Hypersensitivity of human subjects to environmental electric and magnetic field exposure: a review of the literature.

Hypersensitivity to exposure to electric and magnetic fields (EMFs) has been reported for nearly 20 years; however, the literature on the subject is still very limited. Nearly all the literature published concerns a dermatological syndrome that consists of mainly subjective symptoms (itching, burning, dryness) and a few objective symptoms (redness, dryness) appearing after individuals begin working with video display units and decreasing during absence from work. Case-control studies as well as some good but limited double-blind trials have not found any clear relationship between this syndrome and exposure to EMFs. A "general syndrome" with more general symptoms has been rarely described but seems to have a worse prognosis. The symptoms often associated with skin disorders are mainly of neurasthenic type and can cover a lot of nonspecific symptoms present in other atypical syndromes such as multiple chemical sensitivity or chronic fatigue. Most of these symptoms are allegedly triggered by exposure to different sources of EMFs, but there have been no valid etiological studies published on this more general syndrome. It appears that the so-called hypersensitivity to environmental electric and magnetic fields is an unclear health problem whose nature has yet to be determined.

As implied by the title of this review, herein I use the term proposed by the California Department of Health Services: hypersensitivity to electric and magnetic fields (HSEMF). It seems preferable because in the ELF range where electric and magnetic fields are considered separately (12). HSEMF is then defined in this review as a phenomenon where individuals experience adverse health effects while using or being in the vicinity of devices emanating electric and/or magnetic fields of extremely low frequency.
The clinical portraits are sometimes complex, but it seems that two general characteristics are associated with HSEMF (3,5,13): a) a group of symptoms (syndrome) usually appears or worsens during exposure to a specific source of EMFs, usually at work; b) these symptoms are reported to diminish when patients are away from the source and especially during absences from work (weekends, holidays, etc.).
The dermatological syndrome was the first to be described in the literature (2,(14)(15)(16). It is mainly related to exposure to VDUs and mostly has a good prognosis. The symptoms are primarily subjective (itching, burning, stinging, etc.) and sometimes objective but nonspecific (rashes, dry and rosy skin), and are mostly localized to the face. Clinical diagnoses of VDU users with skin disorders were quite commonly facial dermatoses such as seborrheic eczema, acnea vulgaris, mild rosacea, and atopic dermatitis (16)(17)(18)(19). Only Environmental Health Perspectives • VOLUME 110 | SUPPLEMENT

Hypersensitivity of Human Subjects to Environmental Electric and Magnetic
Field Exposure: A Review of the Literature one case has been reported in North America, with dermatological symptoms on hands and forearms that were associated with VDU use (20).
The general syndrome was more recently reported and seems less defined. Patients are described with various health symptoms associated with or without skin problems: functional symptoms of the nervous system (dizziness, fatigue, headache, difficulties in concentration, memory problems, anxiety, depression, etc.), respiratory problems (difficult breathing), gastrointestinal symptoms, eye and vision symptoms, palpitations, and so on (5,13). All are present without any indication of organic lesion. These symptoms are triggered with exposure to different electrical devices and appliances (office equipment, fluorescent lights, household appliances, televisions, etc.), and often seem to worsen with time, with relatively poor prognosis (3).

Occupational Studies
Four studies have been conducted by selfadministered questionnaire to assess the frequency of dermatological symptoms and signs in relation to VDU use in different companies in Europe and Asia. The two more detailed studies in Sweden were supplemented with clinical examination of a sample of the respondents. Those four studies are summarized in Table 1 with their main results. Two were conducted in Sweden (17,18), one in Singapore (21), and one in the United Kingdom (22). Participation rate was excellent, except in the U.K. study. Facial complaints were found to be common among office workers and were more frequent for VDU users in the two Swedish studies (17,18). Operators who complained of skin problems were generally more likely to report other health problems related to VDU use: eye discomfort or irritation (17,18), musculoskeletal symptoms (17,21), and headache (17). Symptoms were associated with duration of work on VDUs but not with the type of VDU used (18).

Population Studies
No published epidemiologic study has been found on this issue. However, a group from the European Commission tried to assess the extent of electromagnetic hypersensitivity in European populations (3). Questionnaires were sent to 138 centers for occupational medicine (COMs) and similar organizations and 15 self-aid groups (SAGs) from 15 different European countries. Response rate was 49% for the COMs and 67% for the SAGs.
Questions were asked about the frequency, type, and severity of cases of electromagnetic hypersensitivity. Although it is difficult to draw statistics from such a semiquantitative survey, the report of the European Commission (3) stated that the prevalence estimated ranges from less than a few per million (COM estimates from United Kingdom, Italy, and France) to a few tenths of a percent (SAGs in Denmark, Ireland, and Sweden) and with severe cases with generally one order of magnitude of lower occurrences.
Details of the European survey were described in the appendix of the report. It was found that the cases from Northern European countries in particular were associated mostly with work exposure, whereas cases in Germany and Ireland were associated only with sources at home. Other countries, such as France, reported mixed exposure. Nervous system and skin symptoms were more frequently reported, and ELF as well as radio frequency source exposures were reported to be associated with these symptoms. We recently reported the results of a population study of HSEMF done by telephone survey in California. The details of this study are presented in this issue (23). Out of a sample of 2,037 Californians, about 3% reported HSEMF, and 0.5% had to change jobs because of it. But no validation of the answers of the respondents was provided. Compared with power lines and distribution lines, hair dryer use was found to be the source of EMFs the most strongly associated with self-reported HSEMF.

Etiological Epidemiologic Studies
Most of the etiological studies conducted on HSEMF and published in peer-reviewed journals have focused on skin symptoms and VDU use. Case-control and experimental studies (provocation studies) have tried to assess the role of exposure to EMFs as well as other environmental factors.

Case-Control Studies
Three case-control studies, focusing on skin disorders in relationship to VDUs, have been published. All were conducted in Sweden and are summarized in Table 2.
One (24) was rather limited regarding the assessment of cases, sample size, and environmental evaluation. Despite their statistical significance, the results of hormonal changes are difficult to interpret. They might be normal variation or due to factors not controlled in the experiment. The two other studies (19,25) have higher quality, with dermatological assessment of skin lesion, environmental assessment with EMF measurements, and organizational and psychological evaluation by questionnaire. These two studies found association of nonspecific skin disorders with VDU use and also with workload. One found association with exposure to background electric field intensity but no direct relation with magnetic fields emitted by the VDUs (25). These three studies observed some type of relationship of health status with the VDU use, but no direct link was found with EMF exposure from VDUs.

Experimental Studies
Five "provocation" studies on subjects with skin disorders associated with VDU use and whose results were published in peer-reviewed journals were found. Two are from Norway (26,27) and three from Sweden (7,28,29); they are summarized in Table 3. The five used a double-blind crossover design and only one used a control group (29). In three studies (26)(27)(28), exposure was produced by real VDUs during a working session on either VDUs with modification of exposure by screen filter (26,27) or a different type of VDU (28). The other two studies (7,29) used the VDUs ("on" or "off") only as a source of EMFs without having the subjects work with them. All assessed real exposure from EMFs [ELF or very low frequency (VLF; 3-100 kHz)] at a distance of 30-50 cm from the VDUs. All used standardized questionnaires for symptom evaluation; two used dermatologists for clinical evaluation (26,28), and two used blood sampling for hormone evaluation (7,29). The quality of the methodology in these experimental studies is considered good, but sample sizes are limited (16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35), and simple statistical analyses are provided by the authors. All the studies were negative except one, which gave an equivocal result (26) but were not reproduced in a more robust study in terms of number of subjects and duration of the experiment (27). Globally, all these studies confirm that skin disorders alleged to be associated with VDU use are not related to EMF emission from VDUs. No study found that reaction of subjects was related to field intensity. But skin disorders were associated with perception that the source emission (7,29) was "on" and with duration of work (26) and low humidity (28).
Very few studies were done on subjects with a more general syndrome. Rea et al. (30) presented the results of a study that they labeled preliminary. One hundred patients treated for some type of environmental sensitivity (the authors briefly mentioned they had been previously evaluated and treated for biological inhalant, food, and chemical sensitivities) and who complained of being EMF sensitive were evaluated in a single-blind screening. They were challenged for 3 min at different frequencies from 0.5 Hz to 5 MHz. The mean intensity of the fields was presented as approximately 2,900 nT at floor level and 350 nT at the level of the chair in Unexplained Symptoms • Review of hypersensitivity to EMFs which the patient sat while being exposed. The imprecision of the exposure measurements and the adequacy of the exposure settings were settled in a letter to the editor from Bergqvist and Franzén (31). Of the 100 patients first challenged, 25 reacted positively to exposure, with only one reaction to exposure to a placebo. These 25 were compared with 25 healthy volunteers for a double-blind challenge. No detail was given on those volunteers or on the double-blind setting. Of the 25 hypersensitive patients, 16 (64%) reacted positively, the majority (53%) reacting to exposure compared with a few (7.5%) who reacted to a blank challenge. In fact, most of the results presented are incomplete, and it was stated that no reaction to any challenge, active or placebo, was found in the volunteer group. The major symptoms reported by the patients tested were mainly neurological, cardiological, and respiratory. In fact, many of the data are imprecise; therefore, it is difficult to give credence to these results. The same group tried to reproduce these results with an improved design but without success (3,32 Table 3. Experimental studies on subjects with skin disorders associated with VDU use.

Swanbeck and 30 patients referred to dermatologists for Double-blind crossover design: Two PCs (A and B)
Most of the patients experienced their usual skin Bleeker skin problems felt to be caused by VDU use with different emissions were used successively. problems (mostly heat or reddening, itching, (1989) (28) Half of them had one of the following skin Magnetic field (1-300 kHz) intensity at 30 cm stinging) when working with VDU but no difference Sweden problems before starting to work on VDU: in front of the VDU: 50 nT (A) and 800 nT (B) between exposure to computer Aor B: 22 reacted eczema, seborrhea, dryness, psoriasis, electrostatic field: 0.2 kV/m (A) and 30 kV/m (B). to computer A and 23 to computer B. Symptoms rosacea, or ictyosis Three hours work randomly assigned on A or B on 2 remarkably reduced when relative humidity was consecutive days. Patients filled out questionnaires increased from 25 to 60% with no difference about symptoms. Dermatologist evaluation 20 min regarding the type of VDU. Of the 13 reacting at and 4-20 hr after the exposure.
high humidity, 11 reacted when exposed to a VDU switched off and covered with a cloth. Only slight redness was found on some of the patients with complaints, but one patient had a Quincke's edema.
Oftedal et al. 20 subjects with skin symptoms associated Double-blind crossover design: Subjects working with Symptoms were less frequent when filters (1995) (26) with work on VDU. 5 subjects had already VDU during 6 weeks at their own workstation: 2 weeks were used and the reduction was stronger for Norway a facial skin disorder (acne, seborrheic without screen filter, 2 weeks with one filter, 2 weeks the first filter used. Tingling, pricking or itching dermatitis, and atopic dematititis). 12 with an other filter. Exposure to each filter was randomly were significantly less pronounced when active subjects had already experienced fewer selected: one supposed to be active and the other filters were used but not other symptoms. symptoms after using a screen filter.
inactive. The two filters reduced significantly the electro-Dermatologic evaluation: no difference between static field but active filters reduced more effectively active and passive filter. Symptoms more ELF and VLF electric fields. No reduction of magnetic pronounced on days with long duration of work field (ELF or VLF) was provided by the filters. Symptoms with VDU. evaluated by questionnaire each day and signs evaluated by dermatologist at the end of each exposure period.
Andersson et al. 16 patients referred by occupational Double-blind crossover design: After a rest period of 15 Subjects could not discriminate between the (1996) (7) physicians and dermatologic clinics. min, patient seated for 30 min in front of a personal two exposure conditions ("on" or "off"). No Sweden Inclusion criteria: clear subjective reactions computer (PC) at a 50-to 60-cm distance. Each subject relationship between subjective symptoms ratings in the skin of the face with exposed to tested 4 times (twice with PC on and twice off). and the actual presence of the field. Symptoms environments with electricity for at least 6 Magnetic fields intensity at 50 cm of the VDU: 245 nT were significantly more intense when subjects months and reacting within 30 (27) distributed to office workers, telephone without filter, then two periods of 3 months of work periods with active or passive filters. Statistical Norway interview, and electric field measurements in with active or passive filter chosen at random. significant reduction of symptoms compared with front of the worker's VDU. Inclusion criteria: Average reduction between active and passive filter at period without filter for: skin symptoms (heat, at least one facial symptom reported in 60 cm of the VDU: 4.3 V/m for ELF and 0.23 V/m for VLF. burning sensation or stinging; tingling, prickling connection with VDU and reduction by a Questionnaire on severity of skin symptoms as well as or itching, sensation of tightness or dryness; filter of ELF or VLF electric field by 40% eye discomfort and nervous system symptoms at the redness or flushing), eye discomfort (stinging or or more. end of each day and a questionnaire on physical and dryness, pain, redness, tiredness, and light psychological factors at the end of each exposure sensitivity), nervous system (headaches, tiredness, period. or fatigue). Symptoms were constant during exposure period regardless of the order of the filter or the intensity of the reduction of the electric fields by the filter.
Lonne-Rahm 24 patients recruited by advertisements in Two double-blind experiments with 12 cases and 12 Patients reported increased skin symptoms when et al.
newspapers or referred by dermatologists. controls: Both groups exposed to 30-min periods of they believed that electromagnetic field was (2000) (29) Inclusion criteria: minimum of 6 months of high or low stress, with and without exposure to turned on. No differences were found between Sweden skin symptoms, reported to appear within 30 electromagnetic fields from a VDU. Matched controls "on" and "off" blind exposure. Inflammatory minutes of exposure to EMF. 12 controls were tested twice with similar exposure but with the mediators and mast cells in the skin were not matched to cases for age, gender, and fields turned on every time. Stress induced by requiring affected by the stress exposure or by exposure to pigmentation ability. participants to act with random sequence of flashing EMFs. No effect of the fields on hormone levels lights while solving mathematical problems. Magnetic and no difference between cases and controls fields intensity measured at 50 cm from the VDU: 198 nT for blood hormones. (ELF), 18 nT (VLF). Electric field intensity: 12V/m (ELF) and 10 V/m (VLF). Blood samples for adrenocorticotrophic hormone, prolactin, growth hormone, melatonin. Skin biopsies analyzed for the occurrence of mast cells.
Sandström et al. (9) presented a report of a challenge with flickering light in 10 patients with HSEMF symptoms and 10 controls. Patients were found to react more intensely than controls to the exposure, as assessed by visual evoked potentials. The authors concluded that the patients labeled as HSEMF are hyperreactive to environmental stimulation such as flickering. Because of its sample size, this study should be considered preliminary, and no relation was evident between the findings and the symptoms reported by HSEMF patients.
More recently, Trimmel and Schweiger (33) reported the results of a double-blind trial to evaluate the effect of a 1-hr exposure to ELF (50 Hz, 1 mT) on concentration and memory. They found that among 66 volunteers, subjects self-rating themselves as sensitive to EMFs tended to perform less well than others when exposed to noise and EMFs. Exposure to noise only had no effect, but the effect of EMFs only was not evaluated, and few details are given on the exposure setting.
In summary, most of the experimental literature is concerned with VDU skin disorders. At present there is no scientific evidence for a link of these disorders with exposure to EMFs, either ELF or VLF. The general syndrome of HSEMF has not been seriously evaluated by researchers. Two recent preliminary studies found that patients labeled as HSEMF reacted differently to different environmental exposures (flickering light, noise plus EMFs) from non-HSEMF patients.

Discussion
The result of the literature review is rather meager. Few studies have been published on the subject of HSEMF in peer-reviewed journals. Most of the studies published on HSEMF come from Nordic countries and are concerned with nonspecific skin disorders related to VDUs. Few studies have been conducted in other countries and almost nothing comes from North America. The evidence of the existence of a more general syndrome associated with HSEMF (including such different nonspecific symptoms of the nervous system as fatigue, dizziness, headache, and depression) is still very weak.
No evidence exists of a link between VDU skin disorders and exposure to EMFs, but there is some evidence of a link with organizational factors and possibly physical factors such as humidity. Moreover, the provocation studies aimed at evaluating the effect of EMF exposure in a double-blind setting failed to reproduce the symptoms of labeled HSEMF patients, and several indicators demonstrated the important psychological factors in the emergence of such a health problem.
Globally, the largest amount of the evidence pleads against a role of EMFs in the reported symptoms and, moreover, its existence in North America has yet to be demonstrated. But the quality of the research on this subject is limited. No good descriptive study is available on the burden of the health problem at a population level, and most of the etiological research on HSEMF has concentrated on VDU exposure. Methodological problems are also an issue.
First, most if not all the cases reported are of subjects who diagnosed themselves as HSEMF cases. No clear case definition exists, and no recognizable criteria are available to confirm this diagnosis. Presentation of symptoms and the alleged causes for the symptoms vary greatly from one country to another, and there is doubt about the specificity of the cases reported. Developing a case definition for such a symptom-based condition is not a simple task, but it is a necessity to improve study quality (34). Some authors have speculated on the possible relation to multiple chemical sensitivity and other related clinical portraits (24). This relationship certainly should be clarified to evaluate the specificity of the HSEMF syndrome.
Most studies on HSEMF are also limited by the data available on the exposures reported by subjects or evaluated in studies. The descriptions of the exposure triggering the symptoms is usually rather vague. In general, the exposure reported refers to sources such as VDUs, which are not usually recognized as important sources of high-intensity exposure to EMFs (35,36). However, the importance of computer use on personal exposure to 60-Hz magnetic fields when considering 24-hr exposure was recently demonstrated (37). Most of the controlled studies did not evaluate the effect of different kinds of exposure to EMFs (e.g., varying frequency, intensity and time course of exposure) but instead focused on a simple exposure setting corresponding to what was usually reported by patients. Usually, no data on quality control of the exposure setting were provided.
Because of the absence of a good case definition and the limited methodology of the studies on HSEMF, it is difficult to determine completely the nature of this possible health problem. The fact that SAGs seem to attract a large number of people who claim that they suffer from HSEMF is rather intriguing (38). More studies are certainly needed to clarify the nature of the health problem labeled HSEMF.
To my knowledge, few expert groups have reviewed the literature on this topic. In 1991, the International Radiation Protection Association, via its Non-Ionizing Radiation Committee, issued a statement regarding the "alleged radiation risks from visual display units" (39). It concluded its review: "Based on current knowledge, there are no health hazards associated with radiation or fields from VDUs." Further research on the possibility that skin disorders may be related to VDU work was recommended.
In 1994 an advisory group of the National Radiological Protection Board of the United Kingdom published a report on health effects related to the use of VDUs (40). The report focused mainly on reproductive outcomes, but a section was devoted to skin problems. It concluded, Skin diseases do not appear to be caused by the electric fields from VDU, although there is anecdotal evidence unsupported by epidemiology that in conditions of low humidity the associated electrostatic fields may aggravate existing skin problems.
In 1997, the European Commission presented a report on the "possible health implications of subjective symptoms and electromagnetic fields" (3). It concluded, "The review was unable to establish a relationship between low or high frequency fields and electromagnetic hypersensitivity." They recommended adequate handling of seriously afflicted individuals. Because of "the inability to clearly describe the syndrome and causation of electromagnetic hypersensitivity," further research was also recommended.
Finally, in its Working Group report on EMF health effects, the National Institute of Environmental Health Sciences presented a brief review of the topic of electromagnetic hypersensitivity (4). This section concluded: Some individuals have subjective symptoms apparently related to [VDU] use in the office environment. The evidence is inadequate to relate such symptoms to the EMF associated with that use. . . . No high-quality double-blind challenge studies have been conducted that conclusively establish the existence of sensitivity to EMF.
In other respects, I consider that the issue of hypersensitivity should not be limited to the HSEMF studies reviewed in this article. In a broader sense, hypersensitivity could mean the greater susceptibility of an individual to EMF effects. This could potentially be found for different outcomes possibly related to EMF exposure. For instance, some studies found that certain subjects might be more sensitive to the effect of EMFs on melatonin secretion (41,42). Although this is still preliminary evidence and not synonymous with adverse health effects, it seems to support the possibility of individual susceptibility to EMF exposure. Research on such a topic should focus not only on the rather nonspecific symptoms of hypersensitivity described in HSEMF reports but also on well-diagnosed illness.
Individual variations to electric field perception have been described previously Environmental Health Perspectives • VOLUME 110 | SUPPLEMENT 4 | AUGUST 2002 but at higher intensities than those usually found in the environment and without reference to symptoms of HSEMF (4). As a matter of fact, the field intensities used in the controlled studies reviewed were not perceived by the patients suffering from HSEMF. Recently, Leitgeb (32) described variability in the perception of induced currents in 606 subjects. Although 2% of the sample seemed particularly sensitive to the currents, no individual reported symptoms of HSEMF. Although the issue of hypersensitivity is still open, it seems clear that there are variations of perception of EMF exposure, but this does not appear to be related to HSEMF symptoms.

Conclusion
To date, the literature on hypersensitivity to EMFs is rather meager and suffers from methodological problems. Most of the published studies were done in the Scandinavian countries and focused on dermatological disorders. The other clinical portraits are rarely well described. Globally, case definition is unclear, and few population studies have evaluated the prevalence of this disorder. The most-studied clinical portraits (dermatological problems associated with VDU work) were evaluated in case-control and in controlled studies, and no consistent relationship was found with EMF exposure.
In conclusion, I found no substantial grounds on which to build a framework for helping a risk assessor to take into account the alleged "HSEMF syndrome." Our knowledge of the nature of the problem seems too vague to integrate it into an EMF risk assessment protocol. But there are certainly grounds for further research to assess more carefully its nature and its possible burden in North America.