For decades, movement was taught as a binary: one muscle contracts, its opposite fully releases. Recent research (2011–2026) dismantles that myth. The reality is subtler, more interesting, and it changes how we should train strength, speed and mobility.

The myth of the "switching off" antagonist

Reciprocal inhibition was first described by Charles Sherrington in the early 20th century: when the agonist contracts, spinal interneurons inhibit the antagonist so it releases. True — but incomplete. Ia sensory fibres from muscle spindles detect stretch, fire the agonist motoneurons, and at the same time hyperpolarise the antagonist motoneurons. On paper, movement should be perfectly smooth.

That only works under narrow conditions: light load, moderate velocity, athlete at rest. Move away from those and the nervous system shifts strategy. Measuring antagonist EMG during effort (Hirabayashi et al., 2019) shows they're never truly silent — just modulated, sometimes strongly, sometimes barely.

The 30% effort tipping point

Hirabayashi's 2019 study (Frontiers in Human Neuroscience) is one of the clearest. Researchers measured reciprocal inhibition between tibialis anterior and soleus across different maximal voluntary contraction (MVC) intensities. Below 15% MVC, Ia inhibition works beautifully: the antagonist quiets down, the joint moves smoothly and precisely.

At 30% MVC, everything shifts. Reciprocal inhibition collapses. The brain switches to another mechanism: presynaptic D1 inhibition, further upstream, which gates neurotransmitter release before it reaches the motoneuron. Why? To allow a protective co-contraction. Under heavy load, fully releasing the antagonist would be dangerous — the joint would lose stiffness, the ligaments would be exposed.

🔑 Key takeaway

  • • < 15% MVC: peak Ia inhibition → smooth, precise movement.
  • • ~30% MVC: switch to presynaptic D1 inhibition.
  • • Heavy load: protective co-activation stiffens the joint — useful, not a flaw.

Why velocity changes everything

Second decisive variable: velocity. In a sprint, throw or jump, the movement unfolds in hundredths of a second. The spinal reflex loop (stretch → spindle → cord → motoneuron) is simply too slow to keep up. If reciprocal inhibition fired at full strength, it would brake the agonist exactly when it needs to hit 100%.

The brain knows this. It actively suppresses the Ia loop during fast movements — a velocity-dependent suppression documented by wireless sEMG on elite sprinters. That's what lets the best sprinters reach rate-of-force-development peaks beginners can't touch. Tam et al. (2016) showed in elite Kenyan runners that the athletes with the lowest co-activation had the best running economy.

Stability vs speed: the continuum

The modern view (Latash, 2018) drops the inhibition / co-activation dualism. They're two ends of a continuum the brain constantly tunes. The more the task demands stability (heavy load, landing, unstable posture), the higher the co-activation. The more it demands pure speed (sprint, strike, throw), the more reciprocal inhibition returns and co-activation is suppressed.

The elite athlete isn't the one who "releases" antagonists better — they're the one who modulates them better depending on context. And that's what we train: the nervous system's ability to flow between modes.

💡 Red flag

An athlete who "locks up" on every sprint start, never finds fluidity in change of direction, or complains of chronic stiffness despite endless stretching, often has an excessive co-activation problem — not a strength problem. Software, not hardware.

Field applications: APS, oscillatory, PNF

🔥 Dynamic warm-up

A-skips, B-skips, carioca, athletic drills. Primes Ia afferents, sets up pre-activation. Long static stretching pre-effort is out before explosive work.

🔁 Agonist-antagonist paired sets (APS)

Chained agonist/antagonist. Rest ≥ 3 min for max strength; < 1 min for metabolic hypertrophy. Higher density, controlled fatigue.

⚡ Oscillatory & AFSM

Cal Dietz: pulsed micro-reps 7–10 cm around the sticking point, or violent reversal after active braking. Loads 25–55% 1RM. Sets < 10 s to stay alactic.

🧘 PNF in recovery

Contract-relax 5–10 s then stretch. Real effect: increased neurological tolerance (gate control), not structural lengthening. Great for rehab, avoid before explosive work.

🎯 Map your co-activation profile

Our 2-hour Full Assessment measures your RFD (rate of force development), peak isometric strength, left/right asymmetries and reactive stiffness index. The numbers reveal whether your neurological brake is tuned for stability or speed — and where to focus the next training blocks to shift it.

Discover the Full Assessment

Key takeaways

  • ✅ The antagonist never fully "switches off" — it's modulated by task.
  • ✅ Around 30% MVC, Ia inhibition collapses and presynaptic D1 inhibition takes over.
  • ✅ At high velocity, the brain suppresses the reflex loop to free the agonist to 100%.
  • ✅ Co-activation stabilises the joint under load; it's an asset, not a flaw.
  • ✅ Pure speed and running economy come from less co-activation, not more.
  • ✅ APS, oscillatory and AFSM are the modern tools to reprogram that dosing.

FAQ

Why don't my antagonists release when I want to be explosive?

Because your brain keeps a safety brake on. Until it trusts the structure (poor technique, unstable joint, past injury), it maintains a protective co-activation that caps your speed. This is reprogrammed through specific velocity work, not by adding raw strength.

Is co-activation always bad?

No, the opposite. Under heavy load or at landing, antagonist co-activation is what stabilises the joint and protects ligaments. The goal isn't to suppress it but to dose it: maximal when you need stability, minimal when you need to accelerate.

Do PNF stretches really lengthen the muscle?

Not structurally, at least short-term. The immediate gains come mainly from greater neurological tolerance to stretch, not from a longer muscle. It's useful in recovery and rehab, but real architectural change takes consistent practice over many weeks.

Fifth episode of our series "Physics & biomechanics of movement, explained simply". Next episode: "Why your ankles decide your sprint speed", 28 May 2026.