Exposure to Mobile Phone Electromagnetic Fields and Subjective Symptoms:
A Double-Blind Study
CATERINA CINEL, PHD, RICCARDO RUSSO, PHD, ANGELA BOLDINI, PHD, AND ELAINE FOX, PHD
Objectives: The objective of this study was to examine whether acute exposure to radio frequency electromagnetic fields (REFs)
emitted by mobile phone may affect subjective symptoms. Methods: Three large groups of volunteers (total 496) were exposed to
REFs emitted by mobile phones in one session and sham signals in a different session. REF and sham exposure sessions were
counterbalanced and double blinded. Participants were exposed to either Global System for Mobile Communication (GSM) or
unmodulated signals, and the mobile phone was positioned either on the left or on the right side of the head. Before and after REF
and sham exposure participants completed a questionnaire to rate five symptoms. Any changes in the severity of the symptoms after
REF exposure were compared with changes after sham exposure. Results: For one group of participants (N ? 160), it was found
that dizziness was affected by GSM exposure, but this was not consistently found with the other two groups of participants. No other
significant effects were found. Conclusions: We did not find consistent evidence suggesting that exposure to mobile phone REFs
affect subjective symptoms. Even though we acknowledge that more research is needed, we believe that our results give an
important contribution to the research on mobile phone use and subjective symptoms. Key words: radio frequency electromagnetic
fields, REF exposure, subjective symptoms, mobile phones.
REF ? radio frequency electromagnetic field; GSM ? Global
System for Mobile Communication; CW ? continuous wave;
SAR ? specific energy absorption rate; SD ? standard deviation.
rash, or warmth sensations on the skin (particularly in the area
around the ear). Thus, in recent years, scientists have been
trying to establish whether symptoms such as these can be
triggered specifically by the radio frequency electromagnetic
fields (REFs) emitted by mobile phones (1). As it is often the
case with studies on the effects of electromagnetic fields on
human health, this is a very controversial area of research and
there are no conclusive studies as yet (2). A way of investi-
gating this issue is by examining symptoms or sensations
experienced by mobile phone users (3–5). Hocking (5) inter-
viewed 40 mobile phone users and found that the majority
(88%) suffered of cranial symptoms such as heat or dull pain
on temple, ear or occiput when using a mobile phone. Typi-
cally symptoms started a few minutes after commencing a
call. Other symptoms frequently reported were transient ef-
fects on vision, such as blurring (31%) and sensation of nausea
or dizziness (43%). The same symptoms did not occur when
using ordinary handsets. In a more recent survey, Balikci et al.
(3) compared symptoms reported by mobile phone users and
nonusers and found that the use of mobile phone may increase
the severity of symptoms such as headache, irritation, clicking
sounds in ears and others. While studies examining the symp-
toms of mobile users in everyday life are essential to reveal
effects related to prolonged daily exposure to mobile phones,
he use of mobile phones has often been associated with
particular subjective symptoms such as headache, fatigue,
a) they do not necessarily reveal an association between
subjective symptoms and REFs exposure, b) they may lack
accurate experimental control, as in those kind of studies it is
very difficult to have control of variables other than REFs that
can trigger particular symptoms, and c) volunteers can be
biased when judging the severity of the symptoms related to
mobile phone use. An alternative way of testing whether or
not mobile phones REFs produce subjective symptoms is by
exposing individuals, under strict experimental control and
under double-blind conditions (i.e., when both participants
and experimenter are unaware of the exposure condition) to
REFs and sham signals, in different sessions, for a given
period of time, and then ask them to report about particular
symptoms they are experiencing (6). Hietanen et al. conducted
a study using a similar method (7). They recruited a group of
volunteers who reported themselves as being sensitive to
cellular phones. In several sessions, participants were exposed
to either REFs or sham signals and were then questioned about
any particular symptoms. The symptoms reported were not
aggravated by REFs exposure. The study also investigated
whether participants were able to sense mobile phone REF
emission and it was found that no one could distinguish
between exposure conditions. Recently, other studies have
shown no relations at all between exposure to mobile phone
REFs and subjective symptoms (6,8,9), not even those studies
that have focused on individuals claiming to be hypersensitive
Even though some scientists came to the conclusions that
nocebo effects (8) or psychological factors (9) may affect
reports of symptoms, generally it is acknowledge that more
research is needed (10).
In the present study, we investigated whether exposure to
mobile phone REFs can affect five symptoms (headache,
dizziness, fatigue, itching or tingling of skin, and sensation of
warmth on skin). The choice of symptoms was based on what
has been previously reported in the literature. In two different
sessions, volunteers were exposed to active REF and sham
signals, and asked to rate the above symptoms. Exposure
conditions (real exposure and sham exposure) were double
blinded. Testing under double-blind conditions seems impor-
tant since many effects reported under single-blind conditions
From the Department of Psychology, University of Essex, United Kingdom
(C.C., R.R., E.F.); and Department of Psychology, University of Valencia,
Address correspondence and reprint requests to Caterina Cinel, Department
of Psychology, University of Essex, Colchester CO4 3SQ, United Kingdom.
This work was supported by a grant from the Mobile Telecommunications
and Health Research Programme (Grant ref. RUM9) to Riccardo Russo and
Elaine Fox. The views expressed in the publication are those of the authors
and not necessarily those of the funders.
Received for publication July 23, 2007; revision received October 10, 2007.
345 Psychosomatic Medicine 70:345–348 (2008)
Copyright © 2008 by the American Psychosomatic Society
have not stood up to the test of replication under double-blind
REF exposure (11–13). The large sample included in the study
and the very high standards of the experimental method used
make our study an important and valuable contribution to the
research on the effects of electromagnetic fields on public
Four hundred ninety-six volunteers took part in the study (330 were
female; average age was 23 years, with SD ? 4.4, minimum age ? 18 years
and maximum age ? 42 years), between March 2003 and September 2005.
All were studying or working at the University of Essex, United Kingdom.
Volunteers were originally recruited to participate in one of three different
groups of experiments (studies 1, 2, and 3), whose main aim was to test
possible effects of exposure to mobile phone REF on cognitive functions
(attention, memory, and auditory perception). In addition to performing their
cognitive tasks, participants were asked to evaluate a group of symptoms both
before and after REF and sham exposure. Details of the three studies are
shown in (13,14). In each study, participants were tested in two sessions
(about 1 week apart): in one session they were exposed to REF signals,
whereas in the other session exposure was sham (the order of REF and sham
sessions was counterbalanced between participants). REF and sham condi-
tions were double blinded. In both conditions, exposure lasted for about 40
minutes. During exposure, participants had to perform a series of cognitive
tasks. Although the nature of the tasks performed was different in the three
studies, REF and sham exposure conditions were identical. Half of partici-
pants were exposed to Global System for Mobile Communication (GSM)
signal and half to continuous waves (CW) unmodulated signals. The phone
was attached to a cap that was then positioned on participant’s head. Its
position was adjusted so that the antenna of the phone was slightly behind and
above the ear. For half of participants, the mobile phone was mounted on the
left side of the head, and on the right side for the other half. The mobile phone
used could emit GSM modulated and CW unmodulated signals at 888 MHz
as well as a sham signal. The level of specific energy absorption rate
(SAR) was the same for both CW and GSM signals (with SAR within the
International Commission on NonIonising Radiation Protection guide-
lines). The average SAR in both modes was 1.4 W/Kg (?30%). For the
GSM mode, the peak SAR was 11.2 W/Kg (CW does not have a peak).
The SAR in the sham condition was less than 0.002 W/Kg. The above features
correspond to the approved exposure system made for the Mobile Telecommu-
nication and Health Research Programme (see http://www.mthr.org.uk/meetings/
nov_2002/summaries/human_exposure.htm) in the United Kingdom.
In each session, participants were given the same questionnaire where
they rated five particular sensations or symptoms experienced at that very
moment, as shown in Table 1. Each symptom was assessed on a scale ranging
from 1 (no sensation) to 5 (strong sensation). The questionnaire was given
both at the beginning (before positioning the phone on the head) and at the end
of the session (once the phone was removed). Therefore, each participant
completed the same questionnaire four times.
The procedures used in the three studies complied with the relevant
safeguards and regulations in place for studies testing human participants at
the University of Essex and the Ethics Committee of the University of Essex
approved the study.
Participants were briefed about the nature of the study and made aware
that they could withdraw at any time, if they wish.
In the analysis of the data, changes in symptom ratings
after real REF exposure were compared with changes after
sham exposure. The number of participants was 168 in studies
1 and 3, and 160 in study 2*. The data of the three studies
were initially separately analyzed, as the tasks performed
during REF exposure were different. In a second analysis, the
data of the three studies were analyzed together to test possi-
ble differential effects of REF exposure due to the type of task
In the first analysis, for each symptom and each study, a
separate mixed factorial analysis of variance was performed
with the following four factors: type of exposure (REF versus
sham; within subjects), when symptom was rated (before
versus after exposure; within subjects), mode (GSM versus
CW; between subjects), and side of exposure (left versus right;
between subjects). To evaluate possible effects of exposure to
mobile phone REFs, the most critical factors were whether
REF exposure was real or sham and whether a symptom was
rated before or after exposure. Significant interactions of the
two factors may suggest an effect of REF exposure on partic-
ular symptoms. Mean ratings of each symptom are shown in
Table 2. Since the statistical analyses were conducted per each
symptom in each experiment, the ? level for each set of
statistical analyses for each symptom was set at 0.01 to control
for type I statistical errors within each experiment.
In all studies each symptom worsened at the end of the
session (all F values ?9), independently of any other factor.
In study 1 and study 2, there were no significant interac-
tions involving the two relevant factors† [in study 1 all F
values(1,163) ? 5.5; in study 2 all F values(1,155) ? 3.7].
In study 3, we found that for dizziness there was a signif-
icant interaction between type of exposure and when symptom
*Note, however, that in each study, a few participants were discarded
before data analysis because of errors in data recording or missing data.
Therefore, in study 1, N ? 167; in study 2, N ? 159; and in study 3, N ? 160.
†With a less stringent ? level at .05, in study 1, there was a significant
interaction for fatigue between type of exposure and when symptom was rated
[F(1,163) ? 5.5, p ? 0.02; see Table 2]. However, note that the symptom
worsened more at the end of the sham session. A three-way significant
interaction was found for symptom “itching or tingling of skin” between type
of exposure, when symptom was rated, and which side the phone was
positioned: when on the left, the difference between before and after exposure
ratings was 0.1 with REF exposure and 0.3 with sham exposure, whereas
when the phone was on the right, the differences were 0.3 and 0.2, respec-
tively [F(1,163) ? 4.7, p ? 0.03]. These results were not replicated in studies
2 and 3, hence, more importantly, it is likely to reflect a type I statistical error.
We also estimated the effect size and found that it was very small: d ? 0.18,
for study 1, and d ? 0.02, for all three studies.
TABLE 1. Symptom Questionnaire Completed by the Participants of
Could you, please, rate the strength of the sensations or symptoms,
reported below, you are experiencing at the moment?
In the scale, next to each symptom/sensation, 1 ? no sensation,
5 ?strong sensation
d) Itching or tingling of skin
e) Sensation of warmth on skin
If you are experiencing any of the sensations in d) or e), could
you please specify exactly where (e.g., back of the head,
temple regions, ear, face, etc.):
Report if you have any other symptom/sensation (with rate)
C. CINEL et al.
346Psychosomatic Medicine 70:345–348 (2008)
was assessed [F(1,156) ? 11.1, p ? .001]. The difference
between before and after exposure ratings was 0.6 with REF
exposure and 0.3 when exposure was sham. No other signif-
icant interactions were found [all F values ?3.3].
The data from the three studies were also analyzed to-
gether. One significant interaction only was found for dizzi-
ness [F(1,474) ? 5.5, p ? .01] showing that while in studies
1 and 2 differences between before and after exposure ratings
remained constant with both types of exposures (0.5), in study
3 that difference decreased when exposure was sham (0.6 with
REF exposure and 0.3 with sham exposure).
In a further analysis, suggested by an anonymous reviewer,
we analyzed the data of the three studies for possible gender
effects. A significant interaction of Exposure by When by
Gender occurred for the symptom of dizziness [F(1,457) ?
6.6, p ? .01; ratings are shown in Table 3]. Inspecting Table
3 it appears that the significant interaction was due to a
relatively higher score of males at the end of the REF expo-
sure session. However, for the remaining four symptoms none
of the exposure by when-symptom-was-rated interactions
were modulated by gender factor [all F values ?2].
In the symptom questionnaire, participants were also asked
to report any other symptom or sensation they were experi-
encing other than those listed. The most frequently reported
symptom was “strained or tired eyes.” Among the 23 volun-
teers reporting the symptom, 11 reported the symptom at the
end of the REF exposure session, 8 reported the symptom at
the end of the sham session, 3 at the end of both sessions, and
1 at the beginning of the sham session. Other symptoms were
reported, but with a very low frequency: “difficulty in con-
centrations” (2 participants), “tension in neck or shoulder” (7
participants), earache (4 participants), and sleepiness (1 par-
ticipant). There were not relevant differences between REF
and sham sessions.
DISCUSSION AND CONCLUSIONS
In the present study, we investigated possible effects of mo-
bile phone REF exposure on subjective symptoms. Under dou-
ble-blind conditions, participants were exposed, in two different
sessions, to active REF mobile phone signals and sham signals
and, in both sessions, they had to rate five subjective symptoms
before and after being exposed to the signals. All symptoms
significantly worsened at the end of the sessions, probably be-
cause of the difficulty of the tasks performed during exposure. Of
all symptoms examined, headache, fatigue, itching or tingling of
skin, and sensations of warmth on skin were not found to be
related to mobile phone exposure, in any of our three studies.
This is consistent with recent studies (6,8,9). Participants in study
3 showed a larger increase in sensation of dizziness after REF
exposure, compared with sham. However, this was not found in
study 1 (where levels of dizziness were numerically higher after
sham exposure) and study 2 (where levels of dizziness were
almost identical in both exposures), and, when the data from all
three studies were analyzed together, the increase in dizziness at
TABLE 2.Mean Ratings of the Five Symptoms in Studies 1–3
Study 1 (N ? 167) Study 2 (N ? 159)Study 3 (N ? 160)
BeforeAfterp BeforeAfterp BeforeAfterp
Itching/tingling of skin
Sensation of warmth on skin
0.9 1.25 (0.5)
0.06 1.19 (0.58)
0.3 1.11 (0.39)
0.8 1.12 (0.36)
0.02 1.72 (0.84)
0.3 1.5 (0.72)
0.06 1.14 (0.46)
0.7 1.34 (0.67)
aStatistically significant at ? level ? 0.01.
SDs are reported in parenthesis. Ratings are reported according to whether participants were exposed to REF or sham signals, and according to when symptoms
were rated (before or after exposure); p-values are reported for the interactions between the two factors (when symptom was rated and type of exposure) in the
ANOVA (? level applied ? 0.01).
TABLE 3.Ratings for Dizziness According to Gender (p ? .01).
GenderExposure When Mean
MOBILE PHONES AND SUBJECTIVE SYMPTOMS
347Psychosomatic Medicine 70:345–348 (2008)
the end of the session did not significantly vary in the two Download full-text
exposure conditions. This pattern of results is quite puzzling
as REF and sham exposure conditions were identical in the
three studies, including the duration of exposure. What
changed from one study to the others was the kind of cognitive
tasks performed during exposure—they were typical cognitive
tasks assessing either attention or memory, where participants
have to respond to stimuli shortly presented on a computer
screen (details of tasks are in Refs. 13,14,15). There are no
reasons to believe that REF effects may depend on the par-
ticular tasks in which participants are involved during expo-
sure. Because recent studies gave no evidence of dizziness
being affected by exposure to mobile phones (3,6), we suggest
that our results may simply reflect type I statistical errors.
However, to further understand the nature of the significant
effect we found on dizziness, we estimated the effect size (as
in Ref. 16) and found that d ? 0.25, when considering study
3 only, and d ? 0.06, when considering the data from all three
studies, which suggests that the effect is indeed very small and
further support our view that there are no real effects of REF
On a post-hoc analysis we also assessed gender effects.
Some modulation occurred for dizziness where a slight incre-
ment in dizziness occurred among males after REF exposure.
It is to be noted that this modulation was not detected for any
of the other symptoms. Hence, this outcome may simply
reflect a type I statistical error.
The present study does not provide consistent evidence that
active exposure to mobile phone REFs can affect subjective
symptoms. The results not only provide reliable evidence but
are also consistent with recent studies (6–9). However, we do
not consider our study to be conclusive, as only a small
number of symptoms were examined and participants were
not asked to rate other symptoms that in the literature have
been reported as being affected by REFs (3) (note however
that, in the present study, participants were encouraged to
report any other symptom not listed in the questionnaire).
Moreover, for our study we did not select volunteers who
reported to be hypersensitive to electromagnetic fields, and
this group might show effects to REF exposure—even though
a recent review has shown that this might not be the case (1).
Also, in our study, exposure duration was about 40 minutes,
which is certainly longer than the duration of typical phone calls.
However, the possibility that frequent daily REF exposure might
have real effects on subjective symptoms cannot be ruled out.
Even though we believe that our results give a much-needed
contribution to the research on mobile phone use and subjective
particular, we believe that there should be more studies on effects
of long-term exposure and effects on children.
We thank three anonymous reviewers for their valuable comments.
1. Rubin GJ, Das Munshi J, Wessely S. Electromagnetic hypersensitivity: a
systematic review of provocation studies. Psychosom Med 2005;67:
2. Maier M, Blakemore C, Koivisto M. The health hazards of mobile
phones. BMJ 2000;320:1288–9.
3. Balikci K, Cem Ozcan I, Turgut-Balik D, Balik HH. A survey study on
some neurological symptoms and sensations experienced by long term
users of mobile phones. Pathol Biol 2005;53:30–4.
4. Santini R, Seigne M, Bonhomme-Faivre L, Bouffet S, Defrasne E, Sage
M. Symptoms experienced by users of digital cellular phones: a study of
a French engineering school. Electromagn Biol Med 2002;21:81–8.
5. Hocking B. Preliminary report: symptoms associated with mobile phone
use. Occup Med 1998;48:357–60.
6. Koivisto M, Haarala C, Krause C, Revonsuo A, Laine M, Ha ¨ma ¨la ¨inen H.
GSM phone signals does not produce subjective symptoms. Biolelectro-
7. Hietanen M, Ha ¨ma ¨la ¨inen AM, Husman T. Hypersensitivity symptoms
associated with exposure to cellular telephones: no causal link. Bioelec-
8. Oftedal G, Straume A, Johnsson A, Stovner LJ. Mobile phone headache:
a double blind, sham-controlled provocation study. Cephalalgia 2007;27:
9. Rubin GJ, Hahn G, Everitt BS, Cleare AJ, Wessely S. Are some people
sensitive to mobile phone signals? Within participants double-blind ran-
domised provocation study. BMJ 2006;332:886–9.
10. Lin JC. Subjective symptoms and hypersensitivity to cellular telephones.
IEEE Antenn Propag M 2004;46:167–9.
11. Haarala C, Bjo ¨rnberg L, Ek M, Laine M, Revonsuo A, Koivisto M,
Hamalainen H. Effect of a 902MHz electromagnetic field emitted by
mobile phones on human cognitive function: a replication study. Bioel-
12. Haarala C, Ek M, Bjo ¨rnberg L, Laine M, Revonsuo A, Koivisto M,
Ha ¨ma ¨la ¨inen H. 902 MHz mobile phone does not affect short term
memory in humans. Bioelectromagnetics 2004;25:452–6.
13. Russo R, Fox E, Cinel C, Boldini A, Defeyter MA, Mirshekar-Syahkal D,
Mehta A. Does acute exposure to mobile phones affect human attention?
14. Cinel C, Boldini A, Russo R, Fox E. Effects of mobile phone electro-
magnetic fields on an auditory order threshold task. Bioelectromagnetics
15. Cinel C, Boldini A, Fox E, Russo R. Does the use of mobile phones affect
short-term memory or attention? Appl Cognitive Psych. In press.
16. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed.
Hillsdale, NJ: Lawrence Earlbaum Associates; 1988.
C. CINEL et al.
348 Psychosomatic Medicine 70:345–348 (2008)