigating
the association of AAC and cancers of the digestive tract (an important location for
manifestation of mucosal inflammation) have had mixed results (12). A meta-analysis of
8 prospective and 8 case-control studies (including both incidence and mortality studies,
published 1985-2006) found no statistically significant evidence of an association
between AAC, asthma, and atopy with CRC risk (13). A study comprised of Taiwanese
individuals reported an inverse association between allergic rhinitis and rectal cancer,
37
but not colon cancer (14). A more recent study within the Cancer Prevention-II (CPS-II)
cohort reported a non-significant 10% decrease in CRC incidence, and statistically
significant 21% decrease in CRC mortality for individuals who have both hay fever and
asthma, with weaker associations for those individuals reporting only one of these
conditions (10,15). The Iowa Women’s Health Study also reported that having two or
more atopic conditions was associated with a 42% reduction in CRC risk (16). Together,
results from these more recent prospective studies have provided evidence for a
protective association of allergic conditions with CRC risk in primarily White populations.
A study is required to examine the association between atopic conditions and CRC in a
multi-ethnic population, as the prevalence of both CRC and atopic diseases differ by
race/ethnicity.
To further understand the relationship between AAC and CRC, we conducted a
prospective analysis among White, African American, Latino, Japanese American and
Native Hawaiian men and women within the Multiethnic Cohort (MEC). In this study, we
examined the association of AAC on CRC risk across populations and whether the
association varies by CRC disease subgroups defined by site, location and stage as well
as known risk factors of this common cancer.
2.3 Methods
The Multiethnic Cohort (MEC) includes over 215,000 men and women from Hawaii and
California, and is comprised primarily of five self-reported racial/ethnic populations:
Whites, African Americans, Latinos, Japanese Americans, and Native Hawaiians (17).
38
Potential cohort members were recruited primarily through Department of Motor
Vehicle license files, using Health Care Financing Administration (Medicare) files for
additional African American individuals. Recruited participants were between 45 and 75
years of age, and completed a detailed, 26-page self-administered questionnaire at
entry into cohort (baseline data, 1993-1996), which included basic demographic
information, lifestyle factors (e.g. diet, exercise, medication use, and smoking history),
and chronic medical conditions. AAC status was defined based on a self-report of having
been previously told by a doctor that the respondent had asthma, hay fever, skin
allergy, food allergy, or any other allergy (asked as a single question on the baseline
questionnaire). Cohort linkage with the Surveillance, Epidemiology, and End Results
(SEER) registries, which covers all of Hawaii and California, was used to identify incident
cancer cases within the cohort, as well as additional details about CRC status (including
location and stage). Deaths within the cohort were determined via linkage with death
certificate files and supplemented with data from the National Death Index (NDI) in both
Hawaii and California. In this study, participants were followed until one of the following
events, the identification of incident CRC cases, death, or end of follow-up (December
31, 2010 in both Hawaii and California). Individuals were excluded if they were not one
of the five main ethnic groups (n=13,988), had a previous diagnosis of CRC by
questionnaire or tumor registry linkage (n=2554), resulting in 199,112 individuals in the
final analysis. 2,554 male and 2,280 female incident cases of invasive CRC were
identified and 735 men and 628 women died of CRC during the follow-up period (1993-
2010).
39
2.3.1 Statistical Analysis
Age-standardized CRC incidence rates were computed based on the number of cases
and person-years accumulated within 5-year age groups, using the US 2000 standard
population. Cox regression was used to estimate all adjusted hazard ratios (reported as
relative risks (RRs)) for the association of AAC with CRC incidence and mortality.
Individuals were censored at the time of death due to other causes, or most-recent
follow-up period. All models were adjusted as strata variables for age at entry into the
cohort, sex, and ethnicity (White, African American, Native Hawaiian, Japanese
American, Latino). The following covariates were included as terms in all log-linear
models: smoking status (never, past, current), pack years (<20, >20, or missing),
education level (<12 years, some college/vocational, college graduate), body mass index
(BMI) (<23, 23 – 24.9, 25 - 29.9, 30 – 34.9, >35kg m
-2
), and aspirin usage (none, past,
current). Individuals missing information for these covariates were assigned a missing
indicator variable in the analysis. Additionally, the following covariates were also
examined: vigorous physical activity, alcohol intake, saturated fat intake, non-saturated
fat intake, dietary fiber intake, energy intake, and a family history of CRC. These
variables were not included in the final model because they did not alter the estimated
relationship between CRC risk and AAC status.
RRs were examined separately for males and females, for colon and rectal cancer, by
site (left vs. right), and by stage (localized vs. regional/distant). Tests of heterogeneity by
tumor characteristic were performed using a fixed effects meta-analysis. Individuals
were excluded from sub-group analyses if they had multiple tumor site diagnoses on the
40
same visit. When examining mortality, sub-group analyses were separated by incident
cancer location only. We also examined whether the associations between AAC status
and CRC risk were heterogeneous by sex, ethnicity, family history, body mass
index(BMI), age, smoking status, aspirin usage and level of education group using
likelihood ratio tests for interaction between AAC and each these variables. Finally, we
also examined AAC status and CRC risk stratified by history of endoscopy (defined as
colonoscopy or sigmoidoscopy), which was self-reported on the second questionnaire
(approximately 6 years after baseline, 1999-2001). This analysis of those who returned
the second questionnaire and answered the question about history of endoscopy
contained 155,915 individuals (78.3% of the cohort), including 2,723 cases that were
diagnosed after the collection of the second questionnaire (56.3% of total cases). All p-
values reported are two-sided.
41
2.4 Results
The mean age at entry for men (n = 89,496) was 60.7 years (SD 8.9), and for women (n =
109,616) was 60.0 years (SD 8.8), and was similar between AAC and non-AAC groups.
The prevalence of AAC varied among ethnicities, from 16% and 24% in the Latino male
and female populations to 24% and 37% in the White male and female populations.
Age-standardized CRC incidence rates were higher in all non-AAC versus AAC groups
(Table 2.1).
In both men and women, AAC individuals were more likely to have higher education
levels (P < 0.001 amongst all ethnicities). Age and race-adjusted family history of CRC
was higher in all AAC groups than non-AAC groups (8.0% vs. 6.9% in men and 9.1% vs.
8.1% in women). In both sexes, AAC sufferers were more likely to be past smokers, and
less likely to be current smokers (P < 0.001 for all ethnic groups). Among both men and
women, those with AAC reported a greater frequency of previous endoscopy than non-
AAC individuals in each population (men: AAC, 34.9% vs. non-AAC, 27.4%, P<0.003;
women: AAC, 32.1% vs. non-AAC, 24.4%, P<0.001; Table 2.1).
42
Table 2.1. Baseline Characteristics of Analyzed Individuals by Sex and AAC Status in the MEC (1993 – 2010)
All Individuals Males only Females only
AAC N AAC % Non-AAC N Non-AAC % AAC N AAC % Non-AAC N Non-AAC % AAC N AAC % Non-AAC N Non-AAC %
No. of subjects 51973 26 147139 74 17997 20 71499 80 33976 31 75640 69
Age at baseline
a
59.38 (8.9) 60.81 (8.8) 59.53 (9.0) 60.97 (8.8) 59.30 (8.8) 60.66 (8.8)
Ethnicity
White 15142 31 33882 69 5472 24 17151 76 9670 37 16731 63
African American 8841 26 25511 74 2227 18 10286 82 6614 30 15225 70
Native Hawaiian 4064 28 10304 72 1307 21 4930 79 2757 34 5374 66
Japanese American 14709 26 41333 74 5538 21 20853 79 9171 31 20480 69
Latino 9217 20 36109 80 3453 16 18279 84 5764 24 17830 76
CRC cases 1053 2.0 3781 2.6 436 2.4 2118 3.0 617 1.8 1663 2.2
CRC incidence rates
b
109.4 131.1 139.0 153.6 94.3 111.1
Family history of CRC
c
8.7 7.5 8.0 6.9 9.1 8.1
Endoscopy Prevalence
c,d,e
Yes 33.1 25.9 34.9 27.4 32.1 24.4
No 47.7 51.6 43.9 47.9 49.8 55.0
BMI
c,e,f
<23 23.8 22.2 16.0 15.7 28.4 28.8
23 – 24.9 17.9 18.1 19.9 19.3 16.8 17.0
25 – 29.9 35.7 38.6 45.1 46.4 30.7 31.1
30 – 34.9 13.9 13.6 13.6 13.5 14.0 13.6
≥ 35 7.1 5.6 4.4 4.0 8.3 7.0
Smoking status
c,e
Never 45.5 42.6 30.3 29.2 53.9 55.5
Past 40.5 38.6 54.4 49.9 32.6 27.6
Current 13.2 17.0 14.8 19.6 12.4 14.6
Educational level
c,e
≤ 12 yrs. 38.4 46.4 34.2 43.0 40.9 49.7
Some college/vocational 31.4 28.0 30.7 28.8 31.8 27.2
College graduate 29.5 24.3 34.6 26.9 26.7 21.6
Physical activity
(hours/day)
c,e
Never 46.9 42.8 31.9 31.1 54.9 53.9
> 0 - 0.21 16.0 15.5 16.1 15.4 15.8 15.5
> 0.21 – 0.71 17.3 17.7 22.8 22.0 14.4 13.6
> 0.71 15.6 18.6 26.4 27.6 9.9 10.0
Aspirin use
c,e
43
Never 57.1 58.2 55.1 56.8 58.3 59.7
Past 18.7 17.2 17.8 17.0 19.1 17.4
Current 20.5 20.0 24.4 22.7 18.3 17.3
AAC, Atopic Allergic Conditions; BMI, Body-mass index; CRC, Colorectal Cancer; SD, Standard Deviation.
a
Expressed as mean (SD)
b
Expressed as
incidence rate, age standardized (5-year groups) to 2000 US standard population with ages 45 years old or above, per 100 000 person-years.
c
Age standardized (5-year age groups) and ethnicity/race standardized to the total population included in the study.
d
Reported on the second
questionnaire (see methods).
e
Does not total to 100% due to missing values.
e
BMI expressed as kg/m
2
.
44
The risk of CRC, after adjustment for potential confounders, was 14% lower among
those with AAC as compared to those without AAC (RR 0.86, 95% CI: 0.80, 0.92; Table
2.2). This inverse relationship was found in each ethnic population, ranging from 0.72
(95% CI: 0.54, 0.96) in Native Hawaiians to 0.96 (95% CI: 0.81, 1.14) in Latinos. The
association was statistically significant in Whites (RR 0.85, P=0.03), African Americans
(RR=0.81, P=0.01) and Japanese Americans (RR 0.87, P=0.01) and the test of
heterogeneity in the risk estimates between ethnicities was not statistically significant
(P eth = 0.36). The association with AAC was observed in both men (RR 0.86, 95% CI: 0.77,
0.95) and women (RR 0.86, 95% CI: 0.78, 0.94, P het=0.94) (Table 2.2).
When stratifying by cancer site, a greater reduction in risk was noted for rectal cancer
(RR 0.78, 95% CI: 0.67, 0.90) and left-sided CRC (RR 0.79, 95% CI: 0.70, 0.91) compared
to right-sided CRC (RR 0.94, 95% CI: 0.85, 1.04, P het for site = 0.04, Table 2.2). An
association with AAC was also found for both localized (RR 0.87, 95% CI: 0.78, 0.96) and
regional CRC (RR 0.84, 95% CI: 0.76, 0.92, P het for stage = 0.64).
In analyses stratified by CRC risk factors, we observed no significant interaction between
AAC status and the following CRC risk factors: age of entry, family history of CRC, body
mass index (BMI), aspirin usage, smoking status, and level of education (Table A.1).
However, there was suggestive evidence of effect modification by smoking status, with
a greater reduction in risk observed in current smokers (RR 0.74, 95% CI: 0.61, 0.90)
compared to past smokers (RR 0.86, 95% CI: 0.77, 0.95) and never smokers (RR 0.90,
95% CI: 0.80, 1.00, P int = 0.23).
45
Table 2.2. Estimated Relative Risk of CRC Incidence Associated with AAC Status by
Ethnicity and Subgroup in the Multiethnic Cohort (1993-2010).
Subgroup AAC Cases
a
Non-AAC
Cases
a
RR
b
95% CI P eth
c
All
All 1053 3781 0.86
d
0.80, 0.92 0.36
White 241 709 0.85 0.73, 0.98
African American 214 783 0.81 0.70, 0.95
Native Hawaiian 63 242 0.72 0.54, 0.96
Japanese American 363 1325 0.87 0.78, 0.98
Latino 172 722 0.96 0.81, 1.14
Men
e
All 436 2118 0.86
d
0.77, 0.95 0.65
White 100 404 0.85 0.68, 1.06
African American 55 340 0.76 0.57, 1.01
Native Hawaiian 26 133 0.79 0.51, 1.20
Japanese American 175 797 0.89 0.76, 1.05
Latino 80 444 0.94 0.74, 1.19
Women
e
All 617 1663 0.86
d
0.78, 0.94 0.49
White 141 305 0.85 0.70, 1.04
African American 159 443 0.84 0.70, 1.01
Native Hawaiian 37 109 0.65 0.45, 0.96
Japanese American 188 528 0.86 0.72, 1.01
Latino 92 278 0.99 0.78, 1.25
Right Colon
f
All 501 1600 0.94
d
0.85, 1.04 0.23
White 120 332 0.88 0.71, 1.09
African American 115 400 0.84 0.68, 1.04
Native Hawaiian 26 83 0.84 0.54, 1.32
Japanese American 153 493 0.99 0.82, 1.19
Latino 87 292 1.19 0.93, 1.52
Left Colon
f
All 290 1122 0.79
d
0.70, 0.91 0.94
White 66 187 0.89 0.67, 1.19
African American 56 207 0.82 0.61, 1.10
Native Hawaiian 21 75 0.81 0.49, 1.32
Japanese American 104 439 0.74 0.59, 0.92
Latino 43 214 0.79 0.56, 1.10
Rectum
f
All 238 977 0.78
d
0.67, 0.90 0.53
White 51 170 0.76 0.56, 1.05
African American 36 149 0.74 0.51, 1.07
Native Hawaiian 14 80 0.48 0.27, 0.85
Japanese American 97 379 0.84 0.67, 1.05
Latino 40 199 0.85 0.60, 1.20
Localized
g
All 481 1683 0.87
d
0.78, 0.96 0.89
White 115 315 0.90 0.72, 1.12
African American 95 311 0.90 0.71, 1.13
Native Hawaiian 31 118 0.72 0.48, 1.07
Japanese American 170 624 0.85 0.72, 1.01
Latino 70 315 0.90 0.69, 1.17
Regional
g
All 532 1970 0.84
d
0.76, 0.92 0.16
White 118 371 0.80 0.65, 0.99
African American 101 430 0.71 0.57, 0.88
46
Native Hawaiian 32 115 0.76 0.51, 1.14
Japanese American 185 674 0.89 0.75, 1.05
Latino 96 380 1.01 0.80, 1.26
AAC, Atopic Allergic Conditions; BMI, Body-mass index; CI, Confidence Interval; CRC,
Colorectal Cancer; RR, Relative Risk; SD, Standard Deviation.
a
Sub-group totals may not
add up to total number of cases as individuals were excluded if they had multiple
tumors at time of diagnosis.
b
Adjusted for age at entry (continuous), smoking status
(status and pack-years), education level, BMI, NSAID use and sex (except for sex-specific
analyses).
c
Adjusted additionally for ethnicity.
d
Test for heterogeneity by ethnicity using
likelihood ratio tests.
e
P het by gender = 0.94 (test for interaction via likelihood ratio test).
f
P het by stage = 0.04 (via fixed effects meta-analysis).
g
P het by location = 0.64 (via fixed
effects meta-analysis).
47
Table 2.3. Estimated Relative Risk of CRC Mortality Associated with AAC Status by
Ethnicity and Subgroup in the Multiethnic Cohort (1993-2010).
Subgroup AAC
Cases
a
Non-AAC
Cases
a
RR
b
95% CI P eth
c
All
All 275 1088 0.80
d
0.70, 0.91 0.14
White 63 226 0.74 0.56, 0.98
African American 70 284 0.74 0.57, 0.97
Native Hawaiian 10 65 0.44 0.23, 0.87
Japanese American 84 297 0.93 0.73, 1.19
Latino 48 216 0.91 0.66, 1.24
Men
e
All 101 634 0.67
d
0.55, 0.83 0.65
White 26 138 0.67 0.44, 1.02
African American 16 125 0.58 0.35, 0.98
Native Hawaiian 4 39 0.40 0.14, 1.13
Japanese American 37 192 0.81 0.56, 1.15
Latino 18 136 0.69 0.42, 1.14
Women
e
All 174 454 0.91
d
0.76, 1.08 0.21
White 37 88 0.82 0.55, 1.20
African American 54 155 0.82 0.60, 1.12
Native Hawaiian 6 26 0.46 0.19, 1.12
Japanese American 47 105 1.09 0.77, 1.55
Latino 30 80 1.15 0.75, 1.75
Right Colon
f
All 128 434 0.89
d
0.73, 1.09 0.39
White 34 90 0.97 0.65, 1.45
African American 33 141 0.68 0.46, 0.99
Native Hawaiian 5 24 0.54 0.20, 1.43
Japanese American 35 105 1.12 0.75, 1.65
Latino 21 74 1.11 0.68, 1.81
Left Colon
f
All 68 259 0.85
d
0.65, 1.12 0.78
White 12 48 0.70 0.37, 1.33
African American 16 57 0.83 0.47, 1.45
Native Hawaiian 3 15 0.73 0.21, 2.57
Japanese American 25 86 0.93 0.59, 1.47
Latino 12 53 0.96 0.51, 1.81
Rectum
f
All 51 278 0.62
d
0.46, 0.84 0.51
White 11 49 0.64 0.33, 1.25
African American 9 52 0.60 0.29, 1.23
Native Hawaiian 1 19 0.13 0.02, 0.97
Japanese American 19 91 0.70 0.43, 1.16
Latino 11 67 0.70 0.37, 1.34
Localized
g
48
All 32 134 0.75
d
0.51, 1.11 0.06
White 9 29 0.88 0.41, 1.87
African American 7 34 0.59 0.26, 1.34
Native Hawaiian 0 10
Japanese American 12 32 1.21 0.62, 2.39
Latino 4 29 0.59 0.21, 1.69
Regional
g
All 209 821 0.80
d
0.69, 0.94 0.36
White 44 163 0.72 0.52, 1.01
African American 49 208 0.71 0.52, 0.98
Native Hawaiian 10 47 0.59 0.30, 1.18
Japanese American 68 242 0.92 0.70, 1.22
Latino 38 161 0.95 0.67, 1.36
AAC, Atopic Allergic Conditions; BMI, Body-mass index; CI, Confidence Interval; CRC, Colorectal
Cancer; RR, Relative Risk; SD, Standard Deviation.
a
Sub-group totals may not add up to total
number of cases as individuals were excluded if they had multiple tumors at time of diagnosis.
b
Adjusted for age at entry (continuous), smoking status (status and pack-years), education level,
BMI, NSAID use and sex (except for sex-specific analyses).
c
Adjusted additionally for ethnicity.
d
Test for heterogeneity by ethnicity using likelihood ratio tests.
e
P het by gender = 0.041 (test for
interaction via likelihood ratio test).
f
P het by stage = 0.015 (via fixed effects meta-analysis).
g
P het
by location = 0.68 (via fixed effects meta-analysis
49
Table 2.4. Estimated Relative Risk of CRC Incidence Associated with AAC Status in Individuals with Self-Reported Endoscopy with
Follow-up Beginning with the Second Questionnaire (1999-2003).
No to endoscopy Yes to endoscopy
Ethnicity Cases (AAC/non-AAC) RR
a
95% CI Cases (AAC/non-AAC) RR
a
95% CI
All Groups 383/1493 0.86 0.77, 0.97 229/618 0.88 0.75, 1.02
White 89/257 0.95 0.75, 1.22 66/138 0.94 0.70, 1.26
African American 70/258 0.86 0.66, 1.12 39/136 0.68 0.47, 0.97
Native Hawaiian 24/99 0.71 0.45, 1.11 7/27 0.57 0.24, 1.33
Japanese American 138/594 0.81 0.67, 0.98 81/210 0.99 0.76, 1.29
Latino 62/285 0.95 0.72, 1.25 36/107 0.93 0.63, 1.36
AAC, Atopic Allergic Conditions; CI, Confidence Interval; CRC, Colorectal Cancer; RR, Relative Risk; SD, Standard Deviation.
a
Adjusted
for sex, age (continuous), smoking status (status and pack-years), education level, BMI, and NSAID use (including recency of
endoscopy, before-1994 vs. 1995 and after, for those that reported a history of endoscopy)
50
To examine the impact of endoscopy on these results, given the higher endoscopy rates
in those reporting AAC, we performed an analysis stratified by a report of endoscopy on
the second questionnaire (1999-2001, see Methods). The relative risk in this analysis
subgroup was similar to that in the entire cohort (RR 0.86). Amongst individuals that
reported a history of having had an endoscopy (N = 54,431, 847 cases), the relative risk
was 0.88 (95% CI: 0.75, 1.02), while in individuals that reported never having a history of
endoscopy (N = 100 328, 1,876 cases) the relative risk was 0.86 (95% CI: 0.77, 0.97)
(Table 2.4).
The inverse association between AAC status and CRC risk was also noted with CRC
mortality (RR 0.80, 95% CI 0.70, 0.91) (Table 2.3). The association was similar across
ethnic groups (P eth = 0.14), and was statistically significant in European Americans and
Native Hawaiians. When stratified by sex, males experienced a greater reduction in risk
(RR 0.67, 95% CI 0.55, 0.83), compared to females (RR 0.91, 95% CI 0.76, 1.08, P het =
0.041). Across all ethnic populations, risk of mortality due to regional/aggressive CRC
associated with AAC was significantly reduced (RR 0.80, 95% CI: 0.69, 0.94, P eth = 0.36).
Mortality risk estimates between regional/aggressive and localized CRC (RR 0.75, 95%
CI: 0.51, 1.11) were similar (P het = 0.68), though there was a limited number of cases in
the localized group (N=166).
2.5 Discussion
In this prospective analysis of five racial/ethnic populations, past atopic allergic
conditions (asthma, allergies or hay fever) were found to be associated with a
51
statistically significant 14% decrease in risk of developing CRC. We observed a similar
inverse association among both men and women, for colon and rectal cancer as well as
for localized and regional disease, with some suggestion of lower risk for left-sided CRC
and rectal cancer versus right-sided colon cancer (P het = 0.04). The inverse association
was noted in all ethnic populations (RRs 0.72-0.96), although the magnitude of the
association was weaker in Latinos for overall disease as well as for all disease subgroup
analyses. Consistent patterns of association were also observed with CRC mortality, with
AAC conferring a reduction in risk of CRC-related death in all populations. Males had a
greater reduction in risk (RR 0.67) compared to females (RR 0.91).
In previous studies, a meta-analysis which combined studies of CRC incidence and
mortality reported no statistically significant association between asthma or any allergy
and CRC incidence/mortality. Findings in the current study on CRC incidence, overall and
by subsite, are also directionally consistent and similar in magnitude to more recent
associations between CRC incidence and hay fever and/or asthma reported in the CPS-II
Nutritional and Taiwanese cohorts. These studies had substantially fewer incident CRC
cases (442 cases in exposed groups within the CPS-II Nutrition Cohort and ~400 cases in
the Taiwanese cohort) and not all results were statistically significant (14,15). A 26%
reduction in CRC risk associated with allergies was also reported in the Iowa Women’s
Health Study (16). Our findings with CRC mortality were also similar to that in the CPS-II
mortality cohort, which reported a 21% decrease in CRC-related mortality due to both
hay fever and asthma (with weaker associations among individuals that reported only
one of these conditions) (15). Despite there being evidence of similarities in results
52
across studies, it is difficult to draw direct comparisons between the above studies and
the current study, as the methods of categorizing exposures is different. For example,
the CPS-II and Taiwanese studies examined one or more of the following: allergic rhinitis
(hay fever), atopic dermatitis, or asthma, although neither study examined all allergies
together, as was done in the current study.
The biological basis for the apparent protective association of AAC with CRC is not
known but could be due to anti-tumor effects of Type I IgE-mediated immune activity, in
which migrating eosinophils, mast cells, and other immune cells are present within the
highly vascularized and rapidly-overturning intestinal mucosa. Atopic allergens can enter
the system at numerous points, including the respiratory system (resulting in asthma)
and the digestive system (promoting food allergies), where mast cells and eosinophils
are ubiquitous (18). This can lead to Type I IgE-mediated hypersensitivity reactions such
as asthma or allergic rhinitis (hay fever), due to the over-stimulation of eosinophils and
mast cells, as well as an excessive T-helper cell type II response (through activation by
IgE and tumor cell complexes). Mast cells can release factors that enhance permeability
and inflammation when degranulated by IgE and tumor cell complexes. Other immune
responses activated in this scenario can include IgE antibody-dependent cellular
cytotoxicity directed at pre-cancerous and cancerous cells, and direct anti-cancer effects
(19). For example, eosinophil levels, which are inversely associated with CRC, are
thought to have cytotoxic effects and can mediate anti-tumor like activity on pre-
cancerous cells which arise in rapidly developing tissues such as the gut lining
(16,20,21). Asthmatics are already known to have increased eosinophilic presence in the
53
respiratory mucosal lining, and together with mast cell degranulation, this may reflect
an increased presence of leukocytes in the gut of individuals with atopic conditions (22).
Another potential theory, known as the prophylaxis hypothesis, contends that allergic
responses and inflammation in mucosal surfaces can lead to more rapid clearance of
mutagenic triggers (12).
A limitation of this study is that the exposure was non-specific, and based on one’s self-
report of any one of the following atopic allergic conditions: asthma, hay fever, skin
allergy, food allergy, or any other allergy. Consequently, we were unable to analyze each
atopic allergic condition separately or examine whether a participant with more than
one condition may represent a more exposed group. However, asthma, hay fever, and
other allergic conditions have a similar biological process based on atopy
(hypersensitization towards an external trigger), and are conditions that are highly
correlated in adult subjects (23–25). Another limitation of the current study is that the
analysis utilized baseline data in the cohort, and thus did not capture new adult-onset
AAC, which would most likely result in non-differential misclassification of AAC
individuals into the non-AAC group and a subsequent under-estimate of the association
of AAC on CRC risk.
Individuals with AAC were noted to have higher endoscopy rates than those without
AAC (33.1% vs. 25.9%). As endoscopy usage has the potential to both reduce the risk of
CRC and increase the rate of CRC capture, we performed an analysis stratified by
endoscopy status reported on the second questionnaire. However, when examining the
population that did not report having an endoscopy, the risk of incident CRC was still
54
significantly decreased (RR 0.86), with an inverse association observed in all ethnic
groups. This indicates that the reduction in risk due to AAC is independent of endoscopy
status.
We conducted analyses stratified by CRC risk factors to elucidate subgroups in which
AAC may be more protective as well as potential biological CRC disease mechanisms
that may be influenced by AAC. In these analyses, AAC status was associated with lower
risk in current smokers compared to past and never-smokers (RR 0.74 vs. 0.86 and 0.90).
One potential explanation may involve the modulation of the T-helper cell type I and T-
helper cell type II responses and resultant modulation of inflammatory cytokine profiles
in response to smoke exposure. Another possibility is that this association may be due
to a greater number of premalignant cells in the gut serving as targets for the
inflammation process to act upon in the environments of smokers, compared to those
where the burden of carcinogens is lower as in past/never smokers. This suggestive
effect modification needs additional investigation. In contrast, the CPS cohorts reported
no evidence of interaction by smoking status, though the authors only reported on the
combined condition of hay fever & asthma (15).
Future work to understand the mechanistic role of AAC in CRC development should
include examining the association between AAC and polyps (detailed data on which are
not available in the MEC). Specifically, it would be interesting to examine whether AAC
reduces CRC risk by preventing polyp formation, or if AAC reduces CRC risk by
preventing polyp progression to CRC. Finally, the analysis of serum specific immune
markers such as IgE or eosinophil count may provide additional insight into the
55
association between AAC and CRC (12). However, self-reported history of allergic
diagnoses may capture aspects of immune function that are not highly associated with a
single measurement of IgE (26).
In this large prospective study among five racial/ethnic populations, we observed a
significant inverse association between AAC and the development of CRC in both men
and women. These findings support results from previous prospective studies linking
atopic allergic conditions such as asthma, hay fever, and food allergies, and the
development of colorectal cancer, and further highlight the importance of the immune
system in the pathogenesis of CRC.
56
2.6 References
1. Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA. Cancer J. Clin.
[electronic article]. 2011;61(2):69–90. (http://doi.wiley.com/10.3322/caac.20107)
2. Howlader N, Noone AM, Krapcho M, et al. SEER Cancer Statistics Review, 1975-
2011. Natl. Canncer Institute, Bethesda, MD [electronic article].
2014;(http://seer.cancer.gov/csr/1975_2011/)
3. Prentice RL, Pettinger M, Beresford S a a, et al. Colorectal cancer in relation to
postmenopausal estrogen and estrogen plus progestin in the Women’s Health
Initiative clinical trial and observational study. Cancer Epidemiol. Biomarkers Prev.
[electronic article]. 2009;18(5):1531–7.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2689377&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
4. Li N, Xi Y, Tinsley HN, et al. Sulindac selectively inhibits colon tumor cell growth by
activating the cGMP/PKG pathway to suppress Wnt/beta-catenin signaling. Mol
Cancer Ther [electronic article]. 2013;12(9):1848–1859.
(http://www.ncbi.nlm.nih.gov/pubmed/23804703)
5. Hoffmeister M, Chang-Claude J, Brenner H. Individual and joint use of statins and
low-dose aspirin and risk of colorectal cancer: A population-based case–control
study. Int. J. Cancer [electronic article]. 2007;121(6):1325–1330.
(http://doi.wiley.com/10.1002/ijc.22796)
6. He J, Stram DO, Kolonel LN, et al. The association of diabetes with colorectal
cancer risk: the Multiethnic Cohort. Br. J. Cancer [electronic article].
2010;103(1):120–126.
(http://www.nature.com/doifinder/10.1038/sj.bjc.6605721)
7. Fung KYC, Fung K, Ooi CC, et al. Colorectal Carcinogenesis: A Cellular Response to
Sustained Risk Environment. Int. J. Mol. Sci. [electronic article].
2013;14(7):13525–13541. (http://www.mdpi.com/1422-0067/14/7/13525/)
8. Bingham SA, Day NE, Luben R, et al. Dietary fibre in food and protection against
colorectal cancer in the European Prospective Investigation into Cancer and
Nutrition (EPIC): an observational study. Lancet [electronic article].
2003;361(9368):1496–1501.
(http://linkinghub.elsevier.com/retrieve/pii/S0140673603131741)
9. Ollberding NJ, Nomura AM, Wilkens LR, et al. Racial/ethnic differences in
colorectal cancer risk: The multiethnic cohort study. Int. J. Cancer [electronic
article]. 2011;129(8):1899–1906. (http://doi.wiley.com/10.1002/ijc.25822)
10. Turner MC, Chen Y, Krewski D, et al. Cancer Mortality among US Men and Women
with Asthma and Hay Fever. Am. J. Epidemiol. [electronic article].
2005;162(3):212–221.
(http://aje.oxfordjournals.org/cgi/doi/10.1093/aje/kwi193). (Accessed January
57
10, 2014)
11. Turner MC. Epidemiology: allergy history, IgE, and cancer. Cancer Immunol
Immunother [electronic article]. 2012;61(9):1493–1510.
(http://www.ncbi.nlm.nih.gov/pubmed/22183126)
12. Josephs DH, Spicer JF, Corrigan CJ, et al. Epidemiological associations of allergy,
IgE and cancer. Clin. Exp. Allergy [electronic article]. 2013;43(10):1110–23.
(http://www.ncbi.nlm.nih.gov/pubmed/24074329). (Accessed January 14, 2014)
13. Vojtechova P, Martin RM. The association of atopic diseases with breast,
prostate, and colorectal cancers: a meta-analysis. Cancer Causes Control
[electronic article]. 2009;20(7):1091–1105.
(http://link.springer.com/10.1007/s10552-009-9334-y)
14. Hwang CY, Chen YJ, Lin MW, et al. Cancer risk in patients with allergic rhinitis,
asthma and atopic dermatitis: a nationwide cohort study in Taiwan. Int J Cancer
[electronic article]. 2012;130(5):1160–1167.
(http://www.ncbi.nlm.nih.gov/pubmed/21455988)
15. Jacobs EJ, Gapstur SM, Newton CC, et al. Hay Fever and Asthma as Markers of
Atopic Immune Response and Risk of Colorectal Cancer in Three Large Cohort
Studies. Cancer Epidemiol. Biomarkers Prev. [electronic article]. 2013;22(4):661–
669. (http://cebp.aacrjournals.org/cgi/doi/10.1158/1055-9965.EPI-12-1229)
16. Prizment AE, Folsom AR, Cerhan JR, et al. History of allergy and reduced incidence
of colorectal cancer, Iowa Women’s Health Study. Cancer Epidemiol. Biomarkers
Prev. [electronic article]. 2007;16(11):2357–62.
(http://www.ncbi.nlm.nih.gov/pubmed/18006924). (Accessed January 10, 2014)
17. Kolonel LN, Henderson BE, Hankin JH, et al. A multiethnic cohort in Hawaii and
Los Angeles: baseline characteristics. Am. J. Epidemiol. [electronic article].
2000;151(4):346–57. (http://www.ncbi.nlm.nih.gov/pubmed/10695593)
18. Kita H. Eosinophils: multifunctional and distinctive properties. Int Arch Allergy
Immunol [electronic article]. 2013;161(Suppl 2):3–9.
(http://www.ncbi.nlm.nih.gov/pubmed/23711847)
19. Jensen-Jarolim E, Achatz G, Turner MC, et al. AllergoOncology: the role of IgE-
mediated allergy in cancer. Allergy [electronic article]. 2008;63(10):1255–1266.
(http://www.ncbi.nlm.nih.gov/pubmed/18671772)
20. Prizment AE, Anderson KE, Visvanathan K, et al. Inverse Association of Eosinophil
Count with Colorectal Cancer Incidence: Atherosclerosis Risk in Communities
Study. Cancer Epidemiol. Biomarkers Prev. [electronic article]. 2011;20(9):1861–
1864.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3175810&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
58
21. Legrand F, Driss V, Delbeke M, et al. Human eosinophils exert TNF-α and
granzyme A-mediated tumoricidal activity toward colon carcinoma cells. J.
Immunol. [electronic article]. 2010;185(12):7443–51.
(http://www.ncbi.nlm.nih.gov/pubmed/21068403). (Accessed January 10, 2014)
22. Straumann A, Aceves SS, Blanchard C, et al. Pediatric and adult eosinophilic
esophagitis: similarities and differences. Allergy [electronic article].
2012;67(4):477–490. (http://www.ncbi.nlm.nih.gov/pubmed/22313241)
23. Leynaert B, Neukirch C, Kony S, et al. Association between asthma and rhinitis
according to atopic sensitization in a population-based study ☆. J. Allergy Clin.
Immunol. [electronic article]. 2004;113(1):86–93.
(http://linkinghub.elsevier.com/retrieve/pii/S0091674903024473)
24. Brinton LA, Sakoda LC, Frederiksen K, et al. Relationships of uterine and ovarian
tumors to pre-existing chronic conditions. Gynecol. Oncol. [electronic article].
2007;107(3):487–494.
(http://linkinghub.elsevier.com/retrieve/pii/S009082580700577X)
25. Togias A. Rhinitis and asthma: Evidence for respiratory system integration. J.
Allergy Clin. Immunol. [electronic article]. 2003;111(6):1171–1183.
(http://linkinghub.elsevier.com/retrieve/pii/S0091674903014490). (Accessed
January 10, 2014)
26. Hoppin JA, Jaramillo R, Salo P, et al. Questionnaire predictors of atopy in a US
population sample: findings from the National Health and Nutrition Examination
Survey, 2005-2006. Am J Epidemiol [electronic article]. 2011;173(5):544–552.
(http://www.ncbi.nlm.nih.gov/pubmed/21273397)
59
CHAPTER 3: RACIAL AND ETHNIC DIFFERENCES IN CARDIOVASCULAR
DISEASE MORTALITY: THE MULTIETHNIC COHORT
Neal Tambe, Peggy Wan, Daniel O. Stram, Lynne R. Wilkens, Gertraud Maskarinec,
Loic Le Marchand, Christopher A. Haiman
AUTHOR AFFILIATIONS: Department of Preventive Medicine, Norris Cancer Center,
University of Southern California, Los Angeles, California (Neal A. Tambe, Peggy Wan,
Daniel O, Stram, Christopher A. Haiman); Department of Epidemiology, University of
Hawaii Cancer Center, Honolulu, Hawaii (Lynne R. Wilkens, Gertraud Maskarinec, Loic Le
Marchand).
Correspondence to Neal A. Tambe, Norris Comprehensive Cancer Center, Department of
Preventive Medicine, Keck School of Medicine, University of Southern California, 1441
Eastlake Avenue #4405, Los Angeles, CA 90089 (e-mail: ntambe@usc.edu)
60
3.1 Abstract
Previous studies have shown racial/ethnic differences in mortality from acute
myocardial infarction (AMI), other heart diseases (OHD), and stroke, with known risk
factors, accounting for some, but not all of these disparities in risk. To further our
understanding of racial/ethnic differences in mortality, we examined the contribution of
known risk factors to population differences in mortality of AMI, OHD, and stroke
among African Americans, Native Hawaiians, Japanese Americans, White and Latino
men and women in comparison to Whites in the Multiethnic Cohort (MEC). Over a 14-
year period there were 1,820 deaths due to AMI, 4,814 due to OHD, and 1,992 due to
stroke in 139,404 men and women. For AMI and OHD, African American and Native
Hawaiian men and women had >50% excess age-adjusted risk of mortality, compared to
Whites (P<0.05), while Japanese Americans had reduced age-adjusted risks, and Latinos
had similar levels of age-adjusted risk. For stroke, age-adjusted risk of mortality was
increased across all racial/ethnic groups of men (P<0.05), and increased across all
racial/ethnic groups of men except Japanese Americans and Latinos. Adjusting for
known risk factors accounted for a large portion of racial/ethnic differences in death
due AMI, OHD and stroke. Compared to Whites, African Americans had an increased risk
of death due AMI (RR 1.51, P=<0.001) and OHD (RR 1.22, P=0.007) in women, and stroke
in men (RR 1.49, P<0.001). Native Hawaiian men and women also had a 60% (P<0.001)
and 73% (P<0.001) greater risk of OHD mortality, respectively, while risk of AMI and
OHD mortality in Japanese Americans were statistically significantly reduced (RRs=0.66-
0.80; P<0.001) compared to whites. In this prospective analysis, established risk factors
61
of AMI, OHD, and stroke do not fully account for racial/ethnic specific differences in risk
of mortality. Additional analysis is necessary to identify the source of the remaining
discrepancies in risk.
62
3.2 Introduction
Cardiovascular disease (CVD), including acute myocardial infarction (AMI), other heart
diseases (OHD), and stroke, is the most common cause of mortality worldwide(1,2). As
of 2013, stroke and heart disease are responsible for one-third of the total deaths within
the United states(2). Mortality due to CVD differs by ethnic/racial group, with Asians
and Latinos having lower death rates in comparison to Whites(3–5). In contrast, African
Americans experience the highest rates of CVD and stroke(6–8).
Risk factors for cardiovascular disease are well-defined and have been shown to
contribute to population differences in CVD mortality. Studies within the Behavioral
Risk Factor Surveillance System (BRFSS) and of National Health and Nutrition
Examination Survey (NHANES) data have estimated that about one-half of CVD-related
mortality can be attributed to the top five risk factors: elevated cholesterol levels,
diabetes, smoking, hypertension, and obesity(9). The prevalence of these and other CVD
risk factors differ by ethnicity; in the 2013 National Health Interview Survey (NHIS),
Asian (14.7%) and Hispanic (16.6%) men were less likely to be current smokers than
White (21.7%) and African American (21.1%) men(6). NHANES data from 2011-2012 also
reported a higher prevalence of hypertension in African American adults (42.1%)
compared to Whites (28.0%), Asians (24.7%) and Hispanics (26.0%)(10). Prevalence of
diabetes also varies widely across racial/ethnic groups, with the highest rates in African
Americans (13.2%), and the lowest rates in Whites (7.6%)(11). Hispanic individuals have
higher rates of CVD risk factors, including elevated LDL and triglycerides levels and
reduced levels of HDL, compared to Whites and African Americans, but paradoxically
63
have lower rates of total CVD(6,12,13). Reduced physical activity, poor diet, and other
risk factors of CVD have greater prevalence in non-White minorities, and may contribute
to racial/ethnic differences in risk of disease(6,12,14). Use of cholesterol-lowering
therapies has been reported to be highest in Asians, and lowest in Hispanics which is
likely to also contribute to ethnic differences in CVD mortality(15,16).
Some, but not all previous studies have reported that a large fraction of the disparity in
CVD mortality across racial/ethnic groups can be accounted for by the varying
prevalence of known risk factors(13,17–21). In a previous study among African
Americans, Latinos, Japanese, Native Hawaiians and Whites in the Multiethnic Cohort
(MEC) that included 946 deaths due to acute myocardial infarction (AMI), and 2,232
deaths due to other heart disease (OHD) we found racial/ethnic differences in CVD
mortality remained after accounting for known risk factors; Native Hawaiians had the
greatest risk of AMI, African American women had greater risks of AMI and OHD, and
Japanese Americans had lower risk of AMI and OHD. In the current study, we have
expanded our investigation of racial/ethnic differences in CVD and stroke mortality in
the MEC, with >8,600 deaths due to AMI, OHD or stroke observed among 139,404
White, African American, Latino, Native Hawaiian and Japanese American participants
with over 14 years of follow-up. In addition to investigating the contribution of all risk
factors we also evaluated the relative contribution of each risk factor individually to
better understand which risk factor(s) are most important to specific racial/ethnic
groups.
64
3.3 Methods
The MEC is composed of more than 215,000 men and women from Hawaii and
California, and includes five major self-reported racial/ethnic populations: Whites,
African Americans (AA), Native Hawaiians (NH), Japanese Americans (JA), and Latinos
(LA). Cohort members were recruited through Department of Motor Vehicle license
files, and supplemented by voter registration and Health Care Financing Administration
(Medicare) files. Participating individuals were between 45 and 75 years of age, and
completed a 26-page self-administered, detailed questionnaire at cohort entry (baseline
data, 1993-1996). The questionnaire included basic demographic factors (including
race/ethnicity and education), lifestyle factors (e.g. diet, medication use and smoking
history), chronic medical conditions, and reproductive history/hormone usage for
women(22). A modified baseline questionnaire was re-administered between 2003-
2007, which contained additional information about usage of cholesterol medication.
Risk Factors
Risk factors evaluated in this study include: race/ethnicity, body mass index (weight
(kg)/height(m2)), <23, 23 - <25, 25 - <30, 30 - <35, 35+), history of hypertension
(ascertained from self-report as well as hypertension medication usage at baseline),
self-reported history of diabetes; smoking status (never, past, current) and pack-years
(<20, >20, or missing), educational level (< 12 years, some college/vocational, college
graduate), physical activity (hours (hr.)/week; none, <1.5, 1.5 - <5.0, 5.0+), alcohol
(grams (g) ethanol/day; <12, 12 - <24, 24+), and diet. Dietary factors included
65
percentage of calories from: cholesterol, saturated fat, unsaturated fat, dietary fiber
from fruits, grains, and total legumes, omega-3 and -6 fatty acids, sodium, folate, and
isoflavones. For women, information about hormone therapy usage (HRT; estrogen
and/or progesterone use, never/past or current use) and menopausal status (natural,
oophorectomy, hysterectomy; age of onset) was also included.
Outcomes
The vital status of each cohort member was obtained through linkage with state death
certificate files in Hawaii and California, supplemented with the National Death Index.
The primary cause of death was used to determine outcome, coded according to the
Ninth and Tenth Revisions of the International Classification of Disease (ICD-9, ICD-10).
Here, cause of death was divided into 3 categories: acute myocardial infarction (AMI:
410, I21), other heart disease (OHD: 411, 413-414, 425-429, I20, I22-I24, I42-I52), and
stroke (430-438, I60-I69).
3.3.1 Statistical Analysis
Hazard ratios (reported as rate ratios (RRs)) and confidence intervals (CI) for AMI, OHD,
and stroke were estimated using Cox proportional hazards regression modeling via
STATA/IC 14.0 (StataCorp, College Station, TX). All RRs are reported with Whites as the
reference group unless specifically stated otherwise. The time variable used was age
from date of enrollment (1993-1996) to age of death or censoring (ending December
31
st
, 2010 in California, and December 31
st
, 2009 in Hawaii). Risk factors were included
as categorical variables, while dietary factors were included as quartiles of nutrient
66
densities. Individuals were excluded from this analysis if they previously reported a
heart attack or stroke (n = ~21,000), or were missing primary covariate information
(men: n=~8,900; women: n=~14,400). Approximately 18,000 additional women were
excluded if they were pre-menopausal, or missing covariate information for menopausal
status or HRT usage. Three models were evaluated: (1) a basic age-adjusted model, (2) a
model including known risk factors as covariates, and (3) a complete model including
dietary and nutritional factors.
Ethnic/racial specific risk factor variation
In order to assess which risk factors contribute to the greatest differences in CVD risk by
ethnicity, risk factors were sequentially added to a baseline model containing ethnicity
with the order determined by which risk factors had the greatest absolute value change
in the risk estimate for ethnic group. In addition, we examined the change in the
racial/ethnic difference in risk adjusting for each risk factor alone.
Cholesterol-Lowering Medication Analysis
Statins and cholesterol-lowering medication may reduce the risk of all-cause and
CVD/CHD related mortality and vary in prevalence by ethnicity, though questions
remain as to their efficacy as a primary prevention measure(23,24). Secondary analyses
were conducted stratified by cholesterol medication use in 31,565 men and 30,785
women to examine the impact of lipid medication use on racial/ethnic differences in
CVD mortality. Analyses used self-reported cholesterol medication use from a follow-up
67
questionnaire administered between 2003-2007, and as such, person-time used in this
subset analysis begins with the follow-up questionnaire date.
Serum Lipid Analysis
We calculated lipid levels among 6,024 non-diabetic and non-cancer men and women,
almost half of which were Latino (n = 2842). Lipid samples were taken as previously
described (25). Briefly, samples were drawn, refrigerated, and processed into
components within 4 hours of collection, and stored in liquid nitrogen (-186C). A short
questionnaire was administered at the time of blood draw including medication us.
Lipids were analyzed via an automated analyzer (Cobra Mira Plus, Roche Diagnostics,
Switzerland). Measurements included total cholesterol (mg/dL), HDL-c (mg/dL), LDL-c
(calculated via the Friedewald formula, mg/dL), log transformed triglyceride values, and
the log-transformed ratio of total cholesterol to HDL-c. After excluding individuals with
missing covariates or lipid values, the final analytical sample included 5,635 – 5,655
individuals based on the lipid biomarker examined. Least squared means were
calculated using generalized linear models for each racial/ethnic groups and were
adjusted for age at blood draw (years), lipid medication use at time of blood draw
(within 2 weeks), hypertension status, fasting time (hours), phase and time of blood
draw, BMI, smoking status, alcohol consumption, and saturated fat/cholesterol
consumption in quartiles(26–28). Analysis was performed separately by sex and
pairwise comparisons were made with Whites. A sensitivity analysis was done by
eliminating Latinos chosen from a pool of T2D controls (n=1,864), but the results were
not significantly different and are not presented here.
68
3.4 Results
Over >14 years of follow-up there were 1069 AMI, 3026 OHD, and 1031 stroke related
deaths amongst 70,738 men. Compared to White men, the age-adjusted risk of AMI-
mortality was highest in African Americans (RR 1.55, 95% CI 1.29-1.86, P < 0.001) and
Native Hawaiians (RR 1.58, 95% CI 1.22-2.03, P < 0.001), and lowest in Japanese
Americans (RR 0.77, 95% CI 0.65-0.92, P = 0.004). The age-adjusted risk of death due to
OHD was highest in Native Hawaiians (RR 2.10, 95% CI 1.82-2.42, P < 0.001), elevated in
African Americans (RR 1.67, 95% CI 1.50-1.87, P < 0.001), and reduced in Japanese
Americans (RR 0.84, 95% CI 0.76-0.94, P = 0.001). Compared to Whites, the age-adjusted
risk for stroke-related mortality was statistically significantly higher in all other
populations, ranging from the highest risk in African Americans (RR 1.86, 95% CI 1.53-
2.27, P < 0.001) to a RR of 1.21 in Japanese Americans (95% CI 1.01-1.45, P = 0.037).
[Table 3.1]
Among 68,666 women in the MEC, there were 751 AMI, 1788 OHD, and 961 stroke
related deaths. African American women had the highest risk of death due to AMI (RR
2.33, 95% CI 1.90-2.85, P < 0.001), while Japanese American women had a RR of 0.77
compared to White women (95% CI 0.61-0.96, P = 0.023). Native Hawaiians also had a
statistically significant increased risk of AMI (RR 1.62, 95% CI 1.15-2.27, P = 0.005).
Compared to Whites, risk of death due to OHD was highest in Native Hawaiians (RR
2.10, 95% CI 1.74-2.53, P < 0.001) and African Americans (RR 1.85, 95% CI 1.62-2.10, P <
69
0.001). For stroke, only African American women had a statistically significantly higher
risk of death than White women (RR 1.48, 95% CI 1.23-1.77, P < 0.001). [Table 3.1]
70
Table 3.1. Mortality rates from AMI, OHD and stroke by race and gender in the MEC
Men W AA NH JA LA Total
# 18,446 8,981 5,020 21,753 16,538 70,738
AMI: Deaths 246 219 80 266 258 1069
MR 77.35 130.25 116.49 58.58 88.78
RR 1.00 1.55 1.58 0.77 1.04
(1.29-1.86) (1.22-2.03) (0.65-0.92) (0.87-1.24)
P <0.001 <0.001 0.004 0.669
OHD: Death 640 629 268 770 719 3026
MR 200.17 386.52 412.15 166.37 247.98
RR 1.00 1.67 2.10 0.84 1.11
(1.5-1.87) (1.82-2.42) (0.76-0.94) (1-1.24)
P <0.001 <0.001 0.001 0.048
Stroke: Deaths 189 211 60 334 237 1031
MR 58.32 122.21 94.92 70.93 80.78
RR 1.00 1.86 1.65 1.21 1.24
(1.53-2.27) (1.23-2.21) (1.01-1.45) (1.03-1.51)
P <0.001 0.001 0.037 0.026
Women W AA NH JA LA Total
# 17,742 12,240 4,616 20,242 13,826 68,666
AMI: Deaths 146 266 44 157 138 751
MR 36.16 99.92 59.70 27.83 48.02
RR 1.00 2.33 1.62 0.77 1.25
(1.9-2.85) (1.15-2.27) (0.61-0.96) (0.99-1.58)
P <0.001 0.005 0.023 0.061
OHD: Death 398 577 155 338 320 1788
MR 101.23 223.80 207.98 62.32 116.28
RR 1.00 1.85 2.10 0.62 1.07
(1.62-2.1) (1.74-2.53) (0.53-0.71) (0.93-1.25)
P <0.001 <0.001 <0.001 0.338
Stroke: Deaths 219 256 49 267 170 961
MR 55.04 102.91 71.76 50.74 62.52
RR 1.00 1.48 1.24 0.87 1.05
(1.23-1.77) (0.91-1.69) (0.73-1.04) (0.86-1.29)
P <0.001 0.177 0.13 0.606
W – Whites; AA – African Americans; NH – Native Hawaiians; JA – Japanese Americans; LA –
Latinos; AMI – acute myocardial infarction; OHD – other heart disease; MR – mortality rate; RR –
relative risk. Mortality rate is per 100,000 persons per year, age-standardized to the US Standard
Population aged 45-79 for the year 2000. Rate ratios are from a crude Cox regression model (see
Statistical Analysis)
71
Established risk factors in men
Differences in the age-standardized distribution of established CVD risk factors within
the MEC have been reported previously(29). Obesity (BMI > 30) was highest in Native
Hawaiians (31.9%) and lowest in Japanese Americans (7.8%). Extreme obesity (BMI > 35)
was associated with increased risk of AMI (RR 1.48, 95% CI 1.10-1.98) and as well as
OHD mortality (RR 2.00, 95% CI 1.68-2.37). Risk of death due to stroke was elevated in
obese individuals but was not statistically significant (RR 1.33, 95% CI 0.94-1.88).
Prevalence of hypertension was highest in African Americans (51.8%) and Native
Hawaiians (48.8%), and lowest in Whites (31.0%) and Latinos (33.2%). Diabetes
prevalence was 15.7% in Native Hawaiians, 14.2% in Latinos, and was lowest in Whites
(5.4%). Hypertension and diabetes were both strongly and highly statistically
significantly associated with AMI, OHD, and stroke mortality (RRs 1.71-2.42, P-values <
0.001). African Americans were most likely to be current smokers (27.6%) and, current
smoking of more than 20 packs-per-day was significantly associated with RR’s of 2.06 for
AMI (95% CI 1.67-2.54), 2.41 for OHD (95% CI 2.13-2.72), and 1.79 for stroke mortality
(95% CI 1.44-2.23) compared with never smokers. Past smokers of more than 20 packs-
per-day also had significantly greater risks of AMI and OHD mortality (Table 3.2).
Greater alcohol consumption and vigorous activity (>5.0 hours/week) were associated
with significantly lower risks of AMI, OHD and stroke (only vigorous activity) (Table 3.2).
Whites reported consuming more alcohol, with 27.8% consuming more than 24
grams/day, while African Americans consumed the least (17.1%). Native Hawaiians
reported the most amount of physical activity, with 37.4% reporting more than 5
72
hours/week, followed by Whites (33.3%) and Latinos (32.6%), while African Americans
(34.9%) and Japanese (32.8%) were the least physically active. College education was
associated with a 19%, 27%, and 22% reduced risk of AMI, OHD, and stroke, respectively
(P trend < 0.009), with 49.8% of Whites reporting a college education compared to 13.1%
of Latinos and 18.0% of Native Hawaiians (Table 3.2)
Dietary patterns also varied widely across racial/ethnic groups, with Japanese American
men consuming the least amount of saturated fat and cholesterol (41.8% and 31.3%
respectively, in the lowest quartile). African Americans and Latinos reported the
greatest amount of saturated fat consumption (38.9% and 37.3% in the highest quartile)
as well as cholesterol (41.4% and 29.1%). Greater intake of saturated fat was associated
with statistically significant increased risk of OHD and AMI and a non-significant
increased risk of stroke). Increased cholesterol intake was non-significantly associated
with AMI and stroke (Table 3.2)
73
Table 3.2. Age-standardized distributions of risk factors (%) with associated RRs of mortality from AMI, OHD, and stroke amongst men in the
MEC
Men W AA NH JA LA AMI P OHD P Stroke P
BMI (kg/m
2
)
≤23 17.4 13.4 10.0 23.9 10.2 1.06 (0.87-1.29) 1.18 (1.05-1.33) 1.14 (0.94-1.38)
>23-25 21.6 16.8 12.8 26.2 16.5 1.00 1.00 1.00
25- 45.5 48.4 45.4 42.1 53.8 0.88 (0.74-1.03) 1.02 (0.92-1.12) 0.93 (0.79-1.09)
30- 12.2 16.5 21.4 6.7 15.7 1.05 (0.85-1.30) 1.30 (1.14-1.47) 1.03 (0.82-1.29)
>=35 3.3 4.9 10.5 1.1 3.9 1.48 (1.10-1.98) 0.668 2.00 (1.68-2.37) <0.001 1.33 (0.94-1.88) 0.442
Hypertension 31.0 51.8 48.8 42.0 33.2 1.86 (1.64-2.11) <0.001 1.71 (1.59-1.84) <0.001 1.77 (1.56-2.01) <0.001
Diabetes 5.4 13.2 15.7 10.3 14.2 2.42 (2.10-2.79) <0.001 2.09 (1.92-2.28) <0.001 1.90 (1.63-2.22) <0.001
Smoking status (packs per day)
Never 33.6 26.9 32.3 30.4 34.2 1.00 1.00 1.00
Past<20 31.5 35.0 30.8 36.4 39.8 0.97 (0.83-1.14) 0.712 0.99 (0.90-1.09) 0.773 0.88 (0.75-1.03) 0.099
Past>20 19.0 10.5 16.2 17.3 7.5 1.22 (1.01-1.46) 0.039 1.36 (1.22-1.52) <0.001 1.00 (0.83-1.20) 0.976
Current<20 4.8 16.3 7.5 5.9 12.5 1.93 (1.55-2.40) <0.001 1.91 (1.67-2.18) <0.001 1.37 (1.08-1.75) 0.011
Current>20 11.1 11.3 13.1 10.0 6.0 2.06 (1.67-2.54) <0.001 2.41 (2.13-2.72) <0.001 1.79 (1.44-2.23) <0.001
Alcohol (g/day)
0 27.1 40.9 40.6 42.8 33.9 1.00 1.00 1.00
>0-12 31.7 32.2 28.5 29.7 38.4 0.88 (0.77-1.02) 0.79 (0.72-0.86) 0.93 (0.80-1.08)
>12-24 13.5 9.8 10.4 10.7 11.0 0.70 (0.56-0.89) 0.63 (0.55-0.73) 0.90 (0.72-1.13)
>24 27.8 17.1 20.5 16.8 16.6 0.79 (0.66-0.95) 0.001 0.92 (0.83-1.01) 0.001 0.96 (0.81-1.15) 0.535
Vigorous Activity (hrs./week)
0 27.1 34.9 22.4 32.8 30.3 1.00 1.00 1.00
>0-1.5 15.0 16.6 13.9 18.9 15.1 0.78 (0.66-0.94) 0.89 (0.80-0.98) 0.84 (0.70-1.00)
>1.5-5.0 24.6 22.8 26.3 24.4 22.0 0.82 (0.69-0.96) 0.91 (0.83-1.01) 0.84 (0.71-0.99)
>5.0 33.3 25.7 37.4 23.9 32.6 0.73 (0.62-0.86) <0.001 0.79 (0.71-0.87) <0.001 0.81 (0.69-0.96) 0.009
Education
< High School 21.4 37.3 53.1 33.3 62.4 1.00 1.00 1.00
Further education 28.8 37.9 28.9 31.2 24.5 0.92 (0.80-1.06) 0.94 (0.87-1.03) 0.80 (0.69-0.93)
74
Some College 49.8 24.7 18.0 35.6 13.1 0.81 (0.69-0.95) 0.009 0.73 (0.66-0.80) <0.001 0.78 (0.67-0.92) 0.001
Saturated fat (quartiles)
1st 20.4 14.3 27.3 41.8 12.9 1.00 1.00 1.00
2nd 23.8 19.8 28.5 31.0 20.2 1.11 (0.91-1.36) 1.09 (0.97-1.22) 0.97 (0.80-1.17)
3rd 26.5 27.0 26.3 19.1 29.6 1.25 (1.01-1.54) 1.34 (1.18-1.51) 1.16 (0.95-1.41)
4th 29.3 38.9 17.9 8.1 37.3 1.49 (1.20-1.85) <0.001 1.46 (1.28-1.67) <0.001 1.07 (0.86-1.34) 0.269
Cholesterol (quartiles)
1st 28.0 15.1 22.5 31.3 19.2 1.00 1.00 1.00
2nd 26.4 18.9 25.2 27.0 24.2 1.10 (0.91-1.35) 0.96 (0.86-1.08) 1.06 (0.88-1.28)
3rd 24.6 24.7 25.7 23.4 27.5 1.18 (0.96-1.46) 0.98 (0.87-1.11) 1.12 (0.91-1.37)
4th 21.0 41.4 26.6 18.3 29.1 1.21 (0.98-1.51) 0.08 1.09 (0.96-1.23) 0.11 1.03 (0.83-1.28) 0.746
W – Whites; AA – African Americans; NH – Native Hawaiians; JA – Japanese Americans; LA – Latinos; AMI – acute myocardial infarction; OHD –
other heart disease; RR – relative risk; BMI – body mass index. Age-standardized (5-year age groups) to the total male population in the study.
RRs for risk factors are shown estimated while simultaneously adjusted for each other.
75
Table 3.3. Age-standardized distributions of risk factors and associated RRs of mortality from AMI, OHD, and stroke amongst women in the MEC.
Women W AA NH JA LA AMI P OHD P Stroke P
BMI (kg/m
2
)
<23 35.0 13.3 19.7 49.1 16.0 1.03 (0.81-1.30) 1.03 (0.89-1.20) 1.31 (1.07-1.60)
23- 19.5 13.9 16.1 21.0 17.7 1.00 1.00 1.00
25- 28.6 37.4 33.9 23.7 39.5 1.13 (0.91-1.41) 0.96 (0.83-1.11) 1.14 (0.93-1.39)
30- 11.2 22.1 18.3 5.1 18.4 1.11 (0.85-1.44) 1.08 (0.91-1.27) 1.22 (0.96-1.54)
>=35 5.8 13.4 12.1 1.2 8.3 1.48 (1.10-1.99) 0.042 1.44 (1.20-1.74) 0.023 1.30 (0.96-1.74) 0.531
Hypertension 34.5 59.3 50.0 37.9 39.1 2.17 (1.85-2.55) <0.001 1.96 (1.76-2.17) <0.001 1.93 (1.68-2.21) <0.001
Diabetes 5.0 13.3 13.6 8.5 13.7 2.97 (2.51-3.50) <0.001 2.78 (2.49-3.10) <0.001 1.35 (1.12-1.63) 0.001
Smoking status (packs per day)
Never 44.8 46.0 48.1 68.4 66.7 1.00 1.00 1.00
Past<20 27.8 28.2 24.2 19.1 21.3 1.26 (1.05-1.51) 0.014 1.26 (1.12-1.43) <0.001 1.22 (1.04-1.44) 0.016
Past>20 10.7 5.3 8.1 3.3 1.8 1.28 (0.93-1.78) 0.133 1.87 (1.56-2.24) <0.001 1.73 (1.34-2.23) <0.001
Current<20 6.1 14.5 11.0 5.5 8.1 2.42 (1.90-3.08) <0.001 2.54 (2.17-2.97) <0.001 2.10 (1.68-2.63) <0.001
Current>20 10.6 6.0 8.7 3.7 2.2 2.76 (2.11-3.63) <0.001 3.05 (2.57-3.61) <0.001 2.13 (1.63-2.78) <0.001
Alcohol (g/day)
0 39.2 61.0 64.5 77.2 63.7 1.00 1.00 1.00
>0-12 37.9 29.7 26.3 19.4 31.0 0.86 (0.72-1.04) 0.78 (0.69-0.88) 0.81 (0.69-0.95)
>12-24 9.7 4.3 4.7 1.8 2.8 0.79 (0.52-1.20) 0.99 (0.79-1.25) 0.80 (0.57-1.13)
>24 13.2 5.0 4.6 1.6 2.5 0.94 (0.66-1.34) 0.201 0.89 (0.71-1.11) 0.03 0.88 (0.65-1.18) 0.049
Vigorous Activity (hrs./week)
0 52.7 63.4 48.7 63.5 61.6 1.00 1.00 1.00
>0-1.5 16.3 17.6 18.0 16.1 16.2 0.86 (0.70-1.07) 0.84 (0.73-0.97) 1.05 (0.88-1.26)
>1.5-5.0 16.9 11.8 19.1 13.2 13.0 0.82 (0.64-1.06) 0.81 (0.69-0.96) 0.85 (0.69-1.07)
>5.0 14.2 7.2 14.2 7.3 9.2 0.75 (0.54-1.03) 0.02 0.70 (0.57-0.87) <0.001 0.76 (0.58-1.01) 0.052
Education
< High School 30.8 37.1 59.1 40.6 70.9 1.00 1.00 1.00
Further education 33.8 38.1 26.6 30.4 20.2 0.92 (0.77-1.09) 1.00 (0.90-1.12) 0.97 (0.83-1.13)
Some College 35.4 24.8 14.2 29.0 8.9 0.82 (0.66-1.01) 0.083 1.04 (0.92-1.18) 0.371 0.83 (0.69-1.00) 0.094
76
Saturated fat (quartiles)
1st 21.3 16.2 24.0 40.3 14.0 1.00 1.00 1.00
2nd 23.8 21.7 27.7 31.0 20.6 0.92 (0.73-1.15) 1.02 (0.88-1.19) 1.09 (0.90-1.33)
3rd 25.7 26.9 27.7 20.6 28.7 1.05 (0.82-1.34) 1.15 (0.98-1.35) 1.15 (0.93-1.43)
4th 29.3 35.2 20.6 8.2 36.7 1.09 (0.84-1.42) 0.353 1.19 (1.01-1.42) 0.026 1.35 (1.07-1.70) 0.3
Cholesterol (quartiles)
1st 26.6 14.8 23.6 32.7 19.5 1.00 1.00 1.00
2nd 25.7 18.7 25.6 28.5 24.8 0.93 (0.73-1.17) 1.17 (1.00-1.37) 1.17 (0.96-1.41)
3rd 24.8 26.2 27.2 23.2 26.9 0.97 (0.76-1.25) 1.29 (1.09-1.52) 0.90 (0.73-1.12)
4th 22.9 40.3 23.5 15.6 28.8 1.14 (0.89-1.48) 0.193 1.49 (1.25-1.76) <0.001 1.00 (0.80-1.26) 0.426
Menopausal status
Nat<=44 8.9 9.8 11.6 7.6 13.6 1.00 1.00 1.00
Nat 45-49 21.2 16.6 19.7 19.5 22.2 0.92 (0.71-1.19) 0.542 0.90 (0.76-1.06) 0.194 0.88 (0.69-1.11) 0.281
Nat 50-54 26.2 19.5 23.8 32.9 23.6 0.82 (0.64-1.05) 0.113 0.74 (0.63-0.88) <0.001 0.83 (0.66-1.04) 0.098
Nat >=55 6.3 5.7 8.0 8.4 5.3 0.70 (0.50-0.99) 0.043 0.67 (0.54-0.84) 0.001 1.00 (0.75-1.32) 0.981
Ooph <=44 10.1 13.0 10.4 8.6 8.0 0.94 (0.68-1.29) 0.698 0.91 (0.75-1.12) 0.383 0.71 (0.52-0.97) 0.034
Ooph 45-49 5.2 5.2 5.4 6.3 3.9 0.60 (0.38-0.97) 0.035 0.81 (0.62-1.05) 0.114 0.80 (0.55-1.16) 0.235
ooph >=50 2.8 2.3 3.3 3.1 2.2 0.88 (0.56-1.40) 0.6 0.63 (0.45-0.88) 0.007 0.65 (0.41-1.03) 0.065
Hyst<=44 14.2 22.2 12.5 9.2 15.7 0.76 (0.57-1.02) 0.07 0.87 (0.73-1.05) 0.142 0.99 (0.76-1.27) 0.907
Hyst45-49 3.6 4.4 3.7 2.8 3.7 0.67 (0.42-1.05) 0.077 0.84 (0.64-1.09) 0.195 0.80 (0.55-1.17) 0.257
Hyst >=50 1.4 1.4 1.7 1.5 1.7 0.91 (0.53-1.54) 0.714 0.76 (0.52-1.10) 0.142 1.28 (0.84-1.93) 0.247
Hormone Therapy Usage
Never 34.3 54.0 49.9 39.1 52.9 1.00 1.00 1.00
Past Estrogen 12.8 18.1 13.8 9.7 14.3 0.92 (0.80-1.05) 0.198 1.01 (0.94-1.09) 0.721 0.95 (0.85-1.06) 0.374
Past Prog. 1.1 1.1 1.1 0.8 1.1 1.21 (0.62-2.38) 0.576 0.69 (0.38-1.26) 0.226 0.60 (0.27-1.37) 0.229
Past EP 6.6 4.4 5.2 5.3 5.2 0.88 (0.65-1.20) 0.416 0.91 (0.75-1.10) 0.339 1.09 (0.88-1.35) 0.435
Cur. Estrogen 25.1 16.4 17.5 22.3 16.7 0.86 (0.79-0.93) <0.001 0.90 (0.86-0.95) <0.001 0.97 (0.91-1.03) 0.273
Cur. Prog. 0.6 0.4 0.5 0.8 0.6 0.08 (0.00-6.00) 0.255 0.70 (0.36-1.34) 0.282 1.18 (0.73-1.90) 0.499
Cur. EP 19.5 5.7 12.0 21.9 9.1 0.81 (0.68-0.97) 0.02 0.79 (0.70-0.89) <0.001 0.93 (0.82-1.06) 0.278
77
W – Whites; AA – African Americans; NH – Native Hawaiians; JA – Japanese Americans; LA – Latinos; AMI – acute myocardial infarction; OHD –
other heart disease; RR – relative risk; BMI – body mass index. Age-standardized (5-year age groups) to the total female population in the study.
RRs for risk factors are shown estimated while simultaneously adjusted for each other.
78
Established risk factors in women
Age-standardized proportions of obesity (BMI > 30) were highest in African American
women (35.5%) and lowest in Japanese American women (6.3%). Increased BMI was
associated with increased risk of AMI (P trend = 0.042), OHD (P trend = 0.023), and stroke
(P trend = 0.53), though the only category reaching statistical significance was individuals
with extreme obesity (BMI > 35). Hypertension and diabetes were highest in African
American (59.3% and 13.3%) and Native Hawaiian women (50.0% and 13.6%) and were
associated with a 1.4 to 3.0-fold statically significant increase in the risk of AMI, OHD
and stroke. 68.4% of Japanese American women and 66.7% of Latina women reported
never having smoked compared to only 45-48% in the other populations with African
American and Native Hawaiian women having the highest percentage of current
smokers (~20%). Both light (< 20 pack-years) and heavy (>= 20 pack-years) smoking were
associated with statistically significant increases in risk of death for all outcomes (RR’s
2.10 – 3.05), with the strength of the associations being greatest for current smokers
(Table 3.3).
Inverse associations were also noted with alcohol consumption which were significant
for OHD (P trend = 0.03) and for moderate consumption of alcohol (0-12g/day) with
stroke. 13.2% of White women report >24grams/day versus <5% for the other
populations. Vigorous activity was also inversely associated with OHD, with African
American, Japanese American, and Latina women reporting a greater percentage (>61%)
of inactivity. As in men, White women report more years of schooling than other
79
racial/ethnic groups, though this association was only non-statistically significantly
inversely associated with AMI and stroke (Table 3.3)
Like for men, African American and Latina women reported the highest consumption of
saturated fat and cholesterol which were associated with statistically significant
increases in risk of OHD and stroke (for saturated fat) among women (RRs, 1.19-1.49).
Increasing consumption of saturated fat and cholesterol also led to increased risks of
OHD (P trend = 0.026 and <0.001, respectively). African Americans were more likely to
have a hysterectomy prior to the age of 44 as opposed to natural menopause or
hysterectomy/oophorectomy at any other age, whereas all other racial/ethnic groups
were more likely to have a natural menopause between 50-54 (23.8% - 32.9%) than
natural menopause or hysterectomy/oophorectomy at any other age. Compared to
women with a natural menopause <44 years, having a natural menopause beyond 55
years of age was inversely associated with AMI (RR 0.70, 95% CI 0.50-0.99) and OHD (RR
0.67, 95% CI 0.54-0.84), but not stroke (RR 1.00, 95% CI 0.75-1.32). Hormone therapy
use was greater for White and Japanese American women with current estrogen or
estrogen plus progesterone use associated with significantly reduced risk of AMI and
OHD (P < 0.02) (Table 3.3).
80
Table 3.4. Observed and adjusted RR and 95% confidence interval of mortality in men by race in the MEC.
AMI W AA NH JA LA
RR 1.00 1.55 1.58 0.77 1.04
(1.29-1.86) (1.22-2.03) (0.65-0.92) (0.87-1.24)
RR + risk factors 1.00 0.99 1.12 0.63 0.79
(0.81-1.20) (0.86-1.46) (0.52-0.76) (0.66-0.96)
RR + risk factors + 1.00 1.00 1.20 0.66 0.79
nutrients (0.82-1.21) (0.91-1.59) (0.53-0.82) (0.65-0.97)
OHD W AA NH JA LA
RR 1.00 1.67 2.10 0.84 1.11
(1.50-1.87) (1.82-2.42) (0.76-0.94) (1.00-1.24)
RR + risk factors 1.00 1.14 1.52 0.73 0.91
(1.02-1.28) (1.31-1.77) (0.65-0.81) (0.81-1.02)
RR + risk factors + 1.00 1.16 1.60 0.80 0.91
nutrients (1.03-1.31) (1.37-1.88) (0.70-0.91) (0.80-1.03)
Stroke W AA NH JA LA
RR 1.00 1.86 1.65 1.21 1.24
(1.53-2.27) (1.23-2.21) (1.01-1.45) (1.03-1.51)
RR + risk factors 1.00 1.45 1.27 1.02 1.06
(1.18-1.78) (0.94-1.72) (0.84-1.24) (0.87-1.31)
RR + risk factors + 1.00 1.49 1.27 0.99 1.11
nutrients (1.21-1.83) (0.93-1.74) (0.79-1.24) (0.89-1.38)
W – Whites; AA – African Americans; NH – Native Hawaiians; JA – Japanese Americans; LA –
Latinos; AMI – acute myocardial infarction; OHD – other heart disease; RR – relative risk. See
methods for risk factors and nutrients included.
81
Table 3.5. Observed and adjusted RR and 95% confidence interval of mortality in women by race in the
MEC
AMI W AA NH JA LA
RR 1.00 2.33 1.62 0.77 1.25
(1.90-2.85) (1.15-2.27) (0.61-0.97) (0.99-1.58)
RR + risk factors 1.00 1.47 1.08 0.65 0.89
(1.18-1.83) (0.77-1.53) (0.51-0.84) (0.70-1.15)
RR + risk factors + 1.00 1.51 1.22 0.75 0.88
nutrients (1.21-1.89) (0.85-1.76) (0.56-1.01) (0.67-1.14)
OHD W AA NH JA LA
RR 1.00 1.85 2.10 0.62 1.08
(1.62-2.10) (1.75-2.53) (0.53-0.71) (0.93-1.25)
RR + risk factors 1.00 1.19 1.58 0.61 0.89
(1.03-1.37) (1.30-1.92) (0.51-0.71) (0.76-1.04)
RR + risk factors + 1.00 1.22 1.74 0.68 0.90
nutrients (1.06-1.41) (1.41-2.13) (0.57-0.83) (0.76-1.06)
Stroke W AA NH JA LA
RR 1.00 1.48 1.24 0.87 1.05
(1.24-1.77) (0.91-1.69) (0.73-1.04) (0.86-1.29)
RR + risk factors 1.00 1.15 1.02 0.83 0.95
(0.95-1.41) (0.74-1.41) (0.67-1.02) (0.76-1.18)
RR + risk factors + 1.00 1.16 1.06 0.88 0.94
nutrients (0.95-1.42) (0.76-1.47) (0.69-1.12) (0.74-1.18)
W – Whites; AA – African Americans; NH – Native Hawaiians; JA – Japanese Americans; LA –
Latinos; AMI – acute myocardial infarction; OHD – other heart disease; RR – relative risk. See
methods for risk factors and nutrients included.
82
Figure 3.1a. Crude and adjusted RRs of mortality from AMI, OHD, and stroke by race in men in
the MEC. †: P<0.05; ‡: P < 0.001
Figure 3.1b. Crude and adjusted RRs of mortality from AMI, OHD, and stroke by race in women
in the MEC. †: P<0.05; ‡: P < 0.001
0
0.5
1
1.5
2
2.5
3
AA NH JA LA AA NH JA LA AA NH JA LA
AMI OHD STROKE
Crude HR + Known Risk Factors + Nutrients
‡
† †
†
‡
‡
‡
‡
‡
†
†
†
† †
‡
‡
‡
‡
‡
‡
‡
‡
‡
0
0.5
1
1.5
2
2.5
3
AA NH JA LA AA NH JA LA AA NH JA LA
AMI OHD STROKE
Crude HR + Known Risk Factors + Nutrients
‡
†
†
‡
‡
‡
‡
†
‡
‡
‡
‡ ‡
‡
‡
‡
‡
83
Racial/Ethnic Comparisons of CVD Mortality
AMI: The highly statistically significant 55% excess risk in African American men when
compared to Whites was eliminated after adjustment for known risk-factors, while the
58% excess risk in Native Hawaiian men was reduced to 20% (P=0.19). The reduced risk
in Japanese American men further decreased from 0.77 to 0.66 (95% CI 0.53-0.82,
P<0.001) after adjustment (Table 3.4). Similarly, in women, differences in risk of death
due to AMI in African American and Native Hawaiian women compared to Whites were
attenuated, but the excess risk in African Americans remained highly statistically
significant (RR 1.51; p<0.001), and was reduced in Native Hawaiians from a 62% excess
risk to a non-statistically significant increase (RR=1.22, P=0.28). The reduction in risk in
Japanese American women remained the same (Table 3.5).
OHD: In men, ethnic differences in risk due to OHD were reduced across populations,
but remained elevated in African Americans (RR 1.16, 95% CI 1.03-1.31, P=0.012) and
Native Hawaiians (RR 1.60, 95% CI 1.37-1.88, P<0.001) (Table 3.4). In women, excess risk
of mortality due to OHD remained 73% higher in Native Hawaiians (95% CI 1.41-2.13,
P=0.007), and 22% higher in African American women (95% CI 1.06-1.41), compared to
Whites. Japanese American women continued to have a statistically significant reduced
risk of OHD (RR 0.68, 95% CI 0.57-0.83, P<0.001) (Table 3.5).
Stroke: In men, the age-adjusted risk of death due to stroke was significantly elevated
across all populations (including Japanese American men) when compared to Whites (all
races/ethnicities P<0.037). Adjustment for known risk factors reduced these differences
84
(all comparisons versus Whites, P>0.05), except for a statistically significant greater risk
that was still observed in African Americans (RR 1.49, 95% CI 1.21-1.83, P<0.001) (Table
3.4). In women, following adjustment for known risk factors, only non-significant
modest differences were observed in African Americans (RR=1.16, P=0.14) and Japanese
(RR=0.88, P=0.28) (Table 3.5).
Cholesterol-Lowering Medication Sub-Group Analysis
There were 31,565 men and 30,785 women with information about use of cholesterol
lowering medication (see Methods), and 167 AMI-related deaths, 436 OHD related
deaths, and 133 stroke related deaths in men, and 85 AMI, 245 OHD, and 144 stroke
related deaths in women in this subgroup. Among men, 46.8% of Japanese Americans
and 43.6% of Native Hawaiians reported cholesterol medication usage, which was ~10%
greater than all other ethnicities. Similar trends were seen in women, where 44.6% of
Japanese Americans and 41.6% of Native Hawaiians reported current cholesterol
medication usage (Table B.1). AMI and OHD cases were combined for additional power,
and revealed that ever-use of cholesterol medication (past and current use) was not
associated with AMI+OHD mortality in men (RR 1.06, 95% CI 0.90-1.25, P=0.51) or
women (RR 1.10, 95% CI 0.88-1.37, P=0.42). Ever use of cholesterol medication was also
not statistically significantly associated with stroke-related mortality in men (RR 0.86,
95% CI 0.61-1.22, P=0.40) or women (RR 0.93, 95% CI 0.66-1.31, P=0.68) (Table B.2).
As in the overall cohort analysis, adjusting for known risk factors attenuated ethnic-
specific differences in risk of AMI, OHD, and stroke across all ethnicities when compared
85
to Whites. Furthermore, adjusting for cholesterol medication usage did not affect the
risk estimates for any outcome, with general trends of excess risk compared to Whites
that are similar to that observed in the overall cohort. Stratification by never/ever-use
of cholesterol medication did not change conclusions for AMI and OHD mortality by
ethnicity (data not shown).
Risk Factor Contributions to Racial/Ethnic Differences
The results from a systematic examination of each risk factor and their relative effect on
racial/ethnic differences in risk are shown in Appendix Table B.3. In African American
and Native Hawaiian men and women, hypertension, diabetes, and education status
accounted for the vast majority of the ethnic differences in risk of AMI, OHD and stroke
compared to Whites (9%-17% change in RR per risk factor). In women, HRT usage also
accounted for a large fraction (2%-13%) of the ethnic differences. In Latino men and
women, where crude RRs for CVD risk were similar to those in Whites, adjustment for
education, diabetes and smoking status, actually increased CVD risk by 10%-17%
compared to Whites. In Japanese Americans, the crude RRs indicated a reduced risk of
AMI, OHD, and stroke relative to Whites. In men, adjusting for saturated fat intake
reduced the difference in risk with respect to Whites by 14% for AMI, 12% for OHD, and
4% for stroke. However, adjustment for all other risk factors in aggregate (especially
education status and diabetes) resulted in a substantial reduction in CVD risk in
Japanese men. Similarly in women, adjusting for smoking status and saturated fat
consumption reduced ethnic differences in risk, however further adjustment for alcohol
86
consumption, diabetes, and other risk factors resulted in reduced risk in Japanese
American women versus Whites (Table B.3).
Analysis of Lipid Levels
Next, in a sample of ~5,600 men and women from these 5 racial/ethnic groups we
examined whether mean lipid levels may be correlated with CVD risk (Table 3.65).
African American men and women had statistically significantly elevated levels of total
cholesterol and LDL-c with respect to Whites (P<0.05). In Japanese Americans, both men
and women had statistically significantly elevated levels of total cholesterol and
triglyceride (P<0.05). Latino men and women had statistically significantly reduced levels
of HDL-c, elevated levels of total cholesterol, triglycerides, and a greater TC:HDL ratio
(P<0.05). Native Hawaiians have a similar lipid profiles to Whites (Table 3.6 & Figure
3.2).
87
Table 3.6. Least-squared lipid means by race/ethnicity and sex in the MEC.
Men N
1
W AA NH JA LA P (F)
2
TC (SE)
3
2330 170.1 (2.4) 182.6 (2.3)* 170.4 (2.4) 179.7 (2.0)* 178.7 (1.3)* <0.001
HDL-c (SE)
3
2332 44.1 (0.9) 45.3 (0.8) 43.1 (0.9) 45.8 (0.7) 39.1 (0.5)* <0.001
TRIG (SE)
3
2331 94.1 (1.0) 88.2 (1.0) 101.6 (1.0) 115.7 (1.0)* 123.7 (1.0)* <0.001
LDL-c (SE)
3
2321 103.7 (2.3) 117.8 (2.1)* 104.3 (2.3) 107.4 (1.9) 110.7 (1.2)* <0.001
TC:HDL 2329 3.9 4.1 4.0 4.0 4.7* <0.001
Women N
1
W AA NH JA LA P (F)
2
TC (SE)
3
2235 188.4 (2.5) 193.6 (2.3) 188.5 (2.5) 197.3 (2.2)* 195.6 (1.4)* <0.001
HDL-c (SE)
3
2234 52.3 (1.1) 55.6 (1.0)* 49.4 (1.1) 54.0 (1.0) 47.1 (0.6)* <0.001
TRIG (SE)
3
2234 104.1 (1.0) 84.0 (1.0)* 118.3 (1.0)* 125.5 (1.0)* 136.8 (1.0)* <0.001
LDL-c (SE)
3
2227 112.2 (2.4) 118.8 (2.2) 111.6 (2.3) 115.1 (2.1) 117.8 (1.4) 0.002
TC:HDL 2233 3.7 3.6 3.9 3.7 4.3* <0.001
TC – total cholesterol; HDL-c – high-density lipoprotein cholesterol; TRIG – triglyceride; LDL-c –
low-density lipoprotein cholesterol; TC:HDL – TC to HDL ratio; W – Whites; AA – African
Americans; NH – Native Hawaiians; JA – Japanese Americans; LA - Latinos. Least-squared means
calculated using generalized linear models, adjusting for age, fasting hours, phase of blood draw,
time of blood draw, BMI, hypertension status, lipid medication usage, smoking status, alcohol
consumption, saturated fat and cholesterol consumption. *Pairwise difference with Whites as
baseline, P<0.05.
1
Sample size not similar for all lipids due to missing values.
2
P-value for Type III
sum of squares F-test.
3
mg/dL
88
Figure 3.2. Calculated least-squared means of lipid levels by sex in the MEC.
TC – total cholesterol; HDL-c – high-density lipoprotein cholesterol; TRIG – triglyceride; LDL-c –
low-density lipoprotein cholesterol; TC:HDL – TC to HDL ratio; W – Whites; AA – African
Americans; NH – Native Hawaiians; JA – Japanese Americans; LA - Latinos. Least-squared means
calculated using generalized linear models, adjusting for age, fasting hours, phase of blood draw,
time of blood draw, BMI, hypertension status, lipid medication usage, smoking status, alcohol
consumption, saturated fat and cholesterol consumption. *Pairwise difference with Whites as
baseline, P<0.05.
0
20
40
60
80
100
120
140
160
180
200
TC HDL-c TRIG LDL-c
Lipid levels in Men (mg/dL)
W AA NH JA LA
*
* *
*
*
* *
*
0
50
100
150
200
250
TC HDL-c TRIG LDL-c
Lipid levels in Women (mg/dL)
W AA NH JA LA
*
*
*
*
*
*
*
*
89
3.5 Discussion
In this large prospective study of five racial/ethnic groups, we found highly significant
differences in age-adjusted risk of CVD (AMI and OHD) and stroke mortality across
racial/ethnic groups. Age-adjusted mortality of CVD was greater in African American and
Native Hawaiian populations (RRs 1.55 – 2.10), reduced in Japanese Americans (RRs 0.62
- 0.84), and similar in Latinos when compared to Whites (Table 3.1). Most, but not all of
these differences were greatly attenuated, and in some cases eliminated, when
adjusting for known CVD risk factors, many of which are modifiable. When compared to
Whites, excess risk of AMI in African American men, and Native Hawaiian men and
women, was largely eliminated. However, excess risk of OHD remained in African
American men (16% excess) and Native Hawaiian men (60% excess) and women (73%
excess). Japanese American men and women had significantly reduced risks of AMI and
OHD mortality when compared to Whites. Latinos had a similar risk to that in Whites
after adjustment for known risk factors. For stroke-related deaths, the age-adjusted
mortality in men was statistically significantly greater for all racial/ethnic groups while in
women a significant difference was only observed in African Americans. After
adjustment for known risk factors, a statistically significant excess risk was only
observed in African American men (49% excess).
The differences in risk of death by race/ethnicity in this study are consistent with
national statistics and several studies summarized by the American Heart Association, as
well as a previous smaller study within the MEC (6,29). While most studies examine all-
cause mortality or CVD/stroke incidence, very few studies have specifically examined
90
the effect of known risk factors on racial/ethnic specific differences in CVD and stroke-
related mortality. Among African Americans in the Atherosclerosis Risk in Communities
(ARIC) study, elevated risk factors including blood pressure, cholesterol levels, diabetes
status, and smoking status could account for approximately 85% of the elevated risk of
CVD/stroke, when compared to Whites(18). In African Americans in the MEC risk of
stroke was greatly reduced in men and eliminated in women, similar to that found in
ARIC and the Reasons for Geographic and Racial Differences in Stroke Study (REGARDS),
where adjustment for stroke risk factors resulted in a more than 40% decrease in the
disparity of incident stroke(19,20,30). In NHANES (1999 – 2004) CVD and stroke-related
mortality risk was similar between Whites and African Americans after adjustment for
age, gender, glomerular filtration rate (GFR), comorbidities, BMI, and income/marital
status(17). In the REACH cohort, African Americans had the highest risk of all-CV related
death, while Asian individuals had the lowest risk (P<0.001), though this difference was
not significant with respect to AMI and stroke(21). A meta-analysis of 17 studies
performed between 1950-2009 and encompassing ~100 million individuals reports an
OR of 0.67 (95% CI 0.57-0.78) between Hispanic and Non-Hispanic Whites, though most
of those studies do not adjust for known CV risk factors(13). Three studies that adjust
for known CVD risk factors report reduced risks of CV mortality similar to what we have
found in the MEC, though two of those studies returned non-statistically significant
results. An examination of the mortality database from the National Center for Health
Statistics (NCHS) observed reduced standardized mortality rates (SMR) of AMI and OHD,
91
and similar SMRs for stroke between Japanese and Whites, similar to that seen in the
MEC(5).
In this analysis, BMI, hypertension, diabetes, and smoking status, were all statistically
significant risk factors for AMI, OHD, and stroke in both men and women, as has been
extensively described (6,31). Increased risks of CVD and stroke were only apparent in
the obese category (BMI > 30), which is similar to research summarized by the AHA(32).
In the MEC population, these risk factors differ greatly by race/ethnicity with African
Americans and Native Hawaiians having greater proportions of obesity, which also
increases their risk of other CVD-related risk factors. African Americans and Native
Hawaiians have the highest rates of hypertension, followed by Japanese Americans,
with Whites and Latinos having the lowest. In contrast, NHANES data shows that Asians
have the lowest prevalence of hypertension(10). In the MEC, prevalence of type 2
diabetes is highest in African Americans, Native Hawaiians, and Latinos, followed by
Japanese Americans and Whites, trends which are in line with national survey data(33).
African Americans also reported smoking the most compared to other populations,
while Japanese Americans smoked the least.
The effect of physical activity as a CVD risk factor has been extensively studied in the
past, and varies across racial/ethnic groups(34). In the MEC, Whites and Native
Hawaiians reported the greatest amount of vigorous activity, with the lowest levels
reported in African Americans and Japanese, groups at opposite ends of CVD risk. In
concordance with previous results, alcohol consumption appeared to be protective in
this population, leading to a 30% decreased risk of AMI in men and a non-statistically
92
significant 21% decrease in women that consumed 12-24 g/day. A systematic review of
the effects of alcohol on CVD summarized a reduced risk of CVD related mortality
(pooled adjusted RR 0.75 in drinkers versus non-drinkers), with the lowest risk occurring
in those consuming 0.5-2 drinks per day(35,36). Years of education, here used as a proxy
for socioeconomic status (SES), was highest in Whites and Japanese Americans, while,
Latinos reported the fewest years.
Later age at natural menopause and oophorectomy was associated with reduced risks of
AMI and OHD while no statistically significant associations were noted with stroke.
White and Japanese American women were more likely to be current estrogen HRT
users (in 1993-1996) which was also found to be significantly associated with an
inversed risk of AMI and OHD, but not stroke. This conflicts with evidence from the
Heart and Estrogen/Progestin Replacement Study (HERS) and Women’s Health Initiative
(WHI), which highlights increased risk of stroke and cardiovascular events among
current HRT users(37,38). This may be explained by the “timing hypothesis”, which
suggests that HRT use at a younger age, “healthier” cardiovascular status, and proximity
to menopausal age all have positive cardiovascular benefits(39,40).
In general, African Americans and Native Hawaiians had a higher prevalence of CVD risk
factors when compared to Whites, while Japanese Americans had a lower prevalence
(except for diabetes). Hypertension, diabetes, and education status contribute the
greatest towards differences in risk of African Americans and Native Hawaiians versus
Whites. As described above, Latinos had higher frequencies of many risk factors
(diabetes, obesity, and less years of education) than Whites, yet they had a similar
93
mortality risk, an observation that has been noted previously and dubbed the “Hispanic
paradox”(13,41). In Japanese Americans, education, hypertension, diabetes status, and
alcohol consumption were the major risk factors leading to an even greater reduction in
risk. These results further strengthen the idea that targeting these top risk factors (e.g.
hypertension and diabetes) could make a major contribution in reducing ethnic
differences in risk of CVD.
As reported in the general population, cholesterol medication and statin usage varies by
racial/ethnic group with Japanese Americans reporting the highest level of cholesterol
medication usage, and African Americans reporting the lowest(42,43). Additional
adjustment for cholesterol medication usage did not account for the observed
differences in CHD-specific mortality. We also observed similar results when stratifying
by cholesterol medication usage, further suggesting that racial/ethnic differences in
usage cannot explain the lower risks in Japanese Americans, nor the increased risk in
African Americans. A possible explanation that requires further examination could be
the due to the slower metabolization of statins in Asian populations when compared to
other populations(44).
Given the well-established link between circulating lipid levels and CVD risk, we also
compared mean lipid levels in these racial/ethnic groups as possible factors that may
explain ethnic differences in CVD mortality(6,45,46). The patterns observed across
White, African American, and Latino populations were comparable to those reported in
NHANES, where Latinos also had a poor lipid profile (in both men and women), despite
having similar risk of CVD death to that in Whites(47). Unlike NHANES data, where
94
Asians had similar levels of total cholesterol to non-Hispanic Whites, Japanese
Americans in the MEC had higher total cholesterol and triglycerides levels in men and
women, compared to Whites(48). The general lack of a correlation between lipid
profiles and CVD mortality in these groups suggests that lipid levels are not a major
factor driving racial/ethnic differences in CVD risk.
There are a number of limitations in this study. Risk factors, such as hypertension and
type 2 diabetes, were from self-report at baseline and were not updated over the 14
years of follow-up. This potential misclassification is unlikely to affect the estimates or
racial/ethnic differences unless there was under-reporting that was differential with
respect to ethnicity which is unlikely. Information regarding quality of and access to
healthcare was not collected in the MEC. Insurance status, treatment strategies and
medication management/adherence differ by race/ethnicity, and can impact
racial/ethnic differences in outcome(49). Ethnic-specific categorizations of BMI to
classify risk not yet defined in all populations examined here (or for the use of
alternative markers, such as waist-to-hip ratio), which can lead to misclassification of
BMI in our analysis, accounting for some of the difference in Native Hawaiians and
Japanese Americans(50). Alternative markers/risk factors that differ by race/ethnicity
and could be included in future studies include waist-to-hip (WHR) ratio, and
cardiorespiratory fitness (which is a distinct, independent risk factor of CVD and
stroke)(51,52). Finally, despite being a large cohort with a large amount of person-time,
it was difficult to examine specific types of OHD due to small numbers of events.
95
In this large population-based prospective, multiethnic study we found that known CVD
risk factors account for a large fraction of racial/ethnic differences in death due to AMI,
OHD, and stroke. However, significantly greater mortality in African Americans and
Native Hawaiians and lower mortality in Japanese remain and deserve further
examination.
3.6 References
1. World Health Organization. Global Status Report On Noncommunicable Diseases 2014.
2014.
2. CDC/NCHS. LCWK9. Deaths, percent of total deaths, and death rates for the 15 leading
causes of death: United States and each State, 2013. 2014 1-26 p.
3. CDC/NCHS. LCWK10. Deaths, percent of total deaths and rank order for 113 selected
causes of death and Enterocolitis due to Clostridium difficile, by race and sex, United
States, 2013. 2013 1-15 p.
4. CDC/NCHS. LCWK11. Deaths, percent of total deaths and rank order for 113 selected
causes of death and Enterocolitis due to Clostridium difficile, by Hispanic origin, race for
non-Hispanic origin and sex, United States, 2013. 2014 1-15 p.
5. Jose PO, Frank ATH, Kapphahn KI, et al. Cardiovascular disease mortality in Asian
Americans. J. Am. Coll. Cardiol. 2014;64(23):2486–2494.
6. Mozaffarian D, Benjamin EJ, Go AS, et al. Heart Disease and Stroke Statistics—2015
Update. Circulation [electronic article]. 2015;131(4):e29–e322.
(http://circ.ahajournals.org/cgi/doi/10.1161/CIR.0000000000000152)
7. Williams JE, Massing M, Rosamond WD, et al. Racial disparities in CHD mortality from
1968-1992 in the state economic areas surrounding the ARIC study communities.
Atherosclerosis Risk in Communities. Ann Epidemiol. 1999;9(8):472–480.
8. Pandey AK, Blaha MJ, Sharma K, et al. Family history of coronary heart disease and the
incidence and progression of coronary artery calcification: Multi-Ethnic Study of
Atherosclerosis (MESA). Atherosclerosis. 2014;232(2):369–376.
9. Patel SA, Winkel M, Ali MK, et al. Cardiovascular Mortality Associated With 5 Leading Risk
Factors: National and State Preventable Fractions Estimated From Survey Data. Ann.
Intern. Med. [electronic article]. 2015;(http://annals.org/article.aspx?articleid=2362308).
(Accessed July 1, 2015)
10. Nwankwo T, Yoon SS, Burt V, et al. Hypertension among adults in the United States:
National Health and Nutrition Examination Survey, 2011-2012. NCHS Data Brief
[electronic article]. 2013;(133):1–8. (http://www.ncbi.nlm.nih.gov/pubmed/24171916)
96
11. Centers for Disease Control and Prevention. National Diabetes Statistics Report:
Estimates of Diabetes and Its Burden in the Epidemiologic estimation methods. Atlanta,
GA: 2014 2009-2012 p.
12. Graham G. Population-based approaches to understanding disparities in cardiovascular
disease risk in the United States. 2014;393–400.
13. Cortes-Bergoderi M, Goel K, Murad MH, et al. Cardiovascular mortality in Hispanics
compared to non-Hispanic whites: A systematic review and meta-analysis of the Hispanic
paradox. Eur. J. Intern. Med. [electronic article]. 2013;24(8):791–799.
(http://dx.doi.org/10.1016/j.ejim.2013.09.003)
14. Kurian AK, Cardarelli KM. Racial and ethnic differences in cardiovascular disease risk
factors: a systematic review. Ethn. Dis. 2007;17(1):143–152.
15. Gu, Qiuping Paulose-Ram, Ryne Burt, Vicki Kit B. Prescription Cholesterol-lowering
Medication Use in Adults Aged 40 and Over : United States , 2003 – 2012. NCHS Data Br.
■ No. 177 ■ December 2014 [electronic article]. 2014;(177):1–8.
(http://www.cdc.gov/nchs/data/databriefs/db177.pdf)
16. Mochari-Greenberger H, Liao M, Mosca L. Racial and ethnic differences in statin
prescription and clinical outcomes among hospitalized patients with coronary heart
disease. Am. J. Cardiol. [electronic article]. 2014;113(3):413–7.
(http://www.sciencedirect.com/science/article/pii/S0002914913021395). (Accessed June
30, 2015)
17. Kovesdy CP, Norris KC, Boulware LE, et al. Association of Race with Mortality and
Cardiovascular Events in a Large Cohort of US Veterans. Circulation [electronic article].
2015;CIRCULATIONAHA.114.015124.
(http://circ.ahajournals.org/lookup/doi/10.1161/CIRCULATIONAHA.114.015124)
18. Hozawa A, Folsom AR, Sharrett AR, et al. Absolute and attributable risks of cardiovascular
disease incidence in relation to optimal and borderline risk factors: comparison of African
American with white subjects--Atherosclerosis Risk in Communities Study. Arch. Intern.
Med. [electronic article]. 2007;167(6):573–9.
(http://archinte.jamanetwork.com.libproxy2.usc.edu/article.aspx?articleid=412002).
(Accessed July 6, 2015)
19. Howard VJ, Kleindorfer DO, Judd SE, et al. Disparities in stroke incidence contributing to
disparities in stroke mortality. Ann. Neurol. [electronic article]. 2011;69(4):619–627.
(http://doi.wiley.com/10.1002/ana.22385)
20. Howard G, Cushman M, Kissela BM, et al. Traditional risk factors as the underlying cause
of racial disparities in stroke: lessons from the half-full (empty?) glass. Stroke [electronic
article]. 2011;42(12):3369–3375.
(http://stroke.ahajournals.org/content/42/12/3369.full.pdf)
21. Meadows TA, Bhatt DL, Cannon CP, et al. Ethnic differences in cardiovascular risks and
mortality in atherothrombotic disease: insights from the Reduction of Atherothrombosis
for Continued Health (REACH) registry. Mayo Clin. Proc. [electronic article].
2011;86(10):960–7.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3184025&tool=pmcentrez
97
&rendertype=abstract)
22. Kolonel LN, Henderson BE, Hankin JH, et al. A multiethnic cohort in Hawaii and Los
Angeles: baseline characteristics. Am. J. Epidemiol. [electronic article]. 2000;151(4):346–
57. (http://www.ncbi.nlm.nih.gov/pubmed/10695593)
23. Ray KK, Seshasai SRK, Erqou S, et al. Statins and all-cause mortality in high-risk primary
prevention: a meta-analysis of 11 randomized controlled trials involving 65,229
participants. Arch. Intern. Med. [electronic article]. 2010;170(12):1024–31.
(http://archinte.jamanetwork.com/article.aspx?articleid=416105). (Accessed February
24, 2015)
24. Taylor F, Huffman MD, Macedo AF, et al. Statins for the primary prevention of
cardiovascular disease. Cochrane database Syst. Rev. [electronic article].
2013;1:CD004816. (http://www.ncbi.nlm.nih.gov/pubmed/23440795). (Accessed January
31, 2015)
25. Morimoto Y, Conroy SM, Ollberding NJ, et al. Erythrocyte membrane fatty acid
composition, serum lipids, and non-Hodgkin’s lymphoma risk in a nested case–control
study: the multiethnic cohort. Cancer Causes Control [electronic article].
2012;23(10):1693–1703. (http://link.springer.com/10.1007/s10552-012-0048-1)
26. Mora S, Rifai N, Buring JE, et al. Fasting compared with nonfasting lipids and
apolipoproteins for predicting incident cardiovascular events. Circulation [electronic
article]. 2008;118(10):993–1001.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2574817&tool=pmcentrez
&rendertype=abstract)
27. Sidhu D, Naugler C. Fasting Time and Lipid Levels in a Community-Based Population.
Arch. Intern. Med. [electronic article]. 2012;172(22):1707.
(http://archinte.jamanetwork.com/article.aspx?doi=10.1001/archinternmed.2012.3708)
28. Collaboration* TERF. Major lipids, apolipoproteins, and risk of vascular disease. Jama
[electronic article]. 2009;302(18):1993–2000.
(http://dx.doi.org/10.1001/jama.2009.1619)
29. Henderson SO, Haiman CA, Wilkens LR, et al. Established Risk Factors Account for Most
of the Racial Differences in Cardiovascular Disease Mortality. PLoS One [electronic
article]. 2007;2(4):e377. (http://dx.plos.org/10.1371/journal.pone.0000377)
30. Suzuki T, Voeks J, Zakai NA, et al. Metabolic Syndrome, C-reactive Protein, and Mortality
in U.S. Blacks and Whites: The Reasons for Geographic and Racial Differences in Stroke
(REGARDS) Study. Diabetes Care [electronic article]. 2014;37(August):dc13–2059–.
(http://care.diabetesjournals.org/content/early/2014/05/21/dc13-2059.abstract)
31. Buse JB, Ginsberg HN, Bakris GL, et al. Primary prevention of cardiovascular diseases in
people with diabetes mellitus: A scientific statement from the American Heart
Association and the American Diabetes Association. Circulation. 2007;115(1):114–126.
32. Lewis CE, McTigue KM, Burke LE, et al. Mortality, health outcomes, and body mass index
in the overweight range: A science advisory from the american heart association.
Circulation. 2009;119(25):3263–3271.
98
33. Facts F, Diabetes ON. National Diabetes Fact Sheet , 2011. Centers Dis. Control Prev. U.S.
Dep. Heal. Hum. Serv. 2011;CS217080A(Division of Diabetes Translation):1–12.
34. Pedersen BK, Saltin B. Exercise as medicine - evidence for prescribing exercise as therapy
in 26 different chronic diseases. Scand. J. Med. Sci. Sports [electronic article]. 2015;25:1–
72. (http://doi.wiley.com/10.1111/sms.12581)
35. Fernández-Solà J. Cardiovascular risks and benefits of moderate and heavy alcohol
consumption. Nat. Rev. Cardiol. [electronic article]. 2015;12(10):576–587.
(http://dx.doi.org/10.1038/nrcardio.2015.91)
36. Ronksley PE, Brien SE, Turner BJ, et al. Association of alcohol consumption with selected
cardiovascular disease outcomes: a systematic review and meta-analysis. Bmj [electronic
article]. 2011;342(feb22 1):d671–d671.
(http://www.bmj.com/cgi/doi/10.1136/bmj.d671)
37. Hulley S. Randomized Trial of Estrogen Plus Progestin for Secondary Prevention of
Coronary Heart Disease in Postmenopausal Women. Jama [electronic article].
1998;280(7):605. (http://jama.jamanetwork.com/article.aspx?articleid=187879)
38. Wassertheil-smoller S, Hendrix SL, Limacher M, et al. Effect of Estrogen Plus Progestin on
Stroke in Postmenopausal Women. 2003;289(20):2673–2684.
39. Harman SMS, Vittinghoff E, Brinton E a. E, et al. Timing and Duration of Menopausal
Hormone Treatment May Affect Cardiovascular Outcomes. Am. J. … [electronic article].
2011;124(3):199–205.
(http://www.sciencedirect.com/science/article/pii/S0002934310009265)
40. Rosano G, Vitale C, Spoletini I, et al. Cardiovascular health in the menopausal woman:
impact of the timing of hormone replacement therapy. Climacteric [electronic article].
2012;15(4):299–305. (http://www.ncbi.nlm.nih.gov/pubmed/22424090)
41. Ruiz JM, Steffen P, Smith TB. Hispanic mortality paradox: A systematic review and meta-
analysis of the longitudinal literature. Am. J. Public Health. 2013;103(3).
42. Miyauchi K, Ray K. A review of statin use in patients with acute coronary syndrome in
Western and Japanese populations. J. Int. Med. Res. [electronic article]. 2013;41(3):523–
536. (http://imr.sagepub.com.libproxy2.usc.edu/content/41/3/523.full). (Accessed June
30, 2015)
43. Lewey J, Shrank WH, Bowry ADK, et al. Gender and racial disparities in adherence to
statin therapy: a meta-analysis. Am. Heart J. [electronic article]. 2013;165(5):665–78,
678.e1. (http://www.sciencedirect.com/science/article/pii/S0002870313001385).
(Accessed June 30, 2015)
44. Pu J, Romanelli R, Zhao B, et al. Dyslipidemia in Special Ethnic Populations. Cardiol. Clin.
[electronic article]. 2015;33(2):325–333.
(http://linkinghub.elsevier.com/retrieve/pii/S0733865115000065)
45. Sarwar N, Danesh J, Eiriksdottir G, et al. Triglycerides and the Risk of Coronary Heart
Disease: 10 158 Incident Cases Among 262 525 Participants in 29 Western Prospective
Studies. Circulation [electronic article]. 2007;115(4):450–458.
(http://circ.ahajournals.org/cgi/doi/10.1161/CIRCULATIONAHA.106.637793)
99
46. Arsenault BJ, Boekholdt SM, Kastelein JJP. Lipid parameters for measuring risk of
cardiovascular disease. Nat. Rev. Cardiol. [electronic article]. 2011;8(4):197–206.
(http://dx.doi.org/10.1038/nrcardio.2010.223)
47. Carroll MD, Kit BK, Lacher D a, et al. Trends in lipids and lipoproteins in US adults, 1988-
2010. JAMA [electronic article]. 2012;308(15):1545–54.
(http://www.ncbi.nlm.nih.gov/pubmed/23073951)
48. Carroll MD, Kit BK, Lacher DA. Total and high-density lipoprotein cholesterol in adults:
National Health and Nutrition Examination Survey, 2009-2010. NCHS Data Brief
[electronic article]. 2012;92(92):1–8. (http://www.ncbi.nlm.nih.gov/pubmed/24165064)
49. Davis AM, Vinci LM, Okwuosa TM, et al. Cardiovascular health disparities: a systematic
review of health care interventions. Med. Care Res. Rev. [electronic article]. 2007;64(5
Suppl):29S–100S. (http://www.scopus.com/inward/record.url?eid=2-s2.0-
34648831043&partnerID=tZOtx3y1)
50. Rao G, Powell-Wiley TM, Ancheta I, et al. Identification of Obesity and Cardiovascular
Risk in Ethnically and Racially Diverse Populations. Circulation [electronic article].
2015;CIR.0000000000000223.
(http://circ.ahajournals.org/lookup/doi/10.1161/CIR.0000000000000223)
51. DeFina LF, Haskell WL, Willis BL, et al. Physical Activity Versus Cardiorespiratory Fitness:
Two (Partly) Distinct Components of Cardiovascular Health? Prog. Cardiovasc. Dis.
[electronic article]. 2015;57(4):324–329.
(http://linkinghub.elsevier.com/retrieve/pii/S0033062014001406)
52. Myers J, McAuley P, Lavie CJ, et al. Physical Activity and Cardiorespiratory Fitness as
Major Markers of Cardiovascular Risk: Their Independent and Interwoven Importance to
Health Status. Prog. Cardiovasc. Dis. [electronic article]. 2015;57(4):306–314.
(http://linkinghub.elsevier.com/retrieve/pii/S0033062014001431)
100
CHAPTER 4: MULTIETHNIC GENOME-WIDE ASSOCIATION STUDY OF
GALLBLADDER DISEASE
4.1 Abstract
Background: Gallbladder disease (GBD) prevalence differs by race/ethnicity and has a
heritable component.
Methods: We conducted a genome-wide association study (GWAS) in the Multiethnic
Cohort (MEC) and SIGMA consortium which includes 3,683 GBD cases and 13,339
controls in African American, Japanese American and Latino men and women. In regions
with significant markers, we performed conditional analyses to identify secondary
signals.
Results: We identified 66 genome-wide significant (P<5.0x10
-8
) risk variants that were
limited to five regions. The strongest signal genome-wide was with a missense SNP
(D19H) at 2p21 in the ABCG5/8 gene region that had a heterogeneous effect across
populations (rs11887534, OR meta = 1.62, P meta=1.05x10
-20
, risk allele frequency (RAF) in
Latinos=0.1 and African Americans=0.07, P het=0.008), and where multiple independent
signals were observed (rs4963026, OR meta=1.49, P meta=1.29x10
-15
). We also detected a
Japanese-specific locus in the intron of the SDK1 gene at 7p22.2 (rs13226493,
OR JA=2.32, P JA=2.39x10
-8
, RAF JA=0.03) as well as two African-specific risk loci at 14q21.3
101
(rs72672032, OR AA=2.43, P AA=2.01x10
-8
, RAF AA=0.03) and 15q21.3 (rs7177754,
OR JA=1.64, P JA=3.95x10
-8
, RAF AA=0.12) in intergenic regions.
Conclusion: In a meta-analysis of three races/ethnicities the ABCG5/8 region was found
to be the strongest susceptibility region for GBD, however the impact of this locus with
respect to effect size and frequency varied widely across populations. Additional
racial/ethnic-specific loci on chromosomes 7, 14, and 15 highlight racial/ethnic
differences in the genetic contribution to GBD risk.
102
4.2 Introduction
Gallbladder disease (GBD) is an increasingly common set of conditions including
cholelithiasis and cholecystitis, whose prevalence has increased to over 20% of the adult
population in the past few decades. Almost 2 million ambulatory care visits and over
750,000 elective cholecystectomy operations occur as a result of GBD in the United
States (1,2). In the U.S., cholesterol gallstones account for the majority of GBD, with the
prevalence ranging between 10%-40% across racial/ethnic populations. Increasing
Amerindian ancestry is associated with greater rates of GBD in comparison to African
Americans and Asians (3,4). Female sex and parity are associated with risk of GBD, as are
various metabolic conditions including diabetes, dyslipidemia, and metabolic syndrome
(abdominal obesity, high blood pressure, increased triglycerides and/or decreased high
density lipoprotein (HDL) levels) (5,6). Other, less common risk factors include cirrhosis,
cystic fibrosis, Crohn’s disease, and various medications such as statins and thiazides (6).
Twin and family-based studies have demonstrated that the additive genetic heritability
of gallstones ranges from 15%-25% (3,7). GWAS of GBD have revealed a small number of
susceptibility loci (8–10). The strongest association has been linked to common variants
in the ABCG5/8 genes that codes for a heterodimeric cholesterol transporter. The
ABCG5/ABCG8 transporters increase cholesterol secretion into the bile, and decrease
cholesterol absorption in partnership with NPC1L1 (11). The strongest risk variant in this
region is a single nucleotide polymorphism (SNP) at rs11887534, a D19H missense
polymorphism, was discovered in 895 cases of Germans and Chileans, with the signal
replicated in a Swedish twin study (12,13). In analyses conditional on the top SNP, as
103
many as four other independent risk SNPs (at p<10
-4
) have been reported in the same
region (9,10). This D19H coding variant varies in frequency across populations, from 10%
in Amerindian populations to 1% in east Asians, and confers a two-fold risk of GBD
(14,15). A handful of GBD risk variants have also been reported (at P<0.5) in studies of
known lipid susceptibility regions, including a enterohepatic cholesterol transporter
associated with levels of plasma cholesterol (SLC10A2, rs9514089) in a German
cohort(8,16), variation at the UGT gene region associated with bilirubin secretion in a
meta-analysis of 3 different ethnic studies(8,17), and variants in genes GCKR and
TTC39B, associated with dyslipidemia in women of European ancestry (18,19). Given the
increased susceptibility of GBD in females, a candidate gene study in a North Indian
population of 230 gallstone patients identified a potential risk variant at ESR1
(rs2234693, rs9340799), and another in 226 patients identified a risk SNP in SLCO1B1
(20,21).
Additional genetic studies in diverse populations are necessary to replicate known and
identify novel susceptibility variants for GBD that are important globally. In this study,
we conducted a large-scale genome-wide analysis of GBD using data from the
Multiethnic Cohort Study (MEC) and Slim Initiative for Genomic Medicine in the
Americas (SIGMA), including 4,433 African Americans, 3,696 Japanese Americans, and
8,893 Latinos. Given that elevated lipid levels (or reduced HDL-c levels) are associated
with risk of GBD, we also examined 157 known risk SNPs associated with triglycerides,
LDL cholesterol, HDL cholesterol, and total cholesterol levels as markers of GBD
risk(4,5,10).
104
4.3 Methods
Study participants
This investigation of genetic contributions to risk of GBD includes cases and controls
from the Multiethnic Cohort and the SIGMA consortium. Epidemiological and clinical
characteristics of the studies are summarized in Table C.1 and described briefly below.
In total, we analyzed 3,683 GBD cases and 13,339 GBD controls among 3 studies
encompassing African Americans, Japanese Americans, and Latinos.
The Multiethnic Cohort (MEC)
The MEC is a population-based cohort study comprised of ~215,000 individuals of
different races/ethnicities from California and Hawaii. Details of this cohort are provided
elsewhere (22). A baseline questionnaire (1993-1996) collected information about risk
factors for chronic disease including dietary and medical history. A short follow-up
questionnaire was administered in 2001 to update medical conditions, and a third
detailed questionnaire was re-administered between 2003-2007. Blood specimens from
~67,000 individuals were obtained for genetic and biomarker studies. This study
includes data from case-control studies created previously for genetic studies of breast
and prostate cancer in African Americans (2,905 males and 1,528 females), Japanese
Americans (1,985 males and 1,711 females), and Latinos (1,743 males and 957
females)(23–28). Individuals with cancer were identified via linkage with Surveillance,
Epidemiology, and End Results (SEER) cancer registries in Hawaii and California.
Additional health outcomes were determined through linkage with California Hospital
105
Discharge Data (CHDD) and Medicare & Medicaid Services Provider Analysis and Review
(MedPAR, outpatient) claim data.
Slim Initiative in Genomic Medicine for the Americas (SIGMA)
Additional genomic data were included from the SIGMA type 2 diabetes (T2D) initiative
which included Latino T2D cases and controls from the MEC, the Diabetes in Mexico
Study (DMS), and the Mexico City Diabetes Study (MCDS) (29). The primary goal of this
consortium was to characterize the genetic basis of type 2 diabetes (T2D) in populations
of Mexican ancestry. The DMS includes T2D cases 18 years or older with a previous
diagnosis of T2D or fasting glucose levels above 125 mg/dL, and T2D controls 45 years
and older with fasting glucose levels below 100 mg/dL (29). GBD and genetic
information was available in 689 T2D cases and 469 T2D controls. The MCDS is a
population based study of men and women between 35-64 from 6 low-income census
tracts of Mexico City. It includes 287 T2D cases diagnosed based on fasting glucose
above 125 mg/dL, 2 hour post-75 gram glucose challenge above 200 mg/dL, or physician
diagnosed diabetes, and 613 controls (30). Individuals from the MEC include 2,055 T2D
cases and 2,082 controls with T2D status defined based on a self-report and use of
medication for T2D (29,31).
Outcome definition
In the MEC, GBD cases were defined by self-reported cholecystectomy or cholelithiasis
on the baseline questionnaire, the follow-up questionnaire, or the third questionnaire.
In addition, MEC linkage with CHDD (1990-2012) and Medicare (1999-2011) included
106
claims of incident cholecystectomy or cholelithiasis via International Classification of
Diseases (ICD) and current procedural terminology (CPT) codes. In the SIGMA studies,
GBD was defined based on physician’s report via ICD coding or ultrasonography. ICD, 9
th
revision (ICD-9) codes used to define GBD cases include 51.2 (cholecystectomy), 574
(cholelithiasis), and 575 (cholecystitis).
Genotyping and Quality Control
All African American samples in the MEC were genotyped using the Illumina Human1M-
Duo v3.0 BeadChip as previously described (25,28,32,33). Japanese American and Latino
samples in the MEC were genotyped using the Illumina Human660W-Quad v1.0
BeadChip array (25,28). All SIGMA studies were genotyped using the denser Illumina
HumanOmni2.5 array at the Broad Institute Genetic Analysis Platform (29). Individuals
included in both the MEC GWAS and SIGMA consortium were excluded from the MEC
nested case-control studies listed above, as the SIGMA data was genotyped on a higher
density array.
In the MEC, haplotype phasing was performed using SHAPEIT v2.r644, and imputation
for non-genotyped SNPs was performed using IMPUTE2 v2.3.0 to the March 2012 1000
Genomes cosmopolitan reference panel of 1092 individuals (34,35). For SIGMA studies,
pre-phasing was conducted with HAPI-UR version 1.01, and imputation with IMPUTE 2.2
using the 1000 Genomes Phase I set (build 37, August 2012) (29). For all studies, indels
were excluded as well as SNPs with a minor allele frequency (MAF) < 1%, or imputation
INFO scores < 0.3. Greater than 14 million SNPs were examined in the analysis of
107
African Americans, and more than 7 million SNPs in Japanese Americans and Latinos.
Table C.2 outlines the number of SNPs analyzed in each sub-study, after filtering.
4.3.1 Statistical Analysis
Principal components analysis (PCA) for ancestry were estimated for each ethnicity
using EIGENSTRAT(36) (by sex, ethnicity/study-specific phenotype status), based
autosomal ancestry informative markers (AIMs) in typed data. SNPs in regions with long-
range LD were not used. For SIGMA studies, PCA was performed on standard genotypes
as previously described (29).
Single association analyses were performed on each SNP by GBD status using
unconditional logistic regression on the expected number of copies of the risk allele, or
allelic dosage, assuming a log-additive genetic model in PLINK v1.07 (37,38). We report
odds ratios (OR), 95% confidence intervals (95% CI), and P-values from logistic
regression adjusted for age (age of entry into the cohort for MEC samples, and age of
diabetes onset in SIGMA samples), the first 10 principle components, and sex (in SIGMA
studies). ORs and P-values for each racial/ethnic group are designated with the subscript
of the ethnic group (i.e. OR AA for African American OR, or P meta for multiethnic meta-
analysis P-value). The P value threshold for genome-wide significance in the GWAS
analysis was set at the traditional value of 5 x 10
-8
. Analyses were done separately by
phenotype (cancer or T2D status) as well as separately by race/ethnicity and sex, and
results were combined using meta-analysis with inverse-variance weighting via METAL
(39). Quantile-quantile (Q-Q) plots were examined to assess general concordance with a
null distribution of P-values. Overdispersion and underdispersion was examined with
108
genomic inflation factors calculated using the median χ
2
value for each study divided by
the chi-squared value for the expected median with 1 degree of freedom (~0.455).
Manhattan plots were constructed using the R package qqman(40). LocusZoom plots
were used for local visualization, and LD was calculated using the appropriate
race/ethnicity and 1000 Genomes March 2012 data (41). Locuszoom plots in the
multiethnic analysis use AmerIndian ancestry data since results are driven by the large
sample of Latinos. All analysis and data manipulation was done in R 3.2.3 (R Core Team,
Vienna, Austria) unless otherwise noted. Functional annotation was performed using
ANNOVAR to scan multiple databases, including the University of California, Santa Cruz
genome browser (42). SNPs in LD with candidate SNPs were identified and LD r
2
values
were calculated using rAggr (http://raggr.usc.edu/), using the 1000 Genomes Project
Phase 3, Oct. 2014 hg19 database.
Secondary signals in each region were examined via conditional analyses. Conditional
logistic regression was performed in each region conditioning on the top SNP in the
region. In regions identified in multiethnic populations, the top SNP was selected based
on the lowest P-value in the meta-analysis of all three ethnic populations. For regions
identified in only one race/ethnic population, the top SNP was selected based on the
lowest P-value of that particular racial/ethnic group.
In addition to novel risk variants, we examined the association between 19 previously
reported susceptibility alleles with MAF ≥ 1% in our study and GBD risk, collected from a
review of GBD genetic studies (8). ORs and P-values were extracted from the original
papers (9,10,15,20,43). We considered replication of a SNP if the direction of effect was
109
consistent to the previously published result, and if the P-value was nominally
statistically significant (<0.05). We also examined 157 SNPs associated with lipid trait
levels from the Global Lipids Genetics Consortium (GLGC) (44). Finally, given these
subjects were initially selected from studies of breast cancer, prostate cancer, or T2D,
sensitivity analyses were performed by only examining study-specific controls (i.e.
breast/prostate cancer and T2D controls). This analytical subset included 1,684 GBD
cases and 7,118 GBD controls.
4.4 Results
In total, 3,683 GBD cases and 13,339 controls were evaluated in this GWAS analysis
(African American: 669 cases, 3764 controls; Japanese Americans: 687 cases, 3009
controls; Latinos: 2327 cases, 6566 controls) (Table C.1). We observed little evidence of
inflation in the test statistics in any racial/ethnic population (African American
λ GIF=0.999, Japanese Americans λ GIF=0.996, Latino λ GIF=1.012, multiethnic meta-analysis
λ GIF=1.011), indicating a low likelihood of cryptic relatedness, or residual population
stratification (Supplementary Figure 1).
110
Table 4.1. Top single SNP associated with GBD by region in the multiethnic meta-analysis and ethnic-specific analyses
META AA JA LA
LOC SNP A1/A2 OR P RAF OR P RAF OR P RAF OR P GENE
2:44066247 rs11887534 C/G 1.62 1.05E-20 0.07 1.18 0.17 0.01 1.07 0.87 0.10 1.76 2.19E-22 ABCG5/8
3:65728243 rs17370322 T/C 2.21 1.70E-08 0.01 1.33 0.34 0 NA 0 0.02 2.55 4.07E-09 MAGI1
7:4293524 rs13226493 C/A 2.70 0.011 0.14 1.14 0.14 0.03 2.33 2.39E-08 0.20 1.02 0.65 SDK1
14:49517763 rs72672032 T/A 1.85 1.84E-06 0.03 2.44 2.01E-08 1 NA NA 0.01 1.08 0.75 Intergene
15:53251332 rs7177754 G/A 1.64 3.95E-08 0.12 1.64 3.95E-08 1 NA NA <0.01 NA NA Intergene
P-values for multiethnic meta-analysis. fixed-effects meta-analysis within each race/ethnicity or across all three race/ethnicities. Models are
adjusted for age, sex, study, study phenotype, and first 10 principal components.
Table 4.2. Top SNP in conditional analysis in the ABCG5/8 region of chromosome 2
SNP BP A1/A2 OR
1
P
1
OR
2
P
2
OR
3
P
4
OR
4
P
rs11887534 44066247 C/G 1.62 1.05E-20
rs4953026 44097367 T/C 1.49 1.29E-15 1.43 1.12E-12
rs185263492 43996767 T/G 1.22 1.12E-05 1.21 1.87E-05 1.20 6.14E-05
rs6544715 44089928 C/A 1.08 2.00E-03 1.02 0.44 1.11 1.30E-03 1.12 2.00E-04
P-values for meta-analysis across 3 ethnicities.
1
Baseline model adjusted for age, sex, study, first 10 principal components.
2
Model 2 additionally
adjusted for rs11887534.
3
Model 3 additionally adjusted for rs4953026.
4
Model 4 additionally adjusted for rs185263492.
111
We detected 66 SNPs that were genome-wide significant (P < 5x10
-8
) in multiethnic or
ethnic-specific analyses in five distinct regions across. Of these, 53 were in the 2p21
region involving the genes THADA, DYNC2LI1, and ABCG5/8, 3 were in the 7p22.2 region
near the SDK1 gene, 8 were in a gene desert at 14q21.3, 1 was in a gene desert region at
15q21.3 and 1 was found in isolation 3p14.1. Table 4.1 summarizes the top SNPs in each
of the five regions identified and Figure 1 shows Manhattan plots from the ethnic-
specific and multiethnic meta-analyses (Figure 1a, 1d).
2p21
The 2p21 region which includes the genes THADA, DYNC2LI1, and ABCG5/8 is an
established GBD susceptibility region (C.3). The most significant SNP in this region was
rs11887534 (OR LA = 1.62, P meta=1.05x10
-20
), which has been reported previously, and is a
nonsynonymous missense SNP located in exon 1 of ABCG8, a cholesterol transporter
protein (12). This polymorphism results in a replacement of aspartic acid with histidine,
which has been shown to influence cholesterol excretion from enterocytes and
hepatocytes (11,12,15,45). The risk allele frequency (RAF) of rs11887534 is common in
Latino populations (RAF 0.10), African Americans (RAF 0.08), and Whites (RAF 0.08) and
rare in Japanese Americans (≤1%). We also observed evidence of significant effect
heterogeneity with rs11887534 across ethnicities (Latinos: OR LA=1.76; P LA=2.19x10
-22
;
African Americans: OR AA=1.18, P AA=0.17; Japanese Americans: OR JA=1.07, P JA=0.87;
P eth=0.008) (Appendix Table C.9.xlsx). Two other genome-wide significant signals were
112
also found in THADA, which are variants detected further away than previously reported
rs6730757 in intron 31 (OR meta=1.18, P meta=3.99x10
-8
) and rs4952986 in intron 29
(OR meta=1.25, P meta=6.67x10
-9
) (Table C.3).
Figure 2a shows a LocusZoom plot of this region from the multiethnic analysis, revealing
the potential presence of a secondary signal. Following conditioning on rs11887534 in
multiethnic and ethnic-specific analyses, rs4953026 remained genome-wide significant
in this region (OR meta=1.43: P meta=1.12x10
-12
, RAF AA=0.89, RAF LA=0.85, P het=0.09; Table
4.2). Variant rs4953026 is in LD with a previously reported secondary signal found in
Whites in this region (rs4299376 conditional on rs1118: OR meta=1.40, P meta= 3.87x10
-15
;
r
2
= 0.70 in AMR populations) (9). Rs11556157, a common non-synonymous variant
(RAF AA=0.67, RAF JA=0.98, RAF LA=0.82) in DYNC2LI1 (A691T) also remained genome-wide
significant in Latinos (P LA=3.36x10
-8
) following conditioning on rs11887534, but this
variant was not associated with risk in any other population (P>0.6).
Analysis conditioning on rs11887534 and rs4953026 eliminated all signals below P≤10
-8
;
the next strongest associations were with rs185263492 (OR meta=1.20, P meta=6.14x10
-5
,
RAF AA=0.72, RAF JA=0.98, RAF LA=0.90) in the ME meta and rs77105521 in Latinos
(OR LA=1.35, P LA=1.54x10
-7
, Allele C RAF AA=0.87, RAF JA=0.83, RAF LA=0.88) (Table 4.2). The
non-synonymous variant in DYNC2LI1 (A691T, rs11556157) was no longer genome-wide
significant in Latino populations after conditional analysis with both rs11887534 and
rs4953026 (P LA=1.69x10
-6
).
3p14.1
113
We detected an isolated signal in intron 1 of MAGI1 with an imputed SNP, rs17370322,
at 3p14.1(RAF AA=0.01, OR AA=1.33, P AA=0.34; RAF LA=0.02, OR LA=2.55, P LA=4.07x10
-9
) (46).
No other variant had a P<1x10
-5
in the 500kb region. Rs17370322 was imputed in all
populations except Japanese American men, with INFO scores 0.69-0.81.
7p22.2
At 7p22.2, we identified a signal that was limited to Japanese Americans defined by 3
SNPs in intronic regions of SDK1 that are highly correlated with one another (r
2
>0.8 in
1KGP3 JPT) (Table C.4). The most significant variant rs13226493 (OR JA=2.33, P JA=2.39x10
-
8
), is less common in Japanese Americans (RAF 3%) compared to African Americans (RAF
13%-14%) and Latinos (RAF 18%-22%). As this signal appeared only in Japanese
Americans, we searched for SNPs in the region in 1000 Genomes Phase 3, as our
imputation used Phase 1, that might be in LD in Japanese populations (JPT, 1000
Genomes: r
2
> 0.4) with rs13226493 and rare or monomorphic in other populations
(Table C.5). Unfortunately, we were unable to find any SNPs in a window within ±500kb
of rs13226493 that displayed such a pattern. Of the 11 SNPs that were in weak LD with
rs13226493 (r
2
>0.4, 1KGP3 JPT), none were significant in African Americans or Latinos
(P>0.05), 3 were genome-wide significant, and 8 were marginally significant in Japanese
Americans (P JA<5.00x10
-5
). We observed only weak evidence of a secondary signal in an
intergenic region close to SDK1 when conditioning on rs13226493 (rs138769605;
OR JA=2.38, P JA=5.32x10
-5
, Allele C RAF AA=0, AF JA=0.03, AF LA=0.04).
14q21.3 and 15q21.3
114
We observed two genome-wide signals that were limited to African Americans (Table
C.6). At 14q21.3, 8 variants were found at P<5x10
-8
(Figure 2c) in an intergenic region
1.3Mb away from LINC00648 and over 500kb away from RPS29, with the top signal
rs72672032 (OR AA=0.41, P AA=2.01x10
-8
). These variants were present only in African
Americans (RAFs of 0.023-0.027), and Latinos (RAF 0.010-0.012) however none of these
variants were associated with risk in Latinos (P LA>0.6).
A signal was also observed at 15q21.3, with risk variants present only in African
Americans (Table C.6). The strongest associated variant was rs7177754 (OR AA=0.62, P AA=
3.93x10
-8
, Allele G RAF AA=0.12) followed by 2 other highly correlated SNPs (r
2
>0.8 in
1KGP3 AFR populations) at P < 10
-7
all of which are intergenic. The closest gene is
ONECUT1, which is over 100kb away from any SNP.
Examination of previously reported loci
Previous candidate gene studies have reported 19 variants in 15 regions (Table C.7)
(8,10,15,20,43). In this study, only two of these previously reported SNPs, located in
intronic regions of UGT1A1 and TTC39B, had ORs that were directionally consistent with
previous reports and were nominally statistically significant (P<0.05) (47). Of the 157
lipid-trait loci discovered by the GLGC, 37 were nominally significant in at least one of
our three racial/ethnic populations, or in the multiethnic meta-analysis (P<0.05, Table
C.8). Of these, only rs4299376 in the 2p21 region (discussed above) was genome-wide
significant (P meta=3.87x10-
15
, P LA=3.57x10
-16
).
115
Figure 4.1. Manhattan plots of GWA inverse-weighted meta-analysis. Blue line = 1.0 x 10
-5
, red line = 5.0
x 10
-8
.
(a) Meta-analysis across all races/ethnicities.
(b) Meta-analysis across African Americans.
116
(c) Meta-analysis across Japanese Americans.
(d) Meta-analysis across Latinos.
117
Figure 4.2. Regional associations for findings P < 5.0 x 10
-8
.
Centered on the top SNP in the region ±500kb unless otherwise noted. The most
significant SNP in each region is marked as a purple diamond. Bottom portion of the plot displays genes in the region. Plots produced using
LocusZoom v1.1.
a) Region at 2p21 encompassing genes ABCG5/8. Colors represent pairwise correlation (r
2
) for the SNPs in the region with the top SNP,
rs11887534, calculated using Amerindian (AMR) ancestry from 1000 genomes March 2012. AMR ancestry was chosen as this SNP is most
118
common in Latino populations.
119
b) Region at 7p22.2, present only in Japanese Americans in the MEC. Colors represent pairwise correlation (r
2
) for the SNPs in the region with the
top SNP, rs13226493, using Asian (ASN) ancestry from 1000 genomes March 2012.
120
c) Region at 14q21.3, present only in African Americans in the MEC. Colors represent pairwise correlation (r
2
) for the SNPs in the region with the
top SNP, rs72672032, using African (AFR) ancestry from 1000 genomes March 2012.
121
d) Region on chromosome 15, present only in African Americans in the MEC. Colors represent pairwise correlation (r
2
) for the SNPs in the region
with the top SNP, rs7177754, using African (AFR) ancestry from 1000 genomes March 2012.
122
Figure 4.3. Association plots of conditional analyses in the 2p21 region, surrounding ABCG5/8. See figure 2a for regional plot without
conditional SNPs. Color scheme indicates pairwise comparisons (r
2
) of SNPs with the top SNP in the region, using AMR ancestry from 1000
Genomes March 2012.
123
a) Local association plot conditioning on the top SNP in the region, rs11887534.
124
b) Local association plot, conditioning on both rs11887534 and rs4953026.
125
c) Third step of local association plot, conditioning on rs11887534, rs4953026, and rs7599564.
126
4.5 Discussion
In this large multi-ethnic GWAS of GBD, we identified five regions with genome-wide
significant signals. The most significant association was with the missense variant,
rs11887534, on chromosome 2p21 which has been reported in previous GWAS and is
the strongest known common genetic risk factor for GBD (9,12,13,20). In a conditional
analysis, we also identified secondary signals at the ABCG5/8 region. Three other
regions contained race/ethnic-specific signals that were genome-wide significant and
have not been previously reported.
We found 53 risk SNPs at genome-wide significance in the 2p21 region, encompassing
THADA, DYNC2LI1, and ABCG5/8. The top SNP, rs11887534 (D19H), codes for a missense
variant in exon1 of ABCG5/8, which is known as an ATP-binding cassette transporter,
and is expressed in enterocytes, the small intestine, and hepatocytes, and is responsible
for limiting sterol uptake, and promoting biliary secretion. The replacement of aspartic
acid with histidine is tolerated (SIFT score = 0.16) and neutral (PROVEAN score = -1.15),
and occurs in the initial intracellular portion of the 6-transmembrane protein, near the
N-terminus, potentially harming ATP-binding (48–51). In the scenario of impairment of
ABCG5/8 function due to mutation, ezetimibe can be used as a sterol absorption
indicator, potentially allowing therapy in affected individuals (45). We observed
evidence of effect heterogeneity with a larger per allele effect in Latinos than in African
Americans (OR LA = 1.76, OR AA = 1.18, P het=0.008). The second signal in this region,
rs4953026, is in intron 6 of 12 and linked with a variant associated with LDL and total
127
cholesterol levels, rs4299376 (r2=0.7 in 1KGP3 AMR) and more recently reported as a
risk variant for GBD in ~15,000 women of European ancestry (9). The previous GBD
study also reported 2 other weaker signals at rs6544718 and rs6720173, only one of
which replicated at nominal significance in our study (rs6544718, P meta=0.81;
rsrs6720173, P meta=0.046).
Neither of the two signals we observed in THADA are in LD with a previously reported
T2D and colorectal cancer variant, rs7578597, however both are in weak LD with
rs1465618 in Amerindian populations (rs6730757 r
2
= 0.44, rs4952986 r
2
= 0.59), which
has been associated with prostate cancer risk (52). DYNC2LI1 forms the dynein 2 light
intermediate chain, and along with ABCG5/8 is involved with sitosterolemia, or
increased uptake of sterols. This study successfully replicates the two top signals in the
ABCG5/8 susceptibility region in a multiethnic study of non-European ancestry, and
reveals putative additional signals may be present further upstream from the known
susceptibility region (ABCG5/8).
The loci at 7p22.2 was identified only in Japanese Americans, and has not previously
been reported in association with GBD. Three SNPs were found to be genome-wide
significant, all in the SDK1 region. SDK1 codes for a cell adhesion molecule that is
expressed in many tissues, and may have a role in podocyte-dysfunctional
glomerulosclerosis (53). A risk variant for childhood obesity in Hispanics has been
reported in this region (rs314590, P=4.00x10
-6
), however this variant is not correlated
with any of our genome-wide significant SNPs in our study (54). The strongest
associated SNP, rs13226493 (OR JA=0.43, P JA=2.39x10
-8
), is rare in Japanese Americans,
128
but common in both African Americans and Latinos. The lack of an association in these
other populations is not clear and may be due to a genetic or environmental modifier or
a functional allele that is limited to Japanese. Using 1000 Genomes Phase 3 data, we
identified 11 SNPs in LD with rs13226493 at r2>0.4, all of which were examined in our
study, however none of these SNPs were limited to Japanese. Replication testing of this
signal as well as additional genotyping and sequencing in this region is warranted to
understand the population-specificity of these findings.
Two regions were identified in African American populations only. The first, at 14q21.3
which is an intergenic region over 500kb away from RPS29, which codes for the 40S
subunit of a ribosomal protein (46). Two SNPs of weak significance in this region have
been reported to be associated with visceral fat in women, however neither of them are
in LD with the SNPs in this current GWAS (r
2
=0.0001 in 1000 Genomes phase 3 AFR) (55).
The other region at 15q21.3 is unique to African Americans. It is 169kb away from the
next closest gene, ONECUT1 which codes for a member of a family of homeobox
transcription factors where expression is increased in the liver, stimulating liver-
expressed genes associated with glucose metabolism, and cell cycle regulation (46,56).
Polymorphisms at ONECUT1 have been associated with cholesterol, HDL, lipid, and body
mass index levels in previous studies (57).
Aside for variants at ABCG5/8, we analyzed 19 SNPs from candidate gene studies across
15 regions previously reported as GBD risk variants. A multi-ethnic study examining 49
lipid-trait associated GWAS markers and their association with GBD risk (852 cases and
4980 controls from the EAGLE study) reported 13 to be significantly associated with GBD
129
risk, (58). Our analysis of previously reported GBD risk variants (19 SNPs in 15 regions)
found only nominally significant associations at UGT1A1 and TTC39B (P<0.05) (8–10,15–
17,21,43). While the function of TTC29B is still being elucidated, TTC39B has also been
found to be associated with increased serum HDL cholesterol levels (18,19). UGT1A1
encodes a UDP-glucuronosyltransferase, which is involved in the transformation of
lipophilic molecules such as steroids, bilirubin, and hormones (56). In addition,
examination of lipid trait associated risk variants revealed 38 (of 157) SNPs to be
nominally associated with GBD risk (P<0.05) in at least one racial/ethnic populations and
in the multiethnic meta-analysis. Based on a significance level of P = 0.05, we would
expect ~24 SNPs to be discovered by chance in any of the racial/ethnic populations.
There are a few limitations of this study. GBD disease classification was done using self-
report, CHDD and Medicare linkage, physician’s report, and imaging studies. Individuals
classified as controls may have non-symptomatic gallstone disease which may become
symptomatic in the future, resulting in underreporting and misclassification that would
bias the estimates towards the null. Another potential concern is the study design as
GBD cases and controls were defined and analyzed within previous studies focusing on
other phenotypes, including T2D in the SIGMA studies, and prostate and breast cancer
in the MEC. While this creates GBD case and control groups that are not representative
of those in the general population, we don’t believe it had an impact on the study as we
performed all analyses stratified by study and study phenotype (T2D/cancer status) to
eliminate this effect. In a sensitivity analysis limited to just T2D/cancer case/control
strata we observed generally consistent results, though power was reduced.
130
In addition, the use of imputed SNPs may result in non-differential misclassification and
less significant results, based on poor accuracy of imputation. However, it should be
noted that all signals where the top SNPs were imputed, INFO scores were very high
(>0.95 for Latinos and African Americans, >0.8 for Japanese Americans). All imputation
was done to 1000 Genomes phase 1 data, and future work will include imputation and
analysis of variation in 1000 Genomes phase 3 data which may help to discover new loci.
The next step would be to replicate the novel GBD risk loci identified here. Examination
of gene and environment (GxE) interaction with respect to BMI, lipids, and other dietary
factors would help elucidate the function and effects behind the GWAS hits identified
here. Fine mapping of the regions examined in this study, especially in the 7p22.2 region
in Japanese, may highlight additional ethnic-specific risk loci. Finally, as lipid traits are
associated with GBD risk, the genome-wide significant risk variants examined here
should be analyzed in relation to lipid levels as well.
In conclusion, in this large, multiethnic GWAS meta-analysis of three races/ethnicities,
the ABCG5/8 region was found to be the strongest susceptibility region for GBD,
however the impact of this locus with respect to effect size and frequency varied widely
across populations. Additional racial/ethnic-specific loci on chromosomes 7, 14, and 15
highlight racial/ethnic differences in the genetic contribution to GBD risk.
131
4.6 References
1. Everhart JE, Khare M, Hill M, et al. Prevalence and ethnic differences in
gallbladder disease in the United States. Gastroenterology. 1999;117(3):632–639.
2. Shaffer EA. Epidemiology and risk factors for gallstone disease: has the paradigm
changed in the 21st century? Curr. Gastroenterol. Rep. 2005;7:132–140.
3. Shaffer E a. Epidemiology of gallbladder stone disease. Best Pract. Res. Clin.
Gastroenterol. 2006;20(6):981–996.
4. Stinton LM, Myers RP, Shaffer E a. Epidemiology of gallstones. Gastroenterol. Clin.
North Am. [electronic article]. 2010;39(2):157–169.
(http://dx.doi.org/10.1016/j.gtc.2010.02.003)
5. Portincasa P, Moschetta A, Palasciano G. Cholesterol gallstone disease. Lancet.
2006;368(9531):230–239.
6. Stinton LM, Shaffer EA. Epidemiology of gallbladder disease: Cholelithiasis and
cancer. Gut Liver. 2012;6(2):172–187. (http://pdf.medrang.co.kr/ekjg/ekjg006-02-
03.pdf). (Accessed May 6, 2015)
7. Katsika D, Grjibovski A, Einarsson C, et al. Genetic and environmental influences
on symptomatic gallstone disease: A Swedish study of 43,141 twin pairs.
Hepatology. 2005;41(5):1138–1143.
8. Krawczyk M, Miquel JF, Stokes CS, et al. Genetics of biliary lithiasis from an ethnic
perspective. Clin. Res. Hepatol. Gastroenterol. [electronic article].
2013;37(2):119–25.
(http://www.sciencedirect.com/science/article/pii/S2210740112002495).
(Accessed May 6, 2015)
9. Rodriguez S, Gaunt TR, Guo Y, et al. Lipids, obesity and gallbladder disease in
women: insights from genetic studies using the cardiovascular gene-centric 50K
SNP array. Eur. J. Hum. Genet. [electronic article]. 2015;(April 2014):1–7.
(http://www.nature.com/doifinder/10.1038/ejhg.2015.63)
10. Goodloe R, Brown-Gentry K, Gillani NB, et al. Lipid trait-associated genetic
variation is associated with gallstone disease in the diverse Third National Health
and Nutrition Examination Survey (NHANES III). BMC Med. Genet. [electronic
article]. 2013;14:120.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3870971&tool=pmc
entrez&rendertype=abstract)
11. Renner O, Lütjohann D, Richter D, et al. Role of the ABCG8 19H risk allele in
cholesterol absorption and gallstone disease. BMC Gastroenterol. [electronic
article]. 2013;13:30.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3598676&tool=pmc
132
entrez&rendertype=abstract)
12. Buch S, Schafmayer C, Volzke H, et al. A genome-wide association scan identifies
the hepatic cholesterol transporter ABCG8 as a susceptibility factor for human
gallstone disease. Nat Genet [electronic article]. 2007;39(8):995–999.
(http://www.ncbi.nlm.nih.gov/pubmed/17632509)
13. Katsika D, Magnusson P, Krawczyk M, et al. Gallstone disease in Swedish twins:
Risk is associated with ABCG8 D19H genotype. J. Intern. Med. 2010;268(3):279–
285.
14. Auton A, Abecasis GR, Altshuler DM, et al. A global reference for human genetic
variation. Nature [electronic article]. 2015;526(7571):68–74.
(http://www.ncbi.nlm.nih.gov/pubmed/26432245)
15. Buch S, Schafmayer C, Vlzke H, et al. Loci from a genome-wide analysis of bilirubin
levels are associated with gallstone risk and composition. Gastroenterology
[electronic article]. 2010;139(6):1942–1951.e2.
(http://dx.doi.org/10.1053/j.gastro.2010.09.003)
16. Renner O, Harsch S, Schaeffeler E, et al. A Variant of the SLC10A2 Gene Encoding
the Apical Sodium-Dependent Bile Acid Transporter Is a Risk Factor for Gallstone
Disease. PLoS One. 2009;4(10):0–6.
17. Johnson AD, Kavousi M, Smith A V, et al. Genome-wide association meta-analysis
for total serum bilirubin levels. Hum Mol Genet [electronic article].
2009;18(14):2700–2710. (http://www.ncbi.nlm.nih.gov/pubmed/19414484)
18. Rees MG, Raimondo A, Wang J, et al. Inheritance of Rare Functional GCKR
Variants and Their Contribution to Triglyceride Levels in Families. Hum. Mol.
Genet. [electronic article]. 2014;23(20):1–30.
(http://www.ncbi.nlm.nih.gov/pubmed/24879641)
19. Santoro N, Caprio S, Pierpont B, et al. Hepatic de novo lipogenesis in obese youth
is modulated by a common variant in the GCKR gene. J. Clin. Endocrinol. Metab.
2015;100(8):E1125–E1132.
20. Srivastava K, Srivastava A, Sharma KL, et al. Candidate gene studies in gallbladder
cancer: a systematic review and meta-analysis. Mutat Res [electronic article].
2011;728(1-2):67–79. (http://www.ncbi.nlm.nih.gov/pubmed/21708280)
21. Srivastava A, Srivastava A, Srivastava N, et al. Organic anion transporter 1B1
(SLCO1B1) polymorphism and gallstone formation: High incidence of Exon4 CA
genotype in female patients in North India. Hepatol. Res. 2011;41(1):71–78.
22. Kolonel LN, Henderson BE, Hankin JH, et al. A multiethnic cohort in Hawaii and
Los Angeles: baseline characteristics. Am. J. Epidemiol. [electronic article].
2000;151(4):346–57. (http://www.ncbi.nlm.nih.gov/pubmed/10695593)
133
23. Morimoto Y, Conroy SM, Ollberding NJ, et al. Erythrocyte membrane fatty acid
composition, serum lipids, and non-Hodgkin’s lymphoma risk in a nested case–
control study: the multiethnic cohort. Cancer Causes Control [electronic article].
2012;23(10):1693–1703. (http://link.springer.com/10.1007/s10552-012-0048-1)
24. Rand KA, Rohland N, Tandon A, et al. Whole-exome sequencing of over 4100 men
of African ancestry and prostate cancer risk. Hum. Mol. Genet. [electronic article].
2015;25(2):ddv462–.
(http://hmg.oxfordjournals.org/content/early/2015/12/08/hmg.ddv462.abstract?
maxtoshow=&hits=1&RESULTFORMAT=&title=prostate+cancer&andorexacttitle=
and&andorexacttitleabs=and&andorexactfulltext=and&searchid=1&usestrictdate
s=yes&resourcetype=HWCIT&ct)
25. Han Y, Hazelett DJ, Wiklund F, et al. Integration of Multiethnic Fine-mapping and
Genomic Annotation to Prioritize Candidate Functional SNPs at Prostate Cancer
Susceptibility Regions. Hum. Mol. Genet. [electronic article]. 2015;ddv269.
(http://www.hmg.oxfordjournals.org/lookup/doi/10.1093/hmg/ddv269)
26. Haiman CA, Han Y, Feng Y, et al. Genome-Wide Testing of Putative Functional
Exonic Variants in Relationship with Breast and Prostate Cancer Risk in a
Multiethnic Population. PLoS Genet. [electronic article]. 2013;9(3):e1003419.
(http://dx.plos.org/10.1371/journal.pgen.1003419)
27. Chen F, Stram DO, Le Marchand L, et al. Caution in generalizing known genetic
risk markers for breast cancer across all ethnic/racial populations. Eur. J. Hum.
Genet. [electronic article]. 2011;19(2):243–5.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3025800&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
28. Siddiq A, Couch FJ, Chen GK, et al. A meta-analysis of genome-wide association
studies of breast cancer identifies two novel susceptibility loci at 6q14 and 20q11.
Hum. Mol. Genet. 2012;21(24):5373–5384.
29. Williams AL, Jacobs SBR, Moreno-Macías H, et al. Sequence variants in SLC16A11
are a common risk factor for type 2 diabetes in Mexico. Nature [electronic
article]. 2013;506(7486):97–101.
(http://www.ncbi.nlm.nih.gov/pubmed/24390345\nhttp://www.nature.com/doif
inder/10.1038/nature12828)
30. Michael P. The Mexico City diabetes study : A population-based approach to the
... 1999;(May).
31. Waters KM, Stram DO, Hassanein MT, et al. Consistent association of type 2
diabetes risk variants found in europeans in diverse racial and ethnic groups. PLoS
Genet. [electronic article]. 2010;6(8).
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2928808&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
134
32. Haiman CA, Chen GK, Blot WJ, et al. Characterizing Genetic Risk at Known
Prostate Cancer Susceptibility Loci in African Americans. PLoS Genet. [electronic
article]. 2011;7(5):e1001387. (http://dx.plos.org/10.1371/journal.pgen.1001387)
33. Haiman CA, Chen GK, Blot WJ, et al. Genome-wide association study of prostate
cancer in men of African ancestry identifies a susceptibility locus at 17q21. Nat.
Genet. [electronic article]. 2011;43(6):570–573.
(http://www.nature.com/doifinder/10.1038/ng.839)
34. Howie BN, Donnelly P, Marchini J. A flexible and accurate genotype imputation
method for the next generation of genome-wide association studies. PLoS Genet.
2009;5(6).
35. Delaneau O, Howie B, Cox AJ, et al. Haplotype estimation using sequencing reads.
Am. J. Hum. Genet. 2013;93(4):687–696.
36. Price AL, Patterson NJ, Plenge RM, et al. Principal components analysis corrects
for stratification in genome-wide association studies. Nat. Genet.
2006;38(8):904–909.
37. Purcell S, Neale B, Todd-Brown K, et al. PLINK: a toolset for whole-genome
association and population-based linkage analysis. Am. J. Hum. Genet. 2007;81.
38. Jiao S, Hsu L, Hutter CM, et al. The use of imputed values in the meta-analysis of
genome-wide association studies. Genet. Epidemiol. 2011;35(7):597–605.
39. Willer CJ, Li Y, Abecasis GR. METAL: Fast and efficient meta-analysis of
genomewide association scans. Bioinformatics. 2010;26(17):2190–2191.
40. Turner SD. qqman: an R package for visualizing GWAS results using Q-Q and
manhattan plots. 2014 559-575
p.(http://biorxiv.org/lookup/doi/10.1101/005165)
41. Pruim RJ, Welch RP, Sanna S, et al. LocusZoom: Regional visualization of genome-
wide association scan results. Bioinformatics. 2010;26(18):2336–2337.
42. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants
from high-throughput sequencing data. Nucleic Acids Res. 2010;38(16):e164.
43. Srivastava A, Sharma KL, Srivastava N, et al. Significant role of estrogen and
progesterone receptor sequence variants in gallbladder cancer predisposition: A
multi-analytical strategy. PLoS One. 2012;7(7):1–10.
44. Willer CJ, Schmidt EM, Sengupta S, et al. Discovery and refinement of loci
associated with lipid levels. Nat. Genet. [electronic article]. 2013;45(11):1274–
1283. (http://www.nature.com/doifinder/10.1038/ng.2797)
45. Ajagbe BO, Othman RA, Myrie SB. Plant sterols, stanols, and sitosterolemia. J.
AOAC Int. 2015;98(3):716–723.
135
46. Flicek P, Amode MR, Barrell D, et al. Ensembl 2014. Nucleic Acids Res.
2014;42(D1):749–755.
47. Kovacs P, Kress R, Rocha J, et al. Variation of the gene encoding the nuclear bile
salt receptor FXR and gallstone susceptibility in mice and humans. J. Hepatol.
2008;48(1):116–124.
48. Kumar P, Henikoff S, Ng PC. Predicting the effects of coding non-synonymous
variants on protein function using the SIFT algorithm. Nat. Protoc.
2009;4(7):1073–1081.
49. Choi Y, Sims GE, Murphy S, et al. Predicting the Functional Effect of Amino Acid
Substitutions and Indels. PLoS One. 2012;7(10).
50. Wang J, Grishin N, Kinch L, et al. Sequences in the nonconsensus nucleotide-
binding domain of ABCG5/ABCG8 required for sterol transport. J. Biol. Chem.
2011;286(9):7308–7314.
51. Sun Back S, Kim J, Choi D, et al. Cooperative transcriptional activation of ATP-
binding cassette sterol transporters ABCG5 and ABCG8 genes by nuclear
receptors including Liver-X-Receptor. BMB Rep. 2013;46(6):322–327.
52. Eeles RA, Kote-Jarai Z, Al Olama AA, et al. Identification of seven new prostate
cancer susceptibility loci through a genome-wide association study. Nat. Genet.
[electronic article]. 2009;41(10):1116–1121.
(http://www.nature.com/doifinder/10.1038/ng.450)
53. Kaufman L, Potla U, Coleman S, et al. Up-regulation of the homophilic adhesion
molecule sidekick-1 in podocytes contributes to glomerulosclerosis. J. Biol. Chem.
2010;285(33):25677–25685.
54. Comuzzie AG, Cole SA, Laston SL, et al. Novel Genetic Loci Identified for the
Pathophysiology of Childhood Obesity in the Hispanic Population. PLoS One.
2012;7(12).
55. Fox CS, Liu Y, White CC, et al. Genome-wide association for abdominal
subcutaneous and visceral adipose reveals a novel locus for visceral fat in women.
PLoS Genet. 2012;8(5).
56. Tatusova T, Ciufo S, Fedorov B, et al. RefSeq microbial genomes database: New
representation and annotation strategy. Nucleic Acids Res. 2014;42(D1):553–559.
57. Kathiresan S, Manning AK, Demissie S, et al. A genome-wide association study for
blood lipid phenotypes in the Framingham Heart Study. BMC Med. Genet.
[electronic article]. 2007;8 Suppl 1:S17.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1995614&tool=pmc
entrez&rendertype=abstract)
58. Goodloe R, Brown-Gentry K, Gillani NB, et al. Lipid trait-associated genetic
136
variation is associated with gallstone disease in the diverse Third National Health
and Nutrition Examination Survey (NHANES III). BMC Med. Genet. [electronic
article]. 2013;14(1):120. (http://www.biomedcentral.com/1471-2350/14/120)
137
CHAPTER 5: CONCLUSIONS AND FUTURE DIRECTIONS
5.1 AAC & CRC
In chapter 2, we examined the relationship between AAC as an indicator of a
hyperactive immune response, and the risk of CRC incidence and mortality in five
racial/ethnic populations in the MEC. Primarily, we found a statistically significant 14%
reduction in the risk of CRC incidence in individuals with AAC compared to those
without. These reductions in risk were statistically significant across all racial/ethnic
populations examined except for Japanese American mortality risk and Latino incidence.
African Americans had a greater proportion of right-sided CRC incidence (51.6%) than
Whites (47.6%) (Table 2.3). The inverse association of AAC with CRC related mortality is
similarly reduced.
This study provides some support for the immune surveillance theory, in which the
innate immune system can either eliminate cancerous cells that have bypassed tumor
suppressor mechanisms, or can target susceptible tumor clones (1). A hyperactive
immune system may increase immune surveillance along the lining of the gut, and IgE
cross-linked to tumor cells may result in tumor specific memory cells that protect from
future antigenic stimulation (2). From this study in the MEC we can conclude that the
inverse association between AAC and CRC incidence/mortality is present and similar
across all five racial/ethnic populations.
138
5.1.1 Future studies with AAC and CRC
AAC status is self-reported (though physician diagnosed) and validation using
biomarkers is a necessary next step. Markers such as eosinophil content, C-reactive
protein (CRP) concentration, and IgE titers would allow for validation of the assessment
of AAC (3,4). Previous studies examining associations between CRC and inflammatory
markers have been inconclusive due to variability based on an individual’s immediate
state of health at the time of blood draw (5,6). However, the MEC has a larger set of
biospecimens in nested case-control studies of breast and prostate cancer. Using the
larger sample set in the MEC will allow for validation of AAC status.
The biology behind the specific protective activity conferred by AAC can be further
informed through the examination of genetic susceptibility markers across AAC and
CRC. CRC genetic susceptibility studies have identified numerous risk alleles, and the
MEC is part of a consortium with over 20 other cohorts conducting genetic studies on
CRC tumors (7,8). Linkage studies have identified important germline mutations, such as
those in APC and DNA mismatch repair genes, as explored in Chapter 1. It is estimated
that up to 35% of CRC risk can be explained by heritable factors, and common genetic
variants could explain 17% of that association (8,9). Additional loci from GWA studies,
including three loci in the 8q24 region (which is also associated with prostate cancer),
and variants near the c-MYC oncogene. Many other genes have been discovered that
require additional examination to understand the biological mechanisms (10).
Susceptibility risk loci have been discovered for AAC as well, though additional studies
are necessary (11). Existing markers of AAC susceptibility (such as ORMDL3, CTNNA3,
139
DPP10, and KCNMA1) could be specifically examined in the MEC CRC cases to see if
these risk alleles are present and increase risk. These analyses would provide additional
information as to the mechanisms of action of protection of AAC on CRC risk.
Finally, the conflicting evidence in the literature of a link between AAC and CRC is not
limited to colorectal cancer, and should be examined in other types of cancer as well.
Associations between AAC and non-Hodgkin lymphoma (NHL) have already been
examined in the MEC, where there was an increased risk of NHL in Latinos with AAC(12).
AAC have also been inversely associated with pancreatic cancer and gliomas, with
conflicting associations with breast, prostate, and lung cancers (13–15).
5.1.2 Summary
Through this analysis in the Multiethnic Cohort we have added additional information to
the inverse association of atopic conditions on CRC risk, and this protection occurs
across all racial/ethnic populations. The observations provided here add support to the
immune surveillance theory.
5.2 Ethnic Differences in Cardiovascular Disease & Stroke risk
In chapter 3 we examined the racial/ethnic specific differences in risk of AMI, OHD, and
stroke-related mortality, and how known risk factors account for some of the
differences in risk. While previous studies either did not include stroke or examined only
one other race/ethnic group at a time, this analysis in the MEC allowed for the
simultaneous comparison of multiple racial/ethnic populations (16–19). In general, the
ethnic disparities in CVD and stroke risk disappeared after adjustment for known risk
140
factors. Most notably, the effect of adding nutrients to models involving Native
Hawaiians and Japanese Americans had great effect on risk estimates. Native Hawaiians
continued to have elevated risk of AMI and OHD after adjustment, and Japanese
Americans continued to have a significant decreased risk of AMI and OHD with respect
to whites within the MEC. The large, multiethnic analysis of stroke provides new
information, as the differences in risk were greatly attenuated in all populations except
for African Americans, who maintained a higher risk of mortality due to stroke.
Using the MEC has allowed us to examine the contribution of known risk factors on the
disparities in risk of disease between multiple races/ethnicities. This analysis helps to
inform both policy makers as well as future researchers on which risk factors to focus
on. The remaining risk/ethnic differences could be due to unmeasured environmental
exposures, socioeconomic factors, or perhaps genetic differences in susceptibility.
In this examination of racial/ethnic differences in cardiovascular disease, we found that
adjusting for known risk factors attenuated most of the differences in risk, though some
disparities remained.
5.2.1 Future Analysis
It is important to elucidate further causes of remaining racial/ethnic specific differences
in risk. Lipid levels are thought to increase cardiovascular disease risk through the
progression of the atherosclerotic process, and is included in a number of risk
calculators (20–25). Mean lipid values differ by ethnicity, and these differences may
require different cutoffs to evaluate risk in these populations (26,27). Though we
141
examined lipid levels and the potential association with differences in disease risk in a
small subset of individuals in chapter 3, it would be beneficial to examine lipid levels in a
larger portion of the cohort. Based on the CVD risk profiles we observed, we would
expect Japanese Americans to have better lipid profiles, which was not seen in this
analysis. Additional analysis in non-Latino samples may provide additional information
as to why this occurs (26). A prospective analysis of lipid levels using AHA cut-points for
dyslipidemia would allow us to extend the analysis performed in chapter 3, and the
current sample set is too small to perform this analysis.
Additionally, a logical next step would be to examine genetic susceptibility in these
populations, and whether or not known risk variants can account for the remaining
differences in risk. As lipids are known to be associated with risk of CVD, an examination
SNPs with known associations with lipid levels is warranted (28). As there is a great
amount of correlation between the lipid markers, an analysis of SNPs should take in to
account the effect on all the lipid markers examined (29). In addition, with multiple
ethnic samples in the MEC genomic database, a GWAS analysis can be performed to
identify novel genetic variants across the genome (similar to what was performed with
GBD in chapter 4).
As the MEC was begun over two decades ago and not initially designed for the study of
atherosclerotic diseases, there are a few risk factors that should be examined in a
multiethnic context, as well as a few additional markers that should be considered in
place of ones currently in use. In this study, BMI is used as the primary indicator of
excess weight and body mass-related heath related effects. However, BMI does not take
142
into account varying body types and distribution of fat within the body, and may
underreport the level of individuals with excess body fat (30,31). In addition, various
locations of fat deposits have differing effects on CVD risk. A more effective alternative
measure that has been selected includes the waist-hip ratio (WHR), which can quantify
central versus peripheral obesity (32,33). The MEC has collected data on waist and hip
sizes in more recent questionnaires, which should be examined in future studies. In
addition, the use of years of education as a marker for socioeconomic status may be
inadequate in assessing the presence of healthcare-related disparities (34). Disparities in
access to and quality of healthcare, including health insurance status, are major
contributors to racial/ethnic differences in risk, and should be measured in subsequent
studies (35–38).
5.3 GBD and genetic susceptibility
In chapter 4, we discussed the differing prevalence of gallbladder disease by ethnicity,
and performed a multiethnic genome-wide association study to identify risk variants for
GBD. In a large, multiethnic study, we were able to replicate known risk variants of GBD
in the ABCG5/8 region using the large set of samples in the MEC and SIGMA. Using a
multiethnic population, we identified two risk variants that were present only in African
Americans, and one cryptic region in Japanese Americans. Without using a multiethnic
population, we would miss those risk SNPs that are common in particular racial/ethnic
groups, but rare or monomorphic in others. As discussed in the Introduction chapter,
143
the MEC has been used numerous times in this way to detect ethnic-specific risk
variants (39–44).
5.3.1 Future examination of GBD risk
Analysis in the MEC was able to identify racial/ethnic specific risk alleles. The next step
would be to replicate the risk variants identified here in additional populations. The two
risk variants unique to African Americans (rs72672032 and rs717754) should be
replicated in populations of African-descent. The significant locus found in chromosome
7 (peak SNP: rs13226393) was rare in Japanese Americans and very common in African
Americans and Latinos. Genotyping of risk SNPs in LD with the top SNP, but not analyzed
in this GWA study, is warranted. Analysis in Whites and Native Hawaiians of the MEC (as
well as other racial/ethnic populations) may reveal additional success in identifying
ethnic-specific GBD risk variants. Including the MEC samples in a meta-analysis of other
racial/ethnic GWA studies would also assist in replication and signal discovery. As most
of the risk SNPs identified here are in intronic or intergenic regions, the next step would
be to perform functional analyses of these variants.
Additionally, the link between lipid levels and GBD risk is clear (45,46). Therefore, an in-
depth analysis of known risk variants that result in poorer lipid profiles and their
association with GBD risk in a multiethnic population is necessary (28,29).
Finally, GBD is very common in the general population, but not all individuals suffering
from GBD become symptomatic (47). As such, GBD is typically underdiagnosed, and
prevalence estimates vary widely. The use of biomarkers that can better detect the
144
presence of gallstones would help to properly classify cases. In addition, integration of
these biomarkers with the genetic risk variants highlighted here would help to define
those individuals that may benefit from targeted prevention measures such as
prophylactic therapy or dietary management.
5.4 Conclusion
In this dissertation, we have examined multiple aspects of the multiethnic nature and
epidemiology of chronic disease. Most chronic conditions have varying degrees of
incidence and mortality by race/ethnicity, which can primarily be explained by
differences in prevalence of known risk factors and markers of genetic susceptibility.
Further study is perpetually needed to fully understand the contributions to racial and
ethnic differences in risk.
5.5 References
1. Swann JB, Smyth MJ. Review series Immune surveillance of tumors. J. Clin. Invest.
2007;117(5):1137–1146.
2. Jensen-Jarolim E, Achatz G, Turner MC, et al. AllergoOncology: the role of IgE-
mediated allergy in cancer. Allergy [electronic article]. 2008;63(10):1255–1266.
(http://www.ncbi.nlm.nih.gov/pubmed/18671772)
3. Abbas AK, Lichtman AH. Basic Immunology. 3rd ed. Elsevier Saunders; 2009.
4. Loutsios C, Farahi N, Porter L, et al. Biomarkers of eosinophilic inflammation in
asthma. Expert Rev. Respir. Med. [electronic article]. 2014;8(2):143–50.
(http://www.ncbi.nlm.nih.gov/pubmed/24460178)
5. Hoppin JA, Jaramillo R, Salo P, et al. Questionnaire predictors of atopy in a US
population sample: findings from the National Health and Nutrition Examination
Survey, 2005-2006. Am J Epidemiol [electronic article]. 2011;173(5):544–552.
(http://www.ncbi.nlm.nih.gov/pubmed/21273397)
6. Wang H, Rothenbacher D, Low M, et al. Atopic diseases, immunoglobulin E and
risk of cancer of the prostate, breast, lung and colorectum. Int J Cancer [electronic
article]. 2006;119(3):695–701. (http://www.ncbi.nlm.nih.gov/pubmed/16506215)
145
7. Kumar V, Abbas AK, Aster JC. Robbins Basic Pathology. 9th ed. Philadelphia:
Elsevier Saunders; 2013.
8. Peters U, Jiao S, Schumacher FR, et al. Identification of Genetic Susceptibility Loci
for Colorectal Tumors in a Genome-wide Meta-analysis. Gastroenterology.
2014;144(4):799–807.
9. Peters U, Hutter CM, Hsu L, et al. Meta-analysis of new genome-wide association
studies of colorectal cancer risk. Hum. Genet. 2012;131(2):217–234.
10. Le Marchand L. Genome-Wide Association Studies and Colorectal cancer. Surg
Oncol Clin N Am. 2009;18(4):663–668.
11. Barnes K. Genomewide association studies in allergy and the influence of
ethnicity. Curr Opin Allergy Clin Immunol [electronic article]. 2010;10(5):427–433.
(http://www.ncbi.nlm.nih.gov/pubmed/20724922)
12. Erber E, Lim U, Maskarinec G, et al. Common immune-related risk factors and
incident non-Hodgkin lymphoma: The multiethnic cohort. Int. J. Cancer [electronic
article]. 2009;125(6):1440–1445. (http://doi.wiley.com/10.1002/ijc.24456)
13. Vojtechova P, Martin RM. The association of atopic diseases with breast,
prostate, and colorectal cancers: a meta-analysis. Cancer Causes Control
[electronic article]. 2009;20(7):1091–1105.
(http://link.springer.com/10.1007/s10552-009-9334-y)
14. Turner MC. Epidemiology: allergy history, IgE, and cancer. Cancer Immunol
Immunother [electronic article]. 2012;61(9):1493–1510.
(http://www.ncbi.nlm.nih.gov/pubmed/22183126)
15. Turner MC, Chen Y, Krewski D, et al. Cancer Mortality among US Men and Women
with Asthma and Hay Fever. Am. J. Epidemiol. [electronic article].
2005;162(3):212–221.
(http://aje.oxfordjournals.org/cgi/doi/10.1093/aje/kwi193). (Accessed January
10, 2014)
16. Graham G. Population-based approaches to understanding disparities in
cardiovascular disease risk in the United States. 2014;393–400.
17. Rosen SE, Henry S, Bond R, et al. Sex-Specific Disparities in Risk Factors for
Coronary Heart Disease. Curr. Atheroscler. Rep. [electronic article].
2015;17(8):523. (http://www.ncbi.nlm.nih.gov/pubmed/26108894). (Accessed
July 4, 2015)
18. Winston GJ, Barr RG, Carrasquillo O, et al. Sex and Racial/Ethnic Differences in
Cardiovascular Disease Risk Factor Treatment and Control Among Individuals
With Diabetes in the Multi-Ethnic Study of Atherosclerosis (MESA). Diabetes Care
[electronic article]. 2009;32(8):1467–1469.
(http://care.diabetesjournals.org/cgi/doi/10.2337/dc09-0260)
146
19. Kovesdy CP, Norris KC, Boulware LE, et al. Association of Race with Mortality and
Cardiovascular Events in a Large Cohort of US Veterans. Circulation [electronic
article]. 2015;CIRCULATIONAHA.114.015124.
(http://circ.ahajournals.org/lookup/doi/10.1161/CIRCULATIONAHA.114.015124)
20. Collaboration* TERF. Major lipids, apolipoproteins, and risk of vascular disease.
Jama [electronic article]. 2009;302(18):1993–2000.
(http://dx.doi.org/10.1001/jama.2009.1619)
21. Arsenault BJ, Boekholdt SM, Kastelein JJP. Lipid parameters for measuring risk of
cardiovascular disease. Nat. Rev. Cardiol. [electronic article]. 2011;8(4):197–206.
(http://dx.doi.org/10.1038/nrcardio.2010.223)
22. Angelantonio E Di, Gao P, Pennells L, et al. Lipid-Related Markers and
Cardiovascular Disease Prediction. Jama. 2014;307(23):2499–2506.
23. Allan GM, Garrison S, McCormack J. Comparison of cardiovascular disease risk
calculators. Curr. Opin. Lipidol. [electronic article]. 2014;25(4):254–65.
(http://www.ncbi.nlm.nih.gov/pubmed/24977979)
24. Howard B V, Robbins DC, Sievers ML, et al. LDL cholesterol as a strong predictor
of coronary heart disease in diabetic individuals with insulin resistance and low
LDL: The Strong Heart Study. Arterioscler. Thromb. Vasc. Biol. 2000;20:830–835.
25. Gordon T, Castelli WP, Hjortland MC, et al. High density lipoprotein as a
protective factor against coronary heart disease. Am. J. Med. 1977;62(5):707–
714.
26. Willey JZ, Rodriguez CJ, Carlino RF, et al. Race-ethnic differences in the
association between lipid profile components and risk of myocardial infarction:
The Northern Manhattan Study. Am. Heart J. [electronic article].
2011;161(5):886–892.
(http://linkinghub.elsevier.com/retrieve/pii/S0002870311000883)
27. Carroll MD, Kit BK, Lacher D a, et al. Trends in lipids and lipoproteins in US adults,
1988-2010. JAMA [electronic article]. 2012;308(15):1545–54.
(http://www.ncbi.nlm.nih.gov/pubmed/23073951)
28. Willer CJ, Schmidt EM, Sengupta S, et al. Discovery and refinement of loci
associated with lipid levels. Nat. Genet. [electronic article]. 2013;45(11):1274–
1283. (http://www.nature.com/doifinder/10.1038/ng.2797)
29. Do R, Willer CJ, Schmidt EM, et al. Common variants associated with plasma
triglycerides and risk for coronary artery disease. Nat. Genet. [electronic article].
2013;45(11):1345–1352.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3904346&tool=pmc
entrez&rendertype=abstract)
30. Okorodudu DO, Jumean MF, Montori VM, et al. Diagnostic performance of body
147
mass index to identify obesity as defined by body adiposity: a systematic review
and meta-analysis. Int. J. Obes. [electronic article]. 2010;34(5):791–799.
(http://www.nature.com/doifinder/10.1038/ijo.2010.5)
31. Romero-Corral A, Somers VK, Sierra-Johnson J, et al. Accuracy of body mass index
in diagnosing obesity in the adult general population. Int. J. Obes. [electronic
article]. 2008;32(6):959–966.
(http://www.nature.com/doifinder/10.1038/ijo.2008.11)
32. Huxley R, Mendis S, Zheleznyakov E, et al. Body mass index, waist circumference
and waist:hip ratio as predictors of cardiovascular risk--a review of the literature.
Eur. J. Clin. Nutr. [electronic article]. 2010;64(1):16–22.
(http://www.ncbi.nlm.nih.gov/pubmed/19654593)
33. Consultation WHOE. Waist Circumference and Waist-Hip Ratio Report of a WHO
Expert Consultation. World Health. 2008;(December):8–11.
34. Laiyemo AO, Doubeni C, Pinsky PF, et al. Race and Colorectal Cancer Disparities:
Health-Care Utilization vs Different Cancer Susceptibilities. JNCI J. Natl. Cancer
Inst. [electronic article]. 2010;102(8):538–546.
(http://jnci.oxfordjournals.org/cgi/doi/10.1093/jnci/djq068)
35. Romero CX, Romero TE, Shlay JC, et al. Changing trends in the prevalence and
disparities of obesity and other cardiovascular disease risk factors in three
racial/ethnic groups of USA adults. Adv. Prev. Med. [electronic article].
2012;2012:172423.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3518078&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
36. Kurian AK, Cardarelli KM. Racial and ethnic differences in cardiovascular disease
risk factors: a systematic review. Ethn. Dis. 2007;17(1):143–152.
37. Davis AM, Vinci LM, Okwuosa TM, et al. Cardiovascular health disparities: a
systematic review of health care interventions. Med. Care Res. Rev. [electronic
article]. 2007;64(5 Suppl):29S–100S.
(http://www.scopus.com/inward/record.url?eid=2-s2.0-
34648831043&partnerID=tZOtx3y1)
38. Mochari-Greenberger H, Mosca L. Differential Outcomes by Race and Ethnicity in
Patients with Coronary Heart Disease: A Contemporary Review. Curr. Cardiovasc.
Risk Rep. [electronic article]. 2015;9(5).
(http://link.springer.com/10.1007/s12170-015-0447-4). (Accessed April 29, 2015)
39. Al Olama AA, Kote-Jarai Z, Giles GG, et al. Multiple loci on 8q24 associated with
prostate cancer susceptibility. Nat. Genet. [electronic article]. 2009;41(10):1058–
60. (http://www.ncbi.nlm.nih.gov/pubmed/19767752). (Accessed January 10,
2014)
148
40. Haiman CA, Patterson N, Freedman ML, et al. Multiple regions within 8q24
independently affect risk for prostate cancer. Nat. Genet. [electronic article].
2007;39(5):638–644. (http://www.nature.com/doifinder/10.1038/ng2015)
41. Haiman CA, Stram DO, Pike MC, et al. A comprehensive haplotype analysis of
CYP19 and breast cancer risk: the Multiethnic Cohort. Hum Mol Genet [electronic
article]. 2003;12(20):2679–2692.
(http://www.ncbi.nlm.nih.gov/pubmed/12944421)
42. Haiman CA, Han Y, Feng Y, et al. Genome-Wide Testing of Putative Functional
Exonic Variants in Relationship with Breast and Prostate Cancer Risk in a
Multiethnic Population. PLoS Genet. [electronic article]. 2013;9(3):e1003419.
(http://dx.plos.org/10.1371/journal.pgen.1003419)
43. Cheng I, Chen GK, Nakagawa H, et al. Evaluating genetic risk for prostate cancer
among Japanese and Latinos. Cancer Epidemiol. Biomarkers Prev. [electronic
article]. 2012;21(11):2048–58.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3494732&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
44. Waters KM, Le Marchand L, Kolonel LN, et al. Generalizability of associations from
prostate cancer genome-wide association studies in multiple populations. Cancer
Epidemiol. Biomarkers Prev. [electronic article]. 2009;18(4):1285–9.
(http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2917607&tool=pmc
entrez&rendertype=abstract). (Accessed January 10, 2014)
45. Portincasa P, Moschetta A, Palasciano G. Cholesterol gallstone disease. Lancet.
2006;368(9531):230–239.
46. Rodriguez CJ, Daviglus ML, Swett K, et al. Dyslipidemia Patterns among
Hispanics/Latinos of Diverse Background in the United States. Am. J. Med.
[electronic article]. 2014;127(12):1186–1194.e1.
(http://linkinghub.elsevier.com/retrieve/pii/S0002934314006640)
47. Stinton LM, Shaffer EA. Epidemiology of gallbladder disease: Cholelithiasis and
cancer. Gut Liver. 2012;6(2):172–187. (http://pdf.medrang.co.kr/ekjg/ekjg006-02-
03.pdf). (Accessed May 6, 2015)
149
APPENDIX A:
AAC AND CRC
150
Table A.1. Multivariate Risks of Colorectal Cancer Associated with AAC Status Stratified by Risk Factors in the MEC (1993-2010).
White African American Native Hawaiian Japanese American Latino All Ethnicities
Case
a
RR
b
95%
CI
Case
a
RR
b
95%
CI
Case
a
RR
b
95%
CI
Case
a
RR
b
95%
CI
Case
a
RR
b
95%
CI
Case
a
RR
b
95%
CI
Pint
c
Age
<60
79/195 0.89
0.68,
1.16
69/242 0.75
0.57,
0.98
38/117 0.79
0.54,
1.14
119/340 0.86
0.69,
1.06
55/231 0.85
0.63,
1.15
360/1125 0.83
0.74,
0.94
Peth
d
= 0.90
60-69
107/295 0.95
0.76,
1.19
83/286 0.91
0.71,
1.17
17/93 0.57
0.34,
0.97
152/573 0.89
0.74,
1.06
93/358 1.12
0.89,
1.41
452/1605 0.93
0.83,
1.03
Peth
d
= 0.20
>70
55/219 0.68
0.50,
0.91
62/255 0.81
0.61,
1.07
8/32 0.88
0.40,
1.95
92/412 0.88
0.70,
1.11
24/133 0.78
0.50,
1.21
241/1051 0.80
0.69,
0.92
Peth
d
= 0.69 0.19
FH
Yes
28/90 0.74
0.48,
1.15
23/86 0.67
0.42,
1.07
3/17 0.49
0.14,
1.73
66/156 1.11
0.83,
1.49
8/35 0.63
0.29,
1.37
128/384 0.86
0.70,
1.06
Peth
d
= 0.08
No
213/619 0.86
0.74,
1.01
191/697 0.83
0.71,
0.98
60/225 0.74
0.55,
0.99
297/1169 0.83
0.73,
0.95
164/687 0.98
0.83,
1.17
925/3397 0.86
0.79,
0.92
Peth
d
= 0.34 0.73
BMI
<25
109/271 0.95
0.75,
1.18
49/186 0.83
0.61,
1.15
8/56 0.38
0.18,
0.81
195/755 0.81
0.69,
0.95
48/183 1.05
0.76,
1.45
409/1451 0.85
0.76,
0.95
Peth
d
= 0.07
25-30
81/279 0.80
0.63,
1.03
82/315 0.84
0.65,
1.07
24/87 0.82
0.52,
1.31
128/441 1.00
0.82,
1.22
69/349 0.86
0.66,
1.12
384/1471 0.88
0.79,
0.99
Peth
d
= 0.73
30+
51/159 0.75
0.54,
1.03
83/282 0.78
0.61,
1.00
31/99 0.84
0.55,
1.26
40/129 0.82
0.57,
1.19
55/190 1.03
0.76,
1.40
260/859 0.83
0.72,
0.96
Peth
d
= 0.45 0.56
Smoking
Never
92/226 0.92
0.72,
1.18
88/292 0.90
0.71,
1.15
16/76 0.57
0.33,
0.99
170/533 0.92
0.77,
1.10
67/269 0.92
0.70,
1.20
433/1396 0.90
0.80,
1.00
Peth
d
= 0.54
Past
115/348 0.81
0.65,
1.00
90/332 0.74
0.58,
0.94
38/110 0.88
0.61,
1.29
153/570 0.85
0.71,
1.02
92/324 1.09
0.86,
1.38
488/1684 0.86
0.77,
0.95
Peth
d
= 0.20
Current
33/127 0.82
0.55,
1.21
34/153 0.76
0.52,
1.11
9/52 0.61
0.30,
1.25
37/210 0.77
0.54,
1.10
10/114 0.51
0.27,
0.98
123/656 0.74
0.61,
0.90
Peth
d
= 0.64 0.23
Aspirin
151
Never
117/387 0.81
0.66,
1.00
96/385 0.74
0.59,
0.92
33/149 0.63
0.43,
0.92
240/934 0.82
0.71,
0.94
101/383 1.10
0.88,
1.37
587/2238 0.83
0.75,
0.91
Peth
d
= 0.04
Past
47/116 0.85
0.60,
1.20
54/207 0.76
0.56,
1.03
15/38 1.01
0.54,
1.88
47/135 1.01
0.72,
1.42
34/160 0.79
0.54,
1.15
197/656 0.86
0.73,
1.01
Peth
d
= 0.59
Current
68/186 0.88
0.67,
1.17
51/135 1.06
0.77,
1.47
15/47 0.86
0.47,
1.57
64/219 0.95
0.72,
1.27
28/142 0.74
0.49,
1.12
226/729 0.91
0.78,
1.06
Peth
d
= 0.65 0.24
Education
<12yrs
69/258 0.87
0.67,
1.14
99/347 1.01
0.81,
1.27
29/162 0.58
0.39,
0.86
106/678 0.71
0.57,
0.87
110/482 1.05
0.85,
1.29
413/1927 0.86
0.77,
0.96
Peth
d
= 0.01
Some
69/223 0.70
0.53,
0.92
60/269 0.59
0.44,
0.78
30/50 1.40
0.88,
2.23
113/370 0.83
0.67,
1.02
46/155 0.96
0.68,
1.33
318/1067 0.78
0.69,
0.89
college Peth
d
= 0.02
Grad
102/221 0.97
0.76,
1.23
52/156 0.82
0.60,
1.13
4/27 0.27
0.09,
0.79
142/265 1.16
0.95,
1.43
16/71 0.68
0.39,
1.18
316/740 0.96
0.84,
1.09
Peth
d
= 0.02 0.20
AAC, Atopic Allergic Conditions; BMI, Body-mass index; CI, Confidence Interval; CRC, Colorectal Cancer; RR, Relative Risk; SD, Standard Deviation.
The number of cases may not add up to case totals due to missing questionnaire information.
a
Cases displayed as cases in the AAC group/cases
in non-AAC group
b
All models adjusted for sex, age, smoking status (status and pack-years), education level, BMI, NSAID usage
c
P-value for test
of interaction by associated risk factor using likelihood ratio tests.
d
P-value for test of interaction by ethnicity using likelihood ratio tests.
152
APPENDIX B:
CVD RISK FACTORS
153
Table B.1. Prevalence of cholesterol medication use and number of events by sex in the MEC.
Men W AA NH JA LA Total
Never % 57.77 55.03 49.81 47.62 55.58 52.94
Past % 4.79 9.57 6.84 5.62 9.48 6.5
Current % 37.44 35.4 43.35 46.75 34.94 40.56
Total 9,563 2,633 2,355 11,221 5,793 31,565
AMI 39 25 13 52 38 167
OHD 112 69 32 136 87 436
Stroke 30 15 15 45 28 133
Women W AA NH JA LA Total
Never % 61.7 57.16 50.39 48.92 56.75 55.2
Past % 5.63 10.17 8.02 6.5 10.19 7.44
Current % 32.67 32.66 41.59 44.59 33.06 37.37
Total 9,180 3,971 2,195 10,324 5,115 30,785
AMI 16 24 2 29 15 86
OHD 61 56 24 61 45 247
Stroke 41 25 5 48 25 144
W – Whites; AA – African Americans; NH – Native Hawaiians; JA – Japanese Americans; LA –
Latinos; AMI – acute myocardial infarction; OHD – other heart disease.
154
Table B.2. Main effect RR of cholesterol medication usage on disease risk, by sex in the MEC.
Men
CVD n=603 P AMI n=167 P OHD n=436 P STROKE n=133 P
Deaths 603 167 436 133
Never 1.00 (ref) 1.00 (ref) 1.00 (ref) 1.00 (ref)
Ever 1.06 0.51 1.24 0.18 0.99 0.95 0.86 0.40
Women
CVD n = 333 P AMI n=86 P OHD n=247 P Stroke n=144 P
Deaths 331 86 247 144
Never 1.00 (ref) 1.00 (ref) 1.00 (ref) 1.00 (ref)
Ever 1.10 0.42 1.14 0.57 1.08 0.55 0.93 0.68
CVD – cardiovascular disease (includes AMI and OHD); AMI – acute myocardial infarction; OHD –
other heart disease; RR – relative risk. See methods for adjusted known risk factors.
155
Figure B.1a-B.1h. Sub-analysis of individuals reporting cholesterol medication usage, and the effect on
risk/ethnic specific risk differences in the MEC using White individuals as the baseline.
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
AA NH JA LA
Male Statin Exploration - CVD
Crude HR - Answered Cholesterol Medication RR + Known risk factors RR
+ Cholesterol medication status RR
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
AA NH JA LA
Female Statin Exploration - CVD
Crude HR - Answered Cholesterol Medication RR + Known risk factors RR
+ Cholesterol medication status RR
156
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
AA NH JA LA
Male Statin Exploration - AMI
Crude HR - Answered Cholesterol Medication RR + Known risk factors RR
+ Cholesterol medication status RR
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
AA NH JA LA
Female Statin Exploration - AMI
Crude HR - Answered Cholesterol Medication RR + Known risk factors RR
+ Cholesterol medication status RR
157
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
AA NH JA LA
Male Statin Exploration - OHD
Crude HR - Answered Cholesterol Medication RR + Known risk factors RR
+ Cholesterol medication status RR
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
AA NH JA LA
Female Statin Exploration - OHD
Crude HR - Answered Cholesterol Medication RR + Known risk factors RR
+ Cholesterol medication status RR
158
W – Whites; AA – African Americans; NH – Native Hawaiians; JA – Japanese Americans; LA – Latinos; CVD -
cardiovascular disease (including AMI and OHD); AMI – acute myocardial infarction; OHD – other heart
disease. Baseline risk is Whites (RR 1.00)
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
AA NH JA LA
Male Statin Exploration - Stroke
Crude HR - Answered Cholesterol Medication RR + Known risk factors RR
+ Cholesterol medication status RR
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
AA NH JA LA
Female Statin Exploration - Stroke
Crude HR - Answered Cholesterol Medication RR + Known risk factors RR
+ Cholesterol medication status RR
159
Table B.3. Percent change in racial/ethnic specific risk estimates and percent of crude RRs by sex and race/ethnicity.
Male Ethnicity Ethnicity Ethnicity Ethnicity
AMI AA NH JA LA
Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude
Crude 1.55 Crude 1.58 Crude 0.77 Crude 1.04
Hypertension 1.34
-
14% 1.34 -14% Diabetes 1.36
-
14% 1.36 -14% Sat. Fat 0.88 14% 0.88 14% Education 0.88
-
15% 0.881 -15%
Diabetes 1.22 -9% 1.37 -12% Education 1.20
-
12% 1.37 -13% Diabetes 0.80
-
10% 0.70 -9% Diabetes 0.78
-
11% 0.91 -13%
Cholesterol 1.13 -7% 1.42 -8% Hypertension 1.10 -8% 1.39 -12% Education 0.73 -8% 0.71 -9% Smoking 0.81 4% 1.06 2%
Education 1.07 -6% 1.42 -8% Sat. Fat 1.18 6% 1.67 6% Hypertension 0.70 -4% 0.71 -8% Sat. Fat 0.79 -3% 0.99 -5%
Alcohol 1.03 -4% 1.44 -7% Vig. Activity 1.21 3% 1.63 3% Alcohol 0.68 -4% 0.71 -8% Alcohol 0.77 -2% 0.99 -4%
Smoking 0.99 -4% 1.49 -4% Cholesterol 1.19 -2% 1.54 -2% Cholesterol 0.66 -3% 0.79 3% BMI 0.78 1% 1.04 0%
Vig. Activity 0.98 -1% 1.49 -4% Alcohol 1.16 -2% 1.49 -6% BMI 0.64 -2% 0.79 2% Hypertension 0.80 2% 1.02 -2%
BMI 0.98 1% 1.51 -2% BMI 1.14 -1% 1.46 -7% Vig. Activity 0.63 -2% 0.74 -4% Cholesterol 0.79 -1% 0.99 -5%
Sat. Fat 0.99 1% 1.48 -4% Smoking 1.12 -2% 1.55 -2% Smoking 0.63 0% 0.78 1% Vig. Activity 0.79 0% 1.03 -1%
Male Ethnicity Ethnicity Ethnicity Ethnicity
OHD AA NH JA LA
Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude
Crude 1.67 Crude 2.10 Crude 0.84 Crude 1.11
Hypertension 1.47
-
12% 1.47 -12% Education 1.79
-
15% 1.79 -15% Sat. Fat 0.94 12% 0.94 12% Education 0.93
-
17% 0.93 -17%
Education 1.35 -9% 1.52 -9% Diabetes 1.61
-
10% 1.87 -11% Education 0.85
-
10% 0.76 -10% Diabetes 0.85 -9% 1.00 -10%
Cholesterol 1.26 -6% 1.56 -7% Hypertension 1.50 -7% 1.89 -10% Diabetes 0.79 -7% 0.78 -7% Smoking 0.91 8% 1.17 5%
Diabetes 1.19 -5% 1.52 -9% Sat. Fat 1.58 6% 2.21 5% Hypertension 0.76 -4% 0.78 -7% Sat. Fat 0.89 -3% 1.07 -4%
Alcohol 1.17 -2% 1.59 -5% BMI 1.52 -4% 1.92 -9% Vig. Activity 0.74 -2% 0.81 -4% Hypertension 0.90 1% 1.10 -1%
Smoking 1.14 -2% 1.64 -2% Vig. Activity 1.56 3% 2.15 2% Cholesterol 0.73 -1% 0.86 2% BMI 0.91 2% 1.10 -1%
Vig. Activity 1.13 -1% 1.62 -3% Cholesterol 1.55 -1% 2.06 -2% Smoking 0.74 1% 0.85 1% Cholesterol 0.91 0% 1.07 -4%
Sat. Fat 1.14 1% 1.61 -4% Alcohol 1.53 -1% 2.00 -5% Alcohol 0.72 -3% 0.79 -6% Alcohol 0.91 0% 1.09 -2%
160
BMI 1.14 0% 1.62 -3% Smoking 1.52 -1% 2.07 -2% BMI 0.73 2% 0.87 3% Vig. Activity 0.91 1% 1.10 -1%
Men Ethnicity Ethnicity Ethnicity Ethnicity
Stroke AA NH JA LA
Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude
Crude 1.86 Crude 1.65 Crude 1.21 Crude 1.24
Hypertension 1.64
-
12% 1.64 -12% Education 1.47
-
11% 1.47 -11% Education 1.12 -7% 1.12 -7% Education 1.09
-
13% 1.09 -13%
Diabetes 1.56 -5% 1.73 -7% Hypertension 1.34 -9% 1.49 -10% Hypertension 1.05 -6% 1.13 -7% Diabetes 1.01 -7% 1.15 -8%
Education 1.49 -4% 1.76 -5% Diabetes 1.26 -6% 1.52 -8% Sat. Fat 1.10 5% 1.26 4% Smoking 1.05 3% 1.28 2%
Smoking 1.47 -2% 1.84 -1% Vig. Activity 1.29 2% 1.68 2% Diabetes 1.05 -4% 1.14 -6% Hypertension 1.06 1% 1.23 -1%
Vig. Activity 1.45 -1% 1.82 -2% Sat. Fat 1.31 2% 1.67 1% BMI 1.03 -3% 1.21 0% BMI 1.08 2% 1.25 0%
Alcohol 1.44 -1% 1.83 -2% Smoking 1.29 -1% 1.63 -1% Vig. Activity 1.01 -1% 1.18 -2% Sat. Fat 1.07 -1% 1.23 -1%
BMI 1.45 0% 1.84 -1% BMI 1.28 -1% 1.59 -4% Smoking 1.02 1% 1.22 1% Vig. Activity 1.06 0% 1.24 -1%
Cholesterol 1.44 0% 1.83 -2% Alcohol 1.27 0% 1.62 -2% Alcohol 1.02 0% 1.18 -2% Alcohol 1.06 0% 1.24 0%
Sat. Fat 1.45 0% 1.85 -1% Cholesterol 1.27 0% 1.64 -1% Cholesterol 1.02 0% 1.22 1% Cholesterol 1.06 0% 1.23 -1%
Women Ethnicity Ethnicity Ethnicity Ethnicity
AMI AA NH JA LA
Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude
Crude 2.33 Crude 1.62 Crude 0.77 Crude 1.25
Hypertension 1.92
-
17% 1.92 -17% Diabetes 1.36
-
16% 1.36 -16% Alcohol 0.66
-
14% 0.66 -14% Diabetes 1.04
-
17% 1.04 -17%
HRT use 1.69
-
12% 2.05 -12% Education 1.22
-
10% 1.41 -13% BMI 0.72 9% 0.83 7% Education 0.90
-
14% 1.05 -16%
Diabetes 1.53 -9% 1.96 -16% Hypertension 1.13 -7% 1.43 -12% Smoking 0.78 8% 0.85 11% Smoking 0.98 10% 1.36 9%
Cholesterol 1.50 -2% 2.21 -5% HRT use 1.06 -6% 1.48 -9% Diabetes 0.73 -7% 0.70 -9% HRT use 0.92 -6% 1.12 -10%
Alcohol 1.46 -2% 2.14 -8% Vig. Activity 1.09 3% 1.65 2% Education 0.69 -5% 0.70 -9% Alcohol 0.89 -3% 1.15 -8%
Smoking 1.43 -2% 2.33 0% Smoking 1.11 2% 1.64 1% Hypertension 0.66 -4% 0.75 -3% Hypertension 0.90 1% 1.22 -3%
Menopause 1.46 2% 2.36 1% Alcohol 1.07 -4% 1.47 -9% Sat. Fat 0.68 2% 0.81 5% Menopause 0.90 -1% 1.22 -2%
161
Vig. Activity 1.43 -2% 2.25 -3% Sat. Fat 1.08 1% 1.65 2% Vig. Activity 0.67 -1% 0.74 -4% BMI 0.90 1% 1.18 -6%
BMI 1.44 0% 2.09 -10% Cholesterol 1.08 0% 1.62 0% HRT use 0.66 -1% 0.75 -2% Vig. Activity 0.90 -1% 1.21 -3%
Education 1.44 0% 2.25 -3% Menopause 1.07 0% 1.60 -1% Cholesterol 0.66 0% 0.79 3% Cholesterol 0.90 0% 1.23 -2%
Sat. Fat 1.47 2% 2.30 -1% BMI 1.08 1% 1.50 -7% Menopause 0.65 -1% 0.77 0% Sat. Fat 0.89 -1% 1.22 -2%
Women Ethnicity Ethnicity Ethnicity Ethnicity
OHD AA NH JA LA
Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude
Crude 1.85 Crude 2.10 Crude 0.62 Crude 1.08
Hypertension 1.56
-
16% 1.56 -16% Diabetes 1.79
-
15% 1.79 -15% Smoking 0.72 16% 0.72 16% Diabetes 0.91
-
16% 0.91 -16%
Diabetes 1.39
-
10% 1.58 -14% Hypertension 1.65 -7% 1.89 -10% Alcohol 0.61
-
16% 0.54 -13% Smoking 1.04 14% 1.23 15%
HRT use 1.27 -9% 1.65 -10% HRT use 1.55 -6% 1.96 -7% Sat. Fat 0.66 10% 0.69 12% HRT use 0.97 -7% 0.98 -9%
Cholesterol 1.20 -6% 1.68 -9% Sat. Fat 1.61 4% 2.17 3% BMI 0.70 6% 0.65 6% Alcohol 0.92 -5% 1.00 -7%
Smoking 1.23 3% 1.89 2% Smoking 1.65 3% 2.15 2% Diabetes 0.66 -7% 0.56 -9% Education 0.90 -3% 0.98 -9%
Alcohol 1.18 -4% 1.72 -7% Alcohol 1.57 -5% 1.93 -8% Hypertension 0.64 -3% 0.60 -3% Menopause 0.88 -2% 1.05 -2%
Vig. Activity 1.16 -2% 1.77 -4% Vig. Activity 1.61 2% 2.15 2% Vig. Activity 0.62 -2% 0.59 -4% Cholesterol 0.87 -2% 1.03 -4%
BMI 1.17 1% 1.68 -9% Cholesterol 1.58 -2% 2.10 0% Cholesterol 0.61 -2% 0.66 6% Hypertension 0.88 1% 1.05 -3%
Menopause 1.17 -1% 1.82 -1% Menopause 1.57 -1% 2.09 -1% Education 0.60 -1% 0.58 -5% BMI 0.89 2% 1.03 -4%
Sat. Fat 1.17 0% 1.80 -3% Education 1.57 0% 1.96 -7% Menopause 0.61 1% 0.63 2% Vig. Activity 0.89 -1% 1.04 -3%
Education 1.19 2% 1.81 -2% BMI 1.58 1% 1.97 -6% HRT use 0.61 0% 0.61 -1% Sat. Fat 0.89 0% 1.03 -4%
Women Ethnicity Ethnicity Ethnicity Ethnicity
Stroke AA NH JA LA
Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude Factor RR %Δ Single
%
Crude
Crude 1.48 Crude 1.24 Crude 0.87 Crude 1.05
Hypertension 1.28
-
14% 1.28 -14% Hypertension 1.13 -9% 1.13 -9% Smoking 0.97 12% 0.97 12% Education 0.95
-
10% 0.95 -10%
HRT use 1.20 -6% 1.39 -6% Education 1.05 -7% 1.14 -8% Alcohol 0.86
-
11% 0.79 -9% Smoking 1.05 11% 1.16 10%
162
Alcohol 1.16 -3% 1.41 -5% HRT use 1.02 -3% 1.18 -5% Sat. Fat 0.91 5% 0.93 7% Alcohol 0.99 -5% 1.00 -5%
Vig. Activity 1.14 -2% 1.44 -3% Alcohol 0.99 -3% 1.17 -6% Education 0.87 -4% 0.82 -6% HRT use 0.96 -3% 0.99 -6%
BMI 1.16 2% 1.43 -4% Sat. Fat 1.02 3% 1.26 2% Vig. Activity 0.86 -2% 0.85 -3% Diabetes 0.93 -3% 1.01 -5%
Smoking 1.15 -1% 1.50 1% Diabetes 1.00 -2% 1.18 -4% BMI 0.87 1% 0.88 1% Menopause 0.92 -1% 1.03 -2%
Menopause 1.13 -1% 1.46 -1% Vig. Activity 1.01 1% 1.25 1% Hypertension 0.84 -3% 0.85 -2% Sat. Fat 0.91 -1% 1.03 -3%
Diabetes 1.13 -1% 1.42 -4% BMI 1.03 1% 1.21 -2% Diabetes 0.83 -1% 0.85 -2% Hypertension 0.92 1% 1.03 -2%
Education 1.12 0% 1.44 -3% Smoking 1.02 -1% 1.26 2% HRT use 0.82 -1% 0.86 -1% BMI 0.94 2% 1.04 -1%
Cholesterol 1.12 0% 1.45 -2% Menopause 1.01 -1% 1.23 -1% Cholesterol 0.83 0% 0.88 1% Vig. Activity 0.94 0% 1.03 -2%
Sat. Fat 1.15 3% 1.46 -2% Cholesterol 1.02 1% 1.24 0% Menopause 0.83 0% 0.87 0% Cholesterol 0.95 1% 1.04 -1%
W – Whites; AA – African Americans; NH – Native Hawaiians; JA – Japanese Americans; LA – Latinos; AMI – acute myocardial infarction; OHD – other heart
disease; RR – relative risk; BMI – body mass index
163
APPENDIX C:
GBD GWAS
164
Table C.1. Characteristics of Studies included in GWA analysis of Gallbladder Disease
RACE/
ETHNICITY
STUDY STUDY PHENOTYPE
1
N GBD CA/CO
2
AGE (SD)
3
GWAS ARRAY
AA MEC Breast cancer 538 cases 113/425
67.0 (9.5)
Human1M-Duov3-0
990 ctrls. 203/787
67.5 (9.3)
Human1M-Duov3-0
MEC Prostate cancer 1236 cases 165/1071
68.9 (7.2)
Human1M-Duov3-0
1669 ctrls. 188/1481
68.5 (7.8)
Human1M-Duov3-0
JA MEC Breast cancer 882 cases 162/720
59.4 (8.2)
Human660W-Quad_v1
829 ctrls. 135/694
59.2 (8.0)
Human660W-Quad_v1
MEC Prostate cancer 980 cases 198/782
71.5 (7.3)
Human660W-Quad_v1
1005 ctrls. 192/813
71.9 (8.1)
Human660W-Quad_v1
LATINO MEC Breast cancer 520 cases 192/328
66.0 (8.0)
Human660W-Quad_v1
437 ctrls. 162/275
64.1 (7.8)
Human660W-Quad_v1
MEC Prostate cancer 1034 cases 270/764
69.0 (6.5)
Human660W-Quad_v1
709 ctrls. 173/536
68.6 (7.1)
Human660W-Quad_v1
SIGMA MEC T2D 2055 cases 784/1271
59.2 (6.9)
HumanOmni25-4v1
LATINO 2082 ctrls. 509/1573
59.3 (7.0)
HumanOmni25-4v1
DMS T2D 689 cases 50/639
55.8 (11.1)
HumanOmni25-4v1
469 ctrls. 29/440
52.5 (7.7)
HumanOmni25-4v1
MCDS T2D 286 cases 65/221
64.2 (7.5)
HumanOmni25-4v1
612 ctrls. 93/519
62.5 (7.7)
HumanOmni25-4v1
1
Phenotype definition by nested case-control status (i.e. Breast cancer cases and controls)
2
Case-control definition used in this analysis (i.e. GBD cases vs. GBD controls)
3
age indicator
AA, African American; JA, Japanese American; LA, Latino
165
Table C.2. Number of typed and imputed SNPs analyzed by substudy.
STUDY TYPED IMPUTED
AA BREAST CANCER 1,003,188 14,467,603
AA PROSTATE CANCER 1,020,005 14,461,691
JA BREAST CANCER 455,442 7,925,542
JA PROSTATE CANCER 459,715 7,422,608
LA BREAST CANCER 520,273 8,939,225
LA PROSTATE CANCER 512,589 8,895,866
SIGMA STUDIES 1,380,981 7,344,244
SNPs excluded if MAF<1% and INFO < 0.3.
166
Table C.3. Genome wide significant SNPs in the 2p21 region.
Meta AA JA LA
Gene SNP BP A1/A2 OR P Freq OR P Freq OR P Freq OR P Anno
THADA rs6730757 43541598 C/T 1.18 3.99E-08 0.49 1.15 0.02663 0.67 1.18 0.01502 0.60 1.20 8.93E-06 Intronic
THADA rs4952986 43574298 T/C 1.25 6.67E-09 0.04 1.37 0.08447 0.34 1.13 0.1215 0.34 1.28 3.66E-08 Intronic
DYNC2LI1 rs17031488 44001139 C/T 1.22 4.56E-06 0.72 1.06 0.4122 0.02 1.17 0.4467 0.86 1.37 2.18E-08 upstream
DYNC2LI1 rs17031494 44004099 T/G 1.20 2.3E-05 0.25 1.00 0.9634 0.02 1.09 0.6889 0.86 1.37 1.66E-08 Intronic
DYNC2LI1 rs58290925 44005458 A/G 1.21 1.5E-05 0.21 1.00 0.9918 0.02 1.13 0.578 0.85 1.37 1.26E-08 Intronic
DYNC2LI1 rs112767530 44008987 C/T 1.18 0.000129 0.22 1.00 0.9917 0.04 1.22 0.1993 0.87 1.37 3.07E-08 Intronic
DYNC2LI1 rs28473566 44012206 G/A 1.22 9.03E-06 0.79 1.01 0.8814 0.02 1.12 0.5954 0.87 1.39 8.47E-09 Intronic
DYNC2LI1 rs1025447 44022970 T/C 1.22 1.01E-05 0.19 1.01 0.8562 0.02 1.10 0.6673 0.86 1.40 3.72E-09 Intronic
DYNC2LI1 rs7594731 44027302 G/A 1.19 5.69E-06 0.67 1.02 0.7431 0.02 1.09 0.6945 0.82 1.33 1.15E-08 Intronic
DYNC2LI1 rs11556157 44028013 A/T 1.19 5.45E-06 0.67 1.02 0.7561 0.02 1.10 0.6675 0.82 1.34 8.75E-09 Exonic Nonsyn
DYNC2LI1 rs78451356 44033754 T/G 1.23 3.07E-06 0.79 1.01 0.9241 0.02 1.13 0.5767 0.86 1.42 9.37E-10 Intronic
DYNC2LI1 rs59513006 44035123 C/G 1.22 5.36E-06 0.22 1.00 0.9735 0.02 1.13 0.5629 0.86 1.41 1.49E-09 Intronic
DYNC2LI1 rs58795814 44035503 G/C 1.23 3.19E-06 0.79 1.01 0.9413 0.02 1.13 0.5556 0.86 1.41 1.22E-09 Intronic
DYNC2LI1 rs8302 44036913 T/C 1.19 3.69E-05 0.26 1.04 0.608 0.02 1.13 0.5506 0.85 1.40 1.97E-09 UTR3
ABCG5 rs77105521 44039733 C/T 1.22 5.26E-06 0.22 1.00 0.9872 0.02 1.15 0.4905 0.86 1.41 1.3E-09 UTR3
ABCG5 rs112686079 44045005 C/T 1.29 5.1E-06 0.04 1.11 0.5331 0.03 1.14 0.5216 0.88 1.42 2.9E-08 Intronic
ABCG5 rs111559090 44045402 G/A 1.29 7.14E-06 0.04 1.15 0.3914 0.03 1.13 0.5272 0.88 1.42 2.44E-08 Intronic
ABCG5 rs56266464 44054991 A/G 1.56 8.02E-19 0.08 1.20 0.08815 0.01 1.38 0.4185 0.10 1.68 1.12E-19 Intronic
ABCG5 rs114938914 44055510 T/C 1.57 9.94E-20 0.09 1.22 0.05363 0.01 1.41 0.3899 0.11 1.69 2.12E-20 Intronic
ABCG5 rs111617668 44055922 T/C 1.54 2.56E-18 0.09 1.15 0.1802 0.01 1.43 0.3727 0.11 1.68 4.64E-20 Intronic
ABCG5 rs79835740 44056148 T/A 1.57 7.79E-20 0.09 1.22 0.05344 0.01 1.43 0.3651 0.11 1.70 1.69E-20 Intronic
ABCG5 rs3923912 44057570 C/A 1.52 8.76E-18 0.11 1.15 0.1436 0.01 1.66 0.3991 0.11 1.68 1.41E-19 Intronic
ABCG5 rs77725792 44058201 A/G 1.53 3E-18 0.11 1.16 0.1074 0.01 1.67 0.3956 0.11 1.68 5.61E-20 Intronic
ABCG5 rs115445558 44063098 C/G 1.59 4.72E-20 0.08 1.18 0.1187 0.01 1.15 0.7406 0.10 1.74 8.27E-22 Intronic
ABCG5 rs6756629 44065090 A/G 1.60 2.15E-20 0.08 1.20 0.09541 0.01 1.55 0.4595 0.10 1.74 1.06E-21 Exonic Nonsyn
ABCG8 rs11887534 44066247 C/G 1.62 1.05E-20 0.07 1.18 0.1667 0.01 1.07 0.8738 0.10 1.76 2.19E-22 Exonic Nonsyn
ABCG8 rs75331444 44069772 A/G 1.60 2.69E-20 0.08 1.21 0.07986 0.01 1.27 0.5694 0.10 1.74 1.24E-21 Intronic
ABCG8 rs4299376 44072576 T/G 1.40 3.87E-15 0.82 1.16 0.05614 NA NA NA 0.81 1.52 3.57E-16 Intronic
ABCG8 rs41360247 44073656 C/T 1.52 4.75E-18 0.11 1.16 0.1157 0.01 1.28 0.5517 0.10 1.69 5.57E-20 Intronic
167
ABCG8 rs6544713 44073881 C/T 1.42 1.77E-15 0.83 1.19 0.03706 NA NA NA 0.81 1.52 7.35E-16 Intronic
ABCG8 rs4953023 44074000 A/G 1.58 6.89E-20 0.08 1.21 0.07567 0.01 1.28 0.5494 0.10 1.72 4.93E-21 Intronic
ABCG8 rs4245791 44074431 T/C 1.44 1.41E-15 0.86 1.17 0.08746 NA NA NA 0.81 1.53 2.4E-16 Intronic
ABCG8 rs76866386 44075483 C/T 1.56 4.89E-19 0.09 1.14 0.1934 0.01 1.29 0.5472 0.10 1.73 2.67E-21 Intronic
ABCG8 rs56132765 44078853 A/G 1.61 2.23E-20 0.08 1.20 0.1101 0.01 2.49 0.1242 0.10 1.74 2.37E-21 Exonic Syn
ABCG8 rs72875462 44079310 A/C 1.53 8.94E-20 0.16 1.22 0.01609 0.01 2.53 0.1063 0.10 1.70 1.44E-20 Intronic
ABCG8 rs6709904 44080324 G/A 1.30 3.4E-11 0.26 1.16 0.02938 0.04 1.09 0.577 0.14 1.40 1.82E-11 Intronic
ABCG8 rs182725015 44081561 C/G 2.62 6.22E-11 NA NA NA NA NA NA 0.02 2.62 6.29E-11 Intronic
ABCG8 rs4076834 44081627 G/T 1.45 1.57E-17 0.19 1.22 0.01042 0.05 1.10 0.5276 0.11 1.66 3.64E-19 Intronic
ABCG8 rs9753033 44084600 T/A 1.51 4.54E-18 0.11 1.23 0.02515 0.04 1.09 0.5847 0.10 1.71 8.98E-20 Intronic
ABCG8 rs4507142 44086591 A/G 1.53 3.29E-19 0.11 1.30 0.00519 0.04 1.10 0.5675 0.10 1.71 8.42E-20 Intronic
ABCG8 rs4614977 44087024 G/C 1.52 2.64E-18 0.11 1.24 0.02101 0.04 1.10 0.5665 0.10 1.71 7.26E-20 Intronic
ABCG8 rs60668987 44088039 A/G 1.85 6.38E-18 0.05 1.37 0.03994 NA NA NA 0.06 2.00 4.78E-18 Intronic
ABCG8 rs58762754 44088041 A/C 1.67 2.71E-15 0.05 1.25 0.161 0.03 1.02 0.9185 0.07 1.94 5.25E-18 Intronic
ABCG8 rs60567145 44088096 G/T 1.54 4.04E-14 0.08 1.13 0.3227 0.01 1.46 0.396 0.09 1.69 1.15E-15 Intronic
ABCG8 rs4603816 44092066 T/C 1.47 3.38E-14 0.91 1.11 0.3357 NA NA NA 0.85 1.58 9.09E-16 Intronic
ABCG8 rs6733452 44094845 A/G 1.43 5.2E-11 0.10 1.25 0.02318 0.04 1.11 0.4978 0.06 1.63 1.65E-11 Intronic
ABCG8 rs150478002 44095585 T/C 1.69 7.15E-10 0.98 1.37 0.2662 NA NA NA 0.05 1.83 9.44E-12 Intronic
ABCG8 rs6740545 44096139 A/G 1.48 1.32E-15 0.89 1.24 0.03227 NA NA NA 0.84 1.56 1.94E-15 Intronic
ABCG8 rs6544716 44096259 G/T 1.47 5.02E-15 0.88 1.20 0.07252 NA NA NA 0.85 1.58 1.3E-15 Intronic
ABCG8 rs6755809 44096336 A/T 1.49 1.42E-15 0.89 1.24 0.03883 NA NA NA 0.85 1.58 1.35E-15 Intronic
ABCG8 rs6544717 44096402 A/G 1.49 3.57E-14 0.94 1.13 0.3539 NA NA NA 0.85 1.57 4.52E-15 Intronic
ABCG8 rs4952688 44096770 A/T 1.48 1.1E-14 0.91 1.16 0.201 NA NA NA 0.85 1.58 1.22E-15 Intronic
ABCG8 rs4953026 44097367 T/C 1.49 1.29E-15 0.89 1.24 0.03631 NA NA NA 0.85 1.58 1.34E-15 Intronic
NA indicates that the SNP was not tested in this analysis, in that racial/ethnic population.
168
Table C.4. Genome-wide significant SNPs in the 7p22.2 region in Japanese Americans in the MEC.
Meta AA JA LA
Gene SNP BP A1/A2 OR P Freq OR P Freq OR P Freq OR P FUNC
SDK1 rs4723539 4292528 T/C 1.11 0.009009 0.14 1.16 0.09 0.03 2.29 2.92E-08 0.21 1.02 0.6188 Intronic
SDK1 rs12531029 4293056 G/A 1.12 0.003815 0.12 1.17 0.08 0.03 2.28 3.28E-08 0.20 1.04 0.4254 Intronic
SDK1 rs13226493 4293524 C/A 1.10 0.01095 0.14 1.14 0.14 0.03 2.31 2.39E-08 0.20 1.02 0.6256 Intronic
NA indicates that the SNP was not tested in this analysis, in that racial/ethnic population.
169
Table C.5. Allele frequencies of SNPs in LD with rs13226493, by race/ethnicity in the MEC
SNP r
2
D' Distance
1
P AA Freq AA P JA Freq JA P LA Freq LA Annotation
rs56250749 0.74 1.00 14995 0.08699 0.12 1.4E-06 0.04 0.5916 0.19 Intronic
rs4723539 1.00 1.00 996 0.09 0.14 2.92E-08 0.03 0.6188 0.21 Intronic
rs6966643 1.00 1.00 532 0.6627 0.34 3.76E-07 0.04 0.9536 0.22 Intronic
rs12531029 0.88 1.00 468 0.08365 0.88 3.28E-08 0.97 0.4254 0.80 Intronic
rs13226493 1.00 1.00 0 0.1358 0.86 2.39E-08 0.97 0.6256 0.80 Intronic
rs10951486 1.00 1.00 -452 0.6229 0.66 4.86E-05 0.95 0.9784 0.78 Intronic
rs4723545 1.00 1.00 -2265 0.1323 0.59 4.86E-07 0.96 0.884 0.77 Intronic
rs67819850 1.00 1.00 -4518 0.09283 0.88 1.32E-07 0.97 0.665 0.80 Intronic
rs71527479 1.00 1.00 -6811 0.07241 0.86 5.13E-08 0.96 0.7668 0.80 Intronic
rs13247068 1.00 1.00 -7666 0.1038 0.14 1.31E-07 0.03 0.796 0.21 Intronic
rs4720208 0.74 1.00 -9463 0.8502 0.01 1.23E-06 0.02 0.8165 0.05 Intronic
500kb window, R2 > 0.2.
1
Distance in base pairs (bp) from index rs13226493.
2
P-value not available as SNP not analyzed in current analysis.
3
Allele frequency pulled from 1000 Genomes Phase 3 data by race/ethnic group using rAggr (http://raggr.usc.edu).
170
Table C.6. Genome-wide significant SNPs in the 14q21.3 and 15q21.3 regions in African Americans in the MEC.
Meta AA JA LA
SNP BP A1/A2 OR P Freq OR P Freq OR P Freq OR P FUNC
rs72672032 49517763 T/A 1.87 1.84E-06 0.03 2.45 2.01E-08 NA NA NA 0.01 1.07 0.75 intergenic
rs72672033 49519111 A/C 1.85 1.83E-06 0.03 2.40 2.59E-08 NA NA NA 0.01 1.08 0.72 intergenic
rs72672035 49520230 T/C 1.87 1.93E-06 0.03 2.45 2.87E-08 NA NA NA 0.01 1.09 0.70 intergenic
rs56412753 49523861 T/C 1.85 2.03E-06 0.03 2.38 4.3E-08 NA NA NA 0.01 1.11 0.66 intergenic
rs55808624 49524829 A/T 1.85 1.98E-06 0.03 2.38 4.32E-08 NA NA NA 0.01 1.11 0.65 intergenic
rs57317656 49526512 C/T 1.86 1.87E-06 0.03 2.38 4.21E-08 NA NA NA 0.01 1.11 0.65 intergenic
rs72672038 49527489 G/C 1.85 1.89E-06 0.03 2.37 4.23E-08 NA NA NA 0.01 1.11 0.65 intergenic
rs72672040 49527790 C/T 1.85 1.95E-06 0.03 2.37 4.33E-08 NA NA NA 0.01 1.11 0.65 intergenic
rs7177754 53251332 G/A 1.63 3.95E-08 0.12 1.63 3.93E-08 NA NA NA NA NA NA intergenic
NA indicates that the SNP was not tested in this analysis, in that racial/ethnic population.
171
Table C.7. Known GBD risk SNPs and associations in MEC
GENE
1
SNP A1/A2 Reported OR
2
Reported P
2
OR
3
P
3
Anno
1
Exonic Effect
DOCK7 rs10889353
4
A/C 2.46 0.0173 0.96 0.2973 Intronic
APOB rs754523
4
A/G 2.77 0.0241 0.89 0.08974 Intergenic
GCKR* rs1260326
5
T/C 0.86 5.88×10−7 0.95 0.1715 Exonic Non-synonymous
UGT1A1 rs6742078
6
G/T 1.27 0.003 1.13 0.001226 Intronic
ESR1 rs2234693
7
T/C 1.50 0.001 0.98 0.559 Intronic
ESR1 rs9340799
7
A/G 1.40 0.005 1.04 0.5996 Intronic
TTC39B rs661048
5
A/G 1.31 2.04×10−6 1.11 0.04841 Intronic
TTC39B rs686030
5
A/C 1.31 6.95×10−7 1.09 0.1378 Intronic
ABCA1 rs4149268
4
C/T 3.00 0.0153 0.94 0.1466 Intronic
ABCA1* rs1883025
4
C/T 2.97 0.033 1.11 0.1116 Intronic
BUD13 rs28927680
4
C/G 3.42 0.0192 1.07 0.211 UTR3
SLCO1B1 rs11045819 C/A 2.22 0.034 1.11 0.3005 Exonic Non-synonymous
HNF1A rs2650000
4
A/C 0.98-1.50 0.0373 0.93 0.02774 Intergenic
CETP* rs3764261
4
C/A 2.59 0.0047 0.99 0.2255 Intergenic
LIPG rs4939883
4
T/C 3.60 0.0271 1.10 0.2055 Intergenic
SUGP1* rs10401969
4
T/C 3.94 0.0189 1.09 0.2894 Intronic
CILP2 rs17216525
4
C/T 2.14 0.0445 0.87 0.09719 Intergenic
PCIF1 rs7679
4
T/C 2.64 0.0048 1.01 0.8698 UTR3
SLC10A2 rs9514089
8
C/T 2.04 0.00767 1.00 0.46 Intronic
1
Gene containing/closest to based on annotation via ANNOVAR.
2
Strongest previously reported P-value and associated OR.
3
OR and P-value for
race/ethnicity of most significant result in this study.
4
Source from Goodloe 2013.
5
Source from Rodriguez 2015.
6
Source from Buch 2010.
7
Source from Srivastava 2012 and Srivastava 2011.
8
Source from Renner 2009.
172
Table C.8. Known lipid-related risk variants and associations with GBD in the MEC.
Meta AA JA LA
SNP CHR:BP Lipid trait
1
OR P RAF OR P RAF OR P RAF OR P
rs1077514 1:23766233 TC 1.03 4.20E-01 0.54 1.04 5.18E-01 0.32 1.03 6.38E-01 0.19 1.05 3.15E-01
rs12027135 1:25775733 LDL-c, TC 1.04 1.83E-01 0.46 1.01 8.31E-01 0.68 1.02 7.61E-01 0.45 1.06 9.13E-02
rs12748152 1:27138393 LDL-c, HDL-c, Trig 1.02 8.32E-01 0.02 1.27 2.09E-01 NA
2
NA
2
NA
2
0.97 1.10 3.62E-01
rs4660293 1:40028180 HDL-c 1.00 9.22E-01 0.94 1.13 3.88E-01 0.16 1.01 9.02E-01 0.14 1.02 7.09E-01
rs2479409 1:55504650 LDL-c, TC 1.03 3.15E-01 0.73 1.07 3.52E-01 0.66 1.05 5.58E-01 0.59 1.04 2.95E-01
rs2131925 1:63025942 LDL-c, Trig, TC 1.05 8.56E-02 0.64 1.06 3.94E-01 0.23 1.03 7.26E-01 0.40 1.06 1.33E-01
rs7515577 1:93009438 HDL-c 1.04 4.77E-01 0.08 1.60 4.52E-06 0.02 1.03 9.07E-01 0.11 1.13 5.33E-02
rs629301 1:109818306 LDL-c, TC 1.09 1.10E-02 0.35 1.04 5.87E-01 0.07 1.18 1.45E-01 0.20 1.11 2.11E-02
rs267733 1:150958836 LDL-c 1.04 4.93E-01 0.93 1.05 6.70E-01 0.04 1.22 1.84E-01 0.09 1.03 6.34E-01
rs12145743 1:156700651 HDL-c 1.02 6.39E-01 0.91 1.06 6.32E-01 0.16 1.03 7.55E-01 0.20 1.03 5.87E-01
rs4650994 1:178515312 HDL-c 1.04 1.43E-01 0.22 1.14 6.98E-02 0.53 1.08 2.15E-01 0.62 1.07 9.30E-02
rs1689800 1:182168885 HDL-c 1.05 1.22E-01 0.33 1.01 8.39E-01 0.25 1.07 3.30E-01 0.31 1.05 1.78E-01
rs2642442 1:220973563 LDL-c, TC 1.01 8.57E-01 0.23 1.01 8.60E-01 0.86 1.15 1.74E-01 0.20 1.01 7.70E-01
rs4846914 1:230295691 HDL-c, Trig 1.02 4.89E-01 0.83 1.03 6.93E-01 0.24 1.08 3.17E-01 0.41 1.02 6.11E-01
rs514230 1:234858597 LDL-c, TC 1.07 2.97E-02 0.26 1.04 5.64E-01 0.30 1.03 6.07E-01 0.69 1.09 2.80E-02
rs1367117 2:21263900 LDL-c, TC 1.04 2.29E-01 0.14 1.05 6.12E-01 0.09 1.03 8.16E-01 0.72 1.07 8.77E-02
rs1260326 2:27730940 Trig, TC 1.03 3.18E-01 0.16 1.10 2.41E-01 0.45 1.05 4.71E-01 0.66 1.05 1.72E-01
rs4299376 2:44072576 LDL-c, TC 1.40 3.87E-15 0.82 1.16 5.61E-02 NA
2
NA
2
NA
2
0.19 1.52 3.57E-16
rs2710642 2:63149557 LDL-c 1.02 5.38E-01 0.15 1.11 2.29E-01 0.77 1.03 6.40E-01 0.35 1.02 6.53E-01
rs10490626 2:118835841 LDL-c, TC 1.11 2.26E-01 0.02 1.26 2.73E-01 NA NA NA 0.96 1.08 4.02E-01
rs2030746 2:121309488 LDL-c, TC 1.02 4.68E-01 0.50 1.04 5.08E-01 0.45 1.00 9.63E-01 0.59 1.02 5.60E-01
rs7570971 2:135837906 TC 1.12 4.93E-03 0.83 1.03 7.06E-01 NA
2
NA
2
NA
2
0.21 1.17 8.20E-04
rs12328675 2:165540800 HDL-c 1.01 9.00E-01 0.15 1.04 6.22E-01 NA
2
NA
2
NA
2
0.09 1.02 8.19E-01
rs2287623 2:169830155 TC 1.02 4.05E-01 0.57 1.03 6.71E-01 0.22 1.02 8.04E-01 0.55 1.04 2.34E-01
rs11694172 2:203532304 TC 1.03 4.70E-01 0.08 1.05 6.40E-01 0.86 1.11 2.35E-01 0.21 1.02 6.58E-01
rs1047891 2:211540507 HDL-c 1.07 7.25E-02 NA NA NA 0.85 1.04 6.99E-01 0.70 1.08 7.23E-02
rs1250229 2:216304384 LDL-c 1.04 2.97E-01 0.19 1.05 5.45E-01 0.94 1.08 5.83E-01 0.76 1.05 2.80E-01
173
rs2972146 2:227100698 HDL-c, Trig 1.04 3.47E-01 0.82 1.06 4.59E-01 0.09 1.06 5.75E-01 0.20 1.05 3.25E-01
rs11563251 2:234679384 LDL-c, TC 1.05 2.41E-01 0.35 1.12 6.62E-02 0.87 1.01 8.86E-01 0.91 1.29 1.63E-04
rs2606736 3:11400249 HDL-c 1.02 5.01E-01 0.57 1.07 2.82E-01 0.32 1.03 6.32E-01 0.56 1.00 9.62E-01
rs2290159 3:12628920 TC 1.08 3.43E-02 0.29 1.00 9.69E-01 0.04 1.25 1.38E-01 0.83 1.11 2.76E-02
rs7640978 3:32533010 TC 1.03 5.14E-01 0.74 1.01 8.75E-01 0.95 1.02 8.84E-01 0.93 1.06 4.43E-01
rs2290547 3:47061183 HDL-c 1.02 5.84E-01 0.96 1.24 1.78E-01 0.79 1.09 2.43E-01 0.88 1.04 5.24E-01
rs2013208 3:50129399 HDL-c 1.01 6.42E-01 0.57 1.05 3.95E-01 0.83 1.06 4.62E-01 0.31 1.06 1.31E-01
rs13326165 3:52532118 HDL-c 1.03 4.82E-01 0.32 1.09 2.01E-01 0.02 1.09 7.15E-01 0.84 1.01 8.42E-01
rs13315871 3:58381287 TC 1.04 5.53E-01 0.94 1.10 4.66E-01 NA
2
NA
2
NA
2
0.95 1.02 8.14E-01
rs6805251 3:119560606 LDL-c, TC 1.01 7.80E-01 0.85 1.16 1.43E-01 0.57 1.07 2.85E-01 0.64 1.06 1.15E-01
rs17404153 3:132163200 LDL-c, HDL-c 1.04 3.00E-01 0.97 1.04 8.33E-01 0.15 1.04 6.49E-01 0.84 1.05 3.05E-01
rs645040 3:135926622 HDL-c 1.06 1.07E-01 0.29 1.02 7.36E-01 0.19 1.11 1.57E-01 0.20 1.05 2.67E-01
rs6831256 4:3473139 HDL-c 1.01 6.08E-01 0.40 1.05 4.65E-01 0.37 1.12 6.44E-02 0.54 1.00 9.82E-01
rs10019888 4:26062990 HDL-c 1.07 1.65E-01 0.21 1.09 2.37E-01 NA
2
NA
2
NA
2
0.12 1.05 3.98E-01
rs442177 4:88030261 Trig, 1.02 4.81E-01 0.50 1.02 7.11E-01 0.47 1.10 1.23E-01 0.32 1.07 9.18E-02
rs3822072 4:89741269 HDL-c 1.03 2.63E-01 0.50 1.10 1.32E-01 0.35 1.07 2.84E-01 0.58 1.00 9.24E-01
rs2602836 4:100014805 HDL-c 1.03 2.87E-01 0.65 1.11 1.11E-01 0.16 1.04 6.58E-01 0.44 1.01 8.20E-01
rs13107325 4:103188709 HDL-c 1.06 5.26E-01 0.98 1.11 6.60E-01 NA
2
NA
2
NA
2
0.96 1.05 6.12E-01
rs6450176 5:53298025 LDL-c, TC 1.06 6.81E-02 0.71 1.09 2.07E-01 0.55 1.05 4.27E-01 0.74 1.05 2.46E-01
rs9686661 5:55861786 LDL-c, HDL-c, Trig, TC 1.03 4.70E-01 0.22 1.01 9.17E-01 0.10 1.04 6.95E-01 0.79 1.03 5.03E-01
rs12916 5:74656539 LDL-c, TC 1.00 8.82E-01 0.27 1.04 5.25E-01 0.46 1.01 9.05E-01 0.40 1.00 9.35E-01
rs4530754 5:122855416 LDL-c, TC 1.05 9.07E-02 0.77 1.08 2.96E-01 0.33 1.02 7.10E-01 0.52 1.05 1.65E-01
rs6882076 5:156390297 LDL-c, Trig, TC 1.05 1.62E-01 0.61 1.12 6.44E-02 0.83 1.01 9.05E-01 0.76 1.03 5.26E-01
rs3757354 6:16127407 LDL-c, TC 1.03 2.60E-01 0.66 1.04 5.26E-01 0.45 1.05 4.12E-01 0.67 1.05 1.73E-01
rs1800562 6:26093141 LDL-c, TC 1.21 1.50E-01 0.02 1.09 7.13E-01 NA
2
NA
2
NA
2
0.98 1.35 5.33E-02
rs3177928 6:32412435 HDL-c, TC 1.12 2.88E-01 0.91 1.12 2.88E-01 NA
2
NA
2
NA
2
NA
2
NA
2
NA
2
rs2814982 6:34546560 TC 1.03 4.61E-01 0.73 1.09 2.05E-01 0.92 1.10 4.09E-01 0.91 1.04 5.88E-01
rs2758886 6:39250837 TC 1.05 2.45E-01 0.18 1.15 7.89E-02 0.99 1.14 8.06E-01 0.79 1.02 7.22E-01
rs998584 6:43757896 HDL-c, TC 1.02 5.73E-01 0.78 1.06 4.56E-01 0.46 1.06 3.75E-01 0.43 1.01 8.45E-01
rs9488822 6:116312893 TC 1.03 2.98E-01 0.38 1.00 9.57E-01 0.18 1.01 9.40E-01 0.45 1.05 2.19E-01
174
rs1936800 6:127436064 HDL-c, Trig 1.00 9.34E-01 0.52 1.05 4.14E-01 0.51 1.01 8.79E-01 0.50 1.03 4.93E-01
rs9376090 6:135411228 HDL-c 1.05 1.76E-01 0.94 1.10 4.84E-01 0.35 1.01 8.48E-01 0.15 1.09 9.02E-02
rs605066 6:139829666 HDL-c 1.00 9.60E-01 0.53 1.09 1.88E-01 0.74 1.01 8.43E-01 0.45 1.04 3.41E-01
rs1564348 6:160578860 LDL-c, TC 1.02 6.48E-01 0.12 1.04 6.77E-01 NA
2
NA
2
NA
2
0.24 1.01 7.55E-01
rs1997243 7:1083777 TC 1.01 8.67E-01 0.96 1.02 9.20E-01 NA
2
NA
2
NA
2
0.08 1.02 8.24E-01
rs702485 7:6449272 LDL-c, Trig, TC 1.05 1.36E-01 0.76 1.06 4.14E-01 0.91 1.10 3.55E-01 0.64 1.04 3.16E-01
rs4142995 7:17919258 HDL-c 1.02 5.76E-01 0.70 1.04 5.12E-01 0.50 1.08 2.19E-01 0.53 1.04 2.74E-01
rs12670798 7:21607352 LDL-c, TC 1.03 3.48E-01 0.32 1.14 4.36E-02 0.52 1.01 8.95E-01 0.18 1.01 8.34E-01
rs4722551 7:25991826 LDL-c, Trig, TC 1.00 9.26E-01 0.94 1.11 4.30E-01 0.01 1.40 2.36E-01 0.16 1.00 9.87E-01
rs2072183 7:44579180 LDL-c, TC 1.03 4.16E-01 0.22 1.05 5.34E-01 0.37 1.02 8.13E-01 0.74 1.03 5.63E-01
rs4917014 7:50305863 HDL-c 1.03 2.73E-01 0.11 1.04 6.95E-01 0.51 1.13 4.35E-02 0.53 1.01 7.85E-01
rs13238203 7:72129667 Trig 1.07 7.07E-01 0.99 1.69 2.11E-01 NA
2
NA
2
NA
2
0.98 1.03 8.70E-01
rs17145738 7:72982874 HDL-c, Trig, 1.01 8.21E-01 0.90 1.24 4.50E-02 0.10 1.13 2.31E-01 0.93 1.06 4.33E-01
rs38855 7:116358044 Trig 1.04 1.83E-01 0.23 1.10 1.96E-01 0.59 1.05 4.29E-01 0.60 1.02 5.77E-01
rs4731702 7:130433384 HDL-c 1.04 1.78E-01 0.28 1.07 3.02E-01 0.31 1.01 8.32E-01 0.58 1.04 3.06E-01
rs17173637 7:150529449 HDL-c 1.24 1.21E-03 0.07 1.42 1.44E-03 NA
2
NA
2
NA
2
0.05 1.15 1.03E-01
rs9987289 8:9183358 HDL-c, Trig, 1.04 2.85E-01 0.82 1.26 6.55E-03 0.99 3.69 2.04E-01 0.77 1.01 7.74E-01
rs11776767 8:10683929 Trig 1.03 3.50E-01 0.64 1.04 5.04E-01 0.20 1.20 1.30E-02 0.74 1.03 5.27E-01
rs1495741 8:18272881 Trig, TC 1.01 6.95E-01 0.37 1.03 6.82E-01 0.67 1.04 5.50E-01 0.36 1.04 2.63E-01
rs12678919 8:19844222 HDL-c, Trig, 1.02 6.51E-01 0.90 1.09 3.75E-01 0.12 1.03 7.75E-01 0.07 1.08 2.85E-01
rs10102164 8:55421614 LDL-c, TC 1.02 5.37E-01 0.83 1.02 7.82E-01 0.21 1.09 2.10E-01 0.87 1.01 9.10E-01
rs2081687 8:59388565 LDL-c, TC 1.10 7.43E-03 0.26 1.04 5.81E-01 0.20 1.21 8.15E-03 0.79 1.08 1.11E-01
rs2293889 8:116599199 HDL-c 1.02 5.32E-01 0.13 1.05 6.05E-01 0.73 1.01 8.35E-01 0.65 1.03 3.90E-01
rs2954029 8:126490972 LDL-c, HDL-c, Trig, TC 1.01 6.83E-01 0.66 1.03 6.07E-01 0.51 1.01 8.91E-01 0.38 1.04 3.60E-01
rs11136341 8:145043543 LDL-c, TC 1.01 8.34E-01 0.50 1.02 7.52E-01 0.85 1.09 3.17E-01 0.25 1.02 5.86E-01
rs3780181 9:2640759 LDL-c, TC 1.00 9.39E-01 0.81 1.02 8.21E-01 0.10 1.05 6.30E-01 0.11 1.00 9.98E-01
rs581080 9:15305378 HDL-c, TC 1.02 5.92E-01 0.47 1.16 1.75E-02 0.93 1.24 9.81E-02 0.13 1.04 4.99E-01
rs1883025 9:107664301 HDL-c, TC 1.02 4.47E-01 0.32 1.11 1.12E-01 0.28 1.16 2.88E-02 0.72 1.05 1.89E-01
rs1832007 10:5254847 Trig, 1.02 5.57E-01 0.95 1.02 8.60E-01 0.91 1.01 9.24E-01 0.18 1.04 4.37E-01
rs10904908 10:17260290 TC 1.02 5.39E-01 0.64 1.00 9.54E-01 0.28 1.13 7.31E-02 0.35 1.08 6.16E-02
175
rs970548 10:46013277 Trig, 1.08 2.82E-02 0.20 1.07 3.64E-01 0.05 1.32 3.69E-02 0.31 1.06 1.41E-01
rs10761731 10:65027610 Trig, 1.05 1.03E-01 0.33 1.10 1.43E-01 0.53 1.02 7.45E-01 0.28 1.06 1.30E-01
rs2068888 10:94839642 Trig, 1.02 4.38E-01 0.70 1.05 4.91E-01 0.61 1.04 4.87E-01 0.55 1.04 3.31E-01
rs2255141 10:113933886 LDL-c, TC 1.01 7.76E-01 0.08 1.08 4.91E-01 0.21 1.01 9.32E-01 0.71 1.00 9.73E-01
rs2923084 11:10388782 HDL-c 1.08 7.85E-03 0.55 1.03 6.25E-01 0.42 1.19 5.50E-03 0.41 1.06 1.32E-01
rs10128711 11:18632984 TC 1.03 3.53E-01 0.75 1.04 5.60E-01 0.47 1.03 6.84E-01 0.65 1.04 2.52E-01
rs3136441 11:46743247 HDL-c 1.07 3.68E-02 0.96 1.11 5.32E-01 0.61 1.04 5.77E-01 0.35 1.10 1.96E-02
rs11246602 11:51512090 HDL-c 1.03 5.95E-01 0.94 1.11 4.27E-01 0.14 1.10 2.81E-01 0.10 1.03 7.45E-01
rs174546 11:61569830 LDL-c, HDL-c, Trig, TC 1.08 1.68E-02 0.10 1.06 5.65E-01 0.41 1.05 4.14E-01 0.41 1.10 2.40E-02
rs12801636 11:65391317 HDL-c 1.04 1.90E-01 0.26 1.06 3.75E-01 0.49 1.12 6.85E-02 0.56 1.00 9.29E-01
rs499974 11:75455021 HDL-c 1.02 6.20E-01 0.33 1.04 5.06E-01 0.30 1.09 2.12E-01 0.86 1.04 4.35E-01
rs964184 11:116648917 LDL-c, TC 1.01 7.81E-01 0.19 1.06 4.56E-01 0.71 1.01 8.71E-01 0.30 1.03 5.16E-01
rs11603023 11:118486067 TC 1.01 6.60E-01 0.49 1.04 4.74E-01 0.74 1.12 1.14E-01 0.73 1.01 8.75E-01
rs7941030 11:122522375 LDL-c, TC 1.02 4.24E-01 0.42 1.08 2.03E-01 0.38 1.06 3.71E-01 0.34 1.01 7.70E-01
rs11220462 11:126243952 LDL-c, TC 1.05 2.29E-01 0.05 1.16 2.61E-01 0.42 1.09 1.53E-01 0.87 1.00 9.70E-01
rs4883201 12:9082581 TC 1.01 8.71E-01 0.03 1.06 7.66E-01 0.77 1.09 2.64E-01 0.08 1.08 2.52E-01
rs7134375 12:20473758 HDL-c 1.06 5.10E-02 0.66 1.04 5.75E-01 0.23 1.01 8.91E-01 0.54 1.08 2.79E-02
rs11613352 12:57792580 HDL-c, Trig 1.01 6.83E-01 0.11 1.09 3.84E-01 0.08 1.03 7.66E-01 0.62 1.04 3.53E-01
rs7134594 12:110000193 HDL-c 1.01 6.23E-01 0.29 1.05 4.34E-01 0.71 1.06 3.96E-01 0.52 1.01 7.88E-01
rs11065987 12:112072424 LDL-c, TC 1.04 3.05E-01 0.90 1.01 9.15E-01 NA
2
NA
2
NA
2
0.25 1.05 2.86E-01
rs1169288 12:121416650 LDL-c, TC 1.09 6.39E-03 0.87 1.08 4.44E-01 0.48 1.07 2.67E-01 0.35 1.10 1.69E-02
rs4759375 12:123796238 HDL-c 1.03 4.81E-01 0.89 1.09 4.23E-01 0.27 1.16 3.30E-02 0.83 1.01 8.05E-01
rs4765127 12:124460167 HDL-c, Trig 1.04 1.86E-01 0.35 1.08 2.15E-01 0.17 1.01 9.23E-01 0.72 1.04 3.95E-01
rs838880 12:125261593 Trig 1.02 5.91E-01 0.65 1.03 6.80E-01 0.46 1.08 2.24E-01 0.40 1.01 8.19E-01
rs4942486 13:32953388 LDL-c 1.04 1.21E-01 0.51 1.02 7.30E-01 0.35 1.02 7.17E-01 0.51 1.06 1.10E-01
rs8017377 14:24883887 HDL-c, TC 1.02 6.47E-01 0.82 1.01 8.59E-01 0.06 1.02 8.96E-01 0.75 1.02 6.24E-01
rs4983559 14:105277209 HDL-c 1.01 7.23E-01 0.43 1.06 3.16E-01 0.23 1.02 7.50E-01 0.48 1.01 7.54E-01
rs2412710 15:42683787 Trig 1.15 7.31E-02 0.93 1.06 6.05E-01 NA
2
NA
2
NA
2
0.97 1.32 5.91E-03
rs2929282 15:44245931 Trig 1.06 1.76E-01 0.70 1.14 5.73E-02 0.02 1.14 4.93E-01 0.09 1.02 7.02E-01
rs1532085 15:58683366 HDL-c, Trig, TC 1.01 7.79E-01 0.51 1.02 7.99E-01 0.60 1.03 6.55E-01 0.63 1.00 9.54E-01
176
rs2652834 15:63396867 HDL-c 1.03 5.55E-01 0.29 1.01 8.52E-01 0.02 1.03 9.22E-01 0.87 1.03 5.48E-01
rs3198697 16:15129940 Trig 1.02 7.00E-01 0.90 1.03 7.40E-01 NA
2
NA
2
NA
2
0.78 1.03 5.71E-01
rs11649653 16:30918487 Trig 1.09 6.50E-03 0.10 1.13 2.12E-01 0.11 1.03 7.53E-01 0.45 1.14 5.50E-04
rs1121980 16:53809247 HDL-c, Trig 1.07 3.05E-02 0.46 1.00 9.95E-01 0.23 1.15 5.31E-02 0.68 1.07 7.61E-02
rs3764261 16:56993324 LDL-c, HDL-c, Trig, TC 1.02 6.10E-01 0.32 1.02 8.13E-01 0.79 1.01 9.40E-01 0.68 1.02 5.82E-01
rs16942887 16:67928042 HDL-c 1.03 5.16E-01 0.21 1.09 2.42E-01 0.05 1.00 9.87E-01 0.85 1.09 9.97E-02
rs2000999 16:72108093 LDL-c, TC 1.03 3.73E-01 0.91 1.11 3.43E-01 0.61 1.05 4.72E-01 0.82 1.01 8.22E-01
rs2925979 16:81534790 HDL-c 1.06 9.16E-02 0.31 1.06 3.41E-01 0.34 1.07 3.11E-01 0.82 1.05 3.25E-01
rs314253 17:7091650 LDL-c, TC 1.01 8.11E-01 0.61 1.05 4.77E-01 0.49 1.05 4.63E-01 0.49 1.01 7.41E-01
rs11869286 17:37813856 HDL-c 1.02 5.33E-01 0.28 1.01 9.37E-01 0.48 1.02 7.41E-01 0.41 1.02 5.20E-01
rs8077889 17:41878166 LDL-c 1.03 4.78E-01 0.81 1.14 1.06E-01 NA
2
NA
2
NA
2
0.13 1.12 4.53E-02
rs7206971 17:45425115 HDL-c 1.02 5.46E-01 0.53 1.03 6.14E-01 0.27 1.11 1.21E-01 0.63 1.05 1.73E-01
rs1801689 17:64210580 LDL-c 1.21 8.20E-02 0.01 1.02 9.71E-01 NA
2
NA
2
NA
2
0.03 1.22 7.64E-02
rs4148008 17:66875294 HDL-c 1.00 9.61E-01 0.53 1.02 8.05E-01 0.55 1.09 1.83E-01 0.32 1.03 4.64E-01
rs4129767 17:76403984 HDL-c 1.05 7.88E-02 0.65 1.09 1.71E-01 0.44 1.13 5.10E-02 0.43 1.10 7.81E-03
rs7241918 18:47160953 LDL-c, TC 1.05 3.89E-01 0.05 1.09 5.26E-01 0.10 1.04 7.13E-01 0.08 1.04 5.80E-01
rs12967135 18:57849023 HDL-c 1.01 6.96E-01 0.27 1.19 8.79E-03 0.78 1.09 2.78E-01 0.88 1.05 4.10E-01
rs7248104 19:7224431 HDL-c, TC 1.01 7.15E-01 0.32 1.02 8.02E-01 0.34 1.01 8.17E-01 0.62 1.03 4.52E-01
rs7255436 19:8433196 Trig 1.04 2.21E-01 0.65 1.00 9.48E-01 0.83 1.11 2.22E-01 0.50 1.04 3.10E-01
rs6511720 19:11202306 Trig, 1.01 8.28E-01 0.86 1.14 1.42E-01 NA
2
NA
2
NA
2
0.91 1.05 4.23E-01
rs737337 19:11347493 HDL-c, Trig 1.06 7.96E-02 0.58 1.08 2.23E-01 0.73 1.05 5.09E-01 0.27 1.05 2.51E-01
rs10401969 19:19407718 LDL-c, Trig, TC 1.04 4.10E-01 0.83 1.09 2.89E-01 0.10 1.02 8.23E-01 0.05 1.04 6.39E-01
rs731839 19:33899065 HDL-c 1.02 4.24E-01 0.64 1.09 1.56E-01 0.51 1.12 6.25E-02 0.40 1.03 4.50E-01
rs4420638 19:45422946 LDL-c, HDL-c, TC 1.10 3.80E-02 0.80 1.15 8.17E-02 0.90 1.22 8.50E-02 0.10 1.04 5.83E-01
rs492602 19:49206417 TC 1.07 3.83E-02 0.50 1.11 7.82E-02 NA
2
NA
2
NA
2
0.32 1.05 1.86E-01
rs17695224 19:52324216 HDL-c 1.01 8.42E-01 0.16 1.03 6.74E-01 0.26 1.03 6.20E-01 0.67 1.03 4.76E-01
rs386000 19:54792761 HDL-c 1.04 1.69E-01 0.18 1.06 4.66E-01 0.19 1.05 5.06E-01 0.53 1.07 1.03E-01
rs364585 20:12962718 LDL-c 1.01 6.27E-01 0.17 1.03 7.51E-01 0.58 1.03 6.80E-01 0.66 1.03 4.67E-01
rs2328223 20:17845921 LDL-c 1.09 1.56E-02 0.82 1.13 1.21E-01 0.76 1.19 1.60E-02 0.24 1.04 4.02E-01
rs2277862 20:34152782 TC 1.06 1.34E-01 0.19 1.08 3.38E-01 0.11 1.08 4.13E-01 0.81 1.04 3.59E-01
177
rs2902940 20:39091487 LDL-c, TC 1.02 4.37E-01 0.53 1.10 1.25E-01 0.79 1.05 5.09E-01 0.29 1.01 7.37E-01
rs6029526 20:39672618 LDL-c, TC 1.03 3.53E-01 0.24 1.06 4.34E-01 0.15 1.28 3.64E-03 0.50 1.02 5.88E-01
rs1800961 20:43042364 HDL-c, TC 1.52 8.72E-07 0.99 1.08 9.00E-01 0.99 1.53 5.07E-01 0.96 1.56 3.08E-07
rs6065906 20:44554015 HDL-c, Trig 1.01 8.40E-01 0.17 1.02 8.36E-01 0.01 1.89 2.50E-01 0.11 1.00 9.81E-01
rs181362 22:21932068 HDL-c 1.03 3.23E-01 0.43 1.06 3.65E-01 0.52 1.05 4.44E-01 0.55 1.05 2.18E-01
rs5763662 22:30378703 LDL-c 1.01 7.95E-01 0.99 1.07 8.04E-01 0.06 1.03 8.04E-01 0.86 1.02 7.37E-01
rs138777 22:35711098 TC 1.01 8.10E-01 0.24 1.01 8.75E-01 0.51 1.15 1.79E-02 0.60 1.05 2.19E-01
rs5756931 22:38546033 Trig 1.03 3.05E-01 0.87 1.04 6.91E-01 0.83 1.16 9.17E-02 0.43 1.08 3.69E-02
rs4253772 22:46627603 LDL-c, TC 1.02 7.54E-01 0.97 1.71 1.07E-02 NA
2
NA
2
NA
2
0.94 1.05 5.47E-01
Lipid trait associated SNPs reported in Willer et al. 2013.
1
Lipid traits associated with risk variant.
2
Variant not tested in this analysis due to
monomorphic or rare risk allele.
178
Figure C.1. Quantile-quantile (QQ) plots of race/ethnic specific inverse weighted meta-analyses.
a) Multiethnic meta-analysis. λ GIF=1.011. Most of the extreme values are driven by Latino
populations (see Results).
179
b) African Americans in the MEC. λ GIF=0.999.
180
c) Japanese Americans in the MEC. λ GIF=0.996.
181
d) Latinos in the MEC and SIGMA studies. λ GIF=1.012. Most of the extreme values are driven by
SNPs from the ABCG5/8 region (see results).
Abstract (if available)
Abstract
Racial/ethnic disparities exist in incidence and mortality of colorectal cancer (CRC), cardiovascular diseases (CVD), stroke, and gallbladder disease (GBD). The prevalence of risk factors for these diseases also differ by race/ethnicity. Here we examined common chronic diseases that vary across racial/ethnic populations in the prospective Multiethnic Cohort (MEC) in Los Angeles and Hawaii to define risk factors that are important in each population which may contribute to differences in risk across Whites, African Americans (AA), Native Hawaiians (NH), Japanese Americans (JA), and Latinos. We used a prospective cohort analysis to examine the relationship between atopic allergic conditions (AAC) and CRC and found that individuals with AAC had a 14% decreased CRC risk across all races/ethnicities. In a mortality analysis of CVD and stroke risk, we found that known risk factors accounted for much of the difference in risk of mortality when compared to Whites, except for excess risk in AA women, NH men and women, and decreased risk in JA men and women. A genome-wide association study (GWAS) replicated a known risk locus for GBD in the gene ABCG5/8, and identified in a JA-specific locus (7p22) and two AA-specific loci (14q21.3 & 15q21.3) associated with increased risk of GBD. Both environmental and genetic risk factors contribute to varying risks of incidence and mortality of common chronic diseases across race/ethnicity.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Diet quality and pancreatic cancer incidence in the multiethnic cohort
PDF
Chronic eye disease epidemiology in the multiethnic ophthalmology cohorts of California study
PDF
Extending genome-wide association study methods in African American data
PDF
Identifying genetic, environmental, and lifestyle determinants of ethnic variation in risk of pancreatic cancer
PDF
Utility of polygenic risk score with biomarkers and lifestyle factors in the multiethnic cohort study
PDF
Examining the relationship between common genetic variation, type 2 diabetes and prostate cancer risk in the multiethnic cohort
PDF
Genomic risk factors associated with Ewing Sarcoma susceptibility
PDF
Genetic studies of cancer in populations of African ancestry and Latinos
PDF
Body size and the risk of prostate cancer in the multiethnic cohort
PDF
Genes and environment in prostate cancer risk and prognosis
PDF
An analysis of disease-free survival and overall survival in inflammatory breast cancer
PDF
Arm lymphedema in a multi-ethnic cohort of female breast cancer survivors
PDF
Effects of stress and the social environment on childhood asthma in the children' s health study
PDF
Predictive factors of breast cancer survival: a population-based study
PDF
Origins of the gender disparity in bladder cancer risk: a SEER analysis
PDF
Effects of post-menopausal hormone therapy on arterial stiffness in the ELITE trial
PDF
Risk factors of pelvic floor disorders in the multiethnic cohort study
PDF
Association of chronic obstructive pulmonary disease and mortality following thoracic and complex endovascular aortic repair: a retrospective cohort study
PDF
Disparities in gallbladder, intra-hepatic bile duct, and other biliary cancers among multi-ethnic populations: a California Cancer Registry study
PDF
Association of comorbidity with prostate cancer tumor characteristics in African American men
Asset Metadata
Creator
Tambe, Neal Atul
(author)
Core Title
The multiethnic nature of chronic disease: studies in the multiethnic cohort
School
Keck School of Medicine
Degree
Doctor of Philosophy
Degree Program
Epidemiology
Publication Date
07/05/2016
Defense Date
04/19/2016
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
allergies,asthma,atopy,cancer,cardiovascular disease,colorectal cancer,Epidemiology,gallbladder disease,genome-wide association study,mortality,multiethnic,OAI-PMH Harvest,stroke
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Haiman, Christopher (
committee chair
), Coetzee, Gerhard (
committee member
), Cozen, Wendy (
committee member
), Setiawan, Veronica Wendy (
committee member
), Stram, Daniel (
committee member
)
Creator Email
neal.tambe@gmail.com,ntambe@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-260611
Unique identifier
UC11281103
Identifier
etd-TambeNealA-4503.pdf (filename),usctheses-c40-260611 (legacy record id)
Legacy Identifier
etd-TambeNealA-4503.pdf
Dmrecord
260611
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Tambe, Neal Atul
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
allergies
atopy
cardiovascular disease
colorectal cancer
gallbladder disease
genome-wide association study
mortality
multiethnic
stroke