back to CNR | radiation | rat haus | Index | Search | tree

plain HTML | ASCII text formats

Preventing an Exercise in Self-Defeat:
The Relevance of Medical Radiation to Nuclear Pollution

John W. Gofman, M.D., Ph.D., Professor Emeritus of
Molecular & Cell Biology, Univ. of Calif., Berkeley;
Chairman of Committee for Nuclear Responsibility; April 2000.

  • Part 1 -- An Extraordinary Note, Received Dec. 1999
  • Part 2 -- What Do All Radiation Issues Have in Common?
  • Part 3 -- The Biologically Unnatural Feature
  • Part 4 -- Key Insights on Health-Harm: Xray-Based
  • Part 5 -- Warning: Studies Where Bias Can Enter
  • Part 6 -- Joy over Two Totally Trustworthy Databases
  • Part 7 -- Preventing an Exercise in Self-Defeat

Part 1  o  An Extraordinary Note, Received Dec. 1999 something important

        o  In November 1999, CNR Books published my study entitled Radiation from Medical Procedures ... (details in the box, below). In December 1999, I received an extraordinary note from a scientist who is a leading figure in the effort to contain or recapture the radioactive poisons already created by the production of nuclear electricity, and created by the production of nuclear weapons. His message, stripped to its bone and re-stated, was:

        o  "That study of medical radiation has tied you up for quite a while. Now that you've finished it at last, we certainly hope that you'll return as soon as possible to the really important work of analyzing the latest data from the A-Bomb Survivor Study!"

        o  His assumption was that a study of medical xrays must be irrelevant to nuclear pollution, and that his mission depends on studies like the A-Bomb Survivor Study.

        o  Because his assumption (which we have encountered more than once) will be self-defeating for the important work against nuclear pollution, and because that work must not be defeated, I must express my worries about that kind of thinking.

Radiation from Medical Procedures in the Pathogenesis of Cancer and Ischemic Heart Disease:
Dose-Response Studies with Physicians per 100,000 Population
by John W. Gofman. November 1999. Library of Congress 99-045096.
Available from online booksellers, library distributors, and the publisher. 699 pages.
Hardcover $35, ISBN 0-932682-97-9.      Softcover $27, ISBN 0-932682-98-7.

Part 2  o  What Do All Radiation Issues Have in Common?

        o  There is one thing which all radiation issues have in common, whether your own current interest is:

        o  The unifying concern is the harm to health, to present and future generations, which results from exposure to ionizing radiation --- regardless of the particular path of exposure.


Part 3  o  The Biologically Unnatural Feature

        o  Ionizing radiation, whether it be from xrays, gamma rays, beta particles, alpha particles, or other forms, has a unique and biologically unnatural feature.

        o  The common feature is this: Biological damage from ionizing radiation is always initiated by the travel through cells of a high-energy high-speed charged particle (the "primary" particle) --- something which simply does not occur within the natural biochemistry of cellular operations.

        o  Along its track, the primary particle kicks additional particles ("secondary electrons") into traveling too --- which also does not happen in natural biochemistry. As the primary and secondary particles "shed" their energy onto other substances, they slow down. In general, the average distance which the particles travel, between energy deposits, becomes shorter.

        o  These energy deposits are comparable to small bombs or grenades, causing local havoc inside human cells. As the distance between these grenades becomes shorter, the frequency rises of complex, non-repairable injuries to the genetic DNA molecules and chromosomes in the irradiated cells. The non-repaired injuries have the causal role in the subsequent cases of Cancer and Coronary Heart Disease.


Part 4  o  Key Insights on Health-Harm: Xray-Based

        o  Because all types of ionizing radiation do their damage in the same way (Part 3, above), insights gained from the study of people exposed to one type of ionizing radiation apply also to the other types of radiation (with minor modifications).

        o  So, it would be scientific lunacy for persons working on nuclear pollution to reject the knowledge gained from studying the health-harm from medical xrays. For example, some of the most important insights in the whole scientific effort have been based on medical xrays:

        o  Medical xrays provided Alice Stewart's evidence, in 1956-1958, that pre-birth exposure to ionizing radiation increases the risk of childhood Cancers and Leukemia.

        o  Exposure to medical xrays, in therapy of the painful spinal disease "Ankylosing Spondylitis," provided the first evidence in the 1960s that nearly all kinds of human Cancer would probably turn out to be inducible by ionizing radiation. This early warning was correct.

        o  Human exposure to medical xrays provided the type of evidence which made it possible in 1990 for us to prove, by any reasonable standard of biomedical proof, that there is no safe (risk-free) dose-level or dose-rate of ionizing radiation with respect to causing Cancer. And much of the same evidence was subsequently used in 1995 by the British National Radiological Protection Board, to conclude that "the weight of the evidence falls decisively" against any threshold (safe dose).

        o  Now, in 1999, data from human exposure to medical xrays provides the first powerful evidence that exposure to ionizing radiation is a cause of Coronary Heart Disease. This discovery is clearly as important as the discovery that exposure to ionizing radiation is a cause of Cancer.

        o  How could anyone assume that all the discoveries above have no relevance to exposure by nuclear pollutants? Such an assumption would be an irrational denial of the essential feature of ionizing radiation. Everyone concerned about nuclear pollution has a huge stake in knowledge gained from studies of medical xrays.


Part 5  o  Warning: Studies Where Bias Can Enter

        o  One of the most exciting scientific aspects of my new study, based on medical xrays, is that it uses immensely trustworthy databases. There is no chance that the databases are biased with respect to finding a strong or a weak effect of ionizing radiation.

        o  Few people seem to appreciate how unusual and important this is. (Clearly the issue was not recognized as important by the scientist who urged me to return to analysis of the A-Bomb Survivor Database.)

        o  So, I am going to try to explain the importance, and I hope that the explanation will be treated as a very serious warning for future years, too.

o  Human Radiation Exposure: Rarely Measured

        o  It has not been easy, over the past 30 years, to find reliable data to analyze, about the health effects of ionizing radiation. Why not?

        o  One of the big reasons is that the amount of exposure (dose) has rarely been fully measured, if measured at all. That statement is true for the Atomic-Bomb Survivors in Japan, the nuclear workers in the USA, Britain, France, and the USSR, the military personnel exposed by weapons-related activities, the clean-up workers at Chernobyl and the near-by populations, the population around Three Mile Island, the people living "downwind" of the Nevada Test Site and other nuclear activities, and people exposed to medical xrays.

        o  If the radiation dose for each participant in a study was not measured properly or not measured at all, it is quite easy for database-makers to assign high doses to some people who really received low doses, and to assign low doses to some people who really received high doses. Or, database makers may assume that all participants of a certain type received the same dose, when the dose for those participants really had a ten-fold range. It does not require many mistakes of these sorts (either innocent, or intentional) to make ionizing radiation appear to be a "weak" carcinogen.

o  The First Duty of Any Objective Scientist

        o  "Even Einstein himself will get false results from false databases." So, it is the first obligation of all objective scientists to assure themselves, before they start an analysis, that the data themselves are trustworthy. If they have doubts, they have a duty at least to explain the basis for doubt to the readers.

        o  It would be irresponsible to ignore the fact that many (most) important radiation databases have been assembled and "managed" thanks to funds from sources which are far from neutral about the findings which arise out of those databases. It would not be responsible to rely, for information about the hazards of smoking, on databases sponsored by the tobacco industry.

        o  And yet such reliance is tacitly accepted by my colleagues, when it comes to radiation databases. During the past decade, I have often expressed my deep disappointment that so many analysts have been willing to use databases on the nuclear workers, after those databases were under the control and management of the Dept. of Defense and other sponsors who have had motive, opportunity, and means to manipulate the data before releasing the databases for "open" analysis.

        o  Under those circumstances, analysts cannot possibly meet their duty to assure that the data are not tainted. By touching such data, they debase the standards of credible epidemiological research, in my opinion. By contrast, the last time I analyzed data on nuclear workers (1979), I was able to use Dr. Thomas Mancuso's Hanford database before his data were seized and impounded by the Dept. of Energy.

o  Broken Rules: Destroying the Barriers against Bias

        o  Many rules of research exist in order to create barriers to bias. The most basic rule of credible biomedical research is that the input (e.g., dose-estimates) cannot be replaced after any results are known. If people are free to re-arrange the input, obviously they can produce whatever new "result" they like. Therefore, I have protested use of the Spondylitic Study after its data were retroactively altered with respect to dose-estimates.

        o  The Atomic-Bomb Survivor Study has, since 1986, undergone the most massive retroactive alteration one can imagine. All the doses have been re-evaluated, participants have been suddenly added from a "reserve," and the original cohorts of participants have been shuffled ... and not just once! Retroactive tinkering with the database has become chronic. Who funds this tinkering? The U.S. and Japanese governments.

        o  It is disgraceful that almost no analyst except myself has tried to stop this mishandling of a unique human resource --- one which was providing health information from events which will never be repeated.

        o  In the journal Health Physics, and in my 1990 book, I demonstrated how it would be possible to preserve the credibility of the A-Bomb Survivor Database, while also exploring the impact of revised dose-estimates. The method is called "constant-cohort, dual-dosimetry analysis."

o  The Price of "Going Along" --- A Health Tragedy?

        o  But other analysts prefer to use only the altered databases. They consent, by their silence, to send the unaltered database to oblivion. Perhaps they pardon themselves by saying, "But the study is still giving answers which I like!"

        o  So what! Genuine scientific inquiry is about truth. No scientist would wish to make radiation appear more harmful, or less harmful, than it really is. strings attached

        o  If today my colleagues permit oblivion for the legitimate A-Bomb Survivor Database and tolerate the destruction of the standard barriers against bias, tomorrow they will have to accept as valid whatever results come out of the database's hundredth version in some future year (say 2020) --- yet they will be powerless to know whether the results are valid or contrived (biased).

        o   What or who is going to prevent the alteration of potentially all the important radiation databases? Silence on the issue of trustworthy databases could result in a tragedy for human health.


Part 6  o  Joy over Two Totally Trustworthy Databases

        o  You can hardly imagine my joy when finally I figured out a way to evaluate the impact of medical radiation upon cancer mortality in the USA, from two databases which I could totally trust to be unbiased with respect to the topic.

    "PhysPop" --- Defined    

        o  Because medical xray doses were not measured in the past, and are rarely measured today, both past and current dose-estimates are highly uncertain. But we recognized that the problem (of estimating the impact of medical radiation on national age-adjusted cancer mortality-rates) might be solved by using data which are available --- namely, the number of Physicians per 100,000 Population ("PhysPop") in each of the nation's Nine Census Divisions.

        o  Physicians (not patients) order xray procedures. As the density of physicians goes up per 100,000 population, more xrays will be ordered per 100,000 population. This common-sense premise is supported by surveys reported in 1988. In each Census Division, the population's average per-capita xray dose will be approximately proportional to the Census Division's PhysPop value for the same year.

        o  Thus, the nine PhysPop values are a valid indicator of the relative magnitude of annual per-capita xray dosage received by the nine populations of the nation's Nine Census Divisions. However, we had to ascertain that PhysPop would be a valid indicator of accumulated per-capita xray doses, not just the doses received in a single year. It turns out that the relative magnitude of PhysPop values was remarkably stable, among the Nine Census Divisions, from 1921 to 1990.

o  Two Totally Neutral, Trustworthy Databases

        o  Because we were able to use the PhysPop database for dose, we are totally confident that we have been able to use data which are absolutely free from bias with respect to xray dose. After all, the PhysPop values in every state were collected and published from 1921 to the present day by the American Medical Association, for completely different purposes. The AMA statisticians clearly had no idea that anyone would ever use the PhysPop values to evaluate the impact of medical radiation upon cancer mortality, by Census Divisions.

        o  In addition, we are totally confident that the mortality rates by Census Divisions, collected for Vital Statistics by the U.S. Government, also are absolutely free from bias with respect to whether medical xrays have a big or a small impact on cancer mortality, by Census Divisions.

o  Probably the Strongest Study Ever Done on Radiation

        o  These two databases permit our study to begin in 1940, by "enrolling" all 150 million inhabitants of our Nine Census Divisions into the study. (The study ends in 1990, when the combined population was 247 million.) By contrast, the A-Bomb Study has about 90,000 participants --- which severely limits its power to discern provable differences between dose-groups.

        o  Our PhysPop study is probably the largest, statistically strongest, prospective study of radiation health-effects ever done --- and it was done with assuredly neutral data. One should go where the trustworthy data are, and we did.


Part 7  o  Preventing an Exercise in Self-Defeat

        o  Our PhysPop study revealed not only that medical xrays are an extremely important cause of the nation's past and present cancer mortality, but it produced the first powerful evidence that ionizing radiation is also an extremely important cause of the other biggest killer in the nation: Ischemic/Coronary Heart Disease.

        o  To assume that these striking findings are irrelevant to issues of nuclear pollution, because the insights derive from medical xrays, would be a serious mistake (Part 4, above). It would be self-defeating to ignore undeniably strong findings from any immense, neutral, highly credible database --- in favor of perpetual dependence on marginal findings from the A-Bomb Survivor Database, the nuclear worker databases, and other databases with unreliable dose-estimates, retroactively altered input, and pro-nuclear management.

        o  It would make good sense to avoid an exercise in self-defeat. The stakes for posterity are very high.

# # # # #

Dead Center

Committee for Nuclear Responsibiity, Inc. (CNR)
POB 421993, San Francisco, CA 94142, USA
An educational group since 1971.

This document is also available in plain HTML and ASCII text formats
and in online at

back to CNR | radiation | rat haus | Index | Search | tree