The Central Nervous System Tightly Controls Rollator-Supported Transitions Between Sitting and Standing

doi:10.21203/rs.3.rs-5030023/v1

Download PDF

Research Article

The Central Nervous System Tightly Controls Rollator-Supported Transitions Between Sitting and Standing

https://doi.org/10.21203/rs.3.rs-5030023/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Older individuals are at risk of falling. Assistive devices like rollators help to reduce that risk, especially by compensating for decreased leg muscle strength and balance problems. Paradoxically, rollators have been found to be a fall risk as well as being difficult to use. To investigate the causes, this study examines how different levels of rollator support (no assistance, light touch, and full support) and balance demands (standard lab floor, balance pads) affect movement coordination during standing up and sitting down movements.

Methods

Twenty young participants stood up and sat down while full-body kinematics and muscle activity (30 channels) were recorded. Participants stood up and at down using different movement strategies (e.g., forward leaning, hybrid, and vertical rise standing up movement strategies). For each movement strategy, spatial and temporal muscle synergies were extracted from the muscle activity patterns. Temporal muscle synergies provided a more compact, low-dimensional representation than spatial muscle synergies, so they were subsequently clustered with k-means++. The activation duration of the temporal muscle synergies was assessed with full-width at half-maximum at the main peak. Multivariate linear mixed models were used to investigate if the muscle activation vectors associated with the temporal muscle synergies differed across the support conditions.

Results

The activation profiles of the temporal muscle synergy timing, but not the shape, differed across the movement strategies for both types of movement. Across all tasks, temporal muscle synergies showed a narrower width of activation around the time of seat-off and seat-on than at the movement start and end. No support-specific temporal muscle synergies were found, but lower-limb muscle activity decreased while upper-limb muscle activity increased with increased support.

Conclusion

The narrow shape of the temporal synergy activation profiles suggests that the central nervous system controls the movements tightly, especially around seat-off and seat-on and in challenging conditions with increased balance demands. Furthermore, rollator support increases upper-body and decreases lower-body muscle activity, especially around seat-off and seat-on. Future studies may further investigate how the loss of tight movement control may cause falls in older individuals.

assistive device

rollator

walker

sit-to-stand

stand-to-sit

muscle synergies

movement primitives

Worldwide, at least one in four individuals over 65 falls every year [1, 2] and gets seriously injured [3]. Consequently, falls and associated injuries often lead to insecurities and restricted mobility, thereby making daily activities challenging and reducing older people’s independence [1, 4]. Assistive devices, including canes, crutches, walkers, or rollators, intend to reduce falls by providing stability and facilitating daily activity independence [5]. Rollators, i.e., four-wheeled walkers, are often prescribed to patients who require an assistive device to compensate for muscular weaknesses. By lessening lower limb loading and enhancing balance, they can help to reduce pain and injuries[5–8]. Paradoxically, studies have shown that rollators are associated with falls [5, 9, 10]. Also, Mann et al. (1995) reported that 57% of the problems with using walkers relate to a “difficult and/or dangerous” use. Although rollators are used widely, the reasons for the difficulties and the increased fall risk have remained unclear due to the lack of thorough biomechanical studies [11].

Rollators are prescribed primarily to help with walking but are also used to stand up and sit down, especially when other assistance, like a handrail or armrest, is missing [12]. Standing up and sitting down are crucial movements to live an independent life but are demanding due to the dynamic balance requirement during the transitions between sitting and standing [13–15]. Lower body muscular strength and balance are two main predictors of successful sit-to-stand movements [16]. However, these decline with age. In particular, knee extensor strength has been found to decline annually by approximately 2–4% after age 50 [17], and a meta-analysis [18] found lower-body weakness to be a statistically significant risk factor for falls. A rollator may help by providing load transfer from the lower to the upper limbs, reducing the strength demand for the hip and knee extension musculature [5, 19]. Furthermore, the additional contact points through the rollator handles might provide extra positional information next to the feet, potentially improving balance [20]. Indeed, a related study found that rollator support increases movement stability during standing up and sitting down movements in young adults [21]. A possible explanation for the increased fall risk in rollator use could be that they interfere with the movement strategies of different tasks. Literature shows that individuals use different movement strategies, e.g., with or without upper-body momentum, to stand up and sit down without a rollator [14, 16, 22, 23], and we have previously found, based on kinematic analyses, that young individuals change their kinematic movement strategy when provided with rollator support, especially when the balance is challenged [24]. These changes in movement strategies indicate that rollators affect the underlying movement coordination patterns, which could be key to understanding human-rollator interactions and safe use.

With approximately 700 muscles and 300 mechanical degrees of freedom, the musculoskeletal system allows countless movement possibilities [25, 26]. To reduce the inherent complexity, the central nervous system (CNS) may employ a modular control architecture [25, 27, 28] which synergies have been proposed as a possible representation [26, 29, 30]. Regardless of whether synergies are identified in spinal reflexes [31], as muscle synergies [32–34], or as kinematic synergies [35, 36], they may generate movement by activating a few functional groups rather than specifying each single element independently [37]. Muscle synergies are often used as a compact low-dimensional representation of a set of recorded muscle activity [26]. Through flexible recruitment and combinations of synergies, the CNS can generate an extensive movement repertoire [32]. According to a hierarchical organization, movement is generated through a combination of a preexisting, trial-independent part and a flexible, trial-dependent part [38]. The trial-independent part is reused across movements and presumably stored in subcortical areas of the CNS; whereas the trial-dependent part, which accounts for variations across trials, is presumably under cortical control [32, 39]. Different models of muscle synergies have been proposed and differ in terms of which part of the decomposition is trial-dependent and which is trial-independent [40–42]. In spatial muscle synergies, also called time-invariant or synchronous synergies, the trial-independent part is made up of vectors. These vectors represent the activations of multiple muscles, relative in magnitude to each other. These fixed vectors are combined with trial-dependent activation functions (or time-varying coefficients), representing the amount and timing of the muscle vectors. In contrast, temporal muscle synergies (or temporal components, basic patterns), consist of muscle activation profiles, invariant over muscles and conditions, and trial-dependent muscle activation vectors. Hence, the relative weightings of muscle activations vary across trials. Regardless of the model, muscle synergies are robust against the highly variable and stochastic nature of EMG patterns. Since they therefore reveal the neural organization underlying behavior and functional outcomes of muscular activation [43–45], they have been used in various settings to investigate questions regarding movement coordination [46], as well as in sit-to-stand movements.

Three to four spatial muscle synergies typically explain 87–94% of the variance in unassisted sit-to-stand movements across various age groups [47–52]. Usually, each synergy represents one biomechanical function, such as moment transfer and postural stabilization. While muscle synergy structure seems robust in unassisted standing up, even with visual or vestibular disturbances, the activation timing changes with these disturbances [53]. Furthermore, synergy activation timings have been found to differ across movement strategies [52]. Consequently, analyzing the temporal structure is crucial. In contrast to unassisted standing up, assistive devices seem to impact not only the activation timing in standing up, but also the number of synergies increased when participants stood up while being pushed up by the chair [54], and studies with Nordic walking sticks [55] or exoskeletons [56, 57] show that assistive tools alter movement coordination; such that, for example, synergies specific to the device emerge. Consequently, when studying rollator-assisted standing up, it is essential to consider upper-body involvement to account for the involvement of the arms. This is particularly relevant as the aging process leads to muscle weakening [55, 56], and individuals thus often need to push on armrests to master transitioning between sitting and standing. Consequently, it remains an open question how rollator usage influences full-body movement coordination in standing up movements, especially considering the different movement strategies and upper-body involvement.

Like in standing up, assistive devices may alter sitting down coordination. Only a few biomechanical and movement coordination studies have analyzed sitting down [48, 57, 58], let alone with a rollator. This is surprising as sitting down is also a complex movement and not simply the opposite of standing up [16, 59] for example, the gluteus maximus works concentrically when standing up but eccentrically when sitting down [57]. Also, upper body muscles may act differently when sitting down using a rollator, such that the arm extensor muscles presumably work eccentrically rather than concentrically. Accordingly, this implies that rollator usage may influence stand-to-sit movement coordination in a different manner from sit-to-stand.

To improve our understanding of human-rollator interactions and safe use, we apply muscle synergy analysis and examine the movement coordination underlying different movement strategies for standing up and sitting down. In particular, we explore how different rollator support conditions (no assistance, light touching with supposed proprioceptive cues through the rollator handles, and full support with supposed lower body load reduction) affect muscle synergies. As thorough studies on movement coordination underlying rollator-assisted movement are scarce [11], and the rollator-prone population is heterogeneous regarding their underlying deficits, this study investigates young participants as a baseline measure. However, to make standing up and sitting down more challenging, we placed balance pads underneath their feet to evaluate the effect of rollator support while experiencing increased postural instability.

We hypothesized that (1) temporal muscle synergies represent sit-to-stand and stand-to-sit EMG patterns more compactly than spatial muscle synergies. Furthermore, we hypothesize that (2) the temporal muscle synergies differ across movement strategies and that (3) rollator support influences the relative activation between upper-body and lower-body muscle activity for both standing up and sitting down.

Our previously published articles on this dataset describe our analysis of the kinematic and kinetic data [21, 24]. Here, we introduce all the steps regarding the EMG analysis.

Participants

Twenty young and healthy volunteers (10 women, 10 men; 25.5 ± 3.8 years, 1.71 ± 0.08 m height, 67.6 ± 10.9 kg mass) gave written informed consent and participated in the study. The participant shown in Fig. 1 gave informed consent to publish the image. The Ethics Committee of the Medical Department of Heidelberg University (S-105/2021) approved the study, which was then performed according to the Declaration of Helsinki.

Experimental protocol

The participants sat still and, after hearing “stand up” by the experimenter, they stood up at their own pace. Then, the experimenter said, “stand still”. After standing still for at least two seconds, the experimenter said, “sit down,” and the participants sat down at their own pace. The seat height was set to the height of the participant’s lateral epicondyle of the femur, and a custom-built robot rollator simulator was used to provide rollator support (Fig. 1). Following recommendations in the health care literature [60–62], the handle height was set at the participant’s standing wrist height. According to the support condition, they did not use the rollator handles at all (unassisted, UA), only with a light touch of the hand (light touch, LT) or with a power grip (full support, FS). These support conditions were combined with two floor conditions: the standard lab floor and a more "challenging ground", which was created by positioning a circular rubber balance pad with a compliant surface (Dynair® Ballkissen®, diameter 33 cm, height 8 cm, TOGU GmbH, Prien-Bachham, Germany) under each foot. Participants familiarized themselves with the task by performing two repetitions in each condition combination (support: unassisted, light touch, full support; floor: non-challenging, challenging). No further instructions on the movement execution were given, allowing the participants to stand up and sit down as naturally as possible. All participants performed three valid, non-consecutive repetitions in each condition combination, resulting in a total of 18 trials per participant. The order of the support and floor conditions was randomized across participants.

<Please insert Fig. 1.pdf here>

Figure 1: Experimental setup and data analysis. A: The participant stands up from an instrumented chair with the custom-made robot rollator simulator. Full-body passive markers for motion tracking and EMG electrodes were placed on the body. Two movements were studied: sit-to-stand and stand-to-sit. Two floor conditions were used (middle): non-challenging (lab floor) and challenging (balance pads). Three different support conditions were used (right): unassisted (handles not used), light touch (palm on the handles), and full support (power grip). B: Participants used different movement strategies and switched between them, as exemplarily shown for the non-challenging sit-to-stand task. The bottom plot shows the distribution of the trials among strategies. One dot represents one trial. The row indicates to which movement strategy it belongs. The column shows to which participant it belongs. The support conditions are color-coded as indicated by the legend. The labels on the right y-axis show how many trials were associated with the strategy written on the left y-axis. The figure is adapted from [24]. C: EMG data from the trials of the same movement strategy were arranged into a matrix (for example, Mvertical rise). Temporal muscle synergies were extracted from each matrix with NMF [63–65], resulting in trial-independent activation profiles and trial-dependent muscle activation vectors. The activation profiles were matched across the movement strategies and ordered chronologically using k-means + + clustering, ensuring a correlation coefficient above 0.9. Bottom: The duration of each activation profile was assessed as full-width at half-maximum. Linear mixed models were used to investigate how rollator support affects the muscle activation vectors. Abbreviations: P1: participant 1, UA1: first unassisted trial, LT1: first light touch trial, FS1: first full support trial, ES: M. erector spinae, RA: M. rectus abdominis, NMF: non-negative matrix factorization [63, 64].

Data collection

Full-body 3D kinematics were obtained using the IOR full-body marker model [66, 67] and ten cameras (150 Hz; Type 5+, Qualisys, Gothenburg, Sweden). Ground reaction forces (GRF; 1000Hz; Bertec Corp., Columbus, OH, USA) and forces on the seating surface (142 Hz; Phidgets Inc., Calgary, AB, Canada) were measured.

Thirty surface EMG electrodes (two 16-channel systems at 1500 and 4000 Hz; Noraxon USA, Scottsdale, AZ, USA) captured full-body muscle activity of the following muscles bilaterally: pectoralis major (Pec), latissimus dorsi (Lat), trapezius (Tra), deltoideus (Del), biceps brachii (Bic), triceps brachii (Tri), gluteus medius (GM), tensor fasciae latae (TF), rectus femoris (RF), vastus medialis (VM), biceps femoris (BF), tibialis anterior (TA), peroneus longus (PL), and gastrocnemius (GA). Additionally, erector spinae (ES) and rectus abdominis (RA) activities were recorded. The participants’ skin was prepared by shaving, abrasion, and cleansing with alcohol to ensure good electrode-skin contact before the Al/AgCl electrodes were attached according to SENIAM guidelines [68] and [69].

Data processing

To reconstruct the 3D coordinates of the markers, raw kinematic data were processed offline with Qualisys Track Manager (v 2018.1). Subsequently, force and kinematic data were filtered with a 4th -order zero-lag low-pass Butterworth filter at 10 Hz. Using Visual3D (v6, C-Motion Inc., Germantown, MD, USA), full-body kinematics and the center of mass (CoM) were then calculated. Further data analyses were done in Matlab (R2023b, Natick, MA, USA). Raw EMG data were bandpass (20–500 Hz) and notch (50 Hz) filtered with a 4th -order zero-lag Butterworth filter [49]. ECG artefacts apparent in the trunk muscle recordings were removed with a template-matching procedure [70]. For robustness, we created a muscle- and participant-specific template of ECG artefacts using the data from the sit-to-stand and stand-to-sit recordings and an additional 9 minutes of still-standing recordings. Subsequently, filtered EMG data were full-wave rectified and smoothed with a 4th -order zero-lag low-pass Butterworth filter at 10 Hz [49]. Movement start, seat-off/on, and movement end were identified using a k-means + + algorithm on the GRF and CoM data [71]. As muscles need to be active before a visible movement starts, data from 200 ms before the detected start of the movement were included [72, 73]. Afterward, data were segmented and time-normalized to 101 time points (100%) using a spline interpolation. Finally, EMG data were amplitude-normalized per muscle and participant to their maximum activity across the 18 trials [64]. Of the 720 trials, 24 were not included in the analyses as some recordings of a few muscles were corrupt (Supplementary Table 1, Additional File 1).

Movement strategies

Our previous investigation found that participants switched movement strategies when introduced to a rollator [24]. In particular, three movement strategies were identified for the sit-to-stand non-challenging task (“forward leaning”, “hybrid, and “vertical rise”) and two for the challenging task (“exaggerated forward leaning” and “forward leaning”). Likewise, three and two movement strategies were identified respectively in the stand-to-sit tasks (“vertical lowering”, “hybrid”, and “backward lowering”; and “exaggerated forward leaning” and “forward leaning”). The naming of these strategies was inferred by visual inspection of their movement progression, the different hip, knee, and ankle sagittal angle courses, and the relative movements between the CoM and the heel. The grouping of trials into the movement strategies was used in the current analysis to extract muscle synergies specific to the movement strategies. Supplementary Fig. 1, Additional File 1 shows the distribution of trials among strategies.

Data analysis

3.1.1 Muscle synergy analysis

We extracted muscle synergies with respect to our previous findings that participants switched their movement strategies when introduced to a rollator ([24]; Fig. 1). As stated in the introduction, the two models (spatial and temporal muscle synergies) describe different aspects of movement coordination [42, 74–77]. Temporal muscle synergies allow us to follow the underlying assumption that the CNS uses a fixed temporal sequence (trial-independent activation profiles) for the different movement strategies and that muscle weightings vary (trial-dependent muscle activation vectors) across the support conditions. Therefore, the EMG signals (30 channels) of all trials belonging to the same strategy (strat) were horizontally concatenated into a data matrix $\:{\mathbf{M}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t}}\in\:{\mathbb{R}}_{\ge\:0}^{101\:\times\:\:30\bullet\:\text{t}\text{r}}$. NMF decomposed M_strat into a set of N_strat trial-independent activation profiles $\:{\text{C}}_{\text{s}\text{t}\text{r}\text{a}\text{t},\text{n}}\in\:{\mathbb{R}}_{\ge\:0}^{101\:\times\:\:1}$, and trial-dependent muscle activation vectors$\:\:{\mathbf{W}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t},\:\mathbf{n}}^{\mathbf{s}}\in\:{\mathbb{R}}_{\ge\:0}^{30\:\times\:\:1}$ [63–65]. Thus, EMG data from a single trial (s is the trial index) $\:{\mathbf{M}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t}}^{\mathbf{s}}\left(\text{t}\right)$ were decomposed with:

$$\:{\mathbf{M}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t}}^{\mathbf{s}}\left(\text{t}\right)\approx\:{\sum\:}_{\text{n}\in\:{\text{N}}_{\text{s}\text{t}\text{r}\text{a}\text{t}}}{\text{C}}_{\text{s}\text{t}\text{r}\text{a}\text{t},\:\text{n}}\left(\text{t}\right)\:\cdot\:{\mathbf{W}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t},\mathbf{n}}^{\mathbf{s}}$$

To investigate hypothesis 1 (that temporal muscle synergies represent sit-to-stand and stand-to-sit EMG patterns more compactly than spatial muscle synergies), we also extracted spatial muscle synergies. Therefore, the EMG signals (30 channels) of all trials belonging to the same strategy (strat) were horizontally concatenated into a data matrix $\:{\mathbf{M}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t},\:\:\:\mathbf{s}\mathbf{p}\mathbf{a}\mathbf{t}\mathbf{i}\mathbf{a}\mathbf{l}}\in\:{\mathbb{R}}_{\ge\:0}^{30\:\times\:\:101\bullet\:\text{t}\text{r}}$. NMF decomposed M_{strat, spatial} into a set of N_{strat, spatial} trial-independent muscle activation vectors $\:{\mathbf{W}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t},\mathbf{n},\:\mathbf{s}\mathbf{p}\mathbf{a}\mathbf{t}\mathbf{i}\mathbf{a}\mathbf{l}}\in\:{\mathbb{R}}_{\ge\:0}^{30\:\times\:\:1}$, and trial-dependent activation profiles $\:{\text{C}}_{\text{s}\text{t}\text{r}\text{a}\text{t},\:\text{n},\text{s}\text{p}\text{a}\text{t}\text{i}\text{a}\text{l}}^{\text{s}}\in\:{\mathbb{R}}_{\ge\:0}^{1\:\times\:\:101}$ [63–65]. Thus, EMG data from a single trial (s is the trial index) $\:{\mathbf{M}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t},\:\mathbf{s}\mathbf{p}\mathbf{a}\mathbf{t}\mathbf{i}\mathbf{a}\mathbf{l}}^{\mathbf{s}}\left(\text{t}\right)$ were decomposed with:

$$\:{\mathbf{M}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t},\mathbf{s}\mathbf{p}\mathbf{a}\mathbf{t}\mathbf{i}\mathbf{a}\mathbf{l}}^{\mathbf{s}}\left(\text{t}\right)\approx\:{\sum\:}_{\text{n}\in\:{\text{N}}_{\text{s}\text{t}\text{r}\text{a}\text{t},\text{s}\text{p}\text{a}\text{t}\text{i}\text{a}\text{l}}}{\mathbf{W}}_{\mathbf{s}\mathbf{t}\mathbf{r}\mathbf{a}\mathbf{t},\mathbf{n},\mathbf{s}\mathbf{p}\mathbf{a}\mathbf{t}\mathbf{i}\mathbf{a}\mathbf{l}}\cdot\:{\text{C}}_{\text{s}\text{t}\text{r}\text{a}\text{t},\text{n},\text{s}\text{p}\text{a}\text{t}\text{i}\text{a}\text{l}}^{\text{s}}\left(\text{t}\right)$$

NMF’s iterative decomposition was limited to 3000 iterations and started 50 times to avoid convergence to a local minimum [48, 78].

A 5-fold cross-validation procedure was used to increase the confidence that the extracted muscle synergies were robust and generalizable rather than due to characteristics of single trials. In line with the literature, a training/test split of 80:20 was employed [32, 79]. Muscle synergies were extracted from a random 80% portion of the trials. Then, the trial-independent parts were fixed and fitted to the remaining 20% of the trials. Finally, the reconstruction quality R²_CV of the fits to the test sets was used to identify the number of synergies. R² is a multivariate measure allowing assessment of the reconstruction quality: R² = 1 − SSE/SST, with SSE being the sum of the squared residuals and SST the sum of the squared residuals from the mean vector [80]. The numbers of synergies N_strat and N_{strat,spatial} were chosen at the R² knee point, after which the R² curve remained approximately straight [80]. Therefore, a series of linear regressions were fitted to the R² curve, starting with the interval [N₁, N₃₀] and iteratively removing the smallest N from the interval. Then, the regressions’ mean squared residual errors (MSE) were calculated, and N_strat and N_{strat,spatial} were selected for the first number N with an MSE smaller than 10^− 4. The number of synergies to extract must be chosen carefully to obtain a good low-dimensional representation of the data with minimum noise [81, 82], and numerous criteria have been proposed [83]. To avoid the results being specific to the choice of N_strat, rather than reflecting physiological patterns, an additional criterion was used to compare the results. Therefore, N_strat^* was chosen to be the minimum number fulfilling both a global ($\:{R}_{{N}_{strat}}^{2}\ge\:0.9$) and a local criterion ($\:{R}_{n}^{2}-{R}_{n-1}^{2}<0.05;\:n=1...{N}_{strat}$).

To justify the choice of temporal muscle synergies beyond the assumption that the CNS uses a fixed temporal sequence (trial-independent activation profiles) for the different movement strategies and that muscle weightings vary (trial-dependent muscle activation vectors) across the support conditions (degree of handle support), we extracted spatial synergies and evaluated the dimensionality reduction (R²-knee criterion) against the dimensionality reduction of the temporal extraction (N_strat and N_{strat, spatial}). Furthermore, the number of trial-dependent parameters (temporal: synergies ∙ muscles, spatial: time samples ∙ synergies) and the number of trial-independent parameters (temporal: time samples ∙ synergies, spatial: synergies ∙ muscles) were compared between the temporal and spatial extraction, with the number of synergies selected at the respective R² knee [41]. All steps regarding muscle synergy analyses were done separately for the two movements and floor conditions.

3.1.2 Matching of similar synergies across the movement strategies

To investigate hypothesis 2, that the activation profiles of the temporal muscle synergies differ across movement strategies, the activation profiles were matched with k-means + + clustering (Fig. 1; [84]). A correlation coefficient [84] of at least 0.9 per match was ensured. Only one activation profile per strategy was included within each cluster. The clustering was repeated ten times to confirm the robustness of the cluster assignments [85].

3.1.3 Statistics

Timing and duration of temporal synergies

To investigate hypothesis 2 that the activation profiles of the temporal muscle synergies differ across movement strategies beyond their similarity (see 3.5.2), we assessed the duration of activation by measuring the full-width at half-maximum of the main peak (FWHM; Matlab findpeaks; [84]. Therefore, the time difference between the two points at half-height on either side of the main peak was calculated. As sit-to-stand and stand-to-sit are sequential movements, we cannot assume that the boundary synergies (synergies with the peaks close to the movement start and end) are symmetrical. In the case of boundary synergies, we measured the difference between the movement start and the point of half-height of the descending synergy or the difference between the point of half-height and the movement end.

Linear mixed model to assess differences in the activation vectors

To investigate hypothesis 3, i.e., to identify the changes in the muscle weightings between the support conditions (unassisted, light touch, full support), a linear mixed model (LMM) was used (Matlab fitlme). The LMM considers that repeated measures of a single participant are likely correlated [86, 87] allowing us to manage the distribution of participants’ trials over several movement strategies. To account for the simultaneous changes in multiple muscle activations (trial-dependent muscle activation vectors) with support condition, a multivariate LMM approach was used [88].

Single trials (level 1) were nested into participants (level 2). The support conditions were included as dummy variables using reference coding (LT or FS set to 1, or both to 0 for the unassisted condition). The first multivariate LMM was calculated with UA as the reference group, and the second multivariate LMM with LT as the reference group, allowing investigation of all pairwise comparisons within the support conditions. Each muscle activation vector belongs to one temporal synergy, i.e., the reference frames in which the muscle activation vectors lie are different across the temporal synergies, so one multivariate LMM was used for each muscle vector. For example, for the non-challenging sit-to-stand task using the forward leaning strategy, seven synergies led to 14 multivariate LMM tests within this task (seven synergies tested once with UA and once with LT as reference group). Sex was added as a within-participant control variable at level 1 but was not included in the final model as it did not improve the model based on the change in the − 2 log-likelihood or the Akaike's information criterion (Matlab’s linearmixedmodel.compare function; [89, 90]). The residual plots were inspected to assess linearity and homoscedasticity as prerequisites for LMM, and no gross violations were found [91].

The multivariate LMM regression formula was:

Level 2: $\:{{\beta\:}}_{0\text{p}}={{\gamma\:}}_{00}+{\text{u}}_{0\text{p}}$

$$\:{{\beta\:}}_{1\text{p}}={{\gamma\:}}_{10}$$

$$\:{{\beta\:}}_{2\text{p}}={{\gamma\:}}_{20}$$

$\:\text{A}\text{c}\text{t}\text{i}\text{v}\text{a}\text{t}\text{i}\text{o}{\text{n}}_{\text{t}\text{p}}$ represents the mean activation across the left and right limb of each bilaterally assessed muscle and the single muscle activation of M. erector spinae and M. rectus abdominis of a given synergy on the t^th trial for the p^th participant. The $\:{{\beta\:}}_{0\text{p}}$ represents the intercept, $\:{{\beta\:}}_{1\text{p}}$ and $\:{{\beta\:}}_{2\text{p}}$ the fixed effects for the support conditions, and $\:{\text{ϵ}}_{\text{t}\text{p}}$ represents the trial- and participant-specific residual. The variable $\:{\text{u}}_{0\text{p}}$ is a support-specific random component of $\:{{\beta\:}}_{0\text{p}}$ and the $\:{\gamma\:}$ fixed effect parameters. Accordingly, the following formula was used for the function specification of the fitlme function: ‘Activation ~ LT + FS + (1 | Participant)’. To account for the multivariate nature, muscle activities were vertically concatenated into $\:\text{A}\text{c}\text{t}\text{i}\text{v}\text{a}\text{t}\text{i}\text{o}{\text{n}}_{\text{t}\text{p}}$. For the second multivariate LMM test, ‘LT’ was exchanged with ’UA’ to have LT as the reference group. The t-statistic on the ß coefficients was used for hypothesis testing with the significance set a priori at a two-sided α = 0.05 [90]. To reduce the probability of type I errors, the level of significance was adjusted according to the number of tests within one movement strategy (e.g., for the seven synergies in the non-challenging, sit-to-stand forward leaning strategy, the level of significance was adjusted for 14 tests with Bonferroni correction).

If the multivariate LMM showed significance regarding the support groups [88], MscName (a categorical variable representing the muscle names) was used in the second step as a dummy variable of the LMM analyses using effect coding. Therefore, the interaction between the muscles and the support condition was added to the LMM. This revealed the muscles in which the support condition affects their activation with respect to the mean of the group mean of the reference muscle (M. Rectus abdominis) and support group (effect coding). Consequently, the following model was used:

Level 1: $\:\text{A}\text{c}\text{t}\text{i}\text{v}\text{a}\text{t}\text{i}\text{o}{\text{n}}_{\text{t}\text{p}}={{\beta\:}}_{0\text{p}}+{{\beta\:}}_{1\text{p}}\text{L}{\text{T}}_{\text{t}\text{p}}+{{\beta\:}}_{2\text{p}}\text{F}{\text{S}}_{\text{t}\text{p}}+{\varvec{\beta\:}}_{3\mathbf{p}}^{\mathbf{T}}{\mathbf{M}\mathbf{s}\mathbf{c}\mathbf{N}\mathbf{a}\mathbf{m}\mathbf{e}}_{\mathbf{p}}+\:{\varvec{\beta\:}}_{4\mathbf{p}}^{\mathbf{T}}\text{L}{\text{T}}_{\text{t}\text{p}}{\mathbf{M}\mathbf{s}\mathbf{c}\mathbf{N}\mathbf{a}\mathbf{m}\mathbf{e}}_{\mathbf{p}}+{\varvec{\beta\:}}_{5\mathbf{p}}^{\mathbf{T}}\text{F}{\text{S}}_{\text{t}\text{p}}{\mathbf{M}\mathbf{s}\mathbf{c}\mathbf{N}\mathbf{a}\mathbf{m}\mathbf{e}}_{\mathbf{p}}+\:{\text{ϵ}}_{\text{t}\text{p}}$

Level 2: $\:{{\beta\:}}_{0\text{p}}={{\gamma\:}}_{00}+{\text{u}}_{0\text{p}}$

$$\:{{\beta\:}}_{1\text{p}}={{\gamma\:}}_{10}$$

$$\:{{\beta\:}}_{2\text{p}}={{\gamma\:}}_{20}$$

$$\:{\varvec{\beta\:}}_{3\mathbf{p}}={\varvec{\gamma\:}}_{30}$$

$$\:{\varvec{\beta\:}}_{4\mathbf{p}}={\varvec{\gamma\:}}_{40}$$

$$\:{\varvec{\beta\:}}_{5\mathbf{p}}={\varvec{\gamma\:}}_{50}$$

Matlab model function: Activation ~ (LT + FS) * MscName + (1 | Participant)

To reduce the probability of type I errors, the significance level (α = 0.05) was adjusted to the number of tests with the second model, at maximum three if a significant effect was found for each pairwise comparison of UA, LT, and FS. The LMM was implemented using the maximum likelihood method.

Temporal muscle synergies result in a more compact representation than spatial muscle synergies in sit-to-stand and stand-to-sit EMG patterns

To investigate if temporal muscle synergies represent sit-to-stand and stand-to-sit EMG patterns more compactly than spatial muscle synergies, both types of synergy were extracted from the EMG data. In each movement task, six to ten temporal synergies reconstructed the EMG patterns Fig. 2) with an R² between 0.84 and 0.89 (Fig. 2, Table 1). In particular, 5.8 ± 2.4 fewer synergies were necessary for the temporal than for the spatial extraction with the R²-knee criterion. Figure 3 shows the EMG and reconstruction for an exemplary participant. The number of trial-independent parts is smaller for the temporal muscle synergies than for the spatial muscle synergies, and fewer trial-dependent parts are necessary for every trial (Table 2). Accordingly, the total number of the trial-independent and trial-dependent parts is smaller for the temporal muscle synergies, which indicates that the higher R² of the temporal muscle synergies is not simply due to the number of parts. Thus, temporal muscle synergies indeed reduced the dimensionality more than spatial muscle synergies, supporting hypothesis 1.

<Please insert Fig. 2.pdf here>

Figure 2: Reconstruction quality (R²) for every movement (A, B, C, D) and movement strategy (color-coded). Thick lines: Temporal muscle synergy reconstruction quality, thin lines: spatial muscle synergy reconstruction quality. The figure shows mean and standard deviation over the 5-fold cross-validation runs. The vertical dotted lines, color-coded with respect to the movement strategies, indicate the choice for the number of synergies according to Table 1.

Table 1

The numbers and R² of synergies according to the applied criterion and choice of model.
Movement	Sit-to-stand, non-challenging			Sit-to-stand, challenging
Strategy	Forward leaning	Hybrid	Vertical rise	Exaggerated forward leaning	Forward leaning
Temporal R² knee (N_strat)	7 (0.888 ± 0.008)	6 (0.874 ± 0.012)	6 (0.879 ± 0.005)	10 (0.873 ± 0.012)	8 (0.874 ± 0.011)
Temporal R² > 0.9 (N_strat^*)	8 (0.902 ± 0.005)	8 (0.902 ± 0.009)	8 (0.905 ± 0.003)	13 (0.905 ± 0.009)	11 (0.906 ± 0.009)
Spatial R² knee (N_{strat, spatial})	15 (0.952 ± 0.005)	12 (0.931 ± 0.007)	16 (0.953 ± 0.002)	11 (0.882 ± 0.010)	13 (0.920 ± 0.006)

Movement

Stand-to-sit, non-challenging

Stand-to-sit, challenging

Strategy

Backward lowering

Hybrid

Vertical lowering

Exaggerated forward leaning

Forward leaning

Temporal

R² knee (N_strat)

(0.865 ± 0.009)

(0.868 ± 0.017)

(0.864 ± 0.013)

(0.842 ± 0.004)

(0.842 ± 0.007)

Temporal

R² > 0.9 (N_strat^*)

(0.902 ± 0.004)

(0.901 ± 0.014)

(0.905 ± 0.013)

(0.902 ± 0.005)

(0.901 ± 0.005)

Spatial

R² knee (N_{strat, spatial})

(0.929 ± 0.005)

(0.915 ± 0.013)

(0.914 ± 0.011)

(0.895 ± 0.005)

(0.896 ± 0.004)

<Please insert Fig. 3.pdf here>

Figure 3: Kinematics (top), original EMG (gray areas), and reconstruction (red solid lines). Data from one exemplary participant using the forward leaning strategy to stand up. The dashed line indicates the movement start, i.e., the first visible CoM movement. The dotted line indicates seat-off. Left: unassisted, middle: light touch, right: full support condition. Muscle activity of the right limb is in the positive direction, and the left limb is in the negative direction.

Table 2

Amount of dimensionality reduction in terms of trial-independent and trial-dependent signals according to [41].
Movement		Sit-to-stand, non-challenging			Sit-to-stand, challenging
Strategy		Forward leaning	Hybrid	Vertical rise	Exaggerated forward leaning	Forward leaning
Number of trial-independent parameters	Temporal: time samples ∙ synergies	707	606	606	1010	808
Number of trial-independent parameters	Spatial: synergies ∙ muscles	450	360	480	330	390
Number of trial-dependent parameters	Temporal: synergies ∙ muscles	210	180	180	300	240
Number of trial-dependent parameters	Spatial: time samples ∙ synergies	1515	1212	1616	1111	1313
Sum temporal		917	786	786	1310	1048
Sum spatial		1965	1572	2096	1441	1703

Movement		Stand-to-sit, non-challenging			Stand-to-sit, challenging
Strategy		Vertical lowering	Hybrid	Backward lowering	Forward leaning	Exaggerated forward leaning
Number of trial-independent parameters	Temporal: time samples ∙ synergies	808	808	808	707	707
Number of trial-independent parameters	Spatial: synergies ∙ muscles	480	390	390	360	360
Number of trial-dependent parameters	Temporal: synergies ∙ muscles	240	240	240	210	210
Number of trial-dependent parameters	Spatial: time samples ∙ synergies	1616	1313	1313	1212	1212
Sum temporal		1048	1048	1048	917	917
Sum spatial		2096	1703	1703	1572	1572

The timing of temporal muscle synergies differs across movement strategies

To investigate hypothesis 2 that the temporal muscle synergies differ across movement strategies, temporal synergy activation profiles were clustered using k-means + + across the movement strategies. Accordingly, similar temporal muscle synergies are characterized by a similar shape of activation and timing.

4.1.1 Sit-to-stand movement strategies

In the non-challenging task (Fig. 4), the activation profiles of the seven forward leaning synergies and the six hybrid and six vertical rise strategies were grouped into eight clusters. Timing differed across the movement strategies between movement start and shortly before seat-off. Also, there is one distinct forward leaning synergy active right after seat-off. At movement start, seat-off, and at movement end, the clusters contained synergies from all three movement strategies. Hence, at these times, the activation profiles do not differ.

In the challenging condition (Fig. 5), the ten synergies of the exaggerated forward leaning and eight from the forward leaning strategy were grouped into eleven clusters. Right after seat-off, the timing of the synergies differs between the two movement strategies, with three distinct exaggerated forward leaning temporal synergies. This differs from the non-challenging task, where we additionally found differences between movement start and seat-off. Across all movement strategies, the main synergy peaks widths (FWHM) were larger at movement start and end than in the middle, especially around seat-off (Table 3).

The findings that the timing of the temporal muscle synergies differs across the movement strategies are robust regarding the criterion for selecting the numbers of synergies (Supplementary Fig. 2, Additional File 1). Therefore, we accept our first hypothesis that muscle synergies differ across movement strategies regarding the sit-to-stand movements.

<Please insert Fig. 4.pdf here>

Figure 4: Temporal muscle synergies in the sit-to-stand movement and non-challenging condition. Top: Exemplary movements in the unassisted, light touch, and full support conditions (in rows) and movement strategies (color-coded). Bottom, left: Temporal muscle synergies sorted chronologically and color-coded according to the movement strategy. The dashed line indicates the movement start, i.e., the first identified CoM movement. The dotted line indicates seat-off. Bottom, right: Bar graphs show mean muscle activation vectors across all trials within a strategy and support condition, and single dots represent trial-specific activation in two columns for the left and right limbs. The different color shades indicate the support conditions. The bars indicate statistically significant differences according to the LMM statistics (details in Additional File 2).

<Please insert Fig. 5.pdf here>

Figure 5: Temporal muscle synergies in the sit-to-stand movement and challenging condition. Top: Exemplary movements in the unassisted, light touch, and full support conditions (in rows) and movement strategies (color-coded). Bottom, left: Temporal muscle synergies sorted chronologically and color-coded according to the movement strategy. The dashed line indicates the movement start, i.e., the first identified CoM movement. The dotted line indicates seat-off. Bottom, right: Bar graphs show mean muscle activation vectors across all trials within a strategy and support condition, and single dots represent trial-specific activation in two columns for the left and right limbs. The different color shades indicate the support conditions. The bars indicate statistically significant differences according to the LMM statistics (details in Additional File 2).

4.1.2 Stand-to-sit movement strategies

In the non-challenging task (Fig. 6), the activation profiles of the eight backward lowering, hybrid, and vertical lowering strategies were grouped into ten clusters. Most of the time, the timing across the three movement strategies did not differ. However, there was one distinct synergy for each of the movement strategies: the distinct backward lowering synergy is active after seat-on, the distinct hybrid synergy at seat-on, and the distinct vertical lowering synergy between movement start and seat-on.

In the challenging condition (Fig. 7), seven synergies of the exaggerated forward leaning and the forward leaning strategy were grouped into eight clusters. As in the non-challenging task, the timing did not differ between the two strategies most of the time. However, each strategy shows a distinct synergy. For the exaggerated forward leaning strategy, the synergy is active at seat-on, and for the forward leaning synergy after seat-on.

Similar to the sit-to-stand movements, the widths of the main synergy peaks (FWHM) were larger at movement start and end than in the middle, especially around seat-off (Table 3). Also, the finding that the timing of the temporal muscle synergies differs across the movement strategies is robust regarding the criterion for selecting the numbers of synergies (Supplementary Fig. 2, Additional File 1). Therefore, we accept our first hypothesis that muscle synergies also differ across movement strategies for the stand-to-sit movements.

<Please insert Fig. 6.pdf here>

Figure 6: Temporal muscle synergies in the stand-to-sit movement and non-challenging condition. Top: Exemplary movements in the unassisted, light touch, and full support conditions (in rows) and movement strategies (color-coded). Bottom, left: Temporal muscle synergies sorted chronologically and color-coded according to movement strategy. The dashed line indicates movement start, i.e., the first identified CoM movement. The dotted line indicates seat-off. Bottom, right: Bar graphs show mean muscle activation vectors across all trials within a strategy and support condition, and single dots represent trial-specific activation in two columns for the left and right limbs. The different color shades indicate the support conditions. The bars indicate statistically significant differences according to the LMM statistics (details in Additional File 2).

<Please insert Fig. 7.pdf here>

Figure 7: Temporal muscle synergies in the stand-to-sit movement and challenging condition. Top: Exemplary movements in the unassisted, light touch, and full support conditions (in rows) and movement strategies (color-coded). Bottom, left: Temporal muscle synergies sorted chronologically and color-coded according to the movement strategy. The dashed line indicates movement start, i.e., the first identified CoM movement. The dotted line indicates seat-off. Bottom, right: Bar graphs show mean muscle activation vectors across all trials within a strategy and support condition, and single dots represent trial-specific activation in two columns for the left and right limbs. The different color shades indicate the support conditions. The bars indicate statistically significant differences according to the LMM statistics (details in Additional File 2).

Table 3

Full-width at half-maximum (FWHM) of the temporal synergies. Trials are time-normalized to 101 points. Therefore, the numbers in the table represent fractions of the whole trial durations.
Movement	Sit-to-stand, non-challenging			Sit-to-stand, challenging		Stand-to-sit, non-challenging			Stand-to-sit, challenging
Strategy/ synergy	Forward leaning	Hybrid	Vertical rise	Exaggerated forward leaning	Forward leaning	Vertical lowering	Hybrid	Backward lowering	Forward leaning	Exaggerated forward leaning
1	29.1	29.0	25.3		24.4	16.4	15.7	19.4	24.1	24.2
2			9.4	33.2	12.6	23.4	21.3	22.0	19.9	18.7
3	8.7	9.8		11.1	9.0	11.0	13.7	12.7	14.6	14.1
4	9.0	8.4	8.8	8.0	6.3			5.8	7.4	11.3
5	9.5	15.8	18.9	6.7	8.4	8.8	10.8	6.7	6.6
6	13.5			5.8	12.4	10.9	8.4	9.3	14.2	8.9
7	23.1	33.3	32.1	8.2			6.9			8.8
8	22.8	27.3	25.0	7.1		10.2	10.8	11.4	30.9	30.0
9				12.0	25.5	15.1
10				9.6		20.0	28.3	31.2
11				13.1	20.5

The degree of support influences muscle weighting

To test for differences in muscle activation vectors (right-hand columns of Fig. 4, Fig. 5, Fig. 6, and Fig. 7) across the support conditions (UA, LT, FS) with regard to hypothesis 3, that rollator support influences the relative activation between upper-body and lower-body muscle activity, LMMs were calculated. Detailed statistical results with p-values corrected for multiple testing are presented in Additional File 2.

4.1.3 Sit-to-stand movement strategies

In the non-challenging task, muscle activation vectors changed from the start of the movement onwards. The main change was an increase in arm muscle activities in the full support condition compared to unassisted across all movement strategies. Shortly before and at seat-off, muscle activation vectors showed a change with increased arm and decreased lower-limb muscle activations in full support compared to the unassisted condition in the forward leaning and vertical rise strategy. After seat-off, minor changes were only observed in the vertical rise strategy.

More changes were observed in the challenging task than in the non-challenging task. At movement start, there was greater arm muscle activity in the full support condition compared to unassisted for all movement strategies. Shortly before and at seat-off, muscle activation vectors showed increased arm and decreased lower-limb muscle activations in full support compared to the unassisted condition, as in the non-challenging task. After seat-off, we observed decreased activity of the muscles acting around the ankle with light touch and full support, which is a specific finding of the challenging condition.

With regard to the sit-to-stand task, we accept the hypothesis that rollator support influences the relative activation between upper-body and lower-body muscle activity.

4.1.4 Stand-to-sit movement strategies

In the non-challenging task, no significant differences between the support conditions were observed for the backward lowering strategy. However, some changes were observed in the hybrid and vertical lowering strategy: at and after seat-on, the activity of M. pectoralis major, latissimus dorsi, and arm muscles increased in the light touch and full support conditions.

In the challenging task, no significant changes in muscle activation vectors were observed in the forward leaning strategy after the movement began. However, muscle activity vectors changed at seat-on and movement end in the exaggerated forward leaning strategy. Comparable to the non-challenging task, the activity of M. pectoralis major, latissimus dorsi, and arm muscles increased in the light touch and full support conditions. In addition, the activity of M. gluteus medius and tensor fasciae latae decreased with support.

With regard to the sit-to-stand task, we accept the hypothesis that rollator support influences the relative activation between upper-body and lower-body muscle activity.

The study investigated the influence of rollator handle support on movement coordination in sit-to-stand and stand-to-sit movements while considering that people employ different movement strategies. Three support conditions were investigated: no assistance, light touch (supposed proprioceptive cue), and full support (supposed load reduction in the lower limbs). Furthermore, balance pads were placed underneath the young participants’ feet to make the movements challenging by increasing postural instability (non-challenging vs. challenging condition). We hypothesized that (1) temporal muscle synergies represent sit-to-stand and stand-to-sit EMG patterns more compactly than spatial muscle synergies. Furthermore, we hypothesized that (2) temporal muscle synergies differ across movement strategies and that (3) rollator support influences the relative activation between upper-body and lower-body muscle activity.

Temporal muscle synergies represent sit-to-stand and stand-to-sit EMG patterns more compactly than spatial muscle synergies

With six to ten temporal muscle synergies, low-dimensional representations of the sit-to-stand and stand-to-sit movement strategies with 30 EMGs were obtained with R² values between 0.84 and 0.89. The number of synergies found was higher than in studies investigating sit-to-stand movements, in which three to four spatial muscle synergies explained 87–94% of the variance [47–52]. However, in these studies, no upper-body muscle activity was measured, and spatial rather than temporal synergies were extracted. Furthermore, no assistive device was used there. Therefore, comparisons regarding the reconstruction quality are limited. Nevertheless, the main functional groups, i.e., groups of coactivated muscles, identified with spatial synergies in the studies mentioned above largely align with the muscle weightings in our study.

We found that temporal muscle synergies provide a more compact representation than spatial synergies, which is supported by studies of upper-limb and postural tasks [42, 74, 77] and our finding that the change in slope (“knee”) in the R² curve was more pronounced with temporal synergies was also found recently during postural tasks [74]. Spatial and temporal muscle synergies are two of several possible representations of modular control [41]. In general, the CNS might generate motor commands by a few sample-independent modules, which are thought to be shared across tasks and conditions [32] and are activated sample-dependently. The sample-independent part is thought to be stored in subcortical areas [41]. Consequently, with temporal synergies, it is assumed that activation profiles are stored and muscle weightings are sample-dependently composed [92–94]. The good reconstruction quality with temporal synergies hints that the CNS may store activation sequences for sit-to-stand and stand-to-sit movements. The observed differences in the timing of the temporal muscle synergies across the movement strategies further support this.

The activation profiles of the temporal muscle synergies show similar shapes but differ in timing across movement strategies

Across all conditions, the activation profiles of the temporal muscle synergies mostly have a symmetric bell shape with a single point of maximum activation, similar to those found in the literature [75, 95]. The shapes are similar across the movement strategies, with a narrower width at seat contact than at movement start and end. Furthermore, no support-specific synergies emerged with only high lower-limb or upper-limb muscle weightings. However, the timing of the temporal muscle synergies differs across the movement strategies, which supports hypothesis 2. Similarly, Yang et al. (2017) found that their momentum transfer and stabilization movement strategy only differed in the activation timing of the spatial muscle synergy, which was predominantly active around seat-off. This aligns with the activation times of the forward leaning, hybrid, and vertical rise strategies (synergies 2, 3, and 6).

Accordingly, the differences may stem from strategy-specific biomechanical requirements during this phase, e.g. in the forward leaning strategy to transfer the momentum generated by the upper body from the trunk to the thigh to stand up [96–98]. Also, comparing the exaggerated forward leaning to the forward leaning strategy in the sit-to-stand challenging movement, there are distinct activation profiles after seat-off while the body is erected, probably reflecting different biomechanical demands when participants keep their CoM over the base of support for longer in the exaggerated forward leaning strategy [24]. Hence, different movement strategies may emerge through different activation sequences, and rollator support does not provoke a different pattern of impulses.

Temporal synergies reveal that tightness of control is increased in balance-critical phases

A recent study related temporal muscle synergies to the model of intermittent control of movement [74]. According to this model, the CNS sends pulsed commands that are transformed into muscle activation profiles [99]. The muscle activation profiles are adjusted in amplitude (here: sample-dependent muscle activation vectors) and finally result in motor output [100]. Sit-to-stand and stand-to-sit movements consist of multiple phases. Indeed, the literature has identified four to six phases, and even up to eleven[101–103], each of which has different requirements for motor control.

Numerous temporal synergies have been observed in this study, and their peaks are narrow, especially around the balance-critical seat-off and seat-on (see Table 3). Regarding the intermittent control model, these narrow, i.e., tight in time, peaks could reflect that each “control action” has a short duration. Accordingly, the CNS might tightly control the movements using multiple, successive pulsed commands to fulfill the demands of the phases’ different requirements. This tight CNS control is further supported by the increased number of temporal synergies in the challenging sit-to-stand condition, where the balance difficulty was increased. The exaggerated forward leaning strategy revealed three more synergies than the forward leaning strategy in the challenging sit-to-stand task. We previously aligned the former with the “stabilization strategy," suggesting it increases stability and safety when coping with the increased balance challenge [24, 104]. The higher number of temporal synergies found here may reflect this coping mechanism. The CNS might, therefore, be able to increase the tightness of control to enable successful movements. Furthermore, the high number of required narrow synergies (and thus a high demand for CNS control) may be critical for older people or people with disabilities. Although this cannot be addressed here, future studies may investigate whether people with reduced balance capacity in standing up and sitting down movements lack tight CNS control.

Rollator support increases upper body and decreases lower body muscle activity

In support of hypothesis 3, we found that arm muscle activity increased with support across all movement strategies, especially at seat-off and seat-on. Likewise, lower-limb muscle activity decreased with support. In particular, the activity of muscles acting around the ankle decreased with support in the challenging sit-to-stand task.

Generally, these shifts find support in the literature. Suica et al. [105] found that walking with a rollator significantly reduced lower-limb muscle activity in healthy subjects, and Ijmker et al. [106] EMG analysis with stroke survivors showed a drop in lower-limb muscle activity when participants touched or held a handrail while walking on a treadmill. However, in these studies, participants walked rather than performing sit-to-stand or stand-to-sit movements, and only lower limb EMG was measured. Chihara and Seo [107] compared the activity of the anterior deltoid, triceps brachii, rectus femoris, and tibialis anterior in sit-to-stand movements with different handle heights positioned at either side of the participant and found that triceps activity increased while tibialis anterior activity decreased with higher handle height, while the anterior deltoid and rectus femoris were unaffected. This partly supports our findings, yet comparisons are limited as they did not compare muscle activity to unassisted standing up.

Interestingly, the lower limb muscles that often showed reduced activity with increased support in our study, namely rectus femoris, gluteus medius, tibialis anterior, and peroneus longus, are highly relevant for both knee extension and balance [108–111], exemplifying the easing support of the rollator. These results could be very relevant to balance in older adults. Amiridis et al. [112] found that older adults rely more on hip muscles to maintain balance than younger adults in a number of static balance tasks, and a systematic review [113] found that hip abductor strength is critical for balance and to avoid falls. In particular, gluteus medius was among those with decreased activity with support in our study. Upper-body muscle activity, most often in triceps and biceps brachii, as well as latissimus dorsi, increased with support across all movements. Their contributions likely enabled our participants to stand up and sit down with less lower-limb muscle activity. The biceps flexes the elbow, and the latissimus dorsi adducts and medially rotates the humerus at the glenohumeral joints, bringing the body to the arms. The triceps is an elbow extensor [114], assisting in bringing the body upward against gravity and helping to balance the trunk during sit-to-stand [115]. Thus, the rollator-induced decrease in lower-limb muscle activity can be associated with a reduction in leg strength required to fulfill both standing up and sitting down, following the roughly linear EMG/force relationship [116]. Therefore, if less force can or needs to be generated by the lower limb muscles, a rollator can help people with problems standing up.

Limitations

Several potential limitations of this study need to be considered. First, our rollator-simulator device is heavier than a commercially available rollator and cannot dip, which might influence how much people can lean and pull on the rollator handles. Secondly, we restricted the foot placement to be parallel and underneath the knees for standardization purposes, in line with other biomechanical studies [117]. While this is standard practice, this hindered the participants from pulling their feet backward to stand up, which is another everyday common movement strategy while standing up [22]. Thirdly, we included only young people. Even though we used balance pads to challenge them, the generalizability of the results to persons who are older or physically limited and dependent on a rollator may be limited. Nevertheless, Hanawa et al. [49] found that the synergies in standing up remain similar regardless of the participant’s age. Fourthly, the number of participants included in the study is rather low with regards to using LMM statistics, which is due to the time-consuming data collection. Lastly, we observed symmetric activation of the muscles between the right and left limbs. Thus, future studies could focus on unilateral assessment of muscle activity and increase the resolution, e.g., by additionally measuring the hip adductors, M. gluteus maximus, and all deltoid parts.

This study investigated the rollator's influence on the movement coordination for standing up and sitting down, while accounting for different movement strategies. Our temporal muscle synergy analysis found that the timing, but not the shape, of the temporal muscle synergies differs across the movement strategies. Further, we found that the CNS tightly controls standing up and sitting down movements, especially during the balance-critical phases around seat-off and seat-on. Additionally, no support-specific synergies were found, suggesting that the CNS does not need to alter the control if a rollator is used. However, lower-limb muscle activity decreased, and upper-limb muscle activity simultaneously increased with increased support. This leads us to suggest that people struggling with rollator use practice movement strategies requiring less tight CNS control. Further studies should investigate if these findings hold for older and fall-prone individuals, to enable safe and efficient recommendations on rollator usage.

• BF

• musclus biceps femoris

• Bic

• musclus biceps brachii

• CNS

• central nervous system

• CoM

• center of mass

• Del

• musclus deltoideus

• EMG

• electromyography

• ES

• musclus erector spinae

• FS

• full support

• FWHM

• full-width at half-maximum

• GA

• musclus gastrocnemius

• GM

• musclus gluteus medius

• GRF

• ground reaction force

• LMM

• linear mixed model

• LT

• light touch

• Lat

• musclus latissimus dorsi

• MSE

• mean squared residual errors

• NNMF

• non-negative matrix factorization

• PL

• musclus peroneus longus

• Pec

• musclus pectoralis major

• RA

• musclus rectus abdominis

• RF

• musclus rectus femoris

• TA

• musclus tibialis anterior

• TFL

• musclus tensor fasciae latae

• Tra

• musclus trapezius

• Tri

• triceps brachii

• UA

• unassisted

• VM

• musclus vastus medialis

Competing interests

The authors declare that they have no competing interests.

Ethics approval and consent to participate

The Ethics Committee of the Medical Department of Heidelberg University (S-105/2021) approved the study, which was then performed according to the Declaration of Helsinki.

Consent for publication

The participant shown in Fig. 1 gave informed consent to publish the image.

Funding

The research project is funded by the HEiKA – Heidelberg-Karlsruhe strategic partnership, funded by the Federal Ministry of Education and Research (BMBF) and the Baden-Württemberg Ministry of Science as part of the Excellence Strategy of the German Federal and State Governments. It is further part of the projects JuBot and HeiAge, both funded by the Carl-Zeiss-Foundation (Germany). We acknowledge support by the KIT-Publication Fund of the Karlsruhe Institute of Technology.

Author Contribution

MH, FCK, LHS, and TS were involved in the study design. MH and FCK carried out all data collection. MH carried out the data analysis. MH, FCK, DJB, JF, LHS, ADA, and TS were involved in interpreting and discussing the results. MH took the lead in writing the manuscript. MH, FCK, DJB, JF, LHS, ADA, and TS provided critical feedback, revised, and contributed to the final manuscript.

Acknowledgements

Not applicable.

Data Availability

All data generated or analyzed during this study are included in this published article and its supplementary information files.

Rubenstein LZ. Falls in older people: epidemiology, risk factors and strategies for prevention. Age Ageing. 2006;35:ii37–41.
Salari N, Darvishi N, Ahmadipanah M, Shohaimi S, Mohammadi M. Global prevalence of falls in the older adults: a comprehensive systematic review and meta-analysis. J Orthop Surg Res. 2022;17:334.
Ellis AA, Trent RB. Do the risks and consequences of hospitalized fall injuries among older adults in California vary by type of fall? J Gerontol Biol Sci Med Sci. 2001;56:M686–92.
Gaxatte C, Nguyen T, Chourabi F, Salleron J, Pardessus V, Delabrière I, et al. Fear of falling as seen in the Multidisciplinary falls consultation. Annals Phys Rehabilitation Med. 2011;54:248–58.
Bateni H, Maki BE. Assistive devices for balance and mobility: Benefits, demands, and adverse consequences. Arch Phys Med Rehabil. 2005;86:134–45.
Bradley SM, Hernandez CR. Geriatric Assist Devices AFP. 2011;84:405–11.
Ko C-Y, Kim S-B, Chang Y, Choi H-J, Kang S, Heo Y et al. Assessment of forearm and plantar foot load in the elderly using a four-wheeled walker with armrest and the effect of armrest height. CIA. 2014;1759.
Mann WC, Hurren D, Tomita M, Charvat B. An analysis of problems with walkers encountered by elderly persons. Phys Occup Therapy Geriatr. 1995;13:1–23.
Ambrose AF, Paul G, Hausdorff JM. Risk factors for falls among older adults: A review of the literature. Maturitas. 2013;75:51–61.
Kallin K, Jensen J, Olsson LL, Nyberg L, Gustafson Y. Why the elderly fall in residential care facilities, and suggested remedies. J Fam Pract. 2004;53:41–52.
Mundt M, Batista JP, Markert B, Bollheimer C, Laurentius T. Walking with rollator: a systematic review of gait parameters in older persons. Eur Rev Aging Phys Act. 2019;16:15.
Leung C-Y, Yeh P-C. Vertical Force and Wrist Deviation Angle When Using a Walker to Stand UP and Sit down. Percept Mot Skills. 2011;113:229–41.
Dall PM, Kerr A. Frequency of the sit to stand task: An observational study of free-living adults. Appl Ergon. 2010;41:58–61.
Hughes MA, Weiner DK, Schenkman ML, Long RM, Studenski SA. Chair rise strategies in the elderly. Clin Biomech. 1994;9:187–92.
Pai Y-C, Naughton B, Chang R, Rogers M. Control of body centre of mass momentum during sit-to-stand among young and elderly adults. Gait Posture. 1994;2:109–16.
Frykberg GE, Häger CK. Movement analysis of sit-to-stand – research informing clinical practice. Phys Ther Rev. 2015;20:156–67.
Goodpaster BH, Park SW, Harris TB, Kritchevsky SB, Nevitt M, Schwartz AV et al. The Loss of Skeletal Muscle Strength, Mass, and Quality in Older Adults: The Health, Aging and Body Composition Study. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2006;61:1059–64.
Moreland JD, Richardson JA, Goldsmith CH, Clase CM. Muscle Weakness and Falls in Older Adults: A Systematic Review and Meta-Analysis. J Am Geriatr Soc. 2004;52:1121–9.
Joyce BM, Kirby RL. Canes, crutches and walkers. Am Family Phys. 1991;43:535–42.
Jeka JJ. Light touch contact as a balance aid. Phys Ther. 1997;77:476–87.
Krafft FC, Herzog M, Stein T, Sloot LH. Young individuals are more stable and stand more upright when using rollator assistance during standing up and sitting down. Front Bioeng Biotechnol [Internet]. 2022 [cited 2022 Jul 11];10. https://www.frontiersin.org/articles/10.3389/fbioe.2022.876349
Dolecka UE, Ownsworth T, Kuys SS. Comparison of sit-to-stand strategies used by older adults and people living with dementia. Arch Gerontol Geriatr. 2015;60:528–34.
Komaris D-S, Govind C, Murphy A, Ewen A, Riches P. Identification of movement strategies during the sit-to-walk movement in patients with knee osteoarthritis. J Appl Biomech. 2018;34:96–103.
Herzog M, Krafft FC, Stetter BJ, d’Avella A, Sloot LH, Stein T. Rollator usage lets young individuals switch movement strategies in sit-to-stand and stand-to-sit tasks. Sci Rep. 2023;13:16901.
Bernstein NA. The co-ordination and regulation of movements. Oxford: Pergamon; 1967.
Bruton M, O’Dwyer N. Synergies in coordination: a comprehensive overview of neural, computational, and behavioral approaches. J Neurophysiol. 2018;120:2761–74.
Bizzi E, Cheung VCK, d’Avella A, Saltiel P, Tresch M. Combining modules for movement. Brain Res Rev. 2008;57:125–33.
Giszter SF. Motor primitives—new data and future questions. Curr Opin Neurobiol. 2015;33:156–65.
Berniker M, Jarc A, Bizzi E, Tresch MC. Simplified and effective motor control based on muscle synergies to exploit musculoskeletal dynamics. Proceedings of the National Academy of Sciences. 2009;106:7601–6.
Bizzi E, Mussa-Ivaldi FA, Giszter S. Computations underlying the execution of movement: a biological perspective. Science. 1991;253:287–91.
Grillner S. Control of locomotion in bipeds, tetrapods, and fish. In: Brooks VB, editor. Handbook of physiology: section I The nervous system. Bethesda: American Physiological Society; 1981. pp. 1179–236.
d’Avella A, Saltiel P, Bizzi E. Combinations of muscle synergies in the construction of a natural motor behavior. Nat Neurosci. 2003;6:300.
Lee WA. Neuromotor Synergies as a Basis for Coordinated Intentional Action. J Mot Behav. 1984;16:135–70.
Tresch MC, Saltiel P, Bizzi E. The construction of movement by the spinal cord. Nat Neurosci. 1999;2:162–7.
Daffertshofer A, Lamoth CJC, Meijer OG, Beek PJ. PCA in studying coordination and variability: a tutorial. Clin Biomech. 2004;19:415–28.
Santello M, Flanders M, Soechting JF. Postural Hand Synergies for Tool Use. J Neurosci. 1998;18:10105–15.
d’Avella A. Modularity for Motor Control and Motor Learning. In: Laczko J, Latash ML, editors. Progress in Motor Control: Theories and Translations [Internet]. Cham: Springer International Publishing; 2016 [cited 2020 Jul 27]. pp. 3–19. https://doi.org/10.1007/978-3-319-47313-0_1
Ting LH, McKay JL. Neuromechanics of muscle synergies for posture and movement. Curr Opin Neurobiol. 2007;17:622–8.
Ting LH, Chvatal SA. Decomposing muscle activity in motor tasks. Motor Control: Theor Experiments Appl. 2010;102–38.
Berger DJ, Masciullo M, Molinari M, Lacquaniti F, d’Avella A. Does the cerebellum shape the spatiotemporal organization of muscle patterns? Insights from subjects with cerebellar ataxias. J Neurophysiol. 2020;123:1691–710.
Delis I, Panzeri S, Pozzo T, Berret B. A unifying model of concurrent spatial and temporal modularity in muscle activity. J Neurophysiol. 2014;111:675–93.
Russo M, D’Andola M, Portone A, Lacquaniti F, d’Avella A. Dimensionality of joint torques and muscle patterns for reaching. Front Comput Neurosci [Internet]. 2014 [cited 2020 Feb 14];8. https://www.frontiersin.org/articles/10.3389/fncom.2014.00024/full
Bizzi E, Cheung VC. The neural origin of muscle synergies. Front Comput Neurosci [Internet]. 2013 [cited 2020 Feb 14];7. https://www.frontiersin.org/articles/10.3389/fncom.2013.00051/full
Dominici N, Ivanenko YP, Cappellini G, d’Avella A, Mondì V, Cicchese M, et al. Locomotor primitives in newborn babies and their development. Science. 2011;334:997–9.
Safavynia SA, Torres-Oviedo G, Ting LH. Muscle Synergies: Implications for Clinical Evaluation and Rehabilitation of Movement. Top Spinal Cord Inj Rehabil. 2011;17:16–24.
Singh RE, Iqbal K, White G, Hutchinson TE. A Systematic Review on Muscle Synergies: From Building Blocks of Motor Behavior to a Neurorehabilitation Tool. Appl Bionics Biomech. 2018;2018:1–15.
An Q, Ikemoto Y, Asama H. Muscle Synergy Analysis Between Young and Elderly People in Standing-Up Motion. J Robot Mechatron. 2013;25:1038–49.
Carey HD, Liss DJ, Allen JL. Young adults recruit similar motor modules across walking, turning, and chair transfers. Physiol Rep. 2021;9:e15050.
Hanawa H, Kubota K, Kokubun T, Marumo T, Hoshi F, Kobayashi A, et al. Muscle synergies underlying sit-to-stand tasks in elderly people and their relationship with kinetic characteristics. J Electromyogr Kinesiol. 2017;37:15–20.
Kogami H, An Q, Yang N, Wang R, Yoshida K, Hamada H, et al. Analysis of muscle synergy and kinematics in sit-to-stand motion of hemiplegic patients in subacute period. Adv Robot. 2021;35:867–77.
Ranaldi S, Gizzi L, Severini G, Marchis CD. Optimal Identification of Muscle Synergies from Typical Sit-to-stand Clinical Tests. IEEE Open J Eng Med Biology. 2023;1–8.
Yang N, An Q, Yamakawa H, Tamura Y, Yamashita A, Asama H. Muscle synergy structure using different strategies in human standing-up motion. Adv Robot. 2017;31:40–54.
Yoshida K, An Q, Yozu A, Chiba R, Takakusaki K, Yamakawa H, et al. Visual and Vestibular Inputs Affect Muscle Synergies Responsible for Body Extension and Stabilization in Sit-to-Stand Motion. Front Neurosci. 2019;12:1042.
Wang T, Okada S, Guo A, Makikawa M, Shiozawa N. Effect of Assist Robot on Muscle Synergy during Sit-to-Stand Movement. 2021 IEEE International Conference on Intelligence and Safety for Robotics (ISR). 2021. pp. 367–8.
Candow DG, Chilibeck PD. Differences in size, strength, and power of upper and lower body muscle groups in young and older men. J Gerontol Biol Sci Med Sci. 2005;60:148–56.
Lynch NA, Metter EJ, Lindle RS, Fozard JL, Tobin JD, Roy TA et al. Muscle quality. I. Age-associated differences between arm and leg muscle groups. Journal of applied physiology (Bethesda, Md: 1985). 1999;86:188–94.
Ashford S, De Souza L. A comparison of the timing of muscle activity during sitting down compared to standing up. Physiother Res Int. 2000;5:111–28.
Jeon W, Whitall J, Griffin L, Westlake KP. Trunk kinematics and muscle activation patterns during stand-to-sit movement and the relationship with postural stability in aging. Gait Posture. 2021;86:292–8.
Dubost V, Beauchet O, Manckoundia P, Herrmann F, Mourey F. Decreased trunk angular displacement during sitting down: an early feature of aging. Phys Ther. 2005;85:404–12.
Boelen MP. Health professionals’ guide to physical management of parkinson’s disease. Champaign, IL: Human Kinetics; 2009.
Lockette K. A caregiver’s complete guide for safe mobility and independence in the home. Kevin Lockette; 2011.
Wilkins LW. Perfecting clinical procedures. Lippincott Williams & Wilkins; 2008.
Lee DD, Seung HS. Learning the parts of objects by non-negative matrix factorization. Nature. 1999;401:788–91.
Lee DD, Seung HS. Algorithms for Non-negative Matrix Factorization. In: Leen TK, Dietterich TG, Tresp V, editors. Advances in Neural Information Processing Systems 13 [Internet]. MIT Press; 2001 [cited 2020 Mar 5]. pp. 556–62. http://papers.nips.cc/paper/1861-algorithms-for-non-negative-matrix-factorization.pdf
Russo M, Scano A, Brambilla C, d’Avella A, SynergyAnalyzer. A Matlab toolbox implementing mixed-matrix factorization to identify kinematic-muscular synergies. Comput Methods Programs Biomed. 2024;251:108217.
Cappozzo A, Catani F, Della Croce U, Leardini A. Position and orientation in space of bones during movement: anatomical frame definition and determination. Clin Biomech. 1995;10:171–8.
Leardini A, Biagi F, Merlo A, Belvedere C, Benedetti MG. Multi-segment trunk kinematics during locomotion and elementary exercises. Clin Biomech. 2011;26:562–71.
Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol. 2000;10:361–74.
Perotto AO, ANATOMICAL GUIDE FOR, THE ELECTROMYOGRAPHER. The Limbs and Trunk. 5th ed. Springfield: Charles C Thomas; 2011.
Peri E, Xu L, Ciccarelli C, Vandenbussche NL, Xu H, Long X, et al. Singular Value Decomposition for Removal of Cardiac Interference from Trunk Electromyogram. Sensors. 2021;21:573.
Sloot LH, Millard M, Werner C, Mombaur K. Slow but steady: similar sit-to-stand balance at seat-off in older vs. younger adults. Front Sports Act Living. 2020;144.
Dehail P, Bestaven E, Muller F, Mallet A, Robert B, Bourdel-Marchasson I, et al. Kinematic and electromyographic analysis of rising from a chair during a Sit-to-Walk task in elderly subjects: Role of strength. Clin Biomech Elsevier Ltd. 2007;22:1096–103.
Kim MH, Yi CH, Yoo WG, Choi BR. EMG and kinematics analysis of the trunk and lower extremity during the sit-to-stand task while wearing shoes with different heel heights in healthy young women. Hum Mov Sci. 2011;30:596–605.
Brambilla C, Atzori M, Müller H, d’Avella A, Scano A. Spatial and Temporal Muscle Synergies Provide a Dual Characterization of Low-dimensional and Intermittent Control of Upper-limb Movements. Neuroscience. 2023;514:100–22.
Chiovetto E, Berret B, Delis I, Panzeri S, Pozzo T. Investigating reduction of dimensionality during single-joint elbow movements: a case study on muscle synergies. Front Comput Neurosci [Internet]. 2013 [cited 2020 Feb 14];7. http://journal.frontiersin.org/article/10.3389/fncom.2013.00011/abstract
Delis I, Hilt PM, Pozzo T, Panzeri S, Berret B. Deciphering the functional role of spatial and temporal muscle synergies in whole-body movements. Sci Rep. 2018;8:8391.
Safavynia SA, Ting LH. Task-level feedback can explain temporal recruitment of spatially fixed muscle synergies throughout postural perturbations. J Neurophysiol. 2012;107:159–77.
Bach MM, Daffertshofer A, Dominici N. Muscle Synergies in Children Walking and Running on a Treadmill. Front Hum Neurosci [Internet]. 2021 [cited 2024 Mar 7];15. https://www.frontiersin.org/articles/10.3389/fnhum.2021.637157
Muceli S, Boye AT, d’Avella A, Farina D. Identifying Representative Synergy Matrices for Describing Muscular Activation Patterns During Multidirectional Reaching in the Horizontal Plane. J Neurophysiol. 2010;103:1532–42.
d’Avella A, Portone A, Fernandes L, Lacquaniti F. Control of fast-reaching movements by muscle synergy combinations. J Neurosci. 2006;26:7791–810.
Banks CL, Pai MM, McGuirk TE, Fregly BJ, Patten C. Methodological Choices in Muscle Synergy Analysis Impact Differentiation of Physiological Characteristics Following Stroke. Front Comput Neurosci. 2017;11:78.
Tresch MC, Cheung VCK, d’Avella A. Matrix factorization algorithms for the identification of muscle synergies: evaluation on simulated and experimental data sets. J Neurophysiol. 2006;95:2199–212.
Zhao K, Zhang Z, Wen H, Liu B, Li J, Andrea d’Avella, et al. Muscle synergies for evaluating upper limb in clinical applications: A systematic review. Heliyon. 2023;9:e16202.
Ivanenko YP. Coordination of Locomotion with Voluntary Movements in Humans. J Neurosci. 2005;25:7238–53.
van Drongelen S, Stetter BJ, Böhm H, Stief F, Stein T, Meurer A. Identification of patients with similar gait compensating strategies due to unilateral hip osteoarthritis and the effect of total hip replacement: a secondary analysis. JCM. 2021;10:2167.
Aarts E, Verhage M, Veenvliet JV, Dolan CV, Van Der Sluis S. A solution to dependency: using multilevel analysis to accommodate nested data. Nat Neurosci. 2014;17:491–6.
Tirrell TF, Rademaker AW, Lieber RL. Analysis of hierarchical biomechanical data structures using mixed-effects models. J Biomech. 2018;69:34–9.
Twisk JWR, editor. Multivariate Mixed Model Analysis. Applied Mixed Model Analysis: A Practical Guide [Internet]. 2nd ed. Cambridge: Cambridge University Press; 2019 [cited 2024 May 29]. pp. 151–65. https://www.cambridge.org/core/books/applied-mixed-model-analysis/multivariate-mixed-model-analysis/24F30835C3D1A00EA835A8AA9251F27D
Ippersiel P, Preuss R, Fillion A, Jean-Louis J, Woodrow R, Zhang Q, et al. Inter-joint coordination and the flexion-relaxation phenomenon among adults with low back pain during bending. Gait Posture. 2021;85:164–70.
Russell DM, Haworth JL. Walking at the preferred stride frequency maximizes local dynamic stability of knee motion. J Biomech. 2014;47:102–8.
Hox J, Moerbeek M, van de Schoot. Rens. Multilevel Analysis: Techniques and Applications, Third Edition. 3rd ed. London: Routledge; 2017.
Alessandro C, Delis I, Nori F, Panzeri S, Berret B. Muscle synergies in neuroscience and robotics: from input-space to task-space perspectives. Front Comput Neurosci [Internet]. 2013 [cited 2020 Feb 13];7. https://www.frontiersin.org/articles/10.3389/fncom.2013.00043/full
Hart CB. Modular Premotor Drives and Unit Bursts as Primitives for Frog Motor Behaviors. J Neurosci. 2004;24:5269–82.
Hart CB, Giszter SF. A Neural Basis for Motor Primitives in the Spinal Cord. J Neurosci. 2010;30:1322–36.
Torricelli D, De Marchis C, d’Avella A, Tobaruela DN, Barroso FO, Pons JL. Reorganization of Muscle Coordination Underlying Motor Learning in Cycling Tasks. Front Bioeng Biotechnol. 2020;8:800.
Anan M, Ibara T, Kito N, Shinkoda K. The clarification of the strategy during sit-to-stand motion from the standpoint of mechanical energy transfer. J Phys Ther Sci. 2012;24:231–6.
Jeon W, Jensen JL, Griffin L. Muscle activity and balance control during sit-to-stand across symmetric and asymmetric initial foot positions in healthy adults. Gait Posture. 2019;71:138–44.
Riley PO, Schenkman ML, Mann RW, Hodge WA. Mechanics of a constrained chair-rise. J Biomech. 1991;24:77–85.
Leib R, Russo M, d’Avella A, Nisky I. A bang-bang control model predicts the triphasic muscles activity during hand reaching. J Neurophysiol. 2020;124:295–304.
Gross J, Timmermann L, Kujala J, Dirks M, Schmitz F, Salmelin R et al. The neural basis of intermittent motor control in humans. Proceedings of the National Academy of Sciences. 2002;99:2299–302.
Etnyre B, Thomas DQ. Event Standardization of Sit-to-Stand Movements. Phys Ther. 2007;87:1651–66.
Kotake T, Dohi N, Kajiwara T, Sumi N, Koyama Y, Miura T. An analysis of sit-to-stand movements. Arch Phys Med Rehabil. 1993;74:1095–9.
Schenkman M, Berger RA, Riley PO, Mann RW, Hodge WA. Whole-body movements during rising to standing from sitting. Phys Ther. 1990;70:638–48.
Papa E, Cappozzo A. Sit-to-stand motor strategies investigated in able-bodied young and elderly subjects. J Biomech. 2000;33:1113–22.
Suica Z, Romkes J, Tal A, Maguire C. Walking with a four wheeled walker (rollator) significantly reduces EMG lower-limb muscle activity in healthy subjects. J Bodyw Mov Ther. 2016;20:65–73.
IJmker T, Lamoth CJ, Houdijk H, Tolsma M, van der Woude LHV, Daffertshofer A, et al. Effects of handrail hold and light touch on energetics, step parameters, and neuromuscular activity during walking after stroke. JNER. 2015;12:70.
Chihara T, Seo A. Evaluation of multiple muscle loads through multi-objective optimization with prediction of subjective satisfaction level: Illustration by an application to handrail position for standing. Appl Ergon. 2014;45:261–9.
Donath L, Kurz E, Roth R, Zahner L, Faude O. Different ankle muscle coordination patterns and co-activation during quiet stance between young adults and seniors do not change after a bout of high intensity training. BMC Geriatr. 2015;15:19.
Gottschalk F, Kourosh S, Leveau B. The functional anatomy of tensor fasciae latae and gluteus medius and minimus. J Anat. 1989;179–89.
Louwerens JWK, Linge BV, De Klerk LWL, Mulder PGH, Snijders CJ. Peroneus longus and tibialis anterior muscle activity in the stance phase: A quantified electromyographic study of 10 controls and 25 patients with chronic ankle instability. Acta Orthop Scand. 1995;66:517–23.
Salzman A, Torburn L, Perry J. Contribution of rectus femoris and vasti to knee extension. An electromyographic study. Clin Orthop Relat Res. 1993;236–43.
Amiridis IG, Hatzitaki V, Arabatzi F. Age-induced modifications of static postural control in humans. Neurosci Lett. 2003;350:137–40.
Lanza MB, Arbuco B, Ryan AS, Shipper AG, Gray VL, Addison O. Systematic Review of the Importance of Hip Muscle Strength, Activation, and Structure in Balance and Mobility Tasks. Arch Phys Med Rehabil. 2022;103:1651–62.
Kholinne E, Zulkarnain RF, Sun YC, Lim S, Chun J-M, Jeon I-H. The different role of each head of the triceps brachii muscle in elbow extension. Acta Orthop Traumatol Turc. 2018;52:201–5.
Munro BJ, Steele JR. Facilitating the sit-to-stand transfer: a review. Phys Ther Rev. 1998;3:213–24.
Alkner BA, Tesch PA, Berg HE. Quadriceps EMG/force relationship in knee extension and leg press. Med Sci Sports Exerc. 2000;32:459.
van der Kruk E, Silverman AK, Reilly P, Bull AMJ. Compensation due to age-related decline in sit-to-stand and sit-to-walk. J Biomech. 2021;122:110411.

No competing interests reported.

Download PDF

Reviewers agreed at journal
11 Nov, 2024
Reviewers agreed at journal
06 Nov, 2024
Reviewers agreed at journal
05 Nov, 2024
Reviewers agreed at journal
04 Nov, 2024
Reviewers invited by journal
04 Nov, 2024
Editor assigned by journal
05 Sep, 2024
Submission checks completed at journal
05 Sep, 2024
First submitted to journal
04 Sep, 2024

You are reading this latest preprint version

The Central Nervous System Tightly Controls Rollator-Supported Transitions Between Sitting and Standing

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Background

Material & methods

3.1.1 Muscle synergy analysis

3.1.2 Matching of similar synergies across the movement strategies

3.1.3 Statistics

Results

4.1.1 Sit-to-stand movement strategies

4.1.2 Stand-to-sit movement strategies

4.1.3 Sit-to-stand movement strategies

4.1.4 Stand-to-sit movement strategies

Discussion

Conclusion

Abbreviations

Declarations

Competing interests

Funding

Author Contribution

Acknowledgements

Data Availability

References

Additional Declarations

Supplementary Files

Status:

Version 1