Most people who worry about their memory do not have dementia. That is worth saying first, because the worry itself is common, distressing, and frequently attached to something else entirely — often something treatable.
It is also not a reason to dismiss the question. Some memory change needs assessing, and the difference between the two is more knowable than most people expect. Here is what normal age-related change looks like, what doesn't fit that pattern, and the list of reversible causes worth ruling out before anything else.
What normal age-related memory change actually is
Memory does change with age, and the change has a characteristic shape. It is mostly about retrieval speed and efficiency rather than loss of the information itself.
- Names arrive late. You know exactly who the person is, you can picture their house and their dog, and the name surfaces twenty minutes later. The record was intact; the index was slow.
- You need more cues. Prompted recall works far better than unprompted. "Was it Thursday or Friday?" gets an answer that "when was it?" didn't.
- Divided attention costs more. Being interrupted mid-task derails you more than it used to. The car keys go down while you're answering the door and the moment simply wasn't encoded.
- Retracing works. You walk back through your steps and find the keys, the thread, the word.
- New learning takes longer, but it still happens. Acquiring a new phone or a new system is slower, not impossible.
- Nothing gets worse, week to week. This is the most reassuring feature of the lot. Normal ageing is a slope measured in years, not a decline you can track month by month.
And the single most reassuring sign: you are the one who noticed. Being troubled by your own memory, monitoring it, and being able to give specific examples requires the very faculties people are afraid of losing.
What doesn't fit that pattern
These are the things worth taking to a doctor rather than watching. None of them means dementia on its own — they mean assessment.
| Typical of ageing | Worth assessing |
|---|---|
| Forgetting a name, recalling it later | Not recognising a close family member |
| Forgetting part of a conversation | Not remembering the conversation happened, and repeating it |
| Misplacing keys and retracing | Finding objects in places they could not plausibly have been put |
| Taking longer to find a word | Substituting wrong words, or losing the thread of a sentence mid-way |
| Momentary disorientation in an unfamiliar place | Getting lost somewhere familiar, or on a routine route |
| Making an occasional poor decision | A pattern of uncharacteristic judgement — finances, safety, trust |
| Being aware of, and bothered by, the changes | Not noticing, while others around you clearly have |
| Difficulty with a new device | Difficulty with a long-familiar task — a recipe cooked for years, a well-known route |
Two patterns on the right-hand side carry particular weight. Losing function in something previously automatic matters more than difficulty with anything new. And a family member being more worried than you are is, statistically, a more useful signal than the reverse.
There is also a middle category clinicians use: mild cognitive impairment, meaning a measurable deficit on testing that is greater than expected for someone's age, while day-to-day independence remains intact. It is not dementia, a proportion of people with it remain stable or improve, and it is one reason getting assessed is more informative than guessing.
The reversible causes — rule these out first
This is the most useful part of this article, because several of these are common, all are treatable, and every one of them can produce a convincing impression of memory decline.
Depression. The most important item on this list. Depression impairs concentration, motivation and recall, and in older adults it can present so much like dementia that the pattern has its own clinical description. It is also treatable, and the cognitive symptoms substantially improve when it is. Low mood that arrived before or alongside the memory problem is a strong reason to raise both together.
Grief. Worth separating from depression, because it is so often dismissed. Bereavement produces genuine, measurable difficulty with concentration and memory — the research on what people call "grief brain" is more substantial than you'd expect — and it is a common reason someone in their fifties suddenly can't hold a thought.
Anxiety and chronic stress. Sustained high arousal consumes the attention that encoding requires. If your mind is elsewhere, nothing is being filed — which reads exactly like forgetting.
Sleep. Memory consolidation happens during sleep, so poor sleep produces genuine memory impairment rather than the appearance of it. Worth separating two very different problems here: insomnia, and undiagnosed sleep apnea, which fragments sleep hundreds of times a night without the person remembering any of it. Anyone who sleeps a full night and wakes unrefreshed should screen for the second.
Thyroid dysfunction. An underactive thyroid classically produces slowed thinking, poor concentration, low mood and fatigue. It is a single blood test and it is treatable — we've covered how thyroid problems affect the brain separately.
B12, folate and iron deficiency. B12 deficiency in particular can cause cognitive impairment, and it is more common in older adults, in vegetarians and vegans, and in people on long-term acid-suppressing medication. All are cheap to test. Here is the short list worth asking for.
Alcohol. Both directly, and by depleting thiamine. It is also the item most consistently under-reported at appointments, which is unhelpful, because it is one of the more reversible. The "moderate drinking is good for you" story has not aged well either — the newer research is worth reading.
Medication. Several widely-used drug classes affect memory and attention, and the effect is often mistaken for ageing: sedating antihistamines, some bladder and overactive-bladder medications, benzodiazepines and Z-drugs, some antidepressants, and opioids. The combined anticholinergic load across several medicines matters more than any one of them. Do not stop a prescribed medication on the strength of a web page — bring the full list, including anything over-the-counter, to whoever prescribes it and ask the question there.
Perimenopause. Memory and word-finding complaints are common through the menopausal transition and are frequently misread by the person experiencing them as early dementia. Worth naming, because the fear it generates is out of proportion to what is happening — the actual mechanism is here.
What an assessment actually involves
Worth knowing, because the fear of the process keeps people away from it.
A first appointment is usually a conversation: what you've noticed, when it started, whether it has changed, what else is going on, and your full medication list. It often includes a short pen-and-paper cognitive screen taking ten to fifteen minutes. Bloods are commonly ordered to cover the reversible causes above. Depending on that picture, next steps may involve fuller neuropsychological testing or imaging.
Two things people expect that are usually wrong. It is not a pass/fail test — a screening score is a data point read alongside everything else. And it is not a one-way door: a large share of these assessments identify something other than dementia, which is precisely the outcome that makes them worth having.
When not to wait
Go sooner rather than later if: the change has developed over weeks rather than years · there is confusion that comes and goes, especially with a fever or a recent infection · personality or behaviour has changed noticeably · there is new difficulty with language · there are physical symptoms alongside it, such as weakness, numbness, unsteadiness or headache · or it follows a head injury.
Rapid onset in particular points away from the slowly progressive causes and towards something that may need treating now.
Where we sit
No supplement prevents, treats or slows dementia, and there is no supplement on the Lancet Commission's list of 14 modifiable risk factors. If you are worried about your memory, the useful step is an appointment and a blood test, not a product — and we would rather say that plainly than sell you something while you wait.
What does have a genuine evidence base is the risk-factor list itself: blood pressure, lipids, blood sugar, hearing and vision, alcohol, movement and social contact. The full breakdown of all fourteen factors is here, and what the headline 45% figure actually means is here. Hearing in particular is more actionable than most people realise.
This article is not medical advice and cannot assess anyone's memory. If you or someone close to you is worried, that is a sufficient reason to see a doctor — the outcome is more often reassurance or a treatable cause than it is a diagnosis of dementia.