Neuropsychological testing is a key diagnostic tool for assessing people with dementia and mild cognitive impairment, but can also help in other neurological conditions such as Parkinson’s disease, stroke, multiple sclerosis, traumatic brain injury and epilepsy. While cognitive screening tests offer gross information, detailed neuropsychological evaluation can provide data on different cognitive domains (visuospatial function, memory, attention, executive function, language and praxis) as well as neuropsychiatric and behavioural features. We should regard neuropsychological testing as an extension of the neurological examination applied to higher order cortical function, since each cognitive domain has an anatomical substrate. Ideally, neurologists should discuss the indications and results of neuropsychological assessment with a clinical neuropsychologist. This paper summarises the rationale, indications, main features, most common tests and pitfalls in neuropsychological evaluation.
- cognitive neuropsychology
- alzheimer-s disease
Statistics from Altmetric.com
Neuropsychological testing explores cognitive functions to obtain information on the structural and functional integrity of the brain, and to score the severity of cognitive damage and its impairment on daily life activities. It is a core diagnostic tool for assessing people with mild cognitive impairment, dementia and Alzheimer’s disease,1 but is also relevant in other neurological diseases such as Parkinson’s disease,2 stroke,3 4 multiple sclerosis,5 traumatic brain injury6 and epilepsy.7 Given the relevance and extensive use of neuropsychological testing, it is important that neurologists know when to request a neuropsychological evaluation and how to understand the results. Neurologists and clinical neuropsychologists in tertiary centres often discuss complex cases, but in smaller hospitals and in private practice this may be more difficult. This paper presents information on neuropsychological testing in adult patients, and highlights common pitfalls in its interpretation. A recent paper in Practical Neurology focused on neuropsychological assessment in epilepsy.7
Neuropsychological testing and its clinical role
Why is neuropsychological testing important?
From early in their training, neurologists are taught to collect information on a patient’s symptoms and to perform a neurological examination to identify clinical signs. They then collate symptoms and signs into a syndrome, to identify a lesion in a specific site of the nervous system, and this guides further investigations. Since cognitive symptoms and signs suggest damage to specific brain areas, comprehensive cognitive assessment should also be part of the neurological examination. Neuropsychological testing may be difficult to perform during office practice or at the bedside but the data obtained nevertheless can clearly complement the neurological examination.
When is neuropsychological testing indicated and useful?
Neuropsychological assessment is indicated when detailed information about cognitive function will aid clinical management:
To assess the presence or absence of deficits and to delineate their pattern and severity.
To help to establish a diagnosis (eg, Alzheimer’s disease or frontotemporal dementia) or to distinguish a neurodegenerative condition from a mood disorder (eg, depression or anxiety).
To clarify the cognitive effects of a known neurological condition (multiple sclerosis, stroke or brain injury).
Neuropsychological testing may address questions about cognition in helping to guide a (differential) diagnosis, obtain prognostic information, monitor cognitive decline, control the regression of cognitive–behavioural impairment in reversible diseases, guide prescription of a medication, measure the treatment response or adverse effects of a treatment, define a baseline value to plan cognitive rehabilitation or to provide objective data for medicolegal situations (table 1). When requesting a neuropsychological assessment, neurologists should mention any previous testing and attach relevant reports, so that the neuropsychologist has all the available relevant information.
Conversely, there are situations when cognitive evaluation should not be routinely recommended, for example, when patient is too severely affected, the diagnosis is already clear, testing may cause the patient distress and/or anxiety, the patient has only recently undergone neuropsychological assessment, there is only a low likelihood of an abnormality (though the test may still bring reassurance) and when there are neuropsychiatric symptoms (table 2). Neuropsychological assessment is time-consuming (1–2 hours) and demanding for the patient, and so neurologists much carefully select subjects for referral.
How is neuropsychological testing done?
Neuropsychological evaluation requires a neurologist or a psychologist with documented experience in cognitive evaluation (ie, a neuropsychologist). The clinician starts with a structured interview, then administers tests and questionnaires (table 3) and then scores and interprets the results.
The interview aims to gather information about the medical and psychological history, the severity and the progression of cognitive symptoms, their impact on daily life, the patient’s awareness of their problem and their attitude, mood, spontaneous speech and behaviour.
Neuropsychological tests are typically presented as ‘pencil and paper’ tasks; they are intrinsically performance based, since patients have to prove their cognitive abilities in the presence of the examiner. The tests are standardised, and so the procedures, materials and scoring are consistent. Therefore, different examiners can use the same methods at different times and places, and still reach the same outcomes.
The scoring and analysis of the test results allow the clinician to identify any defective functions and to draw a coherent cognitive picture. The clinician should note any associations and dissociations in the outcomes, and use these to compare with data derived from the interview, including observation of the patient, the neuroanatomical evidence and theoretical models, to identify a precise cognitive syndrome.
What information can neuropsychological testing offer?
Neuropsychological assessment provides general and specific information about cognitive performance.
Brief cognitive screening tools, such as the Mini-Mental State Examination (MMSE), the Montreal Cognitive Assessment (MoCA) and the Addenbrookes Cognitive Examination revised (ACE-R), provide a quick and easy global, although rough, measure of a person’s cognitive function,8 9 when more comprehensive testing is not practical or available. Table 4 gives the most common cognitive screening tests, along with scales for measuring neuropsychiatric and behavioural problems, and their impact on daily life. This type of screening test may suffice in some cases, for example, when the score is low and patient’s history strongly suggests dementia, or for staging and following-up cognitive impairment with repeated testing. However, neurologists should be aware of the limitations of such cognitive screening tools. Their lack of some subdomains may result in poor sensitivity, for example, MMSE may give false negative findings in ‘Parkinson’s disease-related mild cognitive impairment’ because it does not sufficiently explore the executive functions that are the first cognitive subdomains to be involved in Parkinson’s disease. The MMSE is particularly feeble in assessing patients with frontotemporal dementia, many of whom score within the ‘normal’ range on the test, yet cannot function in social or work situations.10 Also, young patients with a high level of education may have normal screening tests because these are too easy and poorly sensitive to mild cognitive alterations. Such patients therefore need a thorough assessment.
A comprehensive neuropsychological evaluation explores several cognitive domains (perception, memory, attention, executive function, language, motor and visuomotor function). The areas and subdomains addressed in neuropsychological examination and the tests chosen depend on the referral clinical question, the patient’s and caregiver’s complaints and symptoms, and the information collected during the interview. Observations made during test administration may guide further exploration of some domains and subdomains. Failure in a single test does not imply the presence of cognitive impairment, since it may have several reasons (eg, reduced attention in patients with depression). Also, single tests are designed to explore a specific domain or subdomain preferentially, but most of them examine multiple cognitive functions (eg, clock drawing test, table 5). For these reasons, neuropsychological assessment is performed as a battery, with more than one test for each cognitive domain.
The main cognitive domains with their anatomical bases are reviewed below. Table 5 summarises the most widely used cognitive tests for each domain. The neuropsychologist chooses the most reliable and valid test according to the clinical question, the neurological condition, the age and other specific factors.
Parallel forms (alternative versions using similar material) may reduce the effect of learning effect from repeated evaluations. They may help to track cognitive disorders over time, to stage disease severity and to measure the effect of pharmacological or rehabilitative treatment.
Main cognitive domains and their anatomical bases
Most cognitive functions involve networks of brain areas.11 Our summary below is not intended as an old-fashioned or phrenological view about cognition, but rather to provide rough clues on where the brain lesion or disease may be.
This process allows recognition and interpretation of sensory stimuli. Perception is based on the integration of processing from peripheral receptors to cortical areas (‘bottom-up’), and a control (‘top-down’) to modulate and gate afferent information based on previous experiences and expectations. According to a traditional model, visual perception involves a ventral temporo-occipital pathway for objects and faces recognition, and a dorsal parieto-occipital pathway for perception and movement in space.12 Acoustic perception involves temporal areas.
The classical neurological examination involves evaluation of strength, coordination and dexterity. Neuropsychological assessment explores other motor features ranging from speed to planning. Visuomotor ability requires integration of visual perception and motor skills and is usually tested by asking the subject to copy figures or perform an action. Apraxia is a higher order disorder of voluntary motor control, planning and execution characterised by difficulty in performing tasks or movements when asked, and not due to paralysis, dystonia, dyskinesia or ataxia. The traditional model divides apraxia into ideomotor (ie, the patient can explain how to perform an action, but cannot imagine it or make it when required) and ideational (ie, the patient cannot conceptualise an action or complete the correct motor sequence).13 However, in clinical practice, there is limited practical value in distinguishing ideomotor from ideational apraxia—see recent review in this journal.14 15 Apraxia can be explored during routine neurological examination, but neuropsychological assessment may offer a more detailed assessment.
Motor control of goal-orientated voluntary tasks depends on the interplay of limbic and associative cortices, basal ganglia, cerebellum and motor cortices.
Memory and learning are closely related. Learning involves acquiring new information, while memory involves retrieving this information for later use. An item to be remembered must first be encoded, then stored and finally retrieved. There are several types of memory. Sensory memory—the ability briefly to retain impressions of sensory information after the stimulus has ended—is the fastest memory process. It represents an essential step for storing information in short-term memory, which lasts for a few minutes without being placed into permanent memory stores. Working memory allows information to be temporarily stored and managed when performing complex cognitive tasks such as learning and reasoning. Therefore, short-term memory involves only storage of the information, while working memory allows actual manipulation of the stored information. Finally, long-term memory, the storage of information over an extended period of time, can be subdivided into implicit memory (unconscious/procedural; eg, how to drive a car) and explicit memory (intentional recollection; eg, a pet’s name). Within explicit memory, episodic memory refers to past experiences that took place at a specific time and place and can be accessed by recall or by recognition. Recall implies retrieving previously stored information, even if they are not currently present. Recognition refers to the judgement that a stimulus presented has previously occurred.
The neuroanatomical bases of memory are complex.16 The initial sensory memory includes the areas of the brain that receive visual (occipital cortex), auditory (temporal cortex), tactile or kinesthetic (parietal cortex) information. Working memory links to the dorsolateral prefrontal cortex (involved in monitoring information) and the ventrolateral prefrontal cortex (involved in maintaining the information). Long-term memory requires a consolidation of information through a chemical process that allows the formation of neural traces for later retrieval. The hippocampus is responsible for early storage of explicit memory; the information is then transmitted to a larger number of brain areas.
Attention includes the ability to respond discretely to specific stimuli (focused attention), to maintain concentration over time during continuous and repetitive tasks (sustained attention), to attend selectively to a specific stimulus filtering out irrelevant information (selective attention), to shift the focus among two or more tasks with different cognitive requirements (alternating attention) and to perform multiple tasks simultaneously (divided attention). Spatial neglect refers to failure to control the spatial orientation of attention, and consequently the inability to respond to stimuli.17
The occipital lobe is responsible for visual attention, while visuospatial analysis involves both the occipital and parietal lobes. Attention to auditory stimuli requires functioning of the temporal lobes, especially the dominant (usually left) one for speech. Complex features of attention require the anterior cingulate and frontal cortices, the basal ganglia and the thalamus.
Executive functions include complex cognitive skills, such as the ability to inhibit or resist an impulse, to shift from one activity or mental set to another, to solve problems or to regulate emotional responses, to begin a task or activity, to hold information in mind for completing a task, to plan and organise current and future tasks, and to monitor one’s own performance.18 Taken together, these skills are part of a supervisory or meta-cognitive system to control behaviour that allows us to engage in goal-directed behaviour, prioritise tasks, develop appropriate strategies and solutions, and be cognitively flexible. These executive functions require normal functioning of the frontal lobe, anterior cingulate cortex, basal ganglia, and many inward and outward connections to the cortical and subcortical areas.
Language includes several cognitive abilities that are crucial for understanding and producing spoken and written language, as well as naming. Given its complexity, we usually explore language with batteries of tests that use different tasks to investigate its specific aspects (table 5). According to the traditional neuroanatomical view, language relies primarily on the dominant brain: specifically comprehension lies on the superior temporal lobe, language production on the frontal regions and frontoparietal/temporal circuits, and conceptual–semantic processing on a network that includes the middle temporal gyrus, the posterior middle temporal regions and superior temporal and inferior frontal lobes.19 However, recent data from stroke patients do not support this model, but instead indicate that language impairments result from disrupted connectivity within the left hemisphere, and within the bilaterally distributed supporting processes, which include auditory processing, visual attention and motor planning.11
Regardless of the theoretical model, there is agreement that intellectual ability—or intellectual quotient (IQ)—is a multidimensional construct. This construct includes intellectual and adaptive functioning, communication, caring for one’s own person, family life, social and interpersonal skills, community resource use, self-determination, school, work, leisure, health and safety skills. The Wechsler Adult Intelligence Scale revised (WAIS-R) is the best-known intelligence test used to measure adult IQ. WAIS-R comprises 11 subtests grouped into verbal and performance scales (table 5). Any mismatch between verbal and performance scores might suggest different pattern of impairments, that is, memory and language versus visuospatial and executive.
Comparing to normative values
A person’s performance on a cognitive test is interpreted by comparing it to that of a group of healthy individuals with similar demographic characteristics. Thus, the raw score is generally corrected for age, education and sex, and the corrected score rated as normal or abnormal. However, not all neuropsychologists use the same normative values. Furthermore, there are no clear guidelines or criteria for judging normality of cognitive testing. For example, the diagnostic guidelines for mild cognitive impairment in Parkinson’s disease stipulate a performance on neuropsychological tests, that is, 1–2 SDs below appropriate norms, whereas for IQ, a performance that is significantly below average is defined as ≤70, that is, 2 SD below the average score of 100.2 Sometimes, the neuropsychological outcome is reported as an equivalent score, indicating a level of performance (figure 1). Understanding how normality is defined—how many SDs below normal values and the meaning of an equivalent score—is crucial for understanding neuropsychological results correctly and for comparing the outcomes of evaluations performed in different clinical settings. Furthermore, estimating the premorbid cognitive level, for example, using the National Adult Reading Test (table 4), helps to interpret the patient score. ‘Crystallised intelligence’ refers to consolidated abilities that are generally preserved until late age, compared with other abilities such as reasoning, which show earlier decline. In people with a low crystallised intelligence—and consequently a low premorbid cognitive level—a low-average neuropsychological assessment score may not represent a significant cognitive decline. Conversely, for people with high premorbid cognitive level, a low-average score might suggest a significant drop in cognitive functioning.
Reaching a diagnosis through neuropsychological testing
Although the score on a single test is important, it is only the performance across the whole neuropsychological test battery that allows clinicians to identify a person’s patterns of cognitive strengths and weaknesses; together with motor and behavioural abnormalities, these may fit into known diagnostic categories (tables 6 and 7).
The neuropsychologist reports the information collected through neuropsychological evaluation in a written clinical report that usually includes the scores of each test administered. The conclusions of the neuropsychological report are important to guide further diagnostic workup, to predict functionality and/or recovery, to measure treatment response and to verify correlations with neuroimaging and laboratory findings.
As well as these quantified scores, it is critically important to have a patient’s self-report of functioning, plus qualitative data including observation of how the patient behaved during the test.
Psychiatric confounders require particular attention. Neuropsychologists apply scales for depression (eg, Beck’s depression inventory, geriatric depression scale) or anxiety (eg, state–trait anxiety inventory) during testing; these may offer information on how coexisting conditions may influence cognition through changes in mood or motivational state. For example, it may be difficult to distinguish between dementia and depressive pseudodementia, because depression and dementia are intimately related.20 Table 8 shows some of the features that may help. Note that antidepressants may ameliorate cognitive deficits, particularly attention and memory, and that opioids may worsen cognitive symptoms.
Knowing that there are other potential factors that may influence neuropsychological testing (and usually worsening performance) should help clinicians to avoid misinterpreting the results (table 9). For example, in Parkinson’s disease, it is important to pay particular care to motor fluctuations, neuropsychiatric symptoms, pain and drug side effects that can worsen cognitive performance.21
Conversely, patients with long-lasting psychiatric disease, such as bipolar disorder or schizophrenia, are often referred for neurological and cognitive assessment when they begin to perform worse in daily activities. Frontal changes are common in bipolar disorders and so finding prefrontal dysfunction in such patients should not lead clinicians to suspect an ongoing neurological disorder. Discussion with the clinical neuropsychologist and the psychiatrist may help to understand potential drug side effects and, eventually, to revise treatment.
Further recommended reading
Lezak MD, Howieson DB, Bigler ED, Tranel D. Neuropsychological Assessment. Fifth edition. Oxford University Press 2012.
For many neurological diseases, neuropsychological testing offers relevant clinical information that complements the neurological examination.
Neuropsychological tests can identify patterns of cognitive strengths and weaknesses that are specific to particular diagnostic categories.
Neuropsychological testing involves tests that investigate different cognitive functions in a standardised way, and so the procedures, materials and scoring are consistent; it also involves an anamnestic interview, scoring and interpreting the results, and comparing these with other clinical data, to build a diagnostic hypothesis.
Neuropsychological evaluation must be interpreted in the light of coexisting conditions, in particular sensory, motor and psychiatric disturbances as well as drug side effects, to avoid misinterpreting the results.
Contributors CZ, AF, AM and ST: designed the article, collected and interpreted the data, drafted the manuscript and revised it. MT and MB: designed the article, collected and interpreted the data, and revised the manuscript for important intellectual content. CZ and ST: take full responsibility for the content of this review. All authors approved the final version of the article.
Funding This research received no specific grant from anyfunding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent Not required.
Provenance and peer review Commissioned; externally peer reviewed. This paper was reviewed by Nick Fox, London, UK.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.