by M J O'Carroll
This paper aims to give balanced and rational account of public health concerns about exposure to electromagnetic fields (EMFs) from electricity supply systems. It is written to advise non-scientific members of the public and local and national government from a scientific perspective. It was first produced for a BBC Radio 4 Helpline and a Public Local Inquiry. Unlike official and industrial advisory papers it describes the public concerns and their associated field levels, assesses those concerns and puts them in perspective, recognising the official and Industrial position.
Contents
1. Which exposures are of concern?
2. What are EMFs?
3. What is ELF?
4. What levels of fields are of public concern?
5. How do natural and artificial EMFs compare?
6. What health effects are of concern?
7. How does cause differ from association?
8. How convincing is the evidence of association of EMFs and cancer?
9. Is there a plausible biological mechanism linking EMFs and cancer?
10. Are there any established biological effects of EMFs?
11. Are there scientific conceptual barriers to recognising bioeffects?
12. Now do EMFs compare with other hazards?
13. What is the overall authoritative assessment?
14. What are the policy options?
15. Conclusions
Sources and Abbreviations
1. Which exposures are of concern?
The general public is concerned about exposure to EMFs at home, at school and in public places. There is a growing body of evidence suggesting health hazards from residential exposure, and complementary but different evidence on exposure at work, both in roughly the same state of inconclusiveness. While there is also concern about exposure to EMFs at other frequencies, from VDUs, appliances, telecommunications, and natural sources, this paper focuses on fields at supply frequency (50 cycles per second, i.e. 50 Hz, in UK, and 60 Hz in USA, with similar effects). The fields arise most markedly from high voltage power lines, and also from substations and house wiring. EMFs from household appliances typically become insignificant beyond half a metre. Exposure to them is generally short and controlled. While internal domestic EMFs are of increasing interest, the main concern is about long- term exposure to externally imposed fields particularly from powerlines.
2. What are EMFs?
EMF stands for electromagnetic field. There are separate electric and magnetic fields, but they can occur together dynamically. Static magnetic fields can exist separately, e.g. around a magnet or the earth. There are also electrostatic fields, e.g. between battery terminals. Electricity supply systems handle alternating currents (AC) which reverse direction backwards and forwards 50 or 60 times a second. They generate alternating EMFs around them. They have an electric component and a magnetic component, both alternating at the supply frequency. The electric component is screened off by walls so it doesn't penetrate buildings, whereas magnetic fields do. Much of the public health concern is about magnetic fields, although there is increasing evidence implicating electric fields.
Measurements of the magnetic fields are usually made in units of Magnetic Flux Density, often in the international unit called the microTesla. In the USA milliGauss is more common (10 milliGauss = 1 microTesla). Some reports use nanoTesla (1000 nanoTesla = 1 microTesla).
3. What is ELF?
The frequency of EMFs is very important. Many effects are special to frequency bands. There is a range of frequencies from zero (i.e. simply DC or static fields) to over 1020Hz (that's 1 with 20 zeros after it!).
Other well known EMFs occur at frequencies much higher than the 50 Hz of electricity supply, e.g. MW radio around 106Hz, microwaves around 1010Hz, visible light 1015 Hz, X Rays 1017 Hz and so on. They are examples of radiation, which transmits energy. We are concerned with near-fields, not radiation.
ELF stands for Extremely Low Frequency and refers to the range 30 - 300 Hz. This is comparatively puny slice, the potential for harm from radiation increases with frequency. EMF radiation has a "quantum" energy level on an atomic scale; it depends on frequency, not on field strength. Above about 1016 Hz it becomes strong enough to break molecular bonds and is then called ionizing radiation, a recognised potential cause of cancer.
4. What levels of fields are of public concern?
Both the frequency band and the field strength are important. While there are wider concerns, e.g. with fields for mobile phones, EMFs from electricity supply are in the ELF band at 50 Hz. In that band the field levels of official concern in UK are at 1600 microTesla. They are associated with large scale induced currents in the body and are not relevant to the public concern about cancer and many other possible health effects. A landmark study in Sweden associated exposure above 0.2 microTesla with childhood leukaemia. Similar levels have occurred in other studies, some finding associations and some not. This level has been referred to in calls for cautionary policies, and adopted by the Swedish National Energy Administration. A draft major report of the American National Council for Radiation Protection, while not confirmed, recommends firm restrictions at 0.2 microTesla, an indicative level of public concern.
In summary
Description level (microTesla)
NRPB investigation level 1600
under 400 kV powerline 40
public health concern 0.2
average domestic exposure 0.05
background <0.01
5. How do natural and artificial EMFs compare?
EMFs occur naturally across the frequency spectrum and include the light and warmth from the sun, which are much stronger than EMFs from powerlines.
However common fluorescent light tubes have lit up (dimly) when held in the hand under powerlines. The earth's magnetic field gives about 50 microTesla, but that is static and not to be compared with 50 Hz fields. In the supply frequency range there is very little natural background EMF. Artificial fields are easily measurable in the vicinity of powerlines at more than 1000 times background levels, as indicated in the table above.
6. What health effects are of concern?
Most people living or working in relevant exposures report no ill effects. It would be unreasonable to suppose that all people are equally sensitive. There is a body of evidence suggesting that some people are hypersensitive to EMFs, though such a condition is not widely acknowledged. Formal research has concentrated on cancer, though the increased incidence is slight and if there is a risk it is small. Several forms of cancer have been cited, notably childhood leukaemia with residential exposure and male breast cancer in power workers. Apart from cancer, published studies have suggested associations with depression, suicide and Alzheimer's disease. Minor ill effects, such as headaches and sleeplessness, are much more commonly reported at the anecdotal level; in some cases they are sufficiently debilitating to cause prolonged absence from work. Moving home has removed the illness. Continual buzzing in the ears, interfering with sleep and normal life, has also been reported; there is an audible buzzing from powerlines which is pronounced in damp weather.
7. How does cause differ from association?
Two things are "associated" if they tend to occur together, for example ill health and unemployment. The tendency may be only partial, reflected in small increases in percentages of occurrence. This doesn't tell us that one thing causes the other, or vice versa. "Cause" is in any case an artificial idea, with a long philosophical and scientific background. In practice, we can never be sure of absolute cause, in which an effect will certainly follow every time, even though in some cases it is nearly so, like switching on the light. Causation of illness is often probabilistic (merely increasing the incidence of it, while many people still feel no effect) and multifactorial (with several alternative or combined causal factors). For example, many smokers do not get lung cancer, and many non-smokers do, with asbestos being among other causal factors. The effects can also be multiple, as smoking is associated with other forms of cancer and with other diseases. We can define cause as we please; it is not an absolute. For practical purposes for health effects there is no hard and fast agreed definition. There are the Bradford-Hill criteria for causation, which are points for guidance requiring expert and subjective interpretation. They mainly relate to two things: association and mechanism.
8. How convincing is the evidence of association of EMFs and cancer?
In its thorough review in 1992, the NRPB-AG described the evidence as "weak", though without any scientific definition of the term. After considering new Scandinavian studies, notably the 1993 landmark study by Feychting and Ahlbom in Sweden, the NRPB-AG recognised they were "well controlled" and provided "some evidence" of the possibility of a cause of childhood leukaemia, while noting the number of cases was small. The main evidence is statistically significant, at a high level of 99.5%. The incidence of leukaemia is increased 3.8 times for the population exposed at 0.3 microTesla, and about 2 times at 0.2 microTeslas, the latter reinforced over a few independent studies taken together. Some studies have failed to find any association, and others are not consistent in the types of associated cancer. New negative results from Finland heralded in the BMJ last month (October 1996) are practically worthless since they fail to identify high-exposure groups. The landmark Scandinavian studies are more reliable. Further major studies from other countries are due to report in the next few years. No alternative explanation of the associations has been shown to be more likely than cause by EMF exposure.
9. Is there a plausible biological mechanism linking EMFs and cancer?
Scientists would like to see a clear step by step mechanism, at the level of biochemical and cellular activity, verified by independently reproduced experimental evidence at each step, not violating established scientific theory, and confirmed experimentally as a whole-animal process, before accepting a causal link. Cancer is not generally that well understood. A "plausible" mechanism might allow minor gaps and incomplete verification, while not violating established science The NRPB-AG, taking a stricter view like the previous paragraph, concludes (1994) "there is no persuasive biological evidence that ELF EMFs can influence any of the accepted stages in carcinogenesis", but stresses the need for urgent research. The NRPB stresses the inability of ELF fields to damage DNA directly, which is accepted, and considers cancer promotion (accelerating an existing cancer) but not indirect initiation of cancer. Some biochemists suggest initiation through chain reactions involving free radicals. Research suggests relevant effects in cell signalling, cell regulation, immune suppression, melatonin synthesis and calcium uptake. NRPB-AG recognizes some positive findings but insufficient verification to be persuaded. In 1996 Henshaw demonstrated a new mechanism in which radon products in the air are attracted to EMF sources and deposited more readily on surfaces in the presence of EMFs. The products are accepted causes of cancer and can be inhaled and deposited on skin, so they provide another possibility of indirect initiation? from EMFs. NRPB hastily dismissed the findings on the grounds that greater surface deposits would deplete the presence of products in the air, but Henshaw's experiments contradict this.
10. Are there any established biological effects of EMFs?
Physical effects, such as a hand-held fluorescent tube lighting up under a powerline, are clear. Some 5% of people can feel effects of electric fields at levels found under high voltage powerlines (PL-EMFs). Spark discharges are perceived more easily and become annoying at field levels under powerlines. Magnetic fields are not normally perceived at these levels. Slight changes in heart rate and EEG recordings have been reported, along with effects on arousal, reasoning and circadian rhythms, from PL-EMFs. The brain is electrically active in the ELF range with beta rhythms close to power frequency. Like the cancer-related effects mentioned in Section 9, these suggested effects are not accepted by NRPB as yet established (1992). In a review paper drawing on 89 references Hendee and Boteler concluded in 1994 that PL-EMFs are "biologically active' but no mechanism completely explains the bioeffects. Ross Adey, chairman of the NCRP-SC, suggested in 1995 there was a "growing internationa1 consensus. that non-ionising EMFs can induce athermal bioeffects. Later the same year the NCRP-SC Draft Report concluded "findings are sufficiently consistent and form a sufficiently coherent picture to suggest plausible connections between ELF EMF exposures and disruption of normal biological processes, in ways meriting detailed examination of potential implications in human health".
11. Are there scientific conceptual barriers to recognising bioeffects?
An early sceptical argument was that typical magnetic fields were much lower than the ambient earth's field, and so would have no significant effect. That overlooks the important frequency effect. Power frequency fields can be a thousand times-stronger than background levels. Similarly internally induced electric fields in the body may be masked by thermally generated electrical noise, although the same is not true of induced current density and internal magnetic fields. The frequency focus of applied fields may still result in subtle effects. The inability of ELF fields to interact on a molecular scale rules out the direct hit approach to cancer initiation However biochemists have identified weak chemical bonding which can be affected, for example releasing hydroxyl radicals from a caged configuration resulting in powerful oxidative chain reactions. Present exposure limits are based on levels of large scale induced currents in the body. Currents induced by residential exposure are far too weak to have the gross effects envisaged. However those large scale effects are not relevant to concerns about cancer and other subtle effects. The purported barriers (comparison with background levels, molecular scale limitations of ELF and large scale strength limitations) are unfounded. No sound scientific reason has been advanced to suggest it would be impossible for EMFs to cause or promote cancer, let alone other bioeffects.
12. Now do EMFs compare with other hazards?
Smoking is accepted as a cause of lung cancer even though the complete process is not fully understood. The link with nicotine is very specific and the epidemiological evidence is very strong, with vast data showing a relative risk (RR) of about 20. By contrast the link of EMFs to childhood leukaemia depends on very few cases (drawn from wide reaching studies), is not supported by some earlier less thorough studies, has an RR of about 2 (rising to 4 or more in a few studies at higher exposures) and has patchy and not well verified ideas on mechanisms. Passive smoking is also accepted as a cause of lung cancer. The biological process is still not fully understood. The RR at 1.2 is below the level some epidemiologists accept as meaningful. EMFs have a stronger case except for the knowledge of nicotine, though that is offset by the lower exposure levels. The risk levels are also similar: about 10 per 100,000 non smokers get lung cancer per year, rising to about 12 among passive smokers; in the Swedish studies about 3 per 100,000 children get leukaemia per year, rising to about 11 among those closest to powerlines. BSE infected beef is not accepted as a cause of CJD, though after a new strain of CJD was recognised in 1996 following the BSE epidemic in UK herds since the late 1980s it is recognised as the most likely explanation. No transfer mechanism is known and previous evidence from other species suggested transfer to humans was unlikely. Epidemiological evidence is very weak, associating about 10 new cases with youth rather than beef.
13. What is the overall authoritative assessment?
The NRPB-AG is the UK's most authoritative body for relevant scientific assessment. As a government-appointed quango, like the bodies advising on BSE the NRPB does not attract the same level of public confidence as a fully independent body, but it is the best we've got. Its summative assessment reported in 1995 in the House of Lords was that "the evidence does not establish that exposure to EMFs is a cause of cancer although it does provide some evidence to suggest the possibility exists". A similar position is taken by the leading international bodies IRPA and ICNIRP and some other national bodies. On the other hand the Swedish NESB in 1993 reported a strong suspicion of a link with childhood leukaemia. The NCRP-SC, USA counterpart to the NRPB-AG, produced a draft report in 1995 which recognised " epidemiological findings that suggest significantly enhanced cancer risks" and biological findings which "suggest plausible connections ..." (see 10 above), while also judging that available evidence can not "yet establish well defined thresholds for safety guidelines". NCRP dismisses the (leaked) draft report since it faces formidable consultative hurdles before adoption. There is broad agreement that cause of cancer is not proven or established, that there is some reason for suspicion which cannot be dismissed, that the evidence does not support firm exposure limits, and that further research is needed. The differences are in the strength of that suspicion and what to do about lt.
Industry and professional engineering bodies are not best placed to give an authoritative independent assessment of health effects. The Electricity Association has relied upon the NRPB. So has the IKE although its own words "not only must an association be proven but the latter three factors must be established to prove conclusion" (referring wrongly to some of the Bradford-Hill factors) added to NRPB's conclusions go too far.
14. What are the policy options?
The first area of policy is public information. Research papers and NRPB detailed documents are publicly available but depend on interpreters to reach the public and politicians meaningfully. Public information leaflets from NRPB refer to its own guidelines and the 1600 microTesla level but not to the 0.2 microTesla level of suspicion and possible risk. Government statements quote NRPB and rely on the conclusion that cause of cancer is not established. In contrast Swedish government public information in 1994 admitted suspicion of risk and reason for caution. The second area is in precautionary policy. New planning proposals continue to come forward, for powerlines near houses and for new houses near or under powerlines. In 1993 the IEHO (now the CIEH) issued a Policy Guidance Statement urging local authorities to take a precautionary approach, but found this approach was not upheld in planning inquiries. After consulting NRPB the CIEH abandoned this policy in 1995, saying that the precautionary principle cannot be sustained on health grounds alone, and if adopted must be based on access, amenity and public acceptability. The NRPB has not explicitly recommended or rejected a precautionary approach. It says it does not have a basis for exposure limits, but that is different. Government statements, and the CIEH, have taken NRPB's silence as rejection of precaution. The precautionary principle is enshrined in EU and UK policy to limit harm when scientific knowledge is not conclusive. One kind of precaution, "prudent avoidance", envisages avoidance only when the cost is small while accepting the uncertain hazard otherwise. On a personal level people can make their own decisions if they are informed. In the case of Simon Studholme, who died of leukaemia after continued EMF exposure, simply moving his bed to the other side of the room would have avoided exposure, and the posterior probability is firmly odds-on that it would have avoided the leukaemia. On a planning level it might mean seeking a minimum distance between new lines and houses where feasible at least until pending research results appear in the next year or two. The draft report of the NCRP-SC goes further, proposing interim guidelines (pending further research) with firm restrictions at 0.2 microTesla for new developments and progressive implementation to reduce existing exposures. The Swedish NESB has recommended restrictions at this level since 1993 and in 1995 the five relevant Swedish government agencies jointly endorsed a policy of prudent avoidance. The EU Parliament in 1994 called for exposure limits on a precautionary basis and corridors imposed for powerlines. Precautionary limits on new developments do not imply costly removal of existing exposures, which industry and government may fear.
15. Conclusions
Public health fears about exposure to EMFs from electricity supply systems are not supported by scientifically established proof, yet such proof is very demanding. Public concern is expressed well before proof is reached, with good reason. There are rational grounds for suspicion of associated health risks of cancer and neurological disease as well as minor illness. The grounds are much stronger than those linking BSE-infected offal (even eating the diseased brains) with human CJD, statistically (if not biologically) stronger than the link of passive smoking with lung cancer, but much weaker than the evidence against active smoking. Hundreds of research papers have been published, with some positive findings in both epidemiological and experimental studies. Major studies are in progress in several countries. More and better focused evidence should become available over the next two or three years. In the interim, there is concern that:
(a) public information should assist people to judge for themselves whether to limit their own exposure, with reference to 0.2 microTesla;
(b) new developments should be subject to reasonable precautionary policy;
(c) local authorities should be entitled to respond to public concern by adopting prudent avoidance in their planning policies.
Sources include
NRPB, Amer J Epidem, Electricity Association, Microwave News, Powerwatch, Hansard and other Journals and documents. Detailed referencing will be available from the author.
Special Abbreviations
BMU British Medical Journal
CIEH Chartered Institute of Environmental Health
ELF Extremely Low Frequency
EMF Electro Magnetic Field
IKE Institute of Electrical Engineers
IRPA International Radiation Protection Association
ICNIRP International Commission on Non-Ionising Radiation Protection
NCRP National Council on Radiation Protection and Measurements (USA)
NCRP-SC NCRP Scientific Committee 89-3
NESB National Electrical Safety Board (Sweden)
NRPB National Radiological Protection Board (UK)
NRPB-AG NRPB Advisory Group on Non-Ionising Radiation
PL-EMF EMF under high voltage power line
RR Relative Risk
M J O' Carroll Garden House, Welbury, Northallerton DL6 2SE, UK
Home Index Economy & Environment Health Issues Press Releases
back to ElectricalConcerns Home Page