482

The Association between White Matter Lesions on Magnetic Resonance Imaging and Noncognitive Symptoms

JOHN O’BRIEN,a ROBERT PERRY, ROBERT BARBER, ANIL GHOLKAR,AND ALAN THOMAS

Institute for the Health of the Elderly, Newcastle General Hospital,Newcastle upon Tyne, UK

ABSTRACT: A number of studies have suggested that cerebral changes, particu-larly deep white matter lesions (WML) visualized on magnetic resonance im-aging (MRI), may be involved in the genesis of late life depression. This hasbeen confirmed in a prospective study which also found a relationship betweenthe presence of WML and poor 3-year outcome in elderly depressed subjects.Most studies find these lesions to predominate in frontal lobe and basal ganglia,supporting the hypothesis of “fronto-striatal” dysfunction in depression. To in-vestigate whether WML are associated with mood disturbance in dementia,proton density and T2-weighted images were obtained in 80 subjects withdementia (dementia with Lewy bodies, n = 27; Alzheimer’s disease, n = 28;vascular dementia, n = 25) and 26 age-matched normal controls. Periventricu-lar lesions (PVL), white matter lesions (WML), and basal ganglia hyperinten-sities (BG) were visually rated blind to diagnosis using a semiquantitative scale.Frontal WML were associated with higher depression scores in patients withdementia, implying a common pathophysiology of depression irrespective ofdiagnosis. Further study of the neurobiological basis of WML is needed. Thiscan best be achieved by serial clinical assessment combined with in vivo and invitro MRI and neuropathological examination.

WHITE MATTER LESIONS IN DEPRESSION

Neuroimaging studies have shown evidence of structural changes in subcorticaland frontal areas in both unipolar and bipolar affective disorder. Individual studieshave shown ventricular enlargement, sulcal widening, a reduction in overall frontallobe, and specifically left subgenual prefrontal cortex volume and caudate atrophy.1–4

These findings are still controversial, though frontal and caudate abnormalities havealso been demonstrated using functional imaging techniques.5,6 However, the mostconsistent abnormality described in depression comes from studies which have re-peatedly shown an increase in the number and/or severity of signal hyperintensitiesin the white matter visualized on magnetic resonance imaging (MRI).1,2,7,8 Signalhyperintensities can be divided into those adjacent to the ventricular system

aAddress for correspondence: Dr. John O’Brien, Wolfson Research Centre, Institute for theHealth of the Elderly, Newcastle General Hospital, Westgate Road, Newcastle upon Tyne,NE4 6BE, UK. Tel.: +44 (0)191 256 3323; fax: +44 (0)191 219 5051.

e-mail: j.t.o’brien@ncl.ac.uk

483O’BRIEN et al.: WHITE MATTER LESIONS VISUALIZED ON MRI

(periventricular lesions or PVL) and those deep in the white matter (WML).9 An in-crease in hyperintensities, particularly WML, has been described in both unipolarand bipolar disorder but is most marked in elderly depressed subjects, particularlythose with late onset depression.8

Major depressive disorder in older individuals often has a poor prognosis, and ce-rebral organic factors may be predictive of poor outcome. Severe WML on MRI maypredispose to the onset of first ever depressions in some elderly subjects and havebeen reported to be associated with poor initial treatment response.10 An adverse ef-fect of severe WML was recently demonstrated in a prospective follow-up study of54 elderly subjects with depression, all of whom underwent MRI scans at baselinewhich were rated for the presence and severity of WML and PVL.11 Outcome wasrated in accordance with established practice12 as i) continuously well, n = 11(severe WML = 0); ii) well after relapse, n = 8 (1 with severe WML); iii) depressiveinvalid, n = 13 (5 with severe WML); iv) currently in relapse, n = 2 (1 with severeWML); v) continuously ill, n = 5 (0 with WML); vi) demented, n = 7 (3 withsevere WML); and dead, n = 8 (3 with severe WML). Data were analyzed categoriz-ing patients in group (i) as having a “good” outcome and those in groups (ii)–(vii)as having a “bad” one.

Subjects with severe WML had a significantly worse outcome than others, andnone remained continuously well (Fishers exact probability test, p = 0.048). Therewas no association between PVL and outcome. Survival analysis confirmed the ef-fects of severe lesions on poor outcome. Survival functions for the severe and non-severe lesion groups were significantly different (Log rank test statistic 3.63, df = 1,p = 0.04). Median survival time for the 13 subjects with severe lesions was only136 days (95% CI 0,309) compared to 315 days (95% CI 0,813) for those withoutlesions.11

These results demonstrated a significant effect of WML on outcome and, in keep-ing with previous research, show that some elderly patients with depression have ev-idence of damage to deep white matter structures which may play a role in theinitiation, maintenance, and outcome of late life depression. Studies of the basal gan-glia have shown the existence of five parallel, segregated frontal-subcorticalcircuits.13 Two of these are highly relevant to affective disorders, because they in-volve reciprocal links between the caudate nucleus and the dorsolateral prefrontalcortex and the cingulate cortex, the two main frontal areas identified as abnormal byimaging studies in depression.16 There is also good clinical evidence from diseasesaffecting these structures to suggest that these regions are involved in moodregulation. These findings have led to the hypothesis that disruption of the frontal-subcortical pathways, probably related to white matter lesions visualized on MRI,underpins the pathophysiology of depression.

WHITE MATTER LESIONS IN DEMENTIA

Introduction

Dementia is know to be associated with a high prevalence of both WML on MRIand noncognitive symptoms, particularly depression and psychosis. Because of therecognized link between WML and depression discussed above, we wished to test

484 ANNALS NEW YORK ACADEMY OF SCIENCES

the hypothesis that white matter changes in dementia are associated with noncogni-tive symptoms, in particular depression. Full results from this study are presentedelsewhere.14

Methods

Eighty subjects over the age of 60 years who fulfilled DSM-IV criteria for de-mentia were recruited. Diagnosis was made after a detailed clinical assessment. Thisincluded an interview with the subject and the most knowledgeable informant usingthe Geriatric Mental State/History Aetiology Schedule, review of clinical records,full psychiatric and medical history, and mental state and physical examination.14 Astandard dementia screen was completed which included hematology and biochem-istry analysis, thyroid function tests, syphilis serology, B12 and folate levels, andcomputer tomography (CT) scan.

Standardized clinical diagnostic criteria were used to characterize the type ofdementia. Diagnoses of Alzheimer’s disease (AD), vascular dementia (VaD), and de-mentia with Lewy bodies (DLB) were made in accordance with NINCDS/ADR-DA,15 NINDS/AIREN,16 and DLB Consensus criteria17 blind to MRI scan findings.Pathological confirmation of clinical diagnosis has since been acquired in four pa-tients. Applying these criteria, 28 subjects had AD (definite n = 2, probable n = 24,and possible n = 2), 25 had VaD (probable n = 15 and possible n = 10), and 27 hadDLB (definite n = 2, probable n = 24, and possible n = 2).

Within three months of completing an MRI scan all subjects underwent a furtherassessment. Cognitive function was measured using the Cambridge CognitiveExamination which incorporates the Mini Mental State Examination (MMSE).18

Depressive symptoms were rated using the Montgomery and Asberg DepressionRating Scale (MADRS).19

Twenty-six age-matched controls (Con) were recruited from among spouses andfriends of dementia subjects. A detailed history and examination was undertaken toinclude demographic data and physical and psychiatric status. All control subjectscompleted the same assessments as listed above. Exclusion criteria were evidence ofcurrent depression (from history or MADRS >10) or dementia (from history orscore <80 on the CAMCOG) and a history of any other significant neurological,physical, or psychiatric disorder including drug and alcohol abuse.

All scans were performed on a 1.0 Tesla Siemens Magnetom Impact MRIScanner. Whole brain axial images of 5 mm thickness (0.5 mm gap) were obtainedusing proton density weighted and T2-weighted turbo/fast spin echo sequences to al-low detailed visualization of white matter lesions (RARE technique-rapid acquisi-tion with relaxation enhancement: TR = 2800 ms; TE 14/85 ms; matrix 256 × 256;field of view = 230 mm giving pixel size = 0.92 × 0.92 mm; acquisition time = 4 min,13 sec).

White matter lesions were rated from hard copies of proton density and T2-weighted axial images using an established scale which provided a semiquantitativemeasurement of the type, size, frequency, and location of PVL and WML.9

485O’BRIEN et al.: WHITE MATTER LESIONS VISUALIZED ON MRI

Results

Subject characteristics are summarized in TABLE 1. Groups were well matchedfor age, sex, length of history, and years of education. As would be expectedCAMCOG and MMSE scores were significantly lower in all dementia groups com-pared to controls (p <0.001). Subjects with DLB were significantly more impairedthan those with VaD on MMSE (13.6 vs 18.2; p <0.05) and CAMCOG (46.1 vs 62.0;p <0.05). There were no differences between dementia groups with regard toMADRS scores though all dementia groups had significantly more depressive symp-toms than controls (p <0.05). PVL were positively correlated with age in all subjects(total PVL score r = 0.41, p <0.001) but were not associated with cognitive impair-ment or depressive symptoms.

A link between regional deep white matter changes and depression was observed.Those with frontal WML had significantly higher depression scores than those with-out lesions (mean MADRS score for subjects with frontal WML = 8.3 vs without =3.4; p <0.05).

Conclusion

Depression is a frequent and clinically important feature of all dementias. As pre-viously discussed, frontal and subcortical WML are implicated in late-life depres-sion and influence outcome, possibly by disruption of frontal-subcortical circuits.The findings of this study suggest frontal WML may also be relevant in understand-ing depression in dementia, and imply a common pathophysiology of depression ir-respective of diagnosis.

TABLE 1. Subject characteristics

DLB(n = 27)

AD(n = 28)

VaD(n = 25)

Con(n = 26) p Value

Age (mean (SD)) 75.9 (7) 77.4 (5) 76.8 (7) 76.2 (5) ns

Sex (M:F) 19:9 10:18 15:10 14:12 ns

Education (years (SD)) 9.1 (1) 9.7 (3) 9.9 (1) 10.1 (2) ns

Length of history (mo (SD)) 38.2 (19) 42.4 (25) 39.7 (25) na ns

MMSE (mean (SD)) 13.6 (7) 15.5 (5) 18.2 (4) 28.1 (2) < 0.001*

CAMCOG (mean (SD)) 46.1 (26) 55.5 (16) 62.0 (13) 97.2 (5) < 0.001*

MADRS (mean) 8.0 7.6 7.7 3.4 < 0.05**

ABBREVIATIONS: ns, not significant; na, not applicable; SD, standard deviation; DLB, dementiawith Lewy bodies; AD, Alzheimer’s disease; VaD, vascular dementia; Con, controls.

*post hoc Scheffé test showed significant differences between Con and AD, Con and VaD,Con and DLB (p < 0.001) and between VaD and DLB (p < 0.05). **Significant differencesbetween Con and DLB, AD and VaD.

486 ANNALS NEW YORK ACADEMY OF SCIENCES

FIG

UR

E1.

(A)

Pro

ton

dens

ity

(TR

= 2

800

ms,

TE

= 1

4 m

s) a

nd (

B)

T2-

wei

ghte

d (T

R=

2800

ms,

TE

= 8

5 m

s) a

xial

in

vivo

sca

ns o

f el

derl

ypa

tien

t w

ith

depr

essi

on. N

ote

the

pres

ence

of

peri

vent

ricu

lar

and

scat

tere

d de

ep w

hite

mat

ter

lesi

ons,

esp

ecia

lly

in f

ront

al l

obe

(arr

ow).

487O’BRIEN et al.: WHITE MATTER LESIONS VISUALIZED ON MRI

FIG

UR

E2.

(A)

Axi

al p

roto

n de

nsit

y po

stm

orte

m s

can

(TR

= 2

800

ms,

TE

= 1

4 m

s) o

f w

hole

hem

ibra

in o

f sa

me

pati

ent a

fter

for

mal

in f

ixat

ion.

Not

e th

at t

he s

ame

peri

vent

ricu

lar

and

deep

whi

te m

atte

r le

sion

s ca

n be

vis

uali

zed

as s

een

on t

he i

n vi

vo s

can.

Arr

ow d

enot

es s

ame

fron

tal

lesi

on.

(B)

T2-

wei

gthe

d im

age

(TR

= 2

800

ms,

TE

= 8

5 m

s) o

f sa

me

slic

e.

488 ANNALS NEW YORK ACADEMY OF SCIENCES

NEUROPATHOLOGICAL BASIS OF LESIONS

Introduction

These clinical studies have demonstrated the potential importance of deep whitematter changes in understanding important noncognitive symptoms like depression.However, our understanding of the neuropathology of such lesions is limited. Somestudies have examined the pathology of hyperintensities seen on MRI, but none hasincluded any patients with affective disorder.20–25 WML can be found as punctatelesions (first degree) with reduced myelination and atrophy of the neuropil aroundfibrohyalinotic arterioles as well as confluent lesions (second and third degree) withgliosis and lacunar infarcts. They are associated with vascular risk factors. In con-trast PVL (unless extensive) are probably not associated with vascular disease andusually represent areas of demyelination associated with discontinuity of the sub-ependymal lining.24

To further investigate the neuropathological basis of white matter change in af-fective disorder we are currently undertaking longitudinal clinicopathological stud-ies in depression and dementia combining in vivo and in vitro MRI withneuropathological analysis. In vivo scans are acquired as above on a 1.0 T Siemensscanner (FIG. 1). In cases where autopsy brain donation is achieved whole hemi-brain MRI is acquired using the same imaging sequences, firstly in the axial planeto allow comparison with the in vivo scan (FIG. 2) and then coronally to facilitate ac-curate localization of lesions after coronal sectioning. Finally, after coronal section-ing slices are rescanned to determine that lesions are still present, to ensure artefactshave not appeared, and to accurately localize lesions to particular slices.

Conclusion

WML on MRI, particularly in frontal and subcortical areas, are associated withdepressive disorder and depressive symptoms in dementia, implying a commonpathophysiology of depression caused by disruption to fronto-striatal circuits. Theselesions predict poor outcome in major depression, and understanding their pathogen-esis will be important in terms of understanding the neurobiology of late life depres-sion and may offer important new insights into therapeutic options for the treatmentand prevention of depression.

REFERENCES

1. VIDEBECH, P. 1997. MRI findings in patients with affective disorder: a meta-analysis.Acta Psychiatr. Scand. 96: 157–168.

2. RABINS, P.V., G.D. PEARLSON, E. AYLWARD, A.J. KUMA et al. 1991. Cortical magneticresonance imaging changes in elderly inpatients with major depression. Am. J.Psychiatry 148: 617–620.

3. DREVETS, W.C., J.L. PRICE, J.R. SIMPSON, R.D. TODD et al. 1997. Subgenal prefrontalcortex abnormalities in mood disorders. Nature 386: 824–827.

4. COFFEY, C.E., W.E. WILKINSON, R.D. WEINER, L.A. PARASHOS et al. 1993. Quantitativecerebral anatomy in depression. Cereb. Anat. 50: 7–16.

5. BEACH, C.J., K.J. FRISTON, R.G. BROWN, L.C. SCOTT et al. 1992. The anatomy ofmelancholia—focal abnormalities of cerebral blood flow in major depression.Psychol. Med. 22: 604–615.

489O’BRIEN et al.: WHITE MATTER LESIONS VISUALIZED ON MRI

6. GOODWIN, G.M., M.P. AUSTIN, N. DOUGALL, M. ROSS et al. 1993. State changes inbrain activity shown by the uptake of 99mTc-exametazine with single photon emis-sion tomography in major depression before and after treatment. J. Affect. Disord.29: 243–253.

7. O’BRIEN, J.T., D. AMES & I. SCHWEITZER. 1996. White matter changes in depressionand Alzheimer’s disease: a review of magnetic resonance imaging studies. Int. J.Geriatr. Psychiatry 11: 681–694.

8. O’BRIEN, J., P. DESMOND, D. AMES, I. SCHWEITZER, S. HARRIGAN & B. TRESS. 1996.A magnetic resonance imaging study of white matter lesions in depression andAlzheimer’s disease. Br. J. Psychiatry 168: 477–485.

9. SCHELTENS, P., F. BARKHOF, D. LEYS et al. 1993. A semiquantitative rating scale for theassessment of signal hyperintensities on magnetic resonance imaging. J. Neurol. Sci.114: 7–12.

10. SIMPSON, S., A. JACKSON, R.C. BALDWIN & A. BURNS. 1997. Subcortical hyperintensi-ties in late-life depression: acute response to treatment and neuropsychologicalimpairment. Int. Psychogeriatr. 9: 257–275.

11. O’BRIEN, J.T., E. CHIU, I. SCHWEITZER, P. DESMOND & B. TRESS. 1998. Severe deepwhite matter lesions on MRI brain scan predict poor outcome in elderly patients withmajor depressive disorder. Br. Med. J. 317: 982–984.

12. AMES, D. & N. ALLEN. 1991. The prognosis of depression in old age: good, bad orindifferent? Int. J. Geriatr. Psychiatry 6: 477–481.

13. ALEXANDER, G.E. & M.D. CRUTCHER. 1990. Functional architecture of basal gangliacircuits neural substrates of parallel processing. TINS 13 (7): 266–271.

14. BARBER, R., A. GHOLKAR, C. BALLARD, P. SCHELTENS, I. MCKEITH, R. PERRY, P. INCE,D. BARER & J.T. O’BRIEN. 1999. White matter lesions on MRI in dementia withLewy bodies, Alzheimer’s disease, vascular dementia and normal ageing. J. Neurol.Neurosurg. Psychiatry 67: 66–93.

15. MCKHANN, G., D. DRACHMAN, M. FOLSTEIN et al. 1984. Clinical diagnosis ofAlzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspicesof Department of Health and Human Service Task Force on Alzheimer’s Disease.Neurology 34: 939–944.

16. ROMAN, G.C., T. TATEMISCHI, T. ERKINJUNTTI et al. 1993. Vascular dementia: diagnos-tic criteria for research studies. Report of the NINDS-AIRENS internationalworkshop. Neurology 43: 250–260.

17. MCKEITH, I.G., D. GALASKO & K. KOSAKA. 1996. Consensus guidelines for the clinicaland pathological diagnosis of dementia with Lewy bodies (DLB): report of the con-sortium on DLB international workshop. Neurology 47: 1113–1124.

18. ROTH, M., E. TYM, C. MOUNTJOY et al. 1986. CAMDEX. A standardised instrument forthe diagnosis of mental disorder in the elderly with special reference to the earlydetection of dementia. Br. J. Psychiatry 149: 698–709.

19. MONTGOMERY, S.A. & M. ASBERG. 1979. A new depression scale designed to be sensi-tive to change. Br. J. Psychiatry 134: 382–389.

20. LEIFER, D., F.S. BUONANNO & E.P. RICHARDSON, JR. 1990. Clinicopathologic correla-tions of cranial magnetic resonance imaging of periventricular white matter. Neurology40: 911–918.

21. VAN SWIETEN, J.C., H.W. VAN DEN HOUT, B.A. VAN KETEL et al. 1991. Periventricularlesions in the white matter on magnetic resonance in the elderly. Brain 114: 761–774.

22. WALDEMAR, G., P. CHRISTIANSEN, H.B.W. LARSSON et al. 1994. White matter resonancehyperintensities in dementia of the Alzheimer type: morphological and regional cere-bral blood flow correlates. J. Neurol. Neurosurg. Psychiatry 57: 1458–1465.

23. CHIMOWITZ, M.I., M.L. ESTES, A.J. FURLAN et al. 1992. Further observations on thepathology of subcortical lesions identified on magnetic resonance imaging. Arch.Neurol. 49: 747–752.

24. FAZEKAS, F., R. KLEIMERT, H. OFFENBACHER et al. 1993. Pathologic correlates of inci-dential MRI white matter hyperintensities. Neurology 43: 1683–1689.

25. SCHELTENS, P., F. BARKHOF, D. LEYS et al. 1995. Histopathologic correlates of whitematter changes on MRI in Alzheimer’s disease and normal aging. Neurology 45:883–888.