Stable gait requires active control of the mediolateral (ML) kinematics of the body center of mass (CoM) and the base of support (BoS) in relation to each other. Stance leg hip abductor (HA) muscle spindle afference may be used to guide contralateral swing foot placement and adequately position the BoS in relation to the CoM. We studied the role of HA spindle afference in control of ML gait stability in young and older adults by means of muscle vibration. Healthy young (n = 12) and older (age > 65 years, n = 18) adults walked on a treadmill at their preferred speed. In unperturbed trials, individual linear models using each subject’s body CoM position and velocity at mid-swing as inputs accurately predicted foot placement at the end of the swing phase in the young [mean R2 = 0.73 (SD 0.11)], but less so in the older adults [mean R2 = 0.60 (SD 0.14)]. In vibration trials, HA afference was perturbed either left or right by vibration (90 Hz) in a random selection of 40% of the stance phases. After vibrated stance phases, but not after unvibrated stance phases in the same trials, the foot was placed significantly more inward than predicted by individual models for unperturbed gait. The effect of vibration was stronger in young adults, suggesting that older adults rely less on HA spindle afference. These results show that HA spindle afference in the stance phase of gait contributes to the control of subsequent ML foot placement in relation to the kinematics of the CoM, to stabilize gait in the ML direction and that this pocess is impaired in older adults.

The aim of this study was to examine the effect of narrowing step width on mediolateral (ML) center of mass (COM) kinematics and margin of stability (MOS) in young and older adults. Fourteen young and 18 healthy older adults were asked to walk on a treadmill at preferred speed, stepping on projected lines at their predetermined preferred step width (PSW) and at a 50% narrowed step width (NSW). Linear trunk accelerations were recorded by an inertial sensor, attached at the level of the lumbar spine and foot placement was determined from force sensors in the treadmill. Mediolateral peak-to-peak COM displacement, COM velocity and MOS within strides were estimated. Mean ML-COM displacement and velocity, which were significantly higher in older compared to young adults, were significantly reduced in the NSW condition while the variability of ML-COM velocity was increased in the NSW condition. A significant interaction effect of step width and age was found for ML-COM velocity, showing larger decreases in older adults in the NSW condition. Walking with NSW reduced the ML-MOS significantly in both groups while it was smaller in the older group. Although reductions of ML-COM displacement and velocity may occur as direct mechanical effects of reduced step width, the larger variability of ML COM velocity in the older adults suggests active control of ML COM movements in response to the reduced base of support. Given the effects on MOS, narrowing step width might challenge ML-balance control and lead to less robust gait especially in older adults.

We experimentally investigated whether unilateral hip abductor muscle fatigue affected gait control and hip position sense in older adults. Hip abductor muscles were fatigued unilaterally in side-lying position in 17 healthy older adults (mean age 73.2 SD 7.7 years). Hip joint position sense was assessed by an active-active repositioning test while standing and was expressed as absolute and relative errors. Participants walked on a treadmill at their preferred walking speed, while 3D linear accelerations were collected by an inertial sensor at the lower back. Gait parameters, including step and stride time, local divergence exponents and harmonic ratio were quantified. In fatigued gait, stride time variability and step-to-step asymmetry in the frontal plane were significantly increased. Also a significantly slower mediolateral trunk movement in fatigued leg late stance toward the non-fatigued leg was observed. Despite these temporal and symmetry changes, gait stability in terms of the local divergence exponents was not affected by fatigue. Hip position sense was also affected by fatigue, as indicated by an increased relative error of 0.7° (SD 0.08) toward abduction. In conclusion, negative effects of fatigue on gait variability, step-to-step symmetry, mediolateral trunk velocity control and hip position sense indicate the importance of hip abductor muscles for gait control.

Previously it has been shown that constraining step width in gait coincides with decreased trunk displacements. Conversely, external stabilization of the upper body in gait coincides with decreased step width, but this may in part be due to changes in passive dynamics of the leg. In the present study, trunk kinematics during gait were constrained without external stabilization by using an orthosis, to investigate whether step width and dynamic gait stability in the ML direction are changed in relation to trunk kinematics. Nine healthy young adults walked on a treadmill at three different speeds with no intervention and while wearing a thoracolumbar orthosis. Based on marker trajectories, trunk COM displacement, body COM displacement and velocity, step width, and margin-of-stability in ML direction were calculated. The results showed that the orthosis significantly reduced trunk and body COM displacements. As hypothesized, the restriction of trunk movement coincided with significantly decreased step width, while the margin-of-stability was not affected. These findings indicate that, when trunk movements are constrained, humans narrow step width, while maintaining a constant margin-of-stability. In conclusion, the present results in combination with previous work imply that in gait a reciprocal coupling between trunk kinematics and foot placement in the frontal plane subserves control of stability in the frontal plane.

Proprioceptive acuity of the knee and hip joints in healthy older adults can be reliably assessed with an active-active procedure in a standing position with respect to relative and absolute error.