Subjects and methods

Fifty five patients with Alzheimer's disease and 66 sex and age matched healthy control subjects were randomly selected from outpatients. The diagnosis of dementia and Alzheimer's disease was made according to the DSM-IIIR criteria and those of the National Institute of Neurological and Cognitive Disorders Association, respectively. We determined the apoE genotype by the procedure described previously.13 Leucocyte DNA was extracted from the subjects' blood samples using standard phenol extracts. The DNA was amplified using the polymerase chain reaction with the specific primers, forward (5′-AGACAGTCT CCCTCTTGCTGA-3′) and reverse (5′-CC TCCTTTCCTGACCCTGTCCTT-3′). The amplified DNA was cleaved with the restriction enzyme Hha I, and then analysed by Southern blotting.

The white matter lesions found by T2 weighted MRI (1.5 tesla) were evaluated in all 55 patients with Alzheimer's disease. The lesions were graded on the following four class scale: grade 1 (no lesions), grade 2 (punctate lesions), grade 3 (lesions reaching confluence), and grade 4 (confluent or irregular periventricular hyperintensity), according to the procedure described by Schmidt et al.14 The evaluation of MRI was blind to other information about the patients.

A patient was considered to have hypertension if the history contained one of the following: treatment for hypertension, medication with antihypertensive drugs, or a systolic/diastolic blood pressure higher than 150/90 mm Hg, disclosed by at least three measurements. Patients were considered to have diabetes mellitus if they were taking medication (sulfonyl urea or insulin therapy), or if the blood sugar was higher than 180 mg/dl at any sampling. A patient was considered to have hyperlipidaemia if: they were presently being treated with antihyperlipidaemia drugs, or if the plasma total cholesterol was higher than 240 mg/dl or the triglyceride was higher than 180 mg/dl from at least two samples of fasting blood.

Firstly, the association of apoE genotype with the presence of white matter lesions was investigated and statistically examined using the χ² test. Then the odds ratio (OR) of various factors, including possession of the ε4 allele, hypertension, diabetes, and hyperlipidaemia to the presence of the lesions was determined.

The association of factors was examined by Fisher's exact test using a 2x2 contingency table. The relation between the lesion grades and these factors was examined by the non-parametric Mann-Whitney's test. The relation between age and the lesion grades was examined by the non-parametric Spearman's rank correlation test.

Results

Cerebral white matter lesions were detected in 36 (65.4%) of the 55 patients with Alzheimer's disease. Twenty nine patients (44.8%) had patchy lesions, but neither confluent nor diffuse lesions in the white matter. Confluent or diffuse lesions were seen in seven of 55 patients (14.3%). The mean age (SD) of the patients with patchy lesions and diffuse or confluent lesions was 77.8 (1.4) years old and 80.6 (2.3) years old, respectively, both of which were significantly higher than that of patients without lesions (71.1 (2.3) years old). The frequency of the possession of the ε4 allele was 36.8% in patients without lesions, 44.8% in those with patchy lesions, and 14.3% in those with confluent or diffuse lesions. There was no statistically significant relation between the ε4 allele frequency and the lesion grades (p=0.642). The female proportion of patients without white matter lesions was 47.6%, whereas that of the patients with patchy and confluent or diffuse lesions was 69.0% and 85.7%, respectively. There was a significant positive relation between the female proportion and the lesion grades (p=0.047, table1).

In the control subjects, the frequency of the ε4 allele was 0.053, which is consistent with previous studies in the Japanese population.11 The allele frequency was 0.237 and 0.222 in the groups of patients with Alzheimer's disease with and without white matter lesions, respectively, and was significantly higher than that in the control subjects. However, the allele frequency was not significantly different in patients with Alzheimer's disease but without white matter lesions (p=0.986, table 2). Furthermore, the OR for the possession of the ε4 allele to the presence of white matter lesions was 1.09 so that the possession of the ε4 allele was not a risk for the presence of white matter lesions (table3).

By contrast with the apoE genotype, aging was a significant risk for the presence of white matter lesions. The OR of ages older than 70 years and older than 80 years was 7.20 (p=0.006) and 4.27 (p=0.041), respectively. As mentioned above, the proportion of women was significantly higher in the group with white matter lesions, and the OR for women was 2.89 (p=0.084). The ORs of the other vascular risk factors, the presence of hypertension, diabetes mellitus, and hyperlipidaemia, were low and not significant risk factors for the lesions (table 3).

Discussion

Although the pathogenetic mechanism causing cerebral white matter lesions has not been resolved, the deep cerebral white matter is vulnerable to chronic ischaemia, because arteries from the leptomeningeal border zones to the deep white matter, and medullary arterioles must pass through this region. The periventricular white matter is thought to be particularly susceptible to ischaemia because the arteries irrigating these areas have few anastomoses. Therefore, chronic ischaemic damage to the white matter could cause white matter lesions. The possession of the ε4 allele has been disclosed as a significant vascular risk factor, because homozygous ε4 alleles are closely related to high total and LDL cholesterol,5 and are a significant risk for fatal myocardial infarction.15However, the relation between the ε4 allele and cerebral vascular lesions is controversial.7-9

Cerebral amyloid angiopathy is associated with ischaemic white matter lesions in Alzheimer's disease,16 and it has been shown that the extent of the cerebral amyloid angiopathy pathology is related to the apoE genotype.17 Although the presence of white matter lesions is not related to the apoE genotype in Dutch hereditary amyloid angiopathy,18 the relation between the presence of white matter lesions and cerebral amyloid angiopathy in Alzheimer's disease has not been elucidated. The present study showed no association between the apoE genotype and white matter lesions in Alzheimer's disease and suggests that there is no correlation between white matter infarction and the extent of amyloid angiopathy; however, further studies are needed to show the relation between white matter lesions and the extent of amyloid angiopathy in Alzheimer's disease.The present study demonstrated that aging was a significant risk for white matter lesions in Alzheimer's disease, consistent with previous studies of Alzheimer's disease.19 In normal brains, however, there is a significant negative correlation between white matter volume and age but no significant association between white matter hyperintensity area, shown by MRI, and age.20Hypotension as a result of postural changes may produce a reduction in cerebral blood flow and can cause repetitive ischaemia in the cerebral white matter, especially in elderly patients.21 Moreover, changes in vascular structure may provide this association. In addition to arteriosclerotic changes, the collagenosis of veins in the periventricular white matter is associated with aging and the degree of white matter lesions.22 Compared with normal brains, these changes in vascular structure may be enhanced in the brains of patients with Alzheimer's disease.

Another significant risk factor for the presence of white matter lesions is sex—that is, a predominance of women, consistent with previous research.1 23 The reason for female predominance is not the age difference between the sexes because there was no significant difference in age between the men and women in the present study. In a community based study, there was no significant difference in the frequency of ischaemic white matter lesions between men and women.24 Recent epidemiological studies have indicated that women have a higher risk of the disease, especially at ages older than 80.25 26 The decline in estrogen has been reported to be related to the risk of Alzheimer's disease in postmenopausal women, and replacement of estrogen reduces the morbidity of the disease. Estrogens have protective effects against neuronal death in in vitro studies.27 29 Therefore, the estrogen decline in menopausal women may be related to the degenerative damage in the cerebral white matter, although further studies are needed to elucidate the effect of sex difference on the presence of white matter lesions.

In the present study, there was no association between ε4 allele frequency and grade of white matter lesion, and lesion grade increased with age and in women. The reason of no association between apoE genotype and white matter lesions in Alzheimer's disease is still undetermined; however, the white matter lesions may be caused by changes in small vessels, which could be affected by age rather than lipid metabolism. Therefore, an age adjusted study would be needed to elucidate the relation of ε4 and white matter lesions; this would be difficult to perform in the current study because of its small size and statistical power issues.