Studies claiming to find a relationship between abortion and subsequent mental health problems often suffer from serious methodological limitations that invalidate their conclusions. In a new analysis, Julia Steinberg, of the University of California, San Francisco (UCSF) and Lawrence Finer, of the Guttmacher Institute show that the findings of a 2009 study by Priscilla Coleman et al—which claimed that women who had reported an abortion were at an increased risk of several anxiety, mood and substance use disorders, compared with women who had never had an abortion—are not replicable.

Steinberg and Finer’s analysis, just published online in Social Science & Medicine, examined the same dataset as Coleman et al. (the National Comorbidity Survey) and found that in every case, the proportions of women experiencing mental health problems reported by Coleman were much larger, sometimes more than five times as large, as Steinberg and Finer’s results. The Coleman findings were also inconsistent with several other published studies using the same dataset and sample.

“We were unable to reproduce the most basic tabulations of Coleman and colleagues,” says Steinberg, postdoctoral fellow at UCSF. “Moreover, their findings were logically inconsistent with other published research—for example, they found higher rates of depression in the last month than other studies found during respondents’ entire lifetimes. This suggests that their results are substantially inflated.”

“Antiabortion activists have relied on questionable science in their efforts to push inclusion of the concept of ‘post-abortion syndrome’ in both clinical practice and law,” says Finer, director of domestic research at the Guttmacher Institute. “Our inability to replicate the findings of the Coleman study makes it clear that research claiming to find relationships between abortion and poor mental health indicators should be subjected to close scrutiny.”

Steinberg and Finer also examined other well-established risk factors for post pregnancy mental health problems, such as preexisting mental health disorders and sexual or physical violence before the abortion, and found that women who had had multiple abortions were more likely to have experienced these risk factors prior to the abortion than women who had had one or no abortions. Once they controlled for these factors, they found no significant relationship between abortion history and subsequent mood or anxiety disorders. These findings support the view that previous mental health status, and not abortion experience per se, is the strongest predictor of post-abortion mental health.