campusreview.com.au
antidepressants – they are more likely to be overweight, have
diabetes, smoke, use multiple drugs, use illicit drugs and drink
alcohol. Virtually all of those we know have potentially negative
outcomes on pregnancy.
You have to control for all of these things; some studies do,
some studies don't. Immediately you have these biases and meta-
analyses try to overcome – as best they can depending on the
information given – these issues.
The meta-analysis that we’re talking about [was done] because over
time we get more and more data, so they have – unlike previous ones
– more data access. They were fairly rigorous with excluding ones
where they didn't think the methodology was adequate.
You also have to take into account how
many babies with cardiac defects are born; it’s
roughly eight for every 1000 deliveries.
Their outcomes show that if you take fluoxetine, it gives you a
slightly increased risk of having a baby with a cardiac malformation.
[This outcome] is not a surprise, because some of those studies
they analysed also found that. What it is doing is just giving more
weight to that. It's not something we didn't know about; we've
known about fluoxetine. There have been other studies, including
ones they used, that found fluoxetine to be problematic.
I would have issue with their belief that, for an Australian
population, fluoxetine is the most commonly prescribed one
in pregnancy. We haven't been doing that – certainly, perinatal
psychiatrists in Victoria or Australia-wide haven't been doing
that – for some time. We don't use it because of the findings we
already knew about, and also because of its long half-life. We really
don't want to expose the babies in pregnancy and post-partum in
particular if they want to breastfeed, with something that's difficult
to change because it hangs around so long.
I think the other important thing with this study is that, yes,
there is an increased relative risk, but it is still quite low. Their
meta-analysis is still not perfect. For instance, some of the studies
they used did calculate for alcohol use, some did not. If you just
looked at the studies that did allow for alcohol usage, the relative
risk rate went down from, I think, 1.3 to 1.18. That sort of shows
that by controlling for these confounding variables it can alter that
risk rate. In lay terms, that means that for every 100 women that
have a baby with a cardiac defect – every 100 women not taking
antidepressants – there will be 118 or 130, depending on which
figures you're looking at, that will have a cardiac defect.
You also have to take into account how many babies with cardiac
defects are born; it’s roughly eight for every 1000 deliveries. So
again, to get 100 women not taking antidepressants to have babies
with cardiac defects, we're talking about a lot of births. It has to be
put into that context.
What might these defects constitute in terms of a major defect and a
cardiovascular defect?
They're only pinpointing cardiovascular ones as being linked to
fluoxetine, although this group broadly does increase risk because
of those compounding variables. The cardiac ones are two types:
one is the septal defects, which are the holes in the heart, and
INDUSTRY & RESEARCH
many of those holes in the heart are very minor and don't need
any intervention and fix themselves by adulthood. The non-septal
defects can be more serious, and these are ones with problems
with valves and the actual vessels of the heart. They can require
surgery and potentially be not compatible with life.
So there is the full range, but many of those septal defects are at
the minor end.
I also just want to add that the thing that needs to be taken into
account for the women who have to balance this risk, and for
the doctors that are trying to advise them, is that being unwell
in pregnancy – so being off medication and unwell – actually
constitutes a risk as well.
We also know women don't look after themselves if they're
unwell, and therefore they won't be looking after their pregnancy;
they won't be eating well, either not putting on weight or putting
on too much weight, smoking more – all of those things which will
impact on the health of the fetus. Also, we know that very anxious
women through pregnancy have babies who are born with higher
cortisol levels, and that higher cortisol stays with them all their
lives, and is probably a marker for later illnesses and mental health
disorders in those women.
I've noticed that there are also potential risks associated with other
types of antidepressants, such as SSRIs. Are they more, less or the
same in terms of the risk posed to the fetus?
The only two SSRIs that have come up as having a specific risk
are paroxetine and fluoxetine. All of the other ones are linked in
together with no specific risk attached to them. The increased risk
is probably more associated with all those compounding variables,
and the other thing that this paper didn't control for and which
is largely unknown is dosage. We're looking at probably a big
difference between the risks for the fetus if you're taking 20mg of
fluoxetine versus taking 80mg of fluoxetine. That exposure is going
to be very different. So dose is important to consider as well.
Some of the other antidepressants, other than those two I've
mentioned, in the SSRIs ... are very clearly safer than fluoxetine
and paroxetine.
What advice would you give women who are concerned about this
analysis?
Not to panic. If they are already pregnant, then go see their doctor
as soon as possible if they haven't already discussed it with their
doctor. Each individual has to weigh up different sets of risks. If
someone has become very unwell, unable to work, acutely suicidal
when they're depressed, then the risks of being off the medication
are going to be much, much greater than staying on it. Those are
the sorts of things that ideally you discuss with your doctor before
getting pregnant.
I do lots of these sorts of consultations to plan these
pregnancies, and if, let's say, someone was on fluoxetine, I would
look at whether we could switch them to something less risky.
What risks would be involved with that? How long is it going to
take and is it worth doing? Can we reduce the dose of whatever
they’re on? Can we help them lose weight, make sure their diabetes
is controlled? Try to help them stop smoking? Do all of the other
things as well as just looking at the medication.
And making sure they're doing things like yoga, exercising
regularly, good sleep habits – all the things that will help maintain a
healthy pregnancy. ■
15