What a step down the stairs tells us about muscle testing in pain and injury
Step off a stair in the dark and you're trusting your life to a reflex that works faster than you can think.
Most of us were taught to test that reflex with a rubber hammer. But what if we've been asking the wrong question.
The Myotatic Reflex on a Staircase
Picture the moment your foot lands on the step below.
Your body weight and downward speed drive the knee into flexion. The quadriceps are stretched, fast. If they don't contract hard enough, right now, the knee folds and you tumble down the stairs.
Here's what happens. The primary endings of the quadriceps muscle spindles are exquisitely sensitive to the speed of stretch. They fire a volley up the Ia afferents. Those fibres synapse directly onto the alpha motor neurons of the same muscle. One synapse. No brain required.
At the same time, Ia inhibitory interneurons turn down the hamstrings. That's reciprocal inhibition, and it stops the antagonist fighting the save.
The short-latency response in the quadriceps arrives in around 20 milliseconds. A voluntary reaction takes 150 milliseconds or more.
By the time you could decide to tighten your thigh, you've already either caught yourself or fallen.
The Knee Jerk Is a Lab Sample
That circuit is the knee jerk.
The patellar tap is a tiny, synchronous stretch delivered to a relaxed muscle in a seated patient. It tells you the arc is intact. It gives you a rough grade.
What it doesn't tell you is whether that same arc can stabilise the knee with 80kg coming down on it at speed.
The stair is the real-life version of the test. And on the stair, the muscle isn't relaxed.
The Nervous System Doesn't Wait for Contact
In downward steps, leg muscle activity starts well before the foot lands. Melvill Jones and Watt recorded calf muscle activity beginning around 140 milliseconds before contact.1 Dietz and colleagues showed that this pre-activity and the stretch reflex work together on landing.2
So the step is a stretch reflex acting on a muscle that is already contracting. That matters, because of something Matthews called automatic gain compensation.3
The stretch reflex response grows with the level of existing activity. The more the motor neuron pool is already firing, the bigger the reflex response to the same disturbance. Matthews traced this to the organisation of the pool itself: more active motor neurons to speed up, and more of them sitting close to threshold.
That's how the reflex copes with a heavy shopping bag or a faster descent. More stretch, more spindle firing. More drive, more gain.
Vision helps you choose where to put your foot. It's far too slow to help once the foot lands.
The Examiner Is the Step
This is what we try to reproduce in every Protective Reflex Test (PRT).
The patient contracts the muscle in its test position. The examiner applies an external force against that contraction, with a timing and size the patient can't predict. The examiner is the step.
The question isn't "how strong is this muscle?" It's "is there still capacity in this contraction to stabilise against a force from outside?"
A normal muscle holds. An inhibited muscle gives way.
Look at that through Matthews' lens. In a normal pool, the added stretch from the examiner meets motor neurons near threshold and gets a proportionate response. Now add a persistent abnormal afferent input, the nail in the foot. The net input to that pool shifts towards inhibition. The motor neurons sit further from threshold. The same added stretch recruits less.
The muscle has been turned down, not switched off. The patient can still contract it voluntarily. What's missing is the reserve to protect the joint when the outside world pushes back.
From the 20-Millisecond Catch to the Two-Second Test
The catch on the stair takes about 20 milliseconds. A PRT takes a second or two. So what is the test actually measuring?
The first 20 milliseconds is the monosynaptic response on its own. Over the rest of the test, three more things join in.
The patient's own drive builds. Motor neurons are recruited in order of size, smallest first, and the ones already firing speed up.4
Long-latency stretch responses follow, partly routed through the motor cortex.5
And as the background drive rises, Matthews' gain compensation makes the reflex response to the examiner's force bigger.
Every one of these runs through the same motor neuron pool. An abnormal afferent input that holds the pool away from threshold turns all three down together.
So the test isn't trying to isolate one synapse. It reproduces the whole job the stair demands, and it reports on the same thing the catch depends on: how close to threshold that pool is sitting.
Why the Dynamometer Misses It
Isokinetic dynamometry fixes the speed and lets the subject push against a predictable resistance. The force is voluntary, planned and steady. It's a good measure of voluntary output. It isn't a step.
A hand-held "make" test has the same problem. The patient pushes into a still device.
A "break" test comes closer, because the examiner overpowers the muscle. But it's designed to record peak force at failure, as a number in newtons. It doesn't ask whether the muscle locked at the moment an unexpected load arrived.
The physiology itself is measurable. Laboratories record reflex responses to sudden perturbations with EMG, and they find changes after injury. Blasimann and colleagues found altered short-latency hamstring responses after an acute ACL rupture, using stair descent and artificially induced tibial translation.6
But those are research rigs. You can't run one on twenty muscles in a 30-minute consultation.
The standard clinical instruments measure what the patient can do on purpose. The stair tests what the nervous system does without asking.
Kinesiology: Same Hands, Different Question
This is where muscle testing gets a bad name, so let's be precise.
Some forms of kinesiology use a muscle test as an indicator. A vial is placed in the hand, a food is held against the body, a statement is spoken, or a parent is tested on behalf of a child. The muscle's response is read as an answer about something else.
That's dowsing with an arm.
There's no known neural pathway by which a sealed vial in the hand changes motor neuron excitability. The evidence agrees. When three experienced applied kinesiologists assessed nutrient status, they didn't agree with each other, didn't match laboratory tests, and couldn't tell nutrients from placebo under double-blind conditions. The authors concluded it was no more useful than random guessing.7
PRT asks a different question. Can this muscle protect this joint against an external force, right now?
The answer is about the muscle. And it can be falsified. Find the source of abnormal afferent input, remove it, and retest. The muscle should lock. Reintroduce it, and the muscle should give way again.
The fair challenge to any manual test is reliability between examiners. That's the right question to ask, and PRT should be held to it like anything else.
What Changes in the Treatment Room
If the quadriceps give way on testing, ask yourself what that patient does on the stairs.
An inhibited muscle on the treatment table is a predisposition to injury on the stairs. The test finds it before the joint pays the price.
Test before you treat. Find out why the muscle is turned down. Restore the neural control first.
The stair doesn't care how much your patient can leg-press. It only cares whether the reflex arrives in time.
To your strength!