Leisure exercise was linked to 24% lower dementia risk and physical work to 20% higher across 74 studies

Public health advice on exercise has settled on a generous message: every move counts. The World Health Organization’s guidelines hold that some physical activity is better than none, and they make no distinction between a jog in the park and a shift spent lifting boxes. For the heart, researchers have doubted that equivalence for more than a decade. A new review asks the same question about the brain.

Natan Feter, David Raichlen and colleagues at the University of Southern California gathered every long-term study they could find that linked physical activity to later dementia and sorted the results by where the activity took place. Their analysis, published in The Lancet Public Health in August, covers 74 studies and more than four million adults. Leisure-time physical activity was associated with a reduced risk of dementia, whereas occupational physical activity was linked with an increased risk of dementia.

Four kinds of physical activity, and why the type matters for dementia research

Researchers divide physical activity into four domains: leisure, work, household tasks and getting from place to place. Almost all the dementia research has looked at the first. That is a narrow slice of how humans move. According to the review, leisure-time activity represents only 12% of total physical activity worldwide, with work and housework making up about half and travel most of the rest. In low-income countries the leisure share falls to 4%.

So when studies report that active adults get less dementia, they are mostly describing those who choose to exercise in their free time. Whether the same holds for a builder, a care assistant or someone who walks an hour to work because there is no alternative has rarely been examined, and the answers that existed were scattered across individual papers.

How the review compared leisure, work and other activity

The team searched six research databases for cohort and case-control studies that measured physical activity in adults whose thinking had been assessed at the start, and then followed them for at least a year. They combined 17 newly found studies with 57 from an earlier review by some of the same authors. The result included data from more than 4 million adults across 30 countries, followed for a median of ten years, and more than 64,000 cases of dementia.

Each study compared its most active participants with its least active. The reviewers pooled those comparisons, first for all activity together and then separately by domain. Leisure was the best covered, with 20 studies. Work activity was examined in five, commuting in two and household activity in one.

Leisure activity went with less dementia, physical work with more

Taken as a whole, the results repeated what earlier reviews had found: the most active participants had about a fifth lower risk of dementia than the least active. The rates of Alzheimer’s disease and of vascular dementia, the form caused by damage to the brain’s blood supply, were lower too.

Split by domain, the picture divided. High leisure-time activity went with a 24% lower risk of dementia. High activity at work went with a 20% higher risk. The single household study pointed toward lower risk and the two commuting studies toward slightly higher risk, though the authors caution that results for household and commuting physical activity were limited to a total of three studies.

The two best-studied domains also behaved as mirror images when the reviewers looked at dose. Dementia risk fell as leisure activity rose and then leveled off. With work activity it climbed and then leveled off.

“The findings challenge the long-held assumption that ‘every move counts’ for brain health,” Feter said in a statement reported by Newsweek.

Why would physical work be linked to more dementia?

The authors do not think lifting and carrying damage the brain. Their explanation is about what comes with the job. They write that many physically demanding jobs involve heightened psychosocial stress, hazardous occupational exposures, and reduced autonomy, along with more exposure to air and noise pollution, both of which have been tied to dementia. Such work also leaves less time and energy for exercise done by choice.

A second idea concerns cognitive reserve, the brain’s capacity to keep functioning as damage accumulates, which is thought to be built up by education and mentally demanding activity. “Occupational physical activity often involves repetitive, low-complexity tasks that may limit the accumulation of cognitive reserve,” Feter told Newsweek.

Put together, the authors conclude that the link might partly reflect an accumulation of social, economic, and environmental adversities and cannot be pinned on the physical effort itself. Physical work, on this reading, is a marker of a harder life.

A Norwegian study gives a sense of which jobs are involved. Ekaterina Zotcheva and colleagues at the Norwegian Institute of Public Health used national registers to trace the working lives of 7,005 Norwegians from their thirties to their sixties, then assessed them for dementia after 70. Those who had spent decades in physically active work had about a third higher risk of dementia than those in the least active jobs. The common occupations in that group were sales staff, nursing and care assistants, and farmers. Zotcheva’s team described these jobs as often marked by a lack of autonomy, prolonged standing, hard work, rigid working hours and stress.

There is also a statistical trap in this research, and it can flip the result. If those in physical jobs are compared with everyone else, the comparison group includes some who are not working at all, often because of poor health, and workers look healthy by contrast. The review describes an earlier analysis in which work activity appeared protective until the comparison was limited to those with jobs, at which point the association was inverted in women and attenuated in men. The studies of work activity in the new review were all restricted to working participants.

The physical activity paradox was first described in heart disease

The idea that work and leisure activity part ways has a name, the physical activity paradox, and it comes from cardiovascular research. In 2021 Andreas Holtermann, of Denmark’s national institute for research on the working environment, reported on 104,046 adults in the Copenhagen General Population Study in the European Heart Journal. Those who moved more in their leisure time had fewer heart attacks and strokes and lower death rates over ten years. The same study found that higher occupational physical activity associates with increased risks, and that the two patterns held independently of each other.

Dementia had been examined in single cohorts before. In the Copenhagen Male Study, Kirsten Nabe-Nielsen and colleagues followed 4,721 male employees who had described their jobs in 1970 and 1971. Men in the most physically demanding work were about half again as likely to develop dementia as men in sedentary jobs. The Southern California review is the first to pull such studies together and set them beside the leisure evidence.

That leisure evidence is itself fairly sturdy for an observational finding. A worry with dementia research is that the disease begins years before diagnosis and slows its sufferers down, so inactivity can look like a cause when it is an early symptom. The 2022 review that the new one builds on, led by Paula Iso-Markku at the University of Helsinki, addressed this by looking at study length, noting that much of the evidence comes from short follow-ups prone to reverse causation. Activity was still associated with less dementia in studies that had followed their participants for 20 years or more.

How strong is the evidence on work activity and dementia?

Weaker than the headline numbers suggest, and the authors say so. Their own grading found that certainty of evidence ranged from very low to low across domains.

Several things pull it down. Most studies asked participants to report their own activity, and recall of that kind is unreliable. The work finding rests on five studies. When the reviewers removed one study at a time and recalculated, leisure activity stayed protective throughout, while the work result kept its direction but in some versions was no longer distinguishable from chance. The individual studies also disagreed with one another far more than chance would explain, which means “leisure activity” and “work activity” are each covering very different things in different places.

Where the research was done matters as well. All but five of the 74 studies came from high-income countries, while the largest rises in dementia are expected elsewhere, in places where most activity is work, housework and travel on foot.

And because these are observational studies, they cannot separate a job from the income, schooling and neighborhood that tend to go with it. The Norwegian and Danish analyses adjusted for education and social position, and the association remained, but adjustment of that kind is never complete.

Where this leaves advice on physical activity and dementia

For anyone who exercises in their free time, nothing changes: the evidence that leisure activity goes with less dementia is as it was. The review does undercut the assumption that a physically hard job provides the same protection, and Newsweek reported that the researchers stressed that the results do not mean people should avoid physically active jobs.

The authors direct their conclusions at policy. They call for more studies in working populations and for measures that expand meaningful access to leisure-time physical activity. Raichlen put it this way: “Improving access to safe parks, walking trails and recreational facilities, ensuring people have adequate leisure time for exercise, and reducing harmful workplace exposures may ultimately prove as important as promoting physical activity itself.”

On evidence its own authors grade as low, activity chosen in free time goes with less dementia and activity required by a job goes with more.