Lifestyle trials show small gains on thinking tests, not fewer diagnoses

The advice sounds settled. People who walk most days, eat more vegetables, keep learning, control their blood pressure and stay connected are supposed to be protecting their brains. If each habit helps a little, doing them all at once should help a lot.

Researchers have tested exactly that, in trials that change several habits together and follow thousands of people for years. The biggest US trial built to test the idea, published in 2025, found that a structured program raised a thinking-test score by 0.029 standard deviations a year more than a do-it-yourself version did. A standard deviation measures how widely scores vary between people, and the yearly edge was about one-thirty-fifth of one.

That edge is real. It is also small, and it was measured on tests, not on diagnoses. Across a decade of these trials, the honest summary sits between two popular claims: that a healthy lifestyle prevents dementia, and that nothing does. The evidence supports neither.

Why researchers started bundling habits

Trials of one habit at a time have mostly disappointed, which is part of why the field moved to combinations. In brief:

• Exercise. Two-year programs for older adults haven’t produced lasting thinking gains over active comparison groups; the exercise-and-thinking report has the trial-by-trial detail.
• Diet. The MIND diet, in its first randomized trial, did no better than a calorie-cut comparison diet, as the diet article explains.
• Sleep. Evidence tying sleep to dementia comes mostly from observation, and trials of treating sleep problems show modest effects; see the sleep piece for the medication cautions.
• Social contact. Loneliness and isolation turn up alongside dementia in study after study, but trials of adding company are few and small, and the isolation article untangles which way the arrow points.
• Hearing. The largest hearing-aid trial found no overall slowing of decline, with a benefit only in a higher-risk subgroup, a result unpacked in the hearing coverage.
• Brain training. The ACTIVE trial’s courses sharpened the specific skills people practiced, yet its main 20-year analysis found no drop in dementia diagnoses; the cognitive-stimulation article reads that result closely.

The common late-life forms of dementia are thought to stem from several causes and risk factors at once, so the logic of bundling is that small effects might add up. A Finnish trial set the pattern the later ones copied.

FINGER set the pattern, with a small result

The FINGER trial, published in The Lancet in 2015, enrolled 1,260 Finns aged 60 to 77 whose scores on a dementia risk tool built from cardiovascular and aging factors were elevated, and whose thinking was average or slightly below average for their age. Half got two years of dietary guidance, exercise, computer-based brain training and monitoring of vascular risk factors. The other half got general health advice.

Both groups’ test scores rose over two years, by 0.20 standard deviations with the program and 0.16 without it. The difference worked out to 0.022 standard deviations a year. That was statistically significant and small, and it was a test score rather than a count of anyone developing dementia. The program also had a cost: adverse events hit 7 percent of the intervention group and 1 percent of the controls, mostly musculoskeletal pain, reported by 5 percent against none.

The investigators called FINGER a proof of concept. Over the next decade, others tried to repeat it.

The trials that found nothing

France and Monaco. MAPT, published in Lancet Neurology in 2017, enrolled 1,680 people 70 and older who had a memory complaint, difficulty with one everyday task or slow walking. Over three years, group sessions in cognitive training, physical activity and nutrition, with or without omega-3 capsules, produced no significant effect on cognitive decline in any arm. (The fish-oil arm’s miss is part of the supplements story told separately.) Part of the funding came from industry.

The Netherlands. preDIVA, published in The Lancet in 2016, went straight for diagnoses. It enrolled 3,526 people aged 70 to 78 through 116 general practices and gave half of them six years of nurse-led care for cardiovascular risk factors. After a median 6.7 years, dementia had developed in 7 percent of each group, and disability, heart disease and deaths didn’t differ either. The authors’ verdict: “A nurse-led, multidomain intervention did not result in a reduced incidence of all-cause dementia in an unselected population of older people.” They suggested that participants’ fairly modest cardiovascular risk and the high quality of ordinary Dutch care left little room for a program to help.

Japan. J-MINT, published in 2024, enrolled 531 people aged 65 to 85 already diagnosed with mild cognitive impairment, the in-between stage before dementia. Eighteen months of vascular risk management, exercise, nutrition counseling and cognitive training did not significantly change the main thinking score. Hints of benefit in people who attended exercise sessions faithfully, or who carried APOE e4, a gene variant linked to higher Alzheimer’s risk, came from secondary analyses and need testing on their own.

What pooled evidence says about dementia itself

In 2021, Cochrane, an international network that pools clinical trials using standardized methods, reviewed nine randomized trials with 18,452 participants, each with at least 400 people and at least a year of follow-up. Only two counted new dementia cases, and together they showed no difference, with the evidence rated high certainty. The reviewers wrote: “We found no evidence that multi-domain interventions can prevent incident dementia based on two trials.”

On detailed test batteries, the programs produced a small gain of about 0.03 standard deviations, also high certainty. On a brief screening test, there was no effect. The test-battery gain was strongest in trials that included cognitive training, so the reviewers could not rule out that people had simply learned the tests. Two of the review’s three authors were also preDIVA investigators. The review’s search ended in April 2021, before the newest trials.

Diagnoses remain the open question. A 2026 analysis of Look AHEAD, a 10-year US lifestyle trial in adults with type 2 diabetes and excess weight, tracked mild cognitive impairment and dementia for 12 to 14 years after the trial ended. The program’s link with those diagnoses depended on starting weight: lower risk among people who began overweight, no difference among those with obesity, and higher risk among those with the most severe obesity. That is a pattern for future trials to test, not a settled answer.

The newest trials: structured beats self-guided

US POINTER

POINTER, published in JAMA in 2025, enrolled 2,111 sedentary adults aged 60 to 79 with less-than-ideal diets plus at least two other risk factors, at five US sites. About 69 percent were women. The Alzheimer’s Association, which has put nearly $50 million into the trial, describes the structured program on its website: 30 to 35 minutes of moderate-to-intense aerobic exercise four times a week, strength and flexibility work twice a week, 30 minutes of computer-based brain training three times a week, the MIND diet, and regular checks of blood pressure, weight and lab results. The self-guided group was encouraged toward the same goals with less structure and accountability.

Both groups’ scores climbed over two years: 0.243 standard deviations a year with structure, 0.213 without. The structured group’s yearly edge of 0.029 was statistically significant. It looked similar in carriers and non-carriers of APOE e4 and larger in people who started with lower scores. The structured group also reported fewer serious adverse events, 151 against 190.

The Alzheimer’s Association’s results page says both programs improved cognition. The trial can’t show that, because it had no group that did nothing.

LatAm-FINGERS

In July 2026, The Lancet published LatAm-FINGERS, run in 11 Latin American countries, a population rarely included in such trials. It enrolled 1,065 people aged 60 to 77 with elevated risk scores and below-par thinking; three-quarters were women. A structured, supervised program beat flexible health advice by 0.11 standard deviations a year, with scores rising 0.31 and 0.20 a year in the two groups. Musculoskeletal symptoms hit 21 percent of the structured group against 2 percent, and serious adverse events 9 percent against 5 percent, none judged related to the program. The Alzheimer’s Association funded it, too.

Brain scans

A POINTER imaging study, published in JAMA Neurology in 2026, followed 983 participants with MRI or PET scans. It found no difference between the programs in amyloid, the protein that forms plaques in Alzheimer’s disease, in tau tangles, in hippocampal volume or in white matter lesions. People who started with a smaller hippocampus gained more from the structured program, an exploratory finding.

Why both groups improved

Three things lift scores in a trial like this. The flexible groups were encouraged toward the same habits, so they were never untreated. People improve on tests they have taken before. And expectation sharpens effort. Without an untreated group, nobody can say how much of either group’s rise came from the programs. What these trials answer is narrower: whether structure and accountability add something beyond good intentions. In two large trials, they did, by a little.

Blood pressure, the strongest lead

One factor has trial evidence on diagnoses rather than test scores. SPRINT MIND, a US trial of more aggressive blood-pressure lowering, cut new cases of mild cognitive impairment without significantly reducing probable dementia, a result our risk-factor piece covers in full.

A larger test came from rural China. In a trial published in Nature Medicine in April 2025, led by researchers at UT Southwestern Medical Center in Dallas and China Medical University, 326 villages were randomly assigned. In half, trained community health workers without medical degrees started and adjusted blood-pressure medicines under physician supervision, aiming below 130/80. The other half got usual care. Across 33,995 people aged 40 and older with uncontrolled hypertension, systolic pressure fell 22 points (mm Hg) more in the program villages over four years. All-cause dementia was 15 percent lower in relative terms, and serious adverse events were slightly less common.

The authors concluded that intensive blood-pressure reduction lowers dementia risk. Three caveats matter before carrying that to an American clinic. The trial was open-label, so villagers and health workers knew who was in the program, though dementia was judged by assessors who did not. The setting was rural China. And the care came from community health workers following a simple stepped protocol. A 15 percent relative drop also says nothing by itself about how many fewer people per hundred were affected.

Where official bodies stand

The national academies of Sciences, engineering, and medicine, 2017. Its committee, sponsored by the National Institute on Aging and chaired by Alan I. Leshner, judged the evidence for three approaches encouraging yet inconclusive: cognitive training, blood-pressure management in people with hypertension (with emphasis on midlife, roughly 35 to 65), and more physical activity. That was not enough for a public health campaign, the committee said, but enough to justify telling the public about blood-pressure management, especially in midlife, and about where the physical-activity evidence stood.

The national institute on aging, June 2026. Its alzheimers.gov page states that no approach has been proven to prevent Alzheimer’s disease and related dementias, and that the cause-and-effect behind observational links remains uncertain.

The world health organization, July 2026. Its second-edition guideline on reducing the risk of cognitive decline and dementia makes three strong recommendations: physical activity for adults with normal cognition (moderate-certainty evidence), tobacco cessation (low certainty), and against vitamin B, vitamin E, omega-3 and multivitamin supplements in people without a diagnosed deficiency (moderate certainty). Most of the rest are conditional, WHO’s softer grade for options people may reasonably weigh differently. That includes cognitive training (low certainty), social activity (very low), hearing aids (low), a healthy balanced diet (moderate) and managing hypertension specifically to reduce dementia risk (very low). WHO says this edition also adds guidance on putting tailored multidomain programs into practice.

Graded with the library’s five evidence verdicts:

• Combined lifestyle programs raise thinking-test scores. Measured, but below the threshold that matters. The gains are consistent and small, and their clinical meaning is unknown.
• Combined programs prevent dementia. Tested and not supported in the two trials that counted diagnoses; insufficient for the newer, more intensive programs, none of which has run long enough to count them.
• Structured programs beat self-guided ones. Positive evidence, contested interpretation, since both groups improved and neither trial had an untreated arm.
• Lowering high blood pressure reduces dementia. Positive evidence, contested interpretation: one large open-label trial positive, one large US trial that missed its dementia endpoint but cut mild impairment, and a conditional, very-low-certainty WHO grade.
• Physical activity for brain health. Recommended by name in a guideline: WHO, 2026, strong, moderate certainty, for adults with normal cognition.

Acting on evidence this uncertain

This sorting guide draws only on the trials and statements above.

1. Separate the habit from the promise. Several of these habits have payoffs that don’t depend on dementia. SPRINT, the parent of SPRINT MIND, stopped early because intensive treatment reduced cardiovascular events and deaths. A program can be worth doing without being dementia insurance.
2. Blood pressure carries the most weight. It is also the item where targets and medicines belong to a prescriber. SPRINT MIND excluded people with diabetes or a past stroke, so its target doesn’t automatically fit everyone.
3. Expect structure to matter more than intention. In POINTER and LatAm-FINGERS, supervision, schedules and check-ins beat advice alone. When comparing programs, those features are the ones the evidence favors.
4. Discount any program that promises prevention. No combined lifestyle trial has shown fewer dementia diagnoses.
5. Count the costs. Time, money and injury risk are real. Muscle and joint pain led the side-effect lists in FINGER and LatAm-FINGERS, and supplements earned WHO’s strong recommendation against them.

When to bring in a clinician

• A new memory or thinking change calls for an evaluation, not a prevention program; a workup’s steps appear in our piece on when memory problems need assessment.
• High blood pressure, diabetes or high cholesterol in midlife raise questions about targets and medicines that belong to the clinician managing them. WHO’s management recommendations here are conditional, and individual risks vary.
• Heart disease or joint problems make the intensity of a new exercise program a decision to make with a clinician first, given how often structured programs in these trials produced aches and injuries.

The takeaway

A healthy lifestyle prevents dementia overstates what a decade of trials found. Nothing works understates it. Programs that bundle exercise, diet, brain training and risk-factor checks reliably nudge thinking-test scores, more so when they come with structure and accountability, but none has yet shown fewer diagnoses. The strongest trial evidence for fewer dementia cases involves treating high blood pressure, which makes it the number in this field that most deserves a conversation with whoever manages it.

Sources

1. National Academies of Sciences, Engineering, and Medicine. “Evidence Supporting Three Interventions That Might Slow Cognitive Decline and the Onset of Dementia Is Encouraging but Insufficient to Justify a Public Health Campaign Focused on Their Adoption (news release on Preventing Cognitive Decline and Dementia: A Way Forward, June 22, 2017).” nationalacademies.org
2. National Institute on Aging (alzheimers.gov). “Can I Prevent Dementia? (last updated June 16, 2026).” alzheimers.gov
3. World Health Organization (via NCBI Bookshelf, National Library of Medicine). “Risk reduction of cognitive decline and dementia: WHO guidelines, second edition (2026).” ncbi.nlm.nih.gov
4. World Health Organization. “Risk reduction of cognitive decline and dementia: WHO guidelines, second edition (publication page, 15 July 2026).” who.int
5. Alzheimer's Association. “U.S. POINTER Study Results.” alz.org
6. The Lancet. “Ngandu T, Lehtisalo J, Solomon A, et al. "A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): a randomised controlled trial." Lancet, 2015. PubMed. DOI: 10.1016/S0140-6736(15)60461-5.” pubmed.ncbi.nlm.nih.gov
7. The Lancet Neurology. “Andrieu S, Guyonnet S, Coley N, et al. "Effect of long-term omega 3 polyunsaturated fatty acid supplementation with or without multidomain intervention on cognitive function in elderly adults with memory complaints (MAPT): a randomised, placebo-controlled trial." Lancet Neurol, 2017. PubMed. DOI: 10.1016/S1474-4422(17)30040-6.” pubmed.ncbi.nlm.nih.gov
8. The Lancet. “Moll van Charante EP, Richard E, Eurelings LS, et al. "Effectiveness of a 6-year multidomain vascular care intervention to prevent dementia (preDIVA): a cluster-randomised controlled trial." Lancet, 2016. PubMed. DOI: 10.1016/S0140-6736(16)30950-3.” pubmed.ncbi.nlm.nih.gov
9. Cochrane Database of Systematic Reviews. “Hafdi M, Hoevenaar-Blom MP, Richard E. "Multi-domain interventions for the prevention of dementia and cognitive decline." Cochrane Database Syst Rev, 2021. PubMed. DOI: 10.1002/14651858.CD013572.pub2.” pubmed.ncbi.nlm.nih.gov
10. Alzheimer's & Dementia. “Sakurai T, Sugimoto T, Akatsu H, et al. "Japan-Multimodal Intervention Trial for the Prevention of Dementia: A randomized controlled trial." Alzheimers Dement, 2024. PubMed. DOI: 10.1002/alz.13838.” pubmed.ncbi.nlm.nih.gov
11. Alzheimer's & Dementia. “Espeland MA, Neiberg RH, Rapp SR, et al. "Does lifestyle intervention lower clinically significant cognitive impairment risk?" Alzheimers Dement, 2026. PubMed. DOI: 10.1002/alz.71668.” pubmed.ncbi.nlm.nih.gov
12. JAMA. “Baker LD, Espeland MA, Whitmer RA, et al. "Structured vs Self-Guided Multidomain Lifestyle Interventions for Global Cognitive Function: The US POINTER Randomized Clinical Trial." JAMA, 2025. PubMed. DOI: 10.1001/jama.2025.12923.” pubmed.ncbi.nlm.nih.gov
13. The Lancet. “Crivelli L, Suemoto CK, Sosa AL, et al. "Multidomain lifestyle intervention for the prevention of cognitive decline in at-risk older adults in Latin America (LatAm-FINGERS): a single-blind, multicentre, randomised controlled trial." Lancet, 2026. PubMed. DOI: 10.1016/S0140-6736(26)01278-X.” pubmed.ncbi.nlm.nih.gov
14. JAMA Neurology. “Harrison TM, Harvey DJ, Chadwick T, et al. "Brain Imaging Biomarkers and Cognitive Outcomes in a Multidomain Lifestyle Intervention: The POINTER Imaging Ancillary Study." JAMA Neurol, 2026. PubMed. DOI: 10.1001/jamaneurol.2026.0832.” pubmed.ncbi.nlm.nih.gov
15. JAMA. “SPRINT MIND Investigators for the SPRINT Research Group; Williamson JD, Pajewski NM, Auchus AP, et al. "Effect of Intensive vs Standard Blood Pressure Control on Probable Dementia: A Randomized Clinical Trial." JAMA, 2019. PubMed. DOI: 10.1001/jama.2018.21442.” pubmed.ncbi.nlm.nih.gov
16. Nature Medicine. “He J, Zhao C, Zhong S, et al. "Blood pressure reduction and all-cause dementia in people with uncontrolled hypertension: an open-label, blinded-endpoint, cluster-randomized trial." Nat Med, 2025. PubMed. DOI: 10.1038/s41591-025-03616-8.” pubmed.ncbi.nlm.nih.gov