Underlying mechanisms of fall risk on stairs with inconsistent going size

Serious falls occur frequently on stairs with inconsistent dimensions. Inconsistent smaller goings are thought to reduce user's foot clearances and foot contact lengths since individuals do not detect and alter their behaviour prior to the inconsistency, increasing the risk of a trip, heel-catch or over-step and potential slip on the stairs. So far, these mechanisms for a stair fall remain theoretical only. The aim of this paper was to identify the underlying mechanisms by which steps with inconsistent going size increase the risk of falls. For this study twenty-seven younger adults (24 ± 3 y, 1.74 ± 0.09 m, 71.41 ± 11.04 kg) and thirty-three older adults (70 ± 4 y, 1.68 ± 0.08 m, 67.90 ± 14.10 kg) ascended and descended a seven-step instrumented staircase in two conditions: 1) consistent dimensions with 200 mm risers and 250 mm goings and 2) inconsistent going dimensions where the going of the third step was reduced by 10 mm, and consequently the going of the second step was larger by 10 mm. Five repeated trials on the inconsistent stairs were performed to assess if there was an adaptation effect after first exposure. In descent in the first inconsistent trial, foot contact lengths were not significantly different between conditions for the younger and older adults on the inconsistently shorter step (∼1%, p = .121). Foot trajectories were pulled further back in the last 22% of swing before contact (p = .025), contradicting previous expectations. Younger adults then had reduced clearances over the next step (∼5 mm, p = .027), which was inconsistently longer, increasing the risk of a heel-catch, whereas foot clearances for older adults were not different. With repeated inconsistent trials the foot contact length of older adults reduced on the shorter step (p = .024). In ascent, in the first inconsistent trial, interaction effects were detected between groups and conditions on three steps: the inconsistently longer step (p = .003), the shorter step (p = .004), the next step (p = .006), as well as on the walkway (p = .048). Older adults positioned themselves further away from the stairs on the walkway compared to younger adults and then had a reduced foot contact length on the inconsistently shorter step (∼2.8%, p = .026), increasing the chances of under-stepping and slipping off the shorter step. Whereas younger adults were positioned closer to the stairs on the walkway, had increased foot contact lengths on the inconsistently longer step and contact lengths that were not different on the inconsistently shorter step. With repeated inconsistent trials, foot contact lengths were reduced on the longer step (p = .006) and then on the shorter step (p = .018). These findings contradict previous assumptions that individuals do not adapt to inconsistent goings on stairs. In descent on the first trial, both groups adjusted their stepping behaviour late in the swing prior to contact with the first inconsistent step. In ascent younger adults made changes to their position and stepping behaviour before stepping on the stairs. These behaviours to mitigate the risk of the inconsistent step, did not persist in the repeat trials. Future investigations should establish the magnitude at which inconsistencies are detectable and can be acted upon and should include a wider range of individuals. This type of research could help inform future initiatives to prevent serious stair falls.