Nerve Conduction Studies

In patients with radiculopathy, nerve conduction studies typically are normal, and the electrodiagnosis is established with needle EMG (Box 29–1). Although some motor abnormalities are occasionally seen in radiculopathy, the more important reason to perform nerve conduction studies is to exclude other conditions that may mimic radiculopathy, especially entrapment neuropathy and plexopathy. In cases of upper extremity lesions, ulnar neuropathy at the elbow and CTS must be excluded. Ulnar neuropathy and C8 radiculopathy both can present with pain in the arm associated with numbness of the little and ring fingers. Likewise, pain in the arm with paresthesias involving the thumb, index, and middle fingers may be seen in C6–C7 radiculopathy and CTS. In the case of lower extremity symptoms, one must exclude peroneal neuropathy at the fibular neck. Both peroneal palsy and L5 radiculopathy may present with pain in the leg, accompanied by footdrop and paresthesias over the dorsum of the foot and lateral calf. In more severe cases, the clinical differentiation between a radiculopathy and a common entrapment usually is straightforward. In mild or early cases, however, the distinction often is more difficult, and nerve conduction studies are useful to either demonstrate or exclude an entrapment neuropathy.

Depending on the underlying pathophysiology and the level of the lesion, abnormalities occasionally may be seen on routine motor conduction and F response studies in radiculopathy. If the pathophysiology is predominantly demyelinating, the underlying axons remain intact. In that case, any motor study, stimulating and recording distally, will show a normal latency, conduction velocity, and compound muscle action potential (CMAP) amplitude. The only possible abnormality will be in the F responses. Because the F responses assess conduction both distally and proximally, abnormal F responses with normal distal conduction studies suggest a proximal lesion, either in the proximal nerve, plexus or roots. Of course, F waves will be abnormal only if the recorded muscle is innervated by the affected nerve roots.

In the upper extremity, F waves are routinely recorded only for the median and ulnar nerves, which are C8–T1 innervated. Thus, median and ulnar F-wave abnormalities may be seen in C8–T1 radiculopathy; however, these roots are infrequently affected by disc or bone impingement, the most common causes of radiculopathy. A radiculopathy at C5, C6, or C7, which are more common sites of root impingement, will not be reflected in the median or ulnar F responses. The situation is different in the lower extremities. The distally recorded peroneal and tibial muscles (extensor digitorum brevis, abductor hallucis brevis) are innervated predominantly by the L5 and S1 nerve roots, respectively. These levels are often affected by radiculopathy. Thus, in L5–S1 radiculopathies, peroneal and tibial F responses may be prolonged, especially in comparison with the contralateral side.

The H reflex occasionally is helpful in evaluating lower extremity radiculopathy. However, the H reflex, recorded from the soleus, can be used to evaluate only a possible S1 radiculopathy and is most useful when the symptomatic side is compared with the asymptomatic side. The H reflex is the electrical correlate of the ankle reflex; accordingly, it may be delayed or absent in any lesion that depresses the ankle jerk, including polyneuropathy, sciatic neuropathy, lumbosacral plexopathy, and S1 radiculopathy. Unfortunately, the combination of normal distal motor nerve conduction studies and an abnormal H reflex cannot help differentiate between plexopathy and radiculopathy, but can only suggest a proximal lesion.

If the pathophysiology also involves axonal loss, nerve conduction abnormalities may be seen in the motor conduction studies. Here again, abnormalities are seen only if the recorded muscle is innervated by the affected nerve root. Axonal loss may result in a decreased CMAP amplitude, with some slowing of conduction velocity and distal latency, especially if the largest fibers are involved. For instance, in an L5–S1 radiculopathy associated with axonal loss, the ipsilateral peroneal and tibial motor responses may have slightly slowed conduction velocities, slightly prolonged distal latencies, and reduced CMAP amplitudes, especially in comparison with the contralateral side. The distal latency prolongation and conduction velocity slowing, however, should never drop into the demyelinating range.

Sensory studies are the most important part of the nerve conduction studies in the assessment of radiculopathy. The sensory nerve action potential (SNAP) remains normal in lesions proximal to the dorsal root ganglion (Figure 29–3). Nearly all radiculopathies, including those caused by compression from herniated discs and spondylosis, damage the root proximal to the dorsal root ganglion (Figure 29–4). Conversely, lesions at or distal to the dorsal root ganglion result in decreased SNAP amplitudes if they are associated with axonal loss. Thus, lesions of the plexus and peripheral nerve (proximal and distal nerve) are associated with abnormal SNAPs, whereas lesions of the nerve root result in normal SNAPs.

FIGURE 29–3 Sensory and motor potentials in axonal loss lesions distal and proximal to the dorsal root ganglion.

A: Normal. B: Lesion proximal to the dorsal root ganglion. C: Lesion at or distal to the dorsal root ganglion. In axonal loss lesions, both proximal and distal to the dorsal root ganglion, degeneration of motor fibers results in decreased compound motor action potential amplitudes. If larger motor fibers are lost, conduction velocities and distal latencies also may slow slightly. The situation is different for sensory fibers. Lesions proximal to the dorsal root ganglion result only in degeneration of sensory fibers proximally into the spinal cord. Because the dorsal root ganglion is a bipolar cell, it remains in continuity with the distal sensory fibers. Therefore, sensory nerve action potentials (SNAPs), when stimulated and recorded distally, remain normal. In axonal loss lesions at or distal to the dorsal root ganglion, distal sensory fibers degenerate, as do the motor fibers. Accordingly, SNAPs are reduced in lesions of the plexus and peripheral nerve but are normal in radiculopathies and other lesions proximal to the dorsal root ganglion.

FIGURE 29–4 Radiculopathy and sparing of the dorsal root ganglion.

Herniated intervertebral discs are a common cause of cervical and lumbosacral radiculopathy. Herniated discs are most often lateral and posterior, with the dorsal root ganglion located distal to the herniation. In disc herniations, this anatomic relationship results in injury to the roots but sparing of the dorsal root ganglion and the peripheral sensory nerves. Consequently, sensory conduction studies remain normal in radiculopathy.

(From Wilbourn, A.J., 1993 Radiculopathies. In: Brown, W.F., Bolton, C.F. (Eds.), Clinical electromyography, 2nd ed. Butterworth, Boston. With permission.)

It is always imperative to check the SNAP that is in the distribution of the sensory symptoms (Table 29–5). For instance, if a patient has pain down the arm with tingling and paresthesias of the middle finger, the median sensory response to the middle finger should be checked. In such a case, if the lesion is at or distal to the dorsal root ganglion (e.g., in the brachial plexus or median nerve) and there is axonal loss, the SNAP amplitude will be abnormal, if enough time has passed that wallerian degeneration has taken place. On the other hand, if the lesion is proximal to the dorsal root ganglion (e.g., C7 radiculopathy), the SNAP amplitude will be normal. The presence of a normal SNAP yields important diagnostic information. A normal SNAP in the same distribution as sensory symptoms and signs should always suggest a lesion proximal to the dorsal root ganglion (although a proximal demyelinating or acute peripheral nerve lesion also can result in a normal SNAP). One important rare exception to his rule is discussed below.

Table 29–5 Sensory Potentials to Check in Radiculopathy

SNAP, sensory nerve action potential.

Note: SNAPs are normal in lesions proximal to the dorsal root ganglion, including lesions resulting in radiculopathies. When evaluating a possible radiculopathy, one should examine at least one SNAP in the distribution of the suspected radiculopathy. For example, the ulnar SNAP to the little finger should be normal in a C8 radiculopathy. If it is abnormal, the lesion likely is not at the root, unless there is another reason for the SNAP to be abnormal, such as a superimposed ulnar neuropathy at the elbow.

Electromyographic Approach

The needle EMG strategy in radiculopathy is straightforward. Distal, proximal, and paraspinal muscles in the symptomatic extremity are sampled, looking for abnormalities in a myotomal pattern that are beyond the distribution of any one nerve (Box 29–2). It is important to exclude a mononeuropathy, polyneuropathy, or more diffuse process that might account for the signs and symptoms.

Table 29–6 Electromyography in Upper Extremity Radiculopathy: Most Useful Muscles to Sample

Note: Green squares indicate “marker” muscles that are most often abnormal for that root in an isolated radiculopathy. Blue squares indicate muscles that may be involved, but are abnormal less frequently. This chart shows those muscles that are most helpful in making the electrodiagnosis of radiculopathy but does not indicate the entire myotomal representation of the individual muscle (see Table 29–3).

From Wilbourn, A.J., 1993. Radiculopathies. In: Brown, W.F., Bolton, C.F. (Eds.), Clinical electromyography, 2nd ed. Butterworth, Boston, with permission.

Table 29–7 Electromyography in Lower Extremity Radiculopathy: Most Useful Muscles to Sample

It may be Difficult to Localize a Radiculopathy to a Single Root Level

Although the EMG study is a sensitive test for identifying radiculopathy, it may still be difficult to identify the specific segmental level because most muscles are innervated by more than one myotome. For instance, the finding of fibrillation potentials with decreased recruitment of MUAPs in the biceps, deltoid, infraspinatus, and mid-cervical paraspinal muscles is consistent with a C5–C6 myotomal pattern. In such a case, one can be certain that the lesion is not due to a single peripheral nerve injury because abnormalities are present in muscles innervated by the dorsal rami as well as by the musculocutaneous, axillary, and suprascapular nerves. However, in such a case, it is more challenging and sometimes impossible to differentiate between a C5 and C6 radiculopathy.

In such a case, one would next sample muscles that belong to one but not the other myotome. For example, it would be helpful to sample muscles with partial C5 but without C6 innervation, as well as muscles with partial C6 but without C5 innervation. For instance, if the rhomboids (C4–C5) were sampled and found to be normal while the pronator teres (C6–C7) showed fibrillation potentials, then a C6 lesion would be more likely than a C5 lesion. The same approach is used to identify radiculopathies at other levels. One can see that it is just as important to identify which muscles are normal as which are abnormal to try to identify the specific root level involved. Often multiple muscles must be sampled to try to define the level of the involved myotome.

In studies of patients who had a surgically defined single-level radiculopathy, the correct level often could be deduced from extensive needle EMG studies (Figures 29–5 and 29–6). However, not infrequently there was significant overlap between adjacent segments, making a single root localization difficult. The most difficult levels to differentiate were C6 from C7.

FIGURE 29–5 Cervical radiculopathy: needle electromyographic results in 50 patients grouped by surgically defined root level of involvement.

Closed circles represent positive waves or fibrillation potentials, with or without neurogenic recruitment and motor unit action potential changes. Half-closed circles represent neurogenic recruitment changes only. Open circles represent normal examination. ADM, abductor digiti minimi; ANC, anconeus; APB, abductor pollicis brevis; BIC, biceps; BRAC, brachioradialis; DEL, deltoid; EDC, extensor digitorum communis; EIP, extensor indicis proprius; FCR, flexor carpi radialis; FDI, first dorsal interosseous; FPL, flexor pollicis longus; INF, infraspinatus; PSP, paraspinal muscle; PT, pronator teres; SUP, supraspinatus; TRIC, triceps.

(From Levin, K.H., Maggiano, H.J., Wilbourn, A.J., 1996. Cervical radiculopathies: comparison of surgical and EMG localization of single-root lesions. Neurology 46, 1022–1025. With permission.)

FIGURE 29–6 Lumbosacral radiculopathy: needle electromyographic results in 43 patients grouped by surgically defined root level of involvement.

Closed circles represent positive waves or fibrillation potentials, with or without neurogenic recruitment and motor unit changes. Half-closed circles represent neurogenic recruitment changes only. Open circles represent normal examination. ADQ, abductor digiti quinti; AH, abductor hallucis; AL, adductor longus; BFLH, biceps femoris, long head; BFSH, biceps femoris, short head; EDB, extensor digitorum brevis; EHL, extensor hallucis longus; GMED, gluteus medius; GMX, gluteus maximus; H, H-reflex; IL, iliopsoas; LG, lateral gastrocnemius; MG, medial gastrocnemius; PL, peroneus longus; PSP, paraspinal muscles; PT, posterior tibialis; RF, rectus femoris; ST, semitendinous; TA, tibialis anterior; TFL, tensor fascia latae; VL, vastus lateralis.

(From Levin, K.H., 2002. Electrodiagnostic approach to the patient with suspected radiculopathy. Neurol Clin 20, 397–421. With permission.)

If the Lesion is Acute, the Electromyographic Study may be Normal

As noted earlier, during the first 10 to 14 days after the onset of an acute radiculopathy, there are no needle EMG abnormalities except for decreased recruitment of MUAPs in weak muscles. Because it is unusual to find significant weakness in radiculopathy, the EMG study often is completely normal in the acute setting. Fibrillation potentials take several weeks to develop in the more distal limb muscles; therefore, it often is best to wait several weeks before sending a patient for an EMG study, unless one is willing to repeat a normal study after several weeks to look for new changes.

If the Radiculopathy is Purely Demyelinating, the Electromyographic Study will be Normal

If the nerve root is compressed, resulting in demyelination without axonal loss, the needle EMG study may be completely normal. Diagnosing radiculopathy with EMG usually rests on the identification of denervation and reinnervation, signs of axonal loss. If there is no axonal loss, the study usually is normal. Only if demyelination results in significant conduction block, with accompanying weakness, will MUAP recruitment be diminished. This situation, however, is rarely seen in radiculopathy.

If the Sensory Nerve Root is Predominantly Affected, the Electromyographic Study will be Normal

Most patients with radiculopathy have prominent sensory symptoms, including pain and paresthesias, indicating dysfunction of the sensory nerve root. If the sensory nerve root is preferentially affected and the motor nerve root is spared, the EMG study will be normal. Unfortunately, there is no good way to assess the proximal sensory segments using routine nerve conduction studies. Somatosensory evoked potentials are often used to assess the proximal segments, but they share many of the limitations of the F response. Most areas of skin are innervated by multiple dermatomes. Accordingly, although a single spinal nerve root may be severely damaged with resultant slowing, the somatosensory evoked latencies may be normal because the adjacent nerve roots (and the overlapping dermatomes they innervate) are not affected.

Different Fascicles may be Preferentially Affected or Spared

Just as in other entrapment syndromes, it is not unusual for some fascicles within a myotome to be affected while others are spared. In fact, some muscles of a particular myotome may be markedly involved, whereas others are affected only minimally or not at all. For example, in some C7 radiculopathies, the triceps may show fibrillation potentials and reduced recruitment of MUAPs, whereas the flexor carpi radialis is essentially normal, although both receive substantial C7 innervation. Thus, the yield of abnormal EMG findings in a radiculopathy clearly increases as more muscles are sampled. As always, however, the electromyographer must balance patient comfort, the length of the test, and the goal of obtaining as much useful information as possible.

The Paraspinal Muscles may be Normal

One expects the paraspinal muscles to be abnormal in radiculopathy, and they often are (Figures 29–5 and 29–6). In some cases, however, they are normal. This may be due to fascicular sparing of fibers to the dorsal rami or may simply be due to sampling error. In addition, some patients have difficulty tolerating the paraspinal examination and consequently may not be able to relax those muscles. The paraspinal needle examination is best done with the patient lying on his or her side in the fetal position, with the side to be studied facing up. This position often will relax the paraspinal muscles. If relaxation is incomplete, however, it may be difficult or impossible to exclude denervation. This situation is encountered most often when studying the thoracic paraspinal muscles.

In addition, reinnervation, like denervation, occurs first in the most proximal muscles. Accordingly, the paraspinal muscles are the first to be reinnervated, often resulting in a pattern of denervation in the limb muscles with sparing of the paraspinals, a pattern equally consistent with plexopathy. In such a case, the finding of normal SNAPs in the distribution of sensory complaints can help to differentiate plexopathy from radiculopathy. One also can look for reinnervated MUAPs in the paraspinal muscles, although it is sometimes difficult for patients to activate these muscles.

Abnormal Paraspinal Muscles are Useful in Identifying a Radiculopathy but Not the Segmental Level of the Lesion

The paraspinal muscles (also known as the erector spinae muscles) run along the spine from the occipital bone in the skull down to the sacrum. Functionally, they are divided into three groups: (1) the iliocostalis (superficial, lateral); (2) the longissimus (superificial, medial); and (3) the multifidus (deep, adjacent to the spinous process and lamina). In the superficial layers of the iliocostalis and longissimus, which are most often sampled during needle EMG, there is marked overlap in the innervation. Thus, a radiculopathy at one level, determined by examination of the limb muscles, may show denervation in the paraspinal muscles not only at the involved level but also at one or more levels above and below. Therefore, EMG abnormalities in these paraspinal muscles can be used only to mark the lesion as at or proximal to the nerve root level, but cannot be used to identify which specific nerve root level is involved.

In contrast, the deep layer of paraspinal muscles, the multifidus, are invariably innervated by only a single nerve root – the one below the spinous process from which they originate. Thus, if these muscles are sampled and are abnormal, the abnormalities are specific to the root (and level) supplying that muscle. The multifidus muscles laterally flex and rotate the spine to the opposite side. The following technique to identify and sample the multifidus has been described:

FIGURE 29–7 Needle EMG sampling of the multifidus muscle.

The paraspinal muscles consist of three groups: the iliocostalis which is superficial and lateral, the longissimus which is superficial and medial, and the multifidus which is deep. Multiple root levels innervate the superficial iliocostalis and longissimus muscles. However, the multifidus is only supplied by a single nerve root. Techniques to sample the multifidus have been described (see text). In this MRI illustration of the lumbar spine, the EMG needle is in the multifidus muscle (red). Note that the multifidus is deep and adjacent to the spinous process and lamina.

Using this technique, one samples the multifidus muscle above the spinous process that was originally marked. For example, if one identifies the L4 spinous process, this technique would then sample the L3 multifidus muscle, innervated by the L3 root.

Although this technique is appealing because it tries to determine the level of the radiculopathy based on the paraspinal exam, in practice it has several limitations. In cadaver studies, the accuracy of sampling the correct multifidus was approximately 80% (i.e., 20% false-positive rate). Second, it relies on identifying the spinous processes and their correct corresponding levels on the basis of anatomic landmarks – either counting up from the lowest spinous process (L5), and/or identifying the L4 spinous process as being at the highest level of the iliac crests. In many individuals, these landmarks can be difficult to determine; in overweight patients, it is nearly impossible. Thus, with these limitations in mind, it is prudent to use abnormalities in the paraspinal muscles as a marker that the lesion is at or proximal to the root level, but leave the determination of the actual root level to the pattern of abnormalities seen in the limb muscles.

There is no Difference between a Radiculopathy/Polyradiculopathy and Focal/Diffuse Motor Neuron Disease on the Electromyographic Study

This is a very important concept often overlooked by electromyographers. Based on EMG results alone, an abnormality of the nerve root cannot be distinguished from an abnormality of the motor neurons supplying that root. The EMG and nerve conduction studies are identical in both conditions. First, all sensory nerve conduction studies are normal in both conditions. In radiculopathy, the sensory roots may be involved, but the SNAPs remain normal because the lesion is proximal to the dorsal root ganglion. In motor neuron disease, the sensory nerves are not involved. Second, motor conduction studies are normal in both radiculopathy and motor neuron disease, unless the muscles being recorded are innervated by the involved roots or anterior horn cells. If there is axonal loss from nerve root or anterior horn cell damage, the CMAP amplitudes may be decreased, with slight prolongation of distal latency and slight slowing of conduction velocity in both situations. Third, in both conditions the EMG shows denervation, reinnervation, or both in the involved myotomes and paraspinal muscles. Consequently, there is no electrophysiologic difference between a polyradiculopathy and diffuse motor neuron disease, or between an isolated radiculopathy and motor neuron disease affecting a single segment.

Although EMG cannot differentiate between a disorder of the nerve roots and a disorder of the motor neurons, there are clear and unequivocal clinical differences that allow the distinction to be made. For example, there likely would be no difference between the EMG of a patient with motor neuron disease, such as amyotrophic lateral sclerosis (ALS), and that of a patient with malignant lymphoma diffusely infiltrating multiple nerve roots and cranial nerves. The F responses might be abnormal in polyradiculopathy, in contrast to motor neuron disease, but otherwise the studies could be identical (normal SNAPs, diffuse denervation and reinnervation). However, the clinical presentation and neurologic examination certainly would be markedly different. In motor neuron disease, there are no sensory signs or symptoms. In contrast, pain and paresthesias are prominent in polyradiculopathy. Deep tendon reflexes usually are depressed or absent in polyradiculopathy, whereas they are increased or present in ALS (although they may be depressed in the progressive muscular atrophy variant of ALS). These points once again underscore that EMG and nerve conduction studies can be properly interpreted only in the context of the clinical history and physical examination.

Fibrillation Potentials may Persist in the Paraspinal Muscles after Spinal Surgery

Patients with recurrent or persistent pain after disc surgery often are referred to the EMG laboratory. However, the interpretation of fibrillation potentials in the paraspinal muscles of such patients is not straightforward. Patients who have undergone successful disc surgery and no longer have symptoms or signs of radiculopathy have been demonstrated to have persistent fibrillation potentials in the paraspinal muscles, often for several years. It is not clear why this occurs, but it may be related to the surgical scar through the paraspinal muscles. For this reason, the paraspinal EMG examination no longer assumes the same diagnostic importance in postsurgical patients, and it is questionable whether sampling the paraspinals is worthwhile in such patients (i.e., the absence of denervation cannot exclude a radiculopathy, and the presence of denervation may be a “normal” finding many years after spinal surgery and is of no clinical significance).

Only the Distal Muscles may be Abnormal in Radiculopathy

The diagnosis of radiculopathy is based on demonstrating neuropathic changes in distal, proximal, and paraspinal muscles in a myotomal pattern. Reinnervation, like denervation, occurs in proximal before distal muscles. In addition, the more proximal the denervation, the more successful the reinnervation. Thus, if the proximal limb and paraspinal muscles have been successfully reinnervated, chronic radiculopathies may show denervation only in the distal muscles. Once this reinnervation has occurred in proximal muscles, it may not be possible to differentiate a radiculopathy from a plexopathy or distal neuropathy by the needle EMG examination alone.

There may be Few or no Electromyographic Abnormalities in Spinal Stenosis

Lumbosacral spinal stenosis is a common condition, especially in the elderly. Patients often develop neurogenic claudication (pain and paresthesias in the legs with standing and walking, relieved by sitting). This condition results from intermittent compression of the lumbosacral nerve roots. Because the symptoms are intermittent and occur only when the nerve roots are compromised in the upright posture, fixed EMG changes seldom occur. More often, the EMG findings in spinal stenosis, especially in mild to moderate cases, are normal or, at most, equivocal.

Fibrillation Potentials in the Paraspinal Muscles do not Necessarily Imply Radiculopathy

Too often, the presence of fibrillation potentials in the paraspinal muscles is automatically interpreted as evidence of radiculopathy. Although fibrillation potentials in the paraspinal muscles are an important finding in radiculopathy, they are frequently seen in other conditions as well. Most important among those conditions are (1) proximal myopathies with inflammatory or necrotic features (e.g., polymyositis), (2) disorders of the motor neurons (e.g., ALS, intrinsic spinal cord disease), (3) botulism, and (4) neuropathies affecting the dorsal rami (e.g., diabetic polyneuropathy). As in any electrophysiologic study, the final impression should never be based on a single finding but rather on a combination of all nerve conduction and EMG data in conjunction with the clinical information.

In addition, sparse fibrillation potentials and especially positive sharp waves are occasionally seen in the paraspinal muscles in normal individuals. Although this is rare in patients younger than 40 years of age, in one study of normal individuals, approximately 40% of older individuals had brief runs of fibrillation potentials or positive sharp waves in the lower lumbosacral paraspinal muscles. In this study, the denervating potentials were counted if they lasted longer than 0.5 seconds. Remember, traditionally a muscle is graded as +1 fibrillation potentials if you see fibrillation potentials and/or positive waves that last more than 3 seconds, at two or more locations. The take home message is the following: increased insertional activity or a few brief runs of fibrillation potentials or positive sharp waves limited to the paraspinal muscles, especially in an older individual, may not have any clinical significance, and needs to be interpreted with caution. Again, remember one of the cardinal rules of EMG: when in doubt, do not overcall a diagnosis. You do not want to make a type I error (i.e., make a diagnosis of an abnormality when one is not really present).

In the Elderly, it may not be Possible to Differentiate a Mild Chronic Distal Polyneuropathy from Mild Chronic Bilateral L5–S1 Radiculopathies

Normally, conduction velocities decrease slightly and sensory amplitudes fall with advancing age. In addition, both polyneuropathy and degenerative disc disease of the lumbosacral spine are common conditions in the elderly. Consider the following pattern in an elderly patient:

If this pattern is present bilaterally in an elderly patient, a definite conclusion cannot be reached. There may be bilateral L5–S1 radiculopathies with successful reinnervation in the paraspinal and proximal muscles, and the borderline SNAPs may be related to advanced age. However, this pattern is also consistent with a distal polyneuropathy that is mild enough that it reduces only the lower extremity SNAPs into the borderline range and spares the upper extremities.