Interferential current, IFC

Order my book in German? Click here. There's an English translation planned for 2025.

This is not the complete IFC chapter. It's a teaser.

Medium frequency therapy is the therapeutic application of alternating currents from the medium frequency range. According to Edel's classic categorisation, this is the range between 1000 and 300 kHz. The somewhat rough distinction is made due to the different physiological effects in the low frequency, medium frequency and high frequency range (Edel 1977).

References can be found in the book, as well as the illustrations (for copyright reasons)

Why the distinction at 1000 Hz?

In the low-frequency range, the principle of period-synchronous stimulation (= cycle-synchronous stimulation) applies. This means that each impulse triggers an action potential, provided that the phase duration and amplitude are sufficient. The action potentials are triggered in the same rhythm as the frequency of the current. 5 pulses trigger 5 action potentials. 50 pulses trigger 50 action potentials. 500 impulses trigger ... I have no idea and actually no one knows.

Explanation below. A little perseverance is required.

A nerve fibre has a maximum depolarisation frequency. This frequency is determined by the refractory period and is theoretically around 300 Hz. In practice, frequencies above 200 Hz are a rarity (del Vecchio et al. 2019). Up to this cut-off frequency range, each pulse will trigger an AP 1:1.

When stimulated at a frequency above the maximum depolarisation frequency, the neural response changes. Firstly, after a few seconds this leads to a conduction block, a so-called High Frequency Electrical Conduction Block, see below. Secondly, it takes several impulses to trigger an action potential. Thirdly, some of the impulses fall within the refractory period, so that not every impulse can cause a depolarisation.

If a nerve is stimulated with 2000 impulses, for example, the other end will not generate 2000 action potentials, but only a few hundred, and irregularly at that. This phenomenon is called asynchronous depolarisation or period asynchronous stimulation. This is why the generously rounded-up distinction between low and medium frequency is made at 1000 Hz.

In the high frequency range, above 300 kHz, electromagnetic waves are used which have a thermal effect in depth (diathermy) and do not trigger any electrochemical and neuromuscular stimulation effects, except those associated with heating.

Summation, Gildemeister effect

At the beginning of the last century, the physiologist Gildemeister discovered that several impulses are necessary to trigger an excitation in MF (Gildemeister 1944). Each negative half-period depolarises the cell membrane a little more until the depolarisation threshold is finally reached after a certain number of alternating current periods (Gildemeister calls this the effective time). An action potential is then triggered.

The higher the amplitude, i.e. the set intensity, the shorter the effective time. Gildemeister recognised this phenomenon as a summation. Bromm and Lullies (1966) investigated the phenomenon further and found that the cathodic half-waves of the sine wave cause the gradual depolarisation and that the anodic sine half-waves that immediately follow are not sufficient to counteract this depolarisation. The effect occurs under both electrodes.

Incidentally, exactly the same thing happens when symmetrically compensated square-wave pulses are used, except that no summation occurs here (Bromm and Lullies 1966). This effect was later called the ‘Gildemeister effect’. Knowledge of this phenomenon makes it difficult to explain the effect of MF currents on tissue.

We can never predict exactly at what frequency the target tissue will be stimulated. This is explained in detail below.

High Frequency Electrical Conduction Block

In 1903, the physiologist Wedensky described an interesting phenomenon that he had already discovered in 1884. When stimulating Aα motoneurons, he had noticed, among other things, that when using medium-frequency currents, the muscle contractions triggered very quickly became weaker or even stopped altogether after a short time. When lower frequencies were used, strong contractions could be triggered again immediately. There was no explanation for this phenomenon at the time; it was later called ‘Wedensky inhibition’ (Bowman and McNeal 1986).

Today we know that the high frequencies have consequences for the stimulation: some of the impulses fall within the refractory period of the nerve. This makes repolarisation of the membrane more difficult or even prevents it. After a few seconds under the electrodes, this causes an equally rapidly reversible tonic depolarisation of the membrane: the resting potential is no longer reached (Bhadra and Kilgore 2004; Kilgore and Bhadra 2004). Consequently, action potentials can no longer be transmitted: A conduction block has been created.

This type of conduction inhibition is now referred to as High Frequency Electrical Conduction Block (Bowman and McNeal 1986). The phenomenon is of clinical-therapeutic interest in the inhibition of unwanted muscle activity, e.g. in spasticity, and perhaps in the inhibition of pain.

In addition to this inhibition, continuous stimulation leads to rapid depletion of the neurotransmitter reserves at the motor end plate. The exhausted endplate is no longer able to transmit action potentials to depolarise the muscle fibre membrane. This phenomenon also manifests itself as inhibition and is an aspect of fatigue.

Why use MF current for treatment at all?

Medium-frequency currents have been known for a long time and, as always, it is only a matter of time before people start to think about their use. In most cases, either the military or the medical sector takes the first step (ultrasound, infrasound, lasers, microwaves, etc.), followed with great interest by industry.

MF currents can be used to stimulate tissue, and Gildemeister's many experiments on humans have shown that the current is not even that unpleasant and only rarely leads to serious side effects despite high intensities (Gildemeister 1944). Furthermore, in purely mathematical terms, the skin resistance would be lower than at low frequency due to the high frequency (more on this topic later).

The problem, however, was the aforementioned Wedensky inhibition. What now?

It should be possible to shape the current in such a way that the nerve fibres have the chance to repolarise so that the aforementioned inhibition mechanisms do not occur. Solution: A continuous MF alternating current is used to form a current that combines the impulse properties of an AF current with the advantages of an MF current.

Solution one

The simplest and most obvious solution in chronological terms is solution two: It was devised in Russia at the end of the 1960s, some time after the Austrian engineer Nemec had developed his method. According to the Russian method, the constant medium-frequency current is interrupted in the device so that the nerve fibres are given the opportunity to repolarise. Wedensky inhibition is thus prevented because it only occurs after a few seconds of continuous stimulation.

For this purpose, so-called MF pulses are formed, whereby the pulse-internal frequency always corresponds to the frequency of the selected MF frequency, which was 5000 Hz for the Russians at the time. This means that when a stimulation frequency of, for example, 50 Hz is set, the MF current is interrupted 50 times per second. This type of current is referred to as a frequency-modulated medium-frequency current or MF pulse method; the pulse shape was sinusoidal (Fig. 3.7, variation C). Today, Ward and Oliver (2007) call it Burst Mode Alternating Current (BMAC) and it is referred to as bursts or trains. These ‘MF pulses’ are therefore not single pulses but ‘blocks’ consisting of several pulses.

Some readers may recognise the current as Russian stimulation avant la lettre, or - in a slightly modified form - as Aussie stimulation. More on this topic in the chapter on muscle stimulation.

Solution two

The second solution is considerably more complicated, but probably lends the current a certain mystique (Johnson 1999). Although the idea has definitely been proven not to work, it is certainly original (Treffene 1983; Beattie et al. 2011).

Initial situation: When two waves meet, interference phenomena occur.
In Fig. 3.8, the superposition on the left leads to a doubling of the wave energy, while on the right the waves cancel each other out: The net energy result is zero. If the waves differ very little in frequency, they can never overlap exactly. This results in a so-called beat: the intensity then increases and decreases at a certain frequency: this is called the beat frequency.

The guitar players among us apply this principle when they tune their instrument with harmonics. The closer the notes, i.e. their frequencies, of two strings are to each other, the lower the beat frequency; the further apart, the higher the beat frequency. If two notes are very close together, for example the A string is tuned to 440 Hz and the string to be tuned ‘almost’ produces an A with 443 Hz, then a beat with a frequency of 3 Hz (443 minus 440) is produced as the difference tone. This becomes audible when tuning the guitar: It is still an A (almost), but the volume (intensity) increases and decreases 3 times per second, you can hear the beat of 3 Hz.

The so-called true interference current (IF) works as follows: An alternating current with a constant frequency of 4000 Hz (the so-called carrier frequency) is combined with an alternating current whose frequency can be set between 4000 and 4200 Hz. The difference between the two frequencies then results in the beat frequency, as with sound vibrations. However, IF is not referred to as beat frequency, but as amplitude modulation frequency (AMF): the frequency remains constant, only the amplitude (intensity) is modulated. Hence the official name: amplitude-modulated medium-frequency current.

Example: One circuit has a fixed frequency of 4000 Hz, the other circuit is set to 4050 Hz, the difference is 50 Hz. This is the AMF: the actual (supposed) stimulation frequency that is supposed to trigger something in the tissue and supposedly has the same properties as a low-frequency 50 Hz pulse current.

The low-frequency interruptions (the beats) actually prevent a conduction block with their creeping on-off phases: the continuous MF current is constantly interrupted. In this way, the cell membrane can repolarise again and again.

According to Nemec, two separate MF circuits with two pairs of electrodes, also known as tetrapolar interference, are used for this purpose in a technically quite complicated way. The two circuits should be joined together in the tissue and thus acquire an AF stimulus character. This is called endogenous interference.

If these two circuits are joined together in the device, this is referred to as exogenous interference, which is bipolar and does not require 4 electrodes. In English-speaking countries, the terms ‘true interferential current’ or ‘premodulated interferential current’ are used.

At the interference localisation (in the so-called interference cross), the amplitude (intensity) of the two converging circuits would theoretically be almost doubled. As with ultrasonic waves, there would be amplifications (increases in intensity) and attenuations (decreases in intensity) of the intensity set on the device in mA. The carrier frequency (tone A in the example above) remains the same. In practice, however, it looks quite different.

Spoiler alert: it doesn't work ...

Special properties of medium-frequency currents

Medium-frequency alternating current is a symmetrically compensated alternating current with a frequency between 1000 and 300,000 Hz. As a rule, frequencies between 2500 and 10,000 Hz are used, most frequently 4000 Hz for unclear reasons (Johnson 1999). Mostly sinusoidal pulses are used. Because of the symmetrical compensation, no electrolytic effects normally occur.

However, this does not mean that MF is harmless! In a study by Ward and Robertson (1998), blisters appeared around the electrodes after application in 3 healthy volunteers. In 2005, Ford et al. described a patient with 3rd degree burns after MF application to the knee following knee replacement surgery. In 2008, Satter described a patient with burns after MF with osteosynthesis material in the treatment area.

The chemical changes that occur in the tissue during the short (125 μs at 4000 Hz), negative half-wave are counteracted by the subsequent positive half-wave, just as with compensated TENS pulses. Without chemical changes, no pH change occurs in the tissue and the risk of tissue damage is low. Because of the rapid polarity changes at the electrodes, the current is referred to as apolar.

Wyss (1975) speaks of ambipolar stimulation, but as the processes under the electrodes do not take place exactly simultaneously, but with a half phase shift, the MF current also follows the laws of polar stimulation as normal (Bromm and Lullies 1966). But that's nitpicking.

The short impulses (125 μs at 4000 Hz) stimulate, like the equally short High TENS impulses, practically no Aδ and C fibers, which is why the current is usually quite comfortable for the patient. That being said, depending on how high the intensity is set, high intensities of MF current can also become unpleasant (Ward and Robertson 1998; Palmer et al. 1999; Ward and Oliver 2007).

According to Nemec (1959, 1960, 1967, 1968), interferential current has three advantages:

1. Due to the high frequency, the skin resistance is lower, making it easier for the current to pass through the poorly conducting skin barrier. This would result in less sensory load for the patient, making the current more pleasant for them. • It's not that simple. First, resistance does not depend solely on frequency but also on the phase duration (Lykken and Venables 1971; Yamamoto and Yamamoto 1977). At a frequency of 4000 Hz, the phase duration is 125 μs, which is in the same range as with TENS applications. Nonetheless, it is a fact that skin impedance is lower with medium frequency than with low frequency. However, the composition of the frequencies depends significantly on the waveform. It is possible for different waveforms to produce the same impedances. Both applications can therefore be equally (un)pleasant depending on intensity. Second, the high skin resistance doesn't really matter. This resistance is caused by the stratum corneum, which contains no sensors. The relevant sensors, responsible for the sensory aspect of the current, are located beneath this layer, where the electrical properties are the same as deeper in the tissue.
2. The two current circuits would interfere with each other, forming a region at the so-called interference cross where the target tissue is maximally stimulated (Fig. 3.9). • That is definitely not true. Interference phenomena occur almost everywhere between the four electrodes, but least in the center. First, the current density decreases with distance from the electrode, making it lowest in the center between the electrodes. Stimulation is always strongest directly under the electrodes. Treffene demonstrated in 1983 that interference phenomena occur almost everywhere between the electrodes in a water-filled plastic basin, but unfortunately weakest in the center where they are supposed to occur; here, they were minimal. The stimulation was greater between two adjacent electrodes than in the area where the circuits were supposed to cross. • Unlike the water basin mentioned above, the human body is not homogeneous, so it is impossible for the voltage lines to cross at the correct point for interference. Beattie et al. conducted measurements on humans in 2011 on the medial quadriceps muscle to determine the voltage distribution of tetrapolar interferential current in the muscle. They placed measurement electrodes at various positions relative to the four stimulation electrodes. The result: the highest voltage was measured near a stimulation electrode, and the lowest in the center at the supposed interference cross. A measurement electrode located 5 cm outside the stimulation area of the four electrodes recorded values almost twice as high as in the supposed interference cross. The authors also compared exogenous IF with endogenous IF and found that with exogenous IF, where the interference occurs in the device rather than in the patient, the highest voltage, as expected, was measured superficially near the electrodes.
3. In the interference cross, a region would form with the actual stimulation frequency, the amplitude modulation frequency (AMF). The AMF, or beat frequency, is determined by the different frequencies of the two current circuits and typically lies, depending on the settings, in the usual low-frequency range of 1 to 200 Hz. This AMF is claimed to have properties different from low-frequency currents, specifically genuine MF stimulation effects due to the MF carrier frequency itself (Edel 1977). • Much of this is incorrect. First: see the point above. Second: The reason for the occurrence of genuine MF stimulation effects is physiologically vague at best. Third: It is repeatedly claimed that the tissue responds to this 50 Hz AMF just as it does to a low-frequency current of the same frequency, with effects like pain relief and improved circulation supposedly occurring with MF current as with LF. More on this later. First, about the AMF: When low frequency stimulates one end of a nerve fiber at 50 Hz, one can expect 50 action potentials at the other end. This works up to a certain physiological limit, which, as mentioned earlier, is determined by the nerve's refractory period. Let us examine this more closely. The part of the pulsed current intended to create stimulation consists of a short series of individual pulses (Fig. 3.7c and d), called a burst. The duration of such bursts is either fixed or variable depending on the device, as long as these 50 Hz fit within a second. For example, if a burst lasts 10 ms, it consists (at a base frequency of 4000 Hz) of 10 ÷ 1000 × 4000 = 40 alternating current pulses. Now the refractory period comes into play. During the absolute refractory period, the fiber cannot generate an action potential, lasting about 1 ms. Then, during the relative refractory period, the nerve fiber can generate an action potential, but only if the stimulus is significantly stronger than normal. This phase lasts about 2 ms. For this reason, the refractory period only affects firing rates higher than about 300 Hz, which is unusual in the human nervous system. These 300 Hz correspond to 1/13 of the 4000 Hz generated by MF current. In other words, at most, every 13th MF pulse could trigger an action potential (Bowman and McNeal 1986). Each burst thus generates at most about 3 action potentials in the nerve with its 40 pulses, and the 50 bursts per second trigger a total of 150 action potentials (40 ÷ 13 × 50). Therefore, the target tissue is not stimulated at 50 Hz but at 150 Hz. Maybe. Some impulses remain subthreshold with the sinusoidal endogenous variant (Fig. 5.7d); they likely trigger nothing, though the Gildemeister effect might play a role here. However, we do not know how this effect influences the process, as we do not know how many individual MF pulses are required to summate and trigger an action potential. Perhaps only 3 pulses per burst are needed. Or 8. Or it varies. It is certainly not exactly 50 Hz but rather a multiple of the set burst frequency.

There is a solid basis for these claims (Bowman and McNeal 1986; Stefanovska and Vodovnik 1985; Laufer et al. 2001; Ward and Robertson 2000; Laufer and Elboim 2008). This tedious calculation is meant to clarify that the tissue is stimulated with quite high frequencies despite the low AMF settings. Whether this is biologically meaningful is unclear and unresearched.
Additionally, the calculation demonstrates that statements like the AMF determines the depolarization frequency and the AMF corresponds to the frequencies used in low-frequency therapy (den Adel and Luykx 2005 in a therapy manual from a device manufacturer) are not true.

The results of TENS research clearly cannot be applied to MF current.

Spectrum, Sweep, and Vector

To optimize the effects of medium-frequency (MF) currents and avoid habituation, most devices offer various adjustment options. Four of these are almost always available, though often under different names. Users can typically select the AMF (amplitude modulation frequency), the so-called spectrum and its sweep method, and a vector.

A spectrum, or sweep in English, is an AMF range that the device automatically cycles through step by step. For example, you might set a lower limit of 70 Hz and an upper limit of 120 Hz, and the device adjusts the AMF within this range in a predetermined rhythm. Alternatively, some devices require you to set a base frequency, such as 70 Hz, along with a spectrum of 50 Hz, causing the AMF to vary from 70 Hz to 120 Hz and back. Other devices may suggest preset programs, like Program 42, which handle all adjustments automatically. It's worth reading the user manual to understand these functionalities fully.

Additionally, every device allows the sweep mode to be selected. These modes determine how the sweep is performed, either more abruptly or smoothly, depending on the patient's condition. For instance, in the 70–120 Hz example above, the AMF could change abruptly—e.g., 70 Hz for one second, then 120 Hz the next—or transition more gradually over six seconds up to 120 Hz and back down. Various frequency combinations can be programmed similarly.

Johnson and Tabasam (2003a, b) conducted two studies on healthy volunteers to investigate how these settings influence sensitivity to cold pain (e.g., immersing the arm in ice-cold water). The setup involved a carrier frequency of 4000 Hz, tetrapolar IF on the forearm, and a strong yet tolerable intensity for 20 minutes. Participants were blinded to the settings.
• In the first study, 60 participants were tested with AMF frequencies of 20, 60, 100, 140, 180, and 220 Hz.
• In the second study, 40 participants were tested with an AMF range of 1–100 Hz and different sweep modes: 1–1 abrupt, 6–6 abrupt, 6–6 gradual, and 100 Hz bursts (burst duration unspecified).

Results: Interferential current (IF) increased cold pain thresholds, but neither the frequency nor the sweep mode influenced the outcome. Fuentes et al. (2010) obtained similar results on healthy subjects, showing that IF increased pressure pain thresholds on the back with or without AMF sweeps.

Then there's the vector. Whether it's labeled as an isoplanar field, dipole vector, rotating vector, or something else, the mechanism generally involves amplitude modulation within the device or varying the intensity across individual channels—or both. This creates the sensation of the current moving around in the treatment area. Practically, it makes electrode placement less critical.
However, Noble et al. (2000) found no added benefit from a rotating vector in their study on mild improvements in blood circulation.

Considerations for Medium-Frequency Current Therapy

Numerous studies affirm the effectiveness of TENS (transcutaneous electrical nerve stimulation). TENS is simple to use, virtually free of side effects, portable, and affordable. While medium-frequency (MF) currents have been shown to be effective—better than placebo—they are no more effective than TENS (Table 3.6). MF devices are costly, often requiring patients to visit a clinic, and the proposed mechanisms of action are vague at best and sometimes entirely unfounded.
The cost argument against IF (interferential current) devices is less valid nowadays, as portable devices for home use are now available in the same price range as high-end TENS devices. The tetrapolar interference cross effect often promoted in IF therapy doesn’t exist where it’s intended. Pre-modulated IF is a more practical choice.
For sensitive patients, a basic setting like 4000 Hz carrier frequency, 80 Hz AMF, 50 Hz spectrum, 6/\6 sweep (gradual increase and decrease over six seconds), and a perceptible but comfortable intensity is generally very effective. The 6/\6 symbol on the device indicates a gradual six-second increase and decrease cycle.

Key points:

• The AMF setting and sweep pattern don’t affect efficacy (Johnson and Tabasam 2003a, b). These adjustments can instead be used to make the treatment more comfortable for the patient.
• The mechanism of action likely involves the Gate Control Theory, as fast-conducting nerve fibers are almost certainly being stimulated. However, unlike TENS, there’s no direct evidence supporting this.
• There’s no data to suggest the involvement of substances like endorphins, dynorphins, or enkephalins in IF’s effects. A search for interferential current AND endorphin on PubMed yields zero results. Much of the theoretical framework for IF appears to be borrowed from TENS research without justification.
One thing seems clear: The harder a therapist tries to localize the interference cross at the right spot, the greater the placebo effect.

Shanahan and Ward (2006) demonstrated that High TENS is objectively more effective than IF for pain relief. However, participants reported that IF was more pleasant to use.
This presents a dilemma: Should we prioritize objective effectiveness or patient preferences? It might not work as well, but it feels better isn’t exactly a ringing endorsement.

Food for thought, folks!