Intended for healthcare professionals

CCBYNC Open access

Rapid response to:


Use of mobile phones and risk of brain tumours: update of Danish cohort study

BMJ 2011; 343 doi: (Published 20 October 2011) Cite this as: BMJ 2011;343:d6387

Rapid Response:

Re: Use of mobile phones and risk of brain tumours: update of Danish cohort study

I am concerned about the validity of the reported outcomes of the
Danish Cohort Study on the relationship of cell phones to cancer that
was published in the British Medical Journal this year. I know your
journal is highly respected worldwide and that people expect utmost
care is taken by peer reviewers and editors in deciding what
constitutes a well-designed and executed study, and what study
conclusions reasonably follow. I believe on both counts the Danish
Cohort Study fails and that this study should be retracted from BMJ.

The study, with a cohort of over 723,000 people, concludes that there
is no causal link between brain cancer and cell phone radiation, as
it is reported in your Journal. This conclusion has been highly

It is, however, highly questionable.

Since cell phone radiation is the "agent " being studied, in order to
determine whether there is a link between exposure and cases, one
would assume the study clearly delineated significantly-to-heavily
exposed people from light-to-non-exposed people. But that is not
the case.

Only length of cell phone subscription was considered. Length of
subscription? Certainly you, as I, have known people who have
maintained a cell phone, but only use it for emergencies. Back when
cell phone minutes were quite expensive (which was during the period
studied, 1982 to 1995), there would have been all the more reason to
be spare in use. But these people, if they owned a cell phone, were
put into the same "exposed" pool with those who used it heavily! How
does that tell us anything useful about the exposure link to cancer?

And who would have been the heaviest users in those times ? Most
likely, those who depended on using the cell phone for their jobs,
and whose cell phone minutes were being paid for by the corporations
they worked for. Does that not make sense? Wouldn't you want to
know what happened to those heavily exposed users? And yet, over
200,000 of these very people were excluded from the study because
they did not have individual subscriptions. They, ironically, became
misclassified as "unexposed".

You must agree that these practices contaminate the groups, and thus,
call the conclusions into serious question. Research Design 101---
this is basic.

And what about people who started using cell phones in 1996--the year
after the subscription cut-off date? They could have used the cell
phone for eleven years, been diagnosed with cancer in 2007 (the last
year for monitoring outcomes), and yet been considered "unexposed."

The amount of confounding is very surprising for a study published in
BMJ. The implications for continuing to lend this study validity by
not retracting it, could be very serious for the Journal, and even
more importantly, for the people who rely on the educated judgment
and expertise of its editors and reviewers when making decisions for
their family, or their nation. What you do makes a difference.

I call upon you to retract the Danish Cohort Study, and to do so in a
way that informs the public of the change.


Margaret Meade Glaser

Competing interests: No competing interests

20 January 2012
Margaret Meade Glaser