Every step we take carries an invisible wager. Most of the time the ground honors its side of the bargain, but a raised pavement edge, an unseen curb, or a sudden shift beneath the feet can turn an ordinary walk into a fight against gravity. For older adults living with mild cognitive impairment, that fight is disproportionately dangerous: this population falls roughly twice as often as their cognitively intact peers, and most of those falls trace back to unexpected perturbations such as slips and trips. A new study published in GeroScience by Jessica Pitts and Tanvi Bhatt of the University of Illinois at Chicago has now pulled back the curtain on what happens, muscle by muscle and millisecond by millisecond, when people with mild cognitive impairment are suddenly tripped while walking. The findings reveal a subtle but consequential story about how the aging, cognitively burdened brain coordinates its emergency response to a forward loss of balance.
The research team recruited fifty adults between the ages of 55 and 90, dividing them into two groups of twenty-five: older adults with mild cognitive impairment, identified through the Montreal Cognitive Assessment and confirmed with a multi-domain neuropsychological battery, and cognitively intact older adults. Anyone with neurological or musculoskeletal diagnoses, recent hospitalization, or an inability to walk ten meters without an assistive device was excluded. Each participant then walked on a motorized ActiveStep treadmill at a self-selected speed of either 0.8 or 1.0 meters per second while wearing a safety harness connected to a load cell. After several unperturbed steps, the treadmill delivered its ambush: the belt abruptly decelerated to a standstill and then accelerated violently forward, yanking the support surface out from under the swinging foot and pitching the body toward a forward fall.
What happened next was remarkably consistent across nearly everyone. Rather than immediately throwing the swinging leg far forward in the classic elevating strategy seen with obstacle trips, almost all participants first slammed the swing limb down almost in line with the stance foot, a rapid lowering step that restored a base of support as quickly as possible. Only then did they take the long forward recovery step that actually arrested the fall, and here the group split into two camps. Some stepped forward with the contralateral limb, others with the ipsilateral limb, and a handful of ipsilateral steppers showed an aborted contralateral step or a weight shift without liftoff before the recovery step began, hinting at a momentary failure of step initiation. Just three participants, one with impairment and two without, immediately executed a long forward step with the swing limb, resembling an elevating strategy, and they were excluded from the primary analysis because their response was biomechanically distinct and too rare to support statistical conclusions.
The headline result was not who fell, but how close everyone came. Falls were defined by a harness load exceeding thirty percent of body weight, verified on video, and the fall rates were strikingly high in both groups: ninety-two percent among those with mild cognitive impairment and seventy-eight percent among the cognitively intact, a difference that did not reach statistical significance. But the margin of stability, a normalized measure of how far the extrapolated center of mass sits beyond the front boundary of the base of support at the moment the recovery foot lands, told a sharper story. Participants with mild cognitive impairment showed significantly higher margins of stability at recovery touchdown, meaning their center of mass had traveled even further past the safe zone than their cognitively intact counterparts. This group difference persisted even after the researchers statistically controlled for baseline mobility, walking speed, and treadmill velocity, and it held regardless of which limb each person used for the recovery step.
The likely culprit was the recovery step itself. Regression modeling showed that the margin of stability at touchdown was predicted primarily by reactive step length, and participants with mild cognitive impairment took recovery steps roughly fifteen percent shorter than the cognitively intact group. A shorter step establishes a smaller base of support and a shorter lever arm against the forward momentum of the trunk, leaving the body’s center of mass stranded further beyond the toes. Notably, the pre-trip picture looked identical between groups: unperturbed walking showed no differences in margin of stability, step length, or trunk angle, suggesting that the deficit was not a feedforward consequence of a cautious or shuffling gait but a genuine failure of the feedback-driven emergency response. The one proactive exception was toe clearance, which was significantly lower in the impaired group, about 5.5 centimeters versus 7.1 centimeters, a detail that matters because low toe clearance is a well-known risk factor for actually catching a foot on an obstacle in the first place.
Beneath these kinematic differences lay something even more intriguing: a rewiring of the muscle coordination itself. Using wireless electromyography from eight lower limb muscles, including the biceps femoris, vastus lateralis, medial gastrocnemius, and tibialis anterior on both legs, the team extracted muscle synergies, the stereotyped co-activation patterns that the nervous system deploys as building blocks for movement. Seven distinct synergies emerged across the cohort. In an exploratory analysis, at least three of these showed low structural similarity between the two groups within each stepping strategy, falling below the similarity threshold the researchers had defined. Cognitively intact older adults recruited highly similar synergy structures regardless of which limb they stepped with, while those with mild cognitive impairment showed more variability between strategies and even recruited a unique synergy not seen in the control group.
The texture of those synergy differences was telling. Cognitively intact participants showed greater contribution from the stepping limb quadriceps in one key synergy, consistent with generating the knee extension needed for a longer, more effective recovery step. Meanwhile, participants with mild cognitive impairment tended to recruit synergies dominated by a single muscle, such as the stance limb tibialis anterior or medial gastrocnemius, whereas the intact group showed richer co-activation of multiple stance limb muscles working together to generate support. The authors interpret this as a shift toward more fractionated, simplistic motor modules rather than the integrated, multi-muscle coordination that efficient balance recovery demands. Such fragmentation could stem from impaired sensorimotor integration, altered executive function, or the structural and functional brain changes, including reduced grey matter volume and disrupted functional connectivity, that are well documented in mild cognitive impairment.
The mechanistic picture the authors sketch runs from the brainstem to the cortex. Reactive balance responses are believed to be initiated by rapid subcortical brainstem loops driven primarily by fast somatosensory signals, and people with mild cognitive impairment often show reduced brainstem grey matter volume and delayed onset latencies of stepping limb muscles, which could slow perturbation detection and response initiation. Later phases of the response, such as scaling step length to the demands of the moment, likely recruit cortical networks, and altered cerebellar functional connectivity has been linked to reactive stability in this population. Muscle synergies themselves may be stored as pre-existing templates in the brainstem or spinal cord, with the cortex priming the appropriate templates through central set, a process that depends on cerebellar-cortical and basal ganglia-cortical loops vulnerable to the neural changes of early cognitive decline. In this framing, the trip response fails not because the legs are weak but because the neural machinery that selects and executes the right motor program at the right moment is degraded.
The clinical implications cut in two directions. On one hand, the results align with earlier work showing that people with mild cognitive impairment also respond poorly to slip-like backward perturbations, suggesting a direction-general deficit in reactive stepping that could undermine recovery from many kinds of everyday destabilization, especially when attention is divided. On the other hand, the study raises a caution about the treadmill paradigms increasingly marketed for balance assessment and training. Surface translations produce continual posterior displacement of the support surface, which appears to favor lowering strategies more often than obstacle-induced trips do, and prior research has found that repeated treadmill slip training did not transfer to improved recovery from a novel overground obstacle trip. Commercialized perturbation systems remain attractive for their scalability, but the authors argue that truly effective fall prevention for this population may need to incorporate more ecologically valid elements such as obstacle avoidance and navigation.
The study has limits the authors acknowledge candidly: a single high-intensity perturbation that pushed fall rates above seventy percent in both groups may have created ceiling effects that masked group differences in the binary fall outcome, the regression and synergy analyses were exploratory with modest subgroup samples, and electromyography captured only eight lower limb muscles when hip and trunk musculature also matter for trip recovery. Still, the core finding stands with unusual clarity. Older adults with mild cognitive impairment do not simply trip more often; when the ground moves beneath them, their recovery steps land shorter, their center of mass drifts further past the point of no return, and their muscles fire in fragmented, less reproducible patterns. It is a vivid demonstration that cognitive decline reaches deep into the body’s most ancient reflexes, and that protecting the aging brain may be inseparable from protecting the aging step.
Subject of Research: Reactive balance and trip recovery strategies in older adults with mild cognitive impairment
Article Title: Recovery strategies during trip-like support surface perturbations while walking in older adults with mild cognitive impairment
Article References: Pitts, J., & Bhatt, T. (2026). Recovery strategies during trip-like support surface perturbations while walking in older adults with mild cognitive impairment. GeroScience. https://doi.org/10.1007/s11357-026-02490-7
Image Credits: AI Generated
DOI: 10.1007/s11357-026-02490-7
Keywords: mild cognitive impairment, reactive balance, trip perturbation, fall risk, muscle synergies, treadmill perturbation, margin of stability, gait, electromyography, aging, sensorimotor integration, GeroScience
News Source: Cassandra Pierce. (October 7, 2026). When the Ground Betrays: How Mild Cognitive Impairment Rewires the Brain’s Emergency Reflexes for Tripping. Scienmag.
