I recently wrote an article, which appears in the November issue of Elle, looking into the aggressive marketing of antidepressants to pregnant women by some doctors and psychiatrists — despite a growing body of scientific research that one popular type of antidepressant, selective serotonin reuptake inhibitors or SSRIs, is associated with adverse birth outcomes. Working with The Investigative Fund, I followed the case of Lyam Kilker, who went into cardiac arrest when he was only six weeks old and eventually had two open-heart surgeries. His mother had taken the commonly prescribed antidepressant Paxil while she was pregnant.

Such stories may soon become more common.

Because of Paxil’s risks, in 2005 the FDA had requested that the drug’s manufacturer, GlaxoSmithKline, change the drug’s pregnancy-use classification from Category C (“potential benefits may warrant use of the drug in pregnant women despite potential risks”) to Category D (“positive evidence of human fetal risk”) — a clear warning for doctors. But next year, as a result of pharmaceutical industry lobbying, the letter categories will be dropped altogether. Instead of seeing a Category D on Paxil’s label, doctors will find a long summary of the medical literature.

Critics of the new labeling fear that doctors won’t read the detailed summaries, and the FDA seems to acknowledge that the new system may lead doctors to more readily prescribe risky medications to pregnant women. But Melissa Tassinari, PhD, a team leader in pediatric and maternal health at the FDA’s Center for Drug Evaluation and Research, defended the change — while acknowledging that labeling is “one of the huge debates.”

“Right now making a presumptive decision based on a letter isn’t serving [patients] well,” she said. But note that Tassinari worked for Pfizer, the makers of Zoloft, a huge-selling SSRI, for 18 years on global regulatory affairs. Such cross-pollination between government regulators and the industries they’re charged with policing is increasingly common. And this is only the beginning of how drug-makers influence the decisions facing pregnant women.

For example, one study often cited by drug industry reps and by physicians who advocate for drug treatment of depression in pregnant women shows that the rate of relapse into depression among pregnant women who discontinued their medications was five times higher than those who remained on them. But the Journal of the American Medical Association, which published the study, was forced to print a correction after it was revealed that the study’s authors, including lead author Lee Cohen, MD, a psychiatry professor at Harvard Medical School, had multiple undisclosed conflicts of interest, having received grants from the likes of Eli Lilly, maker of Prozac, consulted for the likes of GlaxoSmithKline, the makers of Paxil, and a paid speaker for Pfizer, the maker of Zoloft, and many other major drug firms.

Cohen is just one example; high-profile physicians like him are routinely paid by pharmaceutical companies to give talks promoting their drugs, act as consultants, conduct clinical trials, or sit on advisory boards for their drugs. (This is when drug companies recruit physicians, often to a hotel or resort, to give their impressions of promotional materials for their medicines and offer advice about how best to influence consumers and other physicians.) Drug companies refer to them as “opinion leaders,” and the most influential physicians, often found at top universities and hospitals, are called key opinion leaders, or KOLs. Cohen is a KOL for such major drug companies as Pfizer, GlaxoSmithKline, Eli Lilly, AstraZeneca, Berlex, Wyeth, Janssen, and Forest Labs.

Last July, the Department of Justice reached a $3 billion settlement against GlaxoSmithKline, after finding that the firm had, among other fraudulent practices, skewed scientific data in favor of drug safety and efficacy — something that is often accomplished by means of KOLs. Both skewing data and failing to disclose potential conflicts of interest — a common problem with KOLs — are considered scientific misconduct.

Researcher Kara Driscoll, MD, a psychiatrist at Northwestern Memorial Hospital in Chicago who specializes in treating depressed expectant mothers, insists that untreated depression is dangerous to unborn babies, and screens all her pregnant patients, regardless of whether they present with symptoms of depression or not.

But the evidence is thin. The studies Driscoll cites, which show a higher incidence of low birth weight and pre-term birth among mothers with depression, do not control for whether or not the woman was taking an SSRI, meaning they don’t establish whether the poor outcomes were due to the depression or the medication. In fact, the only study controlling for use of SSRIs found a significant increase in low birth weight and pre-term birth among women who took the drugs compared with women with untreated depression.

Driscoll herself doesn’t appear to have any conflicts of interest, but she works directly under Katherine Wisner, who has worked for Pfizer, GlaxoSmithKline and Eli Lilly. And Wisner’s conflicts of interest are all too common, even among prominent experts in the field of fetal health.

Gideon Koren, MD, a toxicologist and pediatrician who runs Motherisk in Toronto, the world’s largest research and counseling program for pregnant women, told me he sits on advisory boards for drug companies, and his CV indicates that he has received vast amounts of research funding from them. Koren would not give specifics about the amount of money he receives from drug companies for sitting on advisory boards and other non-research activities, and when queried about how much of Motherisk’s funding is derived from the pharmaceutical industry, he said, “Not enough.”

“This hospital has hundreds of — not hundreds but in excess of $80 million a year from industry from different countries,” Koren said. “And no one can tell us not to publish. [The companies] have a right to see the paper a month before and make some comments but we will not change results.”

Koren had been criticized for his close ties with pharmaceutical companies in the past. In the late 1990s, he was involved in a scandal after Nancy Olivieri, his research partner on a project to study a drug produced by Canada’s largest generic drug manufacturer, Apotex, exposed him for distorting data in favor of the firm. According to a probe by the College of Physicians and Surgeons of Ontario Olivieri’s colleagues received anonymous hate mail about her, which was later tied to Koren through DNA evidence from saliva on an envelope. The incident resulted in Koren being suspended by the College of Physicians and Surgeons.

Yet in 2007, GlaxoSmithKline involved Koren in performing a meta-analysis of Paxil studies that the company could use to bolster its defense in a birth defects case that was about to go to trial. (The plaintiffs’ attorney, Sean Tracey, filed a motion to disqualify Koren, and GSK later settled the case.)

“I hope to get more from industry, much more than we do,” Koren told me. “Because it doesn’t do anything. It doesn’t affect the way we do science.”

Another prominent researcher in the field riven by conflicts of interest is Anthony Scialli, MD, who studies the causes of birth defects and runs Reprotox, a widely used database that compiles the effects of drug exposures in pregnancy. “I don’t have any connections to the drug industries,” Scialli told me for the Elle piece. “I don’t do a lot of prescribing because I work in labor and delivery, but I do some and I only prescribe generic drugs, which are not promoted, and I don’t meet with drug reps and I don’t read drug advertising and I don’t go to pharmaceutical company-sponsored programs. So I suppose I could be influenced but it would be kind of hard. They’d probably have to put a billboard up outside my apartment.”

Perhaps not. Scialli is actually an expert witness for GlaxoSmithKline, and has given two or three depositions per year for the past quarter century for various drug and chemical companies, including in birth defects cases. In a majority of these depositions, he testified that the drug or pesticide in question was not the cause of a child’s birth defects. Not once did he lay the blame on fetal exposure to a drug. Though Scialli no longer gives talks on behalf of drug companies, he was a key opinion leader while heading the residency program at Georgetown University’s Medical School.

Scialli was also vice-president of Tetra Tech Sciences, a firm that consults for more than 50 corporations in the chemical and pharmaceutical industries. “Litigation support” and “environmental audits” are two of the first services listed on its web site. In fact, Scialli himself has consulted for at least two-dozen major pharmaceutical and chemical companies, yet he maintains that the work he does for these industries does not influence his analyses of their drugs. (Shortly after my article appeared in Elle, the website for Tetra Tech Sciences,www.sciences.com, was taken down.)

Such conflicts of interest exist on the clinical level as well. James Ballenger, MD, was chair of the Medical University of South Carolina’s Department of Psychiatry and now runs a private practice in Charleston, South Carolina. He believes that the benefits of prescribing antidepressants outweigh the risks during pregnancy.

“The downsides to not treating and leaving the woman depressed — there’s risk of suicide and a hard time functioning if she has other children,” Ballenger told me.

Ballenger said he sometimes prescribes benzodiazepines like Xanax or Ativan in addition to SSRIs. The FDA has placed both drugs in Pregnancy Category D, so I asked him about the birth defects associated with these drugs — and again he insisted that risks were worth it. “The average person really doesn’t understand the importance of these medicines in people’s lives,” he said. “Seventy-five percent of depressed people in America don’t get adequate treatment, so the difficulty is really on the other foot… The truth is that far more people should be on antidepressants who are not.”

It may come as no surprise that while he was chair of the MUSC Department of Psychiatry, Ballenger received grants from pharmaceutical companies to conduct clinical trials of their drugs. He did studies on GlaxoSmithKline drugs which were funded by the drug maker, did work for Pfizer, and has been a consultant to Shire, which manufactures Adderall, a psychostimulant (though I was unable to determine the exact nature of his work for Pfizer or Shire and whether it was during his time at MUSC). But he maintains that none of this has affected the way he designed or managed the studies. “We’re essentially not beholden to the drug companies, and if they tried to influence what you published there’d be a gigantic outcry from the researchers,” he said.

Yet Ballenger didn’t push back when GlaxoSmithKline (then SmithKline Beecham) canceled his long-term panic disorder study in 2000. He had been working on the trial for nearly a year, trying to determine whether the rate of relapse into depression was higher among women who stopped taking Paxil compared with women who continued taking the drug. He had already gathered enough data to see something unexpected — severe withdrawal effects, which drug companies call “discontinuation syndrome” — after participants went off Paxil.

But email correspondence between a SmithKline Beecham official and a medical writer Ballenger was collaborating with on his manuscript, Sally Laden, are revealing. Daniel Burnham of SmithKline Beecham wrote to Laden, who was often contracted to ghost-write studies and medical journal articles, “The issue of discontinuation sx [side effects] vs. relapse is obviously a concern of the J Clinical Psychiatry reviewers… Thus we have decided to terminate further work on this manuscript.”

Laden responded, “We understand your reasons for cancelling this project. There are some data that no amount of spin will fix, and these certainly fall into this category.”

Laden went on to e-mail her bosses at the medical education company Scientific Therapeutics Information, Inc. which provided medical ghostwriters for SmithKline Beecham, “Yes, Virginia, there is a God. SB cancelled our project 1059 (long term panic disorder study). Reason: the side effect data was terribly unfavorable to our favorite antidepressant. And we hate when that happens!” (See: Emails between Sally Laden and SmithKline Beecham’s Daniel Burnham, December 2000 [PDF].)

Discussion about withdrawal was already brewing in medical circles related to Paxil and SmithKline Beecham was not eager for the study to confirm such suspicions. “I think they were worried that the data, um, showed side effects,” Ballenger told me. “What that study in retrospect probably showed for the first time was that there’s withdrawal from the medicine…your body might miss it.”

The overwhelming incidence of “discontinuation syndrome” should have been enough to prompt further analysis, and certainly publication of the study, but Ballenger never even put it up for peer-review. I inquired about this discrepancy, since he had insisted that he would never fail to publish a study because a company didn’t like the results.

“In my own personal experience this is the only one that has ever come close to that and there were other reasons,” Ballenger said, citing “so many holes in the data that we couldn’t do it.” When I inquired about what that meant, he said that he had to go and could discuss it further the next day. He did not respond to any of my follow-up calls the next day or the following week. Perhaps these holes were substantial, or there were other weaknesses in his findings, but he never articulated them to me.

Four years later, in 2004, Ballenger published a Paxil study in the Journal of Clinical Psychiatry that was extremely advantageous to GlaxoSmithKline, showing an increase in remission rates following treatment with Paxil. There was no mention of the withdrawal effects.

Researchers have since confirmed that about one-third of infants born to women who take SSRIs during pregnancy suffer what drug companies call “neonatal abstinence syndrome,” more commonly known as withdrawal. Symptoms include tremors, excessive crying, fever, rapid breathing, vomiting, problems feeding and sleeping, and sometimes seizures.

I spoke with Thomas Higerd, PhD, a professor emeritus at MUSC who has spent half of his 40 years there as chair of the Conflict of Interest Review Committee, which reviews conflicts of interest in research and other university projects. Higerd said he was unaware of Ballenger’s decision not to publish the panic disorder study or its cancellation by SmithKline. “I can’t imagine us ever signing off on something like that if we knew about it,” he said. In instances of conflict, he added, “the university can take it all the way up to dismissal.”

Layton McCurdy, MD, MUSC dean emeritus and distinguished university professor of psychiatry, said that while the university expects researchers to help cover their salaries with external grants as Ballenger had, the process is tightly controlled. “We have committees that oversee this,” he explained. “You have to get permission for almost any time you want to blink your eyes.” McCurdy, who was dean of the College of Medicine when Ballenger taught there, said he too was not informed about Ballenger’s decision not to publish his study. Had he been aware of Ballenger’s decision, and found it to be the result of pressure from SmithKline Beecham, it might have merited Ballenger being quarantined from research, he said, “which is a very strong blow.” In Ballenger’s case, no censure ever occurred.

John Gartner, PhD, a clinical psychologist who also teaches psychiatry residents at Johns Hopkins University Medical School, has become a vocal critic of the financial ties that a large majority of psychiatrists have with pharmaceutical companies. “What’s funny is I started out just noticing in my practice that a lot of my patients were overmedicated and were feeling sedated,” he told me. “They were feeling like their psychiatrist was just kind of pushing the medication on them.”

Gartner began to encourage his patients to lower their doses or wean themselves off of antidepressants and anti-anxiety drugs that were causing harmful side effects. As a result, some psychiatrists suddenly refused to see his patients. (Psychologists like Gartner practice talk therapy, but do not prescribe medications, and most of his patients also see a psychiatrist for the sole purpose of filling prescriptions. In fact, according to a 2005 government study, just 11 percent of psychiatrists provide talk therapy to all patients, a figure that has been falling for years and has likely fallen even more since.) It became a regular occurrence, he said, after he wrote a blog post for Psychology Today in which he questioned drug industry influence on medical practice. The issue of key opinion leaders, he says, is of critical importance.

“If we can’t stop them from faking the science, the battle is lost,” Gartner said. “How much scientific misconduct that hurts the public good and the public health do you need to be exposed doing, before you’re not given the mantle of a university position to do it from?”

Updated 11/17/12.