More Activity, Less Dementia—Unless It’s Your Job

We’re often told that moving more is one of the best things you can do for your brain. Walk, cycle, garden, take the stairs.

But a new review of over 4 million people suggests the type of activity may matter. Not all movement was linked to the same benefit.

Exercise in your free time went with a lower risk of dementia. Physical activity at work went with a higher one.

Does exercise lower dementia risk?

Overall, yes, it seems to. People who are more physically active tend to have a lower risk of dementia, and this has been found in many studies.

The NHS lists physical inactivity among the risk factors for dementia. It advises at least 150 minutes of moderate-intensity aerobic activity each week, such as brisk walking, cycling or dancing.

It also recommends strengthening exercises at least twice a week, and sitting less.

The new review adds a twist. Physical activity happens in four main places: free time, work, getting around, and housework.

And they weren’t all linked to the same result.

The study: 74 studies, 4.2 million people

What the researchers did

The review was published in The Lancet Public Health in September 2026. It was funded by the US National Institutes of Health and Brazilian research agencies.

The team combined 57 studies from an earlier review with 17 newer ones. Each followed people over time and recorded who developed dementia.

The four kinds of activity

The researchers split physical activity into leisure-time activity, such as sport and walking for fun, and activity at work. They also looked at commuting, such as walking or cycling to work, and housework.

Most earlier research only looked at total activity or leisure-time activity.

That’s a big gap. Worldwide, leisure time makes up only 12% of people’s physical activity, the authors note.

Work and housework account for 52%, and commuting for 36%.

In low-income countries, leisure is just 4% of all activity, compared with 28% in high-income countries. So for most people, most movement isn’t exercise at all.

What the study found

More activity, lower risk overall

Compared with the least active people, the most active had a 20% lower risk of dementia overall. The risk of Alzheimer’s disease was 21% lower, and vascular dementia 29% lower.

But the type of activity mattered

Leisure-time activity showed a benefit that grew with more activity, then levelled off.

For work, the risk rose with more physical activity, from low levels up to very heavy physical work.

The work result held up

The researchers checked whether one study was driving the result. Removing studies one at a time, the work link kept pointing the same way, though it was weaker in some checks.

The higher risk with physical work also held after taking total activity into account. And they only included working people in those comparisons, to avoid mixing in people too ill to work.

Why would physical work raise the risk?

The authors don’t think lifting boxes harms the brain directly. Instead, they suggest physical jobs often come with other burdens.

Many physically demanding jobs involve high stress, hazardous exposures and little control over your work. They can also leave less time and energy for leisure activity.

These jobs often mean more exposure to air and noise pollution, both linked to dementia. And repetitive, low-complexity tasks may build less “cognitive reserve”, the brain’s resilience.

Active commuting may carry similar burdens. It’s often a necessity rather than a choice, and it can mean breathing more polluted air.

Why the result needs caution

Low certainty

The researchers rated the certainty of evidence as low to very low. The results varied a lot between studies.

Few studies for some kinds of activity

The work result came from five studies. Commuting came from two, and housework from just one.

These are early signals, not firm findings.

Links, not proof

All the studies were observational. The authors say reverse causation “cannot be excluded”: early, undiagnosed brain changes could make people less active years before dementia is found.

Self-reported activity

Most studies asked people how active they were, rather than measuring it. People often misremember or overstate their activity.

What earlier research found

The review it built on

A 2022 meta-analysis in the British Journal of Sports Medicine pooled 58 studies. Physical activity was linked to a 20% lower risk of dementia, even in studies that followed people for 20 years or more.

That longer follow-up matters, because it makes reverse causation less likely.

The physical activity paradox

The idea that work activity behaves differently isn’t new. A 2018 meta-analysis in the British Journal of Sports Medicine pooled data on 193,696 people.

It found men with highly physical jobs had an 18% higher risk of early death.

Researchers call this the “physical activity paradox”: leisure activity helps, but heavy physical work doesn’t seem to.

How it fits together

Exercise in your free time is consistently linked to better brain health. Physical work seems to be a different story, likely because of everything that comes with it.

How much should you trust this?

Early. It’s a very large review, but the underlying studies are observational and the certainty of evidence is low.

What makes it convincing

• It pooled 74 studies and over 4 million people.
• Follow-up lasted a median of 10 years.
• The work result held in several checks.
• It matches earlier findings on the physical activity paradox.

What makes me cautious

• Certainty of evidence was low to very low.
• The work, commuting and housework results rest on one to five studies.
• Most studies relied on people reporting their own activity.
• Nearly all studies came from high-income countries.
• It can’t rule out reverse causation.

What this means for you

• Make time for exercise you choose. Leisure-time activity was linked to the lowest risk.
• Don’t count your job as your workout. A physically demanding job may not bring the same benefits as exercise you enjoy.
• If your work is physically hard, rest and recovery matter. Some leisure-time activity may still help, even if your job is tiring.
• Follow the NHS guidance. At least 150 minutes of moderate activity a week, plus strengthening exercises twice a week.
• Look after other risk factors too. The NHS also lists smoking and air pollution among the factors linked to dementia.

The NHS guide to dementia prevention explains what you can do to lower your risk.

In this short CBS Mornings report, you can see how physical activity can boost brain health and reduce dementia risk:

What we still don’t know

1. Why is physical work linked to higher risk? Stress, pollution and limited leisure time are possibilities, not proven causes.
2. Does commuting really raise risk? Only two studies looked.
3. What about low- and middle-income countries? Only 7% of studies came from them.
4. Would measured activity tell a different story? Devices can’t yet reliably tell work from leisure.
5. Can exercise prevent dementia? That needs long-term trials, which are rare.

My take: free-time movement is the kind that counts

What I like about this study is that it questions an easy assumption: that all movement is equally good. That’s a fresh and useful angle.

I’m less convinced by the precise numbers for commuting and housework. They rest on one or two studies.

Still, the main lesson seems sound. Exercise you choose, in your own time, is the kind most linked to a healthier brain.