Objective: The aim of this first phase of a cross-sectional study from Germany was to investigate whether proximity of residence to mobile phone base stations as well as risk perception is associated with health complaints.
Methods: The researchers conducted a population-based, multi-phase, cross-sectional study within the context of a large panel survey regularly carried out by a private research institute in Germany. In the initial phase, reported on in this paper, 30 047 persons from a total of 51 444 who took part in the nationwide survey also answered questions on how mobile phone base stations affected their health. A list of 38 health complaints was used. A multiple linear regression model was used to identify predictors of health complaints including proximity of residence to mobile phone base stations and risk perception.
Results: Of the 30 047 participants (response rate 58.6%), 18.7% of participants were concerned about adverse health effects of mobile phone base stations, while an additional 10.3% attributed their personal adverse health effects to the exposure from them. Participants who were concerned about or attributed adverse health effects to mobile phone base stations and those living in the vicinity of a mobile phone base station (500 m) reported slightly more health complaints than others.
Conclusions: A substantial proportion of the German population is concerned about adverse health effects caused by exposure from mobile phone base stations. The observed slightly higher prevalence of health complaints near base stations can not however be fully explained by attributions or concerns.
Statistics from Altmetric.com
Two recently released reports by the European Union (EU) reveal the differences between scientific evidence and subjective perception of possible hazards in relation to mobile phone base stations. According to a scientific report by an EU expert group1 there are no established adverse health effects related to exposures to radio frequency electromagnetic fields (RF-EMFs), as long as the respective guidelines for public protection are not exceeded. The expert group however noted that the database for evaluation, especially for long-term low-level exposure, remains limited. The guidelines referred to have been suggested by the International Commission on Non-Ionising Radiation Protection (ICNIRP)2 and are the basis of the EU Council Recommendation of 12 July 1999 for the limitation of exposure of the general public to electromagnetic fields. On one hand, measurements carried out in households in the vicinity of mobile phone base stations have shown that exposures from these antennas hardly reach a few per cent of the ICNIRP protection limits, even in apartments close to the stations.3 On the other hand, a survey among EU citizens in 25 member states showed that 40% of respondents considered that mobile phone base stations affected their health to some extent, while 36% even considered this to be true to a large extent.4 Some experimental laboratory studies have looked into a similar effect, known as the nocebo effect, an inverse of the placebo effect. The adverse symptoms experienced in the nocebo effect occur due to expectations, that is, due to concerns. Results of a systematic review of provocation studies involving electromagnetic hypersensitive (EHS) and non-electromagnetic hypersensitive individuals conducted by Röösli showed that the majority of EHS individuals were not able to detect RF-EMFs under double-blind conditions. In some of the reviewed studies, EHS individuals reported reactions during sham conditions more often than the general population.5 Unfamiliarity with the technology and the involuntary nature of RF-EMF exposure from mobile phone base stations are likely to promote these concerns.
Our knowledge about such worries is also based on numerous anecdotal reports from persons attributing health complaints to various sources of electromagnetic fields. Whereas it seems to be established that many affected persons suffer severely from such health complaints,6 7 the evidence that these complaints are causally linked to the RF-EMF exposure is very weak.1 8 9 This assessment is however mainly based on the absence of consistent effects in human provocation and sleep studies and on the lack of a biologically plausible explanation.1 5 7 10 11 Very little is known about how these health complaints relate to true RF-EMF exposure, and/or to the proximity to mobile phone base stations outside laboratory environments.1
Outside controlled conditions, exposure to RF-EMF is rather difficult to measure and varies considerably in time and space. It has been clearly demonstrated that proximity to a mobile phone base station is a very weak surrogate measurement for true RF-EMF exposure, as there is a mixture of exposures of several magnitudes some 100 m within the vicinity of the mast.12 This is due to the fact that walls and other houses obstruct the emission path and act as shields within the area of a theoretical distance–exposure relationship, which is the main beam of the antenna. Conversely, scattered fields and reflections lead to higher exposures also outside the main beam. Due to the antenna’s side lobe patterns, exposure does not decline monotonously with distance. However, at a distance of more than 500 m, exposures can be assumed to be generally low, providing a crude basis for categorisation of exposure applicable for descriptive purposes in large populations. For formal investigations of associations however, measured RF-EMFs from mobile phone base stations and other relevant sources (TV and radio broadcast towers, cordless phone base stations) are essential.
We performed a comprehensive population-based, cross-sectional study in Germany with the following aims:
To estimate the prevalence of concerns about mobile phone base station-related adverse health effects and attributions of health complaints to RF-EMFs emitted from mobile phone base stations in the general population and particularly among persons living in the proximity of them.
To investigate whether there is an association between health complaints and the exposure to RF-EMFs as measured in the participants’ residences.
The study was conducted in two phases. Phase 1 dealt with the first aim, while the second aim was dealt with in phase 2 (see page 124). This article will report on the results of the first phase.
The study was embedded in a large panel study that is regularly carried out by the Taylor Nelson Sofies (TNS) Infratest/Test Panel Institute (TPI). The panel bases on a large household sample of persons who agreed to participate in future surveys. The current panel of the Institute in Germany comprises more than 73 000 households nationwide, representative for the German population aged 0–79 years. The households are selected on a voluntary basis via e-mail contacts, address publishers or snow-ball systems. The institute has information on the sex, age and income group of all persons residing in the households in their database, as well as the region in which they reside.13 14 The institute’s survey conducted between August 2004 and November 2004 involved 51 444 households with persons aged 14–69 years. In co-operation with the institute and using their data, one person was randomly selected from each household using a randomisation process. This was done such that in the end we had a study population representative for the general German population with regards to age, sex, income group and region of residence. The participants were informed in writing that completing the questionnaire and answering individual questions was voluntary. As is usually the case with this type of panel survey conducted by marketing research companies and in the case of social science, ethical approval was not sought.
A total of 30 161 persons responded to the request (58.6%). After the exclusion of 114 incorrectly filled-in questionnaires, a total of 30 047 questionnaires (58.4%) were available for the analyses. The mailed questionnaire included a list of 38 symptoms that have been reported in previous studies to be possibly associated with RF-EMF exposure (Frick list).15 The symptoms covered in the list range from cognition problems, pain, sleeping problems, skin reactions, gastrointestinal affections, to visual problems. Each of the 38 symptoms were ranked as “not at all”, “little”, “moderate“ or “strong”; hence, a higher summary score indicates poorer well-being. Participants were also asked in two questions whether they are worried about health effects of mobile phone base stations in general (concerns), and whether they believe that their health is adversely affected by mobile phone base stations (attribution). They could answer either “yes” or “no”.
The questionnaire included questions on demographic characteristics. In addition, the subjects were asked to estimate the distance to the next mobile phone base station.
The distance between residence and mobile phone base station was additionally computed based on geo-codes of base station locations and participants’ addresses. Locations of all 51 000 mobile phone base stations in Germany in 2004 provided by the respective German authority (“Bundesnetzagentur”) were geo-coded by the network operators, and household addresses by the Post Direct GmbH, a commercial service provider. The distance estimation based on geo-coding was successful for 29 805 participants (99.2%).
Distance was dichotomised into “not exposed” (>500 m) and “possibly exposed” (⩽500 m).
The study procedure guaranteed data security and anonymity of participants during analyses. The postal questionnaires were sent and collected by the holder of the TPI, which also ordered the geo-coding. All data were forwarded to the TNS Institute and further to the University of Bielefeld and Mainz as anonymous datasets. The linkage for consolidating geo-coordinates and questionnaire data were the so-called TPI-Identity numbers. All requirements of the German Data Protection Act were adhered to.
Concerns about health effects and self-reported adverse health effects attributed to mobile phone base stations (attribution) were combined in a newly created three-level item on risk perception, with the categories “no concern and no attribution”, “concern but no attribution”, and “attribution”. In some analyses, the last two categories were combined.
The prevalence distribution of risk perception and the quartiles of the summary score of the Frick list (38 complaints) were examined by age group, region and by gender. In a non-responder analysis, we compared the educational levels of the responders and non-responders. We also attempted to estimate the extent of any error resulting from non-response. For this purpose we analysed the responses of the participants regarding concern about possible adverse health effects from mobile phone base stations, according to the time span for response to the questionnaire. Participants were placed into three groups: early responders, moderate responders, and late responders. Our model, the delayed response model, is based on the assumption that persons who respond late to questionnaires are more similar to non-responders than early responders.
We also performed regression analyses for these variables to adjust for demographic differences (including age, gender, region and social status in the model). For risk perception we conducted logistic regression analysis with “no concern and no attribution” versus “concern or attribution” as dependent variables. A multiple linear regression model was applied using the summary score of the Frick list as dependent variable, and sex, age and rural areas as independent variables. Comparisons of the responders with population-based numbers show slight differences in age and sex between urban and rural areas, so that additional analyses are based on weighted values. Weights are calculated based on known values of age, sex and community size distribution in the German population and the deviation in the study population. Social status was represented by two variables, namely school education and family income.16
Furthermore, we analysed the relationship between subjective and objective distance to the nearest mobile phone base station. Regression analyses of the relationship between subjective and objective distance and cognition of health risks were also performed, adjusted for demographic differences.
Of the 30 047 participants, 27 376 filled in the questions on concerns and attributions. Eighteen point seven per cent reported that they were concerned about health effects from mobile phone base stations and 10.3% attributed adverse health effects to RF-EMFs from mobile phone base stations. Although the majority of the 2819 participants with attributions also reported being concerned about health effects, 154 (5.5%) reported not to be concerned about health effects.
The prevalence of concern and attribution varies by age, income, education and region, but not by gender (table 1). Participants with higher social status were more concerned than others, but had a lower prevalence of attribution. There were also regional differences with a tendency that participants from south Germany seemed to be more worried and more affected than participants from other parts of Germany.
According to the geo-coding information, 14 503 out of 29 805 residences were 500 m or closer to the nearest mobile phone base station. Hence, the prevalence of living in the proximity of a mobile phone base station, weighted for community size (see Methods), was 51.5%. A comparison of the participants’ responses to distance to the nearest mobile phone base station and information from the geo-coding showed that 12 402 (41.6%) had correctly classified the distance as less or more than 500 m, 3376 (11.3%) had correctly estimated the distance to be ⩽500 m, and 9026 (30.3%) had correctly estimated the distance to be >500 m. A total of 8852 participants (29.7%) wrongly estimated the distance, with 7170 (24.1%) wrongly estimating it to be up to 500 m and 1682 (5.6%) wrongly estimating it to be more than 500 m. Eight thousand three hundred and fourteen (27.9%) were not sure about the distance and 237 (0.8%) did not respond to the question.
Applying a logistic regression model including all relevant variables adjusted for each other, the odds ratios confirmed the findings from the univariate analysis (data not shown), implying that all these factors act independently. Persons stating that they lived in the vicinity of a mobile phone base station significantly reported more worries than other participants (odds ratio 1.35, 95% CI 1.25 to 1.45) (data not shown). However, taking the true distance based on geo-coding into account (two categories, ⩽500 m and >500 m), the odds ratio was 1.00 (0.94 to 1.07).
The quartiles of the summary score of the Frick list of complaints showed little variation by age, region, education, income and other demographic variables, but some differences by gender (higher among women) (data not shown). Participants with concerns had similar quartile values as the reference group, while those who attributed health complaints to mobile phone base stations had higher values. The linear regression model confirmed the gender difference concerning the complaints and suggested little variation with increasing age up to the age of 59 years, and inverse associations with income and education (table 2). It also yielded a slightly higher level of health complaints among concerned participants. Among participants attributing health complaints to the nearby mobile phone base station the score was considerably higher. The summary score of health complaints among people living in the vicinity of mobile phone base stations (⩽500 m) was slightly higher than among those living more than 500 m away from the next mast (table 2).
The non-responder analyses showed only minor differences in educational level between responders and non-responders. We observed a positive trend between time of response to questionnaire and proportion of persons concerned about adverse health effects of mobile base stations. In the early responder group, 24.7% of the participants expressed concern, compared with 31.9% in the late responder group (table 3). Assuming that late responders do have similar response behaviour to non-responders, 6826 (31.9% of non-responders) non-responders would be expected to have reported concerns about adverse health effects. Correcting for this factor would result in a marginal increase in the proportion of participants reporting concern from 28.5% to 29.6% (table 3).
Our study showed that 18.7% of persons aged between 14 and 69 years are concerned about health effects of mobile phone base stations and that 10.3% attribute adverse health effects to the exposure from the masts. There were prevalence differences between north and south Germany (higher in the southern part), across age groups (highest among 30–59 year olds), and across social classes by income and education (lower with increasing level). The score of a list of 38 health complaints (Frick list) showed a clear association with concerns and attributions of health effects to mobile phone base stations and a very weak association with the distance to the nearest mobile phone station.
Our study is so far the largest one in which the possible relationship between proximity of living near a mobile phone base station, concern, attribution and health complaints was investigated. The strength of our survey was that it was not immediately obvious that the questions on health complaints and mobile phone base stations were related to each other, as they were part of a much longer panel questionnaire. Also, our study was representative for the German population and not restricted to areas where persons were already actively involved in actions against the erection of mobile phone base stations.
There is no scientific evidence that radio frequency electromagnetic fields emitted from mobile phone base station antennas are associated with the occurrence of adverse health effects.
There is considerable public concern that living in the vicinity of a mobile phone base station has adverse effects on health.
Until now no large-scale field study addressing this topic has been conducted.
A substantial proportion (about 27%) of the German population is concerned about possible adverse health effects when living in the vicinity of a mobile phone base station and every tenth German attributes health complaints to exposures from these masts. Fifty-one point five per cent of the study population lived in the proximity of mobile phone base stations (⩽500 m distance). Concern about and attribution of adverse health effects to mobile phone base stations are associated with health complaints. The observed slightly higher prevalence of health complaints near base stations can however not be fully explained by attributions or concerns.
Nevertheless, some limitations have to be discussed. As in many postal surveys, the response rate was below 60%. Additionally, persons who had agreed to participate in the panel may have already been a selective subgroup of the general population. However, as the investigation was not only on health aspects and also not introduced as such, it can be assumed that neither attitude towards mobile telecommunication nor specific health problems influenced the response rates. A small group of participants, 5.5%, attributed health effects to mobile base stations but did not report any concerns about health effects. This could imply that some participants had interpreted the attribution item to mean that base stations may cause health problems in general, although they were not affected themselves.
The crude exposure categorisation, namely defining a 500m circle around the nearest mobile phone base station as the exposure area and neglecting exposure from other RF-EMF sources, allows only a descriptive statistical approach. The amount of exposure misclassification introduced by this method is presumably large12 and the resulting bias has to be expected to underestimate possible associations, if there were any.17
In general, cross-sectional studies are limited as exposure and outcome are assessed at the same time, making it difficult to draw conclusions on the temporal relation of cause and effect. In our survey however, the situation concerning proximity to mobile phone base stations and health was expected to be more or less stable for the vast majority of participants, such that many of the investigated complaints were not occurring or vanishing within a short period of time.
For the observed association between attribution of health effects to the mobile phone base stations and the actual occurrence of health effects, it is difficult to say whether persons with persisting health complaints attribute these symptoms to the mobile phone base station in an effort to identify a cause, or whether high levels of anxiety, depression, and stress together with regarding the mast as a hazard promotes health complaints, or a mixture of both. Furthermore, as the attribution item inherently assumes the existence of health complaints, the relationship between attribution and the health score from the Frick list was to be expected. Of more interest is thus the group of participants who reported concerns, but no attributions. This group had a significant regression coefficient of 1.45, lending support to the idea of a general psychological construct, leading to an increased awareness of environmental hazards and a higher likelihood of expression of health complaints.
It can be argued that the investigation of adverse health effects caused by exposure to mobile phone base stations requires a different study design,5 such as a case-control study as used for investigating cancer risk in relation to RF-EMF exposure,18 or a cross-over design.
This comprehensive nationwide, cross-sectional study in Germany shows that a substantial proportion of the population is concerned about possible adverse health effects in relation to mobile phone base stations, and almost every tenth German attributes health complaints to the RF-EMF exposure from nearby masts. More than half of the study population lived in the proximity of mobile phone base stations (⩽500 m distance). Our finding that concern about and attribution of adverse health effects to mobile phone base stations are associated with health complaints supports the findings of the survey carried out among EU citizens.4
The weak association we found between mobile phone base stations <500 m and adverse health effects confirmed the relevance of the second phase of our study, in which actual RF measurements were conducted (see page 124). The observed slightly higher prevalence of health complaints near base stations cannot be explained by attributions or concerns alone.
In conclusion, we believe that the worries and health complaints of people living close to mobile phone base stations need to be taken seriously. Improved risk communication with concerned persons is required.
The full report of this study (in German) is available on request from the authors.
We would like to thank Ernst Schroeder and Claudia Barthold (TNS Healthcare, Munich) and Jörg Riedel (University of Bielefeld) for their support. We would also like to thank Hans-Peter Neitzke and his colleagues from the Ecolog Institute, Hanover, for helpful discussions about how to design exposure metrics. We also thank Florence Samkange-Zeeb for her assistance in editing and submitting this article.
Funding: The study received its funding entirely from the German Federal Ministry for the Environment, Nuclear Safety, and Nature Protection, within the context of the “Deutsches Mobilfunkforschungsprogramm” (German Mobile Telecommunication Research Programme). The study sponsor had no involvement in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Competing interests: None.
We declare that we participated in the study and have seen and approved the final version. MB, GB-B, BS, PP and JS conceptualised the study and developed the study protocol. MB and GB-B were responsible for the conduction of the study. PP was responsible for the survey, which was managed by UR. PP and UR analysed the survey data. GB-B, BK and JB were responsible for data management and, together with SS, for the analysis. JS and MB prepared the manuscript which was jointly finalised by all authors. MB is the guarantor of the work.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.