Selective serotonin reuptake inhibitors in pregnancy and congenital malformations: population based cohort study
BMJ 2009; 339 doi: https://doi.org/10.1136/bmj.b3569 (Published 23 September 2009) Cite this as: BMJ 2009;339:b3569
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Dear Dr Pedersen,
I read your article in the BMJ with great interest, and would like to
ask you for some added information. I assume that by septal defects you
refer to atrial and also to ventricular septal defects. My question is how
and when these were diagnosed. Small atrial septal defects are very
common. Those under 3 mm in size all close within the first year, and many
< 6mm diameter do the same. Therefore it makes a difference how they
were diagnosed. If there were routine or very frequent neonatal
echocardiograms, then a huge number of small and perhaps transient ASDs
will be diagnosed. If they were diagnosed clinically and perhaps after 1
year of age, then probably most would be at least moderately large and
clinically significant. The same problem applies to the VSDs, which have
an even higher incidence of spontaneous closure. It is quite possible that
mothers who had taken SSRIs during pregnancy were more intent on seeking
investigation of their infants, with resultant increase in detection of
minimal lesions.
The issue of the size of the defects is crucial for two reasons. If
they are small, they are usually trivial clinically and do not argue
against the use of SSRIs. Furthermore, developmentally the small ASDs and
VSDs may represent delayed closure of a normal septum rather than
teratological interference with a developmental mechanism.
I would appreciate your comments on these views, and also wonder of
you had data about the size and clinical significance of these defects.
Kindest regards, Julien Hoffman,
Professor of Pediatric Cardiology (Emeritus)
Competing interests:
None declared
Competing interests: No competing interests
We should not fret about the several limitations of yet another
cohort study and instead concentrate on recognising the findings of
existing studies and maintain caution about prescribing SSRIs in
pregnancy.
In the Israeli study, Rachel Levinson-Castiel et al used the Finnegan
scoring to assess neonatal abstinence syndrome (NAS)(1). Almost a third
of these neonates exposed to SSRIs in-utero (either through the entire
pregnancy or at least during the third trimester) had symptoms of NAS.
Four SSRIs were compared (Paroxetine, Fluoxitine, Sertraline and
Citalopram) with the greatest risk of NAS correlating to a higher dose of
Paroxetine. In this study, 3 out of 60 infants with prolonged in-utero
exposure to SSRIs developed congenital anomalies including ventricular
septal defects.
Pedersen et al confirm a similar risk of cardiac malformations but
provided fair assurance that no specific SSRI was associated with major
malformation. The association of septal heart defects was greater with
Sertraline and Citalopram but not, in this study, with Paroxetine (OR
1.99) (2).
A six-fold greater risk of persistent pulmonary hypertension in
newborns with in-utero exposure to SSRIs after the 20th week of gestation
was found by Chambers et al (3). The risk was not, however, increased if
SSRI use was restricted to the first trimester.
Associations are also found with SSRI use and decreased gestational
age, spontaneous abortion (4), reduced APGAR score (5) and decreased birth
weight (5,6,7,8).
Salvatore Gentile draws attention to the possible risk to long term
neurocognitive development of children exposed to SSRIs in-utero (9).
Results however are conflicting. The Stanford Study found significantly
lower scores on neurobehavioral functions and psychomotor indices in
children exposed to SSRIs in-utero compared to children of women with
major depression who were treated with psychotherapy (10). On the other
hand, a study conducted at the University of Toronto showed no differences
in well established neurobehavioural indices between exposed and unexposed
children (11) and prospective studies on Fluoxetine usage during pregnancy
could find no effect on global IQ, language development or behavioural
development in preschool children (12). Untreated depression was
associated with poorer cognitive and language achievement (13).
Clearly, further research is warranted and the clinician should
assess the risk-benefit ratio of continuing SSRIs in pregnancy, consider
the long term harm that may be caused by relapse and promote the
potentially safer modality of psychotherapy.
References:
Levinson-Castiel R, Merlob P, Linder N, et al. Neonatal abstinence
syndrome after in utero exposure to selective serotonin reuptake
inhibitors in term infants. Arch PediatrAdolesc Med. 2006;160:173-176.
Lars Henning Pedersen et al. Selective serotonin reuptake inhibitors
in pregnancy and congenital malformations: population based cohort study.
BMJ 2009; 339: b3569
Chambers CD, Hernandez-Diaz S, Van Marter LJ, et al. Selective
serotonin-reuptake inhibitors and risk of persistent pulmonary
hypertension of the newborn. N Engl J Med. 2006;354:579-587.
Hemels ME, Einarson A, Koren G, et al. Antidepressant use during
pregnancy and the rates of spontaneous abortions: a meta-analysis. Ann
Pharmacother 2005;39:803-9.
Hallberg, P., &Sjoblom, V. (2004). The use of selective serotonin
reuptake inhibitors during pregnancy and breast-feeding: A review and
clinical aspects. Journal of Clinical Psychopharmacology, 25, 59–73.
Gentile S. The safety of newer antidepressants in pregnancy and
breastfeeding. Drug Saf 2005;28(2):137-52.
Simon GE, Cunningham ML, Davis RL: Outcomes of prenatal
antidepressant exposure. Am J Psychiatry 2002; 159:2055–2061.
Oberlander, TF, Warburton, W, Misri, S, et al. Neonatal outcomes
after prenatal exposure to selective serotonin reuptake inhibitor
antidepressants and maternal depression using population-based linked
health data. Arch Gen Psychiatry 2006; 63:898.
Gentile S. SSRIs in pregnancy and lactation: emphasis on
neurodevelopmental outcome. CNS Drugs 2005; 19(7): 623-33.
J. Pediatr. 142[4]:402-08, 2003)
Am. J. Psychiatry 159[11]:1889-95, 2002
Nulman I et al.(1997) NEJM 336:258
Nulman I, et al.. Child development following exposure to tricyclic
antidepressants or fluoxetine throughout fetal life: a prospective,
controlled study. Am J Psychiatry. 2002;159:1889-95.
Competing interests:
None declared
Competing interests: No competing interests
Pedersen et al's paper (1) is a welcome addition to research of this
subject. The importance of such work is confirmed by Durrani and
Cantwell's paper in this month's British Journal of Psychiatry (2) which
found that 43% of patients referred to a community perinatal mental health
team were taking antidepressants at conception, with the most common drugs
taken being SSRIs.
A weakness of the study is that there was no control group of women
with untreated mental illness. Given the population-based design of the
study with the use of national databases and registries, such a control
group should be identifiable. Without such data, it is impossible to be
certain that there is no association between septal heart defects and the
underlying mental illness itself. This study assumes causality of septal
heart defects to citalopram and sertraline but without a true control
group, we can make no such assumption.
Previous research has revealed an association between mitral valve
prolapse and panic disorder (3). It is therefore entirely possible that
other neuroses may be associated with congenital cardiac defects.
References:
(1) Selective serotonin reuptake inhibitors in pregnancy and
congenital malformations: population based cohort study. Lars Henning
Pedersen et al. BMJ 2009;339:b3569
(2) Characteristics of patients seen by a community perinatal mental
health service. Durrani A, Cantwell R. Psychiatric Bulletin 2009;33:368-
370
(3) Prevalence of panic disorder in mitral valve prolapse: a
comparative study with a cardiac control group. Sivaramakrishnan K,
Alexander PJ, Saharsarnamam N. Acta Psychiatrica Scandinavica 1994;89:59-
61
Competing interests:
None declared
Competing interests: No competing interests
Einarson suggests that studies by Malm and Davis were selectively
omitted in the data in the meta-analysis highlighted in my response. This
is not the case. The Forest Plot figures clearly show the studies by Malm
and Davis referred to. The question about why Einarson's groups data was
not included in their meta-analysis should be addressed to GSK. Why this
data was not included in the meta-analysis is puzzling, as it was
available at the time and they have previously provided funding for
Einarsons work. (1)
This meta-analysis helps see the wood among the trees of the single
underpowered studies. The point of the original letter is that this would
appear to be a class effect of SSRI antidepressants. These data need to be
clearly communicated to women using or considering using this class of
drugs in order that they can make informed decisions.
1 Einarson Editors Note on error re disclosures Am J Psychiatry
165;11, November 2008 p1488
Competing interests:
Competing interests declared in previous response
Competing interests: No competing interests
To Dr Mangin,
You appear to be using this forum to state your opinions about
paroxetine in
pregnancy, which is understandable as you are a witness for the plaintiff
in
the GSK/Paxil lawsuiit. However, this study was not about paroxetine per
se,
it was about SSRI's in general and DID NOT find an increase risk for heart
defects associated with the use of paroxetine in pregnancy, which as I
stated
in my previous response was pretty much ignored. I would suggest that if
you
are going to make a statement, you should quote all the evidence-based
research, not only data that supports your agenda. You neglected to
mention
several large studies that have been published recently, that did not
find an
increase risk. My own study with a comparison group which included 1170
women exposed and unexposed found the rates of heart defects the same in
each group.(1) Other groups also did not find an increased risk which also
were not mentioned.(2,3)
Understandably, due to the nature of this research, and the conflicting
results
of the studies, individuals do have strong opinions. However, we must
keep
in mind that the reason this research is conducted is to assist women and
their health care providere in making the right choices for the mother and
unborn fetus and this should be made with evidence-based information, not
opinions.
References
1) Einarson A, Pistelli A, DeSantis M, Malm H, Paulus WE, Panchaud A,
Kennedy D, Einarson TR, Koren G. Evaluation of the risk of congenital
cardiovascular defects associated with use of paroxetine during pregnancy.
Am J Psychiatry. 2008 Jun;165(6):749-52
2) Risks of congenital malformations and perinatal events among infants
exposed to antidepressant medications during pregnancy.Davis RL,
Rubanowice D, McPhillips H, Raebel MA, Andrade SE, Smith D, Yood MU, Platt
R; HMO Research Network Center for Education, Research in Therapeutics.
Pharmacoepidemiol Drug Saf. 2007 Oct;16(10):1086-94
3) Malm H, Klaukka T, Neuvonen PJ. Risks associated with selective
serotonin reuptake inhibitors in pregnancy. Obstet Gynecol 2005; 106:1289-96.
Competing interests:
None declared
Competing interests: No competing interests
We read with interest the article by Pedersen et al (1) published in
the BMJ on selective serotonin reuptake inhibitors (SSRIs) in pregnancy
and congenital malformations. In a study using data from Danish nationwide
registries, the authors followed the standard approach of previous
population-based studies on this topic. They arrived at the same
conclusion of a 2-fold increased risk for septal heart defects after first
trimester exposure to SSRI's (odds ratio 1.99, 95 % confidence interval
1.13 - 3.53). However, contrary to other studies they found an increased
prevalence of heart defects following exposure to citalopram or sertraline
but not to paroxetine or fluoxetine. The greatest risk was observed for
infants exposed in-utero to more than one type of SSRI.
We have recently published our data on first trimester exposure to SSRI
and cardiac malformations (2). Our aim was to compare the rate of non-
syndromic, non-chromosomal congenital heart malformations in SSRI's
exposed newborns to that of unexposed controls. We used a unique screening
approach that included an examination by a pediatric cardiologist and
echocardiography for every newborn with a persistent cardiac murmur (even
mild) on the second or third day of life. This methodology has not been
reported in any of the previous studies on SSRI exposure. Non-syndromic
congenital heart defects were identified by echocardiography in 3.4% of
exposed newborns and in 1.6% in non-exposed babies (relative risk, 2.17,
95% confidence interval, 1.07 – 4.39). All cardiac defects in the study
group were mild: ventricular septal defect, bicuspid aortic valve and
right superior vena cava to coronary sinus. Although our sample size was
too small for analysis of specific drugs, it is interesting to remark that
all four SSRI's (paroxetine, fluoxetine, citalopram and sertraline) were
involved in the exposed newborns with heart defects. On the basis of our
data and clinical experience, it seems appropriate to reassure the women
who require treatment with SSRI's during early pregnancy that the risk is
small and that possible heart malformations are usually mild and often
resolved spontaneously. Early pregnancy monitoring, a late-targeted
ultrasound and fetal echocardiography at 22-23 weeks gestation are
advised. Further larger studies using our unique approach or other
methods are still needed.
References
1.Pedersen LH, Henriksen TB, Vestergaard M, Olsen J, Bech BH.
Selective serotonin reuptake inhibitbors in pregnancy and congenital
malformations: population based cohort study. BMJ 2009;339:b3569.
2.Paul Merlob, Einat Birk, Lea Sirota, Nechama Linder, Michael
Berant, Bracha Stahl, Gil Klinger. Are Selective Serotonin Reuptake
Inhibitors Cardiac Teratogens? Echocardiographic Screening of Newborns
with Persistent Heart Murmur – Birth Defects Res Part A 2009 Aug 18.
[Epub ahead of print]
Competing interests:
None declared
Competing interests: No competing interests
Professor Mangin notes "a key reason for induced abortion is
congenital malformations", but this in the UK at least is a very small
fraction of all terminations performed. Far commoner is that the pregnancy
is not desired, and I would suggest that those women who are depressed are
less likely to wish to face the burden of pregnancy and a child. One can
argue that the likelihood of not being able to continue with SSRI during
pregnancy, or of women being poorly treated on low-doses of SSRIs might
affect the numbers choosing for terminations.
Whilst an abnormality is a disaster for the individuals concerned, we
risk losing sight of the absolute numbers who do NOT get septal heart
defects - 99.5% in those unexposed, 99.1% children of women with one or
more redemption of SSRI and in the case of fluoxetine (0.6% prevalence) a
99.4% chance of no effect. Whilst I think many women will choose to stop
SSRI and see how things fare, at least for the first trimester, we should
be careful not to imply they are always dangerous and can never be taken
at all in pregnancy; the paper concludes number needed to harm for SSRIs
overall is 246, but for fluoxetine this seems more like 1000.
Competing interests:
None declared
Competing interests: No competing interests
After reading the latest study regarding the safety of SSRI's in
pregnancy and
concuring with Dr Chambers editorial, and as we are familiar with the
limitations of these types of studies, we are not going to criticize the
methodology or the results per se. What was interesting to us was how the
results were presented,especially in the abstract.
At The Motherisk Program, a Teratogenic counseling service at The Hospital
for Sick Children, in Toronto, Canada, we answer calls from more than
30,000
women and their health care providers every year about the safety of drugs
in
pregnancy and have published hundreds of papers in this field, including
many on the safety of antidepressants in pregnancy(www.motherisk.org).
Most pregnant women call us because they were advised by their physician,
who often have only read the abstract of papers, and do not know how to
advise their patients. Because of this, it should behoove us to ensure
that
important data is included and it was not in this abstract, most notably
in the
conclusion which was quite misleading. The authors stated that there was
an
increased risk for septal defects "particularly" with sertraline and
citalopram,as if there was also with other SSRI's, which there were not,
and it
should have said "only sertraline and citalopram." In addition, they
should
have stated that there was no association with paroxetine, which is very
important new data, as there have been many warnings about paroxetine and
heart defects and not other antidepressants. Women with depression and
their health care providers,deserve information that is unambiguous
containing all the important evidence-based information, even if it is in
only
250 words.
Thank you
Competing interests:
None declared
Competing interests: No competing interests
The Pedersen article links anti-depressants to birth defects (1).
Anti- depressants are cousins of anti-psychotics. Both types of drug
cause sexual dysfunction, and this may possibly tip over into unfortunate
teratogenic consequences. Many older anti-psychotics will unnaturally
augment prolactin, so as to induce both galactorrhea and amenorrhea.
Breast-feeding can be unsafe if the mother is being given an anti-
psychotic.
Surprisingly, the research into the dangers of anti-psychotics for
pregnancy is not extensive. There is 'very little evidence...regarding the
use of antipsychotics during pregnancy'(2). We have fallen into a murky
complacency over anti- psychotics and teratogenic risk, with large-scale
retrospective studies being lamentably absent.
REFERENCES:
(1) Selective serotonin reuptake inhibitors in pregnancy and
congenital malformations:population based cohort study. Lars Henning
Pedersen et al. BMJ 2009;339:b3569.
(2) Safety of antipsychotic drugs for pregnant and breastfeeding
women with non- affective psychosis. Louise Howard et al. BMJ 2004;329:933
-4.
Competing interests:
None declared
Competing interests: No competing interests
Authors' reply
Various very important points and questions have been raised in
response to our paper; ”selective serotonin reuptake inhibitors in
pregnancy and congenital malformations: population based cohort study”
(1).
Overall, caution is warranted in causal interpretation of
observational studies if the exposure in question is potentially
confounded by unobserved factors. We may have a strong potential
association between the degree of underlying psychiatric disease and use
of antidepressant medication. We have discussed this potential confounding
by indication in the paper but this paramount limitation deserves much
attention. With this in mind we still interpret the difference between the
different SSRIs as suggestive of a drug effect but other explanations
apply, as discussed in the paper, and further studies are needed. No
randomized controlled trials are possible on most types of malformations,
due to ethical and practical considerations, so we are stuck with the
fairly foggy results from epidemiological studies. We present such data
but believe that the results, interpreted with caution in the context of
the existing literature, represent important information in the evidence-
based treatment of depression during pregnancy.
Professor Einarson raises the important point of how data is
presented in the abstract. With English as a second language we are humble
to the fact that our wording could have been more elegant. However, we
still believe the conclusion in the abstract is in line with what our data
show. The estimate of the association between any SSRI and septal heart
defects is not solely based on sertraline and citalopram but also on the
(larger) association with the more than one type of SSRI. In the latter
group some women did not use sertraline or citalopram, hence the wording
“particularly sertraline and citalopram”. We do not present the result on
paroxetine in the abstract, as the OR 0.76 was based on only one case and
we were unable to consider dose as suggested by a previous study (2).
The study is based on information on malformation coded by hospital
physicians and we used data from the first year after birth (with an
additional analyses on malformations coded up to two years after birth).
Professor Hoffman’s questions are essential in the understanding of the
nature of the heart defects. The septal heart defects were defined as
defects of either the atrial or ventricular wall (or both), excluding the
atrioventricular canal defects (3). There was no routine use of
echocardiography during the period and, because the children were born
before the warning on the use of paroxetine, we expect no increased
clinical investigation of the SSRI exposed children. Unfortunately, we
have no further information on the nature of the septal heart defects.
Regardless, even if all the defects were small we believe that the
information is important to consider. Even ASDs with no symptoms for many
years might eventually cause atrial hypertrophy and potentially pulmonary
hypertension later in life. Importantly, more severe septal heart
malformations do mandate surgical treatment but further studies with a
larger sample size are needed to disentangle the potential differences in
severity of the outcome.
References
1. Pedersen LH, Henriksen TB, Vestergaard M, Olsen J, Bech BH.
Selective serotonin reuptake inhibitors in pregnancy and congenital
malformations: population based cohort study. BMJ 2009;339:b3569.
2. Berard A, Ramos E, Rey E, Blais L, St-Andre M, Oraichi D. First
trimester exposure to paroxetine and risk of cardiac malformations in
infants: the importance of dosage. Birth Defects Res.B Dev.Reprod.Toxicol.
2007;80(1):18-27.
3. Louik C, Lin AE, Werler MM, Hernandez-Diaz S, Mitchell AA. First-
trimester use of selective serotonin-reuptake inhibitors and the risk of
birth defects. N.Engl.J Med. 2007;356(26):2675-83.
Competing interests:
None declared
Competing interests: No competing interests