Which gait parameters are linked to falls in older adults?

How common are falls?

Around one in four adults aged 65 and over reports falling each year — more than 14 million people in the United States, according to current CDC figures. The age-adjusted fall death rate among older adults rose 21% between 2018 and 2024, from 64.7 to 78.4 per 100,000. The National Safety Council’s Injury Facts records 43,020 fall deaths among adults aged 65 and over in 2024, and more than 3.85 million fall injuries treated in emergency departments in 2023.

The CDC also notes that falling once roughly doubles the chance of falling again.

You will still see an older pairing quoted widely — “28%, about 36 million falls a year.” It traces back to a 2018 state-level dataset and has been superseded. The figures above are the current ones.

What the gait studies actually found

A 2025 study of 303 older adults (Kim et al., Scientific Reports, DOI 10.1038/s41598-025-02128-2) compared walking in a group identified as at higher fall risk against a group that was not. The fall-risk group showed increased stance time and increased double-stance time, together with reduced stride length, walking velocity and cadence.

Two of those — stride length and velocity — are worth flagging immediately, because Stride One does not measure either of them. They are reported here because they are part of the published finding, not because they are something this device can show you.

A separate study of 446 participants (Park & Bae, PMC12471427) asked a sharper question: which single parameter best separates the two groups? The answer was the variability of step width, with an area under the curve of 0.715, marginally ahead of the Timed Up and Go test at 0.708. Stride One does not derive step width either, and does not claim to.

The limitation nobody mentions

Both studies were conducted on a treadmill with a pressure platform. So are most of the studies in this literature, for the obvious reason that this is where the measurement equipment is. That creates a gap between what has been established and where falls happen: a laboratory corridor at a chosen pace, on a known surface, being watched, is not a kitchen at 6am.

This cuts both ways. It is a genuine reason to be cautious about applying laboratory findings directly to everyday walking. It is also the strongest argument for measuring continuously in ordinary life, which is what wearable sensors make possible — and which nobody has a large fall-risk literature for yet, because the measurements have only recently become practical.

Which of these parameters can be measured in a shoe?

Of the parameters that appear in these studies, three are measurable with in-shoe sensors and are recorded by Stride One: stance time, double-support time and cadence. Step-time variability — how consistent the rhythm is from step to step — is recorded as well.

The parameter that discriminated best in the published work, step-width variability, is not one of them. Stating that plainly is more useful than implying otherwise: the honest position is that in-shoe measurement covers part of what the literature points at, continuously and in real life, rather than all of it, briefly, in a laboratory.

What this means, and what it does not

It means that certain measurable features of walking are, at a population level, associated in published research with fall risk in older adults. It does not mean that Stride One, or any insole, predicts falls, assesses fall risk, screens for it, or can tell you whether you personally are likely to fall. Those are clinical judgements, made by clinicians, using far more than gait parameters. If falls are a concern for you or someone you care for, that is a conversation to have with a doctor.

What Stride One does and does not do

Stride One is a monitoring device. It records how load is shared between the feet, how the foot rolls through each step, and how pressure is distributed across the zones of each foot. It gives you, and anyone you choose to share it with, a record of how those things change.

It does not diagnose anything, detect any condition, predict any outcome, or tell you what to do about what it shows. It does not measure skin temperature, gait speed, step length, stride length or step width. It does not replace an appointment, an examination or your clinician’s judgement.

Sources

• Kim et al. Gait characteristics in older adults at risk of falls, n = 303. Scientific Reports, 2025. DOI 10.1038/s41598-025-02128-2.
• Park & Bae. Discriminative gait parameters for fall risk, n = 446. PMC12471427.
• Centers for Disease Control and Prevention, older adult fall data and fall death rates, 2018–2024.
• National Safety Council, Injury Facts — fall deaths and emergency-department-treated fall injuries.

Read next