Treadmill and Running Speed Effects on Acceleration Impacts: Curved Non-Motorized Treadmill vs. Conventional Motorized Treadmill

e in the popularity of running can be seen over the last decades, with a large number of injuries on it. Most of the running injuries are related to impact accelerations and are due to overuse. In order to reduce the risk of injury or to improve performance and health new treadmill designs have been created, as it can be the curved non-motorized treadmill. The aim of this study was to analyse impact accelerations, spatio-temporal parameters and perceptual differences while running on curved non-motorized treadmill (cNMT) compared to motorized treadmill (MT) at different speeds. Therefore, 27 recreational runners completed two tests consisting of 10 min warm-up and three bouts of 8 min running at 2.77 m/s, 3.33 m/s and self-selected speed on cNMT and MT, previously randomised. Although the surface did not in uence spatio-temporal parameters, a reduction in

while running on curved non-motorized treadmill (cNMT) compared to motorized treadmill (MT) at different speeds. Therefore, 27 recreational runners completed two tests consisting of 10 min warm-up and three bouts of 8 min running at 2.77 m/s, 3.33 m/s and self-selected speed on cNMT and MT, previously randomised. Although the surface did not in uence spatio-temporal parameters, a reduction in impact accelerations, head acceleration rate (mean effect size [ES] = 0.86), tibia peak (mean ES = 0.45) and tibia magnitude (mean ES = 0.55), was observed while running on cNMT in comparison with running on MT. Moreover, higher heart rate (HR) (mean ES = 0.51) and rating of perceived effort (RPE) (mean ES = 0.34) were found while running on cNMT. These ndings demonstrated that higher intensity training and lower impact accelerations are experimented on cNMT, what can be used by trainers and athletes while planning training sessions. Keywords:biomechanics; accelerometer; treadmill; locomotion 1. Introduction The popularity of running has been increasing during the last years due to its bene ts for health, accessibility and low cost, becoming one of the most common ways to exercise [1]. Despite its numerous bene ts, an increase in the prevalence of running injuries can be observed. It has been suggested that around the 42.7% of runners will get injured each year [2], being the majority of this injuries due to overuse [3]. Repeated and accumulated exposure to impact accelerations during long-distance running can overload and fatigue the musculoskeletal system, reducing its ability to absorb them and increasing the risk of injury [4]. As a result, impact acceleration analysis has received considerable scienti c interest in order to reduce these accelerations during running and decrease the incidence of overuse injuries [5]. Several factors can affect these impact accelerations during running, including stride parameters (length and frequency) [6], fatigue [7,8], running mechanics [9], foot strike pattern [10], sports equipment (footwear, compression socks or plantar supports) [11], running surface [12–14] and running speed [15–18]. Int. J. Environ. Res. Public Health2021,18, 5475.

these impact accelerations during running, including stride parameters (length and frequency) [6], fatigue [7,8], running mechanics [9], foot strike pattern [10], sports equipment (footwear, compression socks or plantar supports) [11], running surface [12–14] and running speed [15–18]. Int. J. Environ. Res. Public Health2021,18, 5475.

Int. J. Environ. Res. Public Health2021,18, 5475 2 of 10 Among all the factors that can in uence impact accelerations, running surfaces have been shown to in uence acceleration impacts during: overground running vs. motorized treadmill [7,12,13], concrete vs. grass [14], woodchip trail vs. synthetic track/concrete [15]. On the other hand, it has been shown that impact accelerations increase with higher velocities [17,18]. Runners, whenever they can and if the weather conditions allow it, train outdoors on different surfaces: asphalt, grass and even in the mountains. However, there are sports modalities in which it is necessary to train within a facility, and the treadmill is the only training alternative available due to its characteristics and the possibility of maintaining a certain speed and slope. When comparing overground vs. treadmill, running on a treadmill results in higher stride frequencies and shorter stride lengths [19], being this parameters closely related to running economy. According to Hunter & Smith [20] novice runners select lower stride frequencies than the optimal one, while experienced runners choose unconsciously higher frequencies, optimizing energy expenditure and improving running economy. Moreover, running on treadmill produces lower tibial peak acceleration and impact rates compared to the overground running [7,13]. Nowadays, new treadmill designs, such as the curved non-motorized treadmill (cNMT), have demonstrated to be a valid and reliable tool for rehabilitation, training and laboratory based research [21–23]. cNMT have been designed to evaluate the strength, maximum speeds, and power of the athlete, allowing a more speci c running evaluation. cNMT have a curved non-motorized surface, which requires the runner to impact the surface and propel the band with each stride [23,24]. The main difference compared to motorized treadmills (MT) is that cNMT allows participants to self-select the speed and allows a more valid and ecological laboratory assessment of running performance [25,26]. While different studies have been analysing the cNMT on sprints [27], endurance run- ning [26], cardiometabolic demands [28,29] and team-sport running [30]; other studies have focused on physiological and perceptual variables comparing cNMT with MT [21,25], and overground running [23]. However, only few studies have observed biomechanical changes during

a more valid and ecological laboratory assessment of running performance [25,26]. While different studies have been analysing the cNMT on sprints [27], endurance run- ning [26], cardiometabolic demands [28,29] and team-sport running [30]; other studies have focused on physiological and perceptual variables comparing cNMT with MT [21,25], and overground running [23]. However, only few studies have observed biomechanical changes during walking [29,31] or running on cNMT [32], observing shorter stride length compared to MT [33]. In terms of impact accelerations, just one research have analysed tibial impact acceleration during running on cNMT vs. MT [34], but no study of the impact transmission from the tibia to the head was carried out. To our knowledge, no previous research has analysed the effect on head and tibial accelerations during running on cNMT. Therefore, the aim of the present study was to analyse impact accelerations, spatio-temporal parameters and perceptual changes during running on cNMT vs. MT at different speeds in recreational runners. It was hypothesized that: (a) running on cNMT would reduce impact acceleration parameters, increasing stride frequency and reducing stride length in comparison with motorized treadmill running; and (b) the effect of running speed would affect impact acceleration, increasing when running at higher velocities. 2. Materials and Methods 2.1. Participants Twenty-seven recreational runners: 22 males and 5 females (age 25 7 years, height 170.30 8.09 cm, body mass 64.4 10.3 kg, running training 37 19 km/week) agreed to participate in the study and gave written informed consent. Inclusion criteria included to be physically active (to run a minimum of twice a week in the last year), a training volume of at least 20 km per week, no history of lower limb injuries within the last six months, no suffering of heart failure, neurological or musculoskeletal disorders affecting normal locomotion and to not be taking medication that interferes with stability during running. Exclusion criteria included injury, surgery or illness within the previous six months, and overweight or obesity (BMI > 24.9 kg/m 2 ). Based on previous studies and a general linear

neurological or musculoskeletal disorders affecting normal locomotion and to not be taking medication that interferes with stability during running. Exclusion criteria included injury, surgery or illness within the previous six months, and overweight or obesity (BMI > 24.9 kg/m 2 ). Based on previous studies and a general linear

Int. J. Environ. Res. Public Health2021,18, 5475 3 of 10 model (GLM) of two-way Repeated Measures design, a total sample size of 22 participants was needed to detect signi cant differences associated with a minimum detectable effect size (moderate) f = 0.253 ( = 0.05, = 0.05, power = 0.952) for acceleration impact. The study procedures complied with the Declaration of Helsinki and were approved by the University ethics committee (UV-INV-ETICA-1245207). 2.2. Study Protocol We carried out an experimental study with a quantitative approach without a control group and with a repeated measures design. Participants performed two randomised running test in different treadmills, one on cNMT (Bodytone ZRO-T, Bodytone Interna- tional Sport S.L., Molina del Segura, Spain) and another on MT (h/p/cosmos pulsar ® 3p, h/p/cosmos sports & medical gmbh. Nußdorf, Germany) with 1% incline to replicate the energy cost of outdoor running [35]. Similarly on both surfaces, participants warmed-up for 10 min at self-selected speed, which also served as familiarization time on the treadmill [7]. Then, participants ran 24 min in three separated bouts of 8 min at 2.77 m/s, 3.33 m/s and self-selected speed, in a random order (Figure). A completely randomized design protocol, using opaque envelopes for allocation concealment, was used to the treadmill order selection, and running speed. Envelopes were equal in weight, similar in appearance, and tamperproof [36].Int. J. Environ. Res. Public Health 2021, 18, x FOR PEER REVIEW 3 of 10 of at least 20 km per week, no history of lower limb injuries within the last six months, no suffering of heart failure, neurological or musculoskeletal disorders affecting normal lo- comotion and to not be taking medication that interferes with stability during running. Exclusion criteria included injury, surgery or illness within the previous six months, and overweight or obesity (BMI > 24.9 kg/m 2 ). Based on previous studies and a general linear model (GLM) of two-way Repeated Measures design, a total sample size of 22 participants was needed to detect significant differences associated with a minimum detectable effect size (moderate) f = 0.253 (α = 0.05, β = 0.05, power = 0.952)

six months, and overweight or obesity (BMI > 24.9 kg/m 2 ). Based on previous studies and a general linear model (GLM) of two-way Repeated Measures design, a total sample size of 22 participants was needed to detect significant differences associated with a minimum detectable effect size (moderate) f = 0.253 (α = 0.05, β = 0.05, power = 0.952) for acceleration impact. The study procedures complied with the Declaration of Helsinki and were approved by the University ethics committee (UV-INV-ETICA-1245207). 2.2. Study Protocol We carried out an experimental study with a quantitative approach without a control group and with a repeated measures design. Participants performed two randomised run- ning test in different treadmills, one on cNMT (Bodytone ZRO-T, Bodytone International Sport S.L., Molina del Segura, Spain) and another on MT (h/p/cosmos pulsar ® 3p, h/p/cos- mos sports & medical gmbh. Nußdorf, Germany) with 1% incline to replicate the energy cost of outdoor running [35]. Similarly on both surfaces, participants warmed-up for 10 min at self-selected speed, which also served as familiarization time on the treadmill [7]. Then, participants ran 24 min in three separated bouts of 8 min at 2.77 m/s, 3.33 m/s and self-selected speed, in a random order (Figure 1). A completely randomized design pro- tocol, using opaque envelopes for allocation concealment, was used to the treadmill order selection, and running speed. Envelopes were equal in weight, similar in appearance, and tamperproof [36]. The self-selected speed was chosen by the participants during the warm-up for each treadmill and it was used then as a condition speed. Participants ran with their own shoes in both testing days to reduce biomechanics variability. The tests were separated by at least 48 h and were carried out at the same time of the day (±1 h). Acceleration parameters were collected during the last minute of each bout with the purpose of reducing the measurement error due to stride variability [37]. A total of 2.430 strides on each treadmill and speed condition were analysed in the study. The rating of perceived exertion (RPE, 6–20 Borg scale) was reported after the end of

of the day (±1 h). Acceleration parameters were collected during the last minute of each bout with the purpose of reducing the measurement error due to stride variability [37]. A total of 2.430 strides on each treadmill and speed condition were analysed in the study. The rating of perceived exertion (RPE, 6–20 Borg scale) was reported after the end of each run [38]. Fi- nally, heart rate (HR) was also registered during the last minute using a portable HR belt (Polar V800, Polar Electro, Kempele, Finland). Participants did not know neither the ve- locity that they were running nor the moment when measurements started or finished in order to avoid running alterations [39]. Figure 1. Schematic representation of the study protocol. 2.3. Data Collection Acceleration parameters were measured by two lightweight triaxial wireless accel- erometers (Pikkulab, Blautic Design, Valencia, Spain; total mass: 50 g; dimensions: 50 × 20 × 10 mm; range: ±16 g) firmly attached to the skin with double-sided adhesive tape [7]. Accelerometer-based analysis systems have been used routinely to continuously assess Figure 1.Schematic representation of the study protocol. The self-selected speed was chosen by the participants during the warm-up for each treadmill and it was used then as a condition speed. Participants ran with their own shoes in both testing days to reduce biomechanics variability. The tests were separated by at least 48 h and were carried out at the same time of the day ( 1 h). Acceleration parameters were collected during the last minute of each bout with the purpose of reducing the measurement error due to stride variability [37]. A total of 2.430 strides on each treadmill and speed condition were analysed in the study. The rating of perceived exertion (RPE, 6–20 Borg scale) was reported after the end of each run [38]. Finally, heart rate (HR) was also registered during the last minute using a portable HR belt (Polar V800, Polar Electro, Kempele, Finland). Participants did not know neither the velocity that they were running nor the moment when measurements started or nished in order to avoid running alterations [39]. 2.3. Data Collection

was reported after the end of each run [38]. Finally, heart rate (HR) was also registered during the last minute using a portable HR belt (Polar V800, Polar Electro, Kempele, Finland). Participants did not know neither the velocity that they were running nor the moment when measurements started or nished in order to avoid running alterations [39]. 2.3. Data Collection Acceleration parameters were measured by two lightweight triaxial wireless ac- celerometers (Pikkulab, Blautic Design, Valencia, Spain; total mass: 50 g; dimensions: 50 20 10 mm; range: 16 g) rmly attached to the skin with double-sided adhesive tape [7]. Accelerometer-based analysis systems have been used routinely to continuously assess acceleration peaks during activities such as running and human gait, demonstrating excellent validity and reliability [10]. The accelerometers were placed on the forehead and the distal and anteromedial portion of the tibia [40] and secured by elastic belts. The vertical axis of the accelerometer was aligned to be parallel to the long axis of the shank, as the location of the tibial accelerometer does in uence the acceleration signal [40]. Vertical

Int. J. Environ. Res. Public Health2021,18, 5475 4 of 10 acceleration data were registered at 180 Hz using software Pikkulab APP (Blautic Design, Valencia, Spain). For cNMT, distance running was registered during minutes 2–3, 4–5 and 6–7 to calculate the exact speed and spatio-temporal parameters were obtained from the impact accelerations and the exact speed. Acceleration data were analysed using Matlab (MathWorks, MA, USA). For the spatio- temporal and impact acceleration analysis, the acceleration signal was ltered (Butterworth, second-order, low-pass, cut-off frequency = 50 Hz) and stride length, stride frequency, tibia and head acceleration rate (slope from ground contact to peak acceleration), tibia and head peak acceleration (maximum value of the acceleration signal), tibia and head acceleration magnitude (difference between the positive and the negative acceleration peak) and shock attenuation (reduction in impact acceleration from the tibia to the head) were calculated from the acceleration signal [40]. 2.4. Statistical Analysis Statistical analyses were carried out using SPSS.25 statistics software package (SPSS Inc., Chicago, IL, USA). The normality of the data was veri ed using the Shapiro–Wilk test (p= 0.274). Then, a general linear model of two-way repeated-measures design was per- formed. Post hoc comparisons were performed using the Bonferroni test to identify the location of speci c differences. RPE and HR were analysed through a Friedman test. In those cases where signi cant differences were found (p< 0.05), the Wilcoxon test was performed for pairwise comparison. For parametric and non-parametric analysis, running treadmill (cNMT and MT) and running speed (2.77 m/s, 3.33 m/s and self-selected speed) were considered as within-subject factors. The level of signi cance was set atp< 0.05. For signi cant pair differences, Cohen's effect sizes (ES) were computed and 95% con dence intervals of the differences (95% CI) were provided [41]. 3. Results No statistically signi cant differences (p> 0.05) were found between women and men. Therefore, data analysis was carried out as a homogeneous sample. 3.1. Treadmill Differences Running on cNMT provoked signi cantly lower impacts in head acceleration rate in comparison with MT at self-selected speed (p= 0.000, ES = 0.917, mean difference: 15.488, 95%

CI) were provided [41]. 3. Results No statistically signi cant differences (p> 0.05) were found between women and men. Therefore, data analysis was carried out as a homogeneous sample. 3.1. Treadmill Differences Running on cNMT provoked signi cantly lower impacts in head acceleration rate in comparison with MT at self-selected speed (p= 0.000, ES = 0.917, mean difference: 15.488, 95% CI [9.064–21.913]), 2.77 m/s (p= 0.000, ES = 0.706, mean difference: 20.422, 95% CI [11.378–29.467]) and 3.33 m/s (p= 0.000, ES = 0.951, mean difference: 23.463, 95% CI [13.311–33.615]) (Figure). In terms of tibia peak acceleration, differences between treadmills were observed at self-selected speed (p= 0.008, ES = 0.37, mean difference: 0.489, 95% CI [0.141–0.838]), 2.77 m/s (p= 0.001, ES = 0.477, mean difference: 0.685, 95% CI [0.338–1.033]) and 3.33 m/s (p= 0.001, ES = 0.495, mean difference: 0.865, 95% CI [0.420– 1.309]) (Figure). Finally, differences in tibia acceleration magnitude at self-selected speed (p= 0.022,ES = 0.398, mean difference: 0.699, 95% CI [0.109–1.290]), 2.77 m/s (p= 0.001, ES = 0.568, 1.016, 95% CI [0.464–1.568]) and 3.33 m/s (p= 0.000, ES = 0.67, mean difference: 1.362, 95% CI [0.732–1.991]) were found when cNMT was compared to MT (Figure). However, no other statistically signi cant (p> 0.05) differences in impact accelerations were found between treadmill conditions (Table).

The study compares impact accelerations on different treadmill types in recreational runners.