Examining the Validity and Utility of Montreal Cognitive Assessment (MoCA) Domain Scores for early Neurocognitive Disorders

Abstract

Objectives:

MoCA total scores have been widely used to identify individuals with neurocognitive disorders (NCDs), but utility of its domain-specific scores have yet to be thoroughly interrogated. This study aimed to validate MoCA’s six domain-specific scores (i.e. Memory, Language, Attention, Executive, Visuospatial, and Orientation) with conventional neuropsychological tests and explore whether MoCA domain scores could discriminate between different aetiologies in early NCDs.

Design:

Baseline data of a cohort study.

Setting and Participants:

Study included 14,571 participants recruited from Alzheimer’s Disease Centers across United States, aged ⩾50 years, with global Clinical Dementia Rating of ≤1, and mean age of 71.8±8.9 years.

Methods:

Participants completed MoCA, conventional neuropsychological tests, and underwent standardized assessments to diagnose various aetiologies of NCDs. Partial correlation coefficient was used to examine construct validity between Z-scores of neuropsychological tests and MoCA domain scores, while multinomial logistic regression examined utility of domain scores to differentiate between aetiologies of early NCDs.

Results:

MoCA domain scores correlated stronger with equivalent constructs (r=0.15–0.43, ps<0.001), and showed divergence from dissimilar constructs on neuropsychological tests. Participants with Alzheimer’s disease were associated with greater impairment in Memory, Attention, Visuospatial, and Orientation domains (RRR=1.13–1.55, ps<0.001). Participants with Lewy body disease were impaired in Attention and Visuospatial domains (RRR=1.21–1.47, ps<0.001); participants with frontotemporal lobar degeneration were impaired in Language, Executive, and Orientation domains (RRR=1.25–1.75, ps<0.01); and participants with Vascular disease were impaired in Attention domain (RRR=1.14, p<0.001).

Conclusions and Implications:

MoCA domain scores approximate well-established neuropsychological tests, and can be valuable in discriminating different aetiologies of early NCDs. While MoCA domain scores may not fully substitute neuropsychological tests especially in the context of diagnostic uncertainties, they can complement MoCA total scores as part of systematic evaluation of early NCDs, and conserve the use of neuropsychological tests to patients who are more likely to require further assessments.

Brief summary:

MoCA domain scores approximate well-established neuropsychological tests. They can complement MoCA total scores in routine practice, and allow right-sitting of patients who require further neuropsychological assessment.

INTRODUCTION

Neurocognitive disorders (NCDs) involves the loss of one or more cognitive abilities and affects one’s everyday functionality.¹ Older adults are particularly susceptible to NCDs, with the presence of cognitive impairment as an accelerator to mortality² at substantial socioeconomic cost.³ Montreal Cognitive Assessment (MoCA) has been used in clinical settings as a tool to detect NCDs due to its many desirable characteristics.^4,5 Compared to the widely-used Mini-Mental State Examination (MMSE), MoCA is available for use without licensing fee, possesses robust measures of visuospatial and executive function, and has better diagnostic performance in detecting Mild Cognitive Impairment (MCI).^6,7

There is growing interest in the literature to examine the usefulness of different profiles of cognitive deficits in MoCA, for the purpose of diagnosing various NCDs aetiologies. Recently, Julayanont and colleagues⁸ derived six MoCA domain scores (MDS; comprising Memory, Language, Attention, Executive, Visuospatial, and Orientation) based on published neuropsychological and neuroimaging evidence, and found that individuals with MCI who showed impairment in MoCA–Memory domain were at greater risk of Alzheimer’s disease. Other researchers extended the utility of MDS to discriminate amnestic MCI,⁹ Alzheimer’s disease,¹⁰ and aphasic dementia¹¹ from normal cognition. Specifically, individuals with Alzheimer’s disease scored poorer in memory and orientation domains, while individuals with progressive aphasia showed greater impairment in language and attention domain scores.¹² Goldstein and colleagues¹⁰ evaluated incremental validity of MDS and MoCA total scores between normal cognition, MCI, and Alzheimer’s disease. They revealed a combination of Memory, Executive, and Orientation domain scores yielded stronger incremental validity than total MoCA scores in differentiating Alzheimer’s disease from cognitively normal persons. However, their results were limited to MCI and Alzheimer’s disease, and not generalizable to other NCDs.

Notwithstanding the promise of MDS, literature remains sparse with focus mostly on MCI and Alzheimer’s disease. Some noteworthy limitations in previous studies include relatively small sample sizes, lack of NCDs representation in conditions such as frontotemporal or vascular dementia, and absence of control groups. This study sought to bridge the literature gap and validate MDS in a large cohort of older persons—spanning from normal cognition to various aetiologies in MCI and early dementia. Specifically, we aimed to: (1) examine construct validity between MDS and conventional neuropsychological tests; and (2) evaluate clinical utility of MDS in discriminating various aetiologies of early NCDs (i.e. MCI and early dementia).

METHODS

Study Cohort

The study was conducted using National Alzheimer’s Coordinating Center (NACC) database, based on participants recruited from Alzheimer’s Disease Centers (ADCs) across United States between March 2015 (the date when MoCA was first introduced in NACC database) and August 2021. Participants who fulfilled the following criteria at baseline were included: (1) aged ⩾50 years; (2) provided data on MoCA; and (3) had global Clinical Dementia Rating of ≤1 (indicating the range from normal cognition to early NCDs). The ADCs obtained informed consent from participants, and received approval by their local institutional review boards.

Measures

MoCA is a brief cognitive screening tool with a total score of 30, with scores ≤26 suggestive of cognitive impairment. MDS were derived based on the calculation proposed by Julayanont et al.,⁸ by summing respective items in each domain. Calculation of MDS is shown in Table 1, and briefly described below:

1. Memory domain score is calculated based on the number of words correctly remembered in a 5-word list in free delayed recall (3 points per word), followed by category-cued recall (2 points per word), and recognition recall (1 point per word); with a maximum score of 15.

2. Language domain score is obtained by adding raw scores for animals naming, sentence repetition, and letter fluency; with a maximum score of 6.

3. Attention domain score is calculated by adding raw scores for the words recalled in both trials of immediate recall, digit span forward and backward, letter tapping, serial-7 subtraction, and sentence repetition; with a maximum score of 18.

4. Executive domain score is obtained through addition of modified Trail-Making Test Part B, clock drawing, digit span forward and backward, letter tapping, serial-7 subtraction, letter fluency, and abstraction; with a maximum score of 13.

5. Visuospatial domain score is scored by adding raw scores of cube copy, clock drawing, and animals naming; with a maximum score of 7.

6. Orientation domain score is computed based on total points from MoCA’s orientation section; with a maximum score of 6.

Table 1.

Calculation of Montreal Cognitive Assessment (MoCA) domain scores, adapted from Julayanont et al.¹²

MoCA Domain Scores (MDS)
Items	Memory	Language	Attention	Executive	Visuospatial	Orientation
Trails Making	-	-	-	1	-	-
Cube Copy	-	-	-	-	1	-
Clock Drawing	-	-	-	3	3	-
Animals Naming	-	3	-	-	3	-
5-Word Registration	-	-	2 trials of immediate recall (max=10)	-	-	-
Digits Span	-	-	2	2	-	-
Letter A Tapping	-	-	1	1	-	-
Serial-7 Subtraction	-	-	3	3	-	-
Sentence Repetition	-	2	2		-	-
Letter Fluency	-	1	-	1	-	-
Abstraction	-	-	-	2	-	-
Delayed Recall Without Cue	3 points per word recalled; (max=15)	-	-	-	-	-
Delayed Recall, Category Cue	2 points per word recalled (max=10)	-	-	-	-	-
Delayed Recall, Recognition	1 point per word recalled (max=5)	-	-	-	-	-
Orientation-Date, Month, Year, Day, Place, City	-	-	-	-	-	6
Totals	15	6	18	13	7	6

Conventional neuropsychological tests measure cognitive performance across different domains, namely, immediate memory (Craft Story 21 Immediate Recall),¹³ visuospatial abilities (Benson Complex Figure Copy),¹⁴ delayed memory (Craft Story 21 Delayed Recall and Benson Complex Figure Recall),^13,14 language (Multilingual Naming Test),¹⁵ and Verbal Fluency–L-words),¹⁶ attention (Number Span Test Forward and Backward),¹⁷ processing speed (Trail Making Test Part A),¹⁸ and executive function (Trail Making Test Part B).¹⁸ Scores for each neuropsychological test were converted to age-, sex- and education-adjusted Z-score based on published calculator for NACC cohort.¹⁹

Clinical Dementia Rating (CDR) (CDR^® Dementia Staging Instrument)²⁰ is a well-validated and widely-used scale for staging of cognitive impairment.²¹ CDR employs a semi-structured interview with both participant and informant to rate performance in 6 domains (memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care). Based on published rules,²⁰ responses from the 6 domains are used to assign a global CDR score to indicate severity of cognitive impairment: 0=normal cognition, 0.5=questionable cognitive impairment, 1=mild dementia, 2=moderate dementia, and 3=severe dementia. It noteworthy that this study only included participants with global CDR of ≤1 given the study’s primary aim to examine the utility of MDS in discriminating various aetiologies of early NCDs (i.e., MCI and early dementia).

Diagnoses of the clinical groups (namely normal cognition, MCI, or dementia) were derived from standardized assessments, which included clinical history, physical examination and conventional neuropsychological testing. Majority of the diagnoses (88.2%) were determined through consensus conference, while the remaining were made by single clinicians. MCI was diagnosed using modified Petersen criteria,²² while Dementia was diagnosed following McKhann and colleagues’ criteria.²³ For those diagnosed with MCI and dementia, clinicians at ADCs were required to employed their best judgment to further classify the presumptive primary aetiology of NCD when participants fulfilled the established criteria for a specific aetiology (i.e. Alzheimer’s disease^23,24, Lewy Body disease,^25–27 Frontotemporal lobar degeneration (FTLD),^26,28–33 Vascular disease, or other aetiologies of NCDs); and the specific aetiology was deemed by clinicians to be the primary cause of the observed cognitive impairment. In particular, the presence of Vascular disease was confirmed by clinical evidence of symptomatic stroke (i.e. abrupt onset of focal neurological signs) or neuroimaging evidence of one or more of the following: (1) cystic infarcts (large or small); (2) significant white matter changes (Grade 7 to 8+ on Cardiovascular Health Study White Matter Grade Scale);³⁴ (3) intraparenchymal hemorrhage; or (4) multiple microbleeds.

Statistical Analysis

For Aim (1), construct validity of MDS was evaluated on the full sample (including normal cognition, MCI and early dementia), using partial correlation coefficient (r) to adjust for the influence of various MDS on Z-scores of conventional neuropsychological tests. We expected MDS to show convergence (i.e. stronger correlation) with similar cognitive domains, and show divergence (i.e. weaker correlation) with dissimilar cognitive domains of neuropsychological tests.

For Aim (2), clinical utility of MDS was evaluated on the subsample with MCI and early dementia, using multinomial logistic regression to examine the relationship of MDS to various aetiologies of MCI and early dementia, and adjusting for covariates of age, gender, ethnicity, and years of education. We expected the underlying aetiologies of early NCDs to depict differential profile of MDS due to distinctive cognitive impairments affecting different parts of the brain. Based on published literature for early NCDs, those with a primary aetiology of Alzheimer’s disease would show greater impairment in Memory, Attention, and Orientation domains; Lewy body disease group would show greater impairment in Attention, Executive, and Visuospatial domains; FTLD group would show greater impairment in Language and Executive domains; Vascular disease group would show greater impairment in Attention and Executive domains; and other aetiologies of NCDs group would show lesser degree of impairment across all MDS than Alzheimer’s disease, Lewy body disease, FTLD, and Vascular disease due to mixed aetiologies of cognitive impairment.

All analyses were conducted in Stata (version 16).

RESULTS

The Flow diagram on participant selection is presented in Supplement material 1. As shown in Table 2, a total of 14,571 participants were included, with a mean age of 71.8±8.9 years, mean duration of education 16.0±2.9 years, and mean MoCA total score of 23.1±5.4. Mean MDS between the primary aetiologies of NCDs, among participant with mild MCI and dementia is presented in Supplementary material 2.

Table 2.

Characteristics of participants across Neurocognitive Disorders (NCDs) diagnosis (n=14,571).

Variable	Overall sample (n=14,571)	Nonnal cognition (n=7,987)	MCI (n=3,534)	Dementia (n=3,050)
Age, mean (SD)	71.8 (8.9)	71.2 (8.7)	73.2 (8.8)	71.6 (9.7)
Female sex, n (%)	8,460 (58.1)	5,245 (65.7)	1,753 (49.6)	1,462 (47.9)
Years of education, mean (SD)	16.0 (2.9)	16.2 (2.8)	15.9 (3.0)	15.6 (3.1)
Ethnicity, n (%)
White	11,690 (81.1)	6,221 (78.7)	2,795 (79.9)	2,674 (88.7)
African American	1,892 (13.1)	1,206 (15.3)	479 (13.7)	207 (6.9)
Others/Unknown	836 (5.8)	480 (6.1)	224 (6.4)	132 (4.4)
MoCA total score, mean (SD)	23.1 (5.4)	26.1 (2.9)	22.2 (3.6)	16.1 (5.5)
Aetiologies of NCDs, n (%)
Alzheimer’s Disease	4,418 (30.3)	NA	2,185 (62.0)	2,233 (73.2)
Lewy Body Disease	428 (2.9)	NA	218 (6.2)	210 (6.9)
FTLD	602 (4.1)	NA	141 (4.0)	461 (15.1)
Vascular Disease	304 (2.1)	NA	246 (7.0)	58 (1.9)
Other Aetiologies of NCDs	832 (5.7)	NA	744 (21.1)	88 (2.9)

Note: SD, standard deviation; MCI, mild cognitive impairment; MoCA, Montreal Cognitive Assessment; FTLD, frontotemporal lobar degeneration; NA, not applicable.

Construct validity was assessed by examining the relationship between conventional neuropsychological tests and MDS (Table 3). After controlling for the influence of confounding MoCA domains, partial correlation coefficients ranged from −0.13 to 0.43, ps<0.01. As expected, Benson Complex Figure Copy correlated strongest with MoCA–Visuospatial (r=0.19, p<0.001); Craft Story 21 Immediate Recall, Craft Story 21 Delayed Recall, and Benson Complex Figure Recall correlated strongest with MoCA–Memory (r=0.33–0.43, ps<0.001); Multilingual Naming Test and Verbal Fluency–L-words correlated strongest with MoCA–Language (r=0.20–0.25, ps<0,001); Number Span Test Forward correlated strongest with MoCA–Attention (r=0.20, p<0.001); Number Span Test Backward correlated strongest with MoCA–Executive (r=0.17, p<0.001) and MoCA–Attention (r=0.15, p<0.001); Trail Making Test Part A and Part B correlated strongest with MoCA–Executive (r=0.16, ps<0.001).

Table 3.

Partial correlation between the Z-scores of conventional neuropsychological tests and the MoCA domain scores (n=14,571).

Neuropsychological tests (Z-score)	Cognitive domains	MoCA-Memory	MoCA-Language	MoCA-Attention	MoCA-Executive	MoCA-Visuospatial	MoCA-Orientation
Benson Complex Figure Copy	Visuospatial	−0.03**	−0.13***	0.10***	0.11***	0.19 ***	0.09***
Craft Story 21 Immediate Recall	Immediate memory	0.33 ***	0.06***	0.10***	0.04***	0.04***	0.13***
Craft Story 21 Delayed Recall	Delayed memory	0.43 ***	0.04***	0.03***	0.04***	0.04***	0.18***
Benson Complex Figure Recall	Delayed memory	0.42 ***	−0.07***	0.01	0.03***	0.14***	0.26***
Multilingual Naming Test	Language	0.13***	0.20 ***	0.08***	−0.00	0.19***	0.11***
Verbal Fluency–L-words	Language	0.06***	0.25 ***	−0.04***	0.20***	−0.09***	0.01
Number Span Test Forward	Attention	−0.05***	0.12***	0.20 ***	0.07***	−0.06***	−0.03***
Number Span Test Backward	Attention	0.04***	0.04***	0.15 ***	0.17 ***	−0.01	−0.01
Trail Making Test Part A	Processing speed	0.09***	0.01	0.07***	0.16 ***	0.06***	0.01
Trail Making Test Part B	Executive function	0.12***	0.02*	0.10***	0.16 ***	0.06***	0.04***

Note: MoCA, Montreal Cognitive Assessment;

Salient values are in boldface;

^*P<0.05

^**P<0.01

^***P<0.001.

Clinical utility of MDS was assessed using multinomial logistic regression by examining the relationship of MDS to various aetiologies of MCI and dementia (Table 4). Those with a primary aetiology of Alzheimer’s disease were more likely to have impairment in Memory, Attention, Visuospatial, and Orientation; and less likely to have impairment in Language. Those with Lewy body disease were more likely to have impairment in Attention and Visuospatial, and less likely to have impairment in Language. Those with FTLD were more likely to have impairment in Language, Executive, and Orientation; and less likely to have impairment in Attention. Those with Vascular disease were more likely to have impairment in Attention. The relationship between MDS and various aetiologies of NCDs can also be visualized using the predicted probability from multinomial logistic regression (Figure 1). Lower scores in Memory and Orientation domains were associated with the highest probability of Alzheimer’s disease. Lower scores in Language and Executive domains indicated a higher probability of FTLD. Lower scores in Attention and Visuospatial domains indicated a higher probability of Alzheimer’s disease and Lewy Bodies.

Table 4.

Adjusted multinominal logistic regression for the association between MoCA domains and aetiologies of NCDs, among participants who have mild cognitive impairment and dementia (n=6,584).

Aetiologies of NCDs
MoCA Domains	Alzheimer’s Disease			Lewy Body Disease			FTLD			Vascular Disease
	RRRa	95% CI	p values	RRRa	95% CI	p values	RRRa	95% CI	p values	RRRa	95% CI	p values
Memory	1.15	1.12-1.17	<.001	1.00	0.96-1.03	0.804	1.01	0.98-1.05	0.391	1.00	0.96-1.04	0.944
Language	0.85	0.78-0.94	0.001	0.70	0.61-0.80	<.001	1.75	1.54-1.99	<.001	0.95	0.82-1.11	0.548
Attention	1.13	1.08-1.19	<.001	1.21	1.13-1.29	<.001	0.93	0.87-0.99	0.019	1.14	1.06-1.23	<.001
Executive	0.95	0.90-1.01	0.116	1.07	0.98-1.17	0.114	1.26	1.16-1.36	<.001	1.01	0.92-1.12	0.806
Visuospatial	1.21	1.11-1.33	<.001	1.47	1.28-1.68	<.001	0.98	0.87-1.11	0.773	1.03	0.89-1.20	0.664
Orientation	1.55	1.39-1.72	<.001	1.10	0.95-1.28	0.190	1.25	1.10-1.43	0.001	1.08	0.91-1.28	0.365

Note: NCDs, Neurocognitive Disorders; MoCA, Montreal Cognitive Assessment; FTLD, Frontotemporal Lobar Degeneration; RRR, Relative Risk Ratio; CI, Confidence Interval;

Other Aetiologies of NCDs was used as the reference group in the model; Significant values are in boldface.

^aModel included the MoCA domains (Memory, Language, Attention, Executive, Visuospatial, Orientation), and adjusted for the covariates of age, gender, ethnicity, and years of education.

Figure 1.

Predicted probability of the MoCA domain scores across different aetiologies of MCI and early dementia, stratified by (a) Memory domain, (b) Language domain, (c) Attention domain, (d) Executive domain, (e) Visuospatial domain, and (f) Orientation domain. Note: MCI, Mild Cognitive Impairment; AD, Alzheimer’s Disease; LBD, Lewy Body Disease; FTLD, Frontotemporal Lobar Degeneration; VD, Vascular Disease; Other, Other Aetiologies of NCDs.

DISCUSSION

Summary of Findings

Although MoCA has been validated and used to assess a wide array of cognitive dysfunctions, the validity and clinical utility of MDS has not been adequately established in extant literature. Present study examined the validity and utility of MDS across individuals from normal to impaired cognition and in various aetiologies of early NCDs. The results indicate convergence of MDS with equivalent constructs and divergence with dissimilar constructs of conventional neuropsychological tests, which evidence the potential use of MDS as a valid tool to detect specific subsets of cognitive deficits. Further analyses also demonstrated the clinical utility of MDS in bringing to bear the differential profiles of cognitive impairment in individuals with varying underlying aetiologies of early NCDs.

Interpretation of Findings

The construct validity of MDS is supported by the stronger correlation with equivalent constructs of well-established conventional neuropsychological tests, which builds on previous studies for its use as an appropriate measure of cognitive function.^10,12 The clinical utility of MDS is further demonstrated in this study given the ability of MDS to discriminate between individuals of different NCD aetiologies. The distribution in Figure 1 shows that persons with a primary aetiology of Alzheimer’s disease were associated with greater impairment on Memory, Attention, Visuospatial, and Orientation domains. Persons with Lewy body disease were more likely to demonstrate greater impairment in Attention and Visuospatial domains. Persons with FTLD revealed greater impairment in Language and Executive domain. These findings are coherent with the hallmark symptoms of the respective NCDs: Alzheimer’s disease involves memory loss, problems with decision making, and confusion with time and place; Lewy body disease tends to effect deficits in concentration and change in movement; and FTLD is marked by a change in behaviour, executive dysfunction and language impairment.

Two findings differed slightly from our initial hypotheses. First, Alzheimer’s disease was found to be associated with Visuospatial impairment. As younger patients from 50 years of age were included in the sample, there may have been a subset of patients with posterior cortical atrophy—a subtype of Alzheimer’s disease—that could contribute to deficits in the Visuospatial domain. It is also conceivable that the categorisation of clock drawing under both Executive and Visuospatial domains does not distinguish whether the task is impaired due to executive dysfunction, or visuospatial deficit, or both. Hence, patients with Alzheimer’s disease and impaired clock drawing due to executive dysfunction alone may be inappropriately classified to have Visuospatial impairment as well. Second, the Vascular disease group exhibited only one impairment, the Attention domain. Attention is a subset of executive function, which is involved in frontal lobe function. Small vessel disease typically makes up a significant proportion of Vascular cognitive disease and causes frontal lobe dysfunction due to disruption of the frontal-subcortical circuits. It is thus plausible that deficits in the Attention domain can be attributed to frontal executive dysfunction that typifies Vascular cognitive disorders. Nevertheless, the findings could also suggest that MDS alone may not confer a sufficiently granular profiling of cognitive deficits among individuals with Vascular disease. As such, MDS may not be an adequate substitute for more detailed neuropsychological tests if Vascular disease is suspected clinically.

Clinical Implications

In light of recent research suggesting the usefulness of MDS in diagnosing MCI and dementia, our findings demonstrate MDS are valid and approximate neuropsychological tests in detecting cognitive deficits, and may be used to differentiate the aetiologies of early NCDs. In routine clinical practice, MDS can complement MoCA total scores to provide a more granular understanding of specific cognitive functions. Information from MDS can thus serve as a low cost, time-saving, and convenient means to facilitate timely identification of the primary aetiology of NCDs, as well as highlight impaired cognitive domains that warrant closer monitoring and targeted cognitive rehabilitation. In cases where MDS could not confer conclusive profiling of the cognitive domains, it would then justify a referral for detailed neuropsychological assessment. Such approach – which incorporates MDS as part of the systematic evaluation of early NCDs – would conserve the limited healthcare resources related to detailed neuropsychological assessment, and allow right-siting of patients who are more likely to benefit from these neuropsychological assessments.

Limitations

Notwithstanding the implications of current findings, MoCA in general has several limitations compared to conventional neuropsychological assessment, which should caution readers against using MDS as a full substitute for conventional neuropsychological assessment. First, MoCA lacks the rigour and comprehensiveness of neuropsychological assessment, hence may not confer a sufficiently granular profiling of cognitive deficits, especially for ambiguous cases and early deficits. For example, original Trail Making Test Part B has more than twice the numbers presented in MoCA Trail Making Test B that makes the original task more challenging and thorough. Second, some of the MoCA items may assess more than 1 cognitive domain. As a case in point, the clock drawing item in MoCA tests both the Executive and Visuospatial domains. Thus, when there is an impairment in the MoCA task, there can be difficulty in isolating the specific domain implicated. Third, MoCA does not give due consideration to qualitative data that is critical in NCDs assessments. For example, behavioral observation and information provided through collateral interviews are crucial to arrive at an accurate diagnosis, such as the behavioral variant of FTLD. Fourth, there is no performance validity test embedded in the process of MDS development, which is often a key requirement for conventional neuropsychological tests. Fifth, healthcare professionals who administer MoCA are only required to complete a short training and certification process. As such, the administration procedures for MoCA are relatively less rigorous than neuropsychological tests and can introduce some variability in the administration and results.

Besides MoCA limitations, it is worth noting that the results might not be generalizable to population-based samples as the database was drawn from clinical setting of ADCs. Participants were majority Caucasians, so other ethnicities are under-represented. Future research is needed to replicate the findings across different racial and ethnicity groups, as well as establish normative data from a diverse group of individuals. In addition, 11.8% of diagnoses were made by single clinicians, they may not necessarily be as robust as those made via consensus meetings.

CONCLUSIONS AND IMPLICATIONS

Recent research has illuminated the potential use of MDS in clinical practice, but the utility of MDS has yet to be extensively validated. This study showed promising construct validity and enhanced clinical utility of MDS in discriminating between individuals with different aetiologies of early NCDs. Despite being a much briefer assessment tool, distinctive profiles of MDS may be integrated into routine clinical practice to inform the differential aetiologies of early NCDs. While MoCA domain scores may not be a full substitute for detailed neuropsychological tests especially in ambiguous cases or those with very mild deficits, they can complement MoCA total scores as part of systematic evaluation of early NCDs, and conserve the use of detailed neuropsychological tests to patients who are more likely to benefit from further assessments. Future research needs to establish normative data from a diverse group of participants before MDS may be used to support clinical evaluation of early NCDs.