Dementia assessment is often reduced to a number: an MMSE score, a MoCA score, or an ACE-III total. These measures are useful, but two people with similar scores may have different cognitive syndromes, different underlying diseases, and different support needs.

One person may repeatedly forget recent conversations. Another may remember them but struggle to interpret visual information. A third may develop impaired judgement and loss of empathy while performing adequately on conventional memory tasks.

Neuropsychological assessment examines the pattern behind the score. It helps identify which cognitive systems are affected, which abilities remain preserved, and what these findings suggest about the underlying illness.

Its conclusions are probabilistic. A cognitive profile can support a dementia subtype, but it cannot independently establish the underlying brain pathology.

Three levels of diagnosis: functional status, syndrome, and cause

A scientific assessment separates three related questions.

These levels are not interchangeable. An amnestic syndrome raises suspicion of Alzheimer’s disease but does not prove Alzheimer’s pathology. A language-led presentation can arise from different pathological processes.

The Alzheimer’s Association DETeCD-ADRD guideline adopts this layered approach and recommends neuropsychological evaluation when office-based assessment is insufficiently informative. NICE similarly recommends considering testing when the presence, cause, or subtype of cognitive impairment remains unclear.

Why a screening score cannot reliably subtype dementia

Brief screening instruments sample cognition across a limited set of tasks. Their total scores compress different abilities into one figure.

A low total may reflect widespread impairment or a particularly severe difficulty in one domain. Conversely, a person with high previous ability may experience substantial decline while remaining within a broadly normal range.

Behavioural change may also be poorly represented by conventional screening. Someone may perform adequately in a structured consultation while making increasingly unsafe decisions at home.

Distinguishing mild cognitive impairment from dementia therefore requires evidence about everyday independence, not just a cutoff. The history, trajectory of change, informant observations, and cognitive findings must be interpreted together. Education, occupation, culture, and medical or psychiatric conditions also affect performance.

Memory is not a single process

“Poor memory” does not explain why someone is forgetting. A detailed assessment examines several processes:

• Encoding: Was the information adequately registered?
• Learning: Did performance improve with repetition?
• Retention: How much information remained after a delay?
• Retrieval: Could it be accessed without assistance?
• Recognition: Could previously presented material be distinguished from distractors?

These distinctions matter because difficulty retrieving information is not identical to difficulty retaining it.

Controlled-learning procedures, such as the Free and Cued Selective Reminding Test, help examine whether structured encoding and retrieval cues improve performance.

In a 2008 study, impaired free recall occurred in 83% of the Alzheimer group and 79% of the vascular dementia group. After cueing, impaired total recall occurred in 71% and 21%, respectively. The Alzheimer group included mixed Alzheimer–vascular cases.

The study illustrates why free recall alone may be misleading. However, the dementia subgroups were small, and diagnoses were clinical; these figures should not be presented as universal diagnostic-accuracy estimates.

Typical amnestic Alzheimer’s disease

Typical Alzheimer’s disease often presents with progressive difficulty learning and retaining new information. Repeated questions, forgotten recent events, and loss of recently discussed plans may be prominent.

Controlled-learning studies show that memory impairment can persist despite assistance at encoding and retrieval. This supports an interpretation of deficient retention rather than an exclusively strategic retrieval problem. However, cueing responses vary, and no single response pattern establishes or excludes Alzheimer’s disease.

Interpretation must also account for the demands of the task. Poor recall of a complex figure cannot be confidently attributed to memory if the person never perceived or copied it adequately. Similarly, poor verbal learning may partly reflect impaired word comprehension.

Recognition should be assessed through both correct identifications and false-positive responses. Endorsing many items does not necessarily indicate preserved recognition.

Alzheimer’s disease can also underlie language-led and visual-led syndromes. Preserved performance on a conventional memory task therefore does not independently rule it out.

Vascular cognitive impairment

Vascular cognitive impairment has no single profile because vascular lesions differ in location, burden, and timing.

Subcortical vascular disease often affects processing speed, complex attention, and executive control. Strategic infarcts can instead produce prominent memory, language, or spatial deficits.

Testing may reveal difficulty organising responses, switching between rules, maintaining information in working memory, or completing tasks efficiently. Memory retrieval sometimes improves with structured support.

However, executive dysfunction alone does not establish vascular dementia. The cognitive syndrome must plausibly relate to cerebrovascular disease. Functional impairment from weakness or sensory loss must also be distinguished from cognitive loss of independence.

An appropriate formulation connects the cognitive pattern to the clinical course and imaging findings. Neither slow performance nor incidental white-matter changes should automatically determine the diagnosis.

Dementia with Lewy bodies

Dementia with Lewy bodies commonly produces disproportionate difficulties in attention, executive function, and visual processing. Memory may be affected but need not be the dominant early impairment.

The profile becomes more informative when accompanied by the core clinical features:

• Fluctuations in cognition or alertness.
• Recurrent, well-formed visual hallucinations.
• REM sleep behaviour disorder.
• Spontaneous parkinsonism.

The DLB consensus framework integrates these clinical features with diagnostic biomarkers.

Testing may require breaks or more than one session when alertness varies. Motor slowing can also affect timed tasks independently of cognitive impairment.

Variability alone is not diagnostic of DLB. Sleepiness, medicines, anxiety, sensory difficulties, and acute illness can also produce inconsistent performance.

Behavioural variant frontotemporal dementia

Behavioural variant frontotemporal dementia may become apparent through changes in social conduct, motivation, empathy, eating behaviour, or judgement.

The revised diagnostic framework includes early disinhibition, apathy, loss of sympathy or empathy, repetitive behaviours, and dietary changes. Executive impairment with relative preservation of memory and visuospatial abilities is characteristic, but not every patient follows that pattern.

Probable bvFTD requires more than unusual behaviour; functional decline and supportive imaging are relevant components of the criteria.

An informant history is especially valuable because it establishes change from the person’s previous personality. Social-cognition assessment can explore difficulties that conventional memory tasks may miss.

Being able to follow instructions in a quiet consultation does not necessarily demonstrate intact judgement in complex financial or interpersonal situations.

Primary progressive aphasia

Primary progressive aphasia is a syndrome in which language impairment becomes the leading early difficulty. Assessment examines connected speech, naming, comprehension, repetition, grammar, and motor speech.

These associations are probabilistic. The classification distinguishes the clinical syndrome from imaging-supported and pathologically established diagnoses.

A complaint of “word-finding difficulty” is therefore insufficient for subtyping. The examiner must determine whether the difficulty involves accessing a word, understanding its meaning, constructing grammar, retaining a sentence, or planning speech movements.

Posterior cortical atrophy

Posterior cortical atrophy involves progressive impairment of higher visual and other posterior cortical functions.

The presenting difficulties may involve reading, spatial judgement, object perception, calculation, or learned actions. Some other cognitive abilities may remain relatively preserved early in the illness.

The consensus framework explicitly separates the PCA syndrome from its underlying cause. Alzheimer’s disease is a common cause, but PCA and Alzheimer’s disease are not interchangeable diagnoses.

Assessment should avoid counting the same visual impairment repeatedly as evidence of multiple cognitive deficits. A visually presented memory task may fail because the material was not adequately perceived in the first place.

What should a comprehensive battery examine?

There is no single battery suitable for every patient. Test selection should follow the referral question, clinical presentation, language, and functional abilities.

The following are illustrative task types rather than a mandatory protocol.

Appropriate validated versions and normative data should be used, with administration and interpretation by trained professionals.

The scientific importance of error analysis

A cognitive test is rarely a pure measure of one function.

Drawing requires vision, comprehension, planning, and motor control. Verbal fluency requires language knowledge, initiation, speed, and strategy. Poor performance identifies difficulty with the task; understanding its mechanism requires converging evidence.

A defensible report distinguishes observation from inference:

Patterns across several measures usually carry more meaning than an isolated low result. Large batteries generate many scores, and occasional low performances can occur by chance.

These are principles of clinical interpretation, not an automatic algorithm for assigning a dementia subtype.

Language, literacy, and education in Indian practice

Years of schooling do not fully capture educational exposure. Literacy, quality of schooling, occupation, language proficiency, and familiarity with formal testing can differ substantially between people with similar educational histories.

A translated test is not automatically equivalent to a validated adaptation. Word frequency, cultural familiarity, and literacy demands may change task difficulty.

The report should document the testing language, relevant sensory or motor accommodations, and departures from standard administration. When suitable norms are unavailable, that limitation should affect diagnostic confidence.

These issues matter in both directions: limited educational opportunity can be mistaken for decline, while high previous ability can conceal decline within a broadly average score.

Confounders and mixed disease

Depression, sleep disturbance, medication effects, pain, sensory impairment, and medical illness can affect cognition. These conditions may coexist with neurodegeneration.

Sudden cognitive change over hours or days requires prompt medical assessment for delirium and other acute causes. Extensive dementia-subtype testing is not the immediate priority.

Mixed pathology also limits neat classification. Alzheimer’s disease may coexist with vascular injury or Lewy body disease. An overlapping cognitive profile may be clinically meaningful.

NICE emphasises assessment of potentially reversible contributors and integration of cognitive findings with further investigation when indicated.

Integrating neuropsychology with imaging and biomarkers

Neuropsychological testing characterises the clinical expression of disease. Structural imaging assesses anatomy, vascular injury, and other lesions. Selected molecular or functional investigations address additional questions about the underlying cause.

The specialist DETeCD-ADRD guideline places these investigations within a tiered process informed by the clinical syndrome. Testing should resolve meaningful diagnostic uncertainty.

An illustrative formulation could read:

Progressive amnestic multidomain cognitive impairment with loss of independence in medication and financial management. Alzheimer’s disease is a leading consideration; vascular contributions and alternative causes require correlation with clinical and imaging findings.

This communicates what is established, what is suspected, and what remains unresolved.

How the profile improves care

Assessment should guide practical action. The following are individualised applications rather than subtype-specific prescriptions.

Preserved abilities are equally valuable. Familiar skills, established routines, and effective communication channels should shape the care plan.

Repeat assessment may clarify change, but interpretation must consider practice effects, fatigue, illness, and measurement error. Small score changes should not automatically be treated as progression or treatment response.

The value of neuropsychological assessment lies in connecting cognitive performance with the person’s history, everyday abilities, and care needs—while making diagnostic uncertainty explicit.