Revival: Biological Effects of Low Level Exposures to Chemical and Radiation (1992)

32 BIOLOGICAL EFFECTS OF LOW LEVEL EXPOSURES

state. This site was built in 1943-44 to produce plutonium, which mainly
emits 7 -radiation. Gilbert et al.103105 calculated cancer SMRs for workers at
the Hanford site relative to those occurring in the general U.S. population.
In general, cancer SMRs were less than unity (averaging 0.85), indicating
that Hanford workers were less likely to develop cancer than the U.S.
general population. This was attributed to the “healthy worker effect” 106-108
(i.e., employed workers must be able to conduct productive work and are
therefore healthier than the general U.S. population). This occurs because

the latter population contains individuals unable (too sick) to work and
therefore more prone to mortality. Moreover, comparisons of data within
the Hanford site showed no evidence of a positive correlation between
radiation dose and cancer mortality.
Stewart and Kneale 109 have been highly critical of these findings. They
divided the Hanford healthy worker effect into two components:

1. the external healthy worker effect, which asserts that workers were gener­
   ally healthier than the overall population (see above)


2. the internal healthy worker effect, which asserts that, within this site,
   healthier workers were more likely to receive jobs exposing them to higher
   levels of radiation

They point to a strong negative correlation at the Hanford site between

radiation dose and overall mortality. In short, they believe that those work­

ers most at risk from cancer due to radiation exposure were unusually

healthy (internal and external healthy worker effects), and that this inexora­

bly biases the analysis in favor of reduced mortality among the Hanford

workers.

It is not my intention here to sort out all the arguments or to reach any

conclusion about the epidemiological issues. Rather, I merely note that a

negative correlation between radiation exposure and mortality is consistent

with longevity hormesis. If longevity hormesis occurs in human populations

exposed to ionizing radiation, the epidemiological consequences are formi­

dable (to say the least!). How does one assimilate (1) an external healthy

worker effect, (2) an internal healthy worker effect, (3) toxicity, and (4)

longevity hormesis? This conundrum is not exclusive to the radiation indus­

try, but rather permeates most epidemiological studies involving exposures

to potentially toxic agents. For example, onsite Australian petroleum indus­

try workers have an SMR (all causes of death) of 0.63 relative to the Austra­

lian national population.110

CONCLUSION

There is no love like the old love, that we

courted in our pride;

Though our leaves are falling, falling,

and we’re fading side by side,