Validity and Reliability of Flywheel Resistance Technology as an Assessment Method and Its Association With Sports Performance and Asymmetry: A Systematic Review

ABSTRACT: Mak, MCW, Bishop, C, and Beato, M. Validity and reliability of flywheel resistance technology as an assessment method and its association with sports performance and asymmetry: A systematic review. J Strength Cond Res 39(7): e930-e948, 2025-Flywheel technology is commonly used in training but remains underused for monitoring and testing. Flywheel devices can provide valuable data from mechanical outputs during both concentric and eccentric movements. This systematic review assesses its validity and reliability for evaluating sports performance and limb asymmetry. Searches were conducted in PubMed, SPORT-Discus, Google Scholar, and Web of Science following PRISMA DTA guidelines, focusing on keywords related to flywheel testing. Results show that flywheel testing is reliable (intraclass correlation coefficient = 0.66-0.99, r = 0.69-0.97, α = 0.85-0.98) and valid for the athletic population when subjects undergo 2 familiarization sessions. The test can be conducted using rotary encoders, force plates, linear encoders, or inertial measurement units. Subjects may perform 1-2 prerepetitions followed by 5-10 recording repetitions with 1-4 sets and 2-3 minutes of rest between sets. Moment of inertia can be customized based on the athlete's experience and the type of flywheel device used. Key metrics for assessment encompass speed, force, and power, with peak power being the most commonly used parameter. Few samples of evidence showed that increased asymmetry in concentric power may negatively affect change of direction performance, emphasizing the need for more high-quality studies. In conclusion, flywheel technology offers valuable insights across various movements, providing strength and power assessment while potentially improving athletic performance and injury prevention. Continued research is vital to explore its effectiveness in diverse athletic contexts.