Re: linear hypothesis

[Chris Davey <cdavey@med.phys.ualberta.ca>]

On Tue, 17 Oct 1995, David Scherer wrote:

> I think a few points are in order in this discussion.
> 
> (1) The linear hypothesis applies to stochastic risks.  Drug overdoses are
> presumably nonstochastic.  I believe that a linear, non-threshold model is
> used for Pb exposure as the basis of the current drinking water standards,
> so it isn't only radiation.

Presumeably a small enough dose of a drug does so little damage as to not 
be detectable.  Then the body recovers.  That's exactly how radiation 
works.  Hence 100 pain-killers over an entire year do not cause death, 
just as radiation spread over the year would not.  The stochastic element 
of the radiation exposure could be compared to the possibly carcinogenic 
effect of the pain-killer.  At v. low levels, neither can be detected, 
because of statistical limitations (insufficient study group size, too many 
other variables, etc.) but the difference is we don't worry about the 
pain-killer, whereas we baulk at radiation.

And you're right, there are other things that society is paranoid about, 
but I think radiation still wins first prize every time!

> (2)  Public health rules almost never wait for a potential hazard to be
> proven before they are controlled.  When drugs or medical devices are
> developed, they must be proven to be safe, not the other way around.  The
> onus is not on proving radiation risks.

But radiation has not been 'developed', its always been all around us.  
Surely we can accept that the levels present in nature are not hazardous.
 
> (3)  The linear hypothesis simply says that risks are observed to be linear
> at high doses, so they should be presumed to be linear at low doses.  A
> simple statement of the state of knowledge and uncertainity.  It seems to me
> that the main rub is not this principle, but its societal application, the
> ALARA principle.  The debate should be this: How do we apply the ALARA
> principle at very low doses?  I think BRC is a sensible approach, somewhere
> in the 10s of mrem/y range.

BRC is good, but lets keep it high enough, as near as possible to 100 
mrem, so the paranoia is minimised.  The problem with ALARA is when it is 
used to reduce levels below the BRC level, as it already is in some cases 
(if as you say, 10s of mrem/y are BRC).  That does not usually occur with 
other hazards.
 
> (4)  I don't think we should play with scientific interpretation to achieve
> changes in ALARA.  Ultimately this approach is likely to be
> counterproductive, especially if the radiation biologists do not sign on.

Plenty of studies show hormesis.

> Credibility is our only stock in trade.  Environmental groups have it while
> industry does not.  Let's not make matters worse.

How can anyone be credible when they claim that something is hazardous 
all the way down to zero?  The public (me, for instance) does not spend 
each day considering the miniscule (or not so small) hazards of living in an 
industrial society, somehow we need to put radiation in its place amongst 
the other routine, everyday facts of life.

Chris Davey