REF-702 · Gibala et al., 2000 (with Gibala et al., 1995) · Canadian Journal of Physiology and Pharmacology, · Paper

Myofibrillar disruption after identical resistance exercise: strength-trained 45% of fibres vs untrained 82%

Damage resistance is trainable

Blimited evidenceMeasured on trained to highly trained athletesWritten before October 2026, being re-checked

Protocol

Identical in both studies: 8 sets × 8 unilateral arm-curl repetitions at 80% of concentric 1RM, one arm performing only the concentric (lifting) phase, the other only the eccentric (lowering) phase. Needle biopsy of biceps brachii + electron microscopy to quantify the proportion of disrupted fibres.

Trained cohort (2000): n=6 strength-trained men, biopsied ~21 h post, after 5 days without training. Untrained cohort (1995, companion abstract): n=8 untrained men, biopsied immediately post and 48 h post.

Key Findings

Eccentric armConcentric arm
Untrained (1995, companion abstract)82% disrupted fibres immediately post; 80% at 48 h33% post; 37% at 48 h
Strength-trained (2000)45 ± 11% disrupted (baseline 4 ± 2%)27 ± 4% — not different from baseline
The eccentric (lowering) phase is the damaging one in both groups — but the concentric phase damages too in the untrained (33%, companion abstract), and not significantly in the trained.
In the trained, the authors state disruption was "considerably less severe" than previously observed in untrained subjects after an identical bout — both in proportion of fibres and in magnitude of disruption (sarcomere counting).
Mixed muscle protein synthesis over ~21-29 h post-exercise did not differ between the concentric and eccentric arms.
Untrained strength recovery (companion abstract): voluntary and evoked strength back to baseline within 24 h in the concentric arm, still depressed at 72-96 h in the eccentric arm.
The authors also conclude that fibre disruption from habitual weightlifting is essentially repaired after 5 days of inactivity in trained men.

Mechanism

Inference, not tested here: repeated mechanical loading is thought to remodel the muscle toward structural robustness, so the same relative load produces less sarcomere disruption. This is the cellular side of the repeated-bout effect (REF-549, REF-660); this paper measured the outcome (less disruption in the trained), not the remodelling itself.

Practical Application

In 6 strength-trained men, an identical 8 × 8 arm-curl bout at 80% of concentric 1RM disrupted 45 ± 11% of fibres in the eccentric arm (baseline 4 ± 2%) and left the concentric arm not different from baseline — versus 82% eccentric and 33% concentric in untrained men after the same bout (companion 1995 study). Damage resistance is a trainable quality, not a fixed trait; an athlete who "always blows up in the last hour" is not stuck there.
Inference, not tested: the training that plausibly buys it is high-force work — heavy gym lower-body and high-torque/low-cadence cycling efforts. Nothing here was measured on a bike; see limitations.
Not from this paper; the base's own practice: the first heavy gym block of a season (or after a lay-off) is expected to cause disproportionately more damage than the same session a few weeks later — start a strength block with deliberately conservative loads and expect the soreness curve to flatten quickly. This study supports the direction (training attenuates damage), not the timeline.

Limitations

Arm-curl / biceps model, n=6 and n=8. Nothing here was measured in a leg, on a bike, or in an endurance athlete.
The trained-vs-untrained contrast is between two separate studies, not a randomised comparison — cohorts differ in more than training status.
"Strength-trained" ≠ "endurance-trained". Whether an endurance cyclist's legs show the same protection from cycling-specific loading is untested here.
Uncontrolled pre–post design in the trained cohort (each arm vs its own baseline); no non-exercising control group.
How this entry was checked and changed
Why this level

regraded 2026-09-27 from B+: uncontrolled pre–post biopsy study, n=6; the trained-vs-untrained comparison is across two studies, not randomised

Content correction 2026-09-27
Level B+ -> B: uncontrolled pre–post biopsy study, n=6 (rubric); Sport line "strength, cycling" -> "strength" (arm-curl model; cycling transfer is discussed in the text only).
Untrained-cohort numbers (82/80/33/37%, 24 h, 72-96 h, n=8) are now marked as coming from the companion 1995 abstract (DOI 10.1152/jappl.1995.78.2.702), which was checked; added the abstract's protein-synthesis and 5-day-repair findings.
Mechanism (sarcomeres in series, cytoskeletal reinforcement) relabelled as inference; "three weeks later" and the conservative-first-block rule relabelled as the base's own practice.
Lead ▶ line now states the study's own numbers and population before the practical reading.