13-84 11-18-2013 L. Lloyd Morgan 7520958286 11-19-13
Morgan Ltr to Editor 5-Country Interphone Study 4-22-09
1. Letter to the Editor
Reader’s Response: Meningioma and mobile phone use—a collaborative
case–control study in five North European countries
From L LLOYD MORGAN
There are many problems in this recently published
Interphone study by Lahkola et al.1
Among them are
unexplained differences in the number of controls
and cases, a gender mismatch among controls, a
stated concern for the ‘implausibly high reported
mean daily hours of use’, and a large fraction of
odds ratios (ORs) <1.
Control and case differences
The authors report, ‘. . . the slightly smaller numbers
of cases [1] than in national reports are due to revised
diagnosis, date of diagnosis or history of previous
brain tumour’. This would imply that errata should
have been sent to the journals that published these
three national studies.2–4
What is not reported is that
a considerably larger number of controls (55 total, 47
from the Swedish study) reported in the national
reports are missing in this study. Also, this is the
third five-country study. The total number of reported
controls in this study was 3299, in the five-country
glioma study the reported number of controls was
3301,5
and in the five-country acoustic neuroma
study the reported number of controls was 2661.6
These differences in the number of controls is partic-
ularly hard to understand because the paper states:
‘Frequency-matching employed throughout the
Interphone study allowed us to utilize the entire con-
trol group recruited for all intracranial tumours
(glioma, meningioma and acoustic neuroma) in the
matched strata of the meningioma cases, to increase
statistical power’. Finally, it is not clear when the
13-country pooled Interphone data results are finally
published if the data from this paper or the data from
the three previously published national studies will be
used. This should be clarified.
Gender mismatch among controls
The female/male ratio for cases was 3.02 and for con-
trols was 1.16. The authors cite the known fact
that incidence of meningioma is higher among
women than men. In the USA, the female/male
ratio is 2.27.7
The ratio difference between the USA
and these five countries suggests there may be an
increased risk of meningioma from cellphone use in
women relative to men but because no gender risk
data is reported this cannot be known.
Besides the likely introduction of gender bias
throughout the analysis, the use of cut points based
on control use creates a serious problem. Twice the
paper reports using controls to determine cut-points:
‘An additional analysis of the subgroup with the high-
est cumulative number of calls and cumulative hours
of use was performed with the cut-point defined as
the value among the 10% of controls with the heavi-
est mobile phone use (among regular users)’, and ‘In
analyses of categorical exposure variables, the cut-
points were chosen based on the distribution among
controls’.1
Given this is a paper about the possibility
of a risk of meningioma from cellphone use, it is hard
to understand why controls would be used to deter-
mine cut-points particularly when the gender ratio of
controls are so very different from cases for a disease
with a known increased incidence in women.
Concern for implausibly high
reported mean daily hours of use
The paper reports a significantly increased risk of
meningioma per 100 h of cellphone use [OR ¼ 1.005,
95% confidence interval (CI) 1.001–1.010], yet
expresses concern because ‘. . . this was driven by a
small number of very high values which in turn
reflected subjects with implausibly high reported
mean daily hours of use (2.4 hours per day and 3.5
hours a day for controls and cases)’.1
There is no
explanation why this would be implausible. While
this use per day seems quite plausible to me, my
views, similar to the authors of this paper, are irrele-
vant. To verify if this is plausible or not, cellphone
providers could have been queried for aggregate
hours per day of use by percentile of all subscribers.
Large fraction of ORs <1
Overall, there were 72 reported ORs, either
in tables or in the text, 64 of these were <1, 29 of
these 64 were significantly <1. The authors note this
Central Brain Tumor Registry of the United States, Hinsdale, IL,
USA. E-mail: bilovsky@aol.com
Published by Oxford University Press on behalf of the International Epidemiological Association [2009]
all rights reserved.
International Journal of Epidemiology 2009;1–2
1
Int. J. Epidemiol. Advance Access published April 22, 2009
2. when they obliquely state: ‘. . . mobile phone use was
associated with an apparently reduced risk of menin-
gioma’. Stated more directly, they were reporting
results that show use of a cellphone protects the
user from risk of a meningioma with P-values as
low as 0.000022 (OR ¼ 0.68, 95% CI 0.57–0.85 for
<125 h of cumulative use). Similar incredulous find-
ings have been seen in all 14 of the Interphone brain
tumour studies published to date. The authors should
be congratulated, as they are the first to directly com-
ment on these findings of protection. All previous
papers have only reported the OR and CIs in the con-
text of finding ‘no risk’.
Conflict of interest: All statements are mine and
mine alone and do not represent positions or opinions
of the Central Brain Tumor Registry of the United
States.
References
1
Lahkola A, Salminen T, Raitanen J et al. Meningioma and
mobile phone use—a collaborative case-control study in
five North European countries. Int J Epidemiol 2008;37:
1304–13.
2
Lo¨nn S, Ahlbom A, Hall P et al. Long-term mobile phone
use and brain tumor risk. Am J Epidemiol 2005;161:
526–35.
3
Collatz Christensen H, Schu¨z J, Kosteljanetz M et al.
Cellular telephones and risk for brain tumors, a popula-
tion-based, incident case-control study. Neurology 2005;64:
1189–95.
4
Klaeboe L, Blaasaas KG, Tynes T. Use of mobile phones in
Norway and risk of intracranial tumours. Eur J Cancer
Prev 2007;16:158–64.
5
Schoemaker MJ, Swerdlow AJ, Ahlbom A et al. Mobile
phone use and risk of acoustic neuroma: results of the
Interphone case–control study in five North European
countries. Br J Cancer 2005;93(7):842–48.
6
Lahkola A, Auvinen A, Raitanen J et al. Mobile phone
use and risk of glioma in 5 North European countries.
Int J Cancer 2007;120:1769–75.
7
CBTRUS. Statistical Report: Primary Brain Tumors in the
United States, 2000–2004. Hinsdale, IL, USA: Central
Brain Tumor Registry of the United States, 2008.
doi:10.1093/ije/dyp197
2 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY