Wednesday, March 18, 2015

Is There a Connection Between Electrosensitivity and Electrosensibility? A Replication Study

Joel's comments: The electromagnetically sensitive (ES) participants selected for this study were based upon individuals' responses to a self-reported measure. Most ES participants did not experience severe symptoms so it may be inappropriate to consider them to have electromagnetic hypersensitivity (EHS). Like other sham provocation studies, this study assumed that someone with ES knows when they are exposed to an electromagnetic field (EMF) and when they are not. The study protocol assumed that there is no lag between the exposure and the ability to detect the exposure (or the non-exposure). The study assumed that all ES participants would be affected by exposure to a 50 Hz magnetic fieldHowever, it is likely that some people who experience ES may be sensitive to certain radio frequency fields, but not ELF magnetic fields. Despite the questionable assumptions upon which this study was based, the ES participants were significantly (p = .038)  more likely to detect an MF exposure than chance would dictate. This result replicated the finding of an earlier study.


Is There a Connection Between Electrosensitivity and Electrosensibility? A Replication Study

Szemerszky R, Gubányi M, Árvai D, Dömötör Z, Köteles F. Is There a Connection Between Electrosensitivity and Electrosensibility? A Replication Study. Int J Behav Med. 2015 Mar 17. [Epub ahead of print]


BACKGROUND: Among people with idiopathic environmental intolerance attributed to electromagnetic fields (IEI-EMF), a better than random detection ability for a 50-Hz 0.5-mT magnetic field (MF) and a propensity to experience more symptoms than controls was reported in a previous study.

PURPOSE: The current study aimed to replicate and clarify these results using a modified experimental design.

METHOD: Participants of the provocation experiment were 49 individuals with self-reported IEI-EMF and 57 controls. They completed the questionnaires (symptom expectations, Somatosensory Amplification Scale-SSAS, radiation subscale of the Modern Health Worries Scale-MHWS Radiation) and attempted to detect the presence of the MF directed to their right arm in 20 subsequent 1-min sessions. Symptom reports were registered after each session.

RESULTS: Individuals with IEI-EMF as opposed to the control group showed a higher than random detection performance (d' index of signal detection theory), while no difference in their bias (β index) toward the presence of the MF was found. Predictors of reported symptoms were self-reported IEI-EMF and believed as opposed to actual presence of the MF. People with IEI-EMF reported significantly more symptoms particularly in the believed presence of the MF. IEI-EMF was closely related to MHWS Radiation and SSAS scores.

CONCLUSION: People with IEI-EMF might be able to detect the presence of the MF to a small extent; however, their symptom reports are connected to perceived exposure.


Although the domain of extremely low frequency magnetic fields (ELF MFs, 30 to 300 Hz) is considered harmless below the reference levels (B = 0.2 mT for public exposure and B = 1 mT for occupational exposure) [6], there are people who experience and report unpleasant symptoms in the proximity of electric devices working at this frequency range. The phenomenon is called idiopathic environmental intolerance attributed to electromagnetic fields (IEI-EMF) (formerly electromagnetic hypersensitivity, EHS, or electrosensitivity) [8], and it is not restricted to the ELF range. Symptoms are reported with respect to all frequency domains, in fact, many people experience symptoms related to devices operating at very different EMF frequencies (mobile phones, computers, power lines, Wi-Fi routers, etc.). As reported symptoms were not related to the presence of the MF/EMF in the majority of double-blind experiments [9–13], they were also experienced during sham exposure [14–16], and their severity was a function of the believed magnitude of the field [17], we can conclude that IEI-EMF refers to sensitivity to believed as opposed to actual presence of MFs/EMFs. Moreover, reports of symptoms attributed to MF/EMF exposure were related to certain trait-like personality characteristics and cognitive-emotional factors (e.g., somatization tendency, somatosensory amplification, elevated risk perception, MHWs, expectations) [17–20]. In summary, although there were a few studies that reported a connection between symptoms and actual MF/EMF exposure [21–23], the IEI-EMF phenomenon is recently regarded as mainly of psychogenic origin [24, 25] or, in other words, as a nocebo-related phenomenon [9, 10, 17, 19].

The individual ability to perceive electric or electromagnetic exposures (electrosensibility, ES) is not necessarily related to symptom reports [26]: people with IEI-EMF were usually not able to detect the presence of MFs/EMFs [10, 11]. However, in a previous double-blind provocation experiment using the methodology of signal detection theory (SDT), we found a better than random detection performance (i.e., a higher than zero d′ index) with respect to 50 Hz 0.5 mT MF among people with IEI-EMF while controls could not discriminate between presence and absence of the MF during 20 subsequent 1-min sessions [19]. Moreover, IEI-EMF was connected to a higher bias (i.e., a stronger tendency to assume the presence of the MF, or, in technical terms, a lower decision criterion, β), to more reported symptoms, to higher levels of somatosensory amplification, and to higher scores on the radiation subscale of the MHW scale. Unfortunately, symptom reports were obtained in a retrospective manner, following the last detection session, thus symptoms might have been of psychogenic as well as of “toxicogenic” origin (i.e., caused by the exposure). In consequence, although the IEI-EMF group showed a better detection performance and reported more symptoms than the control group, we were not able to directly investigate the link between sensibility and actual sensitivity. The current study aimed (1) to replicate previous findings and (2) to investigate the abovementioned relationship using a modified experimental design with symptom reports registered after each detection session. Concerning the second point, it was hypothesized that symptom reports would be connected to believed but not to actual EM exposure (i.e., reported symptoms would be of psychogenic origin).

... A priori sample size calculation was based on the detection-related results of our previous study. In that study, an effect size of 0.47 was found in the IEI-EMF group. For the present study, considering an effect size of 0.45, 0.05 for type I error, 0.9 for type II error, and one-tailed t-test, a sample size of 44 was calculated as necessary for each group. 

MF was generated by a standard Helmholtz coil apparatus. The device consisted of two coils (inner diameter 42 cm, outer diameter 55 cm) placed on a common axis spaced apart at a distance equal to their radii (21 cm) with equal currents flowing in the same direction. Helmholtz coil produces a homogeneous MF in the space between the coils. The coils were constructed of glaze-insulated copper wire (d  = 1.4 mm) and had 240 turns (DC resistance was 2.9 Ω). A sinusoidal current with 50 Hz frequency, generated by a 230 V, 180 VA adjustable toroidal transformer, was fed into the coils, producing an oscillating magnetic field. The level of magnetic flux density was set at 500 μT (rms). This value represents a relatively high level: magnetic flux density on 50 Hz generated by commonly used household devices at a similar distance (50 cm) ranges from 0.05 μT (refrigerators) to 1.66 μT (microwave ovens) [30]. Moreover, magnetic flux densities from 0.7 to 2.9 μT were measured in the proximity (15 m) of power transmission lines [31].

... Participants were requested to put their right hand between the coils (as symptoms attributed to MF in the ELF range are often reported following the use of handheld household devices, exposure was directed to the hand) ... Ambient magnetic field in the room was 0.030 μT. When the laptop computer used in the experiment was turned on, the stray magnetic flux density measured at the position of participants’ head was 0.036 μT. During the operation of the Helmholtz coil, the latter value was 1.14 μT. 

In a double-blind provocation experiment, individuals with IEI-EMF as opposed to controls were able to detect the presence of a 500-μT 50-Hz MF to a small extent (i.e., their accuracy was not far from random guessing). However, no direct connection between IEI-EMF (electrosensitivity) and detection performance (electrosensibility) was found. As for reported symptoms, (1) actual presence of the MF had no impact on symptom reports, (2) more symptoms were reported in the believed presence of the MF (i.e., hits or false alarms) in both groups, and (3) people with IEI-EMF reported more symptoms than controls, particularly when they believed that the MF was present. Higher levels of somatosensory amplification and higher scores on the radiation subscale of the MHWS were independently related to IEI-EMF even after controlling for sociodemographic variables.

An important thing that should be kept in mind when evaluating these findings is that participants in the IEI-EMF group were not patients with serious complaints. They voluntarily participated in the experiment and reported severity of their symptoms, and impact of their condition on their everyday life lied in the low to medium range. Their most often mentioned complaints (headache, dizziness, crawly feelings, nausea) clearly belong to the so-called non-specific category which means that the source of these complaints is difficult to identify [33–36].

The most surprising finding of our earlier research, namely the better than random detection performance of the IEI-EMF group, was replicated in the current study, even the magnitude of the d′ indices was quite similar (M = 0.296 vs 0.280). Similarly, d′ indices of the control groups did not statistically differ from zero in both studies. It is worth mentioning, however, that detection performance of the IEI-EMF group was only slightly better than random (i.e., it was far from perfect discrimination), and there was no statistically significant differences between the detection performance of the groups.

Sensibility and sensitivity to a 500-μT 50-Hz MF seem to be independent of each other. People with IEI-EMF of low severity might be able to detect the presence of the MF to a small extent; however, their symptom reports are clearly connected to the believed as opposed to the actual presence of the MF.


Joel M. Moskowitz, Ph.D., Director
Center for Family and Community Health
School of Public Health
University of California, Berkeley

Electromagnetic Radiation Safety

News Releases:
Twitter:                 @berkeleyprc

No comments:

Post a Comment