Follow-Up on Antidepressants During Pregnancy: An Exchange of Views

I have several concerns with the content of the recent Journal Watch Women’s Health article entitled “Antidepressants During Pregnancy: Treating the Condition While Acknowledging the Risks.”1 My main concern, however, is that Dr. Claudio Soares, the author of the piece, is paid by several antidepressant manufacturers2 and this information was not made available to readers of the article.

Dr. Soares writes: “Despite the cautionary remarks commonly made by most regulatory agencies and medical societies about the use of psychotropic medications during pregnancy, considerable data supporting the efficacy and reproductive safety of antidepressants have accrued.” First, he does not provide a reference showing efficacy of these antidepressants in pregnancy. What evidence does Dr. Soares have of the efficacy of these drugs for pregnant women? Could he (or you) please provide me with these references? Where has treatment of the pregnant woman with antidepressants been shown to improve maternal or child outcomes? Second, he references only one article on reproductive safety and completely fails to reference most of the studies showing the lack of reproductive safety.3,4,5,6,7,8,9,10,11,12,13

Dr. Soares then writes: “Conversely, evidence suggests that untreated depression has negative consequences for both mother and child.” His use of the word “conversely” here suggests to the reader that untreated depression in pregnancy causes problems, while antidepressant treatment is associated with better outcomes for the mother and child. This, however, is not the case. Antidepressant treatment of pregnant women has not been associated with better pregnancy outcomes in any study of which I am aware.

In fact, to my knowledge, there is only one study3 that has directly compared women with similar levels of depression (some taking antidepressants versus those not taking them) to see what the pregnancy outcomes were. Oberlander and his colleagues in Vancouver published this study in the Archives of General Psychiatry last year. Dr. Oberlander expected that the patients who were being treated with antidepressants would have better pregnancy outcomes, but this was not the case. The women on antidepressants had more babies with low birthweight and respiratory problems. The newborns exposed to SSRIs also had longer hospital stays, feeding problems, and jaundice. So what Dr. Oberlander showed is that if you take two women with equal levels of depression, one of whom is on an antidepressant and the other who is not, the patient on the antidepressant is more likely to have a child with low birthweight, respiratory distress, and other complications.

So, yes, there are risks to untreated depression in pregnancy, but, as far as we can tell, in the only study that has compared two groups with similar levels of depression, the risks of antidepressant use (regarding fetal outcomes) appear to be greater than in the untreated group.

How did Dr. Soares miss Dr. Oberlander’s article? The Archives of General Psychiatry is a well-respected publication from the JAMA group. And, because a randomized-controlled trial on this topic is not considered ethical, the Oberlander study is really the best (and only) study of its kind that we have.

Dr. Soares writes in boldface type the major heading: “Deconstructing the Myths: The Safety and Efficacy of Antidepressants During Pregnancy.” In this section, Dr. Soares very selectively quotes the literature, citing only studies that show no increase in birth defects with antidepressant use by pregnant women. He fails to cite the studies that do show an increased risk of birth defects.4,11,12,13 Even the summary on Up-to-Date, an online source widely used by physicians, states: “First trimester exposure to any SSRI may increase the risk of congenital anomalies.”14

Even more outrageously: how in the world did Dr. Soares write a section on this topic without once mentioning the association between Paxil and cardiac defects? In his four paragraphs of “deconstructing the myths” he doesn’t mention the association between Paxil and cardiac defects even once. This association prompted the FDA and Health Canada to issue major public health warnings.15,16 The American College of Obstetricians and Gynecologists recommends that “paroxetine (Paxil) use among pregnant women or women planning to become pregnant be avoided.”17 Furthermore, Paxil had its pregnancy class changed from C to D by GlaxoSmithKline.15 How did Dr. Soares miss all of this?

I would like to suggest that Dr. Soares consider sitting down with some of the families whose children suffer from severe heart defects after being exposed to Paxil during pregnancy. Dr. Soares might then feel differently about “deconstructing the myths” and the “safety of antidepressants” during pregnancy. I would be happy to try to arrange such a meeting if he is interested.

Most concerning, though, is the complete lack of financial disclosure information to go along with the article. At the end of the article we see only Dr. Soares’ name but no financial disclosure information. As I was reading this piece, I kept thinking to myself “Boy, this sounds like it was written by someone working for the antidepressant makers.” And, sure enough, Dr. Soares is on the Speaker’s Bureau for the Wyeth-Ayerst Pharmaceuticals, GlaxoSmithKline Pharmaceuticals, Forest Laboratories, and Pfizer Pharmaceuticals and has received honoraria as a research consultant for Sepracor, GlaxoSmithKline Pharmaceuticals, Wyeth-Ayerst Pharmaceuticals, and Neurocrine.2

This financial information is absolutely essential to the reader. Numerous studies (and common sense) have established that taking payments from drug companies influences the payee.18,19 Readers have every right to know that Dr. Soares (whose article is favorable to the continued use of antidepressants by pregnant women) is being paid by the antidepressant makers. And, you at Journal Watch know better. Journal Watch must recognize that financial disclosure statements are important to help inform the reader. In this sense, then, by not presenting Dr. Soares’ financial associations, you have misinformed your readers.

You proudly state on your website that you are “from the publishers of the New England Journal of Medicine.” The New England Journal of Medicine has a financial disclosure policy in which the author affiliations are placed with the article. What exactly is the financial disclosure policy at Journal Watch? How can you run a piece like Dr. Soares’ article, that essentially reads like a promotional piece from the antidepressant makers, without noting that the author is paid by at least six different companies?

Also, I noted that Dr. Soares’ piece is made freely available to the public on your website. Why is that? Who provides the funding to Journal Watch to make his article free to the public? Also, given that the article is free, there is a significant potential here that many people will read the article and be misinformed by the lack of financial disclosure. It is quite likely that many women suffering from depression and the providers who will care for them during pregnancy might come across this article. These readers have the right to know about Dr. Soares’ financial associations in order to make informed judgments about potential biases in his article.

Depression during pregnancy is a difficult issue and one that I deal with on a regular basis as a practicing perinatologist. Depression during pregnancy is very concerning, as is the use of antidepressants during pregnancy. There is much to be worried about with maternal and fetal exposure to these drugs. Antidepressants have not been shown to improve maternal or child outcomes during pregnancy. And in various studies antidepressant use in pregnancy has been associated with increased rates of spontaneous abortion,9 congenital malformations,4,11,12,13 preterm birth,4,7 low birthweight,3,4,7 fetal death,7 seizures,7 neonatal withdrawal syndrome,4,5,6 persistent pulmonary hypertension of the newborn (PPHN),4,8 and a possible predisposition to psychopathology.10

Pregnant women who are depressed often have an extremely difficult decision to make about whether to continue their antidepressant during pregnancy. Those pregnant women and the obstetrical providers who care for them deserve advice on their decision from experts who are unbiased and free from even the suggestion of bias that comes when the expert is paid by the antidepressant makers. Journal Watch should seek out experts on depression in pregnancy who are not being paid by the antidepressant manufacturers. If Journal Watch cannot provide that kind of unbiased expert advice, then you should, at the very least, inform your readers about which companies are paying your writers.

Reply from Dr. Ron Walls, Chair of the Editorial Advisory Board forJournal Watch

Thank you for your thoughtful letter in response to our recent piece in Journal Watch Women’s Health. As Chair of the Editorial Advisory Board for Journal Watch, I am pleased to respond with respect to your concerns about conflict of interest and disclosure. Our Journal Watch editorial leadership team has sedulously developed, implemented, and managed a comprehensive conflict-of-interest policy over the past several years. As you know, the challenge is in balancing the required expertise, reputation, and prominence of our physician editors with the fact that many of these same academic leaders receive research, education, and advisory funding from the companies whose products fall within their areas of practice or expertise.

Journal Watch’s conflict-of-interest policy (which is publicly available on our website) stipulates that investment holdings, grant funding, honoraria, or any other source of funds from any single entity in excess of $1000 per year requires disclosure. Each of our editors submits a detailed accounting of all qualifying assets or funds to the Massachusetts Medical Society and updates the report annually. The Editorial Advisory Board sees and evaluates the detail behind the reporting, including stratification of any funding source or holding over $10,000, and the qualifying entries are reported for each editor on our website. For example in Dr. Soares’ case, the reporting is at: http://womens-health.jwatch.org/misc/board_disclosures.dtl#dSoares. In addition, all holdings or funding sources over $10,000 require a narrative explanation and may require further clarification after review. We are also alert to physicians with holdings in or a relationship with companies with very small product lines, for example, a dermatologic company with a single anti-acne agent. In such cases, if we are concerned about potential bias, we may ask the physician editor to divest the holdings or end the relationship. Each Editor-in-Chief also reviews similar data for first-generation family members of each editor, forwarding potential issues to the Editorial Advisory Board. In addition, physician editors are asked to identify any potential conflict that may be in play with respect to a particular subject or story, allowing the Editor-in-Chief of each Journal Watch section to determine whether another author is assigned to write the summary.

Although we continue to revisit and constantly improve our disclosure policy, I am confident that our current process is sound and effective. Your letter pointed out to me, though, that, in our attempts to maximize the “readability” of our pages, we may have made it too difficult for some of our readers to get to the specific data on any one physician editor, even though it is available on the website. I am not aware of any prior issues that were raised in this regard, but I share your sense of the importance of easily accessible disclosure. As a result, we will be reviewing our linking process to conflict of interest information, and you will see more direct access to this in the near future.

In closing, thank you again for sending along your thoughts on both the summary and our conflict-of-interest policy. We are privileged to have a very sophisticated readership, and such discourse is invariably helpful.

Reply from Dr. Claudio Soares, Associate Editor,Journal Watch Women’s Health

I appreciate Dr. Urato’s concerns about the article’s content. The treatment of psychiatric conditions during pregnancy is a constant challenge to clinicians, patients, and their families. Healthcare professionals engaged in women’s clinical care must struggle to balance the obstetric and neurobehavioral risks inherent to untreated depression during pregnancy against the potential risks (both obstetric and teratogenic) associated with the use of psychotropic agents during this period in a woman’s life.20

As an academic psychiatrist actively engaged in clinical care and solely devoted to women’s mental health issues, I have personally witnessed the difficulty that most patients experience with accepting any help (whether medication-based or not) for the treatment of depression during pregnancy; they fear the stigma, they feel a sense of moral failure, and quite often they are reluctant to accept that this might not be a time when everything will be all right. I have also witnessed the fear of becoming pregnant felt by many women who suffer from psychiatric conditions that require maintenance treatment with psychotropic medications. They may have received little if any clarification from their physicians about their options, and often they have no access to information concerning the efficacy and reproductive safety of psychotropic agents. When such patients do become pregnant, they may discontinue their medications abruptly, thus facing severe withdrawal symptoms. The ensuing scenario commonly includes the re-emergence of depression or anxiety,21 which has been associated with poor neonatal and obstetric outcomes22,23 ultimately contributing to a heightened risk for postpartum depression, psychosis, or both.24

The editorial board of JWWH determines the topics and authors for feature articles each year at our board meeting. I was pleased to be asked to contribute this particular feature based on my subspecialty expertise. A brief summary article such as this is not meant to be the definitive word on the risks and benefits of using antidepressants during pregnancy. As pointed out by Dr. Urato and by many well-respected researchers in this field, this is a complex issue that is fraught with unquantifiable risk, regardless of the path that is chosen. My main objectives were first to make healthcare professionals and the general public aware of the potential consequences of not treating depression during pregnancy — a theme that is neglected quite often by the media; and second, to openly address some of the controversies involving the reproductive safety of these agents.

Dr. Urato seems troubled by my lack of commentary on studies that he considers essential in informing clinicians about the potential risks associated with antidepressant use during pregnancy. More specifically, he cites the retrospective studies by Oberlander et al25 and by Wen et al,26 as well as an unpublished, retrospective, prescription-based study about potential risks involving the use of paroxetine during pregnancy. Although relevant, the two published studies have been criticized by experts in the field27,28 because of their methodologic limitations: The studies are retrospective and include unconfirmed diagnoses of depression within a sample population reporting many other risk factors that could have contributed to adverse neonatal outcomes. Regarding the adverse outcomes documented in a retrospective survey involving prescriptions for paroxetine, these have been disputed by other researchers.29,30 It is true that the survey involving the use of paroxetine, despite its preliminary nature and clear methodologic limitations, generated alarm leading to prompt responses from medical societies and regulatory agencies. It is also true that in our field we have seen many cases in which controversial initial reports were followed by reanalyses of accumulating data; in some pivotal examples, the initial concerns have proven to be somewhat overplayed.31,32

Unfortunately, the existing literature concerning teratogenic and neurobehavioral adverse effects of psychotropic medications includes many unresolved confounders.33 Until more definitive studies are available, refocusing our attention on well-designed, well-controlled prospective studies seems prudent. Stowe and colleagues, for example, have recently demonstrated the importance of obtaining comprehensive records from different sources and incorporating objective laboratory methods when investigating the effects of medications during pregnancy before drawing any conclusions about the risk directly attributed to the medication per se. In a federally funded prospective study of more than 800 pregnant women with lifetime diagnoses of depression or anxiety, the investigators evaluated maternal reports of medication adherence, cigarette smoking, or drug use during pregnancy in relation to assays of urine and blood samples. Based on observed discrepancies between the self-reports and the assay results, the authors concluded that maternal self-report is not a viable proxy for actual fetal exposure.34

I strongly believe that, in conjunction with concentrated efforts to improve the research tools and methodology in this field, clinicians should improve the communication of the existing valid data about the risks of treating psychiatric conditions during pregnancy as well as the risks of not treating such conditions. We owe this effort to the families that have been devastated by cases of perinatal suicide or postpartum infanticide.28

Finally, I wish to clarify unequivocally that I have been entirely compliant at all times with Journal Watch’s comprehensive policy regarding disclosure of information relevant to any potential conflict of interest.