Dementia

Dementia is a syndrome of acquired, progressive cognitive impairment that interferes with a person's ability to function independently — not a single disease, but a clinical picture with several distinct underlying causes. Alzheimer disease (AD) is the most common cause, followed by vascular dementia (VD), Lewy body dementia (LBD), frontotemporal dementia (FTD), and mixed pathology (most often combined AD and VD); dementia is also a recognized complication of Parkinson disease (Parkinson disease dementia, PDD).^[1] The formal DSM-5-TR term is "major neurocognitive disorder," but "dementia" remains the term used throughout clinical practice and this chapter.

Dementia must be distinguished from delirium — an acute, fluctuating disturbance of attention and consciousness with an identifiable and often reversible medical or medication-related precipitant (see Delirium for the full diagnostic approach). The two conditions frequently coexist in the same patient, and a delirium that persists or only partly resolves can be genuinely difficult to distinguish from dementia — this overlap is especially relevant to Lewy body dementia, where fluctuating consciousness is itself a core diagnostic feature rather than automatically a sign of superimposed delirium.

How dementia actually presents

Presentation depends heavily on the underlying subtype, but common threads include a decline in memory, language, visuospatial ability, or executive function severe enough to interfere with instrumental activities of daily living — driving, managing medications and finances, cooking, shopping, using a telephone or other devices, housekeeping — before progressing to impair basic self-care (bathing, grooming, dressing, toileting, feeding).^[1]

• Older age — the single strongest risk factor for dementia overall.
• Family history of dementia, and known genetic risk factors (e.g., ApoE epsilon-4 carriage for Alzheimer disease).
• Down syndrome (markedly increased lifetime Alzheimer disease risk).
• Parkinson disease (risk of Parkinson disease dementia).
• Less education (to age 11-12).
• Social isolation and reduced social engagement.
• Untreated hearing impairment.
• Late-life depression.
• Mid-life obesity.
• Physical inactivity.
• Diabetes.
• Mid-life hypertension.
• Smoking.
• Excessive alcohol consumption.
• Traumatic brain injury.
• Air pollution exposure.
• Untreated or unmanaged sleep apnea, and severe sleep deprivation.

Drawn from the Lancet Commission's synthesis of modifiable dementia risk across the life course, alongside established non-modifiable factors.^[2]

History

A careful history — including collateral history from a caregiver or family member, since insight is often impaired — should establish the cognitive domains affected (memory, language, orientation, visuospatial ability, executive function), the speed of onset and progression, mood and behaviour changes (the NPI-Q and PHQ are useful structured tools), noncognitive symptoms (incontinence, gait disorder, dysphagia, constipation, parkinsonism), functional impact on instrumental and personal activities of daily living, and safety concerns such as wandering or leaving appliances on.

Distinguishing dementia from delirium and other reversible causes

New or acutely worsened confusion should always prompt consideration of delirium first — an acute, fluctuating disturbance of attention with an identifiable precipitant (see Delirium) — before attributing the change to progression of an existing or new dementia. A structured medication review and the laboratory workup below are aimed specifically at identifying reversible or contributing causes before a dementia diagnosis is finalized.

Medication review

Anticholinergic medication burden can itself cause or worsen cognitive impairment, and is more likely when a patient has multiple prescribers.^[1] Use an explicit tool — the Beers Criteria, STOPP/START criteria, or the Anticholinergic Cognitive Burden Scale and calculator — or an implicit tool such as the Medication Appropriateness Index to systematically identify contributing medications.

Physical examination

A full examination including a neurological exam helps identify contributing findings such as parkinsonism or apraxia, and should specifically screen for potentially reversible contributors such as findings suggesting hypothyroidism.

Laboratory tests

CBC, fasting glucose, electrolytes, calcium, renal function, TSH, and vitamin B12 in every patient being newly assessed for cognitive impairment. Atypical presentations warrant additional specialized investigation guided by a dementia specialist (e.g., FDG-PET, SPECT, or CSF studies). CSF or plasma amyloid/tau biomarkers to assess for prodromal Alzheimer disease are not recommended outside of a research setting for routine diagnostic workup — their role today is limited to confirming amyloid pathology as an eligibility requirement before anti-amyloid disease-modifying therapy (see Choose Initial Treatment).

Neuroimaging

Consider CT or MRI (MRI preferred where feasible) broadly for most older adults newly presenting with cognitive impairment. When imaging resources are limited, prioritize patients with any of the following: age under 60; new onset (under 2 years of symptoms); rapid progression (1-2 months); unexpected or unexplained decline in a patient with known dementia; recent significant head trauma or repeated mild head trauma; anticoagulant use or a bleeding disorder; unexplained neurologic symptoms, focal or lateralizing signs, or gait disturbance; early urinary incontinence with gait disorder; significant vascular risk factors; a history of cancer, particularly with brain metastasis risk; or atypical cognitive symptoms.

Cognitive screening tools

The Montreal Cognitive Assessment (MoCA) is most useful for screening mild cognitive impairment or early dementia. The Mini-Mental State Exam (MMSE), alone or combined with a clock-drawing test, is sensitive and specific for distinguishing moderate-stage dementia from normal cognition. The Ascertain Dementia-8 (AD-8) is completed by a caregiver. The Depistage Cognitif de Quebec (DCQ) includes a behavioural domain and is validated for atypical dementias. All brief screening tools carry test-specific biases and an increased risk of false-positive results and misclassification in patients of advanced age, those in a nursing home, those with lower educational attainment, and racialized populations — interpret results through an individualized lens rather than as a stand-alone diagnostic threshold, and use a culturally adapted translated version where available.^[1]

Routine screening for cognitive impairment in asymptomatic older adults — even those with vascular risk factors or a family history — is not recommended; screening is for patients with a specific cognitive or functional concern, not a population-wide test.^[1]

• Optimize function and slow the trajectory of cognitive and functional decline to the extent the specific dementia subtype and stage allow.
• Define specific, individualized target symptoms with the patient and caregiver at the outset, and reassess against those same specific targets — "reduction in repetitive questioning" is a more useful target than "improved cognition."
• Manage behavioural and psychological symptoms of dementia (BPSD) with the least restrictive, lowest-risk effective approach.
• Minimize the risk of harm to the patient and to others.
• Minimize medication-related harm — anticholinergic burden, sedative burden, and antipsychotic adverse effects in particular.
• Support caregiver capacity and reduce caregiver burden as an explicit part of the treatment plan, not an afterthought.

• Ongoing physical exercise, cognitive stimulation, and social engagement benefit most older adults with dementia, across subtypes and stages.
• Group cognitive stimulation therapy is beneficial for mild-moderate dementia.^[1]
• Frontotemporal dementia management is primarily nonpharmacologic, with emphasis on family/caregiver support; subtypes with prominent language deficits benefit from early speech-language pathology involvement.
• Address advance care planning early - advance health-care directives (a living will) and durable power of attorney - before disease progression affects decision-making capacity.
• Offer psychosocial and psychoeducational interventions to caregivers, and refer to the local Alzheimer Society as a starting point for resources, which vary by province.
• Counsel against driving once the patient has progressed beyond the earliest stages, particularly with prominent visuospatial or executive deficits, or with hallucinations or delusions - many provinces have mandatory reporting requirements at this stage.
• For BPSD specifically, apply nonpharmacologic strategies first and always: multidisciplinary care (recreation therapy, exercise, outdoor activity), sensory stimulation (music, therapeutic touch, massage), environmental modification (safe spaces to wander, removal of hazards), person-centred communication with the patient, and caregiver/staff education on distraction, redirection, avoiding confrontation, reassurance, and providing structure. Supervised skills training for care staff (person-centred care, communication, dementia care mapping) is effective; light therapy, aromatherapy, and caregiver training in formal behavioural-management therapy have insufficient evidence to expect benefit.^[8]
• Identify and manage sleep apnea, and avoid severe sleep deprivation (under 5 hours/night) - both may improve cognition.
• A Mediterranean-pattern diet - high in mono- and polyunsaturated fats, low in saturated fat, high in fruits and vegetables - is a reasonable, low-risk recommendation for cognitive health.
• When pharmacologic therapy for BPSD is genuinely necessary, continue nonpharmacologic strategies alongside it - this often allows a lower effective medication dose.

Breastfeeding: the same limited-data caveat applies; individualize with specialist input given how uncommon this clinical scenario is.

Renal and hepatic impairment

Memantine is renally cleared - reduce the target dose in significant renal impairment (e.g., target 5 mg PO BID rather than 10 mg BID with moderate-severe impairment). Galantamine should be avoided in severe hepatic or severe renal impairment. Donepezil is hepatically metabolized via CYP2D6/3A4 - use caution with significant hepatic impairment or strong CYP2D6/3A4 inhibitors. Rivastigmine is not hepatically metabolized and has no established interactions through the cytochrome P450 system, which can make it a useful choice when interactions are a specific concern.

Cardiac disease

All cholinesterase inhibitors carry a small risk of QT-interval prolongation and bradycardia. Consider a baseline ECG before starting one in a patient who is already bradycardic, on multiple other rate-limiting drugs (beta-blockers, calcium channel blockers, amiodarone), or has a history of unexplained syncope or presyncope - and discuss with a specialist first in that situation.

Lewy body dementia

Patients with LBD have markedly increased sensitivity to antipsychotic-induced dopamine blockade given the disease's core dopaminergic deficit - this can precipitate a severe, potentially irreversible neuroleptic sensitivity reaction with worsened parkinsonism, further impaired consciousness, and autonomic instability. Reserve an antipsychotic for LBD only when cholinesterase-inhibitor therapy has failed and responsive behaviours are causing severe distress or a safety risk; when unavoidable, quetiapine at the lowest possible dose under close supervision is better tolerated than risperidone, olanzapine, or any first-generation agent.^[3]

Parkinson disease dementia

Cholinesterase inhibitors (rivastigmine has the most trial evidence in this specific population) are generally well tolerated and do not meaningfully worsen motor symptoms beyond a possible mild increase in tremor. Minimize anticholinergic-burden medications, which worsen both motor and cognitive symptoms in this population. If an antipsychotic is required, the same LBD-level caution applies given the shared dopaminergic vulnerability.

Vascular dementia

Aggressive management of the underlying cardiovascular risk factors - blood pressure, lipids, diabetes, smoking cessation, and antiplatelet/anticoagulant therapy as indicated for the underlying vascular disease - is an essential parallel treatment track, not just prevention. Cholinesterase-inhibitor benefit is modest and of uncertain clinical significance in "pure" vascular dementia without a mixed AD component.

Older adults generally

Start low and titrate slowly for every drug class in this chapter. Minimize overall anticholinergic and sedative burden, and weigh fall risk carefully with any sedating or antipsychotic medication.

• Reassess target symptoms 2-4 weeks after starting or increasing the dose of a cholinesterase inhibitor or memantine, then periodically thereafter.
• In an untreated patient with mild-moderate dementia, MMSE typically declines by 2-4 points per year; an annual decline of under 2 points while on cholinesterase-inhibitor therapy is generally interpreted as a beneficial treatment effect, even though MMSE is a poor tool for judging response in an individual patient day-to-day.^[1]
• Monitor for cholinesterase-inhibitor adverse effects - GI symptoms, weight loss, bradycardia - and consider baseline/follow-up ECG in patients with the cardiac risk factors noted in Special Populations.
• Reassess renal function periodically to confirm memantine dosing remains appropriate.
• For anti-amyloid antibody therapy, follow the scheduled MRI monitoring program - baseline, before specific early infusions, and again at 1 year, with closer monitoring in ApoE epsilon-4 carriers or after any prior ARIA event - and ask specifically about new headache, confusion, visual change, nausea, or seizure between scheduled scans.^[7]
• For any antipsychotic started for BPSD, reassess the ongoing need every 3-6 months with a specific goal of tapering or discontinuing once symptoms have been stable - most patients do not need indefinite therapy.
• Review the full medication list at every visit for anticholinergic burden and other deprescribing candidates (see the medication-review table in Assessment and Diagnosis).
• Ask about caregiver wellbeing and burden at every visit - this is a monitoring parameter in its own right, not a courtesy question.

1. Confirm adherence to the current regimen before concluding a drug has failed.
2. If one cholinesterase inhibitor is not tolerated or does not appear effective, trial a different one before assuming the entire class has failed - response is genuinely individual.
3. If dementia progresses from mild-moderate to moderate-severe while a patient is on a cholinesterase inhibitor, add memantine rather than stopping the cholinesterase inhibitor - continuing shows benefit.
4. If BPSD persists despite an adequate nonpharmacologic trial and an optimized cholinesterase inhibitor/memantine regimen, reassess specifically for an untreated precipitant - pain, infection, constipation, a new medication, or superimposed delirium - before escalating to a psychotropic.
5. If an SSRI trial for agitation with mood/anxiety features is inadequate or not tolerated, reassess target symptoms before moving to an antipsychotic - do not escalate automatically.
6. If an antipsychotic response is inadequate at an appropriately titrated dose within a defined trial period, taper and stop it rather than continuing indefinitely or adding a second antipsychotic.
7. If antipsychotic-induced restlessness (akathisia) appears, recognize it as an adverse effect rather than "not enough drug" - reduce or stop the dose rather than increasing it.
8. Reassess the diagnosis itself if rapid or atypical decline occurs at any point in the treatment course - consider a reversible cause, a different dementia subtype, or superimposed delirium rather than assuming simple disease progression.

• Starting or continuing an antipsychotic in a patient with suspected or known Lewy body dementia without recognizing the severe neuroleptic sensitivity risk.^[3]
• Reaching for an antipsychotic as first-line or routine management of agitation without an adequate nonpharmacologic trial first, and without explicit boxed-warning-level mortality-risk counselling to the caregiver or substitute decision-maker.^[9]
• Treating antipsychotic-induced akathisia as inadequate drug effect and escalating the dose rather than recognizing it as an adverse effect.
• Stopping a cholinesterase inhibitor based on a single low MMSE score alone rather than the full clinical picture, including caregiver-reported function and behaviour.
• Adding a new medication to treat a cholinesterase inhibitor's side effect (e.g., an anticholinergic bladder agent for urinary frequency) instead of reducing the dose or switching agents.
• Using a cholinesterase inhibitor or memantine in frontotemporal dementia - not effective, and may worsen agitation.
• Missing an anticholinergic-burden medication - an over-the-counter first-generation antihistamine sleep aid, a bladder antimuscarinic, an older tricyclic antidepressant - as a reversible contributor to cognitive impairment.
• Treating a positive cognitive screening test (MMSE or MoCA) alone as diagnostic, without the supporting history, exam, and laboratory workup.
• Assuming eligibility for lecanemab or donanemab without confirmed amyloid pathology and ApoE genotyping - both are required prerequisites, not optional add-ons.
• Overlooking caregiver burden and support needs as part of the treatment plan.

How to use this table: Doses are adult doses and require patient-specific confirmation (renal/hepatic function, weight, allergy history). Pediatric dosing is weight-based per the current product monograph unless stated. Drug pricing is intentionally omitted. Select any drug name to open its complete Pharmasist Drug Database page.

Cholinesterase inhibitors

Block acetylcholinesterase, increasing synaptic acetylcholine - the mainstay of cognitive/functional therapy for Alzheimer disease, mixed AD/VD, Lewy body dementia, and Parkinson disease dementia. Not effective in frontotemporal dementia. Start at the lowest dose and titrate slowly - starting at a high dose without gradual titration is associated with more serious adverse events.

NMDA receptor antagonist

Anti-amyloid disease-modifying monoclonal antibodies (specialist-initiated only)

Health Canada-approved for mild cognitive impairment or mild Alzheimer dementia with confirmed amyloid pathology and a permitted ApoE genotype; access is currently limited pending public drug plan reimbursement decisions, which are evolving rapidly.^[11][12]

Antipsychotics for BPSD - reserve for severe, dangerous, or significantly distressing symptoms after nonpharmacologic measures fail

The chance of symptomatic benefit from a second-generation antipsychotic in dementia-related agitation is roughly matched by the chance of an adverse effect such as extrapyramidal symptoms, sedation, or worsened confusion - this is genuine equipoise, not a reason to avoid treatment when symptoms are severe and safety is at risk, but it is the basis for explicit informed counselling every time.^[13][10]

Antidepressants - depression, and BPSD with mood/anxiety features

Benzodiazepines - severe acute agitation only, not for routine or chronic use

Data on efficacy for BPSD are conflicting, and use carries a real risk of oversedation, falls, and worsening cognition; reserve for acute severe agitation when other measures fail.

Agents with weak (case-report-level) evidence only - not routine therapy

These agents seem to work best when the target behaviour resembles the psychiatric syndrome the drug is otherwise used for (e.g., lithium for manic-spectrum features), or after agents with a stronger evidence base have failed under specialist direction.

Evidence note: drug selection, dosing, adverse effects, and comments in this table reflect current Canadian practice, cross-checked against the 5th Canadian Consensus Conference on the diagnosis and treatment of dementia.^[1]