Association Between Two Functional Fibrinogen-Related Polymorphisms and Ischemic Stroke: A Case-Control Study

Abstract

Objectives: Currently, there is a debate regarding the roles of two functional fibrinogen-related variants (rs6050 and rs1800790) and ischemic stroke (IS). Materials and Methods: A total of 1402 subjects (834 cases and 568 controls) were genotyped for single-nucleotide polymorphisms rs6050 and rs1800790 with the ligation detection reaction method. Results: We found that the homozygous minor allele genotype (GG) of rs6050 significantly increased IS risk by 66%, whereas that of rs1800790 reduced risk by 59% (rs6050: odds ratio [OR]=1.660, 95% confidence intervals [CI]: 1.141-2.415, p=0.008; rs1800790: OR=0.413, 95% CI: 0.228-0.747, p=0.003). After stratifying IS by three common subtypes, consistent results were found in IS cases with large-artery atherosclerosis (rs6050: OR=2.116, 95% CI: 1.327-3.376, p=0.002; rs1800790: OR=0.191, 95% CI: 0.085-0.430, p=0.000), and we also observed that the homozygous minor allele genotype of rs6050 increased risk by 86% in IS cases with cardioembolism (OR=1.859, 95% CI: 1.243-2.782, p=0.003). However, a paradox of association was shown between these two sites and fibrinogen levels. In haplotype analysis, we found that those with haplotype AA (major allele of rs6050 and minor allele of rs1800790) could reduce susceptibility to IS by 35% (OR=0.650, 95% CI: 0.493-0.858, p=0.002). Conclusion: Our study suggested that minor allele G of rs6050 is a significant risk factor, which can increase IS risk in the Chinese Han population, while minor allele A of rs1800790 and the haplotype AA lower such risk.

Introduction

Stroke is one of the leading causes of mortality worldwide—about 5.7 million patients died of stroke in 2005 (Roger et al., 2011). As a chronic disease, stroke has brought heavy financial burden to patients, especially in low-income and middle-income countries (Reddy, 2004; Strong et al., 2007). Similar to other countries, ischemic stroke (IS) accounts for 43-79% of all strokes in China, and patients tend to be younger than those in western countries (Reddy, 2004; Liu et al., 2011).

It is believed that the increased level of plasma fibrinogen plays a very important role in the pathogenesis of IS and other thrombotic diseases (Wilhelmsen et al., 1984; Meade et al., 1986; Kannel et al., 1987). Fibrinogen is a key protein that has effects in the pathway of a coagulation cascade and is associated with the formation of blood clots. The structure of the fibrinogen molecule consists of two pairs of disulfide-bridged A^α, B^β, and γ-polypeptide chains, which are encoded by the FGA, FGB, and FGG genes on the long arm of chromosome 4 (Mosesson, 2005; Reiner et al., 2006). Although the transcription and translation process of the three genes are independent, they can impact on each other: increase in the expression of one gene leads to elevation of the other two (Fuller and Zhang, 2001; Duan and Simpson-Haidaris, 2003). Previous studies have indicated that the nonsynonymous single-nucleotide polymorphism (SNP) Thr312Ala (rs6050) lies in the FGA gene and relates to the coagulation factor XIII (FXIII)-dependent crosslinking process. This site can affect the stability of the fibrin clot structure, thereby increasing susceptibility for embolization and risk of IS (Carter et al., 2000). Another variant −455 G/A (rs1800790), which is located in the promoter region of the FGB gene, also associates with changes in fibrinogen levels (Reiner et al., 2006). However, recent studies found paradoxical associations among these two SNPs, plasma fibrinogen levels, and thrombotic disease, especially IS (Carter et al., 1999; Cushman et al., 2007; Gohil et al., 2009; Siegerink et al., 2009).

Therefore, we conducted a case-control study in the Chinese Han population, to explore the association between rs6050, rs1800790, and IS.

Materials and Methods

Study design

In our study, we included 834 cases and 568 unrelated controls. These subjects were consecutively recruited in the West China Hospital (Sichuan University, Sichuan, China), between April 2009 and September 2011. We included patients who had a medical history of IS and focal neurologic deficits, and had their cerebral infarction confirmed by CT or MRI image analyses, following the diagnosis criteria of the World Health Organization for ischemic stroke (WHO, 1989). We excluded patients with any of the following diagnoses: subarachnoid hemorrhage, intracerebral hemorrhage, cerebral vascular malformation, brain tumors, or stroke of undetermined etiology (Adams et al., 1993). Subjects were selected as unrelated controls if they had CT/MRI results without brain infarction or no history of IS, which was confirmed by clinical examinations. Each subject provided written informed consent and described themselves to be of ethnic Han origin. In addition, our study was approved by the Ethics Committee of Sichuan University.

Among all 834 IS cases and 538 unrelated controls, we analyzed available plasma fibrinogen level data in 793 IS cases and 270 controls. After subjects were admitted to the hospital, these data were measured and derived from the prothrombin time and detected by a CA-7000 fully automatic coagulation instrument.

Moreover, we collected other information such as age, gender, hypertension, diabetes mellitus, coronary heart disease, atrial fibrillation, hypercholesterolemia, hypertriglycemia, and history of smoking and drinking. Hypertension was diagnosed as systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg, or ongoing therapy of antihypertensive drugs. Diabetes was diagnosed as a fasting plasma glucose level ≥7.0 mM or ongoing therapy of diabetes. CHD was diagnosed if they had stable/unstable angina, MI (electrocardiographic changes or elevated cardiac enzymes), angiographic evidence of 50% stenosis of at least one major epicardial vessel, and/or a history of known CHD. Atrial fibrillation was established on clinical examination and electrocardiogram. Hypercholesterolemia was diagnosed as >200 mg/dL, and hypertriglycemia was diagnosed as >150 mg/dL (Hu and Ding, 2008). Smoking and drinking were recorded by self description.

Genotype analyses

We collected EDTA-anticoagulated venous blood samples from all 1402 subjects, and extracted human genomic DNA from peripheral blood lymphocytes by using the AxyPrep Blood Genomic DNA Miniprep Kit (Axygen, Inc.).

Genotyping for SNP rs6050 and rs1800790 was performed by the ligation detection reaction (Shanghai Biowing Applied Biotechnology Company; www.biowing.com.cn/) (Thomas et al., 2004; Yi et al., 2010). A 241 bp fragment of rs6050 was amplified with the forward primer 5-ACCGGATCAGAGACGGAAAG-3 and the reverse primer 5-CCGGTACTACCAGGTCTAGGG-3, while a 223 bp fragment of rs1800790 was amplified with the forward primer 5-GGGTCTTTCTGATGTGTATTTTTCA-3 and the reverse primer 5-GACCTACTCACAAGGCAACCA-3. The target DNA sequences were amplified by a multiplex PCR method, and PCRs for each subject were carried out in a final volume of 20 μL containing 1×PCR buffer, 3.0 mM MgCl₂, 2.0 mM deoxynucleotide triphosphates, 0.4 μL primers, 0.3 μL Qiagen HotStarTaq Polymerase (Qiagen), 4 μL of 1×Q-solution, and 50 ng genomic DNA. Thermal cycling was performed in a Gene Amp PCR system 9600 (PerkinElmer) with an initial denaturation of 15 min at 95°C, followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 59°C for 1 min 30 s, and extension at 72°C for 1 min 30 s, followed by a final extension at 72°C for 7 min. The ligation reaction was carried out in a final volume of 10 μL containing 1×NEB Taq DNA ligase buffer, 12.5 pmol of each probe mix, 0.05 μL Taq DNA ligase (NEB Biotechnology), and 1 μL of multi-PCR product. A total of 35 cycles for ligase detection reaction was performed using 35 cycles at 95°C for 2 min, 94°C for 30 s, and 50°C for 2 min. Finally, the fluorescent products of the ligase detection reaction were differentiated by an ABI sequencer 377 (ABI).

Statistical analysis

For each SNP, there is one major allele A and minor allele B, an additive model was calculated appropriately based on comparing AB or BB to AA. A dominant model was assessed on AB+BB versus AA. We analyzed data with the statistical analysis software package SPSS 18.0 (SPSS, Inc.). For genotypic association, p-values, odds ratios (OR), and 95% confidence intervals (CI) were computed with Pearson and contingency tables. We also performed multivariate logistic regression analysis to adjust for risk factors (age, gender, smoking, hypertension, diabetes mellitus, coronary heart diseases, atrial fibrillation, hypercholesterolemia, and hypertriglyceridemia), and we used analysis of variance to calculate plasma fibrinogen levels in different SNP genotypes. The Bonferroni method was used to adjust for multiple testing. The PHASE 2.1 program was used to calculate haplotype frequencies based on genotypes, and SHEsis was used for the pairwise linkage disequilibrium (LD) structure and LD test. Power analysis was performed using the software NCSS-PASS 11 (D'Agostino et al., 1988).

Results

Study population

We included a total of 834 cases and 568 controls, and the power values for rs6050 and rs1800790 in our study were 0.9513 and 0.8265, respectively. Both of the two sites were in Hardy-Weinberg equilibrium (HWE) (For rs6050: control p=0.824; For rs1800790: control p=0.052). General characteristics of IS cases and controls are shown in Table 1. The average ages of cases and controls were 59.2 and 55.2 years. Female accounted for 36.6% and 43.7% of cases and controls, respectively. The number of patients who had a history of smoking or other risk factors was significantly larger than controls. Among all the patients who had hypertension, 59.2% (300/507) took antihypertensive drugs before admission.

Table 1.

General Characteristics of the Ischemic Stroke Case-Control Study

	IS case (n=834)	Control (n=568)	p-Value
Age, year^a	59.2±11.2	55.2±10.2	<10⁻³
Female, n (%)^a	305 (36.6)	248 (43.7)	0.008
Smoking, n (%)^a	312 (37.4)	165 (29.0)	0.001
Hypertension, n (%)^a	507 (60.8)	72 (12.7)	<10⁻³
DM, n (%)^a	225 (27.0)	39 (6.9)	<10⁻³
CHD, n (%)^a	56 (6.7)	1 (0.2)	<10⁻³
Atrial fibrillation, n (%)^a	100 (12.0)	1 (0.2)	<10⁻³
Hypercholesterolemia, n (%)	113 (13.5)	83 (14.6)	0.573
Hypertriglyceridemia, n (%)^a	256 (30.7)	146 (25.7)	0.042

Data are shown as mean±SD for quantitative variables and n (%) for qualitative variables.

p<0.05.

CHD, coronary heart disease; DM, diabetes mellitus; IS, ischemic stroke.

Analysis of plasma fibrinogen levels in different genotypes

Data of plasma fibrinogen levels were available in a total of 793 cases and 270 controls. Regarding rs6050, only IS cases with homozygous minor allele genotypes had significantly lower fibrinogen levels compared to the heterozygous minor allele genotypes (p=0.032). We observed the opposite results in rs1800790, which indicated that all subjects with the homozygous minor allele genotype had significantly higher fibrinogen levels (Table 2).

Table 2.

Plasma Fibrinogen Levels Per Genotype of rs6050 and rs1800790 in Ischemic Stroke Cases and Controls

Group		Genotype	IS, no. (%)	Fibrinogen level (g/L)	Difference (g/L)	p-Value
IS cases	FGA Thr312Ala (rs6050)	Homozygote major allele (AA)	237 (29.6)	3.13±0.85	Reference	Reference
		Heterozygote (AG)	378 (47.7)	3.19±0.88	0.06	0.430
		Homozygote minor allele (GG)	180 (22.7)	2.94±0.91	−0.19	0.032^a
	FGB −455G/A (rs1800790)	Homozygote major allele (GG)	490 (61.8)	3.04±0.86	Reference	Reference
		Heterozygote (GA)	272 (34.3)	3.23±0.90	0.19	0.003^a
		Homozygote minor allele (AA)	31 (3.9)	3.39±0.84	0.34	0.031^a
Controls	FGA Thr312Ala (rs6050)	Homozygote major allele (AA)	94 (34.8)	3.26±1.04	Reference	Reference
		Heterozygote (AG)	127 (47.0)	3.10±0.95	−0.16	0.260
		Homozygote minor allele (GG)	49 (18.1)	3.06±1.16	−0.20	0.274
	FGB −455G/A (rs1800790)	Homozygote major allele (GG)	153 (56.7)	3.08±1.10	Reference	Reference
		Heterozygote(GA)	91 (33.7)	3.15±0.90	0.07	0.597
		Homozygote minor allele (AA)	26 (9.6)	3.54±0.84	0.46	0.032^a

Data are shown as mean±SD for quantitative variables and n (%) for qualitative variables.

p<0.05, p-values were calculated by analysis of variance.

Genotypic associations between rs6050, rs1800790, and IS

Genotype distributions of the two SNPs in controls were in accordance with data in HapMap. By using the homozygous major allele genotype as reference, we found that IS cases with homozygote minor allele genotype of rs6050 increased the risk of IS by 66%, while those of rs1800790 decreased the risk by 59% (rs6050: OR=1.660, 95% CI: 1.141-2.415, p=0.008; rs1800790: OR=0.413, 95% CI: 0.228-0.747, p=0.003), after Bonferroni correction (p<0.025, 0.05/2), the genotype frequency for rs6050 and rs1800790 was still significantly different between cases and controls. After stratifying all cases by three common subtypes of IS, and adjusting for significant risk factors and performing Bonferroni correction (p<0.025, 0.05/2), we observed that having a homozygous minor allele genotype of rs6050 in those with large-artery atherosclerosis could double the risk of IS, while those of rs1800790 decreased the risk by 81% (TOAST=1, rs6050: OR=2.116, 95% CI: 1.327-3.376, p=0.002; rs1800790: OR=0.191, 95% CI: 0.085-0.430, p=0.000) (Table 3). However, we did not observe any significant association of genotypes in other subtypes of IS, except that the homozygous minor allele genotype of rs6050 could increase the risk of IS by 86% in those with cardioembolism (OR=1.859, 95% CI: 1.243-2.782, p=0.003).

Table 3.

Analysis of Genotypic Association of Single-Nucleotide Polymorphism rs6050 and rs1800790 with Ischemic Stroke

		Dominant ^a		Additive 1 ^a		Additive 2 ^a
Group	Subtype ^a	Adjusted p-value^b	OR (95% CI)	Adjusted p-value	OR (95% CI)	Adjusted p-value	OR (95% CI)
rs6050	TOAST=1	0.037^c	1.427 (1.022-1.992)	0.011^c	1.644 (1.121-2.413)	0.002^c	2.116 (1.327-3.376)
	TOAST=2	0.123	2.226 (0.806-6.151)	0.141	1.281 (0.921-1.781)	0.003^c	1.859 (1.243-2.782)
	TOAST=3	0.619	1.105 (0.745-1.638)	0.821	1.051 (0.681-1.623)	0.230	1.400 (0.808-2.426)
	Overall IS	0.097	1.277 (0.957-1.703)	0.388	1.144 (0.842-1.555)	0.008^c	1.660 (1.141-2.415)
rs1800790	TOAST=1	0.010^c	0.665 (0.488-0.907)	0.062	0.718 (0.507-1.017)	0.000^c	0.191 (0.085-0.430)
	TOAST=2	0.648	1.206 (0.540-2.695)	0.413	0.689 (0.283-1.680)	0.126	0.047 (0.001-2.361)
	TOAST=3	0.158	0.763 (0.524-1.111)	0.179	0.754 (0.499-1.138)	0.098	0.512 (0.231-1.132)
	Overall IS	0.017^c	0.720 (0.550-0.943)	0.099	0.788 (0.594-1.046)	0.003^c	0.413 (0.228-0.747)

In subtypes of IS, TOAST=1 indicates large-artery atherosclerosis, TOAST=2 indicates cardioembolism, TOAST=3 indicates small-vessel occlusion. Dominant model was assessed on AB+BB versus AA (the homozygous major allele genotype). Additive 1 indicates AB versus AA, and Additive 2 indicates BB versus AA.

Adjusted p-value by multivariate logistic regression analysis for potential confounders, including age, gender, smoking, hypertension, diabetes mellitus, and coronary heart diseases, atrial fibrillation, hypercholesterolemia, hypertriglyceridemia.

p<0.05.

CI, confidence interval; OR, odds ratio.

Haplotype analysis

It is shown that the LD D′ value for rs6050 and rs1800790 was close to 0.9 (0.896), suggesting a strong recombination event (Gabriel et al., 2002). We found the haplotype AA (major allele of rs6050 and minor allele of rs1800790) could reduce the risk by 35% in overall IS (OR=0.650, 95% CI: 0.493-0.858, p=0.002), but haplotype GG (minor allele of rs6050 and major allele of rs1800790) was not significantly associated with the disease, by using PHASE 2.1 to build three haplotypes and setting the most frequent haplotype (AG) as the reference group and adjusting for risk factors and Bonferroni correction (p<0.025, 0.05/2) (Table 4).

Table 4.

Haplotype Frequencies in Ischemic Stroke Cases and Controls

SNP ^a	Haplotype ^b	IS, n (%)	Controls, n (%)	Adjusted p-value^c	Adjusted OR (95% CI)
1,2	AG	559 (33.5)	363 (32.0)	Reference^d	NA
	AA	332 (19.9)	279 (24.6)	0.002^e	0.650 (0.493-0.858)
	GG	765 (45.9)	486 (42.8)	0.253	1.143 (0.909-1.438)

SNP: 1=rs6050, 2=rs1800790.

Only haplotypes with a frequency of p>0.05 were considered in the analysis. A indicates the major allele of rs6050, G indicates the minor allele of rs6050, G indicates the major allele of rs1800790, A indicates the minor allele of rs1800790.

The most frequent haplotype (AG) was set as the reference group.

p<0.05.

NA, not available; SNP, single-nucleotide polymorphism.

Discussion

In our study, we analyzed the association between two SNPs in the fibrinogen module and IS in a Chinese Han population. Using the homozygous major allele genotype as a reference, our results showed that the homozygous minor allele genotype (GG) of the first site rs6050 could increase the risk of IS 66%, especially in large-artery atherosclerosis. This is similar to previous findings, in which Standeven et al. (2003) carried out a functional trial and proved that rs6050 is prone to clot embolization and elevated IS risk. As for the second site rs1800790, we found the homozygous minor allele genotype (AA) of rs1800790 could reduce 59% IS risk, and haplotype analysis showed haplotype AA (major allele of rs6050 and minor allele of rs1800790) could reduce 35% IS risk; some other studies also showed this SNP could reduce the risk in thrombotic diseases, for example, a large-scale meta-analysis showed rs1800790 had significantly protective effects in venous thromboembolism (OR=0.84, 95% CI: 0.72-0.97, p=0.02) (Cushman et al., 2007; Jood et al., 2008; Gohil et al., 2009). Interestingly, Siegerink et al. (2009) in another IS study explored these two sites and came to a conclusion that rs6050 decreased the risk of IS, while rs1800790 increased it, which is opposite to our results. However, their study only investigated these two sites in young women (<50 years old) in the Netherlands, and their controls were not in HWE (p=0.043). Therefore, we considered that selection biases (especially population bias), might have some influences on their results. Another aspect of the different results might be explained by possible genetic variation among individuals from different ethnic or geographic origins.

To some extent, it was perplexing as to why there is the paradox of association between these two sites and fibrinogen levels in our study: the minor allele of rs6050 or the major allele of rs1800790 may lead to lower fibrinogen levels in subjects, which is opposite to the association between these two sites and IS risk. Interestingly, it was worth noting that this kind of results also appeared in some deep vein thrombosis studies; therefore, we consider that these two sites may have influenced IS using another mechanism rather than making changes in fibrinogen levels only (Cushman et al., 2007). Still, there may be some other factors that influence the plasma fibrinogen levels. For example, due to the limited conditions of our hospital, it is a pity that the data of fibrinogen are temporarily unavailable after patients' discharge from hospital, so an acute phase reaction might influence fibrinogen levels since we obtained patients' data after onset. In our future studies, we will make efforts to collect fibrinogen levels at different time points in the course of disease, and rule out other confounding factors of fibrinogen. Another limitation of our study is that some differences exist in age and gender between IS cases and controls (Table 1). To solve this problem, we carried out logistical regression to adjust for these two factors to minimize their impact on our results, and we found the differences were no longer statistically significant after adjustment.

Our study provided evidence to support that the minor allele G of rs6050 is a significant risk factor, while the minor allele A of rs1800790 and the haplotype AA can lower the risk for IS in a Chinese Han population. Future studies are needed to clarify the detailed functions of these two variants and their associated genes in fibrinogen levels, as well as their interaction with other causal fibrinogen variants in different subtypes of IS.

Footnotes

Acknowledgments

We would like to thank all subjects who took part in this study.

Author Disclosure Statement

No competing financial interests exist.

References

Adams

Jr. , Bendixen

, Kappelle

et al. 1993. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke, 24:35-41.

Carter

, Catto

, Grant

. 1999. Association of the alpha-fibrinogen Thr312Ala polymorphism with poststroke mortality in subjects with atrial fibrillation. Circulation, 99:2423-2426.

Carter

, Catto

, Kohler

et al. 2000. Alpha-fibrinogen Thr312Ala polymorphism and venous thromboembolism. Blood, 96:1177-1179.

Cushman

, Cornell

, Folsom

et al. 2007. Associations of the beta-fibrinogen Hae III and factor XIII Val34Leu gene variants with venous thrombosis. Thromb Res, 121:339-345.

D'Agostino

, Chase

, Belanger

. 1988. The appropriateness of some common procedures for testing the equality of two independent binomial populations. Am Stat, 42:198-202.

Duan

, Simpson-Haidaris

. 2003. Functional analysis of interleukin 6 response elements (IL-6REs) on the human gamma-fibrinogen promoter: binding of hepatic Stat3 correlates negatively with transactivation potential of type II IL-6REs. J Biol Chem, 278:41270-41281.

Fuller

, Zhang

. 2001. Transcriptional control mechanism of fibrinogen gene expression. Ann N Y Acad Sci, 936:469-479.

Gabriel

, Schaffner

, Nguyen

et al. 2002. The structure of haplotype blocks in the human genome. Science, 296:2225-2229.

Gohil

, Peck

, Sharma

. 2009. The genetics of venous thromboembolism. A meta-analysis involving approximately 120,000 cases and 180,000 controls. Thromb Haemost, 102:360-370.

10.

, Ding

. 2008. [Guidelines for management of adult dyslipidemia in China] Zhonghua Nei Ke Za Zhi, 47:723-724.

11.

Jood

, Danielson

, Ladenvall

et al. 2008. Fibrinogen gene variation and ischemic stroke. J Thromb Haemost, 6:897-904.

12.

Kannel

, Wolf

, Castelli

, D'Agostino

. 1987. Fibrinogen and risk of cardiovascular disease. The Framingham Study. JAMA, 258:1183-1186.

13.

Liu

, Wang

, Wong

, Wang

. 2011. Stroke and stroke care in China: huge burden, significant workload, and a national priority. Stroke, 42:3651-3654.

14.

Meade

, Mellows

, Brozovic

et al. 1986. Haemostatic function and ischaemic heart disease: principal results of the Northwick Park Heart Study. Lancet, 2:533-537.

15.

Mosesson

. 2005. Fibrinogen and fibrin structure and functions. J Thromb Haemost, 3:1894-1904.

16.

Reddy

. 2004. Cardiovascular disease in non-Western countries. N Engl J Med, 350:2438-2440.

17.

Reiner

, Carty

, Carlson

et al. 2006. Association between patterns of nucleotide variation across the three fibrinogen genes and plasma fibrinogen levels: the Coronary Artery Risk Development in Young Adults (CARDIA) study. J Thromb Haemost, 4:1279-1287.

18.

Roger

, Go

, Lloyd-Jones

et al. 2011. Heart disease and stroke statistics—2011 update: a report from the American Heart Association. Circulation, 123:e18-e209.

19.

Siegerink

, Rosendaal

, Algra

. 2009. Genetic variation in fibrinogen; its relationship to fibrinogen levels and the risk of myocardial infarction and ischemic stroke. J Thromb Haemost, 7:385-390.

20.

Standeven

, Grant

, Carter

et al. 2003. Functional analysis of the fibrinogen Aalpha Thr312Ala polymorphism: effects on fibrin structure and function. Circulation, 107:2326-2330.

21.

Strong

, Mathers

, Bonita

. 2007. Preventing stroke: saving lives around the world. Lancet Neurol, 6:182-187.

22.

Thomas

, Sinville

, Sutton

. 2004. Capillary and microelectrophoretic separations of ligase detection reaction products produced from low-abundant point mutations in genomic DNA. Electrophoresis, 25:1668-1677.

23.

WHO. 1989. Recommendations on stroke prevention, diagnosis, and therapy. Report of the WHO Task Force on Stroke and other Cerebrovascular Disorders. Stroke, 20:1407-1431.

24.

Wilhelmsen

, Svardsudd

, Korsan-Bengtsen

et al. 1984. Fibrinogen as a risk factor for stroke and myocardial infarction. N Engl J Med, 311:501-505.

25.

, Chen

, Yu

et al. 2010. Development of a PCR/LDR/capillary electrophoresis assay with potential for the detection of a beta-thalassemia fetal mutation in maternal plasma. J Matern Fetal Neonatal Med, 23:920-927.