Bioelectromagnetics 00:1^8 (2006)
Author Proof The ‘‘Mainzer EMF-Wachhund’’: Results From a Watchdog Project on Self-Reported Health Complaints Attributed to Exposure to Electromagnetic Fields
A Joachim Schu«z,1,2* Christian Petters,1 Ulrich Tiber Egle,3 Bernd Jansen,4 Renate Kimbel,4 Stephan Letzel,5 Wilfred Nix,6 Lutz Schmidt,7 and Lutz Vollrath8 1
Institute of Medical Biostatistics, Epidemiology, and Informatics, University of Mainz, Germany 2 Institute of Cancer Epidemiology, Danish Cancer Society, Copenhagen, Denmark 3 Clinic for Psychosomatic Medicine and Psychotherapy, University of Mainz, Germany 4 Department of Hygiene and Environmental Medicine, University of Mainz, Germany 5 Institute of Occupational, Social, and Environmental Medicine, University of Mainz, Germany 6 Neurological clinic, University of Mainz, Germany 7 Psychiatrical clinic, University of Mainz, Germany 8 Department of Anatomy and Cell Biology, University of Mainz, Germany
The ‘‘Mainzer EMF-Wachhund,’’ a watchdog project, offered a system of self-notification of health complaints attributed to exposures to electromagnetic fields (EMFs) to a population of a part of Germany with about 4 million inhabitants. By using a self-administered questionnaire, which was provided online and for download from the Internet, 192 persons reported such health complaints in the period from October 2003 to March 2005. Of these, 56% classified themselves as electromagnetic hypersensitive (EH). Predictors of this self classification were being affected by all kinds of EMF rather than single EMF sources and being female. On average, EH subjects reported a high degree of suffering, 77% of whom had already sought advice from physicians. An Internet-based standardized questionnaire is an economic way of offering affected persons a direct link to scientific institutions to establish contact. However, the study base obtained by such an approach is not representative to estimate a population-based prevalence. As a large number of subjects did not classify themselves as EH and reported very specific links between exposure and symptoms, they may provide a very distinct and interesting group for future research. Bioelectromagnetics 00:1–8, 2006. 2006 Wiley-Liss, Inc. Key words: electromagnetic hypersensitivity; adverse effects; unspecific health symptoms; cellular telephones
INTRODUCTION
Extensive research has been conducted into possible health effects of exposures to electromagnetic fields (EMFs). So far, scientific evidence suggests that no adverse health effects result from exposure levels below the international guidelines proposed by the ‘‘International Commission on Radiation Protection’’ [International Commission on Nonionizing Radiation Protection (ICNIRP), 1998]. Nevertheless, in some areas there is scientific uncertainty because some technologies are recent developments and current research cannot rule out effects from long-term exposure or because study results are ambiguous, as in the case of an association between risk of childhood leukemia and exposures to extremely low-frequency magnetic fields. In the latter case an empirical association emerges from large-scale epidemiological > 2006 Wiley-Liss, Inc.
studies but coherence with experimental research is weakQ2[International Agency for Research on Cancer (IARC), 2002]. Most challenging in this field, however, is the manifold anecdotal evidence from people attributing ————— — Grant sponsor: Ministry for the Environment of the federal state of Rhineland-Palatinate, Mainz, Germany. *Correspondence to: Dr. Joachim Schu¨z, Institute of Cancer Epidemiology, Strandboulevarden 49, Dk-2100 Copenhagen, Denmark. E-mail:
[email protected] Received for review 25 July 2005; Final revision received 28 November 2005 DOI 10.1002/bem.20212 Published online in Wiley InterScience (www.interscience.wiley.com).
BEM/941G- 05
2
Schu«z et al.
Author Proof
health complaints such as headache, sleep disorders, or skin irritations, to various exposures to EMF, in clear contrast with results from scientific studies and lacks convincing explanations [Silny, 1999; Levallois, 2002]. Candidate explanations range from true EMF effects with underlying biological mechanisms that scientists have failed to uncover yet, through health effects from other causes which are only attributed to EMF sources or health symptoms due to anxiety about EMF exposures, to psychosomatic diseases expressing themselves in a fixation on EMF as an overwhelming health hazard. While none of these explanations can be ruled out a priori, the evidence for true EMF effects is very weak [Seitz et al., 2005]. Irrespective of whether the association is causal, the concerned or affected subjects require skilled medical help and appropriate risk communication. A recent German survey suggests the prevalence of subjects who attribute health complaints to EMF exposures is not negligible. In a sample of 2500 interviewees, 8% specifically attributed health complaints to exposures from mobile phone base station antennas or the use of mobile or cordless phones [Institut fu¨ r angewandte Sozialwissenschaft (infas), 2004]. In Sweden, 1.5% of the population claimed to be hypersensitive to EMF, as did 3.2% in California [Hillert et al., 2002; Levallois et al., 2002]. Considerable variation across countries, regions within countries, and surveys in the same regions has been noted before. In 1997, a European expert group reported that electrical hypersensitivity had a higher prevalence in Sweden, Germany, and Denmark than in the United Kingdom, Austria, and France [European group of experts, 1997]. All these data suggest the true number is still uncertain and the topic merits further research. The ‘‘Mainzer EMF-Wachhund’’ was initiated in 2003 in the German federal state of RhinelandPalatinate (about 4 million inhabitants), as a watchdog project to first collect information directly from affected subjects. The project is divided into three phases, with a next phase starting only if the previous one was successful. Phase 1 (contact phase) comprised the installation of a registry for self-reports for subjects attributing health complaints to all kinds of exposures to EMF. In phase 2 (clinical phase), registered subjects from phase 1 are invited for clinical and psychosomatic examinations, together with a random control group from the population, to identify shared characteristics of the affected subjects. Additionally, treatment options will be discussed with the patients. Phase 3 (permanent phase) starts with the installation of a permanent advisory service at the University hospital, that offers guidance for persons with complaints attributed to exposure to EMF. This paper presents results from
phase 1. This phase addressed two goals: (1) To explore whether the system for self-notification is accepted by the target group; (2) to explore results from this phase by primarily focusing on differences between subjects who regard themselves as hypersensitive to EMF and those who experience symptoms attributed to EMF without regarding themselves as hypersensitive.
A METHODS
The website of the ‘‘Mainzer EMF-Wachhund’’ went online on October 15, 2003, accompanied by a press conference of the Ministry for the Environment of Rhineland-Palatinate, Mainz, Germany and an article in the medical journal of the area [Schu¨ z and Vollrath, 2003]. All local newspapers and radio stations provided a full presentation of the project and a link from their own websites to the watchdog homepage. The main part of the website was a standardized questionnaire to be either filled in online or downloaded as a PDF file to be filled in by hand and sent to the University of Mainz through mail. Additionally, all clinicians and general practitioners (GPs) had free access to the questionnaire via their professional association, and paper copies of questionnaires were sent out by the University on personal request. The website was open until March 1, 2005. Self-help groups and environmental activist groups of the area were specifically informed about the project. The web-based questionnaire, the downloadable questionnaire, and the paper version provided to the GPs or sent out on request were identical. In principle, every inhabitant of Rhineland-Palatinate who experienced something noticeable in relation to EMF, was invited to participate. This included subjects having technical questions or general concerns. The project was approved by the Data Protection Officer of the University clinic. As an extension, we sent a letter to all households (n ¼ 5500) in two selected suburbs of Mainz and informed them about the project. On a short reply letter we asked whether they had heard about the watchdog project, whether they had experienced symptoms in relation to EMF exposures and, if both questions were answered with ‘‘yes,’’ why they had not registered themselves. The letter was distributed approximately 1 year after the start of the project, that is in December 2004. The watchdog questionnaire contained personal questions (including age, gender, education, and family status), a list of 30 symptoms possibly experienced during the last month, including a rating of the severity of the symptoms (severe, strong, moderate, mild, not at all), a list of potential sources of EMF exposures to which the subject attributed their symptoms, questions
The ‘‘Mainzer EMF-Wachhund’’Q1
Author Proof
on actions undertaken to reduce the symptoms or eliminate the exposures, and a field for comments regarding the subject’s own observations. Furthermore, subjects had to classify themselves as being hypersensitive to EMF or not. Names and addresses of subjects were asked and about 50 subjects were called by phone to verify some of the data and to get feedback on the usefulness of the questionnaire. The symptom list was based on a scale developed in the psychosomatic clinic for anamnesis of their patients, modified for the purpose of the watchdog project and based on experiences from another project involving subjects claiming to be hypersensitive to EMF exposures [Frick et al., 2002]. The symptom list is given below in Table 3. As the project was not a systematic, populationbased survey, we made only internal comparisons within the study base. In our analysis, we first contrasted subjects classifying themselves as hypersensitive to EMF with others reporting health complaints, using logistic regression analysis with self-reported electromagnetic hypersensitivity (EH; yes, no) as the dependent variable and age, sex, education, living situation, and type of EMF exposure as independent variables. Furthermore, we used logistic regression models again to identify predictors of strong to severe health complaints using the same independent variables as above and, additionally EH. Predictors of the total burden from such health complaints were evaluated using a general linear model. For the dependent variable of the latter model, we created a summary score from the 30 symptoms by giving 4 points for every symptom classified as severe, 3 points for strong symptoms, 2 points for moderate symptoms, 1 point for mild symptoms, and 0 point otherwise. A comparison across groups by sex, age, education, living situation, EH, and type of EMF sources was performed using the Wilcoxon test (2 groups) or Kruskal-Wallis test (>2 groups), respectively.
questionnaire had a disclaimer saying that subjects living outside of Rhineland-Palatinate will not be involved in the phase 2 project. One notification was received from abroad. Notifications of children were all filled in by their parents. The survey of 5500 households did not result in one single additional registration. We received positive reply letters from 28 households (0.7%) reporting 41 persons having symptoms attributed to EMF exposures. From these 28 households, 12 indicated they would register in the future (but never did), while 16 never considered to register. The following reasons were given in these 16 reply letters (five options plus an ‘‘other’’ option were provided): ‘‘I see no benefit for me’’ (n ¼ 4), ‘‘I have no internet access’’ (n ¼ 4), ‘‘I don’t want to be in a register’’ (n ¼ 3), ‘‘I don’t trust the institution running the registration’’ (n ¼ 2), ‘‘I feel already sufficiently looked after’’ (n ¼ 0), and no particular reason (n ¼ 3). In one-fourth of all reply letters it was indicated that subjects had heard of the watchdog project. Table 1 illustrates the demographic characteristics of the 192 subjects reporting health complaints in relation to EMF exposures. For each group, the proportion of subjects classifying themselves as hypersensitive to EMF is given as well as the proportions attributing their symptoms to different types of EMF exposures. In Table 2, the results of the logistic regression analysis for being hypersensitive to EMF are shown by odds ratios and respective 95% confidence intervals. By doing this, the potential predictors are adjusted for each other in a multiple analysis. Female gender and reporting health complaints due to various EMF sources (as seen in the reduced odds ratios for the other sources) were the strongest predictors for classifying oneself as being hypersensitive to EMF. Figure 1 shows the summary scores for 182 affected subjects (10 (5%) gave no complaint profile), divided into EH subjects and non-EH subjects. Using a general linear model, in which potential predictors were adjusted for each other, the strongest predictors for high summary scores were EH (P
