![]() |
|||
« Mini-Double Dose: Art, Pain and Illness; Genes and Race Disparity; Through Sickness, Health and Sex Change | Main | Forget Love. What Would You Do for Health Insurance? » April 30, 2008New Study Examines Depression in Women, Including Effects of Abortion and Partner ViolenceA new study in BMJ Public Health examines depression in women and the relationship of past abortions to the condition. This new report is particularly interesting because it attempts to control for the effects of sociodemographic factors and considers the women's experiences of intimate partner violence, recognizing that multiple factors may impact a woman's mental health. Although the study focuses on Australian women, it may be of interest to readers in other countries as well due to recent attempts by anti-choice groups to promote the existence of a so-called "post-abortion syndrome," or causal link between abortion and depression, which has thus far been unsupported by the medical evidence. Similar to unsupported and debunked claims of an abortion/breast cancer link, this tactic frames the pro-choice position as anti-women's health, despite the lack of evidence to support that framing. (For background reading, try this commentary in Ms. Magazine and a lengthy discussion of the issue in the New York Times.) In the new study, researchers surveyed Australian women ages 18-23, collecting data on sociodemographic variables, reproductive history, and intimate partner violence. The women also completed depression scales and questions about past diagnosis of depression, with a follow-up survey conducted four years later. Before controlling for other variables, a previous pregnancy termination or having two or more children was associated with an increased risk of depression. However, when the researchers controlled for those other factors, there was no longer an association between abortion and depression, and the increased odds of depression for those with two or more children dropped to be almost insignificant. Further, the authors found that, even after controlling for sociodemographic factors, experiencing intimate partner violence had a significant effect on the odds of depression, suggesting that is violence a more important factor than either abortion or childbearing with regards to women's mental health. Meanwhile, anti-choice advocate Michaelene Fredenburg has started the "Abortion Changes You" website, inviting people to submit their stories of regret and shame. ImNotSorry.net takes a different approach, offering an alternative story-sharing site for women who need a space to say that they are not ashamed of their choices. |
Welcome to Our Bodies, Our Blog, your daily dose of women's health news and analysis. Send tips to: BLOG SEARCH
Subscribe to this blog's feed [What is this?]
BLOGS WE LIKE
|
| Copyright © 2006, Boston Women's Health Book Collective, Inc. |
Comments
It seems pretty obvious that violence toward a woman (or a man for that matter) could easily aid in the formation of a depressed state.
I'm not sure I buy the "science" related to the de-correlation of a previous abortion or parenting multiple children when it comes to depression. By removing the sociodemographic factors, the researchers are no longer working in reality, they are working in a sterile scientific fantasy world. The fact is we are all somewhat biased and bound by our sociodemographic make-up. If a person is depressed because she happened to grow up in a strict Catholic environment and then had an abortion, you can't just discount her feelings because if she would have been raised agnostic that "guilt" may not be there.
Posted by: mpro | April 30, 2008 12:21 PM
mpro, thanks for commenting. This study isn't "de-correlating" as you suggest - the existing evidence on abortion and depression also suggests that there is no significant correlation. In other words, this study basically replicates what has already been demonstrated by other researchers.
This type of multiple logistic regression is actually an extremely common technique for allowing researchers to sort out the different effects of multiple independent variables. Say we want to know if abortion "causes" depression. In your example it's the Catholic upbringing wrt abortion that is associated with the depression, not a physiological effect of abortion. That is not to say that an individual woman may not experience some regret or depression (which may be related to other issues, such as violence), but it's not a biological given that abortion causes depression, and the researchers found that having had an abortion did not predict or raise the odds of depression. It's a bit complicated, but there's a fairly straightforward example in the first paragraph here: http://www.statsoft.com/textbook/stmulreg.html
Posted by: Rachel | April 30, 2008 12:45 PM
Thanks for posting this.
We just learned about confounders and effect modifiers and stratified analysis in biostats. Yeesh.
Posted by: Hilary | April 30, 2008 05:15 PM