Potential Radiation Exposure in Military Operations: Protecting the Soldier Before, During, and After (1999)

To understand how to protect soldiers from ionizing radiation*, it is necessary to understand its characteristics, how it interacts with tissues in the body, and the effects that these interactions may have on immediate and long-term health.

All matter is made up of atoms, and each atom consists of a nucleus with neutrons and positively charged protons. Negatively charged electrons surround the nucleus. The nucleus of a radioactive atom has excess energy that causes it to be unstable. To become more stable, the radioactive nucleus will eventually release energy in the form of either particles with mass (e.g., alpha and beta particles) or electromagnetic waves (e.g., gamma and x rays).

When these forms of radiation strike atoms of any material, they may have enough energy to eject electrons, thus resulting in the creation of charged ions. This process, called ionization, can result in the breaking of the electron bonds that hold atoms together. Ionization and other radiation-induced effects, such as excitation and free radical formation, cause chemical changes in components of the living cell, including chemicals, such as deoxyribonucleic acid (DNA), the genetic material that is located in the chromosomes within the cell nucleus.

Alpha radiation colliding with atoms gives up its energy in a very short distance, such as the thickness of a sheet of paper, less than the thickness of skin, or a few centimeters of air. Consequently, alpha particles emitted by radioactive materials are not likely to be harmful when striking the outside of a human body

that is protected by clothing and the outermost dead layer of the skin. However, when these same alpha-emitting radionuclides are taken into the body, their emissions can directly irradiate nearby cells of tissues in which they are deposited and may cause cellular changes. Such changes may result in adverse health effects in the short or long term, depending on the nature of the changes. Alpha-emitting radionuclides may be encountered in contamination created by intentional or accidental dispersion of nuclear weapon-source materials (e.g., plutonium-239) or as a result of a nuclear detonation. Alpha-emitting radionuclides, such as radium in soil and radon in air, are also naturally occurring sources of radiation and contribute to normal background levels.

In comparison to alpha radiation, fast-moving electrons, which are known as beta particles, have much smaller mass and electric charge, are more deeply penetrating, and dissipate their energy over a larger volume of tissue. Even high-energy beta particles, however, will transfer most of their energy and come to a stop within about 1 centimeter of plastic, 1 to 2 centimeters of tissue, or 4 to 5 meters of air. Therefore, beta particles that strike the outside of the body will penetrate only a short distance, but they may travel far enough to damage the actively dividing cells of the skin. Beta-emitting radionuclides are of most concern after they have entered the body and can transfer their energy to nearby cells of internal organs. Beta-emitting radionuclides may be found in contamination consisting of fission products from a nuclear detonation or resulting from the dispersion of nuclear reactor waste or radiotherapy sources (e.g., cesium- 137 and cobalt-60).

Gamma rays and x rays, which are emitted from radionuclides as well as produced by machines, are the most penetrating forms of ionizing radiation and consist of electromagnetic energy. While randomly colliding with electrons in the body along a scattered path length, gamma rays may give up all or part of their energy in tissue or, although it is unlikely, they may pass all the way through the body without interacting. Therefore, exposure to gamma or x rays from sources outside the body may cause ionizations in tissues at any location in their path. Gamma rays are characteristic of a wide variety of radioactive contaminants associated with nuclear weapons and nuclear waste and also with radioactive sources used in medicine and industry, whereas x rays are most commonly encountered in the use of radiation-producing equipment used in medical applications (including those in combat medical facilities).

The energy of ionizing radiation is measured and described in a number of ways. One can use a survey meter or other device to measure exposure-ionization in air caused by radiation. Exposure is measured in coulombs per kilogram

(C kg^-1) of air, formerly* the roentgen (R). This measurement of exposure applies only to ionizing electromagnetic radiation, such as gamma and x rays, not to particulate radiation (e.g., alpha or beta particles). In the field (outside the laboratory), exposure is the quantity that is measured, although for convenience, it is commonly assumed that exposure and absorbed dose (see below) are the same when expressed in traditional units (i.e., I R = 1 rad).

Although beta and alpha radiations can be detected in the field, determination of their contribution to tissue dose is a complex process not reasonably implemented except under laboratory conditions. Exposure to alpha- and beta-emitting radionuclides, expressed in terms of their intake, is related to their concentrations in air, food, and water. The primary dose to persons exposed to these concentrations results from ingestion and inhalation of the radionuclides.

A useful quantity in radiation physics is the energy actually deposited in a certain amount (mass) of tissue. This unit is referred to as absorbed dose. The unit of absorbed dose is the gray (Gy), formerly the rad; the gray is equivalent to the absorption of one Joule of energy per kilogram. One gray equals 100 rad; 1 milligray (mGy) equals 100 millirad (mrad). However, the amount of energy deposited in tissue does not account for differences in the biological effects of different radiation types.

The dosimetric quantity that accounts for the differences in biological effectiveness of various types of radiation and that allows doses from different radiations to be combined, through expressing their health effects on a common basis, is called the equivalent dose. It is calculated by multiplying the absorbed dose by the appropriate radiation weighting factor, "w_R" (ICRP, 1991a). For example, the factor for alpha particles is 20 and that for gamma and beta radiation is 1, indicating that it requires the absorption of about 20 times more energy from gamma or beta radiation than alpha radiation to cause a given biological effect. These weighting factors are approximate and the true value for a given type of radiation, radiation effect, or specific population can vary by up to an order of magnitude. The unit of equivalent dose is the sievert (Sv), formerly the rem. One sievert equals 100 rem; 1 millisievert (mSv) equals 100 millirem (mrem).

Just as different radiation types are more or less effective in damaging tissue, different tissue types have various sensitivities to that damage. For a given equivalent dose of radiation, the more sensitive tissues show a larger increase in cancer and leukemia rates than do less sensitive tissues. For radiation protection purposes, the International Commission on Radiological Protection (ICRP) has developed weighting factors for tissues (called "w_T") that describe the relative sensitivities of different tissues to long-term effects. Tissue weighting factors facilitate the combination of doses to allow a quantitative comparison of the long-term risk from partial body exposure to that from total body exposure. Tissues that are very sensitive to long-term effects from radiation have high weighting factors (e.g., bone marrow w_T= 0.12), whereas less sensitive tissues have lower weighting factors (e.g., skin w_T = 0.01).

The effective dose (that is, the dose to the whole body that represents an equivalent risk) is estimated by multiplying the equivalent dose in each tissue type by its corresponding tissue weighting factor and summing these weighted equivalent tissue doses. This composite dose is proportional to the increased risk from cancer and genetic effects. Like the equivalent dose, the effective dose is expressed in units of sievert or millisievert. Dose limits set for occupational exposures are expressed as effective dose and include the sum of the internal and external doses. Table 2-1 compares the characteristics of the three ways in which dose in biological tissue may be expressed.

It is critical that radiation measurement equipment be suited to its measurement task. Important considerations are the accuracy and sensitivity of the instrument chosen. Alpha, beta, and gamma (or x-ray) radiation measurements each require different instruments because of the way in which each radiation type interacts with matter. An instrument designed for alpha-radiation detection, for example, will not give accurate information for the other types of radiation. A radiation safety program specifies the appropriate equipment to be used to estimate an individual's level of exposure to radiation from external sources.

For the direct measurement of individual doses of gamma radiation (and, under some conditions, beta radiation), a thermoluminescent dosimeter (TLD) is often used. The TLD will give a reasonable measure of the dose to the whole body from gamma rays from a broadly distributed source. Because of the short range of beta particles, however, a TLD will indicate only the dose received from this type of radiation in its immediate location.

Doses to individuals may also be calculated indirectly if exposure rates are known from radiation surveys (radiation measurements made in the field). Two types of radiation survey instruments are helpful for assessing the potential for exposure to military personnel in the field. The first type measures the radiation exposure or dose to which personnel may be subjected. This category of instrument includes devices such as microroentgen meters and ion chambers. The second type of meter is represented by Geiger-Mueller (GM) or sodium iodide detectors. These meters are used to find contamination, although the GM detector may be calibrated to provide exposure readings. The conversion must take into account the efficiency of the probe and a number of other factors.

An ion chamber is designed to measure exposure, that is, ionization in air due to gamma rays (in coulombs per kilogram or roentgen). This instrument measures the quantity of radiation energy at a point in the air. Ion chambers normally come equipped with a moveable cover over the detection chamber. When the cover is opened, the instrument will respond to beta, as well as gamma, radiation. However, these instruments are not usually calibrated for beta radiation, so the instrument reading may not be accurate for them.

A GM detector is primarily designed to measure the number of alpha, beta, or gamma rays that emanate from a source and strike the detector in a given time. This meter does not normally provide information about the energy of incident radiation or about exposure. However, it can be calibrated to relate the number of gamma rays to a known ionization in air to give readings in units of coulombs per kilogram (or roentgen).

The devices briefly discussed above are useful for detecting or measuring contamination on surfaces (e.g., on the ground or on a vehicle such as a tank), but they cannot directly detect low levels of airborne radioactivity that might be hazardous. To determine whether airborne contamination is a health problem, an additional device—the air sampler—is required. Through the use of a filter or by

impaction on a collection plate, this device removes solid radioactive particles from the air and concentrates them sufficiently to be measured by a detector similar to those discussed above. Such measurements, however, must frequently be made in the laboratory.

Determination of internal doses resulting from exposures to inhaled or ingested radionuclides is much more difficult and time-consuming than determining external dose. It requires measurements of levels of air (or water or food) contamination, identification of significant radionuclides, measurement of the amounts excreted by the exposed person, such as in the urine and feces, and the application of sophisticated biomathematical models to determine doses to specific organs. Gamma-emitting radionuclides deposited in the body can be detected and measured with instruments external to the body, for example, through use of a whole-body counter. Under battlefield conditions, rough measurements of environmental contamination can be made as a basis for estimating both internal and external doses. If calibration factors are available for open-window ion chambers and GM counters, those instruments may be used to obtain a very crude estimate of airborne contamination. Under less adverse conditions, more sophisticated instrumentation and techniques should be applied.

In this section, the committee provides a perspective for considering the radiation doses soldiers could receive in the course of military operations. Under normal conditions, everyone is exposed to background ionizing radiation from two major sources: continuous, naturally occurring radiation from space and radiation from radioactive elements and technology-enhanced (often referred to as "man-made") radiation sources. Natural sources of radiation constitute the major source of radiation exposure to the populations of most, if not all, countries, with the next largest source being applications of medical technology.

In the United States the average annual effective dose of naturally occurring background radiation is about 3 mSv (0.30 rem) per year (NCRP, 1987a). A major portion of this arises through internal exposures, namely, 2 mSv (0.20 rem) from airborne radon and its decay products, and 0.39 mSv (0.039 rem) from naturally occurring radionuclides in the human body. The remainder comes from external sources, namely, 0.28 mSv (0.028 rem) from cosmic radiation and an equal amount from naturally occurring radioactive materials in the ground (terrestrial). The effective dose from all natural sources during a 70-year lifetime is approximately 200 mSv (20 rem). Levels of background radiation vary significantly across geographic areas. In the United States, for example, the effective dose for natural background radiation from cosmic rays and terrestrial sources in Denver, Colorado, is 50 percent higher (NCRP, 1987b) than the national average.

In addition to the doses from background radiation, some soldiers are engaged in duties in which they are at risk of exposure to higher levels of ionizing

radiation. Examples of such duties include repairing and maintaining radioactive commodities (e.g., ammunition containing depleted uranium and luminescent sights containing tritium), flying at high altitudes, and administering radiation for medical diagnosis and therapy. Table 2-2 shows the distribution of occupational doses for Army radiation workers.

Apart from routine occupational exposures, the only exposure of large numbers of U.S. military personnel to radiation has been to the approximately 400,000 service members who either were in the occupation forces near Nagasaki and Hiroshima, Japan, at the close of World War II or participated in the aboveground nuclear test program conducted between 1945 and 1962. Of the 202,000 military personnel at test sites (VA, 1997), about 1,750 received doses that were estimated to exceed 50 mSv (5 rem) (DSWA, 1995a)—the present annual dose limit set by the U.S. Nuclear Regulatory Commission (CFR, 1991) for individuals occupationally at risk of exposure to radiation. About 20,000 participants (DSWA, 1995b) have been assigned estimated doses that exceed the more conservative annual occupational limit—20 mSv (2 rem)—recommended by the International Commission on Radiological Protection (ICRP, 1991a). A total of 0.07 percent of the doses exceeded 100 mSv (10 rem); the average estimated dose for an Atomic Veteran is 6 mSv (0.6 rem).

There are three primary means of reducing the radiation dose from sources external to the body: time, distance, and shielding.

For a given source of radiation, the amount of radiation energy deposited in the body is related to how long one is exposed. Therefore, reducing the duration of an individual's exposure to radiation will decrease the dose.

Increasing the distance between an individual and a radiation source is an important means of reducing radiation exposure, because the intensity of the radiation is inversely proportional to the square of the distance from the radiation source. For example, when the distance from a localized source is doubled, the intensity of the radiation is reduced by a factor of 4 (22). When dealing with planar sources such as radioactive fallout on the ground, the decrease in dose with distance is much less.

Shielding is useful for absorbing radiation energy. If enough interactions occur in the shielding material, then much of the radiation is prevented from reaching the body's tissues. Alpha radiation can be stopped by a piece of paper. Beta radiation can be blocked by about a centimeter of plastic. Clothing and the outer layers of skin cells provide some protection from beta radiation outside the body. Gamma radiation, however, may require many centimeters of lead or meters of concrete for shielding.

Once a radioactive material is taken into the body, the protective measures of distance and shielding cannot be applied. However, the duration of internal exposure may be reduced by increasing the rate of excretion of the radioactive material through elimination of body fluids or solids. Increasing the rate of elimination is very specific to the radionuclide and its chemical form. It can be done for some radionuclides (e.g., tritium and iodine) by increasing the amount of fluids entering the body. For other radionuclides (e.g., plutonium) potentially toxic cheating agents can be considered.

The primary means of protection from internal radiation exposure is to prevent radioactive materials from entering the body in the first place. Appropriate respiratory protection can prevent the inhalation of airborne radioactive materials. Ingestion is prevented by not eating, drinking, or smoking where radioactive materials are present.

The most critical target of ionizing radiations passing through living tissues is generally accepted to be the DNA that constitutes the genes in the nucleus of every cell. Ionizing radiation can damage DNA directly or indirectly. For direct damage to occur, the radiation must hit this genetic material. Since the volume of the DNA is very small compared with the total volume of the cell, the probability of this occurring is low. Indirect damage occurs when radiation interacts in close proximity to the genetic material—the interaction can create in cellular water a free radical that can subsequently damage DNA. Two-thirds of the tissue damage created by radiation is caused by these indirect processes.

Complete and accurate biological repair of such DNA damage is a normal process that occurs millions of times daily. However, radiation-induced DNA damage can be irreparable, or the repair can be incomplete or inaccurate. This can result in the appearance of acute adverse health effects (within about 2 months) or delayed effects (over many years or even decades after the exposure).

Most radiation at environmental levels (background) does not result in detectable health effects. The reason is that most radiation interactions occur in the water in the cells of the body, producing free radicals that rapidly dissipate without doing biological damage.

Generally, when radioactive contaminants enter the body, the radionuclides are not uniformly distributed. As a result the dose may be highly localized. Uniform irradiation of the entire (whole) body by radionuclides deposited inside it is very rare and occurs only with very soluble, usually beta- or gamma-emitting, radionuclides. Studies with animals have demonstrated that nonuniform distribution of energy through tissues, such as from radioactive particles, is less hazardous than uniform distribution because of the lower number of cells at risk (Bair, 1997; EPA, 1976; Nenot and Stather, 1979). Cancers resulting from intake of radionuclides are more likely to arise in those tissues that contain the highest concentrations of radionuclides. However, some tissues, such as lymph nodes, which can accumulate radionuclides to high concentrations and which can receive high radiation doses, are much less susceptible to radiation-caused cancers than other tissues, such as red bone marrow.

For a tissue to be affected by radiation it must be directly irradiated. For example, radiation to the hand from an x-ray machine cannot cause primary health effects in other parts of the body, except if the radiation is scattered. Most acute effects of radiation are due to cell killing; these are described as deterministic effects. Long-term effects are usually due to gene mutations in exposed cells; this type of effect, such as radiation-induced cancer, is termed a stochastic effect.

The most important cause of deterministic effects is irreparable radiation-induced DNA damage resulting in premature cell death or the inability of the cell to divide. If cells are damaged faster than they can be replaced or repaired, the exposed person's health may be adversely affected. If this damage versus repair differential is present, clinical signs will be detectable and symptoms may develop early in the postexposure period (within about 2 months).

In contrast to stochastic effects, deterministic effects do increase in severity with dose and have a practical threshold dose below which they are not observed. The type and severity of deterministic effects depend upon the type of ionizing radiation involved, the magnitude of the dose, and the rate at which the dose is accumulated (dose rate). As described above, gamma and x-ray radiations emitted by sources outside the body can penetrate several tens of centime-

ters of tissue to interact with DNA in cells deep within the body. Radiation from high-energy beta-emitting radionuclides, on or close to the skin, can penetrate the skin's outer layer of dead and aging cells to reach the actively dividing cells beneath the outer layer. These exposures have the potential to cause local skin injuries and effects according to the penetrating range. Such manifestations of acute radiation-induced health effects can occur alone, in combination with each other, and with non-radiation-induced trauma, including thermal burns, or other serious medical conditions. Combined injuries of these types tend to be synergistic, that is, the combination can have more of an effect on the health of the exposed person than the sum of the effects of the individual contributors.

Accidents involving humans, medical experience, and studies with animals indicate that doses of radiation must exceed a threshold to cause the various types of acute (deterministic) health effects (injuries) that have been described. Thresholds for several radiation effects of interest are presented in Table 2-3.

If the dose is received instantaneously or within a short time, the threshold for early radiation effects may be rapidly reached or exceeded, resulting in acute effects. This can occur in the event that a high dose from a source outside the body (e.g., nuclear weapon detonation) is received at a high dose rate. If, however, the same total dose is accumulated over a longer period of time (i.e., is fractionated or protracted), the types of deterministic health effects due to the exposure are likely to be fewer in number and less severe. For a given total dose, the effects of protracted or fractionated doses are less than those of acute doses for two reasons: (a) the numbers of cells being killed by the radiation over time will be less than the numbers of new cells being produced in the body's tissue systems during the same period; and (b) because repair of radiation injury occurs within most cells. Doses of radiation can be accumulated over long periods as the result of repeated exposures to radiation outside the body, and when long-lived radionuclides (as opposed to short-lived [rapidly decaying] radionuclides)

are deposited in body tissues. Although alpha- and beta-emitting radionuclides can cause early radiation effects in the tissues in which they are deposited, the likelihood that they will cause generalized symptoms of radiation exposure early in the postexposure period is minimized by their limited penetrating power, which restricts their biological effectiveness to nearby cells.

Another key factor in the body's response to ionizing radiation is the relative sensitivity to radiation of the various cell types that make up body tissues. Bergonié and Tribondeau's Law (1906) implies that rapidly dividing cells (e.g., cells of the blood-forming tissues and certain groups of immature sperm cells) are among the most sensitive to the acute effects of radiation. The more highly differentiated cells (e.g., muscle and nerve cells) are less vulnerable to acute injury as a result of radiation. Other factors that influence the expression of the deterministic effects of radiation include the region of the body irradiated and variation between individuals in their physiologic responses to radiation.

A small group of deterministic effects tends to appear beyond the characteristic early (2-month) postexposure period. This group reflects irreparable DNA damage incurred at the time of exposure and subsequent cell death. It includes cataracts, infertility in males and females, suppression of thyroid gland function, and fibroatrophy as a consequence of radiation-induced damage to connective tissue and blood vessels. These effects are associated with practical threshold doses that are typically higher than those of concern in this report.

Of special concern in the modern military should be the radiation-induced damage that could occur in the embryo or fetus as the result of the inadvertent exposure to radiation of a pregnant soldier. A dose in excess of 50 mSv (5 rem) to the embryo or fetus is associated with an increased risk (relative to the risk for the nonexposed embryo or fetus) of nonspecific deterministic effects in the forms of embryonic death, congenital malformations, or mental and growth retardation, depending on the period of gestation during which the exposure occurred (Brent, 1989; NCRP, 1998).

Incomplete repair or misrepair of radiation-induced DNA damage increases the risk of tumors and heritable effects that may appear many years later, unless the damage is inconsistent with cell survival and division. Such damaging effects occur randomly among individuals in exposed populations or their offspring. The frequency and probability of their occurrence, but not their severity, increase with increasing radiation dose. The types of late effects that can occur depend on the types of cells affected.

Radiation-induced gene mutations in some types of cells (somatic cells) can result in abnormal cell growth that may be benign (noncancerous) or malignant (cancerous). In theory, these abnormal growths can be initiated in a single irradiated (and transformed) cell, but a variety of biological factors influence the pro-

gression of every transformed cell to a malignant focus for cancer or leukemia development. Such factors include the age of the individual at the time of exposure, sex, genetic heritage, and the immune system's ability to resist cancer. Theoretically, and for radiation protection purposes, it is assumed that there is no dose below which the probability that such effects will occur is zero—that is, there are no threshold doses for radiation-induced tumors.

It takes time for damage to DNA to result in a radiation-induced tumor. The interval between the exposure and the detection or diagnosis of a tumor attributable to the exposure is termed the latent period. The latent period is generally accepted to be a minimum of 2 to 5 years for radiation-induced leukemias and 10 years for most solid cancers.

Although all cell types are assumed to be susceptible to malignant transformation by ionizing radiation, cells in certain tissues appear to be more susceptible. Increased risks of benign (noncancerous) nodules in the thyroid gland and female breast tissue, several types of cancer (e.g., lung, thyroid gland, and female breast cancer), and all forms of leukemia except chronic lymphocytic leukemia have been strongly associated with external exposure to ionizing radiations, primarily at high dose rates. Examples of populations in which these associations have been found include the Japanese atomic bomb survivors, some groups of individuals who have had medical diagnostic or treatment exposures, and some occupationally exposed individuals. Although many people were exposed to significant radiation doses after the accident at the Chernobyl nuclear power plant in the former Soviet Union, the follow-up for these individuals is not yet sufficiently long to allow these data to be used to predict the incidence of various cancers induced by radiation exposure. Evidence of the radionuclide intakes that cause harmful effects in populations is relatively scarce and is limited to radium-dial painters, patients treated with radium or thorotrast, uranium and other miners exposed to radon, Pacific Islanders exposed to radioiodine fallout after nuclear weapons tests, and to individuals downwind of Chernobyl who were exposed to radioiodine. Health effects that can definitely be attributed to radionuclide intakes have not been identified in nuclear or medical workers.

The most useful data have been obtained from the atomic bomb survivors in Japan (Pierce et al., 1996; Preston et al., 1994; Thompson et al., 1994), and patients with ankylosing spondylitis or with certain tumors, including carcinoma of the cervix, who received radiation therapy (NRC, 1990). Data from a number of other studies have been used, including those involving patients receiving fluoroscopy for tuberculosis and in utero diagnostic radiation exposure. This material has been collated in a number of reports, including a series of reports on the biologic effects of ionizing radiation (BEIR) beginning in 1972, the latest being BEIR VI, published in 1998 by the National Research Council (National Research Council, 1998). Important analyses of Japanese atomic bomb survivor data have been reported in, among others, four publications of the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR, 1977, 1982, 1988, 1994),

as well as the 1990 Recommendations of the International Commission on Radiological Protection (ICRP, 1991a).

The risks associated with radiation exposure in the range of 50 to 700 mSv (5 to 70 rem) are confined primarily to the risk of an increased incidence of malignant diseases, including many solid tumors and leukemias. The exposure and disease incidence and mortality data have been analyzed in depth and have been converted to various estimates of risk that are generally based on a model calculation that predicts the number of cancer deaths per 10,000 persons per sievert of exposure. A number of models have been used to project from baseline data for the atomic bomb survivors to the risk of excess deaths. For most organ sites, the currently preferred model is the multiplicative one, in which the relative risk resulting from the exposure to two risk factors is the product—rather than the sum—of the relative risk for the two factors taken separately (National Research Council, 1990). For a working-age population (25 to 64 years of age), this model predicts 700 to 800 excess fatal cancers over the lifetime of 10,000 persons externally exposed acutely to I Sv (1 Gy of whole-body low linear energy transfer [LET] radiation) (UNSCEAR, 1988; Upton, 1991).

Predictions of risk of leukemia and nonleukemia derived for BEIR V (National Research Council, 1990) are presented in Table 2-4 for men and women as well as for acute and chronic exposure.

Some information on the shortening of life span from causes other than cancer or leukemia as a result of whole-body radiation exposure is available. However, the information is not sufficient for quantification of this risk (Shimizu et al., 1992).

Risks estimated by a commander are based upon estimated doses for that mission; however, the commander should be aware than an individual's radiogenic cancer risk is a function of his or her cumulative radiation doses including those incurred prior to an anticipated mission. As expressed later, in Chapter 5, the committee has taken the dose categories listed in the Allied Command Europe table to be cumulative doses (NATO, 1996).

A number of factors have been shown to influence the risk of radiation-induced malignancies.

The risk of radiation-induced leukemia is highest among the youngest exposed, including those exposed in utero, infants, and young children, and decreases to a constant risk by age 15 years. The risk of radiation-induced thyroid cancer is higher among infants and children than among individuals exposed at older ages. Breast cancer, although observed later in life, is more common when the individuals are exposed to radiation in childhood or as adolescents; the risk decreases up to the age of menopause, beyond which the risk of radiation-induced breast cancer is not detectable against the rate of spontaneous breast cancer in nonexposed groups (UNSCEAR, 1994). For many other tissues, there are not enough data to establish a relationship between age at exposure and the risk of subsequent malignancies.

With the exception of leukemias, the overall risk of radiation-induced malignancy is generally considered to be higher for females than for males. These differences are proposed to be the result of hormone-dependent promoting factors and differences in cofactors rather than differences in radiation sensitivity according to sex (ICRP, 1991 a).

The biological effects of radiation depend upon the energy transferred to the tissue, and these effects are a function of the type of linear energy transfer (LET). LET refers to the amount of energy deposited in a unit of the distance along the track of radiation. The amount of energy deposited into tissue can be measured as a function of this distance. Various types of ionizing radiation are divided into high-LET and low-LET radiation (Mettler and Upton, 1995). For a given absorbed dose, high LET types of radiation, such as neutrons and alpha particles, are more effective than low LET types, such as gamma and x rays, in inducing malignancies.

Dose rate and dose magnitude have significant effects on malignancy induction, particularly for low LET radiation. It is generally accepted that small repetitive doses or exposures at low doses and low dose rates are associated with a lower

risk of radiation-induced malignancy than a single large exposure. The exact adjustment for dose rate or small-fraction exposure is not known; however, adjustments in the range of a factor of 2 to 10 have been suggested (National Research Council, 1990; UNSCEAR, 1994). Thus, adjustments for dose rate and exposure magnitude should be considered when estimating risk on the basis of data from the usual tables of single acute exposures. A dose delivered continuously at a low dose rate or in multiple small fractions will be significantly less effective than the same total dose delivered instantaneously.

For a certain absorbed dose, the risk of radiation-induced cancer varies by tissue. The comparative susceptibilities of different tissues to radiation-induced cancers can be grouped into high, moderate, low, and very low or absent categories (Mettler and Upton, 1995), as illustrated in Table 2-5.

Ranking by cancer deaths rather than incidence is displayed in a table (Table 2-6) adapted from ICRP Publication 60 (ICRP, 1991a). As calculated by using probability coefficients for fatal cancers, an estimated 5 excess cancer deaths would occur among 10,000 people receiving 0.01 Sv (I rem). The differences by tissue type are evident.

Radiation-induced malignancies occur only in those organs, tissues, or parts of the body that have been irradiated. Risk assessment as a result of internal exposure is much more problematic due to the complexity of estimating the organ dose and its distribution within the tissues, but it must be considered. Some radionuclides are preferentially deposited in specific tissues instead of throughout the body and are associated with the development of specific cancers in those tissues. For example, when radioactive iodine enters the body, it is deposited primarily in the thyroid gland, exposing the thyroid tissue to radiation over a long period. Groups of people exposed to radiation in this way have a higher risk of developing thyroid cancer than unexposed groups.

On the basis of a review of data for the children and grandchildren of Japanese atomic bomb survivors, there is no significant evidence of an increased incidence of heritable abnormalities following radiation doses. Some heritable abnormalities are probably induced, although the incidence is too low to have ever been directly observed; consequently, they are not a major consideration in the estimates of risk.

The effects of radiation on the fetus and embryo have been observed with exposures in the range of 50 to 700 mSv (5 to 70 rem) during the 8th to 25th week of gestation. Mental retardation and decreased IQ are some of the major effects. In addition, nonspecific teratogenesis, which may be fatal to the fetus, is associated with gestational exposures.

In the therapeutic medical setting, there are chemicals that interact with radiation. Some are radioprotective, some have no effect, others have additive ef-

fects, synergistic effects, or effects in-between. However, the effects of chemicals that soldiers might encounter in military operations are largely unknown.

Although the risks in Tables 2-4 and 2-5 are by specific dose, risks can be scaled to other doses on the basis of the linear relationship assumption. Scaling down from a 1-Sv (1,000-mSv) dose should be done with some caution, however, because the effects at very low doses remain to be unambiguously defined. For radiation protection purposes, it is assumed, in the absence of data for humans exposed to very low radiation doses, that there is no dose of radiation below which there is no increased cancer risk. Thus, the risk for 100 mSv (10 rem) would be 10 percent of the risk at 1,000 mSv (100 rem). To project incidence rather than mortality, one would divide the excess mortality by the lethality fraction. For example, thyroid cancer has an estimated lethality fraction of 0.1; thus, the cancer incidence would be 10 times the mortality rate. This calculation and interpretation is useful in counseling people as to what the relative risk of the

incidence of cancer would be after a given exposure to a specific organ. With regard to exposures to military personnel, total individual risk is the cumulative risk from both the current mission and prior radiation exposures while either in the military or in civilian life.

One can summarize the available data on risk of malignancy induction in a number of ways. For example, for a population of 100 25-year-old males instantaneously exposed to 100 mSv (10 rem), there will be an excess of approximately one fatal cancer. Thus, the excess risk is approximately I percent (I in 100 chances) in addition to the natural rate of fatal cancer in the general population of about 20 percent (20 in 100 chances). If this were chronic exposure, the risk would be half of the above.

In his conclusions in the Annals of the ICRP, Upton states, ''On the basis of the latest evidence summarized in the reports from UNSCEAR and BEIR V, the task group concludes that the life time excess risk of fatal cancer for a member of the general population exposed to low dose rate whole-body irradiation can be assumed to average approximately 5 per cent per sievert" (Upton, 1991, pp. 26-27).