(Stroke. 1995;26:1593-1597.)
© 1995 American Heart Association, Inc.
Articles |
From the Department of Neurology, Masuda Red Cross Hospital, Shimane, Japan.
Correspondence to Hitoshi Fukuda, MD, I-103-1 Otoyoshi, Masuda, Shimane 698, Japan.
| Abstract |
|---|
|
|
|---|
Methods The extent of PVH observed on 238 MRI scans was assessed retrospectively in 238 patients older than 40 years with the use of a five-point (0 to 4) classification scale. Hypertensive subjects were divided into two groups: patients receiving no or irregular treatment and patients receiving regular treatment for hypertension.
Results Age, hypertension, treatment of hypertension, and multiple cerebral vascular lesions on MRI correlated significantly and independently with the extent of PVH. Sex and diabetes mellitus did not correlate with PVH. The Mann-Whitney U test showed significantly more extensive PVH in subjects with no or irregular treatment of hypertension compared with normotensives (2.5±0.7 versus 1.9±0.6, P<.0001) and compared with subjects receiving regular treatment of hypertension (2.5±0.7 versus 2.1±0.5, P=.0019). The latter patients had more extensive PVH than normotensives, but the difference was not significant. Mean systolic and diastolic blood pressure differed significantly among the three groups.
Conclusions Hypertension and age were major predictors of the extent of PVH. Regular treatment for hypertension appeared to prevent the progression of PVH.
Key Words: drug therapy hypertension magnetic resonance imaging white matter
| Introduction |
|---|
|
|
|---|
| Subjects and Methods |
|---|
|
|
|---|
|
|
Methods
Arterial blood pressure was measured with the
patient seated. If a patient had received medical treatment, the blood
pressure recorded was that measured during treatment. In most cases
we used the blood pressure data obtained on the day of MRI examination.
When the MRI had been performed only in the acute stage of stroke, we
used in our analysis the blood pressure measured at the chronic
stage more than 1 month after stroke onset.
Hypertension was considered present if the patient's blood pressure was 160/95 mm Hg or higher in any visit or if the patient had received medical treatment for hypertension.
Conditions of treatment for hypertension were defined as follows. If a hypertensive patient had not received medical treatment or had a history of discontinuing medical treatment, that subject was classified as receiving no or irregular treatment. If a hypertensive patient regularly continued taking medical treatment since the hypertension had been diagnosed, he was classified as receiving regular treatment.
MRI studies were performed on a 1.0-T superconductive scanner (Shimadzu SMT100X). Spin-echo pulse sequences were used to generate both T1-weighted (TR, 400 ms; TE, 15 ms; NEX, 3; or TR, 500 ms; TE, 20 ms; NEX, 3) and T2-weighted (TR, 3000 ms; TE, 90 ms; NEX, 1) axial brain images. Matrix was 256x256; scanning was done in the orbitomeatal plane with multipolarization radiofrequency head coil; these sections were 8 mm thick and separated by a 2-mm interscan gap; they began at the medullocervical cord junction and extended superiorly to the inner table of the skull.
Definition of Vascular Lesions on MRI
We counted two types of lesions. One type showed hyperintensity
on T2-weighted imaging and obvious low intensity on
T1-weighted imaging; its diameter exceeded 3 mm. We planned
to include cerebral infarction on MRI, but a large état
criblé can be included in such a lesion.22 23 The
other type of lesion showed a low or high intensity, circumscribed by
low intensity on T2-weighted imaging. We considered this
lesion as intracranial hemorrhage in the chronic
stage.24 Vascular lesions were classified as absent,
solitary, or multiple according to a total count of these lesions.
Method of Grading PVH on MRI
We excluded from analysis punctate hyperintensities in the deep
white matter. Images were read by one of the authors (H.F.) without
knowledge of the clinical data. All MRI sections of each subject were
used to evaluate PVH grade. We used a newly developed five-point scale
to classify PVH (Figure
): 0, absent; 1, caps only on
anterior horns of the lateral ventricle at the level of the basal
ganglia; 2, thin lining, smooth halo, or irregular PVH within an inner
half area of the white matter at the level of the body of the lateral
ventricle; 3, PVH extending into the outer half area of the white
matter at any region around the lateral ventricle; and 4, PVH covering
the entire white matter. An MRI study that gave borderline results
between two grades was automatically assigned to the lower grade. To
dichotomize the PVH score for use in logistic regression
analysis, PVH was defined as mild if the score was
2 and as
severe if the score was
3.
|
We studied 238 MRI scans of 238 patients, that is, one MRI scan per patient. Of the 238 patients included in the study, 51 patients underwent more than one MRI scan. When a patient had serial scans, the MRI scan done on the same day as the laboratory examinations or the day closest to the examinations was included in the study.
We determined the interrater reliability of the two raters (H.F. and M.K.) using 35 MRI head scans selected to represent the entire range of PVH grades in the study population. Spearman's rank correlation was 0.921 (P<.0001).
Statistical Analysis
Data are reported as mean±SD. One-way ANOVA was used to compare
means of ages and blood pressures among three groups. The Mann-Whitney
U test was used to test for significant differences in PVH
grade between two groups, and the Kruskal-Wallis rank test was used to
test for significant differences in PVH grade among three groups. The
2 test was applied to evaluate significant
differences between frequency distributions. Spearman's rank
correlation was used to make univariate comparisons of PVH
grade with age and blood pressures. Multivariate linear
regression analysis was used to estimate the independent
effects of the predictor variables on the PVH grade. We also
performed logistic regression analysis with the PVH score
dichotomized and coded as severe (1) or mild (0). A level of
P<.05 was accepted as statistically significant.
| Results |
|---|
|
|
|---|
Differences in PVH grade and clinical parameters in various
conditions of treatment for hypertension are shown in Table 3
. The Mann-Whitney U test showed a
significantly more extensive PVH in the hypertensives receiving no or
irregular treatment than either the normotensives (P<.0001)
or the hypertensives receiving regular treatment (P=.0019).
The hypertensives receiving regular treatment had a more extensive PVH
than the normotensives, but the difference was not significant
(P=.0545).
|
The mean systolic or diastolic blood pressure differed significantly among the three groups; it was highest in the hypertensives receiving no or irregular treatment, lower in hypertensives receiving regular treatment, and lowest in the normotensives.
Vascular lesions on MRI were significantly associated with the conditions of treatment for hypertension (P=.0046).
The relative contributions of age (range, 41 to 91 years), sex (female [0], male [1]), hypertension (absent [0], present [1]), treatment of hypertension (no or irregular [0], regular [1]), DM (absent [0], impaired glucose tolerance [1], present [2]), and vascular lesions on MRI (absent [0], solitary [1], multiple [2]) to the PVH score (0 to 4) were assessed by multivariate linear regression analysis. Age, hypertension, treatment of hypertension, and vascular lesions on MRI significantly and independently correlated with PVH. The standardized coefficients were 0.30 (P<.0001), 0.33 (P<.0001), -0.24 (P=.0010), and 0.19 (P=.0036) for these four variables, respectively. Neither sex (P=.3787) nor DM (P=.8556) correlated with PVH scores. The relative contributions of these variables to the PVH score, which was dichotomized and coded as severe (1) or mild (0), were assessed by logistic regression analysis. Age, hypertension, treatment of hypertension, and vascular lesions on MRI significantly and independently correlated with the dichotomized PVH. The r statistics were .23 (P=.0001), .24 (P<.0001), -.17 (P=.0020), and .17 (P=.0018) for these four variables, respectively. Sex (P=.8718) and DM (P=.5721) did not correlate with the dichotomized PVH scores. These results agreed with those of multivariate linear regression analysis.
| Discussion |
|---|
|
|
|---|
Vascular lesions on MRI were significantly associated with PVH in our study. Multiple MRI lesions probably show the progressive state of arteriosclerosis. PVH is, therefore, closely connected with cerebrovascular risk factors. Since DM is an important cerebrovascular risk factor, it is probably associated with PVH or PVL. However, our study failed to establish a significant association between PVH and DM. Some studies16 17 18 have suggested an association between DM and PVH or PVL, while others1 3 7 8 12 15 19 21 33 have not. Several possibilities could explain these conflicting results. First, DM is probably a weaker predictor of PVH than hypertension. Several studies on the association of DM with silent infarction have also given conflicting results.34 35 36 37 Reed et al38 reported that serum glucose is marginally associated with atherosclerosis in the large arteries of the circle of Willis, but it is not associated with that in the small arteries; diastolic blood pressure is the only variable that is consistently associated with both atherosclerosis in the large and small arteries and clinical stroke. Second, there may be differences in DM severity and duration among study populations. Third, the number of patients having abnormal glucose metabolism may have been too small in some studies to assess the association of DM with PVH or PVL, since DM is apparently a weak predictor of PVH or PVL. The number was 156 in the study by Manolio et al15 and 142 in the study by Jorgensen et al33 but was less than 38 in other studies.1 3 7 8 12 16 17 18 19 21
Our main purpose was to study differences in the extent of PVH among hypertensives receiving no or irregular treatment or regular treatment for hypertension. Our study showed significantly more extensive PVH in hypertensives receiving no or irregular treatment than in those receiving regular treatment. Of 7 patients with extensive PVH covering the entire white matter, 6 had received no or irregular treatment. Five of the 6 had clinical signs resembling those of subcortical arteriosclerotic encephalopathy (Binswanger's disease).2 25 Although the association of white matter hyperintensities with cognitive decline and other neurological deficits remains controversial,1 2 3 4 5 6 7 8 9 10 11 20 39 40 41 42 when the lesions cover the entire white matter, cognitive decline and other neurological deficits become obvious.2 6 9 43 The fact that normotensives did not show PVH covering the entire white matter suggested that PVH does not extend to cover the entire white matter in the absence of hypertension. If regular treatment of hypertension prevents the progression of white matter hyperintensity and decreases the possibility of obvious cognitive decline, the regular treatment of hypertension is extremely important. Our study has shown that the difference in the extent of PVH between the hypertensives receiving regular treatment and normotensives is not significant. Although this study is retrospective, these findings suggest that regular treatment of hypertension may prevent the progression of white matter hyperintensity.
Regular treatment of hypertension was defined as regular, continuous medical treatment since the diagnosis of hypertension. Blood pressure need not have been maintained within the normal range. Diastolic and systolic blood pressure differed significantly between patients receiving regular treatment of hypertension and normotensives. The differences in the extent of PVH between the two groups, although not significant, may originate from this difference in blood pressures. This suggests that maintaining the blood pressure of hypertensive patients within the normal range from the time of diagnosis may lessen the difference in the extent of PVH between the two groups.
| Selected Abbreviations and Acronyms |
|---|
|
| Acknowledgments |
|---|
Received January 18, 1995; revision received June 1, 1995; accepted June 1, 1995.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
B. van Harten, F.-E. de Leeuw, H. C. Weinstein, P. Scheltens, and G. J. Biessels Brain Imaging in Patients With Diabetes: A systematic review. Diabetes Care, November 1, 2006; 29(11): 2539 - 2548. [Full Text] [PDF] |
||||
![]() |
R. Schmidt, L. J. Launer, L.-G. Nilsson, A. Pajak, S. Sans, K. Berger, M. M. Breteler, M. de Ridder, C. Dufouil, R. Fuhrer, et al. Magnetic Resonance Imaging of the Brain in Diabetes: The Cardiovascular Determinants of Dementia (CASCADE) Study Diabetes, March 1, 2004; 53(3): 687 - 692. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. D. Taylor, D. C. Steffens, J. R. MacFall, D. R. McQuoid, M. E. Payne, J. M. Provenzale, and K. R. R. Krishnan White Matter Hyperintensity Progression and Late-Life Depression Outcomes Arch Gen Psychiatry, November 1, 2003; 60(11): 1090 - 1096. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. D. Taylor, J. R. MacFall, J. M. Provenzale, M. E. Payne, D. R. McQuoid, D. C. Steffens, and K. R. R. Krishnan Serial MR Imaging of Volumes of Hyperintense White Matter Lesions in Elderly Patients: Correlation with Vascular Risk Factors Am. J. Roentgenol., August 1, 2003; 181(2): 571 - 576. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Dufouil, A. de Kersaint-Gilly, V. Besancon, C. Levy, E. Auffray, L. Brunnereau, A. Alperovitch, and C. Tzourio Longitudinal study of blood pressure and white matter hyperintensities: The EVA MRI Cohort Neurology, April 10, 2001; 56(7): 921 - 926. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. S. van der Knaap, S. Naidu, B. K. Kleinschmidt-DeMasters, W. Kamphorst, and H. C. Weinstein Autosomal dominant diffuse leukoencephalopathy with neuroaxonal spheroids Neurology, January 25, 2000; 54(2): 463 - 463. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Adachi, M. Takagi, H. Hoshino, and T. Inafuku Effect of Extracranial Carotid Artery Stenosis and Other Risk Factors for Stroke on Periventricular Hyperintensity Stroke, November 1, 1997; 28(11): 2174 - 2179. [Abstract] [Full Text] |
||||
![]() |
R. Mantyla, T. Erkinjuntti, O. Salonen, H. J. Aronen, T. Peltonen, T. Pohjasvaara, and C.-G. Standertskjold-Nordenstam Variable Agreement Between Visual Rating Scales for White Matter Hyperintensities on MRI : Comparison of 13 Rating Scales in a Poststroke Cohort Stroke, August 1, 1997; 28(8): 1614 - 1623. [Abstract] [Full Text] |
||||
![]() |
T. L. Strassburger, H.-C. Lee, E. M. Daly, J. Szczepanik, J. S. Krasuski, M. J. Mentis, J. A. Salerno, C. DeCarli, M. B. Schapiro, and G. E. Alexander Interactive Effects of Age and Hypertension on Volumes of Brain Structures Stroke, July 1, 1997; 28(7): 1410 - 1417. [Abstract] [Full Text] |
||||
![]() |
N. Watanabe, Y. Imai, K. Nagai, I. Tsuji, H. Satoh, M. Sakuma, H. Sakuma, J. Kato, N. Onodera-Kikuchi, M. Yamada, et al. Nocturnal Blood Pressure and Silent Cerebrovascular Lesions in Elderly Japanese Stroke, August 1, 1996; 27(8): 1319 - 1327. [Abstract] [Full Text] |
||||
![]() |
H. Fukuda and M. Kitani Cigarette Smoking Is Correlated With the Periventricular Hyperintensity Grade on Brain Magnetic Resonance Imaging Stroke, April 1, 1996; 27(4): 645 - 649. [Abstract] [Full Text] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1995 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |