Correspondence
Cellular Telephones and Brain Tumors
N Engl J Med 2001; 344:1331-1332April 26, 2001
- Article
To the Editor:
The article by Inskip et al. (Jan. 11 issue)1 provides reassuring findings that cellular-telephone use is not associated with an increased risk of brain tumors, but the study has some limitations that are due to its retrospective design. Exposure was assessed by interviews, and recall bias, one of the main pitfalls in case–control studies, may be a problem because patients with brain tumors can have impaired memory.
Since telephone companies keep very accurate records of the calls of their customers, it would have been possible to document cellular-telephone use objectively if these data were accessible. Although legal problems might arise, most subjects would probably consent to the retrieval by investigators of data on the total duration of their calls, with no further details. The telephone companies should certainly be blinded to the health status of the customers whose data they provide. Because of the objectivity of billing records, such a design could provide more reliable data on the association between cellular-telephone use and brain tumors.
1 ReferencesMustafa Erman, M.D.
Ismail Celik, M.D.
Ayse Kars, M.D.
Hacettepe University, 06100 Ankara, Turkey1
Inskip PD ,Tarone RE ,Hatch EE , et al. Cellular-telephone use and brain tumors. N Engl J Med 2001;344:79-86
Full Text | Web of Science | Medline
To the Editor:
The study by Inskip et al. has a number of serious deficiencies. The hypothesis that was tested was that there is no association between the frequency of brain tumors (as determined by hospital admissions) and ever having used cellular telephones regularly. Consider the possibility that before the diagnosis of a brain tumor, a physician had advised each of these patients to purchase a mobile telephone for use in emergencies. If that had been the case, the study would have found an increased relative risk in users of cellular telephones, but the causality would have been reversed. I wonder whether the authors have considered the possibility that there might be confounding due to the use of a cellular telephone on the advice of a physician. (Among the controls, a considerable proportion of the patients had cardiovascular diseases — a circumstance in which the purchase of a cellular telephone is often recommended.)
It is indispensable in case–control studies to ensure that the condition under investigation precedes the disease outcome. The most important factor in analyzing the possible contribution of cellular-telephone use to brain tumors is latency. For some brain tumors, the interval between malignant transformation and clinical symptoms or diagnosis can exceed 10 or even 20 years.1-4 The study did not have sufficient power to detect a substantially increased risk if exposure to a cellular telephone is considered as contributing to malignant transformation. Such a contribution of exposure to tumor development was not even addressed, let alone tested, by the authors.
4 ReferencesMichael Kundi, Dr.Phil., Dr.Med.Habil.
University of Vienna, A-1095 Vienna, Austria1
Simmons NE ,Laws ER . Glioma occurrence after sellar irradiation: case report and review. Neurosurgery 1998;42:172-178
CrossRef | Web of Science | Medline2
Daentzer D ,Boker DK . Strahleninduziertes Meningeom 20 Jahre nach Operation und hochdosierter Radiatio eines Ependymoms. Zentralbl Neurochir 1999;60:27-32
Web of Science | Medline3
Loning L ,Zimmermann M ,Reiter A , et al. Secondary neoplasms subsequent to Berlin-Frankfurt-Munster therapy of acute lymphoblastic leukemia in childhood: significantly lower risk without cranial radiotherapy. Blood 2000;95:2770-2775
Web of Science | Medline4
Strojan P ,Popovic M ,Jereb B . Secondary intracranial meningiomas after high-dose cranial irradiation: report of five cases and review of the literature. Int J Radiat Oncol Biol Phys 2000;48:65-73
CrossRef | Web of Science | Medline
To the Editor:
The results of the study by Inskip et al. include a possible 60 percent increase in the incidence of glioma resulting from reported short-term use of cellular telephones. This increase may not rise to the level of statistical significance, but given the low power of the study, which is due, in part, to the truncated nature of the study period, the authors' reassurances are hard to accept. The broad confidence intervals underscore the deficiencies and uncertainties that undermine the authors' conclusions.
Robert C. Kane, M.S.
P.O. Box 133, Blanchardville, WI 53516Author/Editor Response
The authors reply:
To the Editor: Our study was conducted in response to the concern that brain cancers diagnosed in the early 1990s might have been caused by cellular telephones. The upper confidence limits in our study indicate that if there was such an effect, it was small. We found a relative risk of glioma of 0.9, with an upper 95 percent confidence limit of 1.6, associated with more than 100 hours of cellular-telephone use, and a relative risk of 0.5, with an upper limit of 1.3, associated with more than 500 hours of use. The absence of a dose–response relation for any measure of the level of use argues against an association.
We acknowledged the possibility that our study was conducted too early to detect an effect. Insofar as it is not known at what stage in carcinogenesis, if any, radio-frequency radiation might act, it is not clear what we should expect in terms of an induction period, nor that ionizing radiation should serve as the model, as implied by Kundi.
Kundi's suggestion that there might have been a high rate of cellular-telephone use among the controls with cardiovascular disease is not supported by our data. The relative-risk estimates were insensitive to the inclusion or exclusion of any subgroup of controls. The duration of symptoms before admission to the hospital was less than one year for more than 90 percent of the controls, and the exclusion of cellular-telephone use within the year preceding admission did not materially change our findings.
Erman et al. suggest that cellular-telephone use might have been underreported among patients with tumors because of mental impairment, and they advocate the use of billing records to assess cellular-telephone use. Mental impairment, rare in acoustic neuroma and less common in younger than in elderly patients with glioma or meningioma,1 is unlikely to have affected substantially the responses of the heaviest cellular-telephone users — namely, young and middle-aged patients. Billing records have been judged inadequate for the assessment of exposure in case–control studies.2,3 Cellular-telephone service providers typically maintain detailed billing records for a maximum of one year; those records often include only outgoing calls and do not identify the user of the telephone.2,3 Regardless of the limitations of billing records as compared with data from interviews, it is reassuring that a recent cohort study that used service-provider records4 had similar findings.
We see the timing of our study relative to the explosive growth in the use of cellular telephones as its most important limitation. A recently launched, multicenter study3 will have greater statistical power to assess risks associated with long induction periods and the use of digital telephones. However, given the extent of exposure to cellular telephones in the population, one would want to identify any excess risk at the earliest possible time.
4 ReferencesPeter D. Inskip, Sc.D.
Robert E. Tarone, Ph.D.
Martha S. Linet, M.D.
National Cancer Institute, Bethesda, MD 208921
Kleihues P, Cavenee WK. Pathology and genetics of tumours of the nervous system. Lyon, France: International Agency for Research on Cancer, 1997.
2
Dreyer NA ,Loughlin JE ,Rothman KJ . Epidemiological safety surveillance of cellular telephones in the US. Radiat Prot Dosim 1999;83:159-163
Web of Science3
Cardis E ,Kilkenny M . International case-control study of adult brain, head and neck tumours: results of the feasibility study. Radiat Prot Dosim 1999;83:179-183
Web of Science4
Johansen C ,Boice J Jr ,McLaughlin J ,Olsen J . Cellular telephones and cancer -- a nationwide cohort study in Denmark. J Natl Cancer Inst 2001;93:203-207
CrossRef | Web of Science | Medline
- Citing Articles (2)
Citing Articles
1
Soo Chan Kim, Ki Chang Nam, Deok Won Kim. (2006) Estimation of relative exposure levels for cellular phone users using a neural network. Bioelectromagnetics 27:6, 440-444
CrossRef2
Robert C. Owens, Gilles L. Fraser, Patricia Stogsdill. (2004) Antimicrobial Stewardship Programs as a Means to Optimize Antimicrobial Use. Pharmacotherapy 24:7, 896-908
CrossRef
