I'm taking antidepressants while pregnant. Does that mean my child will develop autism?

Recently, an article linking antidepressant use in pregnancy with the risk of autism spectrum disorder in children was published in JAMA Pediatrics. What followed was a flurry of news reports like this and this, claiming that women who take SSRIs in pregnancy increase their child's risk of autism "by as much as 87%" (according to the Newsweek article). Scary news. In our clinic, many women who are either pregnant or preparing for pregnancy have asked us about whether taking antidepressant medications during pregnancy will increase the risk of autism in their children exposed in utero, so we think it's an important topic to cover.

ASD is a developmental disorder characterized by impaired communication and social interactions. More information can be found here. Right now, there is no known single cause, but we do know that risk factors for ASD include particular genetic variants, de novo mutations, maternal disease (like diabetes), and--importantly--maternal history of psychiatric disorders.   

Here's a brief summary of the JAMA article: it was a study that looked at all pregnancies that occurred in Quebec between 1/1/98 and 12/31/09. Data from from several databases were linked with unique personal identifiers to provide information about medical services, prescription drugs, hospitalization, and demographics. The researchers defined antidepressant (AD) exposure as having at least 1 prescription filled at any time during pregnancy or a prescription filled before pregnancy that overlapped with the first day of gestation. The outcome variable was whether or not children had a medical service claim or hospitalization with a diagnosis of ASD between birth and the end of follow-up. The researchers determined crude and adjusted hazard ratios (HRs) which are the expression of the hazard, or chance, of events occurring in the treatment arm as a ratio of the hazard, or chance, of the events occurring in the control arm. Basically, the HR could be expressed as follows:

The researchers found that the hazard ratio for the use of antidepressants during the second and/ or third trimester was 1.87, which is where various news sites got the 87% increased risk of ASD statistic. On the surface, this HR indicates that there is a higher chance of children being diagnosed with ASD in women who used ADs in the second and/ or third trimester than the chance of children being diagnosed with ASD in women who did not. The study authors suggest that this means that the use of antidepressants is associated with an 87% increased risk of ASD, even after taking into account particular confounders. (Of note, the hazard ratio for the use of ADs in the first trimester was 0.84.) However, when we look more closely at the study, we see a few problems with this conclusion: 

1.  The biggest problem is that maternal psychiatric illness, in particular, depression, is known to be associated with an increased risk of ASD in offspring. The study had no reliable measures of depression severity. Therefore, there is no way to tell whether the children were at higher risk for developing ASD because their mothers were taking antidepressants, or because their mothers had more severe depression. In Alison Stuebe's fantastic article on this topic in the Huffington Post, she writes the following: "The key problem is that women who take a medication when they are pregnant have a reason for taking it. Blaming the outcome on the medication, without considering the underlying disease, is like saying that umbrellas cause flooding. Taking away the umbrellas -- "not treating the rain" -- does not prevent flooding; it just means that people get soaked."
   
2.  Another problem with the author's conclusion about an 87% increased risk of autism is that the number is very misleading. If the average child has a 1.5% risk of autism (according to the CDC, the prevalence was 1.5% in 2010), then the children exposed to antidepressants in the second and/ or third trimester in this study had a risk of 2.8%, which is an absolute risk increase of 1.3%, a much more modest value than 87%. 
3. A sensitivity analysis that restricted the study population to children with ASD diagnosis confirmed by a psychiatrist of a neurologist found that the association between the use of antidepressants in the second and or third trimester and risk of ASD was no longer statistically significant (the confidence interval included 1). This means that it's important to realize that the study did not show an increased risk of ASD diagnosis in children of mothers who took antidepressants, but rather, an increased risk of a billing code (which could just indicate an evaluation) for ASD. 

Whether or not to continue taking antidepressants during pregnancy is a very personal decision, and best discussed with your clinician. It's important to remember that there are significant risks of untreated depression or anxiety in pregnancy, including prematurity, intrauterine growth restriction, and low birth weight, as well as higher rates of impulsivity, and cognitive, emotional, and behavioral problems in children. Mothers with untreated depression in pregnancy have significantly higher rates of relapse of their depression if they do not take their medications (68% compared to 26% in women who did continue taking medications during pregnancy), according to a large NIH study looking at mothers with severe depression. To put this in context, if 1000 women with severe depression stop taking their medications during pregnancy, then 680 of them would experience a relapse in major depression, while only 13 of the women would avoid an evaluation of their child for autism.