Resting Membrane Potential

All living cells generate an electrical potential across their membranes, with the intracellular region relatively negative compared with the extracellular region.²⁸ This potential difference across the cell membrane is referred to as the resting membrane potential. The development and maintenance of the resting membrane potential can be explained by a simple model.

We can partition a beaker into right and left halves with an impermeable membrane containing multiple closed potassium channels. Assume this membrane separates two different concentrations of a potassium chloride (KCl) solution with a 10-mM concentration on one side and 100-mmol/L concentration on the other (Figure 9-1).¹⁷ Potassium chloride in solution exists as positive potassium (K⁺) ions (cations) and negative chloride (Cl^–) ions (anions). A voltmeter (a device that detects potential differences) measuring the two solutions fails to detect a potential difference because there is a lack of physical continuity between the left and right halves of the beaker when the partition’s potassium channels are closed. In other words, the two solutions are electrically independent.

FIGURE 9-1 A beaker containing two different concentrations (10 mmol/L and 100 mM) of a potassium chloride (KCl) solution existing as potassium (K⁺) and chloride (Cl^–) ions. An impermeable partition containing closed potassium channels separates the two different concentration solutions. A voltmeter placed across the partition fails to register any voltage difference. If the partition is now made selectively permeable to just the potassium ions (potassium channels are opened), the concentration gradient difference will drive potassium ions into the lower-concentration side of the beaker until the electrical attraction from the accumulating negatively charged chloride ions prevents any further net K⁺ ion movement, thus establishing a dynamic equilibrium. At this point, a potential difference exists across the partition and represents the equilibrium potential.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

If we now open the membrane’s potassium channels, K⁺ cations will flow “down” their concentration gradient from the high (100 mm/L) to low (10 mm/L) ion concentration side of the beaker (see Figure 9-1). The potassium ions will continue this directional flow until there is a balance between (1) the forces of the physical concentration gradient difference driving potassium to the lower-concentration region, and (2) the electrical gradient opposing this directional ion flow. Because there are only potassium channels, the negative chloride ions cannot pass through the membrane and remain on their respective sides of the beaker. As more and more positive potassium ions leave one side of the beaker, there begins to develop an unbalanced or “excess” amount of negative charges (Cl^–) on the high-concentration side of the beaker, with an equal buildup of “excess” positive charges (K⁺) on the other side of the beaker. The increasing net negative charge of the beaker half with the increasing number of unbalanced chloride ions begins to make it increasingly difficult for the positive potassium charges to leave the high-concentration side of the beaker. This is because the progressively building net negative charge increasingly attracts those remaining positive potassium ions. Similarly, a growing amount of positive potassium ions on the formerly low potassium concentration side of the beaker begin to increasingly repel additional potassium ions attempting to enter this side of the beaker.

At some point, the opposing electrical charges on the two sides of the beaker prevent any more potassium from leaving the beaker’s high-concentration side, even though there is still a higher potassium concentration on this side compared with the other. A balance is established between the ionic concentration forces driving potassium ions from the high- to low-concentration regions, and the electrical forces (positive repelling and negative attracting) tending to keep potassium ions in the more concentrated portion of the beaker. Any potassium ions that randomly enter the lower-concentration side of the beaker are balanced by other potassium ions similarly crossing in the opposite direction. The situation where balance between electrical and concentration forces exists is said to be a dynamic equilibrium. Placing a voltmeter across the partition measures a negative potential difference, as electrical continuity is now present between the halves of the beaker through the open potassium ion channels.

The above simple example can be applied to all cells in the body, and in particular to nerve and muscle cells. A nerve’s axon will be used as a prototypical example, although the same principles apply to muscle cells. The nerve cell is known to have a specialized cell membrane permeable primarily to potassium and chloride ions in the resting state, because of proteinaceous transmembranous ion channels rendering the membrane selectively permeable (i.e., semipermeable).³¹ The potassium channels are referred to as passive, because they are always open and permit the free flow of this ion. Contained within the axon but incapable of crossing the cell membrane are large, negatively charged protein molecules. This negative charge attracts potassium ions from outside the cell to enter through the passive potassium channels, resulting in a buildup of potassium ions within the axon. This process continues until there is so much potassium within the axon that the continued entry of more potassium ions is prevented by the high intracellular potassium concentration, even though all the negative charges have not yet been electrically balanced. This is because the force of the accumulated potassium ions’ concentration gradient now attempting to drive potassium out of the cell is just large enough to balance the intracellular anions’ electrical force attracting potassium ions into the cell.

Similar to the above beaker example, a dynamic equilibrium eventually develops between (1) intracellular negative charges attracting potassium ions, and (2) a high intracellular potassium concentration impeding further potassium ion entry. This dynamic equilibrium occurs at a point when the intracellular negative potential is reduced by the inflowing potassium ions to a value approximating –80 to –90 mV compared with the extracellular environment.

Nernst Equation

The above examples of beakers and cells can be expressed mathematically by simply stating that the net “work” of both the electrical gradient work (W_elec) and the concentration gradient work (W_con) is zero in the resting state (W_con + W_con = 0). This means that the work, or energy, of developing the concentration gradient (W_con) is balanced by, or equal but opposite to, that developed by the electrical gradient (W_elec) during the resting state: W_elec = –W_con. The negative sign is present to denote the “opposite” or balanced aspect of the work. The electrical work is expressed as W_elec = Z_iFE_m. The symbols designate specific aspects defining electrical ion work: Z_i is the ion’s charge, F is Faraday’s constant, and E_m is the transmembrane potential. The work required to move ions across the membrane can be expressed as the natural logarithm of the ionic concentration differences between the intracellular ([I]_i) and extracellular ([I]_e) ions. Universal gas (R) and temperature (T) constants are conversion factors necessary to balance units between all the variables. Substituting the above-noted variables into the balanced work equation results in the following equations:

This formula can be rearranged to find the potential at which the electrical work just balances the concentration work (i.e., the dynamic equilibrium or resting membrane potential) by solving the above equation for E_m

The above equation is more commonly known as the Nernst equation, and is a mathematical statement of the potential at which all the electrical and concentration forces are balanced in the cell’s resting state (i.e., the resting membrane potential).^28,31 By substituting the actual values for the different variables and using the more familiar base 10 logarithm, the equation converts to the more recognizable form. Also, the approximate concentration ratio of intracellular to extracellular potassium is 20:1. The Nernst equation then becomes:

Sodium Pump

Experimentation has shown that the Nernst equation predicts the resting membrane potential quite well with different extracellular potassium concentrations, as long as the potassium concentration is relatively high. When the extracellular potassium concentration is low, there is a deviation of the resting membrane potential from that predicted, with a less-negative potential achieved. This finding suggests that another ion (sodium: Na⁺) has some influence on the resting membrane potential. It turns out that sodium has a very high extracellular concentration compared with intracellular ion concentration, which results in small quantities of sodium ions leaking into the membrane through a few passive sodium channels present in the cell’s membrane. This relative impermeability of positive sodium ions, combined with a high extracellular but low intracellular sodium ion concentration, and the resulting electrical drive to enter the cell (negative inside), would tend to “run down” the cell resting membrane potential over time. Fortunately, the cell has developed a mechanism whereby this run down is prevented. Located within the cell membrane is an energy-dependent sodium-potassium pump that pumps in potassium ions and pumps out sodium ions in just the right ratio to match the sodium ions entering and the compensatory potassium ions exiting the cell. Two sodium ions are extruded for every three potassium ions taken into the cell. The sodium-potassium pump maintains the exact ionic balance necessary to sustain a stable resting membrane potential by maintaining this 2:3 ratio.

Goldman-Hodgkin-Katz Equation

An important modification of the Nernst equation is the inclusion of ion permeability as being the primary influence on the cell’s transmembrane voltage. The greater an ion’s permeability, the more likely it is to influence the transmembrane potential. This is because the equilibrium potential of the most permeable ion in effect becomes the cell’s resting membrane potential. The equation accounting for the different permeability (designated as “p” in the equation below) factors is known as the Goldman-Hodgkin-Katz equation^28,30,31:

In this equation, it can be seen that the cell’s transmembrane voltage (E_m) is primarily dependent on which ion has the greatest permeability. For example, in the resting state, the permeability of potassium is known to be relatively high because of non–voltage-gated potassium passive leak channels, while the permeability of sodium ions is very low. Chloride ions are distributed through chloride channels that are always open at the resting membrane voltage, and thus adjust to whichever ion’s permeability predominates between potassium and sodium as dictated by the net transmembrane potential. With a high potassium and low sodium permeability, the above equation simplifies to the Nernst equation; that is, potassium is the predominant ionic species, and E_m becomes –75 mV. If sodium permeability were to increase dramatically, then E_m would approach the sodium ion equilibrium potential of +55 mV. It would not quite reach this value because potassium and chloride ions continue to have some influence and would hold the maximum potential to a less positive (more negative) value of about +40 mV.

Action Potential Generation

In addition to the passive (always open) potassium ion channels and relatively few passive sodium ion channels, there are also sodium and potassium ion channels within nerve and muscle cell membranes modulated by transmembrane voltage differences. These channels are voltage-gated because they open and close depending on the transmembrane voltage.²⁵ Muscle and unmyelinated nerve contain both sodium and potassium voltage-gated channels. These voltage-gated sodium and potassium ion channels are closed at the resting membrane potential. If the transmembrane voltage changes toward a more depolarized direction (less negative) and reaches about 15 to 20 mV less negative than the resting membrane potential, the voltage-gated sodium channels open. This results in an increased permeability of the sodium ion, a process known as sodium activation. As noted above, the Goldman-Hodgkin-Katz equation predicts that the transmembrane potential shifts toward the sodium ion equilibrium potential. This massive shift in transmembrane potential is referred to as depolarization. After staying open for a short period, the sodium gates automatically close (sodium inactivation), with a return of the resting membrane potential again dictated by the resting state ion permeabilities (repolarization). In muscle and unmyelinated nerve membranes, a delayed opening of potassium voltage-gated channels occurs secondary to the depolarization and assists in repolarization.

Very few ions actually cross the membrane for sodium-induced depolarization or potassium-mediated repolarization. The region of membrane where there are large numbers of voltage-gated sodium channels in the open conformation acts as a so-called current sink for sodium ions to “sink,” or enter, the cell’s interior. This implies that a nearby source for the sodium ions must be present. The surrounding extracellular membrane region acts as the current source, thus permitting an ionic flow from the region surrounding the current sink. Sodium ions are thereby removed from the outside of the membrane surrounding the sink (making this region relatively more negative) and deposited on the inside of the cell (Figure 9-2). The newly introduced intracellular ions migrate to help neutralize some of the cell’s interior negative charges. This current flow, or charge transfer, from the extracellular to intracellular space is referred to as a local circuit current.¹⁷ The net effect of charge transfer is to make the cell’s interior less negative and exterior more negative, thereby shifting the transmembrane voltage in the more depolarized direction about the current sink. If the charge transfer is sufficient to depolarize the cell by approximately 15 to 20 mV, the membrane surrounding the sink is induced to permit sodium activation and hence undergo a significant depolarization. This process can then continue along the length of the cell.

FIGURE 9-2 A, Sodium (g_Na) and potassium (g_K) ion conductances are depicted over time resulting in an alteration of the transmembrane potential, creating an action potential. B, The spatial relationship of the sodium and potassium ion flux during an action potential is schematically depicted. Note the alteration of the transmembrane ionic potential differences corresponding to the depolarization and repolarization. C, Local circuit currents described the pathways of extracellular sodium ions entering the cell and then migrating longitudinally within the cell. D, Triphasic extracellular waveform associated with the intracellular monophasic action potential.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

The above process creates an action potential spike at the original site of sodium activation. A mechanical, chemical, or electrical stimulus that causes the membrane potential to reach threshold over a localized region is all that is required to serve as the initiating stimulus of action potential generation. Once the process begins, it is self-sustaining as long as there are sufficient ion channels to repeat the depolarization process. The intracellular action potential is essentially a monophasic positive spike: –75 mV resting potential; +40 mV spike; return to resting –75 mV with sodium inactivation and potassium activation (see Figure 9-2). The membrane’s threshold value must be reached to generate the self-sustaining action potential that is the same at all regions of the membrane. This concept is referred to as the all-or-none phenomenon.

The nodes of Ranvier in myelinated nerves lack voltage-gated potassium channels and contain only voltage-gated sodium channels.^47,48 The action potential actually “jumps” from one node to the next, creating an efficient means of action potential propagation that is referred to as saltatory conduction. Repolarization in myelinated nerve does not require a delayed potassium current. As noted above, the resting membrane potential is restored once the permeability of sodium is reduced. Passive “back-leak” sodium and potassium currents are thought to mediate the discharge of the membrane’s capacitance, which accumulates over time with multiple action potential discharges.

Gender

Only a few studies have attempted to investigate the difference in nerve conduction studies between men and women.³ Noted is a slight increase in the antidromic sensory nerve amplitudes for both the median and ulnar nerves recorded from the digits in women. Also, women demonstrate a greater nerve conduction velocity (NCV) for upper and lower limb nerves compared with men. Both of these differences, however, are eliminated when one considers limb length and digit circumference (see below).⁴⁰

Aging

Several generalizations can be made regarding peripherally evoked sensory nerve action potentials (SNAPs) and aging. The conduction velocity demonstrates a consistent decline approximating 1 to 2 meters per second per decade.³⁶ The duration of SNAPs is about 10% to 15% longer in persons aged 40 to 60 years, and 20% longer in those aged 70 to 88 years, compared with persons aged 18 to 25 years.⁸ Compared with the 18- to 25-year-old group, the SNAP amplitude is one half and one third, respectively, for the 40- to 60-year-old and 70- to 88-year-old groups. The distal sensory latencies reveal a similar prolongation with age. There is a suggestion that the median and radial nerves do not demonstrate considerable alteration with age.¹⁹ At present, there is disagreement as to the magnitude of change in SNAP parameters induced by the aging process.

The results of aging on conduction velocity have been examined in a number of upper and lower limb nerves. Motor NCVs reveal similar changes to those described for sensory nerves. The newborn’s motor NCVs are about half of adult values, which are reached by 3 to 5 years of age.² After the age of 50 years, the conduction velocity of the fastest motor fibers progressively declines by approximately 1 to 2 m/s per decade. There is a concurrent increase in the distal motor latency and decrease in the motor response’s amplitude with advancing age. H-reflex latency demonstrates little alteration with aging in the healthy elderly.²⁰ The decrease in amplitude is difficult to ascertain clinically because of the wide range of normal H-reflex amplitudes and the central effects on amplitude that are difficult to control and quantify.

Digit Circumference

Females consistently demonstrate significantly higher antidromic SNAP amplitudes for the median and ulnar nerves recorded from the second and fifth digits.³ A negative linear correlation exists between finger circumference and amplitude for these two nerves. It is known that as the distance between the recording electrode and neural generator increases, the amplitude precipitously declines. Increasing the circumference of the finger displaces the electrode further from the nerve. Because men have significantly larger finger circumferences than women, this appears to explain the difference in SNAP amplitudes. There is no evidence that this difference is due to an intrinsic neural difference between male and female nerves.

Height

Several investigations have documented slower lower limb NCVs in taller compared with shorter individuals.^9,33 This difference is independent of the limb’s temperature or subject’s age. The etiology of this finding is unknown. Distal nerve tapering or an abrupt change in axon diameter has been speculated to account for this finding, but the exact etiology has not been definitively identified.^14,46

Temperature

Temperature is one of the most important factors influencing nerve conduction studies. As the temperature of the nerve is lowered, the amount of current required to generate an action potential increases. This decreased excitability is a direct temperature effect on the nerve’s action potential–generating mechanism at the nodes of Ranvier, and not a result of membrane resistance changes. The transmembrane resistance is not increased by a drop in temperature.^26,27 In addition to excitability, the configuration of an action potential is profoundly affected by a drop in temperature.

The action potential’s amplitude, rise time, and fall time all increase as the nerve’s temperature declines. The time required for the action potential of a cold nerve to reach its peak depolarization from the resting membrane level increases approximately 33%.³⁸ Similarly, the time necessary for the action potential to return to its resting level is also increased, but much more so than the rise time (69%). Because both the SNAP duration and the spike height increase, the area of the SNAP increases dramatically at lower temperatures.

The compound muscle action potential (CMAP) arising from cooled muscle tissue demonstrates similar changes as those noted for SNAPs. The CMAP amplitude, duration, rise time, and area all increase as the muscle’s temperature is reduced. Intramuscular recordings also reveal that those motor units in close proximity to the recording electrode are also increased in the same parameters noted above.

Temperature also has an impact on NCVs. Based on the prolongation of the rise and fall time noted above, it should be possible to infer the nerve’s response to cooling with respect to NCV. Because propagation is saltatory in myelinated nerves, decreased temperature results in an increase in the amount of time necessary to reach the action potential’s peak at each node of Ranvier, thereby slowing action potential propagation and reducing conduction velocity.

The first detailed investigation of temperature effects on NCV in human nerves revealed an NCV to temperature correlation of 2.4 m/s per °C for median and ulnar motor conduction.²⁴ With every 1° C drop in temperature, there was a 2.4 m/s decrease in the conduction velocity. Reductions in conduction velocity for upper limb motor nerve fibers have also been found to approximate a decrease of 4% or 5% per °C.^10,29 Correction factors using subcutaneous and intramuscular readings are equally correct, but it is more convenient and less painful to use surface measurements.²²

In the upper limb, the relationship between temperature and NCV has been investigated for the surface temperature range of 26° C to 33° C, measured at the midline of the distal wrist crease. Calculations reveal that for median motor and sensory nerves, NCV is altered 1.5 and 1.4 m/s per °C, respectively, while the distal latency for both nerves changes about 0.2 ms/°C.^21–23 The ulnar nerve demonstrated motor and sensory temperature relationships of 2.1 and 1.6 m/s per °C, respectively, and a distal motor and sensory latency correlation of 0.2 ms/°C.

Because of the profound effects of temperature on NCV, it is clear that reliable nerve studies require temperature control. A cool limb, irrespective of the ambient room temperature, can result in latencies, NCVs, and amplitudes that are not in the “normal” range. A normal limb study can yield results that are spuriously thought to be abnormal, but that are due only to the low temperature. This is an especially important issue when an abnormal nerve is being studied. Correction factors can be used for normal nerves, but serious questions remain about how abnormal nerves respond to temperature variations. Although applying a correction factor is less time-consuming than heating the limb, it is doubtful how accurately correction factors can be applied to abnormal nerves. Until more data are available regarding the best correction factor for diseased nerves, warming of the limb should be considered superior to using correction factors. The surface temperature approximating the recording electrodes should be at least 32° C in the upper limbs and 30 °C in lower limbs. Caution must be exercised when attempting to warm the limbs of patients with ischemic limbs, or those with altered sensation, to avoid injury.

Nerve Potentials

Sensory Nerve Action Potentials

Clinical Recordings

SNAPs can be obtained with either antidromic or orthodromic techniques.⁸ The term antidromic implies that the induced neural impulse propagates along the nerve in a direction opposite to its physiologic direction; for example, stimulating the median nerve at the wrist and recording the ensuing response from the second or third digit. A nerve will conduct an impulse proximally and distally from the point of stimulation by a depolarizing current. On the other hand, stimulating the median sensory fibers on the second digit and recording from the wrist is an example of an orthodromic technique. In orthodromic recordings, the sensory fiber impulses are detected at a site proximal to the stimulus as they travel physiologically from a distal to a more proximal region. Antidromically obtained SNAPs are usually larger and hence easier to assess, because the recording electrodes are in close proximity to the proper digital nerves than when a recording electrode is placed over the median nerve at the wrist for orthodromic techniques. Mixed nerve responses can have a component of both orthodromic and antidromic waveforms. For example, midpalm stimulations recorded at the wrist are a combination of orthodromic sensory and antidromic motor responses.

SNAP Configuration

Antidromic and orthodromic bipolar SNAP waveform recordings are typically biphasic rather than triphasic. The biphasic, negative-positive potential is a result of the bipolar recording technique and not a violation of volume conductor theory. Biphasic SNAP waveforms can best be understood by use of bipolar and referential recording montages for median nerve stimulation at the wrist (Figure 9-4).¹³ The median SNAP is indeed a triphasic waveform and conforms to the principles of volume conduction, but only appears biphasic because of the recording montage used.

FIGURE 9-4 The median nerve is stimulated at the wrist with an antidromic sensory recording montage. Active (a) and reference (r) are located on fingers as designated above. A, Bipolar recording with the commonly observed biphasic median nerve sensory nerve action potential (SNAP). B, The same active electrode location in (A) is referenced to the fifth digit, resulting in a triphasic SNAP. C, An active electrode placed on the fifth digit but referenced to the third digit permits one to observe what this electrode records when the median nerve is stimulated. An inverted triphasic potential of small magnitude is detected. It is inverted because of its connection to the inverting amplifier port. D, Electronically summating the potential recorded in (B) and (C) yields that recorded with a bipolar montage in (A).

(Redrawn from Dumitru D. Volume conduction: theory and application. In: Dumitru D, editor: Physical medicine and rehabilitation state of the art reviews: clinical electrophysiology, Philadelphia, 1989, Hanley & Belfus, with permission.)

It is also possible to predict the optimal interelectrode separation to maximize the biphasic potential’s amplitude in the bipolar recording.¹⁷ The critical factor in this instance is the rise time, baseline to negative peak, of the biphasic potential, because it is this parameter that defines the waveform’s maximum amplitude. The recording electrodes must be located at a distance greater than the action potentials’ spatial extent, represented by the rise time duration. The rise time of most SNAPs approaches 0.8 ms, which represents a longitudinal extent of 40 mm for an action potential conducting at 50 m/s (50,000 mm/1000 ms = D/0.8 ms; D = 40 mm). If the two recording electrodes are separated by a distance less than 40 mm, some similar information regarding the main peaks of the waveforms recorded by both recording electrodes (active and reference) results in mutual cancellation of data, producing a potential with a smaller amplitude. A portion of the nerve will be depolarizing under the reference electrode while still in some degree of depolarization under the active electrode. At interelectrode separations greater then 40 mm, the biphasic potential’s negative peak amplitude will no longer grow, but the terminal positive phase will enlarge slightly and might change its configuration. These findings can be demonstrated by varying the distance between recording electrodes and observing the ensuing results (Figure 9-5). In effect, as the recording distance decreases below 40 mm, the amplitude of the recorded waveform declines and the peak latency shortens.

FIGURE 9-5 The effect of interelectrode separation can be easily demonstrated by evoking an antidromic median sensory nerve action potential (SNAP) and progressively increasing the interelectrode separation between the active and reference electrode. The active electrode remains in the same location, while the reference electrode is sequentially displaced more distal on the digit. As can be seen, the SNAP amplitude increases, peak latency increases, and the onset latency remains the same.

(Redrawn from Dumitru D, Walsh NE: Practical instrumentation and common sources of error, Am J Phys Med Rehabil 67:55-65, 1988, with permission.)

Muscle Potentials

Needle Insertional Activity

Normal Insertional Activity

Placing a needle (monopolar or standard concentric) recording electrode into healthy muscle tissue and advancing it in quick but short intervals results in brief bursts of electrical potentials referred to as insertional activity (Figure 9-6, A).¹⁴ The observed electrical activity is believed to result from the needle electrode mechanically depolarizing the muscle fibers surrounding its leading edge as it pierces and deforms the tissue. Minimal and localized muscle tissue damage may occur from direct needle trauma and is the basis for the synonymous term of injury potentials. The purpose of including insertional activity analysis as part of the electromyographic examination is that the probing needle might provoke transient and/or sustained abnormal potentials associated with membrane instability before this abnormal activity with the muscle at rest.

Figure-9-6 A, Inserting a monopolar needle into healthy muscle tissue results in mechanical depolarization of muscle tissue, which generates a brief burst of electrical activity designated as insertional activity. B, Inserting the same needle into fibrotic muscle or subcutaneous fatty tissue results in decreased insertional activity. C, Inserting a monopolar needle into denervated muscle tissue produced not only the initial burst of electrical activity but also associated positive sharp waves and fibrillation potentials that abate over several hundred milliseconds.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

Decreased Insertional Activity

Muscle that has been replaced by fibrous tissue, or that is otherwise electrically inexcitable, can no longer generate electrical activity. Consequently, the needle electrode is incapable of mechanically depolarizing this tissue. The result is that few, if any, electrical waveforms will be detected after needle movement (Figure 9-6, B).

Increased Insertional Activity

Practitioners have noted that insertional activity can appear to persist after needle movement cessation. This finding has led to the term increased insertional activity. In disease states where the muscle is no longer connected to its nerve or the muscle membrane is inherently unstable from primary muscle pathology or abnormal ion channels, the increased insertional activity completes a temporal continuum from the previously normal insertional activity to the development of sustained membrane instability (Figure 9-6, C). Practitioners differ as to the significant differences if any between increased insertional activity and positive sharp waves.

End-Plate Potentials

Miniature End-Plate Potentials

An active needle electrode located in the end-plate region can record two distinct waveforms. One of the waveforms is a short-duration (0.5 to 2 ms), small (10 to 50 μV), irregularly occurring (once every 5 seconds per axon terminal), monophasic negative waveform.¹⁴ These potentials represent the random release of acetylcholine vesicles. Volume conduction theory would suggest that for a potential to be monophasic and negative, the current sink would have to start and finish within the active electrode’s recording region (Figure 9-7, A).Clinically, multiple miniature end-plate potentials (MEPPs) are usually observed with an intramuscular recording electrode, and the sound is referred to as end-plate noise resembling a “seashell murmur” (Figure 9-7, B).

FIGURE 9-7 A, Monopolar needle electrode located over a muscle’s end plate records the spontaneous depolarization of a miniature end-plate potential (MEPP). As the electrode is located over this potential’s subthreshold central current sink, and hence does not propagate, a monophasic negative potential is recorded. B, The large recording electrode is usually positioned over several end plates, thus recording multiple MEPPs and end-plate spikes.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

End-Plate Spikes

A second waveform that can be detected with an active needle electrode placed in the end-plate region is relatively short in duration (3 to 4 ms), of moderate amplitude (100 to 200 μV), irregularly firing, and biphasic with an initial negative deflection.¹⁴ The biphasic potential has an initial negative phase, produced when a current sink originates in the vicinity of the active electrode and then propagates away (Figure 9-8). Triphasic end-plate spikes might also occur if the active electrode induces an action potential in the terminal axon but the electrode’s recording surface is some distance from the end plate. End-plate spikes and MEPPs are frequently observed together, as they arise from the same region (Figure 9-9).

FIGURE 9-8 (A) Irritation of the terminal axon results in a suprathreshold end-plate depolarization, thus generating a single muscle fiber potential. Because the electrode is located over the end-plate zone, an initial negative deflection is recorded. (B) The terminal positive current sources are then recorded, with action potential propagation thereby generating a biphasic, initially negative potential referred to as an end-plate spike.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

FIGURE 9-9 A, Monopolar needle located in a healthy muscle at rest. B, Slight needle movement positions the electrode in an end-plate region with the recording of multiple miniature end-plate potentials of a negative spike configuration. C, Repositioning the needle electrode to a slightly different region primarily records biphasic, initially negative end-plate spikes. D, Advancing the needle electrode slightly permits the simultaneous recording of both potentials noted individually in (B) and (C).

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

Single Muscle Fiber

The single muscle fiber generates three waveforms that are at times difficult to separate: end-plate spikes, positive sharp waves (PSW), and fibrillation potentials. End-plate spikes can have the identical morphology of PSW and fibrillation potentials. The major key to separating them is that PSW and fibrillation potentials have a regular firing rate and can slowly trail off. End-plate spikes usually are irregular. The single muscle fiber’s extracellular waveform configuration, like nerve tissue, depends on the characteristics of the muscle’s intracellular action potential. A muscle’s action potential is approximately 4 to 20 times longer than a nerve’s, because of the prolonged repolarization process.¹⁴ Aside from the longer duration of local circuit currents compared with neural tissue, the concept of a current sink surrounded by two source currents (source-sink-source) remains unchanged. A triphasic waveform with a small terminal phase should then be recorded from an extracellular active electrode placed adjacent to a propagating single muscle fiber action potential at some distance from the end-plate region (Figure 9-10).

FIGURE 9-10 A single muscle fiber action potential propagating past a needle recording electrode results in a triphasic waveform. This is because the voltage distribution creates an initial and terminal positive voltage source surrounding a negative current sink zone.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

Motor Unit Action Potential Configuration

Anatomy

One anterior horn cell gives rise to a peripheral axon that splits into multiple terminal axons, each of which innervates a single muscle fiber. One anterior horn cell, its axon, and all the single muscle fibers supplied by that nerve are referred to as a motor unit. The electrical activity from all these muscle fibers summates to produce a motor unit action potential (MUAP).

Amplitude and Rise Time

The MUAP configuration can be described in terms of its amplitude (maximum peak-to-peak CRT trace displacement), rise time (temporal aspect of a potential’s peak), duration (departure from and return to baseline), and number of phases (baseline crossings plus one) (Figure 9-11). The amplitude of potentials declines exponentially with increases in distance from the current generator. This occurs because the surrounding muscle and its supportive tissues act as a high-frequency filter to impede potentials that change rapidly over a short period. As a result, the peak-to-peak MUAP amplitude is believed to arise from fewer than 12, and possibly just 1 or 2, single muscle fibers located within 0.5 mm of the electrode’s recording surface. Small movements in the needle electrode frequently change the amplitude dramatically as the needle gets closer or farther away from the motor units.

FIGURE 9-11 A motor unit action potential is depicted, with various morphologic aspects measured.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

Duration

The MUAP duration depends on the shortest and longest lengths of terminal axons from the point they separate from the parent nerve to the end plate. The duration of the MUAP also depends on the conduction velocities of the terminal axons and individual muscle fibers with respect to the recording electrode, and the muscle fiber length (see Figure 9-11).^6,7 The duration of an MUAP is the most sensitive clinical parameter that can be routinely quantified to diagnosing disease. Small movements in needle electrode position have a negligible effect on the duration of the MUAP, which is in contrast to the amplitude and rise time.

Phases

As previously stated, the single muscle fiber action potential usually has a triphasic appearance when recorded outside the end-plate zone and away from the tendinous insertion. The voltages from all the single muscle fibers belonging to one motor unit summate to yield an MUAP that is also usually triphasic: positive-negative-positive. This voltage summation does not always produce a smooth result, and small serrations, or turns, can occasionally be seen as part of the major phase of an MUAP (see Figure 9-11). The number of phases is defined as the number of CRT trace baseline crossings plus one. Normal MUAPs are considered to have four or fewer phases. MUAPs with five or more phases are called polyphasic potentials. Recordings of multiple MUAPs from normal muscle tissue can have between 12% (concentric needle) and 35% (monopolar needle) polyphasic potentials, depending on the type of recording electrode used.¹⁴ Slightly different MUAP configurations can be expected, depending on the exact location of the recording electrode with respect to different single muscle fibers within the motor unit territory (Figure 9-12). Satellite potentials are late waves that are linked to the rest of the waveform.

FIGURE 9-12 A, Muscle fibers belonging to a single motor unit are randomly distributed within an approximately 6-mm oval territory. B, Three individual motor unit action potential waveforms are recorded from the same motor unit, and depend on the location of the recording electrode with respect to different muscle fibers within that motor unit’s spatial territory.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

Pathology can also alter the number of phases an MUAP waveform can have. The motor unit can be affected in two general ways after injury or disease: a pathologic condition may affect either the anterior horn cell or peripheral nerve (including the neuromuscular junction), or the muscle fibers comprising the motor unit. If the neural component of a motor unit is compromised severely enough to experience degeneration, all the muscle fibers innervated by the parent nerve will become denervated. These denervated muscle fibers induce nearby terminal axons of intact nerves to send out neural projections to reinnervate the orphaned muscle fibers most likely through neurotrophic factors. Through collateral sprouting, the total number of muscle fibers belonging to a specific motor unit may increase dramatically (Figure 9-13). Neurogenic diseases can lead over a period to larger-amplitude, longer-duration, and highly polyphasic MUAPs.

FIGURE 9-13 Several muscle fibers are denervated in the above example. Collateral terminal nerve sprouts from an intact motor unit (dotted lines) grow to reinnervate those denervated muscle fibers. The end result is a large-amplitude, long-duration potential with more phases.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

If the muscle fibers comprising a motor unit undergo random degeneration, such as occurs in some myopathies, there is a decrease in the total number of fibers belonging to that motor unit (Figure 9-14). A decrease in muscle fiber numbers would result in a reduction of the overall MUAP amplitude, including its initial baseline departure and subsequent return. If the initial baseline departure of the original MUAP and its subsequent return to baseline decline in magnitude, this portion of the waveform can no longer be observed (assuming the amplifier’s gain is not increased). As a result, the total duration of the MUAP would correspondingly appear to decrease. Finally, the dropout of single muscle fiber waveforms would produce less voltage with respect to the spatial summation of single fiber potentials. Fewer muscle fibers could lead to “gaps” in the MUAP waveform, causing an increase in the number of phases. Consequently, a primary myopathic process tends to yield shorter-duration, highly polyphasic, low-amplitude MUAPs.

FIGURE 9-14 Loss of single muscle fibers from a motor unit results in the generation of motor unit action potentials with smaller amplitude, shorter durations, and possibly more phases.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

Compound Muscle Action Potential

To elicit a CMAP from a particular muscle, the active electrode is located on the skin’s surface directly over the muscle’s end-plate region, or “motor point.”¹⁴ The end-plate region typically lies midway between the muscle’s origin and insertion. The reference electrode is usually placed on or distal to the tendinous insertion of the muscle, so as not to record electrical activity from the activated muscle. Stimulating the peripheral nerve innervating the muscle under investigation will result in a relatively large, biphasic waveform with an initial negative deflection (Figure 9-15).

FIGURE 9-15 A, Locating an active recording electrode over the motor point of a muscle places this electrode over the central region of the negative current sink generating all the muscle’s action potentials. An initial negative deflection is recorded because of this location, thereby producing the compound muscle action potential’s characteristic morphology. B, A terminal positive phase is then recorded as the action potential propagates away from the electrode.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

Occasionally, a positive deflection can precede the negative phase of the CMAP. Volume conductor theory can explain this observation (Figure 9-16). The active electrode may not be located directly over the motor point but displaced longitudinally away from it. An active electrode located off the motor point will first record some portion of one of the source currents “feeding” the current sink. Recall that the leading portion of the source current will result in a positive deflection of the CRT trace. As propagation ensues in the muscle, the current sink will eventually reach the active electrode, resulting in a negative deflection. Finally, the terminal source current is detected, producing a positive deflection. Instead of the anticipated biphasic potential, a triphasic positive-negative-positive waveform is recorded. Relocating the active electrode over the anticipated motor point region will usually remedy the situation.

FIGURE 9-16 A, Relocating the active electrode in Figure 9-15 off the motor point results in a compound muscle action potential with an initial positive deflection, as some of the muscle’s action potentials no longer originate under the electrode but propagate toward it. B, When the main negative sink reaches the electrode, the potential’s main negative spike is detected. C, Finally, the terminal positive source currents are recorded, generating the potential’s terminal positive phase.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

This situation is complicated by the interaction of near-field and far-field waveforms. At times, a far-field waveform might predominate, depending on the active and reference electrodes’ respective locations. This can lead to a waveform with an initial positive deflection or a varying negative waveform onset (so-called false motor points). The interested reader is encouraged to explore these fascinating concepts of near-field and far-field waveform interactions.¹⁴

Fibrillation Potentials

In vitro observations have shown that approximately 6 days after denervation, a skeletal muscle fiber’s resting membrane potential decreases to a less negative level of −60 mV compared with the normal value of –80 mV.^5,45 Additionally, the resting membrane potential begins to oscillate. Because the threshold level for initiating the all-or-none action potential is now closer to the new resting membrane potential, an oscillating membrane potential will eventually reach the threshold level. Once threshold is achieved, a propagating action potential is induced, referred to as a fibrillation potential. The repolarization phase of denervated muscle results in a temporarily more hyperpolarized (–75 mV or more) level than the newly established resting membrane potential of –60 mV. When the hyperpolarized membrane level (–75 mV or more) returns to its resting membrane level of –60 mV, an action potential is again produced. This process regularly repeats, resulting in the commonly observed spontaneous and regularly repetitive discharge of fibrillation potentials. Irregularly firing fibrillation potentials occur at times, and are less well understood but thought to arise from spontaneous depolarizations within the transverse tubule system.

Fibrillation potentials are simply spontaneous depolarizations of a single muscle fiber, and demonstrate waveform configurations similar to those of single muscle fibers that are voluntarily activated (Figure 9-17). Fibrillation potentials are typically short in duration (less than 5 ms), less than 1 mV in amplitude, and fire at rates between 1 and 15 Hz. They have a typical sound, likened to a high-pitched tick like “rain on a tin roof,” when amplified through a loudspeaker. When the recording electrode is located in the previous end-plate zone of a denervated muscle, fibrillation potentials can be biphasic with an initial negative deflection. A recording electrode outside the end-plate zone, but far from the tendinous region, will detect fibrillation potentials that are triphasic (positive-negative-positive).

FIGURE 9-17 A, Monopolar needle recording of positive sharp waves (PSWs) (p) and fibrillation potentials (f)B, Only PSWs are depicted.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ. Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

It is possible to generate waveforms with similar configurations to fibrillation potentials while examining healthy muscle tissue. Specifically, inadvertently irritating the terminal axon or end-plate zone with the electrode’s shaft, and evoking end-plate potentials while simultaneously recording at some distance from the end-plate zone, will generate triphasic end-plate spikes that are identical in appearance to fibrillation potentials (Figure 9-18).¹⁴ This is understandable because both waveforms are single muscle fiber discharges. The key to identification is the rapid and irregular discharge of end-plate spikes, while fibrillation potentials fire much slower and very regularly.

FIGURE 9-18 A needle electrode is purposefully located in an end-plate region where prototypical as well as atypical end-plate spikes are evoked. Note that end-plate spikes may appear biphasic, initially negative, as anticipated, as well as triphasic and biphasic, initially positive, resembling fibrillation potentials and positive sharp waves, respectively. The important distinction between normal and abnormal waveforms in this case is not waveform configuration but rather firing rate (highly irregular for end-plate spikes).

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

Positive Sharp Waves

PSWs are waveforms that can be recorded from a single muscle fiber having an unstable resting membrane potential secondary to denervation or intrinsic muscle disease. Typically, they have a large primary sharp positive deflection followed by a small negative potential. These potentials are called PSWs (see Figure 9-17). This waveform is thought to have the same clinical significance as a fibrillation potential, in that it is a single muscle fiber discharge. Amplified through a loudspeaker, PSWs have a regular firing rate (1 to 15 Hz) and a dull thud sound. Their durations range from several milliseconds to 100 ms or longer. Although commonly observed to fire spontaneously, PSWs can be provoked by electrode movement, as can fibrillation potentials.

A number of other waveforms can be observed that have the configurations resembling a PSW. An MUAP recorded from the tendinous region can also have an initial positive deflection followed by a negative potential, because the current sink cannot pass beyond the recording electrode.¹⁴ It is also possible for the recording electrode to damage a number of muscle fibers in close proximity to the recording surface, again preventing an action potential from passing its recording surface, resulting in a primarily positive potential. These two potentials can be distinguished from a PSW in that they are MUAPs and subject to voluntary control, whereas a PSW is not. Asking the individual to contract and relax the muscle under investigation should demonstrate that the waveform has a variable firing rate. A PSW typically fires at a regular rate and is not under voluntary control. If any doubt remains, the electrode should be repositioned until successful recordings are obtained. PSWs and fibrillations will usually disappear when the needle electrode is moved a small distance such as 1 to 5 mm, while a motor unit will usually remain on the screen and change shape. This is because the fibrillation potential is generated by a very small electrical charge from only one muscle fiber, whereas an MUAP is generated by hundreds of muscle fibers covering an area of 1 to 3 cm².

Transient runs of “PSW-appearing potentials” can be seen in healthy skeletal muscle, particularly in the paraspinal and hand or foot intrinsic muscles. The “nonpathologic” PSWs are thought to arise because the needle electrode is oriented in such a manner as to irritate an end plate’s terminal axon, but extend along the muscle fiber while compressing the tissue and preventing action potential conduction.¹⁴ The induced end-plate spike appears like a PSW, but it displays a relatively rapid and irregularly firing rate characteristic of end-plate spikes (see Figure 9-18). Sustained PSWs are more significant than unsustained PSWs.

There is debate as to whether fibrillation potentials are distinct from PSWs, and if so, whether this has any clinical significance. Fibrillations potentials have been characterized as the spike form or biphasic waveform.¹² There is strong evidence that the PSW is a fibrillation that has a different shape primarily because the needle electrode is touching the fibrillating muscle cell, altering the muscle membrane and consequently the shape of the waveform.^15,16 It is unlikely that there is any clinically significant difference between the two.

Complex Repetitive Discharge

A complex repetitive discharge (CRD) is a spontaneously firing group of action potentials (formerly called bizarre high-frequency discharges or pseudomyotonic discharges) and requires a needle recording electrode for detection.¹⁴ The configuration of these waveforms is that they are continuous runs of simple or complex spike patterns (fibrillation potentials or PSWs) that regularly repeat at 0.3 to 150 Hz. The repetitive pattern of spike potentials has the same appearance with each firing, and bears the same relationship with its neighboring spikes (Figure 9-19). A distinct sound, likened to that of heavy machinery or an idling motorcycle, is produced by the firing of CRDs. In addition to the sound and repetitive pattern, a hallmark of these waveforms is that they start and stop abruptly. CRDs might begin spontaneously or be induced by needle movement, muscle percussion, or muscle contraction. Nerve block and curare do not abolish CRDs, suggesting that the origin of these potentials is within the muscle tissue. Single-fiber and standard electromyographic studies suggest CRDs are generated by an ephaptic activation of adjacent muscle cells.¹² Detection of these waveforms usually suggests that a chronic process has resulted in a group of muscle fibers becoming separated from their neuromuscular junctions. They may be associated with a currently active disease or be the residua of a past disorder.

FIGURE 9-19 Several examples of complex repetitive discharges. Note how the same pattern of potentials repeats. The sound is like that of heavy machinery, and they start and stop abruptly.

(Redrawn from Dumitru D, Amato AA, Zwarts MJ: Electrodiagnostic medicine, ed 2, Philadelphia, 2002, Hanley & Belfus, with permission.)

Myotonic Discharges

The phenomenon of delayed muscle relaxation after muscle contraction is referred to as myotonia or action myotonia.¹⁴ The finding of delayed muscle relaxation after reflex activation, or induced by striking the muscle belly with a reflex hammer, is called percussion myotonia. Clinical myotonia is usually accentuated by energetic muscle activity after a rest period. Continued muscle contraction lessens the myotonia and is known as the “warmup.” It is thought that cooling the muscle accentuates myotonia, but this finding has only been objectively documented in paramyotonia congenita.

Myotonic discharges can present in one of two waveform types (Figure 9-20). The myotonic potential induced by needle electrode insertion usually assumes a morphology similar to that of a PSW or a fibrillation potential. It is thought that the needle movement induces a repetitive firing of the unstable membranes of multiple single muscle fibers. This is because the recording needle is thought to have damaged that portion of the muscle fiber with which it is in contact. Regardless of the waveform type, the hallmark of myotonia is the waxing and waning in both frequency and amplitude. The myotonic discharge has a characteristic sound, likened to that of a dive bomber and easily recognized. Amplitudes range from 10 to 1 mV, and firing rates from 20 to 100 Hz.⁴²

FIGURE 9-20 A run of myotonic potentials demonstrating the waxing and waning nature of the discharge.

Myotonic discharges can occur with or without clinical myotonia. The observation of these potentials requires needle movement or muscle contraction. These waveforms persist after nerve block, neuromuscular block, or frank denervation. This suggests that their site of origin is the muscle membrane itself, and they appear to be due to a channelopathy. Although the exact mechanism of myotonic discharge production remains unclear, it is proposed that decreased chloride conductance is responsible at least in part for the findings in myotonia congenita.¹⁴ Myotonic discharges are not specific for any one disorder. In addition to the myotonic disorders of muscle, myotonic discharges can also be detected occasionally in acid maltase deficiency, polymyositis, drug-induced myopathies, severe axonal neuropathies, and at times in any nerve or muscle disorder.

Fasciculation Potentials

The visible spontaneous contraction of a portion of muscle is referred to as a fasciculation. When these contractions are observed with an intramuscular needle recording electrode, they are called fasciculation potentials.^4,14 A fasciculation potential is the electrically summated voltage of depolarizing muscle fibers belonging to all or part of one motor unit. Occasionally, fasciculation potentials can be documented only with needle electromyography, because they lie too deep in muscle to be seen.

Fasciculation waveforms can be characterized with respect to phasicity, amplitude, and duration (Figure 9-21). Their discharge rate (1 Hz to many per minute) is irregular and occur randomly. They are not under voluntary control, nor are they influenced by mild contraction of the agonist or antagonist muscles. The site of origin of fasciculation potentials remains unclear, although it appears that there are three possible sites: the spontaneous discharge might arise from the anterior horn cell, or along the entire peripheral nerve (particularly the terminal portion), or at times within the muscle itself.

FIGURE 9-21 Fasciculation potentials recorded from a healthy individual’s abductor hallucis.

Fasciculation potentials occur in almost all normal persons in the foot intrinsic or gastrocsoleus muscles, and in patients with a variety of diseases. Typical diseases in which fasciculation potentials can be found include motor neuron disorders, radiculopathies, entrapment neuropathies, and cervical spondylotic myelopathy. Fasciculation potentials have also been described in metabolic disturbances, including tetany, thyrotoxicosis, and anticholinesterase overdoses. Studies have unsuccessfully attempted to distinguish between benign (normal) and pathologic fasciculation potentials. There is no reliable way to categorize whether fasciculation potentials indicate a disease state just by considering their inherent characteristics based on routine needle electromyography. Perhaps the best way to evaluate fasciculation potentials is to analyze the “company they keep.” A careful analysis of voluntary MUAP morphology, combined with a search for abnormal spontaneous potentials, is required before concluding that fasciculation potentials are either a normal or an abnormal finding.

Myokymic Discharge

Myokymia is frequently readily observable as vermicular (bag of live worms) or rippling movement of the skin. It is usually associated with myokymic discharges.¹⁴ The myokymic discharge consists of bursts of normal-appearing motor units with interburst intervals of electrical silence. Typically, the firing rate is 0.1 to 10 Hz in a semirhythmic pattern. Two to 10 potentials within a single burst can fire at 20 to 150 Hz. These potentials are not affected by voluntary contraction (Figure 9-22). The sound associated with these potentials is a type of sputtering often heard with a low-powered motorboat engine. The actual discharge may be distinguished from CRDs in that myokymic discharges do not display a regular pattern of spikes from one burst to the next, nor do they typically start and stop abruptly. Myokymic discharges are groups of motor units, while CRDs represent groups of single muscle fibers. The groups of motor units within a burst may fire only once or possibly several times. The sputtering bursts of myokymic discharges sound quite different than the continuous drone of a CRD.

FIGURE 9-22 Myokymic discharge from a person with radiation plexopathy. Note the bursts of motor unit activity with electrical silence between bursts.

Myokymic potentials can be observed in the face (facial myokymia), arising from multiple sclerosis or a brainstem neoplasm. Segmental myokymic discharges can be noted in syringomyelia or radiculopathies. Generalized myokymic discharges have been detected in uremia, thyrotoxicosis, inflammatory polyradiculoneuropathy, and Isaac’s syndrome. Limb myokymic discharges have also been described associated with radiation plexopathy and chronic compressive neuropathies.¹

Continuous Muscle Fiber Activity

A number of relatively rare syndromes producing continuous muscle fiber activity associated with muscle stiffness have been reported.¹⁴ Portions of both the central and the peripheral nervous system have been implicated in generating the sustained firing of motor units. One syndrome with continuous muscle fiber activity is known as “stiff man syndrome.” The motor unit discharges in this condition are believed to have a central origin, as they are abolished or attenuated by peripheral nerve block, neuromuscular block, spinal block, general anesthesia, and sleep. The continuous motor unit firing is diminished by diazepam but not by phenytoin or carbamazepine. The patient can voluntarily control motor unit activity, but the overriding involuntary firing returns when the patient relaxes. Progressive muscle stiffness involving all muscles (including the chest wall and pharynx) eventually occurs, resulting in contractures and profound impairment. A needle electrode recording reveals normal motor unit potentials producing a sustained discharge pattern in both the agonists and the antagonists.

Neuromyotonia is a “peripheral” form of continuous muscle fiber activity originating in the peripheral motor axon referred to as Isaac’s syndrome. The continuous motor unit activity is eliminated by neuromuscular block but not peripheral nerve block, spinal or general anesthesia, or sleep. The motor unit activity usually begins in the lower limbs in the late teens and progresses to all skeletal muscles. Needle recordings demonstrate motor unit discharges with frequencies up to 300 Hz, including doublets, triplets, and multiplets (see later discussion).

Cramps

A sustained and possibly painful muscle contraction of multiple motor units, lasting seconds or minutes, can appear in normal individuals or specific disease states.³⁴ In healthy subjects, a cramp usually occurs in the calf muscles or other lower limb muscles after exercise, abnormal positioning, or maintaining a fixed position for a prolonged period. Cramps might also be induced by hyponatremia, hypocalcemia, vitamin deficiency, or ischemia. They also occur in early motor neuron disease and peripheral neuropathies. Familial syndromes have been reported that involve fasciculations and cramps; alopecia, diarrhea, and cramps; and simply an autosomal dominantly inherited cramp syndrome.

A needle recording electrode placed into a cramping muscle shows multiple motor units firing synchronously between 40 and 60 Hz, and occasionally reaching 200 to 300 Hz (Figure 9-23). A large portion of the muscle is simultaneously involved in a cramp, as opposed to the asynchronous excitation of motor units during voluntary activation. Cramps are believed to arise from a peripheral portion of the motor unit. A cramp that results in a taut muscle with electrical silence is the physiologic contracture seen in McArdle disease.

FIGURE 9-23 A characteristic muscle cramp as recorded with an intramuscular needle electrode. There is an initial burst of motor unit activity that eventually subsides as the cramp dissipates.

(Redrawn from Daube JA. AAEM minimonograph 11: Needle examination in electromyography, Rochester, 1979,  AAEM, with permission.)

Multiplet Discharges

A clinical syndrome manifested by spontaneous muscle twitching, cramps, and carpopedal spasm is known as tetany.¹⁴ This entity usually results from peripheral and/or central nervous system irritability associated with systemic alkalosis, hypocalcemia, hyperkalemia, hypomagnesemia, or local ischemia. Clinically, one can induce tetany by tapping the facial nerve (Chvostek’s sign), the peroneal nerve at the fibular head (peroneal sign), and inducing limb ischemia (Trousseau’s sign).

In the above conditions, characteristic motor unit potentials can be observed. A single motor unit potential might fire rather rapidly, with an interdischarge interval of 2 to 20 ms. If the motor unit fires twice, it is referred to as a doublet; three times denotes a triplet; and more than three firings is called a multiplet (Figure 9-24). These potentials can be seen after voluntary contraction, or can be observed spontaneously from the induction maneuvers noted above (Chvostek’s sign or Trousseau’s sign), in which MUAPs can fire in long trains or short bursts of 5 to 30 Hz (tetany). Motor units with an interdischarge interval of 20 to 80 ms are called “paired discharges” but can arise in similar states, as previously described.

FIGURE 9-24 Doublets (upper trace) and multiplets (lower trace), motor unit action potentials resulting from voluntary contraction.

(Redrawn from Daube JA: AAEM minimonograph 11: Needle examination in electromyography, Rochester, 1979, AAEM, with permission.)

Seddon’s Classification

The degree to which a nerve is damaged has obvious implications with respect to its present function and potential for recovery. There are essentially two general classification systems.^39,43 One classification is that of Seddon, and considers neural injury from the perspective of a combination of functional status and histologic appearance. In Seddon’s scheme, there are three degrees or stages of injury to consider: neurapraxia, axonotmesis, and neurotmesis (Table 9-1).

Sunderland’s Classification

A second popular and somewhat more detailed classification is that proposed and subsequently modified by Sunderland. This classification of nerve injury is based on the results of trauma with respect to the axon and its supporting connective tissue structures. Basically, Sunderland’s classification is divided into five types of injury, based exclusively on which connective tissue components are disrupted (Figure 9-25). Type 1 injury corresponds to Seddon’s designation of neurapraxia. Seddon’s axonotmesis is subdivided by Sunderland into three forms of neural insult (types 2 through 4). A type 2 injury involves loss of axonal continuity with preservation of all supporting neural structures, including the endoneurium (this closely corresponds to Seddon’s axonotmesis). Type 3 and 4 injuries result in progressively more neural disruption. Sunderland’s type 5 injury corresponds to Seddon’s neurotmesis (complete neural disruption).

FIGURE 9-25 The five degrees of neural injury as classified by Sunderland. Type 1, conduction block; 2, Wallerian degeneration occurring secondary to a lesion confined to the axon, with preservation of the endoneurial sheath; 3, there is noted to be disruption of the axon and endoneurial tube, within an intact perineurium; 4, disruption of all neural elements except the epineurium; and 5, complete discontinuity of the entire nerve trunk.

(Redrawn from Sunderland S: Nerve injuries and their repair: a critical appraisal, Edinburgh, 1991, Churchill Livingstone, with permission.)

These are the most commonly used classifications. Unfortunately, they do not describe the pathology for the most common type of nerve injury, a compression neuropathy such as carpal tunnel syndrome (CTS). The lesion in CTS usually causes thinning of the myelin and widening of the nodes of Ranvier, causing the slowing of latency and velocity seen in NCS. Neurapraxia and axon loss can also occur in CTS, but this is much less common than the unclassified thinning of myelin.

Electrodes

The two types of electrodes are surface and needle. Surface electrodes are manufactured in various sizes and shapes so as to best conform to the body part under investigation. The electrode must be firmly secured to the patient to preclude any movement. Well-secured electrodes minimize movement artifact that could contaminate the desired signal.

Two commonly used needle recording electrodes are monopolar and concentric (Figure 9-27). The monopolar needle is a solid, stainless steel shaft coated completely with Teflon except for the bare metal tip. It is this bare metal tip that acts as the recording surface. The needle is typically 12 to 75 mm in length and 0.3 to 0.5 mm in diameter, with a recording surface of 0.15 to 0.6 mm². Separate reference and ground electrodes are required. The concentric needle electrode is a hollow, stainless steel hypodermic needle with a central platinum or Nichrome-silver wire about 0.1 mm in diameter, surrounded by epoxy resin acting as an insulating material from the surrounding cannula. The cannula has a similar length and diameter to that of the monopolar needle. A separate ground is required, but the cannula serves as the reference electrode.

FIGURE 9-27 Three needle electrodes and their recording areas. A, A monopolar needle electrode requires a separate reference and ground electrode. B, The concentric needle electrode uses the cannula as the reference, while a separate ground is required. C, The single-fiber electrode is essentially a modified concentric needle electrode.

(Redrawn from Dumitru D, Walsh NE. Electrophysiologic instrumentation. In: Dumitru D, editor: Physical medicine and rehabilitation state of the art reviews: clinical electrophysiology, Philadelphia, 1989, Hanley & Belfus, with permission.)

There has been considerable discussion about the merits of each of these electrodes compared with the other. Both electrodes have advantages and disadvantages depending on the clinical circumstances. Monopolar needle electrodes have a wider recording territory and a distant reference, which makes the recording more “noisy” with respect to distant activity and interference. On the other hand, the Teflon coating probably reduces patient discomfort depending on technique. The concentric needle electrode has the active and reference electrodes close together, making them electrically quieter than monopolar needles. Concentric electrodes typically cause more patient discomfort, but again, this is likely technique dependent. Concentric needle electrodes yield the following as compared with monopolar needle electrodes: smaller potential amplitudes, possibly fewer phases, comparable durations, and less distant motor unit activity. The introduction of commercially available, disposable monopolar and concentric needle electrodes has eliminated such worries as Teflon peeling back on the monopolar needles, and hook formation on the tip of the concentric needle electrodes. The quality of disposable needle electrodes has improved, eliminating the need to use nondisposable needle electrodes. If electrodes are reused, they should be properly sterilized with presoaking in sodium hypochlorite and steam autoclaving.¹⁴ There are concerns about the possibility of spreading prion disease using sterilized, reusable needles.³⁷ Documentation of this, however, is lacking.

Single-fiber electrodes are essentially modified concentric needle electrodes. A small, 25-μm recording port is placed opposite the electrode’s bevel and several millimeters from the tip. This makes this special electrode capable of recording the electrical activity from a single muscle fiber. The uptake area for this electrode is approximately 300 μm (see Figure 9-27). Recent studies show comparable jitter values comparing disposable concentric needles and single-fiber electrodes.³²

Amplifier

Biologic signals range in size from microvolts to millivolts, thereby requiring significant amplification before being analyzed. An amplifier is simply a device with the ability to magnify the desired signals and minimize unwanted signals or noise. Amplification is expressed as gain or sensitivity. Gain is a ratio of the signal’s output divided by the input. For example, an output of 1 V for an input of 10 mV implies that the amplifier has a gain factor of 100,000 (output/input= 1V/0.00001V= 100,000). Sensitivity is the ratio of the input voltage to the size of deflection on the CRT, and is usually measured in centimeters. For example, an amplifier that produces a 1-cm deflection for an input of 10 mV has a sensitivity of 10 mV/cm or 10 mV/division. The sensitivity or gain setting used is important because it can influence the onset latency. Increasing the sensitivity for a given waveform results in the instrument displaying the potential’s initial departure from baseline earlier in time, because progressively earlier and hence smaller baseline deviations can now be observed.

Understanding amplification is simplified with two concepts: (1) differential amplification in which different signals are amplified, and (2) common mode rejection where common signals are eliminated. The standard electromyograph has two amplifiers. The amplifier connected to the active electrode is known as the noninverting amplifier, while the reference electrode is connected to the inverting amplifier. The inverting amplifier magnifies the signal presented to it in essentially the same manner as the noninverting amplifier, with the exception of inverting the signal. Both amplified signals (inverted and noninverted) are then electronically summated, and like signals are cancelled. When an identical signal is presented to both amplifiers, theoretically there should be no output from the instrument because there is elimination of the same or so-called common signals. For example, 60-Hz interference recorded by the active and reference electrodes is eliminated as a common mode signal. It is impossible to build two amplifiers with identical properties, so common mode rejection can never be perfect. The ratio of the instrument’s differential to common output (common mode rejection ratio) should exceed 100,000:1. For example, suppose the noninverting amplifier is set with a gain of 100, and the inverting amplifier is also set with a gain of 100 but in reality has a gain of 99.9, remembering that the two amplifiers can be very similar but not identical. A 100-μV signal of interest is presented to our noninverting amplifier as recorded by the active recording electrode. The reference electrode presents no signal to the inverting amplifier because it is located in an “electrically silent” area compared with the signal of interest. The result is a signal that has been magnified to 10,000 μV. Additionally, there is a 1-μV noise signal present that is detected by both the active and the reference electrodes, resulting in this signal amplified to 100 μV and 99.9 μV for the noninverting and inverting amplifiers, respectively. The difference between these two signals, as electrically subtracted from each other by the two amplifiers, is 0.1 μV. The common mode rejection ratio is then the difference signal (10,000 μV) divided by the common mode signal (noise presented to both amplifiers) of 0.1 μV (10,000 μV/0.1 μV), yielding the anticipated ratio of 100,000:1.

Filters

Perhaps the most misunderstood and ignored aspect of the electrodiagnostic instrument is the filters. The main purpose of filters is to form a window or bandwidth of frequencies containing the desired waveform, but exclude those frequencies not comprising the signal of interest (“noise”). Low- and high-frequency filters are used to prevent those frequencies below and above the respective filter settings from being amplified and subsequently interfering with the signal of interest. A high-frequency filter eliminates those frequencies higher than its numeric designation and permits frequencies lower than its numeric designation to pass. That is why a high-frequency filter is also known as a low-pass filter: it permits low frequencies to pass through and be displayed. Similarly, a low-frequency filter extracts low frequencies from the total recorded signal below its numeric designation. A low-frequency filter permits frequencies greater than its numeric designation to pass.

Any biologic signal can be conceptualized as a series of sine waves of various frequencies and amplitudes. The combination of “appropriate” sine wave amplitudes and frequencies can result in the formation of essentially any waveform. In this way, the biologic signal recorded by the instrument consists of multiple subcomponent waveforms with specific frequency and amplitude characteristics. Eliminating any of these subcomponent waveforms results in a distortion of the waveform’s appearance. That is exactly what can happen if the high- or low-frequency filters are set such that the desired biologic waveform has various subcomponent frequencies eliminated. The examples provided below apply equally well to nerve and muscle potentials.

In our example, a recorded median SNAP and CMAP are sequentially distorted by altering the low- and high-frequency filter settings. Let us begin with an arbitrary low-frequency filter setting of 1 Hz and a high-frequency filter cutoff of 10,000 Hz. Sequentially elevating the low-frequency filter from 1 to 10 Hz, 100 Hz, and finally 300 Hz, while maintaining a high-frequency filter at 10,000 Hz, results in characteristic waveform distortions (Figure 9-28). The onset latency does not change, the peak latency decreases, amplitude is serially reduced, and the total potential duration decreases. Also, an additional phase is created. The use of sequentially higher low-frequency filter settings removes progressively more low frequencies from the SNAP. In other words, the remaining potential now has a predominance of high frequencies in it as compared with the original potential. The onset of the potential remains unaltered because it is a quick departure from baseline and is not influenced by an alteration in the low-frequency content of the waveform. The remainder of the potential, however, is influenced by the now predominant high-frequency subcomponent waveforms. By taking out the low frequencies, the amplitude declines as subcomponent waveforms are removed. Clearly, the waveform’s magnitude could not possibly increase because we are taking out part of its content. Further, the remainder of the waveform is shifted to an earlier time of occurrence because of the high frequencies remaining in the SNAP. Because comparatively high frequencies occur sooner in time, it is logical that all the waveform shifts toward the potential’s stationary onset, producing decreases in its peak latency, negative spike duration, and total duration. A third phase is created as the potential begins to appear more like a sine wave as the higher frequencies begin to emerge. Remember, high frequencies have more phases per unit time. A similar but more dramatic occurrence is noted for the CMAP, because it comprises comparatively more low frequencies than a SNAP does.

FIGURE 9-28 An antidromic median sensory nerve action potential is evoked from the third digit, while the low-frequency filter is sequentially elevated with a constant high-frequency filter of 10,000 Hz. Note how the onset latency does not change; however, the peak latency decreases, as does the potential’s amplitude. A third phase is also produced at the low-frequency filter setting of 300 Hz.

(Redrawn from Dumitru D, Walsh NE: Practical instrumentation and common sources of error, Am J Phys Med Rehabil 67:55-65, 1988, with permission.)

Eliminating high frequencies results in a somewhat different set of alterations. Because we are removing waveforms from the total potential, a reduction in amplitude can be anticipated. Because the SNAP is now biased toward a potential with more low frequencies, it takes longer to occur in time. This results in a delay of both the onset and the peak latencies (Figure 9-29). Lowering the high-frequency filter while maintaining a constant low-frequency filter results in a waveform with a comparatively smaller amplitude, longer onset latency, and longer peak latency, as well as longer overall duration. Similar findings can be observed for a CMAP.

FIGURE 9-29 An antidromic median sensory nerve action potential is recorded from the third digit, while a constant low-frequency filter of 10 Hz but different high-frequency filters are used. Note how the onset and peak latencies are sequentially delayed with decreasing high-frequency filter settings. The potential’s amplitude also decreases.

(Redrawn from Dumitru D, Walsh NE: Practical instrumentation and common sources of error, Am J Phys Med Rehabil 67:55-65, 1988, with permission.)

There are no universally agreed filter settings for any electrodiagnostic medicine procedure. Arriving at optimal filter settings is highly empiric. The high- and low-frequency filters are lowered and raised, respectively, until waveform distortions are observed. The filters are then expanded until no waveform changes are noted. The goal is to include the major components of the waveforms while eliminating undesired signals or noise. The most important factor is to reproduce all filter settings originally described by those investigators whose reference data are being used (Table 9-2).

Table 9-2 Recommended Filter Settings

From Dumitru D, Walsh NE. Practical instrumentation and common sources of error, Am J Phys Med Rehabil 67:55-65, 1988, with permission.

Sound

After the biologic signal is filtered, it is fed to a loudspeaker. The acoustic analysis of both normal and abnormal potentials is extremely important. It is not uncommon for practitioners to “hear” an abnormality before viewing it on the CRT. The instrument must have a relatively good speaker so as to accurately present the sounds associated with the biologic signals.

Stimulator

Two different types of stimulators are commercially available: constant current and constant voltage. For both types of devices, neural tissue is activated under the cathode (negative pole) while the anode (positive pole) completes the stimulating circuit. A constant current stimulator effectively delivers the desired current output for each stimulus irrespective of the resistance between the skin and cathode or anode. This is accomplished by varying the voltage or current driving force as is necessitated by any alterations in the skin-stimulator interface’s resistance. Similarly, a constant voltage stimulator is designed to deliver the same voltage with each stimulus, even if the resistance between the skin and stimulator changes. A compensatory increase or decrease in current is provided so as to maintain the same voltage level. In short, a constant current stimulator is effectively a variable voltage stimulator, while a constant voltage stimulator is a variable current stimulator. Both stimulators are acceptable for most purposes. The constant current stimulator is preferred when the same current must be delivered for each stimulus in clinical situations requiring quantification of current delivery, such as in somatosensory evoked potentials, research, or evaluating side-to-side stimulation thresholds during facial nerve excitability testing.

An important problem associated with stimulators is the stimulus artifact. It is commonly a large potential recorded during the delivery of the stimulus. At times, the magnitude of the potential is large enough to compromise the desired neural or muscular response. In such situations, it is necessary to minimize the shock artifact. An effective method of reducing the shock artifact is to use fast recovery amplifiers that act to suppress this artifact by quickly recovering from the overwhelming voltage delivered. These amplifiers are not available on all instruments, and so other means must be found to deal with the artifact. The skin surface must be dry, and any perspiration, body lotion, makeup, or other surface conductors should be removed. Wiping a large portion of the body segment under investigation with an alcohol pad usually removes all surface conducting films. A ground electrode is best placed between the stimulus site and the active recording electrode. Wire leads between the patient and stimulator should be separated to avoid any type of capacitive interaction. The stimulator circuit should be isolated from the instrument’s ground circuit, which is true of virtually all commercially manufactured instruments. Perhaps the most effective method of reducing stimulus artifact, once all the above have been addressed, is to rotate the anode about the cathode. This optimizes the stimulator’s voltage output, as recorded by the active and reference electrode, to take advantage of differential amplification and the elimination of the shock artifact as a common mode signal. An attempt is made to have both the active and the reference electrodes record similar voltages, thereby minimizing the stimulator’s signal from being amplified and displayed along with the signal.

Anodal block is an interesting concept that has been widely discussed but has little supporting experimental data. Theoretically, the anode hyperpolarizes the neural tissue in its immediate vicinity and should result in an action potential failing to conduct past the anode. In humans, investigations using bipolar and monopolar anodal current stimulation at the highest current outputs failed to document any type of anodal block. Because the anode is capable of stimulating neural tissue and not blocking it, at this time it appears that anodal block does not occur during the routine electrodiagnostic medicine consultation.