Autism-antidepressant Link Questioned

Study ties autism risk to mother’s illness, not prenatal medication use

Previous studies have suggested an increased risk of autism among children of women who took antidepressants during pregnancy. New research questions that conclusion, saying the link may actually reflect the known increased risk associated with severe maternal depression. Image: iStock

In a study published in Molecular Psychiatry, Harvard Medical School investigators at Massachusetts General Hospital report that while a diagnosis of autism spectrum disorder was more common in the children of mothers prescribed antidepressants during pregnancy compared with those with no prenatal exposure, when the severity of the mother’s depression was accounted for, that increased risk was no longer statistically significant.

An increased risk for attention-deficit hyperactivity disorder (ADHD), however, persisted even after controlling for factors related to a mother’s mental health. 



“We know that untreated depression can pose serious health risks to both mother and child, so it’s important that women being treated with antidepressants who become pregnant, or who are thinking about becoming pregnant, know that these medications will not increase their child’s risk of autism,” said Roy Perlis, HMS associate professor of psychiatry at Mass General and senior author of the report. 


Genetic factors are known to play a substantial role in autism, but exactly how that risk may be exacerbated by environmental factors is not well understood. Some animal studies and investigations based on patient health records have suggested an increased risk associated with prenatal antidepressant exposure, but others found no such association.

Because discontinuing antidepressant treatment significantly increases the risk of relapse—including an increased risk of postpartum depression—the current study was designed to clarify whether any increased autism risk could be attributed to the medication. 



To investigate this possibility, the research team analyzed electronic health record data for children born at Mass General, Brigham and Women’s Hospital, or Newton Wellesley Hospital for whom a diagnostic code for pervasive developmental disorder, a category that includes autism, was entered at least once between 1997 and 2010. They matched data for these almost 1,400 children with more than 4,000 children with no autism diagnoses, born in the same years and matched for a variety of demographic factors. 



The children’s information was paired with their mothers’, noting any factors related to the diagnosis and treatment of major depression or other mental illness, including prescriptions for antidepressants and other psychotropic drugs. A similar analysis was done for almost 2,250 children with an ADHD diagnosis, compared with more than 5,600 matched controls with no ADHD diagnoses. 


Prenatal exposure to antidepressants did increase the risk for either condition, but in the autism-focused comparison, adjusting for factors indicating more severe maternal depression reduced the strength of that association to an insignificant level. Taking antidepressants with stronger action in the serotonin pathway, which has been suspected of contributing to a possible autism risk, did not increase the incidence of the disorder. In addition, the children of mothers who took a serotonin-targeting nonantidepressant drug for severe morning sickness had no increased autism incidence.

Prescriptions for antipsychotic drugs sometimes used to treat severe, treatment-resistant depression, as well as psychotic disorders, did appear to increase the risk for autism. For ADHD, however, the increased risk associated with prenatal antidepressant exposure remained significant, although reduced, even after adjustment for the severity of maternal depression. 



“There are a range of options—medication and nonmedication–for treating depression and anxiety in pregnancy,” said Perlis. “But if antidepressants are needed, I hope parents can feel reassured about their safety.”



Support for the study includes National Institute of Mental Health grant R01MH086026 and support from the Stanley Center for Psychiatric Research.

Adapted from a Mass General news release.