Page 37 - Contributed Paper Session (CPS) - Volume 6

CPS1483 Ken Karipidis et. al
used, the RF energy absorption is at least several times higher in the temporal
lobe than in the frontal lobe.(13) In our data there was a large number of
gliomas with unspecified or overlapping location. Reclassification of these did
reduce the trends for the temporal lobe during the periods before substantial
mobile phone use, and for the frontal lobe during all the periods.
In our study we also compared the observed incidence with a modelled
predicted incidence assuming a causal association between mobile phone use
and glioma as reported in the Interphone and Hardell studies. The results
suggest that, if the effects of mobile phones on glioma risk are real, then the
incidence rates would be far higher than those observed. The present study
has some limitations. The accuracy of the Australian cancer registration system
in the early periods when it began in the 80s is unknown for all the states and
territories.
We estimated mobile phone use using information on mobile phone
accounts, and this may not be a true indicator of actual use as some people
may have multiple accounts and others may use a phone without having an
account. We mitigated this by also using data from a consumer survey
conducted by the national telecommunications regulator on the proportion of
the population using mobile phones. Information from the survey was only
available from the years 2009 to 2013 and this was applied to data on the
annual number of mobile phone accounts from 1987. However, mobile phone
use patterns have likely changed from 1987 to 2009. Further, the exposure
metric is unclear when investigating whether mobile phone use is implicated
in brain cancer risk. Prevalence of phone use is a de facto measure for the
amount of RF energy a person is receiving when using a mobile phone, and
changes in technology and patterns of individual use were not taken into
account in this investigation.
We estimated the prevalence of mobile phone use equally across the 20-
59 age range and both males and females. The use of subscription data in
early years is likely to underestimate prevalence of use in males and
overestimate it in females given that users in early years were middle-aged
working men on company mobile phone subscriptions.(14) In later years
mobile phone use became equal between the two genders.(15)
For information on the proportion of regular and heavy mobile phone users
we used data from the Interphone study, which also included data from
Australia. Mobile phone use in the Interphone study was self-reported, relying
on participants’ recall of past phone use.(2) Sensitivity analyses on the
Interphone methodology reported that for short term recall (up to a year)
there was underestimation of phone use by regular users and overestimation
by heavy users.(16) For longer recall (3 to 5 years) there was an
underestimation of number of calls and an overestimation on the duration of

26 | I S I W S C 2 0 1 9