Time for a fresh look at low-dose radiation

By Craig Piercy, ANS Washington Representative

“How many angels can dance on the head of a pin?”

We will never know whether Thomas Aquinas actually pondered this specific question during the Middle Ages, but today the metaphor has shed its religious overtones to be a general warning against engaging in protracted debate on existential issues while more urgent problems fester.

It is also the first thing that comes to my mind when I think of the linear no-threshold (LNT) vs. hormesis debate.

As scientific theories go, LNT has always been weak. It relies on extrapolation from high-dose radiation exposure data, where the effects are clear, to predict the risk of annual doses below 10 rem, where there is simply no epidemiological evidence of increased cancer risk in human populations. However, with the weight of 60 years of government policy and our societal tendency to “err on the side of caution” behind it, dethroning the LNT will always be a formidable task.

Meanwhile, the science of radiation hormesis gets more interesting by the day. Just recently, I learned about the abscopal effect–a well-documented medical phenomenon in which treating a cancerous tumor with radiation creates a systemic immune response in the body that attacks other metastases, even those untouched by radiation. An amazing and medically beneficial effect of radiation! Adherents of hormesis need to be careful about how they communicate, however. Yes, there is a growing body of evidence for the benefits of low-dose radiation exposure, but the idea that “a gamma a day keeps the doctor away” can sound a little . . . well . . . crazy, especially if the pitch includes a conspiracy theory about how the Rockefellers quashed data in the 1950s.

Let’s call a truce in the radiation science war. We should all be able to agree that the average person’s cancer risk from an added 50–100 mrem of annual radiation exposure is vanishingly small—“not statistically different from zero,” in the words of the Health Physics Society—otherwise we would have evacuated Vail, Colorado, a long time ago.

The real crime is how our radiation standards and policies cause us to misallocate public resources on a colossal scale. Each year, roughly 1.7 million Americans will be diagnosed with cancer. Roughly one-quarter of them will experience financial distress because they cannot fully afford prescribed treatment in our current health care system. Yet, the Environmental Protection Agency quietly issued a memo in 2014 that set the Superfund radiation cleanup standard at 12 mrem instead of 15 mrem per year, basically adding $100 billion—roughly $58,000 per new U.S. cancer patient last year—to the estimated cleanup tab for Department of Energy defense sites, with no proof it would save a single life.

Now let’s consider the threat of unchecked climate change: thousands, if not millions, of lives lost to extreme heat, violent storms, famine, vector-borne disease, and water-borne illness if we don’t bend the emissions curve in the decades to come. That’s a worst-case scenario, of course, but can we agree that it’s within the realm of plausibility? Can we really allow U.S. nuclear innovation to fail because we are afraid that our waste disposal method might expose someone to a CT scan’s worth of radiation 1 million years from now, when presumably cancer will be a thing of the past? There comes a point when intergenerational equity becomes our generation’s stupidity.

We don’t need to solve the LNT vs. hormesis controversy to prove that our national policies on radiation are flawed. It’s time to stop counting angels, fix the problems that are right in front of our face, and start fresh with a public discourse on the issue.—cpiercy@ans.org

(Reprinted with permission from ANS News, July/August 2019, p. 5)


Craig PiercyCraig Piercy has represented ANS in Washington, D.C., for the past 13 years. He works with Congress and Executive Branch officials to advocate for policies that support nuclear science and technology.


 

2 thoughts on “Time for a fresh look at low-dose radiation

  1. William Schenewerk

    CO2 Mitigation Using Atomic Power-2025 Deployment
    Abstract
    CRISPR gene editing will result in 15 billions because human life expectancy will nearly double. This means world energy will continue increasing 2.25%/a through 2100, requiring ~40 terawatts-electric average power generation by 2100, equivalent to 120 terawatts-thermal energy. Sufficient utility-scale energy storage to average 40 terawatts wind and solar energy, ~2 terawatt-a, costing ~2000 trillion USD at 0.10 USD/Wh, will never exist. Wind and solar energy collection cost for ~500 terawatt “electrical” nameplate will add another ~1500 trillion USD, not counting ~500 trillion USD transmission cost. Absent utility-scale energy storage, wind, solar and big hydro will never average more than 2 terawatts electric generation. Atomic power must expand 5%/a from 2020 to 50 TWe nameplate. Otherwise atmospheric CO2 will increase to multiples of preindustrial. Atomic power can be any combination of: (1) seawater-fueled LWR, (2) FBR, or (3) D2O slow-neutron pile. Sufficient D2O will be available from electrolysis and fuel-cell hydrogen consumption. 8.5 terawatt continuous nameplate power is will be needed to pump 15,000 km^3/a water south from Canada and Russia. Atmospheric CO2 modeling assumes ocean continues absorbing 1/3 of industrial CO2 emissions. Fossil fuel is modeled as gasoline, C8H18. Maximum CO2 is ~700 ppm around 2100. After fossil fuel is phased out, CO2/CH4 (GHG) atmospheric half-life is estimated 60 a, resulting in 350 ppm CO2 around 2250. A 1000 year CO2/CH4 half-life is typically used in the literature.
    2.1 LNT Does Not Agree with Nagasaki Bomb Data
    Mortality/1000-person-a arising from malignant neoplasms at Hiroshima and Nagasaki is presented 1950-1974 [13]. Except at zero exposure, Hiroshima Data is different from Nagasaki data, even at extreme of error bars, which generally do not overlap. In the instance of breast cancer, time delay is greater than 10 a [14]. Low Dose analysis requires curve fitting[14]: “Disagreement about the somatic risks from low doses of ionizing radiation stem from two difficulties fundamental to the logic of inference from observational data. First, precise direct estimation of small risks requires impractably large samples. Second, precise estimates of low-dose risks based on high-dose data, for which the sample size requirements are more easily satisfied, must depend heavily on assumptions about the shape of the dose-response curve, even when only a few of the parameters of the theoretical form of the curve are unknown.” There is also the issue that cancer mortality is constantly dropping, with leukemia mortality being half what it was in 1945. Paraphrased: 1.0 Sv mammograms (normal mammogram does is ~0.004 Sv) would have to be given to more than 10,000 women, all the same age, to have a statistically significant test of LNT, and the result would only be available after 2035, 15 years before CO2 doubles.
    2.1.1 Hiroshima Bomb Data
    Hiroshima annual mortality curve fit from [13] data, including neutrons versus exposure, 0 to 280 rad (0 to 2.8 Gy):

    Hiroshima Annual Mortality = 2.0/1000-a + 0.009/1000-rad-a * 100 rad/Sv
    = 0.002/a + .0009/Sv-a * Sv = annual risk per person. (01)

    Using Hiroshima data, each Sv dose increases annual cancer death rate by ~1/2.

    Hiroshima lifetime risk = 0.0009/Sv-a * 70 a/life = 0.063/Sv-life. (02)

    ICRP Publication 103, 2007, LNT whole risk coefficient is 0.057/Sv-life [15]. Result: Curve fit to Hiroshima bomb data, 0 to 2.8 Sv, supports LNT. At 350 rads, Hiroshima mortality drops to ~3.5/1000-a, then rises to 5/1000-a at 550 rad. Hiroshima bomb had neutrons which are difficult to convert to Sv. Hiroshima annual cancer mortality at 1.0 Sv is 0.0029/a, 1.45 times zero radiation exposure rate.
    2.1.2 Nagasaki Bomb Data
    Nagasaki bomb curve fit from [13] data is significantly different from Hiroshima bomb data. Nagasaki Annual Mortality, (0 to 5.5 Gy):

    Nagasaki Annual Mortality = 0.0018/a + 0.000006/Sv^2-a * (100 Rad/Sv)^2/1000-a
    = 0.0018/a + 0.00006/Sv^2-a * Sv^2 = annual risk per person. (03)

    Using Nagasaki data, each Sv dose increases annual cancer death rate by ~3%. Nagasaki annual cancer mortality at 1.0 Sv is 0.00186/a, 1.03 times zero radiation exposure rate. At 1 Sv, Hiroshima elevated mortality rate is more than 10 times Nagasaki rate. Similar Hiroshma versus Nagasaki mortality results is obtained from 1980-CONAES [16], which is a good starting point for any energy research.
    [13] Alvin M. Weinberg, “The Future of Nuclear Energy,” I:Physics Today, Volume 34, Number 3, p48, doai: 10.1063/1.2914469, http://ds.doi.org/10.1063/1.2914469, (1980).
    [14] Charles E. Land, Health statistician, Environmental Epidemiology Branch of the National Cancer Institute, Bethsda, Maryland, 20205, “Estimating Cancer Risks from Low Doses of Ionizing Radiation,” I: Science, B: 209, pp. 1197 – 1199, (September 9, 1980).
    [15] Radiation Protection Recommendations, I:ICRP Publication 103, Nominal risk coefficients, 10^-2/Sv for stochastic effects after low-dose exposure, Annex A, (2007).
    [16] Harvey Brooks and Edward L. Ginzton, co-Chairman, “Energy in Transition , 1985-2010: Final Report of the Committee on Nuclear and Alternative Energy Systems, Committee on Nuclear and Alternative Energy Systems, National Research Council, ISBN: 0-309-66780-1, http://www.nap.edu/catalog/11771.html, (1980).

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>