Free-Time Exercise Went With 24% Lower Dementia Risk, While Hard Physical Work Went With 20% Higher Risk, in 4.2 Million Adults

Original study: Domain-specific physical activity levels and risk of dementia: a systematic review and meta-analysis
Natan Feter, Paula Iso-Markku, Tiladia Markarian, Dylan M Luong, Jack Gunnink, Natália Schröder, Eduardo Lucia Caputo, Jayne Feter, Anamika Nanda, Sarah Hourihan, Gustavo S da Silva, Ashley Ventura, Daniela Benzano Bumaguin, Yann C Klimentidis, Gene E Alexander, Daniel Umpierre, Pedro C Hallal, Airton José Rombaldi, David A Raichlen
The Lancet Public Health August 18, 2026
What the researchers wanted to know
Does it matter where your physical activity happens? The researchers compared activity in your free time, at work, doing household chores and getting from place to place, and asked whether each one goes with a lower or higher risk of dementia.
Why it matters
Most research on exercise and dementia has looked only at free-time exercise, or at all activity lumped together. Yet across the world, most of people's physical activity comes from work, chores and travel. Earlier research on the heart has found that hard physical work does not protect the way free-time exercise does, and no one had checked whether the same is true for dementia.
Dementia is a loss of memory and thinking skills that is bad enough to get in the way of daily life. Alzheimer's disease is the most common kind. Vascular dementia is caused by poor blood flow to the brain.
Researchers often split physical activity into four kinds:
- Free time: exercise and sports you choose to do
- Work: lifting, carrying and being on your feet as part of your job
- Home: housework and yard work
- Getting around: walking or cycling to get somewhere
All the studies combined here followed people over time and watched who developed dementia. This kind of study can show that two things go together, but not that one causes the other.
The researchers built on an earlier review that had gathered every study up to 2021. They searched six research libraries again for newer studies up to May 2026, using the same rules, and found 17 more. In all they combined 74 studies with 4.2 million adults from 30 countries. About half were women, and the typical person was 67 at the start. Everyone was free of dementia at the start, and older people had their memory and thinking checked first.
People were followed for about 10 years on average (anywhere from 5 to 21 years in most studies). In that time, 64,009 developed dementia.
In each study, people had been grouped by how active they were. The researchers compared the most active with the least active, first for all activity together, then separately for each of the four kinds of activity. For activity at work, they used only people who were still working.
Two or more researchers checked each step separately and rated the quality of every study. They also checked whether the results held up when low-quality studies were left out, when any single study was left out, and when overall activity was taken into account.
Being more active in your free time went with a lower risk of dementia. Being more active at work went with a higher risk.
Change in risk of dementia, most active compared with least active:
- Free time: 24% lower (20 studies)
- Home chores: 15% lower (1 study)
- Work: 20% higher (5 studies)
- Getting around: 10% higher (2 studies)
For all kinds of activity together, the most active people had a 20% lower risk of dementia, a 21% lower risk of Alzheimer's disease and a 29% lower risk of vascular dementia.
Studies that grouped people by whether they met the guideline of at least 150 minutes of moderate activity a week found a lower risk in those who met it. For Alzheimer's disease the risk was 35% lower, and for vascular dementia 38% lower.
More free-time activity went with lower risk up to a point, after which extra activity did not lower it further. For work, risk kept rising as people did more physical work.
The higher risk with physical work stayed after taking into account how active people were overall, and after leaving out low-quality studies. 69 of the 74 studies came from high-income countries. The researchers rated the certainty of the evidence as low to very low, mainly because the studies' results varied a lot and few studies looked at work, chores or getting around.
Conclusion
Being active in your free time goes with a lower risk of dementia, while doing a lot of physical work on the job goes with a higher risk.
Therefore
Where your activity happens may matter for your brain, not only how much you do. Making time and safe places for free-time exercise could help prevent dementia, and steps to make physically demanding jobs less harmful may matter as much.
How much can you trust this study?
Two studies can say the same thing and deserve very different amounts of trust. LayMed scores every study out of 100 on:
- Study design
- Bias Control & Transparency
- Size and precision
- Measurement
LayMed Plus members see the Study Quality score for every study, with the reason behind each part.
Get LayMed Plus$7.99 a month or $79 a year
How far does this apply to you personally?
A study can earn a high quality score and still not apply well to you. How much its results carry over to you can change a lot depending on your sex, age, race or ethnicity, body weight, genes, hormones, blood pressure and resting heart rate, health conditions, medications, activity level, training experience, diet, sleep, smoking and drinking, job and gut bacteria.
Hover over or tap any of these to see a real study where it changed the results.
LayMed Plus members get a personal Relevance score for every study. It weighs all of these factors to estimate how well the results apply to you and your life.
Get LayMed Plus$7.99 a month or $79 a year
Important new studies, explained in plain English, sent straight to your inbox.
Almost done. Please check your email for a confirmation sent from "laymed@dr-gains.com".
Last checked against the source: November 18, 2026
Update November 18, 2026: Written to the LayMed article standard from the full published paper.
LayMed articles are plain-English summaries of published research, written for education. They are not medical advice. Always read the original study and talk to a qualified professional about your own situation.