Low reliability and high redundancy in inter-limb asymmetry: Effects of formula and variable selection

Inter-limb asymmetry monitoring is widely used for athletic performance and injury-risk assessment, yet the longitudinal psychometric impact of formula selection remains poorly understood. This study examined the methodological variability and reliability of ten commonly used inter-limb asymmetry formulas. Using a six-week, weekly repeated-measures design, thirty trained youth soccer players (age: 15.6 ± 1.07 years) completed unilateral countermovement jumps (CMJ) and unilateral flywheel leg press assessments, from which jump height and six concentric and eccentric power metrics were derived. Pearson correlation analysis revealed substantial mathematical redundancy, reducing the ten formulas to 3-4 distinct computational groups. The KOBAYASHI, ROBINSON, and WONG formulas showed the highest relative reliability (ICC: 0.152-0.367), although all formulas remained poor in absolute terms (ICC < 0.40). Mean concentric power emerged as the most reliable variable (ICC: 0.184-0.367). CMJ height asymmetries exhibited the lowest absolute thresholds for change (MDC₉₅: 10.08-33.37), indicating comparatively greater sensitivity for detecting longitudinal adaptations than eccentric-phase metrics. Asymmetry magnitude and reliability therefore depend on both the formula and movement phase analyzed. Practitioners should prioritize CMJ height and mean concentric power for stable monitoring, and researchers should explicitly report formulas and their psychometric properties to ensure clarity and comparability across sports science studies.