Diagnostic Approach for Dementia
Core Diagnostic Criteria
Dementia is diagnosed when cognitive or behavioral symptoms interfere with the ability to function at work or usual activities, represent a decline from previous levels of functioning, and are not explained by delirium or major psychiatric disorder. 1, 2
The diagnosis requires assessment across five key cognitive domains: 1, 2
- Memory impairment (repetitive questions, misplacing belongings, forgetting appointments)
- Executive dysfunction (poor judgment, inability to manage finances, poor decision-making)
- Visuospatial deficits (inability to recognize faces/objects, difficulty with spatial orientation)
- Language impairment (word-finding difficulty, speech errors)
- Personality/behavioral changes (uncharacteristic mood fluctuations, apathy, agitation)
Step-by-Step Diagnostic Algorithm
Step 1: Obtain Corroborative Informant History (MANDATORY)
Never rely solely on patient self-report—patients with dementia lack insight into their deficits. 3 Obtain structured history from a reliable informant using validated tools: 2, 3
- AD8 or Alzheimer's Questionnaire for cognitive/functional changes
- IQCODE or ECog for informant-reported cognition
- Lawton IADL Scale for functional assessment (finances, medications, transportation, household tasks, cooking, shopping)
- NPI-Q or MBI-C for behavioral symptoms
Document specific details: 3
- Onset and progression pattern (gradual vs. rapid, stepwise vs. continuous)
- Impact on instrumental activities of daily living
- Behavioral/personality changes
- Risk factors (stroke/TIA, depression, sleep apnea, head injury, family history)
Step 2: Cognitive Screening with Validated Tools
Use the Montreal Cognitive Assessment (MoCA) as first-line screening—it has superior sensitivity for mild cognitive impairment and early dementia compared to MMSE. 2, 3 The MoCA is more sensitive for detecting MCI and mild AD. 3
- MMSE remains valuable for moderate dementia and longitudinal tracking (sensitivity/specificity >80%)
- Clock Drawing Test as useful supplementary screening
If screening is inconclusive (symptoms present but normal examination), order neuropsychological testing to establish extent and severity of impairment objectively. 4, 5 This is particularly important for: 2
- Complex presentations
- Mild or unusual symptoms
- Educational extremes
- Language/cultural considerations
Step 3: Core Laboratory Testing (Rule Out Reversible Causes)
Order the following mandatory tests: 1, 2, 3
- Complete blood count
- Comprehensive metabolic panel (sodium, calcium, glucose)
- Thyroid function tests (TSH, free T4)
- Vitamin B12 and folate levels
- HIV testing if risk factors present
These tests identify reversible causes including hypothyroidism, metabolic encephalopathies (hyponatremia, hyperparathyroidism, hypoglycemia), and vitamin deficiencies. 6
Step 4: Structural Neuroimaging
MRI is preferred over CT, especially for detecting vascular lesions, atrophy patterns, and structural abnormalities. 1, 3
Neuroimaging is recommended when: 1
- Onset of cognitive symptoms within past 2 years (regardless of progression rate)
- Unexpected/unexplained decline in cognition or function
- Recent significant head trauma
- Unexplained neurological manifestations (severe headache, seizures, Babinski sign, gait disturbances)
- History of cancer (risk for brain metastases)
- Risk for intracranial bleeding
- Symptoms compatible with normal pressure hydrocephalus
- Significant vascular risk factors
If MRI is performed, use these sequences: 1
- 3D T1 volumetric sequence with coronal reformations for hippocampal assessment
- FLAIR (fluid-attenuated inversion recovery)
- T2 or susceptibility-weighted imaging (SWI)
- Diffusion-weighted imaging (DWI)
- Prefer 3T MRI over 1.5T if available
Use semi-quantitative scales for interpretation: 1
- Medial temporal lobe atrophy (MTA) scale
- Fazekas scale for white matter changes
- Global cortical atrophy (GCA) scale
If CT is performed, obtain non-contrast CT with coronal reformations to assess hippocampal atrophy. 1
Step 5: Rule Out Dementia Mimics and Contributors
Systematically evaluate for conditions that can mimic or exacerbate cognitive impairment: 3, 4
- Depression (use GDS, Cornell Scale for Depression in Dementia, or PHQ-9)
- Delirium (acute onset, fluctuating course)
- Medication effects (anticholinergics, benzodiazepines, opioids)
- Sleep disorders (particularly sleep apnea)
- Sensory deficits (hearing loss, vision loss)
- Pain and mobility problems affecting function
Step 6: Diagnostic Formulation
Integrate all data to establish: 2, 3
- Cognitive-functional status (does patient meet criteria for dementia vs. MCI?)
- Characterization of syndrome (which cognitive domains are affected?)
- Likely etiology based on clinical features:
- Alzheimer's disease (gradual onset, prominent memory impairment)
- Vascular dementia (stepwise decline, focal neurologic signs, vascular risk factors)
- Dementia with Lewy bodies (visual hallucinations, parkinsonism, fluctuating cognition)
- Frontotemporal dementia (behavioral/personality changes, language impairment, younger onset)
- Parkinson disease dementia (dementia developing >1 year after motor symptoms)
Advanced Testing (Tier 2 - Select Cases Only)
Consider specialized testing for atypical presentations: 2, 4
- Age of onset <65 years
- Rapid symptom onset
- Unusual cognitive profile (impairment in multiple domains but not episodic memory)
- Uncertain diagnosis after standard workup
- Functional imaging (FDG-PET, amyloid PET)
- CSF biomarkers (Aβ42, tau, phospho-tau)
- Genetic testing (if family history suggests autosomal dominant inheritance)
Do NOT routinely use advanced MR sequences (rs-fMRI, MR spectroscopy, DTI, ASL) or quantification software in clinical practice. 1
Common Pitfalls to Avoid
- Failing to obtain informant history leads to missed diagnoses due to patient lack of insight 2, 3
- Relying on patient self-report alone significantly reduces diagnostic accuracy 3
- Using a "shotgun approach" to testing rather than tiered, methodical evaluation 2
- Overlooking reversible causes (thyroid disease, B12 deficiency, medication effects) 2, 7
- Failing to use standardized, validated instruments reduces accuracy and makes longitudinal tracking unreliable 8
- Mistaking dementia as normal aging delays diagnosis and management 7
Longitudinal Monitoring
Schedule follow-up visits every 6-12 months for stable patients to track disease progression. 2, 8, 3 Use a multi-dimensional approach assessing: 1, 2, 8
- Cognition (repeat MoCA or MMSE)
- Functional autonomy (IADL scales)
- Behavioral symptoms (NPI-Q)
- Caregiver burden (Zarit Burden Interview)
More frequent assessment is needed for patients with behavioral symptoms. 8