Neurol Sci DOI 10.1007/s10072-014-1921-3
ORIGINAL ARTICLE
Montreal Cognitive Assessment (MoCA)-Italian version: regression based norms and equivalent scores Silvia Conti • Stefano Bonazzi • Marcella Laiacona Marco Masina • Mirco Vanelli Coralli
•
Received: 27 December 2013 / Accepted: 4 August 2014 Ó Springer-Verlag Italia 2014
Abstract The Montreal Cognitive Assessment (MoCA) is a brief cognitive screening instrument developed by Nasreddine et al. to detect mild cognitive impairment, a high-risk condition for Alzheimer’s disease and other forms of dementia. In this study we report normative data on the MoCA-Italian version, collected on a sample of 225 Italian healthy subjects ranged in age between 60 and 80 years, and in formal education from 5 to 23 years. The global normal cognition was established in accordance with the Mini–Mental State Examination score and with the Prose Memory Test score (Spinnler and Tognoni, Ital J Neurol Sci 6:25–27, 1987). None of the participants had a history of psychiatric, neurological, cerebrovascular disorders or brain injury or took drugs affecting cognition.
S. Conti (&) Scuola di Specializzazione in Neuropsicologia, Universita` degli Studi di Padova, Padua, Italy e-mail: [email protected]; [email protected] S. Bonazzi Dipartimento Cure Primarie, UOSD-ArOA Riabilitativa Bentivoglio, via Marconi 35, Bentivoglio, AUSL Bologna, Italy M. Laiacona Divisione di Neurologia, Fondazione S. Maugeri, IRCCS, Istituto Scientifico di Veruno, via per Revislate 13, Veruno, Novara, Italy M. Masina UOC Geriatria, Ospedale di Bentivoglio, via Marconi 35, Bentivoglio, AUSL Bologna, Italy M. V. Coralli Centro Disturbi Cognitivi, UOC Geriatria, Ospedale di Bentivoglio, via Marconi 35, Bentivoglio, AUSL Bologna, Italy
Linear regression analysis was performed to evaluate the potential effect of age, education and sex on the MoCA total performance score. We provide correction grids to adjust raw scores and equivalent scores with cut-off value to allow comparison between MoCA performance and others neuropsychological test scores that can be administered on the same subject. Keywords Montreal Cognitive Assessment (MoCA) Mild cognitive impairment (MCI) Normative data Screening test
Introduction Montreal Cognitive Assessment (MoCA) is a one-page-30point test that takes 15 min to administer, specifically developed to differentiate mild cognitive impairment (MCI) from cognitive changes of normal ageing [1]. MCI is a clinical condition characterized by cognitive deficits that are not so severe to meet the criteria of dementia. Its importance is due to the fact that often, even if not always, MCI represents a transitional state between normal ageing and Alzheimer’s Disease (AD) or other types of dementia [2, 3]. In 2006, Pirani et al. [4] translated the test into Italian and its instructions for administration and scoring. The use of this version of the test in clinical practice is recommended by the latest review of the ‘‘Stroke prevention and educational awareness diffusion-cerebral stroke: Italian guideline’’ [5], despite the fact that to date there is no published Italian normative data. The aim of the present study is to collect normative data on the MoCA-Italian version in a sample of Italian elderly subjects.
123
Neurol Sci
Methods Subjects
Table 1 Demographic data of the sample Age/sex
Education years 5
Two-hundred and ninety five subjects, aged between 60 and 80 years, and with at least 5 years of formal education, were originally sampled for this study. Participants were recruited among patients’ consorts admitted to a memory clinic (Cognitive Disorders Centre, Geriatric Unit of Bentivoglio, Bologna) and in a senior centre. All subjects came from the city of Bologna and its province and were recruited on a voluntary basis from January 2013 to June 2014. Only those who were classified as healthy in accordance with the following inclusion criteria, were admitted to the normative sample: (a) absence of traumatic brain injury with loss of consciousness for at least 1 h in the history (b) no history of stroke or transient ischemic attacks (c) no history of neurological disorders (Parkinson’s disease, multiple sclerosis, epilepsy, obstructive sleep apnoea syndrome, normal pressure hydrocephalus, brain surgery for tumour or meningioma) (d) absence of psychiatric disorders in the history (mood disorders, obsessive–compulsive disorder, schizophrenia) (e) no use of psychotropic drugs (neuroleptics, benzodiazepines, antidepressants, anticonvulsive medications) within 7 days before testing (f) absence of dementia (adjusted score on the Mini-Mental State Examination (MMSE) [23.8 [6, 7]) (g) normal memory performance in accordance with the Prose memory Test (PMT) [8] (adjusted score [6.25, non-parametric inner tolerance limit with confidence level of 95 % [9]) (h) no major sensory deficits. The inclusion criteria (a)–(e) were assessed by asking the subjects to complete a questionnaire which was subsequently checked with the examiner during a brief interview. To avoid selecting a sample of ‘‘supernormal’’ subjects we did not exclude individuals with pharmacological wellcompensated hypertension and diabetes. Seventy subjects (23.7 % of the whole sample) were discharged because they did not fulfil the inclusion criteria. The normative sample (NS) was composed of 225 subjects (111 men, 114 women), the mean age was 70.1 years (SD ± 5.7, range 60–80) and the mean education was 9.9 years of formal schooling (SD ± 4.6, range 5–23). Subjects’ stratification according to age, years of formal education and sex is reported in Table 1. After providing the subjects with a definition of anterograde episodic memory, the self-reported memory problems were investigated paying particular attention to induce the correct comparison by asking the participants
123
6–8
9–13
C14
Overall (225) by age
60–64 years Males
4
6
7
5
22
Females
4
5
6
5
20
65–70 years Males
5
13
10
6
34
Females
10
14
7
6
37
71–75 years Males
7
8
10
5
30
Females
17
5
5
6
33
Males
9
6
4
6
25
Females
9
3
7
5
24
76–80 years
Overall (225) by education Males
25
33
31
22
111
Females
40
27
25
22
114
‘‘do you have memory problems relative to most people your age?’’ [10]. None of the normative sample subjects replied positively. Nonetheless it was not an inclusion criterion. The reason why we have not inserted the selfreported memory and cognitive complaints among the selection criteria is that we prefer to give more importance to objective measures (memory prose score and MMSE score) rather than subjective ones. This is in line with recent research that has questioned the use of subjective memory complaint as an MCI diagnostic criterion, since some studies have reported that the relationship between subjective memory complaint and objective memory impairment is not clear [11–13]. The study was approved by the local ethics committee and was performed in accordance with the Helsinki Declaration. The informed written consent was obtained from all participants. Materials and procedure Each participant was assessed in a single session structured as follows: first part––administration of MMSE and PMT [8] (the 10-min interval between the immediate recall and the delayed recall was filled with the execution of the Raven Progressive Coloured Matrices 47 [14]); second part––15 min interval to minimize the effects of fatigue; third part—administration of MoCA-Italian version [1, 4].
Neurol Sci
Although there is evidence of a higher sensitivity of the word list learning test in detecting memory deficit due to MCI [15–17] and Alzheimer disease [15] than story recall test, we chose to use the latter for decreasing the potential effect of the proactive interference on the memory recall task of MoCA. This is justified because some studies showed that story recall test is sensitive in distinguishing normal from MCI and AD [16, 18]. Moreover, the story recall test is more ecological than the word list learning test, because it is similar to an everyday meaningful discourse that can be found in a newspaper, television and conversation.
Table 2 The effects of age, education and sex within the linear regression model of MOCA test (for further details see ‘‘Statistical methods’’) Simple regression
Simultaneous regression
Age (linear): F(1,223) = 37.931, p \ 0.0001
Age (linear, education partialled out): F(1, 222) = 30.381, p \ 0.0001
Education (1/education years): F(1,223) = 69.208, p \ 0.0001
Education (1/education years, age partialled out): F(1, 222) = 60.633, p \ 0.001
Sex: F(1,223) = 1.261, ns Adjusted score: Raw score ? 0.175 (age - 70.080) ? 24.300 [(1/ education years) - 0.126]
Statistical methods Control scores were analysed by means of simultaneous multiple regression, to check the influence of age, education (years of schooling) and sex. For the MoCA test we searched the best fitting linear regression model that could be used to adjust original scores according to the demographic variables. To this aim, we estimated a linear model able to calculate the score expected for a given subject on the basis of his/her age, education and sex. The simultaneous regression included the variables that resulted significant when considered one at the time: at this stage the effect of each predictor variable was studied partialling out the effect held in common with the other terms of the model. Taking this model as a basis, we calculated from the raw score an adjusted score, by adding or subtracting the contribution given by each significant concomitant variable in the final correction model. Following this approach, scores can be directly confronted across subjects of different demographic data. Adjusted scores were then ranked, and by means of a non-parametric procedure [19] we set tolerance limits. In this way we found both the outer and inner one-sided tolerance limits [20]. Above the outer tolerance limit one expects to find at least 95 % of the normal population (with 95 % confidence): hence, when a score is below the outer tolerance limit, the subject can be declared ‘‘not normal’’ with 95 % confidence. Above the inner tolerance limit one expects to find at most 95 % of the population (with 95 % confidence): hence, when a score is above the inner tolerance limit, the subject can be declared ‘normal’ with 95 % confidence. When a score falls between the outer and inner tolerance limits, no inferentially controlled judgement is possible. Given our sample size, outer and inner tolerance limits were fixed, respectively, on the basis of the values corresponding to the 6th and 18th ranked scores after demographic adjustments. To avoid errors due to the fixed upper limit of the test scores, no adjustment was made to scores of 30/30. The adjusted scores were classified into five categories
(equivalent scores) endowed at least with an ordinal relationship: 0 = scores lower than the outer 5 % tolerance limits; 4 = scores higher than the median value of the sample; 1, 2 and 3 = intermediate scores between the central value and the pathology threshold on a quasiinterval scale. To illustrate the raw score adjustment, consider the case of a 75-year-old male with 5 years of schooling. The original raw score achieved on MoCA was 19/30; the adjusted score becomes 19 ? 2.659 = 21.659. The corresponding equivalent score is 3.
Results The results of the regression analyses are shown in Table 2. The influence of age and education was always significant: the effect of age was most significant without any mathematical transformations of the original scale, whereas the effect of education was most significant considering the reciprocal of the original value of this variable, i.e. 1/years of schooling. This effect was confirmed on simultaneous regression. No difference between females and males was observed. Table 3 reports the correction grid for the most frequent combinations of age and education; intermediate values can be obtained by interpolation or using the original linear models reported in Table 2. Inner and outer tolerance limits are also reported: subjects with scores below the outer limit should be considered pathological, while those with scores above the inner limit can be considered normal. Subjects with scores included between the outer and the inner limits are better viewed as ‘‘borderline’’ cases. The values delimiting the Equivalent Scores are reported in Table 4. For a descriptive presentation of MoCA raw scores according to age and educational level, see Table 5.
123
Neurol Sci Table 3 Adjustments to be added to, or subtracted from, the raw scores according to age and education (expressed as years of schooling) Education
Age 55
60
65
70
75
80
85
5
-0.841
0.034
0.909
1.784
2.659
3.534
4.409
8
-2.663
-1.788
-0.913
-0.038
0.837
1.712
2.587
13
-3.832
-2.957
-2.082
-1.206
-0.332
0.543
1.418
17
-4.271
-3.396
-2.521
-1.646
-0.771
0.104
0.979
19
-4.422
-3.547
-2.672
-1.796
-0.922
-0.047
0.828
Outer and inner non-parametric one-sided tolerance limits are also reported (see ‘‘Statistical methods’’) Outer limit 17.362, inner limit 19.262
Table 4 Equivalent scores (ES) classification of adjusted scores Equivalent scores
Score interval
ES = 0
0 ? 17.362
ES = 1
17.363 ? 19.500
Density
Cumulative frequency
6
6
16
22
ES = 2
19.501 ? 21.562
36
58
ES = 3
21.563 ? 23.361
54
112
ES = 4
[23.361
113
225
ES = 0 corresponds to an inferentially controlled judgement of being below the norm; 4 is equal or better than the 50th percentile; 1, 2 and 3 are intermediate between 0 and 4 on a quasi-interval scale
Table 6 shows the frequency distributions for each subtest obtained by the normative sample. The correlation between the MoCA scores and the MMSE scores was 0.486 (df = 223, p \ 0.001. The correlation between these variables after adjustment for age and education was 0.321 (df = 219, p \ 0.001).
Discussion The collection of normative data represents the first step to introduce a new test in the clinical practice. MoCA is now added to the list of screening tests that can be used on Italian people to detect subjects with cognitive deficit taking into account the effect of the demographic variables.
Table 5 Descriptive statistics (mean ± standard deviations) of raw scores obtained on MoCA stratified by age and education
123
Age (years)
The mean of MoCA raw scores of 23.28 (SD ± 3.22) observed in the present study is lower than the normal controls in the original work (NC): 26.90 (SD is not reported by the authors) [21]. This may be attributable to the fact that though the mean ages of the two samples are comparable (NC 72.84 ± 7.03 years, NS 70.1 ± 5.7 years), our sample is less educated (mean 9.9 ± 4.6 years) than the NC (mean 13.33 ± 3.40 years). This discrepancy does not change even if the raw scores are adjusted in accordance with the instructions provided by the development group of MoCA, and reported in the Italian version of the test, too (?1 point if the subject’s educational level is B12 years). In doing so our normative sample obtains a mean of 23.96 ± 3.10, whereas the mean of the NC becomes 27.37 ± 2.20. The difference might derive from an incorrect estimation of the effect of education on the MoCA scores in the Canadian study, however, this is just a conjecture. Furthermore, if the cut-off proposed by the authors and reported in the Italian version is used (26), 74 % of our healthy sample (166 out of 225 subjects) would be classified as cognitively compromised. Another possible cause of the discrepancy is the difference in the selection criteria used to enrol subjects in the two studies. Narazaki et al. [22] have hypothesized that the strict criteria of admission used by Nasreddine et al. (no memory or cognitive complaints, normal baseline neuropsychological performance, no abnormality in neurological examination and computed tomography
Education years 5
Overall 6–8
9–13
C14
60–64
23.50 ± 2.27
25.09 ± 2.12
25.08 ± 2.60
26.70 ± 1.95
25.17 ± 2.43
65–70
19.47 ± 3.54
23.59 ± 2.34
26.00 ± 1.54
25.42 ± 1.88
23.61 ± 3.36
71–75
21.04 ± 3.07
23.31 ± 2.98
24.47 ± 2.53
24.73 ± 1.95
22.97 ± 3.14
76–80
20.67 ± 3.09
21.67 ± 2.74
22.09 ± 2.88
22.55 ± 2.07
21.59 ± 2.80
Overall
20.88 ± 3.25
23.52 ± 2.65
24.61 ± 2.69
24.82 ± 2.41
23.28 ± 3.22
Neurol Sci Table 6 MoCA: cognitive domains, subtests and frequency distributions for each subtest obtained by the normative sample (N = 225) Cognitive domains
Subtests
Frequency distributions for each subtest obtained by the normative sample Points scored
Memory recall
Visuospatial construction
Recall 5 words
Clock drawing
Copying a cube Executive functions
Attention, concentration and working memory
Language
0
85 (37.8 %)
1 2
25 (11.1 %) 43 (19.1 %)
3
36 (16.0 %)
4
22 (9.8 %)
5
14 (6.2 %)
0
1 (0.4 %)
1
9 (4.0 %)
2
56 (24.9 %)
3
159 (70.7 %)
0
57 (25.3 %)
1
168 (74.7 %)
Trail making test B modified
0
79 (35.1 %)
1
146 (64.9 %)
Phonemic fluency
0
90 (40.0 %)
1
135 (60.0 %)
Verbal abstraction
0 1
43 (19.1 %) 98 (43.6 %)
2
84 (37.3 %)
Target tapping Serial subtraction
0
22 (9.8 %)
1
203 (90.2 %)
0
2 (0.9 %)
1
2 (0.9 %)
2
25 (11.1 %)
3
196 (87.1 %)
Digit span (forward and backward)
0
12 (5.3 %)
1
54 (24.0 %)
2
159 (70.7 %)
Confrontation naming
0
0 (0 %)
1
2 (0.9 %)
Repeating two sentences Orientation
Frequency count for each point
To time and place
2
30 (13.3 %)
3 0
193 (85.8 %) 6 (2.7 %)
1
63 (28.0 %)
2
156 (69.3 %)
0
0 (0 %)
1
0 (0 %)
2
0 (0 %)
3
0 (0 %)
4
0 (0 %)
5
39 (17.3 %)
6
186 (82.7 %)
scan) may have led to select a ‘‘supernormal’’ sample, that may not represent the target population. For the aforementioned reasons, the Nasreddine et al.’s correction and cut-off values are not adequate for the Italian population. Our results (mean MoCA raw scores ± SD and percentage of individuals who scored below 26) are in line with those reported by studies in which similar selection criteria to ours were used (exclusion of subjects with disorders that may impact on cognition; normal performance on cognitive screening tests; levels of age and schooling comparable to ours): [22, 23]. In the Japanese study the means of the MoCA raw scores obtained by the age categories of 65–75 and 70–80 years were 22.9 ± 3.4 and 21.6 ± 3.7, respectively; and 75.1 % of the normative sample fell below the 26, despite the ?1 point correction. Also the Portuguese study reported in the age interval of C65 year (mean age 71.96 ± 5.43) a lower mean MoCA raw score than the Canadian one: 22.71 ± 3.60. Moreover, our sample shows frequency distributions for each MoCA subtest comparable to those obtained by the Japanese one, except for the verbal abstraction and sentence repeating. Our computed equivalent scores allow comparison between MoCA performance and other neuropsychological measures (e.g. Mental Deterioration Battery [14], Free and Cued Selective Reminding Test [24], Frontal Assessment Battery [25]). Further studies with an MCI patients group will be required for determining the sensibility, specificity, positive and negative predicting values of the Italian version of MoCA. Acknowledgments We are grateful to Erminio Capitani for his comments on a preliminary draft of this study. Daniella Mancini revised the English text. Thanks to Maria Leggieri for logistic help. Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Nasreddine ZS, Phillips NA, Be´dirian V, Charbonneau S, Whitehead V, Collin I, Cummings JL, Chertkow H (2005) The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 53:695–699 2. Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, Rabins PV, Ritchie K, Rossor M, Thal L, Winblad B (2001) Current concepts in mild cognitive impairment. Arch Neurol 58:1985–1992 3. Storandt M, Grant EA, Miller JP, Morris JC (2006) Longitudinal course and neuropathologic outcomes in original vs. revised MCI and in pre-MCI. Neurology 67:467–473 4. Pirani A, Tulipani C, Neri M (2006) italian translation of MoCA test and of its instructions. http://www.mocatest.org
123
Neurol Sci 5. Micieli G, Bottini G, Cerri C et al (2012) Complicanze psicocognitive dell’ictus. In: SPREAD stroke prevention and educational awareness diffusion. Ictus cerebrale: linee guida italiane di prevenzione e trattamento, 7th edn. Pierrel Research Italy Spa, Milano, p 110 6. Folstein MF, Folstein SE, McHugh PR (1975) ‘‘Mini-mental state’’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189–198 7. Measso G, Cavarzeran F, Zappala` G, Lebowittz BD, Crook TH, Pirazzolo FJ, Amaducci LA, Massari D, Grigoletto F (1993) The Mini-Mental State Examination: normative study of an Italian random sample. Dev Neuropsychol 9:77–85 8. Spinnler H, Tognoni G (1987) Standardizzazione e Taratura Italiana di Test Neuropsicolgici. Ital J Neurol Sci 6:25–27 9. Capitani E (1997) Normative data and neuropsychological assessment. Common problems in clinical practice and research. Neuropsychol Rehabil 7:295–309 10. Salthouse TA (1991) Theoretical perspectives on cognitive aging. Lawrence Erlbaum Associates Inc, Hillsdale, p 438 11. Lenehan ME, Klekociuk SZ, Summers MJ (2012) Absence of a relationship between subjective memory complaint and objective memory impairment in mild cognitive impairment (MCI): is it time to abandon subjective memory complaint as an MCI diagnostic criterion? Int Psychogeriatr 24:1505–1514 12. Purser JL, Fillenbaum GG, Wallace RB (2006) Memory complaint is not necessary for diagnosis of mild cognitive impairment and does not predict 10-year trajectories of functional disability, word recall, or short portable mental status questionnaire limitations. J Am Geriatr Soc 54:335–338 13. Roberts JL, Clare L, Woods RT (2009) Subjective memory complaints and awareness of memory functioning in mild cognitive impairment: a systematic review. Dement Geriatr Cogn Disord 28:95–109 14. Carlesimo GA, Caltagirone C, Gainotti G (1996) The Mental Deterioration Battery: normative data diagnostic reliability and qualitative analyses of cognitive impairment. The group of the standardization of the Mental Deterioration Battery. Eur Neurol 36:378–384 15. Baek MJ, Kim HJ, Kim S (2012) Comparison between the story recall test and the word-list learning test in Korean patients with
123
16.
17.
18.
19. 20.
21.
22.
23.
24.
25.
mild cognitive impairment and early stage of Alzheimer’s disease. J Clin Exp Neuropsychol 34:396–404 Rabin LA, Pare´ N, Saykin AJ, Brown MJ, Wishart HA, Flashman LA, Santulli RB (2009) Differential memory test sensitivity for diagnosing amnestic mild cognitive impairment and predicting conversion to Alzheimer’s disease. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn 16:357–376 Tremont G, Miele A, Smith MM, Westervelt HJ (2010) Comparison of verbal memory impairment rates in mild cognitive impairment. J Clin Exp Neuropsychol 32:630–636 Christensen H, Hardzi-Parlovic D, Jacomb P (1997) The psychometric differentiation of dementia from normal aging: a metaanalysis. Psychol Assess 3:147–155 Wilks KK (1941) Determination of sample size for setting tolerance limits. Ann Math Stat 12:91–96 Capitani E, Laiacona M (2000) Classification and modelling in neuropsychology: from groups to single cases. In: Boller F, Grafman J, Rizzolatti G (eds) Handbook of neuropsychology, 2nd edn, vol 1, pp 53–76 Nasreddine ZS, Phillips N, Chertkow H (2012) Comment on: Normative data for the Montreal Cognitive Assessment (MoCA) in a population-based sample. Author replay. Neurology 78:765–766 Narazaki K, Nofuji Y, Honda T, Matsuo E, Yonemoto K, Kumagai S (2013) Normative data for the Montreal cognitive assessment in a Japanese community-dwelling older population. Neuroepidemiology 40:23–29 Freitas S, Simo˜es MR, Alves L, Santana I (2011) Montreal Cognitive Assessment (MoCA): normative study for the Portuguese population. J Clin Exp Neuropsychol 33:989–996 Frasson P, Ghiretti R, Catricala` E, Pomati S, Marcone A, Parisi L, Rossini PM, Cappa SF, Mariani C, Vanacore N, Clerici F (2011) Free and Cued Selective Reminding Test: an Italian normative study. Neurol Sci 32:1057–1062 Apollonio I, Leone M, Isella V, Piamarta F, Consoli T, Villa ML, Forapani E, Russo A, Nichelli P (2005) The Frontal Assessment Battery (FAB): normative values in an Italian population sample. Neurol Sci 26:108–116
ORIGINAL ARTICLE
Montreal Cognitive Assessment (MoCA)-Italian version: regression based norms and equivalent scores Silvia Conti • Stefano Bonazzi • Marcella Laiacona Marco Masina • Mirco Vanelli Coralli
•
Received: 27 December 2013 / Accepted: 4 August 2014 Ó Springer-Verlag Italia 2014
Abstract The Montreal Cognitive Assessment (MoCA) is a brief cognitive screening instrument developed by Nasreddine et al. to detect mild cognitive impairment, a high-risk condition for Alzheimer’s disease and other forms of dementia. In this study we report normative data on the MoCA-Italian version, collected on a sample of 225 Italian healthy subjects ranged in age between 60 and 80 years, and in formal education from 5 to 23 years. The global normal cognition was established in accordance with the Mini–Mental State Examination score and with the Prose Memory Test score (Spinnler and Tognoni, Ital J Neurol Sci 6:25–27, 1987). None of the participants had a history of psychiatric, neurological, cerebrovascular disorders or brain injury or took drugs affecting cognition.
S. Conti (&) Scuola di Specializzazione in Neuropsicologia, Universita` degli Studi di Padova, Padua, Italy e-mail: [email protected]; [email protected] S. Bonazzi Dipartimento Cure Primarie, UOSD-ArOA Riabilitativa Bentivoglio, via Marconi 35, Bentivoglio, AUSL Bologna, Italy M. Laiacona Divisione di Neurologia, Fondazione S. Maugeri, IRCCS, Istituto Scientifico di Veruno, via per Revislate 13, Veruno, Novara, Italy M. Masina UOC Geriatria, Ospedale di Bentivoglio, via Marconi 35, Bentivoglio, AUSL Bologna, Italy M. V. Coralli Centro Disturbi Cognitivi, UOC Geriatria, Ospedale di Bentivoglio, via Marconi 35, Bentivoglio, AUSL Bologna, Italy
Linear regression analysis was performed to evaluate the potential effect of age, education and sex on the MoCA total performance score. We provide correction grids to adjust raw scores and equivalent scores with cut-off value to allow comparison between MoCA performance and others neuropsychological test scores that can be administered on the same subject. Keywords Montreal Cognitive Assessment (MoCA) Mild cognitive impairment (MCI) Normative data Screening test
Introduction Montreal Cognitive Assessment (MoCA) is a one-page-30point test that takes 15 min to administer, specifically developed to differentiate mild cognitive impairment (MCI) from cognitive changes of normal ageing [1]. MCI is a clinical condition characterized by cognitive deficits that are not so severe to meet the criteria of dementia. Its importance is due to the fact that often, even if not always, MCI represents a transitional state between normal ageing and Alzheimer’s Disease (AD) or other types of dementia [2, 3]. In 2006, Pirani et al. [4] translated the test into Italian and its instructions for administration and scoring. The use of this version of the test in clinical practice is recommended by the latest review of the ‘‘Stroke prevention and educational awareness diffusion-cerebral stroke: Italian guideline’’ [5], despite the fact that to date there is no published Italian normative data. The aim of the present study is to collect normative data on the MoCA-Italian version in a sample of Italian elderly subjects.
123
Neurol Sci
Methods Subjects
Table 1 Demographic data of the sample Age/sex
Education years 5
Two-hundred and ninety five subjects, aged between 60 and 80 years, and with at least 5 years of formal education, were originally sampled for this study. Participants were recruited among patients’ consorts admitted to a memory clinic (Cognitive Disorders Centre, Geriatric Unit of Bentivoglio, Bologna) and in a senior centre. All subjects came from the city of Bologna and its province and were recruited on a voluntary basis from January 2013 to June 2014. Only those who were classified as healthy in accordance with the following inclusion criteria, were admitted to the normative sample: (a) absence of traumatic brain injury with loss of consciousness for at least 1 h in the history (b) no history of stroke or transient ischemic attacks (c) no history of neurological disorders (Parkinson’s disease, multiple sclerosis, epilepsy, obstructive sleep apnoea syndrome, normal pressure hydrocephalus, brain surgery for tumour or meningioma) (d) absence of psychiatric disorders in the history (mood disorders, obsessive–compulsive disorder, schizophrenia) (e) no use of psychotropic drugs (neuroleptics, benzodiazepines, antidepressants, anticonvulsive medications) within 7 days before testing (f) absence of dementia (adjusted score on the Mini-Mental State Examination (MMSE) [23.8 [6, 7]) (g) normal memory performance in accordance with the Prose memory Test (PMT) [8] (adjusted score [6.25, non-parametric inner tolerance limit with confidence level of 95 % [9]) (h) no major sensory deficits. The inclusion criteria (a)–(e) were assessed by asking the subjects to complete a questionnaire which was subsequently checked with the examiner during a brief interview. To avoid selecting a sample of ‘‘supernormal’’ subjects we did not exclude individuals with pharmacological wellcompensated hypertension and diabetes. Seventy subjects (23.7 % of the whole sample) were discharged because they did not fulfil the inclusion criteria. The normative sample (NS) was composed of 225 subjects (111 men, 114 women), the mean age was 70.1 years (SD ± 5.7, range 60–80) and the mean education was 9.9 years of formal schooling (SD ± 4.6, range 5–23). Subjects’ stratification according to age, years of formal education and sex is reported in Table 1. After providing the subjects with a definition of anterograde episodic memory, the self-reported memory problems were investigated paying particular attention to induce the correct comparison by asking the participants
123
6–8
9–13
C14
Overall (225) by age
60–64 years Males
4
6
7
5
22
Females
4
5
6
5
20
65–70 years Males
5
13
10
6
34
Females
10
14
7
6
37
71–75 years Males
7
8
10
5
30
Females
17
5
5
6
33
Males
9
6
4
6
25
Females
9
3
7
5
24
76–80 years
Overall (225) by education Males
25
33
31
22
111
Females
40
27
25
22
114
‘‘do you have memory problems relative to most people your age?’’ [10]. None of the normative sample subjects replied positively. Nonetheless it was not an inclusion criterion. The reason why we have not inserted the selfreported memory and cognitive complaints among the selection criteria is that we prefer to give more importance to objective measures (memory prose score and MMSE score) rather than subjective ones. This is in line with recent research that has questioned the use of subjective memory complaint as an MCI diagnostic criterion, since some studies have reported that the relationship between subjective memory complaint and objective memory impairment is not clear [11–13]. The study was approved by the local ethics committee and was performed in accordance with the Helsinki Declaration. The informed written consent was obtained from all participants. Materials and procedure Each participant was assessed in a single session structured as follows: first part––administration of MMSE and PMT [8] (the 10-min interval between the immediate recall and the delayed recall was filled with the execution of the Raven Progressive Coloured Matrices 47 [14]); second part––15 min interval to minimize the effects of fatigue; third part—administration of MoCA-Italian version [1, 4].
Neurol Sci
Although there is evidence of a higher sensitivity of the word list learning test in detecting memory deficit due to MCI [15–17] and Alzheimer disease [15] than story recall test, we chose to use the latter for decreasing the potential effect of the proactive interference on the memory recall task of MoCA. This is justified because some studies showed that story recall test is sensitive in distinguishing normal from MCI and AD [16, 18]. Moreover, the story recall test is more ecological than the word list learning test, because it is similar to an everyday meaningful discourse that can be found in a newspaper, television and conversation.
Table 2 The effects of age, education and sex within the linear regression model of MOCA test (for further details see ‘‘Statistical methods’’) Simple regression
Simultaneous regression
Age (linear): F(1,223) = 37.931, p \ 0.0001
Age (linear, education partialled out): F(1, 222) = 30.381, p \ 0.0001
Education (1/education years): F(1,223) = 69.208, p \ 0.0001
Education (1/education years, age partialled out): F(1, 222) = 60.633, p \ 0.001
Sex: F(1,223) = 1.261, ns Adjusted score: Raw score ? 0.175 (age - 70.080) ? 24.300 [(1/ education years) - 0.126]
Statistical methods Control scores were analysed by means of simultaneous multiple regression, to check the influence of age, education (years of schooling) and sex. For the MoCA test we searched the best fitting linear regression model that could be used to adjust original scores according to the demographic variables. To this aim, we estimated a linear model able to calculate the score expected for a given subject on the basis of his/her age, education and sex. The simultaneous regression included the variables that resulted significant when considered one at the time: at this stage the effect of each predictor variable was studied partialling out the effect held in common with the other terms of the model. Taking this model as a basis, we calculated from the raw score an adjusted score, by adding or subtracting the contribution given by each significant concomitant variable in the final correction model. Following this approach, scores can be directly confronted across subjects of different demographic data. Adjusted scores were then ranked, and by means of a non-parametric procedure [19] we set tolerance limits. In this way we found both the outer and inner one-sided tolerance limits [20]. Above the outer tolerance limit one expects to find at least 95 % of the normal population (with 95 % confidence): hence, when a score is below the outer tolerance limit, the subject can be declared ‘‘not normal’’ with 95 % confidence. Above the inner tolerance limit one expects to find at most 95 % of the population (with 95 % confidence): hence, when a score is above the inner tolerance limit, the subject can be declared ‘normal’ with 95 % confidence. When a score falls between the outer and inner tolerance limits, no inferentially controlled judgement is possible. Given our sample size, outer and inner tolerance limits were fixed, respectively, on the basis of the values corresponding to the 6th and 18th ranked scores after demographic adjustments. To avoid errors due to the fixed upper limit of the test scores, no adjustment was made to scores of 30/30. The adjusted scores were classified into five categories
(equivalent scores) endowed at least with an ordinal relationship: 0 = scores lower than the outer 5 % tolerance limits; 4 = scores higher than the median value of the sample; 1, 2 and 3 = intermediate scores between the central value and the pathology threshold on a quasiinterval scale. To illustrate the raw score adjustment, consider the case of a 75-year-old male with 5 years of schooling. The original raw score achieved on MoCA was 19/30; the adjusted score becomes 19 ? 2.659 = 21.659. The corresponding equivalent score is 3.
Results The results of the regression analyses are shown in Table 2. The influence of age and education was always significant: the effect of age was most significant without any mathematical transformations of the original scale, whereas the effect of education was most significant considering the reciprocal of the original value of this variable, i.e. 1/years of schooling. This effect was confirmed on simultaneous regression. No difference between females and males was observed. Table 3 reports the correction grid for the most frequent combinations of age and education; intermediate values can be obtained by interpolation or using the original linear models reported in Table 2. Inner and outer tolerance limits are also reported: subjects with scores below the outer limit should be considered pathological, while those with scores above the inner limit can be considered normal. Subjects with scores included between the outer and the inner limits are better viewed as ‘‘borderline’’ cases. The values delimiting the Equivalent Scores are reported in Table 4. For a descriptive presentation of MoCA raw scores according to age and educational level, see Table 5.
123
Neurol Sci Table 3 Adjustments to be added to, or subtracted from, the raw scores according to age and education (expressed as years of schooling) Education
Age 55
60
65
70
75
80
85
5
-0.841
0.034
0.909
1.784
2.659
3.534
4.409
8
-2.663
-1.788
-0.913
-0.038
0.837
1.712
2.587
13
-3.832
-2.957
-2.082
-1.206
-0.332
0.543
1.418
17
-4.271
-3.396
-2.521
-1.646
-0.771
0.104
0.979
19
-4.422
-3.547
-2.672
-1.796
-0.922
-0.047
0.828
Outer and inner non-parametric one-sided tolerance limits are also reported (see ‘‘Statistical methods’’) Outer limit 17.362, inner limit 19.262
Table 4 Equivalent scores (ES) classification of adjusted scores Equivalent scores
Score interval
ES = 0
0 ? 17.362
ES = 1
17.363 ? 19.500
Density
Cumulative frequency
6
6
16
22
ES = 2
19.501 ? 21.562
36
58
ES = 3
21.563 ? 23.361
54
112
ES = 4
[23.361
113
225
ES = 0 corresponds to an inferentially controlled judgement of being below the norm; 4 is equal or better than the 50th percentile; 1, 2 and 3 are intermediate between 0 and 4 on a quasi-interval scale
Table 6 shows the frequency distributions for each subtest obtained by the normative sample. The correlation between the MoCA scores and the MMSE scores was 0.486 (df = 223, p \ 0.001. The correlation between these variables after adjustment for age and education was 0.321 (df = 219, p \ 0.001).
Discussion The collection of normative data represents the first step to introduce a new test in the clinical practice. MoCA is now added to the list of screening tests that can be used on Italian people to detect subjects with cognitive deficit taking into account the effect of the demographic variables.
Table 5 Descriptive statistics (mean ± standard deviations) of raw scores obtained on MoCA stratified by age and education
123
Age (years)
The mean of MoCA raw scores of 23.28 (SD ± 3.22) observed in the present study is lower than the normal controls in the original work (NC): 26.90 (SD is not reported by the authors) [21]. This may be attributable to the fact that though the mean ages of the two samples are comparable (NC 72.84 ± 7.03 years, NS 70.1 ± 5.7 years), our sample is less educated (mean 9.9 ± 4.6 years) than the NC (mean 13.33 ± 3.40 years). This discrepancy does not change even if the raw scores are adjusted in accordance with the instructions provided by the development group of MoCA, and reported in the Italian version of the test, too (?1 point if the subject’s educational level is B12 years). In doing so our normative sample obtains a mean of 23.96 ± 3.10, whereas the mean of the NC becomes 27.37 ± 2.20. The difference might derive from an incorrect estimation of the effect of education on the MoCA scores in the Canadian study, however, this is just a conjecture. Furthermore, if the cut-off proposed by the authors and reported in the Italian version is used (26), 74 % of our healthy sample (166 out of 225 subjects) would be classified as cognitively compromised. Another possible cause of the discrepancy is the difference in the selection criteria used to enrol subjects in the two studies. Narazaki et al. [22] have hypothesized that the strict criteria of admission used by Nasreddine et al. (no memory or cognitive complaints, normal baseline neuropsychological performance, no abnormality in neurological examination and computed tomography
Education years 5
Overall 6–8
9–13
C14
60–64
23.50 ± 2.27
25.09 ± 2.12
25.08 ± 2.60
26.70 ± 1.95
25.17 ± 2.43
65–70
19.47 ± 3.54
23.59 ± 2.34
26.00 ± 1.54
25.42 ± 1.88
23.61 ± 3.36
71–75
21.04 ± 3.07
23.31 ± 2.98
24.47 ± 2.53
24.73 ± 1.95
22.97 ± 3.14
76–80
20.67 ± 3.09
21.67 ± 2.74
22.09 ± 2.88
22.55 ± 2.07
21.59 ± 2.80
Overall
20.88 ± 3.25
23.52 ± 2.65
24.61 ± 2.69
24.82 ± 2.41
23.28 ± 3.22
Neurol Sci Table 6 MoCA: cognitive domains, subtests and frequency distributions for each subtest obtained by the normative sample (N = 225) Cognitive domains
Subtests
Frequency distributions for each subtest obtained by the normative sample Points scored
Memory recall
Visuospatial construction
Recall 5 words
Clock drawing
Copying a cube Executive functions
Attention, concentration and working memory
Language
0
85 (37.8 %)
1 2
25 (11.1 %) 43 (19.1 %)
3
36 (16.0 %)
4
22 (9.8 %)
5
14 (6.2 %)
0
1 (0.4 %)
1
9 (4.0 %)
2
56 (24.9 %)
3
159 (70.7 %)
0
57 (25.3 %)
1
168 (74.7 %)
Trail making test B modified
0
79 (35.1 %)
1
146 (64.9 %)
Phonemic fluency
0
90 (40.0 %)
1
135 (60.0 %)
Verbal abstraction
0 1
43 (19.1 %) 98 (43.6 %)
2
84 (37.3 %)
Target tapping Serial subtraction
0
22 (9.8 %)
1
203 (90.2 %)
0
2 (0.9 %)
1
2 (0.9 %)
2
25 (11.1 %)
3
196 (87.1 %)
Digit span (forward and backward)
0
12 (5.3 %)
1
54 (24.0 %)
2
159 (70.7 %)
Confrontation naming
0
0 (0 %)
1
2 (0.9 %)
Repeating two sentences Orientation
Frequency count for each point
To time and place
2
30 (13.3 %)
3 0
193 (85.8 %) 6 (2.7 %)
1
63 (28.0 %)
2
156 (69.3 %)
0
0 (0 %)
1
0 (0 %)
2
0 (0 %)
3
0 (0 %)
4
0 (0 %)
5
39 (17.3 %)
6
186 (82.7 %)
scan) may have led to select a ‘‘supernormal’’ sample, that may not represent the target population. For the aforementioned reasons, the Nasreddine et al.’s correction and cut-off values are not adequate for the Italian population. Our results (mean MoCA raw scores ± SD and percentage of individuals who scored below 26) are in line with those reported by studies in which similar selection criteria to ours were used (exclusion of subjects with disorders that may impact on cognition; normal performance on cognitive screening tests; levels of age and schooling comparable to ours): [22, 23]. In the Japanese study the means of the MoCA raw scores obtained by the age categories of 65–75 and 70–80 years were 22.9 ± 3.4 and 21.6 ± 3.7, respectively; and 75.1 % of the normative sample fell below the 26, despite the ?1 point correction. Also the Portuguese study reported in the age interval of C65 year (mean age 71.96 ± 5.43) a lower mean MoCA raw score than the Canadian one: 22.71 ± 3.60. Moreover, our sample shows frequency distributions for each MoCA subtest comparable to those obtained by the Japanese one, except for the verbal abstraction and sentence repeating. Our computed equivalent scores allow comparison between MoCA performance and other neuropsychological measures (e.g. Mental Deterioration Battery [14], Free and Cued Selective Reminding Test [24], Frontal Assessment Battery [25]). Further studies with an MCI patients group will be required for determining the sensibility, specificity, positive and negative predicting values of the Italian version of MoCA. Acknowledgments We are grateful to Erminio Capitani for his comments on a preliminary draft of this study. Daniella Mancini revised the English text. Thanks to Maria Leggieri for logistic help. Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Nasreddine ZS, Phillips NA, Be´dirian V, Charbonneau S, Whitehead V, Collin I, Cummings JL, Chertkow H (2005) The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 53:695–699 2. Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, Rabins PV, Ritchie K, Rossor M, Thal L, Winblad B (2001) Current concepts in mild cognitive impairment. Arch Neurol 58:1985–1992 3. Storandt M, Grant EA, Miller JP, Morris JC (2006) Longitudinal course and neuropathologic outcomes in original vs. revised MCI and in pre-MCI. Neurology 67:467–473 4. Pirani A, Tulipani C, Neri M (2006) italian translation of MoCA test and of its instructions. http://www.mocatest.org
123
Neurol Sci 5. Micieli G, Bottini G, Cerri C et al (2012) Complicanze psicocognitive dell’ictus. In: SPREAD stroke prevention and educational awareness diffusion. Ictus cerebrale: linee guida italiane di prevenzione e trattamento, 7th edn. Pierrel Research Italy Spa, Milano, p 110 6. Folstein MF, Folstein SE, McHugh PR (1975) ‘‘Mini-mental state’’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189–198 7. Measso G, Cavarzeran F, Zappala` G, Lebowittz BD, Crook TH, Pirazzolo FJ, Amaducci LA, Massari D, Grigoletto F (1993) The Mini-Mental State Examination: normative study of an Italian random sample. Dev Neuropsychol 9:77–85 8. Spinnler H, Tognoni G (1987) Standardizzazione e Taratura Italiana di Test Neuropsicolgici. Ital J Neurol Sci 6:25–27 9. Capitani E (1997) Normative data and neuropsychological assessment. Common problems in clinical practice and research. Neuropsychol Rehabil 7:295–309 10. Salthouse TA (1991) Theoretical perspectives on cognitive aging. Lawrence Erlbaum Associates Inc, Hillsdale, p 438 11. Lenehan ME, Klekociuk SZ, Summers MJ (2012) Absence of a relationship between subjective memory complaint and objective memory impairment in mild cognitive impairment (MCI): is it time to abandon subjective memory complaint as an MCI diagnostic criterion? Int Psychogeriatr 24:1505–1514 12. Purser JL, Fillenbaum GG, Wallace RB (2006) Memory complaint is not necessary for diagnosis of mild cognitive impairment and does not predict 10-year trajectories of functional disability, word recall, or short portable mental status questionnaire limitations. J Am Geriatr Soc 54:335–338 13. Roberts JL, Clare L, Woods RT (2009) Subjective memory complaints and awareness of memory functioning in mild cognitive impairment: a systematic review. Dement Geriatr Cogn Disord 28:95–109 14. Carlesimo GA, Caltagirone C, Gainotti G (1996) The Mental Deterioration Battery: normative data diagnostic reliability and qualitative analyses of cognitive impairment. The group of the standardization of the Mental Deterioration Battery. Eur Neurol 36:378–384 15. Baek MJ, Kim HJ, Kim S (2012) Comparison between the story recall test and the word-list learning test in Korean patients with
123
16.
17.
18.
19. 20.
21.
22.
23.
24.
25.
mild cognitive impairment and early stage of Alzheimer’s disease. J Clin Exp Neuropsychol 34:396–404 Rabin LA, Pare´ N, Saykin AJ, Brown MJ, Wishart HA, Flashman LA, Santulli RB (2009) Differential memory test sensitivity for diagnosing amnestic mild cognitive impairment and predicting conversion to Alzheimer’s disease. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn 16:357–376 Tremont G, Miele A, Smith MM, Westervelt HJ (2010) Comparison of verbal memory impairment rates in mild cognitive impairment. J Clin Exp Neuropsychol 32:630–636 Christensen H, Hardzi-Parlovic D, Jacomb P (1997) The psychometric differentiation of dementia from normal aging: a metaanalysis. Psychol Assess 3:147–155 Wilks KK (1941) Determination of sample size for setting tolerance limits. Ann Math Stat 12:91–96 Capitani E, Laiacona M (2000) Classification and modelling in neuropsychology: from groups to single cases. In: Boller F, Grafman J, Rizzolatti G (eds) Handbook of neuropsychology, 2nd edn, vol 1, pp 53–76 Nasreddine ZS, Phillips N, Chertkow H (2012) Comment on: Normative data for the Montreal Cognitive Assessment (MoCA) in a population-based sample. Author replay. Neurology 78:765–766 Narazaki K, Nofuji Y, Honda T, Matsuo E, Yonemoto K, Kumagai S (2013) Normative data for the Montreal cognitive assessment in a Japanese community-dwelling older population. Neuroepidemiology 40:23–29 Freitas S, Simo˜es MR, Alves L, Santana I (2011) Montreal Cognitive Assessment (MoCA): normative study for the Portuguese population. J Clin Exp Neuropsychol 33:989–996 Frasson P, Ghiretti R, Catricala` E, Pomati S, Marcone A, Parisi L, Rossini PM, Cappa SF, Mariani C, Vanacore N, Clerici F (2011) Free and Cued Selective Reminding Test: an Italian normative study. Neurol Sci 32:1057–1062 Apollonio I, Leone M, Isella V, Piamarta F, Consoli T, Villa ML, Forapani E, Russo A, Nichelli P (2005) The Frontal Assessment Battery (FAB): normative values in an Italian population sample. Neurol Sci 26:108–116
