Use of Mobile and Cordless Phones and Survival of

Original Paper Neuroepidemiology 2013;40:101–108 DOI: 10.1159/000341905

Received: June 15, 2012 Accepted: July 15, 2012 Published online: October 24, 2012

Use of Mobile and Cordless Phones and Survival of Patients with Glioma Lennart Hardell Michael Carlberg Department of Oncology, University Hospital, Örebro, Sweden

Key Words Mobile phones ⴢ Cordless phones ⴢ Glioma ⴢ Survival

Abstract Background: We analysed the survival of patients after glioma diagnosis in relation to the use of wireless phones. Methods: All cases diagnosed between 1997 and 2003 with a malignant brain tumour (n = 1,251) in our case-control studies were included and followed from the date of diagnosis to the date of death or until May 30, 2012. Results: For glioma, the use of wireless phones (mobile and cordless phones) gave a hazard ratio (HR) = 1.1 (95% confidence interval, CI = 0.9–1.2), with 110-year latency HR = 1.2 (95% CI = 1.002–1.5, p trend = 0.02). For astrocytoma grade I-II (lowgrade), the results were, HR = 0.5 (95% CI = 0.3–0.9) and for astrocytoma grade IV (glioblastoma), HR = 1.1 (95% CI = 0.95– 1.4), with 110 year latency HR = 1.3 (95% CI = 1.03–1.7). In the highest tertile (1 426 h) of cumulative use, HR = 1.2 (95% CI = 0.95–1.5) was found for glioblastoma. The results were similar for mobile and cordless phones. Conclusions: Decreased survival of glioma cases with long-term and high cumulative use of wireless phones was found. A survival disadvantage for astrocytoma grade IV, but a survival benefit for astrocytoma grade I-II was observed which could be due to exposure-related tumour symptoms leading to earlier diagnosis and surgery in that patient group.

Introduction

The use of mobile phones has increased rapidly since the early 1980s. The Nordic countries were among the first to adopt this technology, so studies on long-term health effects are especially worthwhile in these countries. Analogue phones were used at first, but digital phones were introduced in the early 1990s and now dominate the market. The analogue phone system was closed down in Sweden in 2007. Another type of wireless phone is the cordless desktop phone. These have been in use in Sweden since the late 1980s and are very common, overtaking telephones connected to landlines. Worldwide, an estimate of 5.9 billion mobile phone subscriptions has been reported at the end of 2011 by the International Telecommunication Union (http://www.itu.int/ITU-D/ict/facts/2011/material/ ICTFactsFigures2011.pdf). Concern over adverse health effects has been increasingly discussed with the expansion of this technology. The brain is an organ for nearfield exposure to radiofrequency electromagnetic fields (RF-EMF) during the use of wireless phones. Even a small increase in the risk for brain tumours would have a major impact on public health due to the widespread use of wireless phones. In previous case-control studies, we have reported an increased risk for glioma and acoustic neuroma associ-

Copyright © 2012 S. Karger AG, Basel © 2012 S. Karger AG, Basel 0251–5350/13/0402–0101$38.00/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/ned

Dr. Lennart Hardell Department of Oncology University Hospital SE–70185 Örebro (Sweden) Tel. +46 19 60 21 000, E-Mail lennart.hardell @ orebroll.se

ated with the use of both mobile and cordless phones [1– 4]. The risk was especially high for ipsilateral use, i.e. the tumour occurred on the same side of the brain that was predominantly used for wireless phone calls. However, we found no consistent pattern of increased risk for meningioma [2]. In May 2011, the International Agency for Research on Cancer concluded that RF-EMF emissions overall, e.g. occupational and wireless phone use, are possibly carcinogenic to humans [5]. The mechanism for a possible carcinogenic effect from RF-EMF emissions is unclear. However, some studies indicate that DNA might be damaged by free radicals that are formed inside the cell during exposure, probably via the Fenton reaction, and might thus lead to an increased risk for cancer [6, 7]. If exposure to RF-EMF is carcinogenic, then it might also have a correlation with survival among exposed cancer patients. To further elucidate that possibility we analysed survival among all cases with malignant brain tumours (n = 1,251) in our case-control studies [1, 3, 8].

Materials and Methods Detailed information on materials and methods has been given in our previous publications [1, 3, 8]. In short, our studies on this topic were of the case-control type. The cases were reported to us from the regional cancer registries in Sweden. Matched controls were selected from the same geographical areas using the Swedish population registry. The regional ethical committees approved all studies. The study included men and women with a histopathologically verified diagnosis of brain tumour between 1997 and 2003 and aged 20–80 years at the time of the diagnosis. The geographical study areas differed somewhat in two time periods. From January 1, 1997 until June 30, 2000, it covered Uppsala-Örebro, Stockholm, Linköping and Göteborg medical areas in Sweden whereas from July 1, 2000 until December 31, 2003 only UppsalaÖrebro and Linköping regions were included. The cases were enrolled after we had received copies of the cancer reports to the regional cancer registries that covered the study areas. One living control person matched on age and sex was drawn to each living case from the Swedish population registry. However, this publication includes the cases only, all with histopathological confirmation of a malignant brain tumour. There was a delay of some weeks or months before the patients could be included in our studies due to a delay in the reporting to the cancer registry from the different clinics. Astrocytoma is the most common type of glioma in adults. The prognosis of high-grade astrocytoma (especially WHO grade IV, glioblastoma) is poor. Thus, a number of patients had died before we could interview them. All of those deceased patients with a malignant brain tumour diagnosed between 1997 and 2003 were instead included in a later study [8]. Persons who had died from other diseases were used as controls. Relatives of both cases and

102

Neuroepidemiology 2013;40:101–108

controls were identified through the Swedish population registry at the Swedish Tax Agency. Tumour localisation was based on information in medical records, i.e. MRI/CT scans, and all tumour types were defined by using histopathology reports. Exposures were assessed by a mailed questionnaire that was sent to the living cases and their controls or to the next-of-kin of the deceased cases and controls. The information was supplemented over the phone by a trained interviewer who did not know whether it was a case or a control that was being investigated. Regarding the use of wireless phones, detailed questions were asked on the following: type, time period, average number of minutes per day over the years, ear mostly used during calls (not for deceased subjects), use of hands-free devices and use of external antenna in a car. Only exposure before the date of tumour diagnosis was assessed thereby using a minimum latency period of 1 year. Thus, exposure starting ^1 year before diagnosis was disregarded. These items were discussed in more detail in the different publications [1, 3, 8]. The aim of this study was to analyse the survival of patients with a malignant brain tumour in relation to the use of wireless phones, i.e. mobile and cordless phones. Survival was followed using the Swedish population registry. That registry covers the whole population with a unique ID number for all inhabitants. Thereby, it is possible to check vital status and, if deceased, the date of death. All cases of both genders diagnosed between 1997 and 2003 and aged 20–80 years who participated in our previous case-control studies were included. All malignant brain tumours were classified according to the WHO classification [9]. Statistical Methods The Cox proportional hazards model was used to calculate hazard ratios (HR) and corresponding 95% confidence intervals (CI). Follow-up time was counted from the date of diagnosis to the date of death or until May 30, 2012 (living cases). Adjustment was made for age (as a continuous variable), gender, year of diagnosis, socioeconomic code and study (material with living cases interviewed and material with next-of-kin interviewed). The proportional hazards assumption was tested using Schoenfeld residuals. A statistically significant violation of the proportionality assumption was detected for age; therefore age was also adjusted for as a time-dependent covariate. The exposed cases were divided according to phone type; analogue, digital and cordless. The use of analogue and digital phones was analysed combined (i.e. mobile phone) and results for all phone types combined (wireless phone) are also presented. Note that some cases could have used both mobile and cordless phones. The unexposed category consisted of cases that reported no use of wireless phones or only exposure ^1 year before diagnosis. Latency (tumour induction period) was analysed using three time periods, 11–5 years, 15–10 years and 110 years since the first use of a mobile or a cordless phone until diagnosis. Tertiles of cumulative lifetime use in hours among the controls were used as the cut-off in the dose-response calculations. They were divided as follows: wireless phone in first tertile 1–91 h, second tertile 92–426 h and third tertile 1 426 h; mobile phone in first tertile 1–36 h, second tertile 37–183 h and third tertile 1183 h, and cordless phone in first tertile 1–122 h, second tertile 123–456 h and third tertile 1 456 h. Lifetime use in hours was also divided in three groups, 1–1,000 h, 1,001–2,000 h and 12,000 h, to further explore the dose-response relationships.

Hardell/Carlberg

Table 1. Age distributions for all and for exposed/unexposed to wireless phones

Exposed to wireless phone

All malignant Glioma Astrocytoma Grade I-II Grade III-IV Grade IIIa Grade IVa Other gliomab Other malignantc

Unexposed to wireless phone

All

n

mean

median min

max

n

mean median min

max

n

mean median min max

719 662 546 92 454 98 355 116 57

51 52 53 42 55 48 57 46 47

54 54 55 40 57 49 59 46 50

80 80 80 69 80 80 79 73 76

514 470 392 39 353 53 298 78 44

60 60 62 48 64 60 64 52 53

80 80 80 71 80 77 80 79 79

1,233 1,132 938 131 807 151 653 194 101

55 55 57 44 59 52 60 49 50

20 20 20 20 20 20 21 20 20

63 63 64 52 65 63 65 56 61

20 21 23 23 27 27 29 21 20

57 58 59 42 60 54 61 50 52

20 20 20 20 20 20 21 20 20

80 80 80 71 80 80 80 79 79

a Information regarding grade III or IV missing for 3 cases. b Oligodendroglioma, other/mixed glioma. c Medulloblastoma, ependymoma, other.

Table 2. Hazard ratios from the Cox proportional model for glioma (n = 1,132; unexposed to wireless phones, n = 470) and use of mo-

bile and cordless phones with adjustments for age, gender, year of diagnosis, socioeconomic code and study, with age as a time-dependent covariate Wireless phone

Total Latencya >1–5 years >5–10 years >10 years Hoursa 1–1,000 h 1,001–2,000 h >2,000 h First tertile Second tertile Third tertile

Mobile phone

exposed, n

HR

95% CI

exposed, n HR

662

1.1

0.9–1.2

524

1.1

268 247 147

0.9 1.1 1.2

0.8–1.1 0.96–1.4 1.002–1.5

247 156 121

475 73 114 182 200 280

1.0 1.1 1.3 0.9 1.1 1.1

0.9–1.2 0.8–1.4 1.04–1.7 0.8–1.1 0.9–1.4 0.9–1.3

425 43 56 152 158 214

Cordless phone 95% CI

exposed, n

HR

95% CI

0.9–1.2

398

1.0

0.8–1.2

0.9 1.1 1.3

0.8–1.1 0.9–1.4 1.0005–1.6

205 150 43

0.9 1.1 1.3

0.7–1.1 0.9–1.3 0.9–1.9

1.0 1.6 1.2 0.9 1.0 1.3

0.9–1.2 1.1–2.2 0.9–1.6 0.7–1.1 0.8–1.3 1.05–1.6

295 48 55 116 110 172

0.9 1.0 1.4 0.9 0.9 1.2

0.8–1.1 0.7–1.4 1.03–1.9 0.7–1.1 0.7–1.1 0.97–1.5

a p, trend: Wireless phone: latency, p = 0.02; hours (1–1,000, 1,001–2,000, >2,000), p = 0.06; hours (tertiles), p = 0.15. Mobile phone: latency, p = 0.04; hours (1–1,000, 1,001–2,000, >2,000), p = 0.03; hours (tertiles), p = 0.01. Cordless phone: latency, p = 0.07; hours (1– 1,000, 1,001–2,000, >2,000), p = 0.03; hours (tertiles), p = 0.01.

In total, 1,251 cases (85%) with a malignant brain tumour participated in our studies [3]. Of these, 18 were excluded from this analysis since their brain tumours were diagnosed at autopsy; survival 0 days. Table 1 gives numbers, age distributions and tumour types for the remaining 1,233 cases. Most of them were diagnosed with glioma (n = 1,132, 92%). Only 129 glioma cases (11%) were

alive on May 30, 2012 (glioblastoma, n = 3 alive). The median survival for glioma was 393 days (glioblastoma, 272 days) and the median duration of follow-up for glioma was 4,703 days (range 3,088–5,628 days). As can be seen in table 1, the median age for the exposed cases with glioma was 9 years lower than for the unexposed cases. In total, the HR was close to unity for glioma (table 2). In the 110-year latency group, an elevated HR for overall survival for all patients with glioma was found for wire-

Wireless Phone Use and Glioma Survival


Results

103

Table 3. Hazard ratios from the Cox proportional hazards model for astrocytoma grade I-II (low-grade, n = 131; unexposed to wireless

phones, n = 39) and use of mobile and cordless phones with adjustments for age, gender, year of diagnosis, socioeconomic code and study, with age as a time-dependent covariate Wireless phone


Mobile phone

Cordless phone

exposed, n HR

95% CI

exposed, n HR

95% CI

exposed, n HR

95% CI

92

0.5

0.3–0.9

68

0.5

0.3–0.9

58

0.5

0.3–0.99

45 35 12

0.5 0.6 0.6

0.3–0.9 0.3–1.1 0.3–1.6

42 18 8

0.5 0.5 0.5

0.3–1.01 0.2–1.1 0.2–1.5

32 21 5

0.5 0.6 0.8

0.3–1.03 0.3–1.1 0.2–2.5

67 13 12 26 26 40

0.5 0.7 0.7 0.3 0.8 0.7

0.3–0.9 0.3–1.7 0.3–1.6 0.1–0.6 0.4–1.6 0.4–1.3

58 6 4 23 22 23

0.5 0.9 0.5 0.4 0.5 0.9

0.3–0.9 0.3–3.1 0.1–2.1 0.2–0.9 0.2–1.03 0.4–1.9

43 9 6 15 18 25

0.5 0.9 0.4 0.2 0.8 0.7

0.3–0.98 0.3–2.4 0.1–1.4 0.1–0.6 0.4–1.7 0.3–1.3

a

p, trend: Wireless phone: latency, p = 0.83; hours (1–1,000, 1,001–2,000, >2,000), p = 0.55; hours (tertiles), p = 0.02. Mobile phone: latency, p = 0.95; hours (1–1,000, 1,001–2,000, >2,000), p = 0.55; hours (tertiles), p = 0.22. Cordless phone: latency, p = 0.78; hours (1– 1,000, 1,001–2,000, >2,000), p = 0.51; hours (tertiles), p = 0.054.

less phones, HR = 1.2 (95% CI = 1.002–1.5, p trend = 0.02), mobile phone HR = 1.3 (95% CI = 1.0005–1.6, p trend = 0.04) and cordless phone HR = 1.3 (95% CI = 0.9–1.9, p trend = 0.07). The cumulative use of cordless phones 12,000 h yielded HR = 1.4 (95% CI = 1.03–1.9, p trend = 0.03) and a statistically significant increasing HR was also found for mobile phone use (p trend = 0.03). Similar results were found using tertiles for cumulative use with statistically significant trends for both mobile and cordless phone use. Wireless phone use gave a decreased HR = 0.5 (95% CI = 0.3–0.9) for low-grade astrocytoma (WHO grade I-II; table 3). Similar results were found for both mobile and cordless phones. These results were not affected by latency, showing no statistically significant trends. Regarding the cumulative number of hours for use of wireless phones, the lowest ratio was found in the category 1–1,000 h yielding HR = 0.5 (95% CI = 0.3–0.9), increasing to HR = 0.7 (95% CI = 0.3–1.6, p trend = 0.55) for the cumulative use of 12,000 h. A similar pattern was found using tertiles of cumulative use (p trend = 0.02). Regarding the separate analysis of mobile and cordless phone use, a statistically significant decreased HR was found in the lowest category of cumulative use. In the third tertile, mobile phone use yielded HR = 0.9 (95% CI = 0.4–1.9, p trend = 0.22) and cordless phone use HR = 0.7 (95% CI = 0.3–1.3, p trend = 0.054). No separate calculations 104


were made for astrocytoma grade I, since few cases had that type of tumour (n = 23). Analysis of astrocytoma WHO grade II yielded similar results to those for the whole group with WHO grade I-II (data not in table). In table 4, results are given for anaplastic astrocytoma (WHO grade III). Overall, there was a slightly decreased HR for both mobile and cordless phones, although not statistically significant. There was no statistically significant increasing or decreasing trend for latency or cumulative number of hours of use. However, in the highest category of cumulative use, 12,000 h, an increased HR was found for wireless phone use, HR = 1.2 (95% CI = 0.6–2.5), although with no statistically significant trend (p = 0.27). The use of wireless phones in total gave HR = 1.1 (95% CI = 0.95–1.4) for astrocytoma WHO grade IV (glioblastoma), increasing with a latency period of 110 years to HR = 1.3 (95% CI = 1.03–1.7, p trend = 0.25; table 5). Also, for the use of mobile and cordless phones, an elevated HR was found in the 110 year latency group, HR = 1.3 (95% CI = 0.9–1.7, p trend = 0.37) and HR = 1.8 (95% CI = 1.2– 2.8, p trend = 0.04), respectively. Regarding cumulative use, no clear pattern of an association was found although a somewhat higher HR was found in the highest tertile. We made separate analyses for other types of glioma (n = 194) and other malignant brain tumours (n = 101) but found no statistically significant increase or decrease Hardell/Carlberg

Table 4. Hazard ratios from the Cox proportional hazards model for astrocytoma grade III (anaplastic astrocytoma, n = 151; unexposed to wireless phones, n = 53) and use of mobile and cordless phones with adjustments for age, gender, year of diagnosis, socioeconomic code and study, with age as a time-dependent covariate

Wireless phone


Mobile phone

Cordless phone

exposed, n HR

95% CI

exposed, n HR

95% CI

exposed, n HR

95% CI

98

0.8

0.5–1.2

82

0.7

0.4–1.1

58

0.7

0.4–1.2

40 35 23

0.8 0.8 0.8

0.5–1.3 0.5–1.5 0.4–1.4

40 25 17

0.7 0.7 0.7

0.4–1.2 0.4–1.5 0.3–1.5

28 23 7

0.7 0.7 0.5

0.4–1.3 0.4–1.4 0.2–1.6

66 14 18 27 32 39

0.8 0.7 1.2 1.1 0.6 0.6

0.5–1.2 0.3–1.5 0.6–2.5 0.6–1.9 0.4–1.1 0.3–1.1

62 9 11 25 22 35

0.7 1.3 1.2 0.6 0.5 1.0

0.4–1.1 0.5–3.1 0.5–2.8 0.3–1.1 0.3–1.1 0.6–1.9

45 7 6 17 20 21

0.8 0.3 1.7 0.8 0.7 0.5

0.4–1.3 0.1–0.9 0.5–6.1 0.4–1.5 0.4–1.4 0.2–1.04

a

p, trend: Wireless phone: latency, p = 0.97; hours (1–1,000, 1,001–2,000, >2,000), p = 0.27; hours (tertiles), p = 0.13. Mobile phone: latency, p = 0.97; hours (1–1,000, 1,001–2,000, >2,000), p = 0.13; hours (tertiles), p = 0.10. Cordless phone: latency, p = 0.89; hours (1– 1,000, 1,001–2,000, >2,000), p = 0.07; hours (tertiles), p = 0.47.

Table 5. Hazard ratios from the Cox proportional hazards model for astrocytoma grade IV (glioblastoma, n = 653; unexposed to wireless phones, n = 298) and use of mobile and cordless phones with adjustments for age, gender, year of diagnosis, socioeconomic code and study, with age as a time-dependent covariate

Wireless phone


Mobile phone

Cordless phone

exposed, n HR

95% CI

exposed, n HR

95% CI

exposed, n HR

95% CI

355

1.1

0.95–1.4

285

1.1

0.9–1.4

212

1.1

0.9–1.4

120 139 96

1.1 1.1 1.3

0.9–1.4 0.9–1.4 1.03–1.7

112 89 84

1.2 1.0 1.3

0.9–1.5 0.8–1.3 0.9–1.7

97 88 27

1.1 1.1 1.8

0.8–1.4 0.8–1.4 1.2–2.8

245 39 71 85 113 157

1.2 1.0 1.1 1.1 1.1 1.2

0.97–1.4 0.7–1.4 0.8–1.4 0.9–1.5 0.9–1.4 0.95–1.5

225 26 34 67 90 128

1.1 1.1 1.1 1.1 1.1 1.2

0.9–1.4 0.7–1.7 0.8–1.7 0.8–1.5 0.8–1.4 0.9–1.5

149 27 36 62 47 103

1.1 1.3 1.2 1.0 1.0 1.2

0.9–1.3 0.9–2.0 0.8–1.7 0.7–1.4 0.7–1.4 0.9–1.6

* p, trend: Wireless phone: latency, p = 0.25; hours (1–1,000, 1,001–2,000, >2,000), p = 0.60; hours (tertiles), p = 0.77. Mobile phone: latency, p = 0.37; hours (1–1,000, 1,001–2,000, >2,000), p = 0.997; hours (tertiles), p = 0.89. Cordless phone: latency, p = 0.04; hours (1–1,000, 1,001–2,000, >2,000) p = 0.64; hours (tertiles), p = 0.46.



105

in HR (data not included in table). Separate calculations for analogue and digital types of mobile phone were also made but several of these results were based on too low a number of cases to be meaningful to report.

Discussion

Astrocytic tumours comprise different neoplasms with different growth potential, progression and clinical course. Malignant transformation is a multistep process reflecting the type and sequence of genetic alternations. Astrocytic tumours are divided in two groups depending on the malignant potential: low-grade (WHO grade I-II) and high-grade (WHO grade III-IV). They comprise more than 60% of all primary intracranial neoplasms and in this study 76% of all malignant brain tumours. Lowgrade astrocytoma has a relatively favourable prognosis. About 5% of all astrocytoma consist of the pilocytic type (WHO grade I), which occurs mainly in children and young adults. Long-term survival is common and malignant transformation is rare. In our case series, only 23 patients were diagnosed with WHO grade I. Diffuse astrocytoma (WHO grade II) represents 10– 15% of all astrocytoma. The peak incidence is among young adults aged 30–40 years with a slight predominance in males. The mean survival after surgery is 6–8 years. The prognosis is, however, mainly influenced by malignant progression to glioblastoma, which has been reported to occur after 4–5 years [10, 11]. Anaplastic astrocytoma (WHO grade III) is a diffuse infiltrating tumour that primarily affects adults with a mean age of about 40 years. It may arise from a diffuse astrocytoma, WHO grade II, or without evidence of a malignant precursor. It has a strong tendency for progression to glioblastoma (WHO grade IV) with a mean time of about 2 years [12]. The most common primary brain tumour is glioblastoma (WHO grade IV), which accounts for about 60–75% of all astrocytic tumours. In our series it was 70%. The peak incidence is between 45 and 75 years of age. Secondary glioblastoma develops from diffuse astrocytoma, WHO grade II, or from anaplastic astrocytoma, WHO grade III. Secondary glioblastoma is less frequent than primary, less than 10%, and typically develops in younger persons, mean age 45 years. The time interval from diffuse astrocytoma grade II to glioblastoma grade IV varies from less than 1 year to more than 10 years, with a mean interval of 4–5 years [11, 13]. Survival for patients with secondary glioblastoma has been reported to be lon106


ger (median 7.8 months) than for primary glioblastoma (median 4.7 months) [11, 12]. The main finding of this study was an elevated risk of shortened survival for cases with glioma associated with long-term use of wireless phones. The HR increased with both latency and cumulative use of both mobile and cordless phones with a statistically significant trend. Regarding astrocytoma, this pattern was also found with latency for the most malignant type, grade IV, whereas the opposite was found for astrocytoma grade I-II, without a statistically significant trend, however. Regarding astrocytoma grade III, the use of wireless phones did not statistically significant increase or decrease survival for latency or cumulative use. The mean and median age among the cases that had used a wireless phone was lower than for those with no use. Younger age is a predictive factor for longer survival among glioma patients, especially glioblastoma [11, 14]. When adjustment was made for age, the cases with glioblastoma who had used wireless phones had an elevated risk of shortened survival compared to unexposed cases in our study. The contrasting finding for low-grade glioma is striking and cannot be explained by the age difference since we had adjusted for that. The overall finding was a statistically significant decreased risk of survival failure among cases with astrocytoma WHO grade I-II (HR = 0.5, 95% CI = 0.3–0.9) for wireless phone use. The fact that there is no clear trend with intensity or duration of wireless phone use does not speak in favour of an effect of RFEMF from such use. The exposure might, however, produce awareness bias in these cases. RF-EMF may give tumour promotion [15], inducing disease-related personality disturbances and habit changes leading to earlier tumour diagnosis than among unexposed patients. This would result in an earlier treatment with a better prognosis after surgery in this patient group [10]. Regarding astrocytoma WHO grade III, no pattern of an increased or decreased risk was found, although a somewhat elevated HR was found for cumulative use of mobile or cordless phone 12,000 h. This result was based on a low number of cases and no trend was seen when cumulative use was divided in tertiles. The genetic profile differs for different types of astrocytoma [16]. For example, overexpression of BAG3 protein in human glioblastoma has been indicated to promote tumour growth and give poorer survival for these patients [17]. Exposure to RF-EMFs may change the genetic profile important for survival, which would be of interest in further studies. Hardell/Carlberg

There is no general accepted mechanism by which RFEMF exposure produces changes in DNA. The energy level associated with exposure is too low to cause direct DNA strand breaks and DNA crosslinks. However, DNA damages can be caused by cellular biochemical activities such as free radicals. Several studies indicate that RFEMFs increase free radical activity in cells, as reviewed by Phillips et al [7]. This process is probably mediated via the Fenton reaction. Hydrogen peroxide is converted into hydroxyl free radicals that are potent cytotoxic molecules. This reaction is catalysed by iron. High levels of iron are found in metabolic active cells such as cancer cells and cells undergoing abnormal proliferation, but also in brain cells. Glial cells might turn cancerous due to DNA damage, but they might also be killed by RF-EMF due to the over-accumulation of genetic damage. Thus, one possibility may be that RF-EMF could retard tumour growth and kill cancer cells. The effect might depend on both the level and time of exposure. Interestingly, an anti-tumour effect of EMF was shown for patients with advanced hepatocellular cancer [18], and the growth of hepatocellular and breast cancer cells was significantly decreased by specific modulation EMF frequencies [19]. It should also be noted that extremely low-frequency (ELF) EMF has been classified as possibly carcinogenic to humans [20]. The mechanism is unclear, although free radicals might be involved [21]. A poorer survival among children with acute lymphoblastic leukaemia exposed to ELF-EMF has been reported in two studies [22, 23]. These findings certainly strengthen a causal association between exposure to ELF-EMF and childhood leukaemia. It should be noted that ELF-EMF is also emitted by digital mobile phones [24]. The strength of our study is that we had histopathology verification of all tumour diagnoses. The follow-up of all cases was complete, using the Swedish population registry. Due to the unique ID number for every Swedish person, all inhabitants can be traced for current address and vital status. For deceased persons, the date of death is registered. However, causes of deaths are not available from that registry. It would have been of value to get information on the clinical cause of death in the brain tumour cases, but it is unlikely that most cases with, for example, glioblastoma had other causes of death. We had no objective verification of the use of wireless phones; such use was assessed by questionnaires and supplementary phone interviews. The response rate was high, and it is unlikely that the results were influenced by selection bias. We have discussed potential bias in more detail in previous publications [1, 3, 4, 8]. It would have been of

interest to assess the use of wireless phones after tumour diagnosis to further elucidate the influence of exposure to RF-EMF emissions on prognosis, at least for low-grade glioma, but that was not part of the case-control studies. The results regarding exposure before diagnosis seem to be of biological relevance and it is unlikely that these results are chance findings. Our findings indicate a complex biological effect from RF-EMF exposure and strengthen a causal association between these tumour types and the use of wireless phones.



Conclusion

This study showed elevated HR, indicating decreased survival of glioma cases with long-term and high cumulative use of wireless phones. The results differed according to WHO grade of astrocytoma: with an increased HR for astrocytoma WHO grade IV, a survival disadvantage. However, a decreased HR was found for astrocytoma WHO grade I-II, indicating a survival benefit in that group of cases. This could be caused by RF-EMF exposure leading to tumour promotion and earlier detection and surgery with better prognosis in that patient group. Further studies are needed to confirm these findings and to investigate cellular genetic profile alterations from RFEMF exposure.

Acknowledgments The authors thank Professor Henry Lai for valuable comments. The study was supported by grants from Cancer- och Allergifonden and Cancerhjälpen.

Disclosure Statement The authors state no conflicts of interest.

References

1 Hardell L, Carlberg M, Hansson Mild K: Pooled analysis of two case-control studies on use of cellular and cordless telephones and the risk for malignant brain tumours diagnosed in 1997–2003. Int Arch Occup Environ Health 2006;79:630–639. 2 Hardell L, Carlberg M, Hansson Mild K: Pooled analysis of two case-control studies on the use of cellular and cordless telephones and the risk of benign brain tumours diagnosed during 1997–2003. Int J Oncol 2006; 28:509–518.

107

3 Hardell L, Carlberg M, Hansson Mild K: Pooled analysis of case-control studies on malignant brain tumours and the use of mobile and cordless phones including living and deceased subjects. Int J Oncol 2011;38:1465– 1474. 4 Hardell L, Carlberg M, Hansson Mild K: Case-control study on cellular and cordless telephones and the risk for acoustic neuroma or meningioma in patients diagnosed 2000– 2003. Neuroepidemiology 2005;25:120–128. 5 Baan R, Grosse Y, Lauby-Secretan B, El Ghissassi F, et al: Carcinogenicity of radiofrequency electromagnetic fields. Lancet Oncol 2011;12:624–626. 6 Lai H, Singh NP: Melatonin and a spin-trap compound blocked radiofrequency radiation-induced DNA strand breaks in rat brain cells. Bioelectromagnetics 1997:18:446–454. 7 Phillips JL, Singh NP, Lai H: Electromagnetic fields and DNA damage. Pathophysiology 2009;16:79–88. 8 Hardell L, Carlberg M, Hansson Mild K: Mobile phone use and the risk for malignant brain tumors: a case-control study on deceased cases and controls. Neuroepidemiology 2010;35:109–114. 9 Louis DN, Ohgaki H, Wiestler OD, et al: The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol 2007;114:97–109.

108

10 Vertosick FT Jr, Selker RG, Arena VC: Survival of patients with well-differentiated astrocytomas diagnosed in the era of computed tomography. Review. Neurosurgery 1991; 4:496–501. 11 Ohgaki H, Kleihues P: Population-based studies on incidence, survival rates, and genetic alterations in astrocytic and oligodendroglial gliomas. Review. J Neuropathol Exp Neurol 2005;64:479–489. 12 Ohgaki H, Dessen P, Jourde B, et al: Genetic pathways of glioblastoma: a populationbased study. Cancer Res 2004;64:6892–6899. 13 Watanabe K, Sato K, Biernat W, et al: Incidence and timing of p53 mutations during astrocytoma progression in patients with multiple biopsies. Clin Cancer Res 1997; 3: 523–530. 14 Ohgaki H, Kleihues P: Epidemiology and etiology of gliomas. Review. Acta Neuropathol 2005;109:93–108. 15 Kundi M: Essential problems in the interpretation of epidemiologic evidence for an association between mobile phone use and brain tumours. C R Phys 2010;11:556–563. 16 Ohgaki H, Kleihues P: Genetic profile of astrocytic and oligodendroglial gliomas. Brain Tumor Pathol 2012; 28:177–183. 17 Festa M, Del Valle L, Khalili K, et al: BAG3 protein is overexpressed in human glioblastoma and is a potential target to therapy. Am J Pathol 2011; 178:2504–2512.


18 Costa FP, de Oliveira AC, Meirelles R, et al: Treatment of advanced hepatocellular carcinoma with very low levels of amplitudemodulated electromagnetic fields. Br J Cancer 2011;105:640–648. 19 Zimmerman JW, Pennison MJ, Brezovich I, et al: Cancer cell proliferation is inhibited by specific modulation frequencies. Br J Cancer 2012;106:307–313. 20 IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Non-Ionizing Radiation. Part 1. Static and extremely lowfrequency electric and magnetic fields. Lyon, IARC, 2002. 21 Lai H, Singh NP: Magnetic-field-induced DNA strand breaks in brain cells of the rat. Environ Health Perspect 2004;112:687–694. 22 Foliart DE, Pollock BH, Mezei G, et al: Magnetic field exposure and long-term survival among children with leukaemia. Br J Cancer 2006;94:161–164. 23 Svendsen AL, Weikopf T, Kaatsch P, Schüz J: Exposure to magnetic fields and survival after diagnosis of childhood leukemia: a German cohort study. Cancer Epidemiol Biomarkers Prev 2007;16:1167–1171. 24 Linde T, Hansson Mild K: Measurement of low frequency magnetic fields from digital cellular telephones. Bioelectromagnetics 1997;18:184–186.

Hardell/Carlberg