Mobile Phone Use and Risk of Brain and Other Cancers
Mobile Phone Use and Risk of Brain and Other Cancers
In this large prospective study of middle-aged UK women, use of mobile phones was not associated with an increased risk of glioma, meningioma, total cancer or cancer at 18 other specific sites. We found an increased risk of acoustic neuroma in women who had used a mobile phone for 5 years or longer, with risk increasing with increasing duration of exposure.
Possible carcinogenic effects of non-ionizing radiofrequency electromagnetic fields from handheld mobile phones have been of concern for many years, with their widespread and rapidly increasing use since the late 1990s. Based on estimates of site-specific radiofrequency field dose, interest has focused on risk of tumours of the head and neck, and in particular on those of the brain and cranial nerves, including glioma, meningioma and acoustic neuroma. It has also been suggested that there may be an increase in risk of leukaemia, through exposure of bone marrow, and of malignant melanoma.
In May 2011, an IARC Working Group concluded that there is 'limited evidence in humans' for the carcinogenicity of radiofrequency electromagnetic fields, based on associations between glioma and acoustic neuroma and exposure to these fields from wireless phones. For meningioma and for non-CNS cancers, the IARC Working Group found the available evidence to be 'insufficient to reach a conclusion on the potential association with mobile phone use'. The epidemiological evidence, which has been extensively reviewed, came largely from retrospective case-control studies, notably the INTERPHONE multi-centre study and studies from the Hardell group in Sweden. Potential limitations of studies that collect exposure information retrospectively are well known, and are particularly pertinent for brain tumours, which may impair cognitive functioning and are often rapidly fatal. Some studies used proxy respondents to report the patient's past exposure. The INTERPHONE study of glioma risk, for example, used proxy reports of mobile phone use for 13% of cases. It is not clear how proxies would affect accuracy of exposure information; the Hardell group reported similar results for living (no proxies) and dead (100% proxies) cases of malignant brain tumours.
Consistent with the findings from the only other study with prospective recording of exposure, we found no increase in the risk of glioma in mobile phone users. Combining results from the two prospective studies gives a RR of 0.98 (95% CI 0.83–1.15, P = 0.76) for 10 or more years of use of a mobile phone, inconsistent with the findings from the Hardell group (RR 2.5, 95% CI 1.8–3.3 in mobile phone users of more than 10 years). An increased risk for glioma (RR 1.40, 95% CI 1.03–1.89) in INTERPHONE was seen only in people with the highest decile of reported call time; the lack of a dose-response relationship and the likelihood of recall bias have meant that the authors and others have cautioned against regarding this finding as strong evidence for a causal relationship. Also there has been no observable increase in glioma incidence during the past decade or so.
For meningioma, our results and those from the Danish prospective study show no increase in the risk related to mobile phone use, with a combined RR of 0.97 (95% CI, 0.72–1.32, P = 0.86) for 10 or more years of use. Studies with retrospective reporting of exposure have also found little evidence for increased risk of meningioma in mobile phone users.
In contrast to the findings from the Danish prospective study, we did find a trend of increasing risk for acoustic neuroma with increasing duration of mobile phone use. Acoustic neuroma is rare; there were relatively few incident acoustic neuromas in mobile phone users in either study (96 in our study and 261 in the Danish study), and confidence intervals surrounding each risk estimate are large. Combining results from the two studies gives a summary RR of 1.16 (95% CI 0.75–1.81, P = 0.50) for mobile phone use for at least 10 years. With retrospective reporting of exposure, the INTERPHONE study found little evidence for increased risk of acoustic neuroma in mobile phone users. As in the analyses for glioma, their elevated odds ratio was found for acoustic neuroma only in those in the top decile of reported call time; and, again, no dose-response relationship was seen. The Hardell group reported a relative risk of 2.9 (95% CI 1.6–5.5) for acoustic neuroma associated with the use of mobile phones for more than 10 years. Acoustic neuroma often causes hearing loss: in the INTERPHONE study, 79% of acoustic neuroma patients reported having hearing problems before diagnosis, with 25% having had these symptoms for more than 5 years before diagnosis. Given the media coverage of possible relationships between mobile phone use and brain tumours, it is possible that some of the observed associations are due to differential diagnosis, as long-term mobile phone users may have been selectively investigated for symptoms of hearing loss.
The rapidly increasing prevalence of mobile phone use in our cohort, from 34% in women reporting in 1999 to 79% in those reporting in 2005, is consistent with the steep increase in numbers of mobile phone subscriptions in the UK from the early 1990s to 2003 and mirrors similar increases in the rest of the world. There is, however, little to suggest an increase in the incidence of acoustic neuroma in England between 1998 and 2008 (Figure 2).
We found a raised relative risk for pituitary tumours in ever users of mobile phones vs never users (RR = 1.52, 95% CI 0.99–2.33, P = 0.06), but no evidence for a trend with increasing duration of use. Previous studies of incident pituitary tumours and mobile phone use have found no increase in risk.
We found no evidence for increased incidence of other cancers in relation to use of a mobile phone, including cancers of the head and neck, all cancers, or cancer at 15 other specific sites, including malignant melanoma, leukaemia, multiple myeloma and non-Hodgkin lymphoma. These results are consistent with the limited published data for non-CNS tumours. As found in the Danish cohort, mobile phone users in our study had a slightly lower incidence of lung cancer and all cancer than non-users; in the Million Women Study, mobile phone users were less likely than non-users to be current smokers at baseline, and it is possible that the slightly reduced risk of lung cancer reflects some residual confounding with smoking.
Mobile phone use was also not consistently associated with increased incidence of stroke or of ischaemic heart disease. The analyses of vascular disease risk were included largely for comparison with those for cancer. Although a case report of an indirect (mechanical) association between using a mobile phone and risk of cerebral ischaemia has been published, we are not aware of any substantial hypothetical or reported direct associations between mobile phone use and vascular disease.
The main strengths of this study lie in the prospective collection of information on use of mobile phones, and the inclusion of large numbers who had used mobile phones for more than 5 years, and many for more than 10 years. As a prospective study with individual participant information on amount of mobile phone use and on possible confounders, this study was prone neither to the shortcomings of retrospective reporting of exposure nor to the limitations of the Danish prospective study, which was based on follow-up of subscription holders and had limited adjustment for other risk factors. In previous analyses we have shown associations between height, body mass index and use of hormone therapy for menopause and risk of CNS tumours. Obesity, physical activity, smoking and alcohol consumption are associated with risk of cancers at other sites. Some of these factors are related to reported mobile phone use and so could potentially confound associations between phone use and cancer risk, but we adjusted for these. Thus, with virtually complete follow-up, we were able to compare risks for a wide range of cancer outcomes in users and non-users of mobile phones, in an analysis free from recall bias and adjusted for potential confounding factors.
The main limitation of the study is that mobile phone use was reported at baseline and may have changed subsequently. Almost all women who reported daily use of mobile phones at baseline were still using a mobile phone at least once a week when asked again 8.8 years later. However, some women who reported not using a mobile phone at baseline began use subsequently; and this might dilute our estimates of relative risk towards the null. Our data suggest that, as expected, this problem is likely to be greatest among women who reported their baseline use of mobile phones in 1999 and 2000, before use became widespread; however, excluding these women did not materially alter our results. We did not have details of handedness of phone use, nor information on tumour laterality. Despite the large study size, the numbers of incident intracranial CNS tumours were still relatively small, especially for rarer tumours such as acoustic neuroma.
In conclusion, in this large prospective study we found no increase in the risk of glioma or meningioma, consistent with findings from the only other prospective study. We did find an increase in the risk of acoustic neuroma among those who had used mobile phones for 5 years or longer; but risk for acoustic neuroma in long-term mobile phone users was not significantly increased when our results were combined with those from the only other published prospective study. In relation to previous studies, our results weaken the evidence for an association between mobile phone use and risk of glioma, but leaves open the possibility of an increased risk of acoustic neuroma in long-term users of mobile phones.
Discussion
In this large prospective study of middle-aged UK women, use of mobile phones was not associated with an increased risk of glioma, meningioma, total cancer or cancer at 18 other specific sites. We found an increased risk of acoustic neuroma in women who had used a mobile phone for 5 years or longer, with risk increasing with increasing duration of exposure.
Possible carcinogenic effects of non-ionizing radiofrequency electromagnetic fields from handheld mobile phones have been of concern for many years, with their widespread and rapidly increasing use since the late 1990s. Based on estimates of site-specific radiofrequency field dose, interest has focused on risk of tumours of the head and neck, and in particular on those of the brain and cranial nerves, including glioma, meningioma and acoustic neuroma. It has also been suggested that there may be an increase in risk of leukaemia, through exposure of bone marrow, and of malignant melanoma.
In May 2011, an IARC Working Group concluded that there is 'limited evidence in humans' for the carcinogenicity of radiofrequency electromagnetic fields, based on associations between glioma and acoustic neuroma and exposure to these fields from wireless phones. For meningioma and for non-CNS cancers, the IARC Working Group found the available evidence to be 'insufficient to reach a conclusion on the potential association with mobile phone use'. The epidemiological evidence, which has been extensively reviewed, came largely from retrospective case-control studies, notably the INTERPHONE multi-centre study and studies from the Hardell group in Sweden. Potential limitations of studies that collect exposure information retrospectively are well known, and are particularly pertinent for brain tumours, which may impair cognitive functioning and are often rapidly fatal. Some studies used proxy respondents to report the patient's past exposure. The INTERPHONE study of glioma risk, for example, used proxy reports of mobile phone use for 13% of cases. It is not clear how proxies would affect accuracy of exposure information; the Hardell group reported similar results for living (no proxies) and dead (100% proxies) cases of malignant brain tumours.
Consistent with the findings from the only other study with prospective recording of exposure, we found no increase in the risk of glioma in mobile phone users. Combining results from the two prospective studies gives a RR of 0.98 (95% CI 0.83–1.15, P = 0.76) for 10 or more years of use of a mobile phone, inconsistent with the findings from the Hardell group (RR 2.5, 95% CI 1.8–3.3 in mobile phone users of more than 10 years). An increased risk for glioma (RR 1.40, 95% CI 1.03–1.89) in INTERPHONE was seen only in people with the highest decile of reported call time; the lack of a dose-response relationship and the likelihood of recall bias have meant that the authors and others have cautioned against regarding this finding as strong evidence for a causal relationship. Also there has been no observable increase in glioma incidence during the past decade or so.
For meningioma, our results and those from the Danish prospective study show no increase in the risk related to mobile phone use, with a combined RR of 0.97 (95% CI, 0.72–1.32, P = 0.86) for 10 or more years of use. Studies with retrospective reporting of exposure have also found little evidence for increased risk of meningioma in mobile phone users.
In contrast to the findings from the Danish prospective study, we did find a trend of increasing risk for acoustic neuroma with increasing duration of mobile phone use. Acoustic neuroma is rare; there were relatively few incident acoustic neuromas in mobile phone users in either study (96 in our study and 261 in the Danish study), and confidence intervals surrounding each risk estimate are large. Combining results from the two studies gives a summary RR of 1.16 (95% CI 0.75–1.81, P = 0.50) for mobile phone use for at least 10 years. With retrospective reporting of exposure, the INTERPHONE study found little evidence for increased risk of acoustic neuroma in mobile phone users. As in the analyses for glioma, their elevated odds ratio was found for acoustic neuroma only in those in the top decile of reported call time; and, again, no dose-response relationship was seen. The Hardell group reported a relative risk of 2.9 (95% CI 1.6–5.5) for acoustic neuroma associated with the use of mobile phones for more than 10 years. Acoustic neuroma often causes hearing loss: in the INTERPHONE study, 79% of acoustic neuroma patients reported having hearing problems before diagnosis, with 25% having had these symptoms for more than 5 years before diagnosis. Given the media coverage of possible relationships between mobile phone use and brain tumours, it is possible that some of the observed associations are due to differential diagnosis, as long-term mobile phone users may have been selectively investigated for symptoms of hearing loss.
The rapidly increasing prevalence of mobile phone use in our cohort, from 34% in women reporting in 1999 to 79% in those reporting in 2005, is consistent with the steep increase in numbers of mobile phone subscriptions in the UK from the early 1990s to 2003 and mirrors similar increases in the rest of the world. There is, however, little to suggest an increase in the incidence of acoustic neuroma in England between 1998 and 2008 (Figure 2).
We found a raised relative risk for pituitary tumours in ever users of mobile phones vs never users (RR = 1.52, 95% CI 0.99–2.33, P = 0.06), but no evidence for a trend with increasing duration of use. Previous studies of incident pituitary tumours and mobile phone use have found no increase in risk.
We found no evidence for increased incidence of other cancers in relation to use of a mobile phone, including cancers of the head and neck, all cancers, or cancer at 15 other specific sites, including malignant melanoma, leukaemia, multiple myeloma and non-Hodgkin lymphoma. These results are consistent with the limited published data for non-CNS tumours. As found in the Danish cohort, mobile phone users in our study had a slightly lower incidence of lung cancer and all cancer than non-users; in the Million Women Study, mobile phone users were less likely than non-users to be current smokers at baseline, and it is possible that the slightly reduced risk of lung cancer reflects some residual confounding with smoking.
Mobile phone use was also not consistently associated with increased incidence of stroke or of ischaemic heart disease. The analyses of vascular disease risk were included largely for comparison with those for cancer. Although a case report of an indirect (mechanical) association between using a mobile phone and risk of cerebral ischaemia has been published, we are not aware of any substantial hypothetical or reported direct associations between mobile phone use and vascular disease.
The main strengths of this study lie in the prospective collection of information on use of mobile phones, and the inclusion of large numbers who had used mobile phones for more than 5 years, and many for more than 10 years. As a prospective study with individual participant information on amount of mobile phone use and on possible confounders, this study was prone neither to the shortcomings of retrospective reporting of exposure nor to the limitations of the Danish prospective study, which was based on follow-up of subscription holders and had limited adjustment for other risk factors. In previous analyses we have shown associations between height, body mass index and use of hormone therapy for menopause and risk of CNS tumours. Obesity, physical activity, smoking and alcohol consumption are associated with risk of cancers at other sites. Some of these factors are related to reported mobile phone use and so could potentially confound associations between phone use and cancer risk, but we adjusted for these. Thus, with virtually complete follow-up, we were able to compare risks for a wide range of cancer outcomes in users and non-users of mobile phones, in an analysis free from recall bias and adjusted for potential confounding factors.
The main limitation of the study is that mobile phone use was reported at baseline and may have changed subsequently. Almost all women who reported daily use of mobile phones at baseline were still using a mobile phone at least once a week when asked again 8.8 years later. However, some women who reported not using a mobile phone at baseline began use subsequently; and this might dilute our estimates of relative risk towards the null. Our data suggest that, as expected, this problem is likely to be greatest among women who reported their baseline use of mobile phones in 1999 and 2000, before use became widespread; however, excluding these women did not materially alter our results. We did not have details of handedness of phone use, nor information on tumour laterality. Despite the large study size, the numbers of incident intracranial CNS tumours were still relatively small, especially for rarer tumours such as acoustic neuroma.
In conclusion, in this large prospective study we found no increase in the risk of glioma or meningioma, consistent with findings from the only other prospective study. We did find an increase in the risk of acoustic neuroma among those who had used mobile phones for 5 years or longer; but risk for acoustic neuroma in long-term mobile phone users was not significantly increased when our results were combined with those from the only other published prospective study. In relation to previous studies, our results weaken the evidence for an association between mobile phone use and risk of glioma, but leaves open the possibility of an increased risk of acoustic neuroma in long-term users of mobile phones.
Source...