Abstract

Age-normative hearing loss is linked with reduced cognitive functioning, mobility decline, and increased falls risk in older adults. One explanation, as yet untested experimentally, is that with age, there are increasing demands for limited cognitive capacity required to successfully perform walking and listening. To address this gap, the current study investigated age-related differences in the allocation of cognitive resources between concurrent walking and listening tasks. Seventeen younger and twelve older adults with normal hearing participated. Three sentences (1 target, 2 maskers) were played simultaneously from apparent auditory scene locations (left, center, right) in a virtual reality street crossing scene. Target location probability (100% versus ≤75%) was varied. Participants reported key elements from the target sentences. Gait during self-paced treadmill walking was assessed with a motion capture system using active markers positioned on the head, sternum, waist, and feet. Participants completed five conditions: walking only, listening only 100% probability, listening only ≤75% probability, walking while listening 100% probability, and walking while listening ≤75%. Key dependent measures were listening accuracy, head and trunk position, and gait parameters (e.g., step width, stride length, stride time, stride time variability, velocity, and cadence). Word recognition accuracy was significantly worse in (a) older than younger adults, (b) dual- than single-task, and (c) less predictable location probability. Kinematic analyses revealed reduced average head pitch and stride time variability, but increased variability in cadence from single-task walking to dual-task conditions in older adults. Younger adults did not exhibit a consistent gait pattern. Overall, older adults displayed more dual-task costs in listening performance, but younger adults showed more performance costs in motor function. A closer examination of performance costs revealed that older adults who demonstrated improved posture through reduced head pitch rotation were also better at listening under dual-task conditions. Altogether, listening in a multitalker situation in old age was particularly hampered when concurrently walking and when the signal location was less predictable. Furthermore, under increased cognitive load, older adults displayed reduced gait variability in stride time, and straightened their head alignment. These findings suggest that older adults redistribute cognitive resources toward gait over listening performance, consistent with theories of postural prioritization and cognitive compensation.