Re: FGR-13 -Reply

[Charles Willis <CAW@nrc.gov>]

Gene,

It may sound evasive, but risk factors actually refer to what the authors intend for
them to address.  Some risk factors, e.g. those in BEIR-5, refer to cancer mortality.
Other risk factors refer to cancer mortality plus serious genetic effects.  Others,
such as ICRP-60, refer to cancer mortality plus serious genetic defects plus a
weighted factor for cancer morbidity (i.e. being sick from cancer).   There is no
standard in this area.  

Also, there is a lack of a complete set of data on any of the risks.  The best we have
are the A-bomb survivor data and they are incomplete (1) because at the time of
the last RERF report (1990) most of  that population was still alive and (2) the data
collection did not start until a few years after the exposures so some leukemias
probably were missed.   The RERF experts make educated guesses about what will
happen to those still alive and thus produce mortality risk factors.  Also there are
incidence data from the A-bomb survivors but these data sets are less complete
than the mortality data.   There are no positive data on human genetic effects; all
the data show is that the A-bomb survivor sample is to small to show genetic
effects in humans; chromosome abortions can be seen but not deleterious health
effects.   Risk estimates largely are based on mouse data and the linear, no
threshold model (which is a bit hard to square with the observation that doses to
female mice have no discernable genetic effect if the doses are delivered 6 or more
weeks before conception).

Most of  the human mortality data are obtained from death certificates, although
there were a number of autopsy checks.  Incidence data, of course is obtained from
cancer registries (without autopsy validation, of course).

This gets to be a big subject.  Hope this bit of it is of some use.

Charlie Willis
caw@nrc.gov