The research gives a surprisingly specific answer. Here's how much, how often, how fast — and what happens in your brain when you do it.
Last reviewed: May 2026
Key takeaways
+Walking is the most studied and most accessible form of exercise for brain health. The evidence base is unusually strong.
+A landmark randomised trial found that walking 40 minutes, 3 times a week, increased hippocampal volume by ~2% — reversing 1–2 years of age-related brain shrinkage.
+Large observational studies suggest 7,000–10,000 steps per day is associated with 38–51% lower dementia risk.
+Pace matters: brisk walking appears more protective than slow walking, even at the same step count.
The short answer
If you want the most direct answer the research currently supports: walk briskly for about 30–40 minutes, at least three times a week. That's the protocol from the strongest experimental evidence we have — and it's the one that actually showed changes in brain structure, not just self-reported cognitive improvement.
But there's an important nuance: the relationship between walking and dementia risk isn't a cliff edge where "enough" walking suddenly protects you and less doesn't. It's a gradient. Benefits start at surprisingly low thresholds and continue to accumulate as you do more — up to a point.
The study that changed the conversation
In 2011, Kirk Erickson and colleagues at the University of Pittsburgh published a study in PNAS that became the most cited piece of exercise–brain research in the dementia prevention field. It's the study behind the headline you've probably seen: "Walking can reverse brain shrinkage."
What they did: 120 sedentary older adults without dementia were randomly assigned to one of two groups. One group walked around a track at moderate intensity for 40 minutes a day, three days a week. The other group did stretching and toning exercises only. Both groups were followed for one year with brain MRI scans at baseline, six months, and twelve months.
~2%
Increase in hippocampal volume after one year of walking
Erickson et al., 2011 — PNAS (n=120, randomised controlled trial)
The walking group showed an average 2% increase in the volume of their hippocampus — the brain's memory centre and one of the first regions affected by Alzheimer's disease. The stretching group, by contrast, showed a ~1.4% decrease over the same period, consistent with the normal rate of hippocampal shrinkage in late adulthood.
Put differently: a year of moderate walking didn't just slow age-related brain shrinkage — it reversed it, effectively turning back the clock by one to two years. The walking group also showed improvements on spatial memory tests, and those improvements were correlated with increases in hippocampal size and in brain-derived neurotrophic factor (BDNF), a molecule involved in the growth and survival of neurons.
Why this study matters
This was a randomised controlled trial — the gold standard in research design. It didn't just observe that active people have bigger hippocampi (they do). It showed that previously sedentary adults who started walking experienced brain changes within a year. The intervention was walking — not running, not gym training, not a complex programme. Walking.
How many steps do you actually need?
The Erickson trial measured time (40 minutes, 3 times a week), not steps. But several large observational studies have since looked at the question in terms of daily steps — which is how most people now track their movement.
3,800
steps per day
25% lower dementia risk
This is the threshold where measurable benefit begins. Roughly 2 miles of walking.
del Pozo Cruz et al., 2022 — JAMA Internal Medicine
7,000
steps per day
38% lower dementia risk
The point of diminishing returns. Most of the benefit is captured here.
University of Sydney, 2025 — global review of 57 studies
9,800
steps per day
51% lower dementia risk
The optimal dose in the largest study. About 5 miles of walking.
del Pozo Cruz et al., 2022 — JAMA Internal Medicine (UK Biobank, n=78,430)
The practical takeaway on steps
If you're currently sedentary, getting to 3,800 steps a day is the first meaningful threshold. If you're already walking regularly, aim for 7,000 — that captures most of the observed benefit. Going beyond 10,000 steps may offer some additional protection, but the incremental gain is smaller. Every step counts, but the biggest jump in benefit comes from moving out of a sedentary baseline.
Does pace matter?
Yes. The UK Biobank study (del Pozo Cruz et al., 2022) found that walking intensity mattered independently of step count. People who walked at a brisk pace — roughly 100–130 steps per minute — had lower dementia risk than those walking the same number of steps at a slower pace.
A useful benchmark: if you can talk comfortably but couldn't sing, you're probably at the right intensity. This is sometimes called "moderate intensity" in the research — it maps to about 3–4 mph for most adults, or a pace where you notice your heart rate is up and your breathing is slightly faster.
That said, slow walking still counts. The UC San Diego study (2023) found that each additional 1,865 daily steps — at any pace — was associated with a 33% lower risk of cognitive impairment in women over 65. If brisk walking isn't realistic for you right now, more steps at any speed is still a meaningful investment.
What's actually happening in your brain
Walking doesn't protect the brain through one single mechanism. The research points to at least four pathways working together.
HV
Hippocampal growth
The Erickson trial directly measured this: aerobic walking increased hippocampal volume. The hippocampus is the brain's memory formation centre and one of the first regions to deteriorate in Alzheimer's disease. Its ability to grow in response to exercise — even in late adulthood — was one of the most important findings in the field.
BF
BDNF production
Walking increases blood levels of brain-derived neurotrophic factor (BDNF), a protein that supports the growth, survival, and repair of neurons. Higher BDNF levels are associated with better memory and slower cognitive decline. In the Erickson study, increases in BDNF correlated directly with hippocampal growth.
VH
Vascular health
Walking improves cardiovascular fitness, lowers blood pressure, and reduces arterial stiffness. This matters because the brain consumes roughly 20% of the body's blood supply. Hypertension, diabetes, and high cholesterol — all Lancet Commission risk factors — are improved by regular walking.
IF
Reduced inflammation
Chronic, low-grade inflammation is implicated in neurodegeneration. Regular moderate exercise reduces systemic inflammatory markers. A 2025 Mass General Brigham study found that walking was associated not only with slower cognitive decline but also with reduced accumulation of tau protein — one of the two hallmark pathological proteins in Alzheimer's disease.
A realistic starting plan
The research supports a clear target: work up to 30–40 minutes of brisk walking, at least three times a week. But if you're currently sedentary, jumping straight to that can feel daunting — and the research shows that smaller amounts still help. Here's a graduated approach.
Weeks 1–2
15 minutes, 3 times a week
Comfortable — faster than a stroll, but you could hold a conversation easily.
The goal is consistency, not intensity. Build the habit of going out the door.
Weeks 3–4
20 minutes, 3 times a week
Slightly brisker — breathing is faster, but you can still talk.
This is the dose used in the Erickson study's ramp-up phase. Benefits to focus and mood start here.
Weeks 5–8
30 minutes, 3 times a week
Brisk — you could talk but not sing. Heart rate noticeably elevated.
You're now in the range where hippocampal changes were observed in the Erickson trial.
Weeks 9+
40 minutes, 3–5 times a week
Brisk — sustained moderate intensity.
The full Erickson protocol. At this point, adding a 4th or 5th day may give more benefit than adding more minutes.
Make it stick
The biggest predictor of whether walking protects your brain isn't intensity or duration — it's whether you're still doing it six months from now. Walk at the same time each day, pick a route you enjoy, and don't increase pace or duration until the habit feels automatic. Consistency beats optimisation.
Common questions
Walking in the bigger picture
Physical inactivity is one of 14 modifiable risk factors for dementia identified by the 2024 Lancet Commission. It's also one of the most accessible to change — you don't need equipment, a subscription, or specialised knowledge. Just shoes and time.
But walking alone isn't the full picture. The FINGER trial — the first randomised trial to show that a multi-domain lifestyle intervention can improve cognitive function in at-risk adults — combined exercise with diet, cognitive training, and vascular risk monitoring. Each domain contributed independently, and the combination outperformed any single change.
Erickson KI, et al. (2011). Exercise training increases size of hippocampus and improves memory. Proceedings of the National Academy of Sciences, 108(7), 3017–3022. PubMed →
del Pozo Cruz B, et al. (2022). Association of daily step count and intensity with incident dementia in 78,430 adults living in the UK. JAMA Internal Medicine, 182(11), 1139–1148. PubMed →
Nguyen SA, et al. (2023). Accelerometer-measured physical activity and sitting with incident mild cognitive impairment or probable dementia among older women. Alzheimer's & Dementia, 19(10), 4457–4465. PubMed →
Erickson KI, et al. (2010). Physical activity predicts gray matter volume in late adulthood. Neurology, 75(16), 1415–1422. PubMed →
Livingston G, et al. (2024). Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet, 404(10452), 572–628. PubMed →
Ngandu T, et al. (2015). A 2-year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER). Lancet, 385(9984), 2255–2263. PubMed →
This content is for educational purposes and is not medical advice. Consult a healthcare professional before making changes to your health routine. Last reviewed: May 2026.
This site uses cookies for analytics to understand how visitors engage with our content.
Privacy policy
