One in six patients experience withdrawal symptoms when stopping antidepressant treatment

One in six patients experience withdrawal symptoms when stopping antidepressant treatment
One in six patients experience withdrawal symptoms when stopping antidepressant treatment

Berlin, Germany – That stopping antidepressant treatment can cause dizziness, nausea or sleep problems is a well-known fact. What was not yet known is the frequency of these withdrawal effects, which are also linked to negative expectations in some patients.

A meta-analysis carried out by the team gathered around Jonathan Henssler (head of the Clinic for Psychiatry and Psychotherapy at the Charité Hospital in Berlin, and head of the Evidence-based mental health working group) and published in the Lancet Psychiatry [1] shows that about 15 percent of them – or about one in six to seven – experience actual withdrawal symptoms.

Researchers analyzed data from 21,002 patients included in a total of 79 studies. 72% of them were women and the average age was 45 years. Participants had a mental, behavioral, or neurodevelopmental disorder. The included studies analyzed the frequency of withdrawal symptoms from the most common antidepressants and placebo.

Risks vary depending on the antidepressant

Across these studies, 31% of patients who stopped treatment reported withdrawal symptoms, and 2.8% (or approximately 1 in 35) reported serious symptoms. These figures rose to 17 and 0.6% respectively in patients who took a placebo, which gives a difference of around 15 and 2.2% to actually be attributed to the antidepressants.

The results of the meta-analysis also suggest that the risk would be higher with certain antidepressants (imipramine, paroxetine, desvenlafaxine and venlafaxine), with no significant difference depending on whether the cessation of treatment would be abrupt or gradual.

Previous studies that showed higher rates of withdrawal symptoms drew criticism, particularly because these high rates were regularly used as an argument against prescribing antidepressants. Whatever the case, Jonathan Henssler said he was not surprised by these results during a press briefing. [2] au Science Media Centre (SMC, UK): “In my personal experience, these figures – including around one in six patients – are roughly in line with what we see in the clinic. »

Very real withdrawal symptoms

According to Pr Christopher Baethge (Clinic for Psychiatry and Psychotherapy at the Medical University of Cologne), one of the co-authors of the analysis: “Withdrawal symptoms are real and patients should be informed, monitored and supported when they occur. » According to him, the study provides an estimate of their frequency, “but we do not assert that they are rare nor, conversely, particularly frequent”

The nocebo effect taken into account

“One of the particular merits of this meta-analysis is that it also focuses on the nocebo effect, excluding it from the incidence of withdrawal phenomena after a real treatment,” underlines the Pre Katharina Domschke, who directs the Psychiatry and Psychotherapy Clinic at Friborg University Hospital. An opinion shared by the Pr Klaus Lieb (Clinic of Psychiatry and Psychotherapy of Mainz University Hospital): “The exclusion of nocebo patients is innovative and the methodology used in this meta-analysis is of very high quality. »

So far, according to the Dr Michael Hengartner (Department of Applied Psychology at the Zurich University of Applied Sciences – ZHAW), the rate of symptoms after stopping antidepressants was considered to be somewhat higher. The heterogeneity of the studies could be a reason for the variations in the effects observed.

The results, according to Katharina Domschke, largely correspond to what seems to be observed in clinical practice. “In reality, the main conclusion of the analysis is that the proportion of patients affected is relatively low and that antidepressants are not addictive. »

These are not, as the study explicitly points out, withdrawal symptoms such as observed with alcohol or benzodiazepines, but symptoms similar to those which can occur when stopping other medications such as cortisone or antihypertensives. The results of the meta-analysis “should in no case lead to stopping prescribing the drugs concerned, but to better informing – both patients and caregivers – about their interruption”, she believes.

Gradual weaning generally prevents the withdrawal phenomenon.

“The incidence of withdrawal symptoms includes the significant nocebo effect which also occurs after stopping a placebo,” note the Dr Eric Ruhé (Department of Psychiatry at the Radboud University Medical Center in Nijmegen, the Netherlands) and the Pr Christiaan Vinkers (University Medical Center Amsterdam). “This suggests that the nocebo effect and anticipation are important factors and that future studies of antidepressant treatment discontinuation need to be placebo controlled to properly identify them.” »

The importance of multidisciplinary guidelines

The Pr Erich Seifritz, who heads the University Psychiatry Clinic in Zurich, believes that the fundamental problem with the original studies analyzed in this work is that “the treatments were stopped abruptly, something no experienced doctor would do today. » In clinical practice, gradual cessation, spread over several months, makes it possible to avoid the phenomenon of withdrawal in most patients.

According to Christiaan Vinkers, the accumulation of knowledge on antidepressant withdrawal is to be welcomed and should lead to multidisciplinary guidelines to help clinicians and patients discuss and manage it.

Informing patients about possible symptoms should be the norm

For Jonathan Henssler, the results of the meta-analysis have direct consequences on clinical practice: “Personally, I think that there must be good reasons to initiate treatment with venlafaxine or paroxetine. » Fluoxetine or sertraline could be considered as alternatives to these two products, as they have comparable effectiveness. In the meta-analysis, venlafaxine and paroxetine caused more withdrawal symptoms than other antidepressants. “A good reason to prescribe them could be, for example, that the patient tolerates them well and responds well to them. »

Christopher Baethge sees no reason to change molecules when a patient responds well to paroxetine: “We must keep in mind that not all patients treated with this drug will develop withdrawal symptoms. These possible symptoms must be taken into account, of course, but the results of the meta-analysis should not constitute a warning signal. »

“Our work also shows the importance of informing patients about the possibility of withdrawal symptoms and the need for monitoring,” adds Jonathan Henssler. This information should be the norm, “because it can limit the patient’s negative expectations and perhaps prevent a pronounced nocebo effect.” »

Finally, Klaus Lieb believes that all large clinics should have specialized antidepressant withdrawal services.

Links of interest

Henssler J, Baethge C, Domschke K, Howes O, Seifritz E: no declaration of conflict of interest.

Hengarter MP: no conflict of interest.

Lieb K: he declares having co-written in 2022 a review on the same subject with the main authors Tom Bschor and Christopher Baethge in The neurologist. Personally, he does not feel influenced by the current work but this joint publication on the same subject formally represents a conflict of interest. He has not accepted personal fees from pharmaceutical companies for 15 years.

Vinkers C: co-founder of the Withdrawal Clinic in Amsterdam, he recently published a study on withdrawal from all classes of psychotropic drugs. Principal investigator of the study versus placebo in progress called Trial Examining Methods for Antidepressant Discontinuation  (TEMPO)supported by ZonMW (Netherlands), he contributed to the Dutch multidisciplinary guideline on depression.

Ruhé E was a member of the working group that drafted the Dutch multidisciplinary document on withdrawal from SSRIs and SNRIs. He recently published an article on withdrawal from all classes of psychotropic medications and is the principal investigator of the TEMPO study. He also contributed to the Dutch multidisciplinary guideline on depression.

This article is a translation-adaptation by Dr Claude Leroy of the article written by Ute Epinger, published by and entitled Meta-analysis of withdrawal reactions after antidepressant therapy: 1 in 6 people experience unpleasant symptoms afterwards



NEXT Crack and cocaine are exploding in Switzerland and that can be explained