How Alzheimer's is diagnosed
Last reviewed November 16, 2026. OwieClear editorial team. Waiting for clinician review. Not medical advice. Talk with your doctor about your own care.
There is no single test. The diagnosis is assembled from what has changed, what testing shows, and what has been excluded.
The exclusion step matters more than people expect: several causes of memory loss are treatable, and some are fully reversible if found.
Related pages: Memory loss and dementia (Alzheimer's)
The bottom line
- Diagnosis combines history, examination, cognitive testing and investigations.
- Several causes of memory loss are reversible.
- Imaging and biomarkers can support the diagnosis.
- Blood biomarker tests have entered clinical use.
- How blood biomarker tests should be used outside specialist settings.
- Who should be tested for APOE status, and what to do with the result.
The story, step by step
What the evidence labels mean
What has changed
What happens
The account of what is different is the foundation of the diagnosis.
What the research shows
- What changed, when, and how fast. Gradual decline over months to years fits Alzheimer's; sudden change does not.
- Information from someone close to the person is essential, because insight is often affected and people compensate well in a short appointment.
- What has become difficult in practice: managing money, cooking a familiar meal, finding the way somewhere known, keeping appointments.
- A full medication review, including over-the-counter sleep aids and anticholinergic drugs, which commonly impair memory in older adults.
- Mood, sleep, alcohol, hearing and vision all affect cognition and all are modifiable.
Why it matters
Bring someone who sees the person regularly, and bring specific examples rather than general impressions. That is what makes the assessment useful.
What's still uncertain
- Early symptoms are easy to attribute to normal ageing in both directions.
Read more
- National Institute on Aging: Alzheimer's diseaseGovernment
- MedlinePlus: Alzheimer's diseaseGovernment
Next: Cognitive testing
What the tests show
What happens
Structured tests measure what conversation cannot.
What the research shows
- Short tests such as the MoCA or MMSE take about ten minutes and give a baseline to compare against later.
- They are screening tools, not diagnoses. Education, language, hearing and anxiety all affect scores.
- Fuller neuropsychological testing takes hours and maps which domains are affected, which helps distinguish Alzheimer's from other dementias.
- The pattern matters as much as the score: Alzheimer's typically affects recent memory and learning first, while other dementias start with behaviour, language or visual processing.
- Repeating the same test later is often more informative than a single result.
Why it matters
A single low score is not a diagnosis, and a normal score in someone clearly changed is not reassurance. The trajectory is what counts.
What's still uncertain
- Brief cognitive tests perform less well in people with low education or limited English.
Read more
- National Institute on Aging: Alzheimer's diseaseGovernment
- NINDS: DementiaGovernment
Next: Ruling out
The reversible causes
What happens
The part of the workup that can change everything.
What the research shows
- Vitamin B12 deficiency, thyroid disease, and in some cases syphilis or HIV testing.
- Depression, which in older adults can produce memory and concentration problems that improve with treatment.
- Delirium, a sudden confusion usually caused by infection, medication, dehydration or pain. It comes on over hours to days and is treatable.
- Medications: anticholinergics, sedatives, some bladder and allergy drugs, and combinations that are individually reasonable.
- Sleep apnoea, heavy alcohol use, and normal pressure hydrocephalus, which causes memory change with walking difficulty and bladder problems and can improve with a shunt.
- A brain scan to look for stroke, bleeding, or a tumour, which occasionally present as memory loss.
Why it matters
If memory changed over days or weeks rather than months, that is a different problem and it needs looking at urgently.
What's still uncertain
- Fully reversible causes are a minority, but they are the ones worth not missing.
Read more
- National Institute on Aging: Alzheimer's diseaseGovernment
- OwieClear: Sudden confusion in an older adult (delirium)OwieClear
Next: Scans and biomarkers
What technology adds
What happens
Investigations that can support or change the diagnosis.
What the research shows
- MRI or CT shows shrinkage patterns, strokes and other structural causes. Shrinkage in the memory regions supports Alzheimer's without proving it.
- Amyloid PET scans show amyloid in the brain. A negative scan makes Alzheimer's very unlikely; a positive scan is less decisive, because amyloid is also found in older people without symptoms.
- Tau PET tracks the protein that correlates better with symptoms, and is used mainly in specialist settings.
- Spinal fluid testing measures amyloid and tau directly and has been used for years.
- Blood tests measuring phosphorylated tau, particularly p-tau217, now perform well enough to have entered clinical use. They are a major practical change, because they are cheap and accessible compared with scans and spinal taps.
- Biomarkers do not replace the clinical picture. A positive result in someone with no symptoms does not mean dementia, and that is why testing people without symptoms outside research is not recommended.
Why it matters
Blood biomarkers are the biggest practical change in Alzheimer's diagnosis in decades, and they still answer only one question: is this pathology present.
What's still uncertain
- How to use blood biomarkers in primary care is still being worked out.
Read more
- National Institute on Aging: Alzheimer's diseaseGovernment
- NINDS: DementiaGovernment
Next: Genetics
APOE and the rest
What happens
What genetic testing can and cannot tell you.
What the research shows
- APOE4 is the main genetic risk factor for late-onset Alzheimer's. One copy raises risk; two copies raise it considerably more.
- It is a risk factor, not a verdict. Many people with APOE4 never develop Alzheimer's and many people with Alzheimer's have no APOE4.
- Routine testing is not recommended for people without symptoms, because the result does not change what anyone should do and can cause considerable distress.
- It has become relevant clinically for one reason: APOE4 carriers, especially those with two copies, have a higher risk of brain swelling and bleeding on the new anti-amyloid treatments, so testing is done before those drugs are started.
- Rare inherited forms, caused by mutations in APP, PSEN1 or PSEN2, cause early-onset Alzheimer's in families and are the situation where genetic counselling is genuinely indicated.
- Consumer genetic tests report APOE status. Receiving that result without counselling is a known source of harm.
Why it matters
Knowing your APOE status changes nothing you should be doing anyway, unless anti-amyloid treatment is being considered.
What's still uncertain
- How much APOE4 raises risk varies by ancestry and is not fully explained.