The Lower Limb Muscle Co-Activation Map during Human Locomotion: From Slow Walking to Running

central nervous system (CNS) controls movements and regulates joint stiffness with muscle co-activation, but until now, few studies have examined muscle pairs during running. This study aims to investigate differences in lower limb muscle coactivation during gait at different speeds, from walking to running. Nineteen healthy runners walked and ran at speeds ranging from 0.8 km/h to 9.3 km/h. Twelve lower limb muscles’ co-activation was calculated using the time-varying multi- muscle co-activation function (TMCf) with global, flexor–extension, and rostro–caudal approaches. Spatiotemporal and kinematic parameters were also measured. We found that TMCf, spatiotemporal, and kinematic parameters were significantly affected by gait speed for all approaches. Significant differences were observed in the main parameters of each co-activation

km/h to 9.3 km/h. Twelve lower limb muscles’ co-activation was calculated using the time-varying multi- muscle co-activation function (TMCf) with global, flexor–extension, and rostro–caudal approaches. Spatiotemporal and kinematic parameters were also measured. We found that TMCf, spatiotemporal, and kinematic parameters were significantly affected by gait speed for all approaches. Significant differences were observed in the main parameters of each co-activation approach and in the spa- tiotemporal and kinematic parameters at the transition between walking and running. In particular, significant differences were observed in the global co-activation (CI glob, main effectF (1,17)= 641.04 , p< 0.001; at the transitionp< 0.001), the stride length (main effectF (1,17)= 253.03 ,p< 0.001; at the transitionp< 0.001), the stride frequency (main effect F (1,17)= 714.22,p< 0.001; at the transition p< 0.001) and the Center of Mass displacement in the vertical (CoMy, main effectF (1,17)= 426.2 , p< 0.001; at the transitionp< 0.001) and medial–lateral (CoMz, main effectF (1,17)= 120.29 p< 0.001; at the transitionp< 0.001) directions. Regarding the correlation analysis, the CoMywas positively correlated with a higher CI glob(r= 0.88,p< 0.001) and negatively correlated with Full Width at Half Maximum (FWHM glob,r=−0.83,p< 0.001), whereas the CoMzwas positively correlated with the global Center of Activity (CoA glob,r= 0.97,p< 0.001). Positive and negative strong correlations were found between global co-activation parameters and center of mass displacements, as well as some spatiotemporal parameters, regardless of gait speed. Our findings suggest that walking and running have different co-activation patterns and kinematic characteristics, with the whole-limb stiffness exerted more synchronously and stably during running. The co-activation indexes and kinematic parameters could be the result of global co-activation, which is a sensory-control integration process used by the CNS to deal with more demanding and potentially unstable tasks like running. Keywords:muscle co-activations; walking; running; spinal map; sEMG 1. Introduction Muscle co-activation is one of the strategies used by the central nervous system (CNS) to simplify movements and ensure adequate joint stiffness by regulating the time Bioengineering2024,11, 288.

walking; running; spinal map; sEMG 1. Introduction Muscle co-activation is one of the strategies used by the central nervous system (CNS) to simplify movements and ensure adequate joint stiffness by regulating the time Bioengineering2024,11, 288.

Bioengineering2024,11, 288 2 of 21 and amplitude of simultaneous activity of a pair or group of muscles [1–4]. Muscle co- activation is thought to maintain effector-level control (low dimensional), removing the need for individual muscle coordination control (high dimensional) [2]. Although muscle co-activation has been extensively studied during gait in both healthy patients and patients with a variety of motor disorders (for a review, see [2]), only a few stud- ies have looked at muscle co-activation during running [5–9]. Walking to running represents a critical transition time that corresponds to a change in CNS activation [10], leg geome- try [11,12], joint compliance [13], and a robust mechanical energytransformation [14–16] . Indeed, walking and running patterns, despite the similarity, differ in spatiotemporal and kinematic characteristics, such as the lower limb joint angles [17,18]. Most humans will voluntarily switch from walking to running at 6.8 and7.5 km/h [ . At these higher speeds, running becomes less expensive than walking by utilizing a mass-spring mechanism that exchanges kinetic and potential energy in very different ways [15,23] . Most studies on muscle activation during running have concentrated on the activity of single muscles [24–26]. Running has been linked to the increased activation of all lower limb muscles in general [5,27–29], despite the fact that some muscles (e.g., gluteus maximus) appear to play a larger role during running than walking [30]. Only a few studies have looked at the activation of muscle pairs while running [9,31]. When compared to a single-joint muscle solution, these studies discovered that a longer dura- tion of muscle co-activation between the rectus femoris and gastrocnemius during stance provided a better metabolic solution to multiple joint stability, implying that biarticular muscles redistribute mechanical power from proximal joints to distal joints, and ultimately to the ground. Moore et colleagues [9] discovered that co-activation in the distal and leg flexor muscles decreases with speed while running, implying that when both legs are off the ground, muscle co-activation must be reduced to allow for more leg propulsion both upwards and forwards. A critical missing piece is whether the descending motor commands for running are

ultimately to the ground. Moore et colleagues [9] discovered that co-activation in the distal and leg flexor muscles decreases with speed while running, implying that when both legs are off the ground, muscle co-activation must be reduced to allow for more leg propulsion both upwards and forwards. A critical missing piece is whether the descending motor commands for running are set up to coactivate the lower limb muscles as a whole entity within a gross motor strategy. Taking previous evidence into account [9,31], it is possible to hypothesize that descending commands for running cause an increase in whole limb stiffness to stabilize the limb during ground contact on the one hand and a decrease in limb stiffness to facilitate forward progression of the leg during air stepping on the other. This basic descending motor control would imply that flexor muscles are activated during air stepping when whole-limb stiffness is reduced and that muscle co-activation occurs via top-down (rostro– caudal) recruitment (from L3 to S2). This strategy would greatly simplify motor control at the low-dimensional effector level during shock absorption, allowing for motor control decentralization while relying on peripheral feedback via muscle reflexes to generate the required muscle co-activations and regulate each muscle’s gain. Muscle co-activation patterns are determined by the supraspinal processes involved in movement control [2].Two main circuits, corticocerebellar–thalamo–cortical and corticobasal– thalamo–cortical, appear to be important in defining co-activation patterns [32]. In particu- lar, studies of subjects with cerebellar degeneration [33], spasticity [34], or extrapyramidal rigidity [35] have found that the cerebellum plays a role in modulating muscle co-activation, implying that processing modules distributed within the brain define the patterns of co- activation observed in the extremities [36]. We recently proposed a novel approach to studying time-varying multi-muscle co- activation function (TMCf) [37,38], which is a good indicator of the CNS’s overall strategy for modulating the simultaneous activation of many lower limb muscles during locomotion. This approach provides a new perspective on the spatiotemporal motor control of the lower limbs, emphasizing how all muscles of the lower limb are synchronously coactivated in an attempt to increase whole-limb stiffness, regardless

activation function (TMCf) [37,38], which is a good indicator of the CNS’s overall strategy for modulating the simultaneous activation of many lower limb muscles during locomotion. This approach provides a new perspective on the spatiotemporal motor control of the lower limbs, emphasizing how all muscles of the lower limb are synchronously coactivated in an attempt to increase whole-limb stiffness, regardless of single-joint antagonist muscles or modular activation of a group of muscles [39]. The objectives of this study were to observe the differences in lower limb muscle co-activation strategies during locomotion across speeds from walking to running. We

Bioengineering2024,11, 288 3 of 21 hypothesized that the behavior of lower limb muscles, in terms of co-activation parameters, may differ across the speeds, particularly between walking and running, and that the co-activation parameters may vary based on the muscular functions (flexion or extensor functions) and across the rostro–caudal recruitment map. The findings of this study may aid clinicians to better understand the mechanisms of muscular behavior during the transition from walking to running, potentially provid- ing information on control abnormalities that lead to muscle injuries, as well as sport professionals in developing tailored training programs. 2. Materials and Methods 2.1. Subjects Nineteen healthy runners were recruited (9 men and 10 women; mean age of 40.95±7.96 years; mean weight of 66.47±14.60 kg). Each runner had declared run- ning for at least 5 years, running at least 3 times per week for at least 5 km per training session [40]. None of the runners had any known diseases that might affect their regular gait or running pattern. This study was authorized by the local ethics committee (N. 0078009/2021) after all participants gave written informed consent and the study’s design adhered to the Declaration of Helsinki. The number of the included subjects was in the range of subjects usually enrolled in similar studies [41–43]. Moreover, a priori power analysis using the G∗Power computer program [44] indicated that a total sample of 14 participants would be needed to detect a medium effect size (Cohen’s F = 0.25) with 80% power using the repeated measurement ANOVA with gait speed as a within-subjects factor with a = 0.05. 2.2. Experimental Procedure Each runner was asked to walk at thirteen various speeds, ranging from 0.8 km/h to 6.8 km/h, and run at five different speeds, ranging from 7.3 km/h to 9.3 km/h, with increased steps of 0.5 km/h, on a treadmill. We chose 7.3 km/h as the transition speed because it is consistent with earlier research [19,21,22] in which this transition was deter- mined considering the metabolic energy cost of locomotion, as well as the human capacity for purposeful gait modulation and the importance of physiologic and

7.3 km/h to 9.3 km/h, with increased steps of 0.5 km/h, on a treadmill. We chose 7.3 km/h as the transition speed because it is consistent with earlier research [19,21,22] in which this transition was deter- mined considering the metabolic energy cost of locomotion, as well as the human capacity for purposeful gait modulation and the importance of physiologic and metabolic demands. Before the recording session, individual speed performance schedules were arranged for each participant, featuring randomized presentations. Each runner underwent a 10 s adaptation period to familiarize with the given speed, followed by a 30 s recording period. Following each trial, the speed was reset to zero, and the runner was granted a minimum of 1 min of stationary rest to prevent fatigue before commencing the subsequent trial. Each trial was performed for 30 s before moving on to the next speed that was chosen at random. 2.3. Data Acquisition A bipolar 16-channel wireless acquisition device (Mini Wave System; Cometa, Bareg- gio, Milan, Italy) was used to record all the surface myoelectric activity at a sampling rate of 2 kHz. Twelve Ag/AgCl pre-gelled electrodes (Kendall ARBO, inter-electrode distance: 2 cm) were placed over each participant’s right side lower limb muscles over the following muscles: gluteus medius (GM), rectus femoris (RF), vastus lateralis (VL), vastus medialis (VM), tensor fascia latae (TFL), semitendinosus (ST), biceps femoris (BF), tibialis anterior (TA), gastrocnemius medialis (GasM), gastrocnemius lateralis (GasL), soleus (SOL), and peroneus longus (P) [45,46], following the European Recommendations for Surface Elec- tromyography [47], the Atlas of Muscle Innervation Zones [48], and best practices [45,46,49]. Before applying the electrodes, the skin was prepared by shaving, if needed, and cleaned with alcohol and dried. A stereo-photogrammetric motion analysis system with optoelectronic technology (SMART-DX 6000 system: BTS, Bari, Italy) and with a sampling rate of 340 Hz was employed to collect the kinematics data. Eight infrared cameras were used to record five passive spher-

Bioengineering2024,11, 288 4 of 21 ical markers, covered with aluminum powder, positioned above the sacrum and bilaterally on the anterior superior iliac spines, heel, metatarsal head, and lateral malleoli [50]. A video camera (BTS Vixta; BTS, Milan, Italy) with a frame rate of 25–30 frames per second and a video resolution of 640×480 pixels was used to visually monitor the acquired tasks. All the data collected were synchronized. 2.4. Data Analysis Three-dimensional reconstruction software (SMART Tracker and SMART Analyzer: BTS, Italy) and MATLAB (R2019b 9.7; MathWorks, Portola Valley, CA, USA) were used to process the sEMG and kinematics data. 2.4.1. Cycle Definition and Temporal Normalization The heel strike (HS) and toe-off (TO) events were determined in this study along the anteroposterior trajectories through the maximum point of the heel and the minimum point of the metatarsal, respectively. The gait and run cycle was defined as the time between two consecutive HSs of the same leg. Then, the electromyographic and kinematic data were time-normalized to the duration of each cycle and reduced to 201 samples using an interpolation procedure to allow a comparison between different cycles that had different durations [31,37–39]. Ten cycles for each gait speed level for each runner were analyzed. 2.4.2. Global, Flexor, Extensor and Rostro–Caudal Co-Activation of Lower Limb Muscles The raw sEMG signals were visually reviewed to remove any artifact-containing cycles. They were then band-pass filtered with a zero-lag fifth-order Butterworth(20–450 Hz) [ to keep only the signal of interest and then full-wave rectified and low-pass filtered with a zero-lag fifth-order Butterworth (10 Hz) [53]. The elaborated sEMG signals [54,55] of each muscle were amplitude-normalized (0–100%) for each runner in relation to the mean of their three highest peak values detected across all gait cycles and velocities [33,56]. The TMCf was used to calculate the simultaneous activation of the lower-limb muscles based on the processed sEMG signals [3,33,37,38,57–59]. The full-wave-rectified, low-pass- filtered, and 0–100% amplitude normalized sEMG signals were used as inputs to this sigmoid-weighted time-dependent co-activation function for the inclusion of multiple muscles during walking and running. This co-activation function’s values ranged from 0 to

velocities [33,56]. The TMCf was used to calculate the simultaneous activation of the lower-limb muscles based on the processed sEMG signals [3,33,37,38,57–59]. The full-wave-rectified, low-pass- filtered, and 0–100% amplitude normalized sEMG signals were used as inputs to this sigmoid-weighted time-dependent co-activation function for the inclusion of multiple muscles during walking and running. This co-activation function’s values ranged from 0 to 100%, and they were calculated as follows: TMC f(d(i),i) =C(d(t))· (∑ M m=1 EMGm(i)/M) 2 maxm=1...M[EMGm(i)] = ȷ 1− 1 1+e −a(d(i)−b) ff · (∑ M m=1 EMGm(i)/M) 2 maxm=1...M[EMGm(i)] (1) whereC(d(t)) is the sigmoid weight reduction coefficient,Mis the number of muscles considered, andEMGm(i) is the sEMG sample value of them-thmuscle at instanti. The functionC(d(t)) , ranging between 0 and 1, takes into account, within the exponential function, the constantsaandbequal to 12 and 0.5 respectively [37,57].d(i) is the mean of the differences between each pair among the 12 muscles (EMGm(i)) values at instanti: d(i) = ∑ M−1 m=1 ∑ M n=m+1 |EMGm(i)−EMGn(i)| L(M!/(2!(M−2)!)) ! = ∑ M−1 m=1 ∑ M n=m+1 |EMGm(i)−EMGn(i)| L(M(M−1)/2) ! (2) whereLis the length of the signal andM!/(2(M−2)!) is the total number of possible differences between each pair ofEMGm(i). TMC f(d(i) ,i)has the following properties: an inverse relationship with the mean of the differencesd(i), i.e., values close to the mean activation of them(i)muscle sample values considered whend(i)is close to 0, and values close to 0 whend(i)is close to 1. The smaller the differences in muscle activations are, the closer thed(i)values are to 0; the closer the sigmoid-coefficient values are to 1, the closer theTMC f(d(i),i) value is to its mean value. In contrast, the greater the differences in muscle activations, and the higher thed(i)and the lower the sigmoid coefficient, the lower theTMC f(d(i),i) values will be.

Bioengineering2024,11, 288 5 of 21 Data over individual strides was calculated for each runner and speed and then averaged across cycles. All the acquired muscles were inserted in the calculation of the TMCf to assess global co-activation (TMCf glob). The co-activation of extensor (TMCfext) and flexor (TMCf flex) muscles were computed (Table) by considering the extensor and flexor muscle subgroups made up according to the “Concentric function”; it indicates that the flexor and extensor subgroups include those muscles which contract concentrically (shortening) during the flexion and extension of the joint [60]. The biarticular muscles were considered flexors or extensors based on their proximal function [61,62]. Furthermore, we calculated the TMCf by separating the muscles on the basis of their spinal segment of innervation (Table): first, all the muscles innervated by L3, then those by L4, then those by L5, then the one by S1, and finally those innervated by S2; thus rostro–caudal organization (TMCfL3; TMCfL4; TMCfL5; TMCfS1; TMCfS2) [41,60,63,64] was assessed using subgroups of muscles (see Table). Muscles were considered flexors or extensors based on their concentric function in the sagittal plane [60]. Table 1.Each dot in the table indicates muscles included in the time-varying co-activation (TMCf) function for each muscle co-activation investigated: global (all the monitored lower limb muscles), extensor (extensor muscle subgroup), flexor (flexor muscle subgroup), and rostro–caudal organization (muscles on the basis of their spinal segment of innervation, see text for further details). Smallest dots indicate a halved weight (amplitude of muscle activity multiplied by 0.5) for that specific muscle in the TMCf function. Maps Global Extensor Flexor L3 L4 L5 S1 S2 Muscles GMBioengineering 2024, 11, x FOR PEER REVIEW 5 of 22 í µí±‘(í µí±–)= F ∑∑ | í µí°¸í µí±€í µí°º à(í µí±–)âˆ’í µí°¸í µí±€í µí°º á(í µí±–)| Æ á @ à > 5 Æ ? 5 à @ 5 í µí°¿(í µí±€!/(2!(í µí±€âˆ’2)!)) G= F ∑∑ | í µí°¸í µí±€í µí°º à(í µí±–)âˆ’í µí°¸í µí±€í µí°º á(í µí±–)| Æ á @ à > 5 Æ ? 5 à @ 5 í µí°¿(í µí±€(í µí±€ − 1)/2) G (2) where í µí°¿ is the

µí±€í µí°º á(í µí±–)| Æ á @ à > 5 Æ ? 5 à @ 5 í µí°¿(í µí±€!/(2!(í µí±€âˆ’2)!)) G= F ∑∑ | í µí°¸í µí±€í µí°º à(í µí±–)âˆ’í µí°¸í µí±€í µí°º á(í µí±–)| Æ á @ à > 5 Æ ? 5 à @ 5 í µí°¿(í µí±€(í µí±€ − 1)/2) G (2) where í µí°¿ is the length of the signal and í µí±€! (2(í µí±€âˆ’2)!)⁄ is the total number of possible differences between each pair of í µí°¸í µí±€í µí°º à(í µí±–). í µí±‡í µí±€í µí°¶í µí±“(í µí±‘(í µí±–),í µí±–) has the following properties: an inverse relationship with the mean of the differences d(i), i.e., values close to the mean activation of the í µí±š(í µí±–) muscle sample values considered when í µí±‘(í µí±–) is close to 0, and values close to 0 when í µí±‘(í µí±–) is close to 1. The smaller the differences in muscle activations are, the closer the í µí±‘(í µí±–) values are to 0; the closer the sigmoid-coefficient values are to 1, the closer the í µí±‡í µí±€í µí°¶í µí±“(í µí±‘(í µí±–),í µí±–) value is to its mean value. In contrast, the greater the differences in muscle activations, and the higher the í µí±‘(í µí±–) and the lower the sigmoid coefficient, the lower the í µí±‡í µí±€í µí°¶í µí±“(í µí±‘(í µí±–),í µí±–) values will be. Data over individual strides was calculated for each runner and speed and then aver- aged across cycles. All the acquired muscles were inserted in the calculation of the TMCf to assess global co-activation (TMCf glob). The co-activation of extensor (TMCfext) and flexor (TMCf flex) mus- cles were computed (Table 1) by considering the extensor and flexor muscle subgroups made up according to the “Concentric function”; it indicates that the flexor and extensor subgroups include those muscles which contract concentrically (shortening) during the flexion and extension of the joint [60]. The biarticular muscles were considered flexors or extensors based on their proximal function [61,62]. Furthermore, we calculated the TMCf by separating the muscles on the basis of their spinal segment of innervation (Table 1): first, all the

Examines muscle co-activation patterns in runners transitioning from walking to running.