A Common Polymorphism in the Methylenetetrahydrofolate Reductase Gene, Homocysteine, and Ischemic Cerebrovascular Disease
Background and Purpose A common polymorphism (T/t) in the gene encoding the methylenetetrahydrofolate reductase (MTHFR) enzyme has been associated with elevated serum homocysteine, itself a risk factor for stroke. Some studies have reported an association with ischemic heart disease, but no published studies have examined its relationship with stroke.
Methods We determined the TT genotype frequency and T allele frequency in 345 patients with ischemic cerebrovascular disease (CVD) and 161 control subjects. In a subgroup we also determined serum homocysteine and folate concentrations.
Results In the patient group there was a significant relationship between TT genotype and homocysteine concentration after we controlled for other risk factors. Controlling for serum folate weakened this relationship, and folate itself was independently related to serum homocysteine. There was no difference between patients and control subjects in either TT genotype frequency (10.7% versus 13.7%; P=.34) or T allele frequency (0.68 versus 0.67; P=.67). There was no association when analysis was limited to individuals deficient in folate (serum folate <25th centile) or to younger individuals (<65 years). There was no association between TT genotype and any stroke subtype or with degree of carotid stenosis.
Conclusions In patients with CVD we confirmed a relationship between the MTHFR genotype and serum homocysteine concentration and an interaction with serum folate concentration. We found no association between CVD and genotype. However, the interaction with serum folate suggests that the genotype could still be a risk factor in populations with a low folic acid intake.
Genetic factors are important in CVD.1 While the genetic basis of some relatively rare metabolic and coagulation disorders predisposing to stroke is understood, the molecular basis of the genetic predisposition in the majority of patients remains unknown. Such genetic influences may act either independently or by predisposing to or modulating the effect of known risk factors such as hypertension.
An important independent genetic risk factor may be serum homocysteine. Homocysteine is a thiol-containing amino acid derived from the metabolism of methionine. Homocysteine itself is metabolized by two pathways: catabolism by cystathionine β-synthase, a pyridoxal 5′-phosphate–dependent enzyme, or remethylation to form methionine, a reaction that in most tissues is dependent on the cofactor activity of folate and vitamin B12. Very high serum homocysteine concentrations (hyperhomocysteinemia) result in homocysteine excretion in the urine (homocysteinuria). This autosomal recessive condition is most commonly due to cystathionine β-synthase deficiency; such individuals present with ocular and skeletal abnormalities as well as premature vascular disease and venous thrombosis. Such cases of severe hyperhomocysteinemia are rare, and on a population basis they are not an important cause of stroke. However, several studies have shown that moderately raised homocysteine concentrations are common in adults with vascular disease,2 particularly CVD, and are an independent risk factor for stroke even in older individuals.3 4 Up to 30% of patients with premature cerebral and peripheral vascular disease have hyperhomocysteinemia. This suggests that there may be common genetic polymorphisms accounting for this moderate elevation that, on a population basis, could be an important risk factor for stroke. One such factor may be genetically determined abnormalities in the enzyme MTHFR.
MTHFR catalyzes the reduction of 5,10-methylenetetrahydrofolate to 5-methyltetrahydrofolate, the predominant circulating form of folate and carbon donor for the remethylation of homocysteine to methionine. A thermolabile form of MTHFR, associated with reduced enzyme activity, has been reported in 17% of patients with coronary artery disease5 and 7% of individuals with premature vascular disease.6 It appears to be predictive of coronary artery stenosis independent of other conventional risk factors.7 Recently, a common polymorphism in MTHFR, which results in this thermolability and occurs at a high allele frequency of approximately 0.38 (or 38%), has been identified.8 A C to T substitution at nucleotide 677 results in an alanine to valine substitution in the MTHFR protein. In a pilot study in normal volunteers, heterozygotes (Tt) had a mean of 65% of normal enzyme activity, whereas homozygotes (TT) had 30% enzyme activity. Individuals homozygous for the polymorphism but without clinical vascular disease had significantly elevated plasma homocysteine concentrations.8 The influence of genotype on homocysteine concentration is greater in individuals with low serum folate and B12 concentrations.9 In the last year conflicting results about the relationship between the MTHFR polymorphism and ischemic heart disease have been published.10 11 12 13 14 However, no published studies have determined the relationship of the MTHFR polymorphism with stroke. One abstract reported an association, but only 50 control subjects were studied, and the control TT genotype frequency was lower than reported in other studies.15 This study was performed with two aims: (1) to confirm that a relationship between MTHFR genotype and serum homocysteine is present in patients with CVD and to examine the effect of serum folate on this relationship and (2) to determine whether either the TT genotype or the T allele is an independent risk factor for CVD or any particular stroke subtype or increased carotid artery atheroma determined ultrasonically.
Subjects and Methods
Three hundred forty-five patients presenting with ischemic stroke or TIA to a neurological CVD service were studied and compared with 161 white control subjects. Patients with cerebral hemorrhage or cerebral venous thrombosis were not included. Control subjects were recruited from consecutive spouses of the same patients. Because of the marked variation in common vascular candidate gene polymorphisms in different ethnic groups,16 only white case and control subjects were studied. Control subjects were included if they had vascular risk factors or a history of myocardial infarction but were excluded if they had CVD (three case subjects).
CT and/or MR head imaging, extracranial duplex ultrasound, and electrocardiography were performed in all patients; echocardiography was performed in approximately 40%. MR angiography of the extracranial and intracranial vertebrobasilar system was performed in approximately 50% of cases of posterior circulation ischemia. On the basis of clinical features and results of the investigation, patients were divided into four subtypes: (1) large-vessel disease: internal carotid or vertebral artery stenosis >50% (diagnosed on carotid duplex for anterior circulation ischemia and carotid duplex and/or MR angiography for posterior circulation ischemia) with symptoms in that arterial territory; (2) lacunar stroke: a clinical lacunar syndrome with an appropriate CT infarct or a typical clinical syndrome lasting >24 hours17 and a normal CT scan; (3) uncertain or probable cardiac embolic source (these two categories were included together because not all patients had echocardiography, and transthoracic echocardiography does not detect all cardioembolic sources); and (4) tandem pathology: more than one cause of cerebral ischemia.
Hypertension was defined as either systolic blood pressure >160 mm Hg, diastolic pressure >95 mm Hg, or current treatment with antihypertensive drugs. A smoker was defined as a current smoker or ex-smoker. Carotid stenosis was measured with the use of internal carotid artery peak systolic velocity and end-diastolic velocity for stenoses >50% and with the use of B-mode imaging for stenoses with no velocity increase. The mean carotid stenosis was determined for each case.
We performed all molecular genetic and biochemical analyses blinded to case-control status. DNA was extracted from leukocytes with the use of a commercially available kit (Nucleon, Scotlab Ltd). The MTHFR C to T 677 substitution was identified with the use of restriction enzyme digestion of the polymerase chain reaction–amplified products, as previously described.8 5′-TGAAGGAGAA GGTGTCTGCG GCA-3′ (exonic) and 5′-AGGACGGTGC GGTGAGAGTG-3′ (intronic) primers were used; these generate a 198-bp fragment. The C to T substitution creates a HinfI recognition sequence, and digestion of the mutant product results in 175- and 23-bp fragments. The fragment size was determined by gel electrophoresis.
Serum homocysteine was measured in a subgroup. Samples were nonfasting. Homocysteine was measured in 160 case subjects and 75 control subjects by reverse-phase high-performance liquid chromatography with precolumn derivatization and fluorometric detection based on the method of Fiskerstrand et al.18 The coefficient of variation was <8%. We initially measured serum folate only in the same case and control subjects. However, after data suggested that the genotype might be a risk factor only in individuals with a folate deficiency, we measured folate in all individuals in whom we had available serum (227 case subjects and 113 control subjects). Folate was measured by an ion capture assay with the Abbott IMX analyzer; the coefficient of variation was 6%.
Differences between groups were examined with the use of the χ2 test or the unpaired Student’s t test when appropriate. Logistic regression analysis was used to determine the relationship between CVD and risk factors. Serum homocysteine and folate were logarithmically transformed to obtain normal distributions. The relationship between genotype and serum homocysteine was determined by means of ANOVA with Scheffé’s test for post hoc analysis. Multiple regression was used to determine the influence of other risk factors on this relationship. Significance was taken as P<.05.
Genotype and Serum Homocysteine
In the patient group there was a significant relationship between genotype and homocysteine concentration (Table 1⇓; P=.01, ANOVA). This relationship was still significant (P=.009) after we controlled for age, sex, hypertension, smoking, and diabetes. There was a significant positive relationship between homocysteine and age (P=.02) but no significant relationship between homocysteine and sex, smoking, hypertension, and diabetes. The effect of genotype on homocysteine concentration was most marked for patients with a homocysteine level in the top 5% and was progressively less for patients in the top 10% and top 25% of the serum homocysteine distribution (Table 2⇓).
There was a significant interaction between MTHFR genotype, serum homocysteine, and serum folate. When we considered only the 144 case subjects in whom both folate and homocysteine were measured, log homocysteine concentration was significantly related to TT genotype independent of age, sex, hypertension, smoking, and diabetes (P=.003). When log folate was entered into the multiple regression, the strength of the association between genotype and homocysteine was reduced (P=.01), and log folate was independently related to log homocysteine (P=.01). Within the 227 case subjects in whom folate was measured, there was no significant relationship between genotype and folate concentration, but there was a trend toward lower log folate in patients with the TT genotype: (mean [SD] log folate: TT, 0.75 [0.26]; Tt, 0.85 [0.24]; tt, 0.87 [0.24] μg/L; P=.1, ANOVA).
A trend toward similar relationships between genotype and homocysteine was found in the 75 control subjects in whom homocysteine concentrations were measured: (mean [SD] concentration: TT, 1.30 [0.24]; Tt, 1.31 [0.14]; tt, 1.23 [0.22] μmol/L; P=.22); however, because of the small numbers, particularly of the TT genotype (n=7), the power of the sample size to detect significant differences was small, and detailed analysis in this report is limited to the case subjects.
MTHFR Genotype and CVD
Genotype frequencies did not differ from the Hardy-Weinberg equilibrium in control subjects, case subjects, or the whole population. There was no difference in genotype frequency or allele frequency between cerebrovascular case subjects or control subjects (Table 3⇓). TT frequency was 10.7% in case subjects and 13.7% in control subjects (TT versus Tt+tt, P=.34). Individuals with TT genotype had an OR of CVD of 0.76 (95% CI, 0.43 to 1.33). The T allele frequency was 0.68 in case subjects and 0.67 in control subjects (P=.67). In contrast, the following were significant independent risk factors for CVD: hypertension (OR, 2.73; 95% CI, 1.81 to 4.14; P<.00001), smoking (OR, 2.80; 95% CI, 1.87 to 4.21; P<.00001), and diabetes (OR, 2.37; 95% CI, 1.13 to 4.96; P=.02).
There was no association between genotype and CVD when only the 157 case subjects and 79 control subjects aged ≥65 years were considered: TT frequency of 13.5% in case subjects and 15.2% in control subjects (P=.70); T-t allele ratio of 0.33:0.67 in case subjects and 0.33:0.67 in control subjects (P=.96). It is possible that the effect of genotype as a risk factor for stroke would only be seen in individuals deficient in folate. Therefore, the relationship between genotype and CVD was examined in individuals with low serum folate. When subjects with a serum folate <25th centile (4.6 g/L) were considered separately, there was no difference in TT frequency in case subjects versus control subjects: 12/71 (16.9%) versus 4/29 (13.8%); OR, 1.27 (95% CI, 0.37 to 4.3) (P=.70).
There was no difference in genotype distribution between subjects with stroke or with TIA without CT infarct (TT frequency, 30/271 (11.1%) versus 7/74 (9.5%); P=.69). There was no association between genotype and any particular stroke subtype (lacunar, large-vessel, or cardioembolic/unknown cause), as shown in Table 4⇓. There was no association between genotype and mean carotid stenosis (mean [SD] stenosis: TT, 27.6% [25.7]; Tt, 33.5% [28.6]; tt, 32.5% [27.8]; P=.51, ANOVA).
Log homocysteine was nonsignificantly higher in 160 case subjects in whom it was assayed than in 75 control subjects (mean [SD]: case subjects, 1.32 [0.19] μmol/L; control subjects, 1.27 [0.19] μmol/L; P=.09). Log folate was nonsignificantly lower in the 227 case subjects in whom it was measured than in the 113 control subjects (mean [SD]: case subjects, 0.85 [0.24] μg/L; control subjects, 0.90 [0.24] μg/L; P=.07).
This study confirms a significant relationship between the MTHFR genotype and serum homocysteine concentration in patients with CVD, as reported previously in normal populations. However, we found no relationship between the TT genotype or the T allele and risk of CVD in an unselected group of white patients with stroke and TIA. The gene frequencies we found in both case and control subjects are similar to those reported in previous studies.10 11 12 13 14 The relationship was unchanged when only younger patients and control subjects (≥65 years) were considered, and there was no relationship between genotype and either stroke subtype or degree of carotid stenosis.
The relationship we found between serum homocysteine and genotype in patients with CVD is similar to that reported in previous studies in normal control subjects and more recently patients with ischemic heart disease.8 9 14 19 Nonfasting homocysteine concentrations are higher and influenced to a greater extent by dietary intake than are fasting concentrations. However, despite measuring nonfasting concentrations in this study, we were still able to demonstrate a highly significant relationship between serum homocysteine and genotype, with higher levels in individuals homozygous for the polymorphism. We also found an interaction between the MTHFR genotype, serum homocysteine, and serum folate. Entering folate into the logistic regression weakened the association between genotype and homocysteine, while serum folate itself was independently related to serum homocysteine. Furthermore, there was a nonsignificant trend toward serum folate being lower in patients with the TT genotype. This interaction has been previously reported; individuals with the TT genotype have lower serum folate, particularly if the analysis is restricted to subjects in the lowest quartile of the folate distribution.9 14 Harmon et al19 found a highly significant relationship between genotype and homocysteine concentration in individuals whose serum folate was below the median value but no relationship in individuals in whom it was above the median. Moderate elevations of serum homocysteine in individuals with premature vascular disease and thermolabile MTHFR and suboptimal intake of folate can be corrected by folic acid supplementation.20 It has been suggested that the region in MTHFR enzyme relating to this common polymorphism is involved in folate binding and that the enzyme might be stabilized in the presence of folate.8 Therefore, the combination of the genetic defect and inadequate folate intake may cause elevated homocysteine and increase the risk of cardiovascular disease. A recent study has reported no association between the TT genotype and myocardial infarction, but when only individuals with a serum folate in the lowest quartile were considered, there was a nonsignificantly increased risk among individuals with the mutation.14 Therefore, we determined serum folate in all individuals in whom serum was available. Even in this subgroup we found no significant relationship between TT genotype and CVD, with an OR of 1.27. However, the interaction between genotype, homocysteine, and folate does raise the possibility that the genotype may still be a significant risk factor for stroke in populations with a very low folate intake. This interaction with folate, as well as a similar more recently reported interaction with vitamin B12,9 may at least partly explain the conflicting results of recent studies examining the relationship between the MTHFR polymorphism and ischemic heart disease.10 11 12 13
This study was designed to determine the relationship between the MTHFR genotype and CVD and not the previously described relationship between homocysteine and CVD. However, in the subgroup of our population in whom it was measured, we found nonsignificantly higher homocysteine levels in CVD patients. Despite this, there was not even a trend toward an increase in the TT genotype in CVD patients. This suggests that factors other than the TT genotype are important in determining much of the variance in homocysteine levels in our population. However, further studies are required in different populations to determine whether the TT genotype can be a risk factor for CVD, particularly in populations and individuals with low folate intake.
Selected Abbreviations and Acronyms
|TIA||=||transient ischemic attack|
This study was supported by a project grant from the Stroke Association.
Presented in part at the 3rd World Stroke Congress and 5th European Stroke Congress, Munich, Germany, September 1-4, 1996, and published in abstract form (Cerebrovasc Dis. 1996;6[suppl 2]:19).
- Received April 14, 1997.
- Revision received June 13, 1997.
- Accepted June 13, 1997.
- Copyright © 1997 by American Heart Association
Brass LM, Shaker LA. Family history in patients with transient ischemic attacks. Stroke. 1991;22:837-841.
Verhoef P, Hennekens CH, Malinow MR, Kok FJ, Willett WC, Stampfer MJ. A prospective study of plasma homocyst(e)ine and risk of ischemic stroke. Stroke. 1994;25:1924-1930.
Kang SS, Wong PW, Susmano A, Sora J, Norusis M, Ruggie N. Thermolabile methylenetetrahydrofolate reductase: an inherited risk factor for coronary artery disease. Am J Hum Genet. 1996;48:536-545.
Kang SS, Passen EL, Ruggie N, Wong PWK, Sora H. Thermolabile defect of methylenetetrahydrofolate reductase in coronary artery disease. Circulation. 1996;88:1463-1469.
Frosst P, Blom HJ, Milos R, Goyette P, Sheppard CA, Mathews RG, Boers GJ, den Heijer M, Kluijtmans LA, Van der Heuvel LP, Rozen R. A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase. Nat Genet. 1996;10:111-113.
Jacques PF, Bostom AG, Williams RR, Ellison RC, Eckfeldt JH, Rosenberg IH, Selhub J, Rozen R. Relation between folate status, a common mutation in methylenetetrahydrofolate reductase, and plasma homocysteine concentrations. Circulation. 1996;93:7-9.
Kluijtmans LA, van den Heuvel LP, Boers GH, Frosst P, Stevens EM, van Oost BA, den Heijer M, Trijbels FJ, Rozen R, Blom HJ. Molecular genetic analysis in mild hyperhomocysteinemia: a common mutation in the methylenetetrahydrofolate reductase gene is a genetic risk factor for cardiovascular disease. Am J Hum Genet. 1996;58:35-41.
Schmitz C, Lindpaintner K, Verhoef P, Gaziano JM, Buring J. Genetic polymorphism of methylenetetrahydrofolate reductase and myocardial infarction: a case-control study. Circulation. 1996;94:1812-1814.
Gallagher PM, Meleady R, Shields DC, Tan KS, McMaster DM, Rozen R, Evans A, Graham IM, Whitehead AS. Homocysteine and the risk of premature of coronary heart disease: evidence for a common gene mutation. Circulation. 1996;94:2154-2158.
Wilcken DEL, Wang XL, Sim AS, McCredie MM. Distribution in healthy and coronary populations of the methylenetetrahydrofolate reductase (MTHFR) C677T mutation. Arterioscler Thromb Vasc Biol. 1996;16:878-882.
Ma J, Stampfer MJ, Hennekens CH, Frosst P, Selhub J, Horsford J, Malinow MR, Willett WC, Rozen R. Methylenetetrahydrofolate reductase polymorphism, plasma folate, homocysteine, and the risk of myocardial infarction in US physicians. Circulation. 1996;94:2410-2416.
Press RD, Beamer N, Coull BM. A common mutation in methylenetetrahydrofolate reductase in stroke. Stroke. 1997;28:265. Abstract.
Donnan GA, Norrving B, Bamford JM, Bogousslavsky J. Lacunar and Other Subcortical Infarctions. Oxford, UK: Oxford University Press; 1995.
Fiskerstrand T, Refsum H, Kralheim G, Ueland PM. Homocysteine and other thiols in plasma and urine: automated determination and sample suitability. Clin Chem. 1993;39:263-271.
Harmon DL, Woodside JV, Yarnell JWG, McMaster D, Young IS, McCrum EE, Gey KF, Whitehead AS, Evans AE. The common ‘thermolabile’ variant of methylene tetrahydrofolate reductase is a major determinant of mild hyperhomocysteinaemia. QJM.. 1996;89:571-577.