Heffernan et al.

Measurement of Electromagnetic Field in the Healing Response

Michael Heffernan, Ph.D.

Pain Control Clinic
1807 HWY 35
Rockport, TX 78382
Phone and Fax: (512)-729-0817

Correspondence should be addressed to: Michael Heffernan, Ph.D.

Submitted for publication: September 1996

Keywords: Electromagnetic Field In Healing, Microcurrent Therapy, Tissue Healing

Abstract Introduction Materials and Methods Results
Discussion Conclusions References Table of Contents

Figure 1 Figure 2 Figure 3 Figure 4
Figure 5 Table I Table II Table III


A theoretical framework is presented identifying low frequency electromagnetic field components, in living tissue, as necessary "first messengers" in tissue healing. Examining the electromagnetic spectra in injured ,and contralateral, non injured tissues, the author finds an increase in the low end, 1-4Hz band, associated with healing. The method used a ten thousand gain, common mode rejection, amplifier, which feed raw signal data into an optical isolated computer, which in turn analyzed, by Fast Fourier Transform, the relative root mean square amplitudes of tissue at selected injured, and non injured joints. This finding is limited to acute injuries in normal or healthy subjects. Injures associated with degenerative conditions are believed, by the author, to require an artificial low frequency band, square wave, microcurrent to promote the healing response.


Constant shifts in the direct current field around injured tissues has been measured and found to be maintained during normal healing and regeneration. Immediately following injury a wave of positive polarization in the site of injury is soon followed by a longer shift to negative potential which lasts for the duration of the healing period (1). Following healing the tissue again returns to a neutral field. Delivery of a small nanoampere, direct current (DC), with the negative pole placed in the area of injury has been shown to facilitate healing in otherwise unhealed, non union bone fractures (1). Other researchers have used 15 Hz electromagnetic fields (EMF) placed in the vicinity of unhealed soft tissue injuries to facilitate and speed the healing response which is otherwise slow to non existent (2).

A transmembrane signaling process has been identified, whereby very weak EMF at certain windows of frequency and intensity, act as first messengers, sensed at the glycoprotein sites in the cell membrane. Following exposure to this EMF, calcium ions are effluxed from the cell interior. The resulting effluxed calcium ions bind to anion sites of hydrphobic membrane protein, further amplifying weak external EMF through intramembranous proteins into the cell interior to increase Cylic-AMP, ATP, DNA transcription, and increased production of cell growth promoters (3). The frequency and amplitude characteristics of EMF needed to promote this cell response were found to be in the low end of the extremely low frequency (ELF) range. Based on evolutionary history, the author speculates that such ELF frequencies would be similar to the Schumann resonance frequencies found naturally in the ionosphere. Schumann frequencies as low as ten microTesla are of the same magnitude as the ELF that was used to initiate cell amplification by positive feedback loops and calcium efflux (3). In this model Schumann frequencies would positively stimulate cells over a broad range of ELF frequencies between 1-10 Hz.

Deducing from the natural evolutionary history of cellular ELF expoure, some range of mixed EMF signal would be expected to be beneficial in promoting normal cellular function and regeneration. The band width of expected beneficial frequencies would exclude those singular frequencies associated with possible cancer promotion. Cancer promotion frequencies have been found in singlar narrow bands, mainly above the low end Shumann resonances of the ionosphere. The cancer promotion frequencies between 15-60 Hz, alter cellular protein synthesis, disrupt mRNA, reduce immune response, and interfer with cellular electromagnetic communication in the "gap junctions" (4). It can not be assumed, that the specific and singular frequencies used experimentally as cancer promoters, are representative of the cellular response from a mix or whole band of external frequencies as found in nature, or produced by the body itself.

A method decided upon by the author in learning more about the amplitude and frequency parameters of beneficial cellular frequencies, was to compare the power spectrum of EMF produced in the body naturally as well as during healing. Evidence from the spectral character of amplified tissue EMF has shown that healthy subjects and pain patients vary dramatically in the frequency and amplitude characteristics of tissue EMF fields made by each (5). Furthermore, microcurrent stimulators which most closely emulate the spectral characteristics of pain free healthy subjects provide the greatest pain relief (5). The spectra of healthy or pain free subjects shows a generally higher amplitude at the lower 2-5 Hz frequency band with a non linear rapid drop in power at increasing frequencies (5).

To continue these observations, the author decided to measure the spectral characteristics of the EMF produced during healing in normal subjects. The hypothesis was that there would be a specific differences in the healing tissue's spectral characteristics when the healing tissue was compared to non injured control tissues in the same subjects. Specifically, healing tissue EMF spectra, measured by a computer analysis of amplified microvoltage fluctuations produced around injuries, would be significantly different when compared to the spectral curve produced from the contralateral, non injured body part on the same subject.


Common mode rejection amplifiers, amplifying microvoltage readings by ten thousand were used to measure the electric field fluctuations in human subjects by placing electrodes over the body region being measured. These amplifiers were then interfaced with an optical isolation device, and linked to computers that preformed the necessary Fast Fourier Transform (FFT) calculations from raw signal input. The author used an amplifier built by J and J Electronics, Poulsbo, Washington, and software imported from Russia, sold by API, Inc. of Washington State, under trade name "Neurodata".

To view signals and FFTs, gains were adjusted to plus or minus 50 microvolts full scale, and FFT's taken over a minimum of at least sixty seconds, thereby eliminating momentary chaos of the electromagnetic signals. Bilateral, silver-sliver chloride, one inch disk electrodes, prepared with standard EEG electrode paste were placed on the subject a few inches apart. When measuring joint areas on the extremities a dorsal/ventral or medial/lateral electrode placement was used.

Reliable measurement of the body's EMF requires elimination of potential electrical artifact from movement, EMG, and EKG. By simultaneous observation of both raw signal and its Fast Fourier Transform (FFT), artifact was easily detected. To identify EKG or EMG artifact, the raw signal, sensitive to plus or minus 2 microvolts, was observed for QRS, slow waves, or higher frequency EMG.

To insure that some form of endogenous electromagnetic signal, independent of EKG, or EMG, was being reliably measured, the author performed a variety of observations of FFT and raw signal under varying conditions. Figure 1 in shows a marked EKG from abdominal placement near a major artery. The corresponding 2 second update (averaged) FFT shows a harmonic spreading of microvoltages from 2 to 35 Hz. This is a characteristic response of FFT to EKG input which comes about from improper electrode placement over arteries or when electrodes are too far apart.

To avoid EKG artifact, locations and electrode placements were selected where no EKG is seen in the raw signal. When EKG is absent from the raw signal, the harmonic spreading and flatness of the FFT will give rise to an FFT which shows a spectrum as a non linear descending function. The non linear descending curve seen in Figure 2 is without artifact, and is the same as that observed in another study done by the author in measuring spectra statistically associated with pain relief resulting from a particular type of microcurrent stimulation (5).

Figure 3 shows an example of a fidgeting subject who moved her foot while recording electrodes were distally located on the right hand. The wide DC shifts in microvoltage are characteristic of movement artifact. Electromyographic (EMG) artifact is easily seen in both the raw signal and FFT. In the raw signal the EMG will appear as a consistent 30-35 Hz frequency pattern which disappears when the subject relaxes and increases when the subject increases muscular tension. The FFT, correspondingly reflects this EMG artifact as a gradual peak in the higher frequency ranges. Figure 4 shows this effect from a right hand recording with muscle tension in the shoulder area. This figure also shows some movement artifact in the first part of the raw signal epoch.

Reliable measurement of the tissue EMF, required electrode placement away from major arteries, and instructions to subjects to remain still and relaxed in a comfortable siting position during spectral readings.

Subjects were selected from the authors pain control practice, on the basis of willingness to participate in a "study to measure the electrical characteristics of soft tissues during normal healing". Median age of the S's, six women four men, was 65, with age range of 35-60 yrs. Median household income was 60K per year with average education being four years beyond twelfth grade.

Subject's (S's) were told that their body was being examined electromagnetically and that they would be given the results following the analysis. All subjects consented to the procedure. S's were selected from a pain management clinic, six women four men, age ranged from 35-60. S's selected were required to evidence an acute injury of not more than one week in duration, and the injury had to be unilateral to the knee, shoulder, or elbow, thus allowing for a non injured contralateral control joint. The ten S's broken down by type of injury consisted of four shoulder, three knee, and three elbow injuries. An additional ten (10) pain free Ss were selected from a population of otherwise similar characteristics to serve as control Ss.

Electrode placement for the knee and elbow was medial-lateral as seen from standard anatomical position. The electrode placement for the shoulder was ventral to dorsal in an axis perpendicular to the insertion of the pectorals muscle. Double blinding was achieved by not communicating with the subjects about any expected differences in recording sites, and by saving all data collected by location and subject in a separate coded computer file, randomly selected for data analysis after all recordings were completed. Following data analysis the computer files were then decoded.

By recording the Fast Fourier Transforms (FFT's) of the same subjects, from the same locations on two successive days a retest reliability quotient was calculated. This allowed for a determination of the stability and reliability of the FFT's shape and magnitude during the healing process. A recording of the root mean square (RMS) microvoltage amplitude at each one Hz band width from 2-30 Hz over the two separate days was made. A Pearson r correlation coefficient of reliability for all data at each band width produced a reliability coefficient of .89, significant at the .001 level.


All ten pain subjects showed elevated, low end 2-4Hz, FFT amplitudes in their injured joint when compared to the non injured joint on the opposite side of the body. The four knee injuries, three shoulder and three elbow injuries, all displayed this same pattern. Higher frequency bands in the FFT of injured side were greater than non injured tissue readings, but the difference did not achieve significance when analyzed by paired T tests. The significant results between injured and non injured low band (2-4Hz) amplitudes are shown in Table I below.

Subject #        RMS Amplitude (2-4Hz)
                 injured   non injured

   1               .55        .38
   2               .43        .22
   3               .67        .40
   4               .74        .33
   5               .64        .23
   6               .78        .33
   7               .85        .45
   8               .64        .37
   9               .87        .42
  10               .73        .28

Data were analyzed for significance using a one tailed paired t test. The statistical table in computing the t statistic is shown below as Table IA:


Mean difference of paired scores	  0.35
Sum of squared deviation scores		132.25
t statistic			 	  2.88
Significance level (one tailed)	 	  >.01

The observed difference in amplitudes between non injured and injured tissues showed injured tissue to have a significant elevation in RMS values. Figure 5 shows a sample FFT taken from the right injured knee of S5 in the top of the figure, the bottom of the figure shows the non injured knee for same S. When measured on one month follow up the significant difference in FFT's between injured and non injured contralateral body parts declined to non significance after healing. These results are shown below as table II.


Subject #       RMS Amplitude (2-4Hz)
                injured   non injured

    1             .29         .30
    2             .38         .36
    3             .33         .43
    4             .22         .28
    5             .30         .28
    6             .37         .33
    7             .39         .43
    8             .28         .31
    9             .35         .38
   10             .28         .29

The average difference between injured and non injured joint at one month follow up was only .04 as compared to .35 found during healing.

Table III (below) shows the comparison of the non injured control group right and left shoulder, knee, and elbow. No significant difference was found between contralateral joints in this non injured population. Mean difference between contralateral, non injured joints being only .03.


Subject#         RMS Amplitude (2-4Hz)
                  Right       Left

   1               .35        .32
   2               .27        .31
   3               .31        .31
   4               .30        .28
   5               .35        .36
   6               .29        .28
   7               .32        .35
   8               .36        .29
   9               .29        .27
  10               .33        .35


From this studies findings there appears to be a significant elevation in 2-4 Hz band amplitude in the EMF produced by the body during healing. Furthermore, this significant increase in the low end of the FFT curve returns to normal with resolution of the injury. This result has a remarkable similarity to that found and reported by Becker (1), in his analysis of DC changes at the site of injury. There appears to be an electromagnetic field of specific frequency and amplitude as well as specific spectral characteristics that accompanies the DC shift found around the site of injuries. The body's perineural cells have been found to obey the conductive characteristics of N-type semiconductors since they display the "Hall Effect" (1). One property of N-type semiconductors is rectification of an alternating EMF into DC current. This fact may explain the origin of Becker's DC "current of injury" found to be required for healing and regeneration (1). If picoamp DC is needed in the healing response, and if this study's findings of a low end spectral increase is the source of the "current of injury", then increased 1-4Hz in the tissue is a necessary condition for healing.

The question then arises, "From where does this 1-4Hz EMF in the body arise?" One probable source of low frequency fluctuations in tissue is the fluctuating cellular membrane potentials. Researchers have measured a consistent membrane potential of somatic cells as being around 100 millivolts (2). As receptor sites on cells bind ionic substances, resulting membrane depolarizations are constantly shifting the magnitude of the membrane potential. Even a ten percent (10%) fluctuation in this membrane potential would produce 100 microvolt EMF pulses. Other research suggests the extracellular gap junctions act as conduits of electromagnetic pulses needed in cellular communication (2). These gap junction EMF pulses are believed to be necessary for cell communication and signaling, by maintaining normal growth and differentiation in living tissue (3). The authors prior work has shown that low frequency, random or chaotic pulses produce the most effective and rapid pain relief when compared to higher frequency microcurrent stimulation (5).

Unfortunately not all injuries undergo normal healing. Most clinicians know there are certain individuals from whom there is an absence of normal healing, or the rate of healing is so slow that it is imperceptible, frustrating both doctor and patient. This lead the author to make some EMF observations of injury sites that were chronic and non healing. Viewing spectra of patients with unilateral knee or shoulder degenerative joint disease, and comparing non injured with injured joints, showed no significant difference anywhere in the spectra between opposing joints. The second, observed difference in the degenerative joint disease (DJD) sample were that all FFT's were typically lower and flatter than that found in normal or healthy subjects with acute injury. There appears to be a distinct and significant difference in the EMF spectra of those persons who heal normally and those persons having chronic disease states like DJD. The normal controls heal by elevating the low frequency component of the spectra, whereas chronic disease sufferers are unable to do this. This brings in the possibility of artificially boosting the low end of the spectra of patient's suffering from chronic, unhealed injury. Some preliminary observation would suggest this is a sensible approach in that both pain reduction and increased range of motion have been consistently observed by the author when DJD suffers are given systemic boosts of proper microcurrent.


  1. Becker, R., Cross Currents: the promise of electromedicine and the perils of electropollution.(1990) New York: Putnam.
  2. Luben, R.A., Effects of low-energy electromagnetic fields(pulsed and dc)on membrane signal transduction processes in biological systems.(1991) Health Physics, Vol 61, No.1, pgs 15-28.
  3. Adey, R.W., Electromagnetics in biology and medicine. (1993) In Modern Radio Science, (ed. H. Matsumoto)pgs 277-245. Oxford University Press, Oxford.
  4. Adey, R.W., ELF magnetic fields and promotion of cancer: experimental studies.(1992) In Interaction Mechanisms of low-level Electromagnetic Fields in Living Systems, (eds. B. Norden and C. Ramel)pgs 23-46. Oxford University Press, Oxford.
  5. Heffernan, M., Effects of Variable Microcurrent on EEG Spectra and Pain Control, (1996) in press ISSSEEM.

© 1996 Epress Inc.