The 2024 Lancet Commission, which calls itself the most comprehensive analysis to date of dementia risk factors that can be changed, lists 14 of them, from high blood pressure and smoking to untreated hearing loss, and estimates that together they account for nearly half of cases worldwide.¹

That estimate is a ceiling, not a promise. It rests mostly on observational researchobservational study: A study in which researchers record what people already do or are exposed to, rather than assigning anyone to anything. It can show that two things go together, not that one causes the other, because the groups being compared may differ in other ways as well.Full entry in the glossary and assumes each factor is a cause, and trials that change several factors at once have so far moved thinking-test scores only slightly. In its 2026 guideline, the World Health Organization rated as high certainty the evidence that such combined programs have little or no effect on new dementia cases in adults with normal thinking, including those at raised risk, alongside consistent but very small gains in thinking scores.²

For a reader, the useful part is short and ordinary: keep physically active, stop using tobacco, and ask a doctor about blood pressure, cholesterol and hearing. The check-up questions further down turn that into a conversation with a doctor. Everyday memory techniques live in the wider memory and brain performance section.

The 14 dementia risk factors on the Lancet Commission’s list

Fourteen potentially modifiable risk factors make up the 2024 Lancet Commission’s list, sorted by the stage of life when they seem to matter most; the international panel of dementia researchers added high LDL cholesterol and untreated vision loss to the dozen in its 2020 report. Early life has one entry, less education. Midlife holds most of the list: hearing loss, high LDL (“bad”) cholesterol, depression, head injury, physical inactivity, diabetes, smoking, high blood pressure, obesity and excessive alcohol. Late life adds social isolation, air pollution and untreated vision loss.¹

The timeline exists because risk builds over decades. The Commission describes cognitive reserve, the brain’s ability to withstand damage before symptoms appear, and cites evidence that physical, mental and social activity build it. It credits greater reserve and less damage to blood vessels with part of the fall in age-specific dementia rates seen in some countries. In its model, hearing loss and high LDL cholesterol in midlife carry the largest single shares, about 7 percent of cases each.¹

1. Early life: less education
2. Midlife: hearing loss, high LDL cholesterol, depression, head injury, physical inactivity, diabetes, smoking, high blood pressure, obesity and excessive alcohol
3. Late life: social isolation, air pollution and untreated vision loss
4. The shaded share: about 45% of dementia cases worldwide in the model: a theoretical ceiling that assumes every factor is a cause

Picture someone in their fifties who strains to follow conversation in noisy meetings and has not had a cholesterol test in years (an illustration, not a case). Two of the biggest items on the list are sitting in routine check-ups they have skipped. Read that way, the list is less a brain program than a general health check, and most of it can be raised with a family doctor. These questions are a place to start.

WiserHours

Five questions for your next check-up

• When was my blood pressure last measured, and is it in a healthy range?
• Is a cholesterol test due for someone my age?
• Would a hearing test be worth doing, and would hearing aids help me?
• What kind of regular physical activity suits my health?
• If I smoke or use tobacco, what help to stop is available here?

Find this checklist online:
https://wiserhours.com/memory/dementia-risk-factors/#checklist-dementia-risk-factors-checkup

Why “nearly half” is a ceiling, not a promise

The Commission’s headline share is a population attributable fraction: an estimate of how many cases would disappear if every risk factor vanished, on the assumption that each one causes dementia. The Commission says plainly that it made that assumption, and that lowering risk helps populations but guarantees nothing to any individual. The method combines how common each factor is with how much it raises risk, then corrects for overlap, because risks cluster in the same people. The Commission estimated that overlap from a large Norwegian health study, and notes that most of the research it drew on comes from high-income countries, leaving low- and middle-income countries under-represented.¹

The share also depends on where people live. A 2019 analysis by the same team, using an earlier list of nine factors, estimated about 40 percent in China and India and 56 percent in a sample from six Latin American countries and territories, against 35 percent worldwide, because less education, high blood pressure, smoking, obesity and diabetes were more common there. The same paper notes that depression might be an early feature of dementia rather than a cause.³

That is the core problem with observational evidence: the arrow can point both ways. WHO’s guideline makes the same point, that some late-life links may reflect early disease rather than causes.² Someone who stops going out in the years before a diagnosis may be isolated because the illness has begun, not ill because they were isolated.

Myth

Nearly half of dementia can be prevented, so healthy habits will halve my risk.

Fact

That share is a theoretical ceiling for whole populations, built on the assumption that every factor is a cause. It says nothing precise about one person's odds, and the few combined-habit trials that counted dementia cases found little or no drop.

So treat “linked to” as a prompt for questions rather than proof, and any single risk figure as a description of a population, not a forecast about you.

What WHO’s 2026 guideline actually recommends

WHO’s 2026 guideline, the second edition of one first published in 2019, gives a strong recommendation to only two steps aimed at protecting thinking: physical activity, and help to stop using tobacco. It also strongly advises against several supplements, and its other recommendations are all conditional, most of them resting on evidence of low or very low certainty. A strong recommendation means the guideline panel was confident the benefits outweigh the downsides for most people; a conditional one means benefits probably outweigh harms, but different choices may suit different people.² Certainty of evidencecertainty of evidence: A rating of how confident we can be that a body of research has an effect about right, graded high, moderate, low or very low. It grades the evidence, not the stakes: low certainty means little is known, not that little is at risk.Full entry in the glossary, graded from high to very low, is how sure the panel is that the estimated effect is close to the true one.

Two patterns stand out. The strongest recommendations go to habits with well-established benefits for overall health. And where the dementia evidence is thin, WHO’s panel stresses that the underlying problem is worth treating anyway: high blood pressure for overall health, and hearing loss for its own sake, with hearing aids given a conditional recommendation backed by low-certainty evidence. Brain training, which gets a conditional nod on low-certainty evidence, is a story of its own, told in what large brain-training trials found.²

In everyday terms, picture a shopper choosing between a bottle of omega-3 capsules and a pair of walking shoes (an illustration). On WHO’s reading, the shoes come with a strong recommendation for protecting thinking and the capsules with a strong recommendation against, unless a vitamin deficiency has been diagnosed.

FINGER and the trials that change several habits at once

Programs that tackle several risk factors together have produced very small but consistent gains on thinking tests and, so far, no clear drop in dementia, according to WHO’s 2026 review of such trials. The idea behind them is that risks cluster in the same people, so a program aimed at several at once should do better than any single change.² Finland’s FINGER trial, published in The Lancet in 2015, tested it on adults aged 60 to 77 at raised risk: two years of diet guidance, exercise, brain training and monitoring of vascular risks such as blood pressure, against general health advice. The program group scored slightly better on a battery of thinking tests. FINGER measured thinking scores rather than new dementia cases, and more people in the program group had side effects, most often muscle and joint pain.⁴

The World-Wide FINGERS network, launched in 2017 to adapt the approach to other countries, had more than 25 member countries within two years, with trials planned from China and Singapore to Australia and across Latin America.⁵ A large recent test is the US POINTER trial from 2025: older adults at raised risk who followed a structured program, with regular coached team meetings and set exercise plans, improved slightly more on thinking tests over two years than a group making changes on their own. Both groups improved, and without an untreated comparison group the investigators could not exclude practice at the tests as one explanation. The Alzheimer’s Association funded the trial and helped design it, and a US blueberry industry council gave the structured group monthly rebates.⁶

Some trials found no benefit at all. In the Netherlands, six years of nurse-led vascular care did not reduce new dementia among people in their seventies⁷, and in France and Monaco a three-year program with or without omega-3 capsules made no clear difference to thinking scores.⁸

The study

Strong evidence

Forty-four combined-risk trials, pooled by WHO in 2026

For adults with normal thinking or raised risk, programs tackling at least three risk factors had little or no effect on new dementia cases and consistently positive but trivial effects on thinking-test scores, all on high-certainty evidence. People with mild cognitive impairment may gain slightly more, with certainty ranging from very low to high. Only two of the 44 trials counted new dementia cases, so the long-term question is still open. WHO adds that tiny average gains can add up across a whole population, and that the programs appear safe.²

Short trials can only show so much. The ones described here ran for two to six years, while the Commission’s model spans a lifetime. For one person, the practical lesson from POINTER is about structure: a retired teacher who joins a twice-weekly walking group and books a yearly blood-pressure check is closer to what the trials tested than someone relying on good intentions (an illustration).

Blood pressure: what the SPRINT MIND trial found

In the US SPRINT MIND trial, published in 2019 in JAMA, aiming for a systolic blood pressure below 120 mm Hg instead of below 140 did not clearly reduce probable dementia, its main outcome for memory and thinking. Fewer people in the intensive group developed mild cognitive impairment, a secondary outcome the authors asked readers to interpret cautiously.⁹

The raw counts show why. Over a median of about five years, probable dementia was diagnosed in 149 of the roughly 4,700 people in the intensive group and 176 of the roughly 4,700 in the standard group, a gap of 27 people that could have been chance. Mild cognitive impairment appeared in 287 against 353. The trial was stopped early because intensive control cut cardiovascular events and deaths, so it probably had too few dementia cases to give a firm answer. It also found no sign that lower blood pressure harmed thinking. The trial enrolled US and Puerto Rican adults aged 50 and over without diabetes, previous stroke, advanced kidney disease or symptomatic heart failure, so its result does not transfer to everyone.⁹

WHO’s panel weighed SPRINT MIND with other trials and rated the effect of intensive control on dementia as uncertain, on very-low-certainty evidence, drawing on reviews published in 2020 and 2021.² A larger 2025 trial in rural China points the other way. It randomized about 34,000 adults with uncontrolled high blood pressure by village, and dementia was diagnosed less often over four years where trained community health workers lowered blood pressure than where usual care continued. Its abstract, the only part we could read, gives only a relative difference.¹⁰

WHO still stresses the treatment gap: worldwide, fewer than half of adults with high blood pressure are diagnosed and treated, and fewer than a quarter have it under control.² Someone who feels well and last had a reading taken years ago could be part of that gap without knowing it (an illustration).

Blood-pressure targets are a decision for you and a doctor, who weighs your other conditions and medicines. The step anyone can take is simpler: know your numbers, starting with the first question on the check-up list.

Memory changes that need a doctor, from urgent to routine

Sudden confusion is an emergency, memory or thinking problems that disrupt daily life deserve a doctor’s visit within weeks, and occasional forgetfulness is common with age. The tiers below draw on WHO, UK and US health services; wherever you live, start with your local emergency number or your usual doctor.

Soon, within days or weeks: WHO lists early signs of dementia that include forgetting recent events, getting lost when walking or driving, confusion in familiar places, losing track of time, trouble with decisions or with finding words, and difficulty with familiar tasks.¹³ The UK’s NHS suggests a family doctor visit once memory problems start to affect day-to-day life, since several common culprits, including stress, anxiety, depression and poor sleep, are treatable.¹⁴

At your next routine appointment: the National Institute on Aging (US) describes mild forgetfulness, such as needing more time to learn something new, as often part of normal aging, and still advises raising any noticeable memory change with a doctor.¹⁵ That visit is also a good moment for the check-up questions above.

The bottom line

The Lancet Commission’s list of 14 changeable dementia risk factors is the best map available, but its “nearly half” is an upper estimate built on observational data, and trials that change several factors at once have so far nudged thinking scores rather than cut dementia. The steps with the firmest backing are ordinary: stay active and do not smoke. Blood pressure, hearing and cholesterol checks from midlife have weaker dementia evidence but are worth having for overall health. Take any memory change that disrupts daily life to a doctor.