Walking and Dementia Risk: What Two Real Studies Actually Found

"Walking is good for your brain" is the kind of advice that gets repeated so often it stops sounding like something researchers actually measured. Two separate studies did measure it, though - one tracking nearly 80,000 adults for almost seven years to see who developed dementia, the other following fewer than 300 older adults for closer to a decade to see how fast cognitive decline showed up once early Alzheimer's changes were already present in the brain. They ask different questions and land on different numbers. Here's what each one actually found, and what to do with both.

Does walking actually lower dementia risk?

The best evidence says yes, with a real dose-response relationship: the more people walked, up to a point, the lower their risk of developing dementia over the following years. That's not the same as proof that walking alone prevents dementia - the underlying research is observational, meaning it tracked what people already did rather than randomly assigning some people to walk more than others. But the pattern is consistent, it shows up at a meaningful scale, and it holds up across more than one independent dataset.

What the 2022 UK Biobank study found

The largest and most-cited evidence here comes from Borja del Pozo Cruz and colleagues, published in JAMA Neurology in 2022. The study followed 78,430 adults in the UK Biobank who wore a wrist-worn accelerometer to record their actual daily step counts, then tracked their health records for a median of 6.9 years. Participants were mostly in their 60s at the start (mean age 61.1), and by the end of follow-up, 866 of them - about 1.1% - had gone on to develop dementia.

The step-count pattern was clear: risk dropped steadily as daily steps rose, leveling off around 9,800 steps a day, which was associated with roughly a 50% lower risk of developing dementia compared with the least-active participants. You didn't need to reach that number to see a benefit, either - around 3,800 steps a day was already associated with about a 25% lower risk. The curve wasn't linear all the way up; benefit accumulated fastest at the lower end and flattened out as step counts climbed into five figures.

Does walking speed matter more than step count?

The same UK Biobank study also looked at intensity - specifically, each person's fastest sustained 30 minutes of walking in a day, a proxy for how brisk their typical walking pace was. Higher intensity strengthened the association independent of total volume: participants who walked at a brisker pace saw a similar-or-better risk reduction at a noticeably lower total step count. About 6,315 steps a day at that higher intensity was associated with a 57% lower dementia risk - a bigger reduction than the 9,800-step figure above, at roughly two-thirds the daily total.

That doesn't mean a slow walk is wasted - the lower-intensity numbers above are real, independent findings on their own. It means pace adds something on top of raw volume, which is worth knowing if you're choosing between a longer easy walk and a shorter brisk one on a day where time is the limiting factor.

What if you already have early signs of Alzheimer's?

The UK Biobank study looked at people developing dementia from a healthy baseline. A separate, newer study asked a different question: among people who are still cognitively healthy but already show early, preclinical Alzheimer's-related changes in the brain, does walking more slow down what happens next? Wai-Ying Wendy Yau and colleagues at Mass General Brigham, publishing in Nature Medicine in 2025, followed 296 participants aged 50 to 90 from the Harvard Aging Brain Study. Each wore a pedometer, underwent PET brain scans to measure amyloid-beta and tau - two proteins tied to Alzheimer's pathology - and completed yearly cognitive testing over an average of 9.3 years.

Among participants who already had elevated amyloid-beta - meaning early Alzheimer's-related changes were already present - those who walked 3,000 to 5,000 steps a day showed cognitive decline that arrived roughly 3 years later, on average, than those who walked less. At 5,000 to 7,500 steps a day, the delay stretched to roughly 7 years. The more active participants also showed slower accumulation of tau, the protein more closely tied to symptom progression; step count wasn't associated with a change in amyloid-beta itself. In plain terms: walking didn't appear to stop the underlying process, but it was associated with a meaningfully slower timeline once it had already started.

How many steps should you actually aim for?

Put both studies side by side and a practical range emerges rather than one magic number. Roughly 3,800 steps a day was already linked to a real reduction in dementia risk in the larger study; the benefit kept climbing through about 9,800 steps, where it leveled off. Separately, in people who already had early Alzheimer's-related brain changes, 5,000 to 7,500 steps a day was tied to the biggest delay in cognitive decline measured. That puts the sweet spot most people should actually aim for somewhere in the 5,000-to-9,800 range - genuinely walkable, and consistent with what the broader steps-per-day research has found for other outcomes, too.

If you're not there yet, both studies still found benefit below their headline numbers - there's no cliff where fewer steps stop counting. And this is a separate question from the one we cover in how many steps a day actually help your mood: that piece looks at depression and anxiety outcomes specifically, which land on a somewhat lower step range. The two aren't in conflict - they're just measuring different things, and it's fine for a single daily walk to be doing both jobs at once.

Making a step goal you can actually sustain

Both studies above are measuring sustained activity over years, not the effect of any single walk - which means the number that matters most isn't today's step count, it's the one you can keep hitting most days for a long time. Stepzy's daily step goal calculator suggests a realistic next target based on your current average rather than dropping you straight at 9,800, and adaptive goals move that target as your real baseline changes, instead of leaving you stuck chasing a number that stopped fitting months ago.

A single flat daily number also tends not to survive a busy week. Weekday and weekend split goals let a packed Wednesday and a slower Sunday carry different targets instead of treating both as the same shortfall, and Streak Skips mean a missed day or two doesn't erase months of consistency - the same reasoning we cover in why a streak with skips beats a perfect streak, applied here to a habit that's worth keeping for years, not weeks. If you're building this habit later in life, our page for seniors and how many steps you should walk after 60 cover how to set a target that fits a week that doesn't look the same every day.

This is general wellness information, not medical advice or a treatment for dementia, Alzheimer's disease, or any diagnosed cognitive condition. Both studies above are observational or based on a relatively small sample - the second involved 296 participants - so they show association and timing, not proof that walking alone changes any individual's outcome, and neither can rule out that healthier brains simply make it easier to walk more in the first place. If dementia risk or cognitive changes are a concern for you or someone you care for, talk to a doctor about what fits alongside your care. For most people without those considerations, the practical takeaway holds across both studies: walking most days, somewhere in the 5,000-to-9,800-step range and at a pace that feels a little brisk when you can manage it, is doing real, measured work.