Healthy Aging Secrets
Evidence-led living for your second fifty years

Brain & Memory

Normal forgetting, mild cognitive impairment, or dementia: how the line is drawn

Losing your keys is not a symptom. Losing the ability to work out what keys are for is a different matter. The clinical distinction rests on function, not on how often you forget.

Senior man in a jacket reads a book indoors in a calm setting.
Senior man in a jacket reads a book indoors in a calm setting. · Photo via Pexels
Medical disclaimer. This site publishes health journalism, not medical advice. Read the full disclaimer.

Almost everyone over sixty has stood in a doorway wondering what they came in for and felt a small cold moment about it. The reassuring fact is that this particular experience — the interrupted intention — is common at every adult age and is a feature of attention rather than memory.

Some cognitive change with age is genuinely normal and reasonably well characterised. Processing speed declines steadily from early adulthood. Working memory capacity falls modestly. Word-finding becomes slower, producing the familiar tip-of-the-tongue experience. Learning new information takes more repetitions.

What is not normal ageing: forgetting entire recent events rather than details of them, repeating the same question within a conversation, getting lost somewhere familiar, or a decline in the ability to plan and sequence a task you have done for forty years.

The distinction that matters clinically

Subjective cognitive decline is when you notice a change and objective testing does not. It is common, frequently related to anxiety, depression, poor sleep or medication, and is associated with a modestly elevated risk of later decline — enough to warrant follow-up, not enough to warrant alarm.

Mild cognitive impairment means objective testing shows performance below what would be expected for your age and education, but daily function is preserved. You still manage your finances, your medications and your household, possibly with more effort or more lists than before.

Dementia means the cognitive decline is sufficient to interfere with independent daily function. That is the whole distinction. Not severity of forgetting, not test scores — whether the person can still run their own life.

MCI is not simply early dementia

Roughly 10–15 per cent of people with MCI progress to dementia each year, which means most do not in any given year. A meaningful proportion — estimates commonly range from 15 to 30 per cent — revert to normal cognition on retesting, usually because the impairment was driven by something reversible. That possibility is the main reason to be assessed rather than to wait and see.

The reversible causes worth ruling out

This is the practical argument for getting an assessment rather than quietly worrying for two years. A meaningful minority of people presenting with cognitive symptoms have something treatable, and the list is unglamorous.

  • Depression. In older adults it frequently presents with cognitive slowing and poor concentration rather than low mood, and it is the single most common mimic. Treating it can substantially restore function.
  • Medication. Anticholinergic burden, benzodiazepines, opioids, and simple polypharmacy. This is the first thing a good clinician reviews.
  • B12 and folate deficiency, and hypothyroidism. Cheap blood tests, occasionally dramatic answers.
  • Obstructive sleep apnoea. Chronic fragmented sleep produces a convincing impression of cognitive decline and is treatable.
  • Alcohol. Consistently under-reported and a substantial contributor.
  • Hearing and vision loss, which degrade test performance and real-world function alike.
  • Normal pressure hydrocephalus — the triad of gait disturbance, urinary incontinence and cognitive change. Rare, and occasionally reversible with shunting.

What an assessment actually involves

Usually a structured history — ideally with someone who knows the person well, because insight is often the first thing to go — a cognitive screening instrument, blood tests, and imaging in most cases.

The brief screening tools are useful and limited. The Mini-Mental State Examination is the best known and is relatively insensitive to early and to executive-type impairment. The MoCA is more sensitive but is affected by education and by language; both were designed as screening instruments, not diagnostic ones. A score is a prompt for further assessment, never a diagnosis.

Neuropsychological testing, where available, is considerably more informative because it profiles which domains are affected. Memory-predominant impairment, executive-predominant impairment and language-predominant impairment point toward different underlying pathologies.

Why the subtype matters

Dementia is a syndrome, not a disease. Alzheimer's is the commonest cause, but vascular contributions are extremely common — and mixed pathology is probably the norm rather than the exception in the over-eighties.

Getting the subtype right changes management in concrete ways. Dementia with Lewy bodies, characterised by fluctuating attention, visual hallucinations and parkinsonism, carries a severe sensitivity to antipsychotic drugs that can be dangerous if unrecognised. Frontotemporal dementia typically presents with personality and behaviour change rather than memory loss, often in people in their fifties and sixties, and is frequently misdiagnosed as a psychiatric condition for years. Vascular contributions shift emphasis onto blood pressure and stroke prevention.

What has changed recently, and what has not

Blood-based biomarkers for Alzheimer's pathology — particularly phosphorylated tau assays — have moved quickly and are entering clinical use. They will make it far easier to establish whether amyloid pathology is present without a lumbar puncture or PET scan.

The anti-amyloid antibodies, lecanemab and donanemab, have shown statistically significant slowing of decline in early Alzheimer's disease. Whether the size of that slowing is clinically meaningful to patients is genuinely debated, the drugs carry a real risk of brain oedema and microhaemorrhage requiring monitoring, and they apply only to a narrow early-stage population with confirmed amyloid.

What has not changed is that the largest determinants of how a person with dementia actually lives are non-pharmacological: whether the environment is adapted, whether the carer is supported, whether hearing and vision are corrected, whether pain is being treated, and whether the medication list has been pruned.

When to seek assessment

If the person themselves is worried, that is enough. If a family member has noticed a change over six months in memory, judgement, language or personality, that is more than enough — informant concern is a better predictor than self-report, precisely because insight is affected.

The instinct to wait, to avoid the appointment, to see whether it settles down, is entirely understandable and it is the wrong instinct. A meaningful fraction of what presents this way is treatable, and none of it is treated by waiting.

dementiamild cognitive impairmentmemorydiagnosis
Dr. Helen Marsh
Medical Editor, Healthy Aging Secrets

Helen is a geriatrician who spent nineteen years on hospital wards before moving into health writing. She reads the primary literature so readers do not have to, and she is unusually blunt about what the evidence does not show.

More from Dr. →

Also by Dr. Helen Marsh

Preventive Care

Hearing loss is a brain problem, and waiting makes it worse

The average person waits around a decade between noticing hearing difficulty and doing something about it. That delay is not neutral — the auditory system reorganises in the meantime.

Dr. Helen Marsh··4 min read

Heart & Metabolic

Heart failure: the symptoms that get mistaken for getting older

Breathlessness on the stairs, swollen ankles and needing an extra pillow are the three commonest presentations, and all three are routinely attributed to age. Half of cases involve a heart that pumps normally.

Dr. Helen Marsh··4 min read

Living Well

Caring for a spouse: the health risks the carer takes on

Spousal carers have measurably worse health outcomes than matched non-carers, and the effect is concentrated in those who report strain. Most of the protective measures are refused as luxuries.

Ingrid Solberg··4 min read