REF-306 · Black MI, Jones AM, Blackwell JR et al., 2017 · Journal of Applied Physiology, · Paper · PubMed
Muscle Metabolic and Neuromuscular Determinants of Fatigue During Cycling in Different Exercise Intensity Domains
Participants
n=11 men (the abstract gives no training status or VO2max)
Protocol
Ramp incremental test to locate the gas exchange threshold (GET) and critical power (CP)
Constant-work-rate cycling to the limit of tolerance (Tlim) in each domain: 4-5 severe-intensity bouts (SEV, above CP), one heavy-intensity bout (HVY, between GET and CP), one moderate-intensity bout (MOD, below GET)
Muscle biopsies at Tlim (pH, [PCr], [lactate]); blood lactate and [K+]
Neuromuscular measures: vastus lateralis M-wave amplitude and neural drive (EMG) during exercise
Key Results
Severe domain: all SEV bouts (Tlim about 2-14 min) ended at a similar muscle metabolic milieu — low pH and [PCr], high [lactate] (P > 0.05 between bouts) — i.e. consistent "limiting" metabolite values
Metabolic perturbation at Tlim: greater after SEV than HVY, and greater after SEV and HVY than MOD (all P < 0.05). The abstract gives no numeric pH / PCr / lactate values
M-wave amplitude (VL) fell to a similar extent in all three domains: SEV −38 ± 15%, HVY −68 ± 24%, MOD −53 ± 29% (P > 0.05 between domains)
Neural drive to VL rose only during SEV (+4 ± 4%, P < 0.05); no change during HVY or MOD
During SEV and HVY, but not MOD, the rates of change in M-wave amplitude and neural drive correlated with changes in muscle [PCr], [lactate] and blood [lactate], [K+] (P < 0.05)
Recovery was not measured
Mechanism
Authors' interpretation: fatigue above CP is associated with reaching consistent limiting values of muscle metabolites; at lower intensities (heavy, moderate) substrate availability and limits to muscle activation may constrain performance instead. The boundaries GET and CP therefore separate qualitatively different ways of reaching exhaustion, not just slower versions of the same one.
Practical Application
In n=11 men cycling to exhaustion in each intensity domain, every bout above critical power ended at the same "limiting" muscle metabolic state (low pH and PCr, high lactate) after 2-14 min, the metabolic perturbation at exhaustion was SEV > HVY > MOD, yet the drop in M-wave amplitude was similar across domains (−38 to −68%, ns) and neural drive rose only above CP — what ends a severe effort is metabolite accumulation to a fixed ceiling; what ends a moderate or heavy effort is something else (substrate availability, muscle activation), so the limits are of different kinds, not just different sizes.
Supports treating CP (and GET) as real boundaries: work above CP draws down a finite metabolic reserve to a consistent end point, which is why time above CP is predictable from W' while time below it is not.
Inference, not tested: because the exhaustion mechanism differs by domain, the recovery demand may differ too — but this study measured nothing after exhaustion. Any recovery-hour rule between sub- and supra-threshold work must come from elsewhere, not from this paper.
Limitations
Time to exhaustion (not TT) as performance measure
n=11 men without stated training status; likely non-athletes
Acute response — not repeated training effect; recovery after exhaustion was not measured
Neuromuscular fatigue was characterised by M-wave and EMG during exercise, not by post-exercise force testing
How this entry was checked and changed
Why this level
regraded 2026-09-26 from A-: mechanistic, n=11
Content correction 2026-09-27
"Neuromuscular fatigue significantly greater after severe vs heavy vs moderate" -> the M-wave amplitude decline was similar across domains (SEV −38%, HVY −68%, MOD −53%; P > 0.05); only neural drive rose, and only in SEV.
Numeric pH / PCr-depletion / muscle-lactate table -> not in the abstract; deleted. The abstract states only that metabolic perturbation at Tlim was SEV > HVY and SEV, HVY > MOD.
"Recovery time proportional to perturbation", "double sessions vs 48 h+", "95-100% vs 105% FTP" -> recovery was not measured; these house rules were removed.