Regular leisure-time exercise such as walking, running, swimming and cycling may be associated with a significantly lower risk of dementia, while physical activity performed as part of a person’s job may be linked to a higher risk of the condition, researchers have found.

Researchers from the University of Southern California (USC) said the findings suggest that, when it comes to dementia risk, the context in which people are physically active may matter as much as the amount of activity they undertake.

The findings, published in The Lancet Public Health, are based on a meta-analysis of 74 cohort and case-control studies involving more than 4.2 million people across 30 countries.

For years, studies have linked regular physical activity with a lower risk of dementia. However, the researchers said treating all forms of physical activity alike could overlook important differences between recreational exercise and activity undertaken at work.

The researchers investigated whether the phenomenon known as the “physical activity paradox”, already documented in cardiovascular research, could also apply to dementia. The paradox refers to the observation that physically demanding occupations may not provide the same health benefits as voluntary exercise and may, in some circumstances, be associated with poorer health outcomes.

Participants in the studies were aged between 45 and 93 and were followed for a median of 10 years. During the follow-up period, researchers tracked diagnoses of all-cause dementia, Alzheimer’s disease and vascular dementia.

The analysis found that people who engaged in higher levels of leisure-time physical activity had a 24 per cent lower risk of dementia than those who were least active.

By contrast, occupational physical activity was associated with a 20 per cent higher risk of dementia.

The study also found indications that household physical activity could be beneficial, although the evidence was based on only one study. Active commuting, such as walking or cycling for transportation, was associated with a modest increase in risk, but those findings were based on only two studies.

“The findings challenge the long-held assumption that ‘every move counts’ for brain health,” said the study’s first author and a postdoctoral fellow in USC’s Evolutionary Biology of Physical Activity Lab, Natan Feter.

Feter said strenuous physical activity at work could affect dementia risk differently from leisure exercise because it often involves repetitive, low-complexity tasks that may limit the accumulation of cognitive reserve.

“It may also occur in environments characterised by hazardous exposures, such as air and noise pollution. In contrast, leisure-time activity is more likely to be self-selected, varied, and socially engaging,” he said.

The authors said the higher dementia risk associated with occupational physical activity was unlikely to be explained by physical movement itself. Instead, they suggested that the relationship could reflect broader social and environmental factors often associated with physically demanding work.

Workers in physically demanding jobs may also have fewer opportunities to participate in recreational exercise because of fatigue, time constraints or limited access to safe exercise spaces.

The analysis showed contrasting patterns between leisure and work-related physical activity. Dementia risk declined as leisure-time exercise increased before eventually reaching a plateau. Occupational physical activity showed the opposite pattern, with higher levels associated with progressively greater risk before the relationship also levelled off.

Feter said the findings called for greater attention to structural factors that influence people’s opportunities to engage in meaningful leisure-time physical activity.

“In the paper, we call for shifting focus away from individual-level behavioural changes and toward structural determinants, prioritising worker-centred policies and equitable urban planning that expand access to meaningful leisure-time physical activity,” he said.

The authors stressed, however, that the findings should not be interpreted as evidence that physically demanding jobs cause dementia.
Because the study was observational, it could identify associations but could not establish cause and effect.

The researchers also acknowledged several limitations. Evidence on housework and active commuting remains limited, most of the available studies were conducted in high-income countries, and nearly all relied on participants’ self-reported physical activity levels, which could introduce recall bias.

Despite the limitations, the researchers said the association remained consistent across multiple sensitivity analyses, strengthening their confidence that the findings warranted further investigation.

They said the results could have important implications for public health efforts to reduce dementia risk by improving access to safe and accessible spaces for recreational exercise.

Feter said the next stage of research should examine not only how often people engage in physical activity and sedentary behaviour, but also when and why they do so.

He called for further studies to investigate how domain-specific physical activity affects different occupational groups and to assess the long-term effects of occupational physical activity, sedentary behaviour and other work-related factors on brain health.

A professor of biological sciences and anthropology at USC and the study’s lead researcher, David Raichlen, said the context in which physical activity occurs may be as important as movement itself in protecting brain health.

“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,” he said.