In a motor NCS, the peripheral nerve is stimulated with a transcutaneous depolarizing square wave electrical pulse (stimulation electrode) of very short duration. The stimulus generates a muscle contraction in the target muscle, and its intensity must be supramaximal in order to excite all A-alpha fibers. Furthermore, a ground electrode is also necessary, providing a zero-voltage reference point. With two self-adhesive surface recording electrodes that are placed on the muscle belly (pick-up electrode) and on the tendon of the target muscle (reference electrode), the compound muscle action potential (CMAP) arising from the activated muscle fibers is recorded. The stimulation can be carried out proximally to the recording electrodes at different sites of the nerve course (e.g., wrist, forearm, elbow, upper-arm, and Erb’s point) [9].

In the case of deeply situated peripheral nerves, such as the lumbar plexus, stimulation is also possible with needle electrodes or, alternatively, by means of transcutaneous magnetic stimulation (TMS). This also applies to deeper lying muscles, where the CMAP can also be recorded with needle electrodes. However, these special techniques are discussed elsewhere [2, 3].

Neurophysiological convention specifies that negative voltage results in an upward deflection of the CMAP. Taking this fact into account, the routine technique described above allows a couple of parameters derived from the CMAP, that are used in the diagnosis of peripheral nerve diseases to be measured and calculated (Fig. 5.1a) [9–12]:

Note that the larger the distance between the stimulation and recording electrodes, the longer the duration of the CMAP owing to an increased range of conduction velocities. This phenomenon is additionally characterized by an increasing decay of the amplitude of the CMAP, also called temporal dispersion (Fig. 5.1b) [13, 14].

Age-matched normal values for NCS, including motor nerve conduction velocity, distal motor latency, and the CMAP amplitude of every single peripheral nerve are either derived from studies in the laboratories or obtained from related literature [2, 3, 9, 15].

There are many physiological factors affecting nerve conduction. Only a few are addressed here. Generally, nerves of the upper extremity conduct faster than those of the lower extremity. Moreover, the nerve conduction velocity in proximal nerve segments is faster than in the distal parts of a peripheral nerve [16]. The most important factors influencing NCS in the clinical routine are age and skin temperature. In infants and patients of advanced age, nerve conduction velocities are diminished. Furthermore, a decrease in skin temperature also decreases the conduction velocities [17–19]. Therefore, standardization and documentation of the skin temperature during electrodiagnostic testing is mandatory. In the case of low skin/room temperatures, a hot water bath or ceiling-mounted radiant heaters can be effective in raising the skin temperature of the limbs.

Demyelinating disorders of the peripheral nerve result in impairment of conduction owing to the focal or generalized pathological features of the myelin sheath [11]. As a consequence, EDX demonstrates a slowing of the motor nerve conduction velocities and/or a prolongation of distal motor latencies. Depending on the type of nerve involvement, different patterns of motor nerve conduction velocity deterioration can be observed [2, 3, 9–12, 15]:

If only the distal nerve segment is affected, prolongation of the distal motor latency will result, whereas nerve conduction velocity and CMAP amplitude are normal. This applies in particular to the early stages of carpal tunnel syndrome, tarsal tunnel syndrome, and Guyon’s canal syndrome. In contrast to focal entrapments, a generalized demyelinating neuropathy leads to prolongation of the distal motor latencies in addition to a slowing of the nerve conduction velocities (Fig. 5.2). Distal motor latency is used for evaluation because the nerve conduction velocity of distal nerve segments cannot be calculated. DML can therefore be explained by the fact that the exact distance between the stimulation electrode and the endplate region of the supplied muscle is unknown. Moreover, distal motor latency contains the indeterminable time needed for neuromuscular transmission and the generation of muscle contraction (CMAP) [2, 3, 9–12, 15].

Apart from prolonged distal motor latencies and lowered conduction velocities, there are two other conditions that typically occur in demyelinating nerve disorders, referred to as “abnormal temporal dispersion” and “conduction block” [13, 14]. As mentioned above, in a healthy peripheral nerve the duration of the CMAP after proximal stimulation normally increases and the CMAP amplitude decreases compared with the CMAP recorded after distal stimulation. This feature is termed “temporal dispersion.” In the case of a demyelinating disorder, temporal dispersion is further increased because of the lowered conduction velocity and increasing heterogeneity of the motor nerve conduction velocities of different A-alpha fibers generating the CMAP. Consequently, besides the prolonged duration of CMAP and the decay of its amplitude, it appears to be split at and above the site of the lesion, whereas the distal CMAP remains normal (Fig. 5.3d) [9–12]. In scientific studies, different consensus criteria regarding the “abnormal temporal dispersion” have been published. According to the American Academy of Neurology, temporal dispersion is abnormal “if the CMAP obtained after proximal stimulation shows more than a 20 % decay of the amplitude and more than a 15 % increase of the duration compared with the CMAP obtained after distal stimulation” [13]. In contrast, the criteria of the American Association of Neuromuscular and Electrodiagnostic Medicine assume an abnormal temporal dispersion “if the duration of the proximal CMAP is prolonged by more than 30 % compared with the duration of the distal CMAP.” It is assumed that segmental demyelination over longer distances is responsible for this phenomenon [14].

On the other hand, a short segmental demyelination is attributed to a “conduction block.” If the conduction block is complete, a normal CMAP can be recorded after distal stimulation, whereas stimulation proximal to the lesion will generate no CMAP (Fig. 5.3c). An incomplete conduction block is assumed if the CMAP amplitude (or area under the curve) after proximal stimulation is reduced only when with the CMAP after distal stimulation [9]. Depending on the affected nerve, there are again different criteria defining an incomplete conduction block. According to the American Association of Neuromuscular and Electrodiagnostic Medicine, a “drop of more than 50 % in proximal CMAP compared with CMAP obtained after distal stimulation in the median and ulnar nerves and more than 60 % in the tibial and peroneal nerves is considered as pathological, while CMAP duration increases by no more than 30 %” [14]. In contrast, the criteria of the American Academy of Neurology require a “20 % drop of the (proximal) CMAP area or amplitude compared after distal and proximal stimulation” [13].

Unfortunately, the diagnosis process for demyelinating disorders of the peripheral nerves can become very challenging. Primary axonal loss is sometimes accompanied by secondary demyelination (Fig. 5.3f). This explains why slowed conduction velocities and prolonged distal motor latencies can occur in an axonal disorder [12]. Moreover, during the first few days after severe axonal nerve damage this condition will present a complete or incomplete conduction block because Wallerian degeneration has nor been completed yet and the segment distal to the damage is still conducting. This phenomenon has frequently been referred to as a “pseudo-conduction-block” [20]. Therefore, different criteria have been suggested, considering motor conduction velocities, distal motor latencies, and F-wave latencies. For instance, Van den Bergh and Pieret assume inter alia a motor conduction velocity slowing of <70 % of the lower limit of normal and a prolonged distal motor latency of >150 % of the upper limit of normal to be characteristic of a demyelinating condition [21].

Axonal loss leads to a reduction of the activated muscle fibers and is thus paramount to a reduction of CMAP amplitude. In a fractionated motor nerve conduction study this applies to all generated CMAPS – distally and proximally. As a consequence, and in contrast to focal demyelinating disorders, a localization of the lesion with motor NCS is impossible (Fig. 5.3e) [10–12]. This important lack of diagnostic power can be solved by needle electromyography and nerve imaging (see Sect. 5.6 and Chap.​ 6) [6, 22]. Furthermore, at an early stage (days 0–14) after onset of an axonal injury, this presentation can be misdiagnosed as a conduction block (see Sect. 5.3.2) [20]. If axonal loss also includes the fastest conducting A-alpha fibers, motor nerve conduction velocities may also be reduced. However, the extent of conduction slowing is less than following nerve fiber demyelination, and it usually ranges above 70 % of the lower limits of normal [2, 3, 21].

As in a motor NCS, the peripheral nerve is stimulated with a transcutaneous depolarizing square wave electrical pulse (stimulation electrode) of a very short duration. However, a lower stimulation intensity is required than in a motor NCS. In contrast to motor NCS, a self-adhesive recording electrode can be placed either distally (antidromic recording) or proximally (orthodromic recording) to the stimulus site because the evoked electrical activity propagates in both directions. Furthermore, a ground electrode is also necessary [9]. In clinical practice, antidromic recordings are preferred, mainly because the amplitude of the sensory nerve action potential (SNAP) is larger than that obtained by orthodromic recordings [10–12]. Sensory nerve conduction studies can also be performed with needle electrodes; however, these special techniques are discussed elsewhere [2, 3]. Compared with CMAPs, SNAPs are smaller and measured in microvolts. Therefore, an averaging of several SNAPs is required to improve the signal-to-noise ratio (SNR). Most important parameters derived from sensory NCS are as follows (Fig. 5.4) [10–12]:

As mentioned above, only the largest and fastest conducting sensory A-alpha fibers, which functionally supply senses of touch, position, and vibration senses, can be assessed using sensory NCS. Usually, small fiber neuropathies affecting C-fibers mediating pain do not show any abnormality in conventional sensory NCS [4, 8]. Quantitative sensory testing and autonomic testing or skin biopsy are necessary to provide the correct diagnosis in this situation. These special techniques are described elsewhere [2, 3].

As in motor NCS, the primary feature of demyelination is a reduction of sensory nerve conduction velocity [10–12]. Additionally, the duration of SNAP increases owing to an enlarged range of nerve conduction velocity. It leads to “temporal dispersion” and “phase cancellation” (Fig. 5.5). Consequently, the SNAP amplitude will also drop in a demyelinating disorder [10–12, 23].

Axonal disorders will primarily reduce the amplitude of the SNAP generated after electrical stimulation, whereas its duration is normal. As in motor NCS, further axonal loss of the fastest conducting fibers will also lead to a modest slowing of the sensory nerve conduction velocities (Fig. 5.5) [10–12, 21, 23].

Sensory NCS are useful to determine whether the site of the nerve lesion is pre- or postganglionic. In the case of a preganglionic lesion the connection between the cell bodies of the dorsal root ganglion and the peripheral nerve remains intact. Consequently, a normal SNAP and conduction velocity can be obtained by sensory NCS, despite anesthesia in the dermatome supplied by that root. However, an absent or pathologically changed SNAP indicates a postganglionic lesion either of the nerve plexus or of a single peripheral nerve [24]. Apart from this electrophysiological dogma, in some recent studies it has been shown that – in a small percentage of subjects (2.4–12.1 %) with preganglionic lesions due to a L5 or S1 radiculopathy – the SNAP amplitude was abnormal. This exception may be due to dorsal root ganglion compression proximal to the spinal foramen or intra-spinally, and should be always kept in mind [25, 26].

F-waves are a type of late motor response. Their name is derived from the foot, where they were first described. With a reversed stimulation electrode, F-waves can be recorded in the same setting as stated in Sect. 5.3.1 concerning motor nerve conduction studies [9]. If a nerve is electrically stimulated above its final segment the action potential propagates in both directions.

The orthodromic propagated action potential generates CMAP, also known as an M-response. On the other hand, the antidromic propagated action potential reaches the anterior horn cell body and depolarizes the axon hillock. As a result, a small portion of the alpha-motor neurons backfire and a second orthodromic action potential reaches the muscle. This causes a late muscle depolarization that involves only a small portion of the muscle fibers, referred to as an F-wave [27–31]. Recording F-waves will therefore allow the electrodiagnostic testing of long and proximal nerve segments (e.g., the lumbosacral plexus or nerve roots) which are inaccessible to conventional electrical nerve stimulation [29]. In contrast to the M-response, F-waves usually vary in latency and shape because different populations of motor neurons backfire with each stimulus. Therefore, normally, “10–20 consecutive stimulations will be needed and the most reliable measure of the F-wave is its minimum latency of 10–20 firings” [29, 31]. It strongly depends on the body height, which should be taken into account. Moreover, the “F-wave persistence” is commonly used and represents the number of F-waves obtained per number of stimulations [27–31]. Regarding other parameters (dispersion, amplitude–ratios) derived from F-wave studies please refer to specialist literature [29].