Abstract
BACKGROUND: Older adults have difficulty maintaining side-to-side balance while navigating daily environments. Losing balance in such circumstances can lead to falls. We need to better understand how older adults adapt lateral balance to navigate environment-imposed task constraints. RESEARCH QUESTION: How do older adults adjust mediolateral balance while walking along continually-narrowing paths, and what are the stability implications of these adjustments? METHODS: Eighteen older (71.6±6.0 years) and twenty younger (21.7±2.6 years) healthy adults traversed 25 m-long paths that gradually narrowed from 45 cm to 5 cm. Participants switched onto an adjacent path when they chose. We quantified participants' lateral center-of-mass dynamics and lateral Margins of Stability (MoS(L)) as paths narrowed. We quantified lateral Probability of Instability (PoI(L)) as the probability that participants would take a laterally unstable (MoS(L)<0) step as they walked. We also extracted these outcomes where participants switched paths. RESULTS: As paths narrowed, all participants exhibited progressively smaller average MoS(L) and increasingly larger PoI(L). However, their MoS(L) variability was largest at both the narrowest and widest path sections. Older adults exhibited consistently both larger average and more variable MoS(L) across path widths. Taken into account together, these resulted in either comparable or somewhat larger PoI(L) as paths narrowed. Older adults left the narrowing paths sooner, on average, than younger. As they did so, older adults exhibited significantly larger average and more variable MoS(L), but somewhat smaller PoI(L) than younger. SIGNIFICANCE: Our results directly challenge the predominant interpretation that larger average MoS(L) indicate "greater stability", which we argue is inconsistent with the principles underlying its derivation. In contrast, analyzing step-to-step gait dynamics, together with estimating PoI(L) allows one to properly quantify instability risk. Furthermore, the adaptive strategies uncovered using these methods suggest potential targets for future interventions to reduce falls in older adults.