The evidence, distilled
Principles
Eight things the research supports about how growing athletes' tissues respond to gymnastics loading. Each links to the full literature summary behind it.
- 01
Tissues adapt to load — each on its own clock
Muscle rebuilds in days; tendon and bone improve their material quality over weeks to months and their size over years; cartilage conditions itself over even longer horizons. Most strikingly, the core collagen of the adult Achilles is laid down before roughly age 17 and never replaced — a teenage gymnast is literally constructing the tendon she will compete on forever. Training plans built around muscle's fast clock quietly outrun every slower tissue.
- 02
Intensity gates adaptation; light work doesn’t count
Tendon only adapts when strain exceeds a threshold — roughly 4.5% strain builds tendon while 3% does nothing, no matter the volume. Bone wants high-magnitude, high-rate, novel loading. High-rep, low-load conditioning is fine for other goals, but it does not build the tissues that landings depend on; genuinely heavy, slow strength work does.
- 03
The adaptive signal saturates; damage never does
Cells stop listening after a few dozen loading cycles per bout — most of the adaptive benefit is captured in the first 30–45 reps — but fatigue damage keeps accumulating linearly with every rep after that. The marginal landing in a long session carries full damage and almost no signal. This asymmetry is the mechanical heart of overuse injury.
- 04
Duration matters: fast loads damage, slow loads build
A tumbling-pass landing loads the Achilles for ~130 milliseconds — too fast for the viscoelastic tissue to transmit the deformation to its cells. Held constant for strain, 130 ms repetitions produced ~0% stiffness gain over 14 weeks while 3-second repetitions produced +57%. Landing-dominated weeks load the tendon hard enough to damage it, but not in a way that tells it to get stronger.
- 05
Recovery windows are where adaptation happens
Collagen synthesis peaks about 24 hours after loading and runs for roughly three days; bone's targeted repair of microdamage takes weeks to months. Stack hard, springy sessions closer together than the tissue's rebuild time, over and over, and breakdown outpaces repair even though no single session was excessive. Rest is not the absence of training — it is when training works.
- 06
The enemy is sudden change, not hard work
Tissues cope well with high loads they are accustomed to. The classic danger moments are spikes: the first weeks back after a layoff, a new surface, a sudden jump in tumbling volume before a competition. Across bone, tendon, and ligament, the injury literature keeps finding the same thing — rate of change of load predicts trouble better than load itself.
- 07
Underloading is not the safe default
Immobilized and unloaded tissue actively weakens — stress-deprived tendon cells increase production of collagen-degrading enzymes, and ligament loses strength within weeks of disuse faster than training rebuilds it. A flaring tendon needs its load changed, not removed: cut the springy, high-energy work, keep the heavy slow work, rebuild gradually.
- 08
Pain is a late and unreliable messenger
Cartilage has no nerves; subclinical tendon damage is silent; degenerated-looking tissue can feel fine while mildly affected tissue hurts. By the time pain arrives, the structural story is usually months old. Load must be managed by counting and progression, not by waiting for symptoms — and when a young athlete does report pain, it must be respected, because it is the system's last warning, not its first.
Principles on their own change nothing. See what they look like in practice →