A bigger question sits behind these numbers: could a slowing walk be an early sign of dementia? A large body of research says it can. The evidence is further down.
Why the number matters: in large studies, walking speed is associated with survival, with falls and with later dementia. That is what gives a reference value its weight. See how to measure it with the 10-Meter Walk Test.
What are the reference values?
The largest summary is a meta-analysis of 41 studies and 23,111 apparently healthy adults, walking at their normal speed over a course of 3 to 30 m. Bohannon and Williams Andrews, Physiotherapy, 2011.
| Age | Men | Women |
|---|---|---|
| 20 to 29 | 1.36 m/s | 1.34 m/s |
| 30 to 39 | 1.43 m/s | 1.34 m/s |
| 40 to 49 | 1.43 m/s | 1.39 m/s |
| 50 to 59 | 1.43 m/s | 1.31 m/s |
| 60 to 69 | 1.34 m/s | 1.24 m/s |
| 70 to 79 | 1.26 m/s | 1.13 m/s |
| 80 to 99 | 0.97 m/s | 0.94 m/s |
These are mean values for each group, rounded from the published figures in centimeters per second. The table is reproduced in the Rehabilitation Measures Database.
What about adults over 65?
A reference dataset built for this age group gives more detail. It pooled 954 healthy older adults, free of cognitive impairment and multiple conditions, walking at their usual speed on an electronic walkway. Beauchet et al., Frontiers in Human Neuroscience, 2017.
| Age | Mean walking speed |
|---|---|
| 65 to 74 | 1.25 m/s |
| 75 to 84 | 1.14 m/s |
| Over 85 | 0.89 m/s |
The two sources agree on the shape: stable through midlife, then a steady decline.
What changes the number?
- The protocol. The authors of the meta-analysis warn that their values may not serve as a standard when gait is measured over a short distance from the command "go", or when the walk includes a turn.
- The instruction. In 230 healthy adults, mean comfortable speed ranged from 1.27 to 1.46 m/s across age and sex groups, and mean maximum speed from 1.75 to 2.53 m/s. Bohannon, Age and Ageing, 1997.
- The person. The same study found that speed fell with greater age, and with lesser height and lower-limb strength.
- The setting. A walkway in a quiet clinic is not a pavement. Compare a patient with values collected the way you tested them.
Compare like with like. A norm is only useful when your test matches the test that produced it.
How much change matters?
Two published estimates give a sense of scale:
- About 0.05 m/s is a small meaningful change, and about 0.10 m/s a substantial one, in older adults and people after stroke. Perera et al., Journal of the American Geriatrics Society, 2006.
- Changes of 0.10 to 0.20 m/s may be clinically important across several patient groups, including stroke, hip fracture and multiple sclerosis. Bohannon and Glenney, Journal of Evaluation in Clinical Practice, 2014.
Use both as a guide. They come from specific populations and methods.
The open question: can a slowing walk warn of dementia?
A large body of research says it can, years ahead.
- Slower walkers develop dementia more often. A meta-analysis of prospective cohort studies, with 14,140 participants in the dementia analysis, found that the slowest walkers had a 66% higher risk of dementia than the fastest. Each 0.1 m/s of slower walking came with a 13% higher risk. Quan et al., Journals of Gerontology Series A, 2017.
- The slowing starts years before. In 3,663 people followed for nine years, gait was already slower up to seven years before the clinical onset of dementia, and a steeper decline came with a higher risk. Dumurgier et al., Journals of Gerontology Series A, 2017. In 204 healthy older adults examined every year for up to 20 years, the decline in gait speed accelerated about 12 years before the onset of mild cognitive impairment. Buracchio et al., Archives of Neurology, 2010.
- The change says more than one reading. In 16,855 older adults measured repeatedly, people whose walking speed and memory both declined had the highest risk of dementia, well above those in whom only one of the two declined. The authors concluded that the findings support including gait speed in dementia risk assessments. Collyer et al., JAMA Network Open, 2022. In 154 older adults measured twice a year for five years, slow gait at a single visit did not predict dementia. A decline in gait speed did, with almost seven times the risk. Montero-Odasso et al., Journal of the American Geriatrics Society, 2018.
- The link reaches Alzheimer's disease too. In brain imaging studies of older adults without dementia, more amyloid, the protein that builds up in Alzheimer's disease, went with a slower walk. del Campo et al., Neurology, 2016; Nadkarni et al., JAMA Neurology, 2017. Across dementias, a meta-analysis of 12 studies found that poor gait performance predicted dementia, most strongly the vascular and other non-Alzheimer's types. Beauchet et al., Journal of the American Medical Directors Association, 2016.
What research is working on now: which measure says the most, and how to use it for one patient. Speed is the best studied; step-to-step variability and walking with a second task are close behind. A slowing walk is a reason to look closer, not a diagnosis.
What it changes today: in the 16,855-person study, decline meant losing 0.05 m/s or more per year. A loss that small is visible only if someone measures the same walk, the same way, year after year.
How should you use a norm with a patient?
- Start with their own baseline. A patient compared with themselves, under the same conditions, tells you more than a patient compared with a table.
- Use the norm to set the scene. "Typical for your age" or "slower than most people your age" is a fair summary. It is not a diagnosis.
- Look behind the speed. A patient can be slow because of short steps, a slow rhythm, or time lost on one side. The number alone does not say which.
- Record the method. Distance, instruction, footwear and walking aid belong beside every result.
To measure it, see how to run the 10-Meter Walk Test.
Where does Baliston Health fit?
Gait speed with its explanation attached. Sensor insoles equipped with AI Mov-Scan capture the walk in suitable shoes. AI Mov-Scan captures. The platform analyzes and generates the Full Clinical Report. Balia, the platform's AI assistant, explains. The practitioner decides.
You see speed beside the measures that produce it: cadence, stride length, stance and swing time and left/right symmetry, depending on the analysis. You compare each session with the patient's own earlier record.
Read how walking changes with age, explore the technology, or book a demo.



