Reading time: about 10 minutes
It usually starts small, and it almost never feels like a brain issue. The television creeps a few notches louder. You catch most of what people say in a quiet room but lose the thread completely once a café gets busy. You find yourself nodding along, filling in gaps, quietly exhausted after a long dinner with friends — not because the conversation was hard, but because hearing it was. Most people file all of this under "getting older" and move on.
So it comes as a genuine surprise to learn that, when a group of the world's leading dementia researchers sat down to rank the things that raise a person's risk of cognitive decline, hearing loss came out near the top of the list — above physical inactivity, above social isolation, above smoking. Not a footnote. The single largest one they could name.
That is a striking claim, and it is worth taking seriously. It is also easy to overstate into something the evidence does not support — a straight line from "your ears" to "dementia," and a promise that a hearing aid will save your memory. The real picture is more interesting than either the alarm or the hype. Let's walk through what the science actually shows, where it stops, and what's genuinely worth doing regardless of how the research shakes out.
The claim, and why researchers take it seriously
The list in question comes from the Lancet Commission on dementia prevention, intervention, and care — a periodic, heavily-cited synthesis by an international group of researchers. Its 2020 report identified a set of modifiable risk factors that together were estimated to account for around 40% of dementia cases worldwide, at least in principle. Of all of them, hearing loss in midlife carried the largest estimated share. The 2024 update kept hearing loss prominent and added two more factors to the list, including untreated vision loss.
Why single out hearing? Because the underlying association is unusually strong and consistent. Long-running observational studies — much of the foundational work led by Frank Lin and colleagues at Johns Hopkins — have repeatedly found that people with hearing loss show more cognitive decline and higher dementia risk over the following years than people with normal hearing. And the relationship tends to be dose-dependent: mild hearing loss is associated with a modest increase in risk, moderate with more, severe with more still. In epidemiology, a clean dose-response gradient is one of the things that makes researchers lean in — it's the kind of pattern you'd expect if the exposure genuinely mattered, rather than being pure noise.
So the starting point is solid: hearing loss and cognitive decline travel together, reliably, across many studies and many populations. The hard question — the one that decides whether any of this is actionable — is why.
Why hearing might matter for the brain
There are several plausible explanations, and they are not mutually exclusive. They're worth separating, because they point in very different directions when it comes to what you can do about it.
Effortful listening steals resources. When the signal reaching your ears is degraded, your brain has to work harder to reconstruct it — guessing at half-heard words, using context to fill gaps, running a kind of constant real-time repair job on incoming speech. That effort has to come from somewhere. The "cognitive load" hypothesis is that resources spent straining to hear are resources not available to understand, remember, and think — and that, sustained over years, this tax may contribute to decline. If you've read how attention actually works, the logic will be familiar: attention is a limited budget, and a brain quietly spending most of it on decoding sound has less left over for everything else.
Hearing loss pulls people out of the world. Straining to follow conversation is tiring and, often, embarrassing. So people with untreated hearing loss frequently withdraw — declining the noisy restaurant, saying less at the family gathering, drifting away from exactly the rich, unpredictable social contact that appears to help keep brains resilient. Social disengagement is itself on the dementia risk-factor list, so hearing loss may do part of its damage indirectly, by cutting off a protective pathway rather than by acting on the brain directly.
Less input may change the brain itself. Your brain is shaped by what it's used for — a core theme of what neuroplasticity actually means. When auditory input dwindles, the theory goes, the brain regions that process sound (and their neighbours) get less of the stimulation that keeps them structurally healthy. Some imaging studies have reported faster brain atrophy in people with hearing loss, consistent with a "use it or lose it" story at the level of tissue. This is suggestive, not proven, but it's a coherent mechanism.
They might share a common cause. Here's the caveat that keeps honest researchers cautious: some of the same underlying problems — small-vessel vascular disease, diabetes, the general biology of ageing — can damage both the delicate machinery of the inner ear and the brain. If so, hearing loss and cognitive decline could be two visible symptoms of one hidden process, rather than one causing the other. Under that story, fixing the hearing wouldn't fix the brain, because the ears were never the real problem.
Notice what this means. If the mechanism is effortful listening or social withdrawal, then treating hearing loss should help. If it's a shared common cause, treating the hearing might do nothing for cognition. The mechanisms aren't just academic — they're the whole question. Which brings us to the trial that tried to answer it.
The honest limits: one big trial, one surprising result
Everything above is built largely on observational evidence — studies that watch what happens to people without intervening. Observational studies are powerful for spotting associations, but they can't cleanly separate cause from consequence, because the people who have hearing loss (or who treat it) differ from those who don't in dozens of other ways. To actually test whether treating hearing loss slows cognitive decline, you need a randomized controlled trial: take a large group, randomly assign some to a hearing intervention and some not, and follow them for years.
That trial exists. It's called ACHIEVE, led by the same Johns Hopkins group, and its main results were published in The Lancet in 2023. Nearly a thousand older adults with untreated hearing loss were randomized to either a full hearing intervention (hearing aids plus counselling and support) or a control program, and their cognition was tracked over three years.
The headline result was not what many people expected: across the whole study population, the hearing intervention did not significantly slow cognitive decline. If you stopped reading there, you'd conclude the dementia case for hearing aids had collapsed.
But there was a crucial wrinkle. The trial had deliberately recruited two different groups: some healthy community volunteers, and some older adults already at higher risk of decline (drawn from a long-running cardiovascular health study, generally older and with more risk factors). In that higher-risk subgroup, the hearing intervention was associated with a substantial slowing of cognitive decline over the three years — roughly cut in half. The healthy volunteers, by contrast, barely declined at all during the study, so there was little decline for any intervention to slow.
How you read this depends on your temperament. The optimistic reading: for the people who actually stand to lose ground — older, higher-risk adults — treating hearing loss may genuinely protect cognition, and the null overall result was diluted by a low-risk group that wasn't going to decline anyway. The cautious reading: the trial's primary, pre-specified outcome was negative, subgroup findings are exactly the kind of result that sometimes fails to replicate, and one three-year trial is not proof. Both readings are legitimate. What you can't honestly say is "a landmark trial proved hearing aids prevent dementia." It didn't. It showed a promising, biologically plausible signal in the people most at risk, inside an overall result that was null — and it made clear we need more evidence, not less.
A few more limits worth holding onto. Reverse causation is still possible: very early, undetected neurodegeneration can impair the brain's processing of sound, so in some people failing cognition might drive apparent hearing loss rather than the other way around. And even the strongest observational associations can't rule out that people who seek out and stick with hearing treatment are simply healthier, wealthier, or more engaged to begin with.
Vision counts too
Hearing isn't the only sense in this story. The 2024 Lancet Commission update added untreated vision loss to its list of modifiable dementia risk factors, on similar logic — reduced sensory input, reduced engagement, a brain getting less of what keeps it stimulated. Several observational studies have found that people who undergo cataract surgery, which restores clear vision, subsequently show lower rates of dementia than those who don't.
The same caveats apply, and just as forcefully: this is observational, the people who get cataract surgery differ from those who don't, and no one has proven that restoring vision prevents cognitive decline. But the broader principle is consistent and reasonable — the senses are the brain's windows onto the world, and a brain that receives clearer, richer input is, plausibly, a brain that stays more engaged.
What to actually do
Here's the reassuring part, and it's why this topic is genuinely worth your attention even amid all the uncertainty: the sensible actions are worth taking regardless of how the dementia question ultimately resolves.
Treating hearing loss is not a gamble you're making purely on a maybe-it-protects-my-memory bet. Hearing well lets you stay in conversations, stay connected to the people around you, stay safe (you hear the car, the alarm, your name), and stay less exhausted at the end of a social day. Those benefits are real and immediate, whatever the long-term cognitive science eventually concludes. The dementia angle, if it holds up, is a bonus — not the whole reason.
So, concretely:
- Get your hearing checked, and don't wait for it to be "bad enough." Hearing loss is gradual and easy to normalize; many people go years compensating without realizing how much they've lost. A baseline test — and periodic ones as you age — turns an invisible drift into something you can actually act on.
- Protect the hearing you have. Much age-related hearing loss is accelerated by a lifetime of loud noise exposure. Turn the volume down (especially on headphones), use protection around power tools, concerts, and machinery, and treat your ears as the finite, non-regenerating resource they are.
- If you have hearing loss, treat it. Modern hearing aids are low-risk, and the case for using them stands on communication and quality of life alone. The higher-risk ACHIEVE subgroup result is one more reason to take treatment seriously, not the only one.
- Look after your vision too. Regular eye exams, and treating correctable problems like cataracts, follow the same logic — keep the brain's inputs clear.
- Keep the other levers going. Sensory health sits alongside the rest of what protects an ageing brain: staying physically active, staying socially and mentally engaged, and building what researchers call cognitive reserve — the buffer of richly-connected, well-used neural networks that helps some brains stay sharp even as they age. No single factor is the whole game; they compound.
Where AI fits
For a brain-health topic, the AI angle here is narrow but real, and it's worth naming precisely so it doesn't get oversold.
The genuinely useful part is accessibility. Real-time captioning and speech-to-text — the kind now built into phones, video calls, and glasses — can dramatically lower the effort of following conversation for someone with hearing loss, and keep them in the room, in the meeting, at the table. AI is also increasingly embedded in hearing aids themselves, helping separate a voice from background noise. Used this way, these tools do exactly what good assistive technology should: they reduce the tax on a struggling sense and help someone stay engaged with the world. That's the same principle we come back to again and again — AI at its best augments a human capability rather than replacing it.
But there's a guard worth stating plainly. A live caption on your screen is a workaround, not a diagnosis. It's easy to let good-enough captions become a reason to keep putting off an actual hearing evaluation — to patch the symptom and never check the underlying loss. Don't. Use the technology to communicate better and get your hearing properly assessed and treated; they're not substitutes for each other.
And if you find yourself researching your own hearing, a family member's, or the dementia question online, the ordinary rules of using AI as a thinking partner apply. Treat what a chatbot tells you about a medical topic as a starting point for better questions, not a verdict — a habit we lay out in how to actually learn with AI — and take the actual decisions to a professional who can examine the person in front of them.
The one-sentence version
Hearing loss is one of the strongest and most consistent modifiable risk factors linked to cognitive decline, the causal case is promising but genuinely unproven — the one big trial was null overall, with benefit only in those most at risk — and the honest move is to protect and treat your hearing (and vision) now, because the everyday payoff of hearing your own life clearly is worth it whatever the long-term brain science eventually concludes.
This article is for general educational purposes and is not medical advice. If you notice changes in your hearing or thinking, or want personalized guidance, please talk to a qualified audiologist or physician. If you'd like to understand the bigger picture of how the choices you make actually shape an ageing brain, our Brain Health 101 course walks through the science of sleep, movement, nutrition, stress, and connection — and how to turn it into habits that hold, using AI as a genuine thinking partner rather than a shortcut.