hormones like testosterone are crucial to the development of the
male reproductive tract in all mammals. Laboratory
experiments with animals have shown conclusively that compounds
interfering with androgen actionanti-androgenscan
disrupt the normal development of males. Some of the effects produced
include hypospadias, cryptorchidism
and polythelia. DDE, a metabolite of
DDT, is one such anti-androgen. Each of these birth defects can
be produced reliably in substantial percentages of exposed animals
via exposure to DDE.
health records analyzed by the US Centers for Disease Control and
sister agencies around the world have suggested that some male reproductive
birth defects, including hypospadias
and cryptorchidism have
increased in frequency during the latter half of the 20th century,
in concert with the widespread use of hormonally-active compounds
like DDT. Hence because of the laboratory work on anti-androgens
in animals, the rise in use of anti-androgenic compounds and the
increases in frequency of male reproductive tract birth defects,
scientists have hypothesized that anti-androgenic compounds are
contributing to these human health effects.
new publication by Longnecker et al. attempts to test this
hypothesis. It is the second in a series that take advantage of
samples of umbilical cord blood stored since the 1960s, combining
sophisticated chemical analysis with data on birth outcome. The
first of these studies revealed
a strong relationship between DDE levels and the likelihood of preterm
birth and low birth weight.
Longnecker et al. present an analysis suggesting inconclusive
links between DDE levels in umbilical cord blood and the likelihood
of three male reproductive tract birth defects, hypospadias, cryptorchidism
and polythelia. There may be elevations in the risk of these birth
defects,... but the statistical analysis cannot rule out the possibility
of no association. The suggestion of links to cryptorchidism and
polythelia are stronger than those to hypospadias. Possible explanations
for this finding are discussed below, after a
description of their methods and results.
did they do?
blood samples were obtained as part of the Collaborative Perinatal
Project, research involving women recruited at 12 different study
centers in the US between 1959 and 1966. Blood samples were collected
every 8 weeks during pregnancy from the approximately 42,000 women
who participated, and then once again 6 weeks post-partum. The children
were then assessed for birth defects and other outcomes through
the age of 7 years.
blood samples had been stored for over 30 yrs when Longnecker et
al. began a chemical analysis in 1997. The samples had been
kept frozen at 20°C since collection; records indicate
there was no thaw during that during storage.
et al. measured serum levels of p,p-DDT and p,p-DDE
in samples taken during the third trimester of pregnancy using electron
capture detection after solid-phase extraction, cleanup and dual-gas
then examined statistical associations between the levels of DDE
and the occurrence of the three birth defects using a case-control
design with one control group. The analysis used statistical procedures
to control for a series of potentially confounding variables, including
race, season of birth, mothers
age, parity, socioeconomic index, body mass index before pregnancy,
weight gain during pregnancy, smoking during pregnancy, hypertension,
age at menarche, history of infertility, menstrual cycle irregularity,
estrogen use during pregnancy,
progesterone use during pregnancy, method of delivery, and
serum sodium level.
did they find? In the 22,347 boys of the study, the authors
found records of 241 cases of cryptorchidism, 214 cases of hypospadias,
and 185 cases of polythelia. They then constructed a control group
of 599 boys without recorded birth defects, selected otherwise at
random from the study. A small number of the case group was subsequently
excluded because of details of the chemical analysis. The authors
were then left with sample sizes of: 219 cryptorchids, 199 with
hypospadias, 167 with polythelia and 552 controls.
median level of DDE measured in the control group, 34.3 µg/liter
(adjusted for proportion recovered) was approximately 5-fold the
current level in the United States.
each of the 3 birth defects, results suggested an increase in risk
associated with higher DDE levels. The adjust odds ratio reflected
an modestly increased risk for the highest exposure groups, but
the 95% confidence intervals were wide and included 1 (that is,
they could not exclude the possibility on a statistical basis that
the risk was not elevated).
following table presents this element of Longnecker et al.'s
findings. It shows the adjusted odds-ratio for the highest exposure
group (> 60 µg/liter DDE) for each birth defect along with
95% confidence intervals.
> 60 µg/L DDE
strongest association was for polythelia: the median level of DDE
in cord blood for boys with polythelia was elevated compared to
controls and the adjusted odds ratio for the highest exposure group
compared to controls was 1.7.
does it mean? These results are inconclusive. There were suggestions
of modest elevations in adjusted odds ratios, but, as noted above,
the confidence intervals were wide and included 1. This means
that the authors cannot exclude the possibility that there is no
association, in other words a negative result.
noted above, the DDE levels found in these cord blood samples is
significantly higher than current levels. This decline has occurred
at the same time as increases in the frequency of hypospadias and
cryptorchidism. The combination of these opposite time trends and
the lack of unambigous association between DDE levels and hypospadias
and cryptorchidism odds-ratios would seem to argue that the hypothesis
linking these should be rejected.
are at least 2 significant challenges to this interpretation, however.
the current result may be a "false negative," i.e., that
the statistics indicate there is no relationship when in fact there
is. Two factors at work here would be consistent with a false negative.
sample sizes were small, rendering the statistics unfortunately
weak even if elegant.
is only one of a series of known compounds with anti-androgenic
characteristics. Without a more complete picture of the exposure
history of the fetus, it is impossible to say that other compounds
did not confuse the results. Some of the instances could have
been caused by one anti-androgen; the others by DDE. Data looking
only at DDE would predict the incidence of a birth defect very
imprecisely, and the small sample size would make it very difficult,
if not impossible, to detect the effect of DDE by itself.
the fact that hypospadias have been increasing while body burdens
have been increasing can also be misleading. Again, there are many
more anti-androgens than just DDE. Exposure to others, particularly
several phthalates, has is likely to have been increasing (because
of increased production volume of phthalates and use in a broader
variety of products, especially such consumer products as cosmetics).
It would be highly desirable to have information on time trends
of exposure to the total pool of anti-androgens. It is possible
that total exposure has increased even as the contribution of one
anti-androgen, DDE, to that total, has decreased.
the authors conclude, their results leave unanswered the question
as to whether DDE or other anti-androgens contribute to the causation
of reproductive birth defects in boys. It remains a plausible hypothesis:
summary, these data alone provide no clear evidence of an
effect of DDE on male development, but given the experimental
and emerging human evidence of DDE as an androgen antagonist,
there remains the suspicion that high maternal levels of DDE
may affect the development of male offspring."