8.2 H Reflexes: Origin, Technique, and Clinical Use

8.2 H Reflexes: Origin, Technique, and Clinical Use

Quick Answer: The H reflex is a true reflex: Ia afferents make a monosynaptic connection onto homonymous motor neurons. The classic study is a tibial H to soleus or gastrocnemius (S1). Use submaximal, long-duration stimuli with the cathode proximal (toward the cord). The H appears before the M at low intensity and falls as the M grows.

If the F wave is a motor axon talking to its own cell body, the H reflex is a stretch-reflex arc run electrically. OpenExamPrep teaches the H reflex here as independent technologist study material: origin, setup, relationship to the M wave, habituation, and the clinical jobs it can and cannot do. Domain IV groups H reflexes with F waves because both are late responses, but they are not interchangeable procedures.

Origin: Ia afferents and a monosynaptic motor-neuron arc

Large Ia muscle-spindle afferents are the fastest sensory fibers from the muscle. When you stimulate a mixed nerve at an intensity that prefers those large fibers, the Ia volley races through the dorsal root into the spinal cord and monosynaptically excites homonymous alpha motor neurons—the motor neurons of the same muscle. Those motor neurons fire orthodromically to the muscle. The resulting CMAP is the H reflex (Hoffmann reflex).

Because the circuit includes a sensory afferent, a synapse, and a motor efferent, the H reflex is a true reflex. That is the opposite of the F wave. If the Ia fibers, dorsal root, or the Ia–motor-neuron synapse is interrupted, the H disappears even when motor axons can still produce an M wave. If you need a one-line contrast for the examination: H needs sensory in; F does not.

The synapse also explains habituation. Repeat the stimulus at a faster rate and the reflex can decrement. Keep the patient tense, talking, or poorly relaxed and the H can jitter or vanish. Those are synaptic and descending-modulation effects, not stimulator folklore.

The classic study: tibial H to soleus or gastrocnemius (S1)

The workhorse H reflex in the electrodiagnostic laboratory is the tibial nerve in the popliteal fossa recorded from soleus or medial gastrocnemius. That arc is an S1 study. It is not an L5 study. Fibular/peroneal motor recording from extensor digitorum brevis does not replace a soleus H when the clinical question is S1.

Some laboratories also record a flexor carpi radialis H reflex for a C6–C7 pathway. Know that it exists; do not let it distract you from the outline's classic tibial study. If the stem says S1 radiculopathy, polyneuropathy, or tibial pathway, think popliteal fossa → soleus/gastrocnemius.

The H reflex is not equivalent to an ankle jerk in all patients. A mechanical Achilles tap stretches spindles, engages gamma bias, and may recruit more than the soleus Ia–motor-neuron pair you isolate electrically. Electrical stimulation bypasses the spindle. Patients can therefore have an ankle jerk without a recordable H, or a recordable H when the bedside jerk is sluggish. Report the H as an electrical reflex study, not as a machine-scored ankle jerk.

Technique: submaximal, long duration, cathode toward the cord

Stimulus intensity and duration

Start low. Large Ia fibers have a low electrical threshold, especially with long pulse duration (commonly 0.5–1.0 ms). Short, high-intensity pulses favor motor axons and will jump you straight into an M wave. The technical goal is to recruit Ia afferents before you recruit the entire motor pool.

Increase intensity in small steps:

1. At the lowest effective intensity, an H may appear with little or no M. That dissociation is the best proof you are not looking at a volume-conducted M.
2. As intensity rises, a small M appears and the H typically grows to Hmax.
3. With still stronger shocks the M continues to grow and the H falls.
4. At supramaximal motor intensity the H is suppressed or reduced to a remnant, and any remaining late potential should be interpreted with F-wave rules, not H-wave rules.

That inverted-U relationship to the M wave is the signature of the H reflex. If your late potential only appears after the M is already maximal, you are not looking at a classic H.

Cathode proximal

Place the cathode toward the cord (proximal) and the anode distal. The H reflex is often obtained only, or most cleanly, with cathode-proximal orientation. Cathode-distal stimulation can send the afferent volley the wrong way relative to anodal block and reduces the Ia barrage heading for the spinal cord. If a skilled operator cannot obtain an H, check polarity before declaring the reflex absent.

Rate, relaxation, and matching sides

Use a slow repetition rate, often 0.5 Hz or slower, to limit habituation. The patient should be relaxed, with the knee slightly flexed and the recorded muscle quiet on the speaker. Precontraction and Jendrassik-type maneuvers can facilitate a small H, but they also add variability; for side-to-side latency work, match position, relaxation, electrode placement, and stimulus site.

Sweep and gain must show both the H and the M. The H is a CMAP, so it is larger than a typical F wave, but it is still easy to clip if you leave the motor-NCS gain in place from an abductor hallucis F study. Label the muscle. Soleus is not abductor hallucis.

Relationship to the M wave and habituation

The M wave is the direct motor response. The H wave is the reflex motor response. They share the same motor axons on the way out, which is why collision matters. When you stimulate strongly, the antidromic motor volley travels toward the cord and collides with the orthodromic reflex discharge coming down those same axons. Axons already depolarized in the M wave are also occluded and cannot contribute a separate H. Additional inhibitory circuits (including Ib and Renshaw effects at higher intensities) help extinguish the reflex.

Habituation is the other suppressor. Polysynaptic circuits habituate more than the oligosynaptic H, but even the H can fade if you hammer the nerve at a high rate. If the H melts away during a fast train and returns after a pause, that is reflex physiology, not a dying battery.

Compare that with the F wave from section 8.1: F persists at high intensity and does not depend on Ia input. A technologist who uses one intensity strategy for every late potential will mislabel H as F and F as H.

Hmax/Mmax ratio is used in some laboratories as a rough index of reflex excitability (higher in some upper-motor-neuron states, lower when afferents or motor axons are lost). For the R.NCS.T outline, latency and presence, especially side-to-side, matter more than memorizing a single ratio cutoff. If you report a ratio, use the laboratory's method and reference range.

Setup table: recording, stimulation, pitfalls

Volume-conducted M is the most important identification trap. An M wave generated in a nearby muscle can appear as a small, earlyish potential in your soleus channel. If the putative H grows in lockstep with a remote M, or never appears in the no-M window, it is not an H. The second trap is the wrong muscle: an abductor hallucis recording after tibial stimulation at the ankle is a routine motor NCS, not an S1 H reflex.

Clinical use

S1 radiculopathy

Unilateral S1 root disease can prolong or abolish the soleus H on the affected side. The most reproducible number is a side-to-side latency difference obtained with matched setups. Laboratories publish their own abnormal difference (often on the order of about 1.0–1.5 ms); use that local reference rather than inventing a universal cutoff from this chapter. Amplitude side-to-side ratios are more variable because relaxation and electrode placement move H amplitude around.

A normal H does not exclude every S1 problem, and an abnormal H does not prove that needle EMG will be silent. The H is an S1 pathway test, not a complete radiculopathy examination. It is more useful when needle access is limited, when you need a numbered side-to-side comparison, or when you are documenting a tibial reflex-arc lesion in a polyneuropathy workup.

Polyneuropathy and the tibial pathway

Length-dependent polyneuropathy can delay or eliminate H reflexes, sometimes when distal motor latencies are still only mildly abnormal, because the arc includes proximal and distal segments plus a synapse. Bilateral absence in an older patient can be technical, neuropathic, or age-related difficulty obtaining the reflex; interpret it with the rest of the study, not as a standalone diagnosis. Unilateral absence with a robust contralateral H and a preserved ipsilateral M is much more localizing to S1, the dorsal root, or the tibial Ia pathway on that side.

What the H reflex does not do

• It does not diagnose L5 radiculopathy.
• It does not replace sensory NCS of the sural or plantar nerves.
• It does not replace needle EMG.
• It does not equal the ankle jerk in every patient.
• It is a poor F-wave substitute: different intensity, different pathway, different interpretation.

Putting a tracing on the page

Imagine a 42-year-old with unilateral S1 pain. Right soleus H appears at low intensity at 32.0 ms, then shrinks as the M grows. Left soleus H appears at 30.4 ms with the same cathode placement, pulse width, and knee angle. Distal tibial CMAPs are symmetric. The finding worth reporting is the side-to-side H latency gap on a technically matched S1 reflex study, not a claim that the machine has confirmed a herniated disc.

If instead both H reflexes are unobtainable, both M waves are tiny, and sensory responses are absent, you are looking at a neuropathic or inexcitable nerve problem, not a clever H-reflex diagnosis. Technique still comes first: confirm cathode polarity, pulse duration, relaxation, and that you are on soleus before you call the reflex absent.

Used with those rules, the H reflex earns its place next to the F wave on the outline: same domain, opposite physiology, and a setup you can defend on an examination stem.