Sample characteristics

Fourteen healthy older adults (66.9 ± 4.3yrs, 1.73 ± 0.08m, 76.9 ± 10.8kg) and sixteen healthy young adults (26.3 ± 3.6yrs, 1.72 ± 0.10m, 66.2 ± 14.0kg) were recruited from the greater St. Pölten (AUT) metropolitan area. Inclusion criteria included individuals aged between 60–75 years old for the older adults and 20–35 years old for the younger adults. Further, all participants were required to be able to walk independently without the use of external support. Exclusion criteria included any medically diagnosed diseases that affected gait, surgeries affecting the musculoskeletal systems in the last six months, musculoskeletal injuries in the past six months that affected gait, and medically diagnosed neurocognitive impairments. This study was approved by the Ethics Committee of Lower Austria in accordance with Declaration of Helsinki and all participants signed an informed consent prior to participating.

Data collection and analysis

Participants walked along a 12m walkway at their preferred speed until three trials each containing one full gait cycle were collected by the force plates. Three-dimensional motion capture was conducted with a 16 camera Vicon system (Nexus, 2.14, Vicon, Oxford UK) and three force plates (AMTI, Watertown, MA) at 120Hz and 1200Hz, respectively. Participants were outfitted with a full-body marker set using the Cleveland Clinic marker set for the lower extremity and the Vicon Plug-In Gait model for the upper body [13]. Raw trajectory and ground reaction force data were filtered with a 4^th order low-pass Butterworth filter with a 10Hz and 15Hz cut frequency, respectively [9]. Data were then exported to Visual 3D (C-Motion, Germantown MD) for reconstruction as a 15-segment full-body model for 3D kinematic and kinetic calculations [23, 24]. The BOS per step was defined as the instantaneous COP position which provides a more accurate representation of the effective BOS [15, 25]. The COM, COM velocity, COP, and joint moments (normalized to body mass) were calculated in Visual 3D prior to processing in custom Matlab 2023b (Mathworks, Natick MA) scripts for further analyses. The MOS was calculated in the sagittal and frontal planes for each foot with the formula:

Where xCOM is the extrapolated COM calculated as:

In which COMp = COM position and COMv = COM velocity. ωθ was calculated as:

In this term, g = 9.81m/s² and l is the length of the inverted pendulum determined as the average distance of the right/left lateral heel marker to the COM [15]. Data for the left and right legs were averaged across the three trials per participant. MOS and joint moments were calculated in the sagittal and frontal planes. For the knee joint, only sagittal plane joint moments were assessed. All data were time normalized to 101 points. Joint moments were calculated across the gait cycle and per step while the MOS was calculated only per step.

Main findings

This study examined if healthy aging affects the relationship between the MOS and leg joint moments in healthy young and older adults. Contrary to our hypothesis, older adults had a larger MOS in the frontal plane than the younger adults while demonstrating no differences in the sagittal plane for this metric. However, in line with our hypothesis, older adults had a stronger relationship between the MOS and hip joint moments in the sagittal plane than the younger adults. The stronger relationship in the sagittal plane between the MOS and the hip joint moment for older adults may indicate that healthy aging increases the coupling between the hip moments and the COM to control dynamic stability during single-support. Additionally, the larger frontal plane hip abduction moment observed in our older adults, compared to young adults may reflect an effort to control the trajectory of the xCOM in the frontal plane as it moves away from the BOS.

Margin of Stability

In contrast to our hypothesis, the healthy older adults in our study had a larger MOS in the frontal plane than the younger adults during the single-support and the beginning of the second double-support phases. This finding is unexpected as a larger value for the MOS in the frontal plane is suggested to indicate greater instantaneous dynamic stability [15, 18, 29–31]. Specifically, previous literature proposes that the further medial the xCOM is from the lateral edges of the BOS (indicated by a larger and more positive MOS), the greater the level of dynamic stability is in the frontal plane [15, 30, 32]. Key methodological differences may account for this discrepancy.

For instance, research thus far quantified the MOS in the frontal plane at discrete time points in the gait cycle (initial-contact and toe-off) and/or utilized foot markers to define the BOS lateral boundaries [16, 17, 21, 22, 30, 31, 33, 34]. Several limitations exist in the quantification and interpretation of the MOS when exclusively examining initial-contact and toe-off. Specifically, initial-contact occurs during the double-support phase when the BOS is largest and corrective action to adjust potential errors in the trajectory of the COM can be larger [6, 9, 10]. Indeed, as both feet are on the ground, the location of the COP can be shifted over a longer and wider surface area to adjust the location of the COM after foot placement is achieved [6, 9, 10]. This corrective action, however, is more limited during the single-support phase as the BOS of the single foot is shorter and narrower [6, 9, 10]. This poses a particular challenge to the neuromuscular system when maintaining dynamic stability as the COM reaches its most lateral position during the single-support phase [9, 10]. Therefore, to control the large inertial load of the upper body, the neuromuscular system generates stabilizing moments from the ankle and hip to avoid a loss of balance [10]. However, the effect of these moments on the MOS would require time for execution and would not necessarily be quantifiable at the exact instant of toe-off. Additionally, previous research has predominately utilized foot markers to define the BOS which may underrepresent the COP excursion underneath the foot during the gait cycle which, in turn, directly impacts the quantification of the MOS [25]. Therefore, in both cases, critical information relating to how aging affects the mechanisms contributing to dynamic stability and fall risk in older adults may be overlooked. Only by quantifying the MOS across a step with the excursion of the COP did we find differences between the older and younger adults.

The larger frontal plane MOS in our older adults compared to the younger adults, however, raises the notion that the MOS quantifies a neuromuscular strategy aimed at avoiding loss of dynamic stability rather than a quantification of an individual’s mechanical stability level that would reflect fall risk. Specifically, the larger frontal plane MOS in our older adults may indicate attempts to reduce the likelihood of lateral balance loss by shifting their xCOM away from the lateral boundaries of their BOS. While this strategy may be aimed at mitigating lateral balance loss, the xCOM is brought closer to the medial borders of the BOS which may cause pre-mature termination of the contralateral swing phase if medially directed destabilization occurs. Therefore, future research on falls should consider examining the effect of mediolateral perturbations applied during the single-support phase to determine whether this motor strategy leads to a heightened fall risk in older adults.

Unlike the frontal plane, no group differences were observed in the sagittal plane for the MOS. The lack of differences in this plane may be because of the distinct neuromuscular mechanisms that control dynamic stability in the sagittal compared to frontal plane [5]. Control of the COM in the sagittal plane while walking is considered predominately regulated by automated sub-cortical and somatosensory mechanisms [5, 35]. In contrast, the mechanics of the frontal plane are controlled by active mechanisms that stem from higher level cortical regions [5, 35]. Within this multifaceted control system, current evidence suggests that the active mechanisms which control frontal plane mechanics are particularly sensitive to the neurodegeneration that occurs from healthy aging while the sub-cortical mechanisms remain relatively unaffected [5]. Indeed, Bruijn et al. [36] found that a lower white matter microstructural organization in cortices of older adults, compared to younger adults, was strongly correlated with narrower steps and a smaller MOS in the frontal plane during treadmill walking. Thus, in our study, no sagittal plane differences in the MOS further supports the notion that healthy aging has a more limited impact on dynamic stability in the plane of progression [5, 36]. Instead, the aging neuromuscular system may strive to execute a more conservative control of the COM in the frontal plane during overground walking by maintaining their upper body closer to the center line of progression. This may be a preemptive strategy to potentially curb the neurodegenerative effects on the active mechanisms that control frontal plane dynamic stability during overground walking.

As such, future research should consider examining the allocation of attentional resources, which stem from higher level cortical regions (ex: pre-frontal, frontal, and primary motor cortices), over the full gait cycle to reveal underlying compensation strategies implemented by healthy older adults. While the literature on dual-tasking is extensive, the evidence thus far is sparse in examining the effects of a secondary dual-task on dynamic stability when assessed with statistical parameter mapping [37, 38]. Additionally, further research should consider applying neuroimaging to these dual-tasking protocols to provide direct evidence for changes related to the allocation of attentional resources when controlling dynamic stability in healthy older adults.

Joint kinetics

Our hypothesis was partially supported as the older adults had a stronger correlation, than the younger adults, between the hip joint moment and the MOS only in the sagittal plane. This finding may indicate that, in the sagittal plane, older adults exert a tighter control on their xCOM to attenuate the destabilizing forces acting on the upper body. Although no group differences were observed in the relationship between the MOS and the joint moments in the frontal plane, the older adults had a larger frontal plane hip moment during the single-stance phase than the younger adults.

In the sagittal plane, the hip joint reaction force is the primary destabilizing force which causes the upper body to flex during weight-acceptance and extend during push-off [6]. Further, during the swing phase, the xCOM is held in front of the BOS to progress the body forward [5, 6, 9, 13]. As the BOS is located behind the xCOM, corrective action by the BOS to contribute to one’s dynamic stability level is only feasible by effective foot placement at upcoming initial contact [6, 8, 9]. Therefore, to control the upper body’s movement, the musculature of the hip executes an opposing moment to maintain upright posture and dynamic stability during the single-support phase [6]. In our study, the role of the hip muscles in controlling dynamic stability would account for the very strong relationship between these parameters in the sagittal plane for both groups. Furthermore, the stronger relationship for these parameters in our older, than younger, adults is in line with the relatively greater contribution from their hip than ankle moments in this demographic while walking [14]. However, the group difference in the sagittal plane hip and MOS pairing was relatively small and did not coincide with any additional group differences in this plane. This may indicate that older adults, compared to young, increase the coupling between the hip and upper body to maintain dynamic stability, without the need to increase the generation of joint moments at the hip.

In contrast to the sagittal plane, gravity is the primary destabilizing force in the frontal plane causing the upper body to accelerate medially toward the center line of progression during the single-support phase [10]. Thus, to avert a fall, the hip abductors generate a nearly equal and opposite counterbalancing moment to control the medial acceleration of the COM [6, 10]. Previous research suggests that this careful balancing of moments depends on the ability of the somatosensory system to sense the kinematic state of the COM during the stance-phase, so that the neuromuscular system can then generate an effective counterbalancing hip moment [6, 10]. This intricate system then maintains an individual’s erect posture and reduces the likelihood of a mediolateral balance loss or the need for corrective stepping [6, 9, 10]. In relation to our study, the increased hip moment and frontal plane MOS for older than younger adults occurred during the single-stance phase. This may indicate that, compared to the young adults, our older adults increased their hip abduction moment to better control the larger medial acceleration of the xCOM while their MOS in the frontal plane is enlarged.

The increased hip abduction moment was, however, brief despite the continued enlargement of the frontal plane MOS into the double-support phase in our older adults. One possible explanation may be that the purpose of the brief increase in the hip abduction moment was to mitigate small deviations from the nominal trajectory of the xCOM in the frontal plane as it separated from the BOS of the stance limb. Indeed, in the frontal plane, small deviations to the nominal trajectory of the upper body exist and must be constantly attenuated by the neuromuscular system [5, 29]. Previous research demonstrates that such deviations also arise during unperturbed walking due to the natural variability in an individual’s gait pattern [5, 29, 39]. In older adults, compared to young, current evidence demonstrates a further increase in the deviations to this nominal trajectory which has been linked to age related increases in gait variability [5, 40–42]. Therefore, in our study, the brief increase in the hip abductor moment in our older adults may have served to attenuate potentially deviations in the monophasic trajectory of the xCOM during the gait cycle [5, 6, 10].

It should be noted, however, that a certain amount of natural variability is still necessary in this control system to facilitate gait adaptation to environmental demands [5, 39]. This would account for the moderate relationship between the hip and MOS in the frontal plane for both groups in our study. The moderate relationship between these parameters may indicate a control system whereby the medial acceleration of the COM is flexibly regulated to maintain its trajectory toward to its position at upcoming initial contact. Further, as the kinematic state of the COM is the primary determinant of step width, a stringently tight system would reduce the natural variability in the neuromuscular system which facilitates gait adaptation to changes in environmental demands [5, 31, 39, 43].

Limitations

When considering the findings of our research several limitations should be considered. In our study, we did not record participants’ fall history nor fear of falling. Both aspects are shown to cause changes in the gait pattern in older adults. Thus, we are unable to rule out any potential adjustments or impacts fall history or fear of falling have on our findings. Additionally, by examining differences between healthy younger and older adults, our results do not examine at which point during the adult lifespan that changes to control of the frontal plane MOS begin to occur. Finally, we conducted our analysis on a convenience sample which may have limited our ability to detect all potentially significant differences between older and younger adults.