Postpartum Mental Health Study Flawed by Fetal Loss Omission
2/5/16
A series of studies on post-partum mental health from Denmark have demonstrated diligent efforts to conceal the negative effects of abortion on women’s mental health, according to an article published in the Scandinavian Journal of Primary Health Care.
The research team, led by Trine Munk-Olsen, specializes in analyses of Denmark’s national medical records, especially pertaining to mental health treatments associated with women’s reproductive health. Notably, of over two dozen such studies, in only two does Munk-Olsen report on the mental health treatments associated with induced abortion.
Munk-Olsen’s most recent study found that women receiving psychiatric care after giving birth were more likely to have sought treatment from a primary care provider before, during and after pregnancy. Based on this finding, Munk-Olsen asserts that childbirth is a psychological stressor that can be “a strong trigger of psychiatric episodes.”
Notably, this is a frequent assertion made in Munk-Olsen’s studies: childbirth triggers mental health problems. On the other hand, in the only two studies she has published reporting on mental health effects associated with abortion, she has argued that abortion has no impact on mental health.
Omitted Variables Produce Predictable Distortions
In a criticism of Munk-Olsen’s latest study published in the Scandinavian Journal of Primary Health Care, David Reardon, director of the Elliot Institute and one of the most published researchers on abortion and mental health issues, questions the validity of Munk-Olsen’s conclusions. He is especially critical of the manner in which she modifies her research methods to produce results to avoid reporting on the negative effects of abortion.
Reardon points out, for example, that in Munk-Olsen’s most recent study examining medical records of women who had postpartum psychiatric treatments, she chose to ignore the effects of prior pregnancy losses (abortions and miscarriages) on treatment rates. At the very least, this was a serious oversight since previous research has shown that women exposed to prior pregnancy losses are at higher risk of post-partum depression and other mental health issues.
Reardon asserts that Munk-Olsen was clearly aware of this connection based on her own prior research, much less that of others in the field. By ignoring prior pregnancy losses in her most recent study, Munk-Olsen builds a platform to argue that the differences in primary care treatments among women with mental health issues following delivery are due to childbirth alone, when in fact a more detailed analysis, as demonstrated by other researchers, is likely to show that the heightened risks are more likely due to prior abortions and miscarriages.
An Ongoing Pattern of Omissions & Refusals
Reardon’s letter goes on to assert that Munk-Olsen has repeatedly altered her research methodology precisely in ways designed to avoid showing the effects of abortion on mental health. For example:
- In a study of bipolar disorders following delivery, Munk-Olsen ignored prior research linking abortion to elevated rates of bipolar disorder, and once again chose to exclude any examination of women’s prior exposure to abortion and miscarriage. When this oversight was called to her attention, she refused to run additional analyses . . . or at least to report on her results from such analyses.
- In a study of the rates of anti-depressant use among pregnant and post-partum women, Munk-Olsen yet again refused a request, published in the same journal, to provide a breakdown of anti-depressant use among the same women relative to their past exposure to abortions.
- In a study of psychiatric interventions following miscarriage, Munk-Olsen revealed in a footnote that she actually did control for prior history of abortion, but once again refused requests to provide any information showing how abortion itself affected the rates of psychiatric treatment.
Munk-Olsen has only reported on mental health treatment rates following abortion in two studies, both of which were used to boldly proclaim that abortion has no effect on mental health. But Reardon asserts that Munk-Olsen chose to deviate from her own prior methodology, and that of other researchers in the field, in order to massage her data in a way that would minimize and obscure the effects of abortion on mental health.
For example, in Munk-Olsen’s first abortion study, published in 2011, she compares inpatient and outpatient psychiatric treatments for nine months before the abortion or first childbirth took place and twelve months afterward. But the time frames and inclusion criteria used for this 2011 study were markedly changed from a similar study she had done in 2006 examining mental health treatment rates following childbirth.
In the 2006 study she excluded all women with any history of inpatient psychiatric care prior to their first delivery in order to measure how childbirth might impact emotionally healthy women. But in the 2011 study, she excluded only women with a history of inpatient psychiatric care nine months prior to their abortion or delivery. The only possible explanation for changing her methodology in this way was that it produced results she preferred to present––or concealed results she did not want to present.
Blended Groups for Concocted Results
An even more serious flaw in her 2011 study was that she included women who had abortions in both groups. Unlike previous researchers who compared women who aborted their first pregnancies to women who gave birth to their first pregnancies, Munk-Olsen included in the childbirth group women with prior abortions who later gave birth. Every researcher knows this mixing would tend dilute the differences between the two groups being compared. It is simply bad technique––but useful if your goal is to obfuscate rather than illuminate.
Despite these study design choices that were clearly meant to minimize findings regarding psychological effects of abortion, her study still showed that women who had abortions had significantly higher rates of psychiatric treatment (15.2 per 1000 person years) compared to both delivering women (6.7 per 1000 person years) and women who have not been pregnant (8.2 per 1000 person years). It also found that the relative risk for psychiatric visits involving neurotic, stress-related, or somatoform disorders was significantly higher for women post-abortion compared to pre-abortion at 2 and 3 months respectively, and that treatment for personality or behavioral disorders was 56 percent, 45 percent, 31 percent, and 55 percent higher at 3, 4-6, 7-9, and 10-12 months respectively.
Despite these findings, Munk-Olsen argues that all of these results can be ignored because, in her view, women with a propensity to mental illness are simply more likely to have abortions, which may, in her view, be a good thing. But as criticisms of her study have elsewhere noted, even her selectively reported data does not support this broad assertion.
Munk-Olsen’s second study of mental health effects associated with abortion was even more severely flawed. But once again, Munk-Olsen refused to provide any additional data that would either confirm or refute her selectively reported results.
Below is the complete version of Reardon’s letter to the Scandinavian Journal of Primary Health Care, containing additional material that was edited out to accommodate the journal’s word count limitations.
Please note: the editor of Scandinavian Journal of Primary Health Care offered Munk-Olsen an opportunity to publish a response to Reardon’s letter, but she declined.
Postpartum Mental Health Study Flawed by Fetal Loss Omission
David C. Reardon
The recent study of postpartum psychiatric illness[1] using the excellent Danish record linkage dataset is unfortunately flawed by Munk-Olsen et al’s decision to, once again, exclude the effects of prior pregnancy losses on their results.
This is a serious oversight. Numerous studies (including one by the Munk-Olsen team) have shown that prior induced abortion or miscarriage increases the risk of postpartum psychiatric disorders.[2,3] It is also known that there is a dose effect, with multiple losses increasing risk, which supports a causal contribution to the higher risk.[3] This association is also on a continuum with studies showing increased mental illness among women who have experienced the death of a child.[4]
Moreover, record linkage has also shown elevated rates of primary care services following abortion.[5] This may help explain why women with postpartum disorders have higher recourse to primary care services both prior to and after subsequent delivered pregnancies.[1]
A better designed analysis of the Danish data would likely have confirmed that women who may require more postpartum care can be identified by identifying prior pregnancy losses. That would be a very actionable finding.
Unfortunately, the Munk-Olsen team is disturbingly inconsistent in their methodologies. For example, in their primary care study,[1] they (a) exclude all women with any history of treatment for mental health disorders prior to childbirth, and (b) also show the consultation rate ratio for two years before childbirth through one year after. This is an excellent study design. It reduces confounding issues and provides an objective health metric pre-pregnancy, during pregnancy, and post-pregnancy.
Similarly, in their fetal death study,[2] Munk-Olsen (a) excluded women with prior psychiatric contact, limiting the subjects to the most psychologically healthy; and (b) used a 12 month period prior to fetal death for their baseline. But then an oddity occurs. Unlike other researchers,[3] they decide to define fetal death to exclude induced abortions. Even more oddly, they modify their results by controlling for exposure to induced abortions while at the same time omitting any statistic showing how abortion exposure affected mental health associated with miscarriage. The rationale for both using and obscuring the effects associated with abortion is not explained but may be deduced from studies indicating that multiple losses, from both abortion and miscarriage, increase the risks to mental health.[3]
Concerns regarding the Munk-Olsen team’s highly erratic treatment of abortion-associated effects is heightened by the very uncharacteristic methodology employed in Munk-Olsen’s two abortion studies.[6,7] In these studies, (a) analyses for women without a prior history of mental disorders are omitted, exactly the opposite of what was done in her other studies,[1,2] and (b) rather than showing the treatment rates for psychological conditions for two years prior to pregnancy outcome,[1] or even just one year,[2] a very convoluted mental health baseline is created based on only first-time contact for mental health care during a nine-month period prior to abortion or childbirth. This baseline choice covers the entire time delivering women were pregnant but includes a mix of pre-conception time and pregnant time for those who had abortions. In her response to a comment raising this concerrn,[7] Munk-Olsen admits this baseline comparison “may not be directly compatible.” But isn’t the whole point of good methodology to find choices that make groups as “directly comparable” as possible?
If the aforementioned differences in methodology are not confounding enough, Munk-Olsen’s abortion studies also mix women who had one or more abortions into the comparison group of women giving birth, an approach contrary to prior research protocols. This cross-adulteration of the samples makes it impossible to compare women who abort their first pregnancies to those who deliver their first pregnancies.
Subsequent refusals by Munk-Olsen to provide any additional information about her data add to our concerns. For example, when a request was made to show the rate of contact for mental health treatments before conception and after the pregnancies, Munk-Olsen dismissed this idea, insisting contact rates had no informative value.[7] But her response is clearly inconsistent with the methodology she employed in her primary care study.[1]
Similarly, when a colleague requested a simple count of the number of women included in the study who had both abortions and childbirth experiences and the percentage of this group who had psychiatric contact before and after each pregnancy outcome, Munk-Olsen emailed that she could not since it would take too much time and effort.
Given these facts, can I be forgiven for imagining that Munk-Olsen’s deviations from protocols used elsewhere[1,2] appear to be intended to obscure rather than elucidate the associations between abortion and mental health in Danish records?
I am not optimistic. But I will again request Munk-Olsen et al to reanalyze the data presented here[1] so as to show the segregated effects of prior voluntary and involuntary pregnancy loss on consultation rates in general, and for mental health consultation rates in particular, both before and after subsequent deliveries. If an association is found, that would provide an actionable screening criteria for early identification of women who may need more postpartum care.
References
[1] Munk-Olsen T, Pedersen HS, Laursen TM, Fenger-Grøn M, Vedsted P, Vestergaard M. Use of primary health care prior to a postpartum psychiatric episode. Scand J Prim Health Care. 2015;33(2):127-133. doi:10.3109/02813432.2015.1041832.
[2] Munk-Olsen T, Bech BH, Vestergaard M, Li J, Olsen J, Laursen TM. Psychiatric disorders following fetal death: a population-based cohort study. BMJ Open. 2014. doi:10.1136/bmjopen-2014-005187.
[3] Giannandrea S a M, Cerulli C, Anson E, Chaudron LH. Increased risk for postpartum psychiatric disorders among women with past pregnancy loss. J Womens Health (Larchmt). 2013;22(9):760-768. doi:10.1089/jwh.2012.4011.
[4] Li J, Laursen TM, Precht DH, Olsen J, Mortensen PB. Hospitalization for mental illness among parents after the death of a child. N Engl J Med. 2005;352(12):1190-1196. doi:10.1056/NEJMoa033160.
[5] Østbye T, Wenghofer EF, Woodward CA, Gold G, Craighead J. Health services utilization after induced abortions in Ontario: a comparison between community clinics and hospitals. Am J Med Qual. 16(3):99-106.
[6] Munk-Olsen T, Laursen TM, Pedersen CB, Lidegaard Ø, Mortensen PB. Induced first-trimester abortion and risk of mental disorder. N Engl J Med. 2011;364(4):332-339.
[7] Munk-Olsen T, Laursen TM, Pedersen CB, Lidegaard O, Mortensen PB. First-time first-trimester induced abortion and risk of readmission to a psychiatric hospital in women with a history of treated mental disorder. Arch Gen Psychiatry. 2012;69(2):159-165. doi:10.1001/archgenpsychiatry.2011.153.
Resources for Help With Postpartum Depression: PostpartumDepression.org
