Basic HTML Version
Nonetheless, the pooled analysis reported in this issue of the Journal offers support
from another line of investigation for use of a linear, nonthreshold relationship in
estimating the risk of indoor radon. The findings of the North American pooling are
mirrored by those of the just-reported pooling of 13 European studies involving 7148
cases and 14,208 controls (Darby et al., 2005). The risk for lung cancer was estimated
to increase by 8.4% per 100 Bq/m3 compared with 11% in the North American data.
The dose-response relationship appeared linear and without indication of a threshold.
The policy implications of a linear, nonthreshold, dose-response relationship are
substantial, and finding support in general population studies further reduces
uncertainty on the risk of residential radon exposure. With completion of the European
pooling and then the global pooling, this part of the radon story will be finished, as
further lung cancer case-control studies are not in progress. There is strong coherence
of the epidemiological data on lung cancer risk from studies of miners and the general
population with the experimental data from animal and cellular models. Policies for
testing and control can be based with confidence in the assumption that risk increases
with concentration and that a “safe” concentration cannot be specified.
Is further research needed? Taken together, the epidemiological studies of miners and
the general population, along with laboratory evidence and biophysical theory, give a
strong foundation for presently used risk models. In fact, there are few environmental
carcinogens for which human data are so abundant and the mechanism of
carcinogenicity so well understood.
There are a few remaining details, however. Follow-up of the miner cohorts should be
maintained so that the excess lung cancer risk caused by radon can be charted across
the life span. These high-risk cohorts might also be informative on the genetic basis of
susceptibility to radon. Presumably, another pooled analysis of the miner studies will
soon be warranted. I also look to the experimentalists to continue to refine our
understanding of how radon causes lung cancer. Work still in progress is refining the
exposure measures in some of the case-control studies using surface radioactivity of
glass as an index of cumulative exposure. There is also the possibility of exploring
genetic modifiers of risk from residential exposure. However, I do not anticipate new
research findings that would move us away from the present strategy of measurement
and mitigation. And finally, no more case-control studies of lung cancer, please!