Appliances and Leukemia

[A.R. Liboff]

A peculiar report appears in the May 1998 issue of Epidemiology (Hatch,
Linet et al, 9: 234-245). The possible correlation of exposure to household
appliances and acute lymphoblastic leukemia (ALL) in children up to 14 was
examined, both pre- and post-natally in nine states for 640 patients and
640 matched controls.

THis report is peculiar for a number of reasons. First, although the
(eleven) authors seem to have been motivated to undertake this study
because of the magnetic fields associated with 60 Hz appliances, they
dismissed whatever positive correlations that were obtained as probably
having little to do with magnetic exposure: "...we think that a causal
relation between magnetic fields from the appliances and [ALL] is
unlikely...". The underlying reason for this conclusion is that such a
finding might seem to contradict a prior study in 1997 by the same group
(Linet et al, New England J. of Medicine, 337: 1-7) which claimed that it
was not possible "to link measured magnetic fields in residences to
leukemia...".

At times, we are all a little guilty of looking at our own work as the last
word, as the most definitive. But somehow I expect more of epidemiologists.
Perhaps it's because they wear their objectivity on their sleeves. They
hardly ever smile or frown for fear that someone will accuse them of
expressing an opinion.

However, judging from the way this group invoked its own prior research in
coming to its conclusions, perhaps they are more human than I thought, as
self-serving as the rest of us.

A more thoughtful reading of their results might have treated their data
set in an independent manner, without resorting to the results of a totally
different study to color their conclusions, particularly in view of the
very different protocols and objectives employed in the earlier work. On
the other hand, there is something to be said for a group seeking the
appearance of consistency in consecutive studies, especially when large
sums of money and time are involved.

Another strange thing about the conclusions from this report is that they
are tied strongly to the premise that for magnetic fields to have any
possible causal relationship to ALL, there must be evidence of a
dose-response. Especially strange is that their prior report on power line
fields seemed to take the opposite point of view, that whatever dose
response was observed, it had little to do their findings. In any
event,this premise has less to do with epidemiology and more to do with
science. It has been pointed out, not only by me, but also by others, that
magnetic intensity as an independent variable is merely one metric among
many others. Investigators looking at the epidemiological aspects of
magnetic fields should rethink the approach that deals with B-fields as if
these will necessarily act in the same manner as a chemical carcinogen,
with magnetic intensity playing a role similar to ppm contamination or
daily milligrams consumed.

There are only three dimensions that comprise the universe of chemical
carcinogens, the formulation, the concentration, and the rate of exposure.
By comparison, the potential number of dimensions for independent B-field
interactions include, among other things, the intensity, the three spatial
directions, the frequency, the various AC/DC combinations, the gradient,
the waveshape, and carrier modulations. For each of these nine variables
there are separate interaction possibilities. If one finds evidence for a
leukemia correlation with magnetic field, it is not correct to discount
this evidence simply because of an "inconsistency in the dose-response
patterns".

A.R. Liboff
Professor
Department of Physics
Oakland University
Rochester, MI 48309 USA
lib...@oakland.edu