The majority of symptoms reported by patients after stopping antidepressants are mild and do not meet the criteria for discontinuation syndrome — a set of more severe withdrawal-like symptoms that some patients and guidelines describe, new research suggests.
The new findings raise questions about the necessity of routinely using long-term tapering strategies and suggest that current deprescribing guidelines may need to be revised, the investigators, with first author Michail Kalfas, MSc, King’s College London, London, England, noted.
“I would hope that the guidelines from various authorities such as the Royal College… should be considered to be changed to reflect this most recent update in terms of the evidence,” the study investigator Allan H. Young, MD, PhD, told reporters attending a virtual press briefing.
Young emphasized the importance of weighing the benefits and risks when considering antidepressant treatment, noting that some patients hesitate to start medication due to fears of difficult discontinuation symptoms.
He said the new findings should help reassure both clinicians and patients, as commonly used antidepressants for depression and anxiety offer clear benefits, and any discontinuation effects are typically mild, limited in number, and tend to lessen over time.
The study was published online on July 9 in JAMA Psychiatry.
Inconsistent Clinical Guidance
Antidepressant discontinuation symptoms have been recognized since the 1950s, but clinical guidance on how to manage them remains inconsistent. While most international guidelines recommend tapering, there is little consensus on how long tapering should last or what symptoms to expect.
Previous studies have reported wide-ranging rates of discontinuation symptoms, but many lacked details on the type or severity and often did not include comparisons with placebo.
Current Royal College guidelines appear to rely on survey studies with self-selected samples, including users of tapering strips. These are “methodologically weak, with unclear response rates and probable selection bias,” senior author Sameer Jauhar, PhD, told Medscape Medical News.
“I am slightly perturbed that the authors of this guidance did not consider our critiques of clear methodological fallacies in this literature — which they were aware of,” he added.
To determine the incidence and nature of symptoms following discontinuation of antidepressants the researchers conducted a systematic review and analysis of 49 randomized clinical trials of patients who discontinued or continued antidepressants or discontinued placebo. Additional unpublished data from 11 randomized trials were also included.
The primary aim of the research was to assess antidepressant discontinuation symptoms using the Discontinuation Emergent Signs and Symptoms (DESS) scale, by directly comparing symptom scores between individuals who stopped taking an antidepressant and those who either continued treatment or discontinued a placebo.
A secondary goal was to evaluate the incidence of specific discontinuation symptoms in patients who stopped antidepressants compared with those who stopped placebo.
In all, the analysis included 17,828 participants. They had a mean age of 44 years and 66.9% were women. The follow-up was between 1 day and 52 weeks after cessation, with discontinuation symptoms measured using the DESS scale, or as reported adverse events.
Most participants had major depressive disorder, but there were a wide range of conditions including generalized anxiety disorder, panic disorder, fibromyalgia, premenstrual dysphoric disorder, posttraumatic stress disorder, and compulsive-shopping disorder.
Kalfas noted during the briefing that the analysis included a broad range of antidepressants, primarily selective serotonin reuptake inhibitors and serotonin-norepinephrine reuptake inhibitors such as paroxetine, duloxetine, sertraline, escitalopram, venlafaxine, fluoxetine, and vortioxetine.
Discontinuation Effects
The investigators found that participants who discontinued antidepressants experienced significantly more withdrawal symptoms at 1 week than those who either continued treatment or stopped placebo.
Using the DESS scale, the standardized mean difference was 0.31 (95% CI, 0.23-0.39; k = 11 studies; n = 3915), which corresponds to approximately one additional symptom.
Investigators also reported increased odds of specific symptoms following antidepressant discontinuation, including dizziness (odds ratio [OR], 5.52; 95% CI, 3.81-8.01), nausea (OR, 3.16; 95% CI, 2.01-4.96), vertigo (OR, 6.40; 95% CI, 1.20-34.19), and nervousness (OR, 3.15; 95% CI, 1.29-7.64) compared with placebo. Dizziness was the most common symptom, with a risk difference of 6.24%.
Despite including participants with major depressive disorder (k = 5 studies), the analysis showed no association between discontinuation and a re-emergence of depressive symptoms.
Kalfas explained that the primary analysis compared symptom levels in individuals who stopped taking antidepressants with those who either continued the medication or stopped placebo.
On average, those who discontinued antidepressants experienced about one additional symptom — 1.09 to be exact — within a week. The range of additional symptoms (0.8-1.36) remained below the threshold for clinically significant discontinuation syndrome, which is defined as four or more symptoms, Kalfas noted.
There were slightly more symptoms in those discontinuing duloxetine (1.61) and desvenlafaxine (1.37) than those discontinuing vortioxetine (0.56), whereas those who discontinued agomelatine did not experience any symptoms.
“This is a very important and interesting finding, those who discontinued antidepressants did not experience worsening moods in the first 2 weeks, although this outcome was measured in people with major depressive disorder,” said Kalfas.
The authors emphasized the importance of recognizing that antidepressant discontinuation can cause real, and sometimes distressing symptoms. However, they cautioned that both clinical practice and media coverage should present these effects in a balanced way.
Overstating the risks, they warned, could unintentionally contribute to or prolong symptoms through psychological factors such as the nocebo effect — where negative expectations lead to real physical symptoms.
Criticism, Kudos
Commenting on the research for Medscape Medical News, Allen Frances, MD, professor emeritus of psychiatry at Duke University School of Medicine, Durham, North Carolina, and former chair of the DSM-IV Task Force, said the authors have “over-interpreted” their findings.
He added that the study does not change his clinical view that gradual deprescribing remains the best approach for patients who are not benefiting from a psychiatric medication.
Frances described the analysis as questionable and highlighted what he believes are three major weaknesses. These include the inclusion of underpowered studies, short durations of antidepressant therapy, and averaging symptoms across the entire study population. He explained that if only 1 in 5 patients experience clinically significant withdrawal with five symptoms, whereas the other four have none, averaging symptoms for the whole group can mask the true extent of withdrawal harms.
Mark Horowitz, MD, a vocal advocate for gradual antidepressant tapering and Visiting Lecturer in Psychopharmacology at King’s College London, also noted a risk for bias in many of the included studies.
He pointed out that many of the authors have close ties to drug manufacturers and argued that focusing on short-term, industry-funded research may underestimate the potential long-term harms faced by antidepressant users. Horowitz discussed these concerns in an article for The Conversation.
Horowitz, who leads the Psychotropic Deprescribing Clinic at the North East London NHS Foundation Trust, pointed out that the study’s average antidepressant exposure was only 8-12 weeks.
He warned this short duration likely underestimates the risk for withdrawal for the millions of patients taking these medications for years. A recent survey by his group found significantly higher rates and greater severity of withdrawal symptoms among individuals with longer-term antidepressant use.
However, Jauhar denied that his group focused on short-term, industry-funded studies.
“We included all eligible randomized controlled trials irrespective of length and funding source, one of which was the ANTLER trial, the most comprehensive primary study on antidepressant discontinuation, that had a follow up of 12 months and 70% of the sample was treated with antidepressants for at least 3 years,” he noted.
Evidence also shows that industry-funded studies do not report lower rates of discontinuation symptoms than studies not funded by industry, he added, and cited a study published last year showing no association.
“The vernacular in regard to funding is misleading and ignores the fact that almost all antidepressants in the UK are generic — i.e. there is no financial gain,” he added.
In a statement from the UK nonprofit Science Media Centre (SMC), additional experts welcomed the study for its clearer, more comprehensive analysis of antidepressant discontinuation symptoms, but also noted key limitations and the need for further research.
Katharina Domschke, MD, PhD, praised the study’s stronger methodology compared with earlier, smaller studies with significant biases. Susannah Murphy, DPhil, highlighted the large dataset of over 17,000 patients and the inclusion of placebo comparisons but pointed out that the analysis only covered symptoms within the first 2 weeks after stopping medication and did not assess symptom severity.
Christiaan Vinkers, MD, PhD, said the findings help temper alarmist claims about widespread, severe withdrawal but acknowledged remaining gaps — particularly concerning long-term users and real-world tapering strategies.
A spokesperson for the Royal College of Psychiatrists noted that the study reinforces current understanding that most patients benefit from antidepressants, though some individuals experience side effects, and that this new evidence will inform future patient resources.
Overall, the consensus presented by SMC experts underscores the study’s valuable contribution while calling for further research on long-term discontinuation effects and individualized tapering approaches.
Kalfas reported receiving personal fees from Neurocentrx Pharma outside the submitted work; employment by King’s College London; and funding by the UK National Institute for Health Research (NIHR) Biomedical Research Centre. Jauhar reported receiving speaker fees from Boehringer Ingelheim, the Dubai Masterclass conference, Janssen, Lundbeck, and Recordati; serving as a nonpaid member of the UK National Institute for Health and Care Excellence Health Technology Appraisal committee; serving as a committee member and on the funding panel for the Wellcome Trust; and receiving advisory board fees from Boehringer Ingelheim and LB Pharmaceuticals outside the submitted work. Horowitz reported he is an investigator on the RELEASE and RELEASE+ trials in Australia, funded by the Medical Research Future Fund and the National Health and Medical Research Council, evaluating hyperbolic tapering of antidepressants against care as usual. He also reported being a co-founder and consultant to Outro Health, a digital clinic which provides support for patients in the US to help stop no longer needed antidepressant treatment using gradual, hyperbolic tapering. He also reported receiving royalties for the Maudsley Deprescribing Guidelines.Frances reported no relevant financial disclosures. Murphy reported receiving consultancy fees from Zogenix, Sumitomo Dainippon Pharma, UCB Pharma, and Janssen Pharmaceuticals and held grant income from Zogenix, UCB Pharma, Janssen Pharmaceuticals and ADM. Domschke reported receiving speaker’s fees from Janssen and is a member of the Neurotorium editorial board, Lundbeck Foundation. Vinkers reported being involved publicly in ZonMW-funded research on antidepressant discontinuation, including the TEMPO and HARMONIE studies, affiliation with the antidepressant discontinuation outpatient clinic in Amsterdam, and membership of the Dutch multidisciplinary guideline committee on psychotropic drug discontinuation. He had also received a speaker’s fee from Tiofarma but no financial ties to pharmaceutical companies relevant to this work.
Kate Johnson is a Montreal-based freelance medical journalist who has been writing for more than 30 years about all areas of medicine.