Peter D. Inskip, Sc.D., Robert E. Tarone, Ph.D., Elizabeth E. Hatch, Ph.D., Timothy C. Wilcosky, Ph.D., William R. Shapiro, M.D., Robert G. Selker, M.D., Howard A. Fine, M.D., Peter M. Black, M.D., Jay S. Loeffler, M.D., and Martha S. Linet, M.D.
Background Concern has arisen that the use of hand-held cellulartelephones might cause brain tumors. If such a risk does exist,the matter would be of considerable public health importance,given the rapid increase worldwide in the use of these devices.
Methods We examined the use of cellular telephones in a case–controlstudy of intracranial tumors of the nervous system conductedbetween 1994 and 1998. We enrolled 782 patients through hospitalsin Phoenix, Arizona; Boston; and Pittsburgh; 489 had histologicallyconfirmed glioma, 197 had meningioma, and 96 had acoustic neuroma.The 799 controls were patients admitted to the same hospitalsas the patients with brain tumors for a variety of nonmalignantconditions.
Results As compared with never, or very rarely, having useda cellular telephone, the relative risks associated with a cumulativeuse of a cellular telephone for more than 100 hours were 0.9for glioma (95 percent confidence interval, 0.5 to 1.6), 0.7for meningioma (95 percent confidence interval, 0.3 to 1.7),1.4 for acoustic neuroma (95 percent confidence interval, 0.6to 3.5), and 1.0 for all types of tumors combined (95 percentconfidence interval, 0.6 to 1.5). There was no evidence thatthe risks were higher among persons who used cellular telephonesfor 60 or more minutes per day or regularly for five or moreyears. Tumors did not occur disproportionately often on theside of head on which the telephone was typically used.
Conclusions These data do not support the hypothesis that therecent use of hand-held cellular telephones causes brain tumors,but they are not sufficient to evaluate the risks among long-term,heavy users and for potentially long induction periods.
Hand-held cellular telephones were introduced to the U.S. marketin 19841 but were not widely used until the mid-1990s. By early2000, the number of subscribers to cellular-telephone serviceshad grown to an estimated 92 million in the United States and500 million worldwide.2,3 Some concern has arisen about adversehealth effects, especially the possibility that the low-powermicrowave-frequency signal transmitted by the antennas on handsetsmight cause brain tumors or accelerate the growth of subclinicaltumors.4,5,6,7,8 It is generally agreed that the heating ofbrain tissue by cellular telephones is negligible, and thatany carcinogenic effect would have to be mediated through anonthermal mechanism, the nature and existence of which remaina matter of speculation.5,7,8,9,10,11,12,13 Direct genotoxiceffects are unlikely.7,14,15,16
Data concerning the risk of cancer associated with the exposureof humans to nonionizing radiation of the frequencies used bycellular telephones are limited,4,6,7,17,18,19 and review panelshave called for additional research.5,8,20,21,22 We report theresults of a case–control study that was initiated in1993, shortly after the possibility of a link between use ofcellular telephones and brain tumors received extensive mediaattention and elicited public concern.
Methods
Study Setting and Population
The study methods have been described in detail previously.23The study was conducted at hospitals in Boston (Brigham andWomen's Hospital); Phoenix, Arizona (St. Joseph's Hospital andMedical Center); and Pittsburgh (Western Pennsylvania Hospital).Each is a regional referral center for the diagnosis and treatmentof brain tumors. Institutional review boards at the NationalCancer Institute and all participating hospitals approved theprotocol, and written informed consent was obtained from eachsubject or from his or her proxy. Enrollment began in June of1994 and ended in August of 1998. The study was restricted toadults 18 years old or older who received care at one of theparticipating hospitals, resided within 50 miles of the hospital(or within Arizona, in the case of the Phoenix center), andcould understand English or Spanish.
Eligible patients with tumors were those in whom a first intracranialglioma or neuroepitheliomatous tumor (codes 9380 to 9473 and9490 to 9506 of the International Classification of Diseasesfor Oncology, 2nd edition [ICD-O-2]24), hereafter referred toas glioma, intracranial meningioma (ICD-O-2 codes 9530 to 9538),or acoustic neuroma (ICD-O-2 code 9560) had been diagnosed withinthe eight weeks preceding hospitalization at a participatinghospital. Microscopical confirmation was required for gliomasand meningiomas, and a confirmatory magnetic resonance imaging(MRI) or computed tomographic (CT) scan was required for acousticneuromas (96 percent of which were also confirmed microscopically).The tumor grade of gliomas was classified according to the guidelinesof Kleihues et al.25,26 The location of the tumor was determinedon the basis of the MRI, CT, and surgical reports. Of the potentiallyeligible patients with tumors, 92 percent agreed to participatein the study. Most (80 percent) were enrolled and interviewedwithin three weeks after the qualifying diagnosis.
The controls were patients who were admitted to the same hospitalsfor a variety of nonmalignant conditions and were frequency-matchedto the total group of patients with tumors according to hospital,age (in 10-year strata), sex, race or ethnic group, and proximityof their residence to the hospital (0 to 8 km, >8 to 24 km,>24 to 48 km, >48 to 80 km, or >80 km [0 to 5 mi, >5to 15 mi, >15 to 30 mi, >30 to 50 mi, or >50 mi]).Priority for approaching candidates for the control group wasdetermined by the relative numbers of patients with tumors andcontrols in the different strata. Of the eligible controls whowere contacted, 799 (86 percent) participated.
Data Collection
A research nurse administered a computer-assisted, personalinterview in the hospital. Proxy interviews were conducted ifthe subject was too ill or functionally impaired or had died.Proxy interviews were necessary for 16 percent of the patientswith glioma, 8 percent of the patients with meningioma, 3 percentof the patients with acoustic neuroma, and 3 percent of thecontrols. The usual proxy was the spouse. Interviews were audiotapedif the respondent consented, and nearly all respondents didso.
The interview questions concerned the use of hand-held cellulartelephones (with the antenna on the handset), car telephones,and transportable cellular telephones with separate batterypacks.4 Participants were asked about the calendar years ofthe first and last use; the duration of "regular" use (definedas at least two calls per week within this period); the usualfrequency of use (in minutes per day); and the hand usuallyused to hold the handset. The questions concerning the durationand frequency of use were asked only of persons who reportedusing a particular type of telephone more than five times intheir lives. A card was shown displaying photographs of differenttypes of cellular and cordless telephones, and the interviewer'sscript included a description of the distinguishing featuresof each type. The managers of the study listened to audiotapesof this portion of the interview for all study subjects andcorrected the data when any discrepancies were found.
Also covered in the interview were the subject's educationallevel, household income, type of health insurance coverage,religion, marital status, history of medical exposure to ionizingradiation, and handedness. The subjects' addresses were linkedwith 1990 U.S. Census data files to determine the median householdincome for the census block or tract in which the person residedat the time of hospital admission.
Statistical Analysis
Conditional logistic regression27,28 was used to estimate oddsratios, compute confidence intervals, and perform likelihood-ratiotests (two-sided tests at an level of 0.05). Odds ratios wereused as estimates of the relative risk. In addition to the matchingvariables, the analyses accounted for the date of the interview(a continuous variable), the type of respondent (the subject,a proxy, or both), educational level (less than high school,high school or general equivalency diploma, one to three yearsof college, or four years of college or more), annual householdincome (six categories, from less than $15,000 to $75,000 ormore), type of health coverage, marital status, religion, historyof radiotherapy to the head or neck, and handedness.
The association between the laterality of the tumor and theself-reported laterality of telephone use was examined amongthe patients with tumors. The relative risk associated withthe use of hand-held cellular telephones was estimated as (OR+1)÷ 2 (where OR denotes odds ratio), on the basis of anapproach described in the Appendix and elsewhere (*). Two-sidedP values for tests of independence were based on Fisher's exacttest.
Results
Characteristics of Patients with Tumors and Controls
The 782 patients with tumors included 489 with glioma, 197 withmeningioma, and 96 with acoustic neuroma (Table 1). Nearly 71percent of the patients with gliomas had been given a diagnosisof glioblastoma or other type of astrocytoma, and 17 percenta diagnosis of oligodendroglioma or mixed glioma (mostly oligoastrocytomas).Patients with high-grade gliomas outnumbered those with low-gradegliomas (354 vs. 135). The ratios of men to women were 1.3 amongpatients with glioma, 0.3 among patients with meningioma, and0.6 among those with acoustic neuroma. The most common reasonsfor hospitalization among the 799 controls were injuries (197)and disorders of the circulatory (179), musculoskeletal (172),digestive (92), and nervous (58) systems. Diagnoses in the controlsvaried among the hospitals, with trauma accounting for the largestfraction in Phoenix (35 percent), circulatory diseases in Boston(29 percent), and musculoskeletal diseases in Pittsburgh (41percent). The patients with tumors tended to be older than thecontrols, more highly educated, and from homes with higher householdincomes (Table 1); these differences were most pronounced inthe patients with acoustic neuromas.
Table 1. Selected Characteristics of Patients with Brain Tumors and Control Patients.
Among the controls, 232 (29 percent) reported having used ahand-held cellular telephone more than five times; the proportionwho reported use at that level was 29 percent among patientshospitalized with trauma, 22 percent among those with circulatorydisease, 31 percent among those with musculoskeletal disorders,and 33 percent among other patients. The proportion of controlswho reported having used a hand-held cellular telephone morethan five times increased during the course of the study, whereasthe corresponding proportions for the use of car phones andtransportable cellular telephones did not (Figure 1A). Use ofhand-held cellular telephones was higher among men than amongwomen, decreased with age for both sexes (but more so for women),and was positively associated with self-reported household incomeand educational level (Figure 1B, 1C, and 1D). The lower prevalenceof use among patients with circulatory disease reflects theirolder age (mean, 57 years, as compared with 47 years for othercontrols).
Figure 1. Percentage of Controls Who Had Used a Cellular Telephone More Than Five Times, According to the Year of the Interview and Type of Cellular Telephone (Panel A), Age at the Time of the Interview and Sex (Panel B), Self-Reported Annual Household Income (Panel C), and Level of Education (Panel D).
Panels B, C, and D show the data for hand-held cellular telephones only.
Risk According to Level of Use
Neither ever having used a hand-held cellular telephone norregular use of one was significantly associated with the relativerisk of glioma, meningioma, or acoustic neuroma (Table 2). Amongregular users, relative risks were not higher among those whobegan using cellular telephones in earlier years, nor did therisk of any type of tumor increase significantly with increasingduration of use, frequency of use, or total cumulative use (Table 2).
Table 2. Relative Risks of Different Types of Intracranial Tumor of the Nervous System, According to the Level of Use of Hand-Held Cellular Telephones.
Seventeen patients with tumors and 28 controls reported havingused hand-held cellular telephones for an average of 15 or moreminutes per day for at least three years. The distribution ofthese patients according to histologic type of tumor was asfollows: four had anaplastic astrocytoma, three had glioblastoma,one had astrocytoma, one had oligodendroglioma, one had ependymoma,two had an unspecified type of glioma, four had meningioma,and one had acoustic neuroma. The relative risk for glioma atthis level of cellular-telephone use was 0.7 (95 percent confidenceinterval, 0.3 to 1.6).
Significantly elevated relative risks were not observed forany of the hospitals, for either sex, or for any age groups,and there was little heterogeneity in risk among subgroups ofthe controls (data not shown). For glioma, the relative riskassociated with using a cellular telephone more than five timesas compared with never using one was 1.2 when the control groupused in the analysis was limited to persons admitted becauseof injuries; the relative risk was 1.0 when the subgroup wasthose admitted for circulatory diseases; 1.0 when the subgroupwas those admitted for musculoskeletal diseases; and 0.9 whenthe subgroup was those admitted for other illnesses or conditions.
The relative risks were not significantly increased for subcategoriesof glioma defined according to histologic type, grade, laterality,or affected lobe of the brain (Table 3). Estimated relativerisks were 1.1 for all astrocytic tumors combined and 0.9 forhigh-grade astrocytoma. Within the latter category, the relativerisk was nonsignificantly greater than 1.0 for anaplastic astrocytomaand nonsignificantly less than 1.0 for glioblastoma. The relativerisk of anaplastic astrocytoma was greater for persons who hadused cellular telephones one to five times in their lives (relativerisk, 2.6) than for persons who had used them more than fivetimes (relative risk, 1.8).
Table 3. Relative Risks of Glioma and Neuroepitheliomatous Tumors Associated with the Use of Hand-Held Cellular Telephones More Than Five Times, as Compared with No Use.
Additional covariates (marital status, religion, type of healthinsurance, history of radiotherapy to the head or neck, andhandedness) had little or no influence on the estimates of relativerisk, nor did age when age categories smaller than 10 yearswere used. Patients with glioma were interviewed, on average,four months earlier than controls, but the analysis was adjustedfor the date of the interview, so the trend over time in theuse of cellular telephones was taken into account.
Laterality of Tumor and of Telephone Use
Among patients with tumors who reported having used hand-heldcellular telephones regularly for at least six months beforediagnosis, the laterality of the tumor was not significantlyassociated with the self-reported laterality of use of the cellulartelephone (Table 4). Relative risks of gliomas involving thefrontal, temporal, and parietal lobes were 1.0 (P=1.00), 0.9(P=0.66), and 0.8 (P=0.55), respectively, and the relative riskof anaplastic astrocytoma was 1.2 (P=1.00).
Table 4. Laterality of Tumor with Respect to Laterality of Telephone Use among Patients with Brain Tumors with Regular Use of a Hand-Held Cellular Telephone for at Least Six Months before Diagnosis.
Discussion
Our results do not substantiate the concern that some braintumors diagnosed in the United States during the mid-1990s werecaused by the use of hand-held cellular telephones. There waslittle or no indication of an increased risk of glioma, meningioma,or acoustic neuroma associated with any use, cumulative use,or the laterality of use of these telephones. There was no significantlyincreased risk associated with the use of cellular telephonesat any of the three centers, and estimates of the relative riskassociated with the use of cellular telephones were insensitiveto the inclusion or exclusion of any of the four principal subgroupsof controls.
Some degree of misclassification of patients' relative levelsof exposure was inevitable, because of inherent limitationsin the ability of the interview to capture historical changesin a patient's pattern of cellular telephone use, inaccuraciesin patients' recall, and variations in the levels of exposureto microwave radiation with different types of hand-held cellulartelephones and under different circumstances of use. In a previousstudy,29 a good correlation (Spearman's r=0.79) was reportedbetween the self-reported use of cellular telephones and billingrecords from cellular-telephone companies. Patients with gliomassometimes have impairments of memory or cognition that mightcompromise their ability to report past events and habits asaccurately as healthy persons. Such impairment is considerablyless common among young and middle-aged patients with gliomasthan among elderly patients and is uncommon among patients withmeningiomas or acoustic neuromas.26 Cellular-telephone use wasmuch more common among persons less than 70 years old than amongthose 70 years old or older (Figure 1B). If cellular telephonescause brain tumors, and the induction period is not long, onewould expect the effect to be evident among case patients youngerthan the ages at which aphasia is a common symptom. It was not.
In addition to the frequency and duration of telephone use,factors that can affect the level of exposure to microwave radiationinclude the distance from the base station, the local topographyand vegetation, whether the phone is used indoors or outdoors,the design of the particular model of telephone, and the positionof the antenna and the telephone in relation to the head.8,20,30,31Failure to account for these variables could result in the misclassificationof the level of exposure, but these factors are unlikely tovary systematically with the frequency or duration of use. Themisclassification of the level of use is more likely than themisclassification of use itself, and the low overall risk amongregular users suggests that if the study failed to detect asubstantially elevated risk, it was confined to a small subgroup.
Microwave radiation is attenuated rapidly with passage thoughtissue, so that the absorption of energy by tissue 5 cm belowthe surface of the skin is less than 10 percent of that at thesurface, and absorption is an order of magnitude less on theside of the head opposite that on which the telephone is usedthan it is on the side of use.1,4,20,31,32,33,34 However, wefound no positive associations between the laterality of thetumor and the side of cellular-telephone use for any of theprincipal categories of tumors. It is difficult to assess correlationsbetween the location of the tumor and the deposition of microwaveenergy at a finer gradation than laterality, insofar as thepattern of energy absorption is sensitive even to small changesin the position and design of the telephone32,35 and the preciseorigin of the tumor is unknown. Parts of the temporal, parietal,and occipital lobes have been described as regions of relativelyhigh exposure, whereas the frontal lobe is generally consideredto be much less exposed.1,4,6,34 We found no evidence that cellular-telephoneuse was more strongly associated with gliomas in the temporalor parietal lobe than it was with gliomas in the frontal lobe.
The most important limitation of our study is its limited precisionfor assessing the risks after a potential induction period ofmore than several years or among people with very high levelsof daily or cumulative use. For example, we cannot rule outa 60 percent increase in the risk of glioma associated withcumulative use of 100 hours or more. If an effect of cellulartelephones was limited to a relatively uncommon type of tumor,or to a very small volume of highly exposed tissue immediatelyadjacent to the handset, a much larger sample would be requiredto detect it. The study was designed to have adequate powerto assess the relative risk of all gliomas combined, but notfor subtypes of glioma.
There have been substantial changes in wireless communicationtechnology since this study began in 1993. The current trendis toward greater use of digital technology and higher frequenciesof transmission.8,20,36,37,38 Because of the timing of thisstudy, we presume that our results pertain primarily to analoguetelephones with frequencies of 800 to 900 MHz. Digital telephonesoperate at a lower average power than analogue telephones,4,8and we would not expect them to carry a higher risk, unlessthere is an important aspect of exposure other than the rateof energy deposition. No increase in the risk of brain tumorsassociated with the use of analogue or digital cellular telephoneswas found in a small, case–control study conducted inSweden.6 Multicenter, international studies that are just gettingunder way39,40 will have greater statistical power to assessthe risks of cancer associated with long induction periods andwill provide information about the risks associated with theuse of digital telephones.
Our results do not support the view that exposure to low-powermicrowave radiation from hand-held, analogue cellular telephonescauses malignant or benign tumors of the brain or nervous system.However, given the fact that widespread use of cellular telephonesis a recent development, the dramatic increases that have occurredin the frequency of use, and the changes in cellular technologyover time, the findings should be seen as an estimate of therisk at an early stage of the use of this technology.
We are indebted to our external advisory panel for their thoughtfulcomments and advice concerning the manuscript; to Emily Khoury,Brian Paul, Patsy Thompson, Donna Houpt, Kelli Williamson, SandraMcGuire, Renee Karlsen, Patricia Yost, Janice Whelan, DouglasWatson, Diane Fuchs, Bob Saal, Christel McCarty, George Geise,and Rebecca Albert for their valuable and dedicated assistanceduring the conduct of the study; and to the University of NorthCarolina Hospitals for providing the opportunity to preteststudy instruments and procedures on a sample of neurosurgicalpatients.
* See the Appendix or NAPS document no. 05579 for 2 pages of supplementarymaterial. To order the NAPS document, contact NAPS, c/o MicrofichePublications, 248 Hempstead Tpke., West Hempstead, NY 11552.
Source Information
From the Epidemiology and Biostatistics Program, Division of Cancer Epidemiology and Genetics (P.D.I., R.E.T., E.E.H., M.S.L.), and the Neuro-Oncology Branch (H.A.F.), National Cancer Institute, Bethesda, Md.; the Epidemiology and Medical Studies Program, Research Triangle Institute, Research Triangle Park, N.C. (T.C.W.); the Department of Neurology, Barrow Neurological Institute, St. Joseph's Hospital and Medical Center, Phoenix, Ariz. (W.R.S.); the Division of Neurosurgery, Western Pennsylvania Hospital, Pittsburgh (R.G.S.); Brigham and Women's Hospital, Boston (P.M.B.); and the Department of Radiation Oncology, Massachusetts General Hospital, Boston (J.S.L.).
Address reprint requests to Dr. Inskip at Executive Plaza S., Rm. 7052, 6120 Executive Blvd., Rockville, MD 20852, or at inskippe{at}mail.nih.gov.
References
Balzano Q, Garay O, Manning TJ Jr. Electromagnetic energy exposure of simulated users of portable cellular telephones. IEEE Trans Veh Technol 1995;44:390-403.
How many people use wireless phones? Washington, D.C.: Cellular Telecommunication Industry Association, 2000. (See http://www.wow-com. com/consumer/faqs/faq_general.cfm#one.) (See NAPS document no. 05580 for 3 pages, c/o Microfiche Publications, 248 Hempstead Tpke., West Hempstead, NY 11552.)
Maier M, Blakemore C, Koivisto M. The health hazards of mobile phones. BMJ 2000;320:1288-1289. [Free Full Text]
Rothman KJ, Chou CK, Morgan R, et al. Assessment of cellular telephone and other radio frequency exposure for epidemiologic research.Epidemiology 1996;7:291-8.
McKinlay A. Possible health effects related to the use of radiotelephones: recommendations of a European Commission Expert Group. Radiol Prot Bull 1997;187:9-16.
Hardell L, Näsman Å, Påhlson A, Hallquist A, Hansson Mild K. Use of cellular telephones and the risk for brain tumours: a case-control study. Int J Oncol 1999;15:113-116. [Web of Science][Medline]
Moulder JE, Erdreich LS, Malyapa RS, Merritt J, Pickard WF, Vijayalaxmi. Cell phones and cancer: what is the evidence for a connection? Radiat Res 1999;151:513-531. [Web of Science][Medline]
Independent Expert Group on Mobile Phones. Mobile phones and health. Chilton, England: National Radiological Protection Board, 2000.
Hermann DM, Hossmann K-A. Neurological effects of microwave exposure related to mobile communication. J Neurol Sci 1997;152:1-14. [Medline]
Anderson V, Joyner KH. Specific absorption rate levels measured in a phantom head exposed to radio frequency transmissions from analog hand-held mobile phones. Bioelectromagnetics 1995;16:60-69. [Medline]
Valberg PA. Radio frequency radiation (RFR): the nature of exposure and carcinogenic potential. Cancer Causes Control 1997;8:323-332. [CrossRef][Web of Science][Medline]
Van Leeuwen GMJ, Lagendijk JJW, Van Leersum BJAM, Zwamborn APM, Hornsleth SN, Kotte ANTJ. Calculation of change in brain temperatures due to exposure to a mobile phone. Phys Med Biol 1999;44:2367-2379. [CrossRef][Web of Science][Medline]
Adey WR. Bioeffects of mobile communications fields: possible mechanisms for cumulative dose. In: Kuster N, Balzano Q, Lin JC, eds. Mobile communications safety. New York: Chapman & Hall, 1997:95-131.
Meltz ML, Walker KA, Erwin DN. Radiofrequency (microwave) radiation exposure of mammalian cells during UV-induced DNA repair synthesis. Radiat Res 1987;110:255-266. [Medline]
Brusick D, Albertini R, McRee D, et al. Genotoxicity of radiofrequency radiation. Environ Mol Mutagen 1998;32:1-16. [CrossRef][Web of Science][Medline]
Malyapa RS, Ahern EW, Straube WL, Moros EG, Pickard WF, Roti Roti JL. Measurement of DNA damage after exposure to electromagnetic radiation in the cellular phone communication frequency band (835.62 and 847.74 MHz). Radiat Res 1997;148:618-627. [Web of Science][Medline]
Dreyer NA, Loughlin JE, Rothman KJ. Cause-specific mortality in cellular telephone users. JAMA 1999;282:1814-1816. [Free Full Text]
Elwood J. A critical review of epidemiologic studies of radiofrequency exposure and human cancers. Environ Health Perspect 1999;107:Suppl 1:155-168.
Morgan RW, Kelsh MA, Zhao K, Exuzides KA, Heringer S, Negrete W. Radiofrequency exposure and mortality from cancer of the brain and lymphatic/hematopoietic systems. Epidemiology 2000;11:118-127. [CrossRef][Web of Science][Medline]
International Commission on Non-Ionizing Radiation Protection. Health issues related to the use of hand-held radiotelephones and base transmitters. Health Phys 1996;70:587-593. [Web of Science][Medline]
Repacholi MH. Low-level exposure to radiofrequency electromagnetic fields: health effects and research needs. Bioelectromagnetics 1998;19:1-19. [CrossRef][Web of Science][Medline]
Electromagnetic fields (300 Hz to 300 GHz). Geneva: World Health Organization, 1993.
Inskip PD, Hatch EE, Stewart PA, et al. Study design for a case-control investigation of cellular telephones and other risk factors for brain tumors in adults. Radiat Prot Dosim 1999;86:45-52.
Percy C, Van Holten V, Muir C, eds. International classification of diseases for oncology, 2nd ed.: ICD-O. Geneva: World Health Organization, 1990.
Kleihues P, Burger PC, Scheithauer BW. The new WHO classification of brain tumours. Brain Pathol 1993;3:255-268. [Web of Science][Medline]
Kleihues P, Cavenee WK. Pathology and genetics of tumours of the nervous system. Lyon, France: International Agency for Research on Cancer, 1997.
Lubin JH. A computer program for the analysis of matched case-control studies. Comput Biomed Res 1981;14:138-143. [CrossRef][Web of Science][Medline]
Funch DP, Rothman KJ, Loughlin JE, Dreyer NA. Utility of telephone company records for epidemiologic studies of cellular telephones. Epidemiology 1996;7:299-302. [Web of Science][Medline]
Cleveland RF Jr, Athey TW. Specific absorption rate (SAR) in models of the human head exposed to hand-held UHF portable radios. Bioelectromagnetics 1989;10:173-186. [Medline]
Gandhi O. Some numerical methods for dosimetry: extremely low frequencies to microwave frequencies. Radio Sci 1995;30:161-77.
Dimbylow PJ, Mann SM. sar calculations in an anatomically realistic model of the head for mobile communication transceivers at 900 MHz and 1.8 GHz. Phys Med Biol 1994;39:1537-1553. [CrossRef][Web of Science][Medline]
Dimbylow P, Mann SM. Characterization of energy deposition in the head from cellular phones. Radiat Prot Dosim 1999;83:139-41.
Kuster N, Balzano Q. Experimental and numerical dosimetry. In: Kuster N, Balzano Q, Lin JC, eds. Mobile communications safety. New York: Chapman & Hall, 1997:13-64.
Dreyer NA, Loughlin JE, Rothman KJ. Epidemiological safety surveillance of cellular telephones in the US. Radiat Prot Dosim 1999;83:159-63.
Ramsdale PA, Wiener A. Cellular phone base stations: technology and exposures. Radiat Prot Dosim 1999;83:125-30.
Millington RJ. Mobile and personal communications in the 90s. In: Kuster N, Balzano Q, Lin JC, eds. Mobile communications safety. New York: Chapman & Hall, 1997:3-9.
Of mobile phones and morbidity. Environ Health Perspect 1998;106:A474-A475. [Web of Science][Medline]
Cardis E, Kilkenny M. International case-control study of adult brain, head and neck tumours: results of the feasibility study. Radiat Prot Dosim 1999;83:179-83.
Appendix
In the following analysis, B denotes brain tumor, P cellular-telephoneuse, R right side, L left side, and Pr probability. The usualestimate of the odds ratio from a two-by-two table for lateralitylike those in Table 4 is an estimate of the quantity
[Pr(BR|PR)Pr(BL|PL)]÷[Pr(BL|PR)Pr(BR|PL)],
where, for example, Pr(BR|PR) is the probability of the developmentof a brain tumor on the right side of the brain when the cellulartelephone was used on the right side. If cellular-telephoneuse is independent of the risk of brain tumors, then
Pr(BR|PR)=Pr(BR|PL)=Pr(BR) and Pr(BL|PL)=Pr(BL|PR)=Pr(BL),
so that the odds ratio is always 1.0, even though there maybe a higher probability that a tumor will develop on one sideof the brain than the other.
Suppose that cellular-telephone use is associated with braintumors and that the relative risk that a tumor will developon the side of cellular-telephone use is the same for tumorson the left side and on the right side. That is, assume that
Pr(BL|PL)= Pr(BL|PR) and Pr(BR|PR)= Pr(BR|PL).
It follows that the usual odds ratio estimated from a two-by-twotable for laterality is an estimate of 2, and thus that thesquare root of the odds ratio is an estimate of . If cellular-telephoneuse does not expose the side of the brain opposite the sideof use to any radiation, then the laterality risk ratio, , quantifiesthe risk of brain tumors due to cellular-telephone use.
The relative risk for the general population is defined as
RR=P(B|P)÷P(B|no P),
where brain tumors and cellular-telephone exposure can occuron either side of the brain. Assume that brain tumors are equallylikely on the left and the right side in the absence of cellular-telephoneexposure. Then it follows that
Myung, S.-K., Ju, W., McDonnell, D. D., Lee, Y. J., Kazinets, G., Cheng, C.-T., Moskowitz, J. M.
(2009). Mobile Phone Use and Risk of Tumors: A Meta-Analysis. JCO
27: 5565-5572
[Abstract][Full Text]
Vrijheid, M, Mann, S, Vecchia, P, Wiart, J, Taki, M, Ardoino, L, Armstrong, B K, Auvinen, A, Bedard, D, Berg-Beckhoff, G, Brown, J, Chetrit, A, Collatz-Christensen, H, Combalot, E, Cook, A, Deltour, I, Feychting, M, Giles, G G, Hepworth, S J, Hours, M, Iavarone, I, Johansen, C, Krewski, D, Kurttio, P, Lagorio, S, Lonn, S, McBride, M, Montestrucq, L, Parslow, R C, Sadetzki, S, Schuz, J, Tynes, T, Woodward, A, Cardis, E
(2009). Determinants of mobile phone output power in a multinational study: implications for exposure assessment. Occup. Environ. Med.
66: 664-671
[Abstract][Full Text]
Bhatti, P., Stewart, P. A., Hutchinson, A., Rothman, N., Linet, M. S., Inskip, P. D., Rajaraman, P.
(2009). Lead Exposure, Polymorphisms in Genes Related to Oxidative Stress, and Risk of Adult Brain Tumors. Cancer Epidemiol. Biomarkers Prev.
18: 1841-1848
[Abstract][Full Text]
Rajaraman, P., Brenner, A. V., Butler, M. A., Wang, S. S., Pfeiffer, R. M., Ruder, A. M., Linet, M. S., Yeager, M., Wang, Z., Orr, N., Fine, H. A., Kwon, D., Thomas, G., Rothman, N., Inskip, P. D., Chanock, S. J.
(2009). Common Variation in Genes Related to Innate Immunity and Risk of Adult Glioma. Cancer Epidemiol. Biomarkers Prev.
18: 1651-1658
[Abstract][Full Text]
McKean-Cowdin, R., Barnholtz-Sloan, J., Inskip, P. D., Ruder, A. M., Butler, M., Rajaraman, P., Razavi, P., Patoka, J., Wiencke, J. K., Bondy, M. L., Wrensch, M.
(2009). Associations between Polymorphisms in DNA Repair Genes and Glioblastoma. Cancer Epidemiol. Biomarkers Prev.
18: 1118-1126
[Abstract][Full Text]
Olsen, J.
(2009). Mobiles phones and health. Scand J Public Health
37: 1-3
Kheifets, L., Oksuzyan, S.
(2008). Exposure assessment and other challenges in non-ionizing radiation studies of childhood leukaemia. Radiat Prot Dosimetry
132: 139-147
[Abstract][Full Text]
Lahkola, A, Salminen, T, Raitanen, J, Heinavaara, S, Schoemaker, M., Christensen, H C., Feychting, M, Johansen, C, Klaeboe, L, Lonn, S, Swerdlow, A., Tynes, T, Auvinen, A
(2008). Meningioma and mobile phone use--a collaborative case-control study in five North European countries. Int J Epidemiol
37: 1304-1313
[Abstract][Full Text]
Wen, P. Y., Kesari, S.
(2008). Malignant Gliomas in Adults. NEJM
359: 492-507
[Full Text]
Kundi, M
(2008). What can be learned from the Japanese study of mobile phone use and acoustic neuroma?. Occup. Environ. Med.
65: 428-428
[Full Text]
Samanic, C. M., De Roos, A. J., Stewart, P. A., Rajaraman, P., Waters, M. A., Inskip, P. D.
(2008). Occupational Exposure to Pesticides and Risk of Adult Brain Tumors. Am J Epidemiol
167: 976-985
[Abstract][Full Text]
Sadetzki, S., Chetrit, A., Jarus-Hakak, A., Cardis, E., Deutch, Y., Duvdevani, S., Zultan, A., Novikov, I., Freedman, L., Wolf, M.
(2008). Cellular Phone Use and Risk of Benign and Malignant Parotid Gland Tumors--A Nationwide Case-Control Study. Am J Epidemiol
167: 457-467
[Abstract][Full Text]
Rajaraman, P., Hutchinson, A., Rothman, N., Black, P. M., Fine, H. A., Loeffler, J. S., Selker, R. G., Shapiro, W. R., Linet, M. S., Inskip, P. D.
(2008). Oxidative response gene polymorphisms and risk of adult brain tumors. Neuro Oncol Duke
10: 709-715
[Abstract][Full Text]
Krex, D., Klink, B., Hartmann, C., von Deimling, A., Pietsch, T., Simon, M., Sabel, M., Steinbach, J. P., Heese, O., Reifenberger, G., Weller, M., Schackert, G., for the German Glioma Network,
(2007). Long-term survival with glioblastoma multiforme. Brain
130: 2596-2606
[Abstract][Full Text]
Hardell, L., Carlberg, M., Soderqvist, F., Mild, K. H., Morgan, L. L.
(2007). Long-term use of cellular phones and brain tumours: increased risk associated with use for >=10 years. Occup. Environ. Med.
64: 626-632
[Abstract][Full Text]
Rajaraman, P., Wang, S. S., Rothman, N., Brown, M. M., Black, P. M., Fine, H. A., Loeffler, J. S., Selker, R. G., Shapiro, W. R., Chanock, S. J., Inskip, P. D.
(2007). Polymorphisms in Apoptosis and Cell Cycle Control Genes and Risk of Brain Tumors in Adults. Cancer Epidemiol. Biomarkers Prev.
16: 1655-1661
[Abstract][Full Text]
Bluhm, E. C., Zahm, S. H., Fine, H. A., Black, P. M., Loeffler, J. S., Shapiro, W. R., Selker, R. G., Inskip, P. D.
(2007). Personal Hair Dye Use and Risks of Glioma, Meningioma, and Acoustic Neuroma among Adults. Am J Epidemiol
165: 63-71
[Abstract][Full Text]
Schuz, J., Jacobsen, R., Olsen, J. H., Boice, J. D. Jr, McLaughlin, J. K., Johansen, C.
(2006). Cellular Telephone Use and Cancer Risk: Update of a Nationwide Danish Cohort. JNCI J Natl Cancer Inst
98: 1707-1713
[Abstract][Full Text]
Rajaraman, P., Stewart, P. A., Samet, J. M., Schwartz, B. S., Linet, M. S., Zahm, S. H., Rothman, N., Yeager, M., Fine, H. A., Black, P. M., Loeffler, J., Shapiro, W. R., Selker, R. G., Inskip, P. D.
(2006). Lead, Genetic Susceptibility, and Risk of Adult Brain Tumors. Cancer Epidemiol. Biomarkers Prev.
15: 2514-2520
[Abstract][Full Text]
Takebayashi, T, Akiba, S, Kikuchi, Y, Taki, M, Wake, K, Watanabe, S, Yamaguchi, N
(2006). Mobile phone use and acoustic neuroma risk in Japan. Occup. Environ. Med.
63: 802-807
[Abstract][Full Text]
Hepworth, S. J, Schoemaker, M. J, Muir, K. R, Swerdlow, A. J, van Tongeren, M. J A, McKinney, P. A
(2006). Mobile phone use and risk of glioma in adults: case-control study. BMJ
332: 883-887
[Abstract][Full Text]
De Roos, A. J., Rothman, N., Brown, M., Bell, D. A., Pittman, G. S., Shapiro, W. R., Selker, R. G., Fine, H. A., Black, P. M., Inskip, P. D.
(2006). Variation in genes relevant to aromatic hydrocarbon metabolism and the risk of adult brain tumors.. Neuro Oncol Duke
8: 145-155
[Abstract][Full Text]
Wood, A W
(2006). How dangerous are mobile phones, transmission masts, and electricity pylons?. Arch. Dis. Child.
91: 361-366
[Abstract][Full Text]
Schuz, J., Bohler, E., Berg, G., Schlehofer, B., Hettinger, I., Schlaefer, K., Wahrendorf, J., Kunna-Grass, K., Blettner, M.
(2006). Cellular Phones, Cordless Phones, and the Risks of Glioma and Meningioma (Interphone Study Group, Germany). Am J Epidemiol
163: 512-520
[Abstract][Full Text]
Propp, J. M., McCarthy, B. J., Davis, F. G., Preston-Martin, S.
(2006). Descriptive epidemiology of vestibular schwannomas. Neuro Oncol Duke
8: 1-11
[Abstract][Full Text]
Hoffman, S., Propp, J. M., McCarthy, B. J.
(2006). Temporal trends in incidence of primary brain tumors in the United States, 1985-1999. Neuro Oncol Duke
8: 27-37
[Abstract][Full Text]
Christensen, H. C., Schuz, J., Kosteljanetz, M., Poulsen, H. S., Boice, J. D. Jr, McLaughlin, J. K., Johansen, C.
(2005). Cellular telephones and risk for brain tumors: A population-based, incident case-control study. Neurology
64: 1189-1195
[Abstract][Full Text]
Lonn, S., Ahlbom, A., Hall, P., Feychting, M., the Swedish Interphone Study Group,
(2005). Long-Term Mobile Phone Use and Brain Tumor Risk. Am J Epidemiol
161: 526-535
[Abstract][Full Text]
Kleinerman, R. A., Linet, M. S., Hatch, E. E., Tarone, R. E., Black, P. M., Selker, R. G., Shapiro, W. R., Fine, H. A., Inskip, P. D.
(2005). Self-reported Electrical Appliance Use and Risk of Adult Brain Tumors. Am J Epidemiol
161: 136-146
[Abstract][Full Text]
Brenner, A. V., Linet, M. S., Shapiro, W. R., Selker, R. G., Fine, H. A., Black, P. M., Inskip, P. D.
(2004). Season of birth and risk of brain tumors in adults. Neurology
63: 276-281
[Abstract][Full Text]
Pueschel, J. K.
(2004). Integrative Tumor Board: Glioblastoma Multiforme: Neurooncology. Integr Cancer Ther
3: 151-152
Kundi, M
(2004). Mobile phone use and cancer. Occup. Environ. Med.
61: 560-570
[Full Text]
Christensen, H. C., Schuz, J., Kosteljanetz, M., Poulsen, H. S., Thomsen, J., Johansen, C.
(2004). Cellular Telephone Use and Risk of Acoustic Neuroma. Am J Epidemiol
159: 277-283
[Abstract][Full Text]
Ahlbom, A., Feychting, M.
(2003). Electromagnetic radiation: Environmental pollution and health. Br Med Bull
68: 157-165
[Abstract][Full Text]
Brenner, A. V., Linet, M. S., Selker, R. G., Shapiro, W. R., Black, P. M., Fine, H. A., Inskip, P. D.
(2003). Polio Vaccination and Risk of Brain Tumors in Adults: No Apparent Association. Cancer Epidemiol. Biomarkers Prev.
12: 177-178
[Full Text]
De Roos, A. J., Rothman, N., Inskip, P. D., Linet, M. S., Shapiro, W. R., Selker, R. G., Fine, H. A., Black, P. M., Pittman, G. S., Bell, D. A.
(2003). Genetic Polymorphisms in GSTM1, -P1, -T1, and CYP2E1 and the Risk of Adult Brain Tumors. Cancer Epidemiol. Biomarkers Prev.
12: 14-22
[Abstract][Full Text]
Wrensch, M., Minn, Y., Chew, T., Bondy, M., Berger, M. S.
(2002). Epidemiology of primary brain tumors: Current concepts and review of the literature. Neuro Oncol Duke
4: 278-299
[Abstract]
Groves, F. D., Page, W. F., Gridley, G., Lisimaque, L., Stewart, P. A., Tarone, R. E., Gail, M. H., Boice, J. D. Jr, Beebe, G. W.
(2002). Cancer in Korean War Navy Technicians: Mortality Survey after 40 Years. Am J Epidemiol
155: 810-818
[Abstract][Full Text]
Muscat, J. E., Malkin, M. G., Shore, R. E., Thompson, S., Neugut, A. I., Stellman, S. D., Bruce, J.
(2002). Handheld cellular telephones and risk of acoustic neuroma. Neurology
58: 1304-1306
[Abstract][Full Text]
Takahashi, S., Inaguma, S., Cho, Y.-M., Imaida, K., Wang, J., Fujiwara, O., Shirai, T.
(2002). Lack of Mutation Induction with Exposure to 1.5 GHz Electromagnetic Near Fields Used for Cellular Phones in Brains of Big Blue Mice. Cancer Res.
62: 1956-1960
[Abstract][Full Text]
Johansen, C., Boice, J. D. Jr., McLaughlin, J. K., Olsen, J. H.
(2001). Response: Cellular Telephones and Cancer--a Nationwide Cohort Study in Denmark. JNCI J Natl Cancer Inst
93: 878a-879a
[Full Text]
Lubin, E., Sullivan, P. R., DeAngelis, L. M.
(2001). Brain Tumors. NEJM
344: 1478-1479
[Full Text]
Erman, M., Celik, I., Kars, A., Kundi, M., Kane, R. C., Inskip, P. D., Tarone, R. E., Linet, M. S.
(2001). Cellular Telephones and Brain Tumors. NEJM
344: 1331-1332
[Full Text]
Kane, R. C., Hardell, L., Mild, K. H., Zimmerman, S. M., Zimmerman, R. W., Muscat, J. E., Stellman, S. D., Malkin, M. G., Thompson, S., Shore, R. E., Neugut, A. I.
(2001). Handheld Cellular Telephones and Brain Cancer Risk. JAMA
285: 1838-1839
[Full Text]
(2001). Cell Phone Use Not Associated with Brain Cancer. JWatch Psychiatry
2001: 14-14
[Full Text]
(2001). Cell Phone Use Not Associated with Brain Cancer. JWatch Gastroenterology
2001: 15-15
[Full Text]
Frumkin, H., Jacobson, A., Gansler, T., Thun, M. J.
(2001). Cellular Phones and Risk of Brain Tumors. CA Cancer J Clin
51: 137-141
[Full Text]
(2001). Cell Phone Use and Risk for Brain Cancer. JWatch Neurology
2001: 5-5
[Full Text]
Park, R. L.
(2001). Cellular Telephones and Cancer: How Should Science Respond?. JNCI J Natl Cancer Inst
93: 166-167
[Full Text]
Johansen, C., Boice, J. D. Jr., McLaughlin, J. K., Olsen, J. H.
(2001). Cellular Telephones and Cancer--a Nationwide Cohort Study in Denmark. JNCI J Natl Cancer Inst
93: 203-207
[Abstract][Full Text]
Gellin, B. G., Schaffner, W.
(2001). The Risk of Vaccination -- The Importance of "Negative" Studies. NEJM
344: 372-373
[Full Text]
(2001). Cell Phone Use Not Associated with Brain Cancer. JWatch General
2001: 4-4
[Full Text]
DeAngelis, L. M.
(2001). Brain Tumors. NEJM
344: 114-123
[Full Text]
Trichopoulos, D., Adami, H.-O.
(2001). Cellular Telephones and Brain Tumors. NEJM
344: 133-134
[Full Text]
level of 0.05). Odds ratios were used as estimates of the relative risk. In addition to the matching variables, the analyses accounted for the date of the interview (a continuous variable), the type of respondent (the subject, a proxy, or both), educational level (less than high school, high school or general equivalency diploma, one to three years of college, or four years of college or more), annual household income (six categories, from less than $15,000 to $75,000 or more), type of health coverage, marital status, religion, history of radiotherapy to the head or neck, and handedness. 
OR+1) ÷ 2 (where OR denotes odds ratio), on the basis of an approach described in the Appendix and elsewhere (*). Two-sided P values for tests of independence were based on Fisher's exact test. 
Pr(BL|PR) and Pr(BR|PR)=
