Rapid cognitive assessment: Accuracy and discriminant validity of Mini-Cog and process-based Clock Drawing Test

Abstract

Background

The Mini-Cog is a brief cognitive examination comprising a three-item memory recall and a simplified Clock Drawing Test (CDT). There is limited research on the effects of detailed scoring criteria for the Mini-Cog on cognitive screening.

Objective

To assess the diagnostic effectiveness of three Mini-Cog versions and a new process-based CDT test in identifying mild cognitive impairment (MCI) and Alzheimer's disease (AD) compared to cognitively normal controls (NC).

Methods

We prospectively enrolled 950 subjects who underwent standardized neuropsychological assessments and the Mini-Cog test. The CDT was scored using an adapted 10-point scale. The accuracy of three Mini-Cog versions (Mini-Cog1: 3-word delayed recall + 2-point CDT; Mini-Cog2: 3-word immediate recall + 3-word delayed recall + 3-point CDT; Mini-Cog3: 3-word immediate recall + 3-word delayed recall + 10-point CDT) was assessed through the receiver operating characteristic analysis. Sensitivity and specificity were determined for each diagnostic threshold.

Results

The optimal cut-off point for Mini-Cog3 is 12/16 for MCI and 10/16 for AD. Mini-Cog3 demonstrated the highest diagnostic efficacy, with AUCs of 0.82 (95% CI: 0.78–0.85) for MCI and 0.95 (95% CI: 0.94–0.97) for AD, with sensitivity of 85% for both MCI and AD. The CDT's AUCs were 0.77 (95% CI: 0.73–0.81) for MCI, and 0.87 (95% CI: 0.84–0.90) for AD, with sensitivity of 89% for MCI, and 82% for AD.

Conclusions

A more elaborate scoring system, such as Mini-Cog3, may serve as an effective screening method for the rapid and accurate detection of cognitive dysfunction in patients with MCI and AD.

Keywords

Alzheimer's disease clock drawing test mild cognitive impairment Mini-Cog

Introduction

With the global population aging rapidly, the prevalence of dementia is increasing at an alarming rate, significantly impacting the physical and mental well-being of older adults.¹ Since there are no effective treatments for moderate to severe dementia, the focus has shifted towards early identification and intervention in patients with mild cognitive impairment (MCI), who are at higher risk of developing dementia.² Given the rising prevalence of dementia and its significant impact, rapid screening is essential, especially in general care settings and among large populations. However, commonly used tools like the Mini-Mental State Examination (MMSE)³ and the Montreal Cognitive Assessment (MoCA)⁴ face limitations in busy Chinese clinical environments due to their lengthy administration time (10–15 min) and the requirement for trained personnel. This highlights the urgent need for a brief, effective cognitive screening tool that can be easily administered without extensive training.

The Mini-Cognitive Assessment (Mini-Cog), a concise and commonly utilized cognitive examination, combines a memory task and a streamlined version of the Clock Drawing Test (CDT).⁵ It has several advantages: it is quick to administer (typically taking 3–5 min), easy to perform, highly acceptable to patients, and requires minimal training.⁶ Additionally, the Mini-Cog is not influenced by cultural or linguistic factors and does not require specialized equipment, making it a feasible tool for use by non-specialists in various settings.⁵ Despite the fact that its precision fluctuates depending on the area and interpretation approach, it has been proven to have high sensitivity and specificity in identifying cognitive dysfunction.⁷ These attributes make the Mini-Cog a promising tool for widespread application in community and outpatient settings, facilitating early detection and intervention for cognitive impairments.

The CDT, recognized as one of the most established and extensively utilized brief cognitive assessments,⁸ is regarded by certain scholars as an optimal tool for cognitive screening due to its inherent simplicity and conciseness.⁹ However, its numerous scoring methods—ranging from quantitative to qualitative analyses based on the clock face, numbers, and hands—pose challenges for clinical practitioners due to their varying complexity and effectiveness.⁹ Borson, in developing the Mini-Cog, selected a simplified, free version of the CDT, focusing on the correctness of number sequence, placement, and hand direction without complex scoring details.⁵ In practice, we found that the execution strategies during the drawing process, including participants’ memory of instructions, varied across different cognitive states. Additionally, research indicates that the incidence of “12-3-6-9” anchoring points in the CDT is significantly higher among MCI patients compared to normal participants, which enhances MCI detection rates.¹⁰ Furthermore, the position and length of the clock hands are crucial in screening for cognitive impairments.¹¹ Therefore, we designed a more elaborate CDT scoring criterion (10 points) based on the drawing process, and we evaluated the diagnostic efficacy of three different Mini-Cog versions, which vary in their word recall and CDT scoring methods, for detecting MCI and Alzheimer's disease (AD) from cognitively normal controls (NC).

Methods

Participants

This prospective cohort study enrolled 950 participants from the Memory Clinic at Shanghai Sixth People's Hospital and Shanghai communities through internet-based and print advertisements from April 2020 to January 2024. Inclusion criteria were age between 40 and 88 years, more than one year of education, possessing normal or corrected vision and hearing, capability of communication in Mandarin Chinese, willingness and ability to complete neuropsychological tests and cranial MRI examination. Individuals were excluded (1) if they had a history of alcohol addiction, substance misuse, or other serious neuropsychiatric conditions; (2) cognitive dysfunction due to causes such as brain trauma, intracranial space-occupying lesions, hydrocephalus, subdural hematoma, infections of the central nervous system (for instance, AIDS, syphilis), or chemical and physical influences (such as drug overdose, alcohol addiction, carbon monoxide poisoning); (3) MRI scans revealing significant localized lesions, such as more than two strokes larger than 2 cm in diameter, or strokes in crucial regions such as the thalamus, hippocampus, entorhinal cortex, perirhinal cortex, angular gyrus, and other cortical and subcortical gray matter nuclei; (4) inability to cooperate with cognitive function assessment or having contraindications for MRI examination.

This study was approved by the Ethical Committee of the Shanghai Sixth People's Hospital, and written informed consent was obtained from all individual participants (shown in Figure 1).

Figure 1.

Flow chart for clinical evaluation.

Neuropsychological assessment and diagnostic criteria

Cognitive assessment was conducted using standardized neuropsychological tests to evaluate overall cognitive function, domain-specific cognitive impairment, and other non-cognitive functions in all participants. The assessments were performed in the Neuropsychology Laboratory of the Geriatrics Department at the Shanghai Sixth People's Hospital by trained raters using standardized evaluation language and scoring criteria. Importantly, these assessors were blinded to the participants’ cognitive status during the Mini-Cog and CDT administration. Tests for overall cognitive assessment included the Montreal Cognitive Assessment Basic (MoCA-B)¹² and the Addenbrooke's Cognitive Examination III (ACE-III).¹³ The cognitive assessment battery encompassed domain-specific tests, including the Auditory Verbal Learning Test (AVLT)¹⁴ for evaluating memory, the Boston Naming Test (BNT)¹⁵ and Animal Verbal Fluency Test (AFT)¹⁶ for assessing language abilities, the Shape Trail Test Part A and B (STT-A, STT-B)¹⁷ for executive function, the Judgment of Line Orientation (JLO)¹⁸ for visuospatial ability, and the Symbol Digit Modalities Test (SDMT)¹⁹ for attention. The non-cognitive assessment scales included the Activities of Daily Living (ADL) scale and the Functional Activities Questionnaire (FAQ).²⁰

After completing the neuropsychological assessments, a separate team of clinicians, who were also blinded to the Mini-Cog and CDT results, conducted a thorough evaluation of each participant's cognitive status. This evaluation included a review of the neuropsychological test scores, medical history, and other relevant clinical information. Based on this comprehensive assessment, the clinicians assigned a clinical diagnosis to each participant. The NC group included participants who denied cognitive decline and exhibited cognitive assessment results within the normal range on all standardized neuropsychological scales. The MCI group was determined according to the Peterson diagnostic criteria,²¹ which involve subjective report of cognitive decline, objective evidence of cognitive impairment, essentially preserved functional activities of daily living, and absence of dementia. The AD group was diagnosed according to the National Institute on Aging-Alzheimer's Association (NIA-AA) criteria,²² requiring the presence of dementia, progressive and substantial impairment in memory and at least one other cognitive domain, with an insidious onset and a clear history of cognitive decline.

Mini-Cog and Clock drawing test

Three different scoring criteria for the Mini-Cog were established based on the Borson diagnostic criteria.⁵ Participants were instructed to remember three words and draw a clock face with all numbers and the hands set to 2:50. The CDT was scored using either a simplified system or a more elaborate process-based scoring system (shown in Textbox 1). Three Mini-Cog versions were scored as shown in Textbox 2.

Textbox 1.

Scoring criteria for process-based CDT.

Component	Scoring Criteria	Score
Executive Function	Correct placement of the 12, 3, 6 and 9 o’clock positions anchors	1
Executive Function	Wrong placement of the 12, 3, 6 and 9 o’clock positions anchors	0
Instructions Recall	Remember setting time without asking	3
	Ask once	2
	Ask twice	1
	Ask more than twice	0
Minute Hand Position	Minute hand correctly positioned at 10 o’clock without any modification	2
	Minute hand correctly positioned at 10 o’clock with modification	1
	No minute hand or minute hand not at 10 o’clock	0
Hour Hand Position	Hour hand correctly positioned between 2 and 3 o’clock, closer to 3, without any modification	3
	Hour hand correctly positioned between 2 and 3 o’clock, closer to 3, with modification	2
	Hour hand positioned between 2 and 3 o’clock, closer to 2	1
	No hour hand or hour hand not between 2 and 3 o’clock	0
Pointer Length	Minute hand longer than the hour hand	1
Pointer Length	No pointers or pointers of equal length	0
Total Score	Sum of the scores from each component	10

Textbox 2.

Scoring criteria for three Mini-Cog versions.

Versions	Immediate Recall	Delayed Recall	Clock Drawing Test	Total scores
Mini-Cog1 (short version)	N/A	3 points for recall of all 3 words	2 points for correct clock drawing	5 points
Mini-Cog2 (medium version)	3 points for recall of all 3 words	3 points for recall of all 3 words	3 points each for correct numbers order, numbers position, and hands orientation	9 points
Mini-Cog3 (long version)	3 points for recall of all 3 words	3 points for recall of all 3 words	10 points for process-based CDT	16 points

Statistical analysis

Statistical analyses were conducted utilizing SPSS version 26.0 (IBM, Armonk, New York, USA) and MedCalc version 19.5.6 (MedCalc Software bvba, Ostend, Belgium). A p-value ≤ 0.05 was statistical significance across all analyses. For normally distributed continuous variables—including age, years of education, overall cognitive screening scores, domain-specific cognitive scores (AVLT delayed recall, AVLT recognition, BNT, AFT, JLO), Mini-Cog scores, and CDT performance—three group differences were assessed using F-tests and one-way analysis of variance (ANOVA). For non-parametric distribution data—including gender, STT-A, STT-B, SDMT, ADL, and FAQ—three group differences were assessed using chi-square tests and Kruskal-Wallis tests. The diagnostic accuracy of the Mini-Cog and the process-based CDT scoring systems was evaluated across different cognitive states and stratified by age and education levels using the area under the receiver operating characteristic (ROC) curve (AUC), constructed with MedCalc. The AUC served as a measure of diagnostic accuracy, with 0.5 indicating no discriminative ability, 0.5 to 0.7 indicating low accuracy, 0.7 to 0.8 suggesting moderate accuracy, and 0.8 to 1.0 reflecting high accuracy. Optimal diagnostic cut-off points were identified using the Youden index, which represents the maximum sum of sensitivity and specificity.

Results

Demographic data and neuropsychological performance

The demographic characteristics, cognitive screening scores, standardized neuropsychological tests, and global function results for the NC, MCI, and AD groups were shown in Table 1. Participants with MCI and AD were significantly older and had lower educational levels than the NC group, and they also exhibited significantly worse performance on standardized neuropsychological tests, with lower scores in the AVLT, BNT, AFT, SDMT, JLO, and higher scores in the STT. Additionally, individuals with MCI and AD demonstrated a more impaired functional status, as reflected by higher scores on the FAQ and ADL scale.

Table 1.

Demographics and standardized neuropsychological tests for NC, MCI, and AD.

Variables	NC (n = 235)	MCI (n = 326)	AD (n = 389)	F/χ²	P
Demographics
Age (y)	64.01 ± 9.18	68.26 ± 7.19	71.65 ± 7.94	66.77	<0.001
Age level (Younger:Older)	168:67	181:145	131:258	88.75	<0.001
Education (y)	12.94 ± 3.43	10.75 ± 3.35	10.25 ± 3.53	47.00	<0.001
Education level (Low:High)	74:161	188:138	242:147	59.76	<0.001
Sex (M:F)	83:152	108:218	140:249	0.67	0.715
Cognitive screening tests
MoCA-BC	26.14 ± 2.34	20.35 ± 3.69	12.51 ± 4.40	1021.79	<0.001
ACE-III-CV	83.52 ± 6.68	68.02 ± 8.44	49.29 ± 11.95	931.03	<0.001
Standardized neuropsychological tests
AVLT delayed recall	6.29 ± 1.95	1.74 ± 1.79	0.42 ± 0.93	714.97	<0.001
AVLT recognition	22.00 ± 1.49	17.49 ± 2.59	15.03 ± 4.57	306.88	<0.001
BNT	25.38 ± 2.36	20.76 ± 4.22	19.16 ± 4.98	135.10	<0.001
AFT	18.70 ± 4.21	12.84 ± 3.96	10.07 ± 3.61	312.64	<0.001
JLO	21.66 ± 4.54	18.09 ± 5.01	16.58 ± 5.72	49.21	<0.001
STT-A	43.00 (36.00, 52.00)	56.00 (45.00, 70.00)	72.00 (58.00, 91.00)	177.72	<0.001
STT-B	116.00 (97.00, 145.00)	161.00 (130.75, 205.50)	213.82 (172.00, 253.50)	222.01	<0.001
SDMT	42.00 (35.50, 50.00)	30.00 (22.25, 37.00)	20.00 (15.00, 27.00)	279.31	<0.001
Global functional tests
ADL	20.00 (20.00, 20.00)	20.00 (20.00, 20.00)	22.00 (20.00, 27.00)	290.02	<0.001
FAQ	0.00 (0.00, 0.00)	0.00 (0.00, 1.00)	3.00 (0.00, 9.00)	225.28	<0.001

MoCA-BC: Chinese version of Montreal Cognitive Assessment-Basic; ACE-III-CV: Chinese version of Addenbrooke's Cognitive Examination III; AVLT: Auditory Verbal Learning Test; BNT: Boston naming test; AFT: Animal verbal fluency test; STT-A and B: Shape Trail Test Part A and B; SDMT: Symbol Digit Modalities Test; JLO: Judgement of Line Orientation; ADL: Activities of Daily Living; FAQ: Functional Assessment Questionnaire; NC: normal control; MCI: mild cognitive impairment; AD: Alzheimer's disease. Age level, Younger adults: < 70 years, Older adults: ≥ 70 years; Education level, Low: < 12 years, High: ≥ 12 years. Bold indicates statistical significance.

Similarly, performance on the Mini-Cog tests and CDT showed a consistent pattern of cognitive decline across the groups, as summarized in Table 2. NC participants scored the highest across all measures and AD group performed the worst.

Table 2.

Comparison of three Mini-Cog tests and CDT for NC, MCI, AD.

Index	NC (n = 235)	MCI (n = 326)	AD (n = 389)	F	p
MiniCog1 (short version)	3.53 ± 1.21	2.39 ± 1.11	1.00 ± 1.14	370.77	<0.001
MiniCog2 (medium version)	7.89 ± 1.03	6.83 ± 1.17	4.61 ± 1.65	483.29	<0.001
CDT	7.61 ± 1.61	5.81 ± 1.86	3.98 ± 2.68	207.4	<0.001
MiniCog3 (long version)	13.10 ± 1.82	10.68 ± 2.03	7.20 ± 3.12	437.49	<0.001

CDT: Clock Drawing Test; Mini-Cog: Mini-Cognitive Assessment; NC: normal control; MCI: mild cognitive impairment; AD: Alzheimer's disease.

ROC analyses for three Mini-Cog Tests and CDT to differentiate MCI and AD from NC

The diagnostic accuracy of three Mini-Cog tests and CDT in differentiating MCI and AD from NC was evaluated using ROC analysis (Table 3). All assessments demonstrated good diagnostic accuracy. For MCI, Mini-Cog3 had the highest AUC of 0.82 with a cutoff of 12, achieving a sensitivity of 0.85 and a specificity of 0.65. In the AD group, the diagnostic efficacy of both Mini-Cog2 and Mini-Cog3, with an AUC of 0.95, surpassed that of Mini-Cog1, which had an AUC of 0.92. The CDT showed variable but moderate to high diagnostic accuracy across all groups (AUC range: 0.77–0.87).

Table 3.

ROC analyses for three Mini-Cog tests and CDT to differentiate MCI and AD from NC.

	MCI
Index	AUC	95%CI	Cutoff	Sensitivity	Specificity	PPV	NPV	PLR	NLR	Accuracy
Mini-Cog1 (short version)	0.75	0.71–0.79	2	0.57	0.83	0.82	0.58	3.25	1.91	0.68
Mini-Cog2 (medium version)	0.75	0.71–0.79	7	0.68	0.70	0.76	0.61	2.29	2.20	0.69
CDT	0.77	0.73–0.81	7	0.89	0.55	0.73	0.78	1.98	4.87	0.75
Mini-Cog3 (long version)	0.82	0.78–0.85	12	0.85	0.65	0.77	0.75	2.43	4.24	0.76
	AD
Mini-Cog1(short version)	0.92	0.90–0.94	2	0.89	0.83	0.89	0.82	5.10	7.47	0.87
Mini-Cog2(medium version)	0.95	0.94–0.97	6	0.88	0.89	0.93	0.81	7.92	7.21	0.88
CDT	0.87	0.84–0.90	6	0.82	0.75	0.84	0.72	3.28	4.22	0.79
Mini-Cog3(long version)	0.95	0.94–0.97	10	0.85	0.91	0.94	0.79	9.09	6.08	0.87

CDT: Clock Drawing Test; Mini-Cog: Mini-Cognitive Assessment; NC: normal control; MCI: mild cognitive impairment; AD: Alzheimer's disease; AUC: area under the curve; PPV: positive predictive value; NPV: negative predictive value; PLR: positive likelihood ratio; NLR: negative likelihood ratio.

ROC analyses among groups of different age and education level

The AUCs for different tests across age and education levels were shown in Table 4. For MCI diagnosis, outside the group of individuals’ aged ≥ 70 years with less than 12 years of education, the Mini-Cog3 showed high diagnostic efficacy in other groups, with AUCs consistently above 0.80. For AD diagnosis, three Mini-Cog tests demonstrated high accuracy across all age and education groups (AUC: 0.87–0.98). In the population with age <70 years and education <12 years, the CDT demonstrated the highest diagnostic performance, with AUCs of 0.81 for MCI and 0.90 for AD.

Table 4.

ROC analyses among groups of different age and education level.

	Age < 70 years,	Age < 70 years,	Age ≥ 70 years,	Age ≥ 70 years,
	Education < 12 years	Education ≥ 12 years	Education < 12 years	Education ≥ 12 years
	MCI
	MCI: n = 107,	MCI: n = 74,	MCI: n = 80,	MCI: n = 65,
AUC (95%CI)	NC: n = 51	NC: n = 117	NC: n = 21	NC: n = 46
Mini-Cog1 (short version)	0.73 (0.64–8.81)	0.78 (0.71–0.84)	0.69 (0.55–0.84)	0.70 (0.60–0.78)
Mini-Cog2 (medium version)	0.72 (0.63–0.80)	0.77 (0.70–0.83)	0.68 (0.53–0.83)	0.74 (0.64–0.83)
CDT	0.81 (0.74–0.88)	0.74 (0.67–0.81)	0.76 (0.63–0.88)	0.70 (0.60–0.80)
Mini-Cog3 (long version)	0.82 (0.75–0.89)	0.80 (0.74–0.86)	0.77 (0.64–0.90)	0.80 (0.71–0.88)
	AD
	AD: n = 105,	AD: n = 26,	AD: n = 136,	AD: n = 122,
AUC (95%CI)	NC: n = 51	NC: n = 117	NC: n = 21	NC: n = 46
Mini-Cog1 (short version)	0.91 (0.86–0.95)	0.93 (0.88–0.98)	0.87 (0.78–0.95)	0.91 (0.86–0.96)
Mini-Cog2 (medium version)	0.93 (0.89–0.96)	0.98 (0.95–1.00)	0.91 (0.85–0.98)	0.95 (0.92–0.99)
CDT	0.90 (0.85–0.95)	0.86 (0.78–0.94)	0.86 (0.79–0.93)	0.84 (0.78–0.90)
Mini-Cog3 (long version)	0.94 (0.91–0.98)	0.95 (0.91–0.99)	0.93 (0.87–0.98)	0.96 (0.93–0.99)

CDT: Clock Drawing Test; Mini-Cog: Mini-Cognitive Assessment; NC: normal control; MCI: mild cognitive impairment; AD: Alzheimer's disease; AUC: area under the curve.

Discussion

In this study, we compared the value of three Mini-Cog tests and a process-based CDT scoring system in the diagnosis of MCI and AD. Our findings indicate that all three Mini-Cog versions demonstrated strong diagnostic performance for AD, with AUCs ranging from 0.92 to 0.95. However, the diagnostic efficacy for MCI was less robust, particularly for Mini-Cog1 and Mini-Cog2, which had AUCs of 0.75. Notably, by enhancing the scoring criteria of the CDT within Mini-Cog3, the diagnostic accuracy for MCI was significantly improved, achieving an AUC of 0.82. Comparing our findings with previous research, Carnero-Pardo reported a sensitivity of 60% and specificity of 90% for the Mini-Cog in diagnosing Cognitive Impairment (CI), which includes MCI or AD, using a cutoff score of ≤2.²³ This is similar to the results of our Mini-Cog1 (short version) for diagnosing MCI. However, our results for AD showed a higher sensitivity of 89% and a slightly lower specificity of 83%. Li compared 119 MCI patients with 110 normal controls, adopting a Mini-Cog scoring system totaling 9 points (consistent with our Mini-Cog2 scoring method). The study reported a sensitivity of 85.71% and a specificity of 79.41% for MCI detection, with an average Mini-Cog score of 5.2 ± 1.6.²⁴ However, in our study, the Mini-Cog2 demonstrated relatively lower sensitivity and specificity for MCI, at 68% and 70%, respectively. This discrepancy may be attributed to the overall better performance observed with the Mini-Cog2 in our cohort, evidenced by an average score of 6.83 ± 1.17, suggesting a potential variation in the diagnostic accuracy of the Mini-Cog2 across different populations or study settings. Furthermore, a recent meta-analysis found that the Mini-Cog is effective in detecting AD and MCI with sensitivities of 0.76 (AD) and 0.84 (MCI), and specificities of 0.83 (AD) and 0.79 (MCI).⁷ These figures align reasonably well with our findings, particularly with Mini-Cog3 showing the highest overall efficacy.

We also examined the diagnostic performance of three Mini-Cog tests across different age and educational level groups (Table 4). Regardless of age and educational background, Mini-Cog2 and Mini-Cog3 demonstrated higher diagnostic performance for AD compared to Mini-Cog1, with AUCs exceeding 0.9. This is consistent with the fact that the Mini-Cog demonstrates robust performance in detecting dementia, and its diagnostic accuracy remains unaffected across diverse demographic subgroups.²⁵ However, there is limited research on the effectiveness of Mini-Cog in diagnosing MCI. A study by Panita involved 150 participants (42 MCI, 108 NC) found that Mini-Cog had an AUC of 0.73, with a sensitivity of 57.4% and a specificity of 85.4%, showing similar performance across groups with different levels of education.²⁶ Simultaneously, our study revealed that the diagnostic efficacy of Mini-Cog1 for MCI, across any age and educational background, tends to be generally low (AUC: 0.69–0.78). Notably, when the 10-point CDT was integrated to form the new Mini-Cog3, its diagnostic efficacy for MCI increased to above 0.8 except for individuals aged ≥70 years with less than 12 years of education (AUC: 0.77). This suggests that a more detailed scoring system can enhance the accuracy of MCI diagnosis, but for older adults with lower educational levels, there may be a slight reduction in diagnostic effectiveness, which could be related to the limited cognitive reserve and the challenges in executing complex tasks among this demographic.

The CDT is widely acknowledged for its utility in cognitive function screening, particularly in the diagnosis of AD. Its effectiveness stems from its ability to assess a range of cognitive domains, such as executive function, visuospatial skills, and numerical understanding.⁹ The CDT has been adapted into numerous variants, each employing different scoring systems to assess diverse aspects of cognitive function, highlighting its versatility and ability to probe various cognitive domains.²⁷ Although the various CDT variants share similarities and correlations, they exhibit different diagnostic values. The CDT is typically effective in identifying moderate to severe dementia, demonstrating a sensitivity of 67% to 98% and a specificity of 69% to 94%.^28,29 However, its sensitivity and specificity for MCI detection have been found to fluctuate significantly, with sensitivities from 30% to 95% and specificities from 50% to 85%.³⁰ The traditional CDT scoring systems, such as those used in studies by Sunderland and Mendez have been identified as producing broader categorical scores that may overlook the nuanced deficits of MCI patients.^31,32 In contrast, our process-based CDT scoring method, which includes the initial placement of anchor numbers and the assessment of introductions recall and pointer length, demonstrated considerable diagnostic accuracy across all participant groups, with an AUC ranging from 0.77 to 0.87, indicating its robust performance. Notably, our process-based CDT achieved a high sensitivity of 89% for detecting MCI, with an AUC of 0.77. This suggests that our scoring framework is capable of detecting the finer impairments associated with MCI.

Neuroimaging studies have associated CDT performance with structural and functional changes in diverse brain regions, encompassing the frontal, temporal, and parietal cortices, as well as subcortical structures.^33,34 Our study introduces a novel process-based quantitative scoring system for the CDT, designed to enhance the diagnostic precision for MCI. This system includes specific numeric values for anchor numbers “12, 3, 6, 9” which are considered to be related to executive function.³⁵ The adept placement of these numbers by “anchoress” is linked to enhanced local network efficiency in key brain regions, including the medial orbitofrontal cortex and anterior cingulate,³⁶ suggesting a potential neuroanatomical basis for executive dysfunction in MCI. Furthermore, the assessment of instructions recall, such as remembering the setting time without asking, is another key feature of our scoring system. Failure to recall the setting time in the CDT not only leads to imprecise clock hand positioning but also serves as a key indicator of the episodic memory deficits commonly associated with AD.³⁵ The evaluation of hand position and pointer length within our scoring system is a pivotal analytical dimension, as it quantifies the test-taker's proficiency in the spatial and motor accuracy required to depict the clock hands. This task is intricately linked to the integrity of spatial cognition and fine motor control. More importantly, deviations in hand positioning, indicative of compromised spatial planning and conceptualization abilities, are often pathognomonic of the cognitive deficits that can arise from dysfunction within the frontal and parietal lobes,³⁷ which are critical for these cognitive processes.

The primary objective of our research was to demonstrate the effectiveness of our process-based CDT scoring system in identifying MCI patients. Our results support our hypothesis, showing high sensitivity for MCI detection. More importantly, the integration of CDT scores with word recall items into a combined assessment tool, termed Mini-Cog3, significantly improved diagnostic efficiency for MCI. This synergistic effect of the combined approach underscores the potential of integrating multiple cognitive assessment tools to enhance the accuracy of MCI diagnosis.³⁸

In addition to evaluating the diagnostic accuracy of three Mini-Cog versions, we also assessed the scoring time for each method. We observed that Mini-Cog1 took an average of 90 s to score in the NC group, whereas Mini-Cog3 required approximately 110 s. Scoring time for both versions increased by an additional 50 s when assessing AD participants. The longer scoring time for Mini-Cog3 is due to its more detailed scoring criteria, which, despite being more time-consuming, offer a more thorough cognitive evaluation that could potentially improve diagnostic accuracy for MCI. Although the increased scoring time is a consideration in resource-limited clinical settings, the enhanced diagnostic benefits of Mini-Cog3 may warrant the additional time investment. Further research is necessary to evaluate its cost-effectiveness in routine clinical practice.

However, several limitations warrant consideration in interpreting our findings. Firstly, the relatively modest sample size may hinder the generalizability of our results, necessitating validation through larger, multi-centric studies across diverse populations and clinical settings. Additionally, the lack of integration with neuroimaging data, such as MRI and PET-CT, leaves unexplored the potential correlation between our scoring system and the neurobiological insights that these modalities could provide. This represents a significant avenue for future research to enhance our understanding of the underlying pathological mechanisms of cognitive impairments. Finally, it is important to acknowledge that our study population primarily consisted of individuals from Shanghai communities, who tend to have higher education levels compared to rural areas. This demographic characteristic may have influenced the cognitive test results. Addressing these challenges will be crucial to ensure the scoring system's broad applicability and effectiveness in clinical practice.

Conclusion

Our study validates the detailed Mini-Cog3 scoring system as a significantly advantageous tool for diagnosing MCI and AD. This validation is achieved through the integration of a more elaborate process-based CDT scoring method and a three-word recall assessment, which collectively enhance the detection of early cognitive impairments. To further substantiate these findings, additional investigation into its clinical application and validation across diverse populations is warranted.

Footnotes

Acknowledgments

We would like to express our sincere gratitude to evaluators Xiangqing Xie and Yun Yang from the Neuropsychology Laboratory for their outstanding work in conducting standardized neuropsychological tests to our participants. We also extend our heartfelt thanks to Dr Lin Huang for her expertise in interpreting MRI reports, which was crucial for the exclusion of ineligible subjects. Additionally, we extend our deepest appreciation to Zhi-Rui Zhou from Huashan Hospital for his invaluable guidance on the statistical analysis included in our manuscript. Their support was vital to our research. Moreover, we wish to acknowledge the assistance of Kimi, a sophisticated language model developed by Moonshot AI (), which was used for language refinement of this manuscript. The precision it brought to our writing process significantly enhanced the quality of our final document.

ORCID iDs

Qian Yang

Hao Yang

Wei Long

Sheng Zuo

Defa Zhu

Qihao Guo

Author contributions

Qian Yang (Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Software; Visualization; Writing – original draft); Hao Yang (Formal analysis; Methodology; Validation; Writing – original draft); Wei Long (Data curation; Resources; Writing – review & editing); Sheng Zuo (Supervision; Writing – review & editing); Defa Zhu (Methodology; Supervision; Validation; Writing – review & editing); Qihao Guo (Conceptualization; Data curation; Funding acquisition; Methodology; Project administration; Resources; Supervision; Validation; Writing – review & editing).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Natural Science Foundation of China (82171198) and the Shanghai Municipal Science and Technology Major Project (2018SHZDZX01).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Date availability

The data supporting the findings of this study are available on request from the corresponding author.

References

Long

Benoist

Weidner

. World Alzheimer Report 2023: Reducing dementia risk: never too early, never too late. London: Alzheimer's Disease International, 2023. https://www.alzint.org/resource/world-alzheimer-report-2023/ .

Pardridge

. Alzheimer's disease: future drug development and the blood-brain barrier. Expert Opin Invest Drugs 2019; 28: 569–572.

Folstein

McHugh

. Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–198.

Damian

Jacobson

Hentz

, et al. The Montreal Cognitive Assessment and the mini-mental state examination as screening instruments for cognitive impairment: item analyses and threshold scores. Dement Geriatr Cogn Disord 2011; 31: 126–131.

Borson

Scanlan

Chen

, et al. The Mini-Cog as a screen for dementia: validation in a population-based sample. J Am Geriatr Soc 2003; 51: 1451–1454.

Borson

Scanlan

Watanabe

, et al. Simplifying detection of cognitive impairment: comparison of the Mini-Cog and Mini-mental state examination in a multiethnic sample. J Am Geriatr Soc 2005; 53: 871–874.

Abayomi

Sritharan

Yan

, et al. The diagnostic accuracy of the Mini-Cog screening tool for the detection of cognitive impairment-A systematic review and meta-analysis. PLoS One 2024; 19: e0298686.

Hazan

Frankenburg

Brenkel

, et al. The test of time: a history of clock drawing. Int J Geriatr Psychiatry 2018; 33: e22–e30.

Shulman

. Clock-drawing: is it the ideal cognitive screening test? Int J Geriatr Psychiatry 2000; 15: 548–561.

10.

Pinto

Peters

. Literature review of the Clock Drawing Test as a tool for cognitive screening. Dement Geriatr Cogn Disord 2009; 27: 201–213.

11.

Rouleau

Salmon

Butters

, et al. Quantitative and qualitative analyses of clock drawings in Alzheimer's and Huntington's disease. Brain Cogn 1992; 18: 70–87.

12.

Huang

Chen

Lin

, et al. Chinese version of Montreal Cognitive Assessment basic for discrimination among different severities of Alzheimer's disease. Neuropsychiatr Dis Treat 2018; 14: 2133–2140.

13.

Pan

Wang

Huang

, et al. Validation of the Chinese version of Addenbrooke's cognitive examination III for detecting mild cognitive impairment. Aging Ment Health 2022; 26: 384–391.

14.

Zhao

Guo

Liang

, et al. Auditory verbal learning test is superior to Rey-Osterrieth Complex figure memory for predicting mild cognitive impairment to Alzheimer's disease. Curr Alzheimer Res 2015; 12: 520–526.

15.

Qiao

Wang

, et al. Culture effects on the Chinese version Boston naming test performance and the normative data in the native Chinese-speaking elders in mainland China. Front Neurol 2022; 13: 866261.

16.

Zhao

Guo

Hong

. Clustering and switching during a semantic verbal fluency test contribute to differential diagnosis of cognitive impairment. Neurosci Bull 2013; 29: 75–82.

17.

Zhao

Guo

, et al. The Shape Trail Test: Application of a new variant of the Trail Making Test. PLoS One 2013; 8: e57333.

18.

Qualls

Bliwise

Stringer

. Short forms of the Benton judgment of line orientation test: development and psychometric properties. Arch Clin Neuropsychol 2000; 15: 159–163.

19.

Benedict

Smerbeck

. Construct validity of the symbol-digit modalities test. Mult Scler 2023; 29: 483–485.

20.

Pfeffer

Kurosaki

Harrah

Jr , et al. Measurement of functional activities in older adults in the community. J Gerontol 1982; 37: 323–329.

21.

Petersen

Doody

Kurz

, et al. Current concepts in mild cognitive impairment. Arch Neurol 2001; 58: 1985–1992.

22.

Hyman

Phelps

Beach

, et al. National Institute on Aging-Alzheimer's Association guidelines for the neuropathologic assessment of Alzheimer's disease. Alzheimers Dement 2012; 8: 1–13.

23.

Carnero-Pardo

Cruz-Orduna

Espejo-Martinez

, et al. Utility of the Mini-cog for detection of cognitive impairment in primary care: data from two Spanish studies. Int J Alzheimers Dis 2013; 2013: 285462.

24.

Dai

Zhao

, et al. Comparison of the value of Mini-Cog and MMSE screening in the rapid identification of Chinese outpatients with mild cognitive impairment. Medicine (Baltimore) 2018; 97: e10966.

25.

Borson

Scanlan

Brush

, et al. The Mini-cog: a cognitive ‘vital signs’ measure for dementia screening in multi-lingual elderly. Int J Geriatr Psychiatry 2000; 15: 1021–1027.

26.

Limpawattana

Manjavong

. The Mini-Cog, Clock Drawing Test, and three-item recall test: rapid cognitive screening tools with comparable performance in detecting mild NCD in older patients. Geriatrics (Basel) 2021; 6: 91.

27.

Rakusa

Jensterle

Mlakar

. Clock Drawing Test: a simple scoring system for the accurate screening of cognitive impairment in patients with mild cognitive impairment and dementia. Dement Geriatr Cogn Disord 2018; 45: 326–334.

28.

Chen

Jin

, et al. A comparison of six clock-drawing test scoring methods in a nursing home. Aging Clin Exp Res 2018; 30: 775–781.

29.

Lin

O'Connor

Rossom

, et al. Screening for cognitive impairment in older adults: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med 2013; 159: 601–612.

30.

Breton

Casey

Arnaoutoglou

. Cognitive tests for the detection of mild cognitive impairment (MCI), the prodromal stage of dementia: meta-analysis of diagnostic accuracy studies. Int J Geriatr Psychiatry 2019; 34: 233–242.

31.

Sunderland

Hill

Mellow

, et al. Clock drawing in Alzheimer's disease. A novel measure of dementia severity. J Am Geriatr Soc 1989; 37: 725–729.

32.

Mendez

Ala

Underwood

. Development of scoring criteria for the clock drawing task in Alzheimer's disease. J Am Geriatr Soc 1992; 40: 1095–1099.

33.

Hirjak

Wolf

Pfeifer

, et al. Cortical signature of clock drawing performance in Alzheimer's disease and mild cognitive impairment. J Psychiatr Res 2017; 90: 133–142.

34.

Talwar

Churchill

Hird

, et al. The neural correlates of the Clock-Drawing Test in healthy aging. Front Hum Neurosci 2019; 13: 25.

35.

Huang

Pan

Huang

, et al. The value of clock drawing process assessment in screening for mild cognitive impairment and Alzheimer's dementia. Assessment 2023; 30: 364–374.

36.

Lamar

Ajilore

Leow

, et al. Cognitive and connectome properties detectable through individual differences in graphomotor organization. Neuropsychologia 2016; 85: 301–309.

37.

Nakashima

Umegaki

Makino

, et al. Neuroanatomical correlates of error types on the Clock Drawing Test in Alzheimer's disease patients. Geriatr Gerontol Int 2016; 16: 777–784.

38.

Ismail

Rajji

Shulman

. Brief cognitive screening instruments: an update. Int J Geriatr Psychiatry 2010; 25: 111–120.