Abstract
ckground/Objectives: Understanding how different fatigue contexts influence mus- cle architecture is essential for optimizing training and recovery strategies in endurance athletes. Ultramarathon running involves prolonged mechanical load and high eccentric demands, which may elicit different acute responses compared to controlled laboratory protocols. This study aimed to examine the effects of time, condition (laboratory vs. race), and muscle on ultrasound-derived muscle architecture in ultratrail runners.Methods: A repeated-measures within-subject design was employed. Forty ultratrail runners completed two fatigue conditions: (1) a standardized laboratory downhill running protocol and (2) an ultramarathon race (CSP 2025; 106 km, +5600 m elevation gain). Muscle thickness and pennation angle of the rectus femoris, vastus lateralis, and medial gastrocnemius were assessed using ultrasound before and after each condition. Linear mixed models were used to evaluate the effects of time, condition, muscle, and their interactions.Results: Forty participants were recruited; 29 completed all assessments. No significant effects of time or condition were observed for muscle thickness, and no interaction effects were detected, indicating that muscle size remained stable across conditions and time points. A signif- icant main effect of muscle was identified (p< 0.001), reflecting inherent morphological differences,
effects of time, condition, muscle, and their interactions.Results: Forty participants were recruited; 29 completed all assessments. No significant effects of time or condition were observed for muscle thickness, and no interaction effects were detected, indicating that muscle size remained stable across conditions and time points. A signif- icant main effect of muscle was identified (p< 0.001), reflecting inherent morphological differences, with greater thickness in the vastus lateralis compared to the rectus femoris and medial gastrocnemius. In contrast, pennation angle showed a significant main effect of condition (p= 0.031) and a significant condition×muscle interaction (p= 0.005), indicating muscle-specific differences between laboratory and race contexts. No significant effect of time was observed for pennation angle.Conclusions: Muscle thickness appears to remain stable following acute fatigue, regardless of the assessment context. In contrast, pennation angle may be more sensitive to condition-specific and muscle-dependent factors. These findings suggest that ultrasound-derived architectural changes observed immediately after exercise likely reflect acute physiological responses rather than true structural adaptations. Therefore, the interpretation of muscle architecture should consider both contextual factors and methodological constraints. Keywords:muscle fatigue; ultrasonography; running Diagnostics2026,16, 1080 https://doi.org/10.3390/diagnostics16071080
Diagnostics2026,16, 1080 2 of 14 1. Introduction Recently, specific assessment techniques have been implemented to evaluate the archi- tectural characteristics of the quadriceps (vastus lateralis and rectus femoris) and medial gastrocnemius muscles, with the aim of improving the diagnosis, prognosis, and perfor- mance monitoring in endurance athletes. Consequently, the precise evaluation of these architectural features has become an important tool in sports medicine, even in injury prevention programs. In fact, ultrasound and other noninvasive imaging techniques have substantially enhanced our understanding of muscle structure–function relationships, in- cluding implications for injury risk and performance [1]. For example, the measurement of muscle fascicle length (FL) via ultrasound has become a widely accepted method for detecting increases associated with training, as well as reductions related to muscle disuse or aging [2]. Recent studies have shown that muscle morphological adaptations to training in the vastus lateralis, rectus femoris, and medial gastrocnemius vary greatly between indi- viduals [3]. These inter-individual differences highlight the importance of individualized monitoring to optimize training outcomes. Furthermore, these data suggest that varia- tions in muscle architecture may influence the behavior of active muscle components and contribute to differences in force generation, fatigue resistance, and injury susceptibility [4]. Given the importance of muscle size and architecture in the muscles actually assessed, it is essential to consider intrinsic muscle properties, such as fiber type composition and fascicle length, which significantly influence maximal power output and energy expendi- ture [5]. Muscle activation and strength development have been shown to differ across exercises targeting the quadriceps and gastrocnemius muscles, influencing adaptations in fascicle length, pennation angle, and muscle volume [6]. These findings contribute to the refinement of musculoskeletal models that aim to predict mechanical muscle output during contraction more accurately [7]. Additionally, fiber type and fascicle length account for substantial variability in maximal power output and energy cost, underscoring the importance of fascicle length in the power–economy trade-off [8]. Ultramarathon races have grown exponentially in recent years; in this context, ultramarathon runners are exposed to unstable environmental and weather conditions for prolonged periods, as well as constant changes in terrain topography and altitude, with
and fascicle length account for substantial variability in maximal power output and energy cost, underscoring the importance of fascicle length in the power–economy trade-off [8]. Ultramarathon races have grown exponentially in recent years; in this context, ultramarathon runners are exposed to unstable environmental and weather conditions for prolonged periods, as well as constant changes in terrain topography and altitude, with significant positive and negative elevation gains [9]. Such prolonged and demanding conditions can lead to temporary muscular fatigue, either due to the accumulation of metabolites within the intracellular space or depletion of energy-producing substrates [10]. Specifically, the challenging terrain in ultramarathon running induces intense muscular activity, particularly during the eccentric contraction phase, resulting in muscle damage, hypercatabolism, elevated inflammatory biomarkers, and oxidative stress [11]. Consequently, muscle damage contributes to deficits in strength, rate of force development, and power output [12]. Additionally, exercise-induced fatigue exacerbates these impairments through fluid redistribution (muscle swelling) and alter- ations in neuromuscular activation, thereby disrupting key regulatory mechanisms [13]. Importantly, these acute responses may reflect transient edema or fluid shifts rather than true long-term architectural adaptations, which must be considered when interpreting post-exercise measurements. However, markers of muscle architecture and composition assessed via ultrasound did not show significant adaptation following the specific training intervention described in the cited study [14]. Notably, in the medial gastrocnemius mus- cle, eccentric contractions (rather than active stretching) appear to induce greater muscle damage than prolonged contractions at extended lengths [15]. These findings highlight the intricate nature of muscle responses to fatigue and mechanical stress, particularly among endurance athletes who encounter prolonged and variable conditions, such as those expe- rienced in ultramarathon events. Consequently, comprehending the subtleties of muscle https://doi.org/10.3390/diagnostics16071080
Diagnostics2026,16, 1080 3 of 14 architectural adaptations in response to diverse fatigue protocols is crucial for optimizing performance and recovery strategies within this population. Despite these advances, previous research has methodological limitations, including the absence of direct comparisons between controlled laboratory protocols and real compe- tition settings, limited use of statistical approaches capable of handling repeated measures and inter-individual variability, and reduced ecological validity in experimental designs. Importantly, there is a lack of studies directly comparing controlled laboratory fatigue pro- tocols (e.g., downhill running) with real ultramarathon competitions regarding their impact on the architecture of the quadriceps and medial gastrocnemius, particularly in the context of fatigue and performance [16]. Furthermore, individual variability, prior lower extremity injuries, or neuromuscular conditions may influence ultrasound assessments, highlighting the need for standardized evaluation protocols [17,18]. The objective of this study was to investigate the effects of time (pre vs. post), condition (laboratory vs. race), and muscle on muscle architecture, specifically focusing on muscle thickness and pennation angle, and to ascertain whether these factors interact within a repeated-measures framework. Furthermore, this study aimed to explore the potential impact of the assessment context on muscle architectural properties by evaluating the stability of muscle thickness and the sensitivity of pennation angle to condition-specific factors across different muscles. 2. Materials and Methods 2.1. Study Design This study employed a repeated-measures within-subject design, in which participants were assessed under two experimental conditions (laboratory versus race) and at two time points (pre- and post-exercise). This methodological approach facilitated the investigation of within-subject alterations in muscle architecture while accounting for inter-individual variability. Given the absence of randomization and the ecological nature of the race condi- tion, this study is more accurately categorized as a quasi-experimental design rather than a true controlled experiment. This study received approval from the Ethics Committee of Universitat Jaume I (Castellón) (reference number CEISH/103/2024). Prior to participation, all participants provided informed consent in accordance with the ethical standards of the Declaration of Helsinki [19]. The trial adhered to the STROBE guidelines [20]. Additionally, the study protocol was prospectively registered in the official ClinicalTrials.gov registry (ID: NCT06969898). 2.2.
experiment. This study received approval from the Ethics Committee of Universitat Jaume I (Castellón) (reference number CEISH/103/2024). Prior to participation, all participants provided informed consent in accordance with the ethical standards of the Declaration of Helsinki [19]. The trial adhered to the STROBE guidelines [20]. Additionally, the study protocol was prospectively registered in the official ClinicalTrials.gov registry (ID: NCT06969898). 2.2. Participants The present study involved ultra-trail runners registered as official participants in the CSP 2025 race, which was conducted in Castellón, Spain, on 11–12 April 2025. A total of 40 participants, comprising both men and women, were included. The study population consisted of healthy adult runners aged between 18 and 60 years who satisfied specific eligibility criteria. Recruitment and data collection were conducted during the pre-race period, with eligible participants being contacted and assessed prior to the event. The inclusion criteria mandated that participants be healthy adults aged between 18 and 60 years with prior experience in ultra-endurance events. Specifically, all participants were required to have completed at least two ultra-trail races of 65 km or longer and to be officially registered as participants in the CSP 2025 race (Castellón, Spain). The exclusion criteria encompassed a history of heart or kidney disease and the ongoing use of any medication at the time of recruitment. https://doi.org/10.3390/diagnostics16071080
Diagnostics2026,16, 1080 4 of 14 2.3. Experimental Design All participants underwent two fatigue-inducing conditions: a controlled laboratory downhill running protocol and an actual ultramarathon race, with assessments conducted both prior to and following each condition, allowing comparison between controlled and real-world fatigue scenarios. 2.3.1. Fatigue-Inducing Protocol Under Controlled Conditions (Laboratory) A controlled laboratory protocol was developed to induce neuromuscular fatigue through a standardized downhill running (DR) session. The first visit, conducted4–6 weeks prior to an ultramarathon race, involved baseline measurements, including blood sampling, muscle morphology assessment via ultrasound, isometric strength tests, and an uphill walking economy test. Participants then engaged in the DR protocol, which entailed 5 km of treadmill running at a 15% decline, designed to replicate descents typical of major ultratrail races [16]. The running speed was individualized to match the heart rate corresponding to the first ventilatory threshold (VT1) obtained during the uphill CPET, ensuring con- sistent relative intensity [21]. Mechanical variables, such as step length, step frequency, ground contact time, and vertical oscillation, were recorded using a foot-mounted power meter [22,23]. This protocol enabled the consistent induction of fatigue within a controlled setting, replicating the mechanical and physiological demands associated with trail running descents. Muscle architecture was evaluated in the laboratory both before and following the participants’ completion of a 5 km downhill running protocol. 2.3.2. Fatigue Assessment During Real Competition: Ultramarathon Race Protocol (Race) The field-based protocol was implemented during the 13th edition of the CSP (Castelló– Penyagolosa), which took place in Castellón, Spain, on 12 April 2025, commencing at 00:00 a.m.from the athletics track at Universitat Jaume I. The course encompassed a total distance of 106 km, featuring a positive elevation gain of 5600 m and a negative elevation of 4400 m. Participants were allotted a maximum time of 25 h and 30 min to complete the race. This authentic ultratrail competition offered an ecologically valid setting for evaluating neuromuscular fatigue and physiological adaptations under prolonged, high- demand endurance conditions. Additional information about the event can be found at: https://www.penyagolosatrails.com/csp/ were performed 6–12 h prior to the race using the same ultrasound device and
allotted a maximum time of 25 h and 30 min to complete the race. This authentic ultratrail competition offered an ecologically valid setting for evaluating neuromuscular fatigue and physiological adaptations under prolonged, high- demand endurance conditions. Additional information about the event can be found at: https://www.penyagolosatrails.com/csp/ were performed 6–12 h prior to the race using the same ultrasound device and assessor as in the laboratory. Pre-race assessments were performed 6–12 h prior to the race, using the same ultrasound device and assessor as in the laboratory. A temperature-controlled tent and portable treatment table were used to ensure reproducibility. At race completion, runners were guided directly to a second mobile testing area located at the finish line, where post-race ultrasound assessments were performed within 2–4 min of finishing. The same assessor performed all measurements to eliminate inter-rater variability. 2.4. Outcome Measures and Muscle Architecture Assessment 2.4.1. Measurement of Anthropometric Variables At the onset of the study, the researchers conducted an initial interview to gather anthropometric measurements from the participants. Weight and height were measured directly using a calibrated digital scale (Seca 813) and stadiometer (Seca 217), respectively. BMI was calculated as kg/m 2 , incorporating the age and sex of the study participants. 2.4.2. Primary Outcome: Muscle Architecture Assessed Using Ultrasound Ultrasound images were acquired using a Philips Lumify ® portable ultrasound system with an L12-4 linear transducer (Philips Healthcare, Best, The Netherlands). Depth, gain, https://doi.org/10.3390/diagnostics16071080
Diagnostics2026,16, 1080 5 of 14 and dynamic range were standardized across sessions. Two architectural features were assessed: pennation angle (PA) and muscle thickness (MT) in the rectus femoris, vastus lateralis, and medial gastrocnemius. Image analysis was performed using the ImageJ DICOM viewer (Pixmeo) and ImageJ (NIH) software (Version 1.53, National Institutes of Health, Bethesda, MD, USA) [24,25]. In both measurement sessions, a generous amount of ultrasound gel was applied to minimize pressure on the skin. Images were obtained by maintaining the transducer in a vertical position relative to the skin. The mean of the three measurements was used [26]. The measurement sites were pre-marked with a surgical skin marker [27]. Medial Gastrocnemius To assess the medial gastrocnemius muscle, participants were positioned in a prone posture with the knee fully extended. While the participants remained relaxed, longitudinal ultrasound measurements were conducted parallel to the orientation of the muscle fibers within the medial gastrocnemius belly. The measurement site was identified at one-third of the distance from the center of the knee joint to the calcaneus, measured proximally along the length of the leg [28]. Participants maintained passive muscle relaxation during imaging to avoid fascicle distortion. Vastus Lateralis and Rectus Femoris Ultrasound imaging of the vastus lateralis (VL) and rectus femoris (RF) was performed to evaluate muscle thickness and pennation angle in the dominant leg. Muscle thickness was defined as the distance between the superficial and deep aponeuroses at each end of the image. The pennation angle was defined as the angle formed between a muscle fascicle and its deep aponeurosis [29]. The measurement sites for the VL and RF were marked and recorded at 36% and 57%, respectively, of the distance from the superior border of the patella to the anterior superior iliac spine [30]. Participants were supine during these measurements. Anatomical landmarks were rechecked at each session before imaging. 2.5. Sample Size The sample size was estimated using GPower software (version 3.1.9.2; Franz Faul, Universität Kiel, Kiel, Germany). The selected effect size was categorized as large (>0.6), which was substantiated by previous and subsequent study markers of oxidative stress conducted specifically
iliac spine [30]. Participants were supine during these measurements. Anatomical landmarks were rechecked at each session before imaging. 2.5. Sample Size The sample size was estimated using GPower software (version 3.1.9.2; Franz Faul, Universität Kiel, Kiel, Germany). The selected effect size was categorized as large (>0.6), which was substantiated by previous and subsequent study markers of oxidative stress conducted specifically on ultramarathon runners [31]. Furthermore, at the level of Type I error (α) 0.05 and Type II error (1−β) 0.95, the estimated total sample size was 32 volunteers. To account for potential attrition during follow-up (20%), the minimum number of participants was determined to be 39. 2.6. Bias A physiotherapist with >5 years of musculoskeletal ultrasound experience executed the ultrasound protocol. Each measurement site was assessed thrice to evaluate intra-rater reliability, specifically using the intra-class correlation. Intra-class reliability pertains to the consistency or stability of measurements conducted by the same observer [32]. To evaluate inter-rater reliability and mitigate the risk of bias due to potential measurement variability, a secondary assessment was conducted on a sample of 12 subjects. This evaluation focused on specific time points and muscles: pre-laboratory measurement (MT of the rectus femoris), post-race (PA of the vastus lateralis), and post-race (MT of the medial gastrocnemius). The intraclass correlation coefficient (ICC) was calculated using the ICC (3,1) model, as classified by Shrout and Fleiss [32]. This model is suitable when the same raters assess all subjects and are considered fixed effects, estimating the absolute agreement between individual https://doi.org/10.3390/diagnostics16071080
Diagnostics2026,16, 1080 6 of 14 measurements. In both scenarios, the ICC was calculated and categorized as follows: poor (<0.4), moderate (0.4–0.75), and good (>0.75) [33]. 2.7. Statistical Analysis Descriptive statistics were computed to characterize the sample. Continuous vari- ables are presented as means±standard deviation. The data distribution was visually inspected to ensure the appropriate interpretation of the results. To examine the effects of time, condition, and muscle on muscle thickness and pennation angle, linear mixed models (LMMs) were employed. This approach appropriately accounts for the hierarchical structure of the data and repeated measurements within participants. For each dependent variable (MT and PA), fixed effects included time (Pre vs. Post), condition (Laboratory vs. Race), and muscle (rectus femoris, vastus lateralis, and medial gastrocnemius), as well as all interaction terms (time×condition, time×muscle, condition×muscle, and time×condition×muscle) . A random intercept for participant ID was included in all models to account for within-subject variability. Random slopes were not included in the final models to avoid overparameterization and potential singular model fits. Models were estimated using restricted maximum likelihood (REML). Statistical inference for fixed effects was conducted using Type III analysis of variance with the Satterthwaite approxi- mation for degrees of freedom. marginal means (EMMs), along with their standard errors and 95% confidence intervals, were calculated to facilitate interpretation and to describe patterns across conditions, time points, and muscles. Model assumptions were assessed through visual inspection of residual plots to evaluate normality and homoscedasticity. All analyses were performed using JASP (version 0.19.1), and statistical significance was set atp< 0.05. 3. Results 3.1. Recruitment, Program Feasibility and Safety: Attendance, Compliance Within the laboratory conditions, attendance was defined as the number of participants who initiated the experimental protocol, whereas compliance referred to the proportion who completed all required assessments. Of the 40 participants who initiated the laboratory conditions, 36 completed the protocol, resulting in a compliance rate of 90%. Similarly, in the race condition, attendance referred to the number of participants who started the protocol, whereas compliance represented those who completed both pre- and post-assessments. Of the 40 participants who initiated this condition, 29 completed all required
required assessments. Of the 40 participants who initiated the laboratory conditions, 36 completed the protocol, resulting in a compliance rate of 90%. Similarly, in the race condition, attendance referred to the number of participants who started the protocol, whereas compliance represented those who completed both pre- and post-assessments. Of the 40 participants who initiated this condition, 29 completed all required measurements, yielding a compliance rate of 72.5%. 3.2. Descriptive Analysis A total of 29 participants were included in the study, each evaluated under two intervention conditions (Laboratory and Race). All available observations were included in the mixed models. The descriptive characteristics of the sample are presented in Table. The participants’ mean age was 45.28±7.55 years, average body weight was68.22±10.93 kg , mean height was 171.67±8.82 cm, and body mass index was 23.02±2.37 kg/m 2 . Table 1.Descriptive characteristics and muscle outcomes across conditions. Variable Total ( n= 29) Laboratory Pre Laboratory Post ∆% Lab Race Pre Race Post ∆% Race Participant characteristics Age (years) 45.28±7.55 – – – – – – Height (cm) 171.67±8.82 – – – – – – Weight (kg) 68.22±10.93 – – – – – – BMI (kg/m 2 ) 23.02±2.37 – – – – – – https://doi.org/10.3390/diagnostics16071080
Description
This study examines muscle architecture changes in ultramarathon runners under different fatigue conditions.