Three medications into Robert J. Brent’s economic analysis of dementia treatments, the pattern wasn’t looking good. FDA-approved Alzheimer’s drugs: net loss. Beta-blockers: net loss. Antipsychotics: the biggest loss of all. Then came the fourth medication class, and for the first time in the entire study, the math actually worked.

That medication class is one almost everyone already has at home: NSAIDs. Aspirin, ibuprofen, naproxen the over-the-counter painkillers sitting in a bathroom cabinet somewhere, largely ignored as anything more than a headache fix.

A modest effect, but a real one

NSAIDs produced a small but genuine reduction in dementia symptoms in Brent’s data, modest compared to the effect of avoiding a nursing home stay or using hearing aids, but real, and in the right direction. That alone was notable, because up to this point, no pharmaceutical intervention in the study had managed to move the needle on dementia symptoms without also making things worse elsewhere.

The bigger story came from mortality. NSAIDs showed a meaningful reduction in death rates, both directly and through their modest effect on dementia. Converting that into dollar terms using the standard value-of-a-statistical-life framework produced total benefits of roughly $46,800 per person. Against a lifetime cost of about $11,000, remember, these are over-the-counter medicines, no doctor visits required; the net gain came out to roughly $35,700 per person. For the first time in the book, a medicine had cleared the bar.

Not all painkillers are created equal.

Because NSAIDs are common enough that individual drugs could be separated out in the data, Brent’s team went a step further and broke the group apart. The results were lopsided. Aspirin did the heavy lifting, with an effect on mortality roughly double the group average on its own; aspirin’s net benefit came out to nearly $59,000 per person. Ibuprofen contributed too, at about a third of the average effect, still clearing a positive net benefit of a couple thousand dollars per person on its own.

Naproxen, on the other hand, showed no statistically meaningful effect at all. And acetaminophen, despite being widely recommended in medical literature as the go-to option for pain in advanced dementia, came back with no benefit whatsoever. Given that it isn’t even technically an anti-inflammatory drug, it’s a useful reminder that “commonly recommended” and “actually effective” aren’t always the same thing.

Why does this matter more than it sounds?

It’s easy to read “aspirin helps a little” and shrug. But step back and look at what this actually represents: a pharmaceutical intervention, available for a few dollars without a prescription, that measurably reduced dementia symptoms in a population where the flagship, purpose-built Alzheimer’s drugs failed to do so and did it while producing a net financial gain instead of a loss.

That runs counter to a persistent assumption in the medical field that no pharmaceutical currently exists that can meaningfully affect dementia symptoms, let alone reverse them. NSAIDs, modestly but measurably, did exactly that. It’s not a cure. It’s not even close to a cure. But it’s evidence that useful tools don’t always look like breakthrough science; sometimes they’re already sitting on a shelf at the pharmacy, just never evaluated through this particular lens before.

A word of caution

None of this is a green light to start a daily aspirin regimen without a conversation with a doctor. Aspirin and other NSAIDs carry their own well-documented risks, including bleeding, especially in older adults, and this analysis is about population-level averages, not individual medical advice. What it does establish is a genuine data point: among five widely used medication classes prescribed to people with dementia, this was the first one where the numbers actually worked in the patient’s favor.