Curriculum/Pillar 2 · Nerve Conduction Studies

Late Responses: F-waves & H-reflex

Module 7 of 36·24 min read

Learning objectives

1Explain F-wave physiology and its proximal reach and chronodispersion
2Describe the H-reflex arc and its role in S1 assessment
3Use late responses in early demyelination and radiculopathy

Routine nerve conduction reaches only as far proximally as a stimulator can be placed, leaving the most clinically consequential territory — the proximal nerve, the root, and the spinal cord — electrically silent. The late responses solve this by exploiting the round trip. By sending an impulse all the way to the anterior horn and reading what returns, the F-wave and the H-reflex interrogate the proximal arc that no distal montage can see, and they do so through entirely different physiological mechanisms that determine exactly what each can and cannot localise.

F-wave physiology: recurrent motor backfiring

A supramaximal stimulus to a motor nerve launches impulses in both directions along the axon. The orthodromic volley descends to the muscle and produces the CMAP (the M-wave). The antidromic volley ascends to the anterior horn cells, where the invading impulse depolarises the initial segment and a small, fluctuating subset of motor neurons backfire — they generate a new action potential that travels orthodromically back down the very same axon to the muscle. This recurrent discharge, recorded as a small late potential after the M-wave, is the F-wave. Its defining features follow directly from this mechanism:

It is a pure motor phenomenon. There is no synapse in the loop — no Ia afferent, no interneuron — only antidromic invasion and recurrent discharge of the motor neuron itself, so the F-wave is unaffected by sensory pathology.
It samples the entire length of the motor axon, including the proximal segment, plexus, and ventral root that distal stimulation cannot reach. The impulse traverses that proximal territory twice, doubling the sensitivity of the latency to proximal slowing.
Each stimulus recruits a slightly different, small population of backfiring neurons (only ~1–5% of the motor pool responds to any given shock), so latency, amplitude, and morphology vary shot to shot. The F-wave is therefore an ensemble measurement, not a single waveform.

Because of that variability, F-waves are acquired as a train of 10–20 stimuli and quantified by three derived parameters:

Minimum F-latency — the shortest latency in the train, representing the fastest-conducting motor axons over the full proximal-to-distal round trip; corrected for limb length or height and compared against normative limits.
Chronodispersion — the spread between minimum and maximum latency, which widens when conduction velocities across the proximal segment become heterogeneous (acquired demyelination).
Persistence — the proportion of stimuli that elicit an identifiable F-wave; reduced persistence or frank absence signals failure of proximal conduction or motor-neuron excitability.

supramax

Stimulus intensity (F-wave)

~1–5%

Motor pool backfiring / shock

10–20

Stimuli per train

proximal

Segment uniquely sampled

What prolonged or absent F-waves mean

A prolonged minimum F-latency or reduced persistence, with normal distal conduction, implicates the proximal segment that routine studies miss. This is the basis of the F-wave's signature clinical role in early Guillain–Barré syndrome, where F-waves may be delayed, dispersed, or absent before distal parameters change — often the earliest electrophysiological abnormality in the illness — and in polyradiculopathy, where proximal demyelination or root compromise lengthens the round-trip time.

H-reflex physiology: the electrical ankle jerk

The H-reflex is the electrical analogue of the ankle jerk and, unlike the F-wave, a true monosynaptic reflex. A submaximal stimulus to the tibial nerve in the popliteal fossa preferentially excites the large-diameter, low-threshold Ia sensory afferents arising from soleus muscle spindles. These afferents conduct orthodromically to the spinal cord and synapse directlyon alpha motor neurons — a single synapse, the same arc as the tendon reflex but bypassing the muscle spindle's mechanical transduction — which then discharge and produce a reflexive CMAP in soleus. Because the loop traverses the S1 root on both its afferent and efferent limbs, the H-reflex is a sensitive, though anatomically non-specific, probe of the S1 reflex arc.

The H-reflex inverts the stimulus–response logic of the F-wave, and the reason is collision. At low intensity, only the large Ia afferents are recruited and the reflex appears with little or no M-wave. As intensity rises, the stimulus begins to excite motor axons directly; the resulting antidromic motor volley collides with and cancels the orthodromic reflex discharge descending from the cord, so the H-reflex shrinks and disappears precisely as the M-wave grows to maximum. The H-reflex is thus elicited at low intensity and abolished at high intensity — the mirror image of the supramaximal F-wave.

In adults the H-reflex is reliably obtainable at only two sites: the soleus (tibial nerve, S1) and the flexor carpi radialis (median nerve, C6–C7). A unilaterally absent or delayed soleus H-reflex, interpreted against the contralateral side, supports an S1 radiculopathy and is one of the few electrodiagnostic measures sensitive to a purely demyelinating or early root lesion that leaves the needle examination normal. Its limitation is non-specificity: any lesion anywhere along the long arc — sciatic, tibial, plexus, or root — can abolish it, and once lost it does not recover, so it cannot date a lesion.

A-waves

An A-wave (axon reflex) is a small, late potential of constant latency and morphology, typically appearing between the M-wave and F-wave at submaximal intensities. It arises when an antidromic impulse reaches a branch point and turns back down a collateral sprout to the muscle — a re-routing along a fixed anatomical path, which is why, unlike the variable F-wave, it is invariant from shot to shot. A-waves are a normal occasional finding but, when frequent, are a marker of collateral reinnervation and are seen in chronic neuropathy, radiculopathy, and early GBS.

Clinical Pearl

F-waves and the H-reflex are complementary, not interchangeable, and the difference is anatomical specificity. The H-reflex tests one defined sensory-to-motor arc (S1, monosynaptic) and is exquisitely sensitive there but is available at only a couple of sites and cannot say where in the long loop the fault lies. The F-wave tests pure motor conduction along any accessible nerve and surveys the whole proximal course, but it localises nothing. Use the H-reflex to ask "is the S1 arc intact?" and the F-wave to ask "is proximal motor conduction normal?"

Common Pitfall

F-waves are the most variable measurement in the laboratory, so a single prolonged sweep proves nothing; only the minimumlatency across an adequate train (and persistence) is interpretable, and an unwarmed limb prolongs every late response non-specifically. For the H-reflex, the cardinal error is driving stimulus intensity too high: because the antidromic motor volley cancels the reflex by collision, an over-strong stimulus abolishes a perfectly normal H-reflex and manufactures a false "absent" result. Start low, build the M/H recruitment curve, and report the H-reflex at the intensity that maximises it — not at supramaximal output.

Key points

F-waves are recurrent antidromic backfiring of anterior horn cells — pure motor, no synapse — that uniquely sample the whole axon including the proximal segment and root.
Quantify F-waves as minimum latency, chronodispersion, and persistence across a 10–20 stimulus train; they vary shot to shot.
Prolonged/absent F-waves with normal distal studies implicate proximal pathology — the earliest sign in GBS and a marker of polyradiculopathy.
The H-reflex is a monosynaptic Ia→alpha-motor reflex (the electrical ankle jerk) testing the S1 arc; elicited at low intensity, abolished at high intensity by antidromic collision; reliable only in soleus and FCR.
A-waves are constant-latency late potentials from collateral re-routing, a marker of reinnervation; the H-reflex is sensitive but non-localising and never recovers once lost.