Journal of Exposure Science and Environmental Epidemiology (2008) 18, 95–108; doi:10.1038/sj.jes.7500551; published online 14 February 2007

Gulf war depleted uranium risks

Albert C Marshalla

aConsultant for Sandia National Laboratories, Albuquerque, New Mexico, USA

Correspondence: AC Marshall, Sandia National Laboratories, Albuquerque, PO Box 52, Sandia Park, NM 87047, USA. Tel.: +1 505 281 1067. E-mail:

Received 17 August 2006; Accepted 4 December 2006; Published online 14 February 2007.



US and British forces used depleted uranium (DU) in armor-piercing rounds to disable enemy tanks during the Gulf and Balkan Wars. Uranium particulate is generated by DU shell impact and particulate entrained in air may be inhaled or ingested by troops and nearby civilian populations. As uranium is slightly radioactive and chemically toxic, a number of critics have asserted that DU exposure has resulted in a variety of adverse health effects for exposed veterans and nearby civilian populations. The study described in this paper used mathematical modeling to estimate health risks from exposure to DU during the 1991 Gulf War for both US troops and nearby Iraqi civilians. The analysis found that the risks of DU-induced leukemia or birth defects are far too small to result in an observable increase in these health effects among exposed veterans or Iraqi civilians. The analysis indicated that only a few (~5) US veterans in vehicles accidentally targeted by US tanks received significant exposure levels, resulting in about a 1.4% lifetime risk of DU radiation-induced fatal cancer (compared with about a 24% risk of a fatal cancer from all other causes). These veterans may have also experienced temporary kidney damage. Iraqi children playing for 500h in DU-destroyed vehicles are predicted to incur a cancer risk of about 0.4%. In vitro and animal tests suggest the possibility of chemically induced health effects from DU internalization, such as immune system impairment. Further study is needed to determine the applicability of these findings for Gulf War exposure to DU. Veterans and civilians who did not occupy DU-contaminated vehicles are unlikely to have internalized quantities of DU significantly in excess of normal internalization of natural uranium from the environment.


analytical methods, cancer, exposure modeling, inhalation exposure, particulate matter, radiation, metals



The United States and Great Britain make extensive use of depleted uranium (DU) metal in armor-piercing military rounds. DU is a by-product of the uranium enrichment process and is equivalent to natural uranium with the percentage of uranium-235 reduced from 0.72% (for natural uranium) to ~0.2%. The isotope uranium-238 constitutes about 99.8% of the uranium in a DU penetrator shell. Uranium-238 is primarily an alpha particle emitter with a half-life of 4.47 × 109 years. The long half-life results in a very small specific activity (<10−6Ci/g). Weapon developers selected DU because of its high density, self-sharpening capability, pyrophoric characteristics, and low cost. However, uranium particulate generated by shell impact with military vehicles may be inhaled or ingested by troops and nearby civilians. As DU is chemically toxic and weakly radioactive, a number of critics assert that wartime use of DU weapons has resulted in a variety of unintended health consequences. Some critics claim that significant increases in leukemia have been observed among individuals exposed to DU and that increases in the rate of birth defects have been observed in their progeny (Catalinotto and Flounders, 2004).

Reports in the popular press of serious health effects from DU exposure often include misinterpretations of scientific reports, unsubstantiated claims, and false claims. Some excellent scientific studies have investigated DU health effects at the cellular level (e.g., Miller et al., 2002) or by testing animals exposed to DU (e.g., Domingo et al., 1989). These studies suggest possible chemically induced cancer risks, birth defects, immune system impairment, etc. Cellular-level experiments can enhance the understanding of biological processes and suggest further study, but they cannot in themselves provide quantitative risk projections. Furthermore, cellular-level experiments do not include all of the complex biochemical interactions required to understand how (and if) these potential health effects will manifest in exposed individuals. Animal tests implicitly address biochemical interactions, but they often involve massive doses of DU rather than realistic doses. Also, animal tests may not provide reliable correlations for human health effects. Nonetheless, animal experiments can provide some guidance for health effects assessments if human exposure levels can be correlated with comparable animal test exposure levels and observed health effects. In order to provide a quantitative health effects assessment for individuals exposed to DU, a comprehensive analysis is needed that includes realistic assessments of human exposure levels, application of validated exposure/health effect correlations, a review of medical statistics for exposed individuals, and consideration of the findings from DU experiments.

Several relatively comprehensive scientific studies have been carried out to assess the health risks associated with the use of DU munitions. The first study was initiated by the US Department of Defense (DoD) in the late 1990s (USACHPPM, 2000). Fetter and von Hippel (1999) carried out an independent analysis of DU hazards and the British Royal Society completed a study on DU munitions health effects and published their report a few years later (Royal Society, 2002a). Pacific Northwest National Laboratory (PNNL) performed extensive DU weapons testing for the DoD Capstone Program to characterize the particulate generated during DU impact with armored vehicles (Parkhurst et al., 2004). The PNNL data were used by Los Alamos National Laboratory to predict possible veteran health effects (Guilmette et al., 2004). The conclusions from these studies did not support assertions of significant and observable increases in cancers as a result of exposure to DU munitions. Each of these studies was a major undertaking that greatly improved the understanding of the health effects of DU munitions; however, some topics were not addressed by these studies and the most recent test data were not available for the early studies.

In order to assess methodologies for predicting dispersion of radiological material and health consequences, Sandia National Laboratories carried out an analysis of DU dispersion and consequences using the 1991 Gulf War as a case study (Marshall, 2005). Because the study included a comprehensive investigation of Gulf War DU dispersion and health effects, including topics not examined in previous studies, the study findings have contributed to the understanding of health effects relating to military use of DU weapons. Subsequent to the release of the Sandia report in July of 2005, the author made several improvements and additions to the analysis. The findings from the Sandia study, including improvements and additions, are presented and discussed in this paper.



The US DoD defines three levels of DU exposure for Gulf War veterans. Level I exposures correspond to friendly fire incidents in which US tanks mistakenly fired DU rounds at other US combat vehicles. The crews of the targeted vehicles were exposed to aerosolized DU particulate, and some of the crew were wounded by DU fragments. In addition, US troops involved in rescue operations were exposed to potentially high concentrations of aerosolized DU. The Level II category includes veterans involved in post-combat evaluation of DU-damaged vehicles, removal of equipment, and preparation of vehicles for transport. Level III exposures correspond to short-term, low-level inhalation exposures to DU during and following battle. This study examines possible DU health effects for all three levels of veteran exposure and included exposures by inhalation, ingestion, and embedded fragments. Exposures to DU aerosols from munitions fires were not investigated because very little DU is converted into respirable aerosols (Fetter and von Hippel, 1999). The types of Iraqi civilian internal exposures considered in this study include inhalation, direct ingestion (hand-to-mouth transfer), and consumption of food and water contaminated by DU particulate. Possible health effects from DU exposure are examined for Iraqi civilians living downwind of the battle zone and for children playing in post-battle zones. Both nominal and maximum exposure cases are assessed for veterans and Iraqi civilians. (Maximum exposure refers to the individual in the group receiving the highest dose, rather than the upper bound. Nominal exposure is the average or typical exposure for the group.) External radiation effects are also examined for gamma and beta radiation emitted from DU shells, fragments, and particulate. A detailed discussion of methods and references for data sources used in this study are provided in the original Sandia document (Marshall, 2005). The following summary of the methodology begins with the approaches used to estimate the quantity of DU internalized (taken into the body). A variety of methods were utilized to compute the internalized DU mass in order to make use of the most reliable data for each exposure scenario.

In-vehicle DU Inhalation

For Level I veterans, a biokinetic analysis was used to find the inhaled DU mass that yielded calculated DU concentrations in urine matching the DU concentrations measured in the urine of Level I veterans who had no fragments or history of fragments in their bodies. The highest measured DU concentration was used to estimate the inhaled mass for veterans with the highest exposure, and the average of the measured DU concentrations was used to obtain the inhaled mass for nominal exposure. The DU particle data used in the biokinetic analysis were based on the detailed particle characterizations (particle size distributions, composition, and solubility) from the Capstone program (Parkhurst et al., 2004).

The inhaled mass of uranium by Level II veterans is mostly from in-vehicle inhalation of resuspended DU. Level II inhaled masses were computed using the estimated air concentration, typical breathing rates, and appropriate exposure times. The quantity of resuspended DU within a vehicle will depend, among other considerations, on the type of vehicle and the number and size of DU rounds striking the vehicle. Using the Capstone test data, a maximum in-vehicle resuspended DU air concentration of 2mg/m3 was obtained. This DU concentration is equivalent to that for a tank struck by two 120mm shells. Adult elevated breathing rates of 2 and 3m3/h were assumed for nominal exposure and maximum exposure, respectively. Using a DoD review of veteran exposures (Rostker, 2003), the exposure duration for nominally and maximally exposed Level II veterans were estimated to be 10 and 100h, respectively. These durations correspond to unprotected (no breathing apparatus) in-vehicle exposure.

The approach used for Level II veterans in-vehicle inhalation was also used for Level III veterans and Iraqi children. Iraqi children were assumed to play in-vehicle for 50h for the nominal case and 500h for the maximally exposed case. Most of the inhaled uranium for a child at play in the battlefield (after hostilities have ceased) will also result from in-vehicle exposure to resuspended DU. Some Level III veterans were also assumed to explore DU-contaminated vehicles for a period of 1h.

DU Fragments

Fragments of DU munitions embedded in the body from friendly fire DU impact will be mostly unoxidized uranium metal. For embedded DU fragments, the fragment dissolution rates for the nominal and maximum exposure cases were inferred from DU urine-concentration data. DU concentrations in urine from fragments were obtained from Level I veterans with fragments or history of fragments in their bodies.

In-vehicle DU Ingestion

Ingestion calculations for veterans used an estimated hand-to-mouth transfer rate of 3mg of particulate per hour as the ingestion rate for adults working in a DU-contaminated environment. For children, much higher hand-to-mouth transfer rates were used than for adults. The nominal and maximum transfer rates for a nominally and maximally exposed child were taken to be 10 and 30mg/h, respectively.

Ex-vehicle DU Inhalation

For veterans and civilians with in-vehicle DU exposures, this analysis found that ex-vehicle inhalation exposures were insignificant compared to in-vehicle exposures. Ex-vehicle DU inhalation must be evaluated, however, for downwind civilians and Level III veterans. To compute DU air concentrations and inhaled DU mass for ex-vehicle exposure, it is necessary to determine the quantity of aerosolized DU released into the environment. Using the Capstone test data, the mass of DU aerosols released into the atmosphere was estimated to be less than 100g/120mm shell impact with a tank. The dispersion of DU particulate following impact was computed using a Gaussian puff model. The dispersal and deposition of DU at the battle location depends on the atmospheric conditions at the time of the battle. To estimate aerosol densities and surface depositions from DU impact with hard targets, calculations were performed assuming both wet and dry deposition, low and high wind speeds, and a range of weather conditions categorized by atmospheric stability classes. The largest integrated air concentrations and the greatest DU deposition densities resulted from wet deposition, stable weather conditions, and low wind speed (2m/s). For the best estimate prediction, slightly unstable weather conditions, wet deposition, and an intermediate wind speed of 8m/s were assumed.

DU aerosols resulting from multiple-target impacts provide multiple source terms. Several different representations were explored for multiple targets; that is, (1) a tight cluster of tanks, (2) a column of more than 50 tanks with a 20m separation distance between tanks, and (3) a long column of densely packed, light target vehicles struck by 30mm DU rounds. DU dispersal for a tight cluster of tanks (case 1) can be modeled at some distance from the targets by a simple Gaussian model with the DU release increased in direct proportion to the number of tanks in the cluster. For the column of tanks (case 2), the Gaussian model was modified to account for multiple separated sources. For case 3, a situation similar to the so-called “Highway of Death” was examined. For simplicity, an infinitely long battlefield was assumed with tightly packed targets along the length of the battlefield. For high-density battlefields of the type considered here, most of the Iraqi vehicles were unarmored and destroyed by a combination of cluster bombs (containing no DU) and 30mm DU penetrators fired from A-10 “tank-killers” and other aircraft. An infinitely long-line source model was used with an estimated release of 15g of DU per meter of column. At long distances, the effect of puff reflection by an atmospheric temperature inversion layer may become important. To examine this potential effect, calculations were repeated assuming a relatively low ceiling height of 150m.

The calculated deposition density and the inhaled DU mass for case 3 are presented in Figure 1 as a function of downwind distance. For an infinite line source, the deposition density (areal density) is almost constant at about 7 × 10−5g/m2 over a 20km distance and is less than 10% of the natural uranium areal density for the top 1mm of typical Iraqi soil. For the inhaled DU mass, however, the concentration drops appreciably with downwind distance. Inversion layer reflection has the effect of maintaining the inhaled DU mass at a nearly constant value beyond ~2km. The deposited DU areal density is the same with and without reflection by a temperature inversion layer. In order for the tight cluster of tanks (case 1) to reach the same DU air concentrations and depositions as case 3 with an inversion layer, we would need to assume a cluster of about 50 tanks, each hit by two 120mm DU rounds. For the column of tanks, each hit by two 120mm shells (case 2), DU air concentrations were predicted to be three to four times lower than for case 3. Given these comparisons, case 3 with a 150m ceiling was selected as a conservative representation for DU dispersal and deposition for multiple targets.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact or the author

DU mass inhaled and deposition (areal) density vs. downwind distance for the infinite line source model and the very high target density scenario. * 10% of the areal density of natural uranium in top 1mm of soil.

Full figure and legend (20K)

A Level III veteran downwind from a vehicle struck by a DU shell is unlikely to be closer than 100m from the vehicle during the battle. Figure 1 shows that at about 100m from the battle zone, the time-integrated inhaled DU mass is about 10−4g, and the deposited DU areal density is essentially position-independent at about 10−4g/m2. These values are selected for the inhaled masses and areal density for maximum exposure Level III veterans. For nominal exposure of downwind veterans, the inhaled DU mass and deposition density at about 3km were used, that is, about 3 × 10−6g and 7 × 10−5g/m2, respectively. A modified Garland resuspension model with a high scale factor (=100) was used to estimate resuspended-DU air concentrations and inhaled DU masses. The battlefield exposure times for Level III veterans were assumed to be 30 and 90 days for the nominal and maximum exposure cases, respectively. The DU-puff air concentrations for downwind civilians are the same as for the nominal case for Level III veterans; however, the downwind civilian exposure time was assumed to be 50 years.

Ex-vehicle DU Ingestion

Ex-vehicle DU ingestion by a child at play in the battlefield was found in this analysis to be much less than in-vehicle DU ingestion. DU may be ingested by downwind civilians as a result of ingesting contaminated food and drinking water or by hand-to-mouth contact if hands are contaminated with DU particulate. The deposited downwind DU concentrations are much less than the natural uranium concentrations in the soil. Based on data from environmental monitoring of DU near battlefields in Kosovo and Kuwait and test areas in the United States (Royal Society, 2002b; Bem and Bou-Rabi, 2003; Bleise et al., 2003; UNEP, 2003), the following conclusions were drawn: (1) mathematical estimates of DU transport to water and food supplies are unlikely to be highly reliable; (2) ground-deposited DU particulate downwind of the battlefield is only a small fraction of the natural uranium soil concentrations; (3) because of the low solubility of uranium metal, significant DU contamination of food and water from buried munitions is unlikely; and (4) no significant DU contamination of the environment was observed near battlefields in Kosovo and Kuwait or near test areas in the United States. Given these observations, the exposure of downwind civilians to DU from contaminated food and water is considered to be insignificant, and a mathematical analysis was not undertaken for this potential pathway.

Biokinetic, Dose, and Risk Methods

Both internal and external radiation effects were considered in this study. External sources include undamaged DU penetrators, fragments, and particulate. Standard calculational methods were used to estimate external radiation effects. Internal radiological exposures result from internalization of DU and are far more important for DU exposure than external radiation. Internal radiation was examined for inhalation and ingestion of DU particulate and for dissolution of DU fragments embedded in the body (for some veterans). A biokinetic model was used to obtain a prediction of the time-dependent passage of DU through the various organs of the body. The biokinetic analysis is based on the established biokinetic models developed by the International Commission on Radiological Protection (ICRP, 1979, 1991, 1994, 1995). The models use sets of differential equations along with basic input data, such as the internalized DU mass, the dissolution rates for uranium, and organ region-dependent particle transport rates. The total DU mass in the various organs is primarily the sum of contributions from the inhaled particulate from the passing DU cloud (puff) produced by penetrator impact, the inhaled DU from resuspended particulate, and (for Level I veterans) inhaled DU particulate from shell penetration and uranium from embedded fragments. DU exposure from ingestion is important for children playing in or near DU-destroyed vehicles.

The doses to various organs were calculated using standard dose calculation methods. Almost the entire equivalent internal dose (~99%) is from alpha particle emission; the remainder is primarily from gamma and beta activity from uranium-238 daughter products Th-234 and Pa-234m. The activity from the small concentrations of other uranium isotopes in DU makes up about 17% of the total DU alpha particle activity. The dose from trace actinides (e.g., Pu) is less than 1%. Internal dose contributions from other uranium isotopes, trace actinides, and daughter products were accounted for by increasing the calculated dose from U-238 alpha decay by 20%. Radiation health risks are readily determined by multiplying equivalent organ doses by risk coefficients. The risks of DU-induced birth defects are estimated using the calculated equivalent doses to the gonads and risk coefficients for genetic effects.

Additions and Modifications

A re-examination of the original Sandia study resulted in some additions and modifications to the original analysis. These changes are as follows: (1) because downwind civilians would not be located close to the targeted vehicles during the battle, the maximum exposure case does not apply to civilians and was dropped from consideration; (2) the analysis now includes the effect of blood-deposited DU in internal organs from ingested DU; (3) the current calculations account for the fact that only about half of the ingested in-vehicle particulate was DU; (4) the original analysis used a maximum in-vehicle exposure time of 300h for both the maximally and nominally exposed child. For this analysis, the nominal and maximum exposure time was changed to 50 and 500h, respectively; (5) the analysis presented in this paper assesses both radiological and chemical effects on the fetus from exposure to DU; (6) this analysis examined the possibility of cancer induction by chemical and synergistic radiation-chemical effects from DU exposure; and (7) the possibility of adverse effects on the immune system and bone formation from the chemical effects of DU is addressed. These modifications did not alter the conclusions drawn from the original analysis.



The principal results from this analysis include the mass of internalized DU, the time-dependent distribution of DU to the body organs, the radiological dose from DU exposure, and the calculated risk of cancer fatalities and genetic birth defects from the radiological effects of DU exposure.

DU Internalization and Distribution

The internalized DU masses for veterans and civilians are presented in Tables 1 and 2, respectively. The mass given for fragments is the total mass dissolved in the blood assuming a constant dissolution rate over a 50-year time span (a generally conservative assumption). The health effect per gram of internalized DU depends on the internalization pathway (e.g., ingestion of DU results in about 15% of the time-integrated body burden of that resulting from inhaling the same quantity of DU). Consequently, care must be exercised when comparing internalized DU masses.

The time dependence of the percent of inhaled DU remaining in the body is presented in Figure 2 for the acute exposure case. All but about 10% of the inhaled DU is either exhaled or rapidly cleared (~1 day) by ciliary action and blood absorption followed by excretion in the urine and feces. Most of the 10% of DU remaining in the lung after 1 day will eventually be absorbed by the blood, and all but about one-third of the DU absorbed by the blood will be excreted rapidly in the urine. Thus, about 3% of the inhaled DU will be redeposited in the kidney, bones, and other organs. The DU deposited in these other target organs will be eventually reabsorbed by the blood and excreted in the urine or redeposited among various organs. Within less than 2 years, only 1% of the inhaled DU is retained. Figure 3 presents the time-dependent DU concentrations in organs (g DU/g organ) following inhalation for nominally exposed Level I veterans. The time-dependent DU mass per gram of organ for the combined effect of DU inhalation and embedded fragments for Level I veterans is presented in Figure 4. When fragment dissolution is included, it is seen that the concentrations of DU in the kidney, bone, and liver do not decline appreciably after 100 days, and the DU concentrations in the soft tissues increase after a few years. (Soft tissues include a number of organs and tissue, such as the brain, gonads, and muscle.) The plots of organ DU concentrations for Level II and III veterans and downwind civilians show the same general trends observed in Figure 3, although the quantities are orders of magnitude smaller than for Level I veterans.

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact or the author

Percent of inhaled DU in body vs. time after exposure for acute inhalation.

Full figure and legend (11K)

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact or the author

DU mass/g organ vs. time for nominal Level I veteran inhalation.

Full figure and legend (19K)

Figure 4.
Figure 4 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact or the author

DU mass/g organ vs. time for Level I inhalation with DU fragments for the nominal case.

Full figure and legend (18K)

For children playing in DU-destroyed vehicles, the trends are somewhat different. Chronic DU exposure results in a gradual buildup of DU concentrations in organs up to 10 years (a child is assumed to play intermittently in the post-battle zone for a period of 10 years). Thereafter, DU concentrations decline relatively quickly as DU is eliminated from the body. Except for the lungs, ingested DU is the primary source of uranium to internal organs for children playing in vehicles, whereas ingested DU is of minor importance for all other exposure scenarios. DU concentrations in the kidney are also obtained from the biokinetic analysis in order to assess the potential for toxic effects on the kidney. The predicted acute and chronic DU concentrations in the kidney for veterans and civilians are presented in Table 3.

Doses and Risks

For Level I veterans, the lifetime doses presented in Table 4 were computed for a period of 50 years following exposure. Equivalent doses are given for nominal and maximum exposures assuming DU inhalation with and without imbedded fragments. The effective dose was obtained as the sum of the weighted organ doses. Effective doses of 0.0025 and 0.038 sievert (Sv) were predicted, respectively, for nominally and maximally exposed Level II veterans. For Level III veterans, the nominal and maximum effective doses are 2.5 × 10−7 and 5.0 × 10−4Sv, respectively. The doses to the major organs for both nominally and maximally exposed children are presented in Table 5. For prolonged ingestion and the high DU ingestion rate estimated for a child at play, the radiation dose to the organs of the large intestines is comparable to the lung dose. The equivalent doses for a civilian downwind of the battle zone were computed for inhaled DU entrained in the air during battle (puff) and DU resuspended subsequent to battle (50-year exposure). The doses to downwind civilians are also presented in Table 5. All radiological organ doses to downwind civilians resulting from inhaled DU are predicted to be extremely small.

Table 6 compares the predicted total cancer risks as well as the risks of lung cancer and leukemia from DU exposure for veterans to the US average for background fatal cancers. Predicted risks for Level I veterans are given for inhalation and ingestion with and without embedded fragments for both the nominal and maximum cases. The predicted radiation-induced fatal cancer risks from DU exposure for civilians are provided in Table 7. The risks in Tables 6 and 7 are incremental risks; that is, they are fatal cancer risks over and above the background fatal cancer risks of 8%, 1%, and 24% for lung cancer, leukemia, and all cancers, respectively. (A background fatal cancer risk of 24% means that out of all fatalities in the United States for 1 year, 24% of the deaths will result from cancer. Note that individual risks for cancer can vary significantly from the average of 24%, depending on life style.) Table 8 provides the incremental genetic risk for veterans for the nominal and maximum cases. These predictions are compared to the US average of ~8% of live births resulting in serious birth defects (Mettler and Upton, 1995). The US average for all serious birth defects includes both genetic and in utero effects and includes birth defects recognized over a period of about 1–2 years following birth. Table 9 provides the incremental genetic risk for Iraqi civilians. The only effects resulting from the modifications discussed in the Methods section was a significant dose and risk decrease for the nominally exposed child scenario and a small dose and risk increase for the maximally exposed child scenario.

External radiation effects were briefly studied for veterans and civilians. Alpha particles from external sources cannot penetrate the dead skin layer of the body and do not present a health hazard to humans. External radiation by beta particles from DU can penetrate the dead skin layer, but cannot reach internal organs. Thus, beta particles from external sources may cause skin burns or increase the risk of skin cancer, but will not contribute to the cancer risk for internal organs. External gamma radiation, however, can penetrate through the body and reach internal organs. Figure 5 presents the gamma dose rate and the incremental whole-body fatal cancer risk from an intact DU penetrator as a function of distance from the penetrator for 100, 1000, and 10,000h of exposure.

Figure 5.
Figure 5 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact or the author

Whole-body dose rate and incremental fatal cancer risk vs. distance from 120mm DU round.

Full figure and legend (21K)


This discussion includes an assessment of the validity of the mathematical analysis, an interpretation of the results of the analysis, and a discussion of other issues that cannot be directly addressed by established calculational methods. These other issues include cancers induced by the chemical characteristics of DU, direct irradiation by embedded fragments, teratogenic effects, neurological effects, bone damage, immune system impairment, and the general disorder known as Gulf War Syndrome.

Analysis Validity and Uncertainties

A detailed discussion of the validity of the analysis used for this paper is provided in the original Sandia document (Marshall, 2005). The validity of the computer models developed for the analysis was confirmed by comparing computer model predictions with a test case provided by the ICRP (1994) and one of the cases studied by the Royal Society (2002a). In addition, an uncertainty analysis was carried out that included uncertainties associated with the basic input data (e.g., particle size and dissolution rate), dispersion modeling, the biokinetic model, and radiological risk coefficients. For Level I and II veterans, the upper uncertainty bound was a factor of 2.5 higher than the predicted cancer risks presented in Table 6 and a factor of 3.0 higher than the genetic risks presented in Table 8. For Level III veterans and downwind civilians, the upper uncertainty bound was a factor of 14 higher for both cancer and genetic risks. The upper uncertainty bound for a child playing in DU-destroyed vehicles was a factor of 3.5 for both cancer and genetic risks. The use of an upper uncertainty bound did not alter the general conclusions on health effects based on best-estimate predictions presented in Tables 6, 7, 8 and 9.

DU Internalization

Level I and II veterans and Iraqi children playing in DU contaminated vehicles were typically exposed to DU in concentrations significantly greater than natural levels. Internalized DU for most exposed veterans and civilians, however, was smaller than exposure levels to natural uranium in the environment. Compared with nominal Level III veterans and downwind civilian total lifetime intake of 6μg of DU or less, natural uranium internalization is typically 460μg/year ingested, 0.6μg/year inhaled, and 90μg in the body at any time point (Arfsten et al., 2001). As a consequence, veterans and civilians who were never inside DU-contaminated vehicles are unlikely to internalize DU in quantities much in excess of normal intake of natural uranium from the environment. In order to determine the accuracy of this prediction, a careful, systematic analysis of DU concentrations in urine is recommended for veterans and civilians. Although some urine analyses for veterans have been carried out by both government and private organizations, these tests have not been extensive or systematic and the accuracy of some of these tests have been questioned.

Doses and Cancer Risks

To provide perspective, the effective radiation dose for DU-exposed veterans can be compared to the average US background radiation dose of about 0.17Sv (50-year dose) and the maximum permitted cumulative dose (35-year dose) for radiation workers of 0.35Sv. The maximum permitted radiation worker dose does not include the exposure contribution from background radiation. Thus, the accumulated dose for a maximally exposed Level I veteran (0.25Sv) is less than the allowed lifetime dose for a radiation worker. Lifetime doses from DU exposure for all other scenarios are lower than the accumulated lifetime background dose.

From Table 6, the risk of fatal cancer for the maximally exposed Level I veteran is 1.4% (upper bound of 3.5%). Almost all of the DU-induced cancer risk is associated with the risk incurred for lung cancer. DU fragments are predicted to increase non-respiratory system cancer risks. However, even for the maximum case with fragments included, the risks for all other cancer types are very small. For the maximally exposed Level I veteran with fragments, the risk of leukemia is 0.03%. The cancer risks predicted for Level III veterans (Table 6) and downwind civilians (Table 7) indicate that the risks among significant populations of veterans and civilians are far too low to result in observable cancers. These findings are consistent with medical records for US troops in the 1991 Gulf War (PAC, 2005) and Italian ground troops in the Balkan War (Italian Ministry of Defense, 2001). McDiarmid et al. (2001), however, found an elevation in sister chromatid exchanges (SCE) for veterans with high DU concentrations in their urine relative to veterans with low DU concentrations. These observations are similar to observations for uranium workers showing an increase in SCE (Martin et al., 1991) with no statistically significant increase of cancers of any type (Waxweiler et al., 1983).

As seen in Table 7, a maximally exposed child playing for hundreds of hours in DU-destroyed vehicles would incur an incremental total cancer risk of about 0.4% (upper bound 1.4%) with predicted lung and colon cancer risks of about 0.23% and 0.14%, respectively. The predicted risk of leukemia for this maximally exposed child is about 0.01%. The cancer risk for the nominally exposed child is much smaller than for the maximally exposed child. For downwind civilians, the risk of cancer is very small.

Chemical and Synergistic Effects on Cancer

If another mechanism associated with DU exposure can induce cancers (in addition to radiation), it is possible that the cancer risks are greater than the risks presented in Tables 6 and 7. Miller et al. (1998), using in vitro tests with DU chloride, found that human osteoblast cells can be transformed into the tumorigenic phenotype. The authors of the study point out that the dose–effect relationship is similar to that observed for nonradioactive heavy metals; however, a later study (Miller et al., 2002) showed an approximately linear relationship between transformation frequency and specific activity (nuclear disintegrations/s·g) for the same uranium concentrations. Also, Miller observed a significant increase in dicentric frequency for in vitro tests with DU that was not observed for nonradioactive heavy metals.

The tests by Miller et al. are useful for demonstrating the potential for carcinogens, but they do not provide data directly applicable to quantitative risk predictions. Furthermore, it is not clear whether the observed transformations associated with DU exposure were due to just radiological effects, or chemical and radiological mechanisms acting separately, or if the effect was synergistic. Some perspective on the DU cancer potential observed in the Miller-1998 tests can be inferred from another set of tests by Miller et al. (2004), in which the authors observed that the magnitude of the effects associated with cancer induction from DU is about the same as for nickel and for other heavy metals. Quantitative guidance has been provided for the risk associated with nickel inhalation by the U.S. Environmental Protection Agency (1999). The Environmental Protection Agency model predicts that the maximum cancer risk from breathing air with a nickel concentration of 0.004μg/m3 for a lifetime (70-year dose) is 10−6. Using an adult average inhalation rate of about 1m3/h, this concentration amounts to a lifetime inhalation of 0.0025g of nickel. Hence, the risk of cancer per gram of inhaled nickel is about 0.04%. Tables 1 and 6 show that for Level I veterans, inhalation of 4g of DU corresponds to a cancer risk of 1.4%, or 0.35% per gram of DU inhaled. This comparison suggests that for DU exposure, the predicted radiological risk for cancer is an order of magnitude greater than the risk inferred from the in vitro experiments.

As chemical and radiation processes are very different, these comparisons do not provide conclusive evidence that the standard cancer risk prediction methodologies are valid. If the observed effects from the Miller DU tests are entirely due to radiation, then the standard radiological risk methodology should be applicable to DU. If the effect observed in the tests is mostly a chemical effect, then the comparison with the chemical effect of nickel should be reasonable and we can conclude that the carcinogenic potential due to the chemical effects of DU is small compared to the radiological effects of DU. If, however, the observed carcinogenic potential of DU is due to synergistic chemical–radiation effects, then the observations based on a comparison to the chemical effects of nickel become ambiguous and it is possible that the standard radiological risk methodology underestimates the cancer risk from DU internalization. On the other hand, medical records do not show an increased incidence of cancer for 1991 Gulf War veterans; consequently, any underestimate of cancer risk should not be large enough to result in a near-term observable increase in cancers from Gulf War exposures. Further testing to better understand the potential for synergistic effects is recommended.

Radiation from Embedded Fragments

Although the foregoing discussion of doses and cancer risks included the effect of dissolution of DU fragments and uranium deposition in organs, the discussion did not address the effect of alpha particle bombardment of tissue near the surface of embedded fragments. The radiation doses at these locations are very high and may result in localized cancer risks greater than those predicted using standard ICRP models. The local alpha particle dose, however, will mostly induce cell killing rather than nonfatal DNA damage to the cell. Nonetheless, cancers may be induced in the less-intense radiation field near the boundary of the alpha particle range. The contribution of beta particles to the local dose must be considered as well. Furthermore, recent studies have demonstrated the so-called bystander effect in which damage may be induced in cells that are in close proximity to radiation-damaged cells.

Tests on rats by Hahn et al. (2002) showed the development of soft-tissue sarcomas in rats at the location of large surgically implanted squares. Although the DU implant size is comparable to the size of DU fragments remaining in the bodies of some Level I veterans, the body mass of a veteran is roughly 100 times greater than the body mass of the rats used in the experiments. Hahn observed a DU-fragment surface area dependence wherein 1 × 2mm pellets did not induce tumors, 2.5 × 2.5 × 1.5mm2-induced tumors in 6% of the test rats, and 5 × 5 × 1.5mm2-induced tumors in 18% of the test rats. Leggett and Pellmar (2003) conducted similar tests and also found that 1 × 2mm pellets did not induce tumors. Hahn points out that rats are more implant-sensitive and more radiation-sensitive than humans. For many materials, tissue reactions that have been observed in rats do not occur in humans. Although sarcomas in the vicinity of embedded fragments have not been reported in the literature for veterans, these animal tests suggest further study and continued monitoring of veterans with embedded fragments.

External Radiation Effects

Beta radiation from DU exposure is very low, and the analysis performed for this study indicated that the effect of long-term contact of DU with the skin should not result in beta radiation burns. However, direct contact with DU (e.g., wearing jewelry made from DU) for decades might result in beta radiation-induced local skin cancer. The incremental fatal cancer risk from gamma radiation is observed in Figure 5 to be very small, even at close proximity for 10,000h. The US DoD measured dose rates of <0.002mSv/h to the occupants of an Abrams tank containing a full load of DU penetrator rounds (USACHPPM, 2000). Based on this measured dose rate, the tank crew could occupy the vehicle 24h a day 365 days a year without exceeding international radiation worker guidelines. Fetter and von Hippel (1999) have estimated that the ground shine from a distribution of 1g DU/m2 would be only 0.01mSv/year. Calculations from the Sandia study indicate that the typical DU areal density is much lower than 1g DU/m2; consequently, the dose from ground shine is insignificant.

Reproductive Issues

The risks of radiation-induced genetic birth defects are predicted to be extremely small (Tables 8 and 9). Furthermore, radiation exposure in utero has not been observed to result in birth defects in humans for doses less than 0.05Gy (Mettler and Upton, 1995). The predicted soft tissue dose for downwind civilians is ~10−9Gy and for Level III veterans the dose is ~10−9 and ~10−6Gy for the nominal and maximally exposed cases, respectively. (The dose in Gy can be obtained by dividing the dose in Sv by a radiation-weighting factor of 20 for alpha particles.) Assuming that the dose to the uterus is equal to the dose to soft tissues, the dose for downwind civilians and female veterans is far less than the presumed threshold dose of 0.05Gy. Tests have shown that implanted uranium pellets in rats can result in uranium accumulation in the fetus; however, the quantities deposited are very small (McClain et al. 2001). Based on these considerations, internalized DU should not result in observable increases in birth defects from genetic or in utero effects of radiation exposure.

It may be possible, however, that chemically induced birth defects could result from in utero exposure to DU. Domingo et al. have shown that very high incidences of birth defects result when pregnant rats and mice are given large doses of uranium acetate (Domingo et al., 1989; Domingo, 1995, 2001). Domingo concludes that the observed developmental effects are due to chemical rather than radiation effects because the effects are observed at much lower doses than could be explained by radiation effects. In one experiment, pregnant mice were given uranium acetate dihydrate by gavage at dosages of 0, 5, 10, 25, and 50mg/kg during gestation days 6–15 (Domingo et al., 1989). A dose of 50mg/kg is equivalent to one-fifth of the acute oral LD50 dose. The DU concentration in the kidney increases linearly for many days (the linear increase is illustrated by the example shown in Figure 6 for the chronic ingestion scenario for child at play). Hence, 50mg/kg for 10 days is equivalent to twice the LD50 dose. Although the pregnant mice survived to scheduled termination, significant weight reduction resulted during the period of dose administration. The fact that the pregnant mice survived these very high chemical doses underscores the variability and uncertainty associated with LD50 guidelines for uranium toxicity. Using the data from this experiment series, an external defect ratio Re and skeletal defect ratio Rs are defined and plotted as a function of the fraction of the LD50 dose f (f=dose/LD50 dose) in Figure 7. The external defect ratio and skeletal defect ratio are defined here as the ratio of the number, respectively, of external and skeletal defects (in excess of the defects at zero dose) to the number of fetuses examined. The fraction of the LD50 dose for downwind civilians and maximally exposed Level III veterans (~10−6) is also shown in Figure 7. Although no guidance exists for chemically induced teratogenic effects from DU exposure for humans, the comparison with animal testing suggests that observable increases in birth defects from the chemical effects of DU exposure would not be expected for US veterans or Iraqi civilians. Additional perspectives can be gained by noting that the inhaled uranium concentrations by downwind veterans and civilians are less than the typical internalization of natural uranium in the environment by inhalation and ingestion (Domingo et al., 1989; Arfsten et al., 2001; WHO, 2001). This analysis of radiation and chemically induced genetic and in utero birth defects does not support the claim that the children of DU-exposed US Gulf War veterans were born with serious birth defects at a rate that significantly exceeds projected incidents for unexposed veterans.

Figure 6.
Figure 6 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact or the author

DU mass/g organ vs. time from chronic ingestion for maximally exposed child at play.

Full figure and legend (18K)

Figure 7.
Figure 7 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact or the author

Ratio of defects to number of fetus examined (in excess of defects at 0 dose) for mice, given uranium by gavage, as a function of the fraction of the LD50 dose.

Full figure and legend (13K)

A number of birth defect studies were performed for children of Gulf War veterans. In one study of 75,000 births, 7.45% of the Gulf War veteran children were born with birth defects, compared to 7.59% for children of veterans not deployed in the Gulf (Cowan et al., 1997). In another study (Araneta et al., 2003), statistically significant excesses were found for a few specific types of birth defects (i.e., tricuspid valve insufficiency, aortic valve stenosis, and renal agenesis) out of 46 birth-defect categories. The authors concluded, “We did not have the ability to determine if the excess was caused by inherited or environmental factors, or was due to chance because of myriad reasons, including multiple comparisons.” None of these statistical analyses distinguished between DU-exposed veterans and veterans exposed to other potential toxins. Data for birth defects in Iraq are of uncertain reliability. Furthermore, if an increase in birth defects in Iraq is validated, association with a particular agent would be very difficult because of the prevalence of a number of potential teratogenic agents in post-war Iraq, including malnutrition, disease, and chemical toxins released into the environment.

On another topic relating to reproductive health, McDiarmid et al. (2001) reported that initial veteran test results for DU-exposed veterans suggested a possible linkage between high levels of the hormone prolactin with DU internalization. This finding, however, was disputed by the National Academy of Science, and the possible association of DU with prolactin levels was not found during a subsequent reassessment (McDiarmid et al., 2001).

Kidney Toxicity

From Table 3, the estimated peak DU concentration for the maximally exposed Level I veteran is very high and is approximately equal to the LD50 concentration of 50μg U/g kidney (DOE, 2000). The peak concentration for the nominal case Level I veteran is, coincidentally, essentially at the 3μg U/g kidney worker limit. The acute kidney concentrations for the nominally and maximally exposed Level II veterans are 0.5 and 8μg DU/g kidney, respectively. The chronic uranium concentrations in the kidney for Level II and III veterans are all less than 0.1μg DU/g kidney. Table 3 shows that for the maximally exposed child at play, chronic and acute exposures are less than 0.1μg DU/g kidney. This concentration is below the minimal value reported for adverse kidney effects from chronic exposures of 0.1μgU/g kidney and acute exposures of 1μg U/g kidney (DOE, 2000). For downwind civilian populations, the estimated kidney concentration is insignificant.

Although adverse kidney health effects have not been reported for Level I veterans, maximally exposed veterans may have experienced undiagnosed transient kidney effects associated with high concentrations of DU in the kidney. Our current understanding suggests no long-term effects on the kidney; nonetheless, the potentially significant kidney damage predicted for some Level I veterans merits careful consideration and continued monitoring. The fact that no kidney failure fatalities or hospitalization have been reported for Level I veterans suggests that exposures significantly higher than the estimated values are unlikely. The calculations for all civilians show low uranium concentrations in the kidney, even when uncertainties are included. These results imply that no adverse kidney effects will result for Iraqi civilians.

Neurotoxic Effects

Because uranium is a heavy metal, DU internalization may be associated with neurotoxic effects. Pellmar et al. (1998) found that for rats implanted with DU pellets, uranium concentrated in the hippocampus area of the brain. Furthermore, excitability of neurons in the hippocampus was reduced for rats with significant internalized DU, although no behavioral differences were observed from a battery of behavioral tests. Among veterans, McDiarmid et al. observed a statistically lower score in one type of neurocognitive test for veterans with high uranium concentrations in their urine; however, normal functioning was not affected and the measured effect appears to be declining (McDiarmid et al., 2001). Although significant cognitive impairment has not been observed in veterans or test animals, sufficient evidence has been found to suggest some neurotoxic effects associated with large quantities of internalized DU. Further research on this topic is warranted to quantify the relationship between DU dose and neurocognitive effects.

Bone Damage

Studies have shown that Wistar rats injected with large acute doses of uranyl nitrate resulted in inhibition of bone formation (Ubios et al. (1998)). This dose appears to have been very high, given that almost all of the rats died within 15 days after receiving the injections. An earlier study showed that the effect of uranium injections on bone formation is dose dependent (Ubios et al. (1991)). The information from these tests is insufficient to determine the effect on bone formation for veteran and civilian exposures during the 1991 Gulf War. Further study is warranted.

Immune System

Radiological doses from DU exposure during the Gulf War are far too small to affect the immune system; however, studies have shown that DU may result in chemically induced immune system impairment (Wan et al. (2006)). Wan et al. performed in vitro experiments on viability and immune function as well as cytokine gene expression in murine (mouse) peritoneal macrophages and splenic CD4+ T cells. The authors conclude that DU can cause inappropriate apoptosis of macrophages, which can lead to both autoimmune problems and immunosuppression. The similarities between murine and human immune system genetics suggest that these findings may also apply to humans. These findings, however, have not been validated for humans using cohort data. Furthermore, the observed effects were found to be dose-dependent with a threshold for observed effects. No method has been established to correlate these in vitro DU exposure levels with human exposure levels by inhalation or ingestion. Because in vitro experiment results have not been correlated with human exposure levels, it is premature to conclude that Gulf War DU exposure did or did not result in immune system impairment. The importance of these findings suggest the need for further study with emphasis on establishing correlations of immune system effects with human exposure levels. Furthermore, cohort studies among Level I, II, and III veterans, Gulf War veterans not exposed to DU, and non-deployed veterans could provide valuable insight into possible adverse immune system effects relative to the quantity of DU internalized.

Gulf War Illness

The most common symptoms of what is now called Gulf War Illness include joint pain, fatigue, headache, rash, musculoskeletal system disease, psychological conditions, and memory loss (PAC, 2005). Although cancers and reproductive health concerns are sometimes included in this category, these health effects are treated separately in this study. The common symptoms of Gulf War Illness have not been identified as expected symptoms from the radiological effects of uranium internalization. However, Wan et al. (2006) have suggested that the chemical effect of DU internalization on the immune system may be related to some of the symptoms identified as Gulf War Illness. As stated above, only a small fraction of the deployed troops received exposures significantly in excess of background exposure to natural uranium. In addition, many other potential agents must be considered for all health effects issues. These other potential agents include pyridostigmine bromide, pesticides, indigenous infectious diseases, exposure to oil well fires, exposure to chemical warfare agents, vaccines, sand, and stress.

Conclusions and Recommendations

This study found that the quantity of DU internalized by most exposed veterans and civilians (those who did not occupy DU-contaminated vehicles) was smaller than typical exposure levels to natural uranium in the environment. Consequently, adverse health effects from DU exposure are unlikely for most veterans and nearby civilians. The risks of DU-induced leukemia or birth defects are far too small to result in an observable increase in these health effects among exposed veterans or Iraqi civilians. The analysis indicated that only a few US veterans (~5 men), who were in vehicles accidentally targeted by US tanks, received high DU exposures. These veterans were predicted to incur about a 1.4% lifetime risk of DU-induced fatal cancer, compared to about a 24% risk of a fatal cancer from all other causes. These veterans may have also experienced temporary kidney damage. Our current understanding suggests no long-term effects on the kidney; nonetheless, the potentially significant kidney damage predicted for some Level I veterans merits careful consideration and continued monitoring. No adverse kidney effects should result for all other veterans or civilians. Iraqi children playing for 500h in DU-destroyed vehicles are predicted to incur a cancer risk of about 0.4%. The only potential risk identified from external exposure to DU was the possibility of an increased risk of skin cancer for a postulated scenario wherein a large piece of DU metal is held in contact with the body for decades (e.g., jewelry). Monitoring of DU in the environment suggests that DU contamination of food and water has not been significant, and widespread DU contamination in the future is unlikely. The additions and modifications included in this analysis did not alter the conclusions reached in original Sandia study.

Although veteran medical records support the prediction of no observable increase in cancers from DU exposure, the latency period for detecting tumors for the principal cancer risk (lung cancer) is 10–15 years. Consequently, continued veteran monitoring and further testing to assess the possibility of synergistic chemical–radiation effects is recommended. Veteran test results suggest that major neurotoxic effects from DU exposure are unlikely; nonetheless, continued research is recommended to quantify the relationship of DU exposure to possible neurotoxic effects. Timely post-battle screening, continued medical examination of DU-exposed veterans, and tracking of veteran medical records are recommended to assure that no unexpected health effects develop. Studies have shown that DU may result in chemically induced immune system impairment. These findings, however, have not been validated for humans using cohort data. The importance of these findings suggest the need for further study with emphasis on establishing correlations of immune system effects with human exposure levels.

Reasonable precautions are recommended to reduce DU exposure from military use of DU. For example, to reduce the risk of civilian exposure, DU-damaged vehicles should be rendered inaccessible to civilians after hostilities have ceased. To assure that long-term contamination of the environment has not resulted from DU munitions use, continued monitoring of the post-battle zones and nearby civilian populations is recommended.



  1. Araneta M.R.G., Scglangen K.M., Edmonds L.D., Destiche D.A., Merz R.D., Hobbs C.A., et al. Prevalence of birth defects among infants of gulf war veterans in Arkansas, Arizona, California, Georgia, Hawaii, and Iowa. Birth Defects Res A Clin Mol Teratol 2003: 67: 246–260. | Article | PubMed | ChemPort |
  2. Arfsten D.P., Still K.R., and Ritchie G.D. A review of uranium and depleted uranium exposure on reproduction and fetal development. Toxicol Ind Health 2001: 17: 180–191. | Article | PubMed | ChemPort |
  3. Bem H., and Bou-Rabi F. Environmental consequences of depleted uranium use in the 1991 Gulf War. Enviro Int 2003: 30: 123–134. | Article | ChemPort |
  4. Bleise A., Danesi P.R., and Burkart W. Properties, use and health effects of depleted uranium (DU): a general overview. J Environ Radioact 2003: 64: 93–112. | Article | PubMed | ChemPort |
  5. Catalinotto J., and Flounders S., In: (Eds). Metal of Dishonor-Depleted Uranium: How the Pentagon Radiates Soldiers & Civilians with DU Weapons, International Action Center: New York NY, 2004.
  6. Cowan D.N., DeFraites R.F., Gray G.C., Goldenbaum M.B., and Wishik S.M. The risk of birth defects among children of Persian Gulf War veterans. N Engl J Med 1997: 336: 1650–1656. | Article | PubMed | ISI | ChemPort |
  7. Domingo J.L. Chemical toxicity of uranium. Toxicol Ecotoxicol News 1995: 2: 74–78. | ChemPort |
  8. Domingo J.L. Reproductive and developmental toxicity of natural and depleted uranium: a review. Reprod Toxicol 2001: 15: 603–609. | Article | PubMed | ChemPort |
  9. Domingo J.L., Paternain J.L., Llobet J.M., and Corbella J. The developmental toxicity of uranium in mice. Toxicology 1989: 55: 143–152. | Article | PubMed | ChemPort |
  10. Fetter S., and von Hippel F.N. The hazard posed by depleted uranium munitions. Sci Global Security 1999: 8: 125–161.
  11. Guilmette R.A., Parkhurst M.A., Miller G., Hahn F.F., Roszell L.E., Daxon E.G., et al. Human health risk assessment of capstone depleted uranium aerosols. Battelle Pacific Northwest Laboratory report PNWD-3442, October 2004.
  12. Hahn F.F., Guilmette R.A., and Hoover M.D. Implanted depleted uranium fragments cause soft tissue sarcomas in the muscles of rats. Environ Health Perspect 2002: 110: 51–59. | PubMed | ChemPort |
  13. ICRP. ICRP publication 30: Limits of intakes of radionuclides by workers. Pergamon: New York, NY, 1979.
  14. ICRP. ICRP publication 60: 1990 recommendations of the International Commission on Radiological Protection. Pergamon: New York, NY, 1991.
  15. ICRP. ICRP publication 66: Human respiratory tract model for radiological protection. Pergamon: New York, NY, 1994.
  16. ICRP. ICRP publication 69: Age-dependent doses to members of the public from intake of radionuclides: part 3 ingestion dose coefficients. Pergamon: New York, NY, 1995.
  17. Italian Ministry of Defense. Preliminary report - Investigative Commission of the Italian Ministry of Defense on the incidence of malignant neoplasia among military personnel involved in operations in Bosnia and Kosovo, March 2001.
  18. Leggett R.W., and Pellmar T.C. The biokinetics of uranium migrating from embedded DU fragments. J Environ Rad 2003: 64: 205–225. | Article | ChemPort |
  19. Marshall A.C. An analysis Of uranium dispersal and health effects using a gulf war case study, Sandia National Laboratories Report SAND2005-4331, July 2005.
  20. Martin F., Earl R., and Tawn E.J. A cytogenetic study of men occupationally exposed to uranium. Br J Ind Med 1991: 48: 98–102. | PubMed | ChemPort |
  21. McClain D.E., Benson K.A., Dalton T.K., Ejnik J., Emond C.A., Hodge S.J., et al. Biological effects of embedded depleted uranium (DU): summary of Armed Forces Radiobiology Research Institute research. Sci Total Environ 2001: 274: 115–118. | Article | PubMed | ChemPort |
  22. McDiarmid M.A., Squibb K., Engelhardt S., Oliver M., Gucer P., Wilson P.D., et al. Surveillance of depleted uranium exposed gulf war veterans: health effects observed in an enlarged friendly fire cohort. J Occup Environ Med 2001: 43: 991–1000. | PubMed | ChemPort |
  23. Mettler F.A., and Upton A.C. Medical effects of ionizing radiation, 2nd ed. WB. Saunders: Philadelphia, PA, 1995.
  24. Miller A.C., Blakely W.F., Livengood D., Whittaker T., Xu J., Ejnik J.W., et al. Transformations of human osteoblast cells to the tumorigenic phenotype by depleted uranium-uranyl chloride. Environ Health Perspect 1998: 106: 465–471. | Article | PubMed | ChemPort |
  25. Miller A.C., Brooks K., Smith J., and Page N. Effect of the militarily-relevant heavy metals, depleted uranium and heavy metal tungsten-alloy on gene expression in human carcinoma cells (HepG2). Mol Cell Biochem 2004: 255: 247–256. | Article | PubMed | ChemPort |
  26. Miller A.C., Xu J., Stewart M., Brooks K., Hodge S., Shi L., et al. Observation of radiation specific damage in human cells exposed to depleted uranium: dicentric frequency and neoplastic transformations as endpoints. Radiat Prot Dosimetry 2002: 99: 275–278. | PubMed | ChemPort |
  27. PAC. Presidential Advisory Committee Presidential Advisory Committee on Gulf War Illnesses, Final Report Chapter 3 2005.
  28. Parkhurst M.A., Szrom F., Guilmette R.A., Holmes T.D., Cheng Y.S., Kenoyer J.L., et al. Capstone depleted uranium aerosols: generation and characterization, Volumes 1 and 2. Pacific Northwest National Laboratory report PNNL-14168 October 2004.
  29. Pellmar T.C., Hogan J.B., Benson K.A., and Landauer M.R. Toxicological evaluation of depleted uranium in rats: six month evaluation point. Armed Forces Radiobiology Research Institute: Bethesda MD, 1998.
  30. Rostker B. Environmental exposure report: depleted uranium in the Gulf (II), Second interim DoD report. 2000179-0000002. Tab B. December. Available:[accessed 1 June 2006] 2003.
  31. Royal Society. The health hazards of depleted uranium munitions Part I. The Royal Society: London, UK, 2002a.
  32. Royal Society. The health hazards of depleted uranium munitions Part II. The Royal Society: London UK, 2002b.
  33. Ubios A.M., Braun E.M., and Cabrini R.L. Effect of biophosphonates on abnormal mandibular growth in rats intoxicated with uranium. Health Phys 1998: 75: 610–613. | PubMed | ChemPort |
  34. Ubios A.M., Guglielmotti M.B., Steimetz T., and Cabrini R.L. Uranium inhibits bone formation in physiologic alveolar bone modeling and remodeling. Environ Res 1991: 54: 17–23. | Article | PubMed | ChemPort |
  35. United Nations Environmental Programme. Depleted uranium in bosnia and herzegovina post-conflict environmental assessment. UNEP: Nairobi, Kenya, 2003.
  36. USACHPPM. Depleted uranium human exposure assessment and health risk characterization in support of the environmental exposure report "depleted uranium in the Gulf" of the Office of the Special Assistant to the Secretary Of Defense for Gulf War Illnesses, Medical Readiness and Military Deployments (OSAGWI), Health Risk Assessment (HRA) Consultation no. 26-MF-7555-00D, September 2000.
  37. US Department of Energy. Guide to good practices for occupational radiological protection in uranium facilities, US DOE Standard SAFT-0072 (DRAFT), 2000.
  38. U.S. Environmental Protection Agency. Nickel refinery dust data from the Integrated Risk Information System (IRIS) database, Available:[accessed 1 June 2006] 1999.
  39. Wan B., Fleming J.T., Schultz T.W., and Sayler G.S. In vitro immune toxicity of depleted uranium: Effect on murine macrophages CD4+ T cells, and gene expression profiles. Environ Health Perspect January 2006: 114: 85–91. | ChemPort |
  40. Waxweiler V.E., Archer, Rosco R.J., Watanabe A., and Thun M.H. Mortality patterns among a retrospective cohort of uranium mill workers. Proc 16th Midyear Topical Symposium of the Health Physics Society 1983, 428–435.
  41. WHO. Depleted uranium: sources, exposure and health effects, World Health Organization Report WHO/SDE/PHE/01.1. April 2001.


This work (Sandia document SAND 2006-4999J) was performed for the US Department of Energy. I am grateful for the review of this analysis by Dr. Celeste Drewien and Dr. Leonard Connell and the support of Dr. Clyde Layne and Dr. Jon Rogers from Sandia National Laboratories National Securities Studies Department. I am also grateful for the review of the medical aspects of this study by Dr. Larry Clevenger from Sandia National Laboratories Medical Department.


These links to content published by NPG are automatically generated.


Radiation dose assessment of exposure to depleted uranium

Journal of Exposure Science and Environmental Epidemiology Article Response

Limited Nuclear War

Scientific American Article (01 Nov 1976)