Childhood cancer mortality and radon concentration in drinking water in North Carolina

Gw Collman,Dp Loomis,Dp Sandler

doi:10.1038/bjc.1991.143

Gw Collman, Dp Loomis + Show 1 more

Open Access

https://doi.org/10.1038/bjc.1991.143

Copy DOI

Abstract

We explored the association between groundwater radon levels and childhood cancer mortality in North Carolina. Using data from two state-wide surveys of public drinking water supplies, counties were ranked according to average groundwater radon concentration. Age and sex-adjusted 1950-79 cancer death rates among children under age 15 were calculated for counties with high, medium, and low radon levels. Overall cancer mortality was increased in counties with medium and high radon levels. The strongest association was for the leukaemias, but risks were also suggested for other sites. These associations could be due to confounding or other biases, but the findings are consistent with other recent reports.

Highlights

We have explored the potential link between radon and childhood cancer mortality in North Carolina using countyspecific levels of radon in groundwater as our indication of exposure
For 25 counties with missing data, radon concentration was imputed by linear regression based on radon concentrations in other North Carolina counties with similar geological characteristics (Loomis et al, 1988)
The relative risk for all childhood cancers combined was slightly increased for both medium (RR = 1.16, 95% confidence interval (CI) 1.05, 1.28) and high radon counties (RR= 1.23, 95% CI 1.11, 1.37)

Summary

Radon measurements

The concentration of radon in groundwater supplies in North Carolina was available from surveys of 308 public water supplies in communities with a population of at least 100 people. These were carried out by the North Carolina Department of Human Resources and the US Environmental Protection Agency in 1975 and 1981-2 (Sasser & Watson, 1978; Horton, 1983). Relative risks were increased for brain and CNS tumours, lymphomas, and bone cancer, but their precision was poor and dose-response gradients were not always present

Cancer mortality rates

Results

Discussion

Bone Other cancers