Figure 5. Mediation models showing indirect effects of internal-load indicators (X) on rectus femoris (RF) electromyography (EMG) outcomes (Y) via respiratory–metabolic variables (M). (E-G) Training impulse per minute (TRIMP·min-1) as X with peak oxygen uptake (VO2 mL·kg-1·min-1) as M and RF root-mean-square amplitude (RF_RMS), RF averaged EMG amplitude (RF_aEMG), or RF integrated EMG (RF_iEMG) as Y. (H) total energy expenditure (EE_total, kcal) as X, average VO2 (mL·kg-1·min-1) as M, and RF_aEMG as Y. Path coefficients (a, b, c′), bootstrapped 95% confidence intervals (CIs), and indirect effects (a×b) are shown. p < 0.05.