SUPPLEMENTAY APPENDIX
SUPPLEMENTARY Appendix I – METHODS
a.Selection of Relevant Studies
b.Data Collection
c.Assessment of Methodological Quality
d.Statistical Analyses
1.Main and Subgroup Analyses
2.Case Control Studies Meta-analysis Interpretation
SUPPLEMENTARY Appendix II – RESULTS
a. Funding of Interphone Studies
SUPPLEMENTARY Appendic III –DISCUSSION
a.Overall assessment using Bradford-Hill’s criteria
SUPPLEMENTARY Appendix IV - TABLES
SUPPLEMENTARY Appendix V - FIGURES
SUPPLEMENTARY APPENDIX I : METHODS
Ia. Selection of Relevant Studies
Two of the authors (PK, MP) independently evaluated the articles for eligibility based on the predetermined selection criteria. The full publications of titles and abstracts of search results were scanned to obtain any relevant articles,(Figure 1) followed by full paper review. All discrepancies were resolved after rechecking the source papers and further discussion among both authors, and consultation of the co-author (KP) where needed.
Ib. Data Collection
Relevant data was extracted from the articles about the number of participants, age range of participants, participation rate, country where study was conducted, type of tumour investigated and funding source of the study by two authors (MP,PK). The 22 papers were written by 15 authors. The authors were contacted through email for identifying any inconsistency between their data and the data extracted from their respective papers. The corrected data was sent back for confirmation and the same was confirmed.
Ic. Assessment of Methodological Quality
The quality of studies was assessed using the scores proposed by modified Newcastle-Ottawa Quality Assessment Scale (NOS).[1] The modified NOS grading for case-control or cohort study included three criteria’s: (i) selection, total score: 4; (ii) comparability, total score: 2; and (iii) exposure (case-control)/outcome (cohort), total score: 3. A high score out of a total of nine points indicates high methodological quality. In the present analysis, study awarded 7 or more stars has been considered as a high quality study, however, all levels of quality were included in the analysis in a hierarchical approach.
The quality assessment was done by the two authors (MP, PK) and the disagreements were resolved by discussion. The scores of individual studies have been shown in Supplementary Table IV.
Id .Statistical Analyses
1.Main and Subgroup Analyses
Case control studies
Studies included in the meta-analysis fall into five data streams for case control study
(a) Risk of brain tumour,
(b) Risk of brain tumour with mobile phone use > = 10 years,
(c) Risk of gliomas,
d) Risk of meningiomas,
(e) Risk of acoustic neuromas.
Two of these analyses (Risk of brain tumour, Risk of brain tumour with mobile phone use > = 10 years) were further done in four parts:
· All studies
· Government-funded studies (where sources of funding were mainly government agencies)
· Industry-funded studies (where sources of funding were mainly mobile phone industries)
· Mixed-funded studies
· A metaregression analysis was introduced to explore heterogeneity between studies.
2. Case Control Studies Metaanalysis Interpretation
I-square statistics was interpreted as suggested in ‘Cochrane Handbook for Systematic Reviews of Intervention’ (version 5.1.0) as follows:
· 0% to 40%: might not be important;
· 30% to 60%: may represent moderate heterogeneity;
· 50% to 90%: may represent substantial heterogeneity;
· 75% to 100%: considerable heterogeneity
SUPPLEMENTARY Appendix II: RESULTS
IIa. Funding of Interphone
The Interphone study was undertaken as a collaborative effort between a numbers of partner institutions, co-ordinated by IARC. [2]
The overall funding assigned to the Interphone study amounts to approx. 19.2 million euros (€). Of this amount 5.5 million € were contributed by industry sources.
1. Of these 5.5 million €,
-3.5 million € were contributed by the Mobile Manufacturers' Forum (MMF) and the GSM Association, each contributing half of that amount, through a firewall mechanism provided by the UICC (InternationalUnion against Cancer) to guarantee the independence of the scientists.
- Rest of the 5.5 million € came indirectly to individual centers from mobile phone operators and manufacturers, for example, through taxes and fees collected by government agencies.
- Only 0.5 million € (2.5%) of the overall study costs were provided directly by the industry, in Canada and France, under contracts which preserved the independence of the study.
2. Other funding was provided by the European Commission (3.74 million €) and national and local funding sources (9.9 million € in total) in participating countries.
3. Additional funding for the extension of the research to younger and older age groups was received directly from mobile phone operators in the UK under contracts which preserved the independence of the study.
Sno. / Funding Source / Comments1 / ‘Quality of Life and Management of Living Resources’ / Contract QLK4-CT-1999901563
2 / The International Union against Cancer (UICC), / Mobile Manufacturers’ Forum and GSM Association.
3 / International Agency for Research on Cancer (IARC) / Lyon, France
SUPPLEMENTARY APPENDIX III : DISCUSSION
IIIa.Overall assessment using Bradford-Hill’s criteria
Applying Bradford-Hill’s criteria [3] to assess causation leads to following consideration:
Strength: our meta-analysis shows a statistically significant increased risk of gliomas, particularly after long-term (>10 years) use or after cumulative mobile phone use of more than 1640 hours. In fact, the odds ratio for use of >1640 hours has magnitude of 2.29, which is less than three (the conventional figure to show a ‘strong association’) but it is not realistic to expect mobile phones to triple the risk of gliomas. Given the world-wide exposure of a large population to mobile phones, even a small increase in risk will mean a large population attributable risk.
Consistency: our forest plot of high quality studies reveals reasonably close point estimates and over-lapping confidence intervals. I-squared also ranges from 0 to 14%, confirming that if quality of 7 or 8 is taken into consideration, there is consistency of results across the studies.
Sometimes, experts consider that there is inconsistency between the results of Hardell’s studies and INTERPHONE study, but this is because of the difference in the age criterion. When Hardell et al recalculated their results using the age group 30-59 years (same as in the INTERPHONE study) for glioma, then the results turned out to be similar. [[4]]
Biological gradient: Biological gradient refers to the presence of a dose-response relationship. Several hypotheses can be formulated to test the dose response relationship.
(a) The risk of tumour is more on the side of regular usage (ipsilateral) than on the opposite side (contralateral) of the brain.
(b) This risk increases with duration of exposure. For example, the risk with ten years of usage is more than the risk with less than ten years of usage.
(c) The risk of tumour is more for the part closer to the ear than for the part remote from it. For example, the risk estimate associated with temporal lobe gliomas is higher than gliomas of other lobes of cerebral hemisphere.
(d) The risk increases with dose of exposure [4]
Our analysis supports all of these hypotheses.
Dose response relationship is also seen in terms of higher risk of gliomas on the ipsilateral side compared to the contralateral side. Higher risk of gliomas of temporal lobe as compared to that of other lobes also strengthens the dose response relationship. [5]
Increasing risks with higher dose or duration of exposure reported by Hardell et al strongly supports dose response relationship. OR increased statistically significant for glioma for cumulative use of mobile phones per 100h; OR=1.014, 95% CI=1.008-1.019, and per year of latency; OR=1.056, 95% CI=1.037-1.075. Separate calculations of mobile phone and cordless phone use yielded similar results with statistically significant increasing risks. [5]
Specificity: This criterion refers to existence of association of exposure to certain diseases while there is absence of association with others. The degree of specificity may vary depending on the nature of exposure, it biological effects and diseases under consideration.
Finding of positive association between mobile phone use with gliomas and possibly acoustic neuromas along with finding of lack of such a finding with meningiomas in the same set of studies supports the fulfillment of criterion of specificity.
Biological plausibility: Mobile phones emit radiofrequency electro-magnetic fields (RF-EMFs). RF-EMFs do not have sufficient energy to directly break chemical bonds like ionizing radiation. However, they may impair DNA-repair mechanisms and introduce epi-genetic changes to DNA. Several studies have shown that RF-EMF increase free radical activity within cells. This is probably caused by Fenton reaction. It should also be noted that possible biological effects might not have linear dose response as indicated in some studies [6] and that the effects are depending on the carrier frequencies. [5], [7]
While these and other mechanisms continue to form subjects of study by biology scientists, there is enough evidence from laboratory to support potential of RF-EMFs to cause brain tumours.
Temporality: This criterion refers to the necessity that the cause precedes the effect in time. The finding of OR in those using mobile phones for >10 years clearly supports temporality as those with pre-existing gliomas would not have survived for 10 years. Also, association of ipsilateral acoustic neuroma in Hardell’s studies suggests that the tumour followed rather than preceded mobile phone use. If the tumour had preceded the phone use, then the cases would not have habitually used the ipsilateral side during use of phone.
Coherence: Coherence refers to lack of conflicting information as regards cause and effect interpretation for an association between an exposure and disease. In the context of brain tumours, a study using ecological data in 165 of 208 countries found a consistent association between brain and nervous system cancer incidence rates and penetration of rate of mobile phone subscriptions. [4] Incidence of brain tumour in different countries also provides coherent information into the cause and effect of brain tumours and mobile phone use. [5]
Experimental evidence: According to Hill, experimental evidence meant decreasing a potentially harmful exposure and demonstrating that the frequency of disease declines. Such evidence is impossible to obtain in the context of mobile phone use and brain tumour. However, some experts consider animal experiments demonstrating the adverse effects of exposure as experimental evidence. As indicted under sub-heading biological plausibility, there is substantial though not conclusive evidence that RM-EMF has an effect on rodent cells.
Analogy: Analogy as one of Hill’s criteria refers to examples of similar association between exposure and disease. However, this criterion is at odds with the specificity criterion and therefore is not sine qua non to establish association between mobile phone use and brain tumour.
The data for ipsilateral use and temporal lobe location could not be retrieved from the papers which might be considered as one of the limitation of this metaanalysis.
Supplementary Appendix IV: DATA TABLES
Supplementary Table I. Funding Sources of Case-Control Studies
Case control StudiesS.No / Study ID(sorted by year of publication) / Funding Source
1. / Hardell 1999 [8] / Swedish Medical Research Council and Orebro Cancer Fund
2. / Muscat 2000 [9] / Cellular Industry Telecommunications Association via the Wireless Technology Research
3. / Inskip 2001 [10] / US Tax Payers
4. / Auvien 2002 [11] / Finnish mobile phone manufacturers and network providers
5. / Hardell 2002 [12] / Swedish Work Environment Fund, Cancer-ochAllergifonden, Orebro Cancer Fund and Telia.
6. / Warren 2003 [13] / -
7. / Hardell 2005 [14] / Cancer-ochAllergifonden, Cancerhjalpen, Orebro Cancer Fund and Nyekelfonden
8. / Lonn 2005 [15] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
9. / Schoemaker 2005 [16] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
10. / Hardell 2006 [17] / Cancer-ochAllergifonden, Cancerhjalpen, Orebro Cancer Fund and Nyekelfonden
11. / Schuz 2006 [18] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
12. / Takebayashi 2006 [19] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
13. / Hours 2007 [20] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
14. / Lahkola 2007 [21] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
15. / Schlehofer 2007 [22] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
16. / Lahkola 2008 [23] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
17. / Takebayashi 2008 [24] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
18. / Hardell 2010 [25] / Cancer-ochAllergifonden, Cancerhjalpen, Fondkistan and Orebro University Hospital Cancer Fund
19. / Interphone study group 2010*[26] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
20. / Interphone Study group 2011*[27] / Mobile Manufacturer’s Forum, Global System for Mobile Communication Association
21. / Aydin D 2011 [28] / The Danish Strategic Research Council, The Swedish Council for Working Life and Social Research ,The Swedish Research Council, The Swedish Childhood Cancer Society ,The Research Council of Norway.
22. / Coureau 2015 [29] / The Fondation de France, the Agence Française de Sécurité Sanitaire de l’Environnement et du Travail, theAssociation pour la Recherche contre le Cancer, the Ligue contre le Cancer, theInstitut National de la Santé Et de la Recherche Médicale
—ATC Environnementet Santé
*The detailed funding for Interphone studies mentioned in Supplementary Appendix II
Supplementary Table II. List of Participating Institutions in Interphone Study
LIST OF PARTICIPATING INSTITUTIONS[26], [27]1 / Australia Centre
-Australian National Health and Medical Research Council
- Bruce Armstrong was supported by a University of Sydney Medical Foundation Program Grant and
- Julianne Brown by an Australian Postgraduate Award -Cancer Council NSW and The Cancer Council Victoria / (EME Grant 219129)
2 / Canada -Canadian Institutes of Health Research
- Partial support from the Canadian Wireless Telecommunications Association.
-DrSiemiatycki’s research team was partly funded by the Canada Research Chair programme and by the Guzzo-CRS Chair in Environment and Cancer. -Dr Parent had a salary award from the Fonds de la recherche en sante´ du Que´bec.
-University–industry partnership grant from the Canadian Institutes of Health Research (CIHR). / (project MOP-42525)
3 / Denmark
- Danish Cancer Society
4 / Finland - The Finnish centre by the Emil Aaltonen Foundation and the Academy of Finland.