Advertisement

A comparison of true and premodulated interferential currents 1

      Abstract

      Ozcan J, Ward AR, Robertson VJ. A comparison of true and premodulated interferential currents. Arch Phys Med Rehabil 2004;85:409–15.

      Objective

      To compare true and premodulated interferential currents (IFCs) in terms of sensory, motor, and pain thresholds; maximum electrically induced torque (MEIT); and comfort.

      Design

      Repeated-measures design.

      Setting

      Laboratory setting.

      Participants

      University student and staff volunteers.

      Interventions

      Participants were exposed to 4 different conditions, chosen to evaluate 2 fundamental differences between true and premodulated IFCs. The conditions were different combinations of (1) premodulated or constant-amplitude currents applied at the skin and (2) crossed or parallel current paths.

      Main outcome measures

      Sensory, motor, and pain thresholds; MEIT; and subjective reports of relative discomfort were recorded for each of the 4 conditions. Motor to sensory threshold ratios were subsequently calculated to assess depth efficiency of stimulation.

      Results

      The major findings were that crossed currents (true IFC) had no advantage over parallel currents (premodulated IFC) in terms of motor to sensory threshold ratio, MEIT, or comfort, and that premodulated currents produced higher torque values and less discomfort than constant-amplitude currents (true IFC). These results contradict the claimed superiority of true IFC.

      Conclusions

      The findings indicate that premodulated IFC, delivered via 2 large electrodes, may be clinically more effective than the traditional true IFC arrangement in terms of depth efficiency, torque production, and patient comfort.

      Keywords

      ELECTRIC STIMULATION IS used extensively throughout the world. Uses include augmenting muscle strength and endurance, controlling pain, promoting tissue healing, managing spasticity, and providing assistance as an orthosis.
      • Low J.
      • Reed A.
      ,
      • Selkowitz D.M.
      Electrical currents.
      The focus of our study is on 1 type of electric stimulation—interferential therapy (IFT)—which uses medium-frequency alternating current (MFAC) for patient treatment. IFT is widely used by physiotherapists in Australia,
      • Lindsay D.
      • Dearness J.
      • Richardson C.
      • Chapman A.
      • Cuskelly G.
      A survey of electromodality usage in private physiotherapy practices.
      ,
      • Robertson V.J.
      • Spurritt D.
      Electrophysical agents implications of their availability and use in undergraduate clinical placements.
      Canada,
      • Lindsay D.
      • Dearness J.
      • McGinley C.
      Electrotherapy usage trends in private physiotherapy practice in Alberta.
      and England
      • Pope G.D.
      • Mockett S.P.
      • Wright J.P.
      A survey of electrotherapeutic modalities ownership and use in the NHS in England.
      to treat a range of clinical conditions, even though there is little evidence for claims of its advantage over other forms of electric stimulation. Decisions to use IFT rather than other forms of transcutaneous stimulation appear to be based on anecdotal evidence and tradition, not on research evidence. Only recently have some of the claims for IFT been called into question.
      • Alon G.
      Principles of electrical stimulation.
      ,
      • Martin D.
      Interferential therapy.
      ,
      • Johnson M.I.
      • Tabasam G.
      A double-blind placebo controlled investigation into the analgesic effects of interferential currents (IFC) and transcutaneous electrical nerve stimulation (TENS) on cold-induced pain in healthy subjects.
      Interferential currents (IFCs) are produced by crossing 2 MFACs (“carrier” currents) of slightly different frequency, so that they interfere with one another and generate an amplitude-modulated, low-frequency (resultant) current. IFT was first described in the English-language literature by its inventor, Hans Nemec.
      • Nemec H.
      Interferential therapy a new approach in physical medicine.
      Nemec’s article includes statements of 2 fundamental aspects of IFT, namely, that (1) by using MFAC, skin impedance is minimized, and (2) if 2 MFACs of different frequency are applied to the body simultaneously, they will combine and produce a low-frequency “beating” effect.
      Both statements are supportable. That skin impedance is minimized at kilohertz frequencies is well documented.
      • Martin D.
      Interferential therapy.
      ,
      • Reilly J.P.
      The dry, outer layer of the skin, the stratum corneum, acts as an insulator. The skin thus acts as a capacitative barrier to the flow of current and, as with any capacitor, its impedance is inversely proportional to the alternating current frequency. Therefore, at low frequencies, the skin impedance is high, whereas at high frequencies, the skin impedance is lower.
      • Low J.
      • Reed A.
      ,
      • Martin D.
      Interferential therapy.
      The second statement—if 2 different MFACs are applied to the body simultaneously, they will combine and produce a low-frequency beating effect—is well documented.
      • Low J.
      • Reed A.
      ,
      • Selkowitz D.M.
      Electrical currents.
      ,
      • Martin D.
      Interferential therapy.
      ,
      • De Domenico G.
      ,
      • Goats G.C.
      Interferential current therapy.
      ,
      • Kloth L.C.
      Interference current.
      A difference in frequency will result in the 2 currents shifting in and out of phase (fig 1A). The resultant stimulus intensity systematically varies from a maximum, when 2 peaks or troughs coincide (constructive interference), to zero amplitude, when the currents cancel each other out (destructive interference). The rate of variation in amplitude, the beat frequency, is equal to the difference between the 2 MFAC frequencies.
      Figure thumbnail GR1
      Fig 1Resultant current produced by (A) interference of 2 sinusoidal currents of different frequencies and (B) interference of 2 rectangular pulsed currents shifted in and out of phase.
      When sinewaves of different frequency are used, the combined current has a sinusoidal wave envelope (fig 1A). This is because the currents drift smoothly in and out of phase. Early IFT machines produced sinewave currents of different frequency. Some modern machines produce a rectangular alternating current waveform, which is phase shifted to produce bursts of current at the beat frequency. The abrupt switching of the currents in and out of phase produces a rectangular wave envelope (fig 1B). Whether phase-switching or different alternating current frequencies are used, the essential features of Nemec’s original design are preserved. The carrier currents are kilohertz-frequency alternating current of constant amplitude, and low-frequency bursts (rectangular or sinusoidal) are generated when the currents interfere (fig 1).
      IFT machines use either 2 or 4 electrodes on the skin. In a 4-electrode arrangement, known as true IFC (also quadripolar or endogenous IFC), the constant amplitude MFAC carrier currents are applied at the skin surface via 2 isolated circuits, and the currents cross and interfere within the tissues (fig 2A). With a 2-electrode arrangement, an amplitude-modulated current is generated by the machine, and the output is described as premodulated IFC (also bipolar or exogenous IFC). Only 1 circuit is used, so there is no crossing of current paths at depth (fig 2B).
      Figure thumbnail GR2
      Fig 2Regions of maximum stimulation (shaded), which are predicted with application of (A) true and (B) premodulated IFCs.
      Some sources claim that true IFC gives better depth efficiency with less sensory stimulation and is more comfortable than premodulated IFC.
      • Low J.
      • Reed A.
      ,
      • Goats G.C.
      Interferential current therapy.
      ,
      • Kloth L.C.
      Interference current.
      ,
      • Savage B.
      True IFCs are mixed within the tissues, and the resultant current is the sum of the 2 different frequency currents (fig 1). Therefore, the amplitude of the resultant (beating) current in tissue is higher than either individual current applied to the skin. With premodulated IFC, because current modulation occurs outside the tissues, no interference occurs with the tissues, and the amplitude of the current is simply that applied to the skin.
      Shaded regions in figure 2 show that maximal stimulation is expected to occur superficially with premodulated IFC and more deeply with true IFC.
      • Low J.
      • Reed A.
      ,
      • Kloth L.C.
      Interference current.
      With premodulated IFC, there is no interference or reinforcement of currents within the tissues, and because of current spread in deeper tissues, maximal stimulation is believed to occur superficially, close to the electrodes.
      • Goats G.C.
      Interferential current therapy.
      However, with true IFC, maximal stimulation occurs where the applied currents cross paths and interfere, and this is theorized to occur maximally deep within the tissues, in and around the geometric center between the 4 electrodes.
      • Kloth L.C.
      Interference current.
      Because tissues are not homogenous with regard to electric conductivity, the real patterns of maximal stimulation are likely to be more irregular,
      • Low J.
      • Reed A.
      ,
      • Savage B.
      ,
      • Lambert H.L.
      • Vanderstraeten G.G.
      • De Cuyper H.J.
      • et al.
      Electric current distribution during interferential therapy.
      but one would still expect greater depth efficiency with true IFC.
      The laws of nerve fiber excitation provide a means to establish or estimate depth efficiency. It is generally observed that nerve fibers are recruited in the order of sensory, motor, and finally pain fibers, as the intensity of an applied stimulus is increased. This is because 2 important factors that determine whether nerve fibers are excited (or not) are fiber diameter and distance of the fibers from the stimulating electrode.
      • Knaflitz M.
      • Merlitti R.
      • DeLuca C.J.
      Inference of motor unit recruitment order in voluntary and electrically elicited contractions.
      ,
      • Binder-Macleod S.A.
      • Halden E.E.
      • Jungles K.A.
      Effect of stimulation intensity on the physiological responses of human motor units.
      The stimulus intensity needed to stimulate a nerve fiber is inversely proportional to its diameter. Hence, the larger diameter fibers are intrinsically more sensitive to electric stimulation. The stimulus intensity required to excite a nerve fiber is also proportional to its distance from the stimulating electrode. The greater the distance, the higher the stimulus intensity required. It is because of this distance effect that, as the stimulus intensity is increased, the sensory threshold is normally reached before the motor threshold. Although the largest α-motoneurons have greater diameters than the largest sensory fibers, the more superficial location of many sensory fibers
      • Low J.
      • Reed A.
      results in their earlier recruitment.
      Several measurable outcomes provide a quantitative indication of depth efficiency. When stimulation is more depth-efficient, such as in figure 2B, it would be expected that α-motoneurons would be more readily recruited, so the motor threshold would decrease relative to the sensory threshold. The motor to sensory threshold ratio would consequently be smaller. Other evidence of greater depth efficiency would include a higher maximum electrically induced torque (MEIT) and more comfortable stimulation.
      The aim of our study was to compare true and premodulated IFC and to establish whether differences in motor to sensory threshold ratio, MEIT, and comfort exist. The 4 conditions compared were chosen to distinguish the effects of current type applied to the electrodes (constant amplitude or premodulated) and current path within tissue (crossed or parallel).

      Methods

      Approval for the study was obtained from the Ethics Committee of the Faculty of Health Sciences of LaTrobe University, Victoria, Australia, before data collection commenced.
      The 12 subjects (7 men, 5 women) participating in this study were volunteers who met the inclusion criteria. That is, they had no orthopedic or neurologic pathology or impairment of their right lower limb and no breaks or irritation of the skin under the electrodes. Another criterion was that subjects have experience with electric stimulation. This ensured that they had an established tolerance to electric stimulation and precluded any need for lengthy familiarization trials. Subjects were recruited from the students and academic staff members of the School of Physiotherapy. Subjects’ ages ranged from 21 to 51 years, with a mean age ± standard deviation (SD) of 26.3±8.8 years. Each subject was tested on 2 occasions, on the same day, with a minimum of 4 hours between test sessions.
      After the procedure to be used was explained to each subject, the subject was seated in a KinCom dynamometer,
      Chattecx Corp, Chattanooga Group, 4717 Adams Rd, Hixson, TN 37343.
      with the right thigh fully supported by the seat and Velcro straps secured across the pelvis and right thigh. The pivot of the KinCom arm was aligned with the subject’s lateral femoral epicondyle. The tibial pad was placed just above the talocrural joint line and was secured firmly with a Velcro strap. The knee was positioned in 45° flexion during testing and was measured by using a conventional long-arm goniometer.
      Four conductive rubber electrodes
      Stimtrode 50×90mm rectangular self-adhering electrodes; Axelgaard Manufacturing Co, 329 W Aviation Rd, Fallbrook, CA 92028.
      (50×90mm) were attached to the skin of the right lower limb. The skin was first prepared by wiping with an alcohol swab, and the line from the superomedial patella to the anterior superior iliac spine was measured with a tape measure. A point 40% of this distance from the patella was marked on the skin. The 4 electrodes were placed equidistantly from this point, running parallel down the thigh. The distance between each medial and lateral pair of electrodes was approximately the width of the subject’s patella. This electrode placement was based on pilot testing, which indicated that it consistently produced a strong, comfortable quadriceps contraction. The electrodes were removed at the end of the first session and were used again for the same subject in the second testing session. The corners of each electrode placement were outlined on the skin with indelible ink, to ensure consistent placement across test sessions.
      An IFT unit,
      Metron Vectorsurge III; Metron Medical Australia Pty Ltd, 57 Aster Rd, Carrum Downs, Victoria 3201, Australia.
      with a constant voltage output, which produced a phase-shifted rectangular pulsed current at a frequency of 4kHz, was used. The beat frequency was set to 50Hz for all test conditions. Stimulus intensity (in volts) was measured by using an oscilloscope,
      Trio CS-1560A II, 15MHz oscilloscope; Jaycar Electronics Pty Ltd, 100 Silverwater Rd, Silverwater, NSW 2128, Australia.
      which was connected directly to the skin-mounted electrodes.
      The 4 test conditions (fig 3) enabled comparisons of the effects of current type (constant amplitude vs premodulated) and current path (crossed vs parallel). Current type was set by using the 2/4-pole switch on the IFT machine. Set to 4-pole, constant-amplitude currents were delivered (conditions A and C), whereas set to 2-pole, the output current was premodulated (conditions B and D). Current path was adjusted by the lead arrangement of the IFT machine. For a crossed-current path (conditions A and B), both outputs of the IFT machine were used, and diagonally opposite electrodes were connected to the same output. Parallel current paths (conditions C and D), however, used only 1 output, with the proximal 2 electrodes connected to 1 socket and the distal 2 electrodes to the other. Although some crossing of current paths would inevitably occur with conditions C and D (due to current spread), the single output ensured no beating effect would be produced in tissue, which ensured that the flow of current was analogous to premodulated IFC.
      Figure thumbnail GR3
      Fig 3The 4 test conditions used in the study. Condition A: constant-amplitude currents, paths crossed. Condition B: premodulated currents, paths crossed. Condition C: constant-amplitude currents, paths parallel. Condition D: premodulated currents, paths parallel.
      Each subject was exposed to the 4 test conditions in a randomly allocated order. The 24 possible orders of the 4 test conditions were assigned to a test session (12 subjects×2 sessions) via a ballot. This meant that every possible test order was used only once, and each test session had a unique order of conditions. It also meant that any variation in current flow, because of changes in skin impedance during the course of the measurements, would not be a confounding variable. First, sensory thresholds were measured for the 4 conditions. Subjects were asked to increase the stimulus intensity until they could just feel an electrically induced sensation. The stimulus voltage, displayed on the oscilloscope, was recorded. Next, motor thresholds were measured. The motor threshold was defined as the lowest stimulus intensity required to produce a just-visible muscle contraction. Finally, MEIT and the associated stimulus intensity (pain threshold) were measured. Subjects were asked to gradually and continually increase the intensity to the maximum tolerable level that they believed they could sustain for the period of time the researcher required to record the voltage and force readings (≈3s). It was emphasized to the subjects that they should remain relaxed and avoid any voluntary contractions of their right quadriceps or hamstrings muscle groups. Two familiarization trials were conducted to help subjects accurately identify their tolerance limit. The 4 conditions were then tested in the predetermined order, with a brief rest between trials. The 4 test conditions were then tested in reverse order (to reduce any order effects) and then again in the original order.
      Subject comments regarding each of the test conditions were solicited before MEIT/pain threshold testing. The subject was asked to tell the experimenter if any test condition felt noticeably more uncomfortable than the others. All comments each subject made about any test conditions were recorded by the experimenter.

      Data analysis

      Data sets (sensory, motor, and pain threshold and MEIT) were analyzed statistically by using SPSS, version 10.
      SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.
      Preliminary tests of normality using Q-Q plots showed that the data sets were normally distributed. A 2-factor analysis of variance (ANOVA) for repeated measures was performed. An α level of .05 was used for the ANOVA tests. Where the Mauchly test for sphericity indicated that the assumption of sphericity had been violated, the degrees of freedom were adjusted by using the Huynh-Feldt ε correction factor. The ANOVAs identified no significant between-sessions effect; therefore, threshold and MEIT data from the 2 test sessions were averaged for each subject, giving 1 value (per subject) for each test condition. The averaged data were used in the calculation of means and SDs for each test condition (as presented in the Results). The use of averaged, rather than pooled, data, for post hoc comparisons ensured that calculated P values were not biased toward significance by an inflated (doubled) number of degrees of freedom.
      Post hoc testing involved multiple paired t test comparisons, so a Bonferroni-adjusted level of significance was used to reduce the risk of a type I error. The overall desired level of significance was .05 for a set of 6 comparisons; therefore, each comparison had to achieve significance at P equal to .008 (.05/6).

      Results

      Threshold measurements

      Figure 4 shows the mean sensory, motor, and pain thresholds for each test condition. The results show that condition C consistently had the highest threshold value, followed in order by conditions D, B, and A.
      Figure thumbnail GR4
      Fig 4Sensory, motor, and pain thresholds for each test condition shown in . Bars represent the mean; error bars show the SD.
      A 2-factor ANOVA of sensory thresholds showed a statistically significant between-conditions effect (F3,33=50.8, P=.000) and between-subjects effect (F1,11=229.1, P=.000), but no significant between-sessions effect (F1,11=.83, P=.381) and no significant interaction between condition and session (F2.77,30.5=.93, P=.430) were found. Post hoc comparisons (shown in table 1) identified a statistically significant difference (Bonferroni-adjusted, P<.008) between each of the test conditions except A and B.
      Table 1Post Hoc Test of Differences Between Conditions for Sensory, Motor, and Pain Thresholds and MEIT
      ComparisonSensory ThresholdMotor ThresholdPain ThresholdMEIT
      t11Pt11Pt11Pt11P
      A vs B2.45.032
      Not significant.
      4.10.0024.35.0012.38.037
      Not significant.
      A vs C10.39.0008.29.0006.88.0003.95.002
      A vs D7.20.0005.39.0008.75.0001.89.085
      Not significant.
      B vs C9.57.0007.16.0004.79.0015.57.000
      B vs D4.19.0023.73.0033.88.0030.33.748
      Not significant.
      C vs D4.77.0018.24.0003.06.011
      Not significant.
      6.36.000
      NOTE. Paired t tests, α=.008.
      Not significant.
      A 2-factor ANOVA of motor thresholds showed a statistically significant between-conditions effect (F1.90,20.9=43.5, P=.000) and between-subjects effect (F1,11=254.3, P=.000), but no significant between-sessions effect (F1,11=1.79, P=.208) and no significant interaction between condition and session (F3,33=.24, P=.871) were shown. Post hoc testing (table 1) identified a statistically significant difference between all test conditions.
      A 2-factor ANOVA of pain thresholds showed a statistically significant between-conditions effect (F2.05,22.5=29.2, P=.000) and between-subjects effect (F1,11=286.5, P=.000), but no significant between-sessions effect (F1,11=.56, P=.470) and no significant interaction between condition and session (F1.76,19.3=.15, P=.840) were seen. Post hoc comparisons (table 1) identified statistically significant differences between each of the test conditions except C and D.

      Motor to sensory threshold ratios

      Motor to sensory threshold ratios were calculated to assess the depth efficiency of stimulation. As figure 5 shows, conditions where currents crossed (A and B) had higher ratios than those in which the currents were parallel (C and D). A 2-factor ANOVA for repeated measures was performed to investigate current type (constant amplitude or premodulated) and current path (crossed or parallel). Between-subject variance was minimized by first standardizing values for each subject to a mean of 100 across conditions. The ANOVA showed a statistically significant between-current path effect (F1,11=18.1, P=.000); the between-current type effect was not significant (F1,11=.86, P=.358), and there was no significant interaction between current type and path (F1,11=.11, P=.742). That is, the motor to sensory threshold ratio was significantly higher (P=.000) for conditions where currents crossed (A and B) than for where they ran parallel (C and D).
      Figure thumbnail GR5
      Fig 5Motor to sensory threshold ratio for each test condition shown in . Bars represent the mean; error bars show the SD.

      Maximum electrically induced torque

      Figure 6 shows mean MEIT values for each condition. Condition B had the highest MEIT value, followed in order by D, A, and C. A 2-factor ANOVA for repeated measures was performed on current type (constant amplitude, premodulated) and current path (crossed, parallel). Between-subject variance in absolute MEIT was minimized by first standardizing values for each subject to a mean of 100 across conditions. The ANOVA showed a statistically significant between-current path effect (F1,11=11.4, P=.002). The between-current type effect was highly significant (F1,11=52.4, P=.000). Also, there was a significant interaction between current type and path (F1,11=14.3, P=.000). To establish the source of the interaction, post hoc comparisons were made (table 1) by using paired t tests at a Bonferroni-adjusted significance level of .008. The MEIT value for condition C was significantly lower than for the other 3 conditions, thus accounting for the highly significant interaction effect. There were no statistically significant differences between conditions A, B, and D.
      Figure thumbnail GR6
      Fig 6Maximum electrically induced torque for each test condition shown in . Bars represent the mean; error bars show the SD.

      Subject comments

      A total of 30 comments were made by subjects about particular conditions producing most discomfort at the pain threshold. Of these comments, 47% related to condition A, 40% to condition C, and 13% to condition B. No comments were made about condition D.

      Discussion

      The major findings of our study were that (1) crossed currents did not have greater depth efficiency than parallel currents and (2) premodulated applied currents produced higher MEIT values and less discomfort than constant-amplitude currents. True IFC is widely accepted as superior to premodulated IFC because of greater depth efficiency and comfort,
      • Goats G.C.
      Interferential current therapy.
      ,
      • Kloth L.C.
      Interference current.
      despite the lack of documentary evidence. However, the findings of our study are that true IFC has no measurable advantage over premodulated IFC in terms of depth efficiency (as assessed by thresholds), torque production, or comfort.
      Our study used a 4-electrode setup to compare true and premodulated IFCs. This was necessary to compare 2 fundamental differences between true and premodulated IFCs: the current type applied to the electrodes (constant amplitude vs premodulated) and the current path within tissue (crossed vs parallel). The experimental design maintained a consistent electrode placement and surface area of tissue stimulation. In addition, subjects were blinded to the test conditions, so that a rigorous and valid comparison of true and premodulated IFCs was possible. This is not possible when true IFC is applied using 4 electrodes and premodulated using 2 electrodes.

      Threshold measurements

      A consistent finding of the threshold data was that condition C had the highest value each time, followed in order by D, B, and A (fig 4). The differences between these conditions were statistically significant at the Bonferroni-adjusted significance level of .008 in 16 of the 18 comparisons (table 1).
      The finding that crossed-current paths (conditions A and B) had lower thresholds than parallel currents (conditions C and D; fig 3) was an expected consequence of the interference of the currents. Interference within tissue means the nerve fibers experience a greater stimulus intensity (fig 1), so a lower stimulus voltage (at the electrodes) is needed to initiate nerve firing. What is perhaps surprising is that the difference between crossed and parallel conditions was small, and that a similar pattern occurred across all thresholds (sensory, motor, pain). The threshold voltages for crossed currents (conditions A and B) were typically 85% to 90% of the parallel current values (conditions C and D). This finding indicates that interference, as such, may not be as clinically effective as has been claimed
      • Nemec H.
      Interferential therapy a new approach in physical medicine.
      or assumed.
      • Goats G.C.
      Interferential current therapy.
      Two factors that might explain the results are current spreading and the lack of electrical homogeneity of living tissue. Extensive current spreading with crossed currents would result in interference occurring in all areas of tissue between the electrodes, not maximally in deep tissues. In other words, current spreading is perhaps much greater than figure 2 suggests. Support for this explanation is found in a study by Treffene,
      • Treffene R.J.
      Interferential fields in a fluid medium.
      who measured interference of currents in a water bath and found that the interference of (crossed) currents occurred not only in the central region but also throughout the whole of the water volume. The lack of electrical homogeneity of tissue would result in complex current flow patterns and therefore irregular patterns of interference within it. The combination of irregular and more widespread current flow patterns would help explain the lower than expected difference between crossed- and parallel-current thresholds.

      Motor to sensory threshold ratios

      The ratio of motor to sensory thresholds provides a measure of depth efficiency, central to claims of the superiority of true IFC. Parallel currents (conditions C and D) had a significantly lower motor to sensory threshold ratio than crossed currents (conditions A and B). Had claims of increased depth efficiency with true IFC been correct, the result would have been the opposite. This is because crossing currents within tissue is expected to favor the stimulation of deeper structures (ie, motor nerves) relative to the more superficial ones (ie, sensory nerves). Instead, a significant between-current path effect was detected (P=.000); parallel currents (conditions C and D) had a significantly lower motor to sensory threshold ratio than crossed currents (conditions A and B). This finding suggests that, contrary to what was expected, crossed currents (true IFC) have less depth efficiency than parallel currents (premodulated IFC).
      The differences in the motor to sensory threshold ratio between premodulated currents (conditions B and D) and constant-amplitude currents (conditions A and C) were insignificant (P=.358). This indicates that applying a constant-intensity stimulus at the electrodes has no advantage, in terms of depth efficiency, over applying a premodulated stimulus waveform.

      Maximum electrically induced torque

      MEIT values provide another indication of depth efficiency and are of clinical importance for motor stimulation. Conditions B (crossed path) and D (parallel path) applied premodulated currents and produced the highest MEIT values. This provides further evidence that current spreading is extensive enough to reduce any possible differences between crossed and parallel current paths.
      Of the 2 constant-amplitude currents, crossed paths (condition A) produced a higher MEIT value than parallel paths (condition C). Condition A allowed for modulation of current within tissue, but, with condition C no modulation was possible, because the stimuli applied to each electrode pair were not phase shifted but identical. This suggests that current modulation is essential for high torque production.
      The difference in MEIT between premodulated currents (conditions B and D) and constant-amplitude currents modulated in tissue (condition A) approached, but did not reach, statistical significance (A vs B, P=.037; A vs D, P=.085). Calculations using the Student t distribution and different degrees of freedom
      • Aitken A.C.
      indicated that sample sizes of 19 and 28 subjects, respectively, would have been needed for these comparisons to achieve significance (P<.008). Our conclusion is therefore that true IFC (condition A) either produces less than or the same amount of torque as premodulated IFC.
      Thus, the results showed that applying premodulated currents produced maximal torque, and their path (crossed or parallel) had no effect. Where constant-amplitude currents were applied, modulation of the currents within tissue produced high torque, and having the currents run parallel (ie, no modulation at all), least torque.

      Subject comments

      Thirty comments were made by subjects about the 4 test conditions. Each related to a condition being particularly uncomfortable at the pain threshold. Of these, 26 comments (90%) concerned constant-amplitude currents (conditions A and C). This is clear evidence that premodulated IFCs produce less discomfort than constant-amplitude currents.
      There are 2 possible explanations for this finding. First, perhaps subjects tolerated a higher pain threshold voltage when constant-amplitude currents were applied (conditions A and C). However, this did not occur. Whereas condition C had the highest pain threshold voltage, condition A had the lowest (fig 4). The second possibility is that constant-amplitude currents (conditions A and C) produced the highest MEIT values, so any discomfort associated with the strength of muscle contraction may explain the more frequent comments about discomfort. Again this was not the case: condition C had the lowest MEIT and condition A the second lowest (fig 6).
      We are left with the observation that constant-amplitude currents (conditions A and C) produced more discomfort than premodulated currents (conditions B and D) and not because of the intensity of applied voltage or strength of muscle contraction. This finding indicates that premodulated currents are superior to constant-amplitude currents in terms of maximizing patient comfort while producing a forceful muscle contraction.

      Clinical implications

      Our findings suggest that the use of premodulated IFC, delivered through 2 large electrodes on either side of the target area, is likely to be more effective than the traditional true IFC arrangement in terms of depth efficiency, torque production, and comfort. Furthermore, premodulated IFC must be preferred from a safety perspective. True IFC applies double the average current of premodulated IFC because it applies constant amplitude and not premodulated currents. The risk of skin burns with electric stimulation increases when high average current is used.
      • Alon G.
      Principles of electrical stimulation.
      ,
      • De Domenico G.
      Therefore, the use of premodulated IFC should decrease the risk of skin burns to patients.

      Conclusions

      The findings of our study cast doubt on some fundamental claims of the superiority of true IFC. Crossing currents within tissue has no identified benefit in terms of the motor to sensory threshold ratio, torque production, and perceived comfort, when compared with currents that do not cross in tissue. Also, the premodulation of applied currents is advantageous in terms of torque production and perceived comfort compared with constant-amplitude currents. These findings contradict those predicted on the basis of conventional wisdom.
      • Low J.
      • Reed A.
      ,
      • Selkowitz D.M.
      Electrical currents.
      ,
      • Martin D.
      Interferential therapy.
      ,
      • De Domenico G.
      ,
      • Savage B.
      Two claims made by Nemec
      • Nemec H.
      Interferential therapy a new approach in physical medicine.
      are that using medium (kHz) frequency alternating current reduces skin impedance and the simultaneous application of 2 different medium-frequency currents produces a low-frequency beating effect. These claims seem reasonable and were not questioned in our study. What was questioned is whether these claims translate into clinical effectiveness.
      The findings suggest that premodulated IFC, delivered via 2 large electrodes, may be clinically more effective than the traditional, true IFC arrangement in terms of depth efficiency, torque production, and patient comfort. This raises the question whether simpler, single-output stimulators might provide safer, better, and more cost-effective patient treatments than the widely used true IFC stimulators.
      Suppliers
      aChattecx Corp, Chattanooga Group, 4717 Adams Rd, Hixson, TN 37343.
      bStimtrode 50×90mm rectangular self-adhering electrodes; Axelgaard Manufacturing Co, 329 W Aviation Rd, Fallbrook, CA 92028.
      cMetron Vectorsurge III; Metron Medical Australia Pty Ltd, 57 Aster Rd, Carrum Downs, Victoria 3201, Australia.
      dTrio CS-1560A II, 15MHz oscilloscope; Jaycar Electronics Pty Ltd, 100 Silverwater Rd, Silverwater, NSW 2128, Australia.
      eSPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

      References

        • Low J.
        • Reed A.
        Electrotherapy explained. 3rd ed. Oxford, Butterworth-Heinemann2000
        • Selkowitz D.M.
        Electrical currents.
        in: Cameron M.H. Physical agents in rehabilitation from research to practice. WB Saunders, Philadelphia1999: 345-427
        • Lindsay D.
        • Dearness J.
        • Richardson C.
        • Chapman A.
        • Cuskelly G.
        A survey of electromodality usage in private physiotherapy practices.
        Aust J Physiother. 1990; 36: 249-256
        • Robertson V.J.
        • Spurritt D.
        Electrophysical agents.
        Physiotherapy. 1998; 84: 335-344
        • Lindsay D.
        • Dearness J.
        • McGinley C.
        Electrotherapy usage trends in private physiotherapy practice in Alberta.
        Physiother Can. 1995; 47: 30-34
        • Pope G.D.
        • Mockett S.P.
        • Wright J.P.
        A survey of electrotherapeutic modalities.
        Physiotherapy. 1995; 81: 82-91
        • Alon G.
        Principles of electrical stimulation.
        in: Nelson R.M. Hayes K.W. Currier D.P. Clinical electrotherapy. 3rd ed. Appleton & Lange, East Norwalk1999: 55-139
        • Martin D.
        Interferential therapy.
        in: Kitchen S. Bazin S. Clayton’s electrotherapy. 10th ed. WB Saunders, London1996: 306-315
        • Johnson M.I.
        • Tabasam G.
        A double-blind placebo controlled investigation into the analgesic effects of interferential currents (IFC) and transcutaneous electrical nerve stimulation (TENS) on cold-induced pain in healthy subjects.
        Physiother Theor Pract. 1999; 15: 217-233
        • Nemec H.
        Interferential therapy.
        Br J Physiother. 1959; 12: 9-12
        • Reilly J.P.
        Electrical stimulation and electropathology. Cambridge Univ Pr, Cambridge1992
        • De Domenico G.
        New dimensions in interferential therapy. Reid Medical Books, Sydney (Aust)1987
        • Goats G.C.
        Interferential current therapy.
        Br J Sports Med. 1990; 24: 87-92
        • Kloth L.C.
        Interference current.
        in: Nelson R.M. Currier D.P. Clinical electrotherapy. 2nd ed. Appleton & Lange, East Norwalk1991: 221-260
        • Savage B.
        Interferential therapy. Faber & Faber, London1984
        • Lambert H.L.
        • Vanderstraeten G.G.
        • De Cuyper H.J.
        • et al.
        Electric current distribution during interferential therapy.
        Eur J Phys Med Rehabil. 1993; 3: 6-10
        • Knaflitz M.
        • Merlitti R.
        • DeLuca C.J.
        Inference of motor unit recruitment order in voluntary and electrically elicited contractions.
        J Appl Physiol. 1990; 68: 1657-1667
        • Binder-Macleod S.A.
        • Halden E.E.
        • Jungles K.A.
        Effect of stimulation intensity on the physiological responses of human motor units.
        Med Sci Sports Exerc. 1995; 27: 556-565
        • Treffene R.J.
        Interferential fields in a fluid medium.
        Aust J Physiother. 1983; 29: 209-216
        • Aitken A.C.
        Statistical mathematics. Oliver & Boyd, Edinburgh1952