American Journal of Epidemiology Vol. 154, No.

12,
Copyright © 2001 by the Johns Hopkins University Bloomberg School of Public Health Supplement
All rights reserved Printed in U.S.A.

Radiation Exposure and Cancer Matanoski et al.

Radiation Exposure and Cancer: Case Study

Genevieve M. Matanoski,1 John D. Boice, Jr.,2 Stephen L. Brown,3 Ethel S. Gilbert,4 Jerome S. Puskin,5 and
Tara O’Toole6

Downloaded from https://academic.oup.com/aje/article-abstract/154/12/S91/64652 by Lincoln University user on 19 June 2020

The long, colorful, and tragic history of ionizing radiation patients developed liver cancer and leukemia at high rates
and the use of epidemiologic approaches to characterize (4). In the 1940s, the first studies on leukemia excesses
cancer risk distinguish this carcinogen from many other among radiologists were published (5–7). While the first
known carcinogens. Knowledge on the health effects of warnings of adverse health effects were seen in radiologists,
exposure of human populations to ionizing radiation is central in radiation protection has been the study of atomic
extensive, having mounted progressively through the 20th bomb survivors. Findings of the initial studies of leukemia
century. After their discovery in 1895, x-rays were used for in atomic bomb survivors and other exposed populations
25 years by enthusiasts in applications ranging from treat- were reported in the 1950s (8, 9). Nearly a decade later, the
ment of acne and “female problems” to removal of first quantitative estimates of lung cancer risk from cohort
unwanted facial hair (1). This enthusiasm persisted despite studies of underground miners exposed to radon were
the scientific community’s early knowledge of the adverse reported (10).
effects of x-ray overexposure at that time (1). The effects of During the past 50 years, many other human studies have
extremely large exposures were first demonstrated by severe quantified cancer risks among persons exposed in military,
x-ray burns and cancer deaths among pioneering radiation occupational, medical, or environmental settings. These
workers. One of these pioneers was Thomas Edison’s assis- studies, along with confirming animal and other experimen-
tant, Clarence Dally, who helped develop the x-ray fluoro- tal data, leave no question as to whether radiation is a car-
scope. After submitting to amputation of his burned and cinogen. Several critical scientific issues remain to be
ulcerated arms, he died from cancer in 1904 (1). resolved, however, including differences in risks for brief
The height of x-ray use was during World War I versus chronic exposures (i.e., is there a dose-rate effect?),
(1914–1918) when primitive mobile x-ray machines were the shape of the dose-response curve at low doses, the life-
used extensively in the field to locate shrapnel and help set time risk following childhood exposures, the possible exis-
broken bones (1). In the 1920s, young women painting tence of radiosensitive subgroups within the population
watch dials with radium paint were found to have high rates (perhaps genetically determined), and the possible interac-
of bone cancer from ingesting large quantities of radium tion between radiation and other carcinogenic exposures.
when licking the brushes to make fine points (2). While this For over 70 years, standing committees have continually
occupational hazard was eventually recognized (1, 3) and reviewed new data on radiation effects and recommended
the practice of licking the paint brushes was prohibited by protection guidelines for workers and the public (11). The
the late 1920s (3), use of ionizing radiation continued in first recommendations for radiologic protection came from
many other applications for decades. For example, individual physicians, with little impact. Following a second
Thorotrast, a radiographic contrast agent consisting of tho- wave of deaths in the 1920s due to leukemia and cancers with
rium dioxide, was used from the 1930s to 1951, and exposed longer latency periods among the early radiologists and
exposed patients, there were renewed efforts to institute
Received for publication April 27, 2001, and accepted for publica- safety standards for the use of radiation (1). Medical and radi-
tion August 29, 2001. ologic societies in various countries finally took a lead in set-
Abbreviations: ALARA, as low as reasonably achievable; BEIR, ting standards for radiation exposure (1). International agree-
Biological Effects of Ionizing Radiation; ICRP, International
Commission on Radiation Protection; NCRP, National Council on ment on guidelines for radiation safety was achieved in the
Radiation Protection and Measurements; UNSCEAR, United second meeting of the International Congress of Radiology in
Nations Scientific Committee on the Effects of Atomic Radiation. 1928, when it formed the International Committee on X-ray
1
Department of Epidemiology, Bloomberg School of Public and Radium Protection (11). This Committee, later to become
Health, Johns Hopkins University, Baltimore, MD.
2
International Epidemiology Institute, Rockville, MD.
the International Commission on Radiological Protection
3
Risks of Radiation and Chemical Compounds, Oakland, CA. (ICRP) in 1950, issued guidelines approximately once every
4
Radiation Epidemiology Branch, National Cancer Institute, 3 years (11). While they were detailed, the guidelines
National Institutes of Health, Rockville, MD.
5
remained arbitrary without the necessary data to assess safe
US Environmental Protection Agency, Washington, DC. levels of exposure (1). The United States counterpart to the
6
Annapolis, MD.
Reprint requests to Dr. Genevieve Matanoski, Johns Hopkins ICRP, the US Advisory Committee on X-ray and Radium
Bloomberg School of Public Health, 111 Market Place, Suite 850, Protection, had its first meeting in 1929 (11). The Advisory
Baltimore, MD 21202-6709 (e-mail:[email protected]). Committee reorganized in 1946 to become the National

S91
S92 Matanoski et al.

Committee on Radiation Protection (NCRP), which then comes, a firm understanding of risks from these various
became a formal council in 1964 with a charter from the US exposures is essential.
Congress and was renamed National Council on Radiation Examples of the questions that society has grappled with
Protection and Measurements (11). around radiation risk include: Are nuclear weapons tests
The ICRP’s recommended limits for workers have essential for national defense, and, if so, what are the ranges
decreased by a factor of 6 over the years, from 0.1 roent- of risks from fallout or accidental detonations? Is the soci-
gen/day (approximately 300 millisieverts (mSv)/year) in etal gain from electricity generated from nuclear power
1934 to 5 rem/year (50 mSv/year) in 1956 (1). Today, the plants sufficient to balance the risk associated with long-
ICRP would allow 50 mSv in 1 year as long as the yearly term storage of high level radioactive wastes or of low-
average over 5 years is not more than 20 mSv. probability but deadly Chernobyl-like disasters? Is screen-

Downloaded from https://academic.oup.com/aje/article-abstract/154/12/S91/64652 by Lincoln University user on 19 June 2020

Recommended dose limits by the NCRP are similar to those ing asymptomatic women for the early detection of breast
of the ICRP, although differing somewhat in form. The cancer with mammographic x-rays worth the risk of poten-
NCRP also would permit 50 mSv in 1 year as long as the tially inducing malignancies among healthy women? How
cumulative lifetime dose does not exceed 10 mSv multiplied should underground uranium miners be compensated for the
by the person’s age. Legislation by appropriate government likelihood that their lung cancer was caused by radon expo-
regulatory bodies often follows the recommendations set sure experienced in the mines? How does one compute the
forth by the ICRP and NCRP (12). probability that radiation was responsible for cancer in an
Both the ICRP and the NCRP rely heavily on the detailed individual who received exposure from fallout during
scientific reports that come from two sets of periodically con- nuclear weapons testing, as a resident in a fallout impacted
vened committees: the United Nations Scientific Committee area, or as a participant in the weapons test as a soldier? Is
on the Effects of Atomic Radiation (UNSCEAR) and the radon in homes a serious public health hazard, and should
National Academy of Sciences/National Research Council testing be mandatory? What is the risk of using plutonium
committees on the Biological Effects of Ionizing Radiation generators in spacecraft that might explode before leaving
(BEIR). In the 1950s there was growing public concern about the earth’s atmosphere? If genetic testing uncovers groups
radiation risks related in part to extensive nuclear weapons who are sensitive to the induction of cancer by radiation,
testing, such as in the Pacific Ocean where residents of sev- should radiation guidelines be lowered or should such indi-
eral atolls had substantial fallout exposure. UNSCEAR was viduals be prohibited from working with radiation?
created in 1955 by the United Nations and still periodically
reviews the extensive literature on the sources and effects of RADIATION SCIENCE AND POLICY TODAY
radiation exposure, publishing comprehensive reports. The
National Academy of Sciences/National Research Council Scientific data regarding radiation risks
committees on the Biological Effects of Atomic Radiation The basic mechanism by which ionizing radiation causes
published a series of reports during the 1950s. First convened damage to living matter is well known. Ionizing radiation
in 1954 to review the available scientific knowledge on the has sufficient energy to break the electronic bonds that join
effects of atomic radiation among living organisms, commit- atoms into molecules. Consequently, ionizing radiation can
tees on the Biological Effects of Atomic Radiation were ter- cause damage to living cells and organisms resulting in
minated in 1964 (13) and later replaced by BEIR. genetic effects, cancer, and other somatic effects. The mech-
In 1970, the Federal Radiation Council, whose activities anisms follow basic radiophysical principles, and there has
were later transferred to the US Environmental Protection been some understanding of the radiobiologic effects of the
Agency, asked the National Academy of Sciences to prepare agent for almost 100 years (3, 17, 18). Several questions still
a report on information relevant to an evaluation of then cur- remain, but the extent of understanding of this agent’s
rent radiation protection guidelines. The first BEIR report effects is much greater than for most chemicals.
was published in 1972 (14) in response to this request, and Epidemiologic data have driven the assessment of radia-
further reports have been published up to the present to tion risks (20, 21). Epidemiologic studies included occupa-
address the health effects of exposure of human populations tional groups, such as radiologists, radium dial painters, and
to low-dose radiation (4, 15–19). uranium miners, exposed to different forms of radiation (2,
The extensive uses of radioactive materials and of radi- 5–7, 10), as well as individuals exposed to radiation as a
ation across the last century led to development and sup- therapeutic agent (4, 22–25). The types of radiation expo-
port of an expansive community of radiation scientists and sure, the dose, and the duration of exposure differed in these
policy-makers, perhaps in part because of the ubiquitous studies. The studies came from different parts of the world,
but controllable nature of this exposure. Radiation is yet the data indicating a risk of cancer in humans were gen-
everywhere and can be found within the food we eat, the erally mutually compatible and consistent with findings
air we breathe, and the homes we live in. Even during air from animal exposures and laboratory experiments (3).
travel we are exposed to increased levels of cosmic rays. One of the major sources of epidemiologic data for risk
The medical uses of radiation are extensive, from diagnos- assessment has been the longtime prospective cohort study
tic applications to the treatment of malignant diseases. of the atomic bomb survivors in Japan (26). This study has
Nuclear energy continues to be a source of power through- proved invaluable because the population is large and
out the world to generate electricity. In order to balance the includes persons exposed at all ages, including in utero. Of
benefits of radiation uses with potential adverse health out- course, the exposure for this population was almost entirely

Am J Epidemiol Vol. 154, No. 12, Suppl., 2001

 Radiation Exposure and Cancer S93

to a single high dose of radiation. A select group of sur- Thus, the public is typically unquestioning of medical uses for
vivors, numbering approximately 87,000, has been followed screening, diagnosis, and therapy. Exposure to radiation from
since the 1950s; dose estimates have been derived for each many other sources, however, may be regarded with an
individual and the survivors are followed for incident can- almost unreasoning fear, as occurred, for example, with the
cers and other diseases and mortality (27, 28). Follow-up irradiation of food as a method of sterilization. This fear prob-
has been maintained to the present by the Radiation Effects ably originates, in part, with the mystique of radiation and the
Research Foundation and its predecessor, the Atomic Bomb involuntary nature of many exposures and the possibility,
Casualty Commission. The data are periodically analyzed to however remote, of a catastrophic disaster. In fact, the Three
track the cancer risk associated with radiation exposure. As Mile Island episode and the Chernobyl disaster support this
follow-up has lengthened, the data are increasingly informa- concern, providing widely publicized evidence that low prob-

Downloaded from https://academic.oup.com/aje/article-abstract/154/12/S91/64652 by Lincoln University user on 19 June 2020

tive concerning the time course of the excess cancer risk ability, but potentially lethal, events can take place. The
associated with radiation (29, 30). Risk estimates from the extensive scientific evidence on radiation risk has not suc-
cohort have been continually refined and updated as follow- cessfully dispelled these fears. Paradoxically, the information
up has lengthened, statistical analysis methods have become on the potential lung cancer risk from indoor radon has moti-
more sophisticated, and dose estimates have been revised. vated only a small minority of US homeowners to measure
The study remains the strongest source of data for estimat- radon levels in their homes and to carry out mitigation if
ing risks of low linear-energy transfer (LET) radiation. needed (31, 32).
Additionally, as more radiation-exposed populations have
been studied, risk estimates from them have been generally REMAINING ISSUES
consistent with those in the atomic bomb survivors.
Thus, in the case of ionizing radiation, there is a wealth of Science questions remain
scientific data on risk, supported by biologic theory and lab-
oratory data. The cancer risk is well accepted, although there Despite a wealth of epidemiologic data, supplemented by
is controversy over risks at low doses and dose rates. a large body of laboratory evidence on the effects of radia-
Differing risk for different organs is well established. tion at the molecular, cellular, and whole animal levels,
Perhaps the major controversy at present is the shape of the important questions remain with regard to the risk of cancer
dose-response curve at low doses and dose rates; this topic induction by radiation. The questions include a broad range
is being addressed by the current BEIR Committee (BEIR of issues from biologic effects, such as whether normal
VII). The epidemiologic data are increasingly informative as DNA repair mechanisms are sufficient to reverse radiation
follow-up of the existing cohorts continues. damage at the lowest doses and dose rates, to interactive
effects of radiation with other agents, to whether there are
unrecognized non-cancer, non-genetic endpoints.
Committees for science and policy reviews The most contentious issue in radiation risk assessment
revolves around the estimation of risks at very low doses and
Standing committees on radiation have continued to meet dose rates—small increments of exposure only slightly above
periodically and their reports have provided key guidance to natural background radiation. Risks in this dose range cannot
governmental agencies in the United States, Europe, and be studied with sufficient precision by direct epidemiologic
elsewhere. The BEIR and UNSCEAR committees, for investigation, although epidemiologic studies of populations
example, have reviewed and evaluated the scientific evi- accumulating moderate-to-high doses of radiation over an
dence, developing new risk models and risk estimates and extended time can inform the discussion of risks at low doses.
characterizing uncertainties. Committees such as the NCRP For public health protection, another important issue is the
and ICRP have used these risk models to assess whether cur- possible existence of groups more susceptible than average to
rent policies and standards set an acceptable level of safety the harmful effects of radiation. This sensitivity might arise
for the public and for workers. Typically, these committees from a genetic basis, from synergism of radiation with other
have included the full range of scientists involved in under- carcinogens, or from inherent biologic characteristics like age
standing radiation risks—epidemiologists and statisticians, and gender. It is likely that some individuals are genetically
radiobiologists, health physicists and dosimetrists, and oth- predisposed toward radiation-induced cancer. A number of
ers. The work of these committees has generally been neu- genes have been investigated and the new approaches of mol-
tral to policy implications and free of pressure from regula- ecular and genetic epidemiology should provide further
tions, the regulated community, and the public. More insights, particularly when combined with the technologic
recently, however, controversy concerning radiation risk has advances for studying the genetic basis of disease. Sensitivity
become stronger and more strident, even reaching into the through synergism is illustrated by the example of radon and
appointment of the BEIR VII Committee. smoking increasing the risk of lung cancer. Thus, the risks of
the mixed chemical and radiation exposures associated with
Public perception of radiation nuclear waste clean-up need to be better understood. Finally,
there is renewed interest in the possibility that moderate doses
The public has an understanding that radiation poses a can- of radiation might cause delayed, non-cancer health effects
cer risk, but the range of views extends from a general accep- such as cardiovascular disease.
tance of medical exposures to widespread concern about the There will always be some imprecision in radiation risk
risks of some sources, such as nuclear power generation. estimates and uncertainty associated with the models used to

Am J Epidemiol Vol. 154, No. 12, Suppl., 2001

S94 Matanoski et al.

derive the estimates. These uncertainties need to be fully ply unaware of the quantitative information available, it is
characterized so that their implications for risk management widely believed that the public also incorporates character-
are clear, both to policy-makers and the public. There are istics of risk, such as dread or unfamiliarity, that are not cap-
improving techniques for this purpose. Careful review and tured in the quantitative analyses (35). Whether or not such
integration of epidemiologic and radiobiologic results will behavior is “rational” is debated, but clearly it must be con-
lead to more certain models and accurate depiction of sidered in the development of policy.
assumptions and attendant uncertainties. A comprehensive It is also clear that the public is more accepting of radia-
listing of uncertainties offers an appropriate framework for tion risk (as quantitatively determined) in the medical arena
advancing a policy-relevant research agenda. The process of than in the environmental arena. Most of this difference is
uncertainty analysis also can be used to assess the relative probably attributable to the closer coupling of risks and ben-

Downloaded from https://academic.oup.com/aje/article-abstract/154/12/S91/64652 by Lincoln University user on 19 June 2020

importance of various sources of uncertainty, which can efits in the medical uses, involving a voluntary acceptance
help guide future research. of risk rather than the external imposition of risk. Whether
the advice of the physician is appropriately balanced in
regard to risks and benefits is a challenging question.
Policy issues remain
Qualitatively, the public may be less accepting of non-
Harmonizing radiation and chemical risk manage- medical radiation risks than of chemical and other risks,
ment. For historic reasons, radiation risk assessment and based on its relatively strong opposition to nuclear power
management have developed under a different paradigm than and its aversion to nuclear waste depositories. Whether epi-
is applied to risk assessment and management for chemical demiology has a role to play in clarifying this issue is
carcinogens. Data on chemical carcinogens are generally unclear, and educating the public on the extent of the evi-
derived from animal studies, while radiation data are mainly dence available and its subtleties may be difficult. Currently
from human studies of cohorts, such as the atomic bomb sur- available methods are unlikely to have the power to discern
vivors, uranium miners, and patients medically exposed to whether or not a reactor or waste depository is affecting the
radiation. The risk management approaches for chemicals and surrounding public. However, the more important question
radiation differ fundamentally in their goals. Chemicals are to answer is how large an effect might be credible in a given
evaluated individually, and the aim is to set standards to keep situation based on known scientific evidence updated con-
risks for each below 10–6 if possible. Occasionally it is not stantly with new information. If the possible risk is small,
possible to set standards to maintain this low risk because to then a definitive answer on whether an effect exists is not
do so would deny benefits to the population. In that case, meaningful. Unfortunately, model-based predictions that the
standards will be selected to keep risks at least below 10–4 at probability of cancer is low around such facilities do not
the most. The paradigm for radiation risk management is seem to have much weight with the public or even with
based on the principle of “as low as reasonably achievable” some health physicists.
(ALARA), and this principle is subscribed to by the National Lack of effective interaction between epidemiologists and
Research Council and US Department of Energy. After all policy-makers: practical implementation of policies. The
sources of ionizing radiation exposure are considered in set- interaction between epidemiologists and policy-makers
ting a maximum dose limit, ALARA is then applied to reduce could be improved. First, epidemiologists working on radi-
the risk, which involves “making every reasonable effort to ation risk should be cognizant of the likely application of
maintain exposures to ionizing radiation as far below dose their findings by policy-makers. Results should be presented
limits as practical…” (33). in such a way that the implications for policy are clear and
Difficulties have arisen when the US Environmental balanced. Uncertainties should be fully discussed and the
Protection Agency or other regulatory agencies have applied results placed in perspective with respect to those obtained
the paradigm for risk management of chemicals to a radia- from other studies. In fact, the relatively small cadre of epi-
tion issue that would have been handled under the radiation demiologists working on radiation generally have sensitiv-
paradigm. For example, applying US Environmental ity to these issues.
Protection Agency’s paradigm for chemicals and treating the Second, policy-makers may be confused by seemingly
current radiation dose rate limit of 100 millirems per year as conflicting epidemiologic results. Individual studies may be
a standard would result in an unacceptably elevated lifetime inappropriately presented in a dichotomous fashion as “posi-
risk of about 3 in 1,000 for exposures at the standard (34). tive” or “negative.” Epidemiologists can assist the policy-
This approach assumes that population exposure would be maker in evaluating the quality of studies in relation to their
at the dose limit of 100 millirems per year over a lifetime. interpretation for policy issues. Additionally, new approaches
Applying the radiation paradigm, however, exposures and to data synthesis by meta-analysis and pooling of data com-
associated risks would be much lower over the lifespan for bine evidence in a useful way for policy-makers and provide
most people with the ALARA principle in effect (34). a framework for evaluating the contribution of new evidence.
Influence of public perception of acceptable and actual Measurement error models provide insights into the degree to
risks of radiation on policy. Risk perception studies have which risk estimates are biased by the unavoidable misclassi-
consistently shown that the general public ranks radiation fication of exposure in epidemiologic studies.
risks higher than many risks that have lower rankings when Third, epidemiologists can be helpful to policy-makers at
evaluated with existing quantitative risk assessment tech- the point of implementation. In particular, epidemiologists
niques. Although it is clear that many of the public are sim- possess tools for assessing the population impacts of alter-

Am J Epidemiol Vol. 154, No. 12, Suppl., 2001

 Radiation Exposure and Cancer S95

native policy choices and can aid in communicating risk bring together multidisciplinary groups, and this diversity is
information to the public. critical for synthesis. Mechanisms need to be found for
encouraging this same breadth in research groups.
Scientists may confuse their role in policy debates and
LESSONS LEARNED
overstep by proposing what is proper for the public to know
The example of ionizing radiation offers several useful or what risk management option is optimal. Instead, scien-
“lessons” concerning epidemiology and policy. First, there tists should communicate openly and serve as a resource for
is a relative abundance of data, far more than will ever be the community and policy-makers in order to identify their
available for most chemical carcinogens. Nonetheless, new needs and concerns and to help answer those needs. This
interaction is critical for all parties.

Downloaded from https://academic.oup.com/aje/article-abstract/154/12/S91/64652 by Lincoln University user on 19 June 2020

questions have continued to challenge epidemiologists and
epidemiologic data. Over recent decades, emphasis has The field of radiation science has clearly been subject to
shifted from estimating risks to evaluating modifiers of the the dictum that “policy drives funding and funding drives
risks, particularly dose-level (low versus high) and dose- science.” Radiation sciences have advanced over the years
rate. Epidemiologic studies can contribute to these issues producing a wealth of knowledge, new methodologies, and
but complementary laboratory data are needed, particularly innovation and technology. However, support is waning
as laboratory findings point to mechanisms that should be even as questions about risk remain unanswered. The boom
incorporated into the structure of risk models. of research and research funding that was driven by medical
Second, the control of radiation exposures has been and military uses and power generation is long over. Much
improved by informed use of the epidemiologic evidence. The of our understanding of fundamental radiology and health
periodic review of the evidence by scientific review commit- physics was gained during the boom period. Key cohorts
tees has enhanced acceptance of both the scientific data and were initiated at that time. However, subsequent cycles and
regulatory standards arising from the data. Implementation of a general decline of support have reduced the role of
increasingly stringent standards has driven down radiation involved scientists, and few new researchers are entering
exposure levels, especially from occupational and environ- radiation sciences. Given the substantial lag needed to
mental sources. In addition, the development of effective real- renew the pool of scientists, there is the possibility of future
time monitoring systems for use in industry has resulted in crises when critical issues need rapid answers. Methods of
lowered occupational exposures. Although there may be ways providing stable scientific support are needed.
to bypass these monitoring systems and regulations, the more As we enter the next millennium there will undoubtedly be
effectively we can build information systems which provide new and difficult issues surrounding the uses of radiation, and
ready access to exposure data for individuals and businesses, the interface between policy and science will remain critical
the harder it is to escape the protective regulations. for protecting the public’s health. As before, judgments on
Third, we have learned that communication of scientific radiation issues will take into account the best scientific evi-
results and the related uncertainties in risk numbers is a dif- dence available, and there will be a continuing need for evi-
ficult and only partially solved problem. Scientists may be dence on radiation risks. To this end, it is encouraging that the
ill-equipped to approach these issues as few are formally Department of Energy has initiated a new program to train
trained in risk communication, having learned instead radiation scientists at the University of Pittsburgh, that
through experience. Only now is the scientific community Harvard University has expanded its program in radiation
beginning to address the issues of effective communication research, that the National Cancer Institute and the Centers for
and what are the factors, the situations, and the tools by Disease Control and Prevention have retained programs in
which scientific information can be best transmitted (36). radiation epidemiology, and that the ICRP, NCRP, UNSCEAR
Communicating the concept of uncertainty may be particu- and BEIR committees remain strong and influential.
larly challenging for scientists, as well as reaching a “bot-
tom line” in the face of uncertainty (37). Different audiences RECOMMENDATIONS
will have differing needs to understand uncertainty and its
implications (36). Thus, needs of fellow scientists, policy- Changing technologies and applications in radiation sci-
makers, the media, and the public may differ substantially. ence will bring emerging issues to the forefront. Over 50
The media may not be trying to gain knowledge but may be years after the development of massive nuclear weapons
seeking information which can be “used” to justify the complexes in the United States, the former Soviet Union,
thrust of a story, such as the need to clean up an existing and other countries, there is now the challenge of safely
waste dump (37). The community stakeholders may be managing wastes, assuring that workers are not placed at
seeking firm answers and not explanation and calls for fur- undue risk, and that radioactive materials can be safely
ther research. stored for centuries. Because the potential exposures from
As in most areas of research dealing with environmental radioactive waste depositories are likely to be at low doses
problems, a multidisciplinary approach is essential. and dose rates, answering questions about this area of the
However, science has long been compartmentalized into dose-response relation becomes even more important.
disciplines that have their own culture, language, and fund- Meanwhile, the expanding use of radionuclides and external
ing sources. There is little incentive offered to diverge from radiation in diagnostic imaging and nuclear medicine (radio-
this structure and to address problems in a truly multidisci- pharmaceuticals) challenges policy-makers with decisions
plinary fashion. Committees like the BEIR Committees about risk-benefit trade-offs in the medical arena.

Am J Epidemiol Vol. 154, No. 12, Suppl., 2001

S96 Matanoski et al.

With such emerging issues still to be addressed, several ment groups when the public believes that these organiza-
recommendations are offered based on the lessons learned tions have behaved inappropriately in the face of a hazard.
over the past 70 years of radiation science and policy-making. The experience of confronting these angry audiences is dis-
maying to many scientists who may then avoid further such
1. A stronger emphasis on helping epidemiologists and interactions even though there is a pressing need for scien-
other radiation scientists to more effectively tists who can clearly and credibly articulate the public health
communicate with the media, the public, and issues entwined in controversial situations. Epidemiologists
policy-makers would benefit from basic training in how to constructively
engage groups in emotionally charged settings, and from
Epidemiologic research addresses questions about health exposure to “lessons learned” anecdotes from colleagues or

Downloaded from https://academic.oup.com/aje/article-abstract/154/12/S91/64652 by Lincoln University user on 19 June 2020

among human populations; epidemiologic studies are often professionals who work frequently in the public realm.
of interest to the general public, and research results are
increasingly reported in the mass media. Consequently, epi-
demiologists are frequently called upon to present their find- 2. Change organizational structures to support/promote
ings directly to audiences of affected stakeholders or to interdisciplinary teams
print, television, and radio reporters. Government officials
and decision-makers rely heavily on the news media for Many of the problems probed by epidemiology are com-
information related to science and health policy. Arguably, plex, multifaceted issues that require investigation by, and
the most direct route to improving the translation of epi- collaboration among, scientists and experts from several dif-
demiology into sound public policy is to emphasize the ferent technical disciplines. The fruits of such multidisciplin-
importance of communications skills in epidemiologic train- ary efforts can be bountiful, but the logistical and emotional
ing programs and to offer practicing scientists continuing difficulties of working in a team environment is often over-
education in such skills. looked. Scientists from different disciplines may speak dif-
The ability to convey clearly and succinctly the results of ferent languages and work from different assumptions.
scientific research verbally and in writing has become a criti- Communication between radiation epidemiologists and
cal skill for epidemiologists. As reporters with limited techni- health physicists, for example, is often problematic. Even
cal backgrounds cover stories with epidemiologic content or seemingly simple concepts, such as “radiation dose,” may
consult epidemiologists for “expert” views on newsworthy pose significant translation problems that take time and
science, and as policy-makers rely more and more on the mass patience to resolve. Such realities of interdisciplinary work
media as a primary source of scientific knowledge, it is impor- need to be reflected in timetables for research and scientists’
tant that epidemiologists become proficient at interacting with schedules.
members of the media, lest distorted or misinterpreted science In spite of the fact that most epidemiologic studies with
become the basis for public belief or policy decisions. The implications for public policy are multidisciplinary investiga-
idea is not to make every holder of a Ph.D. the host of their tions, the administrative and incentive systems of academia
own Discovery Channel program, but to acquaint students remain narrowly focused on specific disciplines. Existing
with the rules of media engagement, to equip them with basic institutional structures and expectations may discourage
public speaking skills, and afford them some practice address- rather than support the teamwork that is increasingly a central
ing groups in front of microphones and cameras with the aspect of epidemiologic research. Thus, promotion and tenure
opportunity to review videotapes of their performance. depend on publication in specialty journals and citations as
Epidemiologists consider that journalists should be a tar- the “first author” of peer reviewed studies. Contributions
get for education. Teaching opportunities might include made as part of a team that undertakes a significant multidis-
developing classes for journalism students. Symposia tar- ciplinary research project are not accorded the same respect.
geted at science writers are conducted by special groups but Similarly, active participation in policy-making, such as ser-
this effort should be expanded to encompass small newspa- vice on expert panels, providing Congressional testimony, or
pers and media groups. This may be an effective way of pro- aid in drafting legislative language are not generally recog-
viding education before a controversial issue challenges nized as worthy credit or reward within academia. Funding
rational responses. sources are keyed to discipline-specific departments and divi-
Communication is a two-way street. Policy-makers and sions, and it is often difficult to “share” money or research
the public would benefit from more skilled and attentive lis- support among institutional duchies and kingdoms. Scientific
tening on the part of the scientific community. Scientists journals are devoted to specific disciplines and sub-
engaged in epidemiologic research should carefully con- disciplines and rarely feature in-depth examinations of all
sider the impact that proposed studies might have on facets of complex scientific and policy issues.
affected communities and arrange to communicate the intent Universities and funding sources should re-examine how
and limitations of studies before they are initiated. As epi- their organizational structures and practices support or
demiologists pursue questions entwined with controversial penalize multidisciplinary research. Consideration should
policy issues, they must learn to interact constructively with be given to supporting conferences, journals, and seminars
worried community members and skeptical representatives that explicitly consider the multidisciplinary nature of many
of the press. The problem is that public concern may not be important epidemiologic studies and potential studies and
a response only to hazard but a combination of “hazard and seek ways of encouraging and supporting scientists from
outrage.” The outrage arises against companies or govern- different fields to work together efficiently.

Am J Epidemiol Vol. 154, No. 12, Suppl., 2001

 Radiation Exposure and Cancer S97

3. Support forums and journals that address science and then at the National Cancer Institute where he devel-
and policy aspects of complex public health problems oped and headed the Radiation Epidemiology Branch for 12
years. He is an international authority on radiation effects
The absence of a common forum wherein scientists and and serves as advisor to the United Nations. He is on the
policy-makers could examine and review the different facets main commission of the International Commission on
of complicated public health issues has limited the useful- Radiation Protection, on the National Council for Radiation
ness of science to decision-makers and stunted the creation Protection and Measurements, and on the radiation working
of innovative policy. group of the International Agency for Research on Cancer.
It is difficult for anyone to obtain a comprehensive and Dr. Boice holds a faculty appointment as Professor of
coherent picture of the scientific and policy aspects of most Medicine, Vanderbilt University and Vanderbilt-Ingram

Downloaded from https://academic.oup.com/aje/article-abstract/154/12/S91/64652 by Lincoln University user on 19 June 2020

major public health issues. There is, for example, no simple Cancer Center.
way to review the current debate about “mercury toxicity” Dr. Stephen Brown is an independent consultant in risk
that has been occupying many university scientists and assessment and risk management, who has worked on a
policy-makers at the US Environmental Protection Agency wide spectrum of issues involving radiation and chemical
short of reading hundreds of pages of testimony and gov- compounds in the environment. His clients have included
ernment documents. Even then, one is left to infer most of federal and local governments as well as a variety of
the policy issues and options and to chase down unclear sci- private-sector clients in such industries as agribusiness,
entific references. Scientists often present data with only a chemical manufacturing, petroleum, paper, electric power,
sketchy understanding of policy concerns or the available mining and milling, foods, flooring materials, and waste dis-
options for government action. Decision-makers for their posal. He has served as an expert witness for both plaintiffs
part frequently pursue policy “solutions” with an incomplete and defendants. His former employers include the Stanford
or erroneous understanding of the underlying science. If Research Institute, the National Academy of Sciences,
researchers do not fully comprehend the pertinent policy ENVIRON Corporation, and ENSR Consulting and
questions and options, they may pursue questions that are Engineering. He was a member of the US Environmental
not relevant to the pending decisions or are unnecessarily Protection Agency’s Science Advisory Board for 9 years,
detailed, too imprecise, or otherwise off the mark. Decision- with a 2-year term as chair of the Radiation Advisory
makers must have sufficient knowledge of science to engage Committee there. He has served on other scientific advisory
in a constructive dialogue about what studies might be use- committees for the Science Advisory Board, the National
ful in formulating sound public health policy, and must Academy of Sciences, and the State of California.
understand the limits of available knowledge and attendant
Dr. Ethel Gilbert is a biostatistician in the Radiation
uncertainties.
Epidemiology Branch of the National Cancer Institute. Her
research has included studies of nuclear workers, radiation
risk assessment, and studies of second cancers after radio-
and chemotherapy. Dr. Gilbert currently serves on the
DISCLOSURES National Academy of Sciences Committee on Health Risks
from Exposure to Low Levels of Ionizing Radiation (BEIR
Dr. Genevieve Matanoski is Professor of Epidemiology at VII), and previously served on the Committee on Health
the Johns Hopkins Bloomberg School of Public Health. A Effects of Exposure to Radon (BEIR VI). She is a fellow
major portion of her epidemiologic research studies have of the American Statistical Association and a member of
been focused on environmentally-induced cancers, espe- the National Council on Radiation Protection and
cially those related to radiation exposure. The studies have Measurements.
included risks to workers from ionizing radiation in occupa- Dr. Tara O’Toole is currently the Deputy Director of the
tional settings in medical facilities or shipyards where Johns Hopkins University Center for Civilian Biodefense
nuclear-powered ships were overhauled. Her studies of risks Studies and a member of the faculty of the Bloomberg
in children from therapeutic radiation has also offered School of Public Health. The Center is dedicated to inform-
important public health information regarding medical uses ing policy decisions and promoting practices that would
of radiation in treatment of the young. Dr. Matanoski has help prevent the use of biologic weapons. From 1993 to
served on several scientific committees which have 1997, Dr. O’Toole was Assistant Secretary of Energy for
reviewed the hazards associated with radiation and other Environment Safety, serving as principal advisor to the
environmental issues, including the Science Advisory Board Secretary of Energy on matters pertaining to protecting the
of the US Environmental Protection Agency where she environment and workers and public health from US
served as Chair of both the Radiation Subcommittee as well Department of Energy operations. During her tenure, she
as the Executive Committee. She has testified before conducted four major “Vulnerability Studies” that identified
Congressional committees of both the House and Senate on major safety and environmental hazards at the nation’s
the hazards of radon as a public health issue and the imple- nuclear weapons complex, and led a multi-agency, multi-
mentation of recommendations for household testing. million dollar task force that oversaw the government’s
Dr. John Boice is Scientific Director at the International investigations into human radiation experiments conducted
Epidemiology Institute. He served 27 years in the US Public during the Cold War. From 1989 to 1993, Dr. O’Toole was
Health Service, first at the Food and Drug Administration a Senior Analyst at the Congressional Office of Technology

Am J Epidemiol Vol. 154, No. 12, Suppl., 2001

S98 Matanoski et al.

Assessment where she directed and participated in studies of National Academy Press, 1977.
health impacts on workers and the public due to environ- 16. National Research Council, Committee on the Biological
Effects of Ionizing Radiations. The effects on populations of
mental pollution resulting from nuclear weapons produc- exposure to low levels of ionizing radiation, 1980: BEIR III.
tion, among other projects. She has served as a consultant to Washington, DC: National Academy Press, 1980.
industry and government in matters related to occupational 17. National Research Council, Committee on the Biological
and environmental health, worker participation in work- Effects of Ionizing Radiations. Health effects of exposure to
place safety protection, and organizational change. low levels of ionizing radiation: BEIR V. Washington, DC:
National Academy Press, 1990.
Dr. Jerome Puskin is the Director of the Center for 18. National Research Council, Committee on Health Risks of
Science and Risk Assessment in the Office of Radiation and Exposure to Radon. Health effects of exposure to radon: BEIR
Indoor Air at the US Environmental Protection Agency. In VI. Washington, DC: National Academy Press, 1999.

Downloaded from https://academic.oup.com/aje/article-abstract/154/12/S91/64652 by Lincoln University user on 19 June 2020

this capacity, he is responsible for making quantitative esti- 19. National Research Council, Committee on Health Risks of
Exposure to Radon. Health risks from exposure to low levels
mates of the health risks from ionizing radiation and their of ionizing radiation: BEIR VII, Phase I, Letter report.
uncertainties and for developing documentation which Washington, DC: National Academy Press, 2000.
explains the scientific basis for these estimates. These esti- 20. Boice JD Jr, ed. Implications of new data on radiation cancer
mates and documentation are used to support US risk. Proceedings of the 32nd annual meeting of the National
Environmental Protection Agency actions to limit public Council on Radiation Protection and Measurements,
Arlington, Virginia, April 1996 (proceedings no. 18).
exposure to radiation. Bethesda, MD: National Council on Radiation Protection and
Measurements, 1997.
21. Samet JM. Epidemiologic studies of ionizing radiation and
cancer: past successes and future challenges. Environ Health
Perspect 1997;105:883–9.
REFERENCES 22. Shore RE, Hildreth N, Dvoretsky P, et al. Thyroid cancer
among persons given x-ray treatment in infancy for an
1. Caufield C. Multiple exposures: chronicles of the radiation enlarged thymus gland. Am J Epidemiol 1993;137:1068–80.
age. London, UK: Secker and Warburg, 1988. 23. Ron E, Modan B, Preston D, et al. Radiation-induced skin car-
2. Martland HS, Humphries RE. Osteogenic sarcoma in dial cinomas of the head and neck. Radiat Res 1991;125:318–25.
painters using luminous paint. Arch Pathol 1929;7:406–17. 24. Nekolla EA, Kreisheimer M, Kellerer AM, et al. Induction of
3. Stannard JN. Baalman RW Jr, eds. Radioactivity and health: a malignant bone tumors in radium-224 patients: risk estimates
history. Vol 1: Laboratory research. Columbus, OH: Battelle based on the improved dosimetry. Radiat Res 2000;153:93–103.
Press, 1988. 25. Hildreth NG, Shore RE, Dvoretsky PM. The risk of breast can-
4. National Research Council, Committee on the Biological cer after irradiation of the thymus in infancy. N Engl J Med
Effects of Ionizing Radiations. Health risks of radon and other 1989;321:1281–4.
internally deposited alpha-emitters: BEIR IV. Washington, 26. Schull WJ. Effects of atomic radiation: a half-century of stud-
DC: National Academy Press, 1988. ies from Hiroshima and Nagasaki. New York, NY: Wiley-Liss,
5. Henshaw PS, Hawkins JW, Meyer HL, et al. Incidence of 1995.
leukemia in physicians. J Natl Cancer Inst 1944;4:339–46. 27. Pierce DA, Shimizu Y, Preston DL, et al. Studies of the mor-
6. March HC. Leukemia in radiologists. Radiology 1944;43: tality of atomic bomb survivors. Report 12, part I: Cancer,
275–8. 1950–1990. Radiat Res 1996;146:1–27.
7. Ulrich H. The incidence of leukemia in radiologists. N Engl J 28. Thompson DE, Mabuchi K, Ron E, et al. Cancer incidence in
Med 1946;234:45–6. atomic bomb survivors. Part II: Solid tumors, 1958–1987.
8. Folley JH, Borges W, Yamawaki T. Incidence of leukemia in Radiat Res 1994;137(2 suppl):S17–67.
survivors of the atomic bomb in Hiroshima and Nagasaki, 29. Little MP, de Vathaire F, Charles MW, et al. Variations with
Japan. Am J Med 1952;13:311–21. time and age in the risks of solid cancer incidence after radia-
9. Leukaemia and aplastic anaemia in patients irradiated for tion exposure in childhood. Stat Med 1998;17:1341–55.
ankylosing spondylitis. London, UK: Medical Research 30. Ron E, Preston DL, Mabuchi K, et al. Cancer incidence in
Council, 1957. atomic bomb survivors. Part IV: Comparison of cancer inci-
10. Wagoner JK, Archer VE, Carroll BE, et al. Cancer mortality dence and mortality. Radiat Res 1994;137(2 suppl):S98–112.
patterns among US uranium miners and millers, 1950 through 31. Radon testing in households with a residential smoker—
1962. J Natl Cancer Inst 1964;32:787–801. United States, 1993–1994. MMWR Morb Mortal Wkly Rep
11. Stannard JN. Baalman RW Jr, eds. Radioactivity and health: a 1999;48:683–6.
history. Vol 3: Applied aspects, instrumentation, and conclu- 32. Ford ES, Eheman CR, Siegel PZ, et al. Radon awareness and
sions. Columbus, OH: Battelle Press, 1988. testing behavior: findings from the Behavioral Risk Factor
12. Title 10. Energy, Chapter 1. Nuclear Regulatory Commission, Surveillance System, 1989–1992. Health Phys 1996;70:363–6.
Part 20. Standards for protection against radiation, Subpart C. 33. Title 10. Energy, Chapter 1. Nuclear Regulatory Commission,
Occupational dose limits, Sec. 20.1201. Occupational dose Part 20. Standards for protection against radiation, Subpart A.
limits for adults, 10 CFR 20.1201. General provisions, Sec. 20.1003. Definitions, 10 CFR
13. Committees on Biological Effects of Atomic Radiation, 20.1003.
1954–1964. November 2000. (http://www4.nationalacade- 34. Tran NL, Locke PA, Burke TA. Chemical and radiation envi-
mies.org/arc.nsf/44bf87db309563a0852566f2006d63bb/ced1c ronmental risk management: differences, commonalities, and
a6c88f1a9288525689c005a42ae?OpenDocument). challenges. Risk Anal 2000;20:163–72.
14. National Research Council, Committee on the Biological 35. Slovic P. The perception of risk. London, UK: Earthscan
Effects of Ionizing Radiations. The effects on populations of Publications, 2000.
exposure to low-levels of ionizing radiation: BEIR I. 36. National Research Council, Committee on Risk Perception
Washington, DC: National Academy Press, 1972. and Communication. Improving risk communication.
15. National Research Council, Committee on the Biological Washington, DC: National Academy Press, 1989.
Effects of Ionizing Radiations. Considerations of health 37. Cohn V. News and numbers: a guide to reporting statistical
benefit-cost analysis for activities involving ionizing radiation claims and controversies in health and other fields. Ames, IA:
exposure and alternatives: a report: BEIR II. Washington, DC: Iowa State University Press, 1989.

Am J Epidemiol Vol. 154, No. 12, Suppl., 2001