Search This Blog

Monday, October 31, 2011

KISS For LNT-Deniers

KISS - Keep It Simple Sparky!

Based on the barrage of feedback I've received by LNT deniers over the last few days, mostly on Atomic Insights, I thought I'd take the KISS approach, addressing many of the fallacies.

Let's avoid numbers and obscure molecular biology for now.  Let's remember that it's almost 2012.

-------------

A dose response curve looks something like:



So let's talk ionizing radiation.  Let's talk radon which is ubiqutious.  It's a radioactive gas that is a decay product of uranium, which is in the ground.  As more and more uranium was mined to support a growing nuclear power industry, we observed an excess cancer rate in the miners.  These folks were exposed to a high concentration of radon (it doesn't matter what the numerical value was).

The BEIR IV (1988) studied this and determined cancer risk coefficients, and found a linear dose curve for those high range of doses.  In the 1990's, as more energy efficient homes were being built, we noted radon gas accumulation in the homes.  This was significantly above background (as measured outside).  BEIR VI (1999) was published, in order to extrapolate the high range observed linear dose response to the low ranges found in homes.

Actual studies (not extrapolations) were performed throughout the 1990's.  The best were integrated in large meta-studies in 2005:


And there is an increased risk of cancer even at low concentrations (doesn't matter what the numbers are).

So we have a strong linear dose relationship for radon decay products (which include a variety of different kinds of emissions - a good surrogate for "low level radioactivity").  From high concentrations actually experienced by humans in real life....to low concentrations, actually experienced by humans in real life.

Ironically, a  2005 French paper, reached a contrary conclusion to the evidence. 

"In conclusion, this report raises doubts on the validity of using LNT for evaluating the
carcinogenic risk of low doses (< 100 mSv) and even more for very low doses (< 10 mSv). The
LNT concept can be a useful pragmatic tool for assessing rules in radioprotection for doses
above 10 mSv; however since it is not based on biological concepts of our current
knowledge, it should not be used without precaution for assessing by extrapolation the risks
associated with low and even more so, with very low doses (< 10 mSv), especially for
benefit-risk assessments imposed on radiologists by the European directive 97-43."

If one calculates the doses that could be experienced by home dwellers, they are around 50 uSv.

Now, imagine I have magic powers.  I snap my fingers and magically the radon concentrations in all of the homes in the studies are reduced by 50%.    Since today is 2011 and those studies were done in 2005, we don't have a dose response curve going down to the new low levels of radon.  There is a gap there.  Does this mean hormesis exists (where hormesis means that low levels of radioactivity can be beneficial to you)?

Of course not.  To think that, considering the powerful trajectory of the dose-response curve is ridiculous.  It might be possible...but we'd need extraordinary evidence to be convinced.

We have a subgroup of low level radioactivity called LET radiation (beta, gamma).  The dose response of that radiation is discussed in BEIR VII and is based primarily on the long term study of Japanese atomic bomb victims.  We see a linear relationship from the most highly exposed individuals down to those exposed to 100 mSv.  As the study progresses we'll be more confident in our results and we'll be able to discern better below the 100 mSv.  Does the 100 mSv statistical limitation mean hormesis exists below that level?

Of course not!  We have this powerful trajectory going on.  We need strong evidence of hormesis before we accept it.  And note, that part of the reason we have this 100 mSv barrier, is because we want high quality studies.  We are enforcing a 95% confidence level.  If we allowed an 80% CL, we could go even lower. 

I"m not 95% confident that you're wearing underwear, but you probably are. 

We can still retain a sense of reality, even when we're not at 95% certainty.  A trajectory doesn't magically change directions without a good reason.  And no good reasons have been given.

10 comments:

  1. "If one calculates the doses that could be experienced by home dwellers, they are around 50 uSv."

    This doesn't seem quite right.

    UNSCEAR settles on a rate of 9nSv per Bq/h/m3.

    (para 92, p220)
    http://www.unscear.org/docs/reports/2006/09-81160_Report_Annex_E_2006_Web.pdf

    It says average global radon indoors is around 40 Bq/m3

    It says average dose from radon per year is around 1.5mSv.

    The US study covered a period of 25 years, and examined concentrations of 100 Bq/m3 (2.5 times the average). Abstract:
    http://journals.lww.com/epidem/Abstract/2005/03000/Residential_Radon_and_Risk_of_Lung_Cancer__A.1.aspx

    So the extra dose for these people would have been around 2mSv per year, or 50mSv over the period or 25 years.

    50mSv not 50uSv.

    ReplyDelete
  2. Hi Colin G.

    Actually I did the math quickly when responding to a comment from an earlier post, and rounded up here to 50 uSv. I do think I had hit my calculator wrong, though.

    Please double check me (I haven't done this sort of thing in awhile).

    I just calculated 6 mSv.

    I tried to avoid the gross averaging that some of the stuff you site necessitates.

    The study link gives the residential concentration at 3 pCi/L. That was the concentration I used. I wanted the low end, not the average, so we can calculate how low of a dose we can see (we shouldn't fixate on numbers...the larger point is that we see LNT under the conditions people are exposed to, whatever the doses are).

    I used a DCF of 9 mSv/WLM. There are 170 hours at 100 pCi/L per WLM. I assumed a person was at home 10 hours per day everyday of the year...3,650 hours. This way, one could shift the dose estimate based on how many hours you want to assume someone is home per day.

    This should be helpful and sources the dose conversion factor:

    http://www.wise-uranium.org/rdcrnh.html

    ReplyDelete
  3. Hiroshima and Nagasaki survivors tell you almost nothing about high dose, low dose-rate radiation.

    Hiroshima and Nagasaki survivors recieved >100 mSv in ~20 seconds; with the bulk of that being prompt gammas and x-rays in the first few microseconds, neutrons in the first few ms and the rest being gammas from extremely short-lived fission products in the rapidly rising fireball.

    That's a phenomenal dose rate and absolutely overwhelms cellular repair mechanisms.

    You actually have to go and study people or animals who have recieved large doses at low dose rates if you are interested in verifying LNT for low-dose rates.

    You could for instance look at the radium dial painters for a natural experiment on humans in the low dose-rate regime. 64 bone sarcoma cancers occured in 264 cases who recieved more than 10 Gy estimated dose and zero cases of bone sarcoma occured in 2119 people who recieved less than 10 Gy.

    ReplyDelete
  4. I was using an annual dose, not lifetime dose because in the earlier comment section I had discussed something about hours/year.

    That is another source of confusion...so we are closer. 6 mSv/year times x 50 years = 300 mSv.

    But again, it's not the number that matters.

    ReplyDelete
  5. @ Soylent

    Not all those exposed in Japan were in the high dose range...some received low doses. The initial blast was high dose rate, but the residual radiation was intermediate/low dose rate.

    It's not like the study only looks at people who got very high doses and mathematically downgrades them (though there is some of that).

    Chernobyl thyroid cancers show a "strong, approximately linear dose-response relationship."

    The leukemia studies also show excess risk:

    http://chernobyl.cancer.gov/studies.html

    ReplyDelete
  6. Ok, the radon study shows an excess lung cancer risk of 11% for a radon concentration of 100 Bq/m3.

    If the dose response is proportional and linear, what should we expect in an area where the radon concentration is 1000Bq/m3? Everybody gets lung cancer?

    I can see how, theoretically, a single track having the ability to cause cancer means that an increasing number of potential tracks (a higher dose) leads to higher risk. But is this a straight line? Surely no matter now much the probability increases it should never reach 100%.

    (The more dice you roll, the more chances you have of rolling a 6; but no matter how many dice you roll you can never guarantee rolling a 6.)

    Surely the dose response graph should be a curve rather than a straight line?

    ReplyDelete
  7. No, not at all.

    It means that you look at that particular graph (which we don't have) and you see what the cancer risk was for those at the 1000 Bq/m3 concentration.

    The graph will have a certain slope, though I don't know what it is.

    This is why we need to be careful about employing collective dose.

    The slope will be negative (high on the right hand side & low on the left).

    But different studies, which look at different populations under different circumstances, have different slopes.

    If you have an apples-apples situation you can use a particular study and map it to what you are considering. But usually, it isn't apples-apples.

    ReplyDelete
  8. Perhaps I have misunderstood the graph at the top of your post.

    If we know any point on the graph, and we know that it is a straight line, and we "know" it passes through zero, then we can predict all the other points including the point where the dose response reaches 100% mortality. Surely?

    Or is the LNT graph not a straight line? Or does it not pass through zero?

    ReplyDelete
  9. The graph at the top of page is idealized. Life isn't so straight-forward.

    Here is actual graph based on Japanese colon dose:

    http://www.nap.edu/openbook.php?record_id=11340&page=145

    Overall, it's clearly linear. Notice how small region below 0.1 Sv is. Someone would need some powerful evidence to refute LNT in that smudge.

    If the population you are dealing with highly resembles the Japanese population, you could use that graph for colon dose. But likely, a population you are dealing with has many different features... diet, drugs, etc.

    These would introduce errors.

    So, it is more accurate to do another study with your population if you want accuracy....but studies cost money and take time.

    ReplyDelete
  10. Note the homepage is here:

    http://ribjoint.blogspot.com/

    ReplyDelete