[I always open “CC” (closed captions), if it’s available, to help me understand what speakers are saying.]
To refresh everyone’s memory about the “LNT Model,” recall that GEITP has discussed this topic at least a half-dozen times. Ed Calabrese, emeritus in environmental health sciences at University of Massachusetts Amherst, has authored and coauthored more than two dozen papers — starting about 2011 — on questioning the scientific conclusions of the fruit fly experiments leading up to the infamous Linear Non-Threshold (LNT) Model for risk assessment (originally) for ionizing radiation exposure. The LNT Model arose from the 1946 Nobel Prize awarded to Hermann Mueller — and was adopted by the U.S. Environmental Protection Agency (EPA) and many other agencies. And, by extrapolation from ionizing radiation, it was proposed that this same model likely holds true for most (if not all) chemical toxicants and carcinogens.
In his Nobel Prize Lecture on December 12, 1946, Hermann J. Muller argued that the dose-response for ionizing radiation-induced germ cell mutations was linear and that there was “no escape from the conclusion that there is no threshold.” However, a recently discovered (by Calabrese in 2015) commentary by Robert L. Brent indicated that Curt Stern, after reading a draft of part of Muller’s Nobel Prize Lecture, telephoned Muller — strongly advising him to remove reference to the flawed linear non-threshold (LNT)-supportive Ray-Chaudhuri findings and strongly encouraged him to be guided by the threshold supportive data of Ernst Caspari [who, incidentally, was DwN’s genetics mentor in college]. Brent wrote that Mueller refused to follow Stern’s advice, thereby proclaiming (during his Nobel Prize Lecture) support for the LNT dose-response, while withholding evidence that was contrary. This finding is of historical importance, because Muller’s Nobel Prize Lecture then gained considerable international attention in 1946 and was a turning point in acceptance of the linearity model for radiation and chemical hereditary and carcinogen risk assessment.
Ed asserts that the science used — to support the LNT model adopted by the NAS’s 1956 Biological Effects of Atomic Radiation (BEAR) I Genetics Panel — was also tainted by its leaders, who he says deliberately refused to include evidence from NAS’s own Atomic Bomb Casualty Commission (ABCC) human genetic study, also known as the Neel and Schull 1956a report.
Calabrese says Neel and Schull showed “an absence of genetic damage in offspring of atomic bomb survivors, which supports a threshold model (i.e. not the LNT model),” but this was not considered for evaluation by the genetics panel, “therefore, those data could not figure into its decision to recommend the erroneous LNT dose-response model for risk assessment.”
Calabrese suggests that the panel’s work was undermined by Hermann J. Muller and BEAR I chairman Warren Weaver, who “feared that human genetic studies would expose the limitations of extrapolating from animal, especially the fruit fly, Drosophila, to human responses, and admitting this would strongly shift research investments/academic grants from animal to human studies.”
Calabrese adds, “The country expects its scientists to be honest and to follow real data. The BEAR 1 Genetics Panel failed on both counts, being loyal only to their ideology, and then hiding it. They were hailed by many media outlets as the Genetics Dream Team — giving them ‘further cover’ so that their deceptions would never be known. They have gotten away with it, so far, for 68 years.” How much longer will we believe that “inaccurate scientific conclusions do not lead to bad government policy” — which continues to be practiced, today…??
“This history should represent a profound embarrassment to the U.S. NAS, regulatory agencies worldwide — and especially to the U.S. EPA and the risk-assessment community — whose founding principles were so ideologically determined and accepted with little, if any, critical reflection.”
DwN
Number of the Week: “100,000 times greater (than background).” According to Ed Calabrese, Hermann Mueller knew of Caspari’s University of Rochester study — before he gave his Nobel talk in December 1946.
“That [fruit fly] study had shown that at the ‘low’ chronic radiation dose rate (i.e., yet still about 100,000 times greater than background), no radiation-induced mutation effects were found. The Caspari study supported the threshold model, and not the LNT model.”
—DwN
COMMENT:
This 4-minute youtube video [https://www.youtube.com/watch?v=T3IwCjWytiA] is excellent. Very clear, lucid, crisp and brief — just what the busy scientist wants. The bottom line is that the Linear Non-Threshold (LNT) Model is not founded on scientific facts. Instead, wherever possible, we should base all governmental policies on scientific, evolutionary-based data.
Let’s hope that powerful leaders in the U.S. EPA, and other regulatory agencies are listening. ☹
—DwN
Dan,
Please watch this 4-minute video summary by John Cardarelli (President, Health Physics Society) on the LNT story. It is impressive and should be shared widely. Most people are more prone to watch a short (but powerful) conceptual summary than delving into many dozens or hundreds of pages of text, describing this fraud…
Ed,
MORE COMMENTS:
The Radiation Exposure Compensation Act ( “the Act” or “RECA” first passed in Oct 1990)* offers an apology and monetary compensation to individuals who contracted certain cancers and other serious diseases as a result of their exposure to radiation released during above-ground nuclear weapons tests or as a result of their occupational exposure.
This program involves scientists at CDC who attempt to quantify exposures. These estimated exposures are then incorporated within a LNT risk assessment framework to provide a “scientific” basis for “how much compensation may be received”. This same concept is said to possibly occur for people exposed to the Manhattan Project radiation in the recent controversies in St. Louis, MO — involving a school and related areas. Again, this would be driven by the LNT assumption.
There is the built-in assumption of dose-response causality in the low-dose zone.
Even though there is a 25-30% chance of any of us developing cancer over a lifetime within our society, the issue of background radiation appears to be ignored in these cases. The decision to pay the compensation is a political, rather than a scientific one. In this case, the decision to compensate cancer patients — rests with the US Congress.
It may prove to be interesting to provide a listing of how the LNT concept is being applied in many areas within the US and elsewhere. And then to calculate the (excessive and nonscientific estimated) costs to society.
Ed Calabrese
*By a 69-30 vote, the Senate passed legislation just last week that would expand the types of atmospheric nuclear testing and nuclear material exposure covered by the Radiation Exposure Compensation Act as well as the time window for damage claims. Specifically, people exposed to atmospheric nuclear tests in Nevada, and in the Pacific from 1958 to 1962 would now be eligible for damage claims.
The current law limits claims relating to atmospheric testing to parts of Arizona, Nevada, and Utah; the bill would expand that area to include all of those states as well as Colorado, Idaho, Montana, New Mexico, Utah, and Guam. The bill also expands the list of uranium mining-related jobs that qualify for compensation, notably adding workers involved in remediation efforts at uranium mines and mills.
It extends the coverage to include mines that operated through 1990, instead of through 1971. In addition, the bill creates a new claim eligibility for people affected by Manhattan Project waste in Missouri, Tennessee, Alaska, and Kentucky.
The compensation fund itself was set to expire in June, following a previous extension that Congress passed in 2022. The new legislation would extend the fund to 2030. The bill is sponsored by Sen. Josh Hawley (R-MO), who has frequently drawn attention to the enduring effects of radiological contamination in his state. The Biden administration has endorsed the legislation.
MORE COMMENTS:
Just a reminder that we generated some data with respect to chemical carcinogenesis where “the LNT model does not hold”. ☹
Assessment of human cancer risk from animal carcinogen studies is severely limited by inadequate experimental data at environmentally relevant exposures and by procedures requiring modeled extrapolations that are many orders of magnitude below observable data. We used rainbow trout, an animal model well-suited to ultralow-dose carcinogenesis research, to explore dose-response down to a targeted 10-fold excess liver tumors per 10,000 animals (ED001). [“The ED50”, e.g., is the dose/exposure of anything that produces a specific effect in 50% of the population that has been administered that dose/exposure.]
A total of 40,800 trout were fed zero to 225 ppm dibenzo[a,l]pyrene (DBP) for 4 weeks, sampled for biomarker analyses, and then returned to a control diet for 9 months prior to gross and histologic examination. Suspect tumors were confirmed by pathology, and resulting incidences were modeled and compared to the default EPA LED10 linear extrapolation method. The study provided observed incidence data down to two above-background liver tumors per 10000 animals at the lowest dose i.e., an unmodeled ED0002
measurement). [The Linear Extrapolation Model uses a linear equation to predict future outcomes. This method is best suited for predictions close to the given data. One simply draws a tangent line from the last point to extend it beyond the last known value.]
Among nine statistical models explored, three were determined to fit the liver data well — linear probit, quadratic logit, and Ryzin-Rai. None of these fitted models is compatible with the LED10 default assumption, and all fell increasingly below the default extrapolation with decreasing DBP dose. Low-dose tumor response was also not predictable from hepatic DBP-DNA adduct biomarkers, which accumulated as a power function of dose (adducts = 100 x DBP1.31).
Two-order extrapolations below the modeled tumor data — predicted DBP doses producing one excess cancer per million individuals (ED10(-6)) that were 500- to 1500-fold higher than that predicted by the five-order LED10 extrapolation. These results are considered specific to the animal model, carcinogen, and protocol used. However, they provide the first experimental estimation in any model of the degree of conservatism that may exist for the EPA default linear assumption for a genotoxic carcinogen.
Reply: Thanks, Dave. I remember this study well — largely because of the huge number of animals (40,800 trout) that George Bailey and coworkers had used.
MORE COMMENTS:
Dan, Regarding the EPA and “hot showers,” this will be a long story, but it is my history, so please bear with me!
I have always believed that there is a threshold for any toxicant, be it physical or chemical. The dose makes the poison.
When I was a graduate student at NYU we used “whole-body” radiation counters. I used one to study mucociliary clearance, in which subjects would inhale technetium-labeled monodispersed ferric oxide particles. (We had to evaluate the risk of such an experiment and compare it to the risk of chest X-rays. So, this was a benchmark or “threshold” that was considered a minimal risk. We also considered the dose someone would receive if they were living in Denver, or during an airplane flight.)
We had a very good whole-body chamber to block out background radiation, so we could use the lowest dose possible. It was a gun turret from a pre-World War II battleship. The steel was generated before above-ground testing and was several inches thick.
In NYU’s Sterling Forest Laboratory, they built a new chamber to study the disposition of radiation in baboons, and to measure radiation exposure in humans. They improved the chamber to have extremely low background.
So — to get started, they wanted to test as many control subjects as possible, to get a baseline for studies of people with known exposure. In those days, everyone wanted to help, and we all volunteered to sit in the chamber, which took several hours. The graduate students volunteered too, doing their homework and reading manuscripts during the counting. Well, guess what. The graduate students who lived down the road from the lab in student housing, had very high levels of radiation. Levels were higher in the morning tests than in the afternoon tests. This was weird: were the students exposed by drinking beer, water, or smoking pot? It turned out that none of this could explain the difference.
After careful investigation, it was determined that that it was because they were exposed to water during their morning shower, which came from local wells. The region was on the Ramapo fault, which leaked radon into the well water. This led to several follow-up studies to measure radon in basements with our newly developed environmental radon detectors. This was a part of what led to home radon mitigation, which is practice today.
MORE COMMENTS:
Since you are sometimes an expert on risk assessment, here is a slightly related question:
Do you know of any economic study calculating the EXTRA COSTS of clean-up and toxicology prevention, if one accepts the LNT Model, compared with that of a Threshold Model…??
Dan, I’m not aware of any economic analysis review article, but I’m sure the difference (between LNT and Threshold models) would be huge.
I can think of one good example where economics/common sense exceeded LNT: in 2000-01, I served on an NAS panel that evaluated the 50-ppb (parts per billion) drinking water standard for arsenic that had been in place for decades. To make a long story short, 50 ppb was clearly not protective of public health, because there were many studies in arsenic-endemic areas of the world where levels of >200 ppb were clearly associated with substantial increased bladder and lung cancer risks (in people, not rats). If one used the LNT, it is likely that the 1 in 100,000 risk level would have been way less than 1 ppb.
In the end, the EPA adopted a new standard of 10 ppb, because only a small fraction of public water supplies was above 10 ppb. If they had chosen 1 ppb (which still had a theoretical risk of >1 in 10,000), thousands of public water supplies would have required expensive treatment systems. I had the pleasure of briefing Kristie Whitman herself, who was then the EPA Commissioner, in Washington DC at 5 pm on Sept. 10, 2001. You recall what happened the next morning in Washington DC.
Needless to say, the NAS arsenic report didn’t get a whole lot of publicit the next day ☹. The briefing team was just starting to brief a Congressional Committee in the Old Executive Office Building next to the White House, when someone ran into the room screaming that everyone needed to evacuate the building immediately. A few minutes later, a plane crashed into the Pentagon — but people in the know were understandably concerned that it was headed for the White House, not the Pentagon.
MORE COMMENTS:
Dan,
First: My thoughts on the US EPA are generally positive. I think one must be cautious in lumping everything the EPA does today with the faulty reasoning that was used in the past when some risk assessments used the LNT model. My experience of 40+ years, has led me to believe the air pollution standards are based on sound science. These standards have been updated several times based on new findings, especially epidemiological studies. I think the air pollution standards have had a great public health impact on the quality of life for our country and the world.
Second: Your question about my history. I was admitted to New York University (NYU) Department of Environmental Medicine’s graduate program in 1976.
Roy Albert, MD, had started the lab — which my mentor Morton Lippmann, PhD, took over before I arrived. Roy started his studies on measuring particle clearance by generating an aerosol in a food blender and personally inhaling radioactive material thru a straw. (At least that was the rumor told to me when I joined the lab.)
Mort was trained in Engineering at Cooper Union (BS), and Industrial Hygiene at Harvard (MS); then he worked at the Bureau of Occupational and Safety (here in Cincinnati), and in Environmental Health at NYU (PhD) with Roy.
Mort developed a method to generate monodispersed particles (a spinning top generator) of varying sizes so the regional deposition could be quantified. Another scientist in Germany, Willie Stallhofen, also made many measurements on particle deposition. This is why we know that particle size determines where particles land in the nasal-oral, tracheobronchial, or alveolar region, and that the clearance times vary (oral-nasal = minutes to 0.5 hours, tracheobronchial clearance is 3-24 hours, and alveolar 24 h to weeks).
Before and when I arrived at NYU, there was a US Department of Energy Environmental Measurements Laboratory (376 Hudson Street) in lower Manhattan that I visited. The main radon measurement group there was led by Andreas C. George.
I remember meeting David Sinclair there. He is, by many accounts, the “father” of aerosol science. You might want to look up Mort, Andreas, and David.
George Leikauf, Professor DEPHS, Univ Cincinnati
FYI: This is what AI says about LNT assessments.
“Not all risk assessments use a linear no-threshold (LNT) model. The LNT model assumes that any level of exposure to a toxin carries some degree of risk, with risk increasing linearly as exposure increases. However, this model has been increasingly questioned, particularly in the field of toxicology, due to emerging evidence suggesting that low doses of certain stressors may not follow a linear dose-response relationship and may even have beneficial effects (hormesis).
In many cases, risk assessments aim to consider various factors, including the dose-response relationship, exposure pathways, biological mechanisms, and uncertainty factors, to estimate the potential risks associated with exposure to environmental pollutants and chemicals. Common practices in risk assessment today often involve using more sophisticated models and approaches that take into account dose-response curves other than linear, as well as factors such as variability and uncertainty.
Some common practices in contemporary risk assessment include:
Threshold Models: These models suggest that there may be a threshold below which exposure to a toxin does not pose a significant risk. Below this threshold, the body’s defense mechanisms can effectively handle the stressor without adverse effects.
Non-Linear Models: These models acknowledge that the dose-response relationship may not be linear and that the relationship between dose and response may vary depending on factors such as exposure duration, timing, and individual susceptibility.
Mode of Action Analysis: This approach focuses on understanding the biological mechanisms through which a toxin exerts its effects and considers how these mechanisms may influence the dose-response relationship. It can help identify critical events and inform more accurate risk assessments.
Use of Epidemiological Data: Epidemiological studies provide valuable data on the health effects of environmental exposures in human populations. Incorporating epidemiological evidence into risk assessment can help validate dose-response relationships and improve risk estimates.
Overall, contemporary risk assessment practices aim to integrate multiple lines of evidence, consider uncertainties, and use the most relevant and up-to-date scientific knowledge to inform decision-making regarding public health and environmental protection.”