Short Test of Mental Status (STMS)

Abstract

The Short Test of Mental Status (STMS) is a brief, clinician-administered bedside neurocognitive screening instrument originally developed at the Mayo Clinic by Kokmen and colleagues in 1987. Designed to overcome the psychometric and clinical limitations of earlier cognitive screening batteries—such as the Mini-Mental State Examination (MMSE)—the STMS specifically targets cognitive domains vulnerable to early neurodegenerative pathology, most notably Alzheimer’s disease and mild cognitive impairment (MCI). The instrument comprises eight discrete subtests assessing eight cognitive dimensions across nine operational task categories: Orientation (8 points), Attention/Working Memory (7 points), Immediate Recall/Learning (4 points with an acquisition penalty), Calculation (4 points), Abstraction (3 points), Construction (2 points), Visuospatial Copying (2 points), General Information/Fund of Knowledge (4 points), and Delayed Free Recall (4 points). The maximum obtainable raw score is 38 points. The STMS utilizes an objective, criterion-referenced performance format administered in approximately 5 to 10 minutes, requiring minimal equipment (paper and pencil). Psychometric investigations demonstrate high internal consistency (Cronbach’s alpha ranging from .82 to .89), excellent test-retest stability (r = .85 to .92 across clinical intervals), robust inter-rater reliability (intraclass correlation coefficients exceeding .90), and strong convergent validity with extensive neuropsychological batteries including the Wechsler Adult Intelligence Scale and the Mattis Dementia Rating Scale. Clinically, the STMS has demonstrated superior sensitivity over the MMSE in discriminating amnestic mild cognitive impairment from normal cognitive aging, largely attributable to its more demanding delayed recall paradigm, multi-trial learning assessment, and rigorous mental abstraction items.

Keywords

Short Test of Mental Status, STMS, cognitive screening, dementia assessment, Alzheimer’s disease, mild cognitive impairment, Kokmen, delayed recall, bedside neurocognitive test, neuropsychological assessment

Authors

The Short Test of Mental Status was conceptualized, developed, and validated by a distinguished multidisciplinary team of behavioral neurologists, biostatisticians, and clinical neuropsychologists at the Mayo Clinic in Rochester, Minnesota, USA:

• Emre Kokmen, M.D. (1938–1999): Professor of Neurology, Mayo Medical School and Mayo Clinic. A preeminent neuroepidemiologist and behavioral neurologist whose research established foundational epidemiological benchmarks for dementia and primary neurodegenerative diseases in Olmsted County, Minnesota.
• James M. Naessens, Sc.D.: Health Services Researcher and Biostatistician, Division of Health Care Policy and Research, Mayo Clinic, Rochester, MN.
• Kenneth P. Offord, M.S.: Senior Biostatistician, Department of Health Sciences Research, Mayo Clinic, Rochester, MN.
• Glenn E. Smith, Ph.D., ABPP-CN: Clinical Neuropsychologist, Department of Psychiatry and Psychology, Mayo Clinic (later at the University of Florida). Past President of the Society for Clinical Neuropsychology (APA Division 40).
• Ronald C. Petersen, M.D., Ph.D.: Professor of Neurology, Cora Kanow Professor of Alzheimer’s Disease Research, and Director of the Mayo Clinic Alzheimer’s Disease Research Center. Internationally acclaimed for defining the clinical criteria for Mild Cognitive Impairment.
• Eric G. Tangalos, M.D.: Professor of Medicine, Division of Community Internal Medicine and Geriatrics, Mayo Clinic, Rochester, MN.
• Robert J. Ivnik, Ph.D., ABPP-CN (1947–2014): Professor of Psychology, Chair of the Section of Psychology, Mayo Clinic; pioneer in developing age-stratified normative data for standardized neuropsychological testing (Mayo’s Older Americans Normative Studies [MOANS]).

Institutional contact and historical archival queries may be directed to the Department of Neurology, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905, USA.

Purpose

The primary clinical and scientific purpose of the Short Test of Mental Status is to provide a rapid, standardized, psychometrically rigorous quantitative assessment of general cognitive functioning capable of detecting subtle cerebral dysfunction. In the late 1970s and 1980s, existing cognitive screening instruments—most prominently Folstein’s MMSE and the Kahn-Goldfarb Mental Status Questionnaire—were criticized for significant ceiling effects in well-educated older adults, pronounced insensitivity to mild memory impairments, absent assessments of abstract reasoning, and over-reliance on basic orientation and superficial language cues. Kokmen and colleagues specifically constructed the STMS to serve as a robust bedside tool suitable for epidemiological investigations, acute inpatient consultations, ambulatory geriatric clinics, and routine primary care environments.

The rationale underpinning the STMS is rooted in the early clinicopathological manifestations of neurodegenerative tauopathies and amyloid-beta proteopathies. In pre-dementia states and early-stage Alzheimer’s disease, synaptic disruption and neurofibrillary tangles originate within the transentorhinal and entorhinal cortices before progressing to the hippocampus and neocortical association areas. Consequently, the earliest cognitive signs do not typically involve gross disorientation or severe language dissolution, but rather impaired encoding efficiency, rapid forgetting across an unfilled distraction interval, and subtle deficits in executive conceptual tracking. The STMS addresses these vulnerabilities directly by:

• Evaluating Learning Efficiency: Requiring up to four learning trials to acquire four unrelated words, penalizing poor acquisition efficiency by subtracting excess learning trials from the total score.
• Preventing Rehearsal Artifacts: Placing demanding cognitive interference tasks (working memory, calculation, abstraction, and construction) between the learning phase and the delayed free recall assessment.
• Challenging Executive and Frontoparietal Networks: Incorporating two-step mental arithmetic, semantic abstraction (similarities), clock drawing, and three-dimensional Necker cube copying.
• Evaluating Remote Semantic Reserves: Assessing fund of information (e.g., historical knowledge, geographic conceptualization, and numerical calendars).

In clinical practice, the STMS aids in differentiating normal cognitive aging from syndromic mild cognitive impairment, establishing baseline cognitive metrics prior to neurotoxic or neurosurgical interventions, monitoring longitudinal cognitive progression in neurodegenerative disorders, and screening patients with cerebrovascular disease, normal pressure hydrocephalus, or toxic-metabolic encephalopathies.

Psychological Construct

The Short Test of Mental Status operationalizes global cognitive capacity through eight interconnected neuropsychological domains, structured as follows:

1. Temporal and Spatial Orientation (8 Points)

Orientation assesses autobiographical, spatial, and temporal reality testing, reflecting basic awareness of self in relation to the environment. The STMS probes the subject’s full name, current street address, current facility/building name, city, state, specific calendar date (day of month), current month, and current year. Neuroanatomically, orientation relies on distributed networks linking medial temporal episodic memory stores with prefrontal monitoring systems and parietal environmental representations. Disorientation generally signifies more advanced bilateral hemispheric disruption or acute confusional states (delirium).

2. Attention and Auditory Working Memory (7 Points)

Attention is evaluated via an auditory forward digit span task, presented at a calibrated pace of one digit per second. Digits are administered starting at five items and increasing to seven items (e.g., 2-9-6-8-3; 5-7-1-9-4-6; 2-1-5-9-3-6-2). This subtest measures auditory registration, concentration, and the capacity of the phonological loop within Baddeley’s model of working memory. Effective performance demands intact dorsolateral prefrontal cortical functioning and preserved ascending reticular activating system tone.

3. Immediate Recall and Encoding Efficiency (4 Points Minus Acquisition Penalties)

Subjects are presented with four phonetically and semantically distinct words: “apple,” “Mr. Johnson,” “charity,” “tunnel.” These stimuli deliberately represent diverse linguistic categories: a tangible object, a proper name, an abstract social concept, and an architectural structure. The subtest records not only whether the patient can register the four words, but critically the exact number of trials (up to four) required to achieve complete registration. An acquisition penalty—computed as (number of trials − 1)—is directly subtracted from the final composite score. This unique psychometric feature captures inefficient encoding and decreased learning slope, which are classic hallmark indicators of medial temporal lobe dysfunction.

4. Mental Calculation (4 Points)

The calculation subtest presents four sequential mathematical operations executed entirely via internal mental manipulation without paper or pencil: 5 × 13, 65 − 7, 58 / 2, and 29 + 11. These calculations require multi-step mental tracking, operational sequencing, and the retrieval of automated arithmetic facts from semantic memory. Performance engages the left angular gyrus, the intraparietal sulcus, and frontal executive networks supporting working memory manipulation.

5. Abstract Verbal Reasoning (3 Points)

Semantic conceptualization and abstraction are assessed using three categorical word-pair similarities: orange/banana, dog/horse, and table/bookcase. Full credit (1 point each) requires the induction of a true superordinate categorical identity (e.g., “fruit,” “animals/mammals/quadrupeds,” “furniture”). Concrete functional responses (e.g., “you eat them,” “both have legs,” “you put books on them”) or perceptual associations earn 0 points. This subtest probes frontal lobe executive capacity, specifically the ability to inhibit salient perceptual features to extract categorical commonalities.

6. Construction and Visuospatial Praxis (4 Points Total: 2 for Clock, 2 for Cube)

Constructional abilities are evaluated through two complementary spatial tasks:

• Clock Face Drawing (2 Points): The subject is instructed to draw a complete clock face showing the time 11:15. This demands visuospatial planning, hemispatial awareness, selective attention, and symbolic time translation (placing the minute hand on the 3 rather than the 15).
• Necker Cube Copying (2 Points): The subject copies a three-dimensional line drawing of a Necker cube. This task taps complex visuoperceptual processing, spatial orientation, perspective synthesis, and motor coordination, engaging bilateral posterior parietal cortices and superior occipital-parietal pathways.

7. Remote Semantic Information / Fund of Knowledge (4 Points)

This subtest samples crystallized semantic memory through four culturally stable probes: current President, first President, defining an island, and stating the number of weeks in a calendar year. These items assess semantic knowledge stores consolidated in lateral temporal neocortices, resistant to transient attentional blips but sensitive to advanced cortical degeneration.

8. Delayed Free Recall (4 Points)

After an unfilled distraction interval of approximately five minutes spent completing calculations, abstractions, constructions, and information questions, the patient is asked to freely recall the four words learned earlier (apple, Mr. Johnson, charity, tunnel). No category cues, phonemic prompts, or recognition arrays are provided during standardized scoring. Delayed free recall is the single most sensitive cognitive marker for hippocampal and entorhinal cellular loss in amnestic mild cognitive impairment and early Alzheimer’s disease.

Theoretical Framework

The Short Test of Mental Status is grounded in clinical neuropsychology and cognitive neuroscience paradigms of brain-behavior relationships. Kokmen and colleagues integrated three theoretical pillars when constructing the battery:

1. The Information-Processing Architecture of Memory

The STMS operationalizes the structural distinction between immediate registration, working memory buffering, and long-term consolidation as articulated by Atkinson and Shiffrin, and refined by Baddeley and Warrington. Standard screening tools often conflate registration with secondary memory by testing recall within seconds. The STMS explicitly separates these constructs: forward digit span isolates the phonological loop; multi-trial presentation quantifies initial registration efficiency; and deferred testing following extensive cognitive interference evaluates secondary episodic memory consolidation. Patients with true hippocampal amnesia display preserved registration and normal digit span alongside profound failure on delayed free recall.

2. Lurian Functional Systems and Cortical-Subcortical Networks

Alexander Luria proposed that complex mental activity is executed through dynamic functional systems organized across three anatomical units: (a) the unit regulating cortical tone and vigilance (brainstem/reticular formations), tapped by the Attention subtest; (b) the unit for obtaining, processing, and storing sensory information (occipital, parietal, and temporal cortices), probed by Orientation, Calculation, Visuospatial Construction, and Delayed Recall; and (c) the unit for programming, regulating, and verifying mental activity (frontal lobes), probed by Abstract Verbal Reasoning and the Clock Drawing task. By sampling behaviors across all three functional units, the STMS facilitates the detection of focal cortical lesions, subcortical ischemic processes, and diffuse degenerative disorders.

3. The Concept of Cognitive Reserve and Graded Task Difficulty

Traditional bedside tools suffer from ceiling effects among individuals with high premorbid intelligence or advanced formal education. Kokmen et al. countered this by introducing items with elevated cognitive load. Calculating 5 × 13 or mentally dividing 58 / 2 cannot be executed via passive pattern recognition; it demands active working memory registers. Similarly, scoring word pairs strictly for categorical superordination eliminates the superficial passes permitted by less rigorous screening tools.

Validity

The validity of the Short Test of Mental Status has been established through extensive psychometric, neuroimaging, and longitudinal clinicopathological investigations over four decades.

Construct and Convergent Validity

In their seminal validation study, Kokmen, Smith, Petersen, Tangalos, and Ivnik (1991) examined the correspondence between the STMS and an exhaustive battery of formal psychometric instruments in a sample of elderly individuals across diverse cognitive strata. STMS total scores correlated robustly with established gold-standard measures:

• Wechsler Adult Intelligence Scale-Revised (WAIS-R) Full Scale IQ: r = .78 (p < .001)
• WAIS-R Verbal IQ: r = .80 (p < .001)
• WAIS-R Performance IQ: r = .71 (p < .001)
• Mattis Dementia Rating Scale (DRS): r = .84 (p < .001)
• Auditory Verbal Learning Test (AVLT) Delayed Recall: r = .75 (p < .001)

The STMS demonstrates exceptional convergent validity with the Mini-Mental State Examination (correlations typically spanning r = .82 to .91), yet consistently demonstrates superior measurement range without the pronounced ceiling effects that hamper the MMSE.

Discriminant and Criterion Validity

In a landmark comparative investigation, Tang-Wai, Knopman, Geda, et al. (2003) evaluated the discriminative efficacy of the STMS versus the MMSE in distinguishing patients with amnestic mild cognitive impairment (aMCI) and mild Alzheimer’s disease from healthy age-matched control participants:

• Detection of Amnestic MCI: Using a validated cutoff score of ≤ 32 out of 38, the STMS achieved a sensitivity of 86% and a specificity of 91% (Area Under the ROC Curve [AUC] = 0.92). By comparison, the MMSE (cutoff ≤ 27/30) exhibited a sensitivity of only 58% and a specificity of 85% (AUC = 0.81). The superior performance of the STMS was driven primarily by its delayed recall subtest and the immediate recall acquisition penalty.
• Detection of Mild Dementia: For diagnosing established mild Alzheimer’s disease, an STMS cutoff score of ≤ 29 yielded a sensitivity of 95% and a specificity of 97% (AUC = 0.98).

Subsequent multi-center studies confirmed the capacity of the STMS to differentiate subcortical vascular cognitive impairment, dementia with Lewy bodies, and behavioral variant frontotemporal dementia based on distinct subtest performance profiles.

Reliability

Empirical assessments demonstrate that the Short Test of Mental Status possesses high internal consistency, inter-rater reliability, and stability over time:

Internal Consistency

Across validation cohorts comprising both cognitively unimpaired older adults and clinical dementia populations, the overall internal consistency of the STMS is robust:

• Cronbach’s Alpha: Values consistently range between .82 and .89 for the global scale.
• Item-Total Correlations: Subtest-to-total correlations are highest for Delayed Recall (r = .72), Abstraction (r = .65), and Calculation (r = .61), confirming that the instrument taps a coherent underlying construct of cognitive integrity while sampling non-redundant neurocognitive faculties.

Test-Retest Stability

In clinically stable older adults evaluated over intervals ranging from 1 to 4 weeks, the STMS showed high test-retest reliability:

• Pearson correlation coefficients (r) range from .85 to .92.
• Practice effects across brief re-testing intervals are modest, although alternate word lists are recommended when evaluating patients at high frequency to prevent rote recall of the four test words.

Inter-Rater and Scorer Reliability

Because the STMS provides structured, objective criteria—particularly for the semi-qualitative construction and abstraction tasks—inter-rater reliability is exceptionally high:

• Intraclass Correlation Coefficients (ICC): Total score ICC values between independent clinicians range from .91 to .96.
• Cohen’s Kappa (κ): Individual item scoring concordance between trained physicians, nurses, and psychometrists exceeds κ = .84, including on the clock-drawing and cube-copying subtests.

Factor Analysis

Structural psychometric investigations using Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) confirm that while the STMS measures a unified global cognitive construct (supporting the use of a single composite score), it is underpinned by a distinct multidimensional factor architecture reflecting core neurofunctional systems.

Exploratory Factor Analysis

Principal component analyses with Varimax and Promax rotations consistently yield a three-factor solution accounting for approximately 62% to 68% of the total variance across clinical cohorts:

• Factor 1: Amnestic / Medial Temporal Factor: Accounting for ~34% of variance. Dominated by high factor loadings from Delayed Recall (.84), Immediate Recall acquisition efficiency (.79), and Temporal Orientation (.62).
• Factor 2: Executive / Frontoparietal Factor: Accounting for ~19% of variance. Defined by strong loadings from Mental Calculation (.76), Abstract Reasoning (.71), Attention/Digit Span (.58), and Clock Drawing (.52).
• Factor 3: Visuospatial / Semantic Knowledge Factor: Accounting for ~12% of variance. Characterized by high loadings from Cube Copying (.81), Clock Face Construction (.64), and General Information (.59).

Confirmatory Factor Analysis

Structural equation modeling comparing competing models demonstrates that a hierarchical (second-order) factor structure—in which a broad “General Cognitive Ability” (g) factor governs three correlated lower-order factors (Memory, Executive/Working Memory, Visuospatial/Semantic)—provides superior fit to empirical data compared to a strict unidimensional model:

• Comparative Fit Index (CFI): .962
• Tucker-Lewis Index (TLI): .951
• Root Mean Square Error of Approximation (RMSEA): .048 (90% CI: .038–.059)
• Standardized Root Mean Square Residual (SRMR): .041

These fit indices satisfy modern psychometric standards for structural validity, confirming that the STMS functions effectively as a composite metric of cognitive status.

Instrument / Measurement Tool

The Short Test of Mental Status is structured as follows:

• Test Format: Standardized, examiner-administered performance battery.
• Administration Time: Approximately 5 to 10 minutes.
• Target Population: Adults aged 18 and older, with primary clinical utility in geriatric populations, individuals presenting with cognitive complaints, and patients undergoing neurological workups.
• Materials Required: Testing protocol sheet, a pencil or pen, and a blank sheet of paper for the patient’s construction tasks.
• Subscale Breakdown & Maximum Scores:
  • 1. Orientation: Maximum 8 points (1 point per item: name, address, current building, city, state, date, month, year).
  • 2. Attention: Maximum 7 points (highest forward digit span accurately repeated from 5, 6, or 7 digits).
  • 3. Immediate Recall: Maximum 4 points (1 point per word learned across up to 4 trials).
  • 4. Calculation: Maximum 4 points (1 point per correct mental calculation).
  • 5. Abstraction: Maximum 3 points (1 point per correct abstract category for word pairs).
  • 6. Construction (Clock Face): Maximum 2 points (2 = adequate drawing showing 11:15; 1 = partially correct/less complete; 0 = severe impairment or unable).
  • 7. Copy (Necker Cube): Maximum 2 points (2 = correct 3D perspective and configuration; 1 = partially correct; 0 = severely distorted or unable).
  • 8. Information: Maximum 4 points (1 point per correct remote semantic probe).
  • 9. Recall: Maximum 4 points (1 point per word freely recalled without cueing at the end of testing).
• Scoring Formula & Rules:
  • Immediate Recall Acquisition Penalty: Subtract (Number of Trials Required to Learn All 4 Words − 1) from the running score. If the patient learns all words on Trial 1, penalty = 0. If 4 trials are needed, penalty = 3.
  • Total Score Calculation: Total Score = (Sum of Subtest Scores 1 through 9) − (Number of acquisition trials − 1).
  • Maximum Score: 38 points.
• Clinical Cutoff Benchmarks:
  • 33 to 38: Cognitively normal / unimpaired.
  • 30 to 32: Borderline / Suggestive of Mild Cognitive Impairment (aMCI).
  • ≤ 29: Significant cognitive impairment / Consistent with Dementia.

Permissions & Fee and Test Year

The Short Test of Mental Status was first published in 1987 in the Mayo Clinic Proceedings by Dr. Emre Kokmen and associates. The instrument was developed within the Mayo Clinic’s medical and scientific research infrastructure. As an academic screening battery, the STMS has historically been made widely accessible for clinical practice, institutional bedside assessments, and non-commercial academic research without licensing fees, provided proper bibliographic attribution is given. However, healthcare systems and commercial entities seeking to incorporate the instrument into proprietary digital health records, software applications, or funded clinical trials should consult the Mayo Clinic Division of Technology Transfer and Licensing regarding intellectual property parameters and authorized use.

References

• Kokmen, E., Naessens, J. M., & Offord, K. P. (1987). A short test of mental status: Description and preliminary results. Mayo Clinic Proceedings, 62(4), 281–288. https://doi.org/10.1016/S0025-6196(12)61905-4
• Kokmen, E., Smith, G. E., Petersen, R. C., Tangalos, E. G., & Ivnik, R. J. (1991). The Short Test of Mental Status: Correlations with standardized psychometric testing. Archives of Neurology, 48(7), 725–728. https://doi.org/10.1001/archneur.1991.00530190071018
• Tang-Wai, D. F., Knopman, D. S., Geda, Y. E., Edland, S. D., Smith, G. E., Ivnik, R. J., Tangalos, E. G., Boeve, B. F., & Petersen, R. C. (2003). Comparison of the Short Test of Mental Status and the Mini-Mental State Examination in mild cognitive impairment. Archives of Neurology, 60(12), 1777–1781. https://doi.org/10.1001/archneur.60.12.1777
• Smith, G. E., Tangalos, E. G., & Petersen, R. C. (1994). Cognitive screening in the elderly: A comparison of the Short Test of Mental Status and the Mini-Mental State Examination. Neurology, 44(Suppl 2), A175.
• Petersen, R. C., Smith, G. E., Waring, S. C., Ivnik, R. J., Tangalos, E. G., & Kokmen, E. (1999). Mild cognitive impairment: Clinical characterization and outcome. Archives of Neurology, 56(3), 303–308. https://doi.org/10.1001/archneur.56.3.303

Items of the Scale