Changes to margins of stability from walking to obstacle crossing in older adults while walking fast and with a dual-task.

It is not well understood how older adults meet the combined locomotor demands of obstacle avoidance at fast speeds as compared to obstacle avoidance under cognitive loads. The purpose of this study was to quantify changes in locomotor stability (margin of stability, MOS) from walking to crossing obstacles at fast speeds versus with added cognitive demands in older adults. Community-dwelling older adults walked on an unobstructed and obstructed path at their preferred speed (preferred); during a dualtask (verbal fluency); and at their 'fastest comfortable' speed (fast). We used motion capture to calculate MOS in the anteroposterior direction, and compared minimum MOS between crossing foot and support phase (lead single support, lead double support, trail single support, trail double support) and tested for within subject changes using a linear mixed effect regression model [Condition (preferred, fluency, fast) x Walkway (unobstructed, obstructed) x Phase (single support, double support) x Foot (lead, trail)]. We examined crossing kinematics (approach distance, toe clearance, and recovery distance) between conditions. A significant omnibus effect partially supported our predictions. A Condition x Walkway x Phase interaction supported that older adults increased stability under a cognitive load and prioritized stability, demonstrated by not changing MOS from walking to obstacle crossing. During fast obstacle crossing they decreased stability during double support and exhibit more stability in single support, when vulnerable to external perturbations (contacting the obstacle). During a dual-task, older adults took shorter and higher steps over the obstacle to ensure they cleared it safely, but at fast speeds they increased the length of their crossing step without higher toe clearance. The results suggest older adults attempt to preserve stability when crossing obstacles under both cognitive and speed demands, but may be unable to ensure a safer limb elevation to avoid obstacles at fast speeds as they do under cognitive demands.