Understand
The difference between dementia and Alzheimer’s, the main types, the stages, what goes wrong in the brain, the competing theories of cause, and who is most at risk.
Doctors use “dementia” and “Alzheimer’s” as if they were the same thing. They are not. Knowing the difference helps you ask better questions, and it explains why two people with the same label can have very different years ahead of them.
Dementia is an umbrella word. It means memory, thinking or reasoning has declined enough to get in the way of daily life. It describes symptoms. It is not a single disease.
Alzheimer’s disease is one specific brain disease, and the most common cause of dementia. It is defined by two kinds of damage: clumps of a protein called amyloid between nerve cells, and twisted fibres of a protein called tau inside them. It causes an estimated 60–80% of dementia in the United States and 60–70% worldwide.
A comparison that helps many families: “dementia” is like “heart disease”, and “Alzheimer’s” is like a heart attack, one particular cause.
The type matters. It changes what to expect, which medicines are safe, and which treatments are even an option.
| Type | How common | What usually shows first | Worth knowing |
|---|---|---|---|
| Alzheimer’s disease | 60–80% of US cases | Short-term memory: repeating questions, forgetting recent talks, getting lost in familiar places | Slow, steady decline |
| Vascular dementia | 5–10% on its own; far more often mixed with Alzheimer’s | Planning, judgment, attention and slowed thinking, often more than memory | Caused by poor blood flow and strokes. Can start suddenly or worsen in steps. Many experts think it is underdiagnosed. |
| Dementia with Lewy bodiesDLB | About 4% in community studies, 7.5% in clinics | Detailed visual hallucinations, alertness that swings by the hour or day, acting out dreams, stiffness or slowness | Serious reactions to antipsychotic drugs in about half of people. Probably underdiagnosed. |
| Parkinson’s disease dementia | About 1 in 10 people with Parkinson’s in one UK clinic study | Thinking problems at least a year after Parkinson’s movement symptoms | Same protein as Lewy body dementia. If thinking problems come first or within a year, it is called DLB. |
| Frontotemporal dementiaFTD | About 50,000–60,000 people in the US | Personality and behavior (loss of empathy, poor judgment) or language. Memory is often fine early. | Usually starts at 45–65. About one-third of cases are inherited. |
| Mixed dementia | Over half of people diagnosed with Alzheimer’s | A blend of the above | In one autopsy study, 82% had Alzheimer’s changes plus at least one other type. |
| LATE | Mostly people over 80 | Memory loss that looks like Alzheimer’s | Tends to progress more slowly. Often occurs alongside Alzheimer’s. Can only be confirmed after death for now. |
LATE stands for limbic-predominant age-related TDP-43 encephalopathy. Figures on how common it is vary widely and we have not yet confirmed them. Other, rarer causes include Down syndrome-associated Alzheimer’s, Huntington’s disease and Creutzfeldt-Jakob disease. Normal-pressure hydrocephalus, which is treatable, is covered on the getting a diagnosis page.
Mild cognitive impairment (MCI) means thinking or memory problems that are noticeable and measurable, while the person still manages most daily life on their own. It affects about 12–18% of people aged 60 and over.
Many doctors and hospices use Dr. Barry Reisberg’s seven-stage scale. Its functional version (called FAST) gives typical lengths for each stage. Those times come from Alzheimer’s research in the 1980s, and real people vary a great deal. The scale was built for Alzheimer’s. Lewy body, vascular and frontotemporal dementia often follow a different order.
Typically about 7 years
Typically about 2 years
Typically about 1.5 years
Typically about 2.5 years in total
Often 1–1.5 years or more for each step within this stage
Families mostly use the scale for planning. For what to do at each point, see living with Alzheimer’s.
You may have read that the brain cleans itself during sleep. A 2013 mouse study reported faster amyloid clearance during sleep. A 2024 mouse study in Nature Neuroscience found the opposite: clearance was lower during sleep and anesthesia (Miao and colleagues, 2024). Sleep matters for brain health. The “sleep washes out amyloid” story is still argued over and should not be treated as settled.
Scientists do not agree on a single cause. Here are the main ideas, what supports each one, and what counts against it. Most researchers now see late-life Alzheimer’s as having several causes that work together.
The idea: amyloid builds up first and sets off tau tangles, inflammation and the death of nerve cells.
For it: People with Down syndrome carry an extra copy of chromosome 21, which holds the amyloid gene (APP). More than 90% of them develop Alzheimer’s in their lifetime. Drugs that clear amyloid slow decline a little. In the lecanemab trial, 1,795 people were followed for 18 months, and decline was 27% slower, a difference of 0.45 points on an 18-point scale. The inherited early-onset genes are often said to all affect amyloid; we have not yet confirmed that from a primary source.
Clarity AD: lecanemab (Leqembi) in early Alzheimer’svan Dyck CH et al., New England Journal of Medicine 2023 · 1,795 people with mild cognitive impairment or mild dementia and confirmed amyloid · 18 months
Promising4/5Industry
Against it, or complicating it:
Some researchers argue tau, not amyloid, is the main driver of damage, because where tau spreads seems to match symptoms more closely. The 2024 criteria do use tau scans for staging. We have not yet confirmed the key studies behind the stronger claim, so we list it here as an argument, not a finding.
This theory holds that the brain’s immune cells stay switched on and cause harm, driving the disease rather than just reacting to it. It is an active research area. We have not yet confirmed the key studies, so we give no numbers.
Poor blood flow and damage to small vessels may start or speed up the disease. Two facts support it: mixed Alzheimer’s and vascular damage is very common, and high blood pressure, diabetes and high LDL cholesterol are all on the Lancet Commission’s list of risk factors.
This is the idea that the brain becomes resistant to insulin. The label “type 3 diabetes” comes from a 2008 paper by Suzanne de la Monte and Jack Wands. A 2024 review in Ageing Research Reviews found shared mechanisms and reported that people with diabetes have a 65% higher chance of Alzheimer’s. It concluded the label is defensible but still debated.
So far, treatments built on this idea have not worked. Insulin sprayed into the nose did not help. The semaglutide trials (the same drug family as Ozempic), with 3,808 people, did not slow Alzheimer’s either, although some blood markers improved.
Craft S et al., JAMA Neurology 2020 Read the study
Result No benefit for thinking or daily function compared with placebo. This was a direct test of the “type 3 diabetes” idea, and it did not work in this form.
Who paid: National Institute on Aging (reported; not yet confirmed from the paper)
Conflicts: Not reported
evoke and evoke+: semaglutide (the Ozempic family) as a pillNovo Nordisk topline announcement, 24 November 2025 (company release) · 3,808 people with early Alzheimer’s · 2 years
Didn’t work4/5Industry
Diabetes is a real risk factor. “Alzheimer’s is just diabetes of the brain” goes further than the evidence.
This theory says viruses such as herpes simplex (the cold-sore virus) or other germs trigger amyloid as a defense.
For it: in Wales, people who were just old enough to get the shingles vaccine had 20% fewer new dementia diagnoses over 7 years than people just too old, in a study of 282,541 adults. The effect was stronger in women. Against it: the VALAD trial gave the antiviral valacyclovir to 120 people with early Alzheimer’s and the cold-sore virus for 78 weeks, and found no benefit.
Shingles vaccine and dementia: the Wales natural experimentEyting M et al., Nature 2025 · 282,541 adults in Wales · 7 years
Promising4/5Public / nonprofit
VALAD: the herpes antiviral valacyclovir in early Alzheimer’sDevanand DP et al., JAMA 2025 · 120 people with early Alzheimer’s who carried herpes simplex virus · 78 weeks
Harmful4/5Public / nonprofit
The vaccine finding is strong observational evidence but not proof. Treating the virus after symptoms start has not helped so far.
This idea says the decline of the cell’s energy factories (mitochondria) with age comes first. It has less direct evidence in people than the other theories.
Age is the biggest risk factor by far.
| Age | Share | About 1 in… |
|---|---|---|
| 65–74 | 5.2% | 19 |
| 75–84 | 13.8% | 7 |
| 85 and over | 35.8% | 3 |
APOE is a gene everyone carries in some form. The ε4 version (APOE4) is the most important common risk gene for Alzheimer’s. Here is the absolute risk, for people aged 65–69, of developing dementia by their early-to-mid 80s.
| Copies of APOE4 | Risk by early-to-mid 80s | Out of 100 people |
|---|---|---|
| None | 5–7% | 5 to 7 |
| One | 15–16% | 15 to 16 |
| Two | 31–40% | 31 to 40 |
Rare changes in three genes (APP, PSEN1 and PSEN2) cause Alzheimer’s that usually starts in a person’s 30s to 50s. These families are often said to make up less than 1% of cases, with each child of a carrier having a 50% chance of inheriting the gene. We have not yet confirmed those figures from a primary source. If Alzheimer’s started very young in several relatives, ask about genetic counseling.
Nearly two-thirds of the 7.4 million Americans with Alzheimer’s are women, about 4.5 million people. A 65-year-old woman has about a 1 in 5 lifetime risk.
In 2024 the Lancet Commission on dementia listed 14 risk factors people can change (Livingston and colleagues, Lancet 2024). It estimated that about 45% of dementia could in theory be prevented or delayed if all 14 were removed, up from 40% with 12 factors in 2020.
| Stage of life | Risk factors |
|---|---|
| Early life | Less education |
| Midlife | Hearing loss; high LDL cholesterol (new in 2024, about 7% of dementia); depression; head injury; physical inactivity; diabetes; smoking; high blood pressure; obesity; heavy drinking |
| Later life | Social isolation; air pollution; untreated vision loss (new in 2024, about 2% of dementia) |
We have confirmed the Commission’s share for two factors only (LDL cholesterol and vision loss), so we give no share for the others yet.
“45% preventable” is a population estimate from a model. It does not mean any one person can cut their own risk by 45%. It is not a reason to blame your mother, or yourself, for her diagnosis. For what the trials show about lowering risk, see lowering your risk.
| Measure | Figure |
|---|---|
| Americans 65+ with Alzheimer’s dementia | 7.4 million |
| Projected by 2060 without a breakthrough | 13.8 million |
| Official Alzheimer’s deaths, 2024 | 116,022 |
| Rank as a cause of death | 6th overall, 5th at 65+ |
| Rise in Alzheimer’s deaths, 2000–2024 | 134% |
| Cost of care for people 65+, 2026 | $409 billion |
| Unpaid family caregivers, 2025 | More than 12 million |
| Hours of unpaid care, 2025 | 19.6 billion |
| Value of that unpaid care | $446.3 billion |
Death certificates undercount dementia, so the true number of deaths is higher.
| Measure | Figure |
|---|---|
| People living with dementia (2021) | 57 million |
| New cases each year | Nearly 10 million |
| Share caused by Alzheimer’s | 60–70% |
| Rank as a cause of death | 7th |
| Global cost (2019) | US$1.3 trillion |
The Alzheimer’s Association says people live 4–8 years on average after an Alzheimer’s diagnosis, and some live 20 years. The largest review so far, from 2025, shows that age at diagnosis matters more than almost anything else.
Brück et al., BMJ 2025;388:e080636 Read the study
Result Age at diagnosis mattered most. Average life expectancy after diagnosis was 8.9 years for a woman diagnosed at 60 and 2.2 years for a woman diagnosed at 85. For men it was 6.5 years at 60 and about 2 years at 85. People with Alzheimer’s lived up to 1.4 years longer than people with other types. 13% had moved to a nursing home within 1 year, 35% within 3 years and 57% within 5 years. These are averages; any one person can live much longer or shorter.
Who paid: Not yet confirmed.
Conflicts: Not reported
If your mother was diagnosed in her 80s, it is reasonable to plan on a shorter horizon than a family whose mother was diagnosed at 65. The first 30 days page covers the planning that helps in either case.
