in MS:
Maurelli M, Marchioni E, Cerretano R, Bosone D, Bergamaschi R, Citterio A, Martelli A, Sibilla L, Savoldi F. Neuropsychological assessment in MS: clinical, neurophysiological and neuroradiological relationships. Acta Neurol Scand 1992: 86: 124-128.
M. Maurelli, E. Marchioni, R. Cerretano, D. Bosone, R. Bergamaschi, A. Citterio, A. Martelli, L. Sibilla, F. Savoldi Neurological Institute "C. Mondino", University of Pavia, Italy
~
We assessed cognitive performance and its relationship with clinical and anatomic disease severity in MS with mild to moderate handicap; 34 definite MS and 18 healthy subjects matched for age and education were submitted to a neuropsychological test battery. Both groups were examined for anxiety. MS patients underwent magnetic resonance imaging examination. MS performed worse than controls on all WAIS-P subtests and had learning, short- and long-term verbal memory impairment. Cognitive deficits were not related to abnormal emotional states, but were found to be associated with attentional process and information-processing speed impairment. Cognitive impairment did not correlate with severity of physical disability. The most severe memory deficits were found in patients with extensive periventricular damage.
Deficits in cognitive functioning and memory performance ire common findings in patients with MS (1-8). Recent frequency estimates of cognitive dysfunction in MS have ranged from 43 to 65% (5-8). Nevertheless, the relation between these disturbances and severity of disease is still unclear. In fact, results of correlation studies between cognitive impairment with either disability score or degree of cerebral demyelination are conflicting (3, 6, 8-14). The present study investigated cognitive functioning and its relationships with clinical and anatomic disease severity in MS patients with mild to moderate disability. Material and methods
Thirty-four right-handed subjects were selected. Inclusion criteria was: clinically definite MS (15); Kurtzke Expanded Disability Status Scale (EDS S) score (16) equal or inferior to 6.5; clinically stable phase of disease (17); no severe strength/ coordination or visual impairment that might interfere with cognitive testing; educational level of at least five years; no history of alcohol/drug abuse, psychiatric or nervous system diseases other than MS; willingness to participate in the study. 29 of these 34MS with mild (EDDS 0-4) or moderate 124
Key words: multiple sclerosis; cognitive functioning: attention: information processing speed; MRI Maurizia Maurelli, lstituto Neurologico "C. Mondino", via Palestro, 3, 27100 Pavia, Italy
I
Accepted for publication December 13, 1991
(EDSS 4.5-6.5) physical disability, had a relapsing remitting and 5 patients had a relapsing progressive course. Neuropsychological assessment was performed in all 34MS. MRI of the brain was performed in only 21 of 34 MS selected within 2 to 15 days after neuropsychological testing. Eighteen normal controls, matched for average age and education to the MS, were recruited from family members and friends of MS patients or from hospital staff. None had a neurologic history or signs of alcohol/drug abuse, psychiatric ornervous system disorders. No patient or control had been submitted to these psychological tests previously. Depressed subjects were excluded. Characteristics are summarized in Table 1. Neuropsychological assessment
The following tests were administered to MS and controls in the same sequence during two sessions lasting approximately 1.5 h each: Wechsler adult intelligence scale (WAIS); Corsi span (18), a sequential block-tapping task, which is considered the visuospatial equivalent of digit span; Rey auditory verbal learning test (Rey AVLT), a verbal auditory memory task which studies immediate recall (Rl), learning ability (R2 to R5) and retention after 20 min of other
Neuropsychological assessment in MS Table 1. Composition of groups
Table 2. Results of MANCOVA of psychometric measures by cases and controls (data adjusted for age, education) Controls
(n=18) Sex, No M F
MS patients
(n=34)
P Tests
6 males 12 females
11 males 23 females
Age (years) Mean SD Range
37.7 10.0 24-55
38.5 9.8 25-56
n.s.
Education, yr Mean SD Range
10.8 4.5 5-19
9.1 3.9 5-18
ns.
Disease duration, yr* Mean SD Range EDSS Mean SD Range
34 MS meanfSEM
18 CONTROLS meanfSEM
F
P
0.8936 15.3755 36.6316 2.6772 7.0930
0.349 0.000 0.108 0.011
1.2854 5.2096 1.2793 0.6118 0.1899 0.0046 4.2850 17.0057 6.6754 15.6174 14.7842
0.263 0.027 0.264 0.438 0.665 0.946 0.044 0.000 0.013 0.000 0.000
Multivariate Hotellings test F= 1.1081 p=O.OOO
10.1 6.5 3-27 4.1 1.6 1.5-6.5
Verbal I.Q. Performance I.Q. Memory Quotient STAI-X, STAI-X,
94.9f1.89 89.8f2.56 90.8f2.22 45.7f1.84 46.3f1.56
100.5f2.87 105.8f1.93 114.2f3.72 41.8f1.98 38.8f2.02
0.000
Multivariate Hotellings test F=0.8513 p=0.007 WAIS Information Comprehension Arithmetic Digit span Similarities Vocabulay Picture completion Picture arrangement Block design Digit symbol Object assembly
10.0f0.39 7.9f0.46 8.8f0.48 8.2f0.51 10.0f0.46 9.0f0.34 8.5f0.41 6.6f0.53 8.8f0.41 6.0f0.50 7.1f0.53
11.2&0.63 7.5f0.73 10.1f0.56 9.3f0.56 10.2k0.68 9.1f0.58 10.1f0.35 10.1f0.40 10.6f0.54 9.5f0.61 9.5f0.51
* Since onset of symptoms. Multivariate Hotellings test F= 1.8104 p=O.OOO
activities (RR); Wechsler memory scale (WMS); two anxiety inventories for state (STAI-x,) and trait (STAI-x,) anxiety (19). MRI assessment
Brain MRI was performed on a Philips Gyroscan T.5 0.5T. imager using contiguous 8 mm slices, made with Spin Echo and F.F.E. sequences. Especially T,- weighted images were considered for the white matter. Analysis of demyelinized areas was done according to Ormerod’s method (20) by a neuroradiologist unaware of the clinical and neuropsychological findings. This method evaluates presence and size of lesions in seven periventricular and in six parenchymal areas. Data analysis
Student’s t test was performed only for age and education between the MS and controls. Psychometric scores of MS and controls were compared by means of three multivariate analysis of the variance models adjusting for age and education. The variables were thus grouped: general cognitive and memory indicators, WAIS subtests, memory tests (Table 2). Assessment of variances homogeneity was verified by means Mauchly sphericity test. According to Bonferroni’s correction an alpha level equal to 0.05/3 (0.016) was chosen as cutoff of significance for Hotelling T-test of MANCOVA. MANCOVA
WMS Information Orientation Mental control Digit span (forward) Digit span (backward) Logical memory Visual reproduction Associate learning (easy) Associate learning (hard) CORSI SPAN
5.2f0.13 4.9f0.03 5.9f0.32 5.6f0.17 3.9f0.25 6.2f0.39 7.3f0.53
5.8f0.07 4.9f0.04 8.0f0.30 6.1f0.23 4.6f0.23 11.8f0.75 8.0f0.66
7.0227 0.0221 12.9140 0.9329 1.4249 19.7771 0.1307
0.011 0.882 0.001 0.339 0.339 0.000 0.719
8.6f0.10
8.9f0.02
4.4856
0.039
5.0f 0.46 5.3f0.22
8.6k0.51 4.8f0.33
24.7460 2.1289
0.000
5.1f0.29 7.7f0.30 9.4f0.30 10.1f0.05 11.0f0.43 11.0f0.62
5.7f0.44 9.8f0.63 11.6k0.51 12.6f0.35 13.4i0.40 14.7f0.09
0.8910 8.1199 12.1910 17.2296 12.0730 15.9307
0.314 0.006 0.001 0.000 0.001 0.000
0.151
Rey AVLT
R1 R2 R3 R4 R5 RR
test results in Tables 2, 3 represent the adjusted means of age and education standard error mean (SEM). In the MS group some variables (EDSS, age, education, disease duration, MRI scores) were correlated with psychometric scores by means of the Spearman rank correlation test. In the MS group submitted to brain MRI one MANCOVA and one MANOVA models were performed (Table 3). Patients with scores 2SDs below normal control performance on at least two or three WAIS-P subtests, three WMS subtests and on learning/ 125
Maurelli et al. Table 3. Comparison of demographic, clinical and neuropsychological data in patients without (Group 1 ) and with (Group 2)extensive periventriculardemyelination (MANOVA, MANCOVA adjusted for age and education) Group 1
(N=13)
Group 2 (N=8)
F
P
MANOVA (Multivariate Hotellings test F=2.5298p=.008) Age (years) Scolarity (years) Length of disease (years) EDSS Frontal score (MRI) Temporal score (MRI) Verbal 1.Q. Performance I.Q. Memory Quotient STAI-X, STAI-X,
mean (SD)
mean (SD)
39.3(10.4) 8.2 (3.3) 9.7 (4.9) 3.5 (1.4) 3.3 (4.3) 5.2 (5.6) 97.1 (4.9) 96.7(14.1) 95.6 (8.6) 48.1(11.1) 47.6 (9.7)
35.5 (8.2) 0.7963 11.3 (3.8) 3.9723 8.2 (5.5) 0.3897 3.9 (1.6) 0.4047 12.2 (4.9) 19.0842 11.8 (3.2) 9.2477 91.1(17.6) 6.0866 85.5(16.1) 2.5225 82.7(18.6) 10.6404 47.2(13.9) 0.0024 43.8 (9.8) 1.0533
0.383 0.061 0.540 0.532 0.000 0.007 0.025 0.131 0.005 0.961 0.319
MANCOVA (Multivariate Hotellings test k5.1163 p=0.760) meanfSEM
meanfSEM
Information Orientation Mental control Digit span Logical memory Visual reproduction Associate learning (easy) Associate learning (hard)
5.3f0.16 4.9k0.08 6.5f0.16 8.3f0.58 6.5f0.77 7.6f0.66
5.2f0.46 5.0f0.03 5.3k1.00 7.2k1.46 4.8f0.57 6.8f1.42
0.0555 0.0734 2.4736 4.6408 5.8689 1.1829
0.817 0.790 0.134 0.046 0.027 0.292
8.4f0.19
8.5f0.35
0.1740
0.682
5.8f0.77
3.2f0.92
5.1871
0.036
Corsi span
5.8f0.27
4.8f0.42
2.8387
0.110
5.1t0.52 8.2f0.41 9.7f0.47 10.6f0.48 11.6f0.67 11.9f0.97
5.0f0.39 7.8f0.64 8.5f0.53 8.1f0.74 10.5f0.82 9.6f1.60
0.2878 0.5225 3.2678 9.7341 2.3362 3.0570
0.599 0.480 0.088 0.006 0.145 0.098
Rey AVLT
R1 R2 R3 R4 R5 RR
delayed recall of the Rey AVLT were considered to be cognitively impaired. Results
Neuropsychology
No statistical differences in age and years of education were recorded between patient and control groups. Table 2 summarizes all performances parameters. Although in the average range, MS group’s WAIS performance IQ was significantly below the age and education similar control group. This overall difference was due to poorer performance on every WAISP subtest. No statistically significant differences on verbal scale IQ were recorded, however MS patients scored significantly lower on comprehension verbal subtest. The MS group’s memory quotient (MQ) was significantly lower with respect to controls, although 126
remaining in the average range. This discrepancy was due to poorer performance on two short-term memory tasks (logical memory and associate learning) and to significant impairment in attention, as measured by mental control subtest in MS patients. No differences were observed in digit-span and visual reproduction. MS patients performed below the controls in personal and current opinion (information). Learning and long-term memory were impaired in MS patients. In fact they recalled significantly fewer words during learning on the Rey AVLT and also recalled significantly less when long-term recall was required after a 20 min delay. Eleven of 34 MS patients (32.3%) were classified as cognitively impaired when applying a cutoff of 2SDs below normal control performance on at least two or three WAIS-P subtests, three WMS subtests and on learning/delayed recall on Rey AVLT. Although anxiety in MS patients was normally greater (STAI-x,), they faced the neuropsychological tests in the same manner as controls (STAI-x,). In fact state-anxiety scores were not statistically different with respect to controls. In the MS group the Spearman rank correlation test showed a good correlation between both: MQ and WAIS full IQ (r = 0.383, p = 0.02); long-term recall (RR) and R3,R4,R5 (r = 0.366, p = 0.03; r = 0.365, p = 0.03 and r = 0.52, p = 0.002 respectively). However, no neuropsychological score was significantly related to age, education, EDSS and disease duration. Neurophysiology and neuroradiology
MRI abnormalities were detected in all 21 patients studied. This MS sample was representative of the overall MS group for age [ 37.9 (9.6) yr], education [9.4(3.8)yr],diseaseduration [9.1(5)yr] andEDSS [3.7 (1.5)]. As expected, most of the cerebral lesions were located in the periventricular areas. The lesions were quantified in the various categories outlined previously and the four scores obtained (frontal, periventricular, temporal and total) were analyzed with respect to neuropsychological scores, EDSS, age, disease duration by means the Spearman rank correlation test. As expected, the four MRI scores were strongly interrelated. The same analysis failed to reveal any significant correlation between MRI scores and neuropsychological scores, EDS S and disease duration. Patients were then subdivided into two groups, based on presence of extensive, irregular periventricular demyelination (N = 8) or of single or multiple discrete lesions in the periventricular or extraperiventricular regions (N = 13). The patients with extensive periventricular demyelination had a significantly lower memory quotient. This difference
Neuropsychological assessment in MS was due to poorer performances on three subtests: logical memory, associate learning (hard) and digit span (Table 3). Discussion
Cognitive decline was considered very frequent at the more advanced stage of the disease (21) and in chronic progressive MS (22). However, in agreement with same previous reports (2-3, 9, 12), our results proved that cognitive disturbances, especially in the verbal learning and recall areas, also occur in patients with mild or moderate physical disability. The frequency of cognitive dysfunction in our MS population was 32.3 %, lower than estimates of 43 % to 65 % reported in other recent works (5-8). However several factors may concur. We excluded patients in active phase of the disease (5.7) and/or with severe disability (EDSS > 6.5) (5, 6, 8) or unable to perform testing because of severe motor or visual impairment, we studied greater proportion of relapsing remitting than relapsing progressive MS (6, S), finally differences in sensitivity of neuropsychological methods (5-8) may contribute. Memory disturbance is one of the most consistently impaired cognitive functions in MS, but no convincing and univocal explanation concerning the neuropathological basis of this deficit has yet been proposed. Van den Burg (12) showed that memory defects can not be attributed to impairment in attentional process nor to increasing fatigue during testing but only to poor initial learning. More recently Litvan (2) proved that the long-term verbal memory deficit in MS patients was associated with a significantly slowed information processing. This finding was confirmed by Beatty (23-24) and by Rao (25), yet some authors (8,26) emphasized attentional process impairment . In this regard, we noted an impairment in substained attention, as measured by the mental control subtest from the WMS, which didn’t depend on abnormal emotional status. In fact, none of the subjects studied were depressed. Moreover, even if MS patients were habitually more anxious than controls (STAI-x,), they faced testing in the same manner as controls, therefore, poorer performances were not related to pathological anxiety states during testing (STAI-x,). Furthermore, in contrast to substantially unaffected verbal abilities, MS patients scored lower on all WAIS’s performance subtests which greatly depend on attention, psychomotor and visuospatial abilities and require speed of information processing. The same patients, however, scored as well as controls in other neuropsychological tasks similarly dependent on visuospatial or strength/coordination skills (i.e. Corsi span and visual reproduction subtest
of the WMS), while performed worse in others which do not have psychomotor components (i.e. Rey auditory verbal learning test). Therefore, in our opinion, an impairment in attentional process and in information processing speed could almost partially explain the lower performances in WAIS performance subtests and the poorer learning of patients versus controls. While some investigators (5-6, 8, 10) have noted statistically significant group differences on visuospatial and visual memory tasks, our results suggested that these memory abilities were preserved in contrast to verbal memory. This disagreement between our results and other reports could be due either to discrepancies in MS population characteristics or to different sensitivity of visuospatial tasks employed. In this regard, the Corsi span which we used, failed to detect a visuospatial impairment in other two recent reports of cognitive impairment in mild disabilited MS patients (9,27). Reports are also conflicting regarding the relationship between physical and cognitive disabilities (1 1). In agreement with many authors (3, 12-14), cognitive functioning in our study was neither related to disability scores nor to disease duration. This finding is supported by the observation that mild disturbances in intelligence and memory functions were also present in patients with mild disability (3, 9, 12, 27) as well as in early phases of the disease (7, 2829). CT and more recently MRI have been used to determinate whether a particular anatomic pattern correlates with cognitive impairment in MS. Some studies (9-10, 27) reported good correlation with anatomic involvement (especially with callosal atrophy and periventricular lesion burden), while MRI parameters in other studies were not related to any neuropsychological dysfunction (2,3). These discrepancies were very probably due to methodological differences in neuroradiological and neuropsychological parameters. considered. Despite lack of correlation between the MRI and neuropsychological scores, we detected a significantly inferior memory quotient in patients with MRI evidence of extensive periventricular demyelination. This finding seems to support the hypothesis that memory deficit in MS could result from the disconnection between prefrontal and limbic structures secondary to periventricular demyelinized lesions disrupting the white matter fibers pathway (1). In conclusion, given the frequency of cognitive impairment in mildly disabled MS patients, cognitive functioning must also be systematically investigated in early stages of the disease with a comprehensive neuropsychological battery including intelligence and memory tests, measures of speed of mental processing, substained attention, language 127
Maurelli et al. 3. MARSHGG. Disability and intellectual function in multiple sclerosis patients. J Nerv Ment Dis 1980: 168: 758-762. 4. BEATTYWW, GOODKIN DE, HERTSGAARD D, MONSON N. Clinical and demographic predictors of cognitive performance in multiple sclerosis. Arch Neurol 1990: 47: 305-308. 5. POSERCM, PATYDW, SCHEINBERG L et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Anna1 Neurol 1983: 13: 227-231. 16. KURTZKEJS. Rating neurological impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983: 33: 1444-1452. 17. SCHUMAKER GA, BEEBEG, KIBLERR F et al. Problems of experimental trials of therapy in multiple sclerosis: report by the panel on the evaluation of experimental trials of therapy in multiple sclerosis. Am NY Acad Sci 1965: 122: 522-568. Acknowledgement 18. PAGANO G, VALLARG. Verbal and special immediate meWe acknowledge Dr. M. Grassi, of the Biometric Institute, Unimory span. Normative data from 1355 adults and 1112 chilversity of Pavia, for statistical revision. dren. Ital J Neurol Sci 1987: 8: 539-548. 19. SPIELBERGER CD, GORSUCHRL, LUSHENERE. STAI manual for the state-trait anxiety inventory. Polo Alto, CA: References Consulting Psychology Press, 1970. 20. ORMEROD IEF, MILLERDH, MCDONALDWI et al. The 1. RAO SM. Neuropsychology of multiple sclerosis: a critical role of NMR imaging in the assessment of multiple sclerosis review. J Clin Exper Neuropsychol 1986: 8: 503-542. and isolated neurological lesions: a quantitative study. Brain J, VENDRELL P, MARTINEZJM. 2. LITVANI, GRAFMAN 1987: 110: 1579-1616. Slowed information processing in multiple sclerosis. Arch 21. STAPLESD, LINCOLNNB. Intellectual impairment in mulNeurol 1988: 45: 281-285. tiple sclerosis and its relation to functional abilities. Rheu3. LITVANI, GRAFMAN J, VENDRELL P et al. Multiple mematol Rehab 1979: 18: 153-160. mory deficits in patients with multiple sclerosis. Arch Neurol 22. HEATONRH, NELSONLM, THOMPSON DS, BURKSJS, 1988: 45: 607-610. FRANKLIN GM. Neuropsychological findings in relapsing4. BEATTYWW, GOODKIN DE. Screening for cognitive imremitting and chronic progressive multiple sclerosis. J Conpairment in multiple sclerosis. Arch Neurol 1990: 47: 297sult Clin Psycho1 1985: 53: 103-110. 301. DE, MONSONN, BEATTYPA, 23. BEATTYWW, GOODKIN 5. TRUELLE JL, PALISSONE, LE GALLD, STIP E, DERHERTSGAARD D. Anterograde and retrograde amnesia in OUESNE c . Troubles intellectuels et thymiques dans la sckpatients with chronic progressive multiple sclerosis. Arch rose en plaques. Rev Neurol (Paris) 1987: 143: 595-601. Neurol 1988: 45: 611-619. 6. RAOSM, HAMMEKE TA, MCQUILLEN MP, KHATRIBO, DE, MONSONN, BEATTYPA. 24. BETTYWW, GOODKIN LLOYDD. Memory disturbance in chronic progressive mulCognitive disturbances in patients with relapsing remitting tiple sclerosis. Arch Neurol 1984: 41: 625-631. multiple sclerosis. Arch Neurol 1989: 46: 1113- 1119. 7. LYON-CAEN 0, JOUVENTR, HAUSERS etal. Cognitive 25. RAOSM, ST AUBIN-FAUBERT P, LEOGJ. Information profunction in recent-onset demyelinating diseases. Arch Neurol cessing speed in patients with multiple sclerosis. J Clin Exp 1986: 43: 1138-1141. Neuropsychol 1989: 11: 471-477. 8. RAOSM, LEOGJ, BERNARDIN L, UNVERZAGT F. Cogni26. FEALLYDCM, HEATONRK, NELSONLM, BURKSJS, tive dysfunction in multiple sclerosis. I Frequency, pattern FRANKLINGM. A comparison of dementia in Alzheimer’s and prediction. Neurology 1991: 41: 685-691. disease and in multiple sclerosis. Arch Neurol 1988: 45: 61 19. ANZOLAGP, BEVILACQUA L, CAPPASF et al. Neuropsy619. chological assessment in patients with relapsing-remitting 27. POZZILLI C, PASSAFIUME D, BERNARDI S et al. SPECT, multiple sclerosis and mild functional impairment: correlation MRI and cognitive functions in multiple sclerosis. J Neurol with magnetic resonance imaging. J Neurol Neurosurg PsyNeurosurg Psychiat 1991: 54: 110-115. chiat 1990: 53: 142-145. 28. GRANTI, MCDONALD WI, TRIMBEE MR. Neuropsycho10. HUBER JS, PAULSON GW, SHUTTLEWORTH EC et al. Maglogical impairment in early multiple sclerosis. In: JENSENK netic resonance imaging correlates of dementia in multiple et al., eds. Current problems in neurology, 10. Mental disorsclerosis. Arch Neurol 1987: 44: 732-736. ders and cognitive deficits in multiple sclerosis. London: John E, KNUDSEN L, JENSENK. Multiple sclerosis: 11. STENAGER Libbey, 1989: 17-26. correlation of cognitive dysfunction with Kurtzke disability 29. FRANKLIN GM, NELSONLM, FILLEYCM, HEATONRK. status scale. In: JENSENK et al., eds. Current problems in Cognitive loss in multiple sclerosis. Case reports and review Neurology, 10. Mental disorders and cognitive deficits in mulof the literature. Arch: Neurol. 1989: 46: 162-167. tiple sclerosis. London: John Libbey, 1989: 27-37. 30. RAOSM, LEOGJ, ELLINGTON L, NAUERTZ T, BERNARAH, MINDERHOUD 12. VAN DEN BURGW, VAN ZOMEREN DIN L, UVERZAGT F. Cognitive dysfunction in multiple scleJM, PRANGEAJA, MEIJERNSA. Cognitive impairment in rosis, 2. Impact on employment and social functioning. Neupatients with multiple sclerosis and mild physical disability. rology 1991: 41: 692-696. Arch Neurol 1987: 44: 494-501.
and visuospatial skills. However, only uniform methodology studing a very large sample of patients representative of the heterogeneous MS population could lead to comprehension of neuropathological and anatomic basis of cognitive dysfunctioning in MS. This strategy is also necessary to establish guidelines for informed and effective long-term management of MS patients because cognitive dysfunction negatively influences their employment and social functioning (29-30).
128
Maurelli M, Marchioni E, Cerretano R, Bosone D, Bergamaschi R, Citterio A, Martelli A, Sibilla L, Savoldi F. Neuropsychological assessment in MS: clinical, neurophysiological and neuroradiological relationships. Acta Neurol Scand 1992: 86: 124-128.
M. Maurelli, E. Marchioni, R. Cerretano, D. Bosone, R. Bergamaschi, A. Citterio, A. Martelli, L. Sibilla, F. Savoldi Neurological Institute "C. Mondino", University of Pavia, Italy
~
We assessed cognitive performance and its relationship with clinical and anatomic disease severity in MS with mild to moderate handicap; 34 definite MS and 18 healthy subjects matched for age and education were submitted to a neuropsychological test battery. Both groups were examined for anxiety. MS patients underwent magnetic resonance imaging examination. MS performed worse than controls on all WAIS-P subtests and had learning, short- and long-term verbal memory impairment. Cognitive deficits were not related to abnormal emotional states, but were found to be associated with attentional process and information-processing speed impairment. Cognitive impairment did not correlate with severity of physical disability. The most severe memory deficits were found in patients with extensive periventricular damage.
Deficits in cognitive functioning and memory performance ire common findings in patients with MS (1-8). Recent frequency estimates of cognitive dysfunction in MS have ranged from 43 to 65% (5-8). Nevertheless, the relation between these disturbances and severity of disease is still unclear. In fact, results of correlation studies between cognitive impairment with either disability score or degree of cerebral demyelination are conflicting (3, 6, 8-14). The present study investigated cognitive functioning and its relationships with clinical and anatomic disease severity in MS patients with mild to moderate disability. Material and methods
Thirty-four right-handed subjects were selected. Inclusion criteria was: clinically definite MS (15); Kurtzke Expanded Disability Status Scale (EDS S) score (16) equal or inferior to 6.5; clinically stable phase of disease (17); no severe strength/ coordination or visual impairment that might interfere with cognitive testing; educational level of at least five years; no history of alcohol/drug abuse, psychiatric or nervous system diseases other than MS; willingness to participate in the study. 29 of these 34MS with mild (EDDS 0-4) or moderate 124
Key words: multiple sclerosis; cognitive functioning: attention: information processing speed; MRI Maurizia Maurelli, lstituto Neurologico "C. Mondino", via Palestro, 3, 27100 Pavia, Italy
I
Accepted for publication December 13, 1991
(EDSS 4.5-6.5) physical disability, had a relapsing remitting and 5 patients had a relapsing progressive course. Neuropsychological assessment was performed in all 34MS. MRI of the brain was performed in only 21 of 34 MS selected within 2 to 15 days after neuropsychological testing. Eighteen normal controls, matched for average age and education to the MS, were recruited from family members and friends of MS patients or from hospital staff. None had a neurologic history or signs of alcohol/drug abuse, psychiatric ornervous system disorders. No patient or control had been submitted to these psychological tests previously. Depressed subjects were excluded. Characteristics are summarized in Table 1. Neuropsychological assessment
The following tests were administered to MS and controls in the same sequence during two sessions lasting approximately 1.5 h each: Wechsler adult intelligence scale (WAIS); Corsi span (18), a sequential block-tapping task, which is considered the visuospatial equivalent of digit span; Rey auditory verbal learning test (Rey AVLT), a verbal auditory memory task which studies immediate recall (Rl), learning ability (R2 to R5) and retention after 20 min of other
Neuropsychological assessment in MS Table 1. Composition of groups
Table 2. Results of MANCOVA of psychometric measures by cases and controls (data adjusted for age, education) Controls
(n=18) Sex, No M F
MS patients
(n=34)
P Tests
6 males 12 females
11 males 23 females
Age (years) Mean SD Range
37.7 10.0 24-55
38.5 9.8 25-56
n.s.
Education, yr Mean SD Range
10.8 4.5 5-19
9.1 3.9 5-18
ns.
Disease duration, yr* Mean SD Range EDSS Mean SD Range
34 MS meanfSEM
18 CONTROLS meanfSEM
F
P
0.8936 15.3755 36.6316 2.6772 7.0930
0.349 0.000 0.108 0.011
1.2854 5.2096 1.2793 0.6118 0.1899 0.0046 4.2850 17.0057 6.6754 15.6174 14.7842
0.263 0.027 0.264 0.438 0.665 0.946 0.044 0.000 0.013 0.000 0.000
Multivariate Hotellings test F= 1.1081 p=O.OOO
10.1 6.5 3-27 4.1 1.6 1.5-6.5
Verbal I.Q. Performance I.Q. Memory Quotient STAI-X, STAI-X,
94.9f1.89 89.8f2.56 90.8f2.22 45.7f1.84 46.3f1.56
100.5f2.87 105.8f1.93 114.2f3.72 41.8f1.98 38.8f2.02
0.000
Multivariate Hotellings test F=0.8513 p=0.007 WAIS Information Comprehension Arithmetic Digit span Similarities Vocabulay Picture completion Picture arrangement Block design Digit symbol Object assembly
10.0f0.39 7.9f0.46 8.8f0.48 8.2f0.51 10.0f0.46 9.0f0.34 8.5f0.41 6.6f0.53 8.8f0.41 6.0f0.50 7.1f0.53
11.2&0.63 7.5f0.73 10.1f0.56 9.3f0.56 10.2k0.68 9.1f0.58 10.1f0.35 10.1f0.40 10.6f0.54 9.5f0.61 9.5f0.51
* Since onset of symptoms. Multivariate Hotellings test F= 1.8104 p=O.OOO
activities (RR); Wechsler memory scale (WMS); two anxiety inventories for state (STAI-x,) and trait (STAI-x,) anxiety (19). MRI assessment
Brain MRI was performed on a Philips Gyroscan T.5 0.5T. imager using contiguous 8 mm slices, made with Spin Echo and F.F.E. sequences. Especially T,- weighted images were considered for the white matter. Analysis of demyelinized areas was done according to Ormerod’s method (20) by a neuroradiologist unaware of the clinical and neuropsychological findings. This method evaluates presence and size of lesions in seven periventricular and in six parenchymal areas. Data analysis
Student’s t test was performed only for age and education between the MS and controls. Psychometric scores of MS and controls were compared by means of three multivariate analysis of the variance models adjusting for age and education. The variables were thus grouped: general cognitive and memory indicators, WAIS subtests, memory tests (Table 2). Assessment of variances homogeneity was verified by means Mauchly sphericity test. According to Bonferroni’s correction an alpha level equal to 0.05/3 (0.016) was chosen as cutoff of significance for Hotelling T-test of MANCOVA. MANCOVA
WMS Information Orientation Mental control Digit span (forward) Digit span (backward) Logical memory Visual reproduction Associate learning (easy) Associate learning (hard) CORSI SPAN
5.2f0.13 4.9f0.03 5.9f0.32 5.6f0.17 3.9f0.25 6.2f0.39 7.3f0.53
5.8f0.07 4.9f0.04 8.0f0.30 6.1f0.23 4.6f0.23 11.8f0.75 8.0f0.66
7.0227 0.0221 12.9140 0.9329 1.4249 19.7771 0.1307
0.011 0.882 0.001 0.339 0.339 0.000 0.719
8.6f0.10
8.9f0.02
4.4856
0.039
5.0f 0.46 5.3f0.22
8.6k0.51 4.8f0.33
24.7460 2.1289
0.000
5.1f0.29 7.7f0.30 9.4f0.30 10.1f0.05 11.0f0.43 11.0f0.62
5.7f0.44 9.8f0.63 11.6k0.51 12.6f0.35 13.4i0.40 14.7f0.09
0.8910 8.1199 12.1910 17.2296 12.0730 15.9307
0.314 0.006 0.001 0.000 0.001 0.000
0.151
Rey AVLT
R1 R2 R3 R4 R5 RR
test results in Tables 2, 3 represent the adjusted means of age and education standard error mean (SEM). In the MS group some variables (EDSS, age, education, disease duration, MRI scores) were correlated with psychometric scores by means of the Spearman rank correlation test. In the MS group submitted to brain MRI one MANCOVA and one MANOVA models were performed (Table 3). Patients with scores 2SDs below normal control performance on at least two or three WAIS-P subtests, three WMS subtests and on learning/ 125
Maurelli et al. Table 3. Comparison of demographic, clinical and neuropsychological data in patients without (Group 1 ) and with (Group 2)extensive periventriculardemyelination (MANOVA, MANCOVA adjusted for age and education) Group 1
(N=13)
Group 2 (N=8)
F
P
MANOVA (Multivariate Hotellings test F=2.5298p=.008) Age (years) Scolarity (years) Length of disease (years) EDSS Frontal score (MRI) Temporal score (MRI) Verbal 1.Q. Performance I.Q. Memory Quotient STAI-X, STAI-X,
mean (SD)
mean (SD)
39.3(10.4) 8.2 (3.3) 9.7 (4.9) 3.5 (1.4) 3.3 (4.3) 5.2 (5.6) 97.1 (4.9) 96.7(14.1) 95.6 (8.6) 48.1(11.1) 47.6 (9.7)
35.5 (8.2) 0.7963 11.3 (3.8) 3.9723 8.2 (5.5) 0.3897 3.9 (1.6) 0.4047 12.2 (4.9) 19.0842 11.8 (3.2) 9.2477 91.1(17.6) 6.0866 85.5(16.1) 2.5225 82.7(18.6) 10.6404 47.2(13.9) 0.0024 43.8 (9.8) 1.0533
0.383 0.061 0.540 0.532 0.000 0.007 0.025 0.131 0.005 0.961 0.319
MANCOVA (Multivariate Hotellings test k5.1163 p=0.760) meanfSEM
meanfSEM
Information Orientation Mental control Digit span Logical memory Visual reproduction Associate learning (easy) Associate learning (hard)
5.3f0.16 4.9k0.08 6.5f0.16 8.3f0.58 6.5f0.77 7.6f0.66
5.2f0.46 5.0f0.03 5.3k1.00 7.2k1.46 4.8f0.57 6.8f1.42
0.0555 0.0734 2.4736 4.6408 5.8689 1.1829
0.817 0.790 0.134 0.046 0.027 0.292
8.4f0.19
8.5f0.35
0.1740
0.682
5.8f0.77
3.2f0.92
5.1871
0.036
Corsi span
5.8f0.27
4.8f0.42
2.8387
0.110
5.1t0.52 8.2f0.41 9.7f0.47 10.6f0.48 11.6f0.67 11.9f0.97
5.0f0.39 7.8f0.64 8.5f0.53 8.1f0.74 10.5f0.82 9.6f1.60
0.2878 0.5225 3.2678 9.7341 2.3362 3.0570
0.599 0.480 0.088 0.006 0.145 0.098
Rey AVLT
R1 R2 R3 R4 R5 RR
delayed recall of the Rey AVLT were considered to be cognitively impaired. Results
Neuropsychology
No statistical differences in age and years of education were recorded between patient and control groups. Table 2 summarizes all performances parameters. Although in the average range, MS group’s WAIS performance IQ was significantly below the age and education similar control group. This overall difference was due to poorer performance on every WAISP subtest. No statistically significant differences on verbal scale IQ were recorded, however MS patients scored significantly lower on comprehension verbal subtest. The MS group’s memory quotient (MQ) was significantly lower with respect to controls, although 126
remaining in the average range. This discrepancy was due to poorer performance on two short-term memory tasks (logical memory and associate learning) and to significant impairment in attention, as measured by mental control subtest in MS patients. No differences were observed in digit-span and visual reproduction. MS patients performed below the controls in personal and current opinion (information). Learning and long-term memory were impaired in MS patients. In fact they recalled significantly fewer words during learning on the Rey AVLT and also recalled significantly less when long-term recall was required after a 20 min delay. Eleven of 34 MS patients (32.3%) were classified as cognitively impaired when applying a cutoff of 2SDs below normal control performance on at least two or three WAIS-P subtests, three WMS subtests and on learning/delayed recall on Rey AVLT. Although anxiety in MS patients was normally greater (STAI-x,), they faced the neuropsychological tests in the same manner as controls (STAI-x,). In fact state-anxiety scores were not statistically different with respect to controls. In the MS group the Spearman rank correlation test showed a good correlation between both: MQ and WAIS full IQ (r = 0.383, p = 0.02); long-term recall (RR) and R3,R4,R5 (r = 0.366, p = 0.03; r = 0.365, p = 0.03 and r = 0.52, p = 0.002 respectively). However, no neuropsychological score was significantly related to age, education, EDSS and disease duration. Neurophysiology and neuroradiology
MRI abnormalities were detected in all 21 patients studied. This MS sample was representative of the overall MS group for age [ 37.9 (9.6) yr], education [9.4(3.8)yr],diseaseduration [9.1(5)yr] andEDSS [3.7 (1.5)]. As expected, most of the cerebral lesions were located in the periventricular areas. The lesions were quantified in the various categories outlined previously and the four scores obtained (frontal, periventricular, temporal and total) were analyzed with respect to neuropsychological scores, EDSS, age, disease duration by means the Spearman rank correlation test. As expected, the four MRI scores were strongly interrelated. The same analysis failed to reveal any significant correlation between MRI scores and neuropsychological scores, EDS S and disease duration. Patients were then subdivided into two groups, based on presence of extensive, irregular periventricular demyelination (N = 8) or of single or multiple discrete lesions in the periventricular or extraperiventricular regions (N = 13). The patients with extensive periventricular demyelination had a significantly lower memory quotient. This difference
Neuropsychological assessment in MS was due to poorer performances on three subtests: logical memory, associate learning (hard) and digit span (Table 3). Discussion
Cognitive decline was considered very frequent at the more advanced stage of the disease (21) and in chronic progressive MS (22). However, in agreement with same previous reports (2-3, 9, 12), our results proved that cognitive disturbances, especially in the verbal learning and recall areas, also occur in patients with mild or moderate physical disability. The frequency of cognitive dysfunction in our MS population was 32.3 %, lower than estimates of 43 % to 65 % reported in other recent works (5-8). However several factors may concur. We excluded patients in active phase of the disease (5.7) and/or with severe disability (EDSS > 6.5) (5, 6, 8) or unable to perform testing because of severe motor or visual impairment, we studied greater proportion of relapsing remitting than relapsing progressive MS (6, S), finally differences in sensitivity of neuropsychological methods (5-8) may contribute. Memory disturbance is one of the most consistently impaired cognitive functions in MS, but no convincing and univocal explanation concerning the neuropathological basis of this deficit has yet been proposed. Van den Burg (12) showed that memory defects can not be attributed to impairment in attentional process nor to increasing fatigue during testing but only to poor initial learning. More recently Litvan (2) proved that the long-term verbal memory deficit in MS patients was associated with a significantly slowed information processing. This finding was confirmed by Beatty (23-24) and by Rao (25), yet some authors (8,26) emphasized attentional process impairment . In this regard, we noted an impairment in substained attention, as measured by the mental control subtest from the WMS, which didn’t depend on abnormal emotional status. In fact, none of the subjects studied were depressed. Moreover, even if MS patients were habitually more anxious than controls (STAI-x,), they faced testing in the same manner as controls, therefore, poorer performances were not related to pathological anxiety states during testing (STAI-x,). Furthermore, in contrast to substantially unaffected verbal abilities, MS patients scored lower on all WAIS’s performance subtests which greatly depend on attention, psychomotor and visuospatial abilities and require speed of information processing. The same patients, however, scored as well as controls in other neuropsychological tasks similarly dependent on visuospatial or strength/coordination skills (i.e. Corsi span and visual reproduction subtest
of the WMS), while performed worse in others which do not have psychomotor components (i.e. Rey auditory verbal learning test). Therefore, in our opinion, an impairment in attentional process and in information processing speed could almost partially explain the lower performances in WAIS performance subtests and the poorer learning of patients versus controls. While some investigators (5-6, 8, 10) have noted statistically significant group differences on visuospatial and visual memory tasks, our results suggested that these memory abilities were preserved in contrast to verbal memory. This disagreement between our results and other reports could be due either to discrepancies in MS population characteristics or to different sensitivity of visuospatial tasks employed. In this regard, the Corsi span which we used, failed to detect a visuospatial impairment in other two recent reports of cognitive impairment in mild disabilited MS patients (9,27). Reports are also conflicting regarding the relationship between physical and cognitive disabilities (1 1). In agreement with many authors (3, 12-14), cognitive functioning in our study was neither related to disability scores nor to disease duration. This finding is supported by the observation that mild disturbances in intelligence and memory functions were also present in patients with mild disability (3, 9, 12, 27) as well as in early phases of the disease (7, 2829). CT and more recently MRI have been used to determinate whether a particular anatomic pattern correlates with cognitive impairment in MS. Some studies (9-10, 27) reported good correlation with anatomic involvement (especially with callosal atrophy and periventricular lesion burden), while MRI parameters in other studies were not related to any neuropsychological dysfunction (2,3). These discrepancies were very probably due to methodological differences in neuroradiological and neuropsychological parameters. considered. Despite lack of correlation between the MRI and neuropsychological scores, we detected a significantly inferior memory quotient in patients with MRI evidence of extensive periventricular demyelination. This finding seems to support the hypothesis that memory deficit in MS could result from the disconnection between prefrontal and limbic structures secondary to periventricular demyelinized lesions disrupting the white matter fibers pathway (1). In conclusion, given the frequency of cognitive impairment in mildly disabled MS patients, cognitive functioning must also be systematically investigated in early stages of the disease with a comprehensive neuropsychological battery including intelligence and memory tests, measures of speed of mental processing, substained attention, language 127
Maurelli et al. 3. MARSHGG. Disability and intellectual function in multiple sclerosis patients. J Nerv Ment Dis 1980: 168: 758-762. 4. BEATTYWW, GOODKIN DE, HERTSGAARD D, MONSON N. Clinical and demographic predictors of cognitive performance in multiple sclerosis. Arch Neurol 1990: 47: 305-308. 5. POSERCM, PATYDW, SCHEINBERG L et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Anna1 Neurol 1983: 13: 227-231. 16. KURTZKEJS. Rating neurological impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983: 33: 1444-1452. 17. SCHUMAKER GA, BEEBEG, KIBLERR F et al. Problems of experimental trials of therapy in multiple sclerosis: report by the panel on the evaluation of experimental trials of therapy in multiple sclerosis. Am NY Acad Sci 1965: 122: 522-568. Acknowledgement 18. PAGANO G, VALLARG. Verbal and special immediate meWe acknowledge Dr. M. Grassi, of the Biometric Institute, Unimory span. Normative data from 1355 adults and 1112 chilversity of Pavia, for statistical revision. dren. Ital J Neurol Sci 1987: 8: 539-548. 19. SPIELBERGER CD, GORSUCHRL, LUSHENERE. STAI manual for the state-trait anxiety inventory. Polo Alto, CA: References Consulting Psychology Press, 1970. 20. ORMEROD IEF, MILLERDH, MCDONALDWI et al. The 1. RAO SM. Neuropsychology of multiple sclerosis: a critical role of NMR imaging in the assessment of multiple sclerosis review. J Clin Exper Neuropsychol 1986: 8: 503-542. and isolated neurological lesions: a quantitative study. Brain J, VENDRELL P, MARTINEZJM. 2. LITVANI, GRAFMAN 1987: 110: 1579-1616. Slowed information processing in multiple sclerosis. Arch 21. STAPLESD, LINCOLNNB. Intellectual impairment in mulNeurol 1988: 45: 281-285. tiple sclerosis and its relation to functional abilities. Rheu3. LITVANI, GRAFMAN J, VENDRELL P et al. Multiple mematol Rehab 1979: 18: 153-160. mory deficits in patients with multiple sclerosis. Arch Neurol 22. HEATONRH, NELSONLM, THOMPSON DS, BURKSJS, 1988: 45: 607-610. FRANKLIN GM. Neuropsychological findings in relapsing4. BEATTYWW, GOODKIN DE. Screening for cognitive imremitting and chronic progressive multiple sclerosis. J Conpairment in multiple sclerosis. Arch Neurol 1990: 47: 297sult Clin Psycho1 1985: 53: 103-110. 301. DE, MONSONN, BEATTYPA, 23. BEATTYWW, GOODKIN 5. TRUELLE JL, PALISSONE, LE GALLD, STIP E, DERHERTSGAARD D. Anterograde and retrograde amnesia in OUESNE c . Troubles intellectuels et thymiques dans la sckpatients with chronic progressive multiple sclerosis. Arch rose en plaques. Rev Neurol (Paris) 1987: 143: 595-601. Neurol 1988: 45: 611-619. 6. RAOSM, HAMMEKE TA, MCQUILLEN MP, KHATRIBO, DE, MONSONN, BEATTYPA. 24. BETTYWW, GOODKIN LLOYDD. Memory disturbance in chronic progressive mulCognitive disturbances in patients with relapsing remitting tiple sclerosis. Arch Neurol 1984: 41: 625-631. multiple sclerosis. Arch Neurol 1989: 46: 1113- 1119. 7. LYON-CAEN 0, JOUVENTR, HAUSERS etal. Cognitive 25. RAOSM, ST AUBIN-FAUBERT P, LEOGJ. Information profunction in recent-onset demyelinating diseases. Arch Neurol cessing speed in patients with multiple sclerosis. J Clin Exp 1986: 43: 1138-1141. Neuropsychol 1989: 11: 471-477. 8. RAOSM, LEOGJ, BERNARDIN L, UNVERZAGT F. Cogni26. FEALLYDCM, HEATONRK, NELSONLM, BURKSJS, tive dysfunction in multiple sclerosis. I Frequency, pattern FRANKLINGM. A comparison of dementia in Alzheimer’s and prediction. Neurology 1991: 41: 685-691. disease and in multiple sclerosis. Arch Neurol 1988: 45: 61 19. ANZOLAGP, BEVILACQUA L, CAPPASF et al. Neuropsy619. chological assessment in patients with relapsing-remitting 27. POZZILLI C, PASSAFIUME D, BERNARDI S et al. SPECT, multiple sclerosis and mild functional impairment: correlation MRI and cognitive functions in multiple sclerosis. J Neurol with magnetic resonance imaging. J Neurol Neurosurg PsyNeurosurg Psychiat 1991: 54: 110-115. chiat 1990: 53: 142-145. 28. GRANTI, MCDONALD WI, TRIMBEE MR. Neuropsycho10. HUBER JS, PAULSON GW, SHUTTLEWORTH EC et al. Maglogical impairment in early multiple sclerosis. In: JENSENK netic resonance imaging correlates of dementia in multiple et al., eds. Current problems in neurology, 10. Mental disorsclerosis. Arch Neurol 1987: 44: 732-736. ders and cognitive deficits in multiple sclerosis. London: John E, KNUDSEN L, JENSENK. Multiple sclerosis: 11. STENAGER Libbey, 1989: 17-26. correlation of cognitive dysfunction with Kurtzke disability 29. FRANKLIN GM, NELSONLM, FILLEYCM, HEATONRK. status scale. In: JENSENK et al., eds. Current problems in Cognitive loss in multiple sclerosis. Case reports and review Neurology, 10. Mental disorders and cognitive deficits in mulof the literature. Arch: Neurol. 1989: 46: 162-167. tiple sclerosis. London: John Libbey, 1989: 27-37. 30. RAOSM, LEOGJ, ELLINGTON L, NAUERTZ T, BERNARAH, MINDERHOUD 12. VAN DEN BURGW, VAN ZOMEREN DIN L, UVERZAGT F. Cognitive dysfunction in multiple scleJM, PRANGEAJA, MEIJERNSA. Cognitive impairment in rosis, 2. Impact on employment and social functioning. Neupatients with multiple sclerosis and mild physical disability. rology 1991: 41: 692-696. Arch Neurol 1987: 44: 494-501.
and visuospatial skills. However, only uniform methodology studing a very large sample of patients representative of the heterogeneous MS population could lead to comprehension of neuropathological and anatomic basis of cognitive dysfunctioning in MS. This strategy is also necessary to establish guidelines for informed and effective long-term management of MS patients because cognitive dysfunction negatively influences their employment and social functioning (29-30).
128
