Normative data of the Montreal Cognitive Assessment for the blind (MoCA-22) for the Czech population

Overview

Aim: The Montreal Cognitive Assessment (MoCA) is one of the most widely used cognitive screening tests in adults with reference standards for the Czech population. The MoCA-22 variant is designed for individuals with visual impairment or upper limb immobility and can be administered over the telephone. This study presents the Czech MoCA-22 normative standards. Materials and methods: The sample (N = 1,049) consists of participants from four studies conducted in the Czech Republic. The subjects included were aged 19–98 years, and were without neurodegenerative, psychiatric, or other serious illness. Data for the MoCA-22 were derived from data obtained by the standard version of MoCA. Following established clinical practice and statistical analysis, the population and derived norms are divided into three age categories: 19–50 years, 51–74 years, and 75 years and older. Results: For these age categories above, which were further subdivided by educational status (lower, higher), we present mean scores and estimated percentile thresholds. Performance in the MoCA-22 is affected by demoraphic factors, such as educational status and age but not sex, as reflected by the regression equation. Conclusions: Normative data for MoCA-22 will complement the clinical armamentarium in Czechia and allow adequate cognitive screening in people whose health status limits them when using standard methods.

Keywords:

Montreal Cognitive Assessment – screening – immobility – Telemedicine – normative data – blind – phone

This is an unauthorised machine translation into English made using the DeepL Translate Pro translator. The editors do not guarantee that the content of the article corresponds fully to the original language version.

 

Introduction

Screening (also orientation) cognitive tests refer to a group of currently widely used instruments for the assessment of cognitive status in people with psychiatric, neuropsychiatric and neurological diseases [1]. The Montreal Cognitive Assessment (MoCA) is currently one of the most commonly used screening tests to assess memory, as well as visual-spatial, executive, attentional, and speech abilities, including the Czech version [2,3], which can also be psychometrically well evaluated using normative data [4,5].

In previous years, a number of validation studies from the Czech population have shown to be very effective in detecting persons with cognitive deficits compared to persons without cognitive deficits due to various neurodegenerative diseases [6-8].

Furthermore, there are Czech abridged versions of the MoCA showing similar validity characteristics to the standard [9], but with faster administration, two parallel versions for retests [10], or studies providing a reliable change index (RCI) for repeated measures and differentiation of cognitive decline in older persons over time [11].

However, in terms of flexible adaptation of the test, the version for people with visual impairment, upper limb immobility or administered over the telephone, which entered the literature as "MoCA-22", "MoCA for the blind" (MoCA-blind) or "MoCA over the telephone" (telephone-MoCA), is missing in the Czech environment because it excludes items in the test that require visual processing or eye-hand coordination (according to the domains of visual-constructive ability and executive function; animal naming) [12-15]. For the MoCA-22, similar to the standard version, a number of cut-off scores (cut-off ≤ 18 or ≤ 19) have been proposed; however, it is unclear to what extent these scores, which were derived from foreign studies, are transferable to differential diagnosis in the Czech population [13-15] and to what extent they depend on demographic variables.

Therefore, in this report we would like to provide normative data for MoCA-22 in the Czech population. Furthermore, our aim is to develop a regression equation to adequately capture demographic factors with a significant influence on MoCA-22 performance in order to extend the evaluation and its applicability to Czech individuals with suspected cognitive impairment.

 

Material and methodology

The current study combines data from a total of four independent projects, namely the National Normative Study of Cognitive Determinants of Healthy Ageing (NANOK) [4,16], Defining č early cognitive deficits in Parkinson's disease in the context of genetic polymorphisms̊ and other biomarkers̊ [18], the Uniform Data Set Adaptation Memory Examination (UDS-III) battery [17], and the Validity of MoCA for detecting mild cognitive impairment in Parkinson's disease [8]. Three of the four projects mentioned above have already resulted in publications [4,8,18]. The study included volunteers aged ≥ 19 years with both genders (female and male) and two levels of education (lower < 12 years and higher ≥ 12 years). Exclusion criteria included the presence of neurodegenerative disease (e.g., dementia, epilepsy), psychiatric disease (e.g., depressive disorder), current or past brain injury resulting in unconsciousness, stroke or undergoing multiple general anesthesia, and alcohol or other drug dependence. In addition, participants who did not indicate their gender as one of the two categories (female/male, n = 2), did not indicate the number of years of education (n = 1), or were unable to calculate their total MoCA-22 score due to failure to complete one of the domains (n = 9) were excluded. After applying all the aforementioned criteria, our research sample comprised a total of 1,049 participants with an age range of 19-98 years (Table 1).

The research team conducting the participant examinations consisted of psychologists, psychiatrists, and others in the health and helping professions. All were fully trained to administer and score the cognitive tests used to ensure greater agreement among raters. Participants were examined individually in a health care facility, office, or home in a standardized testing environment to achieve the most uniform testing conditions. The study protocol consisted of the English standard version for all tests used, with MoCA version 7.1 used in all studies. In addition, the Short Cognitive State Examination (Mini-Mental State Examination; MMSE), Clock Draw -⁠ ing Test (CDT), Rey-Auditory Verbal Learning Test; RAVLT), Philadelphia Verbal Learning Test (PVLT), Brief Visual-Memory Test (BVMT-R), Wechsler Adult Intelligence Scale, Third Revision Subtests (Wechsler Adult Intelligence Scale; WAIS-III) Digit Span -⁠ Forward and Backward, Symbol and Digit Ordering Test, and Symbol-Encoding; Wechsler Memory Scale, Logical Memory Test Third Revision (Wechsler Memory Scale; WMS-III), Semantic Verbal Fluency Test with categories of animals, plants, hand tools, supermarket, and vegetables, Letter Verbal Fluency Test with categories of letters K, N, M, P, R, and S, Prague Stroop Test (PST), Trail Making Test versions A and B (TMT), and Boston Naming Test-30 (BNT-30) in the NANOK study [16]. Craft Story-21 (Craft Story), Multilingual Naming Test (MINT), Number Ordering Test -⁠ Forward and Backward (Number Span Test) and Benson Figure (Benson Figure) were used in the normative study for UDS-III [17].

Data analysis was performed using the statistical software R (2024). Multiple regression was performed to analyze the relationship between total MoCA-22 score, age, education and gender of the participants. ANOVA with post-hoc tests was used to assess age group assignment (based on non-random quota sampling). Because MoCA-22 results correlate with age, it is appropriate to divide the sample into a total of 3 age groups for normative purposes (Table 1).

 

Results

Dependence of the MoCA-22 test on age, education and sex

Associations between MoCA-22 total score and age (continuous factor), education (2 levels), and sex (2 levels) were assessed using a multiple regression model without interactions. We found that total MoCA-22 score was associated with age (regression coefficient b = -0.055; 95% confidence interval [confidence interval; CI]: -0.065, -0.045; p < 0.001). The mean score of participants in the higher education category was 1.2 points higher than that of participants in the lower education category (95% CI: -1.484, -0.904; p < 0.001). We found no effect of gender (female: b = 0.22; 95% CI: -0.047, 0.496; p = 0.106). We performed an ANOVA with age as the independent variable and found a significant effect of age on performance on the MoCA-22 (F[2, 1 045] = 75.91; p < 0.001). Table 2 shows the descriptive statistics of the MoCA-22 score based on two categories of education (lower and higher) and three age groups (19-50, 51-74, and ≥ 75 years). Table 3 shows the estimated percentile bounds for each of the crude MoCA-22 scores stratified by age and education.

The following regression equation can be established to calculate the expected MoCA-22 performance for individuals given age and education:

expected gross MoCA-22 score = 20.25 -⁠ 0.055 × age (in years) + 0.12 × education (in years)

 

The category issues disappear when we use a regression equation that informs us about expected performance given the age and education of the examinee. For example, if we have a 70-year-old patient with 10 years of education, we plug (20.25 -⁠ 0.055 × 70 + 0.12 × 10 = 17.6) into the equation and the expected score after rounding is 18 points.

 

Discussion

The Montreal Cognitive Test has become a popular screening tool to detect cognitive impairment in the elderly population, as well as in neurological and psychiatric patients not only in the Czech Republic but also abroad. We have used the synthesis of large projects that collected data from healthy volunteers of the full version of MoCA [4,8,18] to present to the professional community representative data for the MoCA-22 version, which is used for the blind, visually impaired, individuals with impaired upper limb mobility, or as part of telemedicine cognition testing. As with the baseline version of the MoCA, the MoCA-22 version has been shown to be influenced by age and education [4,8,18], which is consistent with international studies [12-15]. Because the MoCA-22 test is shorter, differences between age and education cohorts were smaller than in the full version but remained significant. In categorical data processing, the definition of categories is central. The largest of the datasets used (NANOK) used two age categories because individuals aged 60 years and older entered the study [4]. Thus, when the sample was expanded to include younger individuals, we addressed the question of whether to make a division into 4 categories of 19-50 years, 51-59 years, 60-74 years, and over 74 years, but the above solution did not prove to be justified after using statistical analysis; instead, the statistical analysis supported dividing the sample into 3 age groups. If we want to evaluate the result of screening according to percentile estimates (Table 3), then the result above the 15th percentile is within the norm (screening is negative), the result ≤ 15th percentile means positive screening (the examinee is in the range of below 1 standard deviation (SD) from the standard of the Czech population). A result between the 2nd-15th percentile (in the band below 1-2 SD) is suspicious for cognitive performance at the level of mild cognitive impairment, and a result below the 2nd percentile is suspicious for cognitive performance at the level of dementia syndrome. If the screening result is grossly inconsistent with the patient's functioning (self-sufficiency), the screening can be repeated with a time interval, but it must be supplemented by a comprehensive neuropsychological examination. The existence of a rehearsal effect at retest excludes the presence of a progressive disease and points to the influence of confounding phenomena such as anxiety, low motivation or unpreparedness for cognitive testing.

The strength of our study is the large sample of individuals without cognitive impairment. A limitation of the study is the absence of data from individuals with visual impairment, but studies conducted abroad indicate that the difference in cognitive performance between individuals with and without visual impairment is negligible [19]. Another limitation of the study is that the testing was conducted under ideal testing conditions in a face-to-face setting, not at a distance as is common with the telephone form. However, foreign data demonstrate the equivalence of MoCA administered by telephone with MoCA-22 tested in person [14]. Another limitation of the study is the absence of data from patients with cognitive impairment that would allow validation of the test and the establishment of cut-off values with sensitivity and specificity data. This is the next logical step in the development of MoCA-22. Our study provides an estimate of percentile norms for the adult population in the Czech Republic stratified by age and education, which may provisionally replace the existence of borderline scores from the validation study or be an alternative to it. The MoCA borderline scores adjusted by age and education have achieved high specificity at all levels of cognitive impairment with a compromise sensitivity [20]. The existence of normative data for MoCA-22 expands the range of neuropsychological tools that can be used with some groups of disabled citizens or allow for remote examination of the patient, either because of immobility or quarantine.

 

Ethical aspects

The work was carried out in accordance with the Helsinki Declaration of 1975 and its revisions in 2004 and 2008.

 

Grant support

This study was supported by the Internal Grant Agency of the Ministry of Health NT13145, as well as by the Cooperatio program, provided by Charles University, Health Sciences area, conducted at the Faculty of Arts, Charles University and the Neuroscience 207038 program, 3rd Faculty of Medicine, Charles University, Prague.

 

Conflict of interest

The authors declare that they have no conflict of interest in relation to the subject of the study.

 

Table 1. Descriptive statistics of the MoCA-22 cohort (n = 1,049).

Variable	19-50 years	51-74 years	≥ 75 years
Sex (M/F)	24/50	284/313	177/201
education (number of years)	15,54 ± 3,36	14,05 ± 3,46	12,86 ± 3,87
Age (number of years)	32,95 ± 11,25	66,23 ± 5,68	82,67 ± 4,74
MoCA-22 domains
attention	5,78 ± 0,63	5,61 ± 0,69	5,39 ± 0,84
Speech	2,86 ± 0,39	2,70 ± 0,57	2,41 ± 0,78
abstraction	1,89 ± 0,35	1,71 ± 0,51	1,61 ± 0,58
remote equipment	3,59 ± 1,58	2,49 ± 1,65	1,79 ± 1,54
orientation	5,93 ± 0,30	5,89 ± 0,40	5,81 ± 0,45
MoCA-22 (total skin)	20,07 ± 2,02	18,40 ± 2,15	17,02 ± 2,46

 

Values (excluding gender) are presented as mean and standard deviation.

M, male; MoCA, Montreal Cognitive Assessment; F, female

 

 

Table 2. Descriptive statistics for MoCA-22 skin corrected for age and education.

	MoCA-22 scores
Age	Education	n	M ± SD	Median (IQR)
19-50 years	< 12	15	18,33 ± 2,87	19 (4)
19-50 years	≥ 12	59	20,51 ± 1,48	21 (2)
51-74 years	< 12	250	17,89 ± 2,28	18 (4)
51-74 years	≥ 12	347	18,76 ± 1,98	19 (3)
≥ 75 years	< 12	199	16,60 ± 2,56	17 (4)
≥ 75 years	≥ 12	179	17,47 ± 2,27	18 (3)
19-98 years	5-29	1 049	18,02 ± 2,42	18 (4)

IQR, interquartile range; M, mean; MoCA, Montreal Cognitive Assessment; n, number of participants in each group; SD, standard deviation

 

 

Table 3. Normative data of the Czech version of MoCA-22 adjusted by age and education.

	MoCA-22 score
	Age 19-50 years		Age 50-74 years		Age ≥ 75 years
Percentiles	RV < 12	RV ≥ 12	RV < 12	RV ≥ 12	RV < 12	RV ≥ 12
95	22	22	21	22	21	21
90	21	22	21	21	20	21
85	21	22	20	21	20	20
80	21	22	20	21	19	19
75	20	22	20	20	19	19
70	20	22	19	20	18	19
65	20	21	19	20	18	18
60	19	21	19	19	17	18
55	19	21	18	19	17	18
50	19	21	18	19	17	18
45	19	21	18	19	16	17
40	18	21	17	18	16	17
35	17	20	17	18	15	17
30	17	20	17	18	15	16
25	16	19	16	17	15	16
20	15	19	16	17	14	16
15	15	19	15	17	14	15
10	14	18	15	16	13	15
5	13	17	14	15	13	14
2	13	17	12	15	11	13

MoCA -⁠ Montreal Cognitive Assessment; RV -⁠ years of education (number of years of schooling)