white_matter_lesions_and_the_risk_of_dementia.pdf

ORIGINAL CONTRIBUTION

Cerebral White Matter Lesionsand the Risk of DementiaNiels D. Prins, MD, PhD; Ewoud J. van Dijk, MD; Tom den Heijer, MD; Sarah E. Vermeer, MD, PhD;Peter J. Koudstaal, MD, PhD; Matthijs Oudkerk, MD, PhD; Albert Hofman, MD, PhD; Monique M. B. Breteler, MD, PhD

Objective: To study the association between white mat-ter lesions (WML) in specific locations and the risk ofdementia.

Design: The Rotterdam Scan Study, a prospective popu-lation-based cohort study. We scored periventricular andsubcortical WML on magnetic resonance imaging andobserved participants until January 2002 for incidentdementia.

Setting: General population.

Participants: We included 1077 people aged 60 to 90years who did not have dementia at baseline.

Main Outcome Measure: Incident dementia by Di-agnostic and Statistical Manual of Mental Disorders, ThirdEdition (DSM III-R) criteria.

Results: During a mean follow-up of 5.2 years, 45 par-ticipants developed dementia. Higher severity of periven-tricular WML increased the risk of dementia, whereas theassociation between subcortical WML and dementia wasless prominent. The adjusted hazard ratio of dementia foreach standard deviation increase in periventricular WMLseverity was 1.67 (95% confidence interval, 1.25-2.24). Thisincreased risk was independent of other risk factors fordementia and partly independent of other structural brainchanges on magnetic resonance imaging.

Conclusion: White matter lesions, especially in the peri-ventricular region, increase the risk of dementia in el-derly people.

Arch Neurol . 2004;61:1531-1534

C EREBRAL WHITE MATTER LE-sions (WML) in elderlypeople are thought to re-sult from small-vessel dis-ease and are considered to

be a risk factor for dementia.1 Evidence re-lating WML to dementia is mainly derivedfrom studies in patients with stroke andfrom cross-sectional studies in patients with

dementia. White matter lesions increase therisk of poststroke dementia and, togetherwith lacunar infarcts, are considered the pri-mary type of brain lesions in subcortical is-chemic vascular dementia.1,2 Small-vesseldisease may also contribute to the devel-opment of Alzheimer disease (AD), be-cause patients with AD were found to havemore WML than controls.3

White matter lesions are also fre-quently seen on magnetic resonance (MR)imaging of elderly patients without demen-tia, but only a few studies investigated theextent to which WML increase the risk ofdementia in the general population.4,5 Weinvestigated the risk of dementia for WMLin specific locations in the Rotterdam Scan

Study. Furthermore, we assessed whetherthe association between WML and demen-tia is independent of other risk factors fordementia and other structural brain changeson MR imaging.

METHODS

STUDY POPULATION

The Rotterdam Scan Study is a prospective,population-based cohort study designed to studycauses and consequences of age-related brainchanges in elderly people. The characteristicsof the 1077 participants have been described pre-viously.6 All participants were free of dementiaat baseline.5 Baseline examination from 1995 to1996 comprised a structured interview, neuro-psychological tests, physical examination, andblood sampling; all participants underwent MRimaging of the brain. From 1999 to 2000, 787of the 973 participants who were alive and eli-gible were reexamined at the research centersimilar to the baseline examination (responserate, 81%). All participants were continuallymonitored for mortality, dementia, and strokeuntil January 1, 2002.

MR IMAGING PROCEDURE

Details of the MR imaging examinations in theRotterdam Scan Study have been published.6 We

For editorial commentsee page 1503

Author Affiliations:Departments of Epidemiologyand Biostatistics (Drs Prins,van Dijk, den Heijer, Vermeer,Hofman, and Breteler) andNeurology (Drs Prins, van Dijk,den Heijer, Vermeer, andKoudstaal), Erasmus MedicalCenter, Rotterdam, theNetherlands; and theDepartment of Radiology,University Hospital Groningen,Groningen, the Netherlands(Dr Oudkerk).

(REPRINTED) ARCH NEUROL / VOL 61, OCT 2004 WWW.ARCHNEUROL.COM1531

©2004 American Medical Association. All rights reserved.

Downloaded From: http://archneur.jamanetwork.com/ on 11/26/2013

considered WML to be in the periventricular region if they weredirectly adjacent to the ventricle; otherwise, we considered themsubcortical. Periventricular WML were scored semiquantita-tively for locations at the frontal and occipital horns, and the lat-eral walls of the ventricles, in order to obtain a total periventricu-lar score (range, 0-9). For subcortical WML, a total volume asappearing on hardcopy was approximated based on the numberand size of lesions in the frontal, parietal, temporal, and occipitallobes (range, 0-29.5 mL).6 We rated cortical atrophy on a semi-quantitive scale (range, 0-15) and assessed subcortical atrophyby the ventricle to brain ratio (range, 0.21-0.45). Cerebral in-farcts were defined as focal hyperintensities on T2-weighted im-ages, 3 mm or larger, and with a corresponding prominent hyp-ointensity on T1-weighted images if located in the white matter.5

ASCERTAINMENT OF INCIDENT DEMENTIA

Participants with dementia were carefully excluded at base-line.5 We screened all participants for dementia at follow-upwith the Mini-Mental State Examination (MMSE),7 and the Geri-atric Mental State Schedule8 Screen positives were subse-quently evaluated using the Cambridge Mental Disorders of theElderly Examination.9 Participants who were then thought tohave dementia were examined by a neurologist and under-went extensive neuropsychological testing. In addition, we con-tinually monitored the medical records of all participants at theirgeneral practitioners’ offices and at the Regional Institute forOutpatient Mental Health Care to obtain information on newlydiagnosed dementia until January 1, 2002.5 A panel that re-viewed all available information diagnosed dementia and its sub-types according to standardized criteria.10-12 We defined the on-set of dementia as the date on which the clinical symptoms firstallowed the diagnosis of dementia to be made.

OTHER BASELINE MEASUREMENTS

The following variables assessed at baseline were used as pos-sible confounders: age, sex, educational status,13 hyperten-sion, diabetes mellitus, smoking, APOE genotype,14 history ofstroke, and incident stroke.5

DATA ANALYSIS

We assessed the association between WML and measures of gen-eralized brain atrophy with Pearson correlation coefficient, and

the association between WML and the presence of cerebral in-farcts with linear regression analysis. To examine the relation-ship between WML and the risk of dementia and AD, we usedCox proportional hazards regression models. We analyzed peri-ventricular and subcortical WML in categories of severity toanalyze the shape of the relationship and as a continuous vari-able (per standard deviation). Adjustments were made for ageand sex, and analyses were repeated with possible confound-ers and measures of other structural brain changes added tothe models. Additionally, we excluded participants with a his-tory of stroke at baseline, and participants with a baseline MMSEscore of 25 or lower. We examined possible effect modifica-tion by APOE genotype through stratified analysis.

RESULTS

Characteristics of the participants are presented inTable1. Periventricular and subcortical WML were posi-tively correlated with cortical brain atrophy (Pearson cor-relation coefficient 0.40, P�.01 and 0.25, P�.01) and sub-cortical brain atrophy (Pearson correlation coefficient0.21, P�.01 and 0.14, P�.01). Presence of cerebral in-farcts was associated with a higher severity of periven-tricular and subcortical WML (age- and sex-adjusted meandifference in periventricular WML severity 1.6 points, 95%confidence interval [CI], 1.3-1.9 points; in subcorticalWML severity 2.0 mL; 95% CI, 1.6-2.4 mL).

During 5572 person-years of follow-up (mean per per-son, 5.2 years), 45 participants developed dementia (in-cidence rate, 8.1/1000 person-years). Alzheimer disease

5

4

3

2

1

00-3 >3-6 >6-9

Haza

rd R

atio

Severity of Periventricular WML (Grade)

5

4

3

2

1

00-1 >1-6 >6-29.5

Haza

rd R

atio

Severity of Subcortical WML (mL)

Figure 1. Severity of white matter lesions (WML) is divided into 3 categoriesbased on the distribution of periventricular WML scores. The risk ofdementia is expressed as age- and sex-adjusted hazard ratios. The numberof dementia cases and total number of participants, respectively, inconsecutive periventricular WML severity categories were 16 and 749 (grade0-3), 18 and 261 (grade �3-6), 11 and 66 (grade �6-9), and in consecutivesubcortical WML severity categories 24 and 763 (0-1 mL), 10 and 244(�1-6 mL), and 10 and 65 (�6-29.5 mL).

Table 1. Baseline Characteristics of Participantsof the Rotterdam Scan Study*

Characteristic N = 1077

Age, y 72.2 (7.4)Women, % 51.5Education, % primary education only 34.8Mini-Mental State Examination score 27.4 (2.2)Hypertension, % 52.0Diabetes, % 5.8Current smoking, % 15.7Periventricular white matter lesions score 2.4 (2.2)Subcortical white matter lesions, mL 1.4 (2.9)Cortical brain atrophy, grade 5.6 (2.9)Subcortical brain atrophy, ventricle to brain ratio 0.316 (0.036)Cerebral infarcts, % 24.0APOE genotype†

�2/�2 or �2/�3 or �3/�3, % 70.9�2/�4 or �3/�4 or �4/�4, % 29.1

*Values are unadjusted means (standard deviation) or percentages.†The APOE genotype was not determined in 106 of the 1077 participants.




was diagnosed in 34 patients (76%), vascular dementia in6 (13%), and another 5 (11%) were diagnosed as havingother types of dementia (Parkinson disease dementia [3],multiple system atrophy [1], and unspecified dementia [1]).One hundred seventy-four participants died. The risk ofdementia increased linearly with severity of periventricu-lar WML (Figure 1 and Table 2). Increasing severityof subcortical WML tended to increase the risk of demen-tia, but this association was less strong (Figure 1 andTable 2). The hazard ratio for dementia per standard de-viation increment in periventricular WML score re-mained largely the same after exclusion of participants witha history of stroke (n=58) (hazard ratio for dementia 1.66;95% CI, 1.22-2.27) and after adjustment for incident stroke(hazard ratio for dementia, 1.63; 95% CI, 1.21-2.19). Af-ter exclusion of participants with a baseline MMSE scoreof 25 or lower (n=173), the association remained (haz-ard ratio for dementia, 1.50; 95% CI, 1.04-2.16).

Participants who developed dementia during fol-low-up had on average more severe WML at baseline inall locations within the periventricular and subcorticalregion (Figure 2). Periventricular WML also increasedthe risk of AD (hazard ratio for AD, 1.41; 95% CI, 1.01-1.98). The association of periventricular WML and ADwas similar for those with and without an APOE ε4 al-lele (data not shown).

COMMENT

Higher severity of periventricular WML increased the riskof dementia, whereas the association between subcorti-cal WML and dementia was less prominent. The associa-tion between periventricular WML and dementia was in-dependent of possible confounders and partly independentof other structural brain changes on MR imaging. Thestrengths of this study are the large number of participat-ing elderly people, its population-based design, and thefact that we had a complete follow-up for dementia throughour monitoring system. Another important feature is thedistinction between WML in the periventricular region andWML in the subcortical region.

A preclinical phase of many years often precedes a di-agnosis of dementia, especially in the case of AD.15 It istherefore likely that our study population contained par-

ticipants with a preclinical stage of dementia that re-mained below detection at baseline. The association of peri-ventricular WML and the risk of dementia did not changeafter exclusion of participants with a low MMSE score atbaseline, which suggests that the association is not con-fined to participants with a preclinical stage of dementia.

Our results are in line with those from previous stud-ies on the relationship between WML and dementia.Cross-sectional case-control studies reported positive as-sociations of the severity of WML on MR imaging withAD and vascular dementia.3,16 In the Cardiovascular HealthStudy, participants with more severe WML had a 2-foldincreased risk of dementia.4

Table 2. Relation Between the Severity of Periventricular and Subcortical White Matter Lesions and the Risk of Dementia*

Variable Adjusted For

Periventricular WML Subcortical WML

Hazard Ratio (95% CI) Hazard Ratio (95% CI)

Age and sex 1.67 (1.25-2.24) 1.20 (0.98-1.46)Age, sex, and education 1.67 (1.25-2.23) 1.20 (0.98-1.46)Age, sex, hypertension, diabetes, and smoking 1.69 (1.25-2.29) 1.21 (0.99-1.48)Age, sex, and APOE genotype 1.66 (1.23-2.25) 1.18 (0.97-1.44)Age, sex, and cerebral infarcts 1.60 (1.17-2.18) 1.13 (0.91-1.40)Age, sex, and generalized brain atrophy 1.51 (1.13-2.03) 1.15 (0.93-1.41)Age, sex, cerebral infarcts, and generalized brain atrophy 1.42 (1.04-1.94) 1.08 (0.86-1.35)Age, sex, and WML in other locations† 1.91 (1.31-2.78) 0.90 (0.68-1.19)

Abbreviations: CI, confidence interval; WML, white matter lesions.*Numbers are hazard ratios and 95% confidence intervals per standard deviation increment in periventricular white matter lesion score (range, 0-9; SD, 2.2),

and subcortical white matter lesion volume (range, 0-29.5; SD, 2.9).†Periventricular WML conditional on subcortical WML and vice versa.

1.5

1

0.5

0Frontal Caps Lateral Bands Occipital Caps

WM

L Se

verit

y (G

rade

)

Periventricular WML Location

1.5

1

0.5

0Frontal Parietal Temporal Occipital

WM

L Se

verit

y, m

L

Subcortical WML Location

No Incident Dementia Incident Dementia

Figure 2. Bars represent age- and sex-adjusted mean white matter lesionseverity (standard error) at baseline for participants without (n=1032) andparticipants with (n=45) incident dementia during follow-up.




Several potential mechanisms may underlie the ob-served associations between WML and dementia. Histo-pathological studies demonstrated that irregular and con-fluent WML correspond to ischemic tissue damage,including infarction, gliosis and rarefaction, and loss ofmyelin.17 This tissue damage is likely to cause discon-nection of functionally related cortical and subcorticalstructures that are important to cognitive functioning.18

It has been suggested that periventricular WML are justan epiphenomenon of brain atrophy and are not inde-pendently related to disease.19,20 We found that the as-sociation between periventricular WML and the risk ofdementia was partly independent of generalized brain at-rophy. Furthermore, we found that the association be-tween periventricular WML and incident dementia waslargely independent of the presence of cerebral infarcts,of which the majority were lacunar in our study, and wasnot mediated by incident stroke.

Subcortical WML were not as strongly associated withdementia as periventricular WML, which is in line with pre-vious reports.3 Several possible pathophysiologic mecha-nisms may explain this finding. First, WML close to theventricles may interrupt bundles of cholinergic fibers, whichextend from the nucleus basalis to the cerebral cortex, re-sulting in cholinergic denervation.21 Second, the white mat-ter in the periventricular region has a high density of longassociation fibers, whereas subcortical white matter has ahigh density of U-fibers. Diffusion tensor MR imaging stud-ies found that white matter pathologic features in patientswith AD selectively involved fiber tracts connecting corti-cal association areas, such as the cingulate bundles and thecorpus callosum.22,23 Periventricular WML may reflect vas-cular damage to these fiber tracts or, alternatively, repre-sent wallerian degeneration of these tracts.

Extensive WML alone are sufficient for a diagnosis ofvascular dementia,12 which leads to circularity when as-sociations between WML and subdiagnoses of dementiaare studied. However, the observed association betweenperiventricularWMLandADsuggests thatWMLmaycon-tribute toclinicalAD.This is compatiblewith theviewthatmost elderly people with dementia have mixed disease.24

Becauseof the smallnumberof caseswithvasculardemen-tia, we cannot provide reliable estimates for the associa-tion between WML and the risk of vascular dementia. Inconclusion, we found that higher severity of periventricu-lar WMLis independentlyassociatedwithan increasedriskof dementia. Longer follow-up with repeated MR imagingis needed to gain insight into whether, and to what extent,progression of WML increases the risk of dementia.

Accepted for Publication: November 25, 2003.Correspondence: Monique M. B. Breteler, MD, PhD, De-partment of Epidemiology and Biostatistics, ErasmusMedical Center, PO Box 1738, 3000 DR Rotterdam, theNetherlands ([email protected]).Author Contributions: Study concept and design (Drs Prins,van Dijk, Vermeer, Koudstaal, Hofman, and Breteler); ac-quisition of data (Drs Prins, van Dijk, den Heijer, Ver-meer, and Oudkerk); analysis and interpretation of data(Drs Prins, van Dijk, den Heijer, Koudstaal, Oudkerk, Hof-man, and Breteler); drafting of the manuscript (Drs Prins,van Dijk, Hofman, and Breteler); critical revision of the

manuscript for important intellectual content (Drs Prins,den Heijer, Vermeer, Koudstaal, Oudkerk, Hofman, andBreteler); statistical expertise (Drs Prins, van Dijk, andBreteler); obtained funding (Drs Koudstaal, Oudkerk, andBreteler); administrative, technical, and material support(Drs Prins, van Dijk, Vermeer, and Oudkerk); study su-pervision (Drs Koudstaal, Oudkerk, Hofman, and Breteler).Funding/Support: This study was supported by grantsfrom the Netherlands Organization for Scientific Re-search (904.61.096), Den Haag.Acknowledgment: We thank the Regional Institute forAmbulatory Mental Health Care, Rotterdam and Voor-burg, and the general practitioners of Rotterdam and Zoe-termeer for their collaboration.

REFERENCES

1. Roman GC, Erkinjuntti T, Wallin A, Pantoni L, Chui HC. Subcortical ischaemicvascular dementia. Lancet Neurol. 2002;1:426-436.

2. Leys D, Henon H, Pasquier F. White matter changes and poststroke dementia.Dement Geriatr Cogn Disord. 1998;9(suppl 1):25-29.

3. O’Brien JT, Ames D. White matter lesions in depression and Alzheimer’s disease.Br J Psychiatry. 1996;169:671.

4. Kuller LH, Lopez OL, Newman A, et al. Risk factors for dementia in the cardio-vascular Health Cognition Study. Neuroepidemiology. 2003;22:13-22.

5. Vermeer SE, Prins ND, den Heijer T, Hofman A, Koudstaal PJ, Breteler MM. Si-lent brain infarcts and the risk of dementia and cognitive decline. N Engl J Med.2003;348:1215-1222.

6. de Groot JC, de Leeuw FE, Oudkerk M, et al. Cerebral white matter lesions andcognitive function: the Rotterdam Scan Study. Ann Neurol. 2000;47:145-151.

7. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state.” J Psychiatr Res. 1975;12:189-198.

8. Copeland JR, Kelleher MJ, Kellett JM, et al. A semi-structured clinical interviewfor the assessment of diagnosis and mental state in the elderly. Psychol Med.1976;6:439-449.

9. Roth M, Tym E, Mountjoy CQ, et al. CAMDEX. Br J Psychiatry. 1986;149:698-709.10. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders, Revised Third Edition. Washington, DC: American Psychiatric Asso-ciation; 1987.

11. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinicaldiagnosis of Alzheimer’s disease. Neurology. 1984;34:939-944.

12. Roman GC, Tatemichi TK, Erkinjuntti T, et al. Vascular dementia: diagnostic cri-teria for research studies. Neurology. 1993;43:250-260.

13. UNESCO. International Standard Classification of Education (ISCED). Paris, France:UNESCO; 1976.

14. Wenham PR, Price WH, Blandell G. Apolipoprotein E genotyping by one-stagePCR. Lancet. 1991;337:1158-1159.

15. Elias MF, Beiser A, Wolf PA, Au R, White RF, D’Agostino RB. The preclinical phaseof alzheimer disease. Arch Neurol. 2000;57:808-813.

16. Barber R, Scheltens P, Gholkar A, et al. White matter lesions on magnetic reso-nance imaging in dementia with Lewy bodies, Alzheimer’s disease, vascular de-mentia, and normal aging. J Neurol Neurosurg Psychiatry. 1999;67:66-72.

17. Pantoni L, Garcia JH. Pathogenesis of leukoaraiosis. Stroke. 1997;28:652-659.18. Filley CM. The behavioral neurology of cerebral white matter. Neurology. 1998;

50:1535-1540.19. Fazekas F, Kapeller P, Schmidt R, Offenbacher H, Payer F, Fazekas G. The rela-

tion of cerebral magnetic resonance signal hyperintensities to Alzheimer’s disease.J Neurol Sci. 1996;142:121-125.

20. Barber R, Gholkar A, Scheltens P, Ballard C, McKeith IG, O’Brien JT. MRI volu-metric correlates of white matter lesions in dementia with Lewy bodies and Alz-heimer’s disease. Int J Geriatr Psychiatry. 2000;15:911-916.

21. Selden NR, Gitelman DR, Salamon-Murayama N, Parrish TB, Mesulam MM. Tra-jectories of cholinergic pathways within the cerebral hemispheres of the humanbrain. Brain. 1998;121:2249-2257.

22. Bozzali M, Falini A, Franceschi M, et al. White matter damage in Alzheimer’s dis-ease assessed in vivo using diffusion tensor magnetic resonance imaging. J Neu-rol Neurosurg Psychiatry. 2002;72:742-746.

23. Takahashi S, Yonezawa H, Takahashi J, Kudo M, Inoue T, Tohgi H. Selective reduc-tion of diffusion anisotropy in white matter of Alzheimer disease brains measuredby 3.0 Tesla magnetic resonance imaging. Neurosci Lett. 2002;332:45-48.

24. Neuropathology Group of the Medical Research Council Cognitive Function andAgeing Study (MRC CFAS). Pathological correlates of late-onset dementia in amulticentre, community-based population in England and Wales. Lancet. 2001;357:169-175.




white_matter_lesions_and_the_risk_of_dementia.pdf

Documents