Not Every Step Counts the Same: What 4 Million People Reveal About Movement and Dementia

For years the advice has been simple and reassuring: move more, and your brain will thank you. Every step counts. A large new review, published in The Lancet Public Health in August 2026, adds a wrinkle to that story. Movement still helps. But where it happens, and why, seems to change what it does for you.

What the researchers did

A team led by David Raichlen at the University of Southern California, with Natan Feter as first author, gathered 74 studies from 30 countries. Together these studies followed more than 4.2 million adults aged between 45 and 93, for a median of ten years, and recorded who went on to develop dementia, including Alzheimer's disease and vascular dementia.

What made this review different is that the researchers refused to treat physical activity as one thing. They split it into domains: activity in your leisure time, such as walking, cycling, swimming, dancing or sport; activity at work; activity around the house; and active commuting, such as walking or cycling to a job. Then they asked, for each domain separately, how the most active people compared with the least active.

What they found

Leisure-time physical activity was associated with a 24 percent lower risk of dementia. That fits everything you have already heard, and it held for all-cause dementia, Alzheimer's disease and vascular dementia alike. The benefit grew with more activity, up to a point, and then levelled off.

Physical activity on the job told a different story. People whose work was physically demanding had a 20 percent higher risk of dementia than those whose work was not. That pattern also grew with the amount of occupational activity before it plateaued.

Household activity looked helpful, but that finding rests on a single study. Active commuting was linked to a slightly higher risk, but that rests on only two studies. The authors are careful to say that neither of these conclusions is firm.

Why work and play might differ

Researchers have seen this so-called physical activity paradox before in heart health: a hard physical job does not seem to deliver the benefits of a brisk walk after dinner, and sometimes seems to do the opposite. This is the first review to ask whether the same holds for dementia.

The authors offer several explanations, and stress that no single one is likely to explain the whole picture. Physical jobs often involve repetitive, low-complexity tasks that give the brain little new to work on. They frequently take place in noisy or polluted environments. They tend to come with more stress, less control over the day, and lower pay. And people who spend the day lifting and carrying often have less energy, time or safe space for the kind of exercise they would choose for themselves. Leisure activity, by contrast, is usually varied, self-chosen and social. Feter put it plainly: the associations may partly reflect a lifetime of social, economic and environmental disadvantage rather than the physical work itself.

What this does not show

This is an analysis of observational studies. It can show that physical work and dementia tend to go together; it cannot show that physical work causes dementia. The authors say so directly, and add that nobody should read this as a reason to avoid a physically active job.

Almost every study in the review relied on people reporting their own activity, which is imperfect. Most of the studies came from high-income countries, while the largest rise in dementia cases over the coming decades is expected elsewhere. The household and commuting results, as noted, rest on one and two studies respectively, which is too thin to lean on.

Why it is still worth your attention

If you are over fifty, the encouraging part is straightforward. The activity that was consistently linked to a lower risk is the kind that is entirely in your hands: the walk you take because you want to, the swim, the dance class, the garden, the bike ride with a friend. That link held even after the researchers accounted for age, education and chronic disease.

The second lesson is about context. A walk in a park, with company and something new to look at, is a different experience from an hour of heavy, repetitive lifting under pressure, even if a fitness tracker counts both. Choose movement that you enjoy, that varies, and that puts you among people. If your working years were physically hard, none of this is a verdict on you. It is, if anything, an invitation to give yourself the kind of movement you may never have had time for.

Keeping an eye on your own thinking

At Treeny we build short daily exercises for adults over fifty, and Flo, our coach, keeps a quiet record of how your thinking moves over weeks rather than on any one day. A study like this is a reminder that your brain responds to what your days actually contain, and that the choices that matter most are often the small, voluntary ones. Whether you use Treeny or not, the walk you choose for yourself is worth taking.

Source: Feter, Natan, Raichlen, David A., et al. "Domain-specific physical activity levels and risk of dementia: a systematic review and meta-analysis." The Lancet Public Health, published online 18 August 2026.

A note

This article describes published research and is not medical advice. A review of observational studies shows patterns, not causes. If changes in your memory or thinking worry you, or if you are unsure what kind of physical activity is right for you, speak to your doctor, who knows your history and can tell you whether anything needs a closer look.