Today, pregnant women are regularly asked about their mental wellbeing during their medical checkups, and such questioning continues during their postnatal visits to their pediatricians. The quickest assessment relies on the “Whooley questions,” a depression screening tool devised by Pfizer in 1993 shortly after its launch of Zoloft. It has just two questions:
During the past month, have you often been bothered by feeling down, depressed or hopeless?
During the past month, have you often been bothered by little interest, or pleasure in doing things?
An answer of yes to one of the two questions yields a positive result requiring “further evaluation.” If a subsequent assessment determines that the woman is depressed, she may be treated for the condition, even though prior to the screening she might not have self-identified as suffering in this way.
In the United States, such perinatal screening is widely recommended. The American Psychiatric Association (APA), the American College of Obstetrics and Gynecologists (ACOG), the United States Preventative Services Task Force, and the Agency for Healthcare Research and Quality (AHRQ) all promote it as an evidence-based practice.
According to the AHRQ, which is part of the Department of Health and Human Services, it provides a proven benefit: “Depression during pregnancy can raise the risk of having a low–birth–weight baby and premature birth. Screening pregnant women for depression enables health professionals to initiate services that can prevent later problems for both the mother and baby.”
However, other national task forces, most notably in Canada and the U.K., have come to an opposite conclusion about the merits of perinatal screening. The U.K. National Screening Committee concluded in its 2019 review that “there was not enough evidence that such screening programmes were of benefit to women and children,” and that, in fact, they could do harm. False positives could lead to unnecessary treatment, and even if the screening “correctly identified [women] with mental health problems,” it could prove harmful to the majority, the committee wrote.
“Of those women who are correctly identified with mental health problems in the antenatal or postnatal period, only a small proportion of them will receive an effective treatment that does no harm. The remainder will find that: the treatment is effective but brings side effects; the treatment is not effective; the treatment is both not effective and brings side effects.”
Such is the debate around perinatal screening. There is disagreement about whether it provides a benefit or does more harm than good. That debate extends to whether the treatment offered to those diagnosed with depression is “effective.”
This debate, in a particularly extreme form, burst into the news a few weeks ago when a Massachusetts woman, Lindsay Clancy, strangled her three children and then tried to kill herself. News articles told of her serious “mental illness,” and of how postpartum depression needs to be recognized and treated—a message to the public that screening for this “illness” is helpful and that there are effective treatments for it.
However, Clancy’s defense attorney quickly turned that narrative in a different direction. He told of how she had been getting intense psychiatric care for the past four months and had been prescribed 13 different psychiatric medications during that time. This left a question hanging in the air: was it the treatment, rather than the “disorder,” that prompted her homicidal behavior?
The different recommendations by the three task forces, and the public attention drawn to postpartum depression by the Clancy case, raise an obvious question. Why does the U.S. describe perinatal screening as providing a proven benefit, while the task forces in the U.K. and Canada see no evidence of such proof? How did they come to such difference conclusions when the research literature to be reviewed is the same?
This exploration of “evidenced-based” psychiatry also leads to two follow-up questions.
First, how do perinatal women experience this formalized inquiry into their mental health?
Second, what opportunities are lost when it is a medical narrative that governs perinatal care? Instead of viewing distress during this period as a “mental illness” located within the woman, what social treatments and policies might be adopted if the distress was seen as arising from the environment? The latter would involve investigating the “social determinants of health.”
Medical Theories and the Rise of Screening
It has long been noted in medical writings that emotional difficulties may arise during pregnancy and following the birth of a child. As Robert Sparks wrote in Sadness and Support: A Short History of Postpartum Depression, Hippocrates devised a hypothesis in the fourth century B.C. that subsequently “survived for over a thousand years. He proposed that lochial discharge—the fluid that comes from the uterus after birth—if suppressed, could flow to the head and result in agitation, delirium and attacks of mania.”
More than 1500 years later, a 13th-century female physician, Trotula of Salerno, came up with an equally fluid hypothesis. In her treatise Passionibus Mulierum Curandorumon (i.e. The Diseases of Women) she wrote that “if the womb is too moist, the brain is filled with water, and the moisture running over the eyes compels them to involuntarily shed tears.”
A further review of medical history could produce dozens of such pronouncements throughout the centuries. The understanding that drives screening today derives from the third edition of the APA’s Diagnostic and Statistical Manual of Mental Disorders (DSM) published in 1980, which conceptualized depression as a brain disease.
Prior to this time, depression was understood primarily to be an episodic disorder, which often arose from difficulties in life and could be expected to subside with time. The emotional lability many women experienced during pregnancy and postpartum could, for the most part, fit within that concept of “normalcy” in response to the environment. Then Prozac was introduced in 1987 as a breakthrough drug that could fix a chemical imbalance in the brain, and the depression-is-an illness boom was on.
As the SSRIs came to market, manufacturers of the new drugs, in concert with the APA, ran “public education” campaigns informing the public that depression was a brain disorder that was “underrecognized” and “undertreated,” and that it could be “fatal if left untreated.” In 1991, the APA helped organize the first “National Depression Screening Day,” with Eli Lilly, the maker of Prozac, one of the funders of the event. A non-profit organization, Screening for Mental Health, formed to organize these annual events, which have been held in October ever since during “Mental Health Awareness Week.”
With this new “disease model” taking hold, societal use of antidepressants steadily increased, and this was particularly so among women. Prior to 1987, fewer than 3% of U.S. adult women took an antidepressant. Three decades later, nearly 20% were doing so.
The increase in antidepressant use in pregnancy is of an even more pronounced sort.
The Edinburgh Postnatal Depression Scale (EPDS) was released in 1987, just as Prozac was coming to market, and today it remains the mostly commonly used questionnaire for perinatal screening. Much like Pfizer’s Whooley questions, it sets a low bar for a positive result. There are ten questions, each one scored from 0 to 3 points, and screening studies typically use cutoff scores from 10 to 13 to identify those with depression.
Here are two of the ten questions:
I have looked forward with enjoyment to things
0 As much as I ever did
1 Rather less than I used to
2 Definitely less than I used to
3 Hardly at all
I have been anxious or worried for no good reason
0 Not at all
1 Hardly ever
2 Yes sometimes
3 Yes, very often
Given that even scoring a 1 on every question can lead to a “positive” result, it is easy to see how its use could spur an increase in the diagnosis of depression in perinatal women and contribute to an increased use of antidepressants during pregnancy. Another instrument often used is the Patient Health Questionnaire-9 (PHQ-9), which asks similar questions and scores answers in the same 0 to 3 manner. On a scale of 0 to 27, a score above four is seen as evidence of “mild depression,” with higher scores categorized as moderate and severe depression.
Prior to Prozac, and the introduction of such rating scales, antidepressant use during pregnancy was quite rare among U.S. women (less than 1 in 300). This figure rose to 8% of pregnant women by 2008. Rates of antidepressant usage during pregnancy are lower in Europe.
In the U.S., much of this exposure to antidepressants results from women becoming pregnant while on the medication. A study of 340,000 women who gave birth from 2006 to 2011 found that 6.5% used antidepressants during their pregnancy, and that 4.4% had been on the medication before becoming pregnant (prevalent users), and another 2.1% were newly prescribed an antidepressant during the nine months (incident users.)
The “Medical Evidence” For Screening
The opposing conclusions drawn by the U.K., Canadian, and U.S. task forces reflect a diversity of perinatal screening recommendations in developed countries around the world. International guidelines disagree regarding the best time to conduct screenings, who should conduct it, what screening tools should be used, and what the follow-up and referral pathways should be post-screening.
The variation tells of a glaring hole in the evidence base for perinatal screenings. The premise of perinatal screening is that undiagnosed depression can be identified and successfully treated. Yet, even though it has been promoted for more than two decades, there has never been a randomized clinical trial of good quality that compared outcomes in “screened” and “non-screened” cohorts, with depressed patients in both groups provided the same “usual care.”
“That’s the only scientific method that would allow you to tease out the effect of screening alone,” said Eddy Lang, co-chair of guidelines issued by the Canadian Task Force on Preventive Health Care (CTFPHC) and head of emergency medicine at the University of Calgary. “There are many studies that combine screening with intensive interventions following positive screens, [but] that’s a different question. The question of just the screening by itself: Does that, in and of itself, make a difference? There we need randomized controlled trials.”
Maternal depression and birth outcomes
There is agreement in the medical community that maternal depression raises the risk of adverse outcomes for mother and child. Antenatal depression is associated with an increased risk for premature birth and low birth weight. Postpartum depression is associated with adverse cognitive and developmental outcomes for children, which are thought to be mediated through impaired mother-child interactions.
There are two critical “evidence-based” questions specific to perinatal screening:
- Can the screening tools identify depression with a high fidelity and low number of false positives?
- Is there evidence that treatment of depression in perinatal women is effective? In particular, are there studies that show that antidepressants are effective in pregnancy and the postpartum year?
Screening tools: the problem of false positives
While women are said to be at increased risk of depression during the perinatal period, prevalence studies do not find that to be necessarily so. A survey of more than 14,000 women in the United States of childbearing age determined that the 12-month “period prevalence” of depression was 8% among pregnant women, 9% among postpartum women, and 8% among non-pregnant individuals.
The usefulness of the EPDS or any other perinatal screening tool depends on how accurately it can identify women who meet DSM criteria for depression (true positives) and minimize the number of false positives—women who score as depressed even though they are not. Both the U.K. and Canadian task forces told of how the EPDS and other screening questionnaires could produce a significant number of false positives, which mitigates their usefulness.
In its 2022 review, the Canadian task force calculated that if the prevalence of depression in a cohort is 8%, screening 100 women with the EPDS will result in 5 true positives, 5 false positives, 87 true negatives, and 3 false negatives. While the 5 true positives would potentially stand to benefit (if subsequent treatment was more effective than usual care), the 5 false positives would be exposed to further assessments of their mood and possibly unnecessary treatment, and thus be harmed by exposure to its adverse effects.
Research is needed, Lang said, to assess this benefit/harm ratio. “It’s not sufficient for us, from the task force standpoint, to just identify more patients, and to catch more people, and to have more positive screens, unless you demonstrate that you can improve outcomes that are important. So less depression, less hospitalization, fewer suicides. Unless you can show that, you haven’t met the bar for making a recommendation that you would apply to a wide population.”
Indeed, he said, such research could tell of how screening leads to worse outcomes were it to show that “more people who are screened end up on medications, end up taking time off work, end up with a lower quality of life, while those who weren’t screened have less depression and a better quality of life because they never went down that rabbit hole of false positives and overdiagnosis.”
In lieu of formal screening, the Canadian task force recommended that health care providers simply ask their patients about their well-being as part of usual care. “That would be an important part of care,” said Barbara Mintzes, a former public health professor at the University of British Colombia, who is now at the University of Sydney School of Pharmacy. “But that’s different from giving them these screening questionnaires, and making a decision, based on the screening questionnaire, on what they’d do next.”
The U.K. screening committee came to a similar bottom line. With the use of the EPDS, “A high proportion of women with a positive screen referral for a full psychological assessment are likely not to have major depressive disorder.” An editorial in Lancet stated that this high false positive rate could do significant harm:
“This situation is potentially dangerous. Results of qualitative studies suggest that women are extremely concerned about depression screening, about the stigma associated with a diagnosis of depression, and that a positive result might lead to an automatic social service referral, and potentially removal of their baby.”
The U.S. task force, in its 2016 report, similarly found that even if a screening used a cutoff score of 13 on the EPDS scale, from one-third to more than half of the positive screens would be false positives.1 However, in contrast to the U.K. and Canadian task forces, the U.S. task force dismissed “false positives” as a significant drawback. “The USPSTF found adequate evidence that the magnitude of harms of screening for depression in adults is small to none,” it wrote.
Effectiveness of screening + treatment
The second question that arises from screening protocols is whether treatments of perinatal depression are effective. If a screening helps identify “unrecognized” depression in perinatal women, this provides a benefit only if the usual treatments—psychotherapy or medication—are effective in reducing depressive symptoms (compared to placebo).
The U.K. task force
The U.K. task force, for its part, concluded that the “evidence on the effectiveness of treatments is not good enough to make recommendations at the current time. It is not clear which of the treatments work for different common mental health problems before and after giving birth.”
The one possible exception, the U.K. committee stated, was that there was some evidence that cognitive behavioral therapy (CBT) “is likely to lead to a small reduction in the severity” of perinatal depression.
The U.K. task force did not find any study that assessed the efficacy of antidepressants for pregnant women identified as depressed through screening. However, in a general review of antidepressant exposure during pregnancy, they found that “negative offspring outcomes were reported in all three systemic reviews” of such studies, and that “no study reported the benefits to women or their offspring from pharmacological interventions.”
The Canadian task force
The Canadian task force focused on finding randomized controlled trials (RCTs) that evaluated screening for depression in perinatal women compared to usual clinical care, with depressed patients in both groups provided the same treatment following diagnosis (e.g., no enhanced treatment for the screened cohort.)
The task force found no such study in pregnant women that met this standard, and only one small RCT in postpartum women to evaluate.
That Hong Kong study randomized women either to screening or usual clinical care. Women identified as depressed were then offered counselling or management by a community psychiatric team.
Twenty-nine percent of those in the screened group were identified as depressed, compared to 6% in the usual-care arm. Thus, the screening led to nearly a five-fold increase in the percentage of women treated for postpartum depression. If treatment led to better outcomes for depressed patients, then increased diagnosis of depression in the screened group should have led to better outcomes for the cohort as a whole.
However, the Canadian task force found that it was “very uncertain” that the screening “improved outcomes of interest” for the screened cohort. Those outcomes of interest included reduction of depressive symptoms, general mental health outcomes, parent-child stress, and marital stress.
As such, with a lack of RCT evidence that perinatal screening provided a benefit, the Canadian task force recommended against the practice.
The U.S. task force
The U.S. task force, in its review of the literature, expanded the inclusion criteria for screening studies to be assessed. It decided that studies “could also include additional treatment elements” for the depressed patients in the screened cohort. This meant that its review would include studies that compared screening + enhanced treatment + usual care to usual care alone.
The U.S. task force identified six studies, five in in postpartum women, that met this looser criterion. Five were deemed fair in quality and one as good. However, none of the six compared usual care in a screened group to usual care in a non-screened group, which was the comparison that the Canadian task force was focused on.
In the six studies, cut-off off scores for a “positive” result ranged from 10 to 13 on the EPDS scale, with 10% to 28% of the screened women identified as depressed. The enhanced treatment in the screened groups included additional training for nurses and midwives in how to best manage depression, person-centered counseling, and improved follow-up care. In one study, the enhanced treatment specifically included CBT.
The U.S. task force concluded that in the five postpartum studies, there was not a meaningful difference in the mean depressive symptoms for the two groups, as a whole, at later follow-up assessments. The difference in mean EPDS scores was less than one point, which the task force noted was clinically meaningless. At a cohort level, screening did not provide a benefit.
However, the U.S. task force did find a small benefit in the subset of patients identified as depressed. “There was a 21 to 33 percent increase in the likelihood of remission or response” for depressed patients in the screened group compared to depressed patients in the usual care cohort.
As for the pregnancy study, although the difference in symptom reduction among the depressed patients in the two groups was not statistically significant, the task force noted that the results “pointed” in the right direction, and thus were seen as providing some evidence for screening during pregnancy.
The U.S. task force also reviewed studies of various treatments for perinatal depression and their potential harms. “The evidence on treatment benefit is primarily for nonpharmacologic interventions (i.e., CBT),” it concluded. “There is evidence of a small risk of harm to fetal health with SSRI use in pregnant women, and there is a lack of evidence on harms of SSRI use in postpartum women.”
With this research in mind, the U.S. task force determined that “with at least moderate certainty that there is a moderate net benefit to screening for depression in pregnant and postpartum women who receive care in clinical practices that have CBT or other evidence-based counseling available after screening.”
Weighing the benefits versus risks
The three task forces pored through the same scientific literature. The differences in their bottom-line recommendations resulted from their different inclusion criteria for screening studies, and their differing assessments of possible harms.
The U.K. and Canadian task forces did not find evidence that screening provided a benefit on any important outcome, either to the entire cohort or to those diagnosed as depressed. In the absence of any benefit, their assessments highlighted possible harms due to false positives and the drain that screening imposes on clinical resources.
In contrast, the U.S. task force minimized the possible harm related to false positives, didn’t consider the drain on resources from screening, and cast a wider net for evidence of efficacy, finding it in studies that provided “additional” elements of care to depressed patients in the screened cohorts. It viewed the evidence through a different lens.
The moderate benefit quantified
Researchers often calculate the “number needed to treat” to describe the efficacy of a treatment. How many patients must be treated with a drug to produce one patient that has a significant response beyond placebo? In this case, the relevant calculation is the “number needed to screen.”
Of the six screening studies reviewed by the U.S. task force, there was one conducted in the United States and it provides data that makes it possible to do this calculation. In the postpartum study, 45% of depressed patients within the screened group had a significant response to treatment, compared to 34% in the usual care group. Here is the calculation that leads to a “number needed to screen”:
- There were 1353 women in the screened cohort.
- 399 screened positive for depression.
- 217 remained in the study to the end and were treated for depression.
- 98 of the 217 had a significant response to treatment (45%).
- If only 34% of the 217 had had a significant response (the response rate in the usual care group), 74 patients would have had a significant response. Thus, screening 1353 women produced 24 additional “significant” responses.
- Divide 1353 by 24, and the number needed to screen to produce one additional response to treatment is 56.
That is the moderate benefit that the U.S. task force concluded flowed from screening all women in a primary care setting for depression. Meanwhile, a Lancet editorial, which took issue with the U.S. task force recommendation, summed up the evidence base for perinatal screening in this way:
“Routine screening for depressive disorders in pregnancy and post-partum could potentially be harmful. The pregnancy and postnatal period is an opportunity to maximise the health of women and their families. But this needs to be done through sensitive enquiry in the context of a broader conversation about the physical and mental health wellbeing of mothers.”
The Risk of Harm With Antidepressants
While none of the three task forces recommended antidepressants for perinatal depression, in the real world of clinical care it remains a common treatment. Both the U.K. and U.S. task forces told of possible harm from their use during pregnancy.
The scientific literature tells of a long list of possible adverse outcomes for the fetus: miscarriage, spontaneous abortion, severe congenital malformations, cardiac malformations, fetal growth retardation, smaller head size, preterm birth, low APGAR scores, persistent pulmonary hypertension, neonatal behavioral symptoms, neonatal seizures, neonatal respiratory distress, mortality during the first year, slower psychomotor development, autism spectrum disorder, developmental and language delays, and higher rates of depression and anxiety.
Meta-analyses of such studies in the literature find that risks may vary according to the type of antidepressant (SSRI, SNRI, or tricyclic), the specific drugs within the SSRI and SNRI class, and what adjustments are made for confounding factors. One comprehensive review reported “odds ratios” for various possible adverse outcomes ranging from 1.0 to 2.5, meaning that there may be very little (or no) increase in the likelihood of such an event at the low end, and more than a doubling at the high end.
As many of the adverse events are rare, the increase in the absolute risk might be quite small. If the background rate for an adverse event is one in 500 births, then a doubling of the risk would put the absolute risk for antidepressant users at one in 250 births. This was why the U.S. Task Force wrote that the “likelihood” of such severe adverse events is “low.”
However, with so many elevated risks present, the cumulative odds that antidepressant usage will lead to an adverse event of some type could be expected to be significantly higher. There is a hint of this in a 2022 study of Medicaid patients. By age 12, 47% of children whose mothers had taken an antidepressant during the second half of their pregnancy had been diagnosed with a neurodevelopmental disorder, compared to 31% of children without such exposure. Among private pay patients, the respective figures were 25% for the antidepressant-exposed children and 15% for the non-exposed group.2
Perhaps the biggest concern with antidepressant usage relates to the use of SSRIs during the first trimester. Serotonin is known to play an important role in “signaling” the migration of brain cells to different areas of the brain during fetal development, and in rat studies, SSRI exposure has been found to lead to disruptions in this process. Although there is always a question of how relevant such animal studies are to humans, a 2018 study of 98 infants, which was published in JAMA Pediatrics, produced this “key points” summary:
Question: Is prenatal exposure to selective serotonin reuptake inhibitors associated with fetal brain development?
Findings: In this cohort study including 98 infants, significant gray matter volume expansion was noted in the amygdala and insula, as well as an increase in white matter structural connectivity between these same regions in selective serotonin reuptake inhibitor–exposed infants, compared with infants exposed to untreated prenatal maternal depression and healthy controls.
Meaning: In line with prior animal studies, these multimodal brain imaging findings suggest that prenatal selective serotonin reuptake inhibitor exposure has a significant association with fetal brain development.
The most common problem that arises with antidepressant exposure during pregnancy is that up to one-third of newborns suffer “neonatal withdrawal syndromes.” Frequently reported symptoms include respiratory difficulties, irritability/agitation, tremors, feeding problems, and seizures. In a 2022 study, 84% of the reported symptoms were classified as “serious,” and much to the surprise of the researchers, tricyclics were more problematic than SSRIs in this regard.
Advice to Pregnant Mothers
While such risks to the fetus and child have drawn considerable attention, the usual advice to pregnant women is that the risk of untreated depression outweighs the risk of harm to the fetus. This advice stems from “discontinuation” trials in women who, finding out that they were pregnant, were interested in stopping their antidepressant.
In 2006, researchers at Massachusetts General Hospital reported that the relapse rate was 68% in the antidepressant-withdrawn group compared to 26% in the drug-maintained group, a finding that is often cited as reason for pregnant women to stay on their medication. However, critics of this advice note that that it is clouded by a failure to distinguish between drug withdrawal symptoms and relapse.
“They didn’t even consider the possibility of withdrawal—so that’s a major issue,” said Barbara Mintzes, a researcher at the University of Sydney’s School of Pharmacy who specializes in the study of pharmaceutical policy. “There was no tapering plan. They were stopping cold turkey, so the likelihood of withdrawal was very high.”
The discontinuation trials also don’t tell of how well “drug maintenance” treatment works. Studies of real-world patients have found that stay-well rates are quite low, and this appears to be true for pregnant women too. A study of 85 women who had a diagnosis of major depression before getting pregnant and continued their antidepressants during the perinatal period found that only 18% remained in “remission” during this time; 50% percent struggled with “mild symptoms;” and 32% experienced “clinically relevant” symptoms.
A WebMD article summarized the findings with this headline: “Antidepressants Often Ineffective During Pregnancy, in New Moms.”
In a similar vein, a study of 187 pregnant women who reported pre-existing depression or new-onset depression during pregnancy found that antidepressants did not protect against “self-harm” ideation. Twenty-two percent of those who took an antidepressant during pregnancy reported frequent “self-harm ideation” during their postpartum year, compared to 15% of those who didn’t take such medication.
In the Prozac era, maternal mortality has soared, increasing from 17 maternal deaths per 100,000 pregnant women in 1990 to 26 per 100,000 in 2015. Postpartum hemorrhage is the leading cause of maternal death, and a study of postpartum hemorrhage in women with mood disorders found that it occurred 1.5 times more frequently in those who took an SSRI in the last month of their pregnancy compared to those who were “untreated” for the disorder, a finding that suggests increased antidepressant use could be a causal factor contributing, in some slight way, to the increase in maternal deaths.
Thus, the bottom line in the evidence base: While it is known that antidepressant usage during pregnancy presents risks to the fetus, there isn’t good evidence that these drugs are “effective” in alleviating depressive symptoms or reducing self-harm in the mother, and their use late in pregnancy increases the risk of postpartum hemorrhage. The selling of antidepressants to the American public over the past 30 years has produced heightened risks for the unborn and, it seems, may have contributed to a rise in maternal deaths.
Women on Screening
Surveys of perinatal women typically find that they welcome general inquiries about their mental health, whereas mandated screenings often leave them baffled and fearful for what might follow from the results. They worry about losing their jobs, their children, and the burden of being labeled as depressed or “mentally ill.”
That was what Karen Tabb, an associate professor in the school of social work at the University of Illinois at Urbana-Champaign, found when she surveyed women in Illinois, which is one of several states that mandate antenatal and postpartum depression screenings.
“The implementation of that mandate is oftentimes nothing less than horrific,” she said. “You hear the experience of someone reporting that they’re suicidal and they’re scared to hold their baby. . . . and they’re opening up and sharing this with a clinician. And in many of these health care systems in my state, the protocol is to call 911. The patient is detained, and an ambulance comes, and they do a psych evaluation.”
Such implementation of mandated screenings creates a space, she added, “where people aren’t open and willing to discuss their level of need.”
Bafflement was the first response that Liz Woytus had to a postpartum screening shortly after her first child was born 16 years ago. “You know, your hormones are crazy after you have a baby,” she said. “I kept getting up in the middle of the night because I was getting hot and cold and hot and cold.” On top of that, she was crying. “I don’t think screening a woman after she gives birth in the first 72 hours is wise in anybody’s eyes. Your body’s just going nuts.” A little later on, maybe—“but give me a week or two to figure out what the hell just happened.”
Each of the next two times Woytus gave birth, she was screened for depression. She is, she stressed, “completely neutral” in her attitude toward such screenings. But four years ago, during the birth of her third child at a large teaching hospital, her uterus ripped, and she was in a lot of pain. Yet, a social work intern entered the crowded room with a clipboard and asked her about her mood. Was she feeling anxious or depressed?
“My view is that they could be doing a better job with the screening process—and maybe send in more qualified people,” she said. “First of all, why would you have that conversation in front of family? That wasn’t appropriate.” The likelihood that it would prompt honest responses is slim, she added.
Last fall, a woman described her postnatal depression screening on Reddit, and that triggered a heated discussion about its problems. The woman recalled the questions she was asked:
“Do you have little energy to do simple tasks?” Ofc I’m awake all night with a screaming newborn.
“Do you still enjoy things you once used to?” Idk I don’t have the time to do them bc my baby will scream if I’m not holding him.
“I feel anxious for no good reason” I’ve got a 7 week old depending on me for food and safety that’s a pretty good reason imo.
“I feel sad for no good reason” I’m having difficulty bonding with my baby that seems like a great reason to be sad.
“That’s all, just a mini rant about how weird these screening tests are.”
In response, commenters chimed in with their own stories, describing the inadequacies and absurdities of the process—especially for those already diagnosed with such disorders. As Prior_Crazy_4990 put it, “I can tell you right now I lied my ass off on every single one and checked that I was doing great with life because I was scared of having my daughter taken away if I admitted anything was wrong.”
Which, in turn, prompted Not_Starlight_Kitsun to reply: “I know I lied my ass off on mine due to my anxiety.”
In her survey, Tabb identified other common patient complaints. Women told of a process that was rushed and unprofessional, was ineffective because providers didn’t explain the purpose or uses of the screening tool, didn’t tell patients anything about the results, and failed to provide any follow-up relating to patient depression scores.
Concerns about screening, Tabb said, are even more pressing for people of color and the poor, who are already wary of existing power structures and afraid of losing their child. Black patients felt especially fearful and nervous about answering screening questions because of worries it might lead to a visit from Child Protective Services or harm them in a custody case. “It’s usually a hard thing to talk about, like, I’m not going to go tell somebody because I don’t want DCFS to come,” said one patient, referring to the Illinois Department of Child and Family Services.
“Their perceptions are steeped in reality,” Tabb said.
At the same time, Tabb is not anti-screening, but believes a validated questionnaire, if properly implemented, could encourage a dialogue between patients and their providers about the difficulties and stresses they may be experiencing.
For Woytus, it’s the dialogue part that would be helpful . . . just “sitting down” with someone who could listen, read between the lines, ask the right follow-ups, and nudge things to the fore. “That would be more productive than anything else, because you might find things out about the person. . . and obviously, if you’re doing better mentally, you’re going to take care of the baby better.”
The Medicalization of Pregnancy and Postpartum Motherhood
In the three task force reports, there is a consensus that it is helpful for clinicians to ask pregnant and postpartum women how they are doing, and that they provide help and support for those are struggling in some way, with referrals to counseling one recommended option. Those who are critical of perinatal screening similarly encourage such clinical care.
However, as critics note, screening recommendations reflect, in part, a “medicalization” of pregnancy, with normal stresses converted into a medical diagnosis—of depression, anxiety, or other psychiatric disorder.
“Often, the problems that can really lead to distress are life problems,” Mintzes said. “You know, that whole side of just framing a lot of things within mental health rather than looking at what’s going on socially.”
She added that conflation of such social-related stresses with “illness” is an example of “disease mongering” that has been promoted by pharmaceutical companies, a topic she addressed in her 2010 book Sex, Lies, and Pharmaceuticals: How Drug Companies Plan to Profit from Female Sexual Dysfunction. Once women with “life problems” get identified as depressed or anxious, then drug treatment may be in their future.
“It’s a very different situation for women at the most severe end of psychiatric disease in terms of decisions about discontinuing drugs or not—versus somebody who might have come into a GP who might have life difficulties that were temporary, where they might not even have met the criteria for a diagnosis for major depression,” she said. “Or they might be at the very mildest end of the spectrum. And they end up on antidepressants.”
Adam Urato, a consulting obstetrician based in Massachusetts who coauthored a 2016 paper, “Making Sense of the Controversy: Use of SSRIs in Pregnancy,” sees perinatal screening in a similar light.
“At first glance, screenings always sound good, and that’s true in medicine, and in healthcare for a variety of things,” he said. But the question that has to be asked: “Are you creating more harm than good when you’re starting to do widespread screenings? And that would certainly apply in the mental-health realm for a variety of reasons.”
One problem with universal screening, he added, is that it casts a wide net in identifying women as suffering from a psychiatric disorder, catching even those who didn’t see themselves as suffering from depression or anxiety. “They came in, they thought they were doing okay,” he said. “But if you check them off in a certain way, then you get labeled with a disease . . . I really have concerns that what, a lot of times, are natural processes are being pathologized.”
Moreover, he sees perinatal screening in a larger context. The pharmaceutical industry’s influence is undeniable, he said, creating “a pharmaceutical conventional wisdom” that leads clinicians to medicate increasing numbers of people, with the use of antidepressants by women a prime example of that. He regularly cares for women who have become pregnant while on an antidepressant. “The question then becomes: How best to manage that? And that is a common scenario, more common than the primary OB screening and initiating treatment.”
Urato emphasized that, in the real world of clinical care, there is no one-size-fits-all practice to follow. The message to women, he said, should not be “an extreme or 100 percent ‘you can’t do this—you can’t take medication.’” Instead, the focus should be on providing women with information: the risks of going on psychiatric drugs, the difficulties of coming off, the potential consequences during pregnancy, the potential risks to the fetus. “That’s got to be on the radar,” he said.
Georgio Schoretsanitis and Chiara Gastaldon, who conducted the research on neonatal withdrawal syndromes mentioned above, see the challenge in a similar way. “One thing these women should have is access, as easy as it gets, to interdisciplinary care teams that would make it a mission to guide these women through pregnancy,” Schoretsanitis said.
As for perinatal antidepressant use, “there’s no single patient who’s identical to another one,” he said. “You know, every woman has a different history. And if we’re answering this question in a binary way, saying ‘yes’ or ‘no,’ ‘should she’ or ‘shouldn’t she,’ you know, I think this is highly reducing it.”
At the same time, Mintzes and others note that viewing perinatal distress through a medical lens turns societal attention away from the “social determinants of health,” which need to be addressed by social policies. A survey of 215 perinatal women found that two-thirds reported material needs of some type—supplies for baby and family; financial help to pay for housing, food, and utilities; transportation to medical care; and so on. More than half told of needing “social support,” with many new mothers telling of “social isolation.”
Before Mintzes pursued an academic career, she worked in nonprofits in the field of women’s health, where they ran support groups for postpartum women. The chance for new mothers to talk to other women with similar experiences, share stories and strategies—that can be hugely helpful, she said. “And that’s very different from the very medicalized model.”
As might be expected, social policies that provide pregnant women and new mothers with more support have been shown to be effective remedies for perinatal stress. A 2017 study found that something as simple as home visits “can influence postpartum depression in a positive way and could improve mothers’ and infants’ health.”
Most developed countries allow new mothers—and at times fathers—to stay at home for extended periods without needing to go to work, a social policy that eases some of the stresses that come in the postpartum year. Sweden mandates six months for mothers and a month for fathers, a policy that has been found to dramatically improve maternal health, both physical and mental. The U.S., for its part, is a rarity among developed nations for not adopting maternal leave mandates.
Such is the mismatch between what perinatal women say they need and want and what mandated screenings provide.
Is it the Disease or the Drug?
The Boston Globe, in its coverage of Lindsay Clancy’s killing of her three children, has told of how her trial could become a test of how “postpartum mental health is treated.” There are in fact three narratives that are now being presented to the public: Clancy planned the killing in a methodical, cold-hearted manner (the prosecution); she was suffering from postpartum psychosis (the disease explanation); and that it was the drug treatment that is to blame (the defense).
All three narratives begin with the story of a woman who, prior to the birth of her third child, was a loving and caring mother. She worked as a “delivery and labor nurse” at Massachusetts General Hospital, her colleagues describing her as a “consummate caregiver.” She gave birth to her third child last May, and the first months went fairly smoothly. But in September, her maternity leave was ending, and faced with returning to work, she became so anxious that she sought psychiatric help.
This is when the disease narrative takes hold. Postpartum depression is a common disorder, and in Clancy’s case, it devolved into postpartum psychosis. It is a rare condition, occurring in one to two per 100,000 new mothers, and raises the risk—however slight—of infanticide.
Mothers who kill their children, researchers have found, frequently have “depression, psychosis, prior mental health treatment, and suicidal thoughts.” Given that Clancy was treated for her difficulties, the implication in this narrative is that she was treatment-resistant.
The defense narrative tells of psychiatric care gone amok. Clancy was initially prescribed a benzodiazepine for her anxiety, and that devolved over the next four months into an onslaught of polypharmacy, as she was prescribed 13 psychiatric drugs: three benzodiazepines, (Klonopin, Valium, and Ativan); five antidepressants (Zoloft, Prozac, Remeron, Trazadone, and amitriptyline); a mood stabilizer (Lamictal), an antipsychotic (Seroquel), and three other drugs for anxiety and insomnia (Ambien, buspirone, and hydroxyzine). Her husband worried that she was turning into a zombie, and on January 23, apparently while under the influence of “command hallucinations,” she strangled her three children and then threw herself from a second story window in an attempted suicide.
As the defense makes this argument, it will be able point to scientific findings to support its case. In the very first trials of fluoxetine (Prozac), there were reports that it could induce psychosis, hallucinations, and suicidal impulses. Clinical trials of other SSRIs also told of how they could cause agitation and other troubling side effects and, in 2010, researchers searched through the FDA’s adverse events database for reports of “prescription drugs associated with reports of violence toward others.” They determined that one smoking cessation drug, varenicline, and “antidepressants with serotonergic effects were the most strongly and consistently implicated drugs.”
An Australian researcher, Yolande Lucire, has investigated a number of violent and homicidal acts by people taking antidepressants, and found an apparent genetic link: often the perpetrators have a gene mutation that hinders the usual metabolism of serotonin, and this leads to a toxic buildup of the neurotransmitter in the brain. Studies also have found that SSRI antidepressants may stir suicidal and violent impulses in healthy volunteers, evidence that such acts can’t always be dismissed as due to the disease, not the drug.
The assistant district attorney prosecuting the case has told the court that this is a case of “first-degree murder, and that Clancy acted with “deliberate premeditation and extreme atrocity and cruelty.” The prosecution’s argument is that it wasn’t the disease that caused her act, but simply an evil act of homicide, and to support this contention, it has noted that psychiatrists in Rhode Island concluded in December that Clancy wasn’t suffering from postpartum depression.
The prosecution is putting forth a narrative that supports a charge of first-degree murder, but it is also one that the defense will likely welcome. If it isn’t the “disease,” then how will a jury comprehend how a loving, devoted mother, who had a history as a “consummate caregiver” as a delivery nurse, committed “maternal infanticide?
The case, in this way, will bring public attention to how postpartum distress is treated, and, at the very least, raise questions about the merits of drug treatments as a solution for that distress.
MIA Reports are supported, in part, by a grant from The Thomas Jobe Fund.
- The U.S. Task Force calculated that if the prevalence of depression was 10% in perinatal cohort, a cutoff score of 13 on the EPDS would have a “positive predictive value” ranging from 47% to 64% for detecting major depressive disorder. A positive predictive value of 47% would mean that 53% of the “positive” results would be false positives, and 47% would be true positives. A positive predictive value of 64% would mean that 36% of the positive results would be false positive. Hence, the U.S. Task Force calculated that one-third to more than one-half of the positive results would be false positives. ↩
- In this Medicaid study, crude results showed a doubling of the risk of neurodevelopmental disorders in children exposed in utero to antidepressants. However, the researchers stated that after they “adjusted” for an extensive list of potential confounders, the excess risk with antidepressants disappeared. In a blog for Mad in America, Vera Wilde detailed the methodological flaws with the “adjustments” the authors made to their crude results. ↩