Reducing Dementia Risk: The Factors the Evidence Supports

Headlines about “preventing dementia” tend to run ahead of the science. The strongest summary of that science comes from the Lancet Commission on dementia prevention, intervention, and care, which estimates that a large minority of dementia cases are associated with risk factors people can potentially change (Livingston et al., Lancet, 2020). That estimate is built on population data, and it describes associations, not certainties. This article walks through what the Commission says, what the randomized trials add, and where the evidence is thinner than the marketing around brain health suggests. None of it is a treatment plan. Your own situation belongs with your clinician.

What the Lancet Commission estimates

The headline number is a population attributable figure. In 2020, the Lancet Commission concluded that around 40 percent of dementia cases worldwide are associated with 12 modifiable risk factors, meaning that if those factors were fully eliminated across the whole population, that share of cases might in theory be delayed or avoided (Livingston et al., Lancet, 2020). The 2024 update added two more factors, untreated vision loss and high LDL cholesterol, and raised the combined estimate to about 45 percent (Livingston et al., Lancet, 2024).

The full list spans a lifetime. Earlier in life it includes less education. In midlife it includes hearing loss, high LDL cholesterol, depression, traumatic brain injury, physical inactivity, diabetes, smoking, high blood pressure, obesity, and excess alcohol. Later in life it adds social isolation, air pollution, and untreated vision loss.

Read the number carefully. A population attributable fraction is not your personal risk. It assumes complete removal of a factor, which no real program achieves, and it draws on observational data, which can show that two things travel together without proving that one causes the other. The Commission is candid that these are associations across populations. It is a map of where the potential lies, not a promise for any single person.

Blood pressure: the strongest randomized signal

Of all the factors, blood pressure has some of the better trial evidence. SPRINT MIND was a randomized trial that assigned older adults to intensive blood pressure control or to a standard target. The intensive group had significantly fewer cases of mild cognitive impairment, a common precursor to dementia (Williamson et al., JAMA, 2019).

The result came with an important caveat. The reduction in probable dementia on its own did not reach statistical significance. The trial’s blood pressure intervention was stopped early for cardiovascular benefit, which left fewer dementia cases to analyze than planned. The authors were clear that the mild cognitive impairment finding was the significant one, and the dementia finding pointed in the same direction without crossing the line.

That trial sat on top of the larger SPRINT study, which showed that intensive blood pressure control reduced cardiovascular events and deaths (Wright et al., NEJM, 2015). So the brain finding fits a broader pattern of vascular benefit. It is still a finding about a controlled intervention in a specific trial population, not a target for you to set on your own.

Hearing loss: a strong association, a mixed trial

Observational studies have repeatedly linked hearing loss to later dementia, which is why it sits high on the Commission’s list. The obvious question was whether treating hearing loss changes the trajectory. The ACHIEVE trial tested exactly that, randomizing older adults with hearing loss to either a hearing intervention with hearing aids or a health education control, then following their cognition for three years (Lin et al., Lancet, 2023).

Overall, the trial found no significant slowing of cognitive decline from the hearing intervention. That is the primary result, and it should temper any claim that hearing aids prevent dementia.

The nuance is in a prespecified subgroup. The trial enrolled participants from two sources. One group, recruited from an existing long-running heart-health study, was older and at higher risk. In that prespecified higher-risk subgroup, the hearing intervention was associated with a meaningful reduction in cognitive decline over the three years. A prespecified subgroup finding is stronger than a fishing expedition after the fact, but it is still a subgroup, and it needs confirmation. The most accurate reading is that hearing treatment may help some higher-risk people, and that the overall trial did not show a benefit.

Multidomain programs: proof of principle, not a finished recipe

No single pill has shown it prevents dementia, so researchers have tested combined lifestyle programs. The landmark example is the FINGER trial, which randomized at-risk older adults to a structured multidomain program (diet guidance, exercise, cognitive training, and management of vascular risk factors) or to general health advice, over two years (Ngandu et al., Lancet, 2015).

The intervention group did modestly better on a composite measure of cognitive performance. FINGER is widely cited as proof of principle that a coordinated program can produce a measurable benefit. It launched a family of similar trials around the world.

Keep the scale in view. The benefit was small, the trial ran two years, and it measured cognitive test scores rather than proving fewer dementia diagnoses. It shows the concept can work. It does not tell you the exact program, intensity, or duration that would protect any given person, and it is not a guarantee of avoiding dementia.

Where the evidence is weak or absent

This field is crowded with products that outrun their data. Ginkgo biloba is a clear example. The large Ginkgo Evaluation of Memory study randomized thousands of older adults and found that ginkgo did not reduce the rate of dementia or Alzheimer’s disease (DeKosky et al., JAMA, 2008), and a companion analysis found it did not slow cognitive decline either (Snitz et al., JAMA, 2009).

Commercial brain-training apps deserve similar caution. Practice tends to improve the specific tasks people train on, but strong evidence that it prevents dementia or transfers to everyday function is lacking. Many supplements marketed for memory rest on observational hints or small studies rather than convincing randomized trials.

The reporting-versus-selling gap matters here. A factor being on the Lancet list means it is associated with dementia in population data. It does not automatically mean a product aimed at that factor has been shown to work. The supplements article in this hub covers that gap in more detail. When a program has good trial evidence, this article names it. When it does not, that silence is the finding.

This piece reports what the Lancet Commission and the major randomized trials have found. It is not medical advice, and it contains no targets, doses, or instructions for you to follow. The risk factors described are population-level associations, and even the interventions with the best trial evidence reduce risk rather than remove it. Whether any of this applies to you, and what to do about it, is a conversation for you and a clinician who knows your history.

Common questions

Does this mean I can prevent dementia?

No. The evidence supports reducing risk at the population level, not guaranteeing prevention for any individual. The Lancet Commission’s 40 to 45 percent figure assumes the complete removal of risk factors across an entire population, which is not achievable in real life, and it rests partly on observational data (Livingston et al., Lancet, 2020; 2024). Some people who do everything associated with lower risk still develop dementia, and some who do little never do.

Do hearing aids prevent dementia?

The ACHIEVE trial did not show that. Over three years, hearing aids produced no significant slowing of cognitive decline overall, though a prespecified higher-risk subgroup did show a benefit (Lin et al., Lancet, 2023). The link between hearing loss and dementia in observational studies is strong, but the trial evidence that treating it changes cognition is mixed and needs confirmation.

Is one factor more important than the others?

The Lancet Commission assigns each factor a share of the population-level estimate, and blood pressure has some of the strongest randomized support through SPRINT MIND (Williamson et al., JAMA, 2019). But the factors overlap and interact, and the Commission frames them as a combined picture across the life course rather than a single lever to pull. Which factors matter most for a specific person depends on their own health, which is a clinician’s call.

Do supplements or brain-training apps help?

The strong trial evidence is not there. The large Ginkgo Evaluation of Memory study found no reduction in dementia or cognitive decline from ginkgo (DeKosky et al., JAMA, 2008; Snitz et al., JAMA, 2009). Brain-training tends to improve the trained tasks without clear proof it prevents dementia. Many memory supplements rest on weak or observational evidence rather than convincing randomized trials.