The study of Dupree and colleagues (1987) found a significant excess risk of nonmalignant respiratory disease based on 32 deaths; Frome and colleagues (1990) also observed a significantly increased risk of nonmalignant lung disease. However, the other reports, including the larger studies of Checkoway and colleagues (1988), Polednak and Frome (1981), and Ritz (1999), which showed SMRs of less than or close to 100, do not confirm those findings. Polednak and Frome (1981) showed a small but statistically significant risk (SMR = 110), but the study is important principally because it provides important evidence against a large excess risk of nonmalignant lung disease. In this cohort, exposure was probably relatively intense because it occurred during the World War II era when the control of uranium dust was less stringent. In addition, the period of follow-up was long, and the number of expected deaths was large. None of the studies were able to control for smoking, a major causal factor in chronic respiratory disease, or other occupational exposures, which limits the interpretation of the findings.

The information on other health outcomes in humans comes primarily from case reports of workers or other individuals accidentally exposed to large doses of uranium compounds. These health outcomes have not been examined in detail in human studies.

Accidental inhalation exposure of one individual to high levels of uranium produced transient gastrointestinal distress, characterized by loss of appetite, abdominal pain, diarrhea, and pus and blood in the stool (Lu and Zhao, 1990). A case of accidental dermal exposure to uranium (Lu and Zhao, 1990) had no reported gastrointestinal effects.

The human literature lacks documentation on adverse immunological or lymphoreticular effects of uranium. The detection of systemic lupus erythematosus (SLE)-typical antibodies in quartz dust-exposed uranium miners indicates a potentially higher risk for the development of systemic autoimmune disease (Conrad et al., 1996, 1998). The authors of one report detected the 16/6 idiotype,