Past research has framed symptom increases during tapering and discontinuation of psychotropic drugs as a relapse of an underlying “illness.” A new preprint article challenges this framing, estimating that a majority of increases in symptoms during antidepressant tapering are related to withdrawal, not relapse. This study, led by Andri Rennwald from the Zurich University of Applied Sciences in Switzerland, reports that antidepressant dose reduction to below 75% of the minimum effective dose was linked to symptom increases, while reductions below that threshold were not.
The authors write:
“Clinically relevant symptom increases during antidepressant tapering most likely emerge following dose reduction below 75% of the minimum effective dose. About 60% of the risk of clinically relevant intra-individual symptom increases is attributable to pharmacological withdrawal effects.”
Michael Hengartner, one of the authors of the current work, has conducted past research finding that antidepressant use was linked to an 81% increased likelihood of worsening severity of depression symptoms. He is also the author of Evidence-biased Antidepressant Prescription, Over-medicalisation, Flawed Research, and Conflicts of Interest, a critical examination of antidepressant use where he argues that contemporary psychiatry has strayed from truly evidence-based medicine into a state of evidence-biased medicine, where the benefits of these drugs are systematically overestimated and their potential harms are minimized. In an interview with Mad in America, Hengartner challenged the assertion that antidepressants are most useful for severe depression, saying “there is scant or at least very insufficient evidence for the claim that they [antidepressants] clearly work better in severe depression.”
Mark Horowitz, another author of the current work, described his own experience of withdrawal from an antidepressant as “the most unpleasant experience of my life” in an interview with Mad in America. He went on to co-author the Maudsley Deprescribing Guidelines where he advocates for a much slower dose reduction when discontinuing psychiatric drugs compared to past clinical guidelines. He has also been critical of the recently issued American Society of Clinical Psychopharmacology deprescribing guidelines, which he believes mistakes withdrawal for relapse, downplays how common and long-lasting withdrawal can be, offers too few practical guidelines, and privileges medical authority over patient experience.
The goal of this research was to examine how symptoms changed during antidepressant tapering, especially once the dose dipped below 75% of the minimum effective dose, the smallest dose supposed to reduce symptoms. The authors used a longitudinal design and assessed symptoms in service users that had taken antidepressants for at least six months and decided to discontinue their use. They assessed symptoms before tapering began, then again at two, four, six, eight, 16, and 26 weeks after beginning the taper. In total, the authors examined data from 32 Swiss adults.
At all follow-ups, only reductions to below 75% of the minimum effective dose were significantly linked to symptom increases. Reductions above 75% of the minimum effective dose, as well as no reduction, were linked to symptom decreases. The average risk of worsening symptoms during intervals when the dose was reduced to below 75% of the minimum effective dose was 33% compared to 8% during intervals with a reduction to above 75% of the minimum effective dose and 13% during intervals with no dose reduction.
The authors estimate that 60% of symptom increases during tapering were due to withdrawal. They arrived at this estimate by comparing the baseline rate of worsening and returning symptoms that occur without a reduction in antidepressant dose to those that occur while tapering antidepressants. Clinically relevant symptom increases occurred in 13% of intervals with no dose reduction compared to 33% of intervals with dose reductions below 75% of the minimum effective dose where withdrawal effects are most likely.
This means there was a 20% (33% – 13%) excess risk of symptom increases above the observed baseline. The authors reason that if 13 out of 100 people worsen even without tapering, but 33 out of 100 worsen after low-dose reductions when withdrawal is most likely, then the extra 20 cases are likely withdrawal-related. This means 20 of 33 excess cases, or roughly 60% of intervals where symptoms increased, were likely related to withdrawal.
In total, 53% of participants reported worsening symptoms linked to a reduction to below 75% of the minimum effective dose during at least one follow-up. This included 34% with moderate increases and 19% with severe increases. Additionally, 22% of participants reported mild (subclinical) symptom increases during at least one interval when their dose dipped below 75% of the minimum effective dose.
This study had three main limitations. The sample size was small, limiting validity of the results. The participants were all Swiss adults, limiting generalizability to other populations. This is a preprint study and has not yet gone through peer review.
The authors conclude:
“Primary care patients are often on antidepressants for many years. GPs should conduct regular drug reviews and evaluate, whether such long-term antidepressant therapy is indicated and in agreement with clinical practice guidelines to avoid unnecessary and potentially harmful treatment … Practitioners should be aware that even small reductions in the dose range well below the minimum effective dose can trigger withdrawal reactions, and in such cases, personalized tapering plans and close monitoring of withdrawal symptoms are recommended.”
A new letter published in the African Journal of Primary Health Care & Family Medicine and also authored by Mark Horowitz details the safest ways to taper and discontinue antidepressants. The authors emphasize that tapering and discontinuing antidepressants should be tailored to fit each service user, as people have varying reactions and withdrawal may even be delayed by four to six weeks in some cases. While each case must be considered individually, they do offer some general advice for tapering.
Antidepressant use for more than four weeks should not be stopped abruptly, but rather tapered down before discontinuing. However, it should be noted that even weeks of use can result in withdrawal from abrupt cessation in some people. Low risk service users, those taking a low risk antidepressant for weeks, should begin with a 25% dose reduction followed by two to four weeks of monitoring at the new dose before further reductions. Low risk users should expect to take six to nine months to discontinue antidepressants.
Moderate risk service users, those taking moderate risk antidepressants for months, should start slower with a 10% dose reduction followed by two to four weeks of monitoring. These service users should expect to spend nine to 18 months discontinuing antidepressant use. High risk service users, those taking antidepressants for years, should start with a 5% taper followed by monitoring and should expect to spend 2 or more years discontinuing antidepressant use.
The authors advise practitioners to explain the process of tapering to their clients, including the need for slow tapering and the risks of tapering too quickly. They emphasize balancing the rate of taper with the risks of long term antidepressant use and recommend adjusting the process to fit individual needs. They generally advise against alternate day dosing, except in the case of fluoxetine due to its longer half-life.
In cases where akathisia occurs, reinstating the antidepressant dose soon after the symptoms start can lead to improvement. While other drugs may help with akathisia, the authors warn that they should be stopped if no improvement is seen. Protracted withdrawal can be difficult to manage and cannot always be treated by resuming the discontinued drug. They recommend waiting for improvements to occur without further psychotropic use. In cases where no improvements happen for months or even years, the original drug should only be restarted in small doses to gauge the response, as sometimes this can lead to worsening symptoms. Generally, the authors advise against using additional psychotropic drugs to help with symptoms of akathisia and protracted withdrawal.
****
Rennwald, A., Horowitz, M. A., Senn, O., Neuner-Jehle, O., & Hengartner, M. P. (2026). Incidence of Antidepressant Withdrawal Reactions: A Prospective Longitudinal Cohort Study in Primary Care Patients. (Link)
Horowitz M, Perry T.L. Stopping antidepressants safely. African Journal of Primary Health Care & Family Medicine 2026;18(1), a5431. (Link)