Abstract:

The ultimate aim of this research is to elucidate the biological and biophysical processes that subserve the empirical link between environmental electromagnetic fields (EMFs) and human disease. The basic hypothesis is that exposure to environmental EMFs can cause impaired immunosurveillance. The specific hypothesis being tested is that long-term exposure of mice to simulated environmental EMFs will cause changes in the immune system.

Comparisons were planned between the exposed and control groups to permit evaluation of 3 immune-system variables: (1) cellularity in the spleen, thymus, and bone marrow; (2) distribution of lymphocyte sub-populations in the spleen, thymus, and bone marrow; (3) functional activity of cytotoxic T lymphocytes and NK cells. The comparisons were based on changes in both the mean and variance, and were evaluated using the L test.

Homogeneous magnetic fields (±5%) were established using an octapole arrangement of the Merritt 4-coil system. The mice were housed in a totally non-metallic environment, and were maintained in cages that had no direct mechanical path to the field coils, thereby minimizing the potential effect of vibration. Each octapole was operated in series resonance to eliminate the possible confounding effects of line-frequency harmonics. Fourier analysis of the coil current established that the strongest harmonics were more than 50 dB below the fundamental. The exposure and control units were operated at different ends of a dedicated room in an approved animal-care facility. The fringing fields of the exposure units were undetectable at the location of the control units. The ambient 60-Hz field at the control locations averaged less than 0.4 mG, and was never nigher than 0.7 mG. Current, magnetic field, and temperature were continually recorded every 2 minutes to provide a historical record. The room was continuously maintained under temperature and humidity control. The room air was replaced 15 times per hour with fresh air.

Cell sub-populations were identified on the basis of antibody staining and flow cytometry as follows: Spleen, pre-B (CD45), immature B (IgM+/IgD-), mature B (IgM+/IgD+), T (CD90+/CD3+), NK (NK1.1); bone marrow, pre-B, immature B, and mature B; thymus, T, CD4+/CD8-, CD4-/CD8+, and CD4+/CD8+. In vitro allogeneic stimulation was used to induce formation of cytotoxic T lymphocytes (CTL) from spleen cells, and their proliferative response and cytotoxic activity were determined. Spleen cells treated in vitro with IL-2 for 18 hours were used to assess NK-cell cytotoxic activity. Serum corticosterone was measured by radioimmunoassay.

Partial results for lymphoid phenotype in male mice exposed for 1-49 days, 5000 mG, showed that EMF exposure resulted in changes in the distribution of immature B cells in the spleen and bone marrow, and T cells in the spleen and thymus. Under the same exposure conditions, the field had no effect on the distribution of NK cells, the proliferation of CTL, or the lytic activity of CTL or NK cells. Serum corticosterone levels in male mice exposed to 5000 mG, 175 days were significantly different than control levels.

One of the threshold considerations regarding the health and safety issue raised by environmental electromagnetic fields involves the choice of the biological model used to explicate putative EMF/organism interactions. The cybernetic (non-linear) model was chosen in this study for several reasons. First, it is well suited to attempts to resolve the most fundamental issue regarding EMF bioeffects - the occurrence of transduction. Second, the cybernetic model can harmonize reports from different laboratories, thereby resolving the apparent controversies. Third, a living organism is expected on theoretical grounds to exhibit non-linear responses to low-energy stimuli such as environmental EMFs because other conditions besides the EMF are important in determining the response of the animal.

Although important issues are not yet resolved in this study, the results showed that EMFs caused changes in the immune system, possibly mediated by the neuroendocrine system. This conclusion suggests the existence of a specific biological pathway linking environmental EMF exposure and human disease. The biological model employed in this study seems capable of providing a framework for explaining the laboratory and epidemiological reports involving EMF-induced effects.

Experimental Design and Exposure Conditions:

Experimental Hypothesis

The ultimate aim of this research is to elucidate the biological and biophysical processes that subserve the empirical link between environmental electromagnetic fields (EMFs) and human disease. The basic hypothesis is that exposure to environmental EMFs can cause impaired immunosurveillance. The specific hypothesis being tested is that long-term exposure of mice to simulated environmental EMFs will cause changes in the immune system.

Because the particular immune-system changes in laboratory animals that indicate immunosuppression (have functional consequences for the animal with regard to disease endpoints) cannot generally be identified unambiguously, any experimental evidence supporting the specific hypothesis will be interpreted as support for the basic hypothesis. Such evidence will also be taken as proof that EMF signal transduction occurred, thereby implying that both a cellular location and a biophysical process must exist to mediate transduction.

 

Statistical Hypothesis

The choice of the statistical hypothesis was guided by the choice of the biological model used to rationalize inferences of cause-and-effect relationships. We assumed that the response of each EMF-exposed animal in the study was partially determined by the applied EMF, and partially determined by other factors that were not controlled explicitly, and that could never be controlled (because of the complexity of the animal). Consequently, the experimental evidence sought was a consistent occurrence of change in the immune system caused by the EMF, not a consistent pattern of changes in particular dependent variables.

A "change in the immune system" was defined in terms of justifiable rejection of hypotheses of the type (µ1,sS(2,1)) = (µ2,sS(2,2)), where µ1 and sS(2,1) denote the mean and variance in a parameter measured in the EMF-exposed group, and µ2 and sS(2,2) are the corresponding quantities in the control group. A rejection was justifiable when the associated error rates were acceptable (see below).

Comparisons were planned between exposed and control groups to permit evaluation of 3 immune system variables: (1) cellularity in the spleen, thymus, and bone marrow; (2) distribution of lymphocyte sub-populations in the spleen, thymus, and bone marrow; (3) functional activity of cytotoxic T lymphocytes (CTL) and NK cells. The variables were assessed by measuring 3, 12, and 3 parameters, respectively. A procedure was developed to permit statistical testing of the chosen hypotheses based on the use of L, a test statistic that can be expressed as the sum of L1 and L2, the log-likelihood ratio statistics for the variance and mean, respectively (3). The distribution of L is approximately chi-square with 2 degrees of freedom. Thus (µ1,sS(2,1)) = (µ2,sS(2,2)) can be rejected if L > cS(2,L(2,a)) where a is the chosen error rate.

 

Chronicity Criterion

We expected that changes in the immune system would be chronic in the sense that the changes would persist during exposure. Additionally, as mentioned, we expected that the consequences of prolonged EMF exposure would not necessarily be reflected in uniform and consistent changes in the parameters that initially manifested the consequences of EMF exposure. These expectations were reconciled by requiring that a particular parameter be statistically significant at two (or more) consecutive recovery times before rejecting the null hypothesis for the parameter (an "effect"). Thus, two consecutive statistically significant pair-wise comparisons for a particular parameter were accepted as indicating an effect of EMF exposure.

 

Family-Wise Error Rate

Each experiment consists of the exposure of different groups of male or female mice to a magnetic field of given strength for 1, 5, 10, 21, 49, 105, and 175 days. Measurements of 3 + 12 + 3 = 18 immune-system parameters were planned at each recovery time, with the intention to declare an effect whenever two statistically significant pair-wise comparisons (successes) were found. It was therefore necessary to ensure that the family-wise error rate remained within acceptable limits. We chose to do this by adjusting the comparison-wise error rate, q. It can be shown that if q < 0.0285, then the family-wise experimental error is P < 0.05 for the case of 7 recovery times. Thus, in a given experiment, we planned to declare an effect when at least 2 consecutive successes were observed at P < 0.0285 in any parameter.

Replicates

Because of practical limitations regarding the number of mice that could be measured on a given day, 3 replicates of each combination of exposure conditions were planned (generally, N=5 in the exposed and in the control group in each replicate), with the intention of combining the data for a given set of conditions in one overall test. For 3 replicates, overall L (the sum of the L values from each replicate) is approximately chi-square with 6 degrees of freedom under the hypothesis of no treatment effect. Overall L1 and L2 are also approximately chi-square, each with 3 degrees of freedom.

 

Methods

Cell sub-populations were identified on the basis of antibody staining and flow cytometry as follows: spleen, pre-B (CD45), immature B (IgM+/IgD-), mature B (IgM+/IgD+), T (CD90+/CD3+), NK (NK1.1); bone marrow, pre-B, immature B, and mature B; thymus, T, CD4+/CD8-, CD4-/CD8+, and CD4+/CD8+. In vitro allogeneic stimulation was used to induce formation of cytotoxic T lymphocytes (CTL) from spleen cells, and their proliferative response and cytotoxic activity were determined. Spleen cells treated in vitro with IL-2 for 18 hours were used to assess NK-cell cytotoxic activity. Serum corticosterone was measured by radioimmunoassay.

Quality Assurance Measures:

Homogeneous magnetic fields (±5%) were established using an octapole arrangement of the Merritt 4-coil system. The mice were housed in a totally non-metallic environment, and were maintained in cages that had no direct mechanical path to the field coils, thereby minimizing the potential effect of vibration. Each octapole was operated in series resonance to eliminate the possible confounding effects of line-frequency harmonics. Fourier analysis of the coil current established that the strongest harmonics were more than 50 dB below the fundamental.

The exposure and control units were operated at different ends of a dedicated room in an approved animal-care facility. The fringing fields of the exposure units were undetectable at the location of the control units. The ambient 60-Hz field at the control locations averaged less than 0.4 mG, and was never nigher than 0.7 mG. Current, magnetic field, and temperature were continually recorded every 2 minutes to provide a historical record. The room was continuously maintained under temperature and humidity control. The room air was replaced 15 times per hour with fresh air.

Discussion and Contribution to Under-standing Biological Effects of EMF:

The health and safety issue raised by environmental electromagnetic fields is insusceptible of resolution until several scientific, sociological, and procedural issues are resolved first. One of the threshold scientific issues involves the choice of the biological model used to explicate putative EMF/organism interactions. A physical model, by definition, is based on the assumption that the strength of the field principally determines the mean of the measured parameter, independent of the animal's genotype, history, and environment. The physical model is used, for example, to predict the toxic effects of chemicals.

A cybernetic model, in contrast, is defined as one based on the assumption that the instantaneous behavior of the organism is a result of the totality of the conditions present at the immediately preceding instant of time (27). Because the EMF is only one such condition, it (at most) only partially determines what can be observed. The biological variability ordinarily found in an ostensibly identical group of animals exemplifies application of the cybernetic model; although each measurement is determined and bounded, it is not predictable. The cybernetic model requires a reconceptualization of EMF/tissue interactions leading to acceptance of the idea that observables are (at most) only partially predictable, and need not fulfill a subjective requirement of consistency.

The cybernetic model was chosen in this study for several reasons. First, the seminal issue in all EMF bioeffects studies is whether or not EMF transduction occurred. An effect rationalized using the physical model is sufficient but not necessary to imply transduction because the physical model is derived from the cybernetic model in the limit. Thus, the absence of an effect based on use of the physical model would not imply that transduction did not occur. In contrast, in the cybernetic model, the absence of a change in the response of the organism to an EMF would reasonably imply that transduction did not occur. Thus the cybernetic model is singularly suited to attempts to resolve the most fundamental (and divisive) issue regarding EMF bioeffects - the occurrence of transduction.

Second, among the studies in which the physical model has been used, virtually every report of an EMF-induced bioeffect has been followed (or preceded) by a negative report involving a reasonably comparable biological system and the evaluation of a reasonably comparable experimental hypothesis. As a result, not a single EMF-induced bioeffect is presently accepted in the sense that general relativity, the theory of radioactive decay, or the law of gravity are accepted. On the other hand, in the case of one line of studies (those involving EMF-induced effects on growth rate) adoption of the cybernetic model led to a consistent answer to the question of whether the field actually caused biologic effects (7). The fact that the cybernetic model harmonized the implications of reports from different laboratories suggests that conflicts which flow from the use of the physical model are artifacts produced by the model, and that adoption of the cybernetic model would resolve the apparent controversies.

Finally, a living organism is perhaps the best possible example of a far-from-equilibrium system, and it is to be expected that responses to relatively low-energy stimuli will be non-linear in nature (because other conditions besides the EMF are important in determining the response of the animal). Indeed, it is probably the case that (with the exception of studies involving the measurement of parameters associated with immediate early stages of transduction) any EMF-induced biological effects that are not non-linear are either trivial or unrelated to the fundamental issue of health risks due to environmental electromagnetic fields. Thus, on theoretical grounds, only the cybernetic (non-linear) model is applicable to effects caused by environmental EMFs.

Although important issues are not yet resolved in this study, the results showed that EMFs can cause changes in the immune system, possibly mediated by the neuroendocrine system. This conclusion suggests the existence of a specific biological pathway linking environmental EMF exposure and human disease, and the model on which it was based seems capable of providing a framework for explaining the laboratory and epidemiological reports involving EMF-induced effects.

Results:

Partial results for male mice exposed for 1-49 days, 5000 mG, are shown in Figure 1. EMF exposure resulted in changes in the distribution of each of the listed sub-populations. Under the same exposure conditions, the field had no effect on the distribution of NK cells, the proliferation of CTL, or the lytic activity of CTL or NK cells.

 

The results of corticosterone determinations in the serum of mice exposed to 5000 mG, 175 days are shown in Figure 2.

The corresponding calculation for the L statistic (Table 1) clearly illustrates how test means can differ differently in each replicate and yet combine to evidence transduction.

To demonstrate the ability of the L test to discern differences between non-linear systems, the logistic equation was used to generate two non-linear data sets, each of which was sampled 15 times in 3 replicates (Figure 3).

The L test recognized the difference between the populations (Table 2) even though the groups in individual replicates differed differently (Figure 3).

Other EMF Publications:

1. Low-frequency electromagnetic fields alter the replication cycle of MS2 bacteriophage, J. Staczek, A.A. Marino, L.B. Gilleland, A. Pizarro and H.E. Gilleland, Jr. Current Microbiology, In Press.

2. Electromagnetic fields can affect osteogenesis by increasing the rate of differentiation. P.S. Landry, K.K. Sadasivan, A.A. Marino and J.A. Albright. Clin. Orthop. 338:262-270, 1997.

3. Electromagnetic fields enhance chemically-induced hyperploidy in mammalian oocytes. J.B. Mailhes, D. Young, A.A. Marino and S.N. London. Mutagenesis 12:347-351, 1997.

4. Piezoelectricity in the human pineal gland. S.B. Lang, A.A. Marino, G. Berkovic, M. Fowler and K.D. Abreo. Bioelectrochem. Bioenerg. 41:191-195, 1996.

5. Low-level EMFs are transduced like other stimuli. A.A. Marino, G.B. Bell and A. Chesson. J. Neurolog. Sci. 144:99-106, 1996.

6. Electromagnetic fields in the classroom. A.A. Marino. in The Healthy School Handbook, N. Miller, ed., NEA Professional Library, Washington, DC, 221-241, 1995.

7. Different outcomes in biological experiments involving weak EMFs: Is chaos a possible explanation? A.A. Marino. Am. J. Physiol. 268 (Regulatory Integrative Comp. Physiol. 37: R1013-R1018, 1995.

8. Time-dependent hematological changes in workers exposed to electromagnetic fields. A.A. Marino. Am. Ind. Hyg. Assoc. J. 56:189-192, 1995.

9. Electrical potential measurements in human breast cancer and benign lesions. A.A. Marino, D.M. Morris, M.A. Schwalke, I.G. Iliev & S. Rogers. Tumor Biology 15:147-152, 1994.

10. Association between cell membrane potential and breast cancer. A.A. Marino, I.G. Iliev, M.A. Schwalke, E. Gonzalez, K.C. Marler & C.A. Flanagan. Tumor Biology 15:82-89, 1994.

11. Frequency-specific responses in the human brain caused by electromagnetic fields. G.B. Bell, A.A. Marino & A.L. Chesson. J. Neurol. Sci. 123:26-32, 1994.

12. Frequency-specific blocking in the human brain caused by electromagnetic fields. G.B. Bell, A.A. Marino & Andrew L. Chesson. NeuroReport 5:510-512, 1994.

13. Electromagnetic fields, cancer, and the theory of neuroendocrine-related promotion. A.A. Marino. Bioelectrochem. Bioenerg. 29:255-276, 1993.

14. Alterations in brain electrical activity caused by magnetic fields: detecting the detection process. G.B. Bell, A.A. Marino & A.L. Chesson. Electroencephalog. Clin. Neurophysiol. 83:389-397, 1992.

15. Electrical states in the rabbit brain can be altered by light and electromagnetic fields. G. Bell, A.A. Marino, A. Chesson & F. Struve. Brain Res. 570:307-315, 1992.

16. Human sensitivity to weak magnetic fields. G. Bell, A.A. Marino, A. Chesson & F. Struve. Lancet 338:1521-1522, 1991.

17. Environmental electromagnetic fields and public health. A.A. Marino. in Foundations of Modern Bioelectricity, A.A. Marino, ed., Marcel Dekker, New York, 965-1044, 1988.

18. Foundations of Modern Bioelectricity. A.A. Marino, ed. Marcel Dekker, New York, 1988.

19. Are powerline fields hazardous to health? A.A. Marino. Public Power 45:1820, 1987.20. Health risks from electric power facilities. A.A. Marino. in Proceedings of International Utility Symposium, Health Effects of Electric and Magnetic Fields, Ontario Hydro, Toronto, 1986.

20. Health risks from electric power facilities. A.A. Marino. in Proceedings of International Utility Symposium, Health Effects of Electric and Magnetic Fields, Ontario Hydro, Toronto, 1986.

21. Electric Wilderness. A.A. Marino and J. Ray. San Francisco Press, San Francisco, 1986.

22. Chronic electromagnetic stressors in the environment: A risk factor in human cancer. A.A. Marino & D.M. Morris. J. Environ. Sci. C3(2):189-219, 1985.

23. Electromagnetic fields and public health. A.A. Marino. in Assessments and Viewpoints on the Biological and Human Health Effects of Extremely Low Frequency Electromagnetic Fields, American Institute of Biological Sciences, Arlington, Va., 205-232, 1985.

24. Weak electrical fields affect plant development. A.A. Marino, F.X. Hart & M. Reichmanis. IEEE Trans. Biomed. Eng. BME 30: 833-834, 1983.

25. Bioelectric considerations in the design of high-voltage power lines. M. Reichmanis & A.A. Marino. J. Bioelectricity 1: 329-338, 1982.

26. ELF dosage in ellipsoidal models of man due to high-voltage transmission lines. F.X. Hart & A.A. Marino. J. Bioelectricity 1: 129-154, 1982.

27. Electromagnetism & Life. R.O. Becker and A.A. Marino. State University of New York Press, Albany, 1982, p. 160.

28. Environmental power-frequency magnetic fields and suicide. F.S. Perry, M. Reichmanis, A. Marino, & R. Becker. Health Phys. 41: 267-277, 1981.

29. Sensitivity to change in electrical environment: a new bioelectric effect. A.A. Marino, J.M. Cullen, M. Reichmanis, R.O. Becker & F.X. Hart. Am. J. Physiol. 239 (Regulatory Integrative Comp. Physiol. 8): R424-427, 1980.

30. Power frequency electric field induces biological changes in successive generations of mice. A.A. Marino, M. Reichmanis, R.O. Becker, B. Ullrich & J.M. Cullen. Experientia 36: 309-311, 1980.

31. Effect of electrostatic fields on the chromosomes of Ehrlich ascites tumor cells exposed in vivo. J.T. Mitchell, A.A. Marino, T.J. Berger & R.O. Becker. Physiol. Chem. Phys. 10: 79-85, 1978.

32. High voltage lines: hazard at a distance. A.A. Marino & R.O. Becker. Environment 20(9): 6-15, 1978.

33. Power frequency electric fields and biological stress: a cause and effect relationship. A.A. Marino, J.M. Cullen, M. Reichmanis & R.O. Becker. Proc. 18th Ann. Hanford Life Sciences Symposium, Dept. of Energy, Washington, D.C., 1978.

34. Electromagnetic pollution. R.O. Becker & A.A. Marino. The Sciences, January, 1978, pp. 14, 15, 23.

35. In vivo bioelectrochemical changes associated with exposure to ELF electric fields. A.A. Marino, T.J. Berger, B.P. Austin, R.O. Becker & F.X. Hart. Physiol. Chem. Phys. 9: 433-441, 1977.

36. Biological effects of extremely low frequency electric and magnetic fields: a review. A.A. Marino & R.O. Becker. Physiol. Chem. Phys. 9: 131-147, 1977.

37. Energy flux along high voltage transmission lines. F.X. Hart & A.A. Marino. IEEE Trans. Biomed. Eng. BME-24: 493-495, 1977.38. Biophysics of animal response to an electrostatic field. F.X. Hart & A.A. Marino. J. Biol. Phys. 4: 123-143, 1976.

38. Biophysics of animal response to an electrostatic field. F.X. Hart & A.A. Marino. J. Biol. Phys. 4: 123-143, 1976.

39. The effect of continuous exposure to low frequency electric fields on three generations of mice: a pilot study. A.A. Marino, R.O. Becker & B. Ullrich. Experientia 32: 565, 1976.

40. Electric field effects in selected biologic systems. A.A. Marino, T.J. Berger, J.T. Mitchell, B.A. Duhacek & R.O. Becker. Ann. N. Y. Acad. Sci. 238: 436-444, 1974.

Publications from this grant:

1. Electromagnetic fields enhance chemically-induced hyperploidy in mammalian oocytes. J.B. Mailhes, D. Young, A.A. Marino and S.N. London. Mutagenesis 12:347-351, 1997.

2. Low-level EMFs are transduced like other stimuli. A.A. Marino. XXXIII International Congress of Physiological Sciences, St. Petersburg, Russia, L047.07 (Abstract), 1997.

 Marino Home Page | Research Interests| Electromagnetic Fields

32696