Maybe a parent’s dementia diagnosis is what got you walking every morning. Or maybe it was a headline promising that exercise cuts the risk of Alzheimer’s disease by 40 percent. Either way, you’re probably counting on those walks to guard your memory.
Researchers have tested that idea directly. A trial called LIFE enrolled 1,635 inactive Americans aged 70 to 89 at eight centers and randomly assigned about half of them to two years of structured walking, strength and flexibility training. At the end, on a timed symbol-matching test of mental speed, the exercise group averaged 46.26 points. The comparison group averaged 46.28.
That result doesn’t make your walks a waste of time. Exercise ranks among the best-supported habits in medicine, and it measurably sharpens thinking for a while after each session. But the promise of long-term memory protection rests on a different kind of evidence than most headlines admit, and that gap explains most of this subject.
The guideline’s promise and the evidence underneath it
The federal Physical Activity Guidelines for Americans, second edition, published by the Department of Health and Human Services in 2018, say plainly that physical activity lowers the risk of developing cognitive impairment, including Alzheimer’s disease. A government summary updated in November 2025 still treats that edition as current.
Behind those guidelines sits a scientific report from an advisory committee that graded each conclusion as strong, moderate, limited or not assignable. On thinking and the brain, its grades split three ways:
- A single session of moderate-to-vigorous activity gives thinking a temporary lift. Graded strong.
- Regular activity programs improve scores on thinking tests in adults over 50. Graded moderate.
- People who do more activity are less likely to develop cognitive impairment, including Alzheimer’s. Graded strong.
The third conclusion is the one headlines use, and the wording shifts between documents. The guidelines say activity lowers risk; the committee’s evidence says more activity is associated with lower risk. An association is a pattern found by following groups of people over years. It shows that two things travel together, not that one causes the other.
Scientists who carried out a review for the committee spelled out the difference in a 2019 paper in Medicine & Science in Sports & Exercise. The evidence on test scores came mainly from randomized trials, studies that assign people to exercise or not by chance. The dementia evidence came from prospective observational studies that mostly relied on people’s own reports of how active they were. On the dementia question, the review’s authors wrote in that journal that “the prospective observational literature was unequivocal…” in linking more activity with lower risk. Yet they also called for longer trials to find out whether exercise lowers the rate of new dementia.
The famous percentages come from that observational side. In meta-analyses the review cited, more activity went with a 38 percent lower risk of cognitive decline across 15 studies of more than 33,000 people. It went with a 40 percent lower risk of Alzheimer’s across 10 studies of more than 20,000. Nobody in those studies was assigned to exercise. The committee’s literature search also ended in February 2017, before several of the long trials below were published.
What happened when trials ran a year or longer
Four bodies of evidence have since tested the protection claim head-on.
Healthy but inactive older adults
LIFE, published in JAMA in 2015, compared its exercise program with health-education workshops that included gentle upper-body stretching. Participants were chosen because they were at risk of losing mobility, and thinking was a planned secondary measure. After 24 months, the groups matched on mental speed and on delayed recall of a 12-word list, 7.22 words against 7.25. New mild cognitive impairment or dementia turned up in 13.2 percent of the exercise group and 12.1 percent of the education group, a difference consistent with chance.
People aged 80 and older, and those with the weakest physical performance at the start, did show better changes with exercise in executive function, the skills used for planning, switching between tasks and holding back impulses. That is a lead for future work, not a finding; a trial’s overall result outweighs any slice of it.
People with mild memory problems
EXERT, published in Alzheimer’s & Dementia in April 2025, enrolled 296 inactive older adults with amnestic mild cognitive impairment, a stage of measurable memory trouble that falls short of dementia. For 18 months, supervised for the first 12, half did moderate-to-high intensity aerobic training and half did low-intensity stretching, balance and range-of-motion work. Attendance was high, 81 and 87 percent of sessions.
Thinking scores did not differ between the groups. Neither group declined over the first year, and brain scans showed small 12-month losses of hippocampal volume that did not differ between the groups, averaging 0.51 percent across both. Because there was no group that stayed inactive, the trial can’t say whether both programs helped or whether these volunteers would have held steady anyway.
People living with dementia
DAPA, published in The BMJ in 2018, enrolled 494 people with mild-to-moderate dementia in England, average age 77. They were randomly assigned to usual care or to four months of supervised aerobic and strength training plus support to keep going. Fitness improved: walking distance on a six-minute test rose by about 18 meters within six weeks.
Cognition did not. At 12 months, scores on a standard impairment scale, where higher means worse, averaged 25.2 in the exercise group and 23.8 with usual care. The authors called the difference small and of uncertain clinical meaning, but it pointed the wrong way, and they concluded the program does not slow cognitive impairment.
The long trials pooled
A 2024 meta-analysis in Ageing Research Reviews gathered 16 trials lasting at least 12 months, covering 11,402 older adults without dementia, average age 73. For exercise alone, the effect on overall thinking was 0.05 standard deviations, about one-twentieth of the natural spread in scores and not distinguishable from zero, with moderate certainty. Exercise did not change the risk of mild cognitive impairment. The estimate for dementia leaned toward fewer cases, but its range ran from 75 percent fewer to 52 percent more, so it can’t be read either way.
Programs combining exercise with other lifestyle changes showed a small edge in the same review; those programs, and the large trials testing them, belong to our report on preserving brain function.
Put together, the narrow claim that a year or two of exercise alone improves thinking in older adults, beyond what an active comparison group gets, has been tested and so far not supported. That is not the same as saying exercise doesn’t help the brain, which none of these trials showed.
Why active people look protected anyway
If the long trials are flat, why do the long-term observational studies keep finding that active people get dementia less often? Three explanations fit the evidence.
The illness may come first. The brain changes behind dementia are under way long before a diagnosis, and that silent phase may itself bring a drop in physical activity. A 2017 report in The BMJ, from researchers at University College London and France’s national health research institute, tested this in 10,308 London civil servants first enrolled at ages 35 to 55. Their activity was logged seven times between 1985 and 2013, and 329 developed dementia over an average of 27 years. Overall, activity showed no link to later dementia. But among people eventually diagnosed, activity started tapering off as much as nine years before the diagnosis.
The authors wrote in The BMJ: “This study found no evidence of a neuroprotective effect of physical activity.” That is one cohort, and the pooled studies above did find lower risk. Its lesson is about timing. A study that measures activity only a few years before diagnosis may be catching an early effect of the disease on daily habits rather than a cause of it.
Comparison groups move too. LIFE’s control group attended workshops and stretched. EXERT’s did balance and flexibility work. Trials have to compare exercise with something, and social contact, attention and light movement may help both arms. That makes a trial a tougher test than a cohort, which simply compares more active people with less active ones.
Short trials report bigger effects. A 2025 umbrella review in the British Journal of Sports Medicine pooled 133 reviews covering 2,724 trials and more than 258,000 participants of all ages. It found exercise improved general cognition by 0.42 standard deviations, memory by 0.26 and executive function by 0.24. The largest gains came from programs lasting one to three months and from exergames, video games that require movement, and effects were larger in children and teens than in adults.
Set that beside the 0.05 from trials lasting a year or more. One plausible reading is that exercise produces early gains on tests that don’t build into lasting protection. Another is that small, short studies tend to report larger effects than big, long ones. Either way, the long trials are the closer test of what most adults want to know.
What exercise measurably does for thinking
The best-established benefit is also the least advertised: thinking improves for a while after a single bout of activity.
The 2019 review found effects across six earlier reviews ranging from barely detectable to moderate. Researchers express these as standardized effect sizes, where 0.2 is usually called small and 0.5 medium. The gains were most consistent for executive function. Bouts of 11 to 20 minutes showed the largest benefit, and the effect was strongest in preadolescent children and older adults.
Regular programs show a modest effect in shorter trials of people over 50. A 2018 meta-analysis in the British Journal of Sports Medicine, covering 39 trials with a literature search that ran to November 2016, found an overall effect of 0.29, a small one. Aerobic, strength, combined and tai chi programs all had positive estimates. An earlier Cochrane review of 12 aerobic-only trials, lasting 8 to 26 weeks, found no benefit in any thinking domain, even in trials where fitness clearly improved. Its search ended in 2013, so it reflects an older, more unsettled field.
A strength-training trial published in Archives of Internal Medicine in 2010 gives a sense of scale. Among 155 women aged 65 to 75 in Vancouver, a year of once- or twice-weekly resistance training improved performance on a test of selective attention by 12.6 and 10.9 percent. A balance-and-tone group slipped by 0.5 percent.
For people with dementia, the committee graded the evidence that activity improves cognition as moderate, citing among other reviews a meta-analysis of 18 small trials with 802 patients. DAPA alone enrolled more than half as many people as those 18 trials combined, and its result went the other way. When a large, well-controlled trial and a pool of small ones disagree, the large trial deserves more weight.
Brain size and BDNF are clues, not results
Two findings travel together in nearly every story about exercise building your brain. Both are real, and neither is a memory you would notice.
The first is BDNF, brain-derived neurotrophic factor, a protein that supports the growth and survival of nerve cells. A 2015 meta-analysis in the Journal of Psychiatric Research pooled 29 studies of 1,111 people. It found a moderate rise in BDNF levels after a single session of exercise, a larger rise in people who exercised regularly, and only a small rise in resting levels after a training program. Studies with more women showed less change. A rising BDNF level is a possible mechanism, not an outcome.
The second is size. In a 2011 trial in Proceedings of the National Academy of Sciences, 120 older adults were randomly assigned to aerobic training or a comparison program. The front part of the hippocampus, a structure central to memory, grew by about 2 percent with aerobic training, which the authors equated to reversing one to two years of age-related shrinkage. The comparison group’s volume declined, and larger volume went with higher blood BDNF. The 2019 review noted that few large studies have tested hippocampal size, which leaves its long-term meaning unclear.
Brain volume isn’t a simple more-is-better score, either. In the Vancouver strength trial, both training groups improved their attention scores while showing greater reductions in whole-brain volume than the comparison group. Whether adults grow new neurons at all is a separate argument, traced in our neuroplasticity piece.
On the library’s five-verdict scale, the BDNF rise and the hippocampal growth count as measured, but they are mechanisms, not outcomes. The outcome that matters, better memory over the long run, is what the long trials tested, and so far it is not supported.
A translation guide for exercise-and-brain headlines
A few phrases recur in nearly every story on this subject. Each tends to signal one kind of study, and each has a question that tests it.
| When the headline says | It is usually reporting | The question that tests it |
|---|---|---|
| Linked to or associated with lower dementia risk | A cohort that followed people for years | Was activity measured long before diagnosis, or close enough that early illness could explain it? |
| Cuts risk by 40 percent | Pooled observational studies | Was anyone assigned to exercise, or were people simply asked how active they were? |
| Improves memory | Often a trial lasting one to six months | What did the comparison group do, and did anyone retest a year later? |
| Grows the brain or boosts BDNF | A scan or a blood marker | Did anything change that the participants or their families would notice? |
| Stalls decline | Sometimes a comparison with people outside the trial | Did the comparison group come from the same randomized trial, or from a different study? |
If a story can’t answer the last column, treat the claim as a hypothesis. On intensity, the evidence points one way. EXERT found moderate-to-high intensity aerobic training no better than stretching, the 2025 umbrella review found larger effects from low- and moderate-intensity programs, and the advisory committee could not grade any dose.
When a thinking problem needs a clinician, not a workout plan
A training plan can’t tell you why your thinking has changed. These situations call for a professional decision.
- Memory or thinking changes that you or your family have noticed. Whether a change needs formal assessment is a clinician’s call. The National Institute on Aging notes that some medicines and combinations of medicines can cause confusion and memory loss, one of several causes a clinician can look for. How that evaluation works is the subject of our coming article on when memory problems call for assessment.
- A chronic condition or new symptoms. The federal guidelines say adults with chronic conditions or symptoms should be under the care of a health care provider. How hard and how much to exercise with heart disease, lung disease or balance problems is a decision for the clinician managing that condition, or a physical activity specialist they recommend.
- A relative living with dementia. DAPA shows that an intensive program may not protect thinking. What a program is meant to achieve, whether fitness, mood or getting out of the house, is worth settling with the person’s care team before it starts.
- Sudden confusion, trouble speaking, one-sided weakness or numbness, trouble seeing or walking, or a severe headache with no known cause. These are stroke warning signs. The Centers for Disease Control and Prevention’s advice is to dial 911 at once, keep track of when the symptoms started, and let an ambulance do the driving.
The reasons to move that don’t depend on memory
Nothing in the long trials argues against exercise, and the federal guidelines list many benefits beyond the brain. For adults, regular activity lowers the risk of early death, heart disease, stroke, high blood pressure, type 2 diabetes, depression and eight types of cancer. It improves bone health, physical function and quality of life, and in older adults it lowers the risk of falls. Some benefits arrive the same day, including less anxiety, lower blood pressure and better sleep.
The National Institute on Aging’s summary is candid: several studies support a connection between activity and brain health, and more research is needed to determine what role exercise plays in preventing cognitive decline. Physical inactivity does appear among the 14 modifiable dementia risk factors in the 2024 Lancet Commission report, which describes the link between exercise and dementia as likely running in both directions. The full list is unpacked in our article on the risk factors people rarely guess. How much activity to aim for is covered in the library’s guides to exercise amounts.
What to take from all this fits in one sorting rule. Count the many benefits the guidelines list and the sharper spell after a brisk walk as the case for moving, and treat long-term memory protection as a reasonable hope the trials have not yet confirmed. When the next headline promises more, check whether anyone in the study was assigned to exercise.
Sources
- US Department of Health and Human Services. “Physical Activity Guidelines for Americans, 2nd edition (2018).” odphp.health.gov
- Office of Disease Prevention and Health Promotion (HHS). “Top 10 Things to Know About the Second Edition of the Physical Activity Guidelines for Americans (last updated November 19, 2025).” odphp.health.gov
- 2018 Physical Activity Guidelines Advisory Committee (HHS). “2018 Physical Activity Guidelines Advisory Committee Scientific Report, Part F, Chapter 3: Brain Health.” odphp.health.gov
- Medicine & Science in Sports & Exercise. “Erickson KI, Hillman C, Stillman CM, et al. "Physical Activity, Cognition, and Brain Outcomes: A Review of the 2018 Physical Activity Guidelines." Med Sci Sports Exerc, 2019. PubMed. DOI: 10.1249/MSS.0000000000001936.” pubmed.ncbi.nlm.nih.gov
- JAMA. “Sink KM, Espeland MA, Castro CM, et al. "Effect of a 24-Month Physical Activity Intervention vs Health Education on Cognitive Outcomes in Sedentary Older Adults: The LIFE Randomized Trial." JAMA, 2015. PubMed. DOI: 10.1001/jama.2015.9617.” pubmed.ncbi.nlm.nih.gov
- Alzheimer's & Dementia. “Baker LD, Pa JA, Katula JA, et al. "Effects of exercise on cognition and Alzheimer's biomarkers in a randomized controlled trial of adults with mild cognitive impairment: The EXERT study." Alzheimers Dement, 2025. PubMed. DOI: 10.1002/alz.14586.” pubmed.ncbi.nlm.nih.gov
- The BMJ. “Lamb SE, Sheehan B, Atherton N, et al. "Dementia And Physical Activity (DAPA) trial of moderate to high intensity exercise training for people with dementia: randomised controlled trial." BMJ, 2018. PubMed. DOI: 10.1136/bmj.k1675.” pubmed.ncbi.nlm.nih.gov
- Ageing Research Reviews. “Reparaz-Escudero I, Izquierdo M, Bischoff-Ferrari HA, Martínez-Lage P, Sáez de Asteasu ML. "Effect of long-term physical exercise and multidomain interventions on cognitive function and the risk of mild cognitive impairment and dementia in older adults: A systematic review with meta-analysis." Ageing Res Rev, 2024. PubMed. DOI: 10.1016/j.arr.2024.102463.” pubmed.ncbi.nlm.nih.gov
- The BMJ. “Sabia S, Dugravot A, Dartigues JF, et al. "Physical activity, cognitive decline, and risk of dementia: 28 year follow-up of Whitehall II cohort study." BMJ, 2017. PubMed. DOI: 10.1136/bmj.j2709.” pubmed.ncbi.nlm.nih.gov
- British Journal of Sports Medicine. “Singh B, Bennett H, Miatke A, et al. "Effectiveness of exercise for improving cognition, memory and executive function: a systematic umbrella review and meta-meta-analysis." Br J Sports Med, 2025. PubMed. DOI: 10.1136/bjsports-2024-108589.” pubmed.ncbi.nlm.nih.gov
- British Journal of Sports Medicine. “Northey JM, Cherbuin N, Pumpa KL, Smee DJ, Rattray B. "Exercise interventions for cognitive function in adults older than 50: a systematic review with meta-analysis." Br J Sports Med, 2018 (published online 2017). PubMed. DOI: 10.1136/bjsports-2016-096587.” pubmed.ncbi.nlm.nih.gov
- Cochrane Database of Systematic Reviews. “Young J, Angevaren M, Rusted J, Tabet N. "Aerobic exercise to improve cognitive function in older people without known cognitive impairment." Cochrane Database Syst Rev, 2015. PubMed. DOI: 10.1002/14651858.CD005381.pub4.” pubmed.ncbi.nlm.nih.gov
- Archives of Internal Medicine. “Liu-Ambrose T, Nagamatsu LS, Graf P, Beattie BL, Ashe MC, Handy TC. "Resistance training and executive functions: a 12-month randomized controlled trial." Arch Intern Med, 2010. PubMed. DOI: 10.1001/archinternmed.2009.494.” pubmed.ncbi.nlm.nih.gov
- Journal of Psychiatric Research. “Szuhany KL, Bugatti M, Otto MW. "A meta-analytic review of the effects of exercise on brain-derived neurotrophic factor." J Psychiatr Res, 2015 (published online 2014). PubMed. DOI: 10.1016/j.jpsychires.2014.10.003.” pubmed.ncbi.nlm.nih.gov
- Proceedings of the National Academy of Sciences. “Erickson KI, Voss MW, Prakash RS, et al. "Exercise training increases size of hippocampus and improves memory." Proc Natl Acad Sci U S A, 2011. PubMed. DOI: 10.1073/pnas.1015950108.” pubmed.ncbi.nlm.nih.gov
- National Institute on Aging. “Cognitive Health and Older Adults (content reviewed June 11, 2024).” nia.nih.gov
- Centers for Disease Control and Prevention. “Signs and Symptoms of Stroke.” cdc.gov
- The Lancet (accepted manuscript, UCL Discovery). “Livingston G, Huntley J, Liu KY, et al. "Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission." Lancet, 2024. DOI: 10.1016/S0140-6736(24)01296-0.” discovery.ucl.ac.uk