Electrical Stimulatio

doi: 10.2522/ptj.20080060Originally published online December 18, 2008

2009; 89:181-190.PHYS THER. Alex R WardAlternating CurrentElectrical Stimulation Using Kilohertz-Frequency

http://ptjournal.apta.org/content/89/2/181found online at: The online version of this article, along with updated information and services, can be

Collections

Perspectives Electrotherapy

in the following collection(s): This article, along with others on similar topics, appears

e-Letters

"Responses" in the online version of this article. "Submit a response" in the right-hand menu under

or click onhere To submit an e-Letter on this article, click

E-mail alerts to receive free e-mail alerts hereSign up

by guest on February 15, 2015http://ptjournal.apta.org/Downloaded from by guest on February 15, 2015http://ptjournal.apta.org/Downloaded from

http://ptjournal.apta.org/content/89/2/181http://ptjournal.apta.org/cgi/collection/electrotherapyhttp://ptjournal.apta.org/cgi/collection/perspectiveshttp://ptjournal.apta.org/letters/submit/ptjournal;89/2/181http://ptjournal.apta.org/subscriptions/etoc.xhtmlhttp://ptjournal.apta.org/http://ptjournal.apta.org/

Electrical Stimulation Using Kilohertz-Frequency Alternating CurrentAlex R Ward

Transcutaneous electrical stimulation using kilohertz-frequency alternating current(AC) became popular in the 1950s with the introduction of interferential currents,promoted as a means of producing depth-efficient stimulation of nerve and muscle.Later, Russian current was adopted as a means of muscle strengthening. This articlereviews some clinically relevant, laboratory-based studies that offer an insight intothe mechanism of action of kilohertz-frequency AC. It provides some answers to thequestion: What are the optimal stimulus parameters for eliciting forceful, yet com-fortable, electrically induced muscle contractions? It is concluded that the stimula-tion parameters commonly used clinically (Russian and interferential currents) aresuboptimal for achieving their stated goals and that greater benefit would be obtainedusing short-duration (24 millisecond), rectangular bursts of kilohertz-frequency ACwith a frequency chosen to maximize the desired outcome.

AR Ward, PhD, is Associate Pro-fessor, Musculoskeletal ResearchCentre, Faculty of Health Sciences,La Trobe University, Victoria 3086,Australia. Address for correspon-dence: School of Human Bio-sciences, Faculty of Health Sciences,La Trobe University, Victoria 3086,Australia. Address all correspon-dence to Dr Ward at: [email protected].

[Ward AR. Electrical stimulationusing kilohertz-frequency alter-nating current. Phys Ther. 2009;89:181190.]

2009 American Physical TherapyAssociation

Perspective

Post a Rapid Response orfind The Bottom Line:www.ptjournal.org

February 2009 Volume 89 Number 2 Physical Therapy f 181 by guest on February 15, 2015http://ptjournal.apta.org/Downloaded from

http://ptjournal.apta.org/

Two forms of electrical stimula-tion are commonly used clini-cally: pulsed current (PC) andburst-modulated alternating current(BMAC). Examples of BMAC areRussian current and interferentialcurrent. Burst-modulated alternat-ing current stimulation is claimedto be more comfortable than PC andcapable of eliciting greater muscletorque.13

The response of nerve and muscle toPC electrical stimulation has beenstudied by physiologists since thelate 19th century.4,5 Consequently,our present understanding of theeffects of PC is relatively good. Thephysiological response to BMACstimulation is less-well understood.

This article reviews the knownphysiology and clinically relevant,laboratory-based studies of electricalstimulation, which offer some in-sight into the mechanism of actionof BMAC and provide some answersto the questions Does BMAC stimu-lation have an advantage over PC?and What are the optimumtreatment parameters for BMACstimulation?

BMAC Stimulus ParametersAlternating current (AC) used clini-cally is normally kilohertz-frequencyAC, delivered in bursts, with theburst frequency in the physiologi-cal range (up to 100 Hz or so). It,therefore, is called burst-modulatedalternating current. Figure 1 illus-trates, for comparison, unmodulatedAC, monophasic PC, and 2 examplesof BMAC.

The currents illustrated in Figures1A, 1C, and 1D are defined as ACbecause the waveforms have alter-nating positive and negative phaseswith no gap between them. The cur-rent shown in Figure 1B is defined asPC because successive phases (thepulses) are separated by an apprecia-ble gap.6

Pulsed current is easily described byspecifying 3 things: (1) the wave-form (eg, rectangular and monopha-sic, as in Fig. 1B), (2) the pulse du-ration (normally in the range of 50microseconds to 1 millisecond), and(3) the pulse frequency (normally inthe range of 1 Hz to about 100 Hz).

The description of AC is more com-plex. Alternating current, by defini-tion, is biphasic, and the biphasicwaveform can be sinusoidal or rect-angular. The current also can bedelivered continuously (Fig. 1A), inrectangular bursts (Fig. 1C), or insinusoidally modulated bursts(Fig. 1D). Thus, when describing thestimulus, there is the potential forconfusion because several parame-ters must be specified to completelydescribe the waveform. Figure 2shows an example of BMAC, withparticular parameters identified.

In Figure 2, the burst duration is 4milliseconds, and because the inter-val between bursts is 16 millisec-onds, the period (the burst repeti-tion time) is 20 milliseconds, or1/50th of a second. Therefore, theburst repetition frequency is 50times per second in this example (ie,the burst frequency is 50 Hz). Eachburst consists of a number of ACcycles. In this example, each4-millisecond burst consists of 4 ACsine waves. Each sine wave has aduration of 1 millisecond, or1/1,000th of a second, so the sine-wave frequency is 1,000 times persecond (ie, 1 kHz). The sine-wavefrequency is sometimes referred toas the carrier frequency.1,2,7 Each1-millisecond sine wave comprises 2phases: one positive phase followedby one negative phase, so eachphase has a duration of 0.5 millisec-onds, or 500 microseconds.

The greater the number of parame-ters, the greater the number of pos-sible permutations and combina-tions. This raises the question of

whether AC stimulators used clini-cally have the best combination ofparameters for achieving the desiredclinical outcome.

BMAC Stimulation TypesUsed ClinicallyRussian CurrentRussian current is 2.5-kHz AC, ap-plied in 50-Hz rectangular burstswith a burst duty cycle of 50%. Thestimulus waveform is shown in Fig-ure 1C. The burst duration is 10 mil-liseconds at 50 Hz. Russian currentis claimed to be beneficial formuscle strengthening (increasingforce-generating capacity). Thechoice of a 2.5-kHz frequency forRussian current appears to be basedon measurements of maximum elec-trically induced torque (MEIT) byKots and co-workers8 using notbursts but a continuous AC stimulus(Fig. 1A) in the frequency range of100 Hz to 5 kHz.8,9 The choice of aburst-modulated, 50% duty cycle(Fig. 1C) is based on the observationthat there was little difference inMEIT between continuous AC andrectangular bursts with a 50% dutycycle but that with a 50% duty cycle,half as much electrical energy is de-livered, so there is less risk for tissuedamage.8,9

Russian currents became popular de-spite an equivocal evidence base dueto the limited number of studies andtheir different findings.3,9 The bal-ance of evidence supports the no-tion that strengthening can be pro-duced, but at one extreme there isthe single-case study reported byDelitto et al,10 which demonstrated asubstantial strength gain, whereas atthe other extreme there is the studyby St Pierre et al,11 which demon-strated no strength gain. Other thanthe original Russian study,8,9 only 2subsequent studies have addressedwhether 2.5 kHz is the best AC fre-quency for muscle torque produc-tion.12,13 These 2 studies used 50-Hzbursts of kilohertz-frequency AC,

Electrical Stimulation Using Kilohertz-Frequency Alternating Current

182 f Physical Therapy Volume 89 Number 2 February 2009 by guest on February 15, 2015http://ptjournal.apta.org/Downloaded from


and both studies showed that maxi-mum torque was elicited at a fre-quency of 1 kHz. It is noteworthythat Andrianova et al8 reported that2.5 kHz is optimum if stimulation isapplied directly (over the muscle)but that if stimulation is applied in-directly (over the nerve trunk), theoptimum frequency is 1 kHz. Thus, itmight be concluded that optimalstimulus parameters may well de-pend on electrode positioning andthat the popular frequency (2.5 kHz)could be suboptimal for commonlyused electrode placements.

Interferential CurrentsInterferential currents are reportedto be the most popular form of elec-trical stimulation used in clinicalpractice in the United Kingdomand other European countries andin Australia.1 Interferential stim-ulators produce 2 independentkilohertz-frequency AC currents ofconstant intensity (Fig. 1A) appliedby 2 separate pairs of electrodes,which are positioned diagonallyopposed to produce an interfer-ence effect (Fig. 1D) in the centralregion of intersection of the cur-rents (Fig. 3).1,2,7

The currents are applied continu-ously at constant intensity (Fig. 1A),but they have different frequencies(eg, 4,000 and 4,050 Hz), and in thetissue between the electrodes, the 2currents interfere. It is stated1,2,7 thatthe currents reinforce in the centralregion of intersection (Fig. 3A) toproduce a stimulus waveform that is

sinusoidally modulated at a fre-quency equal to the difference be-tween the 2 AC frequencies (Fig. 3B,top). The stimulation waveform,therefore, resembles that illustratedin Figure 1D and would have a mod-ulation frequency of 50 Hz in thisexample. This argument is mislead-ing because it ignores the effect of

Figure 1.(A) Steady, unmodulated alternating current; (B) monophasic pulsed current; (C) burst-modulated alternating current with rectan-gular burst modulation; and (D) burst-modulated alternating current with sinusoidal modulation.

Figure 2.An example of burst-modulated alternating current. A minimum of 5 parameters mustbe specified in order to describe the waveform. In this example, the waves are sinusoi-dal, the alternating current (AC) frequency is 1 kHz, the bursts are rectangular, the burstfrequency is 50 Hz, and the burst duration is 4 milliseconds.




tissue inhomogeneity and, perhapsmore importantly, nerve fiber orien-tation.14,15 Nerve fibers orientedalong an axis directly between onepair of electrodes will experiencecontinuous, unmodulated AC, whileonly those angled optimally betweenthe 2 axes will experience fully mod-ulated AC (Fig. 3). The optimum an-gle depends on the relative intensi-ties of the current. If the currentintensities are equal, the optimumangle is 45 degrees to the currentpaths (ie, horizontally or vertically inFig. 3A), but in practice the currentswill not be equal due to the variationwith position (relative to the elec-trodes) and the variation in electricalimpedance of different tissues (fat,muscle, connective tissue, and bone)in the current pathway.15 Unless theorientation of the nerve fibers is op-timal, the stimulus modulation willbe partial. Thus, with interferentialcurrents, the actual stimulus wave-form applied to the nerve fibers isnot known and can vary betweenunmodulated and fully modulatedAC (Fig. 3B), depending on the nervefiber orientation and location rela-tive to the electrode placement.

Premodulated InterferentialCurrentMost interferential stimulators alsooffer premodulated interferentialcurrent. The term premodulated in-terferential is something of a misno-mer because it refers to current thatis fully modulated (as in Fig. 1D) andapplied between one pair of elec-trodes. Thus, by definition, this cur-rent is no longer interferential (ie, nolonger produced by the interferenceof 2 currents). The current is simplykilohertz-frequency AC, modulatedat a low frequency, typically in therange of 1 to 120 Hz.1,2,7 Unliketrue interferential current, theamount of modulation of the stimu-lation waveform does not depend onthe nerve fiber orientation relative tothe electrodes. The stimulus wave-form is simply that provided bythe stimulator and, therefore, ispredictable.

If the premodulated current is si-nusoidally modulated (as producedby traditional interferential stimula-tors and shown in Fig. 1D), someparts of the burst will be belowthreshold while other parts of theburst will be above threshold. Thus,the effective burst duration for any

given nerve fiber is uncertain andwill vary with stimulation intensity,which varies with proximity to theelectrodes. Nerve fibers close to theelectrodes will be stimulated su-prathreshold for a larger part of eachburst than those further away; thus,the effective burst duration will vary.Some modern interferential stimula-tors use rectangular burst modula-tion (Fig. 1C), so there is no uncer-tainty as to the effective duration:the burst is either fully on or off.

Importance of ModulationEffect of Burst Duration onThresholdsAs noted earlier, Russian current isburst modulated with a rectangularenvelope (Fig. 1C). Premodulated in-terferential current may be eitherrectangular burst modulated (Fig. 1C)or, more commonly, sinusoidallymodulated (Fig. 1D), whereas withtrue interferential currents, thestimulus experienced by a nerve fi-ber may be continuous (unmodu-lated), fully modulated, or partiallymodulated, depending on the fiberlocation and orientation relative tothe electrodes.

Figure 3.(A) Interferential currents are claimed to produce maximum stimulation in the region of intersection of the 2 diagonally opposedcurrents, as shown. (B) The actual stimulation intensity experienced by nerve fibers has maximum modulation if the fibers areoriented optimally and zero modulation when fibers are oriented along one of the current pathways.




The first published report of the ef-fect of modulation appears to bethe report by Soloviev published in1963.16 Soloviev used AC stimulationover the frequency range of 2 to 8kHz and found that there was littledifference in motor threshold regard-less of whether the current appliedwas continuous or burst modulatedat 50 Hz with a 50% duty cycle. A2001 study of motor thresholds byWard and Robertson17 again showedlittle difference, this time over thefrequency range of 1 to 25 kHz. Itshould be noted, however, that onlycontinuous AC and 50% duty cycle,50-Hz bursts were compared, so thecomparison was between 10-milli-second bursts and continuous AC.

In 2007, Ward and Lucas-Toumbourou18 reported a study ofsensory, motor, and pain thresholdsusing AC frequencies of 1 kHz and 4kHz applied as 50-Hz bursts. Theyused burst durations in the range of0.25 to 20 milliseconds and foundthat thresholds decreased to a pla-teau with increasing burst duration.An interesting finding of this studywas that the plateau in thresholdwith burst duration depended on theresponse evoked (ie, sensory, motor,or pain threshold, with values of 5 to7, 10, and 20 milliseconds, respec-tively). Thus, motor thresholds de-crease with increasing burst dura-tion, but at burst durations aboveabout 10 milliseconds, there is nofurther decrease. These findings ex-plain the lack of differences found inthe earlier studies, when only 10-millisecond bursts and continuousAC were compared. The most in-triguing finding of the study, how-ever, was that the burst duration pla-teaus were different for sensory,motor, and pain thresholds. Thismeans that there will be optimalburst durations where the pain/sensory threshold and pain/motorthreshold ratios are maximum. Wardand Lucas-Toumbourou estimated anoptimal burst duration for both sen-

sory and motor stimulation as 2 to3 milliseconds. This is appreciablyshorter than the burst durationscommonly used clinically (typically10 milliseconds for Russian currentand greater or similar for interferen-tial currents).

Effect of Burst Duration on MEITand DiscomfortAndrianova et al8 used different ACfrequencies in the range of 100 Hz to5 kHz and compared not thresholdsbut maximum torque productionusing continuous (unmodulated) ACand AC bursts (modulated at 50 Hzwith a 50% duty cycle [ie, 10-millisecond burst duration]). Theyconcluded that there was little differ-ence in MEIT with burst-mode orcontinuous stimulation, but they didnot make any statistical compari-sons. Their published data (repro-duced in Tab. 3 of the perspectivearticle by Ward and Shkuratova9)however, show that across the fre-quency range, the torques producedby burst-modulated currents were,on average, 14% higher (SD12%).Ward and Shkuratova9 conducted apaired t-test comparison across fre-quencies, using Andrianova and col-leagues published data,8 and foundthat this difference is significant(P.03) (ie, torques are significantlyhigher when a rectangular burst-modulated stimulus of 10 millisec-onds duration is used rather than acontinuous AC stimulus).

Bankov,19 in 1980, compared 5-kHzAC, modulated at 60 Hz, using stim-ulation intensities that produced justenough contraction of the bicepsbrachii muscle to maintain the elbowat 90 degrees of flexion with the up-per arm vertical (an antigravity flex-ion level of muscle activity). He com-pared rectangular bursts of 1, 2, and5 milliseconds duration and re-ported that the 1-millisecond burstwas the most comfortable. Anotherstudy reported by Bankov in thesame year20 compared 60-Hz sinusoi-

dally modulated bursts of AC, whichvaried in their modulation depthfrom 0% (steady, continuous AC;Fig. 1A) to 100% (fully modulated;Fig. 1D), and hypermodulated burstsof AC (gaps between bursts). He re-ported that force increased with thedegree of modulation but that theassociated discomfort showed littlevariation. A conclusion is thatshorter burst durations producemore force at the same level of dis-comfort. In 1981, Bankov andDaskalov21 compared 5-kHz AC ap-plied in 2-millisecond bursts with PCof varying pulse widths. Each wasapplied 3 seconds on and 3 secondsoff at an intensity that produced an-tigravity flexion of the biceps mus-cle. The 5-kHz stimulus was found tobe more comfortable. These earlystudies, thus, had 2 major findings:(1) that for a given level of forceproduction, burst-modulated AC ispreferable to continuous AC or PC,and (2) a short AC burst duration (1or 2 milliseconds) is optimal for leastdiscomfort.

A recent study13 measured MEIT andrelative discomfort using 50-Hzbursts of AC in the frequency rangeof 0.5 to 20 kHz. Burst durationsranging from the shortest possible (1cycle) to the longest (continuousAC) were used. Maximum torquewas produced at a frequency of 1kHz and a burst duration of 2 milli-seconds (10% duty cycle). Minimumdiscomfort occurred at a frequencyof 4 kHz and a burst duration of4 milliseconds (20% duty cycle).Continuous AC produced the leasttorque and the greatest discomfortat all frequencies. Single cycles (bi-phasic PC) produced significantlyless torque than 2-millisecond burstsand were more uncomfortable. Alater study22 compared Russian cur-rent (2.5-kHz AC applied in 10-millisecond bursts) and Aussiecurrent (1-kHz AC applied in4-millisecond bursts) with PC of thesame phase duration (200 and 500




microseconds, respectively) in termsof discomfort and torque produc-tion. The AC bursts (Fig. 1C) weremore comfortable than their PCcounterparts. Both Aussie currentand the 2 forms of PC produced sim-ilarly high torques, but, perhaps sur-prisingly, Russian current evokedless.

Thus, it seems reasonable to con-clude that a stimulus waveform thatconsists of kilohertz-frequency AC inshort-duration bursts (24 millisec-onds) is more comfortable and elicitsgreater MEIT than PC, continuousAC, Russian current, or interferentialcurrent stimulation.

The ConventionalWisdomHistorical Claims ConcerningInterferential CurrentsNemec2326 promoted the therapeu-tic use of interferential currents andadvocated the use of sinusoidal AC atfrequencies around 5 kHz. He arguedthat the 2 currents of slightly differ-ent frequency interfere in tissue,producing maximum stimulation inthe region of intersection of the 2current paths, and that the result-ing (endogenous) current at depthwould be modulated at the beatfrequency, which is the differencein frequency of the 2 currents(Fig. 3).1,7

Nemec2326 gave 3 arguments for theuse of interferential current ratherthan PC:

1. Skin impedance is lower at highAC frequencies; therefore, lesselectrical energy is dissipated inthe skin and, consequently, thereis less sensory stimulation and dis-comfort than with low-frequencyPC.

2. When the constant-intensity cur-rents intersect and interfere,the resulting current will be mod-ulated in intensity at the beat

frequency (the difference be-tween the 2 AC frequencies) andwill produce endogenous low-frequency stimulation (ie, atdepth, rather than superficially).

3. Currents interfere in tissue, pro-ducing maximum stimulation atthe region of intersection of the 2current paths, where a clover-like pattern of stimulation isproduced.1,7

The first point is incorrect for 2 rea-sons. First, the skin impedance to PCdepends on the phase duration, notthe pulse frequency.1,2,5,2729 Theskin impedance to low-frequency ACis much higher than to kilohertz-frequency AC because the phaseduration is much longer. If the PChas the same phase duration as thekilohertz-frequency AC, the skin im-pedance is the same even if the pulsefrequency is low.1,2,5,2729 Conven-tional PC typically has a phase dura-tion similar to that of interferentialcurrent. Thus, the argument that in-terferential current would meet witha lower impedance is without anybasis. Second, a lower skin imped-ance does not mean less stimulationof sensory and pain fibers in theskin and, therefore, less discomfort.The high skin impedance with longphase durations (eg, with low-frequency AC) is due to the skin ca-pacitance, which is due almost en-tirely to the stratum corneum: thedead, scaly, relatively dry, outermostlayer.1,2,5,27,28 The stratum corneumhas no sensory, pain, or other kind ofnerve fibers.30,31 These fibers are lo-cated beneath, in the dermis, whichis well hydrated and of similar con-ductivity to the deeper tissues.5,30,31

The second and third points areoversimplifications. There are 3 im-portant things to consider with inter-ferential stimulation:

1. An interference pattern of stimu-lation is produced everywhere,

not just at the predicted region ofintersection of the currents, andthe extent of modulation of theresulting current will depend onthe location and orientation ofthe nerve fibers relative to theelectrodes.1,2,14,15 This means thatthroughout the tissue volume, fi-bers orientated at an optimum an-gle will experience a fully modu-lated current, whereas those atother angles (the majority) will besubject to a partially modulated orunmodulated stimulus.1,2,14,15

2. Current spreading means thatthere will not be a region at thecenter of intersection of the cur-rents where maximum stimula-tion occurs. Although the stimu-lation at depth might be expectedto be greater, current spread-ing would be expected to signifi-cantly reduce the value of any re-inforcement effect.1,14,15

3. It might be expected that the cur-rent intensity at depth would begreater with quadripolar stimula-tion than with bipolar stimulationbecause of interference and rein-forcement. Lambert et al15 dem-onstrated that this is not true.When currents are applied usingconventional interferential stimu-lation, the pattern of stimulationis not focused centrally. It is morediffuse due to current flow be-tween adjacent electrodes be-cause of the shorter-distance,lower-resistance pathways.

Thus, the depth efficiency claims forinterferential current are not sub-stantiated. This, together with theuncertain degree of modulation ofthe stimulus, calls into questionwhether the interference effect ofinterferential current is of any value.Ozcan et al32 addressed this questionwhen they assessed the relative dis-comfort of true and premodulatedinterferential currents (delivered in50-Hz bursts, 10 milliseconds on and




10 milliseconds off). Premodulatedinterferential current was found tobe significantly more comfortablethan true interferential current andmore effective for muscle contraction.

Historical Claims ConcerningRussian CurrentA talk given by Kots,33 of the CentralInstitute of Physical Culture, Mos-cow, at a conference hosted by Con-cordia University, Montreal, in 1977laid the foundation for what becameknown in the Western world as Rus-sian current electrical stimulation.9

Kots reported strength gains of up to40% in elite Russian athletes stimu-lated with 2.5-kHz AC applied in 10-millisecond rectangular bursts at afrequency of 50 Hz. His protocolused currents with a 10-second onperiod followed by a 50-second restperiod, applied 10 times in eachstimulation session (ie, 10-minutetreatment sessions). Treatment wasapplied daily over a period of weeks.

As noted previously, Russian cur-rents became popular despite anequivocal evidence base due to thelimited number of studies and theirdifferent findings.3,9 The choice, byKots group, of 10-millisecond bursts(50% duty cycle) was because oftheir observation that it evoked justas much muscle torque as continu-ous AC but, because of the burstmodulation, the average current ap-plied to tissue was halved. The effectof different burst durations was notexplored. Bankov19,20 and Bankovand Daskalov,21 in the 1980s, exam-ined the effect of burst duration andfound that, for the same level offorce production, short-durationbursts are more comfortable. An in-ference is that greater levels offorce would be produced at thesame level of discomfort if short-duration bursts were used. This issupported by the recent work ofWard et al,13 who measured torqueat the pain tolerance limit and foundthat the greatest MEIT is produced

using 2-millisecond bursts of ACwith a frequency of 1 kHz.

Thus, the rationale for the clinicaluse of Russian current is called intoquestion. The evidence is that stim-ulation with short-duration bursts ofkilohertz-frequency AC would bepreferable and that a burst durationof 2 milliseconds appears to be opti-mal for torque production.

DiscussionThe KnownElectrophysiologyThe available laboratory-based evi-dence indicates that short-durationbursts of kilohertz-frequency AChave advantages over Russian cur-rent, interferential current, and PCand that there are optimal frequen-cies and burst durations for achiev-ing the desired outcome. There are 4interrelated electrophysiological fac-tors that could help explain the em-pirical findings: summation, multiplefiring, high-frequency fatigue, andneural block.

SummationWith kilohertz-frequency AC stimu-lation, there is the possibility ofsummation, a phenomenon first de-scribed by Gildemeister.34,35 Gilde-meister reported that when bursts ofkilohertz-frequency AC are appliedtranscutaneously, the threshold volt-age for sensory nerve excitation de-creases as the burst duration is in-creased. This phenomenon, latercalled the Gildemeister Effect, oc-curs because, with each successivepulse in the AC wave-train, the nervefiber membrane is pushed closer tothreshold. Membrane threshold isreached when successive pulses re-sult in sufficient depolarization toproduce an action potential. Gilde-meister observed a limit to the sum-mation effect. As the number of cy-cles per burst was increased, thethreshold decreased, but only up to apoint. Beyond a certain burst dura-tion, no further decrease in thresh-old was observed. He called this

maximum burst duration (ie, timeover which pulses could summate)the Nutzzeit or utilization time.

As noted previously, a recent studyby Ward and Lucas-Toumbourou18

showed that the apparent utilizationtime was different for sensory, mo-tor, and pain thresholds and, con-sequently, that relative thresholds(pain/motor and pain/sensory) varywith burst duration. These authorsfound that optimum discrimination(biggest separation between thresh-olds [ie, maximum relative thresh-olds]) occurred at burst durations of2 to 4 milliseconds.

High-Frequency FatigueWhen electrical stimulation is ap-plied to elicit a motor response usingPC frequencies higher than physio-logical or at the high end of the phys-iological range (ie, greater thanabout 50 Hz), it is possible to pro-duce a blockage of muscle activitydue to propagation failure or neuro-transmitter depletion.3638 This is re-sponsible for the phenomenon ofhigh-frequency fatigue,38,39 whichis characterized by its associatedrapid recovery. If a stimulus fre-quency of 80 Hz, for instance, is usedto elicit muscle contraction, the re-sulting muscle force declines rap-idly, but if a brief rest period (a fewseconds) is allowed, marked recov-ery occurs.38,39 This is quite differentfrom low-frequency fatigue, whichis much more akin to normal physi-ological fatigue, where the force de-cline is much slower and the recov-ery time is much longer.

One form of high-frequency fatigue,propagation failure, can occur whenaction potentials are induced in mo-toneurons at sufficiently high fre-quency. This can result in action po-tential failure at branch points wherea motor nerve divides to innervateindividual muscle fibers. Failure alsocan occur at the neuromuscularjunction because neurotransmitter




depletion is possible at relativelyhigh stimulation frequencies.40 Be-yond the neuromuscular junction,transmission failure can occur atthe level of the t-tubule system. Nor-mally, the wave of depolarizationof a muscle fiber action potential istransmitted over the muscle fibermembrane and throughout thet-tubule system, activating the con-tractile elements. When sufficientlyhigh frequency action potentials areinduced, the t-tubule membranes donot have time to recover betweenaction potentials and muscle fibercontraction ceases.39,40 Whicheverthe mechanism, whether propaga-tion failure or neurotransmitter de-pletion, a blockage of muscle con-traction at stimulation frequenciesaround and above about 50 Hz is theresult, and the effect is described ashigh-frequency fatigue.

Summation and Multiple FiringWhen the stimulus is PC applied atlow frequency (less than 100 Hz), itcan be confidently concluded that,provided that the pulse intensityis sufficiently above threshold, thenerve fiber firing frequency willequal the pulse frequency. The firingfrequency could be less if successivepulses occur within the relative re-fractory period and the stimulus in-tensity is not sufficiently high, butthe firing frequency could never behigher than the PC frequency. Withbursts of AC, however, there is thepossibility that a single burst will re-sult in multiple action potentials as aresult of summation4144; therefore,the firing frequency could be somemultiple of the burst frequency. Ifthe first few pulses in a burst sum-mate, the nerve fiber could fire, gothrough a brief period of refractori-ness, and then fire again. If this pro-cess happens rapidly and, therefore,is repeated during the burst, thenerve fiber firing frequency will be amultiple of the burst frequency.There is sound experimental evi-dence for this effect.4145

A problem with multiple firing is thatit could detract from the desired out-come. For example, a motoneuronfiring frequency of 50 Hz might elicitan optimally forceful muscle contrac-tion, so 50-Hz PC would be a goodoption. If long-duration 50-Hz burstsare used, however, the induced fir-ing frequency could be a multiple of50 Hz. This would initially result in aslightly greater muscle force, butthe rate of fatigue would be higher.There also would be a greateramount of high-frequency fatigue. Arecent study by Laufer and Elboim44

compared fatigue rates using 50-Hzbursts of 2.5-kHz AC with a burstduration of 10 milliseconds (Russiancurrent), 50-Hz biphasic PC with thesame phase duration (200 microsec-onds), 50-Hz bursts with a burst du-ration of 4 milliseconds, and 20-Hzbursts with a burst duration of 10milliseconds. They reported thatRussian current was the most rapidlyfatiguing, PC was the least rapidlyfatiguing, and the 2 currents ofshorter burst duration were interme-diate and equally fatiguing. A conclu-sion is that for motor stimulation us-ing kilohertz-frequency AC bursts, ifthe duration is greater than 2 milli-seconds, multiple firing is likely tooccur and the fatigue rate will becompromised.

Neural BlockWith kilohertz-frequency AC stimula-tion, another effect can be pro-duced: direct conduction block ofthe nerve fiber. A direct observationof neural block was reported byTanner,46 who measured compoundaction potentials produced in ex-posed sciatic nerve in response todirect, repetitive stimuli from a low-frequency pulse generator and foundthat neural activity could be blockedusing a 20-kHz AC stimulus appliedto the nerve trunk between the pulsegenerator and the recording elec-trodes. As the AC stimulus intensitywas progressively increased, firstthe fast (large-diameter) fiber re-

sponses disappeared, followed bythe slower (intermediate-diameter)fiber responses and then the slowest(small-diameter) fiber responses.

Bowman and McNeal45 examinedthe -motoneuron response toblocking signals in the frequencyrange of 100 Hz to 10 kHz. Withhigh-intensity 2-kHz AC stimulation,they observed that following a briefperiod of firing at a very high rate(about 1 kHz), there was a progres-sive decrease in firing frequency,which occurred over a time frame oftens of seconds, after which activityceased and complete conductionblock occurred. At higher AC fre-quencies (4 kHz or more), the rate ofdecrease in activity was higher, withthe firing frequency dropping tozero in less than a second and withstimulus intensities of 5 times thethreshold. Bowman and McNeal con-cluded that neural block occursmore readily at multiples of thresh-old stimulation intensities and thatthe effects occur more rapidly athigher kilohertz frequencies.

Direct studies of neural block withAC stimulation, to date, have all usedcontinuous AC. There do not appearto be any reported studies of theblocking effectiveness of burst-modulated AC, so it is not known towhat extent neural block contrib-utes to the effects observed. Indirectevidence for neural block was foundby Ward and Robertson,17 who mea-sured motor thresholds using contin-uous kilohertz-frequency AC, 50-Hzbursts, and single sine waves in therange of 1 to 25 kHz. Irregularities inthe graphs of force versus stimulusintensity were consistent with multi-ple firing followed by nerve block.The effects were more pronouncedat higher kHz frequencies and weregreater with continuous stimulationthan with 50-Hz bursts.

Whether neural block is of practicalsignificance with electrical stimula-




tion as used clinically thus remainsuncertain, but it would affect MEIT,as -motoneurons are more suscep-tible to neural block than pain (A-and C) fibers because of their largerdiameter. This means that muscleforce could be diminished withoutany diminution of pain sensation.

ConclusionIn assessing the relative merits ofdifferent forms of motor electricalstimulation, 2 factors are highlyrelevant: relative discomfort ofstimulation and the ability to elicitmaximum muscle torque. Thesefactors, in turn, depend on the neu-rophysiological responses of differ-ent nerve fiber types to electricalstimulation.

With kilohertz-frequency AC stimula-tion, summation and multiple firing,high-frequency fatigue, and neuralblock can potentially affect the neu-rophysiological response. The ef-fects will vary, depending on the ACfrequency and burst duration.

Both the historical evidence andmore recent findings indicate thatthe stimulation parameters com-monly used clinically (Russian andinterferential currents) are subopti-mal for achieving their stated goalsand that greater benefit would beobtained using short-duration (2- to4-millisecond) bursts of kilohertz-frequency AC, with a frequency cho-sen to maximize the desired out-come. For maximum muscle torqueproduction, a frequency of 1 to 2.5kHz is indicated, with a burst dura-tion of 2 milliseconds or so. For min-imal discomfort (but less muscletorque), a frequency of 4 kHz is in-dicated, with a burst duration of 4milliseconds.

This article was received February 24, 2008,and was accepted November 8, 2008.

DOI: 10.2522/ptj.20080060

References1 Robertson VJ, Ward AR, Low J, Reed A.

Electrotherapy Explained: Principles andPractice. 4th ed. Oxford, United Kingdom:Butterworth Heinemann; 2006.

2 Kloth LC. Interference current. In: NelsonRM, Currier DP, eds. Clinical Electrother-apy. 2nd ed. East Norwalk, CT: Appleton& Lange; 1991:221260.

3 Selkowitz DM. High-frequency electricalstimulation in muscle strengthening. Am JSports Med. 1989;17:103111.

4 Geddes LA. A short history of the electricalstimulation of excitable tissue includingtherapeutic applications. Physiologist.1984;27(suppl):s1s47.

5 Reilly JP. Electrical Stimulation and Elec-tropathology. Cambridge, United King-dom: Cambridge University Press; 1992.

6 Electrotherapeutic Terminology in Phys-ical Therapy. Alexandria, VA: AmericanPhysical Therapy Association; 2000.

7 Palmer S, Martin D. Interferential current.In: Watson T, ed. Electrotherapy:Evidence-Based Practice. 12th ed. Lon-don, United Kingdom: Churchill Living-stone; 2008:297315.

8 Andrianova GG, Kots YM, Marmyanov VA,Xvilon VA. Primenenie elektrostimuliatsiidlia trenirovki mishechnoj sili. NovostiMeditsinskogo Priborostroeniia. 1971;3:4047.

9 Ward AR, Shkuratova N. Russian electricalstimulation: the early experiments. PhysTher. 2002;82:10191030.

10 Delitto A, Brown M, Strube MJ, et al. Elec-trical stimulation of quadriceps femoris inan elite weight lifter: a single-subject ex-periment. Int J Sports Med. 1989;10:187191.

11 St Pierre D, Taylor AW, Lavoie M, et al.Effects of 2,500-Hz sinusoidal current onfibre area and strength of the quadricepsfemoris. J Sports Med. 1986;26:6066.

12 Ward AR, Robertson VJ. The variation intorque production with frequency usingmedium-frequency alternating current.Arch Phys Med Rehabil. 1998;79:13991404.

13 Ward AR, Robertson VJ, Ioannou H. Theeffect of duty cycle and frequency on mus-cle torque production using kHz fre-quency range alternating current. MedEng Phys. 2004;26:569579.

14 Treffene RJ. Interferential fields in a fluidmedium. Aust J Physiother. 1983;29:209216.

15 Lambert HL, Vanderstraeten GG, DeCuyper HJ, et al. Electric current distribu-tion during interferential therapy. Eur JPhys Med Rehabil. 1993;3:610.

16 Soloviev NA. Nyekogorii osobynyenostiielektrostimulatsii na povishchennik chas-totak, 1: optimalnii chastoti elektrostimu-latsii i primenenie chastotnoy i ampli-tudnoy modulatsii. Trudy Instytut MVniimio. 1963;3:162170.

17 Ward AR, Robertson VJ. The variation inmotor threshold with frequency usingkHz-frequency alternating current. MuscleNerve. 2001;24:13031311.

18 Ward AR, Lucas-Toumbourou S. Loweringof sensory, motor and pain-tolerancethresholds with burst duration using kHz-frequency alternating current electricalstimulation. Arch Phys Med Rehabil.2007;88:10361041.

19 Bankov S. Medium frequency modulatedimpulse current for electric stimulation ofnon-denervated muscles. Acta MedicaBulgarica. 1980;7:1217.

20 Bankov S. Prouchvaniya vrku elektro-stimulyatsiita na nedenervirani muskuli sssinusoidalni modulirani tokovy s razlichniparametri. Kurortologiia i FizioterapiiaSofia. 1980;1:2833.

21 Bankov S, Daskalov I. Electrostimulationde muscles non denerves par la courantdynatron et un courant dimpulsions afrequence moyenne modulees. RevueElectrodiagnostic Therapie. 1981;18:2332.

22 Ward AR, Oliver W, Buccella D. Wrist ex-tensor torque production and discomfortassociated with low-frequency and burst-modulated kilohertz-frequency currents.Phys Ther. 2006;86:13601367.

23 Nemec H. Interferential therapy: a newapproach in physical medicine. Br J Phys-iother. 1959;12:912.

24 Nemec H. Reizstromtherapie mit interfer-enzstromen. Der Deutsche Badebetrieb.1960;12:320322.

25 Nemec H. Endogene elektrostimulierungdurch mittelfrequente und interferenz-strome. Rehabilitation (Bonn). 1967;20:111.

26 Nemec H. Elektrostimulierung in endoge-ner anwendung: aktionsmechanismus derinterferenztherapie. Physikalische Medi-zin und Rehabilitation. 1968;9:7375.

27 Burton CE, David RM, Portnoy WM, AkersLA. The application of Bode analysis toskin impedance. Psychophysiology. 1974;11:517525.

28 Lykken DT. Square-wave analysis of skinimpedance. Psychophysiology. 1971;7:262275.

29 Yamamoto T, Yamamoto Y. Analysis forthe change of skin impedance. Med BiolEng Comput. 1977;15:219227.

30 Saladin KS. Anatomy and Physiology.Boston, MA: McGraw Hill; 1998:202206.

31 Grimnes S. Electrovibration, cutaneoussensation of microampere currents. ActaPhysiol Scand. 1983;118:1925.

32 Ozcan J, Ward AR, Robertson VJ. A com-parison of true and premodulated interfer-ential currents. Arch Phys Med Rehabil.2004;85:409415.

33 Babkin D, Timtsenko N, trans. Electro-stimulation (notes from Dr YM Kots[USSR] lectures and laboratory periodspresented at the Canadian-Soviet ex-change symposium on electrostimulationof skeletal muscles, Concordia University,Montreal, Quebec, Canada, December615, 1977). Available from Dr Ward.

34 Gildemeister M. Zur theorie des elek-trischen reizes, V: polarisation durchwechselstrome. Ber Sachs Ges Wiss.1930;81:303313.




35 Gildemeister M. Untersuchungen uber diewirkung der mittelfrequenzstrome aufden menschen. Pflugers Arch. 1944;247:366404.

36 Solomonow M, Eldred E, Lyman J, Foster J.Fatigue considerations of muscle contrac-tile force during high-frequency stimula-tion. Am J Phys Med. 1983;62:117122.

37 Sawan M, Hassouna MM, Li JS, et al. Stim-ulator design and subsequent stimulationparameter optimization for controllingmicturition and reducing urethral resis-tance. IEEE Trans Rehabil Eng. 1996;4:3946.

38 Jones DA. High- and low-frequency fatiguerevisited. Acta Physiol Scand. 1996;156:265270.

39 Jones DA. Muscle fatigue due to changesbeyond the neuromuscular junction. In:Porter R, Whelan J, eds. Human MuscleFatigue: Physiological Mechanisms. Lon-don, United Kingdom: Pitman Medical;1981:178196.

40 Otsuka M, Endo M, Nonomura Y. Presyn-aptic nature of neuromuscular depression.Jpn J Physiol. 1962;12:573584.

41 Stefanovska A, Vodovnik L. Change inmuscle force following electrical stimula-tion. Scand J Rehabil Med. 1985;17:141146.

42 Laufer Y, Ries JD, Leininger PM, Alon G.Quadriceps femoris muscle torques pro-duced and fatigue generated by neuromus-cular electrical stimulation with three dif-ferent waveforms. Phys Ther. 2001;81:13071316.

43 Ward AR, Robertson VJ. The variation infatigue rate with frequency using kHz fre-quency alternating current. Med Eng Phys.2000;22:637646.

44 Laufer Y, Elboim M. Effect of burst fre-quency and duration of kilohertz-frequency alternating currents and oflow-frequency pulsed currents on strengthof contraction, muscle fatigue, and per-ceived discomfort. Phys Ther. 2008;88:11671176.

45 Bowman BR, McNeal DR. Response of sin-gle alpha motoneurons to high-frequencypulse trains. Appl Neurophysiol. 1986;49:121138.

46 Tanner JA. Reversible blocking of nerveconduction by alternating-current excita-tion. Nature. 1962;195:712713.




doi: 10.2522/ptj.20080060Originally published online December 18, 2008

2009; 89:181-190.PHYS THER. Alex R WardAlternating CurrentElectrical Stimulation Using Kilohertz-Frequency

References

http://ptjournal.apta.org/content/89/2/181#BIBLfor free at: This article cites 38 articles, 5 of which you can access

Cited by

http://ptjournal.apta.org/content/89/2/181#otherarticles

This article has been cited by 2 HighWire-hosted articles:

Information Subscription http://ptjournal.apta.org/subscriptions/

Permissions and Reprints http://ptjournal.apta.org/site/misc/terms.xhtml

Information for Authors http://ptjournal.apta.org/site/misc/ifora.xhtml

by guest on February 15, 2015http://ptjournal.apta.org/Downloaded from

http://ptjournal.apta.org/content/89/2/181#BIBLhttp://ptjournal.apta.org/content/89/2/181#otherarticleshttp://ptjournal.apta.org/subscriptions/http://ptjournal.apta.org/site/misc/terms.xhtmlhttp://ptjournal.apta.org/site/misc/ifora.xhtmlhttp://ptjournal.apta.org/

Electrical Stimulatio

Documents

frequency acwith

orgcgico stimulation

online version

pulsed current pc

response of nerve

faculty of health sciences

means of muscle strengthening

trobe university