Journal of the Neurological Sciences 299 (2010) 1518

Contents lists available at ScienceDirect

Journal of the Neurological Sciences

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j n s

The Montreal Cognitive Assessment (MoCA) is superior to the Mini-Mental State Examination (MMSE) for the detection of vascular cognitive impairment after acute stroke
YanHong Dong a,d, Vijay Kumar Sharma a, Bernard Poon-Lap Chan a, Narayanaswamy Venketasubramanian a, Hock Luen Teoh a, Raymond Chee Seong Seet a, Sophia Tanicala a, Yiong Huak Chan b, Christopher Chen c,
a

Department of Medicine, National University Health System, Singapore Biostatistics Unit,Yong Loo Lin School of Medicine, National University of Singapore, Singapore c Department of Pharmacology, National University Health System, Singapore d School of Psychiatry, University of New South Wales, NSW, Australia
b

a r t i c l e

i n f o

a b s t r a c t
Background: The majority of patient with post-stroke Vascular Cognitive Impairment (VCI) have Vascular Cognitive Impairment No Dementia (VCIND). The Mini-Mental State Examination (MMSE) has been criticized as a poor screening test for VCIND due to insensitivity to visuospatial and executive function impairments. The Montreal Cognitive Assessment (MoCA) was designed to be more sensitive to such decits and may therefore be a superior screening instrument for VCIND. Methods: Stable patients within 14 days of their index stroke without signicant physical disability, aphasia, dysarthria, active psychiatric illness or pre-existing dementia were eligible. Cognitive and neurological measures were administered after informed consent. Results: 100 patients were recruited. Of the 57 patients with unimpaired MMSE scores, 18 (32%) patients had an impaired MoCA score. By comparison, only 2 out of the 41 (4.9%) patients with unimpaired MoCA scores had impaired MMSE scores. Moreover, MMSE domain subtest scores could not differentiate between groups of differing screening test results, whilst MoCA domain subtest scores (Visuospatial/Executive Function, Attention and Recall) could. Conclusion: The MoCA is more sensitive than the MMSE in screening for cognitive impairment after acute stroke. Longitudinal studies are required to establish the prognostic value of MoCA and MMSE evaluation in the acute post-stroke period for cognitive impairment as dened by the standard method of formal neuropsychological evaluation 36 months after stroke. 2010 Elsevier B.V. All rights reserved.

Article history: Received 26 April 2010 Received in revised form 25 August 2010 Accepted 25 August 2010 Keywords: Vascular cognitive impairment no dementia Stroke Vascular cognitive impairment Dementia

1. Introduction Vascular Cognitive Impairment (VCI), comprising of Vascular Cognitive Impairment No Dementia (VCIND), Vascular Dementia (VaD) and mixed dementia, is a common consequence of ischemic stroke [1]. Most cases of post-stroke VCI are due to VCIND [2] with 46% of VCIND patients developing incident dementia over a 5-year period [3]. As VCIND patients with more severe impairment were found to be at higher risk of conversion to dementia compared to patients with less severe or no cognitive impairment (NCI) [4], early detection of cognitive decits may facilitate intervention to prevent cognitive
Corresponding author. Department of Pharmacology, National University Health System, 5 Lower Kent Ridge Road, Singapore 119074. Tel.: +65 67726051; fax: +65 67794112. E-mail address: phccclh@nus.edu.sg (C. Chen). 0022-510X/$ see front matter 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2010.08.051

deterioration. The feasibility of cognitive screening in the subacute phase of stroke needs to be investigated as screening at the conventional 3 to 6 month period after stroke may be less practical. The widely used Mini-Mental State Examination (MMSE) [5] was found to be inaccurate in screening post-stroke cognitive impairment as it was especially insensitive to complex cognitive decits [6]. By comparison, the Montreal Cognitive Assessment (MoCA) has been designed to be sensitive to mild decits [7], and may detect more cognitive abnormalities after the ischemic stroke or Transient Ischemic Attack (TIA), particularly in executive function, attention and delayed recall [8]. However, a comparison study of both screening tools for patients in the subacute phase of stroke is required. Hence, the primary aim of the present study was to test the hypothesis that MoCA is more sensitive than MMSE for detecting cognitive impairment in a population of subacute stroke patients. The feasibility and the psychometric properties of MoCA (modied for

16

Y. Dong et al. / Journal of the Neurological Sciences 299 (2010) 1518 Table 2 MMSE and MoCA scores and cognitive impairment. MoCA

Singaporean population) were also compared with MMSE for detecting vascular cognitive impairment. 2. Methods 2.1. Participants One hundred stroke patients ( 21 years old) admitted to the stroke neurology service at the National University Health System (NUHS) of Singapore were recruited. Eligible participants had an acute ischemic stroke or TIA within the preceding 14 days and stable clinical status within the preceding 24 h. Exclusion criteria for this study were major physical disability (modied Rankin Scale (mRS) N 4) [9], signicant aphasia or dysarthria that impeded cognitive assessment (National Institute of Health Stroke Score (NIHSS) [10] Best Language (Aphasia) and Dysarthria score N 1). Patients were also excluded if they had a major and active psychiatric illness according to their medical history and pre-existing dementia according to their medical history and a score of N 3.38 on the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) [11,12]. All patients had either CT or MRI scans performed within 48 h of admission and these together with clinical features were used to classify them according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria [13]. This study was approved by the Domain Specic Review Board and Ethics Committee of the National Healthcare Group of Singapore. The written informed consent was obtained from all participants or their legally acceptable representatives. 2.2. Procedure 2.2.1. Demographics and clinical prole Basic demographic information (age, gender, ethnicity and level of education), cardiovascular risk factor data as well as clinical information were collected. 2.2.2. Cognitive status The original MoCA is a 30-point scale with 7 cognitive subtests: visuo-executive, naming, attention, language, abstraction, delayed recall, and orientation. Three Singaporean versions of the MoCA (English, Chinese and Malay) were developed with the following modications: 1) Item 1 (Visuospatial/Executive Functions): Alternating shapes from circle to triangle replaced Roman alphabets in the Trail-making Test Part B as there was no substitute for Roman alphabets in Chinese. The number of steps required for completion of task was retained. 2) Item 2 (Naming): The picture of a rhinoceros was replaced with an elephant to reect local familiarity. 3) Item 3 (Memory): The word velvet was replaced with silk and daisy was replaced with rose to reect local familiarity.

MoCA 21 MMSE MMSE 24 MMSE N 24 Total 41 18 59

MoCA N 21 2 39 41

Total 43 57 100

4) Item 4 (Attention): Arabic numerals were used instead of Roman alphabets for the reasons given above. The numbers and positions of responses were retained. 5) Item 5 (Language-sentence repetition): The name used in the Chinese version was changed to a more common Chinese name to reect local familiarity. 6) Item 5 (Language-verbal uency): Phonemic letter uency was replaced by animal uency as there is no letter-equivalent linguistic unit in Chinese. 2.3. Statistical analysis Descriptive statistics were obtained on the demographical and clinical prole of subjects. Pearson or Spearman correlation coefcients were calculated between MoCA and MMSE scores, and demographics (age, gender, ethnicity and level of education) and clinical prole (qualifying event of ischemic stroke or TIA, IQCODE, NIHSS and mRS). One-way ANOVA and post-hoc multiple comparisons were conducted to compare MoCA and MMSE mean scores and subtests scores among groups with different cognitive screening test results. 3. Results Recruited patients were mostly Chinese (76%) and males (62%) with a mean age of 61.2 11.3 years. 52% had a level of education of primary and below. Most patients (84%) had ischemic stroke presenting with low disability level (median mRS score = 2 and median NIHSS score = 2). The mean interval between stroke event and assessment was 4.2 2.4 days. Most patients were classied as Small Artery Occlusion by the TOAST (50%) with a further 21.4% with Large Artery Atherosclerosis, 17.9% with Cardioembolism, 9.5% with undetermined etiology and 1.2% with Stroke of Other Determined Etiology. The demographic and clinical prole of recruited participants were comparable with the general stroke population (N = 526) admitted to

MoCA vs MMSE
30

*
20

* *

MoCA MMSE

Table 1 Sensitivity, specicity, positive predictive value, and negative predictive value of the MMSE and the Singaporean version of the MoCA for the detection of high-risk Cognitive Impairment (CIND moderate and dementia, n = 62) and low-risk cognitive impairment (CIND mild and NCI, n = 56) at the NUHS outpatient memory clinic. MMSE Cutoff 22/23 23/24 24/25 25/26 26/27 Se 77.4% 83.9% 85.5% 90.3% 93.5% Sp 91.1% 85.7% 82.1% 75% 60.7% PPV 90.6% 86.7% 84.1% 80% 72.5% NPV 78.5% 82.8% 83.6% 87.5% 89.5% MoCA Cutoff 19/20 20/21 21/22 22/23 23/24 Se 77.4% 83.9% 90.3% 91.9% 95.2% Sp 89.3% 83.9% 76.8% 67.9% 60.7% PPV 88.9% 85.2% 81.2% 76.0% 72.8% NPV 78.1% 82.5% 87.8% 88.4% 91.9%

10

0
ild ild M D M od e N C at C M er N D od D er at e I I VC VC IN IN

IN VC

*p<0.05
Fig. 1. Mean (SD) of MoCA and MMSE scores among three groups with different cognitive status (p b 0.05).

Se = Sensitivity; Sp = Specicity; PPV = Positive Predictive Value. NPV = negative predictive value. Optimal cutoff score.

VC

IN

Y. Dong et al. / Journal of the Neurological Sciences 299 (2010) 1518 Table 3 Mean (SD) of MoCA and MMSE subtest scores among three groups with different cognitive screening test results. Cognitive Impairment MoCA subtest scoresa MMSE subtest scoresb

17

VisuoAttention/ Recall/ Orientation/ Abstraction/ Language/ Naming/ Attention/ Recall/ Orientation/ Language/ Registration/ Praxis/ executive/5 6 5 6 2 3 3 Calculation/5 3 10 8 3 1 1.2,# (0.9) 2.7,# (1.5) 0.8,# (1.2) 4.8,# (1.1) 0.2# (0.4) 1.3# (1.1) 2.6# (0.7) 1.7,# (1.5) 0.9,# (0.9) 7.5,# (1.9) 7.1,# (0.8) 2.7,# (0.6) 0.1 (0.3)

a) Acute VCIND Moderate (N = 41) b) Acute VCIND Mild (N = 19) c) NCI (N = 40)

2.4# (1.0)

4.6# (1.1)

2.1# (1.7)

5.6 (0.7)

0.1# (0.3)

1.7 (0.9)

2.9 (0.3)

3.8 (1.2)

2.0 (0.8)

9.4 (0.7)

7.6 (0.5)

3.0 (0.2)

0.3# (0.4)

3.8 (1.0)

5.5 (0.7)

3.8 (1.2)

5.9 (0.2)

1.1 (0.9)

2.4 (0.7)

3.0

4.4 (0.8)

2.2 (1.0)

9.9 (0.4)

7.6 (0.5)

2.9 (0.2)

0.7 (0.4)

a MoCA domain subtest maximum score/details: Visuo-executive/5 Trail B test, Cube copy, Clock drawing; Attention/6 Digit span, Vigilance (tapping at the number 1 in a list of numbers), Serials 7 s; Recall/5 Recall a list of 5 words; Orientation/6 Date, month, year, day, place, city; Abstraction/2 Similarities between 2 items; Language/3 Sentence repetition, verbal uency; Naming/3 Confrontation naming (lion, elephant, camel). b MMSE domain subtest maximum score / details: Attention/Calculation/5 Serial 7 s; Recall/3 Recall a list of 3 words; Orientation/10 Orientation to place and time; Language/8 Sentence repetition, comprehension, reading, writing; Registration/3 Repeat 3 words; Praxis/1 Copy intersecting pentagons. Signicantly different from Acute VCIND Mild. # Signicantly different from NCI.

NUHS over 11 months inclusive of the recruitment period, except that the recruited participants were signicantly younger than the general stroke population who had a mean age of 65.4 13.8 years. Signicant univariate associations were found between MoCA and MMSE scores and age, ethnicity, education, IQCODE, qualifying event (stroke or TIA), NIHSS and mRS scores. Multivariate stepwise regression analysis showed that age, ethnicity, education, IQCODE and NIHSS scores independently predicted for both MoCA and MMSE scores. MoCA scores were normally distributed (median=20) while MMSE scores were skewed towards the higher scores (median 26). MoCA scores also had a wider dispersion (IQR=9, 1524), than the MMSE (IQR=5, 2227). A cutoff score of MMSE 24 and MOCA 21 for the Singaporean elderly was established for moderate-severe (n = 62) and mild-no cognitive impairment (n = 56) patients from the NUHS memory clinic. The moderate-severe cognitive impairment group included patients with dementia and moderate CIND (more than 2 domains impaired). The mild-no cognitive impairment group consisted of NCI and mild CIND (12 domains impaired) [4]. A consensus diagnosis of dementia was established according to DSM-IV criteria at meetings where the formal neuropsychological assessment, neuro-imaging and clinical assessment were reviewed. Receiver Operator Curve (ROC) analysis was performed to determine the optimal cutoffs for MMSE and MoCA (Table 1). In view of the low educational level in the Singaporean elderly population, one point was added to the total score for patients with primary education and below to adjust for an education effect as per the original MoCA design [7] and in a similar Asian study [14]. A comparison between the MMSE and MoCA in detecting cognitive impairment is summarized in Table 2. Of the 57 patients with unimpaired MMSE scores, 18 (32%) patients were found to have cognitive impairment by the MoCA. By comparison, only 2 out of the 41 (4.9%) patients with unimpaired MoCA scores had impaired MMSE. Patients could be classied into three cognitive screening test result groups: a) acute VCIND moderate (screen positive for both MMSE and MoCA), b) acute VCIND mild (screen positive for either MMSE or MoCA) and c) NCI (screen negative for both MMSE and MoCA). One-way ANOVA and post-hoc multiple comparisons showed signicant mean MMSE and MoCA score differences between these three groups (Fig. 1). However, these differences were of different magnitudes with only a 1-point MMSE scores difference between those screened positive for either MMSE or MoCA and those screened negative for both MMSE and MoCA as compared to a 6-point MoCA score difference between the same groups (Fig. 1). One-way ANOVA analysis showed signicant MMSE and MoCA domain subtest mean score differences between the three groups of differing cognitive

screening test results (Table 3). Post-hoc multiple comparisons showed that none of the MMSE subtest scores could differentiate between all these three groups. By contrast, MoCA subtest (Visuoexecutive Function, Attention and Recall) scores could differentiate all three groups. 4. Discussion The principal nding of this study is that the MMSE is less sensitive than the MoCA in detecting VCI after acute stroke. More patients were identied with cognitive impairment by the MoCA (n = 59) compared to MMSE (n = 43). Both the mean MoCA score and several MoCA subtest scores could signicantly differentiate between all three cognitive screening test result groups and had good discriminating properties, while only the mean MMSE scores but none of MMSE subtest scores could do so. Finally, 18 (29.5%) of the 61 patients with either impaired MoCA or MMSE scores could only be detected by impaired MoCA scores. By contrast, only 2 (3.3%) patient presented with only impaired MMSE scores. The poorer performance of the MMSE at detecting VCI may be due to several factors. As shown by an earlier study in acute stroke [6], the MMSE is less capable of testing for complex cognitive impairments in domains such as visuospatial, executive function and abstract reasoning. In addition, the MMSE subtests of Attention and Delayed Recall contain test items which are not as challenging as contained in the MoCA. For example, the only MMSE test for attention is the serial 7 s test while the MoCA includes 2 additional tests (Digit Span and Vigilance). Similarly, the 3-item Delayed Recall in the MMSE is less difcult than 5-item Delayed Recall in the MoCA. In view of the restrictions of the MMSE, a brief executive function assessment (the trail-making test or digit symbol test) has been recommended to supplement the MMSE to improve its sensitivity of bedside cognitive assessment [15]. The visuo-executive function (Trail B, Cube and Clock) tests in the MoCA distinguished between three groups of differing cognitive screening test results and hence may be of use in the cognitive screening of subacute stroke patients. As shown in recent studies, cognitive impairment in visuo-executive function has been found to predict poor survival after stroke [16], whilst the severity of VCI is associated with incident dementia [4], hence, early detection of VCI by MoCA screening may allow clinicians to intervene and improve prognosis. Our ndings that the MoCA is superior to the MMSE particularly for the detection of decits in recall, visuo-executive function and attention are in keeping with recent ndings from a community based sample of stroke patients [8], where the MoCA was shown to pick up more decits in executive function, attention and delayed recall.

18

Y. Dong et al. / Journal of the Neurological Sciences 299 (2010) 1518 [6] Nys GM, van Zandvoort MJ, de Kort PL, Jansen BP, Kappelle LJ, de Haan EH. Restrictions of the Mini-Mental State Examination in acute stroke. Arch Clin Neuropsychol 2005;20:6239. [7] Nasreddine ZS, Phillips NA, Bdirian V, Charbonneau S, Whitehead V, Collin I, Cummings JL, Chertkow H. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 2005;53:6959. [8] Pendlebury ST, Cuthbertson FC, Welch SJV, Mehta Z., Rothwell PM. Underestimation of cognitive impairment by Mini-Mental State Examination versus the Montreal Cognitive Assessment in patients with transient ischemic attack and stroke: A population-based study. Published online doi:10.1161/STROKEAHA.110.579888. [9] Rankin J. Cerebral vascular accidents in patients over the age of 60 II. Prognosis. Scott Med J 1957;2:20015. [10] Brott T, Adams Jr HP, Olinger CP, Marler JR, Barsan WG, Biller BJ, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 1989;20:86470. [11] Jorm AF, Jacomb PA. The Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): socio-demographic correlates, reliability, validity and some norms. Psychol Med 1989;19:101522. [12] Narasimhalu K, Lee J, Auchus A, Chen CPLH. Improving detection of dementia in asian patients with low education: combining the Mini-Mental State Examination and the Informant Questionnaire on Cognitive Decline in the Elderly. Dement Geriatr Cogn Disord 2008;25:1722. [13] Adams HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classication of subtype of acute ischemic stroke. Denitions for use in a multicenter clinical trial. Acute Stroke Treatment Stroke 1993;24:3541 TOAST. Trial of Org 10172 in. [14] Wong A, Xiong YY, Kwan PW, Chan AY, Lam WW, Wang K, et al. The validity, reliability and clinical utility of the Hong Kong Montreal Cognitive Assessment (HK-MoCA) in patients with cerebral small vessel disease. Dement Geriatr Cogn Disord 2009;28:817. [15] O'Sullivan M, Morris RG, Markus HS. Brief cognitive assessment for patients with cerebral small vessel disease. J Neurol Neurosurg Psychiatry 2005;76:11405. [16] Oksala NK, Jokinen H, Melkas S, Oksala A, Pohjasvaara T, Hietanen M, et al. Cognitive impairment predicts poststroke death in long-term follow-up. J Neurol Neurosurg Psychiatry 2009;80:12305.

In conclusion, the MoCA is feasible and superior to the MMSE in screening for cognitive impairment in subacute stroke/TIA patients, as it detects complex cognitive impairments such as executive function and visual perception/construction. However, longitudinal studies are required to establish the prognostic value of MoCA and MMSE in the acute post-stroke period for cognitive impairment as dened by the formal neuropsychological evaluation 36 months after stroke. Acknowledgements The study is supported by a Centre Grant from the National Medical Research Council (NMRC/CG/NUHS/2010). References
[1] Rockwood K, Wentzel C, Hachinski V, Hogan DB, MacKnight C, McDowell I. Prevalence and outcomes of vascular cognitive impairment. Vascular Cognitive Impairment Investigators of the Canadian Study of Health and Aging. Neurology 2000;54:44751. [2] Tham W, Auchus AP, Thong M, Goh ML, Chang HM, Wong MC, Chen CP. Progression of cognitive impairment after stroke: one year results from a longitudinal study of Singaporean stroke patients. J Neurol Sci 2002;203204: 4952. [3] Wentzel C, Rockwood K, MacKnight C, Hachinski V, Hogan DB, Feldman H, stbye T, Wolfson C, Gauthier S, Verreault R, McDowell I. Progression of impairment in patients with vascular cognitive impairment without dementia. Neurology 2001;57:7146. [4] Narasimhalu K, Ang S, De Silva DA, Wong MC, Chang HM, Chia KS, Auchus AP, Chen C. Severity of CIND and MCI predict incidence of dementia in an ischemic stroke cohort. Neurology 2009;73:186672. [5] Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12: 18998.