Backgroud: Simulating the butterfly stroke is a biomechanical assessment tool used in dryland training. While these mechanical conditions affect the execution compared to the aquatic environment—due to the absence of drag and lift—they allow for the controlled study of variables.Aim: To characterize the percentages of muscle activation levels using surface electromyography (EMG) by normalizing the dynamic peaks of four upper body muscles during the propulsion and aerial phases of the butterfly stroke, under simulated (dryland) conditions with elastic resistance. Methods: Twelve master swimmers (6 men and 6 women; 37 ± 8 years) performed a simulated butterfly stroke at maximum intensity for 60 seconds. Swimmers were positioned prone with their lower limbs immobilized, allowing for the ascent and descent of the upper body during the simulated movement on a specially designed bench. Blind goggles were used to minimize external visual cues and facilitate proprioception. Resistance to the movement was generated by two elastic straps attached to palmar splints (hand wraps). Surface electromyography (EMG) was recorded bilaterally using the FREEMG 1000 BTS device (BTS Bioengineering, Italy) in the upper trapezius (TS), serratus anterior (SM), latissimus dorsi (DA), and pectoralis major (PM) muscles. Muscle selection was based on their involvement during the flight and propulsive phases, as well as on methodological criteria specific to surface electromyography, prioritizing superficial muscles with adequate anatomical demarcation and a lower risk of signal contamination (crosstalk). Electrode placement followed the recommendations of the SENIAM project. Signals were normalized using a dynamic peak model. A triaxial inertial sensor (G-sensor 2, BTS Engineering, Italy) positioned between C5 and C7 was used to temporally demarcate the phases (ascent and descent). The choice of this location was based on the recording of the upper trunk's ascent and descent phases, which are mechanical events associated with the transition between the two phases of the stroke. The recorded acceleration variations allowed for the identification of changes in the direction and behavior of the upper trunk's movement, establishing temporal references for the functional segmentation of the movement cycle. The study prioritized evaluation using the 10-second sliding window technique—the first between 20 and 30 seconds and the second between 50 and 60 seconds—focusing on describing a temporal organization. Results: The temporal analysis of the electromyographic curves showed differences in the distribution of muscle activation between the propulsion and aerial phases of the stroke cycle. During the propulsion phase, the upper trunk (TS) showed the highest activation levels (76.8%), followed by the middle trunk (PM) (73.3%), the upper body (DA) (72.9%), and the middle trunk (SM) (68.6%). During the flight phase, the stabilizing muscles (TS and SM) maintained a sustained activation level throughout the cycle, particularly the TS (77.9%) and SM (79.5%); while the propulsive muscles (PM and DA) maintained stable levels (PM: 76.6% and DA: 71.9%). EMG curves showed an absence of muscle relaxation during the transition and development of the flight phase, suggesting continuous neuromuscular recruitment throughout the entire cycle. The stabilizing muscles increased their participation from 49.9% to 51.5%, surpassing the propulsive muscles, whose participation decreased from 50.1% to 48.5% between phases. Interindividual variability was high (CV: 18.6%-28.7%). The PM group exhibited lower variability (CV: 18.6-20.6%), while the DA group presented the highest coefficients of variation (CV: 25.8%-28.7%). Conclusions: Simulation of the butterfly stroke technique allowed for the characterization of differential muscle activation patterns between the propulsion and aerial phases using surface electromyography. During the propulsion phase, activation levels were observed in all evaluated muscles, while in the aerial phase, the stabilizing muscles maintained sustained activation without showing the expected decreases during this phase. The increased percentage of muscle activation in the stabilizing muscles and the interindividual variability observed according to the coefficient of variation (CV) suggest functional differences and muscle recruitment strategies between phases and among the swimmers evaluated. Taken together, the findings provide biomechanical information on the percentages of muscle activation and their functional distribution between phases during dryland simulation of the upper body in the butterfly stroke. As a consequence of using the slings, these have led to the development of movements that adapt to the traction conditions, primarily during the aerial phase.
© Copyright 2026 XVth International Symposium on Biomechanics and Medicine in Swimming Proceedings. Alle Rechte vorbehalten.