Mobile phones and the illusory pursuit of safety.

See pages 1833, 1837 Two papers in today’s Lancet address the controversial subject of safety of mobile telephones. Ken Rothman considers the issue from an epidemiological viewpoint, whereas G J Hyland seeks a possible mechanism for harmful effects. Readers will make up their own minds, both on their own views and on what they should say to anxious patients. However, it is worthwhile rehearsing a few salient points. The deceptively simple question, much loved by television and radio interviewers, “Is it safe?” is the scientist’s banana skin. A Nobel prize awaits the person who first designs an experiment to show that anything is “safe”. The best that can be achieved is an experiment or epidemiological analysis of data that demonstrates, within specified limits of statistical probability, that the risk associated with a drug, a new mode of radiation, or other insult to the body is no greater than a specified figure—for example, one in a million. In the light of experience with ionising radiation and radioactive materials, out-of-hand dismissal of the possibility of subtle effects of low-intensity, pulsed, microwave radiation is most unwise. Early in the 20th century radon and radium-enriched spa waters were “recommended” for a wide range of aches and minor ailments. As knowledge of the harmful effects of ionising radiation has increased and quantitative risk estimates have become possible (notwithstanding rather large error bands), the permitted annual dose limit has been progressively reduced from the 1930s to the present day. Rothman concludes that it is too soon to reach a verdict on the health risks from cellular telephones. Hyland suggests that there may be subtle, non-thermal effects, principally associated with a synergy or resonance between the frequencies generated by various features of the mobile telephone process and natural body frequencies. How should one weigh the various reported effects, whether harmful or not, that Hyland cites? Have the findings been reproducible? In many cases they have not. Hyland suggests that the highly non-linear nature of living systems militates against the realisation of the identical conditions for exact replication, but this argument confuses deterministic effects, which will always occur above a certain threshold of insult, and stochastic ones, which are governed by the laws of chance. Buying one ticket and winning the lottery is an excellent example of a stochastic effect. Biological systems are subject to the same laws, albeit with a higher level of noise. Are the key experiments closely linked to the proposed mechanism? The postulate that exceptionally low intensities of microwave radiation can have a disproportionately high effect at certain frequencies implies that the effect should disappear when the frequency is “off-resonance”. Such experiments may be very difficult to design in vivo and do not seem to have been done. Have anecdotal reports been followed up by other workers? By their nature such reports are very selective and, to gain credibility, must be corroborated by results from an experimental design in which potential bias has been eliminated. Some papers that did not corroborate earlier findings have been published, as Hyland notes. Would a journal of negative results have published even more? On the other hand, I am aware of two epidemiological studies that have become the subject of legal scrutiny, resulting in curtailment or publication delay. Such intervention inevitably weakens the evidence base on which balanced views should be made. Hyland is right to call for further research but there needs to be a consensus about the most likely interaction mechanisms around which the experiments should be designed. At present it is difficult to make a logical case for more stringent limits on exposure to pulsed microwave radiation. Robust limits are based on either a direct quantitative link between cause and effect (eg, thermal heating) or on a proven causative mechanism supported by dose-effect relations, as found with ionising radiation. Neither is available for these postulated very-low-intensity effects; the process of ionisation and free-radical formation that is known to be the basis of cancer induction with, say, X rays, does not occur with electromagnetic radiation in the microwave frequency range. Ultimately the public perception of safety will be heavily influenced by the perceived level of benefit from the activity in question. This level is clearly high in the case of mobile telephones and in many other domains where individuals exercise freedom of choice. The title of this commentary draws on an editorial in the Guardian newspaper on the announcement, back in 1977, that there was to be an inquiry into a cluster of explosions of domestic gas. That editorial said: “Whatever the findings of the three-man enquiry (under an independent chairman) gas will not become safe. Some oaf will always leave a tap on and then go down at the dead of night with a lighted taper”. This highly explosive substance is piped into millions of homes in the country. Is it safe? Of course not but the amenity value is such that people are prepared to live with the risk. Researchers into the pursuit of safety, of mobile telephones or other features of modern living, would be well advised to take this political element into consideration.