As a young doctor, I joked about a general warning that can still be seen in Danish package inserts for drugs: “Caution is advised during pregnancy.” What does that mean? If you take a pill, it is too late to be cautious, and if you don’t take it, you don’t need to be cautious because you will be totally safe. My joke was that caution meant placing the pill between the legs instead of swallowing it, which would also make it more difficult to become pregnant.
The authorities passed the buck. If your child is malformed, they can say that they did warn you.
Official statements that antidepressants are safe to take during pregnancy should be distrusted. No drug is safe. If drugs were safe, they would not be the leading cause of death, ahead of cardiovascular diseases and cancer. In this article, I shall explain why it is wrong to recommend or take antidepressants during pregnancy.
The Role of Serotonin in the Body
SSRIs stands for Selective Serotonin Reuptake Inhibitors, which is a misnomer. They are not selective at all. They have multiple effects throughout the body and are not directed against any chemical abnormality. People do not become depressed because they have too little serotonin in the body but mainly because they live depressing lives.
Serotonin plays a very important role for many processes in the body, also in many primitive organisms. It is usually a very bad idea to change the blood level of a chemical that has proved so useful during evolution.
Foetal development is a delicate process that can easily go wrong, which is why we tell pregnant women to avoid alcohol. A priori, we would expect any substance that affects serotonin levels to be harmful because serotonin is essential for foetal development. This is basic biology, but we live in a world dominated by financial interests, which is why many pregnant women take antidepressants during pregnancy.
How a Drug Company Fooled the Drug Regulators
The first SSRI approved for use in children was fluoxetine from Eli Lilly. It should never have been approved. When psychiatrist David Healy and I reviewed the confidential internal study reports for the two trials that led to approval of fluoxetine for children with depression, we found that fluoxetine is unsafe and ineffective. In the first trial, the investigators had omitted two suicide attempts on fluoxetine in their published paper, and many of the 48 children on the drug experienced restlessness and had nightmares, which increase the risk of suicide and violence.
In the other trial, one child was severely harmed for every 10 children treated with fluoxetine. Fluoxetine increased the QTc interval on the ECG (P = 0.02), which increases the risk of sudden death, increased serum cholesterol, and was an effective growth inhibitor, reducing the increases in height and weight over just 19 weeks by 1.0 cm and 1.1 kg, respectively (P = 0.008 for both).
The public does not have access to animal experiments with drugs because the drug companies know it would be bad for business if people saw the data. When I got access to Merck’s animal studies for their HPV vaccine Gardasil in a US lawsuit where I was an expert witness, I saw that the data supported what the patients had reported: Gardasil can cause serious neurological harms and the vaccine adjuvant is also harmful. However, drug regulators all over the world have declared that both the adjuvant and Gardasil are safe.
The European Medicines Agency (EMA) had serious concerns about approving fluoxetine for use in children, which is clear in an 86-page document about animal studies from August 2005 that is nowhere to be found on the Internet: “Prozac Paediatric Indication. Arbitration Procedure No: EMEA/H/A-6(12)/671. Lilly Response to Questions from EMEA in Document EMEA/CHMP/175191/05”. I have uploaded this document in the public interest. It illustrates the extent to which drug companies are willing to bend the truth for an economic gain and to harm children, including killing some of them, because antidepressants double suicides.
The EMA deemed Lilly’s data insufficient and asked for further studies and explanations. They noted that a study on young rats showed a “very unfavourable profile” of the drug, which included severe effects on body weight gain, sexual maturation in males and females, testes, skeletal muscles, sperm concentration, and reproductive performances, which appeared with no or low safety margins. Moreover, the effects on testes were not reversible.
When EMA noted that S-norfluoxetine, the active metabolite of fluoxetine, had caused testicular degeneration in 6 of 15 rats, Lilly replied that testicular effects were observed in the rat and the mouse, but not in the dog!
About the inhibition of skeletal growth, Lilly replied: “Fundamental differences in bone physiology between rodents and humans (Kimmel 1996) limit the ability of rodent studies to accurately predict the response in the human skeleton. Human skeletal health is monitorable in the clinic and remains a focus of clinical investigation.”
This is bullshit. Inhibition of growth is a fact and no amount of monitoring can prevent this.
The greatest concern related to growth inhibition with brain-active substances is of course that they could also cause irreversible brain injuries. A large Dutch study found that prenatal SSRI use was associated with less cerebral gray matter in children that persisted a decade later and greater increases in volumes of the amygdala and fusiform gyrus that did not persist until early adolescence.
The EMA asked Lilly to take into account “all available non-clinical and clinical data, discuss whether potential effects on brain development and function are adequately addressed, or whether further data can be obtained.”
EMA’s concern should have killed fluoxetine but this is not how drug regulation works. Lilly replied that “Lilly considers the current nonclinical data package acceptable for the assessment of potential effects on brain development and function.” Well, the EMA had just told Lilly that they had a different view, but Lilly didn’t reply at all. They wrote that a study had reported long-lasting behavioural changes after fluoxetine treatment in mice but considered the clinical relevance of these findings for children “questionable.”
Lilly noted that the mice were administered saline or fluoxetine from postnatal day 4 through 21 and that this period of brain development is considered equivalent to a human third trimester foetus through a 2-year-old child, adding that this “does not replicate the recommended age range for fluoxetine administration.”
This is unbelievable. A mouse study showing irreversible changes in behaviour long after the drug is stopped was dismissed with a nonsense argument.
The EMA noted that statistically significant delayed growth and delayed puberty had been reported in placebo-controlled trials in children. In addition, there had been spontaneous reports to Lilly of growth retardation, delay in puberty, menstrual disorders, and sexual
disorders: “These data are representing a signal which cannot be ignored.”
But Lilly ignored all the harms signals and did what drug companies always do when presented with damning evidence of serious harms of their drug: They promise to do additional studies.
Spineless drug regulators approve harmful drugs on that condition, even though they know that it is unlikely that relevant studies are ever carried out. A systematic review of 117 novel drugs the FDA had approved based on limited evidence showed that for one-third of the drugs, a post-approval study was not carried out. Moreover, most of the studies that were carried out were inadequate: 70% used active comparators and 89% used surrogate outcomes as primary endpoints.
I have not seen any relevant fluoxetine studies of the type EMA requested in the medical literature and Lilly’s explanations were pathetic.
Lilly noted that “Protocol HCLT is being developed as a Phase IV commitment with the United States Food and Drug Administration (FDA) to investigate the effects of fluoxetine
treatment on height and weight in paediatric patients.”
A Phase IV trial is a study conducted after a drug has been approved and after it has been marketed. Lilly argued that there would be “considerable recruitment challenges arising out of recent public discussion regarding the use of SSRIs in the child and adolescent population. This concern over recruitment challenges makes the manner and timeframe for this study uncertain.”
Lilly gave the drug regulator the fuck finger again. Their bullshit is short for: “Rest assured that we will never do the study.” It cannot possibly be difficult to recruit children for a study that shall only measure their height and weight and if it is difficult because of safety concerns, the drug should not be used. Moreover, the study is superfluous, as we already knew that fluoxetine is a strong growth inhibitor. When I searched on the name of the protocol for the study, B1Y-MC-HCLT, there was nothing on the Internet.
Lilly conveniently said nothing about the harms related to development of puberty and to people’s sex life. In half the patients, the sex life is impaired, and when I lectured in Australia in 2015, a paediatrician told me about three boys who had all attempted suicide because they could not get an erection the first time they tried to have sex. Some people have committed suicide because the sexual harms can become permanent.
Wearing off EMA’s attack, Lilly even argued that they had agreed with the regulator to insert this statement in the Summary of Product Characterists: “In addition, long-term safety data in children and adolescents concerning growth, maturation, and cognitive and behavioural development are lacking.” This is seriously misleading. Short-term data show substantial growth inhibition and as most children take SSRIs for many years, they will become irreversibly growth-retarded.
In October 2005, two months after Lilly had responded to the EMA, the rapporteurs issued a 39-page assessment report, which concluded:
“It is not recommended to grant an indication to fluoxetine for the treatment of depression in children and adolescents because the benefit/risk balance in the claimed indication is deemed negative.
Concerns about safety issues were not resolved, specifically concerns about suicide related behaviours, including suicide attempt and suicidal ideation, and, from non-clinical data, about the effect on growth, sexual maturation, cognitive and emotional development. The limited evidence concerning longterm safety is a concern as well, especially given these safety signals.”
It is tragic and a huge mistake that Lilly nonetheless succeeded in having fluoxetine approved for use in children, which paved the way for approval of other harmful SSRIs.
Harms When SSRIs Are Used in Pregnancy
In 2012, obstetrician Adam Urato and colleagues published a study that, like many earlier studies, found many harms of SSRIs including elevated risk of miscarriage, preterm birth, neonatal health complications, and possible long-term behavioural abnormalities, including autism. There is also a strong signal of congenital abnormalities, which in 2005 prompted the FDA to ask GlaxoSmithKline to change the labelling for paroxetine so that it showed that there is a demonstrated risk to the foetus.
An increased risk of pregnancy-induced hypertension and preeclampsia, and respiratory distress and withdrawal symptoms in the newborn have also been demonstrated, and neonatal death and stillbirth are more common, both in animal and in human studies.
Urato has told me that the most consistent findings related to growth are lower birth weight and smaller head circumference in the SSRI exposed groups and that the problem with many large epidemiological studies is that they define exposure as having a prescription, which could be just one or two. The studies where the women have taken the SSRI through the whole pregnancy are much more likely to show harms.
Even though around 400,000 children in the US are born each year by mothers who took depression pills during pregnancy, it took another 13 years and a new administration before the FDA was finally asked to do something about it.
On July 21, 2025, the FDA convened a hearing about whether a stronger warning was needed for using antidepressants in pregnancy.
I was shocked to learn that there are few warnings in FDA-approved package inserts. In Denmark, we have extensive warnings. For example, the package insert for citalopram warns that the drug can increase the risk of a serious condition, Persistent Pulmonary Hypertension of the Newborn. Other serious side effects or withdrawal symptoms in the newborn baby include breathing problems, bluish skin/lips, irregular breathing with pauses in breathing, temperature fluctuations, seizures, lethargy (sleep-like drowsiness), difficulty sleeping, feeding problems, vomiting, low blood sugar, stiff or floppy muscles, abnormally increased reflexes, tremors, extreme nervousness or nervous shaking, irritability, persistent crying, and drowsiness. “If your baby has any of these symptoms, contact your doctor immediately.” Moreover, taking citalopram near the end of pregnancy may increase the risk of heavy vaginal bleeding soon after birth.
The US Is Not Denmark
The United States is a strange country when it comes to healthcare. There were howls of outrage after the FDA meeting. The American Psychiatric Association and other medical associations, most notably the American College of Obstetricians and Gynecologists, issued press releases telling the public that the panel was biased and that the real risk during pregnancy was untreated mental illness.
These medical organisations asserted that the increased risk of adverse outcomes for children born to depressed mothers is due to the illness and not the drug, and that there was plenty of evidence that antidepressants were a helpful and even life-saving treatment for maternal depression.
There is no room for diplomacy or soft talk here. The organisations were lying in the most brutal way. Antidepressants don’t save lives, they take lives – many lives. They are an important reason why psychiatric drugs are the third leading cause of death (mostly because elderly people lose balance and break their hip) and they double suicides, with no age limit. Moreover, they have no clinically relevant effects. The effect in placebo-controlled trials is way below the minimally relevant effect.
After the FDA meeting, Urato said in an interview that if we expose neurons to citalopram, there is a reduction in neurites (branches from a neuron); animal models show a disruption of brain development; ultrasound studies show more jittery behaviour, more agitation, and less quiescence in the foetus; and 12 consecutive MRI studies show the impact from these medications on the brain. And when the children grow up, there are impacts on speech and language development, and more diagnoses such as depression, anxiety, autism, and ADHD.
I shall not go into detail about how convincingly all these effects have been demonstrated, as it is not necessary. It suffices for the ensuing discussion to know that serious harms have been demonstrated in high-quality research.
The American College of Obstetricians and Gynecologists were seriously dishonest. They called the FDA panel “alarmingly unbalanced” with many outlandish and unfounded claims and criticised it for not wanting to prevent the potentially devastating effects of anxiety and depression when left untreated during pregnancy.
However, psychotherapy is not leaving patients untreated. It has an enduring effect that clearly outperforms pharmacotherapy in the long run, and it can halve suicide attempts whereas SSRIs double suicides.
The College claimed that robust evidence has shown that SSRIs are safe in pregnancy, which is also a huge lie, and that untreated depression in pregnancy can put patients at risk for substance abuse, preterm birth, preeclampsia, limited engagement in medical care and self-care, low birth weight, impaired attachment with their infant, and even suicide.
So, we are expected to believe that drugs that cause suicide can prevent suicide. It cannot be more absurd than this. In addition, there is no reliable evidence that depression can cause these untoward outcomes. Psychiatrist Joanna Moncrieff was also interviewed and she pointed out that some of the studies that found an association between being depressed and adverse pregnancy outcomes did not even consider whether people were taking depression pills, which most of them undoubtedly did. Apart from this, depression is associated with all sorts of things that are very likely to affect foetal development, e.g. socioeconomic deprivation, alcohol use, and smoking. It is not possible to adjust reliably for such confounding factors.
The American Psychiatric Association talked about suicide being a leading cause of maternal death implying that antidepressants protect against suicide, which is an inexcusable lie for a psychiatric association to make. Ironically, the Association accused the panel of biased interpretations and the use of opinion, rather than the years of research on antidepressant medications, and said this would exacerbate stigma.
Psychiatrists often talk about stigma when they have no arguments. The Association also said that the dissemination of inaccurate and unbalanced information by a federally sanctioned public panel has the potential to cause harm and can undermine public confidence in mental health treatment.
A novel writer could not have made this up. The reader would think it could never happen. It is just too absurd. But this is psychiatry, which is why I entitled my most recent, freely available book, Is psychiatry a crime against humanity? Psychiatrists push children and young women to suicide with happy pills. Can anything be worse than this?
Brain-Dead Media and Industry-Friendly Medical Journals
As usual, the media acted as the useful idiots for those opposed to any criticisms of SSRIs. Story after story came out with the same basic structure, in the New York Times, NBC News, National Public Radio, STAT News, and elsewhere. They quoted “experts” who criticised the panel for “misinformation,” and they used ad hominem arguments, and emphasised the risks of depression, the supposed safety of the drugs, and the purported benefits. As a typical example, the three key points in the Los Angeles Times were:
- An FDA panel recently attacked SSRIs, a class of antidepressants that RFK, Jr. has targeted in the past.
- Doctors say the panel — comprising mostly critics of antidepressant use — spread misinformation about the drugs’ use in pregnancy.
- The risks of not treating depression in pregnancy far outweigh those of SSRIs, healthcare providers said.
The framing is unmistakable. Since Robert F. Kennedy, Jr. has criticised depression pills, surely the FDA panel must be wrong. Right?
There is absolutely zero evidence for the claim that “The risks of not treating depression in pregnancy far outweigh those of SSRIs.” In fact, given the evidence we have, the opposite is true. And the journalist failed to consider psychotherapy instead of drugs even though the panel had mentioned non-drug modalities.
It is not surprising that the NEJM, nicknamed the New England Journal of Medicalisation because of its lucrative role as a lapdog for the drug industry, also got it wrong. In an opinion piece written by three researchers, two of whom reported significant financial ties to the pharmaceutical industry, the authors tried to argue that better statistical adjustment for confounders means that it is the depression and not the drugs that cause the problems.
As Urato argued, the adjustment for confounding in the studies often appears to be done in a mysterious way, in which the excess risks in the SSRI group are made to suddenly disappear and assigned to being due to the disease. I agree with him that “It’s sort of statistical mumbo jumbo,” and this magic cannot dismiss all the research that has found serious harms of SSRIs.
Robert Whitaker has pointed out the sobering fact that there is no confounding in animal studies, as the rats and mice are not depressed, and that these studies have documented numerous serious harms. For example, mice exposed prenatally to fluoxetine had abnormal emotional behaviour as adults, compatible with anxiety and depression.
Whitaker has provided a comprehensive review of animal and human studies and has rebutted the false claims from medical organisations. He furthermore noted that assessing the merits of a drug intervention requires weighing the harms against the benefits, but for the unborn child, foetal exposure to SSRIs only provides harms.
Urato, in his remarks at the FDA hearing, put it into a haunting perspective: “Never before in human history have we chemically altered developing babies like this, especially the developing fetal brain, and this is happening without any real public warning. That must end.”
Conclusions
It is very easy to provide firm and evidence-based advice to pregnant women and women intending to become pregnant. They should not talk to their doctor, as some guidelines recommend, to find out if the benefits of SSRIs outweigh the harms. Their doctor would not know this and would likely just echo what the industry has told the professional organisations they should believe.
Moncrieff and Urato explain that since the thalidomide tragedy, society has generally taken the view that subjecting the developing foetus to foreign chemicals is risky and best avoided if possible. Yet medical leaders are cavalier about the potential risks of antidepressants. Their attitude illustrates a deep-seated attachment to antidepressants that seems to blind many doctors to their harmful effects.
Antidepressants should be avoided, especially for children, and if a pregnant woman or a woman hoping to become pregnant is taking one, she should look up a professional who can help her withdraw from it, slowly and safely. I have a list of people who are willing to help women this way, no matter where they live. I have also published a guidebook to safe psychiatric drug withdrawal.
Join the conversation:
Published under a Creative Commons Attribution 4.0 International License
For reprints, please set the canonical link back to the original Brownstone Institute Article and Author.
