Are we ready for the dementia crisis?

Dementia is the UK's leading cause of death, responsible for around 66,000 deaths per year and having overtaken heart disease in 2011. An estimated 944,000 people in the UK live with the condition now; by 2040, as the population ages, that figure is projected to reach 1.6 million. The total cost to the UK economy is £34.7 billion per year — more than cancer, heart disease, and stroke combined, according to the Alzheimer's Society. Direct NHS costs account for £4.3 billion; social care costs £16.4 billion; and unpaid family carers absorb an estimated £13.9 billion of largely invisible labour. Dementia is not merely a health crisis: it is a fiscal and social one that scales with an ageing society.

Only 63.4% of estimated cases carry a formal diagnosis — below the NHS target of 66.7%, and a level last consistently met before 2019. The COVID-19 pandemic caused diagnosis activity to collapse to 55.7%, creating a backlog that has not fully cleared. Late or absent diagnosis matters because it delays access to treatment, support, and legal planning. Two-thirds of people with dementia live at home rather than in care, supported overwhelmingly by unpaid family carers who provide an average of 71 hours of care per week. This is not a choice made in full knowledge of alternatives: it is the structural consequence of a social care system that cannot absorb the scale of need. The carer burden — physical, financial, and psychological — is itself a significant public health issue.

Two drugs — lecanemab (Leqembi) and donanemab (Kisunla) — received FDA approval in 2023 and 2024 respectively as the first treatments shown to slow the progression of Alzheimer's disease, reducing cognitive decline by approximately 35% in early-stage patients. NICE is evaluating both for NHS use, but the pathway to access is narrow: the drugs require MRI screening, regular IV infusion, and cost around £26,000 per patient per year at US prices. They work only in early-stage disease, which reinforces the urgency of the diagnosis gap. Alongside drug treatment, the Lancet Commission's 2024 update identified 14 modifiable risk factors — including hearing loss, social isolation, depression, and air pollution — estimated to account for around 45% of dementia risk. Population-level public health action on these factors could substantially reduce future caseload, but it requires sustained political commitment across departments with no single owner.

Behind the clinical and fiscal numbers lies a care system sustained overwhelmingly by unpaid labour. An estimated 900,000 people provide informal care for someone with dementia, disproportionately women who reduce working hours or leave employment entirely to do so. The transition from home care to residential placement is typically triggered not by a step-change in clinical need but by carer breakdown—exhaustion, illness, or the point at which night-time wandering and behavioural symptoms become unmanageable alone. Once in a care home, self-funding residents pay on average 41% more than local authority-funded residents for identical services, a cross-subsidy that keeps the sector financially viable but imposes devastating costs on individuals and families. The care workforce itself is in crisis: approximately 250,000 vacancies exist nationally, with turnover rates exceeding 30% in many providers. South Asian populations are significantly underdiagnosed, partly due to cultural frameworks that normalise memory loss in older age and partly due to language barriers in assessment services. Caribbean communities show lower engagement with memory services that remain overwhelmingly white in staffing and cultural orientation. In rural areas, specialist memory assessment may require journeys that a person with dementia simply cannot make. Admiral Nurses—the specialist dementia nursing role—number only around 600 across the entire United Kingdom, against an estimated need of more than 3,000.

The headline prevalence figure of 944,000 people living with dementia in the United Kingdom is not a direct count. It is derived from age-specific prevalence rates—drawn primarily from the MRC Cognitive Function and Ageing Study—applied to ONS population projections. The resulting estimate carries wide uncertainty bands that are rarely reported alongside the central figure. The diagnosis rate of 63.4% uses this modelled prevalence as its denominator, meaning that if actual prevalence differs from the model's assumptions, the rate shifts accordingly—an improvement in diagnosis recording could appear as a deterioration if estimated prevalence is revised downward. Death registration data presents its own challenges: dementia frequently coexists with pneumonia, urinary tract infections, and cardiovascular conditions, any of which may be recorded as the primary cause of death, potentially undercounting dementia's true mortality burden. The widely cited £34 billion annual cost figure includes a monetisation of unpaid care time—valuing informal carers' hours at replacement wage rates—that not all health economists accept as methodologically sound. Care home cost data relies on industry surveys with variable response rates and limited independent verification. There is currently no administrative data infrastructure capable of tracking individual dementia patients longitudinally through primary care, hospital admissions, community services, and residential care, making it impossible to assess whole-pathway costs or outcomes with precision.