The effect of visual sensory interference during multitask obstacle crossing in younger and older adults

Abstract

When older adults step over obstacles during multitasking, their performance is impaired; the impairment results from central and/or sensory interference. The purpose was to determine if sensory interference alters performance under low levels of cognitive, temporal, and gait demand, and if the change in performance is different for younger versus older adults. Participants included 17 younger adults (20.9±1.9 years) and 14 older adults (69.7±5.4 years). The concurrent task was a single, simple reaction time (RT) task: depress button in response to light cue. The gait task was stepping over an obstacle (8 m walkway) in three conditions: (1) no sensory interference (no RT task), (2) low sensory interference (light cue on obstacle, allowed concurrent foveation of cue and obstacle), or (3) high sensory interference (light cue away from obstacle, prevented concurrent foveation of cue and obstacle). When standing, the light cue location was not relevant (no sensory interference). An interaction (sensory interference by task, p<0.01) indicated that RT was longer for high sensory interference during walking, but RT was not altered for standing, confirming that sensory interference increased RT during obstacle approach. An interaction (sensory interference by age, p<0.01) was observed for foot placement before the obstacle: With high sensory interference, younger adults placed the trail foot closer to the obstacle while older adults placed it farther back from the obstacle. The change increases the likelihood of tripping with the trail foot for younger adults, but with the lead limb for older adults. Recovery from a lead limb trip is more difficult due to shorter time for corrective actions. Overall, visual sensory interference impaired both RT and gait behavior with low levels of multitask demand. Changes in foot placement increased trip risk for both ages, but for different limbs, reducing the likelihood of balance recovery in older adults.

Fig 1

Top panel: Framework indicating demands examined with multitask locomotor research. Relatively lower demands are on the left, with increasing demands to the right in red. Demands of the current study are demonstrated with blue vertical arrows (note that the arrow position demonstrates relative differences across conditions and studies, and it not intended to reflect absolute differences or locations on the spectrum). Two examples of previous sensory interference research are also indicated: (1) * foot target task [13] and (2) + obstacle task [19]. Bottom panel: Information about each study. All studies compared low versus high visual sensory interference. The current study focuses on the less difficult side of the framework, to determine if visual sensory interference alters performance under low levels of cognitive, temporal, and gait demand.

Fig 2

(A) Line drawing of the walkway, obstacle, and RT cue. In the walking conditions, the RT cue was triggered 0–1.5 s (corresponds to gray region) after the participant broke the laser beam (positioned at hip height). The RT cue appeared once during approach, either on the obstacle (low sensory interference (SI)) or at eye-level (high SI). Participants pressed a hand-held remote when the RT light cue turned on. HD-1 and HD+1 represent the horizontal distance from the obstacle to the trail toe and the lead heel, respectively. (B) The block-randomized protocol. In the standing conditions, participants stood 1.5 m in front of the obstacle, and the RT cue appeared on either the obstacle or at eye-level. For standing trials, there was no SI regardless of light cue location. SI = sensory interference.

Fig 3

(A) Mean reaction time as a function of task and sensory interference. An interaction of sensory interference by task was observed (p<0.001); post hoc differences noted with *. (B) Mean reaction time as a function of age, task, and block. An interaction of age by task by block was observed (p = 0.004); post hoc differences noted with §. Error bars represent standard error. SI = sensory interference.

Fig 4

(A) Mean trail foot placement before the obstacle (HD-1) as a function of sensory interference in younger and older adults. An interaction of sensory interference by age was observed; post hoc differences noted with #. (B) Mean lead foot placement after the obstacle (HD+1) for young and older adults. An age main effect was observed (p = 0.01) and noted with †. Error bars represent standard error. SI = sensory interference.

Fig 5. Mean gait speed as a function of step, sensory interference, and age group.

The footprint figures indicate the step where speed was calculated for each of the three steps (one-step before the obstacle crossing, the lead step across the obstacle, and the trail step after the obstacle). For each of the three steps, two main effects were observed: (1) age (p<0.001); post hoc differences noted with +, and (2) sensory interference (p<0.001); post hoc differences noted with *. Error bars represent standard error. SI = sensory interference.