Pierre Samozino
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Pierre Samozino has a total of 164 co-authors. Here are the top co-authors by number of shared publications:
- Jean-Benoit Morin
Inter University Laboratory of Human Movement Biology, EA 7424, University Jean Monnet Saint-Etienne, Lyon 1, University Savoie Mont-Blanc, 42023 Saint-Etienne, France
25 shared publications - Baptiste Morel
Inter-University Laboratory of Human Movement Biology, Université Savoie Mont-Blanc, Chambéry, France
17 shared publications - Maximilien Bowen
Laboratoire Interuniversitaire de Biologie de la Motricité LIBM, EA 7424, Université Savoie Mont-Blanc, Le Bourget-du-Lac, France
13 shared publications - Matt Cross
Auckland University of Technology
10 shared publications - Mylène Vonderscher
Laboratoire Interuniversitaire de Biologie de la Motricité LIBM, EA 7424, Université Savoie Mont Blanc, Le Bourget-du-Lac, France.
7 shared publications - Jean Romain Rivière
Laboratoire Interuniversitaire de Biologie de La Motricité, Univ Savoie Mont Blanc, EA 7424, Chambéry, France
5 shared publications
Latest Publications
- 3 November 2026
- Published by Taylor & Francis in Sports Biomechanics
- p. 1-20
Mechanical constraints of the biceps femoris during maximal acceleration sprinting
Hamstring muscle injuries (HMIs) are a key focus in both practice and research on their management and prevention (Buckthorpe et al., 2019; Ekstrand et al., 2023; Hickey et al., 2022). Despite sign...
- 3 November 2026
- Published by Taylor & Francis in Sports Biomechanics
- p. 1-20
Mechanical constraints of the biceps femoris during maximal acceleration sprinting
Hamstring muscle injuries (HMIs) are a key focus in both practice and research on their management and prevention (Buckthorpe et al., 2019; Ekstrand et al., 2023; Hickey et al., 2022). Despite sign...
- 23 August 2026
- Published by Taylor & Francis in Journal of Sports Sciences
- p. 1-11
No differences between cold and hot water immersion for neuromuscular recovery, sprint mechanics and delayed onset muscle soreness after a simulated football game
Football players typically cover approximately 10 km during a match, interspersed with frequent transitions between low-to-moderate and high-intensity efforts (Bengtsson et al., 2013). From a techn...
- 23 August 2026
- Published by Taylor & Francis in Journal of Sports Sciences
- p. 1-11
No differences between cold and hot water immersion for neuromuscular recovery, sprint mechanics and delayed onset muscle soreness after a simulated football game
Football players typically cover approximately 10 km during a match, interspersed with frequent transitions between low-to-moderate and high-intensity efforts (Bengtsson et al., 2013). From a techn...
- 21 April 2026
- Published by Springer Nature in European Journal of Applied Physiology
- Vol. 126 (9) , 4833-4845
RACLET: the Ramp Above Critical Level Endurance Test to evaluate critical velocity in running
- 21 April 2026
- Published by Springer Nature in European Journal of Applied Physiology
- Vol. 126 (9) , 4833-4845
RACLET: the Ramp Above Critical Level Endurance Test to evaluate critical velocity in running
- 1 April 2026
- Published by The Company of Biologists in Journal of Experimental Biology
- Vol. 229 (7)
Moving against inertia, friction and gravity: different modalities, same consequences for animal propulsive force capacity
The ability to generate propulsive force at different velocities is essential for animal locomotion but has often been overlooked. This study explored animal locomotion under varying mechanical constraints by addressing whether force capacity measured during acceleration on level ground is representative of propulsion capacity exerted during steady velocity uphill running or running against a resistance. We hypothesised that locomotion against resistance induced by inertia, friction or gravity would lead to similar propulsive force capacity, step length and step frequency. Nineteen human participants performed three accelerated, six resisted and ten uphill sprints while their instantaneous velocity, step length and step frequency were measured. The propulsive force capacity decreased linearly with velocity. This individual relationship was preserved among the disparate mechanical constraints, with humans just shifting along this curve. Trivial (−2.0±21.7%, P=0.43) and small differences (−6.1±21.5%, P=0.24), and positive correlation (P<0.001) were indeed found between force capacities at similar velocities among uphill/accelerated (r=0.94) and resisted/accelerated (r=0.91) conditions, respectively. Spatio-temporal variables did not differ between conditions (<2%). Conducting similar analysis in a 12-animal dataset from the literature revealed that different experimental modalities are associated with similar propulsive force–velocity relationships within the same species. Extending the analogy between accelerated, uphill and resisted running to the animal kingdom enabled comparisons between species based on propulsive force capacity and allometric scaling. Using humans as an experimental paradigm, we provide a framework for interpreting how environmental stressors affect movement strategy in many terrestrial species. In the field of sports science, this study has practical implications for the design of training and research protocols.
- 1 April 2026
- Published by The Company of Biologists in Journal of Experimental Biology
- Vol. 229 (7)
Moving against inertia, friction and gravity: different modalities, same consequences for animal propulsive force capacity
The ability to generate propulsive force at different velocities is essential for animal locomotion but has often been overlooked. This study explored animal locomotion under varying mechanical constraints by addressing whether force capacity measured during acceleration on level ground is representative of propulsion capacity exerted during steady velocity uphill running or running against a resistance. We hypothesised that locomotion against resistance induced by inertia, friction or gravity would lead to similar propulsive force capacity, step length and step frequency. Nineteen human participants performed three accelerated, six resisted and ten uphill sprints while their instantaneous velocity, step length and step frequency were measured. The propulsive force capacity decreased linearly with velocity. This individual relationship was preserved among the disparate mechanical constraints, with humans just shifting along this curve. Trivial (−2.0±21.7%, P=0.43) and small differences (−6.1±21.5%, P=0.24), and positive correlation (P<0.001) were indeed found between force capacities at similar velocities among uphill/accelerated (r=0.94) and resisted/accelerated (r=0.91) conditions, respectively. Spatio-temporal variables did not differ between conditions (<2%). Conducting similar analysis in a 12-animal dataset from the literature revealed that different experimental modalities are associated with similar propulsive force–velocity relationships within the same species. Extending the analogy between accelerated, uphill and resisted running to the animal kingdom enabled comparisons between species based on propulsive force capacity and allometric scaling. Using humans as an experimental paradigm, we provide a framework for interpreting how environmental stressors affect movement strategy in many terrestrial species. In the field of sports science, this study has practical implications for the design of training and research protocols.
- 27 November 2025
- Published by Wiley in Scandinavian Journal of Medicine & Science in Sports
- Vol. 35 (12) , e70174
Velocity‐Matched Resisted and Unresisted Sprinting Across Broad Loads Show Minor Biomechanical Differences
Many comparisons of unresisted and resisted sprinting do not match running velocity. As such, differences may partly reflect velocity rather than resistance. This cross‐sectional study examined kinetic and kinematic differences between unresisted and resisted sprinting accelerations at matched running velocities. Fourteen amateur sportsmen performed unresisted and resisted maximal overground sprints. Motorized resistance was applied via the waist and set at 20%, 40%, and 60% decreases in maximal unresisted velocity ( V DEC ). Ground reaction forces and segment/joint angles were compared across velocity‐matched steps between the conditions using ANOVA and statistical parametric mapping (SPM). Longer contact times were observed for resisted versus the unresisted condition, specifically 6.4% and 16.8% at 40% and 60% V DEC (ES: −0.423; −0.851, respectively). Vertical impulse was 8.0%, 7.3%, and 14.6% greater in the resisted condition at 20%, 40%, and 60% V DEC , (ES: −0.788; −0.551; and −0.862, respectively). Ankle range of motion was 5.8° greater at 60% V DEC (ES: −1.064), and hip angle at toe‐off was 2.8° and 8.3° greater at 40% and 60% V DEC , (ES: −0.385; −0.987, respectively), compared to unresisted sprinting. No other statistically significant differences were observed. SPM analyses indicated differences only between kinetics at 20% V DEC and ankle and trunk angles at 20% and 60% V DEC . Our findings contrast the larger and consistent changes typically reported in resisted sprinting, indicating movement velocity may be a confounder. The comparatively modest remaining differences detected may reflect load‐specific mechanisms that deserve future study.
- 27 November 2025
- Published by Wiley in Scandinavian Journal of Medicine & Science in Sports
- Vol. 35 (12) , e70174
Velocity‐Matched Resisted and Unresisted Sprinting Across Broad Loads Show Minor Biomechanical Differences
Many comparisons of unresisted and resisted sprinting do not match running velocity. As such, differences may partly reflect velocity rather than resistance. This cross‐sectional study examined kinetic and kinematic differences between unresisted and resisted sprinting accelerations at matched running velocities. Fourteen amateur sportsmen performed unresisted and resisted maximal overground sprints. Motorized resistance was applied via the waist and set at 20%, 40%, and 60% decreases in maximal unresisted velocity ( V DEC ). Ground reaction forces and segment/joint angles were compared across velocity‐matched steps between the conditions using ANOVA and statistical parametric mapping (SPM). Longer contact times were observed for resisted versus the unresisted condition, specifically 6.4% and 16.8% at 40% and 60% V DEC (ES: −0.423; −0.851, respectively). Vertical impulse was 8.0%, 7.3%, and 14.6% greater in the resisted condition at 20%, 40%, and 60% V DEC , (ES: −0.788; −0.551; and −0.862, respectively). Ankle range of motion was 5.8° greater at 60% V DEC (ES: −1.064), and hip angle at toe‐off was 2.8° and 8.3° greater at 40% and 60% V DEC , (ES: −0.385; −0.987, respectively), compared to unresisted sprinting. No other statistically significant differences were observed. SPM analyses indicated differences only between kinetics at 20% V DEC and ankle and trunk angles at 20% and 60% V DEC . Our findings contrast the larger and consistent changes typically reported in resisted sprinting, indicating movement velocity may be a confounder. The comparatively modest remaining differences detected may reflect load‐specific mechanisms that deserve future study.
- 11 September 2025
- Published by Springer Nature in European Journal of Applied Physiology
- Vol. 126 (2) , 939-949
A single decreasing ramp friction sprint for torque-cadence relationship assessment in cycling
Purpose This study aimed to introduce and validate a novel method for assessing dynamic fatigue components through a single-sprint test, addressing the limitations of traditional multi-sprint evaluations. Methods We tested this method on twenty-one participants by computing torque–cadence relationships from two iso-friction sprints at varying friction levels (3% and 9% of body mass), the traditional combination of these iso-friction sprints and a novel decreasing ramp friction sprint (\(\text {Fr}_{\text {D}}\)). The accuracy of this new method through fatigue was also tested with ten 6-s \(\text {Fr}_{\text {D}}\) sprints interspersed with a 24-s passive rest. Results \(\text {Fr}_{\text {D}}\) outperformed single iso-friction sprints and provided accurate and valid torque–cadence relationship’s parameters estimates (\({T}_{\text {0}}\), \({C}_{\text {0}}\), and \({P}_{\text {max}}\)) with systematic bias < 3%, typical error of estimate < 6% and very high \({r}^{{2}}\) (median of 0.962). The quality of the input data from this method was also high, as evidenced by the well-distributed and wide-range cadence spectrum (51.3% of \({C}_{{0}}\); skewness = \(-\)0.51, \(p < 0.05\)) and was maintained throughout the fatiguing exercise. Conclusions Our novel method not only allows the dynamic fatigue components evaluation with only one sprint but also maintains accuracy and validity across varying fatigue states, offering significant advantages for both research and practical applications.- 11 September 2025
- Published by Springer Nature in European Journal of Applied Physiology
- Vol. 126 (2) , 939-949
A single decreasing ramp friction sprint for torque-cadence relationship assessment in cycling
Purpose This study aimed to introduce and validate a novel method for assessing dynamic fatigue components through a single-sprint test, addressing the limitations of traditional multi-sprint evaluations. Methods We tested this method on twenty-one participants by computing torque–cadence relationships from two iso-friction sprints at varying friction levels (3% and 9% of body mass), the traditional combination of these iso-friction sprints and a novel decreasing ramp friction sprint (\(\text {Fr}_{\text {D}}\)). The accuracy of this new method through fatigue was also tested with ten 6-s \(\text {Fr}_{\text {D}}\) sprints interspersed with a 24-s passive rest. Results \(\text {Fr}_{\text {D}}\) outperformed single iso-friction sprints and provided accurate and valid torque–cadence relationship’s parameters estimates (\({T}_{\text {0}}\), \({C}_{\text {0}}\), and \({P}_{\text {max}}\)) with systematic bias < 3%, typical error of estimate < 6% and very high \({r}^{{2}}\) (median of 0.962). The quality of the input data from this method was also high, as evidenced by the well-distributed and wide-range cadence spectrum (51.3% of \({C}_{{0}}\); skewness = \(-\)0.51, \(p < 0.05\)) and was maintained throughout the fatiguing exercise. Conclusions Our novel method not only allows the dynamic fatigue components evaluation with only one sprint but also maintains accuracy and validity across varying fatigue states, offering significant advantages for both research and practical applications.- Open Access
RACLET: the Ramp Above Critical Level Endurance Test to evaluate critical velocity in running
Critical velocity ( V c ) is an important fatigability threshold in running, but only strenuous methods exist to assess it (e.g. time-trials). Based on a novel mathematical model of fatigability, the Ramp Above Critical Level Endurance Test (RACLET), a simple, non-exhaustive, 5-min test to evaluate V c was developed. This study aimed to test the reliability and validity of the RACLET. Thirty-eight participants performed two RACLET (session 1) and three time-trials (sessions 2-4). For the RACLET, velocity target tracking was guaranteed by either a pacing bike or cones and whistle signals. GPS was used to measure the participants’ running velocities. V c , τ , and were determined by adjusting the fatigability model on the measured V max . The RACLET test-retest and RACLET vs. time-trials parameters were compared for reliability and validity testing, respectively. The RACLET V c is reliable and valid (mean difference=-1.7 ± 1.9% and 6.2 ± 6.5%, respectively). Although τ and presented important variability (s.d. difference ~30%), the combination of the RACLET parameters enabled an excellent prediction of time-trials performance (mean error=-8.9 ± 5.1%, 0.9 ± 7.4%, −2.0 ± 5.2% for 400, 1500, and 3000 m, respectively). RACLET is a novel non-exhausting test that enables a reliable and valid assessment of the velocity–time relationship parameters.
Open AccessRACLET: the Ramp Above Critical Level Endurance Test to evaluate critical velocity in running
Critical velocity ( V c ) is an important fatigability threshold in running, but only strenuous methods exist to assess it (e.g. time-trials). Based on a novel mathematical model of fatigability, the Ramp Above Critical Level Endurance Test (RACLET), a simple, non-exhaustive, 5-min test to evaluate V c was developed. This study aimed to test the reliability and validity of the RACLET. Thirty-eight participants performed two RACLET (session 1) and three time-trials (sessions 2-4). For the RACLET, velocity target tracking was guaranteed by either a pacing bike or cones and whistle signals. GPS was used to measure the participants’ running velocities. V c , τ , and were determined by adjusting the fatigability model on the measured V max . The RACLET test-retest and RACLET vs. time-trials parameters were compared for reliability and validity testing, respectively. The RACLET V c is reliable and valid (mean difference=-1.7 ± 1.9% and 6.2 ± 6.5%, respectively). Although τ and presented important variability (s.d. difference ~30%), the combination of the RACLET parameters enabled an excellent prediction of time-trials performance (mean error=-8.9 ± 5.1%, 0.9 ± 7.4%, −2.0 ± 5.2% for 400, 1500, and 3000 m, respectively). RACLET is a novel non-exhausting test that enables a reliable and valid assessment of the velocity–time relationship parameters.
- Open Access
RACLET: the Ramp Above Critical Level Endurance Test to evaluate critical power in cycling
This study introduces the Ramp Above Critical Level Endurance Test (RACLET), a novel method for evaluating critical power model parameters, and tests its reliability and concurrent validity. Twenty-three participants completed several RACLET and time-to-exhaustion tests (TTE). The RACLET involves a decreasing power ramp with intermittent maximal sprints, inducing moderate fatigue without exhaustion. The test demonstrated excellent reliability for the initial power ( P i ) and critical power ( P c ) (ICC > 0.97), with acceptable reliability for the time constant ( τ ) (ICC = 0.70). The concurrent validity against TTE was excellent for P i and P c (systematic errors: 1.7% and 3.0%, respectively), but moderate for τ and maximum work above the critical power (systematic errors ≈ 10%). The RACLET showed excellent predictive capacity for time-to-exhaustion (systematic error = −0.6%; random error = 10.3%). These results suggest that the RACLET is a reliable, valid, and efficient alternative to traditional critical power testing methods, offering comparable accuracy with a single test without inducing exhaustion. This approach could be particularly beneficial for populations in which exhaustive testing is challenging or impractical.
Open AccessRACLET: the Ramp Above Critical Level Endurance Test to evaluate critical power in cycling
This study introduces the Ramp Above Critical Level Endurance Test (RACLET), a novel method for evaluating critical power model parameters, and tests its reliability and concurrent validity. Twenty-three participants completed several RACLET and time-to-exhaustion tests (TTE). The RACLET involves a decreasing power ramp with intermittent maximal sprints, inducing moderate fatigue without exhaustion. The test demonstrated excellent reliability for the initial power ( P i ) and critical power ( P c ) (ICC > 0.97), with acceptable reliability for the time constant ( τ ) (ICC = 0.70). The concurrent validity against TTE was excellent for P i and P c (systematic errors: 1.7% and 3.0%, respectively), but moderate for τ and maximum work above the critical power (systematic errors ≈ 10%). The RACLET showed excellent predictive capacity for time-to-exhaustion (systematic error = −0.6%; random error = 10.3%). These results suggest that the RACLET is a reliable, valid, and efficient alternative to traditional critical power testing methods, offering comparable accuracy with a single test without inducing exhaustion. This approach could be particularly beneficial for populations in which exhaustive testing is challenging or impractical.
- Open Access
RACLET: the Ramp Above Critical Level Endurance Test to evaluate critical force in isometric task
The Ramp Above Critical Level Endurance Test (RACLET) is a novel submaximal test designed to evaluate the parameters of the critical force model without strenuous exercise. This study aimed to validate the RACLET in healthy and pathological populations and to assess its reliability, concurrent validity, and predictive capacity. Sixteen healthy volunteers and ten patients with respiratory pathologies participated in this study. The RACLET consisted of a decreasing ramp force starting at 60% and ending at 15% of the maximum force for a total duration of 425 s with brief regular maximal voluntary contractions. The goodness of fit for the RACLET model on the maximal contraction force was excellent in both populations (median r 2 ≈ 0.95). In patients, RACLET parameters demonstrated excellent reliability (ICC > 0.90). The concurrent validity of the critical force estimate compared with the all-out method was high (error: −0.3 ± 7.4%). The model’s predictive capacity for time-to-exhaustion and fatigue during constant-intensity exercise was excellent (r 2 = 0.910 and 0.907, respectively). The RACLET provides a reliable and valid estimate of critical force model parameters opening up numerous practical applications in vulnerable populations, individualised physical activity programs, and longitudinal monitoring. The feasibility and performance of the test make it a promising tool for assessing muscle function in various contexts.
Open AccessRACLET: the Ramp Above Critical Level Endurance Test to evaluate critical force in isometric task
The Ramp Above Critical Level Endurance Test (RACLET) is a novel submaximal test designed to evaluate the parameters of the critical force model without strenuous exercise. This study aimed to validate the RACLET in healthy and pathological populations and to assess its reliability, concurrent validity, and predictive capacity. Sixteen healthy volunteers and ten patients with respiratory pathologies participated in this study. The RACLET consisted of a decreasing ramp force starting at 60% and ending at 15% of the maximum force for a total duration of 425 s with brief regular maximal voluntary contractions. The goodness of fit for the RACLET model on the maximal contraction force was excellent in both populations (median r 2 ≈ 0.95). In patients, RACLET parameters demonstrated excellent reliability (ICC > 0.90). The concurrent validity of the critical force estimate compared with the all-out method was high (error: −0.3 ± 7.4%). The model’s predictive capacity for time-to-exhaustion and fatigue during constant-intensity exercise was excellent (r 2 = 0.910 and 0.907, respectively). The RACLET provides a reliable and valid estimate of critical force model parameters opening up numerous practical applications in vulnerable populations, individualised physical activity programs, and longitudinal monitoring. The feasibility and performance of the test make it a promising tool for assessing muscle function in various contexts.
- Open Access
- 9 July 2025
- Published by Center for Open Science in OSF Preprints
The 2-point method for a simplified individualization of resisted sprint assessment and training
Purpose. This study compared the concurrent validity of simplified 2-point and traditional multiple-point method to characterize individual resisted sprint load-velocity (L-v) relationship and estimatemaximal unloaded running velocity and maximal theoretical load.Methods. Twenty-three elite male rugby sevens players performed two testing sessions: One forthe multiple-point approach, which included an unresisted sprint plus four resisted sprints againsta spectrum of loads relative to athletes’ average body mass (25%, 50%, 75%, and 100%). Asecond session for the 2-point approach, involving only resisted sprints against the same 25% and75% loads. Individual L-v relationships were compiled using full data from the multiple-point testing(standard approach) and 2-point relationships extracted from the multiple and 2-point testingsessions (indirect and direct approach, respectively). The main outcomes of individual L-vrelationships—maximal theoretical load (L0) and maximal theoretical velocity (v0)—were comparedbetween the standard multiple-point method and both the direct and indirect 2-point methods.Results. Trivial to small differences (Cohen’s d ≤ 0.34), nearly perfect correlations (r ≥ 0.92), minorhomoscedastic errors (r² ≤ 0.03), and no fixed or proportional bias according to ordinary least-products regressions were observed between the standard and indirect and direct 2-point methods.Conclusion. In habituated athletes, the direct 2-point method is a practical tool for routinelyassessing the L-v relationship during resisted sled sprinting and estimating both v0 and L0. Thismethod offers several advantages over the standard multiple-load approach, including reducedtesting time, lower fatigue, and decreased physical burden from unresisted sprints.
- 9 July 2025
- Published by Center for Open Science in OSF Preprints
Open AccessThe 2-point method for a simplified individualization of resisted sprint assessment and training
Purpose. This study compared the concurrent validity of simplified 2-point and traditional multiple-point method to characterize individual resisted sprint load-velocity (L-v) relationship and estimatemaximal unloaded running velocity and maximal theoretical load.Methods. Twenty-three elite male rugby sevens players performed two testing sessions: One forthe multiple-point approach, which included an unresisted sprint plus four resisted sprints againsta spectrum of loads relative to athletes’ average body mass (25%, 50%, 75%, and 100%). Asecond session for the 2-point approach, involving only resisted sprints against the same 25% and75% loads. Individual L-v relationships were compiled using full data from the multiple-point testing(standard approach) and 2-point relationships extracted from the multiple and 2-point testingsessions (indirect and direct approach, respectively). The main outcomes of individual L-vrelationships—maximal theoretical load (L0) and maximal theoretical velocity (v0)—were comparedbetween the standard multiple-point method and both the direct and indirect 2-point methods.Results. Trivial to small differences (Cohen’s d ≤ 0.34), nearly perfect correlations (r ≥ 0.92), minorhomoscedastic errors (r² ≤ 0.03), and no fixed or proportional bias according to ordinary least-products regressions were observed between the standard and indirect and direct 2-point methods.Conclusion. In habituated athletes, the direct 2-point method is a practical tool for routinelyassessing the L-v relationship during resisted sled sprinting and estimating both v0 and L0. Thismethod offers several advantages over the standard multiple-load approach, including reducedtesting time, lower fatigue, and decreased physical burden from unresisted sprints.