The researchers asked a deceptively simple question: what happens to walking balance when the eyes are fed misleading information about self-motion, and does aging make the consequences worse? Everyday life is full of such moments. When we glance at a stationary object while the scenery slides past a train window, or when we track a moving face in a crowd, the visual system generates signals that can be mistaken for movement of our own body. These illusions, known as visually induced self-motion perceptions, can trigger real postural responses even though the body has not actually moved. Younger adults appear to handle these conflicts reasonably well, but the researchers suspected that older adults, whose vestibular and proprioceptive systems degrade with age and who are known to lean more heavily on vision for balance, might be far more susceptible.

To test this, the team recruited fourteen healthy older adults with an average age of seventy-five and compared them with sixteen younger adults averaging twenty-three years old, using data from the younger group that the same laboratory had collected in a previous experiment. All participants walked on an instrumented treadmill at a fixed speed of 3.6 kilometers per hour while a large screen placed 1.60 meters in front of them displayed a background of black-and-white vertical stripes and a red target dot at eye level. Optical motion capture, sampling at fifty frames per second, tracked marker clusters on the heels, pelvis, trunk, and head, allowing the researchers to reconstruct the mediolateral trajectory of the center of mass, approximated from the pelvis markers, with millimeter precision.

The experiment imposed three distinct kinds of visual trickery. In the moving background condition, participants fixated the stationary red dot while the striped background swept forty-five degrees to the right over four seconds, held there for eight seconds, and returned. In the two target-tracking conditions, the background stayed still while the red dot itself moved to the right; in one version participants tracked it with combined head and eye movements, and in the other with eye movements alone while keeping the head stationary. Each five-minute trial contained roughly fourteen perturbations, triggered at right heel strikes and separated by six to eight randomly spaced steps. A normal walking trial without any perturbations served as the baseline against which all effects were measured.

The first major finding concerned overall gait variability. Across entire trials, both age groups showed significantly increased variability in the mediolateral position of the center of mass during all three perturbation conditions compared with normal walking, with large effect sizes. The older adults also showed greater overall center of mass variability than the younger adults regardless of condition, consistent with a general age-related decline in frontal plane balance control. Interestingly, however, the perturbations did not disproportionately inflate variability in the older group relative to the younger one, a result the authors attributed to the discrete, intermittent nature of the perturbations, which left unperturbed stretches of walking that diluted the age difference over the whole trial.

The real drama unfolded within the perturbation epochs themselves. When the background moved, younger adults consistently shifted their center of mass to the left, opposite to the rightward background motion, exactly as the self-motion illusion would predict: the brain interprets the moving background as evidence that the body is drifting left, and the nervous system responds by stepping to extend the base of support in that perceived direction. The older adults initially responded the same way, but something remarkable happened over repeated exposures: their deviation direction gradually reversed, drifting toward the direction of the background movement instead of away from it. This reversal, confirmed by a significant interaction between age and repetition in a linear mixed-effects model, may reflect either a loss of attentional fixation on the target, causing older participants to treat the moving background as the object of interest, or a strategic shift toward faster corrective moments generated by the stance leg rather than slower foot-placement adjustments.

The target-tracking conditions exposed an even sharper age divide. When participants tracked the moving target with head rotation, both groups deviated toward the target, but the older adults deviated earlier, starting at the very beginning of the target’s movement rather than after it had stopped, and their peak deviations were a staggering 8.5 times those of the younger adults during the moving phase and 3.8 times during the stationary phase. In the eye-movement-only condition, older adults again deviated earlier, though the peak magnitudes between groups were statistically similar. The researchers explain this pattern through the mechanics of sensory cancellation. When we actively rotate our head, an efference copy of the neck motor command allows the vestibular nuclei to suppress vestibular signals that are self-generated, a mechanism elegantly demonstrated in rhesus monkeys. If aging degrades the neck proprioceptive information needed to validate that prediction, the suppression fails, and the combined visual, vestibular, and proprioceptive conflict becomes far harder to resolve, which would explain why head rotation made the age effect so much larger.

The adaptation results carried perhaps the most hopeful message. In the eye-movement-only condition, the older adults’ center of mass deviations shrank significantly across the thirteen analyzed perturbation repetitions, while the younger adults, whose initial responses were small to begin with, showed no systematic change. This suggests that the aging nervous system retains a functional capacity for sensory reweighting, the process by which the brain dynamically adjusts the reliability it assigns to each sensory channel, gradually discounting visual information that proves unreliable. Crucially, however, this downweighting did not emerge in the head-rotation condition for either group, presumably because head movement simultaneously corrupted vestibular and somatosensory inputs, leaving no clean alternative channel for the brain to trust instead.

The practical stakes of these findings became vivid during testing itself. Three older participants deviated so far from the treadmill’s center during the tracking conditions, frequently glancing down at their feet for additional visual anchoring, that the measurements had to be terminated for safety. The authors also acknowledge important limitations: the treadmill belts provide proprioceptive and cutaneous cues that may constrain responses, the harness worn by older participants for safety may have supplied subtle sensory feedback, and the healthy, active volunteers who completed the protocol likely represent a fitter segment of the elderly population, meaning the true effects of aging in the general public may be even larger than measured. The study also cannot fully disentangle effects on balance control from effects on heading direction, since gaze and head orientation influence both.

Nevertheless, the conclusions carry real weight for fall prevention. Falls remain a leading cause of injury and loss of independence in older adults, and this work pinpoints a specific, everyday mechanism of vulnerability: when the visual scene moves or when the eyes chase a moving target, the aging brain struggles to correct the resulting self-motion illusion using vestibular and proprioceptive evidence, and the body’s walking path drifts accordingly. The encouraging corollary is that repeated exposure to such sensory conflicts appears to teach the older nervous system to discount misleading vision, at least when the head stays still. That plasticity opens a potential avenue for balance training programs that deliberately expose older adults to controlled visual perturbations, teaching the brain when to stop believing its eyes.

Subject of Research: Age-related differences in balance control during walking under repeated visual perturbations

Article Title: Balance control is more affected in older than younger adults by repeated visual perturbations during walking

Article References: Li, Y., Lambrecht, E., Bruijn, S. M., & van Dieën, J. H. (2026). Balance control is more affected in older than younger adults by repeated visual perturbations during walking. GeroScience. https://doi.org/10.1007/s11357-026-02573-5

Image Credits: AI Generated

DOI: 10.1007/s11357-026-02573-5

Keywords: balance control, aging, visual perturbation, gait, center of mass, sensory reweighting, vestibular system, proprioception, self-motion perception, treadmill walking, fall prevention, GeroScience