The relationship between white matter low attenuation on brain CT and vascular risk factors: a memory clinic study - computed tomography

Khaled Amar

Introduction

White matter low attenuation (WMLA), or leucoaraiosis are hypodense areas in the cerebral white matter described on brain CT scans since 1980 [1, 2]. WMLA are observed in about 10% of normal elderly people and their prevalence increases with age [3, 4]. They are commonly seen in demented patients with a prevalence of approximately 30% in Alzheimer's disease (AD), and 80% in vascular dementia (VAD) [5-7]. On pathological examination, WMLA are areas of demyelination, gliosis, perivascular oedema, and lipohyalinosis of the deep penetrating arteries [8, 9]. The exact cause and significance of WMLA is not understood but a relationship to vascular risk factors particularly hypertension has been demonstrated in a number of studies [6, 10-12, 14]. However Raiha et al. have recently reported a relationship between WMLA and low rather than high blood pressure, suggesting that haemodynamic factors probably play a dominant role in the aetiology of white matter lesions [13].

The purpose of this study was to examine the prevalence of WMLA in our Memory Disorders Clinic patients including those who are borderline for dementia, and to investigate its relationship with vascular risk factors, particularly blood pressure.

Methods

We assessed brain CT scans of 202 patients referred to our Memory Disorders Clinic between January 1991 and December 1992. All patients are assessed with full history (from patients and carers), physical examination which includes cardiovascular and neurological examination, and are tested with a neuropsychological test battery. They are investigated with a laboratory dementia screen and brain CT scans. In appropriate cases EEG, MRI and/or SPECT scans are also performed. Diagnosis for each patient is made in a case conference which involves at feast two physicians with considerable experience in dementia, a psychiatrist and at least two psychologists who are also experienced in this field. McKhan's criteria are used for diagnosing Alzheimer's disease, while the diagnosis of vascular dementia is usually considered in the light of the clinical findings, the Hachinski Ischaemic scale and the DSM III R criteria for diagnosing vascular dementia [15-17].

Brain CTs were interpreted by an experienced neuroradiologist (T.L.) who was not aware of the clinical data. He recorded the presence or absence of WMLA, cerebral infarcts and atrophy. Areas of WMLA are normally seen in the periventricular and subcortical white matter, have an ill defined and patchy appearance (unlike infarcts which are well demarcated and usually follow a specific vascular territory), and have a density between normal white matter and CSF.

The extent of WMLA was recorded on a four-point scale (from 0 to 3) where 0 = no visible WMLA, 1 = WMLA localized to the frontal and/or occipital periventricular regions of the lateral ventricles, 2 = as in 1 but WMLA spreading towards the centrum semiovale, and 3 = extensive WMLA coalescing with the centrum semiovale. Intensity of WMLA was also graded on a four-point scale from 0 to 3, where 0 = absent, 1 = mild, 2 = moderate, and 3 = severe. Using a modified formula devised by Blennow et al., a leucency score, reflecting WMLA severity, was calculated by adding the scores of WMLA extent and intensity (minimum score = 0 and maximum = 6) [11].

The following information was gathered from the patients' notes; age, sex, diagnosis, presence of the vascular risk factors; hypertension, diabetes, hypercholestrolaemia, heart disease and peripheral vascular disease. Blood pressure, focal neurological signs on examination, blood cholesterol and random blood glucose were also recorded Patients were divided into six diagnostic groups: (1) probable Alzheimer's disease, (2) probable vascular dementia (VAD), (3) mixed AD and VAD, (4) other dementias, (5) possible dementia, (6) isolated memory loss, and (7) no dementia.

Patients in the 'other dementia' group included patients with dementia other than AD or VAD such as senile dementia of the Lewy body type. Parkinson's disease and alcoholism. Those in the 'possible dementia' group included patients with minimal cognitive impairment who were borderline for the diagnostic criteria for dementia. The majority of patients with isolated memory loss were suffering from age-associated memory impairment and memory loss secondary to cerebrovascular disease.

For statistical analysis, Student's t test was used to detect the difference between means, and Pearson's [x.sup.2] was used to compare groups. All statistical analyses were performed using SPSS/PC (Statistical Package for Social Studies).

Results

As can be seen from Table I, 45% of our patients were men (91 patients), and 55% were women. The mean age was 71 years (SD 10.2), and ranged from 45 to 93 years. One hundred patients (49.5%), with a mean age of 74.5 years (SD 8.4), had visible WMLA on their brain CTs, compared with the 102 patients (55.5%), with a mean age of 67.5 years (SD 10.6) with no visible WMLA. The difference in age between these two groups did not reach statistical significance (p = 0.067).

Table I. Demographic characteristics of the patients

[TABULAR DATA OMITTED]

Table II shows that the prevalence of WMLA increases from 12% in patients who had no evidence of a cementing illness, to 32% in those with isolated memory loss and 59% in patients suffering from possible or probable dementia. Patients with high leucency score were also more likely to be suffering from dementia, especially vascular dementia, rather than isolated memory loss or no dementia.

Table II. Number (%) of patients in the different diagnostic
groups and the incidence of WMLA
                   No. of       WMLA
                 patients     present     LS> 4
All Dementias        127         76 (59)    19 (25)
Alzheimer's disease   76     40 (53)     7 (18)
Vascular dementia     33     23 (70)     7 (30)
AD and VAD             5      4 (80)     2 (50)
Other dementia        13      9 (69)     3 (33)
Possible dementia     17     10 (59)     0
Isolated memory loss  34     11 (32)     1 (9)
Normal                25      3 (12)     1 (33)
Total                202    100 (49.5)   21 (21)
LS > 4 = leucency score of 5 or 6.

There was no significant difference between patients with and without WMLA with regard to history of smoking, hypertension, diabetes, or hypercholesterolaemia and no correlation between WMLA and the blood glucose or cholesterol levels. However, as can be seen from Table III, raised systolic blood pressure, heart disease, peripheral vascular disease, focal neurological signs on examination and central atrophy were all significantly more common in patients with WMLA (p < 0.05).

Table III. Prevalence of vascular risk factors, findings on examination, blood results, and CT in patients with and without WMLA

[TABULAR DATA OMITTED]

Although there were more visible cerebral infarcts on CT in the WMLA group, compared with patients with no visible WMLA, this did not reach statistical significance. The presence of central atrophy however was significantly more common in patients with WMLA and this remained even after including patients with both central and cortical atrophy. (p = 0.036).

On stepwise regression analysis, only age, diagnostic group and raised systolic blood pressure were predictors of WMLA When the effect of age was eliminated from the equation both diagnostic groups and raised systolic blood pressure remained as predictors of white matter lesions.

Discussion

The older age for the WMLA group is not surprising, since the prevalence of WMLA has been shown to rise with age [3, 4]. This age difference did not reach statistical significance in our sample (p = 0.067).

WMLA was commonest in patients with vascular and combined vascular and degenerative dementia as in previous reports [5, 7, 12]. The prevalence of WMLA in Alzheimer's disease (52%) was higher than that quoted in most similar studies and this may be explained, at least partially, by easier recognition of WMLA with more sensitive CT scanners. The high prevalence of white matter lesions (59%) recorded by Brun and Englund in their post-mortem studies of white matter changes in Alzheimer's disease suggests that this is not an overestimation [18]. Wallin et al. reported a marked difference in the prevalence of WMLA between early onset (onset before the age of 65 years) and late onset AD with a much higher frequency in patients with late onset AD (80% in late onset AD versus 11.5% in early onset AD) [12]. This was to an extent mirrored in our sample of patients with AD, i.e. among 76 AD patients, 22 patients had an early onset disease with a WMLA incidence of 36%, compared with a prevalence of 59% in 54 patients with late-onset disease. This trend was not statistically significant.

To our knowledge there have been no reports on the prevalence Of white matter lesions in patients with borderline dementia (i.e. those with possible dementia and isolated memory loss). There was no difference in prevalence between patients with possible or probable dementia, while patients with isolated memory loss had a WMLA prevalence which was higher than in patients who were cognitively normal. This may indicate a possible role for white matter lesions in causing or contributing to cognitive impairment, or being a by product of the dementia syndrome, and would be in keeping with the findings of De Reuck et al. of frontal leukoaraiosis contributing to the severity of dementia [19]. It is also supported by recent reports of a subtle cognitive impairment in healthy elderly subjects with white matter lesions [20-23]. Our prevalence figure however will need to be cautiously interpreted, since our study is not population-based and prevalence figures therefore could be influenced by the pattern of referral to our unit.

Although the total number of patients with severe WMLA (leucency score 5 or 6) was small (21 patients), patients with severe WMLA were more likely to be suffering from probable dementia, especially vascular dementia, rather than possible dementia or isolated memory loss. This implies that patients with severe white matter lesions could be at risk of developing dementia and therefore need to be carefully evaluated even in the absence of overt cognitive impairment. Whether white matter lesions are a marker for future development of dementia is not clear yet. longitudinal studies of healthy subjects with WMLA are needed in order to answer this question.

Twenty-four per cent of patients with WMLA had a history of previously diagnosed hypertension compared with 18% of patients without WMLA, but this did not reach statistical significance. The mean systolic blood pressure, however, was significantly higher in patients with WMLA (p = 0.036), and this became more significant (p = 0.002) when we excluded patients with a history of hypertension whose blood pressure could have been unrepresentatively low because of antihypertensive medication.

Our findings of a relationship between high blood pressure and WMLA is in agreement with previous reports [4, 5, 10-12], and contradicts the findings by Raiha and co-workers who reported a correlation between WMLA and low rather than high blood pressure [13]. We cannot explain the lower blood pressure in their patients with WMLA, but it is possible that their patients could have been more cognitively impaired with behavioural problems requiring sedative drugs that artificially lowered their blood pressure. It is also possible that both low and high blood pressure are contributing factors to the genesis of white matter lesions.

A history of smoking, diabetes or hypercholesterolaemia did not help to discriminate patients with WMLA from those without WMLA and mean blood glucose and cholesterol did not differ significantly between the two groups. The finding of a correlation between focal neurological signs on examination and WMLA has been noted in previous studies and seems to suggest that white matter lesions may cause specific neurological deficits [4, 5, 9,10].

Patients with heart disease included patients with current or past atrial fibrillation, heart failure, or ischaemic heart disease. These were all found to be more frequent in patients with WMLA contradicting the finding by Raiha et al. who detected a correlation with atrial fibrillation and heart failure but not ischaemic heart disease, but is in agreement with a recent report by Lindgren and co-workers {13, 14]. Heart disease however was net an independent predictor of WMLA on a multiple regression analysis model.

Cerebral infarcts, especially deep infarcts, were more common in patients with WMLA but this did not reach statistical significance in the small number of patients in these subgroups. The association between cerebral infarcts and WMLA is not surprising given the similarity in the risk factors and justifies the description of WMLA by Brun and Englund as areas of incomplete infarction [18]. The higher prevalence of cerebral atrophy in patients with WMLA would also be in keeping with the observations made by Wallin et al. and Kobari et al. [12, 34]. The correlation between central atrophy and WMLA is also in keeping with previous observations of an overlap between Binswanger's disease and normal pressure hydrocephalus [25, 26]. In both conditions gait abnormalities and urinary problems are common.

Finally, we need to be careful in interpreting these data as the testing of multiple hypotheses raises the probability that one or more of our positive results could have occurred by chance, especially those close to the 5% significance level.

In conclusion,we found that the prevalence of WMLa increases with likelihood of cognitive impairment suggesting that white matter lesions probably play an important role in the dementia process. There is a correlation between WMLA and arterial hypertension as well as heart disease although the latter was not an independent predictor of WMLA. This, and the lack of correlation with low blood pressure suggest that thromboembolic rather than haemodynamic factors are probably more important in the pathogenesis of white matter lesions.

Acknowledgement

We are grateful to Mr Tony Highest for his statistical advice.

References

[1.] Hachinski VC, Potter P, Merskey H. Leuko-araiosis. Arch Neurol 1987;44:21-3. [2.] Valentine AR, Moseley IF, Kendall BE. White matter abnormality in cerebral atrophy: clinicoradiological correlations. J Neurol Neurosurg Psychiatry 1980;43:139-42. [3.] George AK, De Leon MJ, Gentes CI, et al. leukoencephalopathy in normal and pathologic aging. Am J Neuroradiol 1986;7:561-6. [4.] Steingart A, Hachinski VC, Lau C, et al. Cognitive and neurologic findings in subjects with diffuse white matter leucencies. Arch Neurol 1987;44:32-5. [5.] Steingart A, Hachinski VC, Lau C, et al. Cognitive and neurologic findings in demented patients with diffuse white matter leucencies on computed tomographic scan. Arch Neurol 1987;44:36-9. [6.] Inzitari D, Diaz F, Fox A, et al. Vascular risk factors and leukoaraiosis. Arch Neurol 1987;44:42-7. [7.] Erkinjuntti T, Ketonen L, Sulkava R, Vuorialho M, Palo J. CT in the differential diagnosis between Alzheimer's disease and vascular dementia. Acta Neurol Scand 1987;75:262-70. [8.] Janota I, Mirsen TR, Hachinski VC, Lee DH, Merkey H. Neuropathologic correlates of leuko-araiosis. Arch Neurol 1989;46:1124-8. [9.] Rezek DL, Morris JC, Fulling KH, Gado MH. Periventricular white matter leucencies in senile dementia of the Alzheimer type and normal aging. Neurology l987;37:1365-8. [10.] Gupta SR, Naheedy MH, Young JC, et al. Periventricular white matter changes and dementia. Arch Neurol 1 1988;45 :637-41. [11.] Blennow K, Wallin A, Uhlemann C, Gottfries CG. White matter lesions on CT in Alzheimer patients: relation to clinical symptomatology and vascular factors. Acta Neurol Scand l 991 ;83: 187-93. [12.] Wallin A, Blennow K, Uhlemann C, Langstrom G, Gottfries CG. White matter low attenuation on computed tomography in Alzheimer's disease and vascular dementia-diagnostic and pathogenetic aspects. Acta Neurol Scand 1 989;80:518-23. [13.] Raiha I, Tarvonen S, Kurki T, Rajala T, Sourander L. Relationship between vascular risk factors and white matter low attenuation of the brain. Acta Neurol Scand 1993;87:286-9. [14.] Lindgren A, Roijer A, Rudling O, et al. Cerebral lesions on magnetic resonance imaging, heart disease, and vascular risk factors in subjects without stroke-a population based study. Stroke 1994;25:929-34. [15.] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stradlan EM. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA work group under the auspices of the Department of Health and Human Services task force on Alzheimer's disease. Neurology 1984;34:939-44. [16.] Hachinski VC, Iliff LD, Zilhka E, et al. Cerebral blood flow in dementia. Arch Neurol 1975;32:632-7. [17.] American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 3rd edn., rev. Washington, DC: American Psychiatric Association, 1987. [18.] Brun A, Englund E. White matter disorder in dementia of the Alzheimer type: a patho-anatomical study. Ann Neurol 1986;19:253-62. [19.] De Reuck J, Decoo D, Strijckmans K, Lemahieu I. Does the severity of leukoaraiosis contribute to senile dementia? Eur Neurol 1992;32:199-205. [20.] Boone KB, Miller BL, Lesser IM, et al. Neuropsychological correlates of white matter lesions in healthy elderly subjects. Arch Neurol 1992,49:549-54. [21.] Schmidt R, Fazekas F, Offenbacher H, et al. Neuropsychological correlates of MRI white matter hyperintensities: a study of 150 normal volunteers. Neurology 1993 ;43:2490-4. [22.] Ylikoski R, Ylikoski A, Erkinjuntii T, Sulkava R, Raininko R, Tilvis R. White matter changes in healthy elderly persons correlate with attention and speed of mental processing. Arch Neurol 1993;50:818-24. [23.] Breteler MB, van Swieten JC, Bots ML, et al. Cerebral white matter lesions, vascular risk factors, and cognitive function in a population based study: the Rotterdam study. Neurology 1994;44:1246-52. [24.] Kobari M, Meyer JS, Ichijo M. Leuko-araiosis, cerebral atrophy, and cerebral perfusion in normal aging. Arch Neurol 1990;47:161-5. [25.] Roman GC. Senile dementia of the Binswanger type: a vascular form of dementia in the elderly. JAMA 1987;258:1782-8. [26.] Vanneste JA. Three decades of normal pressure hydrocephalus: are we wiser now? 7 Neurol Neurosurg Psychiatry 1994;57:1021-5.

Authors' addresses

K. Amar, T. Lewis, G. Wilcock Frenchay Hospital, University of Bristol, Bristol BS16 1LE

M. Scott, R. Bucks Blackberry Hill Hospital, Bristol

COPYRIGHT 1995 Oxford University Press
COPYRIGHT 2004 Gale Group