Friday, 21 July 2017

Antidepressants during pregnancy and autism in offspring (with care)

There are a few topic areas in the quite vast autism research landscape that consistently seem to keep cropping up. The possibility of some kind of *association* between pregnancy antidepressant use and risk of offspring autism is one of those areas (see here) as the results published by Dheeraj Rai and colleagues [1] (open-access) are presented for your attention. I would also draw your attention to an accompanying editorial discussing the Rai findings (see here).

So: "To help to improve the understanding of the association between antidepressant use during pregnancy and autism in offspring, we applied a range of... causal analytical methods on data from a large total population cohort in Stockholm County" was the starting point, as once again one of those very useful Scandinavian registries provided the source study material (indeed, Rai et al are seemingly experts in their analysis of such resources). The added bonus to the Rai study was their attempt to 'unravel' any association between gestational antidepressant exposure and autism from the reason why such medication was being taken in the first place: maternal psychiatric health issues (and whether this variable may in fact account, at least in part, for any association that has previously been identified).

From a starting population approaching three-quarters of a million people, researchers eventually settled on looking at over 250,000 children under 17 years of age (but over 4 years of age "in whom a diagnosis of autism might be less reliable") who were born to over 150,000 mothers. The vast majority of children (239,943 of 254,610) had no history of exposure to antidepressants during pregnancy. The remaining participants were divided up into two groups: one where there was documentation leading to the assumption of exposure to pregnancy antidepressants (n=3342) and one where there was an indication for such exposure ("mothers with a psychiatric disorder") but no recorded use of antidepressants during pregnancy (n=12,325). Researchers summed up how many children were diagnosed with an autism spectrum disorder (ASD) in each group and applied some statistical modelling.

Results: I think it's important to first highlight a statistic that seems to have been missed by many covering the Rai findings: "Of the 238 943 cohort children for whom there was no record of maternal history of psychiatric disorder or antidepressant use during pregnancy, 4889 had autism (2.1%)." That's 2.1% with a diagnosis of autism or ASD; quite a far cry from the 1% [estimate] statistic from just a few years back (at least here in Blighty).

Then: "Exposure to antidepressants during pregnancy was associated with a higher odds of a diagnosis of autism in offspring than exposure to a maternal psychiatric disorder without antidepressants." The authors caution that: "the absolute risk was small, and 4.1% of children exposed to antidepressants in utero had autism compared with 2.9% of those with a maternal history of psychiatric disorder." Further when looking at those children diagnosed with ASD in the groups, authors observed that "autism without intellectual disability" seemed to be over-represented; something also picked up in previous findings from authors on this current paper [2].

Alongside various opinions on these findings (see here for example), the authors caution about the possible meaning of their results. One obviously has to be quite careful when discussing such data to ensure that an important class of medicines is not unduly vilified. No medicine is however without potential side-effects and appropriate clinical decisions and good medicines management [2] is key, particularly when pregnancy is included as variable. The authors talk, for example, about how "if a causal link were robustly established, and if no pregnant women took antidepressants during pregnancy, only 2% of autism cases in this population would be prevented." Alongside they [importantly] mention that antidepressant use during pregnancy is not typically just a 'choice' but rather being clinically indicated: depression does not simply disappear when a woman is pregnant. Interestingly too, they mention about how their data "suggest that there is an increased background risk of autism in children of women with psychiatric conditions, regardless of antidepressant treatment." This follows a trend in other areas of psychiatry (see here for example).

Yet again, the call is further research in this area and, quite a few more investigations into the possible hows-and-whys of any association (with medication use and/or maternal psychiatric presentation) is made. I might also suggest that taking into account other childhood conditions such as attention-deficit hyperactivity disorder (ADHD), potentially over-represented when it comes to a diagnosis of autism or ASD, could be another important step forward in light of other preliminary *associations* being made with pregnancy medication history in mind (see here). This also includes looking at any issues associated with timing of potential exposure and/or dose ("Because of small numbers, we were not able to assess trimester specific or dose response effects.").

Insofar as the suggestion that autism without intellectual (learning) disability might be an important phenotype when it comes to any association, subsequent research in this area seemingly fits in well with increasingly vocal calls [4] to stop using the generic label of autism as a research starting point (see here). Allied to suggestions for more 'bottom-up' research with autism in mind (see here), also coincidentally mentioning "maternal SSRI use during pregnancy" in the context of autism [5], a future research agenda is seemingly emerging. I might also point out that certain sensitivities need to be kept in mind on the basis of suggestions that an 'environmental exposure' might, in whole or part, be *associated* with a particular type(s) of autism.

All such work - present and future - however needs to be done/presented with care, understanding and minus scaremongering, so as not to unduly alarm pregnant mothers, their families or indeed, the physicians providing their care at such a critical time...


[1] Rai D. et al. Antidepressants during pregnancy and autism in offspring: population based cohort study. BMJ. 2017; 358: j2811.

[2] Rai D. et al. Parental depression, maternal antidepressant use during pregnancy, and risk of autism spectrum disorders: population based case-control study. BMJ. 2013 Apr 19;346:f2059.

[3] Angelotta C. & Wisner KL. Treating Depression during Pregnancy: Are We Asking the Right Questions? Birth Defects Res. 2017 Jul 17;109(12):879-887.

[4] Waterhouse L. et al. The ASD diagnosis has blocked the discovery of valid biological variation in neurodevelopmental social impairment. Autism Res. 2017 Jul;10(7):1182.

[5] Unwin LM. et al. A "bottom-up" approach to aetiological research in autism spectrum disorders. Front Hum Neurosci. 2013 Sep 19;7:606.


Thursday, 20 July 2017

Is gluten avoidance linked to a lower risk for depression?

I've talked a few times on this blog about how avoiding dietary gluten both within (see here) and outside of (see here) the context of coeliac (celiac) disease, the archetypal 'gluten is baddie' autoimmune condition, might have some pretty interesting effects on some aspects of a person's psychology. Today's post reflects yet more peer-reviewed science suggesting that there may indeed be something to see in this potentially important area; particularly pertinent to the presentation of depression or depressive symptoms.

So, the findings reported by Haley Zylberberg and colleagues [1] based on data from some 22,000 participants taking part in the US 2009-2014 National Health and Nutrition Examination Survey are the source material today. Some background material related to this cohort can be found here. They specifically looked at the "prevalence of depression, insomnia, quality-of-life variables, and psychotropic medication use in CD [coeliac disease] participants and PWAGs [people who avoid gluten] to controls." People who avoid gluten - PWAG - represent a group who don't have a diagnosis of CD but nonetheless similar to those who were diagnosed with CD, reported avoiding dietary gluten.

Results: "Depression was present in 8.2% of controls compared with 3.9% of participants with CD... and 2.9% of PWAGs." Even after adjustment for various confounding variables ("age, sex, race, income, and access to healthcare") those gluten avoiders (without CD) less frequently presented with depression compared with data from controls.

Added to the previous occasions where gluten consumption seems either to be linked to [some] depression or removal of gluten seems to positively impact on depressive symptoms at least for some, this is interesting work. Yes, quite a few more controlled trials are required to examine such relationships between food and mood. Although we can speculate on possible mechanisms [2] we don't really know why there may be an effect from gluten removal, but this is an emerging area of science; particularly in the context of how disruptive/disabling/damaging depression can be to someone.

Bearing in mind the caveats of this blog - no medical or clinical advice is given or intended - please don't assume that I'm advocating gluten removal for anything (unless clinically indicated) on the basis of this or other posts. If in doubt, consult your medical physician.


[1] Zylberberg HM. et al. Depression and insomnia among individuals with celiac disease or on a gluten-free diet in the USA: results from a national survey. Eur J Gastroenterol Hepatol. 2017 Jun 27.

[2] Pruimboom L. & de Punder K. The opioid effects of gluten exorphins: asymptomatic celiac disease. J Health Popul Nutr. 2015 Nov 24;33:24.


Wednesday, 19 July 2017

"Medical history was associated with increased risk of CFS/ME"

The paper by Berit Feiring and colleagues [1] (open-access) is the source blogging material today covering an area that I was previously unaware of: "to study the association between HPV [human papillomavirus] vaccination and risk of CFS/ME [chronic fatigue syndrome/myalgic encephalomyelitis] among girls eligible for HPV vaccination in the national immunisation programme in Norway."

A quick trawl of PubMed to look-see whether there was any previous peer-reviewed science in this area revealed some inklings about a possible *association* between use of the HPV vaccine and some symptoms that may follow a diagnosis of ME/CFS [2] but not much to say that risk of an actual diagnosis of CFS/ME was significantly heightened [3] following HPV immunisation. On the topic of postural tachycardia syndrome (PoTS) and HPV vaccination, I had been aware of some, limited peer-reviewed discussions (see here).

As it happens, Feiring et al did not observe any population-wide links between HPV vaccination and increased risk of CFS/ME based on their analysis of "176,453 girls born 1997–2002 [who] were eligible for HPV vaccination." They tempered their conclusions slightly given that they "did not have access to patient records for verification of the CFS/ME diagnoses" derived from the Norwegian Patient Registry (NPR) but for all intents and purposes, their data is compelling based on sample size and how well Scandinavian countries tend to treat their population data.

A few other, rather interesting aspects to the Feiring data are also apparent: "we observed an increase in the incidence of CFS/ME among adolescents aged 10–17 in Norway, during 2009–2014" and: "Medical history was associated with both increased risk of CFS/ME and lower uptake of HPV vaccine." The idea that numbers of cases of CFS/ME is on the up in Norway is an interesting one and perhaps contrasts with other data from here in Blighty for example (see here) bearing in mind the age groups analysed. I was however specifically drawn to the idea that: "The risk of CFS/ME increased with increasing number of previous hospital contacts."

Looking at hazard ratios (HRs) for a diagnosis of CFS/ME based on how many times a girl had had hospital contact, a rather interesting dose-dependent pattern emerged. So, examining the entire follow-up period, the (adjusted) HR for CFS/ME diagnosis for those with only one hospital contact was 1.64 (1.22–2.19). For those with seven or more hospital contacts, the HR jumped to 5.23 (3.66–7.49). Also potentially important were the possible reason(s) for hospital contact and risk of CFS/ME: "According to ICD-10 diagnoses, the highest risk of CFS/ME was found among girls with diagnoses in chapters I00-I99 (Diseases of the circulatory system)..., A00-B99 (Certain infectious and parasitic diseases)... and R00-R99 (Symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified)."

Other discussions on data also from Norway (indeed, from some of the same authors) has talked about medical history around CFS/ME and what conditions seem to be 'over-represented' (see here). On that research occasion [3] they concluded that: "Children with CFS/ME were frequently diagnosed with infections, supporting the hypothesis that infections may be involved in the causal pathway." Said previous findings are not so dissimilar from the current ones reported by Feiring and colleagues.

Insofar as the next steps on from the Feiring data, well one might - assuming independent replication and the like - see a way forward where a potential early warning system for possible CFS/ME screening might be put in place on the basis of patterns of hospital visits and the types/patterns of diagnosis that are given. I appreciate that there is quite a lot of heterogeneity when it comes to CFS/ME (possibly even another label ripe for pluralisation or spectrum-ing?) so I don't want to get too ahead of myself here but there is potential. Whilst HPV vaccination did not seem to be connected to later CFS/ME diagnosis in the Feiring cohort, the question about what factors could be driving the increase in cases remains open: "The reasons for the increase in CFS/ME in Norway are unknown."


[1] Feiring B. et al. HPV vaccination and risk of chronic fatigue syndrome/myalgic encephalomyelitis: A nationwide register-based study from Norway. Vaccine. 2017. June 23.

[2] Brinth LS. et al. Orthostatic intolerance and postural tachycardia syndrome as suspected adverse effects of vaccination against human papilloma virus. Vaccine. 2015 May 21;33(22):2602-5.

[3] Donegan K. et al. Bivalent human papillomavirus vaccine and the risk of fatigue syndromes in girls in the UK. Vaccine. 2013 Oct 9;31(43):4961-7.

[4] Bakken IJ. et al. Comorbidities treated in primary care in children with chronic fatigue syndrome / myalgic encephalomyelitis: A nationwide registry linkage study from Norway. BMC Family Practice. 2016; 17: 128.


Tuesday, 18 July 2017

Anxiety disorder is rife in 'high-functioning' autism

"Lifetime prevalence rates of 53.5% for depressive disorder 73.5% for anxiety disorders and 37.5% for ADHD [attention-deficit hyperactivity disorder] were found."

Those were the figures arrived at by Alexandru Gaman and colleagues [1] who set about investigating the "prevalence rates of psychiatric co-morbidities" among other things in a cohort of over a hundred adults diagnosed with "high-functioning" autism via the quite recently revised DSM-5 criteria. I've stressed the words 'high-functioning' to denote this being the authors' words not mine (personally, I'm not so sure that general level of functioning is all that good as a descriptor).

Various other observations were made by authors such as the finding that: "Subjects with psychotic co-morbid symptoms had a more severe social deficit" which might tap into some other discussions being had on how some of the screening instruments talked about with autism in mind are seemingly not adverse from potentially picking up other labels with a psychosis element to them (see here). I say that also with the understanding that at least for some, autism and psychosis are not diagnostically unstrange bedfellows (see here).

I've zoomed in on the anxiety disorder(s) bit to the Gaman findings because of their very high lifetime prevalence and because, day-to-day, anxiety disorders can be absolutely disabling for many people on the autism spectrum (see here). Indeed, with all the very positive talk about things like employment and further education opportunities [slowly] increasing for autistic young people and adults, one of the details that does not seem to be talked about as much is how issues like anxiety can significantly hinder not only efforts to get a job/student place but also keeping that job/student place in the longer term (see here). Talent is being outshone by crushing anxiety in some cases.

Gaman and colleagues concluded by talking about how identification of something like anxiety disorder is "a crucial clinical issue." I would very definitely agree with this viewpoint but more than that, efforts now need to go into what can be done about treating/managing such anxiety to make people's lives easier (see here); accepting that we still have some distance to go in this process [2]. I'd also like to see some kind of research parity being arrived at specifically with regards to the question: how prevalent and what effects does anxiety have for those NOT described as having 'high-functioning' autism?

To close, having recently been party to some interesting debate on social media about the ins-and-outs, rights-and-wrongs and positives-and-negatives of [exclusive] self-diagnosis with autism in mind, I'd like to link to a paper by Ashwood and colleagues [3] on how one of the premier 'are you autistic?' self-report schedules is not necessarily fit for purpose when it comes to a self-diagnosis of autism. Indeed pertinent to today's post, how "generalized anxiety disorder may ‘mimic’ ASD [autism spectrum disorder] and inflate AQ [Autism-Spectrum Quotient] scores, leading to false positives" echos a viewpoint that I championed: identity, emotions and politics aside, there is no substitute for a thorough professional assessment when autism is suspected. Outside of such an assessment being potentially pertinent to the idea that autism rarely appears in some sort of diagnostic vacuum (see here), it is perhaps even more important as the DSM-5 criteria for ASD and SCD [social (pragmatic) communication disorder] start to become even more mainstream and what it means/will mean to the concept of autistic identity too...


[1] Gaman A. et al. Psychiatric co-morbidities in a French cohort of adults with high-functioning autism (HFA). European Psychiatry. 2017; 41: S136.

[2] Lorenc T. et al. Support for adults with autism spectrum disorder without intellectual impairment: Systematic review. Autism. 2017 Jun 1:1362361317698939.

[3] Ashwood KL. et al. Predicting the diagnosis of autism in adults using the Autism-Spectrum Quotient (AQ) questionnaire. Psychological Medicine. 2016;46(12):2595-2604.


Monday, 17 July 2017

Second seizure risk and "idiopathic autism"

Idiopathic autism refers to instances where autism is the primary diagnosis and not something secondary to another - usually genetic - condition. The paper by Asad Qadir and colleagues [1] reviewed the files of some 150 people diagnosed with an autism spectrum disorder (ASD), idiopathic ASD, and a history of at least one seizure in connection to an important issue: recurrence risk of seizure. They concluded that the average age of first seizure in their cohort was around about 7 years old and many, over 90% of participants included for review, were at serious risk of a second seizure on average just over 6 months later.

This is an important finding. We already know that epilepsy and seizure disorder(s) are not uncommon bedfellows to parts of the autism spectrum (see here) (indeed, many parts of the autism spectrum seem to be prone to unusual EEG - electroencephalographic or electroencephalogram - findings). We know that non-febrile seizures (seizures not attached to fever) are quite a bit more common in relation to autism when compared with non-autistic siblings for example (see here). And on top of all that, we know that seizures can in some instances be life-threatening (see here).

The Qadir data highlights the "short time to second seizure" as a window to appropriate management of seizure when coincidental to autism. The data suggest that even after one seemingly isolated incidence of seizure, clinicians might consider being proactive in (i) screening for signs of EEG anomalies for example, and (b) be pretty assured that initiation of appropriate treatment/management of seizure is very likely applicable insofar as the risk of recurrence in those diagnosed with an ASD. I say all that with my blogging caveat of no medical or clinical advice given or intended, in full working order.

I'd also like to think that as our knowledge about the autism spectrum increases, in particular the idea that there may be many different routes to many different types of autism (see here) so science will start to put some further flesh on the bones that what we call idiopathic autism with epilepsy at the moment, does not necessarily mean things will stay 'idiopathic' in future times. Then, other questions need answering, such as whether certain 'epileptic encephalopathy syndromes' might actually be the cause of some autism [2]...

As I've said many times before on this blog, don't mess with seizures and/or epilepsy...


[1] Qadir AA. et al. Risk of Second Seizure in Pediatric Patients With Idiopathic Autism. J Child Neurol. 2017 Jan 1:883073817713906.

[2] Srivastava S. & Sahin M. Autism spectrum disorder and epileptic encephalopathy: common causes, many questions. J Neurodev Disord. 2017 Jun 23;9:23.


Saturday, 15 July 2017

Should anyone be concerned about pathological video game use in relation to autism?

"... the risk for pathological game use appears larger in adults with ASD [autism spectrum disorder] compared with TD [typically developing] adults. These findings point to pathological game use as a potentially important focus of clinical attention in adults with ASD."

So concluded the paper by Christopher Engelhardt and colleagues [1] (open-access) looking at an interesting issue - "whether adults with autism spectrum disorder (ASD) are at higher risk for pathological game use than typically developing (TD) adults" - based on reports from some 120 participants with and without a diagnosis of autism. Including one Micah Mazurek on the authorship list (who has previously talked about the effects of screen time and autism), researchers delved into self-reported time spent playing video games based on data collected from a cohort "participating in a larger study on violent video game effects" [2]. The average age of participants (with and without a diagnosis of autism) was round about the 20 years old mark.

Results: both on weekdays and weekends, the typical hours of game play were different between the groups of those with autism and those without. Interestingly, authors used a slightly different statistical approach to the traditional "null hypothesis significance tests and 95% confidence intervals" via their use of "Bayes factors to state evidence for or against our predictions" to analyse their data. Bayes and Bayesian teachings provide an alternative to the 'frequentist' considerations in treating data. Alongside asking about game play time, researchers also asked participants to complete a questionnaire on "pathological video game use" that was based on "another behavioral addiction (pathological gambling)" and covered various aspects: "salience, euphoria or relief, tolerance, withdrawal, conflict, relapse and reinstatement."

"The results indicated that adults with ASD endorsed more symptoms of video game pathology than did TD adults. This relationship was strong, enjoying 300,000-to-1 odds in Bayesian model comparison." Such sentences kinda say it all. Risk of pathological video game use was quite a bit more likely in those with autism than those without, and 'escapism' was also reported to be an important reason for such extended use.

I do want to be a little careful in this post not to (a) demonise - pathologise - video game playing and/or (b) isolate those on the autism spectrum as being 'unique' in their video gaming habits. Personally, I don't see an issue if people want to spend large sections of their days playing video games as opposed to watching television or sitting down reading a book (reading in particular, is another fine example of escapism) or doing anything else. This was a study of adults doing what many millions of people do every day, so they should have the right to do as they please in this context. With the wonders of modern technology, gaming today is also not necessarily a socially-isolated pastime so one has to be quite careful about making any sweeping generalisations in that respect too.

I have however, always been a little worried about this and other research on screen time use in the context of autism and what detrimental effects it might have for a person's physical health. Everyone is prone to a degree of 'couch-potatoing' (i.e. loafing of the sofa for hours on end) but when people are talking about nearly 3 hours every day gaming, presumably sat on a couch/sofa, this really can't be doing anyone any good in terms of their long-term physical health. We're all being told to move more (see here) and given the data emerging when it comes to the physical activity patterns of rather a lot of people on the autism spectrum (see here), there are some good grounds for encouraging more movement and more physical activity minus any nanny-stating. I appreciate that sports and exercise is not everyone's cup of tea and in particular, more needs to be done in making physical activity more attractive to many groups of people including those on the autism spectrum (see here). Efforts should continue to ensure a balance is struck between times of sedentary behaviours typically associated with gaming and times of engaging in physical activity; as far as I can see, Engelhardt et al did not ask about physical activities or other related pastimes on this research occasion. Indeed, although I am still a little apprehensive about the use of exergaming, I do wonder if this could be a good bridging point on the road to balance...

Music to close: Shape of You by Ed for one of my brood who seems to be looking forward to becoming a kumite specialist (when she graduates from Ippon kumite).


[1] Engelhardt CR. et al. Pathological game use in adults with and without Autism Spectrum Disorder. PeerJ. 2017. e3393.

[2] Engelhardt CR. et al. Effects of Violent-Video-Game Exposure on Aggressive Behavior, Aggressive-Thought Accessibility, and Aggressive Affect Among Adults With and Without Autism Spectrum Disorder. Psychol Sci. 2015 Aug;26(8):1187-200.


Friday, 14 July 2017

Antidepressant use in pregnancy and risk of offspring ADHD?

Here we go again. Antidepressant use during pregnancy comes under the spotlight yet again (see here) as per the findings reported by Takoua Boukhris and colleagues [1], this time in connection to risk of offspring diagnosis of attention-deficit hyperactivity disorder (ADHD) following gestational exposure to this class of medicines.

Drawing on data derived from the Quebec Pregnancy/Children Cohort (QPC), researchers set out to examine the "risk of ADHD associated with overall and class-specific antidepressant exposure in utero." I might add that previous data from some of the authors of the Boukhris paper on the QPC has suggested that it is "an excellent tool for the study of the risk and benefit of drug use during the perinatal period" [2].

Following nearly 150,000 live births between 1998 and 2009, researchers looked at the rates of subsequent diagnosis of ADHD alongside the pattern of medication use during pregnancy and other important variables such as "maternal history of depression/anxiety and ADHD." They concluded that: "AD [antidepressant] use during the 2nd and 3rd trimester of pregnancy, specifically tricyclics, is an independent risk factor for ADHD in children above and beyond the risk associated with maternal depression/anxiety or ADHD." Further: "SSRI [Selective serotonin reuptake inhibitors] and SNRI [Serotonin–norepinephrine reuptake inhibitors] use were not associated with increased ADHD risk."

As per the convoluted research history talking about pregnancy antidepressant 'exposure' and possible risk of offspring autism, there are certain caveats and issues worth mentioning. Antidepressant use during pregnancy is not something to be taken lightly as per any calls to 'restrict use' during the special nine months that make us. Such medicines serve an important purpose in controlling some potentially important clinical symptoms; said symptoms do not just dissipate during the special time called pregnancy as per other clinical examples with other labels (see here).  It's also worth pointing out that such observational studies looking at medication use and offspring outcomes are just that - observational. As per the tenet 'correlation is not the same as causation', one has to be a little careful with such data...

That all being said, I do find the Boukhris results to be intriguing on the basis that this is not the first time that an *association* has been talked about in the peer-reviewed science domain [3] and on more than one occasion [4]. Further investigations are perhaps required into the possible mechanism(s) involved in any such relationship between medicine and offspring development, and indeed, whether such connections could potentially invite new intervention options for labels like ADHD? I'm also more than a little interested in the 'medication type' angle to the Boukhris observations not necessarily being the same as that put forward in relation to offspring risk of autism.

And just before you go, it's worth highlighting the findings reported by Viktorin and colleagues [5] looking at pregnancy antidepressant use and intellectual (learning) disability: "After adjustment for confounding factors... the current study did not find evidence of an association between ID and maternal antidepressant medication use during pregnancy."


[1] Boukhris T. et al. Antidepressant Use in Pregnancy and the Risk of Attention Deficit with or without Hyperactivity Disorder in Children. Paediatr Perinat Epidemiol. 2017 Jun 22.

[2] Bérard A. & Sheehy O. The Quebec Pregnancy Cohort – Prevalence of Medication Use during Gestation and Pregnancy Outcomes. Croy A, ed. PLoS ONE. 2014;9(4):e93870.

[3] Man KKC. et al. Prenatal antidepressant use and risk of attention-deficit/hyperactivity disorder in offspring: population based cohort study. BMJ. 2017 May 31;357:j2350.

[4] Clements CC. et al. Prenatal antidepressant exposure is associated with risk for attention-deficit hyperactivity disorder but not autism spectrum disorder in a large health system. Mol Psychiatry. 2015 Jun;20(6):727-34.

[5] Viktorin A. et al. Association of Antidepressant Medication Use During Pregnancy With Intellectual Disability in Offspring. JAMA Psychiatry. 2017 Jul 12.