case control study brain cancer

Case-control study of the association between malignant brain tumours diagnosed between 2007 and 2009 and mobile and cordless phone use

Affiliation.

• 1 Department of Oncology, University Hospital, SE-701 85 Örebro, Sweden.
• PMID: 24064953
• PMCID: PMC3834325
• DOI: 10.3892/ijo.2013.2111

Previous studies have shown a consistent association between long-term use of mobile and cordless phones and glioma and acoustic neuroma, but not for meningioma. When used these phones emit radiofrequency electromagnetic fields (RF-EMFs) and the brain is the main target organ for the handheld phone. The International Agency for Research on Cancer (IARC) classified in May, 2011 RF-EMF as a group 2B, i.e. a 'possible' human carcinogen. The aim of this study was to further explore the relationship between especially long-term (>10 years) use of wireless phones and the development of malignant brain tumours. We conducted a new case-control study of brain tumour cases of both genders aged 18-75 years and diagnosed during 2007-2009. One population-based control matched on gender and age (within 5 years) was used to each case. Here, we report on malignant cases including all available controls. Exposures on e.g. use of mobile phones and cordless phones were assessed by a self-administered questionnaire. Unconditional logistic regression analysis was performed, adjusting for age, gender, year of diagnosis and socio-economic index using the whole control sample. Of the cases with a malignant brain tumour, 87% (n=593) participated, and 85% (n=1,368) of controls in the whole study answered the questionnaire. The odds ratio (OR) for mobile phone use of the analogue type was 1.8, 95% confidence interval (CI)=1.04‑3.3, increasing with >25 years of latency (time since first exposure) to an OR=3.3, 95% CI=1.6-6.9. Digital 2G mobile phone use rendered an OR=1.6, 95% CI=0.996-2.7, increasing with latency >15-20 years to an OR=2.1, 95% CI=1.2-3.6. The results for cordless phone use were OR=1.7, 95% CI=1.1-2.9, and, for latency of 15-20 years, the OR=2.1, 95% CI=1.2-3.8. Few participants had used a cordless phone for >20-25 years. Digital type of wireless phones (2G and 3G mobile phones, cordless phones) gave increased risk with latency >1-5 years, then a lower risk in the following latency groups, but again increasing risk with latency >15-20 years. Ipsilateral use resulted in a higher risk than contralateral mobile and cordless phone use. Higher ORs were calculated for tumours in the temporal and overlapping lobes. Using the meningioma cases in the same study as reference entity gave somewhat higher ORs indicating that the results were unlikely to be explained by recall or observational bias. This study confirmed previous results of an association between mobile and cordless phone use and malignant brain tumours. These findings provide support for the hypothesis that RF-EMFs play a role both in the initiation and promotion stages of carcinogenesis.

Publication types

• Research Support, Non-U.S. Gov't
• Brain / pathology
• Brain Neoplasms / epidemiology*
• Brain Neoplasms / etiology
• Carcinogenesis / radiation effects
• Case-Control Studies
• Cell Phone / statistics & numerical data*
• Electromagnetic Fields / adverse effects*
• Environmental Exposure
• Glioma / epidemiology
• Glioma / etiology
• Meningioma / epidemiology
• Meningioma / etiology
• Middle Aged
• Neuroma, Acoustic / epidemiology
• Neuroma, Acoustic / etiology
• Registries / statistics & numerical data
• Surveys and Questionnaires
• Telephone / statistics & numerical data*
• Time Factors
• Young Adult

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings
• Advanced Search
• Journal List
• Am J Public Health
• v.77(9); Sep 1987

Brain cancer and nonoccupational risk factors: a case-control study among workers at two nuclear facilities.

In a nested case-control study of nuclear workers, 82 brain cancer cases were compared with 328 matched controls to investigate the possible association with nonoccupational risk factors such as histories of epilepsy or head injury. We observed a moderately strong association between brain cancer occurrence and history of epilepsy (OR = 5.7, 95 per cent CI: 1.0, 32.1), but did not find a positive association with previous head injury (OR = 0.9, 95 per cent CI: 0.2, 4.2).

Full text is available as a scanned copy of the original print version. Get a printable copy (PDF file) of the complete article (593K), or click on a page image below to browse page by page. Links to PubMed are also available for Selected References .

Selected References

These references are in PubMed. This may not be the complete list of references from this article.

• Gold E, Gordis L, Tonascia J, Szklo M. Risk factors for brain tumors in children. Am J Epidemiol. 1979 Mar; 109 (3):309–319. [ PubMed ] [ Google Scholar ]
• White SJ, McLean AE, Howland C. Anticonvulsant drugs and cancer. A cohort study in patients with severe epilepsy. Lancet. 1979 Sep 1; 2 (8140):458–461. [ PubMed ] [ Google Scholar ]
• Hochberg F, Toniolo P, Cole P. Head trauma and seizures as risk factors of glioblastoma. Neurology. 1984 Nov; 34 (11):1511–1514. [ PubMed ] [ Google Scholar ]
• Shirts SB, Annegers JF, Hauser WA, Kurland LT. Cancer incidence in a cohort of patients with seizure disorders. J Natl Cancer Inst. 1986 Jul; 77 (1):83–87. [ PubMed ] [ Google Scholar ]
• Preston-Martin S, Paganini-Hill A, Henderson BE, Pike MC, Wood C. Case-control study of intracranial meningiomas in women in Los Angeles County, California. J Natl Cancer Inst. 1980 Jul; 65 (1):67–73. [ PubMed ] [ Google Scholar ]
• Percy C, Sobin L. Surveillance, epidemiology, and end results lung cancer data applied to the World Health Organization's classifications of lung tumors. J Natl Cancer Inst. 1983 Apr; 70 (4):663–666. [ PubMed ] [ Google Scholar ]
• Choi NW, Schuman LM, Gullen WH. Epidemiology of primary central nervous system neoplasms. II. Case-control study. Am J Epidemiol. 1970 May; 91 (5):467–485. [ PubMed ] [ Google Scholar ]
• Ahlbom A, Navier IL, Norell S, Olin R, Spännare B. Nonoccupational risk indicators for astrocytomas in adults. Am J Epidemiol. 1986 Aug; 124 (2):334–337. [ PubMed ] [ Google Scholar ]
• Polednak AP, Frome EL. Mortality among men employed between 1943 and 1947 at a uranium-processing plant. J Occup Med. 1981 Mar; 23 (3):169–178. [ PubMed ] [ Google Scholar ]
• Checkoway H, Mathew RM, Shy CM, Watson JE, Jr, Tankersley WG, Wolf SH, Smith JC, Fry SA. Radiation, work experience, and cause specific mortality among workers at an energy research laboratory. Br J Ind Med. 1985 Aug; 42 (8):525–533. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rothman KJ. Induction and latent periods. Am J Epidemiol. 1981 Aug; 114 (2):253–259. [ PubMed ] [ Google Scholar ]
• GARCIA JH, OKAZAKI H, ARONSON SM. BLOOD-GROUP FREQUENCIES AND ASTROCYTOMATA. J Neurosurg. 1963 May; 20 :397–399. [ PubMed ] [ Google Scholar ]
• BUCKWALTER JA, TURNER JH, GAMBER HH, RATERMAN L, SOPER RT, KNOWLER LA. Psychoses, intracranial neoplasms, and genetics. AMA Arch Neurol Psychiatry. 1959 Apr; 81 (4):480–485. [ PubMed ] [ Google Scholar ]
• SELVERSTONE B, COOPER DR. Astrocytomas and ABO blood groups. J Neurosurg. 1961 Sep; 18 :602–604. [ PubMed ] [ Google Scholar ]

Log in using your username and password

• Search More Search for this keyword Advanced search
• Latest content
• Current issue
• BMJ Journals More You are viewing from: Google Indexer

You are here

• Volume 71, Issue 7
• Mobile phone use and brain tumours in the CERENAT case-control study
• Article Text
• Article info
• Citation Tools
• Rapid Responses
• Article metrics

• Gaëlle Coureau 1 , 2 , 3 ,
• Ghislaine Bouvier 1 , 2 ,
• Pierre Lebailly 4 , 5 , 6 ,
• Pascale Fabbro-Peray 7 , 8 ,
• Anne Gruber 1 ,
• Karen Leffondre 2 ,
• Jean-Sebastien Guillamo 9 ,
• Hugues Loiseau 10 ,
• Simone Mathoulin-Pélissier 2 ,
• Roger Salamon 2 , 3 ,
• Isabelle Baldi 1 , 2 , 11
• 1 Laboratoire Santé Travail Environnement , Univ. Bordeaux, ISPED, Bordeaux , France
• 2 INSERM, ISPED, Centre INSERM U897-Epidémiologie-Biostatistique, Bordeaux , France
• 3 CHU de Bordeaux, Service d'information médicale, Bordeaux , France
• 4 INSERM, UMR1086-Cancers et Préventions, Caen , France
• 5 Univ. Caen Basse-Normandie, Caen , France
• 6 Centre François Baclesse, Caen , France
• 7 Laboratoire d'Epidémiologie et de Biostatistiques , Univ. Montpellier, Institut Universitaire de Recherche Clinique, Montpellier , France
• 8 Département d'informatique médicale , CHU de Nîmes, Nîmes , France
• 9 Département de neurologie , CHU de Caen, Caen , France
• 10 Service de Neurochirurgie , CHU de Bordeaux, Bordeaux , France
• 11 Service de Médecine du Travail , CHU de Bordeaux, Bordeaux , France
• Correspondence to Dr Gaëlle Coureau, Université Bordeaux Segalen, ISPED, Equipe Santé Travail Environnement, 146 rue Léo Saignat, 33076 Bordeaux, Cedex, France; gaelle.coureau{at}isped.u-bordeaux2.fr

The carcinogenic effect of radiofrequency electromagnetic fields in humans remains controversial. However, it has been suggested that they could be involved in the aetiology of some types of brain tumours.

Objectives The objective was to analyse the association between mobile phone exposure and primary central nervous system tumours (gliomas and meningiomas) in adults.

Methods CERENAT is a multicenter case-control study carried out in four areas in France in 2004–2006. Data about mobile phone use were collected through a detailed questionnaire delivered in a face-to-face manner. Conditional logistic regression for matched sets was used to estimate adjusted ORs and 95% CIs.

Results A total of 253 gliomas, 194 meningiomas and 892 matched controls selected from the local electoral rolls were analysed. No association with brain tumours was observed when comparing regular mobile phone users with non-users (OR=1.24; 95% CI 0.86 to 1.77 for gliomas, OR=0.90; 95% CI 0.61 to 1.34 for meningiomas). However, the positive association was statistically significant in the heaviest users when considering life-long cumulative duration (≥896 h, OR=2.89; 95% CI 1.41 to 5.93 for gliomas; OR=2.57; 95% CI 1.02 to 6.44 for meningiomas) and number of calls for gliomas (≥18 360 calls, OR=2.10, 95% CI 1.03 to 4.31). Risks were higher for gliomas, temporal tumours, occupational and urban mobile phone use.

Conclusions These additional data support previous findings concerning a possible association between heavy mobile phone use and brain tumours.

• Case-control studies
• Electromagnetic fields
• Mobile phone
• Radiofrequency electromagnetic fields

http://dx.doi.org/10.1136/oemed-2013-101754

Statistics from Altmetric.com

Request permissions.

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.

What this paper adds

The potential association between mobile phone use and brain tumour remains controversial, and original data have mostly been provided by studies performed in Sweden and the international Interphone study.

Some studies suggest that long-term (over 10 years) mobile phone use increases the risk of gliomas, and especially of those with temporal location.

This analysis highlights a positive association between heavy use of mobile phone and brain tumour, considering life-long cumulative duration and number of calls.

Risks were higher for gliomas, temporal tumours, occupational and urban mobile phone use.

This study provides additional data supporting a possible association between heavy mobile phone use and brain tumours.

Introduction

The number of mobile phone subscriptions over the last decade has increased ninefold to reach a startling 6 billion users worldwide in 2011, according to the International Telecommunication Union. From the 1980s, mobile phones have evolved over four different generations, and services have developed very fast (text messaging, internet access, etc). These changes have led to a dramatic growth in mobile phone usage. According to the French Telecommunications and Posts Regulator, the individual mean use for calls in France today is 150 min/month (+27% since 2000), excluding other services and specific usages, such as occupational ones.

The potential carcinogenic effects of radiofrequency electromagnetic fields (RF-EMF) remain controversial. In vitro studies have explored various hypotheses including genotoxicity, cell proliferation, apoptosis, gene expression and direct effect on proteins, but there is still no consensus. 1 Owing to direct contact with the head during communications, the potential association between brain tumours and mobile phone use has become a foremost concern.

For 15 years, original data have mostly been provided by case-control studies, including four studies performed in Sweden, 2–6 and the international Interphone study. 7 Only two cohort studies have addressed the issue; one initiated in Denmark in 1982, 8 and one in the UK in 1996 (Million Women Study). 9 Several meta-analyses have also been performed, 10–14 but most of them have been unable to demonstrate any association between regular mobile phone use (yes/no) and brain tumours, A recent meta-analysis performed by Repacholi et al 15 reported no association (OR=1.1; 95% CI 0.9 to 1.3 for gliomas, OR=0.9; 95% CI 0.8 to 1.1 for meningiomas), whatever the delay since first use. However, beyond these overall risks, some results deserve specific attention. The Interphone study showed an increase in the risk of glioma in the group with the longest duration of use (≥1640 h) (OR=1.4; 95% CI 1.0 to 1.9), higher for ispsilateral use and temporal tumours. This association was not observed for meningiomas (OR=1.15; 95% CI 0.81 to 1.62). 16 Similar results were obtained in some meta-analyses, which showed an increased risk for gliomas and acoustic neuromas with long-term (over 10 years), 14 or long duration use (≥1640 h), 12 ipsilateral use, 14 and temporal location of the tumour. 12 By contrast, the two cohorts that did not face recall bias, showed no increased risk of glioma or meningioma. 8 , 9

As mobile phone use is a recent phenomenon, the uncertainties are larger for slow-growing tumours, such as meningiomas and for long-term use, so exposure assessment is a major challenge and may contribute to the heterogeneity between studies. On the basis of these data, the International Agency for Research on Cancer (IARC) classified RF-EMF as possibly carcinogenic to humans (group 2B) in 2011. 17 Since the aetiology of brain tumours is still largely unknown, additional studies are needed.

The objective of our study was to investigate the relationship between brain tumours and mobile phone use among adults in France.

CERENAT is a multicenter population-based case-control study initiated in 2004 and designed to study the role of environmental and occupational factors in the occurrence of primary central nervous system (CNS) tumours in adults.

CERENAT cases were all subjects aged 16 years and over, with a benign or malignant CNS tumour diagnosed between June 2004 and May 2006, and living in one of four French areas (Gironde, Calvados, Manche, Hérault) at diagnosis. Cases were identified with the collaboration of a network of practitioners involved in the diagnosis and therapeutic management of patients and, with the aim of being exhaustive, from population-based cancer registries. All diagnoses were established by either a neuropathological or, for cases with no histological diagnosis, a clinical and radiological assessment. Primary brain tumours with the following ICDO-3 topography codes were included: C70.0-C70.9, C71.0-C71.9 and C72.2-C72.9. Patients with recurrent tumours, metastases, pituitary tumours, genetic syndrome or AIDS were excluded. Cases were grouped according to morphology codes as gliomas, meningiomas, acoustic neuromas, lymphomas and other unspecified primary brain tumours. 18 In this analysis, only cases of gliomas and meningiomas were considered. Medullary tumours were excluded because the exposure of the spinal cord to RF-EMF from mobile phone use is significantly lower.

For each case, two controls with no history of CNS tumour were randomly selected from the local electoral rolls during the period 2005–2008, individually matched on age (±2 years), sex and department of residence.

Data collection

Data were collected through standardised questionnaires delivered as face-to-face non-blinded structured interviews by trained interviewers. When cases were in a severe clinical condition or deceased, a proxy was invited to complete a simplified questionnaire, which was subsequently completed by their matched controls. The questionnaire covered sociodemographic characteristics, medical history, lifestyle and detailed occupational and environmental data.

Assessment of mobile phone use

A detailed questionnaire including a set of questions on phone use was completed by all subjects regarding themselves as regular users (ie, phoning at least once a week for 6 months or more) (see online supplementary appendix 1). For each new mobile phone or major change of use, the same questionnaire was completed again. Information concerning each mobile phone included: phone model (analogue or digital); beginning and end dates for the use of the phone; average number and duration of calls made and received per month during each use period; shared or individual use; occupational or personal use and hands-free kit use. Duration of calls per month was reported by the subjects or assessed from duration of cards or packages that subjects reported to use monthly (4.5% of the mobile phone users; 5.2% and 3.1% for cases and controls, respectively). Only the dates of use and duration of calls were sought from proxies and their matched controls in the simplified questionnaire.

Potential confounders

The following potential confounders were considered: level of education (primary school or less, secondary school, high school and university), smoking (non-smokers, former smokers, current smokers), alcohol consumption (classified as excessive in men over three glasses of wine, cider, beer or spirits per day, and over two glasses per day in women).

Potential occupational confounders were identified from detailed job calendars, and from specific questions about exposure to pesticides, extremely low-frequency electromagnetic fields (ELF-EMF), RF-EMF, and ionising radiation. 19 , 20 Specifically, pesticide exposure was defined as having performed treatment tasks on crops, gardens, wood, or other circumstances in any job during life. Subjects were classified as occupationally exposed to ELF-EMF if they had worked with welding equipment, grinding machines, induction or microwave ovens, electric machines in the medical sector, industrial machinery in the wood, textile, building, food processing and steel sectors; in the electronics industry; or near power lines. Concerning RF-EMF, jobs with exposure to metal detectors, demagnetisers, porticos or transmission devices were taken into account. Subjects reporting exposure to radioactive sources, use of equipment emitting or measuring radiation, or working at a nuclear site, were considered occupationally exposed to ionising radiation.

The date of diagnosis was taken as the index date for each case and its two matched controls. Phone use during the year before the index date was not taken into account in the exposure assessment for accounting for a potential induction period and to eliminate any reverse causality bias due to prodromal effects. The reference category for all exposure variables comprised persons who were not regular phone users. Mobile phone exposure was assessed using the number of years since first regular use, average length of calls per month (hours), average number of calls per day, cumulative lifetime duration of calls (hours) and cumulative lifetime number of calls. Cumulative lifetime duration and number of calls were the sum of duration of calls and number of calls for each mobile phone reported. For all subjects using a hands-free kit or sharing their phone less than 50% of time, 50% of time, more than 50% of time and all the time, cumulative duration and number of calls (for hand-free kit only) were weighted by coefficients of 0.75, 0.50, 0.25 and 0.10, respectively.

Different aspects of mobile phone use were grouped into a number of categories for the analysis: time since first use was in three categories ((1–4), (5–9) and ≥10 years); average time of calls in four categories (<2, (2–4), (5–14) and ≥15 h/month); average number of calls in four categories (≤1, (2–4), (5–9) and ≥10 calls per day); and duration and number of calls into five categories based on the distribution of values observed in controls (<25th, (25–49th), (50–74th), (75–89th), ≥90th percentile). The latter category was retained because of the large range of values and in accordance with previous findings. 16

Conditional logistic regression for matched sets was used to estimate ORs and 95%CIs. All statistical tests were two-sided, and a global test for each categorical indicator was performed. Confounders were selected using the purposeful selection algorithm, 21 which combines the principles of significance and change-in-estimate in selecting variables for a final model. Each indicator was analysed separately and adjusted for confounders. Two sensitivity analyses were performed: the first one excluded proxy-interviews, as information was supposed more uncertain than the one collected from individuals themselves, and the second one excluded non-regular users and used the lowest category of exposure as the reference. Indeed, non-regular phone users were more often men, younger, more educated and more frequently occupationally exposed to RF-EMF, so that they could also differ in other unmeasured factors. Exposure lagging of 2 and 5 years before the index date was also analysed.

Stratified analyses were performed, and adjusted ORs were re-estimated in strata of tumour location, type of use (occupational/personal only), place of use (urban only/urban and rural), side of use (ipsilateral/contralateral) and phone model. The rationale for stratifying on occupational use was the consideration that it corresponded to a specific profile of use, with more frequent and shorter calls, often made in outdoor settings, potentially with different phone technology for devices being working tools. Side of use was considered as ipsilateral if the phone was used on the same side as the tumour or on both sides. It was defined as contralateral if the phone was used on the opposite side to the tumour. No laterality was assigned for median tumour. Analyses were performed for cases with ipsilateral use or no use and their matched controls, and then for cases with contralateral use or no use and their matched controls. Owing to the restricted number of matched case-control sets within each stratum, unconditional logistic regression adjusting for age and sex and for confounders was used for the stratified analyses on different mobile phone uses (type, place, phone model). For each analysis, mobile phone users in a stratum were compared with all non-regular users. We present the adjusted ORs for the last decile of cumulative duration (heavy mobile phone users).

Analyses were performed with the software SAS, V.9.2 (SAS Institute, Cary, North Carolina, Etats-Unis, USA).

Population characteristics

Out of the subjects defined as eligible, 95% of cases and 61% of controls were contacted, and a total of 596 (73%) cases and 1192 (45%) controls were finally included in the CERENAT study. Participation rate was 66% for glioma and 75% for meningioma cases. The main reasons for non-participation were refusals, severe condition or death without proxy. Non-included cases were older than included cases (mean age: 63 vs 58 years for gliomas and meningiomas). After exclusion of acoustic neuromas (n=42), lymphomas and unspecified brain tumours (n=56), medullar tumours (n=50), and persons with missing data on regular mobile phone use (two controls for gliomas plus one meningioma case and his two controls), 1339 subjects were analysed: 253 cases and 504 controls for gliomas; 194 cases and 388 controls for meningiomas. For gliomas and meningiomas, neuropathological assessment represented 96% of diagnoses, and clinical and radiological assessment 4%.

Median time between the index date and interview was 6 months (IQR: 4, 10) for cases, and 21 months (IQR: 16, 30) for controls, similar for gliomas and meningiomas. The proportion of proxy interviews was 25% for gliomas and 6% for meningiomas ( table 1 ). The average age was 56 years for gliomas and 60 years for meningiomas, and women represented 43% and 75% of the population, respectively. The level of education was higher in controls than in cases (p<10 −3 ).

• View inline

Description of study population. CERENAT, 2004–2006, France

Mobile phone use

Regular use was reported by half the total population, and in the same proportion in cases and controls, (55% for gliomas cases and controls, and 44% for meningiomas cases and controls). On average, users reported having used two different phones in their lifetime. One-third of regular users were occupational users, and 27% of users shared at least one of their phones with someone, but most of them (55%) were the main users. A hands-free kit was used by only 14% of the individuals. Among 64% of specified phone models, most (92%) were digital and a few (8%) were analogue. Table 2 presents characteristics of mobile phone use. Only 12% of the individuals used their phone for 10 years or more (45% for 1–4 years, and 43% for 5–9 years). The median cumulative lifetime duration of calls was 115 h (IQR: 41, 383) with values ranging from 0.7 to 18 612 and 0.2 to 7290 for glioma cases and controls, respectively, and from 3.8 to 4845 and 0.8 to 16 000 for meningioma cases and controls, respectively. The median calling time was 2.7 h/month (IQR: 1.2, 7.5) with values ranging from 0.1 to 198 and 0 to 91 h/month for glioma cases and controls, respectively, and from 0.2 to 100 and 0.1 to 200 for meningioma cases and controls.

Mobile phone use among regular users (n=417/757 gliomas strata; n=253/582 meningiomas strata). CERENAT, 2004–2006, France

Mobile phone users were more often men than non-users (49% vs 38%, p<10 −3 ). They were also younger (median 54 vs 66 years old, p<10 −3 ), more educated (university level for 35% vs 16%, p<10 −3 ), and more occupationally exposed to RF-EMF (7% vs 4%, p=0.02).

Univariate analysis

The proportion of regular users was comparable in cases (57%) and controls (54%), and mobile phone use characteristics are presented in table 2 . An association with gliomas was observed in subjects with the longest cumulative lifetime duration of calls (≥90th percentile ie, 896 h) (OR=2.33; 95% CI 1.17 to 4.67). These results were unchanged when weighting the values for shared use and hands-free kit use, and when excluding simplified questionnaire respondents.

Multivariate analyses

A slight positive association was observed for gliomas in users versus non-users (OR=1.24; 95% CI 0.86 to 1.77) ( table 3 ). Risks tended to increase with time since first use. An association was found with average time of calls per month ( global p<10 −3 ) and with average number of calls per day ( global p=0.04). The increase in risk as compared to non-regular users, was observed for an average of more than 15 h of calls per month (OR=4.21; 95% CI 2.00 to 8.87). Risks increased with cumulative duration of calls ( global p=0.02), but not with cumulative number of calls ( global p=0.41). The increase in risk was statistically significant only in the 90th percentile for cumulative duration of calls (OR=2.89; 95% CI 1.41 to 5.93) and cumulative number of calls (OR=2.10; 95% CI 1.03 to 4.31).

Adjusted conditional logistic regression for each mobile phone use indicator. CERENAT, 2004–2006, France

Analyses excluding proxies provided similar results ( table 4 ) and the ORs were almost unchanged when weighting for shared use and hands-free kit use. In sensitivity analyses excluding non-regular users, and considering the first class of regular mobile phone users as reference, we observed a trend for an increased risk with the cumulative duration of calls that was statistically significant for the last decile (OR=0.88 (0.44 to 1.78); OR=1.30 (0.66 to 2.55); OR=1.96 (0.97 to 3.96); OR=2.53 (1.17 to 5.46)).

Adjusted conditional logistic regression after exclusion of simplified questionnaires. CERENAT, 2004–2006, France

Meningiomas

The proportion of regular users was comparable in cases (41%) and controls (45%). A positive association was observed in subjects with the longest cumulative duration of calls (≥896 h) (OR=2.29; 95% CI 0.94 to 5.58), a result also observed in subjects with the highest number of calls (≥18 360 calls) (OR=1.73; 95% CI 0.66 to 4.50).

No association was observed for meningiomas when considering regular phone users versus non-users (OR=0.90; 95% CI 0.61 to 1.34). An increased OR for more than 15 h of calls per month was observed for meningiomas (OR=2.01; 95% CI 0.84 to 5.22). For cumulative duration of calls, a statistically significant association was observed in the last decile OR=2.57 (95% CI 1.02 to 6.44).

The average number of calls per day was not associated with meningiomas, and the risk was not significantly increased with the cumulative number of calls.

As previously, excluding proxies did not have any effect in the results ( table 4 ). In sensitivity analyses excluding non-regular users, no trend for an increased risk with the cumulative duration of calls was observed when restricting to the regular mobile phone users.

Heavy users and stratified analyses

Among heaviest users (cumulative duration ≥896 h), time since first use was occasionally less than 5 years (11%) but mostly 5–9 years (49%) and 10 years and more (40%) ( table 5 ). Thirty-three per cent of them were commercial agents or sales people, and 22% were chief operating officers or production and operation managers. Sixty-two per cent of them reported occupational mobile phone use. Their median cumulative duration of calls was 1925 h, corresponding to 54 min/day (IQR: 30, 96 min), with a maximum of 6.6 h/day.

Associations between heavy mobile phone use (last decile of cumulative duration) and tumours according to censorship, tumour location and characteristics of use. CERENAT, 2004–2006, France

For gliomas, considering a 5-year latency period led to an increased OR for the last decile compared with non-regular users (5.30; 95% CI 2.12 to 13.23). Temporal location of the tumour presented a higher OR compared with the frontal one. The risk of glioma with occupational use was tripled (OR=3.27; 95% CI 1.45 to 7.35) and exclusively urban setting use was associated with an OR=8.20 (1.37, 49.07). A positive association was observed for ipsilateral tumours while it was negative for contralateral tumours (see online supplementary appendix 2).

For meningiomas, CIs were wider because of the smaller sample size. However, extending the latency period to 2 or 5 years before the index date appeared to decrease the risk. The higher OR was observed for temporal meningiomas, and the risk for ipsilateral tumour was slightly higher than for contralateral tumours.

Finally, same trends were observed in men and women and with regards to age, with higher associations for men and for 30–59-year-old subjects (data not shown).

This study provides additional data on the relationship between RF-EMF exposure and brain tumours. No statistically significant increase in brain tumours was observed in regular users versus non-users. In the heaviest users, however, we found a positive association that was stronger for gliomas and that increased with a 5-year latency before diagnosis. This association was more pronounced for occupational users and in urban settings.

This multicentric study was conducted on the general population, and covered various socioeconomic statuses and environmental and occupational exposures. Cases were included from a clinical network supported by population-based cancer registries, thereby ensuring the reliability of the diagnoses. Controls were randomly selected from the electoral rolls, which include 90% of persons over 18 years, and are representative of the French adult population regarding age and sex. 22 The participation rates (66% and 75% for glioma and meningioma cases, respectively, and 45% for controls) were lower than those reported in previous studies, 23–26 but similar to those of the Interphone study. 16 Unfortunately, the lack of a questionnaire for non-participants prevented us from accurately assessing selection bias. However, the study was presented to participants as dealing with environmental and occupational factors and health in general, and was not focused on mobile phone use. Half the study population reported a regular use, and 63% of 30–59-year-old persons (66% in cases and 62% in controls). This prevalence is slightly higher than in the French part of the Interphone study (54% of regular use for 30–59-year-old cases and 56% for controls in 2000–2004 27 ), but comparable with mobile phone use reported in France in 2003, (62% in 40–59-year-old persons). 28 Thus, non-participation had no evident reason to be specially related to mobile phone use.

RF-EMF exposure from mobile phones was assessed with a face-to-face standardised questionnaire, thus limiting a priori misinterpretation of questions by individuals, and missing responses. It was not possible to blind the case/control status of subjects, but the interviewing team endeavoured to standardise data collection in cases and controls at all stages. The delay between index date and interview was longer for controls, but we censured information on phone use after the index date, and no increase in mobile phone use was observed in the period elapsed since index date in controls.

Some interviews had to be conducted with a proxy because of the health status of the cases. A simplified questionnaire was then used in cases and in matched controls to prevent any differential bias related to simplified questions, even though the quality of data obtained from proxies remains questionable for cases. Nevertheless, analyses excluding simplified questionnaires showed comparable results.

As in any retrospective analysis and in other mobile phone studies performed at the same period, we found indication of recall bias regarding exposure data. Several studies have tried to measure recall bias by crossing the individuals’ reports with operators’ data, published after the beginning of our study. The Interphone validation studies concluded that individuals tended to slightly underestimate the number of calls and overestimate call duration, but no difference was observed between cases and controls. 27 , 29–31 A Finnish study on the validity of self-reported mobile phone use confirmed this trend. 32 An exception would exist for a long time before the interview in the Interphone validation study, where an overestimation was observed, more pronounced for cases than for controls. 30 By contrast, two studies found an overestimation of the number and duration of calls, that increased with phone use. 33 , 34 Thus, like in the Interphone study, finding significant results only in the last decile could suggest that some subjects among the heavy users over-reported their use. We individually checked all extreme values (the maximum was 200 h/month, ie, 6.7 h/day) by reviewing together mobile phone use history and occupational calendars. The information was considered consistent when a plausible reason for the duration was given, for instance, working outdoors, or travelling and having the necessity to contact customers or collaborators to manage appointments or prospect affairs. If recall bias is more pronounced in heaviest users, it is likely that exposure values in the last decile are overestimated. Nevertheless, this should not impact the association we observed when considering exposure in categories. 27 Moreover, if this error is non-differential, 30 associations should be underestimated, and although a differential bias cannot be excluded, underestimation seems to be more likely to occur. 35 To improve exposure assessment, we also considered phone sharing, use of a hands-free kit, occupational use, and urban and rural settings. Since some of the additional analyses were limited by the low numbers, even if most of the estimates show acceptable precision, caution should be taken when interpreting the results.

The lack of statistically significant association when comparing users to non-users is consistent with several previous reports. 16 , 25 , 26 , 33 , 36 , 37 Consistent with previous studies, we found an increased risk in the heaviest users, especially for gliomas. 16 , 24 The statistically significant increase we found was for cumulative duration above about 900 h of use, while the threshold was 1640 h in the Interphone study, 16 , 37 and ranged between 65 and 2000 h in the various swedish studies. 16 , 23 , 24 , 37 Such variations in phone use patterns across different studies and populations impede the definition of a reliable threshold and even to be sure of its reality. Actually, a dose-effect relationship would be more consistent with the role of RF-EMF in the development of tumours. In line with this idea, the trend between categories of use we found in sensitivity analyses for gliomas when considering the lowest phone use (and not non-users) as reference, appears more suggestive of a possible role of RF-EMF.

In our study, we found an increased risk in those subjects reporting a prolonged use, making numerous calls, whose use was especially occupational and more often in urban areas (without correlation between these specific uses). To date, it has not been possible to determine whether the increased risk is related to use over many years or to the cumulative duration of calls. In our study, time since first use was not associated with the presence of a tumour, which may be partly due to the low number of users for 10 years and more. This issue remains controversial, as some studies found an increased risk with high use over a short period, 16 while others demonstrated a risk for prolonged use. 23 , 24 , 33 , 37 Moreover, by censoring exposure 2 years and 5 years before diagnosis, we observed higher associations for gliomas. This could be due to an induction effect of exposure on the emergence of the tumour.

Assuming that RF-EMF emitted by mobile phones are a risk factor for brain tumours owing to their proximity to the head, an increased association for temporal tumours and side of phone use was expected, in accordance with the results of some studies, 16 , 23 , 37 , 38 but not all of them. 33 In our study, the increase in risk was more prominent when considering tumour location, especially for meningiomas, than side of phone use, which could be due to uncertainties in subjects reporting side of use.

As expected, we found a higher risk for temporal location than for frontal one, which was more pronounced for meningiomas. Our results for gliomas are difficult to interpret since the risk for ‘other locations’ is the same as for temporal location. Ipsilateral associations were higher than contralateral ones, but we observed this result whatever the level of phone use and the indicator (see results in online supplementary appendix 2). If we consider that reporting bias mainly concerns heavy users, we would expect increased ORs for ipsilateral use mostly in the higher exposure category. In this circumstance, an overall increase may reflect an overall over-reporting of ipsilateral use by cases, or a real effect of ipsilateral use regardless of the level of use.

Finally, we observed increased OR for urban use for gliomas, a result inconsistent with the hypothesis of a higher RF power output during calls in rural areas, documented by some Swedish study. 39–41 However, our results are consistent with a recent international study showing no difference between rural and urban exposition in most countries except in Sweden, 42 and a Hardell study when considering gliomas separately. 43 Several parameters associated with rural/urban setting are susceptible to impact exposure, such as the concomitant residential exposure and the differences in the types of use and characteristics of users. Additional data including the influence of high density of base station allowing to use low output power for calls in urban areas, but leading to a high number of handovers, during which output power is highest, should be investigated further.

An increase in the incidence of primary brain tumours has been observed in the past decades in several countries. It has been explored whether these trends could be related to change in suspected risk factors, including mobile phone use. 44–48 Results obtained to date are not in favour of a correlation. However, temporal trends and differences between countries are not easy to interpret because of methodological limitation in the recording of brain tumours, including changes with time in recording procedure, and the lack of completeness concerning non-malignant tumours.

Conclusions

This case-control study provides additional data on the relationship between mobile phone use and brain tumours. Considering lifetime cumulative duration of calls, an increased risk appears among the heaviest users, often with occupational use and especially for gliomas. While this is consistent with some other studies, it is difficult to define a level of risk, if any, especially as mobile phone technology is constantly evolving. The rapid evolution of technology has led to a considerable increase in the use of mobile phones and a parallel decrease of RF-EMF emitted by the phones. Studies taking account of these recent developments, and allowing the observation of potential long-term effects will be needed.

Acknowledgments

The authors thank JM Constans, O Coskun, S Eimer and A Vital for their radiological or pathological expertise, C Auguin, G Blaizot, AS Lacauve, L Molinari, E Niez, X Schwall and S Schwall for interviewing the individuals, A Jaffré, V Loyant, N Bousquet, E Berteaud and C Dantas for their helpful collaboration in the study, and all the clinicians who helped us to contact the patients.

• ↵ Health Protection Agency . Health Effects from Radiofrequency Electromagnetic Fields—RCE 20 . Report of an independent Advisory Group on Non-ionising Radiation . 2012 . http://www.hpa.org.uk/Publications/Radiation/DocumentsOfTheHPA/RCE20HealthEffectsfromRFElectromagneticFields/
• Hardell L ,
• Pahlson A ,
• Carlberg M ,
• Hallquist A ,
• Soderqvist F ,
• Richardson L ,
• Deltour I ,
• Poulsen AH ,
• Johansen C ,
• Benson VS ,
• Feychting M ,
• Hansson Mild K
• Söderqvist F ,
• Khurana VG ,
• Repacholi MH ,
• ↵ Interphone . Brain tumour risk in relation to mobile telephone use: results of the INTERPHONE international case-control study . Int J Epidemiol 2010 ; 39 : 675 – 94 . OpenUrl Abstract / FREE Full Text
• Lauby-Secretan B ,
• McCarthy BJ ,
• Coureau G ,
• Williams DK ,
• Pan Ké Shon JL
• Inskip PD ,
• Tarone RE ,
• Muscat JE ,
• Malkin MC ,
• Thompson S ,
• Montestrucq L ,
• ↵ Centre de Recherche pour l'Etude et l'Observation des Conditions de Vie. Enquête Conditions de vie et Aspirations des Français. La diffusion des technologies de l'information dans la société française. November 2003. http://www.arcep.fr/uploads/tx_gspublication/et-credoc2003.pdf .
• Samkange-Zeeb F ,
• Vrijheid M ,
• Armstrong BK ,
• Kurttio P ,
• Salminen T ,
• Parslow RC ,
• Hepworth SJ ,
• McKinney PA
• Krewski D ,
• Auvinen A ,
• Hietanen M ,
• Luukkonen R ,
• ↵ Interphone . Acoustic neuroma risk in relation to mobile telephone use: results of the INTERPHONE international case-control study . Cancer Epidemiol 2011 ; 35 : 453 – 64 . OpenUrl PubMed Web of Science
• Hillert L ,
• Neasham D ,
• Sheppard AR ,
• Forssen U ,
• Vecchia P ,
• de Vocht F ,
• Burstyn I ,
• Hoover RN ,
• Little MP ,
• Rajaraman P ,
• Curtis RE ,

Supplementary materials

Supplementary data.

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Files in this Data Supplement:

• Data supplement 1 - Online appendix 1
• Data supplement 2 - Online appendix 2

Contributors GC worked on mobile phone life-cumulated exposure indicators, analysed the data, and wrote the manuscript. GB contributed to the interpretation of the results and to the writing of the manuscript. PL and PF-P co-supervised with Isabelle Baldi the CERENAT field study and revised the manuscript. AG monitored the study in its enrolment and follow-up stages, managed, checked and corrected the data and reviewed the article. KL contributed to the statistical analysis and revised the manuscript. J-SG and HL contributed to the case inclusions and clinical assessment and revised the manuscript. SM-P contributed to the interpretation of the results and reviewed the article. RS supervised this research and interpretation of data. IB designed the CERENAT study, directed the enrolment and follow-up stages, supervised this research, the analytic strategy, interpretation of data, and helped to write the article.

Funding The study was supported by grants from the Fondation de France, the Agence Française de Sécurité Sanitaire de l’Environnement et du Travail, the Association pour la Recherche contre le Cancer, the Ligue contre le Cancer, the Institut National de la Santé Et de la Recherche Médicale—ATC Environnement et Santé.

Competing interests None.

Patient consent Obtained.

Ethics approval Comité consultatif sur le traitement de l'information en matière de recherche dans le domaine de la santé (CCTIRS); Comités Consultatifs de Protection des Personnes dans la Recherche Biomédicale (CCPPRB); Commission nationale de l'informatique et des libertés (CNIL).

Provenance and peer review Not commissioned; externally peer reviewed.

Read the full text or download the PDF:

• Research article
• Open access
• Published: 08 January 2014

UK case control study of brain tumours in children, teenagers and young adults: a pilot study

• Richard G Feltbower 1 ,
• Sarah J Fleming 1 ,
• Susan V Picton 2 ,
• Robert D Alston 3 ,
• Diana Morgan 1 ,
• Janice Achilles 3 ,
• Patricia A McKinney 1 &
• Jillian M Birch 3  

BMC Research Notes volume  7 , Article number:  14 ( 2014 ) Cite this article

1866 Accesses

5 Citations

1 Altmetric

Metrics details

Tumours of the central nervous system are the second most common group of childhood cancers in 0–14 year olds (24% of total cancers) and represent a major diagnostic group in 15–24 year olds. The pilot case–control study aimed to establish methodologies for a future comprehensive aetiological investigation among children and young adults.

Eligible cases were newly diagnosed with an intracranial tumour of neuroepithelial tissue aged 0–24 years. The pilot recruited patients through Leeds and Manchester Principal Treatment Centres. Controls were drawn from general practice lists. Controls were frequency matched by age and gender.

We interviewed 49 cases and 78 controls comprising 85% of the target sample size. Response rates were 52% for cases and 32% for controls. Completion of the questionnaire was successful, with a very small proportion of missing data being reported (5-10%). The age distribution of cases and controls was similar with around three-quarters of interviewed subjects aged 0–14. Half of cases and almost two-thirds of controls reported using a mobile phone with the majority starting between 10–14 years of age. Prevalence of breastfeeding was lower in cases than controls (Odds Ratio 0.4; 95% CI 0.2-1.2), whilst cases were more likely to be delivered by caesarean section (OR 1.6; 95% CI 0.6-4.4). Cases were significantly more likely to have a birthweight > 3.5 kg compared to controls. Cases were also more likely to come from a family with 3 or more siblings than controls (OR 3.0; 95% CI 0.7-13.6). The majority of participants (>80%) were in favour of taking either blood or saliva to aid molecular epidemiological research.

Conclusions

Successful methods were established for identifying and recruiting a high proportion of case subjects, exploiting strong links with the clinical teams at the treatment centres. Control procedures proved more difficult to implement. However, working closely with national clinical and professional research networks will enable improved control identification and recruitment, with good prospects for collecting biological samples in the future.

Tumours of the central nervous system (CNS) are the second most common group of childhood cancers comprising a quarter of all malignancies in patients aged 0–14 years with approximately 350 children diagnosed each year in the UK [ 1 ]. CNS tumours also represent a major diagnostic group in teenagers and young adults aged 15–24 years with around 150 new cases per year diagnosed in the UK representing 10% of all cancers in this age range [ 2 ]. Survival rates for CNS tumours in young people are generally poor when compared to other cancers occurring among 0–24 year olds: around 50% of these patients die from their disease and those who survive are at particular risk of severely debilitating late effects [ 3 ].

CNS tumours presenting in the young differ notably from those in older adults in terms of the cellular origins, pathological subtypes and anatomic site. The most common subtypes in young people are astrocytic tumours (50%) and embryonal tumours including medulloblastoma (25%) [ 4 ]. Apart from an increasing number of cancer predisposition syndromes associated with early onset CNS tumours, the causes of CNS tumours in children and young adults remain largely unknown.

The only established environmental risk factor for CNS tumours is ionising radiation [ 5 – 8 ]. Exposure to N-nitroso compounds through consumption of cured meat during pregnancy has been consistently reported as an aetiological factor [ 6 , 7 ]. Childhood CNS tumours have also been linked to residential pesticide exposure, traffic pollution, and parental occupations [ 8 – 11 ]. There is accumulating evidence of links between infections and CNS tumours particularly in the young. Supportive evidence for the involvement of infections comes from analyses of space-time clustering of incident cases, geographical and demographic variations in incidence and population mixing [ 12 – 16 ].

Pre-school nurseries have a high prevalence and diversity of infectious disease (e.g. [ 17 ]) and attendance can be considered a proxy for early exposure to infections. Exposure to infection in early life has been investigated in the context of a potential infectious aetiology for childhood type CNS tumours although findings have been inconsistent, varying according to tumour type and exposure of interest [ 18 , 19 ]. Atopic diseases, such as asthma, eczema and allergies, can be markers of immune dysfunction. There is evidence to suggest that atopic conditions may confer a reduced risk of CNS tumours in children [ 20 , 21 ]. Furthermore, specific HLA alleles and haplotypes are associated with relatively higher or lower risks of childhood ALL and possibly also CNS tumours [ 22 , 23 ]. Higher birthweight, especially those born weighing over 4000 g, has also been implicated as a possible causal risk factor for childhood brain tumours [ 24 ], whilst a protective association has been described for maternal farm residence during pregnancy and postnatal contact with birds [ 25 ].

As a forerunner to a population-based case control study of neuroepithelial CNS tumours in children, teenagers and young adults we aimed to undertake a pilot study involving a multidisciplinary team comprising paediatric and adolescent oncologists, research nurses, and epidemiologists. The aims were to 1) establish procedures for optimal case and control ascertainment, 2) pilot a questionnaire and study materials, 3) optimise the collection and storage of biological samples 4) develop a protocol and grant application for the full study.

Case–control selection

Eligible cases were those children and young people who were aged 0–24 years at diagnosis presenting with a primary intracranial tumour of neuroepithelial tissue as defined by WHO classification [ 4 ]. Tumours were classified into the following subtypes: ependymoma, astrocytoma, embryonal and other specified tumours. Cases were identified through clinical teams based in the two UK Principal Treatment Centres of Leeds and Manchester, comprising dedicated paediatric and Teenage and Young Adult oncology units. Approaches to patients/parents were made at a time recommended by the clinical teams. Written, informed consent to take part in the pilot study was obtained. Response rates were assessed by age group, gender, CNS subtype and centre.

Controls from the Leeds centre were randomly selected from general practice (GP) lists to identify a population-based sample and provide access to medical records. As part of the feasibility process, controls were frequency matched according to the age (0–24 years) and sex distribution of the case sample. GP practices were selected whose population demographic (age, sex, social class and population density) reflected those of the larger geographical area. Once approval was obtained, a study administrator based themselves in the practice and randomly selected a list of eligible participants. Study invitation letters were distributed on behalf of the person’s GP. Where a control refused to take part, replacement controls were used and the socio-demographic breakdown of response rates monitored to assess the representation of the participants. From the Manchester centre, three friend controls who fulfilled the required age and sex were selected and interviewed. Numerous GP practices were approached but despite extensive efforts and involvement with the Primary Care Research Network (PCRN), a group which supports clinical research studies involving primary care services in England, we were unable to recruit any practices (see Results).

The pilot study set out to recruit and interview 25 cases and 50 controls per centre (50 cases and 100 controls in total).

Interview materials and processes

Exposure prevalence was assessed through information collected from face-to-face interviews. The interview proforma was designed to be compatible with a large parallel international case–control study covering the Nordic countries [ 26 ], a copy of which is provided in the Additional file 1 . For each centre, an experienced research nurse co-ordinated and conducted interviews with participants and their families. Information was collected on the health of the young person, parental health, the index person’s early social habits as a child and the family histories of cancer for cases. Information was captured by a trained interviewer who administered either a Computer Assisted Personal Interview (CAPI) questionnaire adapted from a parallel Nordic study [ 26 ], or a paper-based questionnaire which was then transferred onto a Microsoft Access database. Parental interviews were undertaken for those aged under 12 years; for older subjects, both parents and cases were interviewed (Additional file 1 ).

Ethical approval for the study was granted by the North West Research Ethics Committee in July 2007 (reference number 07/MRE08/46) and informed consent obtained for every participant. The study conformed to the principles embodied in the Declaration of Helsinki. The recruitment periods were September 2007 to March 2009 in Leeds and June 2008 to June 2010 in Manchester.

Statistical analysis

Conditional logistic regression stratified by age (5-year age groups) and sex was undertaken to derive odds ratios (OR) and 95% CI. Adjustment for deprivation was carried out using the Townsend score of the child’s address at diagnosis by the use of Townsend score quintiles based on the UK population distribution. In view of the limited sample size and power and range of possible aetiological factors involved in the development of CNS tumours in children and young people, we undertook a careful regression analysis including a small number of risk factors which were deemed important based on the epidemiological literature (breastfeeding, caesarean section, birthweight, number of siblings, mobile phone usage, contact with animals). A full list of risk factors collected from the interviews is provided in the Additional file 1 . All analyses were carried out using Stata version 12.1.

Recruitment and participant characteristics

Aggregating the data from across both centres yielded 49 cases and 78 controls who were interviewed. Overall, although both centres experienced some problems in terms of recruitment, across both centres we recruited 85% of the target sample size. The flowchart in Figure  1 describes the recruitment pathways and number of subjects identified at each stage for both centres combined.

Recruitment flowchart (all centres combined).

Table  1 summarises the age, sex and CNS subtype distribution by case and control status. 81% of cases were aged 0–14 years at interview compared to two-thirds of controls. There was a notable excess of controls diagnosed aged 15–19 (22%) compared to cases (10%). Slightly more cases were male (59%) with a slightly higher percentage of controls being female (53%).

Response rates were 52% and 32% for cases and controls respectively out of those who were eligible for the study (Figure  1 ). We found recruitment to be a significant challenge notably for cases in Leeds and controls in Manchester. These were largely attributable to changes in NHS governance and the GP contract during the recruitment phase. Main reasons for not taking part were reported as refusal (4 cases and 27 controls from Leeds; 9 cases from Manchester) and untraceable subjects (5 cases and 139 controls from Leeds; 15 cases from Manchester). Completion of the questionnaire however was a success, with a very small proportion of missing data being reported (typically 5-10% for each variable). Where missing data were present, this largely related to the same individuals.

From Manchester deprivation scores were available from cases who did not take part and it was found that there was no significant difference in Townsend score between the two groups. In the Leeds area deprivation was available for interviewed and non-interviewed controls and when comparing deprivation quintile there was found to be a significant trend of reducing participation with increasing quintiles of deprivation (OR: 0.6, 95% CI: 0.5-0.8, p < 0.001).

Half of cases and almost two-thirds of controls indicated that they had used a mobile phone (Table  2 ). The majority of respondents who used a mobile device began doing so when they were 10–14 years of age. Excluding missing data, the prevalence of breastfeeding was lower in cases (73%) compared to controls (83%), whilst cases (25%) were more likely to be delivered by caesarean section than controls (15%) (Table  2 ). 61% of cases had a birthweight in excess of 3.5 kg compared to only 36% of controls. Cases were also more likely to come from a family with 3 or more siblings (31%) than controls (12%). A lower proportion of cases (29%) reported having regular contact with animals outside than controls (40%) (Table  2 ).

Logistic regression modelling for selected birth related and environmental factors (Table  2 ) showed some evidence of a reduced risk for ever having being breastfed (OR: 0.4, 95% CI: 0.2-1.2). There was a statistically significant finding of increased risk with increased birth weight compared to normal weight (3500-3999 g OR: 2.9, 95% CI: 1.0-8.2, ≥4000 g, OR: 3.7, 95% CI: 1.1-12.4). Other factors explored did not appear to have an association with brain tumour risk.

Of the 12 cases and 25 controls who responded about their willingness to provide a blood or saliva sample to carry out future biological research, all cases and controls said they would agree to provide saliva whilst 89% of cases and 81% of controls would agree to provide a blood sample.

Through this pilot study, we have demonstrated that by working closely as a multidisciplinary team, recruitment of participants diagnosed with brain tumours is feasible as part of a ‘case-control’ design to address aetiological questions, despite the huge challenges facing these young people shortly after diagnosis.

In terms of addressing the aims of the study, we developed successful methods for identifying and recruiting a high proportion of case subjects by exploiting our strong links with local clinicians and research nurses. Control procedures proved more difficult; nonetheless, this pilot study was informative and we propose the following recommendations to facilitate the design and recruitment of future UK case–control investigations involving childhood and young adult cancer:

Close collaboration with primary care and the National Institute for Health Research (NIHR) Comprehensive Clinical Research Network (CCRN), a body which oversees all clinical NHS-based research in England and which supports widening research participation to improve patient benefit across all clinical domains. This will help to optimise recruitment for both cases and controls.

Engagement with the PCRN and National Cancer Research Institute (NCRI) Primary Care Clinical Studies Group (CSG) to identify general practices which are familiar with research studies, the latter a professional group helping to develop major primary care oncology research studies in the UK.

Collaboration with the relevant Childhood Cancer and Leukaemia Group (CCLG) sub-group, e.g. the CNS sub-group, a national group of professionals dedicated to improving the delivery of care for young people with cancer.

Collection of saliva samples for the purposes of molecular or genetic epidemiology.

Both Leeds and Manchester experienced a number of problems relating to recruitment of controls via GP practices. In Leeds, procedures for identifying controls were resource intensive leading to a much lower than anticipated recruitment rate of 31%. The delayed start in Manchester meant adhering to new NHS structures and despite full ethical and Research and Development/Caldicott Guardian approval for the control recruitment protocol and acceptance of the study onto the NIHR/PCRN portfolio, little progress was made in identifying GP practices willing to participate in the research. Exhaustive efforts over a long period of time were made by the Manchester staff to engage with general practices both through the PCRN and directly to practices with little success.

We believe that the new NHS General Medical Services contract for General Practice, which was implemented in 2004 and allocated certain resources to GPs based on how well they manage patient care (the Quality and Outcomes Framework), may have influenced the willingness of GP practices throughout Manchester to participate. We have since taken advice from the national NCRI Primary Care CSG to help develop control recruitment procedures for future research by ensuring that we work closely with the regional CCRN. We are also exploring alternative sources of control subjects such as child health records through our existing links with primary care.

Completion of the questionnaire was a success, with a very small proportion of missing data being reported. The collection of biological samples would be an integral part of future epidemiological research in this field. We explored the possibility of exploiting the national CCLG tumour bank samples in conjunction with case control research projects. The Brain Tumour sub-group of the CCLG was fully supportive and indicated willingness to collaborate with future studies. All tumour and blood samples collected by the CCLG adhere to specific protocols which would be closely mirrored in future studies. Although we did not collect biological material, we did however ask participants about their willingness to provide biological samples and there was a clear consensus in favour of taking either blood or saliva to aid molecular epidemiological research. Participants also stated that saliva samples would be more readily donated than blood, particularly from younger controls.

Although our pilot study had limited statistical power, findings agreed with previous aetiological work in showing an increased risk associated with birthweight greater than 3500 g [ 24 ]. However, our reported non-significant protective association with breastfeeding is contrary to previous findings [ 27 , 28 ] and may have been due to chance. The use of friend controls may have led to a degree of overmatching, although the number of participants selected in this manner (n = 3) is unlikely to have had a major effect on the parameter estimates reported in Table  2 . Control participation in Leeds was also related to deprivation, with higher levels of participation from more affluent areas. This potential participation bias may have contributed to the higher observed rates of breastfeeding, smaller sibling size and mobile phone usage in the control sample. Nevertheless, as we reported the results for a selected range of risk factors from a relatively small feasibility study, odds ratios should be interpreted with due caution and not be taken as evidence or absence of any causal association.

Feedback from participants has provided us with key information with which to revise the study questionnaires and recruitment procedures to ensure that participation rates in future case–control studies can be maximised. It has provided a valuable insight into questionnaire design and recruitment procedures, particularly in terms of overcoming problems associated with the identification of suitable healthy control subjects.

In summary, this pilot has provided us with all the elements necessary to produce a full protocol for a future UK study, including extensive documentation on all aspects of recruiting and approaching case and control subjects and their families. Findings from this pilot will provide essential information for refining the methods for a future large, multi-centre case–control study.

Written informed consent was obtained from the patient or their guardian/parent/next of kin for the publication of this report.

Stiller C, Allen M, Bayne A, Brownbill P, Draper G, Eatock E, Loach M, Vincent T: National Registry of Childhood Tumours, England and Wales, 1981–1990. International Incidence of Childhood Cancer, Vol II. Edited by: Parkin DM, Kramárová E, Draper GJ, Masuyer E, Michaelis J, Neglia J, Qureshi S, Stiller CA. 1998, Lyon: IARC Scientific Publications No 144

Google Scholar  

Birch JM, Alston RD, Kelsey AM, Quinn MJ, Babb P, McNally RJQ: Classification and incidence of cancers in adolescents and young adults in England 1979–1997. Br J Cancer. 2002, 87: 1267-1274. 10.1038/sj.bjc.6600647.

Article   PubMed   CAS   PubMed Central   Google Scholar  

Mulhern RK, Merchant TE, Gajjar A, Reddick WE, Kun LE: Late neurocognitive sequelae in survivors of brain tumours in childhood. Lancet Oncol. 2004, 5: 399-408. 10.1016/S1470-2045(04)01507-4.

Article   PubMed   Google Scholar  

Louis DN, Ohgaki H, Eds: World Health Organization Classification of Tumours of the Central Nervous System. 2007, Lyon: IARC Press, 4

Wrensch M, Minn Y, Chew T, Bondy M, Berger MS: Epidemiology of primary brain tumours: current concepts and review of the literature. Neuro Oncol. 2002, 4: 278-299.

PubMed   PubMed Central   Google Scholar  

Linet MS, Wacholder S, Zahm SH: Interpreting epidemiologic research: lessons from childhood cancer. Pediatrics. 2003, 112: 218-232.

PubMed   Google Scholar  

Baldwin RT, Preston-Martin S: Epidemiology of brain tumours in childhood – a review. Tox Appl Pharmacol. 2004, 199: 118-131. 10.1016/j.taap.2003.12.029.

Article   CAS   Google Scholar  

Little J: Epidemiology of childhood cancer. 1999, Lyon: IARC Scientific publication no. 149

Filippini G, Maisonneuve P, McCredie M, Peris-Bonet R, Modan B, Preston-Martin S, Mueller BA, Holly EA, Cordier S, Choi NW, Little J, Arslan A, Boyle P: Relation of childhood brain tumours to exposure of parents and children to tobacco smoke: the SEARCH international case control study. Int J Cancer. 2002, 100: 206-213. 10.1002/ijc.10465.

Article   PubMed   CAS   Google Scholar  

Cordier S, Monfort C, Filippini G, Preston-Martin S, Lubin F, Mueller BA, Holly EA, Peris-Bonet R, McCredie M, Choi W, Little J, Arslan A: Parental exposure to polycyclic aromatic hydrocarbons and the risk of childhood brain tumours. Am J Epidemiol. 2004, 159: 1109-1116. 10.1093/aje/kwh154.

McKinney PA, Fear NT, Stockton D on behalf of the UK Childhood Cancer Study Investigators: Parental occupation at periconception: findings from the United Kingdom Childhood Cancer Study. Occup Environ Med. 2003, 60: 901-909. 10.1136/oem.60.12.901.

McNally RJ, Cairns DP, Eden OB, Alexander FE, Taylor GM, Kelsey AM, Birch JM: An infectious aetiology for childhood brain tumours? Evidence from space-time clustering and seasonality analyses. Br J Cancer. 2002, 86: 1070-1077. 10.1038/sj.bjc.6600228.

McNally RJ, Alston RD, Eden OB, Kelsey AM, Birch JM: Further clues concerning the aetiology of childhood central nervous system tumours. Eur J Cancer. 2004, 40: 266-272.

Nyari TA, Dickinson HO, Hammal DM, Parker L: Childhood solid tumours in relation to population maxing around the time of birth. Br J Cancer. 2003, 88: 1370-1374. 10.1038/sj.bjc.6600880.

McNally RJQ, Eden TOB, Alexander FE, Kelsey AM, Birch JM: Is there a common aetiology for certain childhood malignancies? Results of cross-space-time clustering analyses. Eur J Cancer. 2005, 41: 2911-2916. 10.1016/j.ejca.2005.04.051.

McNally RJQ, James PW, Picton SV, McKinney PA, van Laar M, Feltbower RG: Space-time clustering of childhood central nervous system tumours in Yorkshire. UK BMC Cancer. 2012, 12: 13-10.1186/1471-2407-12-13.

Istre GR, Conner JS, Broome CV, Hightower A, Hopkins RS: Risk factors for primary invasive Haemophilus influenzae disease: increased risk from day care attendance and school-aged household members. J Pediatr. 1985, 106: 190-195. 10.1016/S0022-3476(85)80285-7.

Schmidt LS, Kamper-Jorgensen M, Schmiegelow K, Johansen C, Lähteenmäki P, Träger C, Stokland T, Grell K, Gustafson G, Kogner P, Sehested A, Schüz J: Infectious exposure in the first years of life and risk of central nervous system tumours in children: analysis of birth order, childcare attendance and seasonality of birth. Br J Cancer. 2010, 102: 1670-1675. 10.1038/sj.bjc.6605679.

Harding NJ, Birch JM, Hepworth SJ, McKinney PA: Infectious exposure in the first year of life and risk of central nervous system tumors in children: analysis of day care, social contact, and overcrowding. Cancer Causes Control. 2009, 20: 129-136. 10.1007/s10552-008-9224-8.

Harding NJ, Birch JM, Hepworth SJ, McKinney PA: Atopic dysfunction and risk of central nervous system tumours in children. Eur J Cancer. 2008, 44: 92-99. 10.1016/j.ejca.2007.10.007.

Roncarolo F, Infante-Rivard C: Asthma and risk of brain cancer in children. Cancer Causes Control. 2012, 23: 617-623. 10.1007/s10552-012-9928-7.

Taylor GM, Dearden S, Ravetto S, Ayres M, Watson P, Hussain A, Greaves M, Alexander F, Eden OB, UKCCS Investigators: Genetic susceptibility to childhood common acute lymphoblastic leukaemia is associated with polymorphic peptide-binding pocket profiles in HLA-DPB1*0201. Hum Mol Genet. 2002, 11: 1585-1597. 10.1093/hmg/11.14.1585.

Tang J, Shao W, Dorak T, Li Y, Miike R, Lobashevsky E, Wiencke JK, Wrensch M, Kaslow RA, Cobbs CS: Positive and Negative Associations of Human Leukocyte Antigen Variants with the Onset and Prognosis of Adult Glioblastoma Multiforme. Cancer Epidemiol Biomarkers Prev. 2005, 14: 2040-2044. 10.1158/1055-9965.EPI-05-0136.

Harder T, Plagemann A, Harder A: Birth weight and subsequent risk of childhood primary brain tumors: a meta-analysis. Am J Epidemiol. 2008, 168: 366-373. 10.1093/aje/kwn144.

Christensen JS, Mortensen LH, Röösli M, Feychting M, Tynes T, Andersen TV, Schmidt LS, Poulsen AH, Aydin D, Kuehni CE, Prochazka M, Lannering B, Klaeboe L, Eggen T, Schüz J: Brain tumors in children and adolescents and exposure to animals and farm life: a multicenter case–control study (CEFALO). Cancer Causes Control. 2012, 23: 1463-1473. 10.1007/s10552-012-0020-0.

Aydin D, Feychting M, Schüz J, Tynes T, Andersen TV, Schmidt LS, Poulsen AH, Johansen C, Prochazka M, Lannering B, Klæboe L, Eggen T, Jenni D, Grotzer M, von der Weid N, Kuehni CE, Röösli M: Mobile phone use and brain tumors in children and adolescents: a multicenter case–control study. J Natl Cancer Inst. 2011, 103: 1264-1276. 10.1093/jnci/djr244.

Harding NJ, Birch JM, Hepworth SJ, McKinney PA: Breastfeeding and risk of childhood CNS tumours. Br J Cancer. 2007, 96: 815-817. 10.1038/sj.bjc.6603638.

Schuz J, Kaletsch U, Kaatsch P, Meinert R, Michaelis J: Risk factors for pediatric tumours of the central nervous system: results from a German population-based case–control study. Med Pediatr Oncol. 2001, 36: 274-282. 10.1002/1096-911X(20010201)36:2<274::AID-MPO1065>3.0.CO;2-D.

Download references

Acknowledgments

This work was supported by the Brain Tumour Trust, previously the Samantha Dickson Brain Tumour Trust, who funded the study (grant number 13/42). We are grateful to all the participants for taking part, as well as the clinicians and GPs who facilitated access to them. We thank Catherine Reynolds for help with data collection and Tom Fleming and Andrew Lee for IT support.

Author information

Authors and affiliations.

Division of Epidemiology and Biostatistics, School of Medicine, University of Leeds, Leeds, LS2 9JT, UK

Richard G Feltbower, Sarah J Fleming, Diana Morgan & Patricia A McKinney

Paediatric Haematology and Oncology, The General Infirmary at Leeds, Martin Wing, Leeds Teaching Hospitals NHS Trust, Leeds, LS1 3EX, UK

Susan V Picton

Cancer Research UK Paediatric and Familial Cancer Research Group, Manchester Academic Health Science Centre, University of Manchester, Manchester, UK

Robert D Alston, Janice Achilles & Jillian M Birch

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Richard G Feltbower .

Additional information

Competing interests.

The authors declare that they have no competing interest.

Authors’ contributions

JB and PM devised the study; DM and JA organised the interviews and gathered the data for the study from participants; SF and RF carried out the statistical analysis; RF drafted the manuscript; all authors provided critical comments and approved the contents of the paper prior to submission. All authors read and approved the final manuscript.

Electronic supplementary material

Additional file 1:family questionnaire.(pdf 812 kb), authors’ original submitted files for images.

Below are the links to the authors’ original submitted files for images.

Authors’ original file for figure 1

Rights and permissions.

This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/2.0 ), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and Permissions

About this article

Cite this article.

Feltbower, R.G., Fleming, S.J., Picton, S.V. et al. UK case control study of brain tumours in children, teenagers and young adults: a pilot study. BMC Res Notes 7 , 14 (2014). https://doi.org/10.1186/1756-0500-7-14

Download citation

Received : 16 April 2013

Accepted : 03 January 2014

Published : 08 January 2014

DOI : https://doi.org/10.1186/1756-0500-7-14

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Brain tumour
• Case–control study

BMC Research Notes

ISSN: 1756-0500

• Submission enquiries: [email protected]
• General enquiries: [email protected]

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 07 March 2017

Study designs may influence results: the problems with questionnaire-based case–control studies on the epidemiology of glioma

• Christoffer Johansen 1 , 2 ,
• Joachim Schüz 3 ,
• Anne-Marie Serena Andreasen 2 &
• Susanne Oksbjerg Dalton 2  

British Journal of Cancer volume  116 ,  pages 841–848 ( 2017 ) Cite this article

3963 Accesses

13 Citations

3 Altmetric

Metrics details

• Cancer epidemiology
• Cancer therapy
• Risk factors

Glioma is a rare brain tumour with a very poor prognosis and the search for modifiable factors is intense. We reviewed the literature concerning risk factors for glioma obtained in case–control designed epidemiological studies in order to discuss the influence of this methodology on the observed results. When reviewing the association between three exposures, medical radiation, exogenous hormone use and allergy, we critically appraised the evidence from both case–control and cohort studies. For medical radiation and hormone replacement therapy (HRT), questionnaire-based case–control studies appeared to show an inverse association, whereas nested case–control and cohort studies showed no association. For allergies, the inverse association was observed irrespective of study design. We recommend that the questionnaire-based case–control design be placed lower in the hierarchy of studies for establishing cause-and-effect for diseases such as glioma. We suggest that a state-of-the-art case–control study should, as a minimum, be accompanied by extensive validation of the exposure assessment methods and the representativeness of the study sample with regard to the exposures of interest. Otherwise, such studies cannot be regarded as ‘hypothesis testing’ but only ‘hypothesis generating’. We consider that this holds true for all questionnaire-based case–control studies on cancer and other chronic diseases, although perhaps not to the same extent for each exposure–outcome combination.

Studies of the aetiology of glioma, the commonest malignant brain tumour, with a very poor prognosis, are urgently needed, specifically to identify modifiable risk factors. The main reason that researchers have used the case–control design as the model of choice for epidemiological studies on the causes of glioma is that it a rare cancer, with an incidence of 4 per 100 000 people (World Standard Population) in Denmark, an incidence typical for a high-income country ( Christensen et al, 2003 ). Furthermore, the design limits the time required to obtain data, the cost is lower than that of more time-consuming designs and a wide range of suspected risk factors can be examined in the same study. In case–control studies, questionnaire data, blood samples and tissue specimens can be obtained from both cases and controls, thereby allowing analysis of both environmental and genetic factors and their interactions.

In questionnaire-based case–control studies, it is anticipated that cases can recall past events with sufficient accuracy. This a priori assumption is somewhat naive in the case of glioma in view of the well-known clinical presentation of the disease. The cancer itself, surgery, radiotherapy, chemotherapy and any combination of treatment may strongly influence the overall cognitive capacity of patients. Some have overt cognitive deficits and may therefore be unable to remember past events or have selective recall. Researchers may have to interview a proxy of the patient, as is often the case in case–control studies of risk factors for glioma.

Finding suitable controls presents another challenge. They must be from the same study population as the cases, as the source may influence reported exposures, and selection may be introduced when potential controls decide whether to participate in a study. If these sources of error are systematically different in terms of the exposures of interest from those in the case group, bias will be present. Bias can be addressed partly with statistical tools; however, they require either some idea of the nature and magnitude of bias from validation studies or assumptions about potential bias in sensitivity analyses. Neither necessarily leads to a satisfactory outcome, especially if the results differ substantially according to the assumptions. Despite these potentially serious limitations of case–control studies, there has been no in-depth debate about situations in which questionnaire-based case–control studies are unlikely to provide reliable results. In some narrative syntheses and meta-analytical reviews, the results of such studies contribute equally to the evidence base, even though application of study quality indicators is recommended when summarising evidence. It is therefore important to consider the level of evidence from case–control studies based solely on differential reconstruction of past exposures as compared with that from prospective studies for investigating glioma, when reconstruction of exposure is hampered by the outcome itself.

In this review, we critically appraise the evidence from both case–control and cohort studies of three risk factors for glioma in humans: medical radiation, exogenous hormone use and allergy. The objective is to provide some insight into the difficulty associated with choosing a study design when studying the risk factors for glioma. We also propose considerations for applying scientific weight to the results of case–control studies in this context.

We searched the Medline–PubMed database on 18 November 2015 using the following search strategy:

Search ((‘Glioma/epidemiology’[Majr] OR (glioma AND epidemiology))) AND ((((‘Risk Factors’[Mesh]) OR ‘Environment and Public Health’[Mesh])) OR (Risk OR exposure OR factor* OR cause*)) Filters: Humans; Meta-Analysis; Review; Systematic Reviews

Search (((((‘Glioma/epidemiology’[Majr] OR (glioma AND epidemiology))) AND ((((‘Risk Factors’[Mesh]) OR ‘Environment and Public Health’[Mesh])) OR (Risk OR exposure OR factor* OR cause*))) AND Humans[Mesh])) AND (((‘Case-Control Studies’[Mesh]) OR ‘Cohort Studies’[Mesh]) AND Humans[Mesh]) Filters: Humans

This search provided 3018 hits. Using the inclusion criteria English language paper, adult glioma, case–control study or cohort study and excluding reviews and/or meta-analyses, overview or commentary, qualitative methodology, children and adolescents, genetic exposures and mortality or survival as the outcome, we identified reports of original studies that included the three selected risk factors for glioma. Our search was intended to be neither comprehensive nor systematic for this review. We are aware that we did not identify some studies, such as those in which the word ‘glioma’ was not in the title, abstract or keywords and those in which none of the three risk factors was mentioned in the title or abstract.

From the selected papers, we extracted the characteristics of the study. We then compared the evidence from studies based on recall by cases and controls with that from studies with either a case–control design, with objective (recall-independent) assessment of exposure or a prospective cohort design.

We selected 30 case–control studies and six cohort studies on the association between glioma and medical radiation ( Preston-Martin et al, 1989 ; Neuberger et al, 1991 ; Schlehofer et al, 1992 ; Ryan et al, 1992 ; Zampieri et al, 1994 ; Ruder et al, 2006 ; Blettner et al, 2007 ; Davis et al, 2011 ), exogenous hormone use ( Huang et al, 2004 ; Hatch et al, 2005 ; Wigertz et al, 2006 ; Silvera et al, 2006 ; Benson et al, 2008 , 2015 ; Felini et al, 2009 ; Michaud et al, 2010 ; Kabat et al, 2011 ; Andersen et al, 2013 ; Anic et al, 2014 ; 2015 ) or allergic diseases ( Cicuttini et al, 1997 ; Schlehofer et al, 1999 , 2011 ; Wiemels et al, 2002 , 2004 , 2009 ; Brenner et al, 2002 ; Schwartzbaum et al, 2003 ; Schoemaker et al, 2006 ; Wigertz et al, 2007 ; Scheurer et al, 2008 ; Berg-Beckhoff et al, 2009 ; Il’yasova et al, 2009 ; 2009 ; McCarthy et al, 2011 ; Calboli et al, 2011 ; Turner et al, 2013 ; Cahoon et al, 2014 ). Table 1 lists the key characteristics of the selected studies.

Medical radiation: information from participants only

Ionising radiation is a long-established human carcinogen. Early cohort studies of patients who received radiation treatment to the scalp to treat tinea capitis or skin haemangioma during childhood had an increased risk for glioma, especially after treatment at an early age (see, e.g., Ron et al, 1988 ). Early case–control studies suggested increased risks for glioma after exposure to dental X-rays or X-rays to the head and neck ( Preston-Martin et al, 1989 ; Neuberger et al, 1991 ; Ryan et al, 1992 ; Schlehofer et al, 1992 ). In contrast, the German part of the Interphone study (a multinational interview-based case–control study on mobile phone use and other risk factors for brain tumours, acoustic neuroma and salivary gland tumours) indicated that exposure to any medical ionising radiation significantly reduced the risk for glioma (OR, 0.63; 95% CI, 0.48–0.83) in a study of 366 glioma patients (of whom 11% reported information on exposure through proxies) and 1538 controls ( Blettner et al, 2007 ). Other research groups have reported a similar protective effect of medical ionising radiation. In a study in Italy in 1984, of 195 cases and hospital controls, in which all information was obtained from proxies, the OR for any diagnostic X-ray was 0.4 (95% CI, 0.1–1.0; Zampieri et al, 1994 ). In two studies in the USA with 798 and 205 cases (proportions of proxies not reported), reduced ORs were found after exposure to full-mouth dental X-rays (OR, 0.75; 95% CI, 0.61–0.92; Ruder et al, 2006 ) and after one or more yearly dental X-rays or three or more full-mouth X-rays (0.60; 95% CI, 0.21–1.73 to 0.70; 95% CI, 0.40–1.21; Davis et al, 2011 ). In personal communications, we have been informed that medical radiation appears to be protective against glioma in the entire Interphone data set and that similar results were obtained in the Gliogene study. Although authors usually appropriately discuss the possibility of chance findings, residual confounding and (more importantly) recall bias, use of proxies and selection bias, data are required to estimate the magnitude and direction of the potential error; otherwise, most of the conclusions remain speculative. Most case–control studies continue to rely on self-reported information, whereas validation from records of medical radiation or dental records should be a minimal quality assurance component of studies. This may be difficult in countries where X-ray machines are available in all hospitals, big or small, and even in some general practices, so that it would be virtually impossible to review all the records for false negatives (that is, examinations not reported by study participants). It should, however, be feasible on a small sample.

Exogenous hormones: self-reported use versus prescription data

Two methods have been used to collect information on exposure in studies of the relation between use of exogenous hormones and glioma: self-reported use and prescription data. In a case–control study in the USA with 619 women with glioma and 650 controls, self-reported use of hormone replacement therapy (HRT) was associated with an OR of 0.56 (95% CI, 0.37–0.84; Felini et al, 2009 ). This result is in line with those of a number of other case–control studies of self-reported use of oral contraceptives or HRT, reported separately ( Huang et al, 2004 ; Hatch et al, 2005 ; Wigertz et al, 2006 ; Anic et al, 2014 ), which did not, however, reach statistical significance. In two case–control studies nested in population-based registries, with data on prescriptions collected prospectively and independently of the study hypothesis, use of HRT did not decrease the risk for glioma, based on 689 cases (OR, 1.14; 95% CI, 0.93–1.40) and 658 cases (OR, 0.9; 95% CI, 0.8–1.1; Benson et al, 2015 and Andersen et al, 2013 ). These results, based on administrative sources, corroborated those of several very large prospective cohort studies with self-reported data on use of oral contraceptives or HRT obtained before diagnosis of a glioma ( Silvera et al, 2006 ; Benson et al, 2008 ; Michaud et al, 2010 ; Kabat et al, 2011 ). Overall, therefore, relying on self-reported information on use of exogenous hormones obtained retrospectively resulted in systematically lower risk estimates than when exposure was measured prospectively or from prescription data, when no convincing reductions in risk were observed.

Allergy: same direction in risk irrespective of study design

The search of an immune factor that may have a role in glioma aetiology has led to studies of several different definitions of outcomes—ranging from self-reported allergic conditions or autoimmune disorders, discharge records of allergic disorders and use of serum IgE levels as a measure of a hyperactive immune system. Several case–control studies showed consistently that self-reported allergic conditions protect against glioma. For example, in the International Adult Brain Tumour Study, with 1178 glioma patients (26% of whom reported through proxies) and 2493 population controls, an OR of 0.59 (95% CI, 0.49–0.71) was found for any self-reported allergy ( Schlehofer et al, 1999 ). Other case–control studies found similarly reduced ORs; these often had substantial proportions of proxy informants: 44% ( Cicuttini et al, 1997 ), 24% ( Wiemels et al, 2002 ), 24% ( Brenner et al, 2002 ), 13% ( Wigertz et al, 2007 ), 4% ( Scheurer et al, 2008 ), 3% ( Berg-Beckhoff et al, 2009 ), 24% ( Wiemels et al, 2009 ) and 17% ( Turner et al, 2013 ); others did not provide information on the proportion of proxies ( Wiemels et al, 2004 ; Schoemaker et al, 2006 ; Il’yasova et al, 2009 ; McCarthy et al, 2011 ). Two Swedish cohorts who self-reported allergies had non-significantly reduced risks for glioma: OR, 0.45 (95% CI, 0.19–1.07) among twins born in 1986–1925 but a nonsignificantly increased risk (OR, 1.09; 95% CI, 0.48–2.48) among twins born in 1926–1958 ( Schwartzbaum et al, 2003 ). In a combined analysis of the two twin cohorts and discharge records of immune-related diseases, including both atopic allergic diseases as well as autoimmune diseases such as diabetes, rheumatoid arthritis and so on, as the exposure measure, the risk was reduced but not significantly (HR, 0.46; 95% CI, 0.14–1.48).

The biological marker immunoglobuline E (IgE) may provide more specificity and reduce bias stemming from self-report. In a case–control study from 2004, both self-reported allergies and IgE levels were reversely associated with gliomas in 258 cases and 289 controls but, as expected, concordance between the two outcomes was not high ( Wiemels et al, 2004 ). In a further study from 2009, both self-reported allergies and IgE levels were reversely associated in 535 cases and 532 controls, but analyses showed that IgE levels obtained in glioma patients were affected by treatment with telozomide, underscoring the need for prospectively collected data ( Wiemels et al, 2009 ). A case–control study nested in the EPIC cohort ( Schlehofer et al, 2011 ) and thus with prospectively collected data on serum IgE levels reported a statistically nonsignificant OR of 0.73 (95% CI, 0.51–1.06) based on 275 cases. Another case–control study, nested in four large cohorts in the USA with 181 cases of glioma, found an almost identical OR of 0.72 (95% CI, 0.51–1.03) for a serum IgE level above normal ( Calboli et al, 2011 ). A cohort study of hospital discharge records of 4.5 million men with a mean 12-year follow-up and 4383 events of glioma showed that any allergy was associated with an HR for glioma of 0.85 (95% CI, 0.72–1.01) with a latency of >2 years and 0.6 (95% CI, 0.4–0.8) with a latency of >10 years ( Cahoon et al, 2014 ). In a meta-analyses of the 14 studies in the international Gliogene case–control study, published after our literature search, with 4533 cases and 4177 controls and <10% proxies, respiratory allergy was associated with an OR of 0.72 (95% CI, 0.58–0.90; Amirian et al, 2016a ).

Imprecisely defined exposures such as allergic disease probably affect the validity of the findings of both case–control and cohort studies. The heterogeneous description of allergy in studies, different levels of detail in self-reporting on individual allergies and use of objective measures of serum IgE levels or discharge records further complicate interpretation of the results. Nevertheless, there is no doubt that most studies of any design, type of measure and size indicate that allergy or a hyperactive immune system, through some as yet unidenfied biological mechanisms might be protective against the development of glioma.

Synthesis of the three examples

In two of our examples, medical radiation and HRT, questionnaire-based case–control studies appeared to show an inverse association, whereas nested case–control and cohort studies showed no association. For allergies, the inverse association is observed irrespective of study design. If the inverse associations with medical radiation and HRT use are spurious, possible explanations are over-reporting by controls, under-reporting by cases or selection bias in relation to the exposure of interest. Over-reporting by controls seems unlikely, unless the time between the reference date (censoring of risk time) and the interview date is long, when controls may incorrectly remember the dates of examinations and report those occurring after censoring of the risk time, as observed in a case–control study on paediatric brain tumours in Germany ( Schüz et al, 2001 ). Selection bias may have some role, as medical radiation and HRT use are more common among more affluent people, while participation as a control is often associated with higher education and income. Underreporting is a concern. It might occur because a patient with the very serious diagnosis of a glioma might view other medical events as less important and could easily be forgotten in an interview. The last finding is curious, because, for environmental exposures, validation studies suggest over-reporting or exaggeration by cases (for instance, in studies on mobile phone use or occupational exposure), perhaps because they try to not miss reporting something they may consider relevant in terms of their cancer diagnosis. (For discussions on bias in case–control studies on brain tumours, see, for example, Vrijheid et al, 2006 , 2009 ).

After 30 years of research, we still do not know much about what causes glioma or protects people from the disease. In the search for causality, many researchers who are systematically evaluating the evidence give more weight to that from cohort studies than from case–control studies (e.g., Cochrane reviews); others go as far as considering case–control studies useful only for hypothesis generating because of their retrospective nature ( Mann, 2003 ). In many systematic reviews and meta-analyses in the peer-reviewed literature; however, there is a tendency to categorise the evidence from case–control studies with evidence derived from prospective cohort studies and to give them equal weight. In studying glioma, we consider it critical that studies based on the recall of patients with a disease that affects the brain and possibly cognition should not be given the same weight as nested case–control studies or cohort studies. In addition to the limitations inherent in questionnaire-based case–control studies on other diseases, the risk for recall bias among cases makes it difficult to draw firm conclusions. Validation studies of recall of exposures by glioma cases and by controls often show that cases recall the past differently from controls ( Vrijheid et al, 2006 , 2009 ). The treatment and even the symptoms that arise before treatment, due to the presence of the tumour, may influence cognitive function, underscoring these objections. In studies of glioma, the widespread acceptance of information obtained from the closest relative—a proxy—adds to the problem of the accuracy of self-reported information. Going back to our examples, would proxies really know about the dental X-rays that the patient had during childhood? Recall bias is an issue not only for the exposure of interest but also for potential confounders in analyses of the exposure–disease relationship, as inaccurately measured confounders obviate appropriate adjustment.

As we have illustrated, studies in which information on exposure is obtained from sources other than memory for both cases and controls and in which the information on outcome is from high quality sources, are more reliable, depending on the completeness and quality of the data that can be obtained.

The cohort design is not free of problems, but it is less vulnerable to methodological errors than case–control studies that rely on the memory of cases and controls. The cohort design is therefore the preferred type for observational studies. Nevertheless, because glioma is a rare event, the case–control design may be the only one possible. During critical appraisal of the evidence derived from such studies, however, quality indicators should be applied, as they should for cohort studies. These quality indicators should address the study population (sampling frame, response rates), exposure measures (ideally showing results from validations), and discussion of potential bias affecting the risk estimation.

The superiority of the cohort design and/or access to data obtained independently of the hypothesis in studies of potential risk factors for cancer have been illustrated by cohort studies of various issues, for example, that abortions increase the risk for breast cancer ( Melbye et al, 1997 ) and that our minds cause cancer ( Johansen, 2012 ). One may say that when studying i.e. low-dose radiation and rare outcomes such as gliomas with complicating problems of recall bias and lack of validation the question cannot be reduced to just choosing cohort studies over case–control studies. Cohort studies may actually not be feasible for evaluation of this exposure. One solution might instead be to extrapolate from cohort studies with greater ranges of exposure like atomic bomb survivors or people exposed to nuclear accidents. Poorly conducted studies give rise to risk, as their outcomes often contribute to public concern and may shift the focus from the relevant to the irrelevant, as for instance in the debate about cancer risks and mobile technologies.

Observational studies on the risk factors for glioma, i.e. reports from the early case–control studies conducted at the University of California at San Francisco (USA; see, e.g., Wrensch et al, 2000 ) and the University of California at Los Angeles (USA; Preston-Martin et al, 1989 ), coordinated by the US National Cancer Institute ( Inskip et al, 2001 ), the first international case–control study ( Schlehofer et al, 1999 ), the Interphone study ( Cardis et al, 2007 ) and probably also the most recent Gliogene case–control study ( Malmer et al, 2007 ), do not provide much evidence on what causes this devastating cancer. Thus, despite all the resources that went into those studies, the results did not provide striking evidence on which to base prevention. Nevertheless, as lifestyle and environmental factors were studied comprehensively, the results may suggest that not many of the usual cancer-causing suspects have an important role in glioma aetiology. This is an important finding to be acknowledged and suggests that for the identification of causes novel ideas are needed. Recent reports on genetic risk factors for glioma suggest that these factors do have a crucial role in the risk pattern ( Amirian et al, 2016b ).

The criteria for causality are the strength of the evidence, consistency across populations, specificity, temporality, dose–response and biological plausibility ( Hill, 1965 ). The temporal criterion should always be addressed in evaluating the evidence, whereas in case–control studies, unless secondary data sources can be used, the information is collected after diagnosis of a disease, that is, the reverse sequence in temporality. Furthermore, there are major problems in self-reporting, as cases are aware of having a fatal disease and may unconsciously change their way of looking at past events. Even physical measurements should be evaluated for the representativeness of contemporary measurements of exposure during the aetiologically relevant period, which might have been decades previously.

On the basis of this review, we recommend that the case–control design be placed lower in the hierarchy of studies for establishing cause-and-effect for diseases such as glioma, which pose challenges for accurate collection of retrospective data. A state-of-the-art case–control study should as a minimum, be accompanied by extensive validation of the exposure assessment methods and the representativeness of the study sample with regard to the exposures of interest. Otherwise, such studies cannot be termed ‘hypothesis testing’ but only ‘hypothesis generating’. We consider that this holds true for all questionnaire-based case–control studies on all cancers and chronic diseases, although perhaps not to the same extent for each exposure–outcome combination. For example, case–control studies clearly linked smoking with lung cancer in the 1950s, prenatal radiation to the fetus with childhood leukemia in the late 1950s/early 1960s, postmenopausal oestrogens with uterine endometrial cancer in the 1960s and diethylstilbestrol with vaginal adenocarcinoma in 1971. Almost all known risk factors for breast cancer were identified in case–control studies and much of the evidence that identified smoking and types of tobacco as the cause of about 50% of bladder cancer was based on case–control studies. However, this list does not include risk factors for glioma and these earlier studies, in some cases, showed risk estimates robust to such a degree that even potential bias could not hamper the associations observed.

We hope that the examples we have provided underscore our points and that our recommendation will be taken into account in ranking the evidence obtained from case–control studies and also in the design of such studies in cancer epidemiology.

Amirian ES, Zhou R, Wrensch MR, Olson SH, Scheurer ME, Il'yasova D, Lachance D, Armstrong GN, McCoy LS, Lau CC, Claus EB, Barnholtz-Sloan JS, Schildkraut J, Ali-Osman F, Sadetzki S, Johansen C, Houlston RS, Jenkins RB, Bernstein JL, Merrell RT, Davis FG, Lai R, Shete S, Amos CI, Melin BS, Bondy ML (2016a) Approaching a scientific consensus on the association between allergies and glioma risk: a report from the glioma international case-control study. Cancer Epidemiol Biomarkers Prev 25 : 282–290.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Amirian ES, Armstrong GN, Zhou R, Lau CC, Claus EB, Barnholtz-Sloan JS, Il'yasova D, Schildkraut J, Ali-Osman F, Sadetzki S, Johansen C, Houlston RS, Jenkins RB, Lachance D, Olson SH, Bernstein JL, Merrell RT, Wrensch MR, Davis FG, Lai R, Shete S, Amos CI, Scheurer ME, Aldape K, Alafuzoff I, Brännström T, Broholm H, Collins P, Giannini C, Rosenblum M, Tihan T, Melin BS, Bondy ML (2016b) The glioma international case-control study: a report from the Genetic Epidemiology of Glioma International Consortium. Am J Epidemiol 183 : 85–91.

PubMed   Google Scholar  

Andersen L, Friis S, Halls J, Ravn P, Gaist D (2013) Hormone replacement therapy and risk of glioma: a nationwide nested case-control study. Cancer Epidemiol 37 : 876–880.

Article   PubMed   Google Scholar  

Anic GM, Madden MH, Nabors B, Olson JJ, LaRocca RV, Thompson ZJ, Pamnani SJ, Forsyth PA, Thompson RC, Egan KM (2014) Reproductive factors and risk of primary brain tumors in women. J Neurooncol 118 : 297–304.

Benson VS, Pirie K, Green J, Casabonne D, Berai V the Million Women Study Collaboration (2008) Lifestyle factors and primary glioma and meningioma tumours in the Million Women Study cohort. Br J Cancer 99 : 185–190.

Benson VS, Kirichek O, Beral V, Green J (2015) Menopausal hormone therapy and central nervous system tumor risk: large UK prospective study and meta-analysis. Int J Cancer 136 : 2369–2377.

Article   CAS   PubMed   Google Scholar  

Berg-Beckhoff G, Schüz J, Blettner M, Münster E, Schlaefer K, Wahrendorf J, Schlehofer B (2009) History of allergic disease and epilepsy and risk of glioma and meningioma (Interphone study group, Germany). Eur J Epidemiol 24 : 433–440.

Blettner M, Schlehofer B, Samkange-Zeeb F, Berg G, Schlafer K, Schüz J (2007) Medical exposure to ionising radiation and the risk of brain tumours: Interphone study group, Germany. Eur J Cancer 43 : 1990–1998.

Brenner AV, Linet MS, Fine HA, Shapiro WR, Selker RG, Black PM, Inskip PD (2002) History of allergies and autoimmune diseases and risk of brain tumors in adults. Int J Cancer 99 : 252–259.

Cahoon EK, Inskip PD, Gridley G, Brenner AV (2014) Immune-related conditions and subsequent risk of brain cancer in a cohort of 4.5 million male US veterans. Br J Cancer 110 : 1825–1833.

Calboli FCF, Cox DG, Buring JE, Gaziano JM, Ma J, Stampfer M, Willett WC, Tworoger SS, Hunter DJ, Camargo-Jr CA, Michaud DS (2011) Prediagnostic plasma IgE levels and risk of adult glioma in four prospective cohort studies. J Natl Cancer Inst 103 : 1588–1595.

Cardis E, Richardson L, Deltour I, Armstrong B, Feychting M, Johansen C, Kilkenny M, McKinney P, Modan B, Sadetzki S, Schüz J, Swerdlow A, Vrijheid M, Auvinen A, Berg G, Blettner M, Bowman J, Brown J, Chetrit A, Christensen HC, Cook A, Hepworth S, Giles G, Hours M, Iavarone I, Jarus-Hakak A, Klaeboe L, Krewski D, Lagorio S, Lönn S, Mann S, McBride M, Muir K, Nadon L, Parent ME, Pearce N, Salminen T, Schoemaker M, Schlehofer B, Siemiatycki J, Taki M, Takebayashi T, Tynes T, van Tongeren M, Vecchia P, Wiart J, Woodward A, Yamaguchi N (2007) The Interphone study: design, epidemiological methods, and description of the study population. Eur J Epidemiol 22 : 647–664.

Christensen HC, Kosteljanetz M, Johansen C (2003) Incidences of gliomas and meningiomas in Denmark 1943 to 1997. Neurosurgery 52 : 1327–1334.

Cicuttini FM, Hurley SF, Forbes A, Donnan GA, Salzberg M, Giles GG, McNeil JJ (1997) Association of adult glioma with medical conditions, family and reproductive history. Int J Cancer 71 : 203–207.

Davis F, Il’yasova D, Rankin K, McCarthy K, McCarthy B, Bigner DD (2011) Medical diagnostic radiation exposures and risk of gliomas. Radiation Res 175 : 790–796.

Felini MJ, Olshan AF, Schroeder JC, Carozza SE, Miike R, Rice T, Wrensch M (2009) Reproductive factors and hormone use and risk of adult gliomas. Cancer Causes Control 20 : 87–96.

Hill AB (1965) The environment and disease: association or causation? Proc R Soc Med 58 : 295–300.

CAS   PubMed   PubMed Central   Google Scholar  

Hatch EE, Linet MS, Zhang J, Fine HA, Shapiro WR, Selker RG, Black PM, Inskip P (2005) Reproductive and hormonal factors and risk of brain tumors in adult females. Int J Cancer 114 : 797–805.

Huang K, Whelan EA, Ruder AM, Ward EM, Deddens JA, Davis-King KE, Carreón T, Waters MA, Butler MA, Calvert GM, Schulte PA, Zivkovich Z, Heineman EF, Mandel JS, Morton RF, Reding DJ, Rosenman KD the Brain Cancer Collaborative Study Group (2004) Reproductive factors and risk of glioma in women. Cancer Epidemiol Biomarkers Prev 13 : 1583–1588.

Il’yasova D, McCarthy B, Marcello J, Schildkrau JM, Moorman PG, Krishnamachari B, Ali-Osman F, Bigner DD, Davis F (2009) Association between glioma and history of allergies, asthma and eczema: a case-control study with three groups of controls. Cancer Epidemiol Biomarkers Prev 18 : 1232–1238.

Article   PubMed   PubMed Central   Google Scholar  

Inskip PD, Tarone RE, Hatch EE, Wilcosky TC, Shapiro WR, Selker RG, Fine HA, Black PM, Loeffler JS, Linet MS (2001) Cellular-telephone use and brain tumors. N Engl J Med 344 : 79–86.

Johansen C (2012) Mind as a risk factor for cancer – some comments. Psychooncology 21 : 922–926.

Kabat GC, Park Y, Hollenbeck AR, Schatzkin A, Rohan TE (2011) Reproductive factors and exogenous hormone use and risk of adult glioma in women in the NIH-AARP Diet and Health study. Int J Cancer 128 : 944–950.

Malmer B, Adatto P, Armstrong G, Barnholtz-Sloan J, Bernstein JL, Claus E, Davis F, Houlston R, Il'yasova D, Jenkins R, Johansen C, Lai R, Lau C, McCarthy B, Nielsen H, Olson SH, Sadetzki S, Shete S, Wiklund F, Wrensch M, Yang P, Bondy M (2007) Gliogene: an international consortium to understand familial glioma. Cancer Epidemiol Biomarkers Prev 16 : 1730–1734.

Mann CJ (2003) Observational research methods. Research design II: cohort, cross sectional, and case-control studies. Emerg Med J 20 : 54–60.

McCarthy BJ, Rankin K, Il’yasova D, Erdal S, Vick N, Ali-Osman F, Bigner DD, Davis F (2011) Assessment of type of allergy and antihistamine use in the development of glioma. Cancer Epidemiol Biomarkers Prev 20 : 370–378.

Melbye M, Wohlfahrt J, Olsen JH, Frisch M, Westergaard T, Helweg-Larsen K, Andersen PK (1997) Induced abortion and the risk of breast cancer. N Engl J Med 336 : 81–85.

Michaud DL, Gallo V, Schlehofer B, Tjønneland A, Olsen A, Overvad K, Dahm CC, Kaaks R, Lukanova A, Boeing H, Schütze M, Trichopoulou D, Bamia C, Kyrozis A, Sacerdote C, Agnoli C, Palli D, Tumino R, Mattiello A, Bueno-de-Mesquita HB, Ros MM, Peeters PH, van Gils CH, Lund E, Bakken K, Gram IT, Barricarte A, Navarro C, Dorronsoro M, Sánchz MJ, Rodríguez L, Duell EJ, Hallmans G, Melin BS, Manjer J, Borgquist S, Khaw K-T, Wareham N, Allen NE, Tsilidis KK, Romieu I, Rinaldi S, Vineis P, Riboli E (2010) Reproductive factors and exogenous hormone use in relation to risk of glioma and meningioma in a large European cohort study. Cancer Epidemiol Biomarkers Prev 19 : 2562–2569.

Neuberger JS, Brownson RC, Morantz RA, Chin TDY (1991) Association of brain cancer with dental x-rays and occupation in Missouri. Cancer Detect Prev 15 : 31–34.

CAS   PubMed   Google Scholar  

Preston-Martin S, Mack W, Henderson BE (1989) Risk factors for gliomas and meningiomas in males in Los Angeles County. Cancer Res 49 : 6137–6143.

Ron E, Modan B, Boice JD Jr (1988) Mortality after radiotherapy for ringworm of the scalp. Am J Epidemiol 127 : 713–725.

Ruder AM, Waters MA, Carreón T, Butler MA, Davis-King KE, Calvert GM, Schulte PA, Ward EM, Connally LB, Lu J, Wall D, Zivkovich Z, Heineman EF, Mandel JS, Morton RF, Reding DJ, Rosenman KD Brain Cancer Collaborative Study Group (2006) The Upper Midwest Health Study: a case-control study of primary intracranial gliomas in farm and rural residents. J Agric Saf Health 12 : 255–274.

Ryan P, Lee MW, North B, McMichael AJ (1992) Amalgam fillings, diagnostic dental x-rays and tumours of the brain and meninges. Oral Oncol Eur J Cancer 28B : 91–95.

Article   CAS   Google Scholar  

Scheurer ME, El-Zein R, Thompson PA, Aldape KD, Levin VA, Gilbert MR, Weinberg JS, Bondy ML (2008) Long-term anti-inflammatory and antihistamine medication use and adult glioma. Cancer Epidemiol Biomarkers Prev 17 : 1277–1281.

Schlehofer B, Blettner M, Becker N, Martinsohn C, Wahrendorf J (1992) Medical risk factors and the development of brain tumors. Cancer 69 : 2541–2547.

Schlehofer B, Blettner M, Preston-Martin S, Niehoff D, Wahrendorf J, Arslan A, Ahlbom A, Choi WN, Giles GG, Howe GR, Little J, Ménégoz F, Ryan P (1999) Role of medical history in brain tumour development. Results from the international adult brain tumour study. Int J Cancer 82 : 155–160.

Schlehofer B, Siegmund B, Linseisen J, Schüz J, Rohrman S, Becker S, Michaud D, Melin B, Bas Bueno-de-Mesquitaeri H, Peeters PHM, Vineis P, Tjønneland A, Olsen A, Overvad K, Romieu I, Boieng H, Aleksandrova K, Trichopoulou A, Bamia C, Lagiou P, Sacerdote C, Palli D, Panico S, Sieri S, Tuminoarricarte A, Borgquist S, Manjer J, Gallo V, Allen NE, Key TJ, Riboli E, Kaaks R, Wahrendorf J (2011) Primary brain tumours and specific serum immunoglobulin E: a case-control study nested in the European Prospective Investigation into Cancer and Nutrition cohort. Allergy 66 : 1434–1441.

Schoemaker MJ, Swerdlow A, Hepworth SJ, McKinney PA, van Tongeren M, Muir KR (2006) History of allergies and risk of glioma in adults. Int J Cancer 119 : 2165–2172.

Schüz J, Kaletsch U, Kaatsch P, Meinert R, Michaelis J (2001) Risk factors for pediatric tumors of the central nervous system: results from a German population-based case-control study. Med Pediatr Oncol 36 : 274–282.

Schwartzbaum J, Jonsson F, Ahlbom A, Preston-Martin S, Lönn S, Söderberg KC, Feychting M (2003) Cohort studies of association between self-reported allergic conditions, immune-related diagnoses and glioma and meningioma risk. Int J Cancer 106 : 423–428.

Silvera SAN, Miller AB, Rohan TE (2006) Hormonal and reproductive factors and risk of glioma: a prospective cohort study. Int J Cancer 118 : 1321–1324.

Turner MC, Krewski D, Armstrong BK, Chetrit A, Giles GG, Hours M, McBride ML, Parent M-E, Sadetzki S, Siemiatycki J, Woodward A, Cardis E (2013) Allergy and brain tumors in the Interphone study: pooled results from Australia, Canada, France, Israel, and New Zealand. Cancer Causes Control 24 : 949–960.

Vrijheid M, Deltour I, Krewski D, Sanchez M, Cardis E (2006) The effects of recall errors and of selection bias in epidemiologic studies of mobile phone use and cancer risk. J Expo Sci Environ Epidemiol 16 : 371–384.

Vrijheid M, Armstrong BK, Bédard D, Brown J, Deltour I, Iavarone I, Krewski D, Lagorio S, Moore S, Richardson L, Giles GG, McBride M, Parent ME, Siemiatycki J, Cardis E (2009) Recall bias in the assessment of exposure to mobile phones. J Expo Sci Environ Epidemiol 19 : 369–381.

Wiemels JL, Wiencke JK, Sison JD, Miike R, McMillan A, Wrensch M (2002) History of allergies among adults with glioma and controls. Int J Cancer 98 : 609–615.

Wiemels JL, Wiencke JK, Patoka J, Moghadassi M, Chew T, McMillan A, Miike R, Barger G, Wrensch M (2004) Reduced immunoglobulin E and allergy among adults with glioma compared with controls. Cancer Res 64 : 8464–8473.

Article   Google Scholar  

Wiemels JL, Wilson D, Pater C, Patoka J, McCoy L, Rice T, Schwarzbaum J, Heimberger A, Sampson JH, Chang S, Prados M, Wiencke JK, Wrensch M (2009) IgE, allergy, and risk of glioma: update from the San Francisco Bay Area Adult Glioma Study in the temozolomide era. Int J Cancer 125 : 680–687.

Wigertz A, Lönn S, Schwartzbaum J, Hall P, Auvinen A, Christensen HC, Johansen C, Klæboe L, Salminen T, Schoemaker MJ, Swerdlow AJ, Tynes T, Feychting M (2007) Allergic conditions and brain tumor risk. Am J Epidemiol 166 : 941–950.

Wigertz A, Lönn S, Mathiesen T, Ahlbom A, Hall P, Feychting M the Swedish Interphone Study Group (2006) Risk of brain tumors associated with exposure to exogenous female sex hormones. Am J Epidemiol 164 : 629–636.

Wrensch M, Miike R, Lee M, Neuhaus J (2000) Are prior head injuries or diagnostic X-rays associated with glioma in adults? The effects of control selection bias. Neuroepidemiology 19 : 234–244.

Zampieri P, Meneghini F, Grigoletto F, Gerosa M, Licata C, Casentini L, Longatti PL, Padoan A, Mingrino S (1994) Risk factors for cerebral glioma in adults: a case-control study in an Italian population. J Neurooncol 19 : 61–67.

Download references

Author information

Authors and affiliations.

Oncology Clinic, Finsen Centre, Rigshospitalet, University of Copenhagen, Blegdamsvej 9, 2100, Copenhagen, Denmark

Christoffer Johansen

Unit of Survivorship, Danish Cancer Society Research Center, Strandboulevarden 49, Copenhagen, 2100, Denmark

Christoffer Johansen, Anne-Marie Serena Andreasen & Susanne Oksbjerg Dalton

Section of Environment and Radiation, International Agency for Research on Cancer, 150 Cours Albert Thomas, Lyon Cedex 08, 69372, France

Joachim Schüz

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Christoffer Johansen .

Ethics declarations

Competing interests.

The authors declare no conflict of interest.

Rights and permissions

This work is licensed under the Creative Commons Attribution-Non-Commercial-Share Alike 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/4.0/

Reprints and Permissions

About this article

Cite this article.

Johansen, C., Schüz, J., Andreasen, AM. et al. Study designs may influence results: the problems with questionnaire-based case–control studies on the epidemiology of glioma. Br J Cancer 116 , 841–848 (2017). https://doi.org/10.1038/bjc.2017.46

Download citation

Received : 22 March 2016

Revised : 30 January 2017

Accepted : 02 February 2017

Published : 07 March 2017

Issue Date : 28 March 2017

DOI : https://doi.org/10.1038/bjc.2017.46

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• case–control study
• medical radiation
• exogenous hormone use

This article is cited by

Association of a novel antisense lncrna tp73-as1 polymorphisms and expression with colorectal cancer susceptibility and prognosis.

• Chunhong Fan

Genes & Genomics (2022)

Searching for causal relationships of glioma: a phenome-wide Mendelian randomisation study

• Charlie N. Saunders
• Alex J. Cornish
• Melissa L. Bondy

British Journal of Cancer (2021)

Testing for causality between systematically identified risk factors and glioma: a Mendelian randomization study

• A. E. Howell
• J. W. Robinson
• K. M. Kurian

BMC Cancer (2020)

Update on the effect of exogenous hormone use on glioma risk in women: a meta-analysis of case-control and cohort studies

• Yu-Long Lan

Journal of Neuro-Oncology (2018)

Impact of atopy on risk of glioma: a Mendelian randomisation study

• Linden Disney-Hogg
• Richard S. Houlston

BMC Medicine (2018)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

• Have your say (Opens in a new tab/window)
• What is ionising radiation?
• Achived - What is ionising radiation?
• What is background radiation?
• What is non-ionising radiation?
• Units for measuring radiation
• Glossary of terms
• Sun protection
• Smart meters
• Mobile phone base stations
• Radioactive waste
• More radiation sources
• Occupational exposure
• Ultraviolet radiation monitoring
• Personal Radiation Monitoring Service
• Australian National Radiation Dose Register (ANRDR)
• The Australian National Radiation Dose Register
• Buyer's guide for sun protection
• Ultraviolet services
• Australian Clinical Dosimetry Service
• Radioanalytical services
• Calibration
• Occupational Radiation Exposure (ORE) for medical facilities
• Practical reference dosimetry course
• Radiation Protection of the Patient
• Radiation safety training courses
• Radiation meter hire - order form
• Non-medical import permit details
• Medical import permit details
• Export permit details
• Our regulatory services
• State & territory regulators
• Regulatory integrity
• International best practice
• Information for licence applicants
• Information for licence holders
• Australian Radiation Incident Register
• Holistic safety
• Radioactive waste disposal and storage
• Security of radioactive material
• Transport of radioactive material
• Publications program
• Radiation Protection Series
• Radiation Health Series
• National Directory for Radiation Protection
• Regulatory Guides
• Regulatory forms
• New radiation equipment
• Radiation literature survey
• Survey of residential radio wave exposure 2022
• Brain tumour study
• National Diagnostic Reference Level Service (NDRLS)
• Wi-fi in schools measurement study
• ARPANSA environmental EME reports
• Base station survey
• Survey of residential power frequency magnetic fields
• Australian national diagnostic reference levels for MDCT - ADULT
• Radiofrequency Electromagnetic Energy and Health: Research Needs (TR 178)
• Australasian Radioanalytical Laboratory Network
• Early notification and assistance conventions
• Nuclear-powered vessel visit planning
• Radiation Emergency Medical Preparedness and Assistance Network
• Australian Radon Action Plan
• ARPANSA EME Program Action Plan 2020-2024
• Electromagnetic Energy Research
• Research Framework
• National collaboration
• National uniformity
• International collaboration
• Organisational structure
• Radiation Health and Safety Advisory Council
• Radiation Health Committee
• Nuclear Safety Committee
• Roles and expectations for advisory bodies
• Audit and Risk Committee
• Our service charter
• Corporate plan
• Planning and accountability
• Regulator Performance Framework
• Work health and safety policy statement
• Annual Report series
• Quarterly Report series
• ARPANSA Research Strategy
• Significant regulatory activities
• ARPANSA Graduate Program
• Applying for a job
• EOI – First Nations representative, Reconciliation Action Plan Working Group (paid)
• Working with ARPANSA
• Access to information
• Child Safe Framework compliance statement
• Executive remuneration reporting 2017-18
• Freedom of information requests
• Gifts and benefit register
• Information Publication Scheme
• Legal services expenditure reporting
• Senate Procedural Order of continuing effect No. 12
• Senate Order on Entity Contracts
• Public Interest Disclosure Scheme
• Reserve Bank of Australia's terms and conditions for ARPANSA
• Privacy Policy
• Quality Policy
• Regulatory Activities Policy
• Work Health and Safety Policy
• Determining breaches of the APS Code of Conduct and deciding sanctions

Mobile phone use and brain tumours in the CERENAT case‑control study

Authored By:

This case-control study investigated the association between mobile phone use and brain tumours in adults. The study compared mobile phone use between 447 cases of brain tumour (253 glioma and 194 meningioma ) that were diagnosed between 2004-2006 and 892 matched controls . There was no statistically significant association between regular phone use and brain tumour. However, the authors found statistically significant positive associations for heavy use, when considering life-long cumulative call duration of more than 896 hours ( odds ratio , OR=2.89; 95% confidence interval , CI 1.41 to 5.93 for glioma and OR=2.57; 95% CI 1.02 to 6.44 for meningioma), and when considering more than 18,360 number of calls for glioma (OR=2.10; 95% CI 1.03 to 4.31). The authors conclude that these results support other findings concerning a possible association between heavy mobile phone use and brain tumours.

Link to PubMed abstract

Published In:

The INTERPHONE project , which is coordinated by the International Agency for Research on Cancer (IARC), is a multi-national series of case-control studies (from 13 different countries including Australia) testing whether using mobile phones is associated with an increased risk of various cancers in the head and neck. The INTERPHONE studies were conducted using the same methods to enable the data to be pooled for analysis.

A pooled analysis of the INTERPHONE studies for malignant brain tumours (glioma and meningioma) showed no overall association ( INTERPHONE Study Group, 2010 (PDF 187kb) . There were suggestions of an association (most pronounced for glioma) in the group representing individuals with the highest cumulative call time. The authors note that limitations of the methodology prevent conclusions of causality being drawn from these observations.

In May 2011 IARC assessed the possible carcinogenicity of radiofrequency electromagnetic fields (RF-EMF). Based on positive associations found in INTERPHONE and some other epidemiological studies between glioma and acoustic neuroma and exposure to RF-EMF from wireless phones (mobile and cordless phones) IARC has classified RF EMF as “possibly carcinogenic to humans” (also known as Group 2B carcinogen) (see June 2011 report ). The classification by IARC does not provide estimates of what risk of cancer might by posed by any given level of exposure to RF fields. An assessment of this and other possible health effects is currently being conducted by the World Health Organization.

Accessibility Links

• Skip to content
• Skip to search IOPscience
• Skip to Journals list
• Accessibility help
• Accessibility Help

Click here to close this panel.

The Society for Radiological Protection (SRP)

As a holder of a Royal Charter, SRP is the leading UK professional society for radiation protection. The Society offers professional development events, access to Chartered status for suitably qualified members and extensive networking opportunities. Membership is available to people interested in radiation protection whether they are UK-based or overseas. Membership provides online access to Journal of Radiological Protection.

http://www.srp-uk.org/

Computed tomography of the head and the risk of brain tumours during childhood and adolescence: results from a case–control study in Japan

Noriko Kojimahara 1,2,3 , Takayasu Yoshitake 4 , Koji Ono 5 , Michiaki Kai 6 , Graham Bynes 2 , Joachim Schüz 2 , Elisabeth Cardis 7 and Ausrele Kesminiene 2

Published 22 September 2020 • © 2020 Society for Radiological Protection. Published on behalf of SRP by IOP Publishing Limited. All rights reserved Journal of Radiological Protection , Volume 40 , Number 4 Citation Noriko Kojimahara et al 2020 J. Radiol. Prot. 40 1010 DOI 10.1088/1361-6498/abacff

Article metrics

3043 Total downloads

Share this article

Author e-mails.

[email protected]

Author affiliations

1 Department of Public Health, School of Medicine, Tokyo Women's Medical University, Tokyo, Japan

2 Section of Environment and Radiation, International Agency for Research on Cancer, Lyon, France

3 Research Support Centre, Shizuoka General Hospital, Shizuoka, Japan

4 Shinbeppu Hospital, Oita, Japan

5 Tokyo Healthcare University, Tokyo, Japan

6 Oita University of Nursing and Health Sciences, Oita, Japan

7 Barcelona Institute for Global Health, Barcelona, Spain

Noriko Kojimahara https://orcid.org/0000-0003-4099-6167

Takayasu Yoshitake https://orcid.org/0000-0002-3036-3692

• Received 8 May 2020
• Accepted 6 August 2020
• Published 22 September 2020

Buy this article in print

To clarify whether medical radiation exposure, especially from head computed tomography (CT), increases the risk of brain tumours in young patients in Japan, which ranks the second highest in the world in the number of paediatric CT examinations following the US. From 2011 to 2015, we performed a case–control study of 120 brain tumour patients and 360 appendicitis patients as controls. Reasons, the number of brain and head CT scans date were available from interviews. A cumulative radiation dose to the brain was calculated as a sum of doses received from head CT scans and from conventional x-rays and estimated using a reference table derived from a literature review of published studies. We performed conditional logistic regression to assess the risk of brain tumours from brain and head CT, and from conventional head x-ray procedures. The case group received on average 1.8 CTs to the brain area and 2.2 CTs to the whole head, with a mean estimated brain dose of 32 ± 13 mGy. The odds ratio for developing a brain tumour from having a brain CT was 0.93 (95% confidence interval: 0.38–1.82). This was hardly altered when adjusting for parental educational history and for other diseases (history of neurological disease and attention-deficit disorder/attention-deficit hyperactivity disorder). Neither whole head CT nor cumulative brain dose to the brain increased the risk of glioma or of all brain tumours. Although this study conducted in Japan, where ranks second in the number of CT scans conducted in the world, did not show an increased risk of brain tumours related to CT scans, it should be taken with caution due to a case–control study with limited sample size.

Export citation and abstract BibTeX RIS

Original content from this work may be used under the terms of the Creative Commons Attribution 3.0 license . Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.

1. Introduction

The development of medical x-ray technology has greatly improved the precision of imaging diagnostics but raised concerns about potential health hazards associated with the increasing use of computed tomography (CT) scans, particularly in children when considering paediatric radiation exposure [ 1 ]. Japan has the highest number of CT machines among Organisation for Economic Co-operation and Development countries, with reported 200 CT scans being conducted for every 1000 individuals per year [ 2 ]. Globally, Japan ranks second after the US in the number of CT scans conducted, which is approximately double of that, for example, in the UK [ 3 ] and Netherlands. Epidemiological studies [ 4 , 5 ] have demonstrated an increase in leukaemia and brain tumours owing to diagnostic CT; however, results are not entirely consistent. Since around 2000, several countries have recommended to apply lower doses for CT scans performed on children than those used for adults because the former are overly sensitive to radiation [ 6 ]. A paediatric CT guideline was published in Japan in 2004 [ 7 ], which recommends the use of a dose reduction filter and a tube setting of 100 mAs or less for children weighing 36 kg or less (approximately under 10 years old). Despite the introduction of these low-dose CT guidelines in several countries, recently, increased morbidities have been reported for leukaemia and some solid tumours in South Korea [ 8 ].

The present study aimed to clarify whether medical radiation exposure, especially that from head CTs, increases the risk of brain tumours in young patients in Japan. We conducted a case–control study in Japan, using data from the Japanese part of the Mobi-Kids international study [ 9 ], and collected data using the same questionnaire. All brain tumours and glioma were included in the study since low grade glioma has been reported as the most frequent outcome in childhood [ 10 ].

2. Materials and methods

2.1. study design and participants.

We enrolled 120 patients with primary brain tumours and 360 patients hospitalised with appendicitis between 2011 and 2015. Inclusion criteria were being aged 10–24 years at the time of diagnosis, which was defined as the reference date, and being an inpatient in the collaborating hospitals in Tokyo metropolitan area. Only patients with pathologically confirmed brain tumours were enrolled. Exclusion criteria included having secondary brain tumours, hereditary disease, and severe mental disorders. Face-to-face interviews for patients aged 18 to 24 years were conducted by trained interviewers during the hospital stay. If the patients were aged 10 to 17 years or severely ill at any age, interviews were performed either with their parents (mothers or fathers) and the patient, or with their parents only. On the main questionnaire, social background, medical history, mobile phone, and Wi-Fi usage, and radiological exposures were collected. In addition, parental and clinical questionnaires were collected from parents and neurosurgeons, respectively. Cases were matched to controls at a 1:3 ratio based on sex and age (age difference within 2.5 years), and date of diagnosis (difference less than 1.5 years). Of these patients, 30 with brain tumours and 120 with appendicitis were enrolled in the Mobi-Kids study to evaluate the association between laterality of brain tumour and mobile phone use. Following the eligibility criteria, we excluded brain tumours which were located in the mid-line (figure 1 ). The detailed protocol of MOBI-Kids was previously published [ 9 ].

Figure 1.  Flowchart for patients' inclusion in the study compared to the eligibility of Mobi-Kids study.

Download figure:

2.2. Exposure assessment

The information on CT and x-ray examination area, age at examination, and reason for examination obtained during interviews, was monitored closely. Data on three exposure variables, including 'Number of brain CT (times)' and 'Number of head CT (times)' were collected. A 2-year lag time was applied to reduce the likelihood that the CT scan was done related to early symptoms of the brain tumour [ 11 ] leaving out the CT scans conducted within 2 years before the date of brain tumour diagnosis (Lag 2). 'Number of brain CT was the sum of CTs that include brain area, while 'Number of head CTs' was the sum of CTs conducted at the brain, neck, dental, whole body (including head and neck), and unknown sites. Third variable 'Cumulative brain dose (mGy)' was estimated using a reference table (table 1 ), based on a review of data published by Pasqual et al [ 12 ], who reported the radiation dose to the brain from conventional x-ray and CT. The literatures given time-age frame [ 13 ] and an estimation of brain dose [ 14 ] were used in a reference table. Mean dose values were estimated for new-borns and other age groups for head CTs and conventional x-rays of the head and neck, including whole body and unknown sites, excluding examinations conducted within the past 2 years. Interviews included also information on the number of dental x-rays, including bite wing, full mouth, and panoramic x-rays in the 5-year age categories, but these were excluded from dose evaluation, given the extremely low dose to the brain from these x-ray examinations [ 15 ].

Table 1.  Reference table for mean brain dose estimation from diagnostic radiation procedures by age and year of examinations [ 12 – 14 ].

NA; not applicable

2.3. Sensitivity analysis

In the present study, the equipment and imaging settings of the CT scans were unknown. Therefore, our brain dose estimates used for sensitivity analyses were based on the National Cancer Institute's dosimetry system for CT (NCICT) ( https://ncidose.cancer.gov/#ncict ) [ 16 ], selecting typical CT scanner models used widely in Japan [ 17 ], namely, TOSHIBA XVISION (used until 1999) and TOSHIBA Aquillion16 (used since 2000). We selected a phantom by sex and age (0 years old at the time of imaging: new-born; 1–2 years old: 1 year old; 3–7 years: 5 years old; 8–12 years: 10 years old; 13–17 years: 15 years old; 18 years and older: adult) for the dose estimation [ 18 ]. The following imaging settings were used as default: tube voltage of 120 kV, rotation time 1, computed tomography dose index vol 6. The absorbed dose to the head was determined using the varying tube currents of 100 mAs, 200 mAs, and 400 mAs for the sensitivity analysis. In addition, we assumed two models: 'Mix 1' determined that the tube current was 400 mAs (the same as that for adults) before the release of the paediatric low dose CT guidelines in Japan in 2004, and 100 mAs, after the release of the guidelines, if the patient was aged 10 years or less at the time of examination. 'Mix 2' was set up as 400 mAs before 1999 and 100 mAs after 2000 if the subject was aged 10 years or less, to allow for the scenario that the dose lowering strategy for paediatric patients might have occurred a few years before the guideline was published.

2.4. Statistical analyses

The odds ratios (ORs) for risk of all brain tumours and histologically confirmed gliomas, and 95% confidence intervals (CIs) were calculated for all diagnostic radiation procedures and head CT scans using the chi-squared test for categorical variables with STATA16 [Stata Corp. 2015. Stata Statistical Software: Release 14. College Station, TX]. Conditional logistic regression was conducted for the main analysis. Using the Power software [ 19 ], we performed a post hoc power calculation. Planning a study with three matched control(s) per case, when the probability of exposure among controls is 0.3 and the correlation coefficient for exposure between matched cases and controls is 0.6, and the true OR for disease in exposed subjects relative to unexposed subjects is 2, we will need to study 198 cases to be able to reject the null hypothesis that this OR equals 1 with probability (power) 0.8. The Type I error probability associated with this test of this null hypothesis is 0.05. The present study was approved by the Tokyo Women's Medical University ethics committee (2394 R5, 8 November 2018).

3.1. Baseline characteristics

Table 2 shows the baseline characteristics of cases and controls 1 year before the reference date. Socioeconomic status (SES) was derived from parental education (mother's or father's education, whichever was higher). Main difference between cases and controls was seen in the parental education: 30% of cases but only 16.4% of controls parental education was high school or less (\, p = 0.001). According to the Mobi-kids questionnaire, the past history of neurological disease, such as migraine, epilepsy, convulsions, and hydrocephalus, was collected. Prevalence of past neurological disease was significantly higher in the case group (20.0%). Among cases, 4.2% had attention-deficit disorder/attention-deficit hyperactivity disorder (ADD/ADHD) while it was only 0.8% among the control group (\, p = 0.034). With regard to medical radiation exposure, 24 (20%) cases and 78 (21%) controls had a history of brain or neck x-rays (\, p = 0.699). 13 (11%) cases and 42 (12%) controls had a history of brain CT more than 1 year before the diagnosis date, with the largest number of scans being 11 and 5, respectively (\, p = 0.964). Finally, 21 (18%) cases and 76 (21%) controls had a history of head CT. More than 80% of patient was responded by themselves in controls, where 59.2% in cases, although guardians, mostly mothers, were obligated to stay at the interviews with patients aged 17 years and younger. Glioma represented 39% of all brain tumours, but the majority of the cases were those of schwannoma, which were excluded in the Mobi-Kids international study.

Table 2.  Baseline characteristic between cases and controls at 1 year before diagnostic date.

3.2. Exposure to brain

Among all brain tumour cases, brain CTs were on average conducted 1.8 ± 2.9 times, with the highest frequency for a single case being 11 times (table 3 ). The corresponding figure was 1.3 ± 0.9 times in the control group, with the highest frequency for a single patient being 5 times. The number of brain CTs among glioma cases was on average 1.0, with no significant difference among the glioma and control groups (\, p = 0.324). Further, the number of head CTs, which was the sum of CTs conducted at the brain, dental, neck, whole body, and unknown sites, was 2.2 ± 0.7 (range 1–11) for the case group, which was higher than that for the control group (1.5 ± 0.1) (range 1–6). Cumulative brain dose from all diagnostic radiation procedures to the head and neck (lagged by 2 years), which was computed using the reference table (table 1 ), was 32 ± 13 mGy (n = 36) and 22 ± 5.5 mGy (n = 13) in the all cases and glioma groups, respectively, as compared to 25 ± 3.0 mGy in the control group. However, there were no significant differences among these groups.

Table 3.  Exposure estimation among exposed patients: numbers of brain and head CT, and radiation dose from CT+ conventional x-rays derived from the reference table (lagged by 2 years).

1) Brain CT only2) Brain CT (1)+ the other head CTs including head, such as neck, whole body, and unknown3) Brain dose from both CT and conventional x-rays to brain, neck, whole body, and unknown area

3.3. Brain tumour risk and medical radiation exposure

Table 4 shows the results of conditional logistic analysis conducted using three exposure measures as explanatory variables. The ORs for developing all brain tumours were 0.93 (95%CI: 0.55–1.58) with brain CTs and 0.97 (95%CI 0.66–1.42) with head CTs. In addition, when the cumulative dose to the brain from all diagnostic radiation procedures was considered, the crude and adjusted OR were not significant for either all brains or gliomas. When the analysis was limited to patients with pathologically confirmed gliomas (n = 47), the number of brain CTs and total number of head CTs (Lag 2) were lower than the exposure for all brain tumours reported in table 3 . Crude and adjusted OR of the number of brain CTs, number of head CTs, and cumulative radiation dose were not significant in the glioma group. Within the case group, one patient had undergone 11 brain CT examinations more than 2 years before the diagnosis, which were conducted to monitor the progress of hydrocephalus since infancy. However, the crude or adjusted OR for all brain tumours and glioma did not change after omitting this patient from analyses.

Table 4.  Risk of all brain and glioma by exposure (lagged by 2 years).

a Adjusted for parental education and history of neurological disease and ADD/ADHD

3.4. Sensitivity analysis

As shown in table 5 , dose to the brain estimated with the NCICT ranged from 18 mGy (18 years and older, with application of 100 mAs) to 100 mGy (new-born, with application of 400 mAs). In table 6 , simulation analysis of exposure from brain CT (Lag 2) according to the NCICT, exposures to brain were relatively higher as compared to the control group. When we estimated them using three different shooting conditions, all crude ORs were not significant for the brain tumour group. When we applied 'Mix 1', brain CT exposure in the case (73 ± 8.5 mGy) and control (54 ± 6.9 mGy) groups showed no significant differences. Further, the exposure doses using 'Mix 2' senario which was low doses were applied approximately 5 years before publishing the Japanese CT guideline for children were lower than those in Mix 1. Even Decreasing of ORs was not detected, even the most optimistic scenario.

Table 5.  Brain dose from head CT estimates with NICICT using different age of phantoms.

Table 6.  Brain dose from brain CT estimated with NCICT for sensitivity analysis.

a Mix1 represented the scenario that was applied 400 mAs before 2004 and 100 mAs after 2005 if the patient was under 10 years old at the exam for the CT shooting condition. b Mix2 represented the scenario that was applied 400 mAs before 1999 and 100 mAs after 2000 if the patient was under 10 years old at the exam for the CT shooting condition.

4. Discussion

This is the first case–control study of brain tumours in children and adolescents following medical radiation exposure in Japanese children. Similar to the sub-analysis of German INTERPHONE study [ 20 ], our results indicated that the radiation exposure from CT and x-ray procedures did not increase the risk of developing brain tumours or its common sub-type, gliomas. To evaluate exposure as accurately as possible, we confirmed the reasons for the CT and conventional x-ray examinations through interviews and eliminated examinations conducted 2 years before the diagnosis date. However, since this study analysed information that was primarily collected for the case–control study on the association between mobile phones and brain tumours, recall bias for medical radiation exposure in the brain tumour group is not considerably influent.

A large UK cohort [ 21 ] regarding paediatric CT scans and the risk of brain tumours using radiology information systems databases, the excess relative risk per mGy was reported 0.023 to 0.016, where incidence rate ratio was 1.24 (95% CI 1.20 to 1.29) in Australia [ 22 ]. Recently in EPICT study of the Netherlands participants aged below 18 years conducted which included 84 patients with brain tumour [ 23 ], the mean cumulative brain dose was approximately the same as observed in our exposed group (39 mGy). The excess relative risk per 100 mGy dose was significantly higher, at 0.86 (95% CI: 0.20–2.22), in the cohort study using 5-year lag. Moreover, a study of 120 000 Koreans aged 19 years or younger reported an elevated risk of leukaemia, with an incidence rate ratio (IRR) of 2.14 [95% CI: 1.86–2.46], and all cancers due to low-dose CT scan conducted more than 2 years before the diagnosis date. Both these studies have no information on medical reasons for conducting the CT scans, and thus, inverse causation cannot be ruled out, given the absence of data on factors such as subjective symptoms of patients [ 8 ].

In our study, the exposure from head CT and x-ray distributed extremely right skewed, such that more than 80% of the participants were not exposed. Further, average dose values in our study were lower than those reported in previous studies [ 8 , 24 – 26 ], shown in table 3 that cumulative brain dose was low as approximately equivalent to one to two brain CT scans. In spite of some unknown CT imaging conditions, strength of this study was focused on detailed interviews on mobile phone history and participants were less concerned about medical radiation exposure what reduced the effects of recall bias.

One of the limitations of our study was an imbalance of cases and controls with regard to parental education, with more case families having lower educational level. This may have influenced the accuracy of recall of the numbers and types of radiological examinations among cases, if one assumes more accurate reporting among those with higher education. However, previous studies have shown that the risk of brain tumours from CT examinations was not affected by parental education, used as a proxy for SES [ 24 , 27 ]. Unfortunately, information on living conditions and economic inequality which are among the leading factors of SES was not collected in the study [ 25 ]. Insufficient adjustment for SES could be therefore another limitation of the study because children living in a less affluent household were reported to be more likely to be susceptible to illness and injuries [ 26 ]. We are conscious that recall bias could lead to differential misclassification in the case–control study, despite that the radiological history was asked during well-structured interview [ 28 ]. Our risk estimates have not changed after adjusting for past history of neurological disease and ADD/ADHD [ 29 ],. In our study, past history of allergies was similar between two groups, although most studies have demonstrated inverse associations with glioma risk [ 30 ].

The small sample size is a major limitation of our study, resulting in low statistical power to detect an association between CT scans and brain tumours. The power was insufficient as we indicated in the method section, since our case sample size was 120 which was approximately 60% of the required number. Another limitation was that the present study did not consider 'retakes,' despite the fact that, before 2000, it was a common practice to conduct multiple scans repeatedly to obtain clear images for infants during examinations [ 4 , 31 ]. Because multiphase CT scans (with contrast material and without) are still performed occasionally and we had no possibility to check this, it may be also that brain CT dose could have been underestimated in the study.

A set of paediatric CT guidelines was published in Japan in 2004, but awareness of the potential health risks from CT scanning varies across medical professions and medical institutions. In the sensitivity analysis conducted in the present study, the application of both scenarios that weighted exposure differently before and after the adoption of the low-dose CT guidelines did not have impact on brain tumour risk. Therefore, compliance to the guidelines needs to be examined in Japan. Further, epidemiological studies using a cohort study design, not subjected to recall bias, and individual dose and uncertainty estimates based on the information collected from radiology departments allow to evaluate more precisely the association between the dose from CT scans and risk of brain tumours [ 32 , 33 ].

5. Conclusion

In this matched case–control study conducted in Japan, we found that brain CT scans were not associated with brain tumours. Diagnostic x-rays are an indispensable medical procedure. Nevertheless, the risk–benefit of such diagnostic techniques should be considered in all medical settings. Given that the exposure per image is 30 mGy or more, it is essential to make every effort to keep exposure to the minimum necessary dose, especially for CT scans in children.

Acknowledgments

We would like to thank all participants and their relatives for their time and collaboration in the Mobi-Kids study. We would like to thank specially Dr Lee Choonsik, Dr Elisa Pasqual, Dr Gemma Castaño- Vinyals, and Dr Isabelle Thierry-Chef for their invaluable support. We would also like to thank Dr Matsushita M and Dr Ikuyo M for their contribution to data management. We would like to thank Editage ( www.editage.com ) for English language editing. Where authors are identified as personnel of the International Agency for Research on Cancer/ World Health Organization, the authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy or views of the International Agency for Research on Cancer/ World Health Organization.

This work was supported by the MOBI-Kids study and the work in this study was obtained from the European Community's Seventh Framework Program under Grant Agreements Number 226873—the MOBI-Kids Project—and 603794—the GERoNiMO project. Japanese participation in MOBI-Kids was supported by the Ministry of Internal Affairs and Communications. The research has received funding from KAKENNHI (18K10112), a Grant-in-Aid for Scientific Research (C) from the Japan Society for the Promotion of Science.

Conflict of interest

The authors declare no potential conflicts of interest.

Author contributions

NK was responsible for the organisation and coordination of collected data in Japan. NK and GB were responsible for the analysis. EC was the chief investigator of the international Mobi-Kids study. YT, KO, and MK supervised issues regarding radiological exposure in Japan. AK and JS revised critically for important intellectual content the analysis and manuscript writing. All authors read and approved the final manuscript.