Re: Alpha radiation as tests of LNT

[David Scherer <scherer@uiuc.edu>]

At 10:08 AM 7/30/96 -0500, you wrote:
>	According to most theories, linear - no threshold should be more 
>valid for alpha particles than for beta or gamma rays because it is much 
>more likely that a single alpha particle can provide the two hits 
>required. For beta and gamma, two separate particles providing the two 
>hits might be more important, in which case you would get a quadratic 
>rather than a linear response. The reason failures are reported for 
>alphas is that they are much easier to detect: according to linear -no 
>threshold, natural background gamma rays are responsible for only 1% of 
>all cancers, but radon in homes is responsible for 10% of all lung 
>cancers. Also, radon in homes is very much more variable than natural 
>gamma ray background.
>
I understand the variation and its importance in probing the theory.  How do
you sort out all the parameters in your tests of the theory?  Given values
for Q (or  w-sub R) and cancer incidence per radiation, it has been
estimated that  10% of lung ca is caused by Rn.  If the results fail, how
can you say the problem is with the low LET risk coefficient risk
coefficient rather than the Q?

It has been a long time since I've looked into microdosimetry, so please
bear with me.  Ra, Rn, etc. alphas have energies of 5-7 MeV, with dE/dx
around 100 keV/um.  The smallest dimension of the DNA molecule is about 2
nm.  So the alpha deposits something greater than around 200 eV in the
vicinity of the DNA molecule.  Far more than enough for two hits.  Perhaps
the insult is often great enough to kill the cell rather than transform it.
Wouldn't this affect risk linearity for high LET (specifically alpha)
radiation, but be silent for LET?  I'm sure this has looked into before.

Regards,

Dave Scherer
scherer@uiuc.edu