Are Myocardial Infarction–Associated Single-Nucleotide Polymorphisms Associated With Ischemic Stroke?

Abstract

Background and Purpose

Ischemic stroke (IS) shares many common risk factors with coronary artery disease (CAD). We hypothesized that genetic variants associated with myocardial infarction (MI) or CAD may be similarly involved in the etiology of IS. To test this hypothesis, we evaluated whether single-nucleotide polymorphisms (SNPs) at 11 different loci recently associated with MI or CAD through genome-wide association studies were associated with IS.

Methods

Meta-analyses of the associations between the 11 MI-associated SNPs and IS were performed using 6865 cases and 11 395 control subjects recruited from 9 studies. SNPs were either genotyped directly or imputed; in a few cases a surrogate SNP in high linkage disequilibrium was chosen. Logistic regression was performed within each study to obtain study-specific βs and standard errors. Meta-analysis was conducted using an inverse variance weighted approach assuming a random effect model.

Results

Despite having power to detect odds ratio of 1.09–1.14 for overall IS and 1.20–1.32 for major stroke subtypes, none of the SNPs were significantly associated with overall IS and/or stroke subtypes after adjusting for multiple comparisons.

Conclusions

Our results suggest that the major common loci associated with MI risk do not have effects of similar magnitude on overall IS but do not preclude moderate associations restricted to specific IS subtypes. Disparate mechanisms may be critical in the development of acute ischemic coronary and cerebrovascular events.

Ischemic strokes (IS) comprise 87% of all strokes.¹ Nearly 800 000 individuals in the United States have a stroke each year, making stroke one of the leading causes of serious, long-term disability in developed countries.^1,2 The tendency of stroke to aggregate within families implies a genetic etiology, although few genetic variants have as yet been unequivocally linked to the common forms of IS.

Some of the difficulty in identifying stroke susceptibility genes may arise from heterogeneity within the stroke phenotype, making it imperative to carry out genetic studies within specific subtypes of IS. IS syndromes can, in fact, be recognized by experienced clinicians with a high degree of inter-rater agreement.³ The major categories, which are based on the presumed stroke mechanism, include atherosclerotic (ie, large-artery) stroke, cardioembolic stroke, and small-vessel (lacunar) stroke. Given disparate mechanisms, it is likely that different risk alleles would contribute to these different stroke subtypes. Previous studies showing that a family history of stroke is more common in younger compared with older age of onset strokes^4–6 raise the additional possibility that the genetic contribution to stroke may differ by age of onset.

IS shares many risk factors with myocardial infarction (MI) and coronary artery disease (CAD), including smoking and high blood pressure. The possibility that common susceptibility genes might predispose to both IS and MI/CAD prompted us to examine whether genetic variants previously associated with MI/CAD might also be associated with IS. To address this hypothesis, we evaluated whether DNA sequence variants at 11 different loci recently associated with MI or CAD through genome-wide association (GWA) studies were also associated with IS. Our analysis included 6865 IS cases from 9 different studies, allowing us to test whether the genetic variants were associated not only with overall IS, but also with the major IS subtypes and with age of stroke onset.

Methods

Study Populations

This analysis includes 6865 IS cases and 11 395 control subjects from 9 different studies: Australian Stroke Genetics Collaborative (ASGC), Bio-Repository of DNA in Stroke (BRAINS), Besta Cerebrovascular Diseases Registry (CEDIR), Genes Affecting Stroke Risk and Outcomes Study (GASROS), Genetics of Early Onset Stroke (GEOS), Graz Stroke Study (GRAZ), Ischemic Stroke Genetics Study and Siblings with Ischemic Stroke Study (ISGS/SWISS), Risk Assessment of Cerebrovascular Events (RACE), and the Vitamin Intervention for Stroke Prevention (VISP) study. All stroke cases, except those from VISP study, were adjudicated for ischemic stroke subtype using the original TOAST system,³ which assigns each case to 1 of the following categories: cardioembolic, large-artery, small-vessel, other known causes, and undetermined causes. Control subjects free of stroke were selected based on study-specific criteria. For this report, cases and control subjects with a known history of MI were excluded from the analyses. Details of the study designs for each participating study are included in the online-only Data Supplement Methods (http://stroke.ahajournals.org).

All participating studies were conducted with the consent of study subjects and were approved by the institutional review board in each institution.

SNP Selection

SNPs shown previously to be associated with MI or CAD in whites were identified through a review of genome-wide association studies published before May 2010. MI/CAD-associated SNPs reported in these previously published GWA studies were selected for inclusion in this study only if their associations with MI/CAD reached genome-wide significance (P<5×10⁻⁸) and were replicated in at least 1 large, independent, white cohort. Based on these criteria, we identified 11 loci from nine chromosomes that were consistently associated with MI/CAD (see Table 1).^7–13

Table 1.

Previous Published SNPs Associated With MI or CAD

SNP	Nonrisk/Risk Allele*	Frequency of Risk Allele	Region	Position	Reported Gene	Previous Reported Odds Ratio on MI	Reference
rs11206510	C/T	0.81	1p32	55,268,627	PCSK9	1.15	8
rs646776	C/T	0.81	1p13	109,620,053	CELSR2-PSRC1-SORT1	1.17	10
rs17465637	A/C	0.72	1q41	220,890,152	MIA3	1.13	8,12
rs6725887	T/C	0.14	2q33	203,454,130	WDR12	1.17	8
rs9818870	C/T	0.16	3q22	139,604,812	MRAS	1.15	7
rs12526453	G/C	0.65	6p24	13,035,530	PHACTR1	1.12	8
rs4977574	A/G	0.56	9p21	22,088,574	ANRIL/CDKN2B-AS1	1.28	8,9,11
rs1746048	T/C	0.84	10q11	44,095,830	CXCL12	1.14	8,12
rs3184504	C/T	0.39	12q24	110,368,991	SH2B3	1.13	13
rs1122608	T/G	0.75	19p13	11,024,601	SMARCA4-LDLR	1.15	8
rs9982601	C/T	0.13	21q22	34,520,998	SLC5A3-MRPS6-KCNE2	1.20	8

SNP indicates single-nucleotide polymorphism; MI, myocardial infarction; CAD, coronary artery disease.

^*Risk allele is the allele associated with increased risk of MI or CAD.

Genotyping and Imputation

The 11 MI-associated single-nucleotide polymorphisms (SNPs) selected for inclusion in this study were genotyped by existing commercial genome-wide SNP platforms, KASPar SNP array (KBioscience, Herts, United Kingdom) and/or by Taqman assays (Applied Biosystems, Foster City, CA). For SNPs that were not directly genotyped, genotypes were obtained either through imputation or the targeted SNP was replaced by a surrogate SNP that was in high linkage disequilibrium (LD r² >0.8) with the target SNP. A detailed description on the genotyping and imputation methods for each participating site is provided in online-only Data Supplement Table S1.

Statistical Analysis

Site-specific logistic regression analysis was performed to test for associations of each individual SNP with IS under an additive model, and then site-specific β coefficients and standard errors (SE) were pooled in meta-analyses. A genotype risk score was also computed for each individual by summing the number of MI/CAD-risk alleles across each of the 11 loci (score 1). Each SNP was given equal weights rather than weights based on their effect sizes (β) for MI/CAD in the risk score calculation given that we do not know the effects of these SNPs on stroke and the assumption that they would have similar effects on stroke as they do on MI seems unjustified. Additionally, we computed a genotype risk score that summed the number of MI/CAD-risk alleles across 10 loci, excluding SNP rs4977574 on chromosome 9, because SNPs near this locus have previously been reported to be associated with large-artery stroke (score 2).¹⁴ Each genotype risk score was used as a continuous independent variable in the logistic regression model to obtain the joint additive effects of all MI/CAD-associated SNPs on the risk of IS. The β of the genotype risk score represents the increase in log odds of stroke associated with having one additional MI/CAD-risk allele. The GRAZ study was not included in the score analyses because three of the 11 SNPs were not genotyped directly and could not be imputed with acceptable quality scores (rs6725887, rs9818870, and rs9982601). Overall, about 35% of the individuals in whom 1 or more SNPs were not genotyped or imputed were removed from the genotype risk score analysis. For ease of result interpretation, all odds ratios were designated with the previously published nonrisk allele as the reference so that the odds ratio reflects the effect of the MI-associated risk allele. Association analyses were adjusted for study-specific covariates (eg, age, sex, and population structure; online-only Data Supplement Table S2).

Meta-analysis was performed using a variance-weighted approach assuming a random effect model (DerSimonian and Laird method),¹⁵ which takes heterogeneity of the genetic effects across studies into account. Between-study heterogeneity was estimated using Cochran Q statistic, which is the summation of the squared, weighted difference between the study-specific effect and the overall effect size (estimated by the fixed effect), as well as I² statistics, which represent the percent of total variation across studies due to heterogeneity after accounting for variability due to chance. All meta-analyses were performed using the metan module implemented in STATA 10 (StataCorp, College Station, TX). Analyses were carried out for total IS, by stroke subtype, and by age of stroke onset (≤50 years, 51–69 years, and ≥70 years).

With our total sample size (6865 cases), we had 80% power to detect odds ratios of 1.09≈1.14 for overall IS and odds ratios of 1.20–1.32 for major stroke subtypes (≈1000 cases) for variants with allele frequency 10–50% (Bonferroni-corrected α=0.0045=0.05/11 SNPs). In this report, nominal probability values were provided for all the analyses.

Results

Study Population Characteristics

This study included 6865 cases and 11 395 control subjects recruited from 9 studies across North America, Europe, Australia, and South Asia. Stroke cases ranged in age between 15–101 years. Among cases with subtype information available, the 3 major stroke subtypes, cardioembolic, large-artery (atherosclerotic), and small-vessel (lacunar), accounted for ≈20.9% (n=1216), 19.6% (n=1137), and 18.0% (n=1043) of the cases, respectively; the remaining cases were attributed to either “other known causes” (5.6%; n=324) or “undetermined causes” (35.9%; n=2084). The majority of the study participants were of European ancestry with the exception of those enrolled in the RACE study (1890 cases and 4625 control subjects), which has enrolled participants of South Asian origin. Detailed characteristics of the participating studies are provided in online-only Data Supplement Table S3.

Associations of MI/CAD SNPs and Overall IS Risk

In the total sample, the meta-analysis odds ratios for all 11 SNPs ranged from 0.92–1.11 (Table 2). Only 2 SNPs were nominally associated with IS: rs11206510 at PCSK9 (odds ratio [OR], 0.92; 95% confidence interval [CI], 0.86–0.99; P=0.03; allele T) and rs3184504 at SH2B3 (OR, 1.11; 95% CI, 1.01–1.21; P=0.03; allele T). This association would not have withstood a conservative Bonferroni correction (α=0.05/11, or 0.0045). Although there was no significant evidence for heterogeneity of effect size across studies for rs11206510, the direction of effect was not consistent across studies with ORs ranging from 0.80 (RACE study) to 1.12 (CEDIR study) (Cochran Q statistic=8.1, P=0.4) (online-only Data Supplement Figure S1). rs3184504, on the other hand, showed significance evidence for heterogeneity (Cochran Q statistic=20.9, P=0.01) with 1 study (BRAINS) showing opposite effect as compared with other studies. We also examined the joint effect of these MI/CAD-associated SNPs by testing for association with the genotype scores and found no significant effect for either genotype score 1 or score 2 (OR, 1.02; P=0.2 for both scores). Removing the RACE study, the only study that had participants of South Asian origin, did not change results significantly although the association with rs11206510, and rs3184504 became less significant (OR, 0.94; P=0.10 and 1.10, P=0.06, respectively; online-only Data Supplement Table S4).

Table 2.

Association Results of the 11 MI/CAD Risk Loci and Risk Scores With Ischemic Stroke

SNP	MI/CAD Risk Allele	Heterogeneity Between Studies		Association With Overall Ischemic Stroke		Minimal OR Detectable in Current Study†
		Q (P)	I2, %	OR (95% CI)*	P
rs11206510	T	8.12 (0.42)	1.53	0.92 (0.86, 0.99)	0.03	1.13
rs646776	T	9.02 (0.34)	11.34	0.95 (0.89, 1.01)	0.12	1.13
rs17465637	C	3.09 (0.93)	0	1.03 (0.97, 1.09)	0.37	1.11
rs6725887	C	13.38 (0.06)	47.67	1.03 (0.90, 1.16)	0.69	1.13
rs9818870	T	5.45 (0.61)	0	1.03 (0.95, 1.11)	0.47	1.12
rs12526453	C	8.81 (0.36)	9.2	1.00 (0.95, 1.06)	0.92	1.10
rs4977574	G	4.17 (0.84)	0	1.02 (0.97, 1.07)	0.54	1.09
rs1746048	C	5.37 (0.72)	0	1.03 (0.96, 1.10)	0.38	1.14
rs3184504	T	20.87 (0.01)	61.67	1.11 (1.01, 1.21)	0.03	1.09
rs1122608	G	6.31 (0.61)	0	1.00 (0.95, 1.06)	0.92	1.11
rs9982601	T	6.20 (0.52)	0	1.00 (0.93, 1.09)	0.95	1.13
Score 1	NA	12.31 (0.09)	43.15	1.02 (0.99, 1.05)	0.23	n/a
Score 2	NA	14.44 (0.04)	51.52	1.02 (0.99, 1.06)	0.23	n/a

MI indicates myocardial infarction; CAD, coronary artery disease; SNP, single-nucleotide polymorphism; OR, odds ratio; CI, confidence interval; NA, not applicable; P, probability value.

^*ORs reflect the effect of the MI-associated risk allele with the previously published MI nonrisk allele as the reference allele defined in Table 1.

^†Minimal ORs that current sample size (6865 overall ischemic stroke cases) can detect with at least 80% power (α=0.0045).

We further stratified the associations between MI/CAD-associated SNPs and overall IS according to age of stroke onset: early (≤50 years), intermediate (51–69 years), and late onset (≥70 years). None of the associations were statistically significant in the age-stratified analyses, either (online-only Data Supplement Table S5).

Associations With Major IS Subtypes

We next examined the effects of MI/CAD-associated SNPs on TOAST-defined subtypes (Table 3). Nominally significant associations (0.03<P<0.05) were observed for 4 of the different TOAST subtypes. The MI/CAD-associated risk allele at rs9818870 in MRAS was associated with the cardioembolic stroke subtype (OR, 1.17; 95% CI, 1.02–1.35; P=0.03); the MI-associated risk allele at rs17465637 in MIA3 was associated with the small-vessel stroke subtype (OR, 1.12; 95% CI, 1.00–1.26; P=0.05); and the MI-associated risk allele at rs3184504 in SH2B3 was associated with strokes due to other known causes (OR, 1.33; 95% CI, 1.02–1.72; P=0.04). In addition, rs11206510 was associated with strokes due to other undetermined causes, although the allele conferring protection against MI was associated with stroke risk (OR, 0.86; 95% CI, 0.76–0.99; P=0.03). The ASGC study had a much higher OR and a wide CI for rs11206510 association, but this study contributed only 11 cases of undetermined causes in the analysis. Excluding the ASGC study from the meta-analysis resulted in an OR of 0.86 (95% CI, 0.76 – 0.97; P=0.01) for the association between rs11206510 and undetermined causes of stroke. The ANRIL rs4977574 SNP was very modestly (and not significantly) associated with the large-artery stroke subtype (OR, 1.09; 95% CI, 0.99–1.20; P=0.09), nor was it associated with any other stroke subtype. None of these results withstand correction for multiple testing. Forest plots showing the most strongly associated SNP with each subtype are in online-only Data Supplement Figure S2.

Table 3.

Association Results of the 11 MI/CAD Risk Loci and Risk Scores With Ischemic Stroke, by TOAST Subtypes

SNP	Cardioembolic (n=1216)		Large-Artery (n=1137)		Small-Vessel (n=1043)		Other Known Causes (n=324)		Undetermined Causes (n=2087)
	OR (95% CI)*	P	OR (95% CI)*	P	OR (95% CI)*	P	OR (95% CI)*	P	OR (95% CI)*	P
rs11206510	1.02 (0.89, 1.18)	0.78	0.91 (0.80, 1.04)	0.17	1.05 (0.91, 1.21)	0.50	0.89 (0.72, 1.11)	0.32	0.86 (0.76, 0.99)	0.03
rs646776	0.93 (0.83, 1.05)	0.25	0.91 (0.81, 1.03)	0.12	0.95 (0.84, 1.07)	0.38	1.01 (0.78, 1.31)	0.94	0.93 (0.82, 1.05)	0.24
rs17465637	1.01 (0.89, 1.16)	0.86	0.98 (0.87, 1.09)	0.69	1.12 (1.00, 1.26)	0.05	0.96 (0.79, 1.17)	0.69	1.01 (0.92, 1.10)	0.90
rs6725887	0.98 (0.83, 1.16)	0.84	1.01 (0.85, 1.19)	0.93	0.91 (0.76, 1.09)	0.30	1.45 (0.93, 2.26)	0.10	0.96 (0.83, 1.11)	0.58
rs9818870	1.17 (1.02, 1.35)	0.03	0.97 (0.83, 1.13)	0.66	0.91 (0.77, 1.07)	0.24	1.32 (0.92, 1.89)	0.13	0.99 (0.87, 1.12)	0.90
rs12526453	1.06 (0.95, 1.18)	0.31	0.93 (0.83, 1.03)	0.15	0.90 (0.80, 1.02)	0.09	1.03 (0.85, 1.25)	0.74	1.00 (0.91, 1.09)	0.99
rs4977574	1.02 (0.89, 1.17)	0.82	1.09 (0.99, 1.20)	0.09	1.03 (0.94, 1.14)	0.52	0.91 (0.69, 1.18)	0.47	1.01 (0.94, 1.09)	0.72
rs1746048	0.99 (0.88, 1.12)	0.93	1.05 (0.92, 1.2)	0.44	1.01 (0.86, 1.18)	0.94	0.97 (0.70, 1.33)	0.85	1.00 (0.91, 1.1)	0.96
rs3184504	1.05 (0.91, 1.22)	0.51	1.16 (0.96, 1.39)	0.11	1.03 (0.89, 1.20)	0.68	1.33 (1.02, 1.72)	0.04	1.12 (0.98, 1.27)	0.11
rs1122608	0.98 (0.88, 1.09)	0.67	1.06 (0.95, 1.18)	0.33	1.03 (0.92, 1.15)	0.66	1.03 (0.85, 1.26)	0.75	0.99 (0.91, 1.07)	0.74
rs9982601	1.05 (0.89, 1.25)	0.55	1.04 (0.89, 1.22)	0.63	1.06 (0.88, 1.27)	0.55	1.15 (0.83, 1.60)	0.39	1.02 (0.85, 1.22)	0.85
score 1	1.02 (0.98, 1.07)	0.29	1.02 (0.97, 1.06)	0.47	1.01 (0.95, 1.07)	0.85	1.06 (0.98, 1.14)	0.16	1.00 (0.96, 1.04)	0.97
score 2	1.04 (0.99, 1.08)	0.11	1.01 (0.96, 1.08)	0.63	1.00 (0.94, 1.07)	0.91	1.08 (1.01, 1.17)	0.03	1.00 (0.96, 1.05)	0.92

MI indicates myocardial infarction; CAD, coronary artery disease; TOAST, Trial of Org 10172 in Acute Stroke Treatment; SNP, single-nucleotide polymorphism; OR, odds ratio; CI, confidence interval; P, probability value.

^*ORs reflect the effect of the MI-associated risk allele with the previously published MI nonrisk allele as the reference allele defined in Table 1.

Discussion

It is of interest to consider whether common loci influence susceptibility to MI/CAD and stroke to the extent that such loci could elucidate common genetic and/or biological pathways. Given that the 11SNPs tested in this study were associated with MI/CAD in previous GWA studies, they probably are among the common SNPs having the largest effect sizes. The risk alleles at these loci were associated with increases of 12–20% in MI risk, with the exception of rs4977574 in ANRIL, for which the risk allele was associated with a 28% increase in Europeans.^7,8,13 In contrast, none of the MI/CAD risk alleles at these loci were significantly associated with risk of overall IS, with the estimated odds ratios ranging from 0.92–1.11 (ie, 8% reduction to 11% increase in odds of stroke per allele).

It is possible that failure to detect associations of these SNPs with the combined IS end point might be partially attributable to the heterogeneity of the IS phenotype. The large size of our sample provided us with the opportunity to evaluate the association of these SNPs to subtypes of ischemic stroke. Similar to the combined IS end point, we did not find any significant associations of MI/CAD SNPs (or the joint genotyping risk scores) with any of the stroke subtypes. Given that this study included more than 6800 total cases and more than 1000 cases for the major stroke subtypes, we have sufficient power to detect even small effects on overall stroke although more modest power to detect associations with stroke subtype. Therefore, our findings suggest that these MI/CAD-associated SNPs do not have effects of similar magnitude on overall stroke as they do on MI/CAD. If an association does exist, it is likely to be very modest for overall IS or restricted to a single stroke subtype.

Although the SNP associations with stroke subtypes did not reach statistical significance, it is interesting that the strongest associations we observed in the subtype analysis were genes involved in cell regulation or signaling. For example, the strongest association with cardioembolic stroke was the MI-risk allele at rs9818870 in MRAS (OR, 1.17; P=0.03). MRAS, an intracellular signal transducer, is highly expressed in the cardiovascular system and may be involved in cell adhesion.^7,8,16 The strongest association with small-vessel stroke was the MI-risk allele at rs17465637 in MIA3 (OR, 1.12; P=0.05), which encodes melanoma inhibitory activity 3 and is associated with cell adhesion and migration in inflammatory pathways¹⁷ It should also be noted that rs4977574 near ANRIL was very modestly associated with the large-artery subtype (OR, 1.09; P=0.09). Although the association did not achieve statistical significance in our study, other SNPs in the ANRIL region on chromosome 9p21 have previously been associated with large-artery stroke, with the strongest signal found with SNP rs1537378 (OR, 1.21; 95% CI, 1.09–1.35; P=0.0005).¹⁴ Interestingly, rs1537378 is not in high LD with our target SNP rs4977574 (r²=0.5), suggesting these 2 SNPs may represent different association signals or each of them may be tagging the causal variant imperfectly (assuming the association is indeed true). ANRIL encodes an antisense noncoding RNA that may disrupt other genes in the region, including CDKN2A and CDKN2B, 2 genes that play a key role in regulating cell proliferation, senescence, and apoptosis. We also observed the MI-associated risk allele at rs3184504 in SH2B3 to be nominally significantly associated with strokes due to other known causes (OR, 1.33; 95% CI, 1.02–1.72; P=0.04). It has been hypothesized that this SNP could reduce the anti-inflammatory activity of SH2B3, which is expressed in the vasculature, thereby leading to the development of plaque development and/or progression in coronary arteries.¹³

In contrast, SNPs whose effects on MI are likely to be mediated though lipid metabolism, for example, rs1122608 at SMARCA4/LDLR (19p13), rs646776 at the CELSR2/PSRC1/SORT1 locus (1p13), and rs11206510 at PCSK9 (1p32), have minimal effects on IS or any of its subtypes, and in some cases, even showed opposite effects on stroke. For example, the MI/CAD-associated risk allele of rs11206510 at PCSK9, a gene involved in cholesterol homeostasis, was modestly associated with decreased risk of overall stroke (OR, 0.92; P=0.03) or strokes due to undetermined causes (OR, 0.86; P=0.03). Our analyses suggest that genes involved in lipid metabolism do not predispose individuals to increased stroke risk as they do for MI, although replications will be needed to confirm these associations.

Taken together, our results suggest that the major common loci associated with MI/CAD risk do not have large effects on stroke. However, although our analyses did not reveal any significant associations between MI/CAD SNPs and stroke, the chromosomal regions harboring these loci might still contain stroke-related SNPs that are not in high LD with the MI-associated SNPs, as the example of chromosome 9p region.¹⁴ A limitation of our meta-analysis is that all included studies are case-control–based, and some have included recurrent stroke, thus allowing for the possibility of selection toward milder strokes (due to survival bias). It is possible that these genetic MI/CAD variants could be associated with more severe forms of stroke, but longitudinal studies that follow healthy cohorts for stroke occurrence prospectively would be needed to address this issue. Finally, our analyses were based on 11 SNPs unequivocally associated with MI and/or CAD identified from the first wave of meta-analyses for these traits. A second wave of even larger meta-analyses of these traits is in progress that will probably generate additional associations. We can expect most of these newly associated SNPs to have even smaller effect sizes on MI and CAD. In fact, in September 2011, 17 additional CAD-associated SNPs were identified by even larger meta-analyses, and the reported effect sizes for these newly identified SNPs, as expected, were smaller (ranging from 1.06–1.17) than the original 11 SNPs used in the present analysis.^18–21 However, this does not necessarily mean that these SNPs will have smaller effect sizes on stroke. Indeed, it is very possible that such SNPs may be in gene pathways that are more relevant to stroke than are the pathways associated with the currently known MI SNPs. Thus, it will be important to monitor new SNPs identified, even those with small effect sizes, as they may generate new biological insights into the etiology of stroke.

Conclusions

Common variants previously associated with MI/CAD risk do not have effects of similar magnitude on overall IS but do not preclude moderate associations restricted to specific IS subtypes.