Abstract
BACKGROUND
In recent years there are a number of reports showing a deterioration in male
reproductive health, i.e. diminished semen quality and increases in the incidence of
testicular cancer and the congenital malformations cryptorchidism and hypospadias.
It is hypothesised that these changes have been caused by increasing in utero
exposure to environmental oestrogens and/or anti-androgens
OBJECTIVES
(i) Describe the geographical distributions of three indicators of male
reproductive health in Scotland (i.e. testicular cancer, cryptorchidism and
hypospadias).
(ii) Describe the conjoint geographical distribution of the three indicators
specified in (i).
(iii) Identify explanatory factors that might account for the geographical
distribution of male reproductive health in Scotland.
METHODS
An epidemiology study modelling the geographical distributions using routinely
collected data of the three indicators. The primary assessment of the geographical
distribution of the indicators was by means of the relative risks at postcode sector
level. If geographically varying risk factors (environmental or not) are associated
with these conditions then would expect to see clustering of relative risks. Bayesian
methods were used to estimate the relative risks so as to account for their variability
due to areas with small number of cases. These Bayesian models were developed
further by including potential covariates to assess if these area specific factors
explain the spatial variation of the three indicators. In addition, Bayesian modelling
of individual data pertaining to the cryptorchidism cases was also carried out to
explore whether the spatial variation in risk might also be explained by the nature of
the cases within each postcode sector rather than area specific covariates. Finally, a
Bayesian model which combined all three indicators was developed to examine the
spatial relationships between the three disease/conditions.
RESULTS
There are similarities in the spatial pattern of the cryptorchidism and hypospadias
relative risks, with both conditions having clusters of high relative risks in the East
and South-West of Scotland. The spatial variation of the testicular cancer relative
risks is not similar to the other two conditions nor is it conclusive that it has a distinct
spatial pattern. The relative risks of the postcode sectors for all the indicators are
associated with radon measurements and the rural/urban indicator. The spatial
analysis of individual information concerning the cryptorchidism cases indicate that
the spatial variation of the relative risks might also be explained by individual
information; namely maternal age and co-morbidity with hypospadias.
CONCLUSIONS
There does appear to be geographically varying risk factors associated with these
three conditions. Furthermore, as the spatial variation of cryptorchidism and
hypospadias is similar it is likely that they have some common aetiology. As the
same risk factors were found to be associated with testicular cancer and the
congenital malformations, then this carcinoma appears to share some aetiology with
cryptorchidism and hypospadias. Therefore there are geographically varying risk
factors whose exposure occurs in utero, that are associated with all three conditions,
providing some evidence to support the proposed hypothesis. Flowever, the common
aetiology of these conditions could not only to be environmental but also due to
genetic and life-style factors, that could pertain to the individual cases rather than the
specific area. Therefore, further studies are required to investigate the associations
between all the disease/conditions of male reproductive heath and the various
potential risk factors.
