Monitoring of possible health effects of living in the vicinity of nuclear sites in Belgium. Is there an increased incidence of childhood leukaemia?

Introduction
This study has been carried out at the demand of the Belgian
Minister of Social Affairs and Public Health to assess, by means
of an epidemiological study at the national level, possible health
effects of living in the vicinity of nuclear sites.
Objectives
To investigate (1) whether there is an excess in incidence of
acute leukaemia in children aged 0-14 years within the 20km
proximity area around the nuclear sites in Belgium as compared
to what is expected in a reference area, and (2) whether there
is evidence for an (increasing) leukaemia risk with increasing
‘surrogate’ exposure around the nuclear sites, i.e. (i) residential
proximity to the site, (ii) prevailing winds, and (iii) radio-active
discharge estimates from the plants based on mathematical
modeling.
Methodology
Ecological study around the nuclear installations of Class 1 in
Belgium, i.e. the nuclear power plants of Doel and Tihange,
the nuclear sites of Mol and Fleurus, and the Belgian territory
within the 20km proximity area around the nuclear power plant
of Chooz (France). Cancer incidence data were retrieved from
the Belgian Cancer Registry (BCR), a population-based registry
at the national level. For this study, leukaemia data were available
from 2002 till 2008 for Flanders and from 2004 till 2008
for Brussels and Wallonia. Age-and sex-specific population data
for this period were collected from the Federal Public Service
(FPS) Economy, S.M.E.s, Self-employed and Energy. Dat a on
socio-economic status and urban-rural situation were included
as covariates.
To study a possible excess of childhood leukaemia, age- and
sex-standardized incidence ratios (SIRs) were calculated
and Poisson regression was carried out. In a further step, we
conducted a sensitivity analysis by varying the radius of the
proximity area; SIRs and RRs were then recalculated for each of
these areas of proximity with increasing radii.
The hypotheses of increasing childhood leukaemia incidence
with increasing levels of ‘surrogate’ exposure were tested by
means of focused hypothesis tests (conditional forms), i.e.
Stone’s Likelihood ratio test and Bithell Linear Risk Score
Test. To facilitate the interpretation of the results from the
focused hypothesis tests, the ‘exposure’-response relations were
estimated and visualized using varying coefficient models.
Results
Results of the study will be presented and discussed at
the Congress.