Content uploaded by Toshikazu Shinba
Author content
All content in this area was uploaded by Toshikazu Shinba on Feb 26, 2021
Content may be subject to copyright.
IEEJ TRANSACTIONS ON ELECTRICAL AND ELECTRONIC ENGINEERING
IEEJ Trans 2021
Published online in Wiley Online Library (wileyonlinelibrary.com). DOI:10.1002/tee.23334
Paper
Extremely Low-Frequency Electric Field Exposure Increases Theta Power of
EEG in both Eyes-Open and Eyes-Closed Resting Conditions in Healthy
Male Subjects
Toshikazu Shinba*a,Non-member
Takaki Nedachi**,Non-member
Shinji Harakawa**,***,Member
The effects of extremely low-frequency (ELF) electric field (EF) exposure on brain activity remain poorly understood. This study
sought to examine the acute effects of EF exposure on brain activity by analyzing electroencephalogram (EEG) basic rhythms.
The heart rate variability (HRV) was also measured to evaluate autonomic activity. Using a system that enables measurement
under EF exposure, occipital EEG and electrocardiogram signals were recorded in nine healthy male subjects during daytime in
eyes-open and eyes-closed conditions, with or without EF exposure (applied voltage: 30 kV, 50 Hz for 1 min). The EEG power
and frequency-domain HRV indices were obtained by power spectrum analysis of EEG and the RR interval trend, respectively.
Resulting data revealed that the theta power of EEG was significantly greater during EF exposure in eyes-open and eyes-closed
conditions. Other EEG rhythms and HRV indices were not affected by EF exposure. These findings suggest that exposure to
50-Hz EF for 1 min altered brain activity by increasing the theta power. ©2021 Institute of Electrical Engineers of Japan.
Published by Wiley Periodicals LLC.
Keywords: extremely low-frequency electric field; brain activity; electroencephalogram; theta wave; heart rate variability
Received 14 August 2020; Revised 13 December 2020
1. Introduction
Exposure to a high-voltage electric field (EF) at various fre-
quency ranges is known to cause biological effects [1– 3]. How-
ever, the effects of an extremely low-frequency EF (ELF-EF)
exposure have not been completely clarified. Our previous studies
have shown that ELF-EF exposure can affect stress and pain con-
trol [4–8]. Harakawa et al. [4] reported that a 1 h EF exposure
significantly attenuates restraint-induced elevation of the plasma
adrenocorticotropic hormone (ACTH) and glucose in rats. Fur-
ther, Hori et al. [7] and Harakawa et al. [6] reported that EF
exposure has a similar suppressive effect on restraint-induced ele-
vation of plasma glucocorticoid levels in mice. Harakawa et al. [5]
also reported that, in dogs with tumors or spinal cord injuries, the
plasma ACTH and beta-endorphin levels decrease and increase,
respectively, after 2 h of EF exposure. Reduction of ACTH, glu-
cose, and glucocorticoid levels, and increase in endorphin levels, is
related to the diminution of stress and pain, respectively. A study
aCorrespondence to: Toshikazu Shinba. E-mail:
t156591@siz.saiseikai.or.jp
*Department of Psychiatry, Shizuoka Saiseikai General Hospital, 1-1-1,
Oshika, Suruga-ku Shizuoka 422-8527, Japan
**Hakuju Institute for Health Science Co., Ltd., 1-37-5, Tomigaya,
Shibuya-ku Tokyo 151-0063, Japan
***Bio-Self-Regulating Science Laboratory, Obihiro University of Agricul-
tural and Veterinary Medicine, Inada-cho Obihiro, Hokkaido 080-8555,
Japan
using a visual analog scale also showed that a 20-min EF exposure
reduced the severity of pain in patients with chronic pain [8].
In addition to these effects of ELF-EF on hormonal and sensory
systems, changes in brain activity by EF are of interest, in terms of
EF effects on the neuronal activity related to these systems. When
the EF affects an electroencephalogram (EEG), modulation of
brain activity by the EF may lead to changes in neurophysiological
and psychological functions. However, there are only a few studies
reporting EF effects on EEG. Yamashita et al. [9,10] performed an
EEG during the eyes-closed condition and showed that a 30-min
ELF-EF exposure decreased the EEG total power and alpha ratio.
The results indicated that brain activity is modulated by ELF-EF.
It is known that the EEG profiles are different between the
periods of the eyes-closed and eyes-open conditions. It would be
interesting to further examine the EF effects on EEG during the
eyes-closed and eyes-open conditions. EEG profiles also change
depending on the arousal level of the subject. It is also important to
exclude the influence of arousal fluctuation related to experimental
procedures and to reveal the EF effects.
Based on this background, in this study on healthy subjects, the
effects of ELF-EF on EEG were analyzed during the eyes-open
and eyes-closed conditions. The duration of a single recording
was set to 1 min to reduce the effect of arousal on the results, and
the recordings with the same conditions were repeated separately
three times.
In addition, we also conducted an experiment on a phantom
designed especially for the apparatus used in this study, and
measured the EFs and induced currents in the same situation as the
©2021 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.
T. SHINBA, T. NEDACHI, AND S. HARAKAWA
experiment for human subjects. EFs act on the skin surface [11]
and may also act directly on the brain. By using data of the EFs and
induced currents on the phantom, the EF data could be discussed
in terms of the EF influence on brain activity of human subjects.
2. Materials and Methods
2.1. Ethical approval Regarding the EF exposure, the
standard regimen, which was approved by the Japanese Govern-
ment, Ministry of Health, Welfare and Labor, was used. This
research was approved by the Ethics Review Board of Tokyo Insti-
tute of Psychiatry.
2.2. Subjects Nine healthy men (mean age ±standard
deviation: 48.6 ±10.4 years, ranging from 31 to 62) were included
in this study. The subjects were recruited from among the members
of Hakuju Institute for Health Science and were consecutively
assigned to this study. Written informed consent was obtained from
all subjects.
2.3. EF exposure For EF exposure, the subjects were
seated in a chair connected to an EF generator (HES-A30; Hakuju
Institute for Health Science, Tokyo, Japan; Fig. 1). Two electrodes
Lower
electrode
Amplier
and recorder in
a conductive
mesh bag
Upper
electrode
Electric eld
generator
Fig. 1. Experimental setup of the electric field exposure system.
An electric field was generated between two electrodes. One
was placed over the head and under the feet, respectively. The
applied voltage was 30 kV. The participant was seated with the
recording system in a conductive mesh bag on his lap during
the electric field exposure. The recorded data were analyzed off-
line after the recording was completed. Consent to publish the
photograph during the electric field exposure was obtained from
the participant. The Hakuju Institute for Health Science agreed to
publish this photograph containing their electric field generator in
this publication.
were attached to the chair, one above the head (upper electrode)
and one below the feet (lower electrode). A plastic plate was placed
between the feet and the lower electrode, and the electrodes did
not directly contact the body of the subject. From the generator, a
50-Hz electric potential of 30-kV root mean square amplitude was
applied to the lower electrode, with the upper electrode serving as
the ground for electrical charges.
2.4. Experimental protocol for human subjects In
human subjects, each measurement was conducted for 12 min
during daytime and consisted of 12 1-min periods. There were
four patterns of eye closure and EF exposure: 1) eyes-open and
EF-off, 2) eyes-open and EF-on, 3) eyes-closed and EF-off, and 4)
eyes-closed and EF-on. Each participant underwent each pattern
three times (patterns A-D, presented in Table I). The EF-on and
EF-off patterns were scheduled alternately to limit the time of each
EF exposure to 1 min. The patterns with the eyes-open and eyes-
closed conditions were ordered randomly. The instruction to open
or close the eyes was provided orally by an experimenter.
During the period of EF exposure, an electric potential sinu-
soidally alternating at 50 Hz with a root mean square amplitude of
30 kV was applied to the lower electrodes of the chair for 1 min.
The subjects were unaware of the shifts from EF-off to EF-on or
vice versa, which were performed by an experimenter. The EEG
and electrocardiogram (ECG) signals during the 12 periods were
recorded consecutively during the experiment. White noise and
wind were generated during the experiment using a speaker and
a fan, respectively, to reduce the effects of noise from the EF
generator and that of body hair vibration caused by the EF.
2.5. EEG and ECG recordings The EEG and ECG
signals were recorded by the Ag/AgCl electrode at the occipital
lead (Oz of the international 10–20 system) and the participant’s
chest, respectively. Oz was chosen to examine the basic rhythm of
the brain including the alpha wave [12]. The reference Ag/AgCl
electrodes for EEG and ECG were placed on the left and right
ear lobes, respectively. The electrodes were round with an 8-
mm diameter. The lead wires were connected to a battery-
powered amplifier (BA-1010; Miyuki, Tokyo, Japan). The data
were amplified 100 times with a 0.3-s time constant and a high-
cut filter set at 30 Hz, and stored in a battery-powered data recorder
at a sampling frequency of 1 kHz for offline analyses (es8; TEAC,
Tokyo, Japan). The amplification was set at 100 times because a
greater amplification would cause the data during EF exposure to
exceed the ±5 V input range of the data recorder.
The EEG and ECG electrodes were covered by electrical
shielding caps, 38-mm in diameter. The lead wires were also
Table I. Protocol for the electric field treatment
Order 123456789101112
Min 111111111111
EF Off On Off On Off On Off On Off On Off On
Eyes OOCCOCCOCCOO
PatternABCDADCBCDAB
Four patterns (A, B, C, D) of 1-min recording with respect to the EF exposure (on and off) and eye condition (open and closed). Each pattern was repeated
thrice in a session comprising a total of 12 periods. The eyes-open and eyes-closed conditions were ordered randomly. The EF-on and EF-off patterns were
scheduled alternately to limit the time of single exposure to 1 min.
C, closed; EF, electric field; O, open.
2IEEJ Trans (2021)
ELECTRIC FIELD EXPOSURE INCREASES THETA POWER OF EEG
(a)
(b)
1 s
3 mV
1 s
3 mV
Filter OFF
Filter ON
Filter OFF
Filter ON
Electric field
*
*
Electric field
Fig. 2. Effect of the digital filter. Effect of the digital filter on
(a) EEG and (b) ECG recordings during EF exposure. In the
absence of a filter, a 50-Hz noise waveform hid the EEG and ECG
signals. The filter could remove all noise, except at the beginning
(*) and end (not shown in this figure) of the EF exposure. EEG,
electroencephalogram; ECG, electrocardiogram; EF, electric field
electrically shielded. The amplifier and data recorder were shielded
by an electrically conductive carbon mesh, with a thickness of
0.25 mm (TORAYCA™cloth; Toray, Tokyo, Japan). The shielding
was connected to the ground electrode of the amplifier set on
the human body. The amplifier and data recorder were placed on
participant’s lap during the measurement.
2.6. Data analysis The EEG and ECG data were first
processed using a digital filter to remove the artifacts caused by EF
exposure (Vital Tracer; Kissei Comtec, Matsumoto, Japan). Notch
and bandpass filters were set at 50 and 0.5–30 Hz, respectively.
Using this procedure, the EEG rhythms and ECG R peaks were
clearly observed during the EF exposure (Fig. 2).
The processed EEG data of each 1-min period were analyzed
using a fast Fourier transform with a Hanning window to obtain
the power spectrum (BIMUTAS II, Kissei Comtec). The data at
the beginning and end of the EF exposure had artifacts due to the
EF (Fig. 2, asterisk) and were manually omitted from analysis. The
delta, theta, alpha, and beta ranges were set at 0.5–4, 4–8, 8 –13,
and 13–30 Hz, respectively, and the integrated area corresponding
to these frequency ranges in the power spectrum were used for
0.2 a.u.
0.3 sec
Filter OFF
Filter ON
Signal without noise
Signal with noise
Fig. 3. Influence of filtering on signals. The influence of filtering
on the sine wave with a peak-to-peak amplitude of 0.02 a.u. (signal
without noise) and the sum of two sine waves, with peak-to-peak
amplitudes of 0.02 and 2.0 a.u., respectively (signal with noise).
More details are provided in Section 2. a.u., arbitrary unit
analysis. The sum of the four scores was used to represent the
total EEG power.
In the ECG data, R peaks were automatically detected using
specialized software (BIMUTAS II). When artifacts were observed
by the experimenter, they were removed and manually replaced
by the averaged data of the preceding and following intervals
using the software. Then, the RR intervals were resampled at the
frequency of the average heart rate to generate the RR interval
trend graph. The maximum entropy method was used to obtain
the power spectrum of the RR interval trend graph (BIMUTAS II).
The maximum entropy method was employed in this study, as it
has been successfully used in previous works that performed heart
rate variability (HRV) analysis with limited data points, as short as
30 s [13]. Following the conventional method of frequency-domain
HRV analysis [14], the low-frequency (LF) and high-frequency
(HF) HRV components were calculated as integrated powers
(ms2Hz) at the ranges of 0.04–0.15 and 0.15–0.4 Hz, respectively.
The LF and HF values corresponded to blood pressure-related
autonomic [15] and respiration-related parasympathetic activities
[16], respectively. LF/HF is considered to be related to sympathetic
activity. The respiration rate was determined by observation of
chest movements and was monitored by an experimenter. It was
confirmed to be within the range of 0.15–0.4 Hz.
2.7. Influence of filtering on signals The influence of
filtering in this study on EEG was determined in an additional
experiment using two sine waves (Excel, Microsoft Corp., Red-
mond, Washington, USA), with peak-to-peak amplitudes of 0.02
arbitrary units (a.u.) and 2.0 a.u., respectively (Fig. 3). The former
wave simulated the EEG wave recorded on the head (signal), and
the latter stimulated the noise accompanying the ELF-EF exposure
(noise). The frequency of noise was 50 Hz, which was the same
as that of EF in the actual experiment with human subjects. The
difference in amplitudes between the two sine waves was deter-
mined based on the difference observed after performing the actual
measurement in human subjects (approximately 1:100).
Four types of signals of different frequencies were used (3, 6,
10, and 20 Hz) with a duration of 10 s and used for simulating EEG
recording without EF (signal without noise, Fig. 3). The sum of
signal (10 s) and noise (10 s) durations was also used for simulating
EEG recording during the EF treatment (signal with noise, Fig. 3).
Theoretically, the power of signal was 5.0 ×10−5a.u.2.
Both signals with and without noise were processed by the
same filtering and analysis system as described above (Fig. 3).
The delta, theta, alpha, and beta power of signals with and
3IEEJ Trans (2021)
T. SHINBA, T. NEDACHI, AND S. HARAKAWA
without noise were 1.651 ×10−5and 1.650 ×10−5a.u.2for
the 3-Hz signal, 1.903 ×10−5and 1.903 ×10−5a.u.2for the
6-Hz signal, 1.903 ×10−5and 1.902 ×10−5a.u.2for the 10-
Hz signal, and 1.901 ×10−5and 1.901 ×10−5a.u.2for the
20-Hz signal, respectively. The differences between the signals
with and without noise were 0.06% and 0.05% for the 3-Hz
and 10-Hz data, respectively. Regarding the 6-Hz and 20-Hz
data, the difference was smaller than the limit of the analysis
software.
2.8. Statistical analysis Twelve sets of EEG and HRV
data obtained from the 1-min periods for each participant were
divided into four groups of three data sets each, as follows:
eyes-open and EF-on, eyes-open and EF-off, eyes-closed and
EF-on, and eyes-closed and EF-off. For the statistical analysis,
the median values of each group were used as opposed to the
mean values, as the sample size was small. To consider the
interindividual differences, a repeated-measure two-way analysis
of variance (ANOVA) was performed to determine the effects
of the EF and eye condition on the EEG and HRV data (Prism
5, GraphPad Software, La Jolla, California, USA). The level of
statistical significance was set to p<0.05.
2.9. Dosimetry with a phantom For simulating the
EF distribution on the body surface and current induction in the
human body, dosimetry with the present EF generation system
was conducted independently using a phantom exposed to the
same EF, as described in the EF exposure subsection (Section 2.3)
(Fig. 4). This dosimetry system was originally developed by our
collaborators [17,18] and modified for our study.
The chair connected to the EF generator was placed in an exper-
imental chamber. The temperature and humidity were maintained
at 20◦C and 40%, respectively. The distances from the chair to
the rear and lateral walls were 40 and 80 cm, respectively. The
distance between the high-voltage main unit and the chair was set
to more than 80 cm.
A hollow plastic manikin (P11 basic patient care simulator; 3B
Scientific, Hamburg, Germany; height: 174 cm, weight: 14 kg) was
used as the phantom and seated on the chair (Fig. 4(a)). The
distance between the upper electrode and the parietal region of
the head of the manikin was 16 cm. The manikin was painted
with a conductive material on the surface and sliced horizontally
at 11 planes of the body. These cross-sectional gaps between the
adjacent regions were connected via short-circuit switches inside
the phantom, and the switch was turned off when the induced
current was measured at that plane (Fig. 4(c)). The resistance
between the parts of the surface was below 1 k.
The EF on the body surface was measured by a sensor electrode
composed of three layers of a 35-μm thick flexible film. The outer
and inner copper layers (40 and 57 mm in diameter, respectively)
were separated by a layer of polyimide insulator intermediate
(Hakuju Institute for Health Science; Fig. 4(a)). As represented
in Fig. 4(a) and (b), the sensor electrode was placed on the
parietal region of the head. The inner layer was attached to the
phantom’s surface, and the outer layer was connected to the
ammeter (Fig. 4(b)). For the EF measurement on the body surface,
all short-circuit switches were turned on, and every part of the
body’s surface was in electrical continuity (Fig. 4(b)). The EF
was measured with the sensor electrode at 15 points on the body’s
surface, as shown in Fig. 5(a). The induced currents flowing from
(a)
(b)
(c)
Sensor electrode
Induced
current
Conductive
painting
I/V
converter
Optical analogue
data link
(Transmitter)
Short
circuit
Optical ber cable
Phantom
Head
To receiver
Induced current
Sensor electrode
(at the parietal region)
Cross-sectional gap
Optical ber cable
To receiver
Conductive
painting
Head
Phantom
Short
circuit
I/V
converter
Optical analogue
data link
(Transmitter)
Fig. 4. Measurement systems for dosimetry. (a) Photograph of
the manikin used as the phantom for dosimetry. The head and
the right shoulder are shown. The arrow indicates the sensor
electrode placed on the parietal region of the head. (b) Structure of
the phantom with the sensor electrode on the parietal region of the
head. The sensor electrode is composed of outer and inner copper
layers. The circle indicated by ‘A’ signifies the ammeter. Other
details can be found in Section 2. (c) Structure of the phantom for
induced current measurement. Details are provided in Section 2.
the outer layer of the sensor to the surface of the body were
converted to voltage by an I/V converter (LI-76; NF, Yokohama,
Japan; Fig. 4(b)), and the voltage signals were transferred to the
transmitter of the optical analog data link (PE-18 type; OPTEX,
Shiga, Japan; Fig. 4(b)). Next, through a 10-m optical fiber cable
(OCW-103PC; OPTEX; Fig. 4(b)), the data were further sent to the
4IEEJ Trans (2021)
ELECTRIC FIELD EXPOSURE INCREASES THETA POWER OF EEG
receiver outside the EF exposure system, where frequency analysis
was performed by a fast Fourier transform analyzer (HC8000;
Graphtec, Yokohama, Japan). The EF strength (root mean square)
was calculated by (1) and (2). At the points of contact between
the sensor and the surrounding surface (the footrest and headrest
composed of several materials with different relative permittivity
[εr] such as acryl [3.3–3.9] [19], dry wood [2, 3] [20], and
polyurethane [6–8] [19]; the εrof polyurethane is decreased when
it is processed into foam), ε0was replaced by 3ε0,astheεrof such
materials was approximately 3. The distribution of the surface EF
is represented in Fig. 5(a).
E=I
ωε0S(1)
where Eis the electric field, Iis the induced current, ε0is the
permittivity of the vacuum, ωis the angular frequency, Sis the
surface area of the sensor.
E=E2
1+E2
3+E2
5+··· (2)
where Eis the electric field, E1is the fundamental harmonic, E3,
E5,...: is the higher harmonics.
Induced currents in the phantom were evaluated at 11 cross
sections, as shown in Fig. 5(b). The short-circuit switch was
turned off at the corresponding cross section for the current
measurement (Fig. 4(c)). The data processing was the same as that
for the surface EF. The induced current densities were calculated
by dividing each current by the corresponding cross-sectional
area.
3. Results
3.1. Effects of EF exposure in human subjects The
results are summarized in Table II. Two-way ANOVA revealed that
EF has an effect on theta power (F[1,16] =6.389, p=0.022),
which was greater during the EF-on period than during the
EF-off period. Eye condition had an effect on alpha power
(F[1,16] =10.31, p=0.005), which was greater during the eyes-
closed than during the eyes-open periods, consistent with previous
conventional findings on the alpha wave [12]. The EF or eye
condition had no effect for other EEG or HRV indices (p>0.05).
Further, there was no interaction effect between the EF and eye
condition for any EEG or HRV indices (p>0.05).
3.2. Dosimetry data in the phantom Results of the
dosimetry experiment are presented in Fig. 5. In this measurement,
E1in (2) was dominant, and higher harmonics were less than 1/18
of E1, at most. Thus, remarkable harmonics were not observed.
The scores of electric fields on the body surface (external EF) and
the distribution of induced currents and induced current densities
are shown in Fig. 5(a)–(c), respectively. When the human body
was assumed to be homogeneous, the induced EFs inside of the
phantom were calculated by (3). When the conductivity of the
body was set at 0.2 S/m [21], the induced current densities were
1.0–54mA/m2(Fig. 5(c)), corresponding to the induced EFs of
5 –270 mV/m.
Ei =J
σ(3)
where Ei is the induced electric field, Jis the induced current
density, σis the conductivity.
(a)
(b)
(c)
106
222
62*
25
49
67
48
157
51
97
67
32
56
1073*
47
25
38
33
73
78
238
247
154
64
170
181
1.0
4.4
1.9
1.6
17
5.9
7.1
9.6
28
22
54
Fig. 5. Dosimetry data. (a) Electric fields on the body surface
measured at 15 points on the phantom’s surface. Corrected values
due to contact were calculated using 3ε0in (1) instead of ε0and are
indicated by the asterisks (*). (b), (c) Induced currents and induced
current densities measured at 11 cross sections of the manikin.
4. Discussion
4.1. EEG change by ELF-EF exposure This study
revealed that 50-Hz EF exposure with an applied voltage of 30 kV
for 1 min increases theta power in eyes-open and closed conditions,
suggesting that ELF-EF affects brain activity. Other frequency
bands of delta, alpha, and beta ranges were not altered, indicating
5IEEJ Trans (2021)
T. SHINBA, T. NEDACHI, AND S. HARAKAWA
Table II. Effects of EF treatment on EEG and HRV indices
Eyes-open Eyes-closed
EF-off EF-on EF-off EF-on
EEG
Delta μV29.46 (8.94) 9.89 (10.12) 5.20 (1.61) 5.33 (2.15)
Theta μV27.06 (4.01) 8.42 (5.20)a7.71 (4.21) 8.61 (3.95)a
Alpha μV26.30 (2.70) 6.73 (3.40) 18.73 (9.68)b22.32 (17.32)b
Beta μV29.86 (3.72) 11.17 (6.05) 12.11 (5.77) 14.00 (7.90)
Total μV232.68 (16.08) 36.20 (18.81) 43.75 (18.11) 50.26 (26.25)
HRV
LF ms2Hz 217.7 (126.7) 410.7 (504.5) 256.9 (279.4) 219.8 (144.5)
HF ms2Hz 297.2 (337.0) 231.4 (199.9) 252.2 (234.7) 186.5 (178.1)
LF/HF 2.03 (2.94) 2.50 (2.34) 2.11 (2.03) 3.43 (5.17)
HR/min 69.9 (7.3) 70.0 (6.7) 69.5 (7.3) 70.2 (6.9)
Data are presented as means (standard deviations).
EF, electric field; EEG, electroencephalography; HRV, heart rate variability; LF, low-frequency HRV component; HF, high-frequency HRV component.
aThe difference between EF-off and EF-on is significant (two-way ANOVA).
bThe difference between eyes-open and eyes-closed is significant (two-way ANOVA).
that the effect of EF on brain activity is restricted to that related to
theta rhythm. An increased theta rhythm is often observed when
the arousal level is low [12]. However, HRV parameters, which
are also related to arousal [22,23], showed no difference during
the ELF-EF exposure in the present study. The increased theta
power during exposure was found in the eyes-closed and eyes-
open conditions, when the arousal level was maintained. Further,
our study attempted to reduce the effects of arousal changes
caused by the experimental protocol by shortening the single EF
exposure length to 1 min, and by randomizing the periods of eyes-
open and eyes-closed conditions. White noise and fan wind were
added to decrease the influence of EF generator noise and body
hair vibration caused by EF on the data. The results under these
experimental settings suggested that probably the lowering of the
arousal level due to the experimental procedures was not the cause
of theta augmentation.
Our results suggest that ELF-EF may directly modify brain
activity, leading to increased theta rhythm, although its functional
significance is not clear. Cortical inactivation is often accompanied
by the slowing of EEG. However, in the frontal midline area, theta
rhythm is related to higher brain activity. A recent study has shown
that this theta activity exhibits a wide topographical distribution
[24]. EEG recording of the frontal brain region will be necessary
to interpret our findings in terms of functional aspects.
In our study, the increase in theta power during the eyes-open
and eyes-closed conditions were 19.3% and 11.7%, respectively.
In the simulation experiment (see Materials and Methods), the
artificial signal and noise were processed with the same filtering
system used for the human subjects, and no measurable difference
in power was observed between the filtered data of signal with and
without noise. The effect of the filtering procedure on our results
is less probable.
Delta, alpha, beta, and total powers were not affected by the 1-
min EF exposure, indicating that the EEG change was restricted to
the theta range. The HRV indices were also not affected, suggesting
that EF exposure in the present experiment did not change the
autonomic activity. Longer exposure in the EF may induce changes
in the wider ranges of the EEG spectrum and HRV indices [6].
To the best of our knowledge, this is the first work reporting that
ELF-EF exposure augmented the EEG theta power. An increase
in the EEG power occurred under EF exposure. This finding
would be useful to understand the effects of ELF-EF on human
health. However, this result was not consistent with that reported
by Yamashita et al. [9,10], which showed that EF exposure
decreased the EEG total power and alpha ratio; the EEG in this
study was measured during the eyes-closed condition for 30 min.
The EF generator and the parameters of EF exposure including
output voltage in this study were the same as in our work. The
longer duration of EF exposure in this study could have produced
this discrepancy; the data were also recorded in the eyes-closed
condition. Eye closure for a long period could have led to changes
in the baseline brain activity and might have affected the results.
Future studies with a longer EF exposure duration in the eyes-open
condition are necessary to resolve the issue. Different experimental
situations may lead to changes in the EF effects.
Paucity of previous research in this field may have been due to
the difficulty in EEG measurement under EF exposure. Adequate
blockade of EF at the electrodes and lead wires, and use of
effective filter enabled the EEG recording, as shown in this study.
Another difference was the age of the subjects. Their subjects
[9,10] were in their 20s and were younger than our subjects (mean
age: 48.6 years). The effects of EF may be influenced by age. In
our study, EEG was recorded only at Oz on the head surface.
Interestingly, EEG analysis at other recording sites may exhibit
different results and, thus, warrant future research.
4.2. EF profiles and faraday cage effect It is known
that an ELF-EF is shielded by a conductive object, such as
a human body (Faraday cage effect). According to the World
Health Organization (WHO) Environmental Health Criteria 238,
’Magnitudes of the induced electric fields are typically 10-4 to
10-7 of the magnitude of external unperturbed field’ [3]. Kaune
and Phillips [25] reported that for a grounded human standing in
an unperturbed, vertical, 60-Hz, and 10-kV/m EF, the surface EF
on the parietal region of the head was enhanced to 180 kV/m, and
the induced current densities inside the body ranged from 0.6 to
20 mA/m2. When the conductivity of the body was set at 0.2 S/m
[21], the induced EF was estimated to be 3–100 mV/m, which
was 10−7to 10−5of the magnitude of the external unperturbed
and uniform EF, consistent with the WHO statement [3].
6IEEJ Trans (2021)
ELECTRIC FIELD EXPOSURE INCREASES THETA POWER OF EEG
In our study, the surface EFs at the parietal region of the head,
chest, waist, and leg were 185, 51, 48, and 32 kV/m, respectively
(Fig. 5(a)). The induced EFs at the level of the corresponding
regions were 5, 8, 35.5, and 110 mV/m, respectively and were
10−8to 10−6of the surface EFs. Though it is important to consider
the difference in the exposure conditions (e.g., grounding and EF
uniformity), our data were consistent with the findings of Kaune
and Phillips [25] concerning the Faraday cage effect, confirming
the validity of our measurement system.
4.3. Possible mechanisms underlying the EEG
changes The mechanisms of EEG change due to EF exposure
were not clear from the results of the current study, but EF
exposure may have directly influenced neuronal activity. Saunders
and Jefferys [26] reported that the threshold of EF stimulation
for in vitro activation of neurons in brain slices is 100 mV/m. In
our study, the induced EF was estimated to be between 5 mV/m
and 270 mV/m. The induced EF was higher than 100 mV/m at
the hand and leg recording sites but lower at the head. The EF
exposure may act on the peripheral system.
It is possible that the effects on EEG by the EF exposure may
be caused by its indirect effects on brain activity through other
systems. Kato et al. [11] reported the effect of EF on the skin
hair and found that the skin hair and its movements due to the EF
induced a change in the detection threshold for the external EF. The
EF could potentially act on the nervous system through this effect
on the skin surface. Future studies on these topics are warranted
to further clarify the mechanisms underlying EEG changes due to
the EF.
4.4. Contribution of the magnetic field The magnetic
field caused by the ELF-EF exposure system may also be involved
in the effects of EF exposure. Previous studies have shown
that various cognitive and behavioral changes are induced by
combined electric and magnetic field stimulation [27]. Specifically,
the alpha wave of EEG is increased or decreased by magnetic field
exposure [28–30]. However, the magnitude of the magnetic field
produced by the present system (HES-A30) was approximately
0.2 μT at the chair position connected to the EF generator
(unpublished observation, Hakuju Institute for Health Science),
which was ≤1/100 of the corresponding value in previous studies
that reported the effects of the magnetic field on EEG. Furthermore,
the maximum current density induced in a person by a uniform,
0.2-μT, 60-Hz magnetic field is reported to be 4.2 μA/m2[31],
which is estimated to be ≤1/238 than that induced by the EF of our
device. The magnetic field of the present experimental system was
lower than that reported to cause biological effects (e.g. Shafiei
et al. [30]) and, thus, it is not probable that the current results are
related to magnetic field generation.
4.5. Limitations In our study, the effects of EF on
measurement devices were eliminated by shielding the electrodes,
lead wires, amplifier, and recording systems with electroconductive
materials connected to the human body. However, the effects
of contact impedance between the EEG electrode and the skin,
and that of skin polarization may not be fully excluded in
human measurements. Future studies recording signals from a
phantom incorporating an induced electrical potential generator
for simulating EEG are also necessary to validate that EF reflects
changes in brain activity. Further, this study had a small sample
size and did not include female subjects. Therefore, it is necessary
to conduct more studies in the future based on a larger number of
subjects.
5. Conclusion
Our study revealed that the theta power of EEG was significantly
greater during ELF-EF exposure in the eyes-open and eyes-closed
conditions. The data suggest that exposure to 50-Hz EF with an
applied voltage of 30 kV for 1 min affects brain activity.
Acknowledgments
The authors thank Professor Koichi Shimizu (Waseda University) and
Emeritus Professor Katsuo Isaka (The University of Tokushima) for their
practical supervision during the measurement and evaluation of EFs;
Professor Hiroshi Suzuki (The Obihiro University of Agriculture and
Veterinary Medicine) for constructive advices on the experimental design;
and Mr. Akikuni Hara (President of the Hakuju Institute for Health Science)
for supporting this study. The authors are also grateful to members of the
Hakuju Institute for Health Science who helped conduct this study.
References
(1) International Commission on Non-ionizing Radiation Protection.
Guidelines for limiting exposure to time-varying electric, magnetic,
and electromagnetic fields (up to 300 GHz). International com-
mission on non-ionizing radiation protection. Health Physics 1998;
74:494– 522.
(2) International Commission on Non-ionizing Radiation Protection.
Guidelines for limiting exposure to time-varying electric and mag-
netic fields (1 Hz to 100 kHz). Health Physics 2010; 99:818–836.
(3) World Health Organization (ed). Environmental Health Criteria
238: Extremely Low Frequency Fields. World Health Organization:
Geneva; 2007.
(4) Harakawa S, Takahashi I, Doge F, Martin DE. Effect of a 50 Hz
electric field on plasma ACTH, glucose, lactate, and pyruvate levels
in stressed rats. Bioelectromagnetics 2004; 25:346– 351.
(5) Harakawa S, Takagi K, Yukawa M, Inoue N, Suzuki H, Nagasawa H.
Effects of an electric field on plasma levels of ACTH and β-endorphin
in dogs with tumors or spinal cord injuries. Journal of Alternative and
Complementary Medicine 2005; 11:788 – 791.
(6) Harakawa S, Hori T, Inoue N, Suzuki H. Time-dependent changes in
the suppressive effect of electric field exposure on immobilization-
induced plasma glucocorticoid increase in mice. Bioelectromagnetics
2017; 38:272– 279.
(7) Hori T, Inoue N, Suzuki H, Harakawa S. Exposure to 50Hz electric
fields reduces stress-induced glucocorticoid levels in BALB/c mice in
a kV/m- and duration-dependent manner. Bioelectromagnetics 2015;
36:302– 308.
(8) Shinba T, Takahashi K, Kanetaka S, Nedachi T, Yamaneki M,
Doge F, Hori T, Harakawa S, Miki M, Hara H, Suzuki H, Hara A. A
pilot study on electric field therapy for chronic pain with no obvious
underlying diseases. Japanese Journal of Integrated Medicine 2012;
5:68– 72 (in Japanese).
(9) Yamashita M, Ohsaki K, Shimizu K. Effects of a strong ELF
electric field on the human electroencephalogram. EMC’98 ROMA
International Symposium on Electromagnetic Compatibility 1998;
1:159– 164.
(10) Yamashita M, Ohsaki K, Shimizu K. Development of biotelemetry
technique for EEG/ECG measurement in a strong ELF electric field.
IEEJ Transactions on Electronics, Information and Systems 2002;
122:1589– 1594.
(11) Kato M, Ohta S, Shimizu K, Tsuchida Y, Matsumoto G. Detection-
threshold of 50-Hz electric fields by human subjects. Bioelectromag-
netics 1989; 10:319– 327.
7IEEJ Trans (2021)
T. SHINBA, T. NEDACHI, AND S. HARAKAWA
(12) Niedermeyer E, Lopes Da Sylva F (eds). Electroencephalography:
Basic Principles, Clinical Applications, and Related Fields. 5th ed.
Lippincott Williams and Wilkins: Philadelphia, PA; 2005.
(13) Shinba T. Major depressive disorder and generalized anxiety disorder
show different autonomic dysregulations revealed by heart-rate vari-
ability analysis in first-onset drug-na¨
ıve patients without comorbidity.
Psychiatry and Clinical Neurosciences 2017; 71:135 – 145.
(14) Malik M. Heart rate variability: Standards of measurement,
physiological interpretation, and clinical use. Circulation 1996;
93:1043– 1065.
(15) Goldstein DS, Bentho O, Park MY, Sharabi Y. Low-frequency power
of heart rate variability is not a measure of cardiac sympathetic tone
but may be a measure of modulation of cardiac autonomic outflows
by baroreflexes. Experimental Physiology 2011; 96:1255–1261.
(16) Akselrod S, Gordon D, Madwed JB, Snidman NC, Shannon DC,
Cohen RJ. Hemodynamic regulation: Investigation by spectral anal-
ysis. The American Journal of Physiology 1985; 249:H867–H875.
(17) Shimizu K, Endo H, Matsumoto G. Fundamental study on measure-
ment of ELF electric field at biological body surfaces. IEEE Trans-
actions on Instrumentation and Measurement 1989; 38:779 – 784.
(18) Isaka K, Nishimura R, Arase S, Hayashi N, Takiwaki H, Osaki K,
Takahashi K, Doge F, Sakaino M. Dosimetry and exposure experi-
ments for extremely low frequency high-tension electric field therapy.
EMC’98 ROMA International Symposium on Electromagnetic Com-
patibility 1998; 1:204– 207.
(19) Mathur AB, Bhardwaj IS. Testing and Evaluation of Plastics. Allied
Publishers: Mumbai, India; 2003.
(20) Norimoto M. Dielectric properties of wood. Wood Research
(Bulletin of the Wood Research Institute Kyoto University) 1976;
59(/60):106– 152.
(21) Andreuccetti D, Fossi R, Petrucci C. An Internet Resource for
the Calculation of the Dielectric Properties of Body Tissues in the
Frequency Range 10 Hz −100 GHz. IFAC-CNR: Florence; 1997
(Based on data published by C. Gabriel et al. in 1996. [Online].
Available: http: //niremf. ifac. cnr.it/tissprop/).
(22) Bonnet MH, Arand DL. Heart rate variability: Sleep stage, time of
night, and arousal influences. Electroencephalography and Clinical
Neurophysiology 1997; 102:390– 396.
(23) Graham C, Sastre A, Cook MR, Kavet R. Heart rate variability
and physiological arousal in men exposed to 60 Hz magnetic fields.
Bioelectromagnetics 2000; 21:480– 482.
(24) Eschmann KCJ, Bader R, Mecklinger A. Topographical differences
of frontal-midline theta activity reflect functional differences in
cognitive control abilities. Brain and Cognition 2018; 123:57–64.
(25) Kaune WT, Phillips RD. Comparison of the coupling of grounded
humans, swine and rats to vertical, 60-Hz electric fields. Bioelectro-
magnetics 1980; 1:117– 129.
(26) Saunders RD, Jefferys JGR. A neurobiological basis for ELF
guidelines. Health Physics 2007; 92:596–603.
(27) Cook MR, Graham C, Cohen HD, Gerkovich MM. A replication
study of human exposure to 60-Hz fields: Effects on neurobehavioral
measures. Bioelectromagnetics 1992; 13:261– 285.
(28) Cook CM, Thomas AW, Prato FS. Resting EEG is affected by
exposure to a pulsed ELF magnetic field. Bioelectromagnetics 2004;
25:196– 203.
(29) Cook CM, Thomas AW, Keenliside L, Prato FS. Resting EEG effects
during exposure to a pulsed ELF magnetic field. Bioelectromagnetics
2005; 26:367– 376.
(30) Shafiei SA, Firoozabadi SM, Tabatabaie KR, Ghabaee M. Investi-
gation of EEG changes during exposure to extremely low-frequency
magnetic field to conduct brain signals. Neurological Sciences 2014;
35:1715– 1721.
(31) National Research Council (ed). Possible Health Effects of Exposure
to Residential Electric and Magnetic Fields. National Academy Press:
Washington, DC; 1997.
Toshikazu Shinba (Non-member) He received a M.D. degree
from Nagoya University in 1981 and a Ph.D.
degree in medicine from Juntendo Univer-
sity in 1992, continued researches at Tokyo
Institute of Psychiatry, and is presently an
assistant director of the hospital and chief
of psychiatry department in Shizuoka Sai-
seikai General Hospital. He has worked on
electrophysiological studies using electroen-
cephalogram and heart rate variability measurements both in nor-
mal subjects and in psychiatric patients. He has recently been
working on the effects of electric field application on these elec-
trophysiological measures. Japanese Society of Psychiatry and
Neurology, Japanese Society of Clinical Neurophysiology.
Takaki Nedachi (Non-member) He received a M.S. degree in
multidisciplinary sciences from Tokyo Uni-
versity in 2004, and is presently a researcher
in Hakuju Institute for Health Science Co.,
Ltd. He has worked on studies on effects
of electric fields on human biological sig-
nals and clinical parameters. He has recently
been working on dosimetry of electric fields
using numerical simulation and electric mea-
surements.
Shinji Harakawa (Member) He received a Ph.D. degree in Vet-
erinary Science (Physiology) from Gifu Uni-
versity in 2006, and is presently a manager in
Hakuju Institute for Health Science and also
is a Project Professor of the Bio-Self Regu-
lating Science Laboratory in Obihiro Univer-
sity of Agriculture and Veterinary Medicine.
He has worked on live cell imaging, electric
field imaging for living body surface, phys-
iology in stress response, and stress management. IEEJ, IEEE,
Bioelectromagnetics, The Japanese Society of Veterinary Science
member.
8IEEJ Trans (2021)
A preview of this full-text is provided by Wiley.
Content available from IEEJ Transactions On Electrical and Electronic Engineering
This content is subject to copyright. Terms and conditions apply.
