A meta-analysis of candidate gene polymorphisms and ischemic stroke in 6 study populations: association of lymphotoxin-alpha in nonhypertensive patients.

Abstract

Ischemic stroke is a multifactorial disease with a strong genetic component. Pathways, including lipid metabolism, systemic chronic inflammation, coagulation, blood pressure regulation, and cellular adhesion, have been implicated in stroke pathophysiology, and candidate gene polymorphisms in these pathways have been proposed as genetic risk factors.
We genotyped 105 simple deletions and single nucleotide polymorphisms from 64 candidate genes in 3550 patients and 6560 control subjects from 6 case-control association studies conducted in the United States, Europe, and China. Genotyping was performed using the same immobilized probe typing system and meta-analyses were based on summary logistic regressions for each study. The primary analyses were fixed-effects meta-analyses adjusting for age and sex with additive, dominant, and recessive models of inheritance.
Although 7 polymorphisms showed a nominal additive association, none remained statistically significant after adjustment for multiple comparisons. In contrast, after stratification for hypertension, 2 lymphotoxin-alpha polymorphisms, which are in strong linkage disequilibrium, were significantly associated among nonhypertensive individuals: LTA 252A>G (additive model; OR, 1.41 with 95% CI, 1.20 to 1.65; P=0.00002) and LTA 26Thr>Asn (OR, 1.19 with 95% CI, 1.06 to 1.34; P=0.003). LTA 252A>G remained significant after adjustment for multiple testing using either the false discovery rate or by permutation testing. The 2 single nucleotide polymorphisms showed no association in hypertensive subjects (eg, LTA 252A>G, OR, 0.93; 95% CI, 0.84 to 1.03; P=0.17).
These observations may indicate an important role of LTA-mediated inflammatory processes in the pathogenesis of ischemic stroke.

Full text

Page 1

ISSN: 1524-4628
Copyright © 2009 American Heart Association. All rights reserved. Print ISSN: 0039-2499. Online
Stroke is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX 72514
DOI: 10.1161/STROKEAHA.108.524587
published online Jan 8, 2009;
Stroke
Consortium
Liu, Paul M. Ridker, Robert Y.L. Zee, Nancy R. Cook and for the RMS Stroke SNP
Bernd Ringelstein, Christof Kessler, Jan Luedemann, Klaus Lindpaintner, Lisheng
Mannhalter, Klaus Berger, Wolfgang Lalouschek, Warren S. Browner, Yu Shi, E.
Xingyu Wang, Suzanne Cheng, Victoria H. Brophy, Henry A. Erlich, Christine

Patients
Study Populations. Association of Lymphotoxin-Alpha in Nonhypertensive
A Meta-Analysis of Candidate Gene Polymorphisms and Ischemic Stroke in 6
http://stroke.ahajournals.org
located on the World Wide Web at:
The online version of this article, along with updated information and services, is
http://www.lww.com/reprints
Reprints: Information about reprints can be found online at

journalpermissions@lww.com
410-528-8550. E-mail:
Fax:Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050.
Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters

http://stroke.ahajournals.org/subscriptions/
Subscriptions: Information about subscribing to Stroke is online at
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 2

A Meta-Analysis of Candidate Gene Polymorphisms and
Ischemic Stroke in 6 Study Populations
Association of Lymphotoxin-Alpha in Nonhypertensive Patients
Xingyu Wang, PhD; Suzanne Cheng, PhD; Victoria H. Brophy, PhD; Henry A. Erlich, PhD;
Christine Mannhalter, PhD; Klaus Berger, MD; Wolfgang Lalouschek, MD; Warren S. Browner, MD;
Yu Shi, MS; E. Bernd Ringelstein, MD; Christof Kessler, MD; Jan Luedemann, PhD;
Klaus Lindpaintner, MD; Lisheng Liu, MD; Paul M. Ridker, MD; Robert Y.L. Zee, PhD;
Nancy R. Cook, ScD; for the RMS Stroke SNP Consortium
Background and Purpose—Ischemic stroke is a multifactorial disease with a strong genetic component. Pathways,
including lipid metabolism, systemic chronic inflammation, coagulation, blood pressure regulation, and cellular
adhesion, have been implicated in stroke pathophysiology, and candidate gene polymorphisms in these pathways have
been proposed as genetic risk factors.
Methods—We genotyped 105 simple deletions and single nucleotide polymorphisms from 64 candidate genes in 3550
patients and 6560 control subjects from 6 case–control association studies conducted in the United States, Europe, and
China. Genotyping was performed using the same immobilized probe typing system and meta-analyses were based on
summary logistic regressions for each study. The primary analyses were fixed-effects meta-analyses adjusting for age
and sex with additive, dominant, and recessive models of inheritance.
Results—Although 7 polymorphisms showed a nominal additive association, none remained statistically significant after
adjustment for multiple comparisons. In contrast, after stratification for hypertension, 2 lymphotoxin-alpha poly-
morphisms, which are in strong linkage disequilibrium, were significantly associated among nonhypertensive
individuals: LTA 252A?G (additive model; OR, 1.41 with 95% CI, 1.20 to 1.65; P?0.00002) and LTA 26Thr?Asn
(OR, 1.19 with 95% CI, 1.06 to 1.34; P?0.003). LTA 252A?G remained significant after adjustment for multiple testing
using either the false discovery rate or by permutation testing. The 2 single nucleotide polymorphisms showed no
association in hypertensive subjects (eg, LTA 252A?G, OR, 0.93; 95% CI, 0.84 to 1.03; P?0.17).
Conclusions—These observations may indicate an important role of LTA-mediated inflammatory processes in the
pathogenesis of ischemic stroke. (Stroke. 2009;40:00-00.)
Key Words: embolic stroke ? genetics ? hypertension ? inflammation
I
ual’s genetic background and various environmental compo-
nents. Previous studies have established hypertension, smok-
ing, diabetes mellitus, body mass index, and age as reliable
stroke risk predictors.1,2However, these conventional stroke
risk factors do not fully account for the overall risk of stroke.
Several physiological pathways, including lipid metabolism,
blood pressure regulation, coagulation, and cellular adhesion,
are thought to play critical roles in stroke pathophysiology.
schemic stroke is a complex multifactorial and polygenic
disorder thought to reflect interactions between an individ-
Among the known risk factors for ischemic stroke, hyper-
tension contributes significantly to the onset of disease.
Increased risk of stroke is not, however, limited to those with
hypertension and the conventional stroke risk factors do not
fully explain the risk among normotensives. Strategies to
identify additional risk factors include stratification by hy-
pertension3,4and using blood pressure as a matching criterion
for cases and control subjects.5A key role for inflammation
is suggested by observations that patients with hypertension
have elevated circulating levels of markers of inflammation
Received April 30, 2008; final revision received July 23, 2008; accepted July 25, 2008.
From the Laboratory of Human Genetics (X.W., Y.S., L.L.), Beijing Hypertension League Institute, Beijing, China.; the Department of Human Genetics
(S.C., V.H.B., H.A.E.), Roche Molecular Systems, Inc., Pleasanton, Calif; the Department of Medical and Chemical Laboratory Diagnostics (C.M.) and
the University Clinic of Neurology (W.L.), Medical University Vienna, Vienna, Austria; the Institute of Epidemiology and Social Medicine (K.B.) and
Department of Neurology (E.B.R.), University of Muenster, Muenster, Germany; S.F. Coordinating Center (W.S.B.), California Pacific Medical Center,
Research Institute, San Francisco, Calif; the Department of Neurology (C.K.) and the Institute of Clinical Chemistry and Laboratory Medicine (J.L.), Ernst
Moritz Arndt University, Greifswald, Germany; Roche Center for Medical Genomics (K.L.), F. Hoffmann-La Roche, Ltd, Basel, Switzerland; and the
Center for Cardiovascular Disease Prevention and Division of Preventive Medicine (P.M.R., R.Y.L.Z., N.R.C.), Brigham and Women’s Hospital, Boston,
Mass.
Correspondence to Nancy R. Cook, ScD, Division of Preventive Medicine, Department of Medicine, Brigham and Women’s Hospital, 900
Commonwealth Avenue East, Boston, MA 02215. E-mail ncook@rics.bwh.harvard.edu
© 2009 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.orgDOI: 10.1161/STROKEAHA.108.524587
1
by on May 10, 2011 stroke.ahajournals.org Downloaded from

Page 3

and that some antihypertensive therapies reduce both levels
of proinflammatory markers and the risk of ischemic stroke in
addition to lowering blood pressure.6
Both systemic and local inflammatory processes are impli-
cated in the etiology of ischemic cerebrovascular disease and
in the pathophysiology of cerebral ischemia.7Viral and
bacterial infections are independent risk factors for ischemic
stroke8and increased levels of systemic inflammatory mark-
ers such as C-reactive protein, leukocyte count, and fibrino-
gen are associated with increased risk of ischemic stroke.9
Moreover, many stroke-related diseases such as Alzheimer
disease and atherosclerosis are initiated or worsened by
systemic inflammation.10,11Polymorphisms in the C-reactive
protein gene have been recently associated with both circu-
lating protein levels and cardiovascular events,12demonstrat-
ing the potential impact of genetic variation. Proinflammatory
cytokines are believed to play a pathogenic role in these
diseases, and variations in cytokine genes have also been
shown to influence both predisposition and penetrance by
altering the transcription profile and pattern of proinflamma-
tory cytokine production.13For example, polymorphism in
the lymphotoxin-alpha gene can enhance transcription and
susceptibility to myocardial infarction.14At the local level,
migration of inflammatory cells to the vascular wall is
associated with vascular changes leading to atherosclerosis,
and early atherosclerotic lesions are preceded by inflamma-
tory cell deposition in the subendothelial layer of major
cerebral arteries and in small brain vessels.15Genetic variants
influencing inflammatory processes could potentially contrib-
ute to the etiology of stroke.
The complex etiology of stroke suggests that individual
genetic polymorphisms have modest effects that are difficult
to detect, as has been observed to date.16Large studies are
needed to assess these polymorphisms as risk factors. We
report a 6-study meta-analysis to investigate the associations
of 105 simple deletions and single nucleotide polymorphisms
(SNPs) in inflammatory and cardiovascular system-related
genes with susceptibility to ischemic stroke. To search for
genetic risk factors contributing to ischemic stroke beyond
hypertension, we stratified the study cohort on hypertension
status.
Materials and Methods
Study Sample Description
As part of the Roche Stroke SNP Consortium, results of 6 indepen-
dent studies (Table 1) were pooled for this analysis. All 6 study
samples were comprised of individuals with proven ischemic stroke
status and healthy control subjects. All studies were approved by the
local ethics committees and all participants gave informed consent.
Briefly, the study subjects were recruited as follows:
Physician’s Health Study
A nested case–control sample (319 cases, 2092 control subjects) was
derived from the Physician’s Health Study (PHS) cohort consisting
of 22 071 predominantly white US male physicians initially free of
prior myocardial infarction, stroke, transient ischemic attack, and
cancer who were enrolled in a placebo-controlled trial of aspirin and
?-carotene for the primary prevention of cardiovascular disease
and cancer.17DNA was isolated from baseline blood samples
provided by 14 916 (68%) of the participants. Incident cases of
ischemic stroke were identified during an average 13-year follow-up
and confirmed by medical record review. Control subjects were
selected from study participants remaining free of reported cardio-
vascular disease and matched to cases of any cardiovascular disease
by age, smoking, and time since study entry.18
Study of Osteoporotic Fractures
Ambulatory women were recruited from 4 clinical centers in
Portland, Ore; Minneapolis, Minn; Baltimore, Md; and the Monon-
gahela Valley, Pa.19The Study of Osteoporotic Fractures (SOF)
cohort consists of 9615 white women of at least 65 years of age who
had not had bilateral hip replacement or earlier hip fracture at the
time of recruitment. The stroke subgroup included here consists of
247 who had adjudicated ischemic strokes and 559 control subjects
who remained free of stroke through the mean follow-up of 5.4
years. Individuals who died during follow-up were included in both
cases and control subjects, avoiding survivor bias.
Westphalia, Germany
Cases (n?700) were recruited through the regional Westphalian
Stroke Register in northwestern Germany.20Standardized patient
documentation included sociodemographic characteristics, comor-
bidities, stroke type and severity as well as details regarding the
diagnostic and therapeutic procedures and complications; 96.8% had
at least one CT or MRI of the brain during hospitalization. Control
subjects (n?757) were recruited from the population-based Dort-
mund Health Study, conducted in the same region.21Participants in
this study were randomly drawn from the city’s registration office
within 5-year age groups and stratified by sex. Medical histories
were assessed in face-to-face interviews.
Pomerania, Germany
Cases (n?277) were recruited with a standardized patient assessment
form; 96.5% had at least one CT or MRI during hospitalization.
Control subjects for this region were recruited from the population-
based Study of Health in Pomerania.22Participants in Study of
Health in Pomerania were 20 to 79 year olds randomly sampled from
registration offices in the area. Face-to-face interviews with each
participant included a short stroke symptom questionnaire. A random
sample of 702 Study of Health in Pomerania participants who were
free of self-reported stroke and within the same age range and sex
distribution as the cases formed the control group.
Vienna Stroke Study
In the Vienna Stroke Registry, cases (n?844) consisted of consec-
utive white patients submitted to one of 9 stroke units within 72
hours of symptom onset of acute ischemic stroke. Patients who died
on the way to the hospital or were first admitted to an intensive care
unit were not included.23All patients underwent cranial CT or MRI
and were documented according to a standardized protocol, includ-
ing stroke severity, risk factors, and medical history (with particular
reference to vascular diseases). Control subjects (n?979) were
voluntary participants in a health care program offered by the city of
Vienna, were free of clinically manifest arterial vascular disease, and
reported no arterial vascular diseases in first-degree relatives.
Stroke Hypertension Investigation in Genetics
Individuals were recruited from 6 geographical regions within China;
70% came from in and near Beijing. Cases (n?1163) were individ-
uals who had a stroke within the previous 5 years as diagnosed by
brain CT/MRI. The original goal was to identify SNPs that predis-
pose to stroke independent of blood pressure; thus, randomly drawn
population-based control subjects were initially individually
matched to cases by sex, birth year ?3 years, geographic location,
and blood pressure category (?140/90, ?140/90 and ?180/105,
?180/105 mm Hg). Because some cases could not be matched,
additional control subjects were recruited for a total of 1471 control
subjects.5
Genotyping
A total of 105 polymorphisms from 64 genes were selected based on
reported associations in the literature as well as on evidence of gene
product involvement in cardiovascular disease and inflammatory
processes. As previously described,24,253 separate multilocus poly-
2 Stroke
March 2009
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 4

merase chain reactions were carried out using biotinylated primer
pools (Roche Molecular Systems, Inc). The resulting polymerase
chain reaction product pools were denatured and hybridized to linear
arrays of immobilized, sequence-specific oligonucleotide probes.
Hybridized amplicon was detected using a streptavidin–horseradish
peroxidase conjugate and a chromogenic substrate. Laboratory tech-
nicians were blinded to the case–control status of each sample.
Genotype assignments were made either manually and independently
by 2 researchers or made by capturing images with a flatbed scanner
and then using proprietary software developed by Roche Molecular
Systems to resolve probe signals into genotypes for all poly-
morphisms. Discordant or ambiguous results were resolved by repeat
polymerase chain reaction or hybridization. Twenty polymorphisms
were not available for PHS and 23 polymorphisms were genotyped in
only a subset of the Vienna Stroke participants; for these, genotypes
were obtained for ?90.5% of the individuals typed. For each of the 62
polymorphisms genotyped for all 10 110 subjects across all 6 studies,
the final genotype database was ?97.5% complete. For the LTA [MIM
153440] 252A?G and LTA 26Thr?Asn polymorphisms, in particular,
the database contained 6090 (not available for PHS) and 10 091
genotypes (99.8% complete), respectively.
Statistical Analysis
Individual-level data were provided from each study site to the
Coordinating Center at the Brigham and Women’s Hospital in
Boston. Prespecified inclusion criteria for the meta-analyses were
age at least 20 years and no history of myocardial infarction. Cases
were restricted to those experiencing an ischemic stroke and control
subjects had no history of stroke.
Allele and genotype frequencies were estimated by study site
among cases and control subjects separately using SAS GENETICS.
Tests for Hardy-Weinberg equilibrium (HWE), both large-sample
and exact, were conducted among cases and control subjects for each
site(supplementalTableI,availableonlineathttp://stroke.ahajournals.
org). Results for the effect of each SNP on ischemic stroke were
estimated for each study separately using logistic regression. Each
analysis controlled for age and sex and assessed genetic effects under
3 modes of inheritance: additive, dominant, and recessive. In
addition, analyses were conducted for each site using all 3 genotypes
using a 2-degree of freedom test.
Meta-analyses were conducted based on the summary logistic
regression results for each study site.26The primary analyses were
fixed-effects meta-analyses adjusting for age and sex. These meta-
Table 1. Characteristics of the Study Subjects*
PHS17,18
SOF19
Vienna23
No. of subjects
Study type
Recruitment period
319 cases, 2092 control subjects
prospective
1982; follow-up through 1995
247 cases, 559 control subjects
prospective
1986–1988; follow-up through
1998
Whites from US
Cases
73.9?5.9
0
79.8
19.3
29.8
26.9?4.6
151.9?23.1
844 cases, 979 control subjects
Cross-sectional
1998–2001
PopulationWhites from US
Cases
61.1?8.3
100
48.6
10.7
60.7
25.6?3.3
133.6?12.8
Whites from Vienna, Austria
Cases
66.1?14.4
50.6
72.2
32.8
54.9
26.7?4.6
NA
Control Subjects
58.8?8.5
100
26.8
2.7
56.6
25.0?3.0
127.4?11.9
P‡ Control Subjects
70.4?4.5
0
62.1
4.3
37.5
26.8?4.9
140.4?17.5
P‡ Control Subjects
48.8?13.0
53.4
34.2
2.5
41.7
25.3?3.9
132.3?20.7
P‡
Age,† years
Sex, % male
Hypertension, % yes
Diabetes, % yes
Smoking, % ever
Body mass index,† kg/m2
Systolic blood
pressure,† mm Hg
Diastolic blood
pressure,† mm Hg
?0.0001
?0.0001
?0.0001
0.23
?0.0001
?0.0001
?0.0001
?0.0001
?0.0001
?0.0001
0.15
0.002
?0.0001
?0.0001
?0.0001
0.15
0.83
?0.0001
82.1?7.079.0?7.2
?0.000178.2?11.676.9?8.90.13 NA82.7?11.7
Westphalia20,21
Pomerania22
SHINING5
No. of subjects
Study type
Recruitment period
Population
700 cases, 757 control subjects
Cross-sectional
2000–2003
Whites from northwestern
Germany
Cases
61.0?16.9
56.3
68.0
21
NA
NA
NA
277 cases, 702 control subjects
Cross-sectional
1996–2001
Whites from northeastern
Germany
Cases
62.8?13.2
55.1
66.8
30.1
87.6
NA
NA
1163 cases, 1471 control subjects
Cross-sectional
1997–2000
Chinese primarily from Beijing,
China
Cases§ Control Subjects§
59.3?10.7
60.3
69.6
15.3
39.7
24.4?3.0
145.4?23.2
Control Subjects
58.7?9.9
46.6
41.0
8.9
54.1
NA
145.9?21.4
P‡
0.002
?0.0001
?0.0001
?0.0001
Control Subjects
62.7?11.8
50.8
52.9
14.5
58.3
NA
148.7?21.3
P‡
0.88
0.23
?0.0001
?0.0001
?0.0001
P‡
Age,† years
Sex, % male
Hypertension, % yes
Diabetes, % yes
Smoking, % ever
Body mass index,† kg/m2
Systolic blood
pressure,† mm Hg
Diastolic blood
pressure,† mm Hg
61.1?10.7
60.1
77.1
8.5
32.3
25.0?3.3
143.2?23.7
?0.0001
0.93
?0.0001
?0.0001
?0.0001
?0.0001
0.018
NA89.1?12.2NA 86.1?11.1 87.0?12.8 86.1?13.0 0.087
*Summary of population characteristics for each of the 6 studies used in the meta-analysis. All cases were diagnosed by CT/MRI. Blood pressure was assessed
before stroke.
†Continuous variables are given as mean?SD.
‡Using ?2test for categorical variables, t test for continuous variables.
§Partially matched by age and blood pressure group during recruitment.
NA indicates not available.
Wang et al LTA and Stroke in Nonhypertensives
3
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 5

analyses were also conducted across whites only (data not shown),
because these comprised the majority of participants for 5 of the 6
studies. Effects for each of the 3 modes of inheritance were
estimated. PROC MIXED of SAS was used for effect estimation.
Tests for heterogeneity of the genetic effect across sites were
conducted using the Q-statistic.27For comparison, random-effects
models were estimated that allowed the genetic effect to vary across
sites using study-specific effect estimates and PROC MIXED of
SAS. To adjust for multiple comparisons, the false discovery rate
(FDR)28was computed and stepdown permutation tests were con-
ducted for selected comparisons.29
Other prespecified analyses adjusted for hypertension as well as
age and sex. Across all studies, hypertensives were defined as having
current or past antihypertensive medication, systolic blood pressure
?140 mm Hg, or diastolic blood pressure ?90 mm Hg. Additional
smoking-adjusted analyses were limited to 5 studies due to the
limited availability of smoking data for the Westphalian study
participants. Subgroup analyses were conducted according to age,
sex, presence of hypertension, or smoking (ever versus never).
Results
A total of 3550 patients with stroke and 6560 control subjects
were genotyped with inflammation and cardiovascular SNP
panels in 6 study sites with common methodology and
genotyping software. The characteristics of all participants
from 6 study sites are listed in Table 1. Two studies were
drawn from prospective cohorts; the PHS study followed only
male subjects and the SOF study followed only female
subjects. The Stroke Hypertension Investigation in Genetics
(SHINING) study was comprised of subjects of Han ethnic-
ity, whereas the 5 other study populations were ?99% white.
There was a greater proportion of hypertension among cases
than in control subjects, except in the SHINING study subset,
for which blood pressure had been matched between the
majority of cases and control subjects.
Table 2 lists the results obtained in the primary fixed-
effects meta-analysis for all polymorphic sites under domi-
nant, additive, and recessive genetic modes of inheritance;
similar results were observed under a random-effects meta-
analysis (data not shown). Nine SNPs were nominally signif-
icant (P?0.05) under at least one mode of inheritance:
ADRB3 [MIM 109691] Trp64Arg, CETP [MIM 118470]
(?629)C?A, GNB3 [MIM 139130] 825C?T, IL4 [MIM
147780] (?590)C?T, LIPC [MIM 151670] (?480)C?T,
LPL
[MIM609708] Ser447Ter,
163729](?690)C?T, PON2 [MIM 602447] Ser311Cys, and
TGFB1 [MIM 190180] (?509)C?T. To account for multiple
hypothesis testing, the FDR or permutation testing was
applied and none of these SNPs remained statistically signif-
icantly associated with ischemic stroke. Among white partic-
ipants only, the same GNB3, LPL, NOS3, PON2, and TGFB1
SNPs were nominally significant under at least one mode of
inheritance in addition to 8 others (APOB [MIM 107730]
71Ile?Thr, APOC3 [MIM 107720] 3175C?G, CCR5 [MIM
601373] (?2459)G?A, IL6 [MIM 147620] (?174)G?C,
IL10 [MIM 124092] (?571)C?A, ITGA3 [MIM 192974]
873G?A, NOS2A [MIM 163730] 231C?T, TNF [MIM
191160] (?376)G?A), but none of the SNPs remained
statistically significant after the FDR was applied (data not
shown).
The data were then stratified on age, sex, hypertension, or
smoking status. No statistically significant associations were
NOS3
[MIM
observed in the age- (Supplemental Table IIA) or sex-
stratified (Supplemental Table IIB) analyses nor among those
with current or past hypertension (Table 3) after adjustment
with the FDR. In contrast, a large number of nominally
significant associations with ischemic stroke among normo-
tensives were observed (Table 4). The strongest associations
under the additive and dominant models were for LTA
252A?G and LTA 26Thr?Asn, 2 SNPs in strong linkage
disequilibrium, whereas NOS3 298Glu?Asp had the stron-
gest association under the recessive model. After adjustment
with FDR and permutation testing, only the LTA 252A?G
SNP showed a significant association among those without
hypertension. In the additive mode, the estimated relative risk
across the 3 LTA 252 genotypes was 1.41 (P?0.00002) in the
fixed-effects analysis and the FDR was 0.002 with P?0.01 in
permutation testing. Results for the dominant model were
similar (OR, 1.57; FDR, 0.005). In the random-effects meta-
analysis (data not shown), the LTA 252A?G association with
stroke under the dominant model (OR, 1.56) had an FDR of
0.02. Among whites only, LTA 252A?G was similarly
associated with ischemic stroke among those without hyper-
tension under additive and dominant models (OR, 1.28;
P?0.016 and OR, 1.39; P?0.019, respectively). Minor allele
frequencies among nonhypertensive control subjects are
given in Table 4; frequencies among nonhypertensive cases
were 0.37, 0.38, 0.31, 0.41, and 0.46 in SOF, Vienna,
Westphalia, Pomerania, and SHINING, respectively.
The point estimates for the OR were somewhat higher for
LTA 252A?G, a polymorphism in intron 1, than for the
nonsynonymous polymorphism LTA 26Thr?Asn, although
the CIs overlapped after adjusting for age and sex (Figure
A–D). The FDR values for the Thr?Asn polymorphism were
also ?0.05. The associations with stroke risk for both LTA
SNPs reached statistical significance among normotensives
within the individual studies of SHINING and Pomerania,
whereas among hypertensives, the OR point estimates were
usually just below 1 and were not statistically significant
(Figure). This trend for increased stroke risk associated with
the LTA SNPs among normotensives relative to hypertensives
was observed across the other studies, although none of these
individual associations was statistically significant. We note
that the PHS cohort was genotyped only for the LTA
26Thr?Asn polymorphism under the expectation that this
coding SNP could be functional and would be an effective
“tag” for LTA 252 based on the very strong linkage disequi-
librium between these 2 polymorphisms; furthermore, in the
Vienna and Westphalia studies, some samples had missing
genotypes for LTA 252A?G. When the meta-analysis was
repeated with only those samples that had been genotyped for
both LTA SNPs, the OR estimates were virtually identical
(1.572 and 1.565 among normotensives under the dominant
model for LTA 252 and LTA 26, respectively; data not
shown). In addition, if the LTA 26 result for the PHS was
imputed for the missing LTA 252 data, the additive result for
LTA 252 would remain highly significant (OR, 1.30;
P?0.0001). Alternatively, if a completely null estimate for
the PHS was imputed, the overall result would remain
significant (OR, 1.27; P?0.0004) and would continue to pass
the stringent multiple comparisons testing.
4 Stroke
March 2009
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 6

Table 2. Unstratified Meta-analysis for Risk of Ischemic Stroke: All SNPs Under 3 Modes of Inheritance
AdditiveDominantRecessive
Gene ?MIM No.?
SNP rs No. SNP Namen* Sites†OR (LCL–UCL)‡P‡ Het P§ OR (LCL–UCL)‡P‡ Het P§Sites¶ OR (LCL–UCL)‡P‡ Het P§
Adducin 1 (alpha) ?102680?
rs4961ADD1 460Gly?Trp 99366 1.02 (0.95–1.10)0.6380.9601.04 (0.95–1.15)0.403 0.8766 0.98 (0.84–1.14)0.7620.692
Adrenergic, beta-2-,
receptor, surface ?109690?
rs1042713ADRB2 16Gly?Arg 10 08260.94 (0.89–1.01) 0.0800.2710.95 (0.86–1.05)0.303 0.43760.90 (0.80–1.01)0.064 0.502
Adrenergic, beta-2-,
receptor, surface ?109690?
rs1042714ADRB2 27Gln?Glu10 08761.04 (0.97–1.11)0.3230.121 1.04 (0.94–1.15)0.4610.17761.06 (0.93–1.22)0.381 0.222
Adrenergic, beta-2-,
receptor, surface ?109690?
rs1800888 ADRB2 164Thr?Ile 853561.33 (0.88–2.02)0.175 0.308 1.29 (0.85–1.98)0.236 0.324 Too few variant homozygotes observed
Adrenergic, beta-3-,
receptor ?109691?
rs4994 ADRB3 64Trp?Arg10 0936 0.91 (0.82–1.02) 0.090 0.6790.89 (0.79–1.00)0.0460.8036 1.13 (0.75–1.69) 0.5700.265
Angiotensin II receptor type
1 ?106165?
rs5186AGTR1 1166A?C 99426 1.04 (0.96–1.13)0.3460.0681.03 (0.93–1.14) 0.5460.0156 1.11 (0.92–1.35)0.2840.734
Angiotensinogen ?106150?
rs699 AGT 235Met?Thr99396 1.00 (0.94–1.07) 0.9730.371 0.97 (0.87–1.08) 0.586 0.2796 1.03 (0.93–1.15)0.5810.344
Angiotensin-converting
enzyme ?106180?
rs1799752ACE IVS16 Del?Ins986260.99 (0.93–1.06) 0.8410.069 1.02 (0.91–1.13)0.758 0.47360.97 (0.88–1.07)0.556 0.029
Apolipoprotein A-IV
?107690?
rs675APOA4 347Thr?Ser10 0916 0.94 (0.85–1.03)0.1930.672 0.96 (0.86–1.07)0.4460.55150.76 (0.57–1.03) 0.0740.512
Apolipoprotein A-IV
?107690?
rs5110 APOA4 360Gln?His 10 08860.95 (0.82–1.11)0.545 0.297 0.97 (0.83–1.14)0.7300.2885 0.81 (0.32–2.03)0.6530.999
Apolipoprotein B ?107730?
rs1367117APOB 71Thr?Ile 10 09161.04 (0.97–1.12) 0.2590.807 1.02 (0.93–1.12)0.6750.4166 1.18 (1.00–1.41)0.056 0.849
Apolipoprotein B ?107730?
rs5742904 APOB 3500Arg?Gln 10 0933 6.11 (0.45–82.6)0.173 0.999 6.11 (0.45–82.6)0.173 0.999 No variant homozygotes observed
Apolipoprotein C-III
?107720?
rs2542052APOC3 (?641)C?A10 0646 1.03 (0.97–1.10)0.303 0.8641.04 (0.95–1.15)0.377 0.5956 1.05 (0.93–1.18)0.4340.834
Apolipoprotein C-III
?107720?
rs2854117 APOC3 (?482)C?T10 08961.06 (0.99–1.13)0.0960.2661.06 (0.97–1.16)0.226 0.1376 1.12 (0.97–1.28)0.1140.728
Apolipoprotein C-III
?107720?
rs2854116APOC3 (?455)T?C 10 0876 1.04 (0.97–1.11)0.2440.7751.04 (0.95–1.15) 0.3720.5446 1.06 (0.94–1.20) 0.311 0.871
Apolipoprotein C-III
?107720?
rs4520APOC3 1100C?T10 0896 1.02 (0.96–1.09)0.5170.313 1.03 (0.93–1.13)0.609 0.2446 1.04 (0.92–1.18)0.5620.111
Apolipoprotein C-III
?107720?
rs5128 APOC3 3175C?G 10 0916 1.01 (0.93–1.10)0.7760.1241.05 (0.95–1.16)0.3510.1536 0.84 (0.66–1.08)0.173 0.389
Apolipoprotein C-III
?107720?
rs4225 APOC3 3206T?G 10 07361.04 (0.97–1.11) 0.2740.3721.09 (0.97–1.21) 0.1370.60661.02 (0.91–1.13)0.7800.405
Apolipoprotein E ?107741?
rs429358APOE 112Cys?Arg 10 05061.04 (0.94–1.14) 0.4470.341 1.03 (0.92–1.14) 0.6240.4696 1.25 (0.88–1.78)0.206 0.223
Apolipoprotein E ?107741?
rs7412APOE 158Arg?Cys 10 05560.96 (0.85–1.07) 0.424 0.1710.97 (0.86–1.09) 0.570 0.22160.71 (0.41–1.23)0.223 0.531
CD14 molecule ?158120?
rs2569190CD14 (?260)C?T853661.03 (0.96–1.11)0.3970.4741.02 (0.90–1.14)0.8010.27461.06 (0.95–1.19) 0.2850.952
Chemokine (C-C motif)
ligand 11 ?601156?
rs4795895CCL11 (?1328)G?A61235 0.96 (0.85–1.08) 0.4760.2020.97 (0.84–1.11)0.6020.08350.88 (0.58–1.33)0.536 0.707
Chemokine (C-C motif)
ligand 11 ?601156?
rs3744508 CCL11 23Ala?Thr 853661.08 (0.99–1.19) 0.0990.856 1.09 (0.98–1.21)0.113 0.9856 1.16 (0.86–1.57)0.3410.224
Chemokine (C-X-C motif)
ligand 12 (stromal
cell-derived factor 1)
?600835?
rs1801157 CXCL12 (?800)G?A603250.97 (0.88–1.06)0.4540.5990.96 (0.86–1.07) 0.4460.74850.99 (0.76–1.27) 0.9110.079
Chemokine receptor 2
?601267?
rs1799864CCR2 62Val?Ile 853261.08 (0.98–1.19)0.118 0.6611.07 (0.96–1.20)0.241 0.6706 1.28 (0.96–1.69)0.0900.717
Chemokine receptor 3
?601268?
rs5742906CCR3 39Pro?Leu85425 0.86 (0.37–2.02) 0.730 0.5650.86 (0.37–2.02)0.7300.565 No variant homozygotes observed
Chemokine receptor 5
?601373?
rs1799987 CCR5 (?2459)A?G85416 1.01 (0.94–1.08)0.8500.0401.00 (0.89–1.12)0.9600.0456 1.02 (0.91–1.15)0.7010.095
Chemokine receptor 5
?601373?
rs333CCR5 580Ins?Del3285235 0.94 (0.81–1.08)0.372 0.2640.94 (0.81–1.10)0.4610.17450.82 (0.44–1.51) 0.5170.706
Cholesteryl ester transfer
protein, plasma ?118470?
rs1800776 CETP (?631)C?A 10 08861.00 (0.87–1.15)0.9770.6161.02 (0.88–1.19)0.7920.71850.92 (0.46–1.83)0.8000.565
Cholesteryl ester transfer
protein, plasma ?118470?
rs1800775CETP (?629)C?A10 0766 0.95 (0.90–1.01) 0.1290.203 0.90 (0.81–1.00)0.0460.59560.98 (0.88–1.08) 0.653 0.134
Cholesteryl ester transfer
protein, plasma ?118470?
rs5882 CETP 405Ile?Val10 08660.99 (0.92–1.05)0.6610.748 0.99 (0.91–1.09) 0.8950.63560.96 (0.84–1.09)0.4980.791
Cholesteryl ester transfer
protein, plasma ?118470?
rs2303790CETP 442Asp?Gly10 08951.28 (0.87–1.89) 0.2071.0001.41 (0.93–2.13)0.1081.000 Too few variant homozygotes observed
Coagulation factor
II (thrombin) ?176930?
rs1799963F2 20210G?A 990251.01 (0.74–1.39)0.9410.1081.02 (0.74–1.40)0.9280.099Too few variant homozygotes observed
(Continued)
Wang et alLTA and Stroke in Nonhypertensives
5
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 7

Table 2.Continued
AdditiveDominantRecessive
Gene ?MIM No.?
SNP rs No.SNP Name n*Sites†OR (LCL–UCL)‡P‡ Het P§OR (LCL–UCL)‡P‡ Het P§ Sites¶ OR (LCL–UCL)‡P‡ Het P§
Coagulation factor
V (proaccelerin, labile
factor) ?227400?
rs6025F5 506Arg?Gln 99416 0.96 (0.77–1.20) 0.7120.8670.96 (0.76–1.20) 0.6980.848 Too few variant homozygotes observed
Coagulation factor
VII (serum prothrombin
conversion accelerator)
?227500?
rs5742910F7 (?323) Del?Ins1099416 0.99 (0.89–1.11)0.909 0.0571.00 (0.89–1.12)0.984 0.0726 1.05 (0.67–1.64)0.8320.908
Coagulation factor VII
(serum prothrombin
conversion accelerator)
?227500?
rs6046F7 353Arg?Gln 993860.96 (0.86–1.08) 0.5100.193 0.97 (0.86–1.10)0.638 0.16360.83 (0.53–1.31)0.427 0.736
Complement component 3
?120700?
rs2230199C3 102Arg?Gly6083 50.99 (0.87–1.13) 0.8910.4880.96 (0.83–1.12)0.609 0.42951.18 (0.80–1.76) 0.4080.798
Complement component 5
?120900?
rs17611C5 802Val?Ile 61175 1.02 (0.95–1.10) 0.6040.0871.07 (0.94–1.21) 0.328 0.1625 0.99 (0.88–1.12)0.8970.328
Colony stimulating factor
2 (granulocyte-macrophage)
?138960?
rs25882 CSF2 117Ile?Thr 612751.00 (0.92–1.09) 0.9610.610 1.03 (0.91–1.17)0.6010.28550.96 (0.83–1.12) 0.626 0.535
Cytotoxic
T-lymphocyte-associated
protein 4 ?123890?
rs5742909CTLA4 (?318)C?T61195 0.91 (0.80–1.02)0.1000.688 0.91 (0.79–1.03)0.135 0.54450.81 (0.51–1.30) 0.3870.197
Cytotoxic
T-lymphocyte-associated
protein 4 ?123890?
rs231775CTLA4 17Thr?Ala612851.02 (0.94–1.11) 0.5800.139 0.99 (0.87–1.13)0.862 0.1545 1.07 (0.94–1.21)0.3010.251
Fibrinogen beta ?134830?
rs1800790FGB (?455)G?A 993361.01 (0.93–1.09) 0.9050.877 1.01 (0.92–1.11)0.8660.6466 1.01 (0.81–1.25) 0.9630.503
Group-specific
component (vitamin D
binding protein) ?139200?
rs7041GC 416Glu?Asp6121 51.02 (0.94–1.10) 0.719 0.2240.97 (0.85–1.12) 0.7180.30651.05 (0.93–1.19)0.395 0.336
Group-specific
component (vitamin D
binding protein) ?139200?
rs4588GC 420Thr?Lys612251.02 (0.94–1.11) 0.5980.436 1.01 (0.89–1.14) 0.9280.38251.06 (0.92–1.21)0.4370.443
Guanine nucleotide binding
protein (G protein), beta
polypeptide 3 ?139130?
rs5443GNB3 825C?T 994161.08 (1.01–1.15) 0.0310.7591.06 (0.96–1.16)0.2430.8646 1.19 (1.05–1.36)0.009 0.134
Integrin, alpha 2 (CD49B,
alpha 2 subunit of VLA-2
receptor) ?192974?
rs1062535ITGA2 873G?A 994161.04 (0.97–1.11)0.268 0.1571.04 (0.95–1.14)0.3670.3276 1.06 (0.93–1.21)0.356 0.163
Integrin, beta 3 (platelet
glycoprotein IIIa, antigen
CD61) ?173470?
rs5918 ITGB3 33Leu?Pro 993760.98 (0.88–1.08) 0.6480.3490.97 (0.87–1.10)0.6620.4235 1.00 (0.70–1.44) 0.9930.375
Intercellular adhesion
molecule 1 (CD54), human
rhinovirus receptor
?147840?
rs5491ICAM1 56Lys?Met 853161.04 (0.84–1.29)0.717 0.928 1.05 (0.84–1.31)0.690 0.9403 0.98 (0.26–3.70)0.9800.907
Intercellular adhesion
molecule 1 (CD54), human
rhinovirus receptor
?147840?
rs1799969ICAM1 241Gly?Arg10 08861.09 (0.97–1.23) 0.1390.466 1.09 (0.95–1.24) 0.208 0.39861.31 (0.86–1.99) 0.2030.916
Interleukin 1, alpha
?147760?
rs1800587IL1A (?889)C?T 850860.93 (0.85–1.02)0.1100.7780.91 (0.82–1.01) 0.0830.83460.95 (0.76–1.19) 0.659 0.686
Interleukin 1, beta
?147720?
rs16944 IL1B (?1418)C?T609451.05 (0.97–1.13)0.240 0.5061.06 (0.95–1.19) 0.2880.47351.06 (0.92–1.23) 0.407 0.541
Interleukin 1, beta
?147720?
rs1143634 IL1B 4336C?T 853460.95 (0.86–1.05)0.331 0.415 0.92 (0.82–1.04)0.1880.2866 1.07 (0.81–1.41)0.6420.595
Interleukin 4 ?147780?
rs2243250 IL4 (?590)C?T 854161.05 (0.97–1.15) 0.237 0.0151.19 (1.05–1.35) 0.0070.10160.92 (0.8–1.07)0.2750.081
Interleukin 4 receptor
?147781?
rs1805010IL4R 75Ile?Val853861.00 (0.93–1.07)0.9870.9101.02 (0.91–1.14)0.7220.7906 0.98 (0.87–1.10)0.7200.915
Interleukin 4 receptor
?147781?
rs1805015 IL4R 503Ser?Pro 61305 1.00 (0.89–1.12)0.9450.2861.00 (0.88–1.14)0.9790.5365 1.00 (0.63–1.59)0.9970.168
Interleukin 4 receptor
?147781?
rs1801275 IL4R 576Gln?Arg 852561.04 (0.95–1.13) 0.4370.058 1.06 (0.96–1.18) 0.247 0.08860.92 (0.70–1.21) 0.546 0.273
Interleukin 5 receptor alpha
?147851?
rs2290608IL5RA (?80)G?A85286 1.02 (0.94–1.11)0.6080.1821.02 (0.92–1.12)0.754 0.2756 1.08 (0.87–1.34)0.482 0.114
(Continued)
6 Stroke
March 2009
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 8

Table 2. Continued
Additive Dominant Recessive
Gene ?MIM No.?
SNP rs No.SNP Name n* Sites†OR (LCL–UCL)‡P‡ Het P§ OR (LCL–UCL)‡P‡ Het P§ Sites¶OR (LCL–UCL)‡P‡ Het P§
Interleukin 6 ?147620?
rs1800796IL6 (?572)G?C 61245 0.92 (0.82–1.02) 0.103 0.0610.84 (0.70–1.00)0.0520.1134 0.95 (0.81–1.10)0.480 1.000
Interleukin 6 ?147620?
rs1800795 IL6 (?174)G?C85436 1.08 (0.99–1.18)0.095 0.687 1.08 (0.95–1.23)0.2320.2485 1.14 (0.97–1.34) 0.1210.718
Interleukin 9 ?146931?
rs2069885 IL9 113Thr?Met 854461.04 (0.92–1.18)0.5590.5431.10 (0.95–1.26)0.209 0.77250.73 (0.44–1.19) 0.205 0.131
Interleukin 10 ?124092?
rs1800872IL10 (?571)C?A 85386 0.96 (0.89–1.04)0.307 0.0120.96 (0.86–1.07)0.463 0.02860.94 (0.82–1.08)0.3750.190
Interleukin 13 ?147683?
rs1295686IL13 4045C?T 53664 1.01 (0.92–1.11)0.8940.7510.99 (0.89–1.12)0.9170.8544 1.07 (0.85–1.35) 0.561 0.380
Leukotriene C4 synthase
?246530?
rs730012LTC4S (?444)A?C 60445 0.99 (0.90–1.09)0.7970.2411.01 (0.91–1.13) 0.874 0.19350.88 (0.69–1.13) 0.3080.940
Lipase, hepatic ?151670?
rs1800588 LIPC (?480)C?T 10 09260.93 (0.87–1.00)0.048 0.4680.94 (0.86–1.03) 0.1820.63660.85 (0.71–1.00)0.051 0.091
Lipoprotein, Lp(a) ?152200?
rs1853021 LPA 93C?T10 0826 0.96 (0.88–1.04)0.3090.5240.97 (0.88–1.06)0.4750.58860.89 (0.66–1.19) 0.424 0.155
Lipoprotein, Lp(a) ?152200?
rs1800769LPA 121G?A 10 08060.95 (0.88–1.03) 0.2060.404 0.94 (0.85–1.03)0.1940.64760.97 (0.81–1.15) 0.6800.127
Lipoprotein lipase ?609708?
rs1800590 LPL (?93)T?G 10 0916 1.12 (0.84–1.49)0.4330.460 1.13 (0.83–1.54) 0.4440.45031.64 (0.42–6.36)0.473 0.861
Lipoprotein lipase ?609708?
rs1801177 LPL 9Asp?Asn 10 09161.15 (0.80–1.65)0.464 0.6101.15 (0.80–1.66)0.443 0.611 Too few variant homozygotes observed
Lipoprotein lipase ?609708?
rs268LPL 291Asn?Ser 10 0906 1.24 (0.94–1.64)0.1330.8871.29 (0.97–1.72) 0.080 0.925Too few variant homozygotes observed
Lipoprotein lipase ?609708?
rs328 LPL 447Ser?Term 10 08860.89 (0.80–0.99) 0.0330.4550.88 (0.79–0.99) 0.033 0.5426 0.89 (0.57–1.40)0.6230.244
Low density lipoprotein
receptor ?606945?
rs5742911LDLR NcoI ?/-10 08661.01 (0.94–1.08) 0.882 0.8690.99 (0.91–1.09)0.860 0.86261.05 (0.91–1.21) 0.540 0.472
Lymphotoxin alpha; Tumor
necrosis factor beta
?153440?
rs909253LTA 252A?G 60905 1.02 (0.94–1.10)0.6820.4041.07 (0.96–1.20) 0.226 0.56150.93 (0.80–1.09) 0.3810.294
Lymphotoxin alpha; Tumor
necrosis factor beta
?153440?
rs1041981 LTA 26Thr?Asn 10 09161.01 (0.95–1.08) 0.7420.387 1.02 (0.93–1.11)0.7150.27461.01 (0.88–1.16) 0.8720.344
Membrane-spanning
4-domains, subfamily A,
member 2 (Fc fragment of
IgE, high affinity I, receptor
for; beta polypeptide)
?147138?
rs569108MS4A2 237Glu?Gly85176 1.08 (0.95–1.23)0.2560.4791.10 (0.95–1.27)0.2200.4803 1.03 (0.65–1.64)0.891 0.999
5, 10-methylenetetrahydro-
folate reductase ?607093?
rs1801133MTHFR 677C?T99436 1.03 (0.96–1.10)0.383 0.389 1.05 (0.95–1.15)0.3600.51461.03 (0.91–1.16)0.6290.165
Natriuretic peptide
precursor A ?108780?
rs5063NPPA 664G?A 99406 0.96 (0.83–1.11) 0.5550.016 0.94 (0.81–1.10) 0.4480.0225 1.54 (0.73–3.26)0.257 0.689
Natriuretic peptide
precursor A ?108780?
rs5065 NPPA 2238T?C 993961.03 (0.92–1.15)0.612 0.781 1.05 (0.93–1.18)0.4520.9246 0.89 (0.59–1.33)0.560 0.453
Nitric oxide synthase
2A (inducible, hepatocytes)
asp346asp ?163730?
rs1137933NOS2A 231C?T613150.91 (0.82–1.01)0.0620.6590.91 (0.81–1.02)0.1110.4395 0.81 (0.60–1.09)0.1660.516
Nitric oxide synthase
3 (endothelial cell)
?163729?
rs1800779 NOS3 (?922)A?G 10 09361.02 (0.95–1.10) 0.5750.6761.03 (0.93–1.14)0.562 0.35861.03 (0.88–1.20) 0.7140.129
Nitric oxide synthase
3 (endothelial cell)
?163729?
rs3918226 NOS3 (?690)C?T 99366 1.15 (1.00–1.32) 0.047 0.479 1.16 (1.00–1.34)0.0520.56251.30 (0.70–2.41) 0.4080.726
Nitric oxide synthase 3
?163729?
rs1799983NOS3 298Glu?Asp 10 09461.05 (0.97–1.13)0.224 0.1761.05 (0.95–1.15)0.327 0.14561.10 (0.92–1.31) 0.2930.569
Paraoxonase 1 ?168820?
rs854560 PON1 55Leu?Met10 0906 0.97 (0.91–1.05) 0.4670.5100.97 (0.88–1.08) 0.5840.49560.95 (0.81–1.12)0.554 0.157
Paraoxonase 1 ?168820?
rs662PON1 192Gln?Arg 10 08861.05 (0.98–1.13) 0.155 0.7741.07 (0.97–1.18) 0.1670.46361.05 (0.93–1.20)0.412 0.268
Paraoxonase 2 ?602447?
rs6954345PON2 311Ser?Cys 10 09261.09 (1.01–1.18)0.0250.7931.09 (1.00–1.20) 0.060 0.6286 1.22 (0.99–1.49)0.063 0.460
Peroxisome proliferator
activated-receptor gamma
?601487?
rs1801282 PPARG 12Pro?Ala10 0936 1.05 (0.95–1.16)0.3620.702 1.07 (0.96–1.20)0.2200.42560.88 (0.59–1.32) 0.532 0.490
Secretoglobin, family 1A,
member 1 (uteroglobin)
?192020?
rs3741240 SCGB1A1 (?38)G?A853860.99 (0.92–1.06)0.797 0.103 0.98 (0.89–1.09)0.7010.43161.01 (0.88–1.16)0.884 0.040
Selectin E (endothelial
adhesion molecule 1)
?131210?
rs5361 SELE 128Ser?Arg 10 09461.00 (0.89–1.13) 0.9650.666 1.01 (0.89–1.15)0.903 0.64261.05 (0.61–1.81) 0.8500.673
Selectin E (endothelial
adhesion molecule 1)
?131210?
rs5355 SELE 554Leu?Phe990460.91 (0.76–1.08) 0.266 0.004 0.92 (0.77–1.10)0.362 0.00550.37 (0.07–1.88) 0.2280.998
(Continued)
Wang et alLTA and Stroke in Nonhypertensives
7
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 9

In the smoking-stratified analyses, no associations re-
mained statistically significant after the FDR was applied,
although under the dominant model, CD14 (-260)C?T was
suggestively associated (OR, 1.24; P?0.001; FDR, 0.058)
with ischemic stroke among never-smokers (Supplemental
Table IIC). Both LTA SNPs were associated with a greater
risk for ischemic stroke in never-smokers than ever-smokers
under the dominant model and although these associations
Table 2.Continued
AdditiveDominantRecessive
Gene ?MIM No.?
SNP rs No.SNP Name n*Sites†OR (LCL–UCL)‡P‡ Het P§OR (LCL–UCL)‡P‡ Het P§ Sites¶ OR (LCL–UCL)‡P‡Het P§
Selectin P (granule
membrane protein 140kDa,
antigen CD62) ?173610?
rs6131SELP 330Ser?Asn 85396 1.04 (0.95–1.13) 0.4240.296 1.03 (0.93–1.14)0.565 0.4836 1.12 (0.88–1.43)0.365 0.506
Selectin P (granule
membrane protein 140kDa,
antigen CD62) ?173610?
rs6133SELP 640Val?Leu 85376 1.09 (0.95–1.25)0.229 0.0371.10 (0.94–1.28) 0.226 0.02851.22 (0.71–2.12) 0.471 0.876
Serine (or cysteine)
proteinase inhibitor, clade
E (nexin, plasminogen
activator inhibitor type 1),
member 1 ?173360?
rs1799768SERPINE1
(?675)Del?InsG
993160.99 (0.93–1.06) 0.7690.5411.04 (0.95–1.15)0.387 0.52560.92 (0.82–1.03) 0.126 0.669
Serine (or cysteine)
proteinase inhibitor, clade
E (nexin, plasminogen
activator inhibitor type 1),
member 1 ?173360?
rs7242SERPINE1 11053T?G99406 0.99 (0.93–1.05)0.6990.266 1.02 (0.92–1.12)0.7660.25760.95 (0.84–1.06) 0.3240.413
Sodium channel,
nonvoltage-gated 1 alpha
?600228?
rs5742912 SCNN1A 493Trp?Arg994060.99 (0.76–1.27) 0.9070.0040.97 (0.75–1.26)0.8170.0024 1.73 (0.10–28.9) 0.705 1.000
Sodium channel,
nonvoltage-gated 1 alpha
?600228?
rs2228576 SCNN1A 663Ala?Thr99326 1.00 (0.93–1.07)0.9580.9581.00 (0.91–1.10)1.0000.92161.00 (0.87–1.14) 0.939 0.275
Matrix metalloproteinase
3 (stromelysin 1,
progelatinase) ?185250?
rs3025058 MMP3 (?1171)
Ins?DelA
993260.97 (0.91–1.04) 0.3840.074 0.95 (0.85–1.05)0.279 0.12660.98 (0.87–1.11)0.801 0.273
Transcription factor
7 (T-cell specific, HMG-box)
?189908?
rs244656TCF7 (?1459)A?T611850.96 (0.86–1.07) 0.438 0.0330.97 (0.86–1.10)0.674 0.0375 0.75 (0.50–1.14)0.1740.610
Transcription factor
7 (T-cell specific, HMG-box)
?189908?
rs5742913 TCF7 19Pro?Thr611751.05 (0.88–1.25)0.614 0.404 1.04 (0.86–1.25)0.7240.4984 1.40 (0.65–3.02)0.388 0.570
Transforming growth
factor, beta 1 ?190180?
rs1800469TGFB1 (?509)C?T 84866 0.92 (0.86–0.99)0.0280.7500.89 (0.80–0.98)0.023 0.1486 0.92 (0.81–1.06)0.252 0.658
Tumor necrosis factor (TNF
superfamily, member 2)
?191160?
rs1800750 TNF (?376)G?A 99416 0.72 (0.50–1.04)0.077 0.3710.72 (0.50–1.04)0.081 0.363Too few variant homozygotes observed
Tumor necrosis factor (TNF
superfamily, member 2)
?191160?
rs1800629 TNF (?308)G?A 10 09661.04 (0.95–1.15) 0.3950.2531.04 (0.93–1.16) 0.4790.32661.14 (0.83–1.59) 0.420 0.506
Tumor necrosis factor (TNF
superfamily, member 2)
?191160?
rs673 TNF (?244)G?A 99385 1.51 (0.38–6.01) 0.5620.944 1.51 (0.38–6.01) 0.5620.944 Too few variant homozygotes observed
Tumor necrosis factor (TNF
superfamily, member 2)
?191160?
rs361525 TNF (?238)G?A 10 0506 1.02 (0.88–1.18) 0.8150.231 1.02 (0.87–1.19)0.7970.28061.04 (0.38–2.82)0.9460.694
Vascular cell adhesion
molecule 1 ?192225?
rs1041163 VCAM1 (?1594)T?C 852560.98 (0.89–1.07)0.5950.8910.99 (0.89–1.10) 0.8050.87160.86 (0.64–1.16) 0.325 0.967
Vitamin D (1,25-
dihydroxyvitamin D3)
receptor ?601769?
rs2228570 VDR 1Thr?Met610451.07 (0.99–1.16) 0.078 0.424 1.12 (1.00–1.25)0.057 0.86751.07 (0.93–1.23) 0.3670.237
Vitamin D (1,25-
dihydroxyvitamin D3)
receptor ?601769?
rs1544410VDR 45082 G?A 612051.01 (0.92–1.12) 0.7980.504 1.00 (0.88–1.14)0.9740.1735 1.06 (0.87–1.31)0.558 0.508
*Total no. of genotypes available across all 6 study populations.
†No. of studies observing sufficient polymorphism at SNP site for analysis under additive and dominant modes.
‡LCL indicates lower confidence limit; UCL, upper confidence limit.
§P value for heterogeneity of the genetic effect across all studies.
¶No. of studies observing sufficient polymorphism at SNP site for analysis under recessive mode.
8 Stroke
March 2009
by on May 10, 2011 stroke.ahajournals.org Downloaded from

Page 10

were not statistically significant after accounting for multiple
testing, this trend was consistent across 5 studies (data not
shown); the Westphalian study was excluded due to limited
smoking data.
Discussion
In this meta-analysis, we evaluated the association between
105 polymorphisms in 64 inflammation and cardiovascular-
related genes and ischemic stroke in 3550 cases and 6560
control subjects across 6 different studies. Although we could
not further define subtypes of ischemic stroke, key strengths
of our stroke consortium are that these analyses were not
subject to publication bias and all studies used common
genotyping reagents. In the primary meta-analysis, modest
associations with stroke became nonsignificant after adjust-
ment for multiple testing using the FDR or permutation
testing. Stratification on sex or age also revealed no signifi-
cant associations. Notably, subjects in 2 of our studies were
limited to one sex and the consortium encompassed subjects
recruited from different regions in Europe, North America,
and China. We observed similar results among white partic-
ipants only, but study population differences resulting in
heterogeneity in stroke etiology could have obscured genetic
associations.
Stratification on hypertension status did, however, reveal a
statistically significant association for LTA 252A?G that
remained after adjustment for multiple testing. Across 4 white
populations and one Chinese population, the OR for LTA
252G was consistently greater among normotensive than
hypertensive subjects. LTA 26Thr?Asn yielded similar re-
sults among study participants with genotypes at both sites, as
expected, given the strong linkage disequilibrium between
these 2 LTA SNPs. Although a recent Japanese study ob-
served no association of these SNPs with any subtype of
ischemic stroke,30a smaller Hungarian study had previously
reported LTA 252G as a risk factor for large vessel ischemic
stroke31and an earlier Korean study had identified the LTA
252AA genotype as a risk factor for cerebral infarction.32We
were unable to analyze ischemic stroke subtypes, but there is
some evidence that subtypes may differ depending on hyper-
tensive status.33Although the number of nonhypertensives
cases was limited to 1068, stratification by hypertension
across our 6 populations may have reduced heterogeneity and
thus enabled us to discern the modest risk associated with
LTA polymorphism.
A role for LTA in chronic inflammation has been
suggested by its ability to induce expression of intercellu-
lar adhesion molecule-1 and vascular cell adhesion
molecule-1 on endothelial cells in vitro.34,35LTA expres-
sion results in a localized infiltrate consisting of T cells, B
cells, follicular and interdigitating dendritic cells, and
macrophages.36A recent mouse model study indicated that
LTA was expressed in atherosclerotic lesions whose size
correlated with concentration. Moreover, loss of the adja-
cent gene TNF did not affect development of lesions in
mice fed an atherogenic diet.37The A252G site is intronic
but has been associated with higher transcriptional activity
in a luciferase assay, whereas the variant protein bearing
the LTA 26 threonine to asparagine substitution has been
observed to induce greater expression of vascular cell
adhesion molecule-1 and SELE mRNA in vascular smooth
muscle cells. Because these 2 LTA SNPs are in almost
complete LD, the variant protein level was estimated to be
1.5-fold higher than the wild-type.10An increased level of
the variant protein may contribute to the increased risk for
ischemic stroke through inflammatory processes. Although
the mechanism by which LTA polymorphisms influence
inflammatory pathways is not clear, the meta-analysis pre-
sented here indicated that these LTA variants were associated
with ischemic stroke in nonhypertensive patients.
It is believed that subjects with hypertension tend to
develop chronic, low-grade systemic inflammation.38–40Se-
verity of inflammation caused by genetic variation could
independently modify predisposition to ischemic stroke. Re-
cent reports on the association of PDE4D variants with
ischemic stroke among normotensives3,4are consistent with
the hypothesis that hypertension may obscure or mask the
effect of inflammation-related genetic variants and that such
genetic effects can be most readily observed in the absence of
this major risk factor.
Smoking, like hypertension, can elicit an inflammatory
response.41In our study, the effect of LTA variation on
stroke was more discernable among never-smokers than
ever-smokers. Whether proinflammatory risks for ischemic
Table 3. Nominally Significant (P<0.05) Fixed-Effects Meta-analysis Results Among Those With Current or Past Hypertension
SNPMode* No. of Studies ORLower CL†Upper CL†
P
FDR
LPA 121G?A
TGFB1 (?509)C?T
CTLA4 (?318)C?T
TGFB1 (?509)C?T
LPA 121G?A
APOC3 3175C?G
SERPINE1 (?675)Del?InsG
LTA 252A?G
ITGA2 873G?A
ADD
ADD
ADD
DOM
DOM
REC
REC
REC
REC
6
6
5
6
6
6
6
5
6
0.886
0.894
0.851
0.845
0.872
0.708
0.851
0.806
1.202
0.801
0.812
0.730
0.737
0.768
0.514
0.731
0.657
1.003
0.979
0.983
0.993
0.969
0.992
0.975
0.991
0.989
1.442
0.017
0.021
0.040
0.016
0.037
0.034
0.038
0.039
0.047
0.915
0.915
0.915
0.968
0.968
0.824
0.824
0.824
0.824
*ADD indicates additive; DOM, dominant; REC, recessive mode of inheritance.
†CL indicates confidence limits.
Wang et alLTA and Stroke in Nonhypertensives
9
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 11

Table 4.
Under 3 Modes of Inheritance
SNPs Achieving P<0.05 for Association With Ischemic Stroke Among Normotensives (1068 cases, 3390 control subjects)
SNP
PHS
(163 cs, 1522 ctrl)
SOF
(50 cs, 212 ctrl)
Vienna
(232 cs, 632 ctrl)
Westphalia
(224 cs, 444 ctrl)
Pomerania
(92 cs, 330 ctrl)
SHINING
(307 cs, 250 ctrl)Meta*
ORP MAF†ORP MAF†ORP MAF†ORP MAF†ORPMAF† ORP MAF†ORP FDR Het‡
Additive mode
LTA 252A?G Not genotyped 0.970.913 0.331.30 0.1820.321.11 0.604 0.32 1.770.0030.291.62
?0.001 0.361.41
?0.00010.002. . .
LTA
26Thr?Asn
1.070.6030.320.99 0.970 0.33 1.100.4750.32 0.960.771 0.331.63 0.008 0.291.60
?0.001 0.351.190.003 0.1610.03
ACE Del?Ins 0.830.107 0.45 0.86 0.5080.49 0.930.5180.47 0.690.003 0.55 0.98 0.8840.500.970.8260.650.86 0.0070.244. . .
CETP
(?631)C?A
0.53 0.0360.07 0.93 0.8710.09 0.990.976 0.090.45 0.0030.080.78 0.470.08 Not
polymorphic
0.00 0.71 0.0090.244. . .
SELP
640Val?Leu
1.73
?0.001 0.111.200.6180.120.910.7590.111.02 0.947 0.11 0.920.777 0.11 1.520.637 0.004 1.310.0130.255 . . .
TCF7 19
Pro?Thr
Not genotyped 0.900.788 0.111.940.0240.071.03 0.9220.101.910.0230.08 Not
polymorphic
0.00 1.450.017 0.255. . .
ADD
460Gly?Trp
1.050.758 0.191.700.0630.170.96 0.7970.181.22 0.2030.19 1.560.022 0.201.16 0.2410.511.170.017 0.255. . .
IL10
(?571)C?A
0.650.006 0.24 0.53 0.0340.26 1.150.508 0.23 0.780.249 0.261.020.915 0.22 0.930.5480.630.84 0.020 0.255 . . .
AGT
235Met?Thr
0.690.002 0.431.03 0.9170.401.13 0.309 0.460.910.436 0.45 0.64 0.0110.43 0.940.670 0.810.87 0.021 0.2550.03
PON2
311Ser?Cys
1.07 0.610 0.23 0.900.7060.24 1.270.096 0.231.210.1950.231.11 0.608 0.241.150.3660.19 1.15 0.0380.389 . . .
IL6
(?572)G?C
Not genotyped1.070.894 0.05 0.740.4850.060.540.189 0.061.30 0.4990.050.76 0.0390.720.79 0.0410.389. . .
APOB
71Thr?Ile
1.230.091 0.29 0.840.4930.32 1.250.089 0.28 1.32 0.0360.280.93 0.682 0.330.84 0.3330.141.130.046 0.389 . . .
Dominant mode
LTA 252A?GNot genotyped1.240.542 0.33 1.530.116 0.32 1.090.7460.321.77 0.025 0.291.94
?0.0010.361.57
?0.00010.005 . . .
LTA
26Thr?Asn
1.140.437 0.321.29 0.4610.33 1.13 0.4620.320.88 0.4540.33 1.63 0.0450.291.87 0.0010.35 1.24 0.0070.253 0.06
ADD1
460Gly?Trp
1.140.443 0.191.500.235 0.170.97 0.8600.18 1.40 0.0660.191.570.059 0.20 1.350.161 0.511.25 0.0080.253 . . .
SELP
640Val?Leu
1.94
?0.0010.111.100.8150.12 0.98 0.9360.111.010.9830.11 0.960.896 0.111.520.637 0.004 1.360.011 0.253 . . .
CETP
(?631)C?A
0.540.0450.07 0.990.9820.091.01 0.954 0.090.38 0.0010.08 0.790.494 0.08Not
polymorphic
0.000.71 0.011 0.2530.08
IL4
(?590)C?T
1.450.0310.17 1.160.682 0.140.86 0.570 0.18 1.000.993 0.15 1.350.245 0.154.44 0.002 0.80 1.300.0150.2750.07
TCF7 19
Pro?Thr
Not genotyped0.96 0.9200.112.140.017 0.07 0.96 0.8880.101.84 0.0380.08Not
polymorphic
0.001.46 0.0220.353. . .
IL10
(?571)C?A
0.620.0070.240.480.0420.26 1.240.388 0.230.780.356 0.260.86 0.5690.221.060.8210.63 0.80 0.0280.363. . .
PON2
311Ser?Cys
0.99 0.9480.23 0.890.7200.241.31 0.116 0.231.380.0730.231.12 0.646 0.241.27 0.184 0.191.19 0.0310.363. . .
AGT
235Met?Thr
0.610.0030.431.150.7040.401.170.406 0.460.900.5800.450.57 0.0230.43 1.260.6090.81 0.820.0330.363 0.05
ACE Del?Ins0.860.3900.450.66 0.2410.49 0.9490.778 0.470.740.149 0.550.93 0.7940.50 0.730.2490.650.830.0460.413 . . .
APOE
112Cys?Arg
1.180.369 0.14 1.190.648 0.13 1.230.327 0.111.54 0.025 0.14 1.090.7540.120.920.7000.11 1.20 0.0480.413 . . .
Recessive mode
NOS3
298Glu?Asp
1.350.2440.31 1.800.205 0.31 1.520.125 0.29 1.630.0840.31 2.490.013 0.270.580.4230.111.560.0010.102. . .
LTA
252A?G
Not genotyped0.530.272 0.33 1.150.7440.321.29 0.5480.322.90 0.0050.29 1.720.0290.36 1.55 0.0070.360. . .
ACE Del?Ins0.650.0660.451.040.9260.490.850.420 0.47 0.480.001 0.55 1.010.980.501.090.6350.650.800.0200.5360.06
GNB3
825C?T
0.970.9180.32 2.070.1640.301.030.9280.311.750.0450.31 1.210.6610.281.42 0.1120.441.310.028 0.569. . .
LTA
26Thr?Asn
0.96 0.882 0.32 0.51 0.254 0.331.100.7370.321.200.5330.332.500.0130.291.800.0170.351.310.0320.569. . .
APOB
71Thr?Ile
1.80 0.0160.29 1.140.8050.32 0.880.6760.28 1.510.154 0.281.140.721 0.331.090.8970.14 1.33 0.0390.592 . . .
*Fixed-effect meta-analyses adjusted for age and sex were conducted based on the summary logistic regression results for each study site.
†Minor allele frequency observed among controls. For ACE, the insertion allele was defined as the minor allele based on the PHS population.
‡P values ?0.1 for heterogeneity of the genetic effect across 6 studies.
Cs indicates cases; ctrl, control subjects.
10Stroke
March 2009
by on May 10, 2011 stroke.ahajournals.org Downloaded from

Page 12

stroke caused by hypertension, smoking, or carrying a risk
allele are additive remains to be addressed by a carefully
designed study.
Summary
Our 6-study analysis surveyed inflammatory and cardiovas-
cular gene polymorphisms in examining the risk for ischemic
stroke. Our results indicate that the LTA 252A?G and LTA
26Thr?Asn polymorphisms have significant effects on the
risk for ischemic stroke in nonhypertensive subjects. We
cannot rule out the possible importance of these poly-
morphisms in hypertensive subjects, but a much larger cohort
may be needed to clarify the interaction of hypertension and
inflammation in the etiology of ischemic stroke.
Appendix
RMS Stroke SNP Consortium Investigators
Physicians’ Health Study: Nancy R. Cook, Paul M. Ridker, and
Robert Y. L. Zee from the Center for Cardiovascular Disease
Prevention and Division of Preventive Medicine, Brigham and
Women’s Hospital, Boston, Mass. Study of Osteoporotic Fractures:
Warren S. Browner, Steve R. Cummings, and Li-Yung Lui from the
S.F. Coordinating Center, California Pacific Medical Center, Re-
search Institute, San Francisco, Calif. Vienna Stroke Study: Georg
Endler and Christine Mannhalter from the Department of Medical
and Chemical Laboratory Diagnostics, Medical University Vienna,
Vienna, Austria; and Stefan Greisenegger and Wolfgang Lalouschek
from the University Clinic of Neurology, Medical University Vi-
enna, Vienna, Austria. Pomerania and Westphalia Studies: Klaus
Berger, Institute of Epidemiology and Social Medicine, University of
Muenster, Muenster, Germany; Harald Funke, Department of Mo-
lecular Hemostaseology, University of Jena, Jena, Germany; E.
Bernd Ringelstein, Department of Neurology, University of Muen-
ster, Muenster, Germany; Florian Stoegbauer, Department of Neu-
rology, University of Muenster, Muenster, Germany, and Klinikum
Osnabru ¨ck, Osnabru ¨ck, Germany; Jan Luedemann, Institutes of
Clinical Chemistry and Laboratory Medicine, University of Greif-
swald, Greifswald, Germany; Christof Kessler, and Department of
Neurology, University of Greifswald, Greifswald, Germany. SHIN-
ING: Lisheng Liu, Yu Shi, and Xingyu Wang from the Laboratory of
Human Genetics, Beijing Hypertension League Institute, Beijing,
China. Roche Molecular Systems: Victoria H. Brophy, Suzanne
Cheng, Henry A. Erlich, Andrea M. Johnson, and Brian K. Rhees
from the Department of Human Genetics, Roche Molecular Systems,
Inc, Pleasanton, Calif.
Acknowledgments
The RMS Stroke SNP Consortium (see Appendix for full list of
investigators) thanks the staff and many participants of the PHS,
SOF, Vienna Stroke Registry, Westphalian Stroke Register, Dort-
mund Health Study, Pomeranian Stroke Study, Study of Health in
Pomerania, and SHINING for their dedication and cooperation. We
thank the RMS CVD project team for their expertise in developing
the genotyping reagents used for these studies, the RMS High-
Throughput Genotyping Group for their expert technical assistance
in genotyping the SOF samples, and Michael Janisiw for assistance
in genotyping the Viennese samples.
Figure. Risk of ischemic stroke associated with LTA polymorphism. ORs under additive and dominant modes of inheritance for the LTA
252A?G and LTA 26Thr?Asn polymorphisms among normotensive (square) and hypertensive (diamond) subjects are plotted for each
study (open squares/diamonds) and the fixed-effects meta-analysis (filled squares/diamonds). Horizontal lines extend across the 95%
confidence limits.
Wang et al LTA and Stroke in Nonhypertensives
11
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 13

Sources of Funding
The SOF is supported by National Institutes of Health funding. The
following institutes provide support: the National Institute of Arthri-
tis and Musculoskeletal and Skin Diseases (NIAMS) and the
National Institute on Aging (NIA) under the following grant num-
bers: AG05407, AR35582, AG05394, AR35584, AR35583, R01
AG005407, R01 AG027576-22, 2 R01 AG005394-22A1, and 2 R01
AG027574-22A1. Case assessments in the Westphalian and Pome-
ranian studies are part of the German “Competence Net Stroke,”
which is supported by the German Federal Ministry of Education and
Research (01GI9909/3). Data collection in the Dortmund Health
Study was funded by the German Migraine and Headache Society
and unrestricted grants of equal share of a consortium of 7 pharma-
ceutical companies. The Study of Health in Pomerania is funded by
grants from the German Federal Ministry of Education and Research
(BMBF, 01ZZ96030) and from the Ministry for Education, Research
and Cultural Affairs and the Ministry for Social Affairs of the
Federal State of Mecklenburg-Vorpommern. The Vienna Stroke
Registry is supported by research grants of the Medizinisch-
Wissenschaftlicher Fonds des Bu ¨rgermeisters der Bundeshauptstadt
Wien (project numbers 1540, 1829, 1970), of the Jubila ¨umsfonds der
Oesterreichischen Nationalbank (project numbers 6866, 7115, 8281,
9344), and the Austrian Science Foundation (P13902-MED). The
Vienna Stroke Registry is also supported by the Wiener Krankenan-
staltenverbund. SHINING was funded through the Beijing Hyper-
tension League Institute, in part through the National Infrastructure
Program of Chinese Genetic Resource (2005DKA21300) and an
unrestricted educational grant from F. Hoffmann-La Roche. The
PHS is supported by grants from the National Heart Lung and Blood
Institute (HL-58755, and HL-63293), the Doris Duke Charitable
Foundation, the American Heart Association, and the Donald W.
Reynolds Foundation, Las Vegas, Nev. Additional funding provided
by the Fondation Leducq, Paris, France (P.M.R.).
Disclosures
S.C., V.H.B., and H.A.E. are employees of Roche Molecular
Systems, Inc, which provided reagents and support for genotyping to
all study sites under research collaborations and partial funding for
the meta-analysis. K.L. is an employee of F. Hoffmann-La Roche,
Ltd, which provided an unrestricted educational grant to the Beijing
Hypertension League Institute.
References
1. Rohr J, Kittner S, Feeser B, Hebel JR, Whyte MG, Weinstein A, Kanarak
N, Buchholz D, Earley C, Johnson C, Macko R, Price T, Sloan M, Stern
B, Wityk R, Wozniak M, Sherwin R. Traditional risk factors and ischemic
stroke in young adults: the Baltimore–Washington Cooperative Young
Stroke Study. Arch Neurol. 1996;53:603–607.
2. Zhang LF, Yang J, Hong Z, Yuan GG, Zhou BF, Zhao LC, Huang YN,
Chen J, Wu YF; Collaborative Group of China Multicenter Study of
Cardiovascular Epidemiology. Proportion of different subtypes of stroke
in China. Stroke. 2003;34:2091–2096.
3. Brophy VH, Ro SK, Rhees BK, Lui LY, Lee JM, Umblas N, Bentley LG.
Li J, Cheng S, Browner WS, Erlich HA. Association of phosphodiesterase
4D polymorphisms with ischemic stroke in a US population stratified by
hypertension status. Stroke. 2006;37:1385–1390.
4. Zee RY, Brophy VH, Cheng S, Hegener HH, Erlich HA, Ridker PM.
Polymorphisms of the phosphodiesterase 4D, cAMP-specific (PDE4D)
gene and risk of ischemic stroke: a prospective, nested case–control
evaluation. Stroke. 2006;37:2012–2017.
5. Zhao Y, Ma LY, Liu YX, Wang XY, Liu LS, Lindpaintner K. Rela-
tionship between alpha-ENaC gene Thr663Ala polymorphism and ische-
mic stroke [in Chinese]. Zhongguo Yi Xue Ke Xue Yuan Xue Bao.
2001;23:499–501.
6. Schmieder RE, Hilgers KF, Schlaich MP, Schmidt BM. Renin–angioten-
sin system and cardiovascular risk. Lancet. 2007;369:1208–1219.
7. Lindsberg PJ, Grau AJ. Inflammation and infections as risk factors for
ischemic stroke. Stroke. 2003;34:2518–2532.
8. Bova IY, Bornstein NM, Korczyn AD. Acute infection as a risk factor for
ischemic stroke. Stroke. 1996;27:2204–2206.
9. Grau AJ, Buggle F, Becher H, Werle E, Hacke W. The association of
leukocyte count, fibrinogen and C-reactive protein with vascular risk
factors and ischemic vascular diseases. Thromb Res. 1996;82:245–255.
10. Hansson GK, Robertson AK, So ¨derberg-Naucle ´r C. Inflammation and
atherosclerosis. Annu Rev Pathol. 2006;1:297–329.
11. McGeer PL, Rogers J, McGeer EG. Inflammation, anti-inflammatory
agents and Alzheimer disease: the last 12 years. J Alzheimers Dis. 2006;
9:271–276.
12. Lange LA, Carlson CS, Hindorff LA, Lange EM, Walston J, Durda JP,
Cushman M, Bis JC, Zeng D, Lin D, Kuller LH, Nickerson DA, Psaty
BM, Tracy RP, Reiner AP. Association of polymorphisms in the CRP
gene with circulating C-reactive protein levels and cardiovascular events.
JAMA. 2006;296:2703–2711.
13. Hollegaard MV, Bidwell JL. Cytokine gene polymorphism in human
disease: on-line databases, Supplement 3. Genes Immun. 2006;7:
269–276.
14. Ozaki K, Ohnishi Y, Iida A, Sekine A, Yamada R, Tsunoda T, Sato H,
Sato H, Hori M, Nakamura Y, Tanaka T. Functional SNPs in the
lymphotoxin-alpha gene that are associated with susceptibility to myo-
cardial infarction. Nat Genet. 2002;32:650–654.
15. Endres M, Laufs U, Merz H, Kaps M. Focal expression of intercellular
adhesion molecule-1 in the human carotid bifurcation. Stroke. 1997;28:
77–82.
16. Dichgans M, Markus HS. Genetic association studies in stroke: method-
ological issues and proposed standard criteria. Stroke. 2005;36:
2027–2031.
17. Steering Committee of the Physicians’ Health Study Research Group.
Final report of the aspirin component of the ongoing Physicians’ Health
Study. N Engl J Med. 1989;321:129–135.
18. Zee RY, Cook NR, Cheng S, Reynolds R, Erlich HA, Lindpaintner K,
Ridker PM. Polymorphism in the P-selectin and interleukin-4 genes as
determinants of stroke: a population-based, prospective genetic analysis.
Hum Mol Genet. 2004;13:389–396.
19. Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE,
Cauley J, Black D, Vogt TM. Risk factors for hip fracture in white
women. Study of Osteoporotic Fractures Research Group. N Engl J Med.
1995;332:767–773.
20. Schmidt W-P, Heuschmann P, Taeger D, Henningsen H, Bu ¨cker-Nott HJ,
Berger K. Determinants of IV heparin treatment in patients with ischemic
stroke. Neurology. 2004;63:2407–2409.
21. Evers S, Fischera M, May A, Berger K. Prevalence of cluster headache in
Germany: results of the epidemiological DMKG study. J Neurol Neu-
rosurg Psychiatry. 2007;78:1289–1290.
22. Luedemann J, Schminke U, Berger K, Piek M, Willich SN, Do ¨ring A,
John U, Kessler C. Association between behavior-dependent cardiovas-
cular risk factors and asymptomatic carotid atherosclerosis in a general
population. Stroke. 2002;33:2929–2935.
23. Lang W, Lalouschek W, on behalf of the Vienna Stroke Study Group. The
Vienna Stroke Registry—objectives and methodology. Wien Klin
Wochenschr. 2001;113:141–147.
24. Cheng S, Grow MA, Pallaud C, Klitz W, Erlich HA, Visvikis S, Chen JJ,
Pullinger CR, Malloy MJ, Siest G, Kane JP. A multilocus genotyping
assay for candidate markers of cardiovascular disease risk. Genome Res.
1999;9:936–949.
25. Barcellos LF, Begovich AB, Reynolds RL, Caillier SJ, Brassat D,
Schmidt S, Grams SE, Walker K, Steiner LL, Cree BA, Stillman A,
Lincoln RR, Pericak-Vance MA, Haines JL, Erlich HA, Hauser SL,
Oksenberg JR. Linkage and association with the NOS2A locus on chro-
mosome 17q11 in multiple sclerosis. Ann Neurol. 2004;55:793–800.
26. Whitehead A. Meta-Analysis of Controlled Clinical Trials. New York:
Wiley; 2002.
27. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin
Trials. 1986;7:177–188.
28. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a
practical and powerful approach to multiple testing. J R Stat Soc B.
1995;57:289–300.
29. Westfall PH, Young SS. P value adjustments for multiple tests in multi-
variable binomial models. J Am Stat Assoc. 1989;84:780–786.
30. Hagiwara N, Kitazono T, Kamouchi M, Kuroda J, Ago T, Hata J,
Ninomiya T, Ooboshi H, Kumai Y, Yoshimura S, Tamaki K, Fujii K,
Nagao T, Okada Y, Toyoda K, Nakane H, Sugimori H, Yamashita Y,
Wakugawa Y, Kubo M, Tanizaki Y, Kiyohara Y, Ibayashi S, Iida M.
12Stroke
March 2009
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 14

Polymorphisms in the lymphotoxin alpha gene and the risk of ischemic
stroke in the Japanese population. The Fukuoka Stroke Registry and the
Hisayama Study. Cerebrovasc Dis. 2008;25:417–422.
31. Szolnoki Z, Havasi V, Talian G, Bene J, Komlosi K, Somogyvari F,
Kondacs A, Szabo M, Fodor L, Bodor A, Melegh B. Lymphotoxin-alpha
gene 252G allelic variant is a risk factor for large-vessel-associated
ischemic stroke. J Mol Neurosci. 2005;27:205–211.
32. Um JY, An NH, Kim HM. TNF-alpha and TNF-beta gene polymorphisms
in cerebral infarction. J Mol Neurosci. 2003;21:167–171.
33. Arboix A, Roig H, Rossich R, Martínez EM, García-Eroles L. Differences
between hypertensive and non-hypertensive ischemic stroke. Eur
J Neurol. 2004;11:687–692.
34. Pober JS, Lapierre LA, Stolpen AH, Brock TA, Springer TA, Fiers W,
Bevilacqua MP, Mendrick DL, Gimbrone MA Jr. Activation of cultured
human endothelial cells by recombinant lymphotoxin: comparison with
tumor necrosis factor and interleukin 1 species. J Immunol. 1987;138:
3319–3324.
35. Cavender DE, Edelbaum D, Ziff M. Endothelial cell activation induced by
tumor necrosis factor and lymphotoxin. Am J Pathol. 1989;134:551–560.
36. Kratz A, Campos-Neto A, Hanson MS, Ruddle NH. Chronic inflam-
mation caused by lymphotoxin is lymphoid neogenesis. J Exp Med.
1996;183:1461–1472.
37. Schreyer SA, Vick CM, LeBoeuf RC. Loss of lymphotoxin-alpha but not
tumor necrosis factor-alpha reduces atherosclerosis in mice. J Biol Chem.
2002;277:12364–12368.
38. Kampus P, Muda P, Kals J, Ristimae T, Fischer K, Teesalu R, Zilmer M.
The relationship between inflammation and arterial stiffness in patients
with essential hypertension. Int J Cardiol. 2006;112:46–51.
39. Li JJ, Chen JL. Inflammation may be a bridge connecting hypertension
and atherosclerosis. Med Hypotheses. 2005;64:925–929.
40. Morishita R. Is vascular endothelial growth factor a missing link between
hypertension and inflammation? Hypertension. 2004;44:253–254.
41. Yanbaeva DG, Dentener MA, Creutzberg EC, Wesselingm G, Wouters
EF. Systematic effects of smoking. Chest. 2007;131:1557–1566.
Wang et alLTA and Stroke in Nonhypertensives
13
by on May 10, 2011 stroke.ahajournals.orgDownloaded from

Page 15

Supplementary Table IA.Frequencies of the Less Common or ’Variant’ Allele Among the Controls for Each Study*
SNP Allele
Allele Frequency†Large-Sample Test for HWE (P)Exact Test for HWE (P)
PHSSOF ViennaWest.Pom. SHIN.PHSSOF ViennaWest.Pom. SHIN.PHS SOFViennaWest.Pom.SHIN.
ACE Del?Ins Ins0.446 0.460 0.4690.541 0.4940.626 0.6580.1090.2850.0080.5660.3840.665 0.1090.3010.007 0.604 0.427
ADD1 460Gly?TrpTrp 0.1870.2060.181 0.187 0.2090.5220.937 0.9040.969 0.1720.218 0.8840.9420.8971.000 0.1840.2460.918
ADRB2 16Gly?Arg Arg 0.3780.3890.3740.3860.382 0.5810.0560.945 0.8700.3820.3950.2780.057 1.0000.892 0.3980.426 0.288
ADRB2 27Gln?GluGlu0.4100.408 0.4150.431 0.418 0.1100.1100.484 0.4930.639 0.824 0.0390.1190.5350.5080.6520.876 0.043
ADRB2 164Thr?IleIle 0.0130.00810.012 0.00970.012 0.0003 0.5570.8480.797 0.7920.7580.9901.0001.0001.0001.000 1.0001.000
ADRB3 64Trp?Arg Arg0.0700.089 0.081 0.071 0.0750.164 0.4610.838 0.8720.117 0.2900.510 0.6061.000 1.0000.168 0.418 0.506
AGT 235Met?ThrThr0.4380.398 0.4560.437 0.4180.793 0.719 0.9550.4260.9990.9890.4180.7211.000 0.439 1.0001.000 0.405
AGTR1 1166A?CC 0.2970.2890.2790.275 0.2980.048 0.8290.329 0.2950.747 0.4610.202 0.840 0.3670.301 0.7910.5330.366
APOA4 347Thr?Ser Ser0.194 0.195 0.2020.2170.197 0.0020 0.6340.2480.238 0.7400.480 0.937 0.6320.2860.2320.7620.4801.000
APOA4 360Gln?His His0.0870.079 0.0530.072 0.0700.00030.551 0.1400.649 0.3030.8010.990 0.5790.142 1.000 0.4261.000 1.000
APOB 71Thr?Ile Ile 0.2840.325 0.2870.2920.3270.1260.1430.3310.7990.418 0.3620.8680.1480.3790.815 0.428 0.380 0.814
APOB 3500Arg?Gln Gln0.00020 0.00050 0.00070 0.991.0.987.0.985.1.000.1.000. 1.000. . .
APOC3 (?641)C?AA 0.3860.370 0.4010.3960.3970.453 0.233 0.6090.617 0.091 0.4580.748 0.230 0.6390.641 0.105 0.4850.749
APOC3 (?482)C?TT 0.2660.263 0.2950.3090.2810.430 0.2390.919 0.8760.609 0.1750.3100.247 0.9120.9390.670 0.189 0.314
APOC3 (?455)T?CC0.378 0.370 0.3950.3950.3910.434 0.407 0.7740.6900.0870.293 0.7220.3980.7810.7360.0910.3410.747
APOC3 1100C?TT 0.2620.2680.3000.2630.3050.584 0.1950.394 0.3850.980 0.708 0.3820.1950.383 0.4021.000 0.7170.390
APOC3 3175C?GG0.1040.0970.102 0.0930.1040.3150.457 0.7080.460 0.496 0.8450.569 0.4760.6340.5830.6660.8380.581
APOC3 3206T?GG0.391 0.375 0.4110.3790.418 0.8110.0310.115 0.8450.3710.3490.1030.0300.1290.891 0.3970.349 0.104
APOE 112Cys?ArgArg 0.1390.1180.1140.1400.1270.102 0.4400.9420.049 0.9420.264 0.3100.515 0.845 0.0520.879 0.3080.314
APOE 158Arg?CysCys 0.0830.082 0.0980.085 0.0890.0850.688 0.3160.5510.7870.244 0.391 0.7840.5740.584 0.8110.238 0.503
C3 102Arg?GlyGlyNA0.2110.195 0.1940.197 0.0061. 0.896 0.0870.7170.664 0.813. 1.0000.1040.816 0.7201.000
C5 802Val?IleIle NA 0.4530.4250.4290.452 0.577. 0.6040.5210.585 0.331 0.522.0.6090.5780.6450.3570.525
CCL11 (?1328)G?AA NA 0.1960.1790.1740.1620.059. 0.2220.368 0.1020.788 0.151. 0.2290.3480.1170.7740.230
CCL11 23Ala?Thr Thr0.1780.180 0.1580.1630.182 0.131 0.091 0.7840.3230.1910.3100.1750.098 0.8850.387 0.2170.372 0.173
CCR2 62Val?IleIle 0.0960.089 0.1060.103 0.089 0.2510.367 0.4660.113 0.9070.2450.3180.376 0.6100.1500.8410.345 0.329
CCR3 39Pro?LeuLeu0.0060 0.0027 0.0000.0014 0.00070.0014 0.7830.949.0.9700.985 0.9581.0001.000. 1.0001.0001.000
CCR5 (?2459)A?GG0.454 0.4410.4160.4440.4180.5630.296 0.6930.2670.3280.1660.0650.317 0.7340.300 0.3310.179 0.075
CCR5 580Ins?Del32 Del0.1090.1110.1200.1040.10800.2570.9580.9180.4690.241. 0.2651.000 0.819 0.6920.245.
CD14 (?260)C?TT0.4840.4870.4480.4510.462 0.6200.4860.535 0.7420.8510.0130.8280.4810.5490.7840.8840.0150.867
CETP (?631)C?AA 0.0690.094 0.0880.0820.0780.00100.0440.1290.887 0.0600.087 0.969 0.0560.1360.8450.0820.1111.000
CETP (?629)C?AA0.4950.4450.4840.4750.4850.545 0.2400.2780.4810.1600.0600.616 0.2390.306 0.487 0.168 0.0690.637
CETP 405Ile?Val Val 0.325 0.3120.2940.3200.3130.4690.3510.884 0.614 0.4580.408 0.3840.3670.9240.646 0.4510.4230.403
CETP 442Asp?GlyGly0.00020.00090.00050.000700.0160.9910.9830.9870.985.
?0.0011.0001.0001.0001.000.
?0.001
CSF2 117Ile?ThrThrNA0.1960.218 0.1960.208 0.606. 0.8820.9080.141 0.819 0.043.0.8981.000 0.1500.819 0.040
CTLA4 (?318)C?TTNA 0.1020.1070.103 0.1070.140.0.3150.8380.8000.1260.386.0.3481.0001.0000.1590.377
CTLA4 17Thr?Ala AlaNA0.3680.387 0.382 0.3880.682.0.7780.1870.7280.2560.386.0.7850.2100.7510.2950.405
CXCL12 (?800)G?AA NA0.1980.242 0.2090.182 0.219. 0.3150.000 0.9030.7640.595.0.343
?0.0010.9100.7970.658
F2 20210G?AA0.019 0.00720.0140.0190.0140 0.8040.864 0.0700.609 0.715.0.5551.0000.1721.0001.000.
F5 506Arg?Gln Gln 0.0250.0340.0320.034 0.0400.0022 0.807 0.4050.3060.876 0.269
?0.0011.0001.0000.6260.5880.6200.005
F7 (?323)
Del?Ins10 Ins0.1460.1270.1120.1220.1190.0500.4470.4380.8090.5390.6870.1380.4770.5660.7460.4920.7140.146
F7 353Arg?GlnGln0.1380.1160.1090.1110.1090.0520.7610.3090.4490.3090.507 0.0210.858 0.4050.410 0.340 0.5570.029
FGB (?455)G?AA 0.2070.197 0.2280.233 0.2320.1990.0520.8630.6110.1620.0020.9250.0511.0000.6450.1830.0020.933
GC 416Glu?Asp Asp NA0.4310.4270.4180.4180.736.0.3980.2570.1650.8290.500.0.4370.2970.2000.873 0.540
GC 420Thr?LysLysNA 0.2880.303 0.2670.281 0.679. 0.019 0.8830.403 0.5150.211. 0.0230.9120.4370.500 0.231
GNB3 825C?TT 0.3170.300 0.3020.297 0.3010.468 0.9100.2870.3250.1770.7230.7950.9600.3240.3650.1940.7890.835
LIPC (?480)C?TT0.230 0.2210.2170.2260.2340.386 0.021 0.5660.8370.2340.1020.4310.0250.6220.8500.2600.1090.437
ICAM1 56Lys?Met Met0.0062 0.00540.004 0.0014 0.00150.063
?0.0010.898
?0.0010.9700.9700.4330.0031.000 0.0071.000 1.0000.643
ICAM1 241Gly?Arg Arg0.0970.1290.1050.1300.135 0.00200.8800.9080.9560.8080.0170.9381.0001.0000.8651.0000.0241.000
IL1A (?889)C?TT0.3090.3050.2830.3140.3050.1040.7450.9670.6110.9460.2010.4150.8031.0000.6490.9330.2370.481
IL1B (?1418)C?TT.0.3400.3430.3080.3220.473.0.0480.8440.3320.9200.663.0.0600.8390.3360.9320.677
IL1B 4336C?TT0.2370.2250.2240.2590.2450.0340.4810.3830.4940.3070.4500.5940.5080.3970.5920.3310.4721.000
IL4 (?590)C?TT0.1730.1400.1800.1420.1510.7810.1530.6990.8210.6670.5930.0030.166 0.727 0.8730.6500.665 0.004
IL4R 75Ile?ValVal 0.4490.454 0.4280.453 0.457 0.5160.0330.853 0.0100.776 0.7530.340 0.0360.866 0.0110.768 0.8140.348
IL4R 503Ser?Pro ProNA 0.1370.1450.1660.167 0.081. 0.2120.7330.0970.1470.610.0.2780.8510.1080.1740.599
(Continued)
14Stroke
March 2009
by on May 10, 2011 stroke.ahajournals.orgDownloaded from