Radium in drinking water and the risk of death from bone cancer among Ontario youths.

OBJECTIVE: To determine whether residents of Ontario who are exposed to radium 226 naturally occurring in drinking water are at increased risk of bone cancer. DESIGN: A population-based case-control study of records from death and birth registries. Water samples were obtained from residences at the time of birth and of death. SETTING: Ontario. PARTICIPANTS: All Ontario-born people under the age of 26 years who died of bone cancer between 1950 and 1983. Control subjects were those who died of any other disease matched by age, sex and year of death. OUTCOME MEASURES: Radium exposure distributions and estimation of risk. RESULTS: An association was found between death from bone cancer and exposure to radium at the birthplace residence in concentrations of 7.0 mBq/L or more (odds ratio 1.58, 90% confidence interval [CI] 1.01 to 2.50; p = 0.047). There was a statistically significant exposure-response relation (p = 0.045). The increase in risk was similar for the main types of childhood bone cancer: osteosarcoma, Ewing's sarcoma and chondrosarcoma. CONCLUSIONS: The estimated risk at these exposure levels is much higher than would be predicted. The association may be spurious, the point estimates of risk may be too high, or risk factors derived from other exposure circumstances may not be valid for exposure to radium beginning in the prenatal period. Should the findings be confirmed, consideration might be given to removing radium from drinking-water sources.