Effects of Different Wearable Resistance Placements on Running Stability

tability during running has been recognized as a crucial factor contributing to running performance. This study aimed to investigate the effects of wearable equipment containing external loads on different body parts on running stability. Fifteen recreational male runners (20.27±1.23 years, age range 19–22 years) participated in five treadmill running conditions, including running without loads and running with loads equivalent to 10% of individual body weight placed on four different body positions: forearms, lower legs, trunk, and a combination of all three (forearms, lower legs, and trunk). A tri-axial accelerometer-based smartphone sensor was attached to the participants’ lumbar spine (L5) to record body accelerations. The largest Lyapunov exponent (LyE) was applied to individual acceleration data as a measure of local dynamic stability, where higher LyE values suggest lower stability. The effects of load distribution appear in the mediolateral (ML) direction. Specifically, running with loads on the lower legs resulted in a lower LyE_ML value compared to running without loads (p= 0.001) and running with loads on the forearms (p< 0.001), trunk (p= 0.001), and combined segments (p= 0.005). These findings suggest that running with loads on the lower legs enhances side-to-side local dynamic stability, providing valuable insights for training. Keywords:treadmill running; weight vest; wearable resistance training; recreational runners; running gait; running stability; local dynamic stability; largest Lyapunov exponent; smartphone-based accelerometry 1. Introduction Wearable resistance training is a method that enables individuals to

0.001), and combined segments (p= 0.005). These findings suggest that running with loads on the lower legs enhances side-to-side local dynamic stability, providing valuable insights for training. Keywords:treadmill running; weight vest; wearable resistance training; recreational runners; running gait; running stability; local dynamic stability; largest Lyapunov exponent; smartphone-based accelerometry 1. Introduction Wearable resistance training is a method that enables individuals to attach gear or garments with additional loading to various body parts [1]. Widely employed in athletic training, it aims to find optimal loads that provide effective resistance training without negatively impacting sporting technique, i.e., to explore suitable loads that allow move- ments to occur without unintentionally impacting the technical execution of the action [1]. This method has been applicable for various purposes, e.g., warm-ups [2], training [3], or sports activities [4]. In the context of running, wearable resistance training involves intentionally adding external resistance or weight to running exercises [1,5,6]. This results in greater ground reaction forces and increased power production and velocity during sprint running [4]. Runners employ various load-bearing strategies, e.g., weighted vests, forearm cuffs, or lower leg cuffs [6]. Trunk loading, exemplified by wearing weighted vests, allows an overload to be uniformly distributed close to a person’s center of gravity, potentially enhancing the capacity to generate higher ground response forces and power [7]. Conversely, loads applied to the distal segments of the limb are usually positioned near their end, increasing the moment of inertia and, consequently, the amount of muscular activity needed [8]. Despite these advantages, it remains unclear whether the placement of such equipment on different body parts affects running stability. Stability in locomotion, which refers to an individual’s capacity to maintain balance and control while in motion, is widely recognized as the inherent ability of the motor control Sports2024,12, 45.

Sports2024,12, 45 2 of 12 system to preserve or return to its initial state even when faced with internal factors (e.g., neuro-muscular aspects) and external perturbations (e.g., environmental) [9–11]. Measures of stability offer valuable insights into understanding the inherent variability in motor task performance, allowing for the direct quantification of dynamic error correction [9–11] since human movement is accepted to arise from non-linear interactions among various neuromuscular elements influenced by both internal and external factors [9,10]. In this sense, a non-linear analysis of human movement has been suggested [12]. For instance, the largest Lyapunov exponent (LyE), a method used to assess local dynamic stability, quanti- fies the exponential rate of divergence of trajectories within the state space, specifically of kinematic data acquired from gait [13,14]. In the context of locomotion stability, LyE serves to evaluate the capacity of the neuromuscular system to adapt and manage infinitesimal perturbations to sustain functional locomotion [10,15,16]. A higher LyE indicates greater complexity and unpredictability, signifying that minor differences in initial conditions lead to significant variations in movement patterns over time [14]. Conversely, a lower LyE suggests more consistent and predictable movements, where minor initial differences result in minimal variations over time [14]. Therefore, stability during locomotion encompasses a multifaceted concept, including the body’s ability to maintain balance, reduce deviations, and control movement effectively [17–19]. Less stability during running has been recog- nized as one of the major contributors to decreased running performance and increased risks of injury [20]. It is widely accepted that every aspect of human movement involves acceleration, primarily driven by muscle actions [21,22]. In maintaining a stable running posture, body accelerations result from a combination of muscle actions [23] and the intricate interplay between the neuromuscular system and other factors, such as gravity, ground reaction forces, friction, air resistance, and external forces [24]. Together, these components modulate muscle-activation levels to provide the appropriate acceleration [25]. These body accelerations essentially reflect the ability of the sensorimotor system to control the body’s motion and preserve stability throughout this dynamic activity, and they are crucial to understanding the mechanics of running [17,18]. Comprehending this

as gravity, ground reaction forces, friction, air resistance, and external forces [24]. Together, these components modulate muscle-activation levels to provide the appropriate acceleration [25]. These body accelerations essentially reflect the ability of the sensorimotor system to control the body’s motion and preserve stability throughout this dynamic activity, and they are crucial to understanding the mechanics of running [17,18]. Comprehending this idea via the application of the Lyapunov exponent (LyE) to body acceleration, as shown in earlier review studies [13,26], might be very important for running-related training, injury prevention, and rehabilitation initiatives. When considering external forces, accelerations can inadvertently manifest, especially in the presence of wearable resistance equipment, challenging running stability. Furthermore, carrying weights on different body parts may necessitate posture adjustments and enhanced neuromuscular control to preserve stability while running. In summary, understanding how different load distribution strategies impact running stability may be pivotal for optimizing training regimens and enhancing overall running performance. Hence, the current study aimed to investigate how load distribution across various body parts—whether focused on adding the load to the trunk, lower legs, forearms, or combinations of these segments—affects local dynamic running stability, as quantified using the Lyapunov exponent (LyE) based on body acceleration. Given that applying external loads during running can lead to altered movement patterns and magnitudes [1,5,6], the hypothesis was that differences in running stability would be observed between running with no load and running with loads on different body parts. 2. Materials and Methods 2.1. Participants Fifteen recreational male runners (age range: 19–22 years) with running exercise regularly at least 3 days/week and good experience with treadmill running participated in the study; their characteristics are represented in Table. All participants were university students with a normal body mass index (BMI) who had self-reported no neurological or musculoskeletal problems within the last six months, no medical conditions (e.g., diabetes, high blood pressure, heart disease, or other diseases), and no experience with wearable

Sports2024,12, 45 3 of 12 resistance training. In addition, runners who had wounds or pains or had a history of consuming alcoholic beverages or energy drinks within 24 h before testing were not eligible for the study. All participants were selected from the invitations via electronic flyers through online platforms and then screened through personal contact in order to provide experimental information for self-preparation before performing the running tests. Table 1.Characteristics of participants (mean±SD). Max Min Mean SD Age (years) 22.0 19.0 20.3 1.0 Mass (kg) 77.0 53.0 64.7 6.5 Height (cm) 180.0 163.0 170.3 5.6 Body mass index (kg/m 2 ) 26.6 19.2 22.3 2.1 Weekly mileage (km) 40.0 5.0 20.1 11.1 In order to calculate the sample size of the current study, a priori power analysis through the G*Power software version 3.1.9.4 (Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany) [27] was used based on a previous preliminary study that assessed the effects of wearable resistance placement by measuring acceleration data [27], yielding an average effect size for comparisons between the placement conditions of 0.42. Based on this computation, with a significance level of a = 0.05 and a desired power of 0.95, the suggested sample size was N = 14. However, fifteen young adults volunteered to participate in the current study. The current experimental procedures were approved by the Institutional Review Board of the University of Phayao, Thailand (Approval Code No.: HREC-UP-HSST 1.3/038/66, Approval Date: 20 August 2023) and conducted in accordance with the Declaration of Helsinki. All volunteers provided written, informed consent before participating. 2.2. Equipment and Experimental Procedure On the day before the experiment, each participant was asked to complete a question- naire to self-check their health status and was informed to prepare appropriate clothing and shoes for running, obtain enough rest (at least 6–8 h), eat a meal at least 2–3 h before the test, refrain from consuming alcohol or energy drinks for at least 24 h, and avoid strenuous exercise for at least 24 h. On the experimental day, all participants were asked to check their blood pressure and be screened by the researcher for any wounds

obtain enough rest (at least 6–8 h), eat a meal at least 2–3 h before the test, refrain from consuming alcohol or energy drinks for at least 24 h, and avoid strenuous exercise for at least 24 h. On the experimental day, all participants were asked to check their blood pressure and be screened by the researcher for any wounds or injuries. In the current study, ref. [28] a tri-axial accelerometry sensor embedded in a smartphone (Samsung Galaxy A52s 5G, Samsung Electronics Co., Ltd., Suwon, Republic of Korea) was employed to measure trunk acceleration [28]. As previously reviewed, smartphone-based accelerometry has been accepted as an alternative, valid, and reliable measure of gait and postural control [28,29]. The smartphone was securely positioned using a waist-mounted pouch in the lumbar region (L5), which was close to the body’s center of mass [30,31]. The accelerometer functioned using the Physics Toolbox Sensor Suite application (version 2023.01.07) on the Google Android platform [32,33], enabling the collection and export of acceleration data at a sampling rate of 200 Hz. All participants wore their running shoes for all study runs. Minimalist running shoes were not allowed due to their effects on increased vertical loading rates compared to cushioned running shoes [34]. All participants began the study with a warm-up consisting of a 5 min brisk walk on a Brightway TT-X10 treadmill (Shandong Brightway Fitness Equipment Co., Ltd., Shandong, China) at a speed of 5.5 km/h and a 5 min whole body stretching [35], and then rested for 5 min before the first experiment ran. Subsequently, volunteers were asked to complete five treadmill running trials in random order: no-load running and running with a load equivalent to 10% of their body weight placed on their lower arms, lower legs, trunk, and a combination of forearms, lower legs, and trunk. The choice of a 10% load of individual body weight was based on its effectiveness in improving running performance, as reported

legs, trunk, and a combination of forearms, lower legs, and trunk. The choice of a 10% load of individual body weight was based on its effectiveness in improving running performance, as reported

Sports2024,12, 45 4 of 12 in the systematic reviews [1,6,8]. The load was applied using a weighted vest, forearm cuffs, and lower leg cuffs inserted via the detachable metal plate (Figure). In conditions where the load was distributed across the forearms, lower legs, and combined segments, equal weight was allocated to each segment. The weight distribution was symmetrical for the anterior and posterior sides of the weight vest and evenly distributed for the forearm and lower leg cuffs.Sports 2024, 12, x FOR PEER REVIEW 4 of 12 a load equivalent to 10% of their body weight placed on their lower arms, lower legs, trunk, and a combination of forearms, lower legs, and trunk. The choice of a 10% load of individual body weight was based on its effectiveness in improving running performance, as reported in the systematic reviews [1,6,8]. The load was applied using a weighted vest, forearm cuffs, and lower leg cuffs inserted via the detachable metal plate (Figure 1). In conditions where the load was distributed across the forearms, lower legs, and combined segments, equal weight was allocated to each segment. The weight distribution was sym- metrical for the anterior and posterior sides of the weight vest and evenly distributed for the forearm and lower leg cuffs. During testing, if the participant had any abnormal symptoms during the test, e.g., dizziness, nausea, vomiting, pain, or accidents, it would be considered a withdrawal of participant criteria, and they could also request to terminate their participation if they felt unsafe or did not wish to participate in the research. Figure 1. Wearable resistance equipment ((A): a weighted vest, (B): forearm cuffs, (C): lower leg cuffs, and (D): detachable metal plates). For each running condition, the treadmill’s speed gradually increased from 0 to 10 km/h within 1 min, remained constant for 3 min, and gradually decreased from 10 to 0 km/h within 1 min. The selected running speed of 10 km/h is closely related to the pre- ferred running speed of recreational runners [18]. Each test consisted of 5 min of running per condition, with participants allowed a 5 min rest

0 to 10 km/h within 1 min, remained constant for 3 min, and gradually decreased from 10 to 0 km/h within 1 min. The selected running speed of 10 km/h is closely related to the pre- ferred running speed of recreational runners [18]. Each test consisted of 5 min of running per condition, with participants allowed a 5 min rest period after each test. Borg’s rating of perceived exertion (RPE) scale was used to measure perceived exertion or exercise in- tensity at the baseline period (before performing experiments) and after each running test [36,37]. The Borg RPE scale used in this study ranges from 6 to 20, with 6 representing very light exertion and 20 representing maximum exertion or exhaustion [36,37]. Since the scale is subjective, individuals rate their own perceived exertion based on their feelings of effort, fatigue, and other sensations during exercise [36,37]. In addition, before starting any running trial, all volunteers were asked to check their readiness; if the 5 min rest was not enough, they could stay for a longer rest. All participants a 4ended one testing session in a temperature-controlled (25 °C) laboratory. The acceleration signals during the 3 min treadmill run for each running condition were recorded using a Samsung Galaxy Tab S6 Lite tablet (manufactured by Samsung Electronics Co., Ltd., Suwon, Republic of Korea) to operate the Physics Toolbox Sensor Suite application installed in the smartphone using the Samsung Flow application version 4.9.08.3 (Samsung Electronics Co., Ltd., Suwon, Republic of Korea). 2.3. Data Analysis MATLAB TM (MathWorks Inc., Natick, MA, USA) was used for all data processing. A Fourier analysis was applied to the raw acceleration signals, revealing that the highest power was concentrated in frequencies around 5–10 Hz, with visible power still present in the 15–20 Hz frequency range. Consequently, these signals underwent smoothing through a 4th-order zero-phase 20 Hz low-pass Bu 4erworth filter, similar to that previ- ously reported [38]. For the analysis of acceleration-based variables, the middle two Figure 1.Wearable resistance equipment ((A): a weighted vest, (B): forearm cuffs, (C): lower leg cuffs, and (D): detachable metal plates). During testing,

present in the 15–20 Hz frequency range. Consequently, these signals underwent smoothing through a 4th-order zero-phase 20 Hz low-pass Bu 4erworth filter, similar to that previ- ously reported [38]. For the analysis of acceleration-based variables, the middle two Figure 1.Wearable resistance equipment ((A): a weighted vest, (B): forearm cuffs, (C): lower leg cuffs, and (D): detachable metal plates). During testing, if the participant had any abnormal symptoms during the test, e.g., dizziness, nausea, vomiting, pain, or accidents, it would be considered a withdrawal of participant criteria, and they could also request to terminate their participation if they felt unsafe or did not wish to participate in the research. For each running condition, the treadmill’s speed gradually increased from 0 to 10 km/h within 1 min, remained constant for 3 min, and gradually decreased from 10 to 0 km/h within 1 min. The selected running speed of 10 km/h is closely related to the preferred running speed of recreational runners [18]. Each test consisted of 5 min of running per condition, with participants allowed a 5 min rest period after each test. Borg’s rating of perceived exertion (RPE) scale was used to measure perceived exertion or exercise intensity at the baseline period (before performing experiments) and after each running test [36,37]. The Borg RPE scale used in this study ranges from 6 to 20, with 6 representing very light exertion and 20 representing maximum exertion or exhaustion [36,37]. Since the scale is subjective, individuals rate their own perceived exertion based on their feelings of effort, fatigue, and other sensations during exercise [36,37]. In addition, before starting any running trial, all volunteers were asked to check their readiness; if the 5 min rest was not enough, they could stay for a longer rest. All participants attended one testing session in a temperature-controlled (25 ◦ C) laboratory. The acceleration signals during the 3 min treadmill run for each running condition were recorded using a Samsung Galaxy Tab S6 Lite tablet (manufactured by Samsung Electronics Co., Ltd., Suwon, Republic of Korea) to operate the Physics Toolbox Sensor Suite application installed in the smartphone using

rest. All participants attended one testing session in a temperature-controlled (25 ◦ C) laboratory. The acceleration signals during the 3 min treadmill run for each running condition were recorded using a Samsung Galaxy Tab S6 Lite tablet (manufactured by Samsung Electronics Co., Ltd., Suwon, Republic of Korea) to operate the Physics Toolbox Sensor Suite application installed in the smartphone using the Samsung Flow application version 4.9.08.3 (Samsung Electronics Co., Ltd., Suwon, Republic of Korea). 2.3. Data Analysis MATLAB TM (MathWorks Inc., Natick, MA, USA) was used for all data processing. A Fourier analysis was applied to the raw acceleration signals, revealing that the highest power was concentrated in frequencies around 5–10 Hz, with visible power still present in the 15–20 Hz frequency range. Consequently, these signals underwent smoothing through a 4th-order zero-phase 20 Hz low-pass Butterworth filter, similar to that previously reported [38]. For the analysis of acceleration-based variables, the middle two minutes of each acceleration signal (anteroposterior (AP), mediolateral (ML), and vertical (VT) accelerations) were selected to exclude any movements associated with adjustments to the desired running speed. Figure smoothed tri-axial acceleration data during running with each condition.

Sports2024,12, 45 5 of 12Sports 2024, 12, x FOR PEER REVIEW 5 of 12 minutes of each acceleration signal (anteroposterior (AP), mediolateral (ML), and vertical (VT) accelerations) were selected to exclude any movements associated with adjustments to the desired running speed. Figure 2 provides an example of visual representations of the smoothed tri-axial acceleration data during running with each condition. Figure 2. Example tri-axial acceleration data of the treadmill running with (A) no load, (B) and run- ning with loads on combined (forearms, legs, and trunk) segments, (C) forearms, (D) lower legs, and (E) trunk, respectively. Note: the presented middle 2 min tri-axial acceleration data were re- trieved from each running condition of the first participant. Then, two acceleration-based variables were computed for each acceleration signal. First, the largest Lyapunov exponent (LyE) was applied to individual acceleration data [14] to investigate running stability by calculating the rate of divergence of closely related trajectories in a state space during locomotion [17–19,39]. This metric offers insights into the motor system’s ability to a 4enuate minor perturbations, and LyE was computed using Wolf’s algorithm [40]. The parameters for LyE calculation, including time delay ( ì = 10) and embedding dimension ( I = 4), were determined using the average mutual infor- mation (AMI) and the false nearest-neighbor algorithms [17–19,39]. A higher LyE value indicates a reduced ability of the motor system to counter infinitesimal perturbations [15], ultimately resulting in greater divergence of state space trajectories and, by extension, lower stability during locomotion [12,41]. Figure 3 shows an example of the space–time representation for the calculated LyE of a tri-axial acceleration. Second, the root-mean-square (RMS) was employed as a measure of the magnitude or intensity of an acceleration signal [42]. This metric serves to evaluate the forces and stresses imposed on the body during physical activities. Figure 2.Example tri-axial acceleration data of the treadmill running with (A) no load, (B) and running with loads on combined (forearms, legs, and trunk) segments, (C) forearms, (D) lower legs, and (E) trunk, respectively. Note: the presented middle 2 min tri-axial acceleration data were retrieved from each running

This study investigates how different placements of wearable resistance affect running stability.