A large-scale study looking at the much-discussed relationship between autism and antidepressant use in pregnancy has found no evidence of a link between the two.
Researchers from the Harvard T.H. Chan School of Public Health and Sweden’s Karolinska Institute looked at nearly 1.6 million babies born to 950,000 women in Sweden from 1996 to 2012 and found no increase in the risk of autism or ADHD among those born to women taking antidepressant medication.
It’s a significant contrast to the last well-publicized study on the subject, a 2015 finding that babies exposed to selective serotonin reuptake inhibitors—SSRIs, a common class of antidepressants—in the second and third trimesters were 87% more likely to develop autism than children whose mothers didn’t take those drugs. That study raised more questions than it answered, particularly whether the increased autism rates were related to the drugs or to the depression itself. But the widespread media attention it garnered was enough to scare a lot of women away from medicine that could have improved their health—and consequently, the health of their babies.
Untreated depression in pregnancy is associated with premature birth and lower birth weight for infants, and greater risk of postpartum depression and self-harm for mothers. So the new research is welcome news for women weighing the pros and cons of continuing antidepressants in pregnancy. It is also a frustrating reminder of how confusing and ill-informed the advice given to women on this subject is, at the expense of their health and their children’s.
Good data on antidepressants and pregnancy is scarce for the same reason that good data on the interplay between most drugs and pregnancy is scarce. Women generally are underrepresented in clinical trials. Researchers and drugmakers are especially reluctant to include pregnant women in trials, despite the American Congress of Obstetricians and Gynecologists (ACOG) saying more research on expecting women is needed and pregnancy should not automatically exclude participants.
The research that does exist has found that antidepressants generally have no discernible effect on the vast majority of babies exposed to them in utero. However, individual studies have found links between certain drugs taken at specific times during the pregnancy and problems for babies.
For example, first-trimester exposure to paroxetine, an SSRI marketed as Paxil, has been linked to increased risk of a heart defect known as a right ventricular outflow tract obstruction. Ten to 30% of babies exposed to SSRIs in the third trimester show temporary signs at birth of withdrawal symptoms like rapid breathing, increased crying, and feeding disturbances, issues that typically resolve within the first two weeks. Babies exposed to antidepressants of all categories in utero are also slightly more likely to develop persistent pulmonary hypertension, a serious but treatable breathing disorder.
As the US Centers for Disease Control and Prevention point out, the incidence of these problems are very small. A baby not exposed to antidepressants at all has a 1 in 1,000 chance of being born with that right ventricular heart defect. If the baby is exposed to Paxil early in the pregnancy, that number jumps—to 2.4 in 1,000. The absolute risk is still very low. Yet to a nervous expectant parent, any known increase in risk can feel like too much.
In the absence of total certainty—a luxury rare as a baby unicorn in medicine—ACOG’s official guidelines on treating most cases of depression in pregnancy boil down to this: weigh the risk to the baby of antidepressant exposure against the risk to the mother of untreated depression.
Evaluating the pros and cons of a decision against the available evidence and your own priorities is the best method we have for tough medical decisions. But it’s also a method easily skewed by the biases of doctors, patients, and their families. Doctors are not immune to the prejudices many of us hold against people with mental health issues, and people with active depression can struggle to accurately assess their own health needs and prioritize them accordingly. That pairing produces less than ideal outcomes.
How does this imperfect process affect women’s health? There are millions of stories; I can speak only to the one I know best. During my first pregnancy in 2010, the obstetrician instructed me to wean myself off the SSRI sertraline by the third trimester to minimize the risk of withdrawal symptoms for the baby. When my primary-care doctor learned late into the pregnancy that I was taking an antidepressant, even at a low and diminishing dose, he called both my obstetrician and me to berate us for threatening my baby’s health. By the time I gave birth I was off the drug completely, a decision that felt less like a clinical imperative than a moral one. I had a healthy baby, but the depression relapsed, as it does for 60% to 70% of women who stop treating the condition during pregnancy.
By my second pregnancy, in 2016, I had read the 2015 study on autism. Though it was far from conclusive, I was determined to minimize any risk. I informed my new obstetrician that I would again wean myself off antidepressants during the pregnancy and she didn’t challenge me.
I felt worse as the months wore on, but depression warps the ability to make good decisions about one’s self-care and I didn’t speak up. By the time I gave birth, the depression was serious enough that the hospital installed a 24-hour observer in my room to ensure I didn’t harm myself and refused to release me before an evaluation from a psychiatric consultant.
Weeks after the birth, when I was back on medication and well enough to be stunned by how swiftly I’d fallen ill, a psychiatrist looked up from my chart upon our first meeting and asked, “Why on Earth did you stop taking your medication?” All I could tell him was that I believed, at the time, that I was doing the right thing for my child.
Better data leads to better decisions. It would have been reassuring to know that a study with as wide a scope as the Harvard-Karolinska research found no evidence linking autism and antidepressants. Lacking that, it would have been helpful to have a doctor put the best evidence available into perspective. I understood that taking the drugs gave me a 1.5% risk of having a neurodivergent child, but not that stopping them gave my child a two in three chance of having a mom with active depression in the first weeks of his life. This latest study is not the last word on the intersection of depression and pregnancy, but it’s a vital decision-making tool for those who find themselves there.