Mobile phone use and risk of glioma in adults: case-control study.

doi:10.1136/bmj.38720.687975.55
2006;332;883-887; originally published online 20 Jan 2006; BMJ
 
Swerdlow, Martie J A van Tongeren and Patricia A McKinney
Sarah J Hepworth, Minouk J Schoemaker, Kenneth R Muir, Anthony J
 
case-control study
Mobile phone use and risk of glioma in adults:
http://bmj.com/cgi/content/full/332/7546/883
Updated information and services can be found at:
These include:
References
http://bmj.com/cgi/content/full/332/7546/883#otherarticles
8 online articles that cite this article can be accessed at:
 
http://bmj.com/cgi/content/full/332/7546/883#BIBL
This article cites 17 articles, 10 of which can be accessed free at:
Rapid responses
http://bmj.com/cgi/eletter-submit/332/7546/883
You can respond to this article at:
 
http://bmj.com/cgi/content/full/332/7546/883#responses
for free at:
14 rapid responses have been posted to this article, which you can access
service
Email alerting
box at the top left of the article
Receive free email alerts when new articles cite this article - sign up in the
Topic collections
(977 articles) Cancer:other •
(2674 articles) Other Public Health •
(1694 articles) Other Epidemiology •
 
Articles on similar topics can be found in the following collections
Notes  
To order reprints follow the "Request Permissions" link in the navigation box
http://resources.bmj.com/bmj/subscribers
go to: BMJTo subscribe to
on 28 September 2007 bmj.comDownloaded from

Research
Mobile phone use and risk of glioma in adults: case-control study
Sarah J Hepworth, Minouk J Schoemaker, Kenneth R Muir, Anthony J Swerdlow, Martie J A van Tongeren, Patricia A
McKinney
Abstract
Objective To investigate the risk of glioma in adults in relation
to mobile phone use.
Design Population based case-control study with collection of
personal interview data.
Setting Five areas of the United Kingdom.
Participants 966 people aged 18 to 69 years diagnosed with a
glioma from 1 December 2000 to 29 February 2004 and 1716
controls randomly selected from general practitioner lists.
Main outcome measures Odds ratios for risk of glioma in
relation to mobile phone use.
Results The overall odds ratio for regular phone use was 0.94
(95% confidence interval 0.78 to 1.13). There was no relation
for risk of glioma and time since first use, lifetime years of use,
and cumulative number of calls and hours of use. A significant
excess risk for reported phone use ipsilateral to the tumour
(1.24, 1.02 to 1.52) was paralleled by a significant reduction in
risk (0.75, 0.61 to 0.93) for contralateral use.
Conclusions Use of a mobile phone, either in the short or
medium term, is not associated with an increased risk of glioma.
This is consistent with most but not all published studies. The
complementary positive and negative risks associated with
ipsilateral and contralateral use of the phone in relation to the
side of the tumour might be due to recall bias.
Introduction
Gliomas are the most common malignancy of the central nerv-
ous system in adults, and the prognosis is extremely poor.1 The
distinct histopathology and cellular origin of gliomas are
probably associated with different aetiological pathways and
mechanisms of carcinogenesis than other subtypes of brain
tumours; the aetiology of gliomas, however, remains unclear.
Recently, considerable interest has focused on whether the use of
mobile phones is associated with an increased risk of gliomas
and other brain tumours, even though little is known about
potential mechanisms.2 The energy of the radiofrequency fields
emitted by mobile phones is thought to be insufficient to cause
malignant transformation through direct damage to DNA.3
Most published epidemiological studies on mobile phone
use and gliomas have not generally reported any increased risk
either overall or with long term use.4–7 Individual studies have
found positive associations between high grade astrocytoma
(glioma) and phone use ipsilateral to the side of the tumour,8
brain tumours and phone use in rural areas,9 and use of
analogue mobile phones.8 10
We carried out a large population based case-control study of
966 patients with glioma in the United Kingdom. This study is
part of the Interphone project,11 an international collaboration
of 13 countries investigating mobile phone use and the risk of
intracranial tumours.
Methods and participants
The study took place in the Thames regions of south east
England and four areas to the north (Trent, West Midlands, West
Yorkshire, and southern Scotland). The total catchment popula-
tion (28.4 million) comprised 48.3% of the UK population. All
areas followed a common protocol, with identical methods of
case ascertainment and data collection with controls randomly
sampled from general practitioner lists. The south east slightly
differed in its method of control selection and the age range cov-
ered.
Cases were ascertained from multiple sources, including hos-
pital departments (neurosurgery, neuro-oncology, neuropathol-
ogy, neuroradiology, neurology) and cancer registries. Patients
aged 18-69 years (northern centres) or 18-59 years (south east)
lived in the study areas and had a first diagnosis between 1
December 2000 and 30 June 2003 (northern) or 29 February
2004 (south east) with a glioma (ICD-O-3 (international classifi-
cation of diseases for oncology)12 topography: C71, morphology:
9380-9411, 9420-9460, 9480, 9505). Data on site, laterality (left,
right, central) and grade of tumour (WHO grade high III-IV; low
I-II13) were abstracted from scan and pathology reports.
In the most recent year we have published data for (1992) an
estimated 98% of the UK population was registered with a gen-
eral practitioner.14 Controls were randomly selected from
general practitioners’ lists by a preset algorithm. In the south east
the controls were frequency matched to reflect the age, sex, and
geographical distribution of cases. In the northern centres one
control per case was individually matched on age, sex, and gen-
eral practice after the patient with glioma was interviewed. Non-
participating controls were replaced. Parallel case-control studies
of meningioma, acoustic neuroma, and other brain tumours
were carried out with identical methods and questionnaires; the
controls for these cases were included in the present analyses.
Consultants or general practitioners approached eligible
participants personally or by an invitation letter. The study was
introduced as an investigation of risk factors for brain tumours,
without emphasising mobile phones. With the participant’s
informed consent, trained interviewers conducted a computer
assisted personal interview. For 69 patients with glioma (7%),
interviewers conducted proxy interviews, mainly with spouses.
During the interview, if participants reported that they had
ever made one or more calls each week on average for a period
of six months or longer, they were asked a detailed set of
questions on mobile phone use. For such participants, all makes
and models of phone were recorded, with a comprehensive rep-
Cite this article as: BMJ, doi:10.1136/bmj.38720.687975.55 (published 20 January 2006)
BMJ Online First bmj.com page 1 of 5
on 28 September 2007 bmj.comDownloaded from

ertoire of photographs to prompt recall. For each phone, the
interviewer recorded the network operator, start and stop year,
and the number and duration of calls made and received. If par-
ticipants were uncertain about the calendar years or amount of
use, a range was reported with the mean value taken for analysis.
Additional details were gathered on which side of the head the
phone was mostly used (50% or more of the time) and factors
influencing emitted power levels to the head, including use of
hands-free kits (start and stop dates of use and proportion of
time used) and whether the phone was used mainly in an urban
or rural area or equally in both.15
In the analysis, we defined regular phone use as use for at
least six months in the period more than a year before diagnosis.
We defined diagnostic date as the date of diagnostic pathology
(n = 935, 97%) or date of the first diagnostic scan (n = 31, 3%) if
diagnostic pathology was unavailable. We assessed exposure to
mobile phones using the number of years from first regular use
of a phone until diagnosis or equivalent reference date for con-
trols, lifetime years of regular use, lifetime cumulative use
(hours), and lifetime cumulative number of calls. Long term
users ( > 10 years) were dichotomised into heavy ( ≥ 113 hours)
and light ( < 113 hours) cumulative hours of use in the period 10
or more years before diagnosis, with the cut off point based on
the median hours of use among control participants. We used
data collated by Interphone11 to classify phones as either
analogue or digital, based on make and model of phone, year of
use, and network operator capabilities during that period.
The exposure period for people with glioma was calculated
up to a year before the date of diagnosis. An equivalent reference
date was required for control participants that allowed for the
increase in mobile phone use over the study period and as con-
trols tended to be interviewed after the patients with glioma. For
each area (south east, northern), we constructed case strata by
single calendar year of interview and single year interval between
diagnosis and interview (“interview lag time”). Control
participants interviewed in each calendar year were randomly
allocated to strata of interview lag time, proportionally to the dis-
tribution of the cases in the same calendar year, to obtain a simi-
lar distribution of lag time as in the cases. We then calculated the
reference dates for controls by subtracting the mean interview
lag time in cases in that stratum from the interview dates of the
controls. Exposure indices were calculated up to a year before
this reference date for controls (that is, a one year latency time
was used). Additional analyses were carried out with a five year
latency time.
Statistical analysis
For statistical analysis we used unconditional logistic regression
(StataCorp, College Station, TX) adjusted for nine regions (five
regions within the south east and the four northern regions), age
at reference date (five year categories), sex, deprivation
(Townsend score16), and combinations of interview year and lag
time to account for the fact that controls were, on average, inter-
viewed later in the study period than patients with glioma. We
derived odds ratios for cumulative use with and without modifi-
cation for reported use of headsets or hands-free sets in a vehi-
cle, or both,7 17 with and without proxy case interviews, and
separately for high and low grade tumours and urban versus
rural use. We also performed a conditional logistic regression
analysis on the matched northern case-control dataset.
We used two methods to assess the risk of a tumour ipsilateral
or contralateral to side of phone use. Firstly, we took two groups
of patients with right and left sided tumours17 and randomly
assigned controls to each group and considered them to have a
tumour on that side (50% left, 50% right) for the analysis. The
odds ratio for risk of an ipsilateral tumour was based on the
results of a logistic regression analysis where ipsilateral phone
use was use on the same side of the head as the tumour for cases
or the assigned side for controls. We adjusted the analysis for the
side of the tumour as well as the variables adjusted for in the
main analysis. If the phone was used on the opposite side to the
side of tumour/allocated side the participant was classified as
unexposed. Those who reported using the phone on both sides
of the head were considered exposed on both the left and right
sides.6 17 As the allocation of controls was based on a random
assignment, the logistic regression analysis was repeated 500
times; the results for the analysis that gave the median odds ratio
for ipsilateral regular phone use are presented. A similar analysis
of contralateral use (use on the opposite side of the head to the
tumour or allocated side) was performed. The second method
calculates a relative risk for laterality of reported side of use in
relation to tumour laterality but only in cases.4
Results
Researchers interviewed 966 cases (367 in the south east and
599 in northern areas) and 1716 controls (630 and 1086). The
main reasons for non-participation were the participant was too
ill or had died before interview (cases 30%, controls < 1%), non-
response (cases 2%, controls 21%), refusal (cases 10%, controls
29%), and other reasons (refusal by consultant or general practi-
tioner, non-English speaking, mental impairment) (cases 7%,
controls 5%). Non-responders included those for whom contact
details may have been incorrect. Overall response rates were 51%
for patients with glioma and 45% for controls, representing the
proportion of all eligible cases and controls from the study areas
who were interviewed in the study. Exclusion of the
non-responders (that is, who may never actually have been
asked) gave response rates of 51% and 57%. Interviewed patients
with glioma were broadly representative of the overall set of
those eligible by age and sex but differed by deprivation
category, being significantly more affluent (�2 test for trend,
P < 0.001). People with low grade glioma were significantly more
likely to be interviewed (�2 test, P < 0.001) than those with high
grade glioma. For control participants, those interviewed were
more likely to be women (�2 test, P < 0.001) and more likely to be
affluent (�2 test for trend, P < 0.001) than those who were not
interviewed.
Table 1 shows the demographic distribution of interviewed
cases and controls. The proportion of men was higher among
the patients with glioma than in the control group. There was
also a slight tendency for interviewed controls to live in more
affluent areas than interviewed patients with glioma.
Table 2 shows an odds ratio of 0.94 (95% confidence interval
0.78 to 1.13) for regular phone users compared with those who
never or only occasionally used mobile phones. There was no
association of risk with lifetime years of use, cumulative hours of
use, cumulative numbers of calls, nor cumulative hours of use
over 10 years before the reference date. These findings were
similar after we excluded patients with glioma with proxy
interviews (n = 69), adjusted cumulative hours of phone use and
number of calls for use of hands-free kits, applied a five year lag
time, or restricted analysis to matched case-control analysis of
northern cases. Table 2 also shows no significant associations
with use in urban or rural areas or separately for 650 high grade
and 306 low grade gliomas.
We found a significant odds ratio of 1.24 (1.02 to 1.52) for a
tumour ipsilateral to side of phone use and a reduced odds ratio
Research
page 2 of 5 BMJ Online First bmj.com
on 28 September 2007 bmj.comDownloaded from

for contralateral use (0.75, 0.61 to 0.93) (table 2). Similar respec-
tive excesses and deficits were present for all exposure measures
of mobile phone use, including use for ≥ 10 years (ipsilateral
1.60, 0.92 to 2.76; contralateral 0.78, 0.43 to 1.41). To investigate
this further, we analysed regular use ipsilateral and contralateral
to handedness, which gave odds ratios of 0.78 (0.62 to 0.99) and
1.07 (0.85 to 1.35), respectively. The concordance between
reported side of use and handedness was 59% for cases and 64%
among controls. The method of Inskip et al4 gave an overall rela-
tive risk of 1.3 (Fisher’s exact P < 0.001) for a tumour ipsilateral to
the side of phone use.
We examined use of analogue phones separately but there
were no significant odds ratios with any exposure metric (table
3). Results from the conditional regression analysis of matched
data from the northern centres did not differ from the overall
findings.
Discussion
This large study on associations between mobile phone use and
the risk of developing a glioma in a UK population has nearly
twice as many cases as the largest previously reported studies of
gliomas,4 5 with more long term users. In addition, it was
designed specifically to address exposure to mobile phones, with
comprehensive and relevant collection of data. Overall, we found
no raised risk of glioma associated with regular mobile phone
use and no association with time since first use, lifetime years of
use, cumulative hours of use, or number of calls. Our results are
consistent with findings from investigations of mobile phone use
in the US,4 5 Denmark,7 18 and Sweden,6 though some studies
have found isolated positive associations for particular
variables.8–10
Analogue phones emit higher average power levels than dig-
ital phones.19 If mobile phone use was causally linked to the
development of glioma and risk was related to power level, we
would predict a higher risk for analogue phone use than for dig-
ital phones. As in some6 7 18 but not all8 10 previous reports we
found no association between risk of glioma and use of analogue
Table 1 Demographic distributions in cases and controls. Figures are
numbers (percentages) of participants
Cases (n=966) Controls (n=1716)
Region:
Thames regions 367 (38.0) 630 (36.7)
Southern Scotland 152 (15.7) 277 (16.1)
Trent 199 (20.6) 372 (21.7)
West Midlands 115 (11.9) 207 (12.1)
West Yorkshire 133 (13.8) 230 (13.4)
Age at reference date (years):
18-29 100 (10.4) 112 (6.5)
30-39 199 (20.6) 281 (16.4)
40-49 216 (22.4) 429 (25.0)
50-59 328 (34.0) 645 (37.6)
60-69† 123 (12.7) 249 (14.5)
Men 604 (62.5) 829 (48.3)
Women 362 (37.5) 887 (51.7)
Deprivation score*:
1 (most affluent) 257 (26.6) 513 (29.9)
2 229 (23.7) 386 (22.5)
3 178 (18.4) 334 (19.5)
4 181 (18.7) 292 (17.0)
5 (least affluent) 121 (12.5) 191 (11.1)
*Townsend score (area based measure of deprivation) categorised into five equally sized
groups based on 2001 census data.
†In control group includes seven people aged >69 at reference date.
Table 2 Odds ratios and 95% confidence intervals for risk of glioma in
relation to mobile phone exposure*. Figures are numbers (percentages) of
participants
Factor and level of exposure Cases (n=966)
Controls
(n=1716) Odds ratio† (95% CI)
Frequency of use:
Never/non-regular 456 (47.2) 818 (47.7) 1.00
Regular 508 (52.6) 898 (52.3) 0.94 (0.78 to 1.13)
Not known 2 (0.2) 0 —
Years since first use:
Never/non-regular 456 (47.2) 818 (47.7) 1.00
1.5-4‡ 271 (28.1) 515 (30.0) 0.90 (0.73 to 1.11)
5-9 170 (17.6) 270 (15.7) 1.04 (0.80 to 1.34)
≥10 66 (6.8) 112 (6.5) 0.90 (0.63 to 1.28)
Not known 3 (0.3) 1 (0.1) —
Lifetime years of use:
Never/non-regular 456 (47.2) 818 (47.7) 1.00
0.5-4 342 (35.4) 623 (36.3) 0.93 (0.76 to 1.14)
5-9 115 (11.9) 206 (12.0) 0.88 (0.66 to 1.17)
≥10 48 (5.0) 67 (3.9) 1.14 (0.74 to 1.73)
Not known 5 (0.5) 2 (0.1) —
Cumulative hours of use§:
Never/non-regular 456 (47.2) 818 (47.7) 1.00
≤99 225 (23.3) 444 (25.9) 0.94 (0.76 to 1.17)
99-≤544 128 (13.3) 218 (12.7) 0.87 (0.65 to 1.15)
>544 135 (14.0) 217 (12.6) 0.94 (0.71 to 1.23)
Not known 22 (2.3) 19 (1.1) —
Cumulative number of calls§:
Never/non-regular 456 (47.2) 818 (47.7) 1.00
≤2071 237 (24.7) 444 (25.9) 0.99 (0.80 to 1.23)
2071-≤6909 102 (10.6) 217 (12.6) 0.70 (0.52 to 0.93)
>6909 146 (15.1) 218 (12.7) 0.97 (0.74 to 1.28)
Not known 25 (2.6) 19 (1.1) —
Cumulative hours of use ≥10 years ago¶:
Never/non-regular 456 (47.2) 818 (47.7) 1.00
<10 years 429 (44.4) 772 (45.0) 0.93 (0.77 to 1.13)
≥10 years, ≤113 hours 23 (2.4) 56 (3.3) 0.61 (0.36 to 1.04)
≥10 years, >113 hours 39 (4.0) 54 (3.2) 1.11 (0.70 to 1.75)
Not known 19 (2.0) 16 (1.0) —
Proportion urban/rural at first use:
Never/non-regular 456 (47.2) 818 (47.7) 1.00
Mainly urban 241 (24.9) 471 (27.4) 0.83 (0.66 to 1.03)
Mainly rural 49 (5.1) 84 (4.9) 0.98 (0.66 to 1.46)
Both 215 (22.3) 343 (20.0) 1.05 (0.83 to 1.31)
Not known 5 (0.5) 0 —
According to tumour grade**
Frequency of use in those with high grade tumours:
Never/non-regular 331 (50.9) 818 (47.7) 1.00
Regular 317 (48.8) 898 (52.3) 0.95 (0.77 to 1.17)
Not known 2 (0.3) 0 —
Frequency of use in those with low grade tumours:
Never/non-regular 122 (39.9) 818 (47.7) 1.00
Regular 184 (60.1) 898 (52.3) 0.85 (0.63 to 1.13)
Not known 0 0 —
According to side of phone use††
Frequency of ipsilateral use*:
Never/non-regular 550 (66.3) 1230 (71.7) 1.00
Regular 278 (33.5) 486 (28.3) 1.24 (1.02 to 1.52)
Not known 2 (0.2) 0 —
Frequency of contralateral use*:
Never/non-regular 629 (75.8) 1225 (71.4) 1.00
Regular 199 (24.0) 491 (28.6) 0.75 (0.61 to 0.93)
Not known 2 (0.2) 0 —
*Reference category is never or non-regular use of any type of mobile phone and, in
ipsilateral analysis, phone use only on opposite side of tumour, and in contralateral analysis,
phone use only on same side as tumour.
†Odds ratios adjusted for age at reference date (in 5 year age groups), sex, region, Townsend
deprivation category, and interview reference date category.
Research
BMJ Online First bmj.com page 3 of 5
on 28 September 2007 bmj.comDownloaded from

phones overall or with time since first use, lifetime years of use, or
cumulative hours or number of calls.
In Sweden Hardell et al reported raised risks for mobile
phone use ipsilateral to the side of development of high grade
astrocytomas (the principal subtype of glioma)8 and for rural use
in different analyses of the same study.9 Several other studies,
however, could not confirm these results and the methods were
criticised.3 20 21 Our laterality analyses both showed a significantly
raised risk for ipsilateral phone use. Using the methods of Lonn
et al,17 however, we found a significantly reduced risk for contral-
ateral use. These “complementary risks” above and below unity,
which we observed for all measures of mobile phone use, can
probably be explained by recall bias.21 The patients with glioma,
who were aware of the location of their tumour, may have
considered that mobile phone use was a cause of its
development, resulting in systematically over-reporting of phone
use on the side of the head where their cancer occurred. Gener-
ally, individuals are likely to overestimate their actual use of
mobile phones,22 and this may have exaggerated the effect of dif-
ferential reporting for laterality. To investigate this issue further,
we made case-control comparisons by handedness as people
tend to use the phone on their handed side and because
handedness is a variable reported at interview and not likely to
be subject to recall bias. The clear pattern of significant odds
ratios above and below unity for ipsilateral and contralateral
mobile phone use, respectively, was not seen for handedness in
equivalent analyses.
Potential bias
Case-control studies are subject to certain biases and particularly
participation bias.23 We interviewed 51% of those patients with
glioma who were eligible, mainly because rapid death prevented
us from approaching all of them. As early death is most likely in
patients with high grade tumours, it is not surprising that partici-
pation rates were higher in those with low grade tumours. A bias
in these results would occur only if mobile phone use was related
to severity of tumour, which was not supported by our analysis,
where odds ratios for mobile phone use showed no increased
risk for high or low grade tumours.
There is also potential for the introduction of participation
bias into the control group. All controls were selected to
represent the general population by using the sampling frame of
general practitioners’ lists. Although methods varied between the
south east and northern centres, the analysis strategy of
frequency matching permitted a combined approach. The over-
all response rate for controls was relatively low (45%) compared
with previously published studies from the Nordic countries6 7
on mobile phone use and risk of glioma. Recruitment of controls
was a complicated procedure, and we tried to optimise participa-
tion rates. Prevailing legislation on consent from patients made
the process a resource intensive exercise for general practition-
ers and in some instances the study could not follow-up patients
directly, instead relying on general practitioners to undertake
this process. The principal reason for non-participation of
controls was that they were “uncontactable,” and constraints
imposed by ethical approval bodies prevented more than one
follow-up attempt. It is likely, therefore, that some of the
apparent non-responders were in fact individuals who were
never contacted and therefore had no opportunity to respond.
Our interviewed controls were more affluent than their
non-interviewed counterparts and the interviewed patients with
glioma. Though we adjusted for deprivation in all the analyses,
this cannot completely remove its potential influence.
There is generally a lack of convincing and consistent
evidence of any effect of exposure to radiofrequency field on risk
of cancer.24 25 Overall our findings are consistent with this and
with most studies on mobile phone use. The positive association
found between risk of glioma and ipsilateral mobile phone use
was accompanied by a negative association for the opposite side
of use to the tumour. Although it is possible the ipsilateral asso-
ciation represents a real effect, this finding is probably explained
by recall bias, with patients with glioma systematically
over-reporting use on the same side as their tumour and conse-
quently under-reporting use on the opposite side. This study
suggests that there are no substantially raised risks of glioma in
the 10 years after first mobile phone use. Only future studies will
be able to address longer latency periods for the development of
glioma.
We thank all the individuals who were interviewed for this project; all the
study interviewers, administrators, and computer programmers who
collected and processed the data; and Elisabeth Cardis, the coordinator of
the Interphone study, and her colleagues for support and provision of data.
The Northern UK study acknowledges the support of the membership of
the study steering group chaired by David Coggon and the following neu-
ropathologists, neuroradiologists, neurosurgeons, neuro-oncologists, clini-
cal oncologists, neurologists, specialist nurses, and administrators based in
hospitals located in Scotland (P Barlow, I Bone, J Brown, J Crowther, R
‡Lower limit 1.5 years ago because regular phone use defined as phone use of at least six
months’ duration at least one year before reference date.
§For cumulative number and duration of calls category cut-off points were median and 75th
centile of use for controls who were regular phone users.
¶Use over 10 years before reference date for controls and diagnosis date for cases.
**10 tumours (1.7%) were of undetermined grade.
††In 449 (46.5%) cases tumour was classified as being on right side of head and in 387
(40.1%) on left side, 49 cases (5.1%) were excluded from this analysis because tumour was
central, 81 cases (8.4%) because side of tumour was unknown, six others where side of
phone use was unknown were also excluded.
Table 3 Odds ratios and 95% confidence intervals for risk of glioma in
relation to use of analogue phones
Factor and level of
exposure Cases Controls Odds ratio (95% CI)
Frequency of use*:
Never/non-regular 456 (47.2) 818 (47.7) 1.0
Digital only 378 (39.1) 685 (39.9) 0.96 (0.79 to 1.16)
Regular analogue 128 (13.3) 212 (12.4) 0.87 (0.66 to 1.15)
Not known 4 (0.4) 1 (0.1) —
Years since first use:
Never/non-regular 456 (47.2) 818 (47.7) 1.0
Digital only 378 (39.1) 685 (39.9) 0.96 (0.79 to 1.16)
1.5-4† 15 (1.6) 33 (1.9) 0.59 (0.30 to 1.15)
5-9 56 (5.8) 84 (4.9) 0.98 (0.66 to 1.45)
≥10 56 (5.8) 95 (5.5) 0.87 (0.59 to 1.27)
Not known 5 (0.5) 1 (0.1) —
Lifetime years of use:
Never/non-regular 456 (47.2) 818 (47.7) 1.0
Digital only 378 (39.1) 685 (39.9) 0.96 (0.79 to 1.16)
0.5-4 90 (9.3) 159 (9.3) 0.82 (0.60 to 1.11)
5-9 27 (2.8) 42 (2.4) 0.97 (0.57 to 1.66)
≥10 10 (1.0) 11 (0.6) 1.20 (0.48 to 3.04)
Not known 5 (0.5) 1 (0.1) —
Cumulative hours of use ≥10 years ago‡:
Never/non-regular 456 (47.2) 818 (47.7) 1.0
Digital only 378 (39.1) 685 (39.9) 0.95 (0.79 to 1.16)
<10 years 69 (7.1) 115 (6.7) 0.86 (0.61 to 1.22)
≥10 years, ≤126 hours 23 (2.4) 47 (2.7) 0.70 (0.41 to 1.21)
≥10 years, >126 hours 31 (3.2) 47 (2.7) 0.98 (0.59 to 1.62)
Not known 9 (0.9) 4 (0.2) —
*Of 4055 phones reported by participants, 586 (14.5%) were analogue, 3183 (78.5%) were
digital, and 286 (7.1%) could not be classified.
†Lower limit 1.5 years ago because regular phone use defined as phone use of at least six
months’ duration at least one year before reference date.
‡Use ≥10 years before reference date for controls and diagnosis date for cases.
Research
page 4 of 5 BMJ Online First bmj.com
on 28 September 2007 bmj.comDownloaded from