We assessed 56 healthy adults, 24 individuals with PD on medication and 11 individuals with MS. Healthy adults were recruited via flyers in public facilities. The patients were recruited from the wards of the neurology department of the University hospital of Schleswig-Holstein (UKSH), Campus Kiel, Germany. Exclusion criteria were a Montreal Cognitive Assessment (MoCA) test score below 15 and the need of a walking aid. All participants underwent a clinical examination and a static postural stability assessment (see below). The clinical examination consisted of the motor part of the Unified Parkinson's Disease Rating Scale revised by the Movement Disorders Society (MDS-UPDRS III) and the MoCA. The MS patients were additionally evaluated with the Expanded Disability Status Scale (EDSS). The full study protocol can be found elsewhere (36).

The ethical board of the medical faculty of Kiel University approved this study (project number D438/18) and all participants gave their written informed consent.

The study participants performed the three static postural stability tasks that are included in the short physical performance battery (37). For the first task, participants were asked to stand with their feet together, side-by-side. For the second task, they were asked to stand with the side of the heel of one foot touching the big toe of the other foot, the semi-tandem position. For the third task, they were asked to stand with the heel of one foot in front of the toes of the other foot, the tandem position (Figure 1). The participants were allowed to put either foot in front (for the semi-tandem and tandem position) and were asked to stand still and hold each position for 10 s.

Postural sway was measured using an IMU (MyoMotion, Noraxon Inc, Scottsdale Arizona, USA) worn on the lower back. The parameters were extracted from the acceleration data as described and validated in another study (7). An explanation of the parameters can be found in Table 1. The acceleration, velocity and jerk parameters were calculated in both AP and ML direction.

ANOVAs were used to analyze the differences between the groups per balance task. Homogeneity was corrected according to Welch and Post Hoc testing was performed according to Tukey's method. Repeated measures ANOVAs were used to analyze the differences between balance tasks within each group. If the assumption of sphericity was rejected, a Greenhouse-Geisser correction was used. Post Hoc testing was performed with Holm's correction.

Directly comparing the PD and MS group was not possible because of the mean age difference of 21 years. Therefore, the data of the healthy participants was used to estimate what the value of the parameters would be at each age with the use of a quadratic function (Figure 2A). For each patient and each task, the distance between the quadratic fit and their value at the appropriate age was calculated (Figure 2B). These obtained values were then used to calculate the change between the tasks for both PD and MS patients by subtracting semi-tandem stance from side-by-side stance and by subtracting tandem stance from semi-tandem stance. To be able to see how these changes relate to healthy adults, the average differences between the quadratic fits were calculated (Figure 2C).

During side-by-side stance, significant differences between the groups were found for sway area (p = 0.042), velocity in ML direction (p = 0.035), acceleration in ML direction (p = 0.011) and jerk in ML direction (p = 0.034). Post hoc tests revealed that PD patients showed significantly higher velocity in ML direction (p = 0.001), compared to healthy adults (Figure 3). MS patients showed a significantly higher acceleration (p = 0.023) and jerk (p = 0.040) in ML direction, and a tendency toward a larger sway area (p = 0.055), compared to the healthy adults (Figure 3).

During semi-tandem stance, significant differences between the groups were found for velocity in ML direction (p = 0.020). Post hoc tests revealed that PD patients had a significantly higher sway velocity in ML direction (p = 0.049), compared to the healthy adults. During tandem stance, significant differences between the groups were found for velocity (p = 0.026) and acceleration (p = 0.005) in AP direction. Post hoc tests revealed that the MS patients showed a significantly higher acceleration in AP direction (p = 0.009) and a tendency toward a higher velocity in AP direction (p = 0.053), compared to the healthy adults (Figure 3).

There were no significant differences between PD and MS patients in any of the three tasks (Figure 3).

Healthy adults showed a significant increase in sway area (p = 0.044), velocity in ML direction (p = 0.007), acceleration in ML direction (p = 0.001) and jerk in both AP (p = 0.001) and ML (p = 0.001) direction between the side-by-side and tandem stance (Figure 3 and Table 4). The sway area (p = 0.028), velocity in ML direction (p = 0.007), acceleration in ML direction (p < 0.001) and jerk in both AP (p <0.001) and ML (p < 0.001) direction significantly increased between the semi-tandem and tandem stance (Figure 3).

MS patients showed a significant increase in sway velocity (p = 0.032), acceleration in ML direction (p = 0.023) and jerk in ML (p = 0.025) and AP (p = 0.047) direction between side-by-side and tandem stance. Moreover, they showed a significant increase in sway acceleration in ML direction (p = 0.015) and jerk in ML direction (p = 0.017) between semi-tandem and tandem stance (Figure 3).

PD patients did not show a significant change in any of the sway parameters between the tasks (Figure 3).

Clear differences in adaptations to increasingly complex stance tasks between PD and MS patients were observed (Figure 4). PD patients showed gradual increases in sway parameters from the least to the most complex stance task. MS patients showed a tendency toward smaller values compared to the healthy adults in sway parameters from the side-by-side stance to the semi-tandem stance, except for velocity. Velocity in MS patients was relatively low during side-by-side stance, compared to the other groups and showed small increases from the side-by-side stance to the semi-tandem stance. MS patients showed large increases in sway between the semi-tandem and tandem stance. This increase was for the area and jerk in ML direction even more than two times the values of the side-by-side stance (Figure 4).

The only statistically significant differences between groups could be found when the groups were compared to healthy adults. In line with a previous study that compared PD patients to age-matched healthy adults, we found higher sway velocities of PD patients (26) during the side-by-side stance.

Unlike another study, at our study the larger sway areas and higher jerk of PD patients compared to healthy controls were not significant (25). However, in the previous study participants were untreated and had to stand still for 120 s. In the present study, only PD patients on dopaminergic medication as well as patients with a combination of deep brain stimulation and dopaminergic medication were included. Thus, it could be that dopaminergic medication as well as deep brain stimulation lead to at least some improvements in static postural sway. There are however discordant findings about this effect in the literature (40–43). MS patients showed a higher sway area, acceleration (ML direction) and jerk (ML direction) during side-by-side stance compared to healthy adults, which is comparable with other studies (32, 44).

During the semi-tandem stance there was only a significant difference between healthy adults and PD patients for the velocity in ML direction. The differences that were found between healthy adults and MS patients during the side-by-side stance are not present anymore during the semi-tandem stance. It seems that the MS patients have “decreased” their sway during this slightly more complex task. It could thus be that, in MS, the increase in the base of support in AP direction leads to a more stable static balance.

During the tandem stance, only the velocity and acceleration in AP direction were significantly different between the healthy adults and MS patients. The difficulty with the tandem stance in MS patients could be caused by spasticity. It has been shown that spasticity negatively affects balance in patients with MS, even with low levels of spasticity (45). MS patients with higher levels of spasticity have also been described to have more postural sway in ML direction (46), which is basically in accordance with our findings.

As a general comment, these differences across studies may be due to different (sub)cohorts included in respective studies. However, since the current study does not focus on this comparison but mainly on differences between sway tasks (see below), we do not consider these differences relevant here.

Adaptation of the body posture is required to compensate for the change in the base of support with the increasingly difficult stance tasks. Differences in these adaptation strategies between groups were visible. Both healthy adults and MS patients showed changes in sway parameters. These results suggest that the healthy adults and MS patients seem to be able to adapt their postural sway to the complexity of the task. However, only the healthy adults were able to increase area significantly over increasingly difficult balance tasks, which could at least indirectly speak for the possibility that both MS and PD already sway so much in the simple side-by-side task that an adequate increase of the sway area is no longer possible even in more difficult tasks.

PD patients did not show a significant increase in postural sway parameters with increasing complexity of the task and seem therefore not able to adapt their sway adequately to increasingly difficult balance tasks, or are not able to afford an increase in postural sway without exceeding their limits of stability. This could be related to PD symptoms, but could also be due to the age-related impairments of the somatosensory (9) and the vestibular system (15), making flexible and quick postural responses more difficult.

Our study shows clear differences in postural adaptation strategies between PD and MS patients. PD patients do not seem to adapt more to increasingly difficult static balance tasks than the healthy adults, although PD patients have lower postural stability and increased fall risk compared to age-matched healthy adults (19). It could be that the PD patients are not able to adapt their sway much because of their bradykinesia and rigidity. However, it is also known that rigidity can increase postural sway by preventing the use of flexible responses and thereby aggravating the deterioration of postural reflexes (22, 23). MS patients increased their sway a lot, especially area, acceleration and jerk, when adapting to the most complex task. It is known that more than 80% of the MS patients have spasticity (31, 47) and that spasticity has a negative effect on postural stability (45, 46). When we consider the results of our study and those of another study (48) together, it seems likely that MS patients have substantial sway problems especially when the base of support is narrow (or the ML dimension small), such as during the tandem stance. A timed 3 m tandem walk was also significantly better in separating asymptomatic and symptomatic (in the motor or cerebellar functional system) MS patients compared to a timed 25-foot walk (49).

We suggest that whether the patients can perform control-like (PD patients) or not (MS patients) could have to do simply with the location of the pathology within the central nervous system. In PD, mainly the basal ganglia are affected. Due to the localization of the affected areas in the parenchymal structures and large networks of the central nervous system, there are extensive possibilities for compensation, both by upstream and downstream nodes (50), and by using alternative pathways or even networks (51). In MS patients the involvement of the pyramidal tract explains spasticity. The pyramidal tract is markedly different from the basal ganglia. This tract is the brain's only gateway to motor movement, and if this gateway is damaged, then no compensation can occur. This is well-studied, for example, in paraplegic syndromes (52). This idea is supported by results from exercise training on falls in these diseases. Whereas, PD patients are obviously able to reduce the number of falls after specific trainings (53–55), exercise studies suggest that this may not be the case in MS (55, 56).

Based on the results of this study, the following may be stated with regard to therapies: PD patients should benefit from exercises that are also effective for adults without PD. This is supported by the “control-like” adaptation of sway to increasingly difficult balance tasks. In addition, they should be trained to fully exploit their sway potential. Our data suggest that PD patients tend to underuse the whole arsenal of adaptation possibilities (no significant increase from easy to difficult task exercise). For this type of training, there is certainly a lesson to be learned from the Lee Silverman Treatment Strategy (57), although specific studies in this field are still lacking. In MS, due to the presumed lack of compensation possibilities at the level of the damaged pyramidal tract, preventive training seems more appropriate. Recent studies showed for example that Vojta reflex therapy seems to be improving balance in MS patients (58, 59).

While most studies measured PD patients OFF medication, the present study included PD patients that were ON medication as well as patients with a combination of deep brain stimulation and medication. Both treatment options have been shown to influence postural sway performance (40–43). Some studies even suggest contradictory effects of dopaminergic medication and deep brain stimulation on postural sway (60). Therefore, putting all PD patients in one group could have had some effect on the results. However, by measuring PD patients ON medication and deep brain stimulation, this study provides more realistic information on the postural sway of PD patients in their daily lives.

The PD patients had a slightly lower general cognitive performance based on the MoCA score compared to the healthy adults and MS patients. In the PD patients the MoCA score is negatively correlated with the sway path length (61). Therefore, the lower cognitive performance in PD could have led to slightly higher sway values. Since we are basically presenting an intra-individual comparison of static balance across different difficulty levels, we do not assume that the preceding information is relevant to this study. Moreover, number of patients were relatively small and the day-to-day variability in the static sway parameters can be relatively high in neurological patients (6), therefore the results need to be interpreted with caution especially concerning sensitivity. Vice versa, it is very likely that the detected significances actually represent true differences.