**Age-Related Di**ff**erences in Muscle Synergy Organization during Step Ascent at Di**ff**erent Heights and Directions**

#### **Remco J. Baggen 1,2, Jaap H. van Dieën 2, Evelien Van Roie 1, Sabine M. Verschueren 3, Georgios Giarmatzis 4, Christophe Delecluse <sup>1</sup> and Nadia Dominici 2,\***


Received: 30 January 2020; Accepted: 5 March 2020; Published: 14 March 2020

**Abstract:** The aim of this study was to explore the underlying age-related differences in dynamic motor control during different step ascent conditions using muscle synergy analysis. Eleven older women (67.0 *y* ± 2.5) and ten young women (22.5 *y* ± 1.6) performed stepping in forward and lateral directions at step heights of 10, 20 and 30 cm. Surface electromyography was obtained from 10 lower limb and torso muscles. Non-negative matrix factorization was used to identify sets of (*n*) synergies across age groups and stepping conditions. In addition, variance accounted for (VAF) by the detected number of synergies was compared to assess complexity of motor control. Finally, correlation coefficients of muscle weightings and between-subject variability of the temporal activation patterns were calculated and compared between age groups and stepping conditions. Four synergies accounted for >85% VAF across age groups and stepping conditions. Age and step height showed a significant negative correlation with VAF during forward stepping but not lateral stepping, with lower VAF indicating higher synergy complexity. Muscle weightings showed higher similarity across step heights in older compared to young women. Neuromuscular control of young and community-dwelling older women could not be differentiated based on the number of synergies extracted. Additional analyses of synergy structure and complexity revealed subtle age- and step-height-related differences, indicating that older women rely on more complex neuromuscular control strategies.

**Keywords:** EMG; muscle synergies; forward stepping; lateral stepping; aging; neural control

#### **1. Introduction**

Aging is associated with gradual changes in neuromuscular control [1,2]. Eventually, these changes can have a major impact on fall risk, mobility and independence in older adults [3–5], which may be exacerbated in post-menopausal women due to accelerated muscle wasting [5]. One major challenge with assessing changes in neuromuscular control in healthy community-dwelling older adults is that, due to the gradual nature of age-related neuromuscular deterioration, pre-clinical changes in neuromuscular control of everyday tasks may go undetected [6]. In healthy older adults, changes in neuromuscular control can be revealed by increasing task challenge. For example, by increasing gait speed, cadence, step length, and step height [7–9].

Step ascent, a functional task in daily life, provides a challenge that can easily be modified by increasing step height and has been demonstrated to be an effective training stimulus to improve muscle volume and functional ability in older women [10,11]. However, the effects of aging and task challenge on the neuromuscular control strategies behind step ascent have not yet been thoroughly explored. In a previous study conducted with older women, we found that peak activation of several major lower limb muscles occurred during the ascent phase of stepping and that there is a positive dose-response relationship between step height and peak muscle activation [10]. However, the increase in step height was also accompanied by increased between-subject variance of peak activation magnitudes [10]. This could be attributable to a tendency of older adults to modulate their motor strategies (e.g., shifting joint moment generation from the knee to the ankle), in order to operate within the limits of their physical capacity when ascending steps [12,13]. Additionally, it is unknown if other observed age-related changes in neuromuscular control strategies, such as increased antagonistic co-contraction of quadriceps and hamstrings to help maintain postural control during dynamic tasks [14–17], may interact with increased task challenge.

One way to assess neuromuscular control is through muscle synergy analysis. Previous studies have found that low-dimensional sets of motor modules, also known as muscle synergies, can be used to reconstruct muscle activation patterns during various motor tasks [18–22]. These synergies are composed of groups of muscles that are assumed to be activated by a single neural command [23]. It is thought that the central nervous system employs this modular organization to reduce the large number of degrees of freedom inherent to the redundancy of the human musculoskeletal system [24], and to allow for flexible but accurate response selection during motor tasks [25]. However, some researchers have argued that modular recruitment of muscles might represent predetermined control strategies and could merely be an effect of task constraints or optimized performance criteria, rather than reflecting neural control strategies employed by the central nervous system [23,26]. Regardless of the mechanisms underlying modular organization of muscle activation, extracting muscle synergies from electromyographic (EMG) signals can provide important insights about neuromuscular control strategies used to perform functional tasks [27]. In older adults with a history of falls, declines in neuromuscular control are reflected in a decreased number of extracted synergies from walking tasks that challenge dynamic balance [6]. A decreased number of synergies is indicative of decreased complexity of motor control or a decreased motor repertoire. These underlying changes in synergy complexity can be quantified using variance account for (VAF) by a given set of extracted synergies and defining the number of synergies required to adequately reconstruct the original EMG signals (indicated by a-priori or a-posteriori set thresholds for VAF) [28]. For example, high VAF by a limited number of synergies represents decreased complexity of motor control, which is often associated with neuromuscular pathologies such as cerebral palsy and stroke [6,27,29] and characterized by increased levels of co-activation between individual muscles [30]. However, healthy aging is associated with a more gradual process of physical decline and thus changes in complexity of motor control may not manifest in a decreased number of synergies [8]. Consequently, age-related changes in motor control may be better reflected by comparisons of VAF for a fixed number of synergies [29,31,32] or by assessing changes in spatio-temporal organization of muscle synergies, such as altered module composition and shifts in activation patterns [2,6]. Differences in spatio-temporal organization within a stable number of synergies can arguably be considered more subtle than differences in the total number of synergies as they tend to reflect compensatory or alternative motor strategies in order to overcome increased or altered task challenges, whereas a decreased number of synergies is usually used as an indication of neural impairments. Although it is currently unknown how aging and task intensity affect muscle synergy organization during stair ascent, analyses of underlying differences could provide a basis to improve detection of pre-clinical age-related deterioration in neuromuscular control and more effectively target fall prevention programs at individuals most at risk.

Therefore, the purpose of this study was to explore muscle synergy recruitment during step ascent in forward and lateral directions and with incremental step heights in young and older adults. Specifically, we aimed to assess the effects of age-related changes, task intensity, and their interaction on complexity and organization of motor control by comparing the number of extracted synergies, variance accounted for (VAF) by a fixed number of synergies, and spatio-temporal characteristics of the extracted synergies across conditions.

#### **2. Materials and Methods**

#### *2.1. Participants*

Eleven older women (67.0 *y* ± 2.5, 161.3 cm ± 4.9, 64.4 kg ± 6.8) and ten young women (22.5 *y* ± 1.6, 168.9 cm ± 1.7, 64.2 kg ± 7.9) were recruited for this study. Potential participants were excluded if they suffered from neurological or motor disorders, impaired balance control, or if they had been involved in a structured training program in the last 6 months prior to participation in the study. This study was approved by the Human Ethics Committee of KU Leuven in accordance with the Declaration of Helsinki and registered with the Clinical Trial Center UZ Leuven (S56405). All participants signed informed consent prior to participation in the study.

#### *2.2. Experimental Protocol*

Participants performed a series of stepping tasks consisting of stepping onto a wooden block in forward direction (Fstep) and in lateral direction (Lstep). Task intensity was determined by the height of the block (10, 20 and 30 cm) [10]. A previous study by Singh and Latash revealed that muscle fatigue can cause higher variability of muscle activation patterns and composition of synergy components [33]. In contrast with more common tasks used for muscle synergy analyses (e.g., gait and perturbation trials), step ascent with incremental heights presents a relatively high physical challenge for older adults [10], thus increasing the risk of fatigue with a high number of repetitions. Pilot testing showed that, despite regular breaks, execution of more than three repetitions per trial, combined with the number of trials, was already quite fatiguing for the older women in this study. Therefore, each trial consisted of three repetitions to avoid confounding effects of fatigue on synergy composition. Every repetition was performed with the dominant leg first, followed by the trailing leg, and ending in double support on top of the block. Left-right dominance was determined during familiarization by noting with which foot participants preferred to take the first step. As a control question, participants were asked with which foot they would prefer to kick a ball [34]. Only two participants (both young women) showed left-side dominance. Step ascent was assessed in both forward and lateral directions, as these are functional tasks that require simultaneous coordination of the hip, knee and ankle musculature and can be challenging for older adults [12,13]. Stepping up in forward direction shows close functional resemblance to stair-climbing [34], which is an activity of daily life associated with high fall risk in older adults [35]. Stepping up in lateral direction is a less common task for older adults and can provide an additional challenge to the hip abductors, which play an important role in medio-lateral balance control [36]. The speed of task execution was controlled by a metronome at 1 second for ascent, 1 second stance, and 1 second descent to avoid differences in muscle activation due to explosive movements.

#### *2.3. Kinematics*

Kinematics were recorded at a sampling rate of 100 samples/s by means of 3D motion capture cameras (Vicon®, Oxford Metrics, Oxford, UK). Infrared reflective markers (diameter 14 mm) were placed on both heels and the sacrum. Only data from the ascent phase were used for analysis. Based on the kinematic data, the start of the ascent phase was defined at 200 ms prior to initial vertical displacement of the leading heel marker beyond 2× the standard deviation obtained during normal stance. The end of the ascent phase was defined at 500 ms after maximum knee extension, defined as the maximum relative distance between the heel and sacrum. Two sub-phases (foot clearance and pull-up) of step ascent were defined using the dominant heel marker trajectories. The shift from foot clearance to pull-up phase was defined at 100 ms after vertical displacement of the leading heel marker dropped below 2× the standard deviation obtained during normal stance. Kinematic data from the trailing heel were included to detect and eliminate trials with undesirable events, such as the toe getting caught on the edge of the step. No such events were detected.

#### *2.4. Electromyography*

Muscular activation was collected unilaterally from ten lower limb and trunk muscles on the dominant side using surface electromyography (EMG) (Aurion®, ZeroWire, Milan, Italy) sampled at 1000 samples/s. Activation was recorded from the following 10 muscles: the tibialis anterior (TA), the lateral head of the gastrocnemius (GL), soleus (SOL), vastus lateralis (VL), rectus femoris (RF), biceps femoris (BF), semitendinosus (ST), gluteus maximus (GMAX), gluteus medius (GMED), and the erector spinae (ERS), in accordance with SENIAM guidelines [37]. The skin was shaved and thoroughly rubbed with an alcohol swab to ensure optimal conductivity. Bi-polar Ag/Ag-Cl electrodes (Ambu® BlueSensor P, Ballerup, Denmark) were then placed on the belly of the muscles with an inter-electrode distance of 25 mm. Sampling of kinematic and EMG data was synchronized.

#### *2.5. Synergy Extraction and Data Analyses*

All EMG and kinematic data were processed using custom MATLAB scripts (MATLAB R2014b, MathWorks®, Natick, MA, USA). The EMG signals were high-pass filtered with a 1st order Butterworth filter with a cut-off at 20 Hz, full-wave rectified and smoothed with a 0.1-s moving average window [10,38]. EMG signals from forward and lateral stepping were normalized to the respective maximum activation obtained over all trials performed in the congruent direction so that activation could not exceed 100% [39,40]. The EMG signals were time-synchronized with the kinematically defined start and end points and subsequently normalized over time to define 0%–100% of the step cycle. Finally, because EMG data were only collected during step ascent and therefore represented intervals, rather than continuous activation patterns (as would be the case during gait trials), signals were averaged over the three repetitions performed in each condition. The choice to average signals rather than concatenating them was made in order to obtain the best reconstruction quality for our relatively short intervals, at the risk of losing information on step-to-step variability [40].

Muscle synergies were extracted from the individual average EMG data matrix using non-negative matrix factorization (NNMF). NNMF calculates muscle synergies (W) and their relative temporal activation patterns (C), resulting in muscle activations being represented as W × C + e. W represents the relative muscle co-activation, defined as the relative weight of each muscle per synergy, and is constructed as an *m* × *n* matrix where *m* is the total number of muscles and *n* is the selected number of synergies. C represents the temporal activation patterns and is constructed as an *n* × *t* matrix where *t* represents the number of data points over normalized time (100 per individual trial) and e is the residual error matrix [22,41]. The algorithm was repeated 1000 times for each subject to avoid local minima. The appropriate number of synergies was defined using two criteria. First, using an iterative process where the number of synergies varied between 1 and 10, the minimum number of synergies was selected based on the number required to reach ≥ 85% of group-averaged variance accounted for (VAF). As an additional local criterion, synergies had to account for ≥75% VAF for each individual muscle [42]. This double criterion approach was selected in order to adequately reproduce relevant features of the synergy compositions. VAF was defined as the uncentered Pearson correlation coefficient between W × C and the EMG amplitude time series. To compare spatio-temporal characteristics, individual synergies obtained from different subjects were pooled and matched based on the correlation of their structure (muscle weightings in each synergy of W) using a custom cluster analysis algorithm [43]. If a synergy showed equal correlation to more clusters, that synergy remained in the pool it was initially assigned to. Each synergy of W and C was subsequently averaged over all participants in that age group. For comparisons between age groups, the group-averaged synergies were also matched based on their structure using cluster analysis. Finally, we computed time-averaged standard deviations of

the synergy activation patterns, with a fixed number of synergies, to assess if age-related differences in group-averaged VAF could be affected by between-subject variability of temporal activation patterns.

#### *2.6. Statistical Analyses*

Statistical analyses were performed with SPSS (IBM® SPSS v23 Statistics for Windows, Armonk, NY, USA). For both step directions, two-way repeated measures ANOVA (age × synergy number) for VAF was used to assess the interaction effect of age and number (*n*) synergies on VAF [39]. The number of synergies needed to adequately reconstruct the EMG signals was then determined using the iterative process described in the previous paragraph. Subsequently, two-way repeated measures ANOVA (age × step height) was used to assess the main and interaction effects of age and step height on synergy complexity, which was defined as VAF obtained with *n* synergies [29] fixed to four. Data were tested for normality with a Kolmogorov–Smirnov test. Sphericity was checked using Mauchly's test for sphericity. If a significant main effect was found, post-hoc tests comparing differences between age groups were performed using independent samples *t*-tests, while related-samples *t*-tests were used to compare differences per step height and synergy number. Alpha was set to 0.05 for all statistical tests.

Similarity of muscle synergies (based on muscle weightings, W) was quantified based on Pearson's correlation coefficients where *r* > 0.7 represented significant similarity and *r* > 0.45 represented marginal similarity [20,44]. Correlated synergies within age groups between step heights, and between age groups for each step height, were considered to be shared synergies, while non-correlated synergies were considered task-specific or age-related synergies [44]. Differences in muscle contributions to each synergy (W) between age groups were checked using Mann–Whitney U tests.

#### **3. Results**

Time-normalized kinematic data from the heel and pelvic markers showed high similarity (*r* > 0.9) in averaged vertical displacement over time between young and older women for all step heights. An example of the averaged vertical displacement patterns and standard deviations at 30 cm step height is provided in Figure 1.

Two-way ANOVA (age × synergy number) of VAF revealed significant main effects of synergy number for all step directions and heights (*p* < 0.001), but no interaction effects with age (*p* ≥ 0.05). A significant main effect of age (*p* = 0.028) was detected only for lateral stepping at 30 cm. For the group-averaged VAF, four muscle synergies were required to achieve a threshold level of 85% VAF for reconstructed signals across both age groups, step directions and step heights (Figure 2). This indicates that age, step direction and step height did not affect the number of synergies needed to reconstruct the EMG data. Consequently, the following results were obtained assuming *n* = 4 synergies. For forward stepping, four synergies accounted for 90.5%, 89.8% and 91.8% of variance in young women and 88.5%, 87.3% and 87.4% in older women for step heights of 10, 20 and 30 cm respectively. In lateral stepping, VAF by four synergies was 90.3%, 90.0% and 91.7% in young women and 88.2%, 88.0% and 87.4% in older women for step heights of 10, 20 and 30 cm respectively. Two-way ANOVA (age × step height) on VAF obtained from *n* = 4 synergies revealed a significant main effect of step height (*p* = 0.002) and age (*p* = 0.026) in forward direction, but not in lateral direction (*p* = 0.187 and *p* = 0.138 respectively). No significant age × step height interaction effect was found for either step direction (*p* > 0.05). For forward stepping, related-samples *t*-tests revealed a significant difference between step heights of 10 cm versus 20 and 30 cm (*p* = 0.009 and 0.014 respectively), but not between 20 and 30 cm (*p* > 0.05). Independent samples *t*-tests revealed a significant difference between age groups for each step height (*p* = 0.005, *p* = 0.041 and *p* = 0.019 for 10, 20 and 30 cm respectively).

**Figure 2.** Averaged variance accounted for (VAF) by number of extracted synergies at step heights of 10, 20 and 30 cm in forward (Fstep) and lateral (Lstep) stepping directions for young and older women. The appropriate number of synergies was defined as the least number of synergies required to reach >85% VAF (indicated by dashed lines) and a reconstruction quality of ≥75% VAF for each individual muscle.

Comparisons between muscle weightings (Table 1, Figures 3 and 4) showed that synergy 2 and 4 had high inter-step height similarity for both age groups and step directions. Synergy 4 also appeared to be highly similar between age groups. In synergy 2, a lower similarity between age groups was found, probably due to a difference in quadriceps/hamstring co-activation, as characterized by a significantly decreased contribution of the quadriceps and increased contribution of the hamstrings for most stepping conditions in older women. The composition of synergy 3 appeared to be the most variable. In contrast with forward stepping, which showed a robust synergy organization within and between age groups, lateral stepping resulted in lower correlations between step heights for the young group when compared to the older group.


**Table 1.** Similarity index (Pearson's *r*) of synergy weightings, across step heights for each age group, and across age groups for each step height. Results are displayed separately for forward stepping (Fstep) and lateral stepping (Lstep). Grey = marginal similarity (*r* > 0.45) and black = significant similarity (*r* > 0.7).

**Figure 4.** Muscle weightings and temporal activation patterns for four synergies (**S1**–**S4**) extracted during lateral stepping (Lstep). Each set of two columns represents muscle synergies extracted from step heights of 10, 20 or 30 cm per age group with young women represented in the left column and older women in the right column. \* indicates a significant difference between contributions of individual muscles between age groups. Time-averaged foot clearance (white bars) and pull-up (black bars) phases of ascent are depicted at the bottom. TA = tibialis anterior, GL = gastrocnemius lateralis, SOL = soleus, VL = vastus lateralis, RF = rectus femoris, BF = biceps femoris, ST = semitendinosus, GMAX = gluteus maximus, GMED = gluteus medius, ERS = erector Spinae.

Analyses of the temporal activation patterns (Figures 3 and 4) and the time-averaged standard deviation of these temporal activation patterns (Figure 5) showed that the between-subject variability of activation timing for most step heights and directions was significantly higher in the older cohort.

**Figure 5.** Between-subject variability (time-averaged standard deviation) of temporal activation patterns across time-normalized trials, expressed as a percent of 100 (in arbitrary units, *n* synergies = 4). Data are displayed separately for forward (Fstep) and lateral (Lstep) step directions with step heights of 10, 20 and 30 cm. \* indicates a significant difference at *p* = 0.05, \*\* indicates a significant difference at *p* = 0.01.

#### **4. Discussion**

The purpose of this study was to investigate the effects of age and task challenge on neuromuscular control and resultant complexity of neuromuscular activation patterns of women during step ascent by examining muscle synergy organization during stepping tasks with incremental step heights in both forward and lateral directions.

Our results show that complexity of motor control is quite robust across step heights and age groups for stepping in forward and lateral directions. We found no differences in the number of synergies between age groups and step heights for either age group. Further analyses of VAF by four synergies revealed main effects, but no interaction effect, of age and step height on complexity of motor control during forward stepping, indicating subtle differences that could not be detected by analyses of the number of extracted synergies. These analyses revealed that older women actually exhibited more complex motor control strategies, indicated by lower VAF [29], for each step height. Additionally, forward step heights of 20 and 30 cm yielded a significantly lower VAF at *n* = 4 synergies compared to 10 cm. Comparisons of muscle synergy organization revealed that muscle contributions to individual synergies (e.g., muscle weightings) during forward stepping were quite robust between step heights for both age groups. Finally, age-related differences in synergy organization were characterized by a notably lower similarity index between step heights for lateral stepping in young women and increased between-subject variability of the temporal activation patterns in older women.

The extraction of four synergies from step ascent is in agreement with previous studies of locomotion in healthy adults that included a maximum of ten lower limb muscles [24,40]. An important aspect of this study is the inclusion of EMG signals obtained from 10 lower limb muscles, including the gluteus medius, gluteus maximus and the erector spinae. Our results were in line with a previous study by Oliveira et al., who also found four synergies were sufficient to reconstruct the EMG signals of ten lower limb muscles. However, they also proposed that the addition of EMG measurements from hip extensors and abductors during gait would likely increase dimensionality [40]. Our data show that this is not necessarily true for step ascent. For example, during forward stepping, gluteus medius activation coincided with activation of the triple extensors (gluteus maximus, plantar flexors and rectus femoris). During lateral stepping, inclusion of the gluteus medius and erector spinae did not increase dimensionality in the form of an additional synergy as their activation coincided mainly with tibialis anterior activation during lift-off of the trailing foot and trunk stabilization prior to the double support phase after ascent. In line with previous studies including healthy older adults, our results indicate that age did not affect the number of synergies required to reconstruct the muscle activation, which implies that the complexity of motor control (or motor repertoire) of our healthy older cohort was not reduced [6,8,17,45]. However, we were surprised to find that the VAF at a fixed number of synergies, which can be used as an alternative way to assess complexity of motor control [29], was decreased in older women and with step height, indicating increased rather than decreased complexity. We propose that this may be due to the increased challenge posed by step ascent for older adults, forcing some to adopt different motor strategies to compensate for decreased physical capacity [15,46]. As a consequence, the organization or timing of motor modules may be altered, inevitably leading to higher between-subject variability in older adults compared young adults. The assumption that shifts in muscle synergy organization are attributable to (relative) increases in task challenge is in line with previous findings by Routson et al., showing that changes in speed, cadence, step length, and step height during gait can lead to altered temporal activation patterns [7]. Additionally, other studies have explored the interaction between age and task challenge during gait and found that walking at a higher than preferred cadence revealed small differences in spatio-temporal characteristics of neuromuscular control in older but not in young adults [8,9], although this proposed interaction effect was not confirmed by our results.

More detailed analyses of synergy organization in our older cohort revealed a trend towards increased contribution of the hamstrings and decreased contribution of the quadriceps in synergy 2, indicating an increase in quadriceps/hamstring co-activation during the initial foot clearance phase for both step directions compared to young women. These results are in line with findings from previous studies that have shown elevated muscle co-activation in older adults to increase joint stiffness and enhance stability during activities of daily life, such as stair climbing and single step descent [17,47–49], and directly affect muscle synergy organization [8,17]. Additionally, comparisons of synergy organization across step heights revealed that, in young women, increasing the step height from 20 to 30 cm was associated with an increased contribution of the gluteus medius and maximus to synergy 1 and of the calf muscles to synergy 3 during foot placement and the initial pull-up phase of ascent in lateral direction, while the composition of the remaining synergies remained similar. These changes indicate that some synergies reflect basic motor patterns which are activated during a variety of tasks, whereas other synergies can be flexibly recruited to match task-specific demands [20,44], such as the increased challenge to medio-lateral stability imposed by increased step heights. This is reflected by kinematic analyses of motor strategies for stair negotiation, showing both common and variable patterns, with the highest variability often seen at the hip joint [50]. The age-related differences in neuromotor strategies found in this study were reflected by subtle differences in kinematic profiles of the heel and sacrum. For example, visual inspection (Figure 1) revealed that, for older women, average vertical displacement of the sacrum was more linear in both directions with distinctly higher between-subject variability during the pull-up phase of forward stepping and higher between-subject variability of heel peak height during the foot clearance phase. However, these kinematic profiles were primarily included to define start and end points of each step cycle. As such, they provide limited information about how muscle synergy organization affects strategies such as total lower limb extension patterns and joint kinematics and kinetics and should be a focus of future studies [50].

Additional analyses were included to detect possible age-related differences of temporal activation patterns. Our results did reveal higher between-subject variability of temporal activation patterns, indicating increased heterogeneity of motor control strategies within the older cohort. This may reflect a relative increase in functional demand imposed by step heights of 20–30 cm in this age group, necessitating individual modulations of synergy timing to compensate for decreased physical capacity [7,12]. Higher between-subject variability of temporal activation patterns associated with increased task challenge in young women indicates that synergy organization is likely also associated with differences in motor skill levels [27]. This is illustrated by notable changes in between-subject variability of activation patterns in synergies 1 and 3 between step heights of 20 and 30 cm in lateral direction. Finally, although not the primary focus of this study, the differences in synergy organization between forward and lateral stepping are in line with findings from previous studies, showing that EMG recruitment patterns are task-specific for forward and lateral stepping [34,51].

Some limitations of this study have to be recognized. We chose to include only women in this study. As such, additional research is required to assess if these findings also apply to older males. EMG was only collected from the dominant leg. For this reason, differences in motor strategies involving additional push-off force of the trailing leg could not be analyzed [25]. Additionally, due to the technical limitations of surface EMG recording, no data were collected from the deeper thigh muscles such as the hip adductors. Future studies involving step ascent should consider including the hip adductors in order to increase dimensionality and provide information regarding the effects of antagonistic co-contraction of the hip ab- and adductors. Finally, a possible limitation lies with averaging EMG signals of individual participants over three repetitions of the same trial, rather than concatenating them prior to running the factorization algorithm. This may have led to a decrease of detail in the data [40].

#### **5. Conclusions**

Neuromuscular control of young and community-dwelling older women in stepping up in forward and lateral direction could not be differentiated based on the number of synergies. However, additional analyses of synergy complexity, such as VAF by the given number of synergies, and synergy structure revealed several age- and step-height related differences. These findings show that the ability to modulate synergy composition is well preserved in healthy older women and that they respond to more challenging physical tasks by adapting basic muscle recruitment patterns. This results in more complex motor control patterns despite evidence of increased antagonistic co-activation, likely indicating increased involvement of mechanisms for balance control.

**Author Contributions:** R.J.B., E.V.R., S.M.V., C.D., J.H.v.D. and N.D.: Conceptualization and experimental design; R.J.B. and G.G.: Participant recruitment and data collection; R.J.B., J.H.v.D. and N.D.: Data analysis and interpretation; R.J.B., E.V.R., S.M.V., J.H.v.D., C.D. and N.D.: Manuscript drafting and/or reviewing; R.J.B., J.H.v.D., E.V.R., S.M.V., G.G., C.D. and N.D.: Final approval of submitted manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was funded by the European Commission through MOVE-AGE, an Erasmus Mundus Joint Doctorate programme (#2014-0691) fellowship awarded to R.J.B., the Fund for Scientific Research Flanders (FWO-Vlaanderen, #G0521-05) awarded to S.M.V., and the ERC-Starting Grant "Learn2Walk" (#715945) and the NWO VIDI-Grant "FirSTeps" (#016.156.346) both awarded to N.D.

**Acknowledgments:** The authors would like to thank Tine De Reys for her assistance with data collection and processing and Hans Essers for his advice on the cluster analyses.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Adaptation in Gait to Body-Weight Unloading**

#### **Rakshatha Kabbaligere 1,2,\* and Charles S. Layne 1,2,3**


Received: 30 September 2019; Accepted: 20 October 2019; Published: 23 October 2019

**Abstract:** Modifications in load-related sensory input during unloaded walking can lead to recalibration of the body schema and result in aftereffects. The main objective of this study was to identify the adaptive changes in gait and body-weight perception produced by unloaded walking. Gait performance during treadmill walking was assessed in 12 young participants before and after 30 min of unloaded walking (38% body weight) by measuring lower limb kinematics, temporal gait measures, and electromyography (EMG). A customized weight-perception scale was used to assess perception of body weight. Participants perceived their body weight to be significantly heavier than normal after unloading while walking. Angular displacement about ankle and knee was significantly reduced immediately after unloaded walking, while temporal gait parameters remained unchanged. The EMG activity in some muscles was significantly reduced after unloading. These findings indicate that walking at reduced body weight results in alterations in segmental kinematics, neuromuscular activity, and perception of body weight, which are the aftereffects of motor adaptation to altered load-related afferent information produced by unloading. Understanding the adaptive responses of gait to unloading and the time course of the aftereffects will be useful for practitioners who use body-weight unloading for rehabilitation.

**Keywords:** motor adaptation; body-weight unloading; gait adaptation; treadmill walking; spaceflight; lower-body positive pressure

#### **1. Introduction**

The sensorimotor system is comprised of sensory systems, motor systems, and the central integration processes that help in producing and controlling movements. The vestibular, proprioceptive, and visual systems provide sensory information from the external environment as well as that related to body position and movement. The sensory inputs from these systems are integrated in the central nervous system (CNS) following which appropriate motor commands are generated and sent through the descending pathways to the various body segments. Any changes in the sensory information either due to changes in the environmental condition or body condition will affect movement control. The process that enables us to modify and maintain accurate movements as sensory condition changes is called motor adaptation [1].

Several studies have investigated motor adaptation using a variety of experimental paradigms. One of the ways is by studying the changes in the movement characteristics produced by adding and subsequently removing sensory (visual, proprioceptive, acoustic, and vestibular) distortions or perturbations while performing motor tasks [2–5]. One of the common locomotor adaptation paradigms involves studying locomotor changes by altering the normal interlimb relationship during walking by changing the speeds of treadmill belts relative to each other (commonly called split-belt treadmill walking [6–8]). Adaptive changes in step length and double support time are observed during split-belt

walk. Similarly, in another paradigm, the effect of increased trunk rotation during walking was studied by having the subjects walk along the circumference of a rotating disc for 2 h [9]. Subsequently, after the adaptation session, the subjects were found to produce curved walking trajectories. Likewise, adaptive changes in heading direction in response to modifications of the direction of optical flow was also observed after exposing subjects to a visual scene that gave the perception of walking along the perimeter of a room [10].

Body-weight unloading (BWU) using various types of body-weight support systems is used to study locomotor adaptation in response to reduced load input [11–14]. Load-related sensory information is essential for regulating the timing, phasing, and magnitude of neural activities that generate locomotor patterns during stepping [15–17]. They also help in maintaining balance control during locomotion and gait termination [18–20]. Lower-body positive pressure (LBPP) is an emerging technology that is used to provide body-weight support [21]. It consists of an air chamber that covers the lower part of the body. When inflated with positive pressure, the lower part of the body is lifted upwards from the hips and the body weight is reduced. LBPP is regarded as one of the superior methods of unweighting when compared to upper-body harness [21]. Upper-body harness partially supports the body weight and results in the formation of pressure points. LBPP on the other hand results in uniform distribution of air pressure around the entire surface of the body while still maintaining normal muscle activation and is thus considered superior [22,23]. Body weight as large as 80% can be unloaded in increments as small as 1% using this system. An antigravity treadmill is a special type of treadmill that is equipped with an LBPP air chamber and provides an opportunity to study locomotor adaptation to BWU. Several studies using either the antigravity treadmill or the vertical harness system have investigated the immediate effects of BWU on metabolic energy expenditure and locomotor performance [12,13,24–26]. A linear decrease in stride frequency, vertical contact forces, stance time, peak hip and knee flexion, and extensor and flexor muscles' activity burst during the stance phase with BWU has been reported [24,26–28].

Besides having acute effects on gait performance, exposure to BWU for a short period of time in the order of a few minutes and subsequent reloading can alter movement characteristics [29,30]. Further, prolonged exposure to BWU for a longer period of time in the order of a few months can also produce significant changes in movement characteristics that last for a significant period of time as seen in crewmembers returning from space [20,31–34]. Due to the fact that adaptive changes produced during spaceflight is coupled with changes in structural and functional characteristics of the neuromuscular system, we cannot compare it with short-term changes that we observe on the ground that are solely produced by reduced body load. However, studying the adaptive changes to short-term BWU can improve our understanding of the mechanism adopted by the sensorimotor system during adaptation. Also, with the increase in use of BWU for treadmill training in patients with neurological impairments, studying the adaptive changes to short-term BWU may provide valuable information to practitioners who administer gait or balance rehabilitation using it [21,35,36].

The main objective of this study was to investigate changes in locomotion produced by 30 min of walking at a reduced body weight, which roughly relates to that of Martian gravity (38% of normal body weight). Locomotion was assessed by measuring temporal gait parameters, changes in joint angles, and neuromuscular activation of the lower limbs before, during, and after the adaptation session. We also assessed the adaptive changes in subjective body-weight perception as a function of movement context (walking, standing, and sitting). Body-weight perception was assessed using a customized weight scale. It was hypothesized that the effect of unloaded walking on weight perception would vary as a function of movement context.

#### **2. Materials and Methods**

#### *2.1. Participants*

Twelve young and healthy participants (23.7 ± 3.3 years; 6 males and 6 females) participated in the study. The experimental protocol was approved by the institutional review board at the University of Houston and was conducted in accordance with the Declaration of Helsinki. The participants were screened using a modified physical activity readiness questionnaire (PAR-Q) [37]. This questionnaire was used to assess if the participants were physically fit and for any conditions that might affect gait and balance. The participants were excluded if their response was "yes" to any of the questions in the modified PAR-Q.

All the participants attended three data collection sessions, which were conducted on three separate days. Their gait performance was assessed during one of the three sessions before, during, and after the adaptation sessions. On the other two days, participants performed two different tests before and after the adaptation sessions while being seated and while standing (Figure 1). These tests were part of another study. Weight perception was assessed on all the three days before, during, and after the two adaptation sessions. During the adaptation sessions, the participants walked continuously for 30 min while being loaded (100% body-weight/control condition) or unloaded (38% body-weight/unloading condition). Post-adaptation sessions consisted of either continuing to walk, sitting, or standing without any movements. The order of the three sessions as well as the order of the two adaptation sessions (control and unloading condition) within each session were randomized across the participants (Figure 1). An overview of the experimental design is depicted in the flow chart below.

**Figure 1.** Flow chart describing the experimental design of the study.

#### *2.2. Body-Weight Unloading*

Participants were unloaded using an LBPP antigravity treadmill to unload the body weight. The participants donned a pair of neoprene shorts that were used to zip them into the antigravity treadmill. Once inside the chamber, the inside pressure was calibrated and set to a level that corresponded either to 100% body weight or 38% body weight.

#### *2.3. Gait Assessment*

After a familiarization period, baseline gait performance (preadaptation) was measured for a duration of two minutes prior to each adaptation session as the participants walked inside the chamber at normal (100%) body weight at a speed of 1 m/s (Figure 2). Following the baseline assessment, they performed the adaptation session for a period of 30 min. In the case of the control adaptation

session, they continued to walk at 100% body weight while, during the unloading adaptation session, they were unloaded to 38% body weight. After the adaptation session, the participants performed the post-adaptation session during which their gait was once again assessed at normal body weight (Figure 1). After this, they were given a mandatory break of at least 10 min before starting the other session.

Participants' gait performance was assessed by measuring kinematics, electromyography (EMG), and temporal parameters before, during, and after the two adaptation sessions. Lower limb kinematics along the sagittal plane were measured by using goniometers (Biometrics©) attached unilaterally on the right ankle and knee joint. Joint kinematics were restricted to ankle and knee joints since the chamber frame obscures the pelvis and hip making recording of the hip motion prohibitive. Kinematic data were collected at a sampling rate of 1000 Hz. EMG was recorded using bipolar surface electrodes (Delsys© Trigno TM wireless EMG system) at a rate of 2000 Hz. EMG was recorded from four muscles viz tibialis anterior (TA), medial head of gastrocnemius (GA), rectus femoris (RF), and biceps femoris (BF) muscles of the right leg. A pair of foot switches (Biometrics©) were attached to the sole of the right foot, one under the great toe and the other under the heel. The data from the foot switches were recorded at a rate of 1000 Hz and were used to determine foot-fall events in the gait cycle.

**Figure 2.** Figure illustrating the setup of lower-body positive pressure (LBPP) or antigravity treadmill.

#### *2.4. Weight Perception*

Weight perception was assessed during all three sessions, both during and after the adaptation sessions. We used a 7-point scale, with 1 being very light and 7 being very heavy. The participants were instructed to mentally assign the weight they perceived prior to the adaptation session as 4 on the 7-point scale that translated to "normal" weight. They were instructed to report their perception of weight relative to this during the adaptation session and at three different time intervals after the adaptation session, which were immediately after (T+0), 5 min (T+5) after, and 10 min (T+10) after the session.

#### *2.5. Data Analysis*

Data analysis for temporal gait measures, kinematics, and EMG was primarily focused on 7 epochs for each of the control and unloading conditions (Figure 3). These were:


**Figure 3.** Flow chart indicating the epochs of interest for data analysis.

#### *2.6. Temporal Gait Measures*

Temporal gait measures were calculated using the gyroscope data obtained from the Trigno EMG sensor attached to the anterior part of the lower right leg (TA muscle). We used a gait feature extraction algorithm that was used previously in an earlier study [38] to identify three main gait-events: (1) initial contact (IC), (2) mid swing (MS), and (3) terminal contact (TC). These gait events were then used to compute stride time, stance time, and swing time. Stride time was defined as the duration between successive IC points. Stance time was defined as the time between IC and TC of each gait cycle. Swing time was defined as the time between TC of one gait cycle to the IC of the next gait cycle. The temporal gait parameters were then averaged within each epoch of interest.

#### *2.7. Kinematic Measures*

Joint angular data recorded from the goniometers were low-pass filtered with a cutoff frequency of 10 Hz [39]. Following this, they were divided into individual strides based on heel-strike events obtained from the foot switches and reduced to 100 data points. Then, the joint-angle waveforms were averaged across 30 strides within each epoch of interest. To facilitate comparison across subjects, the joint-angle waveforms of each of the epochs were normalized to the peak joint angle of the T0 waveform by dividing all the points of the waveform by that peak angle value. Range of motion (ROM) of the joints was then computed as an outcome measure at each epoch as the difference between the maximum and minimum angular displacements of the normalized waveforms.

#### *2.8. EMG Measures*

The EMG signals were first band-pass filtered using a second-order Butterworth filter using cutoff frequencies of 10 Hz and 450 Hz, followed by full wave rectification [29]. The rectified signals were then low-pass filtered at a cutoff frequency of 25 Hz to obtain the linear envelope [40]. Following this, the signals were separated into individual strides based on heel-strike events and reduced to 100 data

points per stride. Then, the EMG waveforms were averaged across 30 strides within each epoch of interest [41]. Waveforms of all the epochs were normalized to the mean EMG activity of the T0 waveform by dividing all the points of the waveform by that mean value [42]. After this, within each epoch, the EMG activity across 7 different gait phases was calculated individually. This was calculated by taking the average of the points that represent the respective gait phases. For example, the average of points 1 through 12 would represent the mean EMG activity of the loading phase. A breakdown of the percentage of the time represented by each of the 7 gait phases within a gait cycle is described below [43].


#### *2.9. Statistical Analyses*

All statistical analyses were performed using SPSS V.20 (IBM Corp, Somers, NY, USA). Two-way repeated measures analysis of variance (RANOVA) was performed on each of the temporal (stride time, stance time, and swing time), kinematic (range of motion of knee and ankle joint), and EMG (mean EMG activity) outcome measures separately. The EMG activity of each of the 7 phases of the gait cycle was analyzed separately. The two factors in these analyses were "condition" (control vs. unloading) and epochs (T0, Tduring, and T1–T5). Simple planned contrasts were used to assess changes in the outcome measures across epochs relative to T0 whenever there was a main effect of epoch. Additionally, a simple effects analysis was performed on each of the conditions whenever there was a significant interaction effect. Before performing the ANOVA analysis, data were analyzed to assess whether all the required assumptions were met. Whenever the assumption of sphericity was not fulfilled, the degrees-of-freedom was adjusted using Huyn–Feldt correction.

A nonparametric test (Friedman's ANOVA) was conducted to test for changes in body-weight perception separately for each of the three sessions. This was followed by a Wilcox signed ranked test to compare differences between different time intervals.

#### **3. Results**

#### *3.1. Body-Weight Perception*

Perception of body weight remained unchanged during and after the control adaptation session regardless of the movement context. During the unloading condition, there was a significant decrease in perceived body weight during unloaded walking across all the three sessions (Z < −2.71, *p* < 0.007). There was a significant increase in perceived body weight after unloading at T+<sup>0</sup> (Z = −3.06, *p* < 0.05) and T+<sup>5</sup> (Z = −2.81, *p* < 0.05) during the walking session. Although there was a trend towards increased body-weight perception while seated and standing after unloading, the changes in the scores were not significant (Figure 4).

**Figure 4.** Group median along with the 25th and 75th percentile scores of weight perception during sitting, standing, and walking across different time intervals for the unloading condition. \* Significant difference (*p* < 0.05) relative to T0 of the corresponding condition.

#### *3.2. Temporal Gait Measures*

There was a significant main effect of time (F (3.12,37.49) = 4.56, *p* = 0.007) on swing time and a significant interaction effect between condition and time (F (2.75,33.00) = 4.04, *p* = 0.017). Results from the simple effects analysis indicated that the swing time was significantly increased at Tduring (0.54 ± 0.01) relative to T0 (0.52 ± 0.01) during the unloading condition and that it returned to baseline immediately after the adaptation session. However, swing time remained unchanged across time during the control condition. Stride time and stance time remained unchanged across time for both the control and unloading conditions. The average stance and swing times across different epochs for the two conditions are summarized in Figure 5.

**Figure 5.** Group means (±SEM) of (**A**) swing time and (**B**) stance time across different epochs for the control and unloading conditions. \* Significant difference (*p* < 0.05) relative to T0 of the corresponding condition.

#### *3.3. Kinematics*

Overall, the ROMs of ankle and knee joints during the unloading condition were significantly less when compared to the control condition (main effect of condition, ankle: F (1, 10) = 15.29, *p* = 0.003; knee: F (1, 10) = 8.47, *p* = 0.016). There were no changes in ankle and knee ROMs across time in the control condition. For the unloading condition, the ROM for both the joints was significantly reduced at Tduring (ankle: 2.44 ± 0.20; knee: 1.29 ± 0.04) and T1 (ankle: 2.93 ± 0.17; knee: 1.46 ± 0.05) relative to T0 (ankle: 3.31 ± 0.22; knee: 1.58 ± 0.03). Knee ROM was also reduced at T2 (1.47 ± 0.05). However, it was not different from T0, starting fromT3 through T5 (Figure 6).

**Figure 6.** Group means (±SEM) of normalized range of motion of (**A**) ankle and (**B**) knee joint across different epochs during control and unloading conditions. \* Significant difference (*p* < 0.05) relative to T0 of the corresponding condition.

#### *3.4. EMG*

#### 3.4.1. Rectus Femoris Activity

EMG activity during all seven gait phases remained unchanged after walking in the control condition. For the unloading condition, there was a significant effect of time on EMG activity during loading phase (F (3.07, 30.70) = 7.62, *p* = 0.001), mid-stance (F (3.24, 32.36) = 5.051, *p* = 0.005), pre-swing (F (3.15, 31.53) = 2.94, *p* = 0.046), mid-swing (F (2.94, 29.41) = 4.557, *p* = 0.01), and terminal swing (F (5.38, 53.85) = 8.90, *p* < 0.0001). There was reduced activity during unloaded walking (1.08 ± 0.11) relative to T0 (1.45 ± 0.15) in the loading phase. Immediately after unloading (T1), the activity increased (1.80 ± 0.18), following which the activity returned to baseline at T2 (1.65 ± 1.7) and remained unchanged through the subsequent epochs (T2–T5). EMG activity during the mid-stance phase was significantly reduced during unloaded walking (0.66 ± 0.07) relative to T0 (0.96 ± 0.10) and returned to baseline at T1 (0.97 ± 0.11). However, it was subsequently reduced at epochs T2 (0.83 ± 0.08) and T3 (0.86 ± 0.09), following which it returned to baseline at T4 (0.84 ± 0.1) and remained unchanged at T5 (0.83 ± 0.09). During pre-swing, there was no significant change in activity during unloading and immediately after unloading at T1, even though there was a major trend towards reduced activity. However, the activity was significantly reduced starting from T2 (0.8 ± 0.07) through subsequent epochs (T3: 0.81 ± 0.08, T4: 0.80 ± 0.07, and T5: 0.80 ± 0.06) relative to T0 (1.08 ± 0.11). The EMG activity during mid-swing was significantly reduced during unloaded walking (0.61 ± 0.06) relative to T0 (0.80 ± 0.05) and returned to baseline immediately after unloading at T1 and was not different from that of T0. Subsequently, it was reduced from T2 through T5 (T2: 0.66 ± 0.05, T3: 0.69 ± 0.06, T4: 0.66 ± 0.07, and T5: 0.68 ± 0.07). Similar to the mid-swing phase, EMG activity was also reduced in the terminal swing phase, during unloaded walking (0.85 ± 0.09) relative to T0 (1.25 ± 0.10). However, it was not different from that of T0 after unloading from T1 through T5. These changes are summarized in Figure 7.

**Figure 7.** Group means (±SEM) of EMG activity of the rectus femoris (RF) muscle across different epochs during the (**A**) loading, (**B**) mid-stance, (**C**) pre-swing, (**D**) mid-swing, and (**E**) terminal swing phases for control and unloading conditions. \* Significant difference (*p* < 0.05) relative to T0 of the corresponding condition.

#### 3.4.2. Biceps Femoris Activity

There were no changes in BF activity after the adaptation session during any of the gait phases in the control condition. However, the BF activity changed in the unloading condition during the terminal stance and mid-swing phases as a function of time. During the terminal stance phase, EMG activity remained unchanged during and immediately after unloaded walking at T1 and T2. However, it was reduced relative to T0 (0.58 ± 0.07) starting from T3 (T3: 0.43 ± 0.07; T4: 0.43 ± 0.07) through T5 (0.41 ± 0.05). BF activity during the mid-swing phase was reduced during unloaded walking (0.83 ± 0.13) relative to T0 (1.37 ± 0.17) and was not different from that of T0 from T1 through T5 (Figure 8).

**Figure 8.** Group means (±SEM) of biceps femoris (BF) muscle activity during the (**A**) terminal stance and (**B**) mid-swing phases of the gait cycle during control and unloading conditions across different epochs. \* Significant difference (*p* < 0.05) relative to T0 of the corresponding condition.

#### 3.4.3. Gastrocnemius Activity

EMG activity of GA remained unchanged relative to baseline during each of the seven phases in the control condition. There was a significant effect of time for the unloading condition during the mid-stance (F (2.14, 21.43) = 8.46, *p* = 0.002) and terminal stance phases (F (5.33,53.28) = 32.902, *p* < 0.001). The EMG activity during both phases was significantly reduced during unloaded walking (mid-stance: 0.77 ± 0.15, terminal stance: 0.81 ± 0.13). However, it was not different relative to T0 (mid-stance: 1.58 ± 0.22, terminal stance: 1.93 ± 0.20) from T1 through T5 (Figure 9).

**Figure 9.** Group means (±SEM) of the medial head of gastrocnemius (GA) muscle activity during (**A**) mid-stance and (**B**) terminal stance during control and unloading conditions across different epochs. \* Significant difference (*p* < 0.05) relative to T0 of the corresponding condition.

#### 3.4.4. Tibialis Anterior Activity

As in the case with other muscles, TA activity remained unchanged after the adaptation session in the control condition. There were changes in the muscle activity in the unloading condition only during the mid-swing phase. Specifically, the activity significantly increased immediately after unloaded walking relative to T0 (1.00 ± 0.07) from T1 (T1: 1.19 ± 0.07, T2: 1.20 ± 0.08, T3: 1.18 ± 01, and T4: 1.19 ± 0.1) through T5 (1.24 ± 0.1) (Figure 10). Table 1 provides a summary of changes in EMG activity during and after unloaded walking across different epochs and gait phases.

**Figure 10.** Group means (±SEM) of tibialis anterior (TA) muscle activity during the mid-swing phase of the gait cycle during control and unloading conditions across different epochs. \* Significant difference (*p* < 0.05) relative to T0 of the corresponding condition.

**Table 1.** A summary of changes in EMG activity of all four muscles across different epochs and gait phases: The upward arrow indicates a significant (*p* < 0.05) increase, and the downward arrow indicates a significant (*p* < 0.05) decrease in activity relative to T0, while "-" indicates no significant changes relative to T0.


#### **4. Discussion**

The main objective of this study was to investigate the changes in temporal, kinematics, and neuromuscular activity during locomotion produced by 30 min of walking at 38% bodyweight and the second was to test the adaptive changes in body-weight perception in response to 30 min of walking at reduced body weight as a function of movement context (walking, standing, and sitting). The results indicate that 30 min of unloaded walking modifies body-weight perception as a function of movement context. Further, it also modifies gait performance characterized by alterations in the movement of ankle and knee joints and neuromuscular activation patterns, some of which persist up to 10 min after the adaptation period. These results are individually discussed in detail in this section.

#### *4.1. Body-Weight Perception*

As expected, participants felt lighter during unloaded walking in all three test sessions. Their perception of body weight after unloaded walking varied as a function of the movement context. It significantly increased after unloading only during the gait session, while it returned to baseline immediately after unloading during the sessions when the participants were seated or stood statically. The increase in perceived body weight observed during the gait session aligns with the finding of increased rate of perceived exertion found after 3 min of unloaded running in a similar study [29]. The increase in perceived body weight only during the gait session could be due to the presence of active movements during walking as opposed to during sitting and standing.

In the literature related to sensory psychology, perception of active touch or touching has been shown to be different from passive touch or the act of being touched [44,45]. Active touch involves a combination of active movement and the sensory feedback that results from touching. Thus, the resulting stimulus during active touch contains two components: exterospecific and propriospecific information. Passive touch, on the other hand, involves only the sensory feedback of the stimulus that is applied on the skin. For example, when we use our hand to touch an object, the movement about all the joints present in the hand and the cutaneous inputs arising from contact are continuously integrated while shaping the percept of the object. However, when being touched by an external stimulus, only the sensory inputs from the cutaneous receptors in the skin and its underlying tissue are available. In other words, active touching involves both objective and subjective sensory information, thus making it comparatively a more enriching experience.

Extending the above theory to the context of the current study, the contexts of sitting and standing can be associated to a "passive experience" of the sensory environment. Conversely, walking can be associated with an "active experience" where the participants had the opportunity to "actively explore" the new sensory environment. Thus, it is possible that the perception of body weight during the active experience was modified as a result of changes in the relationship between the sensory and motor elements that were used in forming the percept of body weight. On the other hand, since only the sensory element related to body-load perception was used in forming the percept of body weight during passive experiences such as during sitting and standing, there was no change in perception of body weight. In summary, the findings of the current study support the notion that different movement contexts can have different sensory consequences associated with them.

#### *4.2. Temporal Gait Measures*

The results indicate that there were no significant changes in any of the temporal gait measures during and after unloaded walking, except for swing time, which was significantly increased during unloaded walking. The increase in swing time during unloaded walking corroborates the findings from previous studies [11,18,24,46,47]. This observation of increased swing time with body-weight unloading is in line with the predictions of the ballistic pendulum model of walking, which states that the motion of the swing leg is like that of a pendulum of which the oscillation period is inversely related to gravity. Additionally, the setup of LBPP by itself might also have contributed to this increase in swing time. Specifically, since the lower body is supported in this type of setup, there is less of a threat to postural instability, due to which one can afford to spend a longer time in the swing phase.

The lack of significant changes in temporal parameters following adaptation to unloading could be due to the constraints imposed on the walking performance by the speed and the dynamics of the treadmill and particularly due to the lack of stride to stride variability inherently associated with treadmill walking [48]. Thus, potentially testing the aftereffects of adaptation to unloading during over-ground walking as opposed to treadmill walking could provide additional insights into the adaptive process.

Although not statistically significant, there was a trend towards increased stance time accompanied by a decrease in swing time immediately after unloaded walking (at T1) similar to that reported in another similar study [30]. These changes in stance and swing time after unloading have been proposed as a control strategy adopted by the motor control system to maintain prolonged foot contact with the ground and, hence, to maintain stable balance by increasing the stance time while also decreasing the time spent during the unstable swing phase to prevent the risk of falls following adaptation to unloaded walking. However, since the lower body was securely attached to the antigravity chamber in the current study, there was no risk of falling. Thus, similar trends in the adaptive changes of stance and swing times between the two studies indicate that these changes might have been a direct result of modulation of somatosensory information resulting from changes in body-load sensing mechanisms and not that due to postural instability.

#### *4.3. Kinematics*

The range of ankle and knee joint angular displacement was significantly reduced during unloaded walking as in other locomotor studies that were focused on assessing immediate online changes in kinematics in response to body-weight unloading [24,46,47,49]. There was reduced angular motion of ankle and knee joints in the form of aftereffects following the adaptation session (post-unloading) for up to three minutes. These modifications in kinematics similar to those observed during the adaptation phase were found to slowly decay and return to baseline levels by the end of the testing session. There were no changes in ankle or knee angular movement during or after the control adaptation session. This confirms that the kinematic changes observed after unloading are not caused by muscular fatigue.

Ruttley [40] found a significant increase in total ankle and knee angular movement from heel strike to peak knee motion during post-adaptation to unloaded walking. The disparity with the current results could be related to the differences in the nature of the unloading system. Unloading systems using a vertical harness produce inertial forces as a result of the systems mass. Thus, excessive joint excursions are produced to overcome these inertial forces while walking, which could have resulted in an aftereffect manifested as increases in knee and ankle joint excursion after unloaded walking in Reference [40]. Since unloaded walking inside the antigravity treadmill is not influenced by such mechanical constraints, the modifications observed in lower-limb movements in the current study might represent the pure aftereffects of adaptation to unloaded walking. Although the body-weight unloading techniques and the outcome measures used to quantify joint motion in the two studies were different, the fact that there was a distinct recovery curve in both studies indicates that there was an adaptive change in kinematics that was produced by unloaded walking.

#### *4.4. EMG*

#### 4.4.1. During Adaptation

In line with the findings from other locomotor studies related to body-weight unloading, we found a significant reduction in the EMG activity of the extensor muscle GA during unloaded walking. This change was observed specifically during the mid-stance and terminal stance phases, which (during loaded walking) are the periods of peak muscle activity. GA enables controlled plantar flexion of the ankle joint in order to shift the center of gravity towards the front and allows lifting of the heel from the ground. The need for forward propulsion during push-off is expected to be reduced with reduced body load. A significant decrease in GA activity combined with reduced ankle-joint motion observed during unloaded walking in the current study supports this argument.

Additionally, we found a decrease in RF activity during mid-stance and later swing phases during unloaded walking. RF does not have a dominant role during mid-stance; however, it is known to play a predominant role during the transitory phases from stance to swing phases and vice versa. During the transition from swing to stance phases, it begins to prepare in a feedforward manner for the large ground reaction forces that the limbs will encounter upon heel strike during the next stance phase. Since the magnitude of the ground reaction forces decreases with unloading, the muscular effort required to counteract these forces at reduced body load is also less. This might be the reason

why there was a reduced activity in RF during the later phases of the swing phase during unloaded walking. During the transition from the stance to swing phases, i.e., during the pre-swing phase, it acts as a hip flexor and helps in lifting and propelling the limb into swing. One would expect that the muscular effort required to propel the limb is reduced during unloaded walking. In line with this hypothesis, there was a trend towards reduced RF activity during the pre-swing phase as well, which approached significance (*p* = 0.065).

#### 4.4.2. Post-Adaptation

Neuromuscular changes after unloading were most evident in RF and TA muscles. Additionally, these changes occurred close to two discrete events in the gait cycle which are characterized by large amounts of energy transfers, namely the heel strike and toe off. These observations are similar to those reported in spaceflight-related locomotor studies pointing towards some potentially common adaptive mechanisms [20,33,42]. As in the case with kinematics, there were no significant changes in neuromuscular activity of any of the muscles either during or after the control adaptation session. This further reiterates that the observed changes in neuromuscular activities after unloading are not caused due to fatigue but rather due to adaptation to unloaded walking. Although statistically insignificant, some visually evident changes in activity of some muscles were observed during certain phases of the gait cycle in the control condition. We speculate that these changes are mere random fluctuations as there are no specific methodological or physiological reasons as to why such changes might occur.

Around heel strike, we saw an increase in RF activity during the loading phase immediately after unloaded walking, combined with a reduction in knee and ankle angular excursion. This could be in response to the large ground reaction forces that the body encountered during the loading phase relative to that encountered during unloaded walking. Layne et al. [20] also found an increase in RF activity during the stance phase of the gait cycle after long-duration spaceflight. It has been suggested that modifications in RF activity combined with increased kinematic variability after spaceflight are attempts to attenuate the energy generated by the ground reaction forces around heel strike that are transmitted to the head [42,48,50,51]. Apart from heel-strike-specific modulations, we also observed a small reduction in RF and BF activity during the mid-stance and terminal stance phases of the gait cycle during some of the epochs. Although RF and BF have limited functional roles during these phases, these changes reflect an overall reinterpretation of sensory inputs as a result of adaptation to unloading.

Around toe off, there was a significant reduction in RF activity during the pre-swing phase after the first 100 strides (T2) post-unloading adaptation session. As mentioned earlier, the RF activity during the pre-swing phase is responsible for lifting the leg so that it can be propelled forward during the swing phase. During unloading, as expected, we saw a reduction in activity during this phase as the muscular effort required to propel the leg is less. Extending this logic, we would expect the muscular effort post-unloading to increase due to the increase in body-weight load. As expected, the RF activity increased at T1 relative to Tduring. This increase in RF activity was however transient, as it was significantly reduced from T2 up to 10 min. A reduction in RF activity was also combined with reduced angular excursion of ankle and knee joints during some epochs.

With regards to TA, we observed an increase in activity during the mid-swing phase. This must have been an attempt by the motor control system to compensate for reduced knee flexion by increasing its activity to increase the dorsiflexion about the ankle joint for adequate clearance of the toe. Since we did not observe any foot-scraping events in any of the subjects, it seems that the ankle dorsiflexion during the mid-swing phase was sufficient for toe clearance.

The above patterns of reduced RF activity during the pre-swing phase and increased TA activity during the mid-swing phase of the gait cycle were also observed after long-duration spaceflight during treadmill walking [20]. The fact that a short exposure of 30 min of body-weight unloading on Earth resulted in similar patterns of neuromuscular changes as those observed after long-duration spaceflight

is interesting. This finding suggests that there might be some common adaptive mechanisms regardless of the duration of exposure to unloading. In general, modifications in neuromuscular activations after spaceflight have been suggested to be driven by alterations in the neural drive to the motor neurons. There is compelling evidence in the spaceflight literature that shows alterations in postural muscle activity caused by exposure to weightlessness [52–55]. Muscle disuse as well as muscle loss are also known to cause changes in neural drive after spaceflight. Furthermore, modifications in proprioceptive functioning has also been suggested to contribute to changes in neuromuscular activation.

#### *4.5. Clinical Implications*

Gait-training programs with BWU are designed to provide support to the patient's body weight to allow for the reestablishment of damaged sensorimotor pathways or emergence of new ones to restore normal walking patterns [22]. Gait alterations observed during and after unloaded walking in the current study suggests that lowering body weight to as low as 38% during gait training is not recommendable if we want to reduce the risk of altering normal gait characteristics. Practitioners should be vigilant about choosing the right combination of body-weight level and walking or running speed for training purposes. Particularly, they have to choose a level that does not alter normal kinematic patterns and neuromuscular activities to a large extent.

#### **5. Conclusions**

Alterations in kinematics and neuromuscular activities observed during unloaded walking are a result of the adaptation of the neuromuscular system to the reduction in ground reaction forces, shear forces, foot sole pressure, and joint loads associated with unloading. In particular, these changes are caused due to somatosensory-mediated central changes in the body schema produced by new relationships between sensory and motor elements that are characteristic to an unloaded environment. The continued alterations in kinematics and neuromuscular activities observed after unloading are aftereffects of the adapted state of the body schema. The recovery of kinematics and neuromuscular activity over the course of the post-adaptation phase are indicative of the recalibration process that the body schema undergoes in order to restore the original sensory motor relationships. Similarities in the pattern of changes in neuromuscular activation amplitudes between spaceflight and the current study indicate that there might be some common adaptive mechanisms that are mediated by load-related somatosensory changes. Additionally, these alterations in kinematics and neuromuscular characteristics caution practitioners to choose the optimal level of body-weight unloading for gait training.

#### **6. Future Direction**

Two important methodological limitations of this study are worth noting. Firstly, the choice of the level of body-weight unloading was limited to 38% body weight. This prevented us from capturing the adaptive effects across different levels of unloading. Future studies should aim to capture the dose-response relationship between level of unloading and the magnitude of adaptive changes in movement characteristics. This will allow us to determine the optimal level of unloading that can be useful for improving gait in patients while producing the least amount of alterations in the movement and neuromuscular activation patterns. Secondly, assessing gait performance during treadmill walking as opposed to during over-ground walking must have limited us from capturing true adaptive effects of unloading with regards to temporal gait measures.

Additionally, it will be worth exploring the effects of passive unloading during standing inside the antigravity chamber for extended periods of time. Comparing the effects of passive unloading and unloaded walking (active loading) as in the current study will help isolate the effects of inactivity from that of unloading.

**Author Contributions:**R.K. and C.S.L. participated in the conceptualization, methodology, validation, formal analysis, and writing –review and editing. R.K. participated in the software, investigation, data curation, writing-original draft preparation and visualization. C.S.L. participated in resources, supervision and project administration.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors thank Mai Lee and David Young for their assistance in running experiments, Beom-Chan Lee for providing us access to the EMG recording system, and Ajitkumar Mulavara for providing valuable inputs during the conception of the study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Postural Control in Children with Cerebellar Ataxia**

**Veronica Farinelli 1, Chiara Palmisano 2,3, Silvia Maria Marchese 1,\*, Camilla Mirella Maria Strano 1, Stefano D'Arrigo 4, Chiara Pantaleoni 4, Anna Ardissone 5, Nardo Nardocci 5, Roberto Esposti <sup>1</sup> and Paolo Cavallari <sup>1</sup>**


Received: 5 December 2019; Accepted: 20 February 2020; Published: 28 February 2020

**Abstract:** Controlling posture, i.e., governing the ensemble of involuntary muscular activities that manage body equilibrium, represents a demanding function in which the cerebellum plays a key role. Postural activities are particularly important during gait initiation when passing from quiet standing to locomotion. Indeed, several studies used such motor task for evaluating pathological conditions, including cerebellar disorders. The linkage between cerebellum maturation and the development of postural control has received less attention. Therefore, we evaluated postural control during quiet standing and gait initiation in children affected by a slow progressive generalized cerebellar atrophy (SlowP) or non-progressive vermian hypoplasia (Joubert syndrome, NonP), compared to that of healthy children (H). Despite the similar clinical evaluation of motor impairments in NonP and SlowP, only SlowP showed a less stable quiet standing and a shorter and slower first step than H. Moreover, a descriptive analysis of lower limb and back muscle activities suggested a more severe timing disruption in SlowP. Such differences might stem from the extent of cerebellar damage. However, literature reports that during childhood, neural plasticity of intact brain areas could compensate for cerebellar agenesis. We thus proposed that the difference might stem from disease progression, which contrasts the consolidation of compensatory strategies.

**Keywords:** children; gait initiation; postural control; generalized cerebellar atrophy; cerebellar vermis hypoplasia; progressive ataxia; compensatory strategies

#### **1. Introduction**

Postural adjustments are involuntary muscular activities that accompany the voluntary movement. These activities spread over adjacent muscles and thus create "chains" that reach the available support points (in many cases, the ground). Such chains allow fine-tuning the body equilibrium, in order to adapt it to the mechanical needs of the ensuing movement. For example, when flexing both arms at the shoulder, postural actions develop in a dorsal muscle chain, including Erector Spinae (ES), Biceps Femoris (BF), and Soleus (SOL), to counteract the reaction force due to arm movement [1]. Instead, when rising on tiptoes, involuntary bursts of activity develop in Tibialis Anterior (TA) muscles, so as

to induce a forward fall of the Centre of Mass (COM); in this way, COM reaches the forefoot, and the voluntary contraction of Soleus muscles (SOL) rises the body [2]. Otherwise, simply recruiting SOL would produce a backward fall. Whenever the mechanical needs of action may be estimated beforehand, like when programming a voluntary movement, appropriate postural actions are usually produced in advance of the movement itself, witnessing that such Anticipatory Postural Adjustments (APAs) are programmed in a feed-forward way [3–6].

APAs are particularly evident in gait initiation, in which they maintain the body's dynamic balance and create the propulsive forces to move the COM forwards. In healthy adults performing gait initiation [7–12], the Center of Pressure (CoP), i.e., the barycenter of the ground reaction forces, first moves backward and towards the future swing foot. The onset of such a CoP shift is usually called APA onset, while its time period is called the *imbalance phase*. Indeed, the ensuing horizontal gap between CoP and COM (where the gravity force vector is applied) produces an "imbalance" torque that pushes COM forwards and towards the future stance foot. Then, CoP moves laterally towards the stance foot, continuing to promote the forward acceleration of COM while braking its lateral fall. At the same time, this CoP shift withdraws body weight from the swing foot, hence the name *unloading phase*. Finally, as COM proceeds, CoP travels forwards along the stance foot, from toe-off to heel-strike of the swing foot (*first swing*) [13]. Considering the muscular actions that drive gait initiation, before the beginning of the imbalance phase, an inhibition occurs in the background EMG (electromyographic) activity of both SOL muscles, which are tonically active during quiet stance. SOL inhibitions are shortly followed by the recruitment of TA muscles, which are silent during quiet stance and activate close to the APA onset. In particular, in the stance leg, the SOL inhibition precedes TA excitation by about 100 ms [7]. A drop in the background activity is also observed in other dorsal muscles, like BF and ES, while bursts of activity occur in ventral muscles, like Rectus Femoris (RF) [14].

Less literature is available on APAs during gait initiation in children. A systematic survey by Ledebt et al. (1998) [15] showed that APAs start developing at 2–3 years of age but complete their maturation well after the age of 8, a result in line with the observations carried out on toddlers up to 5 years old children [16] and in 4–6 years old children [17]. Another interesting study was published by Isaias et al. (2014) [18], who analyzed SOL and TA in 10 ± 3 years-old children and reported inhibition-excitation patterns like in adults, but with a lower time interval between SOL inhibition and TA excitation.

Several studies showed altered gait initiation in those neurological diseases characterized by poor motor control, as Rett syndrome [18], Parkinson's disease [19], and cerebellar pathologies [20,21]. In particular, Timmann and Horak [21] reported that adults with cerebellar deficits showed a decreased force production and a significant reduction of the length and peak velocity of the first step, accompanied by impairments in the predictive adaptation of APAs to the mechanical needs of gait initiation. Despite these authors also found that the temporal parameters of APAs were overall preserved in patients with cerebellar disease, several other works [22–26] addressed the role of cerebellum in postural control and provided evidence that such structure is involved not only in modulating rate and force of muscle activities but also in determining their relative timing. Indeed, cerebellar deficits often lead to dysfunctional co-contractions [22,24,26]. In this regard, it is worth recalling the involvement of the cerebellum in building up the temporal pattern of APAs, in particular, its ability to create and store internal models of body mechanics. This is proved by the contribution of the cerebellum in modulating sensory-motor interactions and integrating feed-forward and feed-back modes [27].

The cerebellum is also known to play an important role in many developmental disorders [28]. Nevertheless, very little attention has been given to the linkage between the development of postural control and the maturation of such neural structure. Aiming to elucidate this topic, we explored quiet stance and gait initiation in children affected by Pediatric Cerebellar Ataxia (PCA), used as a model of cerebellar dysfunction vs. a healthy control group of comparable age.

PCAs are a heterogeneous group of cerebellar developmental disorders characterized by dysfunctional motor coordination and very early cerebellar symptoms. The first clinical signs

are marked hypotonia, wobbling gait, dysmetria, dysarthria, and a significant developmental delay. Most children show also marked speech impairment and cognitive deficits. In some cases, the cerebellum degenerates with time, but so slowly that it becomes difficult to classify the disorder as progressive or not [29]. In this framework, we studied a group of children with generalized cerebellar atrophy and clinical evidence of slow progression during follow-up (SlowP). In other cases, the disease has a proven non-progressive course, as in Joubert syndrome, which is characterized by cerebellar hypoplasia limited to the vermis and peduncles [30]. Thus, we also considered the second group of children (non-progressive, NonP) affected by this kind of pathology. It is also important to note that the onset of the SlowP pathology is clinically indistinguishable from that of the NonP forms; therefore, practically, both diseases are present since birth.

In order to characterize quiet stance and gait initiation, we measured the classical posturographic parameters [31,32] and the first step length and velocity (as measures of performance). Besides, we also calculated the shifts of CoP and COM, as well as the horizontal distance (gap) between these two points in the imbalance and unloading phases, to highlight the net effect on COM dynamics. In order to document possible changes in muscular APAs, we evaluated the EMG activities that accompany the APA onset. Should we observe significant differences between each pathological group and healthy children, this would point out the involvement of the cerebellum also during the key phase of human maturation, in which the central nervous system learns gait initiation dynamics and how to optimize this motor process. Moreover, these findings would be fruitful in tailoring rehabilitation for such pathologies. Finally, a different motor pattern in children with SlowP vs. NonP would suggest possible compensation mechanisms. In particular, taking into consideration that children with SlowP suffer from generalized cerebellar damage, better motor behavior in SlowP vs. NonP could suggest the involvement of extracerebellar regions. On the other side, better behavior in children of the NonP group could as well stem from a compensatory involvement of the cerebellar hemispheres, which are unaffected by Joubert syndrome.

#### **2. Materials and Methods**

#### *2.1. Participants*

Thirteen participants with PCA were recruited at the "Istituto Neurologico Carlo Besta" of Milan: seven of them had radiological signs of generalized cerebellar atrophy and clinical evidence of slow progression during follow-up (SlowP, mean age: 12 ± 3 years), while the remaining six suffered from Joubert syndrome, i.e., a proven non-progressive pathology (NonP, mean age: 12 ± 3 years). All of them underwent clinical evaluation, including the administration of the Scale for the Assessment and Rating of Ataxia (SARA, [33]), an MRI scan for establishing the cerebellar alteration (atrophy and/or hypoplasia), and genetic screening. In particular, all children with Joubert syndrome showed a unique cerebellar and brainstem malformation known as the "molar tooth sign" [30]. The demographic and clinical data of each patient are reported in Table 1.

Seven healthy children, free from neurological or psychological pathologies and typically developing, were enrolled as a control group, from the primary school "FAES" in Milan (H, 4 males and 3 females, mean age: 10 ± 3 years). The experimental procedure was carried out in accordance with the standards of the Declaration of Helsinki. The Ethical Committee "Comitato Etico di Ateneo dell'Università degli Studi di Milano" approved the study and the written consent procedure, on 15 February 2016 (counsel 5/16). Before each acquisition, the child neuropsychiatrist and the experimenters explained the aim of the study and the details of the experimental procedure to the parents and to their child. Only those children who agreed with the study participated in the experiments. The parents of each participant, as her/his legal representatives, signed the consent procedure. All children were perfectly aware of the task since no one of them failed in accomplishing it.

**Table 1.** Demographic and clinical characteristics of children with PCA. SlowP: slowly progressive ataxia; NonP: non-progressive ataxia (Joubert syndrome); EXOSC3: exosome component 3; KCNC3: potassium voltage-gated channel subfamily C member 3; ADCK3: aarF domain-containing kinase 3; NPHP1: nephrocystin 1; AHI1: Abelson helper integration site 1; SUFU: negative regulator of hedgehog signaling; PCA: pediatric cerebellar ataxia. Details about a molecular diagnosis can be found at [34].


#### *2.2. Experimental Protocol*

Subjects were asked to perform a gait initiation task several times. They were instructed to stand quietly on a force plate for 30 s and then to walk at their natural speed after a vocal prompt, self-selecting the leading limb [35]. After three to five steps, subjects stopped and returned to the initial position.

Each subject repeated the task until three *valid* trials were collected (i.e., trials in which the subject did not move the feet, arms, or head during the quiet stance preceding gait initiation). Subjects were allowed to rest 2 min before repeating the task. A maximum of nine trials was required to satisfy the above criterion; the average number of trials per subject was 6.6 ± 2.7. At the end of the resting periods between motor tasks, the experimenter asked the child, "Do you feel fatigued? Are you ready to start again?". Moreover, since the parents assisted at the trials, they could report to the experimenter any possible discomfort of their child. Neither children nor their parents complained about fatigue. The width of the base of support was self-selected by each subject in the first trial, then marked on the platform with adhesive tape and kept fixed for all further trials. The distance between lateral malleoli during quiet stance was comparable among groups (SlowP: median = 162.8 mm, range = 115.6 to 205.6 mm; NonP: 165.6 mm, 152.3 to 198.5 mm; H: 150.3 mm, 147.5 to 242.3 mm), with no significant differences (Kruskal–Wallis *p* = 0.566).

#### *2.3. Recordings*

Body kinematics was recorded by means of a six-cameras optoelectronic system (SMART-E, BTS, Milan, Italy) using a full-body marker set [36], which allowed estimating the Centre of Mass (COM) and its trajectory, according to Isaias et al. (2014) [18]. A dynamometric force plate (9286AA, KISTLER, Winterthur, Switzerland) was used to compute the Center of Pressure (CoP) position. Wireless probes (FREEEMG 1000, BTS, Milan, Italy) were employed bilaterally to record the surface electromyographic (EMG) activity of Tibialis Anterior (TA), Soleus (SOL), Rectus Femoris (RF), Biceps Femoris (BF), and Erector Spinae (ES). Electrodes were placed according to the Surface Electromyography for the Non-Invasive Assessment of Muscles (SENIAM) guidelines [37]. Synchronous data acquisition was accomplished by the SMART-E workstation; sampling rate being 60 Hz for optoelectronic cameras, 960 Hz for dynamometric signals, and 1000 Hz for EMGs.

#### *2.4. Data Processing*

During the 30 s quiet standing period, the statokinesigram, i.e., the trajectory of the CoP in the horizontal plane, was used to extract specific indexes of balance control. These indexes were: the area and the eccentricity of the ellipse containing 95% of CoP positions, the total length of CoP trajectory (CoP length), the average CoP velocity, and the peak-to-peak Mediolateral and Anteroposterior CoP displacements (ML and AP ranges, respectively) [31,32]. In particular, the ellipse area (A) and eccentricity (*e*) were calculated according to the following formulae:

$$\mathbf{A} = a \ast b \ast \pi \tag{1}$$

$$\kappa = \frac{\sqrt{\left|a^2 - b^2\right|}}{a} \tag{2}$$

in which *a* and *b* were the semi-major axis (i.e., half of the ellipse longest diameter) and the semi-minor axis (i.e., half of the shortest diameter), respectively.

Gait initiation was subdivided into three phases [13]: the imbalance phase, in which CoP moves backward and towards the future swing foot; the unloading phase, in which CoP moves laterally towards the stance foot, and the first swing, in which CoP moves forwards along the stance foot, from toe-off to heel-strike of the swing foot. The temporal events delimiting each phase were determined by visual inspection of the CoP trajectory; in particular, the onset of the CoP backward shift represented the APA onset.

For the *imbalance* and *unloading* phases, separately, we measured the phase duration, the length of the CoP trajectory, the maximum AP and ML shifts of both CoP and COM, and the distance from CoP to COM projection on the horizontal plane, at the end of the phase.

The *first swing* phase was evaluated by measuring the length of the first step, normalized to the lower limb length (*LL*), and its velocity (*v*), expressed in Froude number (Fr = √ *<sup>v</sup> <sup>g</sup>*∗*LL <sup>g</sup>* being gravity acceleration [38]); also these measurements were obtained from kinematic data.

For each subject, the kinematic and dynamometric variables were averaged over the three valid trials recorded. Data normality was evaluated by means of the Shapiro–Wilk test. Considering that data were not normally distributed, the differences among SlowP, NonP, and H groups were analyzed non-parametrically by using Kruskal–Wallis tests followed by Dunn post hoc, with Bonferroni adjustment. The level of significance was set to 0.05.

The analysis of EMG recordings regarded the timing of muscle activation or inhibition, surrounding the APA onset at gait initiation. Raw EMG data were high-pass filtered (fcut = 50 Hz) with a zero-phase shift 6th-order elliptic filter, to remove movement artifacts, and then the signals were rectified. For each muscle, the traces of the recorded trials were time-aligned to the APA onset and averaged across trials. For each averaged EMG track, the period from 1 s to 0.5 s before the APA onset—where no EMG changes were observed—was assumed as reference. The signal was integrated (time constant = 11 ms), and the mean level in the reference period was subtracted. The onset of an excitatory or inhibitory EMG change was identified by a software algorithm, which searched the first time point in which the track fell above or below 2 SD of the reference signal (excitation or inhibition, respectively) and remained there for at least 50 ms. Whenever the criterion was met, the algorithm searched backward the point in which the signal started to deviate from the mean reference value [39]. No statistical analysis was performed on EMG timings because of the many cases in which no clear inhibitory or excitatory changes could be identified.

#### **3. Results**

#### *3.1. Postural Parameters*

The analysis of quiet stance (static posturography) highlighted alterations of postural control in subjects with cerebellar deficits. In fact, children of both the SlowP and NonP groups showed an ellipse area greater than H (Kruskal–Wallis *p* = 0.039), mainly due to a greater CoP displacement in the mediolateral direction (Kruskal–Wallis *p* = 0.022).

However, post hoc showed that this difference was statistically significant only in subjects with SlowP vs. H (Table 2). In particular, the latter group revealed an inversion of the normal ellipse configuration, with mediolateral oscillation as the preferred direction (Figure 1), which led to a reduction of the ellipse eccentricity.

**Figure 1.** Statokinesigram, with 95% confidence ellipse, for representative children of the three groups: healthy (H, panel **A**), with non-progressive PCA (NonP, **B**), and with slow progressive PCA (SlowP, **C**). ML: mediolateral; AP: anteroposterior; PCA: pediatric cerebellar ataxia.



#### *3.2. Gait Initiation Parameters*

The spatial and temporal parameters during the imbalance and unloading phases were not significantly affected by the pathology (Table 3). First step length and velocity were instead different among the three groups (Kruskal–Wallis, *p* = 0.005 for length and *p* = 0.019 for velocity). However, post hoc tests showed that such difference was statistically significant only in children with SlowP vs. H (Table 4).

**Table 3.** Postural parameters during the imbalance and unloading phases. Data are shown as *median value (range)* for children with slow progressive PCA (SlowP), children with non-progressive PCA (NonP), and healthy children (H). ML: mediolateral, positive towards the swing foot; AP: anteroposterior, positive when forwards; CoP→COM: horizontal distance from CoP to COM.


**Table 4.** First swing parameters. Data are shown as *median value (range)* for children with slow progressive PCA (SlowP), children with non-progressive PCA (NonP), and healthy children (H). LL: lower limb length; Fr: Froude number; \* significant difference between SlowP and H groups, Dunn test *p* < 0.05.


#### *3.3. EMG*

Postural EMG changes accompanying APA onset, defined as the first CoP backward shift, could not be detected in all recorded muscles and for all subjects. A descriptive analysis of electromyographic recordings allowed appreciating the development of an inhibitory postural chain involving ES, BF, and SOL, followed by an excitatory chain in RF and TA. Such a general pattern was observed in both the stance and swing sides, irrespectively from the healthy or pathological status. Nevertheless, a different timing distribution of the muscular actions was found in the three groups (Table 5). In the stance limb side, healthy subjects showed a clear craniocaudal progression, for both the inhibitory (first ES and BF, then SOL) and excitatory (first RF, then TA) chains. Such a progression was lost both in children with NonP and in those affected by SlowP; moreover, in SlowP, the recruitment of the excitatory chain was delayed. On the contralateral side with respect to the swing limb, both chains had a caudocranial progression in the H group (first SOL, then BF and ES; first TA, then RF), while children with NonP displayed a disrupted progression of the inhibitory chain, where SOL de-activated after BF and ES, followed by an almost synchronous activation of RF and TA. Instead, in children with SlowP, the inhibitory chain still maintained a caudocranial progression but was overall delayed. Also, in this group, the excitatory actions in RF and TA were synchronous.

**Table 5.** Latencies (ms) of postural EMG changes with respect to the APA onset (time 0). Median, minimum, and maximum values, together with the number of subjects (n) in which APAs could be identified, for children with slow progressive PCA (SlowP), children with non-progressive PCA (NonP), and healthy children (H). EMG: electromyographic; APA: anticipatory postural adjustment.


Of note, in the control group, the inhibition of the stance leg SOL started about 40 ms prior to TA excitation. While this timing was overall preserved in the children of the NonP group (about 60 ms), it was effectively increased in children with SlowP (about 100 ms, Figure 2). Similar changes were detected also for the swing leg.

**Figure 2.** EMG (electromyographic) of shank muscles of the stance leg. Comparison among healthy children H, panel (**A**), children with non-progressive PCA NonP, (**B**) and children with slow progressive PCA SlowP, (**C**). One representative subject for each group. Time 0: APA (anticipatory postural adjustment) onset, defined as the first backward shift of the CoP (center of pressure). Black arrows show SOL (soleus) inhibition and the following TA (tibialis anterior) excitation. Note that the time delay between these two reciprocal actions gradually increased in children with NonP and SlowP with respect to H children.

#### **4. Discussion**

The aim of this study was to describe the postural control adopted by children with PCA during quiet stance and gait initiation, in order to draw considerations on the role of the cerebellum in the development of postural control. As a main result, we observed that in patients with slow progressive PCA, i.e., SlowP, both the static and dynamic components of postural control were disturbed, while the postural behavior of children with non-progressive PCA, i.e., NonP, was much similar to that of healthy children.

During the maintenance of upright posture, children with SlowP showed an increased ellipse area, mainly due to large mediolateral oscillations of the CoP. Considering CoP oscillations in anteroposterior direction too, this resulted in a general reduction of the ellipse eccentricity, outlining an omnidirectional decrease of stability. This finding was in agreement with the results described in adults with cerebellar lesions [26]. No statistical posturographic differences were instead found between children with NonP and H participants.

Gait initiation parameters during the *imbalance* and *unloading* phases remained substantially unchanged in patients of both pathological groups compared to H controls. Also, this observation fitted with previous results obtained in adults with cerebellar ataxia [21,40]. First step length and velocity showed instead a marked reduction in children with SlowP with respect to H, possibly reflecting a compensatory strategy for their poor balance control, and in agreement with what has been previously described in adults [21].

Electromyographic data, despite the roughness of the descriptive approach, suggested that patients with NonP and SlowP suffered more alterations in the temporal (when) than in the spatial distribution (to what muscle) and in the sign of activity (how, i.e., excitation or inhibition). This aspect agreed with the general view that assigns to the cerebellum the role of a "timing-machine" [41–45] and leaves the pattern selection to other brain structures, like the basal ganglia. Such a perspective has been confirmed also for what regards APAs in adults [46,47].

A short comment also deserves TA and SOL reciprocal activation. Despite our analysis was descriptive, patients with PCA and healthy children displayed the classical anticipatory postural pattern, characterized by SOL inhibition followed by TA activation of the stance limb (see Introduction). However, the latency between SOL and TA activity in the healthy group (about 40 ms) was consistent with what reported by Isaias et al. (2014) [18] and much lower than what has been found in adults (about 100 ms, [7]). This observation supported the choice of devoting a paper to gait initiation in children and, at the same time, suggested that the present H group, despite scarce, well represents the underlying population. On the contrary, indications of altered timing were observed in patients with PCA, in which the SOL-TA latency slightly increased to about 60 ms in children with NonP and attained about 100 ms in children with SlowP. This suggested a framework of abnormal feed-forward muscle synergies [21,24].

In summary, the reported differences in postural behavior between children with typical development and children affected by PCA support our hypothesis that the cerebellum plays a role also during the key phase of human maturation, in particular, in building internal models of gait initiation dynamics. This finding further stresses the importance of including postural training exercises in the rehabilitation programs for these pathologies [48]. Moreover, the observation that the gait initiation protocol allowed distinguishing motor deficits in children with SlowP vs. NonP, although the corresponding SARA scores were comparable, suggests that such protocol may be useful to monitor the evolution of motor deficits over time. The following two subsections are devoted to discussing the putative reasons for the different motor patterns we observed in patients of the two groups, as well as the resulting clues about the compensation mechanisms.

#### *4.1. Disease Progression and Postural Behavior*

When looking to the present results as a whole, children with SlowP seem to have a worse postural behavior with respect to both children of the NonP and H groups. This result is unlikely related to the severity of the pathology since all patients had a homogeneous SARA score, which, in turn, indicates comparable motor deficits in clinical terms. Therefore, the difference might stem either from the kind of cerebellar lesion (generalized atrophy vs. vermian hypoplasia) or from the progressive or non-progressive nature of the pathology. In this regard, children with SlowP suffered from generalized cerebellar atrophy, which represents macroscopic neuronal death, and received an ascertained clinical and/or radiological diagnosis of slow progression. On the contrary, Joubert syndrome, affecting children of the NonP group, is a congenital malformation that causes anomalous organogenesis of both the cerebellar vermis and peduncles. Therefore, it has an intrinsically stable nature along with the growth of the subject. In fact, once the organogenesis is completed, the vermian hypoplasia remains stable throughout the patient's lifetime.

It could be argued that our observation of worse postural control in children with SlowP may be related to the larger extent of their cerebellar compromission (generalized atrophy vs. vermian hypoplasia). However, literature reports an emblematic case that contrasts with this interpretation. In fact, Titomanlio et al. (2005) [49] published a case report in which a 17-year-old subject with complete cerebellar agenesis showed only mild ataxia with slight dysmetria and moderate mental retardation, but no difficulty in attaining very complex motor tasks. Such evident functional compensation could be explained only through the plasticity of the remaining brain areas, which had to cope with a lesion that is stable since embryogenesis. This report suggests restricting the hypothesis to the progressive nature of the pathology.

Returning to the present study, we envisage that children with NonP could use the plasticity of their intact brain areas, which may include the cerebellar hemispheres, to effectively compensate for their stable lesion and attain an almost normal psychomotor development. On the contrary, children with SlowP suffer from a continuous cerebellar degeneration, which conflicts with the consolidation of compensatory functional strategies. This perspective not only fits with the gradual worsening of postural deficits we documented here when passing from healthy children to patients with NonP and to patients with SlowP but would also explain why patients with adult-onset cerebellar lesions show even more pronounced postural deficits [46]. Indeed, neural plasticity gradually but consistently decreases over the lifespan [50].

#### *4.2. Putative Compensatory Network*

Finally, it remains to figure out which neural substrate could be involved in functional compensation. In this regard, it is interesting to highlight recent evidence showing subcortical bidirectional connections between the basal ganglia and the cerebellum [51–53].

The functional role of the basal ganglia to cerebellum connections has been deeply investigated. Indeed, it has been observed that patients with Parkinson's disease (PD) show abnormal functioning also in the cerebellum [54,55]. A SPECT study in patients with PD confirms an increased cerebellar activity when the effect of the anti-parkinsonian drug extinguishes [56]. Considering the reciprocal connectivity, it is of interest that functional MRI has shown increased putamen-cerebellar activity in patients with PD performing simple motor tasks and that greater putamen-cerebellar connectivity is significantly correlated with better motor performance. On the contrary, the administration of levodopa, which compensates the low endogenous dopamine production in patients with PD, has reduced this connectivity, relieving the cerebellum from its compensatory task [57]. It has also been observed that the compensatory role of the cerebellum contributes to preventing the full manifestation of the typical motor symptoms during the initial stage of PD; this compensatory ability saturates with time, leading these patients to develop cerebellar symptoms too [58].

These pieces of evidence allow arguing that, reciprocally, intact basal ganglia may compensate for cerebellar deficits. This hypothesis is still to be demonstrated, but should it be proved, it would provide a straightforward explanation for the graded postural impairments we found in children affected by PCA, as well as for the worse impairments reported in adult patients with cerebellar ataxia [46]. Evidence in this regard might come from functional MRI and diffusion tensor imaging techniques.

#### *4.3. Limitations of the Study*

The two main limits of the present study were the small number of participants and the low number of valid trials recorded in each of them. In particular, we could not observe APAs in all subjects, and this prevented a statistical analysis of EMG data. While it could be feasible to recruit more H subjects, the rarity of PCAs limited the number of children that precisely fall within the SlowP or NonP groups. With regard to the low number of valid trials, it could be increased only by prolonging the experimental session, which would quickly become burdensome for children, and especially for those affected by PCA.

#### **5. Conclusions**

Although all children with PCA showed clinically similar motor impairments, only children with SlowP were less stable in standing and showed a significantly shorter and slower first step than healthy children. Also, the descriptive EMG analysis in lower limb and back muscles pointed to a worse postural control in children of the SlowP group. On the basis of recent literature, we proposed that such different behavior stems from the disease progression, which interferes with the consolidation of compensatory strategies in children with SlowP but not in those affected by NonP.

**Author Contributions:** Conceptualization, funding acquisition, and supervision, P.C.; formal analysis, V.F., C.P. (Chiara Palmisano), S.M.M., C.M.M.S., and R.E.; investigation V.F., S.M.M., and C.M.M.S.; software, V.F., C.P. (Chiara Palmisano), and R.E.; clinical resources, S.D., C.P. (Chiara Pantaleoni), A.A., and N.N.; writing—original draft preparation, V.F., S.M.M., and C.P. (Chiara Palmisano); writing—review and editing, V.F., C.P. (Chiara Palmisano), S.M.M., R.E., P.C., S.D., C.P. (Chiara Pantaleoni), A.A., and N.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was supported by Fondazione Mariani for Child Neurology.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
