End of Life With Dementia: Where People Die Varies Widely Between Countries

Dementia is one of the leading causes of death in the world, yet it is still too often not treated as a condition people die from. Where people with dementia spend their last days, and whether they get palliative care, depends less on what they need than on the country, the services and the planning around them.

For families, the end of life with dementia rarely arrives as a single moment. It tends to come slowly, through months of eating less, recurring infections and growing frailty, often after the person can no longer say what they want. Families face practical, heavy questions, among them whether anyone asked the person, while they still could, where they would want to be. This article brings together international evidence on where people with dementia die, what the last months of life involve, how far palliative care reaches them, and what planning ahead appears to change.

Dementia is a condition people die from, and the need for palliative care is growing fastest

The World Health Organization (WHO) estimates that 57 million people were living with dementia worldwide in 2021, and describes dementia as currently the seventh leading cause of death globally^[4] (see our study of dementia as a leading cause of death). Yet researchers following people with advanced dementia in US nursing homes wrote that dementia, though a leading cause of death in the United States, is “underrecognized as a terminal illness”^[5]. That gap matters, because care that is not seen as end-of-life care is rarely planned as such.

The need is set to grow. A modelling study published in The Lancet Global Health in 2019 combined WHO mortality projections for 2016–2060 with estimates of symptom prevalence in 20 conditions to project how many people will die with “serious health-related suffering”, the measure of palliative care need defined by the Lancet Commission on Palliative Care and Pain Relief. It projected that 48 million people a year (47% of all deaths) will die with this level of need by 2060, up 87% from 26 million in 2016. Dementia had the highest proportional increase of any condition: a 264% rise, to about 6 million people a year^[1]. These are projections, resting on WHO mortality forecasts made before the COVID-19 pandemic, not counts.

Today’s services fall far short even of current need. WHO estimates that 56.8 million people need palliative care each year, including 25.7 million in their last year of life, and that only about 14% of people who need it receive it. In a 2019 WHO survey of its 194 Member States, palliative care funding was available in 68% of countries, and only 40% reported that services reached at least half of the patients who needed them^[2]. These figures cover all conditions; WHO lists dementia among them, but no comparable global figure exists for people with dementia alone.

Where people with dementia die depends heavily on where they live

Place of death is an imperfect measure of good care, but one of the few counted for whole populations. The clearest finding across countries is how much it varies.

A study of death certificates in five European countries looked at people aged 65 and over whose underlying cause of death was a dementia-related disease. Between 50% (Wales) and 92% (the Netherlands) died in a nursing home, and between 3% (the Netherlands) and 46% (Wales) died in hospital. Death at home was rare, at 3–5%, except in Belgium (11%). Differences in the availability of hospital and nursing home beds explained only part of the variation between countries^[3].

Outside Europe the picture can be very different. In Japan, an analysis of death certificates for people aged 65 and over who died between 2018 and 2021 found that deaths from dementia mostly occurred in hospitals (59%), and that people with dementia there are often admitted to psychiatric hospitals for the management of neuropsychiatric symptoms^[6]. A 2026 scoping review of 48 studies found that dying in a long-term care facility was more common in Europe and North America, while hospitals and other medical facilities were more common in East Asia and the Pacific. Most of the studies came from high-income countries^[7], so very little is known about where people with dementia die across most of Africa, South Asia and Latin America.

These studies use different age bands, years and definitions (dementia as the underlying cause of death, or any mention of it), so the rows are not directly comparable. They show the scale of the differences, not a ranking.

In England and the US, fewer people with dementia now die in hospital

England has unusually deep data and serves as a worked case here. A study of 388,899 deaths of adults aged over 60 with dementia mentioned on the death certificate in England between 2001 and 2010 found that most died in care homes (55.3%) or hospitals (39.6%): two in five died in hospital. A pattern of rising hospital deaths reversed in 2006, and hospital deaths then fell by 0.93% a year. Care home deaths were more likely in areas with more care home beds (a proportion ratio of 1.82)^[8].

More recent national figures from the Office for Health Improvement and Disparities (OHID), based on Office for National Statistics death data for people with dementia aged 65 and over, show the shift continuing. In 2016, 30.9% died in hospital; in 2023, 26.3% did (24,030 of 91,510 deaths). Over the same period, the share dying at home rose from 9.7% to 15.5%, while care homes remained the most common place of death, at 56.4% in 2023^[9]. The home share jumped during the pandemic, from 11.2% in 2019 to 15.6% in 2021, and has stayed close to that level since. The two English sources use different age bands and may define a dementia death differently, so they are best read as two separate series rather than one continuous trend.

The United States shows a similar direction. Death certificate data for 2000–2014 showed people who died with dementia becoming more likely to die at home and less likely to die in institutions, with wide variation between states. States that spent more Medicaid money on home and community-based services had higher rates of home death^[10]. Fewer hospital deaths do not by themselves show that care at home is good, and the researchers called for its quality to be monitored.

The last months of advanced dementia bring complications that can be foreseen

One of the most detailed pictures of the end of life with dementia comes from the US CASCADE study, which followed 323 nursing home residents with advanced dementia, and the relatives or friends who made decisions with them, for 18 months in 22 nursing homes. Over that time, 54.8% of the residents died. The probability of an eating problem was 85.8%, of a fever 52.6% and of pneumonia 41.1%, and each of these complications was linked to a high chance of death within six months. Distressing symptoms were common: 46.0% of residents had shortness of breath and 39.1% had pain^[5].

In their last three months of life, 40.7% of residents had at least one burdensome intervention: a hospital admission, an emergency department visit, treatment through a drip or injection, or tube feeding. The study’s most striking finding was about information. Residents whose decision-makers understood the likely prognosis and the complications to expect had much lower odds of such interventions in their last three months (adjusted odds ratio 0.12)^[5]. The complications of advanced dementia are largely foreseeable. This is one US study, describing what happened, not what should happen for any individual.

Time in hospital near the end of life can be long. In Western Australia, linked hospital and death records for 29,884 people with dementia who died between 1990 and 2005 showed an average of 31.8 days in hospital in the last year of life, and 62% of admissions were emergencies. People in remote and very remote areas spent longer in hospital than those in cities (rate ratios of 1.7 and 3.0), and the authors pointed to geographic isolation and access to services outside hospital^[11]. The data are old, but show how geography and gaps in community care can turn the last year of life into repeated hospital stays (see our study of hospital admissions for people with dementia).

Planning ahead and support at home are linked with dying in the place people prefer

Where preferences are recorded, few people with dementia choose hospital. Among 803 people with dementia who died between December 2015 and March 2017 and had a record on Coordinate My Care, an electronic palliative care coordination system, the most common preferred place of death was a care home (58.8%), followed by home (39.0%). Overall, 83.7% died in their preferred place. The odds of this were higher where the agreed goal of treatment was symptom relief only (odds ratio 2.65)^[12]. People with such a record had, by definition, already had a planning conversation, so this group is not representative of everyone with dementia. Families and other unpaid carers usually carry these conversations and much of the care around them (see unpaid carers).

The evidence on advance care planning programmes is thinner than their importance suggests. A 2026 systematic review found only ten studies of advance care planning programmes for people with dementia in nursing homes, all from the US, the UK, Belgium, France and Finland. Programmes increased the recording of do-not-resuscitate decisions (odds ratio 14.12, from two studies of 453 people), but showed no clear effect on whether people died in the nursing home rather than being moved, and the certainty of evidence was very low for both^[13]. No studies came from non-Western countries.

The way services are organised and paid for also matters. Among more than 2 million US Medicare beneficiaries who died in 2019, those who had received home health care in their last three years of life were less likely to die in hospital without hospice care (23.3% against 31.5%), and the link with dying at home with hospice care was strongest for people with dementia whose home health care began at least a year before death^[14]. A 2024 review of 33 studies covering more than 5 million people with advanced dementia found that capitated funding, a fixed payment per person, was associated with fewer hospital deaths, while the number of hospital and nursing home beds was not^[15]. These are associations from observational data, not proof that one model causes better deaths, but they point in the same direction: support that reaches people where they live, early enough to matter.

What the evidence asks of us

The figures point to a simple principle: end-of-life care for people with dementia should be planned for, funded and measured as end-of-life care, rather than left to happen by default in whichever setting has a bed.

• Recognise dementia as a condition people die from, and plan for it. With dementia projected to show the fastest rise in palliative care need of any condition to 2060^[1], health systems and national dementia plans need palliative care for dementia written in, not added later. NICE guideline NG97 for England already recommends that people living with dementia are offered “flexible, needs-based palliative care” from diagnosis, recognising “how unpredictable dementia progression can be”^[16].
• Offer planning conversations early and often. NG97 recommends early and ongoing opportunities to discuss planning ahead, including preferences for place of care and place of death^[16]. The CASCADE and Coordinate My Care studies link a shared understanding of what to expect with fewer burdensome interventions and more deaths in the preferred place^[5][12].
• Fund care where people live. WHO notes that early palliative care reduces unnecessary hospital admissions^[2], and the US and review evidence associates home and community-based services and integrated funding with fewer hospital deaths^[10][15][14]. Care homes, where people with dementia most often die in Europe and North America, need the staffing and palliative skills to match (see our study of dementia and care homes).
• Measure what matters, everywhere. Research on where people with dementia die comes mostly from high-income countries^[7], and studies of advance care planning programmes only from Western countries^[13]. Funders and statistics agencies should collect place of death, symptom control and palliative care use for people with dementia, and researchers should study end-of-life care in low- and middle-income countries, where most future need will be.

Language: this piece uses person-first language ("people living with dementia", "people with dementia"), as most dementia organisations and people with lived experience prefer, and uses clinical terms such as "Alzheimer's disease" or "mild cognitive impairment" only where a study, dataset or diagnosis refers to that specific condition.

Why we're publishing this

Dementia organisations and carers’ groups argue for better care at the end of life, often with too little data to hand. This study brings the international evidence together in one place for them to cite, as part of our Dementia & Health research library. A good death with dementia is possible: free of avoidable pain and breathlessness, in a place the person would have chosen, with the people who matter around them. The evidence here shows how often systems fall short of that, and that the gap lies in services and planning, not in the person or their family.

How to read this data

Place-of-death figures come from death certificates. Some studies count only deaths where dementia was the underlying cause; others count any mention of dementia on the certificate, which captures more people. Whether dementia is recorded on a death certificate depends on certification practice, which differs between countries and has changed over time, so figures from different countries, years and definitions are not directly comparable. The English OHID series covers people aged 65 and over; the earlier English study covers those over 60. The global palliative care figures are WHO estimates covering all conditions, and the 2060 projection is a model built on WHO mortality forecasts and assumed symptom prevalence. The CASCADE, Coordinate My Care and Western Australian studies each describe one group of people in one place. Odds ratios are reported as odds, and all associations are observational unless they come from a trial. Evidence from low- and middle-income countries is very thin.

Use this data

Free to cite with attribution to Understanding Aging. Charities, carers' groups, campaigners and journalists are welcome to reuse these figures and charts. For the underlying figures as a spreadsheet, get in touch.

Important: This article summarises published research and statistics for general information. It is not medical, care or financial advice and cannot be used to diagnose, treat or predict anything about an individual. If you have concerns about your own or someone else's memory or health, please speak to a GP or another qualified health professional.

Sources

1. Sleeman KE, de Brito M, Etkind S, et al. The escalating global burden of serious health-related suffering: projections to 2060 by world regions, age groups, and health conditions. The Lancet Global Health, 2019. europepmc.org/article/MED/31129125
2. World Health Organization. Palliative care: fact sheet. WHO. who.int/news-room/fact-sheets/detail/palliative-care
3. Houttekier D, Cohen J, Bilsen J, et al. Place of death of older persons with dementia. A study in five European countries. Journal of the American Geriatrics Society, 2010. europepmc.org/article/MED/20398157
4. World Health Organization. Dementia: fact sheet. WHO, 2026. who.int/news-room/fact-sheets/detail/dementia
5. Mitchell SL, Teno JM, Kiely DK, et al. The clinical course of advanced dementia. New England Journal of Medicine, 2009. europepmc.org/article/MED/19828530
6. Nakanishi M, Yamasaki S, Sakai M, et al. Place of death from dementia as an underlying cause during the COVID-19 pandemic in Japan: a cross-sectional study from national death certificates. Palliative Care and Social Practice, 2023. europepmc.org/article/MED/37654728
7. Hidalgo-Andrade P, Unda-López A, Pastrana T. Place of death for people with dementia: a scoping review. PLOS ONE, 2026. europepmc.org/article/MED/41511992
8. Sleeman KE, Ho YK, Verne J, et al. Reversal of English trend towards hospital death in dementia: a population-based study of place of death and associated individual and regional factors, 2001–2010. BMC Neurology, 2014. europepmc.org/article/MED/24666928
9. Office for Health Improvement and Disparities. Place of death (care home, hospital, home): people with dementia aged 65 years and older, England, 2016–2023 (indicators 91893, 91894, 91895), based on Office for National Statistics data. Fingertips, accessed 2026. fingertips.phe.org.uk/api/all_data/csv/by_indicator_id
10. Xu W, Wu C, Fletcher J. Assessment of changes in place of death of older adults who died from dementia in the United States, 2000–2014: a time-series cross-sectional analysis. BMC Public Health, 2020. europepmc.org/article/MED/32522179
11. Zilkens RR, Spilsbury K, Bruce DG, Semmens JB. Clinical epidemiology and in-patient hospital use in the last year of life (1990–2005) of 29,884 Western Australians with dementia. Journal of Alzheimer’s Disease, 2009. europepmc.org/article/MED/19363266
12. Wiggins N, Droney J, Mohammed K, Riley J, Sleeman KE. Understanding the factors associated with patients with dementia achieving their preferred place of death: a retrospective cohort study. Age and Ageing, 2019. europepmc.org/article/MED/30806452
13. Lin YJ, Chen YJ, Ling DI, et al. Effectiveness of advance care planning for end-of-life outcomes in nursing home residents with dementia: a systematic review and meta-analysis. Journal of the American Medical Directors Association, 2026. europepmc.org/article/MED/41198038
14. Kim HD, Duberstein PR, Zafar A, et al. Home health care and place of death in Medicare beneficiaries with and without dementia. The Gerontologist, 2024. europepmc.org/article/MED/39392304
15. Tay R, Tan JY, Lim B, et al. Factors associated with the place of death of persons with advanced dementia: a systematic review of international literature with meta-analysis. Palliative Medicine, 2024. europepmc.org/article/MED/39092850
16. National Institute for Health and Care Excellence. Dementia: assessment, management and support for people living with dementia and their carers (NG97), recommendations 1.1.12 and 1.10.1. NICE, 2018. nice.org.uk/guidance/ng97/chapter/Recommendations