Training Specificity in Trail Running: A Single-Arm Trial on the Influence of Weighted Vest on Power and Kinematics in Trained Trail Runners

nts in trail running races must carry their equipment throughout the race. This additional load modi es running biomechanics. Novel running powermeters allow further analyses of key running metrics. This study aims to determine the acute effects of running with extra weights on running power generation and running kinematics at submaximal speed. Fifteen male amateur trail runners completed three treadmill running sessions with a weighted vest of 0-, 5-, or 10% of their body mass (BM), at 8, 10, 12, and 14 km h 1 . Mean power output (MPO), leg spring stiffness (LSS), ground contact time (GCT), ight time (FT), step frequency (SF), step length (SL), vertical oscillation (VO), and duty factor (DF) were estimated with the Stryd wearable system. The one-way ANOVA revealed higher GCT and MPO and lower DF, VO, and FT for the +10% BM compared to the two other conditions (p< 0.001) for the running speeds evaluated (ES: 0.2–7.0). After post-hoc testing, LSS resulted to be higher for +5% BM than for the +10% and +0% BM conditions (ES: 0.2 and 0.4). Running with lighter loads (i.e., +5% BM) takes the principle of speci

and lower DF, VO, and FT for the +10% BM compared to the two other conditions (p< 0.001) for the running speeds evaluated (ES: 0.2–7.0). After post-hoc testing, LSS resulted to be higher for +5% BM than for the +10% and +0% BM conditions (ES: 0.2 and 0.4). Running with lighter loads (i.e., +5% BM) takes the principle of speci city in trail running one step further, enhancing running power generation and LSS. Keywords:endurance; footpod; running power; wearable resistance 1. Introduction Unlike most road races, trail and ultra-trail runners need to carry their technical equipment (i.e., headlamps, spare batteries, rst aid kit, etc.) from the beginning to the end of the race. This adds an additional weight to the athletes' vest that can be up to 3–4 kg extra. Regarding this, it has been reported that running with additional external loads modi es running biomechanics [1]. Previous studies found that an extra load (5–30% bodymass) increases (+4–7%) ground contact time (GCT), vertical oscillation (VO) (+12–20.5%) [1–3], and leg spring stiffness (LSS) (+7–14%) [1,2,4]; however, controversial results have been reported for step length (SL) [1]. Worth noting, no study included duty factor (DF) or running power in their kinematic analyses. DF refers to the ratio of GCT and total stride time and provides a deeper insight into the overall running pattern than when these spatiotemporal parameters are considered independently [5]. For its part, mechanical power output (MPO) is the rate at which mechanical work is performed, being a key variable in endurance sports. Running power has been related to the metabolic cost of running [6,7], and it has been proposed as a predictor of running performance [8]. Its main advantage over other load indicators is that running power assesses the current workload that the athlete is developing regardless of likely external conditions (i.e., wind speed, slope steepness, terrain type, or additional weight). In fact, power output has been shown to be more sensitive to small changes in exercise intensity than other commonly used internal (i.e., heart rate) and external (i.e., speed) workload Sensors2023,23, 6411.

workload that the athlete is developing regardless of likely external conditions (i.e., wind speed, slope steepness, terrain type, or additional weight). In fact, power output has been shown to be more sensitive to small changes in exercise intensity than other commonly used internal (i.e., heart rate) and external (i.e., speed) workload Sensors2023,23, 6411.

Sensors2023,23, 6411 2 of 10 indicators [9]. In this regard, the Stryd running powermeter showed the highest concurrent validity with metabolic power measurements (i.e., VO2m¡x) (r 0.911, SEE = 7.3%), and the most repeatable device for running power estimation (SEM 12.5 W, CV 4.3%, ICC 0.980 ) [6]. Additionally, it has been considered the best wearable tool to analyze running power given its high values of repeatability [6]. As a result, the use of this foot pod has been spread among athletes and coaches of both road and trail running, generating a parallel interest among researchers in sports science. The emergence of running powermeters, and their friendly use and interpretation of the running power metric, have contributed to their spread amongst athletes and coaches, who are now considering running power as a load indicator. By analyzing the effects of overloaded running on running performance parameters, the optimal additional load to optimize the generation of running power without jeopardizing running performance might be determined, which is one of the objectives of wearable endurance training [1]. In this regard, the use of external additional loads (i.e., weighted vests) during running is a practical example of how the concept of training speci city can be applied in trail running. This has been shown to elicit adaptations in the neuromuscular behavior of the legs to reuse elastic energy and improve running economy [10] by optimizing the behavior of the stretch- shortening cycle and lower-limb stiffness. Theoretically, higher “leg-spring” stiffness maximizes potential elastic energy return, improving running economy [11]. However, the in uence of mechanical stiffness on running performance is speci c to the individual, and the assumption is still controversial. Although several studies have evaluated the acute and longitudinal effects of running and training with additional load, this was mainly analyzed in sprint running [12,13]. A recent study from Cerezuela-Espejo and colleagues measured MPO when running with additional load (i.e., +2.5 and +5 kg) with the aim of determining the test–retest reliability of the Stryd running powermeter [6]. Unfortunately, the likely in uence of additional load on running power was not considered. Therefore, the

with additional load, this was mainly analyzed in sprint running [12,13]. A recent study from Cerezuela-Espejo and colleagues measured MPO when running with additional load (i.e., +2.5 and +5 kg) with the aim of determining the test–retest reliability of the Stryd running powermeter [6]. Unfortunately, the likely in uence of additional load on running power was not considered. Therefore, the present study attempts to determine the acute effects of running with additional load (i.e., +5 and +10% of body mass) on running power generation and running kinematics at submaximal speeds (i.e., 8, 10, 12, and 14 km/h). We hypothesized that running with +5% BM would optimize running power generation without impairing running kinematics, whereas running with +10% BM will cause major changes in running technique to be able to produce the required amount of power. 2. Materials and Methods 2.1. Participants A group of fteen male amateur trail runners (age: 37 6 years; height:1.76 0.04 m ; body mass: 72.6 5 kg) voluntarily participated in this study. All participants met the inclusion criteria: (i) men older than 18 years old, (ii) at least 2 years of experience in trail running, (iii) not suffer any lower-limb injury in the last 6 months before the data collection that would make running training impossible for more than 2 weeks (iv), no cardiorespiratory or metabolic abnormality, and (v) not be taking any type of ergogenic aid that could disturb the results of the study. After receiving detailed information on the objectives and procedures of the study, each subject signed an informed consent form prior to participation, which complied with the ethical standards of the World Medical Association's Declaration of Helsinki (2013). It was made clear that the participants were free to leave the study if they saw t. The study was approved by the Ethics Committee of San Jorge University (Zaragoza, Spain). Sample size power calculation was executed using G*POWER 3.1.9.7 (University of Dusseldorf, Dusseldorf, Germany). The following structure was used based on the analysis: F test-ANOVA: repeated measures, within-interaction; A priori. Effect size f = 0.5; error prob = 0.05; power (1- error

if they saw t. The study was approved by the Ethics Committee of San Jorge University (Zaragoza, Spain). Sample size power calculation was executed using G*POWER 3.1.9.7 (University of Dusseldorf, Dusseldorf, Germany). The following structure was used based on the analysis: F test-ANOVA: repeated measures, within-interaction; A priori. Effect size f = 0.5; error prob = 0.05; power (1- error prob = 0.95; The number of groups = 1n and the number of measurements = 3n. The result showed a suitable total sample size of 12 athletes for actual high power (95.23%).

Sensors2023,23, 6411 3 of 10 2.2. Procedures This study was conducted in four sessions (Figure). During the rst session, a consent form was obtained from every participant; moreover, anthropometric measurements were collected, and the participants became familiar with the running protocol. During days 2, 3, and 4, the running protocol was carried out with the corresponding additional load (i.e., 0, 5, and 10% additional body weight). To avoid any bias, the additional load with which each had to be run was randomized. Prior to all testing, subjects refrained from severe physical activity for at least 72 h, and all tests were performed at least 3 h after eating. The tests were performed with the subjects wearing their usual running shoes to measure their typical performance.Sensors 2023, 23, x FOR PEER REVIEW 4 of 11 2.4. Statistical Analysis Descriptive statistics are presented as means ± standard deviations (SD) for all vari- ables. Comparisons between conditions (i.e., +0% BM vs. +5% BM, +0% BM vs. +10% BM, and +5% BM vs. +10% BM) are presented as mean difference ± SD. The middle 30 s of each speed were included in the analyses to avoid errors derived from adaptation to the differ- ent speeds. The Kolmogorov–Smirnov test was conducted to confirm data distribution normality and Levene’s test for equality of variances. A separate one-way analysis of var- iance (ANOVA) was used to identify differences between conditions regarding the extra weight for the different running speeds (i.e., 8, 10, 12, and 14 km·h −1 ). Finally, Gabriel or Games-Howell post-hoc analyses were also conducted when appropriate to determine significant differences between conditions. Effect sizes for all pairwise comparisons were also calculated using Cohen’s d, with 95% confidence intervals. Cohen’s d were classified as follows: small (0.00 < d < 0.49), medium (0.50 < d < 0.79), and large effects (d > 0.8) [16]. Data analysis was performed using SPSS (version 28, SPSS Inc., Chicago, IL, USA). Figure 1. Project design timeline. Figure 1.Project design timeline. The participants performed an incremental running protocol on a motorized treadmill (HPCosmos 20, Nußdorf, Germany) at an initial speed

small (0.00 < d < 0.49), medium (0.50 < d < 0.79), and large effects (d > 0.8) [16]. Data analysis was performed using SPSS (version 28, SPSS Inc., Chicago, IL, USA). Figure 1. Project design timeline. Figure 1.Project design timeline. The participants performed an incremental running protocol on a motorized treadmill (HPCosmos 20, Nußdorf, Germany) at an initial speed of 8 km h 1 for 3 min. Then, the speed was increased by 1 km h 1 every minute until volitional exhaustion. The

Sensors2023,23, 6411 4 of 10 slope was maintained at 1% over the entire protocol to simulate external conditions of air resistance [14]. No feedback was given to participants during data collection and subsequent analysis was performed by a different researcher for all measurements and conditions at the same time. 2.3. Materials and Testing For descriptive purposes, body height (cm), body mass (kg), and body fat (% body weight) were determined using a precision stadiometer and a weighing scale (SECA 222 and 634, respectively, SECA Corp., Hamburg, Germany). All measurements were taken with the participants wearing underwear. Body mass index (BMI) was calculated from the subjects' body mass and height (kg m 2 ) (Table). Table 1.Descriptive characteristics of the participants (mean (SD)).Variable Mean (SD) Age (years) 37.4 (5.8) Height (m) 176.2 (4.5) Weight (kg) 72.6 (4.9) BMI (m/kg 2 ) 23.4 (1.9) Body fat % 11.9 (3.4) Km/week 49.7 (20.7) BMI: body mass index; Km/week: Run kilometers per week. The spatiotemporal variables of GCT, ight time (FT), DF, SF, VO, and SL; MPO (in W), power output normalized to body mass (nMPO), and LSS was estimated with the Stryd running powermeter (Stryd powermeter, Stryd Inc., Boulder, CO, USA). Stryd is a carbon ber-reinforced foot pod based on a 6-axis inertial motion sen- sor (3-axis gyroscope, 3-axis accelerometer), which has been shown reliable (CV 3%, ICC 0.95) [6] and valid (Pearson r: 0.82 to 0.94, compared to a reference system) for running kinematic analyses [15]. Data from Stryd™were obtained from its website (www.stryd.com/powercenter/analysis, accessed on 16 June 2023) into the .csv le. Those les were imported into Excel ® (2016, Microsoft, Inc., Redmond, WA, USA) and further analyzed in SPSS (described below). The additional load (i.e., +5% and +10% of body weight) was added by using a weighted vest. This has a series of pockets on the front and back where 300 g bags can be added, allowing accurate control of the additional load and an even distribution of the extra load. 2.4. Statistical Analysis Descriptive statistics are presented as means standard deviations (SD) for all vari- ables. Comparisons between conditions

weight) was added by using a weighted vest. This has a series of pockets on the front and back where 300 g bags can be added, allowing accurate control of the additional load and an even distribution of the extra load. 2.4. Statistical Analysis Descriptive statistics are presented as means standard deviations (SD) for all vari- ables. Comparisons between conditions (i.e., +0% BM vs. +5% BM, +0% BM vs. +10% BM, and +5% BM vs. +10% BM) are presented as mean difference SD. The middle 30 s of each speed were included in the analyses to avoid errors derived from adaptation to the different speeds. The Kolmogorov–Smirnov test was conducted to con rm data distribu- tion normality and Levene's test for equality of variances. A separate one-way analysis of variance (ANOVA) was used to identify differences between conditions regarding the extra weight for the different running speeds (i.e., 8, 10, 12, and 14 km h 1 ). Finally, Gabriel or Games-Howell post-hoc analyses were also conducted when appropriate to determine signi cant differences between conditions. Effect sizes for all pairwise comparisons were also calculated using Cohen's d, with 95% con dence intervals. Cohen's d were classi ed as follows: small (0.00 < d < 0.49), medium (0.50 < d < 0.79), and large effects (d > 0.8) [16]. Data analysis was performed using SPSS (version 28, SPSS Inc., Chicago, IL, USA).

Sensors2023,23, 6411 5 of 10 3. Results Table ables under additional loads and body mass conditions and the magnitude and direction of the differences between the three conditions (i.e., 0%, +5%, and +10%). Table 2. Acute response of running power and kinematic variables (mean SD) to the different overweighted conditions and speeds. The main differences SD between conditions are also shown along with Cohen's d for effect size. Speed Variable +0% BM +5% BM +10% BM 0 vs. 5% BM ES 0 vs. 10% BM ES 5 vs. 10% BM ES 8 km h 1 Power (w) 179 3 187 4 195 4 9 5 2.7 16 5 * 5.3 7 5 2.0 Power (w/kg) 2.46 0.03 2.47 0.03 2.43 0.03 0.01 0.03 0.3 0.02 0.04 1.0 0.05 0.01 1.3 LSS (kN/m) 11.36 1.17 11.71 1.22 11.34 1.52 0.35 0.05 * 0.3 0.02 0.06 0.0 0.37 0.06 0.0 GCT (ms) 299 17 297 12 316 19 2 1 * 0.2 17 1 * 0.8 20 1 * 0.9 FT (ms) 123 6 123 5 118 5 0.1 0.2 0.0 4.3 0.2 * 0.8 4.3 0.2 * 0.9 DF (%) 35.4 1.0 35.4 0.7 36.4 0.9 0.1 0.0 0.0 0.9 0.1 * 0.9 1.0 0.1 * 1.1 SF (spm) 164 5 165 6 161 7 1 0.3 * 0.2 3 0.3 * 0.5 4 0.3 * 0.7 SL (cm) 81 4 81 4 81 4 0 0 0.0 1 0 * 0.2 1 0 * 0.2 VO (cm) 6.6 0.7 6.5 0.7 6.4 0.7 0.1 0.0 0.1 0.1 0.0 * 0.2 0.1 0.0 0.1 10 km h 1 Power (w) 221 4 232 4 243 5 11 7 2.7 22 7 * 5.5 11 7 2.7 Power (w/kg) 3.04 0.03 3.04 0.04 3.03 0.03 0.00 0.01 0.0 0.01 0.05 0.3 0.00 0.05 0.3 LSS (kN/m) 11.57 1.28 11.89 1.48 11.65 1.49 0.32 0.07 * 0.2 0.08 0.07 0.1 0.24 0.07 * 0.2 GCT (ms) 259 15 262 12 271 16 3 1 * 0.2 12 1 * 0.8 10 1 * 0.7 FT (ms) 139 8 137 6 133 8 2.1

0.04 3.03 0.03 0.00 0.01 0.0 0.01 0.05 0.3 0.00 0.05 0.3 LSS (kN/m) 11.57 1.28 11.89 1.48 11.65 1.49 0.32 0.07 * 0.2 0.08 0.07 0.1 0.24 0.07 * 0.2 GCT (ms) 259 15 262 12 271 16 3 1 * 0.2 12 1 * 0.8 10 1 * 0.7 FT (ms) 139 8 137 6 133 8 2.1 0.3 * 0.3 5.9 0.4 * 0.7 3.8 0.3 * 0.5 DF (%) 32.5 1.2 32.8 0.9 33.5 1.2 0.3 0.1 * 0.3 1.0 0.1 * 0.8 0.7 0.1 * 0.7 SF (spm) 167 6 168 7 167 7 1 0.3 0.1 0.4 0.3 0.1 1 0.3 * 0.2 SL (cm) 100 5 100 5 100 5 0 0 0.0 1 0 0.0 1 0 0.0 VO (cm) 7.4 0.8 7.3 0.8 7.1 0.9 0.1 0.0 * 0.2 0.3 0.0 * 0.4 0.2 0.0 * 0.2 12 km h 1 Power (w) 258 4 272 5 286 7 14 8 3.5 28 8 * 7.0 14 8 2.8 Power (w/kg) 3.56 0.03 3.57 0.04 3.57 0.04 0.01 0.05 0.3 0.02 0.05 0.3 0.01 0.05 0.0 LSS (kN/m) 11.78 1.38 12.25 1.41 11.74 1.33 0.47 0.07 * 0.3 0.04 0.07 * 0.1 0.51 0.07 * 0.4 GCT (ms) 233 23 233 23 242 24 0 1 0.1 10 1 * 0.4 10 0.5 * 0.4 FT (ms) 149 9 148 6 144 8 0.9 0.3 * 0.1 5.4 0.4 * 0.7 4.5 0.3 * 0.7 DF (%) 30.4 1.3 30.5 0.7 31.4 1.1 0.1 0.1 0.1 0.9 0.1 * 0.8 0.8 0.1 * 1.0 SF (spm) 174 6 174 8 173 7 0.8 0.3 * 0.1 0.6 0.3 0.0 1.4 0.3 * 0.1 SL (cm) 116 5 116 6 116 5 0 0 0.0 0 0 0.1 0 0 0.1 VO (cm) 7.4 0.8 7.4 0.9 7.2 0.9 0.0 0.0 0.0 0.2 0.0 * 0.2 0.2 0.0 * 0.2 14 km h 1 Power (w) 295 5 309 5 329 8 15 9 2.8 34 9 * 6.8 19 9 * 4.0 Power (w/kg) 4.06 0.04 4.06 0.04 4.11 0.05 0.00

The study investigates the impact of weighted vests on running biomechanics in trained trail runners.