Most People With Dementia Worldwide Have No Diagnosis: The Global Gap

Around the world, most people living with dementia have never been given a diagnosis. Among the countries that could report a figure to the World Health Organization in 2024, the median diagnostic rate was 26%, and in the low- and middle-income countries that reported it was just 6%^[1].

A diagnosis is not a label for its own sake. For most people it is the door to everything else: an explanation for changes a family has been struggling to make sense of, information and treatment where it helps, support for carers, adjustments at work, and the chance to plan ahead while the person living with dementia can still lead those decisions. Without one, people are often left to cope with symptoms nobody has named, outside the services built to help them. This study brings together what is known about how many people with dementia have a diagnosis, in which countries, how the figure is measured, and why the gap persists. It is part of our Dementia & Health research library.

Most countries cannot yet say how many people with dementia have a diagnosis

In 2017 the World Health Assembly adopted a global action plan on dementia. One of its targets was that, by 2025, at least half of all countries would diagnose at least half of the people estimated to be living with dementia. WHO’s own progress report, published in December 2024, is frank: the 2025 targets will not be met, and none is on track^[1].

The first obstacle is simply knowing. Seventy-eight WHO Member States took part in the 2024 round of the Global Dementia Observatory, WHO’s monitoring system, up from 62 in the first round of 2018–2020. Only 37 of those 78 countries (47%) were able to report a dementia diagnostic rate. Among them, the median rate was 26%. Across 17 high-income countries it was 62%; across 14 low- and middle-income countries it was 6%^[1]. The Observatory’s own web page, which does not state the year its figure refers to, puts the share lower, at one in three participating countries^[4].

These are medians of rates that countries calculated in their own ways, from their own records and their own estimates of how many people have dementia, so they are a broad picture rather than a league table. But the direction is unmistakable. In 2021, over 60% of the 57 million people living with dementia worldwide lived in low- and middle-income countries^[1], and that is where a diagnosis is rarest and where the systems to count diagnoses are weakest. WHO’s report also notes that community dementia services in low- and middle-income countries are often restricted to cities: only 6 of 43 such countries reported rural access, against 26 of 35 high-income countries^[1].

The report put a draft decision to WHO’s Executive Board to extend the action plan to 2031^[1]. That would keep an agreed international benchmark for diagnosis in view: half of countries diagnosing at least half of the people living with dementia.

When researchers assess people directly, most dementia turns out to be undiagnosed

National diagnostic rates depend on records. A different way to see the gap is to assess a sample of older people for dementia directly, then check whether those who have it have been diagnosed. Studies of this kind consistently find that most dementia goes unrecognised.

A 2017 systematic review and meta-analysis in BMJ Open pooled 23 such studies, in community and residential-care settings, and estimated that 61.7% of dementia was undetected (95% confidence interval 55.0% to 68.0%); in the 15 community-only studies the figure was 63.7% (54.8% to 71.8%). Under-detection was higher in China and India than in Europe and North America, and higher in the community than in residential and nursing care. It was associated with younger age (under 70), being male, and diagnosis being left to a general practitioner^[2].

More recent cohort data tell the same story across very different health systems. A 2026 cross-national study used the US Health and Retirement Study (2016, 9,539 people) and the Brazilian Longitudinal Study of Aging (ELSI-Brazil, 2015–16, 3,603 people). Among older adults identified as having dementia, 76.1% in Brazil and 45.1% in the United States had no diagnosis. In both countries, undiagnosed dementia was associated with higher mortality over four years than no dementia^[5].

A diagnosis on record is also not the same as a diagnosis that has been shared. A 2018 US study of 585 adults aged 65 or over with probable dementia in the National Health and Aging Trends Study (2011), linked to their Medicare claims, found 58.7% were either undiagnosed (39.5%) or had a diagnosis they were not aware of (19.2%)^[6]. Hispanic people had higher odds of being undiagnosed (odds ratio 2.48, 95% CI 1.19 to 5.14), as did people who attended medical visits alone (odds ratio 1.98, 1.11 to 3.51); people with at least a high-school education had lower odds (0.54, 0.32 to 0.91)^[6]. Those are associations, not proof of cause, but they point to where the gap falls: on people with fewer advocates beside them and less access to the system, not on anything people with dementia did or did not do. Differences by ethnicity are covered in more depth in our study of ethnic inequalities in dementia.

England, a worked case: about two in three diagnosed, measured against an estimate

England is one of very few countries to publish a national dementia diagnosis rate routinely, which makes it a useful worked case, not a typical one. A 2026 review of England’s dementia data notes that England is one of only eight countries that estimate the prevalence of dementia, and that none of the other national registries it compared publishes a routine national diagnosis rate of this kind^[7].

On 30 June 2026, 516,230 people in England had a recorded diagnosis of dementia in their GP records, 3,095 more than at the end of March. For people aged 65 or over, the diagnosis rate was 66.2%, down slightly from 66.3% on 31 March 2026^[3]. Of those on the register, 34,783 (6.7%) had received their diagnosis before the age of 65^[3]. A review of the data reported a rate of 66.3% in September 2025 against 65.5% a year earlier, close to the national ambition that two-thirds of people in England with dementia should have a diagnosis^[7].

The rate is a ratio, and its denominator is a model. It divides the number of people aged 65 or over with a recorded diagnosis by the number expected to have dementia, estimated from the Cognitive Function and Ageing Study II (CFAS II)^[7]. The review’s authors point out that CFAS II was conducted over 15 years ago and that the estimate is not adjusted for ethnicity, deprivation or rurality, all of which influence dementia risk. As a result, apparent differences in the rate between areas may partly reflect differences in local populations rather than in how well services diagnose^[7].

A 2025 study took a different route. Using GP records linked to hospital records from 1997 to 2018 and a capture–recapture method, which estimates how many cases are missed from the overlap between two sources, it found the share of estimated cases that had been identified rose from 42.4% in 1997 to 84.4% in 2018. The authors suggest underlying prevalence may be lower than previously estimated, but say this needs replication^[8]. The two figures measure different things and should not be set side by side as if they were the same statistic. Together they show how much a diagnosis rate depends on its estimate of the people it cannot see. How different countries record dementia is the subject of our study on measuring dementia.

Why diagnoses are missed: stigma, the “normal aging” myth and thin services

The reasons behind the gap are well documented, and few of them lie with people living with dementia. Alzheimer’s Disease International’s World Alzheimer Report 2021, on the journey through diagnosis, surveyed 1,111 clinicians, 2,325 people living with dementia and carers, and more than 100 national Alzheimer and dementia associations^[9]. Among its findings:

• 46% of people living with dementia and carers identified fear of diagnosis and stigma as barriers to diagnosis.
• 62% of healthcare practitioners still believed that dementia is a normal part of aging.
• 35% of carers worldwide said they had hidden the diagnosis of a person with dementia.
• 90% of clinicians reported additional delays or waits for diagnosis because of COVID-19.

These are survey responses from people who chose to take part, not representative national samples, but they describe a pattern that the epidemiology supports. When the clinicians people see first regard memory loss as an expected part of growing older, symptoms go unexplored. The 2017 meta-analysis found under-detection was more likely where diagnosis rested with general practitioners^[2], which speaks to the time, training and specialist back-up that primary care is given rather than to individual doctors. Where specialist and community services are concentrated in cities, as WHO found in most low- and middle-income countries that reported^[1], distance becomes a barrier in itself.

The quality of a diagnosis matters as well as its existence. In England’s primary-care data, 28.2% of people on the register (143,831) had dementia recorded without a type specified, compared with 21% in Sweden’s SveDem register and 15% in Norway’s NorCog register^[7]. Knowing the type can shape care and eligibility for treatment. Long waits for specialist assessment are a further part of the picture, covered in our study on waiting for a diagnosis.

What the evidence asks of us

The figures point to a small number of clear tasks for governments, health services, commissioners and funders.

• Count, so the gap can be seen. Fewer than half the countries taking part in WHO’s Global Dementia Observatory could report a diagnostic rate in 2024^[1]. With the global action plan proposed for extension to 2031, every country can work towards publishing one, and towards the plan’s benchmark of diagnosing at least half of the people living with dementia.
• Keep the yardstick current. A diagnosis rate is only as good as its estimate of expected prevalence. England’s rests on a study conducted more than 15 years ago, not adjusted for ethnicity, deprivation or rurality^[7]. Refreshing such estimates costs money, but without it, areas cannot tell whether they are serving their own populations well.
• Invest where people first ask for help. Under-detection is highest in the community and where diagnosis rests with primary care^[2], and rural access to dementia services is rare in most low- and middle-income countries that report^[1]. Training, time and specialist support for primary and community care are the levers.
• Tackle stigma and the “normal aging” myth, including among professionals. When most practitioners surveyed believe dementia is a normal part of aging and nearly half of families cite fear and stigma as barriers^[9], awareness work is part of the diagnostic pathway, not an extra.
• Measure whether a diagnosis is shared and acted on. A diagnosis in a record that the person and family do not know about opens no doors^[6]. Diagnosis rates should sit beside measures of whether people are told, given a dementia type, and offered support afterwards, the subject of our study on post-diagnostic support.

None of this asks more of people living with dementia. It asks health systems to find the people they are currently missing, and to make the diagnosis worth having when they do.

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

We compiled this evidence because a diagnosis is where support for people living with dementia and their families usually begins, and too many people never reach it. Dementia charities, carers’ groups and campaigners argue for better diagnosis every day; we want the international figures, and their limits, in one place, cleanly sourced and ready to put in front of a minister, a commissioner or a journalist. The case belongs to people affected by dementia.

How to read this data

Three kinds of figure appear here, and they are not interchangeable.

• National diagnosis rates (England, and the rates countries report to WHO) divide recorded diagnoses by a modelled estimate of how many people are expected to have dementia. England’s uses CFAS II for people aged 65 or over^[7]; other countries use their own estimates, so rates are not directly comparable between countries. England’s current series is not comparable with figures from before October 2022, and from July 2026 it is published quarterly^[3].
• Cohort and survey studies (the Health and Retirement Study, ELSI-Brazil, the studies pooled in the 2017 meta-analysis) assess people directly and compare with their records. Each uses its own definition of dementia and of a diagnosis, and the meta-analysis pools studies of different dates and settings^[2][5].
• Surveys of opinion and experience (the World Alzheimer Report 2021) describe self-selected respondents, not whole populations^[9].

The WHO medians cover only the 37 countries able to report a rate in 2024^[1], so the countries with the least data are largely missing. Associations in observational studies, such as the higher odds of being undiagnosed among Hispanic older adults in the US^[6], do not show cause. The latest England figure here is for 30 June 2026; the next release, for September 2026, is due on 22 October 2026^[3].

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. World Health Organization. Global action plan on the public health response to dementia 2017–2025: report by the Director-General (EB156/36). WHO Executive Board, 2024. apps.who.int/gb/ebwha/pdf_files/EB156/B156_36-en.pdf
2. Lang L, Clifford A, Wei L, et al. Prevalence and determinants of undetected dementia in the community: a systematic literature review and a meta-analysis. BMJ Open, 2017. europepmc.org/article/MED/28159845
3. NHS England. Primary Care Dementia Data, June 2026. NHS England Digital, 2026. digital.nhs.uk/…/primary-care-dementia-data/june-2026
4. World Health Organization. Global Dementia Observatory: diagnosis, treatment and care. WHO Global Health Observatory, accessed 2026. who.int/data/gho/…/diagnosis-treatment-and-care
5. Aliberti MJR, Avelino-Silva TJ, Covinsky KE, et al. Undiagnosed dementia and mortality among older adults in the United States and Brazil: a cross-national cohort study. Alzheimer’s & Dementia, 2026. europepmc.org/article/MED/42056069
6. Amjad H, Roth DL, Sheehan OC, et al. Underdiagnosis of dementia: an observational study of patterns in diagnosis and awareness in US older adults. Journal of General Internal Medicine, 2018. europepmc.org/article/MED/29508259
7. Hazan J, Howard R, Isaacs JD, et al. England’s dementia diagnosis landscape: strengths, limitations, and future directions. International Journal of Geriatric Psychiatry, 2026. europepmc.org/article/PMC/PMC13327468
8. Mukadam N, Marston L, Flanagan K, et al. Estimating undiagnosed dementia in England using capture recapture techniques. BMC Geriatrics, 2025. europepmc.org/article/MED/39748300
9. Alzheimer’s Disease International. World Alzheimer Report 2021: Journey through the diagnosis of dementia. Alzheimer’s Disease International, 2021. alzint.org/resource/world-alzheimer-report-2021