Clinical aspects

Description

The different subtests of cognitive function are usually grouped into composite cognitive domains reflecting different aspects of brain function. Some aspects of cognitive function such as insight into symptoms are generally not covered by tests directly, but are dealt with in a less structured manner (47).

Mental status screening should include an assessment of cognitive and affective states and use standardized instruments. All mental status examinations should be interpreted within the context of the patients background to minimize the risk of overdiagnosing deficits in patients whose premorbid status was itself deficient.

Following are explanations of cognitive functions generally tested in a mental status examination:

Attention. Attention refers to a persons ability to maintain, shift, and focus awareness to a specific stimulus without being distracted by extraneous stimuli. The ability to focus on a single stimulus is in contrast to the concept of alertness, which is a more basic arousal process that allows the awake person to respond to any stimulus in the environment. Vigilance refers to the ability to sustain attention over an extended period (ie, to concentrate). The capacity to concentrate is important for performing intellectual endeavors and may be impaired in both organic and emotional disorders. Reduced speed during testing is often relevant for test outcome. Although many neuropsychological tests assess this change, it is not addressed in the MMSE. Some principal tests of attention include:

The forward digit span test. This test determines the patients ability to focus attention for short periods. The test is performed by reading lists of random single digits at 1 digit per second and asking the patient to repeat each list after the examiner stops reciting. A normal forward digit span is at least 5 numbers long, and most normal persons can repeat lists of 7 or 8 numbers.

Serial 7 subtraction test. The patient counts backward from 100 by 7s: 100, 93, 86, 79, 72, 65, etc.

Orientation to person, place, and time. This last test has a strong memory component as well.

Language. Language must be evaluated early in the course of the examination because some patients with aphasia cannot perform validly on many tests used to assess verbal memory, abstract reasoning, and calculations.

Spontaneous speech. The physician should listen carefully to the patients spontaneous speech and ask open-ended questions. Especially note articulation, fluency, prosody, syntax, and paraphasia.

Comprehension. (1) Pointing commands. The patient is asked to point to room objects, with an increasing span of objects per test item (eg, "Point to the ceiling and the door"). The average nonaphasic patient should accurately point to 4 objects or more in correct sequence. (2) Yes or no commands. The patient is asked a minimum of 7 questions that require "yes" or "no" response to minimize the possibility of chance.

Repetition. The patient should repeat meaningful sentences of increasing length and complexity with increasing difficulty until failure is consistent. Listen for grammar, omission, addition, perseveration, and errors of naming. Neurologically healthy individuals can correctly repeat 20 syllables.

Naming. The ability to name objects is a basic language function. It is frequently disturbed in subjects with brain damage and is almost invariably impaired in aphasia. Anomia is best tested by visual confrontation naming. Point to a range of objects, and ask the patient to name them. A variety of categories should be used because some aphasic patients demonstrate an inability to name items in a specific category while retaining the ability to name other types of objects. Shift between words occurs with high or low frequency in the persons preferred language. The average person should make no paraphasic error (ie, producing a word that would be anomalous in the persons usual language) and show almost no word-finding pauses.

Reading and writing. Ask the patient to read sentences, spell words from dictation, and compose and write a sentence.

Visuospatial skills. Visuospatial skills can be assessed by asking the patient to copy or draw figures.

Reproduction of drawings. The patient is asked to make pen and paper reproductions of a series of increasingly complex geometric shapes with the stimulus figures in view (eg, a vertical diamond, a 2-dimensional cross, and a 3-dimensional box).

Drawing to command. The patient is requested to draw 3 simple drawings on command (eg, clock with numbers, flower, and 3-dimensional house). In demented patients, the drawings on command are more sensitive than reproduction drawings.

Memory. Memory is subdivided into 3 basic types, (1) immediate, (2) recent, and (3) remote, based on the time span between stimulus presentation and memory retrieval.

Immediate memory. This refers to the recall of a memory trace after an interval of a few seconds. Immediate memory may be tested by digit representation and sentence repetition.

Recent memory. This is the capacity to remember day-to-day events (eg, orientation to person, place, and time, or recent news events). More strictly defined, recent memory is the ability to learn new material and to retrieve that material after an interval of minutes, hours, or days.

Remote memory. This refers to the recall of old, stored memories over a span of years to decades (eg, place of birth, family information, school information, famous faces and 5 presidents during the patients lifetime).

New learning ability. This refers to the ability to process and recall new information.

(1) Recall of 4 randomly chosen words, immediately, and at 5-, 10-, and 30-minute intervals after presentation. If words are not retrieved spontaneously, semantic, phonemic, and recognition cueing is used to aid recall. Normals should recall all items at each interval, whereas low-IQ normals recall a mean of 3 or more, and brain damaged patients recall about 1 word per trial.

(2) Standard verbal story for immediate recall. The average person should recall about 50% of the information from each story (43).

Paired associate learning. Word pairs with strong natural associations and those with no natural associations are used. Two trials of presentation followed by recall of the following word pairs are administered. Examples of test items are "Weather" - "Bag;" "High" - "Low;" "House" - "Income;" and "Book" - "Page." The normal patient should recall all "easy" (naturally associated) pairs and at least 1 of the "difficult" pairs.

Executive functions. Executive function refers to the highest level of intellectual development. It accommodates the manipulation of previously learned knowledge, abstract reasoning, and problem solving and is closely related to intelligence, education, and social exposure.

Definition of similarities and differences. The patient should define similarities or differences among items. For example, the patient is asked to identify the class to which 2 items belong: hat and coat (clothes), rose and tulip (flowers), bicycle and train (means of transportation). The patient should also identify differences between superficially similar items: river and canal (one is man-made); lie and mistake (one is intentional). Patients with limited abstraction abilities will fail to see the appropriate similarities or differences or will respond to concrete, rather than abstract aspects of the problems (eg, a bicycle and a train both have wheels).

Interpretation of proverbs. The patient should interpret sayings such as "Dont cry over spilled milk" and "Rome was not built in a day."

MMSE. The best-known mental status examination test battery is the MMSE, developed by Folstein and colleagues (16). This test is the most widely used instrument to rapidly assess the cognitive status of individuals, both in clinical and research settings. This examination may be supplemented by the checklist suggested by Geldmacher and Whitehouse (17) or supported by the Cognitive Mental Status Task (45). The MMSE may be administered at the bedside or in the office. It requires that the patient has access to a hard surface when writing and drawing.

The MMSE comprises 11 questions assessing orientation to time and place, attention, immediate and short-term recall, language, and visuospatial abilities. The score is a weighted sum of the items. The maximum score is 30 points. Some disagreement exists with respect to what constitutes an abnormal score. Most authors suggest that patients scoring below 24 are cognitively impaired (32). Some accept lower scores as normal, and a few find that 18 is the limit below which definite cognitive impairment is present (03). It should be noted that age and education strongly influence test scores, which makes it difficult to interpret a low score obtained by an elderly patient with modest intelligence or limited education (02). It is possible for intelligent, well-educated persons who are cognitively impaired to perform a cognitive test within the normal range because their premorbid performance was so high that a modest reduction is not picked up by a gross test.

One of the questions in the MMSE is the Serial 7 subtraction test, which has been claimed to be biased against people with a lower level of education. Therefore, this question may be substituted for spelling the word WORLD backward. This, however, tends to be an easier task, and studies have revealed modest correlation between these measures (39).

One of the advantages of the MMSE is that unlike many other tests, it can be administered to patients across a broad spectrum of dementia severity. It is fast and easy to administer, and gives reproducible results. When administered repeatedly, it shows a small, but significant retest effect over short intervals (10). In spite of this short-term retest effect, the MMSE has been widely used in longitudinal studies to follow the course of dementia (ie, in Alzheimer patients on choline esterase inhibitors). When administered to healthy older men (mean age of 71 years) over a 6-year period, 80% remained within 2 points of their initial score. An improvement of 3 or more points was seen in 5% of the subjects, whereas a decline of 3 or more points was seen in 15% of the subjects (13). Decliners were less active at follow-up, rated their health more poorly, and reported more depressive symptoms than non-decliners. A study of the effect of physical and mental activity found that MMSE scores were less affected over time in patients with dementia when they were physically or mentally active than in matched controls (05). In a review, the MMSE was described to be a useful measure of disease status in persons with Alzheimer disease performing aerobic exercise (28).

One of the problems with the MMSE is that it only gives a single score, which is used to identify the presence and degree of brain disease; thus, it is only a gross estimate of the patients global cognitive functioning (44). It often fails to detect focal brain injury and resulting specific cognitive deficits (15). Right-hemisphere lesions are especially overlooked by the MMSE. It is important to be aware of the chemo-brain phenomenon of patients with cognitive impairment after chemotherapy. The MMSE is not sensitive enough for this condition (25).

The MMSE has been translated into many languages. Translation, cross-cultural comparison and validation of neuropsychological tests are central issues. In a study of older Mexican Americans and European Americans, it was found that the MMSE, regardless of language, appears to be a valid indicator of cognitive impairment (14). Differential item functioning refers to a more rigorous description of the performance of different population subgroups based on race, gender, educational level, first language, etc. When differential item functioning is applied to the MMSE, translation of the test turns out to be a critical factor (37). One way of dealing with this problem is back-translation to address semantic equivalence.

Extended versions of the MMSE exist, but have only gained limited use, perhaps because they lack the user-friendly administration format of the MMSE. One of the best known modifications is the "Modified Mini-Mental State Examination" (46). It measures orientation registration, attention, recall, and language similar to the MMSE, supplemented by testing for abstraction and delayed recall. The Modified Mini-Mental State Examination is thought to be more sensitive to mild cognitive defects, especially in the right-hemisphere (19). The enhanced sensitivity is obtained at the cost of increased test complexity.

The Montreal Cognitive Assessment (MoCA) is a bedside test also thought to be more sensitive to mild cognitive impairment (31). This test also has a maximum score of 30 based on visuospatial performance, naming, memory, attention, language, abstraction, and orientation and is generally thought to reflect right hemisphere function and abstraction more than the MMSE.

The MoCA takes approximately twice as long to administer as the MMSE, but because not all test elements seem to be equally important, there is potential for an abbreviated MoCA with high sensitivity combined with fast administration for patients with mild cognitive impairment (07).

Although the MMSE has a long history of validation and is widely used, the MoCA is gaining interest. Therefore, a conversion between these 2 scales has been examined (01). Based on double assessment of a large number of patients from 3 different memory clinics, Bergeron and colleagues conclude that converting from the MMSE to the MoCA should be done with caution, whereas converting from the MoCA to the MMSE seems more robust. An unresolved concern is the heterogeneity of patients with various dementia types and the degree of disease progression for the comparative validation of these tests.

A metaanalysis of accuracy for general cognitive screening tests concluded that there was no clear superiority when comparing the MoCA, MMSE, and others (26).

When focusing on specific cognitive domains in stroke, Demeyere and colleagues found that the Oxford Cognitive Screen (OCS) was overall more sensitive than MoCA and detected important cognitive deficits after stroke not assessed in the MoCA (08).

For patients with predominantly severe cognitive defects, other modifications have been made, for instance the Severe Mini-Mental State Examination, which has been found to be useful for assessing severely impaired Alzheimer disease patients (20). In spite of problems and limitations of the MMSE, this test remains a robust and easily administered test, suitable for first-line clinical practice.

The ACE-III takes longer to administer than the MMSE or the MoCA. The questionnaire covers areas of memory, language, word fluency, attention, and visuospatial skills. The highest score is 100, with higher scores indicating better memory and general cognitive function.

The ACE-III has good sensitivity and specificity in the assessment of cognitive deficits in people with Alzheimer disease and frontotemporal dementia (21). It seems to be more sensitive to impairment of everyday activity than the MMSE or MoCA (18).

New screening tests appear frequently, focusing on various problems in the older tests. The HKBC (Hong Kong Brief Cognitive Test) has been developed to provide a simple test to subjects with low educational level (06). It has been validated in Hong Kong on 359 subjects with neurocognitive disorder or normal control. There was a high overlap between the HKBC and MMSE as well as MoCA. The validity at low educational level is still under investigation.

Thus, a number of more or less validated mental status examination tools exist. They vary in ease and speed of administration, sensitivity and specificity on disease progression, and diagnostics as well as relative focus and various cognitive domains. It is worth noting that the most extensive examinations tend to give the most detailed information.

The goal of mental status examination is to identify brain disease and, if possible, point to the likely site of brain dysfunction. A comprehensive mental status examination includes assessment of attention, language, visuospatial skills, memory, and executive functions. It is important to consider the socioeconomic background, age, and sex of the patient because these factors all affect the test results (10).

The choice of cognitive test depends, to a large extent, on the status of the patient. In a study of driving impairment, the visuomotor clock drawing test turned out to better predict driving ability than did the MMSE (33). The combination of these tests has been shown to be more sensitive but at the cost of increased complexity (41). Subjective cognitive complaints may be important in diagnosis among younger patients with dementia (40). The MMSE is still central in many clinical trials; however, this is often combined with other screening tests aimed at specific cognitive issues, for example, the large cohort study of Pfistermeister and colleagues (35).

Testing of the different functions must be systematically ordered because the outcome of higher-level function depends on the basic functions. The testing should start with the most basic functions (ie, the level of consciousness, attention, and vigilance), and then move on to the higher-level functions (ie, abstract reasoning and special cognitive functions) (44). In neuropsychological testing, the Wechsler Intelligence Scales are frequently used to measure IQ. They provide information about the overall level of intellectual ability, demonstrating presence or absence of significant intellectual disability and providing clues to altered functions (27). When adults generate few responses on the Wechsler Adult Intelligence Scale, the test result is uncertain, and other tests may be preferred to grossly estimate cognitive function.

Indications

Physical examination always involves some degree of mental status evaluation. The type and extent of evaluation depends on the problems under consideration and the state of the patient.

A demented patient presenting with clinical signs requires, as a minimum, mental status screening in the form of bedside tests, and testing with a large-scale neuropsychological test battery is often necessary. Mental status screening may also prove useful among depressed patients. A final group of patients for whom mental status screening is important consists of those who initially present vague behavioral complaints that are difficult to clinically quantify in the standard neurologic examination. Complaints such as memory problems, difficulties in concentration, declining interest in family or work, or various physical complaints without identified structural cerebral etiology should alert the clinician to the possibility of structural brain disease (44).

Contraindications

Acute confusional states and coma are common problems in general medicine. It is estimated that over 5% of admissions to the emergency department of large municipal hospitals are due to diseases that impair consciousness (38). In coma and coma-like syndromes, such as severe acute Guillain-Barré syndrome or catatonia, the use of specific mental status examinations is seldom possible. Estimates of conscious abilities are based on observations of spontaneous movements, response to stimuli, and other aspects of the general neurologic assessment.

In confusional states, which are characterized by reduced mental clarity, coherence, comprehension and reasoning, the main early signs are inattention and disorientation. Subsequently, deterioration in memory, comprehension, problem solving, visual-spatial function, language, and perception may depend on the underlying cause and the localization of cerebral involvement. The following conditions are the most common causes of confusional states: dementia, hypercapnia, hypoxia, ischemia, hypoglycemia, electrolyte derangement, ketoacidosis and other osmolarity abnormalities, renal failure, hepatic failure, trauma, tumors, hydrocephalus and other mechanical effects, hypothermia, drug or alcohol intoxication, endogenous metabolites, anesthesia, epilepsy, and infections. When the cause is immediately identifiable and treatable, the acute treatment should gain priority over elaborate mental status examinations; however, a few moments spent evaluating the mental state of a confused patient during the first contact may be well worth the trouble.

Adverse effects

This section will describe the utility of mental examination with identifying progressive illness. In a study of 51 Alzheimer patients who received the acetylcholinesterase inhibitor donepezil hydrochloride, Shimizu and colleagues found that decline in MMSE score in the course of the disease was associated with decrease in regional cerebral blood flow as evaluated by SPECT (42).

The Alzheimers Disease Neuroimaging initiative has selected 523 individuals with various stages of preclinical or manifest Alzheimer disease for MRI scanning (09). They found an association between gray matter atrophy and MMSE score.

A number of studies have looked at the relation between systemic disease and mental state using the MMSE. Often the cognitive changes looked for are relatively small, and the result may be nonsignificant because the MMSE is not sensitive enough. Molander and coworkers found in a group of very old people (85 years of age and older) that those with low systolic blood pressure (120 mmHg) had lower scores on the MMSE than those with systolic blood pressure higher than 140 mmHg (30). Hypothermia is widely used in patients after cardiac arrest to protect cerebral function. The MMSE score was evaluated along with several other variables to evaluate the overall outcome for these patients. No effect could be measured on MMSE score (04). Inflammation seems to be a risk factor for cognitive decline as demonstrated by a significant difference in MMSE in patients with poor oral health as opposed to controls (49). The same study found that periodontitis might increase inflammatory cytokines in rat models of Alzheimer disease. It is generally thought that MMSE results are influenced by premorbid cognitive ability (11).

The results of the MMSE for diagnostic work-up should generally be improved by comprehensive mental status examination and imaging, for instance, in patients suspected of having mild cognitive impairment for which the MMSE is often not sensitive enough (12). Multimodal methods have been developed for predicting the development of Alzheimer disease. The combination of present PET images, MRI images, CSF measures, etc. can predict much later values of continuous clinical variables like the MMSE with high accuracy (48).

Clinical vignette

A 71-year-old widow consulted her general practitioner about feelings of depression. She had lost her job 2 years earlier due to workforce reductions, and since that time, she had only limited social contacts. Medical intervention with an antidepressant was initiated.

Seven months later she returned to her doctor, accompanied by one of her daughters. There seemed to be no effect from the medication. The children had noticed that she had difficulty finding the appropriate words, she seemed to forget appointments, and she often told the same stories. Symptoms seemed to progress. She presented with a normal neurologic examination except for cognitive function.

Because Alzheimer disease was suspected, an MMSE was administered; she scored 24/30. Orientation to time was slightly compromised, but she scored maximum points on orientation to place. Although attention was moderately disturbed, and recall was slightly affected, she received a full score in language and visuospatial abilities. Because conclusions about the individual elements of a patients cognitive function should not be based solely on the MMSE, she was referred for full neuropsychological testing of orientation, memory, attention, abstraction and executive function, language, visual perception, and visuoconstructive abilities. Her blood tests were normal. Her cerebral MRI scan showed slight cortical atrophy.

On diagnosis of mild Alzheimer disease, she started treatment with a cholinesterase inhibitor. MMSE was then administered every 6 months and showed an initial improvement to 26/30 points, with a gradual decline to 16/30 over the subsequent 5 years. She was lost to follow-up for the last 3 years of her life, which were spent in a nursing home.

Scientific basis

The mental status examination not only identifies brain disease, but often points to the likely site of brain dysfunction. Functional specialization of different brain regions is evident; however, it must be noted that total segregation is never present, and the following conclusions should be read with this in mind (45):

Attention. Attention is thought to involve the reticular activating system, basal ganglia, and the frontal lobes. Abnormal performance is seen in disorders involving obtundation, in acute confusional states, and in severe dementias. Patients in acute confusional states or those with frontal lobe dysfunction tend to make errors of omission. Impulsive and anxious patients, as those suffering from ADHD, often make errors of commission.

Language. Disorders in language can represent focal lesions of the left hemisphere, especially focal-cortical lesions. Abnormalities in comprehension usually occur in association with lesions of the posterior left hemisphere and are present in Wernicke aphasia. Impaired repetition occurs in patients with aphasic syndromes and is associated with lesions involving structures bordering the Sylvian fissure. Reading and writing are typically compromised by lesions of the posterior left hemisphere.

Visuospatial skills. Visuospatial deficits are seen in patients with focal brain lesions as well as in those patients with degenerative brain disorders. Deficits are most marked when lesions involve the right posterior hemisphere, as seen in patients with the unilateral neglect syndrome, but lesions may be present in many other locations (eg, the frontal, parietal, or occipital lobe of either hemisphere).

Memory. The impaired ability to learn new material is a diagnostic criterion for dementia and mild cognitive impairment. Patients who fail to learn verbal material may have damage to the fornix, the hippocampus, the mammillary body, the mammillothalamic tract, or the medial thalamus due to stroke, hypoxia, infection, Wernicke-Korsakoff syndrome, or tumors. Patients with intact learning but poor retrieval have such frontal-subcortical disorders as Pick disease, Huntington disease, tumor, stroke, trauma, or infection. Amnesia for nonverbal material may reflect pathology of the right temporal lobe or medial limbic structures.

Executive functions. Executive function involves frontal lobes and related subcortical structures as affected in Pick disease and Huntington disease.