TY - JOUR T1 - Rethinking Margin of Stability: Incorporating Step-To-Step Regulation to Resolve the Paradox JF - bioRxiv DO - 10.1101/2021.12.01.470263 SP - 2021.12.01.470263 AU - Meghan Kazanski AU - Joseph P. Cusumano AU - Jonathan B. Dingwell Y1 - 2021/01/01 UR - http://biorxiv.org/content/early/2021/12/03/2021.12.01.470263.abstract N2 - Maintaining frontal-plane stability is a major objective of human walking. Derived from inverted pendulum dynamics, the mediolateral Margin of Stability (MoSML) is frequently used to measure people’s frontal-plane stability on average. However, typical MoSML-based analyses deliver paradoxical interpretations of stability status. To address mediolateral stability using MoSML, we must first resolve this paradox. Here, we developed a novel framework that unifies the well-established inverted pendulum model with Goal-Equivalent Manifold (GEM)-based analyses to assess how humans regulate step-to-step balance dynamics to maintain mediolateral stability. We quantified the extent to which people corrected fluctuations in mediolateral center-of-mass state relative to a MoSML-defined candidate stability GEM in the inverted pendulum phase plane. Participants’ variability and step-to-step correction of tangent and perpendicular deviations from the candidate stability GEM demonstrate that regulation of balance dynamics involves more than simply trying to execute a constant-MoSML balance control strategy. Participants adapted these step-to-step corrections to mediolateral sensory and mechanical perturbations. How participants regulated mediolateral foot placement strongly predicted how they regulated center-of-mass state fluctuations, suggesting that regulation of center-of-mass state occurs as a biomechanical consequence of foot placement regulation. We introduce the Probability of Instability (PoI), a convenient statistic that accounts for step-to-step variance to properly predict instability likelihood on any given future step. Participants increased lateral PoI when destabilized, as expected. These lateral PoI indicated an increased risk of lateral instability, despite larger (i.e., more stable) average MoSML. PoI thereby explicitly predicts instability risk to decisively resolve the existing paradox that arises from conventional MoSML implementations.Competing Interest StatementThe authors have declared no competing interest. ER -