Search for a command to run...
Abstract Background: Long-term resistance training induces profound neural adaptations that extend beyond initial rapid gains. The nature of these adaptations appears highly specific to the training stimulus (maximal strength vs explosive power). Objective: This cross-sectional investigation quantified motor unit recruitment ceilings and rate coding characteristics in elite powerlifters (strength-dominant), elite Olympic weightlifters (power-dominant), and advanced recreational controls during progressive isometric loading. Methods: Forty male athletes were studied: • Elite powerlifters (PL, n=14; relative squat 2.41 ± 0.28 × BM) • Elite Olympic weightlifters (WL, n=12; Sinclair 138.2 ± 12.1) • Advanced controls (CTRL, n=14; relative squat 1.68 ± 0.19 × BM) All performed trapezoidal isometric knee-extension ramps (20–100% MVC) on a dynamometer with embedded force plates. High-density surface EMG (128 channels total, vastus lateralis) was decomposed (accuracy >92%). Recruitment ceiling defined as %MVC of highest-threshold motor unit. Olympic weightlifters also performed additional ballistic-intent trials. Results: One-way ANOVA revealed highly significant group effects (all p < 0.001). Recruitment ceiling: WL 78.2 ± 5.1% < PL 83.6 ± 4.8% < CTRL 92.8 ± 4.1% MVC (post-hoc all pairs p < 0.001, η²ₚ = 0.79). High-threshold motor unit maximal discharge rate: WL 58.4 ± 7.2 > PL 48.6 ± 6.3 > CTRL 33.9 ± 5.1 pps (all pairs p < 0.001, η²ₚ = 0.82). Normalized RMS-EMG at 80% and 90% MVC showed the same ordinal pattern (WL < PL < CTRL). Ballistic RFD (WL only): RFD 0–100 ms = 11 842 ± 1741 N·s⁻¹ (+30% vs ramp). Conclusion: A clear continuum exists across the force–velocity training spectrum: explosive training (WL) produces the lowest recruitment ceiling and highest rate-coding ceiling, pure strength training (PL) intermediate, and general resistance training (CTRL) the highest ceiling with lowest rate-coding capacity. This represents a task-specific expansion of the usable neural drive bandwidth and explains divergent performance phenotypes despite comparable hypertrophic potential. All primary analyses were conducted using a three-group design (PL vs WL vs CTRL)