When memory changes are worth checking

Families usually wait too long on this one, and for an understandable reason. Raising it feels like accusing someone of losing their mind, and the fear underneath is that an assessment can only confirm the worst.

Both of those assumptions are wrong in ways the evidence is fairly clear about.

"It is just age" is not the safe assumption

The American Academy of Neurology's position is direct, as quoted in StatPearls: "The AAN recommends that all memory concerns be assessed for MCI and not assumed to be due to normal aging."

MCI is mild cognitive impairment, defined as "a decline in one or more cognitive domains, which is both subjectively and objectively observable" where "this decline does not interfere with the individual's ability to perform daily activities independently".

That last clause is the line between MCI and dementia. In MCI, something has measurably changed but the person still runs their own life. In dementia, the change has begun to take that independence away.

The part nobody expects

Here is the reason waiting is the expensive choice. A meaningful proportion of people assessed for memory changes turn out to have something treatable causing them.

StatPearls lists the reversible causes clinicians are meant to look for first, and the list is unglamorous: medications, particularly opioids, benzodiazepines, anticholinergics and some blood pressure drugs; depression, which "can present as MCI with reversible cognitive deficits"; thyroid dysfunction and B12 deficiency, described as easily treated metabolic causes; sleep apnoea, where "Continuous Positive Airway Pressure (CPAP) treatment may improve cognitive symptoms"; and infections and dehydration.

Read that list as a family member and something reframes. A parent who has become forgetful and slow might have a thyroid problem, a vitamin deficiency, untreated sleep apnoea, depression, or an unfortunate combination of prescriptions. None of those improve by waiting, and all of them are found by the assessment people are avoiding.

Reversion is real and documented. StatPearls reports that roughly 15 to 20 percent of people with MCI show improved cognition within one to two years.

And the trajectory is not what most people assume

The fear driving the delay is that a diagnosis means an inevitable slide. The figures are more mixed than that. Annual progression from MCI to dementia ranges from under 5 percent to somewhere between 12 and 20 percent, and StatPearls notes that "40-70% of patients with MCI may not progress to dementia for over a decade".

So MCI is not a euphemism for early dementia. For a substantial share of people it is a stable state, and for some it improves.

What is worth mentioning to a clinician

Specific examples are far more useful than the word "forgetful".

• • Repeating the same question within a single conversation, rather than forgetting a name.
• • Difficulty with a familiar sequence: a recipe they have cooked for years, paying a bill, the route to a regular destination.
• • Missed appointments or medication doses that would not have been missed a year ago.
• • Withdrawing from things they used to enjoy, which can be cognitive, or depression, or both.
• • Whether it came on gradually over months, or noticeably over days. Sudden changes point somewhere different and more urgent.
• • Everything on the medication list, including anything started in the last year.

That timing question matters most. Gradual change over a year or more is the MCI conversation. Confusion that arrived over days is a different and more urgent problem, and belongs in a same-day call rather than a routine appointment.

How to raise it

The framing that works is the honest one. You are not asking whether they have dementia. You are asking whether something treatable is being missed, and the list above is why that is a reasonable thing to check.

It is easier to accept an appointment about thyroid function, vitamin levels, sleep and medications, which is genuinely what a good assessment involves, than one framed around memory loss.

The honest summary

Do not file memory changes under normal ageing without checking, because the professional guidance explicitly says not to. A meaningful share of cases have a reversible cause, roughly one in five improve, and most people with MCI do not progress quickly. The assessment is far less frightening than the thing families imagine they are agreeing to.