Antidepressant withdrawal symptoms may be more common and more severe than some studies suggest

A newly published study in Psychological Medicine raises questions about a widely cited 2024 meta-analysis that downplayed the frequency and severity of antidepressant withdrawal symptoms. The new reanalysis suggests that when studies use systematic and appropriate methods to assess withdrawal effects, more than half of patients report experiencing symptoms—contradicting earlier claims that only about 15% are affected. The findings challenge current perceptions about the risks associated with stopping antidepressant use, particularly after short-term treatment.

Antidepressants are among the most commonly prescribed psychiatric medications in high-income countries, and many patients remain on them for extended periods. Yet questions remain about what happens when people stop taking them.

In 2024, a large meta-analysis published in The Lancet Psychiatry sought to clarify just how often patients experience discontinuation symptoms after stopping antidepressants. The authors of that review, including Jonathan Henssler and colleagues based in Germany, aimed to produce what they described as a more comprehensive estimate of the incidence and severity of withdrawal effects.

Their study found that about 15% of people who stop taking antidepressants experience one or more withdrawal symptoms that can be directly attributed to the medication. These include issues such as dizziness, nausea, insomnia, and irritability. The researchers also reported that only about 3% of patients experienced what they considered severe symptoms—those intense enough to lead to dropout from a study or the resumption of antidepressant treatment.

To conduct the analysis, Henssler and colleagues systematically reviewed data from 79 studies published between 1961 and 2019, encompassing more than 21,000 participants. Their inclusion criteria were broad: the team accepted both randomized and non-randomized trials, as well as observational studies, as long as they assessed withdrawal symptoms following the discontinuation of antidepressants or placebo. They also analyzed withdrawal symptoms reported after stopping placebos in controlled trials, which they used as a baseline to subtract non-specific symptoms that might be misattributed to withdrawal.

Of the participants, more than 16,000 had discontinued antidepressant treatment, while nearly 4,500 had discontinued placebo. Across all groups, the researchers estimated that roughly 31% of patients experienced some form of withdrawal symptom after stopping antidepressants.

After accounting for the incidence of symptoms in placebo groups—estimated at about 17%—the authors concluded that roughly 15% of the symptoms could be directly tied to antidepressant withdrawal. They also identified some antidepressants, such as imipramine, paroxetine, and venlafaxine, as carrying a relatively higher risk of withdrawal symptoms.

The study attracted attention for its size and its conclusion that withdrawal effects, while real, were not particularly frequent or severe. But critics soon raised concerns about how the data had been collected and analyzed. In particular, several researchers questioned whether the methods used in many of the included studies were sufficient to detect withdrawal symptoms accurately.

That skepticism prompted the publication of a new study in Psychological Medicine, led by Joanna Moncrieff of University College London. Along with a team of researchers from Denmark, Switzerland, Austria, and the United Kingdom, Moncrieff sought to examine the data used in the original review and determine whether its conclusions were supported by more rigorous assessments.

“We did this research because we are concerned that some psychiatrists tend to minimize the adverse effects associated with antidepressants, particularly their dependence-inducing effects,” explained Moncrieff, a professor of critical and social psychiatry and author of Chemically Imbalanced. “This may be because psychiatrists see the medical treatment of depression as central to the enterprise of psychiatry, and therefore feel the need to defend the reputation of antidepressants come what may.”

To assess the reliability of the original review, Moncrieff and her colleagues reviewed all 62 study cohorts that Henssler’s team had used to estimate the incidence of withdrawal symptoms. They examined whether those studies had used systematic and specific methods to measure symptoms, or whether they had relied on more casual observations—such as spontaneous adverse event reporting or non-specific clinician notes.

The researchers found that only five of the 62 studies—less than 10%—had measured withdrawal symptoms using structured questionnaires or tools specifically designed to assess the kinds of symptoms typically reported during antidepressant discontinuation. These tools, such as the Discontinuation-Emergent Signs and Symptoms (DESS) checklist, capture a range of common experiences including dizziness, emotional instability, and “brain zaps.” In contrast, most of the studies relied on vague or inconsistent methods, including open-ended questions, clinician impressions, or even retrospective chart reviews.

“The majority of studies were not proper studies of withdrawal but had looked at withdrawal only casually and incidentally,” Moncrieff told PsyPost. “In fact, very few of the studies had used a proper measure of withdrawal symptoms.”

The new team also raised concerns about other methodological features of the included studies. Many of the trials were short in duration, sometimes observing patients for only two weeks after discontinuation—a period that may not capture delayed withdrawal symptoms. The average length of antidepressant use before stopping was also less than six months in most studies, which could reduce the likelihood of detecting symptoms that become more likely with long-term use.

Another issue involved the way that withdrawal symptoms were sometimes misclassified as a return of the underlying mental health condition, especially in studies where symptoms of depression or anxiety were being monitored alongside withdrawal effects. This made it harder to distinguish between a relapse of the original disorder and genuine discontinuation symptoms.

The researchers then conducted their own meta-analysis, including only those five studies that had systematically measured withdrawal symptoms. These studies involved 601 participants, most of whom had used antidepressants for three months or less. They found that 55% of participants experienced at least one withdrawal symptom after stopping their medication. When they excluded one study in which patients did not fully stop their medication, the estimate rose to 61%. Excluding a trial of agomelatine, which is known to have lower dependence potential, pushed the estimate to 69%.

“We found that of the 62 studies included in the original review, only 5 used a systematic measure of antidepressant withdrawal symptoms,” Moncrieff explained. “Four of these involved people who had only used antidepressants short-term (up to 12 weeks). When we looked at the findings of these studies specifically, between 55% and 69% of participants reported experiencing withdrawal symptoms after stopping their antidepressant.”

“So overall, we found that the majority of evidence on antidepressant withdrawal is of poor quality and does not involve people who have taken antidepressants for long periods (those who are most at risk). The evidence that exists, however, suggests antidepressants may induce withdrawal reactions quite frequently, even after short-term use.”

These findings suggest that when withdrawal is measured systematically, a much larger proportion of patients report symptoms than the earlier review had indicated. The authors emphasize that their estimates do not subtract rates of withdrawal-like symptoms in placebo groups, but argue that the types and severity of symptoms reported after antidepressant discontinuation are not equivalent to those seen with placebo.

The new study concludes that the original 2024 review underestimates the incidence and severity of antidepressant withdrawal, in large part because of limitations in the data it relied on. These included inadequate assessment methods, short follow-up periods, and the frequent misclassification of symptoms. The researchers argue that better-designed studies are needed to understand the full scope of the problem—particularly for individuals who have used antidepressants for longer than a few months.

While both reviews agree that withdrawal symptoms do occur, they offer very different pictures of how often they happen and how disruptive they might be. One suggests that withdrawal is relatively rare and mild; the other indicates that it may be common, even after short-term use, and that current clinical guidelines may not reflect the full reality experienced by patients.

Another recent study, published in JAMA Psychiatry, which involved a meta-analysis of 49 randomized clinical trials, found that people who stop taking antidepressants tend to experience only one additional discontinuation symptom, on average, compared to those who discontinue placebo or remain on the medication. The most common symptom within the first two weeks was dizziness, but the overall number and intensity of symptoms remained below the threshold typically used to define clinically significant withdrawal, and no increase in depressive symptoms was observed.

However, this newer study has also faced critics, who argue that it downplayed the risks of antidepressant withdrawal by relying on short-term, industry-funded studies, failing to adequately account for the more severe and long-lasting symptoms experienced by long-term users and potentially repeating past mistakes that delayed recognition of the issue.

“Another review of antidepressant withdrawal was published recently, which also used short-term trials and poor data to come to the conclusion that antidepressant withdrawal is not a significant problem,” Moncrieff said. “However, in fact, the studies included in this latest review revealed that antidepressant withdrawal does exist, even after short-term use.”

“Evidence from surveys and thousands of patient reports points to the fact that antidepressant withdrawal is a significant clinical problem. Attempts to minimize it are worrying because it means that the many people who suffer from severe withdrawal reactions will not receive the understanding and support they need.”

As the debate continues, the need for high-quality, independent research remains urgent—particularly studies that measure withdrawal with the same level of care used to evaluate drug efficacy.

“There is little good quality data on antidepressant withdrawal, so it is difficult to make estimates about its frequency,” Moncrieff told PsyPost. “Also, it likely that some antidepressants have stronger dependence-inducing effects than others. Withdrawal symptoms include common symptoms such as headaches, dizziness, anxiety, and other complaints, and so far there are no studies that try to distinguish between withdrawal symptoms and ‘background’ symptoms. I hope to set up a study that can explore antidepressant withdrawal in more detail and also how to help people avoid severe withdrawal symptoms.”

The new study, “Evidence on antidepressant withdrawal: an appraisal and reanalysis of a recent systematic review,” was authored by Joanna Moncrieff, Harriet Hobday, Anders Sørensen, John Read, Martin Plöderl, Michael Hengartner, Caroline Kamp, Janus Jakobsen, Sophie Juul, James Davies and Mark Horowitz.