Effect of stride count on the validity and reliability of autocorrelation-based gait variability variables in older adults requiring long-term care.

Thirty people from a single facility, all of them certified as needing long-term care and mostly in their eighties. The authors state that many participants had more than one condition and that the sample was too small for disease-specific subgroup analyses. The authors say the results should not be generalised to community-dwelling older adults without a care certificate, or to people in hospital. Ten strides was the ceiling because that was the most every participant could produce, so the study says nothing about longer analysis windows. The study only measured how the numbers behave. It did not test whether acting on these stride counts changes anything for a patient. No treatment was given and no clinical outcome was followed, so this paper cannot show that gait variability or any intervention affects falls, walking independence or anything else.

The authors report no conflicts of interest and no funding relevant to the article.

Do not read this as licence to shorten walking tests. The six-stride benchmark carried over from step-timing and stride-length variability was not enough here: at 6 strides vertical step-to-step variability sat at an ICC of 0.797 and vertical stride-to-stride at 0.657, and those measures only reached 0.843 at 7 strides and 0.840 at 9 strides. Nor do these numbers show that the measure predicts anything for your patients; the study tested how a measurement behaves in 30 care-home residents, not whether patients do better.