WASHINGTONThe latest effort to establish whether exposure to
radiation during nuclear weapons tests caused leukemia in some
military personnel has left the issue largely unresolved. The study
by the Institute of Medicine (IOM) found a 14% increase in leukemia
deaths among test participants, compared with a control group, but
the increase was not significant. The increase was significant,
however, for soldiers exposed to land-based testing.
The study also showed a significantly increased risk of death from
nasal cancer, prostate cancer, and external causes, such as motor
vehicle accidents, for the servicemen at the nuclear tests. The
findings on nasal and prostate cancer were not expected based
on other studies of the health effects of radiation exposure,
the report said.
The study, The Five Series Study: Mortality of Military
Participants in the US Nuclear Weapons Tests, is one of the
largest of its kind. It was funded by the Department of Defense and
carried out by four members of the IOM staff.
Between 1945 and 1963, more than 200,000 US soldiers, sailors,
airmen, and marines took part in atmospheric nuclear weapons tests.
Concern about their potential cancer risks accelerated in 1976 after
a veteran claimed his leukemia resulted from exposure to radiation
during a 1957 test.
The new IOM study, which supersedes one published in 1985, examined
the causes and rates of death among nearly 70,000 servicemen present
during at least one of five groups of atmospheric nuclear tests in
the Nevada desert and South Pacific during the 1950s. These data were
compared with deaths among nearly 65,000 servicemen who were on
active duty at the same time, but who were not present at any nuclear
tests. The reports findings are based on more than 5 million
person-years of mortality follow-up.
Insufficient Dose Data
A major problem for the researchers was that dose data for test
participants were insufficient and judged inappropriate for
epidemiologic analysis. As a result, the study assessed the
hazards associated with test participation, irrespective of dose.
The key findings of the study were that, overall, participants and
referents had similar risks of death and death from cancer, and that
the 14% higher risk of leukemia deaths among participants could
be a chance finding.
The number of estimated excess deaths from leukemia among nuclear
test participants was 25. Only a study cohort four times the
size of the one available would have been likely to identify the
observed leukemia risk as statistically significant, the report said.
The researchers found a statistically significant increased risk of
leukemia deaths among servicemen at the Nevada site tests, compared
with land-based members of the control group. And the estimated
relative risk for leukemia mortality was higher among those
servicemen exposed to radiation on land than those exposed at sea.
There was no articulated a priori basis to have predicted such
a finding, the report said.
However, participants in the South Pacific tests had an
observed but not significant lower risk of dying of leukemia
than members of the control group who were stationed at sea. There
was also a statistically significant increased risk of death from all
causes among the South Pacific test participants, compared with the
The researchers found that among leukemia subtypes, the highest
relative risk existed with lymphoid leukemia, excluding chronic
The leukemia findings are consistent with other studies and do
not resolve the debate over whether either participation in general
or the radiation dose in particular is associated with leukemia
mortality, the report said. It added that the set of
leukemia findings presented here is broadly consistent with a
radiogenic cause, but is not conclusive.