REPORT NO. 3
health implications
of fallout from
nuclear weapons
testing through 1961
May1962
Report of the
FEDERAL RADIATION COUNCIL
-------
REPORT NO. 3
health implications
of fallout from
nuclear weapons
testing through 1961
May1962
Report of the
FEDERAL RADIATION COUNCIL
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington 25, D.C. - Price 15 cents
-------
CONTENTS
Page
Report of the Federal Radiation Council 1
Appendix A 5
Appendix B 7
Acknowledgment 10
111
-------
REPORT OF THE FEDERAL RADIATION COUNCIL
HEALTH IMPLICATIONS OF FALLOUT
FROM NUCLEAR WEAPONS TESTING THROUGH 1961
The Federal Radiation Council has considered available information on radiation doses and possible
health effects of atmospheric nuclear weapons testing. Before discussing the estimates made in this re-
port in detail, it is appropriate to point out the difficulties of being precise in this field.
Although a large and expanding program for measuring radiation levels at a number of locations
throughout the United States has been in effect for a number of years, the application of such data to the
whole country, to an extended time period, or to the entire population involves assumptions than can not
be completely validated. Furthermore, while a considerable body of information has been accumulated
on the effects of radiation on animals and man, the possible effects of low doses delivered at low dose
rates are insufficiently known to permit firm conclusions about the extremely low exposures resulting
from fallout. Current experimental techniques are not good enough to detect biological effects at the
low levels of worldwide fallout from nuclear tests.
Any possible manifestations resulting from fallout radiation will not be unique, for all of the diseases
and disabilities known to be caused by radiation also occur for other reasons. Whatever effects might
be produced by fallout could only be reflected in statistical increases in the number of conditions al-
ready present in the population. Any individual effects would be so diluted by space and time that they
would not be recognizable among the much larger number of identical effects arising from other
causes, among which they would be interspersed.
Finally, any proper understanding of estimates in this field must take into account the many dif-
ferent ways in which similar or even identical data can be expressed. Many of the apparent differences
among scientists arise from different forms of presentation. Two approaches have been used. One
estimates the risk of damage to a single person. This risk is extremely small in comparison with
others which people normally accept. The second approach considers possible effects on a large pop-
ulation for a year or a generation or for several generations totaling hundreds of years. Even a very
small proportion of affected individuals will, in a very large population for a long period of time,
amount to an impressive total number of individuals.
Estimated Radiation Exposure from Testing
Any consideration of possible health effects from fallout must begin with the radiation doses to which
people are exposed as a result of such tests.
A sharp distinction must be made between the devastating effects of "local" fallout in a nuclear at-
tack on an unshielded population and the effects of fallout from weapons testing. Weapons testing
creates far smaller total amounts of fission products so that its fallout is far less than that which
would result from nuclear war. Furthermore, the tests are planned to avoid local fallout or to confine
it to locations where it will have minimal effects. Hence, in weapons testing the problem is largely
confined to delayed fallout which decays greatly in the upper atmosphere and is dispersed at low con-
centrations over the earth's surface. This report is concerned primarily with the effects of such de-
layed fallout.
Dose estimations must take into account exposure from all sources; external, and internal through
ingestion of food and water and inhalation. Some radioactive elements may concentrate to different
extents in various parts of the body. Those which tend to concentrate in a certain organ will selectively
irradiate that organ. Thus a thyroid dose, for example, represents the sum of the whole-body dose
from a variety of substances plus the extra dose from iodine-131, an element which tends to concen-
trate in the thyroid gland. In addition, some elements are taken up more effectively at one age than
another. For example, the proportion of strontium-90 retained in the growing bones of children is
greater than that retained in the bones of adults ingesting the same foods. Furthermore, different
sources of radiation give off different kinds of radiation having different biological effects, so that doses
cannot be directly compared. These points should indicate the difficulty of referring to any one exposure
level from a particular source without identifying what kind of a dose and what part of the body is involved.
Estimates of doses from fallout from tests through 1961 in millirems, a unit of ionizing radiation
dose, are given in Table I and discussed further in Appendix "A". Because of uncertainties and the
variety of necessary assumptions, these estimates are expressed as ranges of values within which the
average exposure over the United States is expected to lie. The values given apply to the United States,
and are somewhat higher than those for most of the rest of the world. Doses to the whole-body and re-
productive cells represent an average for all age groups in the entire population. Doses to bone and
bone marrow are average values for those who were infants at the time of highest concentrations of the
particular isotopes irradiating these organs; values averaged for all age groups will be lower.
1
-------
The half-life of radioactive iodine, the principal source of the thyroid dose, is only 8 days and the
peak dose rates persist for a relatively short period of time. For this reason thyroid doses are not
included in the table. Doses to the thyroid from the major past tests were estimated to have ranged
from 100 to 200 millirems per year during and immediately following periods of testing. These values
apply only to individuals who were infants at the time of highest concentration of radioactive iodine.
The average value for all age groups was about a tenth as much. Although data from which thyroid
doses during 1957-58 can be estimated are limited, it is likely that there was much geographic varia-
tion, and in some limited areas of the United States the average thyroid doses were probably many
times the national average.
The whole-body dose due to the carbon-14 produced by all tests through 1961 has been included but
not separately listed in Table I. It is estimated to total from 10 to 15 millirems during the first thirty-
year period. The dose rate will decrease much more rapidly than would be predicted on the basis of
the carbon-14 radioactive half-life of 5,700 years because of the absorption of the radioactive carbon
dioxide from the atmosphere into the ocean. After about 200 years the dose rate from carbon-14 will
have been reduced to a total of about 0.75 millirem during a thirty-year period.
To put these dose levels in some perspective, Table I compares them with exposures from natural
background and with the Radiation Protection Guides of the Federal Radiation Council. The compari-
sons indicate that doses from fallout have generally been a small fraction of the Guides for population
groups.
Background radiation arises from naturally radioactive materials such as carbon-14 and potassium-
40 in the human body, radium in the earth's crust, and cosmic radiation from outer space. Man has
always been exposed to these radiations. Natural background radiation varies from place to place, both
with elevation and with radioactive content of local materials. In the United States these values have
been observed to range from 70 to 200 millirems per year. The value for background radiation given in
Table I is a weighted average for the entire United States population.
The estimated values given in Table I for whole-body exposures from fallout are considerably less
than the exposures from natural sources. Over a period of 30 years the average whole-body dose from
all testing through 1961 will be between 60 and 130 millirems compared to 3,000 millirems from back-
ground. Thus testing through 1961, including the contribution from carbon-14, will, over this thirty-
year period, increase exposures over natural background by less than five percent. Seventy-year aver-
age bone doses, when similarly compared, are increased less than ten percent. Any further testing
will, of course, increase the exposure.
The fact that exposure from Some sources is generally accepted without question should not in itself
be a reason for accepting exposure to added levels of man-made radiation. However, comparison of
exposure levels with those of natural background does provide some indication of the significance of
increases from fallout. One normally considers variation in exposure from natural sources to be of
little significance. For example, a resident of the East Coast contemplating a move to a high-altitude
location in the West is unlikely to know or attach any importance to the fact that his exposure to back-
ground radiation will be appreciably increased—morethan twenty-five percent at elevations above one
mile.
Another basis of comparison is the radiation exposure received from medical diagnostic procedures
in the United States. It has been estimated that a person in the United States will accumulate a geneti-
cally effective dose of the order of 1,000 millirems over a thirty-year period. There are, however,
wide fluctuations in the exposures to the reproductive cells from the diagnostic procedures.
Estimates of Biological Effects
Much available evidence indicates that any radiation is potentially harmful. However, effects become
increasingly difficult to demonstrate below 10,000 millirems, and impossible to detect by present tech-
niques at the very low dose levels from fallout. Nevertheless, it is virtually certain that genetic effects
can be produced by even the lowest doses. These effects in the children of exposed parents and all fu-
ture generations may be of many kinds, ranging from minor defects too small to be noticed to severe
disease and death.
In the case of somatic effects, i.e., effects directly on the persons exposed, the evidence is insuffi-
cient to prove either that there is a dosage level below which no damage occurs (the "damage threshold"
hypothesis) or that there is some risk of damage at any dosage level, no matter how low (the "no thresh-
old" hypothesis). It may well be that some effects are of one kind, some of the other. Dose rate is im-
portant; a protracted dose is much less effective than the same total dose given in a short time.
Estimates have been made by national and international groups of scientists of the number of possi-
ble adverse health effects that might occur from various exposure levels. Tables II and III apply some
-------
of these estimates to the exposure levels from all testing through 1961 to indicate the possible adverse
health effects in the United States population that might result from this testing. United States figures
have been used because knowledge of dose levels and of health effects occurring in the absence of test-
ing is more complete for this country than on a worldwide basis. For convenience in expressing the
concepts and calculations in this report, the population of the United States has been taken as approxi-
mately one-tenth of the population in the same latitudes of the northern hemisphere, and as one-twenti-
eth of the population of the entire world. The figures in Table II on the possible number of adverse
health effects from testing through 1961 may be multiplied by 10 to provide a rough estimate of com-
parable worldwide effects with the exception of carbon-14, for which a factor of approximately 20 must
be applied.
Table II and Appendix "B"give numerical estimates of the effects of fallout on one category of genet-
ic effects—severephysical and mental defects. This category includes the hereditary component of
such things as congenital malformations, blindness, deafness, feeblemindedness, muscular dystrophy,
hemophilia and mental diseases.
In Table II the estimated numbers of radiation effects are given as three values. The upper figure is
the best estimate based on radiation-induced mutation rates in mice, and on the spontaneous incidence
of these defects in man. The other figures represent the range within which the true value may reason-
ably be expected to lie.
As shown in the table, about ten percent of the number that may result in all time from weapons tests
through 1961 are estimated to occur in the first generation—thashildren of parents exposed to this fall-
out. The remaining ninety percent occur in decreasing numbers in succeeding generations. Somatic
effects appear only in the irradiated individual himself, and not in his offspring. The manifestations of
particular concern are leukemia and other types of cancer.
The radiation dose from carbon-14 is spread over an enormous period of time extending through
many thousands of years. The number of mutations from carbon-14, when exposure over all time is
considered, is estimated to be greater than from other radioactive elements produced in nuclear deto-
nations. These mutations will, however, be distributed over a much longer time with a much smaller
number in any one generation.
In addition to the gross defects listed in Table II, there may be an unknown but probably a consider-
ably larger number of mutations with less obvious effects such as minor physical abnormalities, mild
diseases, impairment of physiological functions, and reduced resistance to infection or other stresses
of life. Part of this damage will result in a lowered probability of survival at various ages.
Reduced viability of this kind has been consistently found in mouse experiments. The best data on
mice are for the infant and embryonic deaths. To the extent that mouse data can be applied to man, the
results indicate that the radiation-induced mortality of embryos and infants in the first generation after
irradiation is probably larger, perhaps five times larger, than the number of induced defects of the type
estimated in Table II. Numerical estimates are not given for such effects because of uncertainties as to
the comparability of these effects in mice and humans. This is the viewpoint of those who have done
much of the experimental work in this field.
Mutations which have a mild effect on the individual may cause substantial damage in the aggregate.
This is because the mildness permits these mutations, such as slight reductions in viability and other
less obvious effects, to persist in the population longer than mutations with severe effects, and thus to
affect a correspondingly greater number of persons. There are no data which would permit these ef-
fects to be assessed with sufficient accuracy to permit numerical estimates.
If, however, numerical estimates are made of all these genetic effects, both those which are likely
and those which are more speculative, the aggregate of these estimates when counted as the total num-
ber of individuals affected throughout the world in future generations leads to very large numbers.
Likewise, large numbers can be obtained when other effects or deaths from any cause are totaled over
large populations and many generations. On the other hand, it must be emphasized again that whatever
the genetic effects of fallout radiation from weapons testing through 1961 may be, the total effect will
certainly be considerably less than that occurring inescapably from background radiation. This, in turn,
is considerably less than the effects from other factors which determine the total natural mutation rate.
Estimates for two kinds of somatic effects, leukemia and bone cancer, are given in Table III. As
mentioned earlier, it is not known whether or not there is a threshold dose below which these diseases
are not produced. If a threshold exists, fallout radiation may produceno additional cases, and the lower
limits of zero reflect this possibility.
The upper estimates in Table III are made by assuming the effect of a low dose, delivered at a low
dose rate, to be proportional to the effect of a high dose delivered at a higher dose rate. The estimates
for the upper limits are probably too high because no allowance had been made for the possibility that a
-------
given dose is less effective when received slowly over a long period of time. Thus the range of num-
bers given in Table III is reasonably certain to bracket the correct value.
There are other possible somatic effects of radiation such as malignancies (other than leukemia and
bone cancer) and general effects such as life shortening. Among these malignancies is cancer of the
thyroid, a possible effect from exposure to radioiodine. Table III includes no data on the possible in-
cidence of this effect because ,estimates, like those recognized by national and international groups of
scientists for possible leukemia and bone cancer effects, have not been made for cancer of the thyroid.
However, from what little is known about the effect of radioiodine, including data obtained from human
exposures at very high levels, the likelihood of any possible thyroid effects has been considered to be
about the same as other malignancies for comparable exposures. Even less information is available as
to possible increases in all these other effects than is available for leukemia and bone cancer.
To put these estimates of possible adverse health effects in some perspective, Tables II and III also
include the total number of these same effects occurring in the United States from all causes.
Conclusions
We cannot say with certainty what health hazards are caused by fallout from nuclear testing. We
expect there will be some genetic effects; other effects such as leukemia and cancer are more specula-
tive and may not occur at all. We can observe that, compared to the number of these same adverse
biological effects occurring wholly apart from testing, the additional cases that might be caused by test-
ing are a very small quantity. We conclude that nuclear testing through 1961 has increased by small
amounts the normal risks of adverse health effects.
-------
APPENDIX "A"
EXPLANATORY MATERIAL ON DOSE ESTIMATIONS
The estimates of radiation doses attributable to fallout from tests of nuclear weapons given in
Table I have been based on extensive observations and studies through 1961. These estimates include
exposures from fallout which already has occurred and from material from past tests yet to be de-
posited. Estimates are based on measurements of radionuclides in air, rain, soil, water supplies, food,
and people.
Table I gives estimates of radiation doses from fallout resulting from tests through 1961. The dose
ranges given in this table represent estimates made using somewhat different but plausible assumptions
concerning such factors as fallout distribution, the effects of weathering and shielding, and the move-
ment of radioisotopes from the environment to man. It is believed that the best estimates that can be
made at the present time would lie within the ranges given.
In the cases of whole body and reproductive cell exposures, radiation doses are relatively independent
of age, except for the fact that children born in the past two or three years will have missed much of the
exposure from earlier tests experienced by older persons. A large fraction of the dose to the whole-
body and reproductive cells from a particular test may be received within a period of months after
fallout occurs. The contribution of radioiodine to the dose to the thyroid gland is much larger in the
case of infants than in older persons and is effectively complete within a few weeks after a nuclear test.
Radiation doses to the bone and bone-marrow from a particular test will be received at decreasing
rates over a period of a lifetime. Early concentrations in the bone will be greatest for those children
who are less than one year of age at the time that peak concentrations of fallout occur in food. The
average bone and bone marrow doses to such children as estimated in Table I are much larger than the
average to the whole population.
It is estimated that carbon-14 resulting from tests through 1961 will produce a radiation dose to the
whole body including the reproductive cells of 10 to 15 millirems in the first 30 years, which is less
than one percent of the 30 year genetic dose to the present population from natural background.
While carbon-14 decays very slowly with a radioactive half-life of 5,700 years, its availability as a
source of radiation exposure initially decreases rather rapidly because of absorption of carbon dioxide
from the atmosphere into the oceans. In a period of one or two hundred years, the exchange between
the atmosphere and the ocean approaches an equilibrium with most of the carbon-14 in the oceans. This
mixing will reduce the carbon-14 due to weapons tests to about two percent of the natural carbon-14
concentration in the atmosphere, biosphere and oceans. The radiation dose rate at this time will be
about 0.025 millirem per year, or 0.75 millirem per generation. Although the dose rate is very small,
it will continue at a rate which decreases with the radioactive decay of carbon-14 through hundreds of
generations.
Doses to the whole-body and reproductive cells were averaged, weighted according to population; bone
and thyroid doses were averaged over that portion of the population who were infants at the time of
highest concentrations of relevant radioisotopes in the diet. Average doses to older children and adults,
and thus to the total population, were smaller. Some local averages, particularly in the case of the
thyroid, were much higher.
All one year doses are for the year, within the period covered, in which the highest yearly doses were
received. The highest one-year doses to the whole-body and skeleton from tests prior to 1961 were
experienced in 1958-1959. The highest one-year doses to the whole-body and to the skeleton from the
1961 tests are expected during 1962 and 1963.
-------
TABLE I
Estimated Radiation Doses in the United States
(Doses expressed in millirem)
Tissue or organ
Whole body
1 Year
30 Years
70 Years
Reproductive cells
1 Year
30 Years
70 Years
Bone
1 Year
70 Years
Bone marrow
1 Year
70 Years
From all tests
through 1961
10- 25
60-130
70-150
10- 25
60-130
70-150
30- 80
400-900
20- 40
150-350
From natural back-
ground
100
3,000
7,000
100
3,000
7,000
130
9,100
100
7,000
FRC Radiation Pro-
tection Guides* for
normal peacetime
operations
Population groups
170
5,000
11,900
170
5,000
11,900
500
35,000
170
11,900
"The Radiation Protection Guide for whole-body exposure of individual radiation workers is 5,000
millirems per year.
-------
APPENDIX "B"
DISCUSSION OF THE NUMERICAL VALUES IN TABLES H AND III
The estimates of genetic effect are based largely on the reports of the Committee on Genetic Effects
of the National Academy of Sciences, contained in the Academy's 1956 and 1960 Summary Reports on
the Biological Effects of Atomic Radiation. The Summary Reports concluded from the available scien-
tific information that the genetic effects of exposure of a population to small doses of radiation are
proportional to the average dose to the reproductive cells of potential parents.
The Committee reported that normally some four to five percent of children born have or will de-
velop a severe physical or mental defect. Of these defective children about half, or two percent of the
total number born, are thought to have traits whose frequency in the population is directly dependent on
the mutation rate.
The Academy Committee utilized data on mutation rates in mice and estimated the effects on human
populations, assuming that human radiation-induced mutation rates are the same as in mice. The 1956
Report estimated that if the parents of the present generation were exposed to 10,000 millirems, this
average dose would give rise to some 50,000 additional defective children among 100 million children
born. The total number for all future generations, assuming no change in the size of population, was
estimated as 500,000.
Recent data have shown that radiation given at a very low rate produces fewer mutations than the
same total dose given quickly. Since the earlier estimates were based on high dose rates, they should
be reduced accordingly. The results from recent experiments with mice indicate that when both parents
are irradiated the best estimate of the number of mutations should be only 1/6 as large as with high
dose rates.
An application of these modified estimates to the reproductive cell exposures estimated to occur
from past weapons tests, approximately 100 millirems over the first 30 years, leads to an estimate of
110 cases of serious inherited defects in the first generation of 130 million births. The estimates of
radiation doses in Table I apply only to radiation received by the present population of the United States.
At least four physical phenomena contribute to making the radiation doses to future generations from
these tests much smaller. In fact, in a few decades the exposure per generation from residual radio-
activity produced by these tests will have dropped to less than one percent of the exposure to the
current population.
In the case of the whole-body and reproductive cells, about 50% of the 30-year dose from tests
through 1961 has resulted from exposure to radiation from relatively short-live gamma-emitting mate-
rials outside the body. As a result of radioactive decay, these will have essentially disappeared within
a few years.
It is estimated that about 20 percent of the 30-year dose is from cesium-137 in the diet. Most of
this results from the direct deposition of fallout on vegetation. When the deposition rate is low, the
availability of cesium-137 is small. This factor, together with its short retention time in the body,
makes this radioisotope a small contributor to internal irradiation. About 25 percent of the 30 year
dose is due to cesium-137 outside the body. The dose rate from this source decreases with time, not
only as a result of radioactive decay with a half-life of 27 years, but also because of decreasing avail-
ability due to migration into the earth or into streams, storm drains, etc. The dose rate from this
isotope may be reduced by 1/2 to 1/10 after 30 years in addition to radioactive decay.
It is estimated that carbon-14 resulting from tests through 1961 will produce a radiation dose of 10
to 15 millirems in the first 30 years, about 10 percent of the 30 year genetic dose from fallout to the
present population. The radiation dose rate, after equilibrium with the oceans has been reached, will be
about 0.025 millirem per year, or 0.75 millirem per generation. Although the dose rate is very small,
it is of interest because it will continue at a rate which decreases with the radioactive decay of carbon-
14 through hundreds of generations.
In addition to its radiation effects, carbon-14 may produce mutations through disruption of the nor-
mal chemical structure of the gene when the atom of carbon-14 is converted into nitrogen. The contri-
bution from this effect appears to be small in comparison to the radiation effect, and is too speculative
to provide a firm basis for numerical estimates.
The current total incidence of deaths due to leukemia in the United States is about 12,000 per year
and that of bone cancer is about 2,000 per year. These amount to average rates for all ages of 7 cases
per one-hundred thousand persons and 1.1 cases per one-hundred thousand persons, respectively.
-------
It is assumed that the incidence of these diseases as a result of exposure of the blood-forming
tissues and the bone, respectively, to radiation is proportional to the exposure. Observations of num-
ber of cases of leukemia resulting from very large doses of radiation suggest that up to ten percent of
the normal incidence of leukemia may be due to exposure to radiation from natural sources, amounting
to an average of 7,000 millirems in 70 years. The same assumption has sometimes been made for bone
cancer. These assumptions were made, for example, by the United Nations Scientific Committee on the
Effects of Atomic Radiation (1958) in estimating an upper limit to the number of cases of leukemia and
bone cancer that might be expected from low levels of exposure such as those from fallout from the
testing of nuclear weapons.
On this basis, one could estimate that if an average lifetime exposure of 7,000 millirems to the
blood-forming tissues of the population of the United States results in a total of about 84,000 cases of
leukemia in the period of an average lifespan of 70 years, the average lifetime exposure to fallout could
be expected to result in a total of up to 2,000 cases of leukemia, averaging about 30 per year. The
average exposure to the population as a whole from fallout is estimated to be about 175 millirems to the
bone marrow, about half the value calculated for infants, as shown in Table I. A corresponding estimate
for the number of cases of bone cancer from a population weighted lifetime dose of about 450 millirems
would give an upper limit of 700 cases in 70 years, averaging about 10 cases per year.
For comparison, there are about 1,700,000 deaths each year in the United States from all causes. Of
these, up to about 1,400, or about 10% of the total due to leukemia and bone cancer from all causes, are
attributed to radiation exposure from natural sources. The possible additional 40 deaths from these
causes, as estimated above, illustrate the degree of risk to an individual from fallout in comparison to
risks already present.
-------
TABLE II
Effect of Fallout on the Number of Gross Physical or Mental
Defects in Future Generations in the United States
(No allowance has been made for future increases in population)
(1)
Estimated number of
cases due to all causes
(hereditary and non-
hereditary) in children
of persons now living
4,000,000-6,000,000
(2)
Estimated number of additional
cases in the first generation
(children of persons now alive)
caused by all tests through
1961
Fallout Carbon -14
100 10
Range (20-500) (2-50)
(3)
Estimated total number for
all future generations from
all tests through 1961
1,000 2,000
(200-5,000) (400-10,000)
(4)
Risk to an in-
dividual of the
next generation
from all tests
through 1961
1/1,000,000
The upper figures in columns 2 and 3 are best estimates based on radiation-induced mutation rates in
mice, and on the spontaneous incidence of these defects in man.
The lower sets of figures represent the range within which the true value may reasonably be expected
to lie.
TABLE III
Certain Malignant Diseases in the Next Seventy Years in the United States
Leukemia
Bone Cancer
Estimated to-
tal number of
cases from all
causes (present
incidence)
840,000
140 000
Estimated num-
ber of cases
caused by nat-
ural radiation
0-84,000
0-14,000
Estimated num-
ber of addition-
al cases from
all tests through
1961
0-2,000
0-700
Risk to an in-
dividual of de-
veloping the
disease due to
all tests through
1961
0-1/100,000
0-1/300,000
-------
ACKNOWLEDGMENT
The Federal Radiation Council has been assisted in this study by several
special advisors, and by the following consultants selected by the National
Academy of Sciences:
Dr. Howard L. Andrews Dr. James V. Neel
Dr. Victor P. Bond Dr. William L. Russell
Dr. James F. Crow Dr. Shields Warren
Dr. Lester Machta
10
------- |