Apolipoprotein E, methylenetetrahydrofolate reductase (MTHFR) mutation and the risk of senile dementia--an epidemiological study using the polymerase chain reaction (PCR) method.

We examined apolipoprotein E (Apo E) polymorphism and methylenetetrahydrofolate reductase (MTHFR) 677 C to T mutation by using the polymerase chain reaction (PCR) method in 100 elderly Japanese aged 60 or more, and assessed whether these genetic factors are associated with an increased risk for the clinical phenotypes of senile dementia, Alzheimer's disease (AD) and vascular dementia (VD) by cross-sectional survey. It was found that the Apo E epsilon 4 allele were associated with an increased prevalence of AD as previously reported. Although, it was not strongly related to the severity of senile dementia, a weak association between the ApoE genotype and the severity of dementia was suggested. The proportion of patients with senile dementia was higher in the group of carriers of MTHFR mutation than in the group of noncarriers. Furthermore, the proportion of male patients with senile dementia was higher in the group of homozygous for the mutation (+/+) than the group without the mutation (-/-). Notably in VD patients, 5 of 7 males had the +/+ genotype. The results suggest that the ApoE epsilon 4 genotype and the MTHFR mutation are associated with the clinical phenotype and the clinical onset of senile dementia.


INTRODUCTION
In Japan the proportion of elderly people over 65 years old is about 16% and increasing. The prevalence of senile dementia among the total population aged 65 or more in 1995 was about 6.9% 1), which represents a social problem. Vascular dementia (VD) is the most common form of senile dementia in Japan, with Alzheimer's disease (AD) next. Individuals affected with either or both of these diseases account for at least 80% of patients with senile dementia 1). The cause of VD is cerebrovascular atherosclerosis, whereas the cause of AD remains definitively unknown. However, recent molecular investigations have elucidated the genetic factors that play a significant role in the development of AD and vascular atherosclerosis zs) A close relation between apolipoprotein-E (apoE) alleles and AD has been reported by many investigators [4][5][6]. It has also been reported that the C to T transition of 5, 10-Methylentetrahydofolate reductase (MRTHFR) gene at nucleotide position 677 results in a decrease of enzymatic activity and increase of the plasma total homocysteine level, and that hyperhomocysteinemia is a risk factor for atheroscleotic cerebrovascular and coronary heart diseases 7,8). Since the association of atherosclerosis with senile dementia is worthy of consideration in not only VD but also AD, we examined both the apoE polymorphism and the MTHFR mutation in elderly people and assessed whether these genetic factors increase the risk for and modify the clinical phenotype of AD and VD .

MATERIALS AND METHODS
We investigated 100 elderly persons (25 men and 75 women) over 60 years old (average age: 76.8 ± 9.26SD), who comprised residents of a special nursing home for the aged and the patients who were referred to the two hospitals for senile dementia or other diseases. Consent for participation in the study was provided by the subjects themselves or their legal guardians. Dementia was defined using the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (DSM-III-R) 9) referring to the diagnosis of Computed Tomography (CT) or Magnetic Resonance Imaging (MRI  To detect the mutant allele of Methylenetetrahydrofolate reductase (MTHFR), the same leukocyte DNA samples were examined. The PCR procedure was basically the same with for the isotyping of apoE, but the primers used for the analysis were 5'-TGAAGGAGAAGGTGTCTGCGGGA-3' and 5'-AGGACGGTGCGGTGAGAGTG-3', and the restriction enzyme Hinfl was used to identify the mutation, as described by Frosst et al. 75. These primers generate a fragment of 198bp. The substitution creates a Hinfl recognition sequence which digests the 198-bp fragment into 175-and 23-bp fragments. As the latter fragment has been run off, the existence of a 175-bp fragment indicates the mutant allele. The mutant allele was designated as "+", and the wild-type that means absence of mutant allele was designated as "-". The three possible genotypes are +/+homozygosity, +/-heterozygosity, and -/ -homozygosity (Fig.2). All the subjects were classified into Apo E e 4-carriers or non-carriers. According to the three types of MTHFR genotypes, they were divided into "+/+", "+/-" and ' -/ -" . Each group was examined by the proportion and the composition of senile dementia, average age, average age of onset, severity of dementia (CDR) and the rate of decline on dementia (CDR/years from onset of dementia). The difference in these information between the different genotypes of apoE and MTHFR were statistically analyzed with the Yates-corrected chi-square test, Welch's t-test, Pearson's correlation coeficient, and the multiple regression analyses using the computer software HALBAU (N88-Japanese BASIC) and Statistical Package for Social Sciences (SPSS, 7.5J).

The frequencies
of Apo E genotypes and the frequencies of carriers and noncarriers among non-dementia subjects, patients with AD, patients with VD, and patients with other forms of dementia are shown in Table 1. The frequency carriers was 12% in non-dementia, 58% in AD, 23% in VD, and 13% in mixed type. As shown in previous reports 3,4,5.0, carriers was significantly higher in AD patients than in both non-dementia (Odds ratio: 9.05, 95% confidence interval: 2.53-32.16, p<0.001) and VD patients (Odds   Table 2. The prevalence of dementia was higher in the carriers (23/27) than in the noncarriers (44/73) (p<0.05). The difference was remarkable especially in the prevalence of AD (p<0.001). Table 3 shows the average age, average age of onset and the severity of senile dementia in E 4 carriers and noncarriers. The average age of the E 4 carriers in non-dementia was higher than that of noncarriers (p=0.05). In respect of the grade of severity of dementia, the average Hughes' clinical dementia rating (CDR) was higher in the carriers than in the noncarriers (p=0.07). We also examined the CDR per year from onset as the rate of decline. The average score of the CDR per year from onset was lower in the noncarriers (p=0.09). These results suggest that there is also an association between Apo E E 4 and severity of senile dementia.
The frequencies of the three MTHFR genotypes are shown in Table 4. In our series, the frequencies of the MTHFR mutation in non-dementia were as follows: +/+ genotype, 15%, + /-genotype, 45%, -/-genotype, 39%, and these were similar to previous reports on both Japanese 12) and French-Canadian 7) populations. The MTHFR mutation was not associated with the senile dementia. Table 5 shows the odds ratio and 95% confidence interval for the phenotype of senile dementia in carriers and noncarriers of MTH FR mutation. The proportion of patients with senile dementia was slightly higher in the group of carriers. OR was higher in AD than VD patients. This suggests that the MTHFR mutaton was more closely associated with the onset of AD than VD. The types of senile dementia and the different MTHFR alleles are shown in Table 6. The frequency of the +/+ genotype was higher and -/-genotype lower among the AD and VD than non-dementia subjects. The types of senile dementia and the different MTHFR alleles by sex are shown in Table 7. The prevalence of male patients with dementia was higher in the +/+ genotype (7/8) than in the -/-genotype (1/7) (p<0.05). The difference was more remarkable in the prevalence of VD patients (p=0.05) than AD patients (p=0. 16), showing that 5 of 7 male patients with VD had the +/+ genotype. On the other hand, in the female patients there were no significant differences. The average age, average age of onset and the severity of senile dementia for the different MTHFR alleles are shown in Table 8. The average age was higher for the +/+ group than -/-group among non -dementia subjects (p<0.05) . In respect to the grade of severity of dementia, the CDR score among the AD patients was higher in the -/-than +/+    group (p=0.08). Conversely, among VD patients, the CDR score was higher in the +/+ than -/-group. The correlation matrix among Apo E, MTHFR subtypes and other variables was obtain only dementia patients (n=67) and is presented in Table 9. Apo E has positive relationships with AD as well as the decline of severity of dementia (CDR/year from onset), while MTHFR +/+ genotype was found to be correlated with sex of patients. Table 10 shows the results of logistic regression analyses by each subtypes of dementia. It revealed that Apo E E 4 was a predictive factor for dementia patients (p<0.05). Specifically, Apo E E 4 was a strong predictor for the onset of AD from this analysis (p=0.0001). On the other hand, the age of subject was the only factor associated with VD, and the VD was seen more in older patients (p=0.001).
Further, age, sex, and Apo E status were used to investigate the relationship with +/+ MTHFR homozygous (Table 11). Male subjects with any subtypes of dementia were related with MTHFR (p<0.005). Interestingly, this relationship did not exist in non-dementia group.

DISCUSSION
Alzheimer disease (AD) and vascular dementia (VD), two major causes of senile dementia, have imposed a major burden on modem public health care. AD and VD, with the exception of very few familial entities with Mendelian inheritance, are regarded as multifactorial genetic disorders. Molecular investigations using the polymerase chain reaction (PCR) method have analyzed their genetic background. These approaches may be effective for the exploration of preventive methods for senile dementia.
In our study, the Apo E E 4 allele was associated with an increased risk for AD, as currently suggested. According to Forno et al., the Apo E genotype does not strongly influence the rate of decline in AD 13). Our study suggested that the Apo E * 4 allele worsens the severity of senile dementia. The effects of E 4 need to be further investigated. Apolipoprotein The possible association of the MTHFR mutation with vascular diseases led us to hypothesize that it is a candidate for a risk factor for VD and AD. A causal relation between the MTHFR mutation and the onset of AD and VD was suggested in our study and the grade of severity of dementia was paradoxically higher in the -/-than +/+ group. There was no strong association between the MTHFR mutation and senile dementia; nevertheless, the mutation was considered to modify the clinical phenotype of senile dementia. Perhaps the most remarkable result from our study was that the proportion of male patients with senile dementia was higher in the group homozygous for the mutation (+l+) than in the group without the mutation (-/-).
This suggests that the MTHFR mutation is more influential in men than women in causing the onset of VD. In the Hisayama study by Yoshitake et.al., the incidence of VD was higher for men than for women in Japan 23). Deloughery et al. reported that the mean homocysteine levels increased by MTHFR homozygotes in the vascular disease subjects were significantly higher than in healthy agematched subjects and it was slight more apparent for men. But there were no reports for the MTHFR genotypes by sex. Therefore, further epidemiological study is necessary to establish the role of the MTHFR mutation as a risk factor for senile dementia. Alzheimer's disease (AD) is the major cause of dementia in most western countries, whereas vascular dementia (VD) is more common in Japan. According to the Honolulu-Asia aging study, the prevalence of AD among elderly Japanese-American men in Hawaii was higher than that in elderly Japanese men, and was comparable to that in an European-ancestry population 25). These observations suggest that environmental factors play a major role in the onset and clinical variations of senile dementia in addition to genetic risk factors. It can be readily understood that vascular atherosclerosis and eventual VD are affected by life style such as alcohol consumption and smoking. The proposed relation of lipid metabolism to AD, as mentioned above, also implies the importance of life style in preventing the onset of AD.
Our study revealed that Apo E may be a risk factor and a phenotypic modification factor for senile dementia, and MTHFR is slightly associated with the onset of senile dementia, especialy among men. These results suggest that cerebral atherosclerosis contributes to the development of not only VD, but also AD. Investigations on a larger number of senile dementia patients will elucidate the relation of these genetic factors to senile dementia in more detail. Through this research, a way to prevent senile dementia might be found.