Skip to content
MyFreud

Antidepressant Discontinuation Syndrome Explained

How common is antidepressant discontinuation syndrome? A 35-study meta-analysis and a critical reanalysis give different answers. What each found and why.

7 min read

Flat vector illustration of a person in a dark green sweater and trousers sitting on a window ledge with one foot up, looking out at pastel apartment buildings under a pale sky.

Key takeaways

  • A 2025 meta-analysis of 35 studies found that about 43% of people stopping an antidepressant had withdrawal symptoms (42.9% overall, 44.4% across 11 randomised trials), usually within 2 weeks of stopping.
  • A 2025 reanalysis argues that an influential earlier review, which put the figure nearer 15% after subtracting placebo, rested on weak data: only 5 of its 62 study cohorts measured withdrawal systematically, and those 5 pooled to 55%.
  • The figures disagree mainly because of how symptoms were counted, how long people had taken the drug and how long they were watched afterwards, not because anyone found that withdrawal does not exist.
  • Most trials followed people for only a few weeks after short courses of treatment, so none of these numbers tells you what to expect after years on a drug.
  • Do not stop or cut an antidepressant on your own because of anything here: symptoms after stopping can resemble depression returning, and your prescriber can help you tell the difference.

About 4 in 10 people who stop an antidepressant report withdrawal symptoms, according to a 2025 meta-analysis of 35 studies that found 42.9% overall and 44.4% across 11 randomised trials, with symptoms usually starting within 2 weeks. [zhang-2025] Other analyses have put the figure much lower, and a 2025 critique argues that those lower figures come from data that were never designed to measure withdrawal. [reanalysis-2025] This guide explains why the numbers disagree, what both sides agree on and what to do with the uncertainty. Our overview of depression covers the wider picture, and our guide to new antidepressants covers what the drugs are and how they are chosen.

What is antidepressant discontinuation syndrome?

It is a set of physical and emotional symptoms that some people develop after stopping or reducing an antidepressant. The research calls it antidepressant withdrawal syndrome or discontinuation symptoms, and the 2025 meta-analysis notes that such symptoms have been reported for almost all antidepressants. [zhang-2025] A critical reanalysis lists anxiety, fatigue, impaired concentration and worsened mood among the most common, alongside dizziness, headache and electric-shock sensations often called “zaps”. [reanalysis-2025]

The label matters less than the practical point: symptoms after stopping are a recognised thing, and the open question is how often they happen and how bad they get.

How common is antidepressant withdrawal?

On the most recent pooled evidence we could verify, somewhere between roughly 15% and 55% of people, and the honest answer depends on how you count. The 2025 meta-analysis searched six databases to December 2023 and included 35 studies; it found a pooled incidence of 42.9% from all available studies and 44.4% from the 11 randomised trials. [zhang-2025]

A second 2025 paper reanalysed an earlier, influential review, which the reanalysis describes as estimating that 31% of people coming off an antidepressant had any withdrawal symptom, and about 15% after subtracting symptoms seen when placebo was stopped. That earlier review also reported that 3% of people had severe symptoms. [reanalysis-2025] We have not read the earlier review itself, so these figures are as the reanalysis reports them.

SourceFigureWhat it counts
2025 meta-analysis (35 studies)42.9% overall; 44.4% in 11 trialsIncidence of withdrawal symptoms
Earlier review (62 cohorts), as described by the reanalysis31%; about 15% after placebo subtracted; 3% severeAny symptom; net of placebo
2025 reanalysis (5 studies)55%; 25% to 39% after a placebo deductionOnly studies with a systematic measure
In the 2025 meta-analysis, the longer people had taken the drug, the more often withdrawal symptoms were reported Reported figures
0 25 50 75 100 Share of people with withdrawal symptoms (%) 6 to 12 weeks 12 to 24 weeks Over 24 weeks Withdrawal symptoms

Pooled incidence of antidepressant withdrawal symptoms by treatment duration, from the meta-analysis by Zhang and colleagues (2025): 35 studies. 6 to 12 weeks 35.1%; 12 to 24 weeks 42.7%; over 24 weeks 51.4%. These are comparisons across different studies, not the same people followed over time.

Why do the estimates disagree?

They disagree mainly because the studies they pooled were not built to measure withdrawal, and the reanalysis argues that is a fundamental flaw in the lower estimates. In the earlier review, the reanalysis found, 52 of 62 study cohorts took their figures from adverse-event reports, open questions, clinician judgement or an unspecified method. Only 5 of 62 (8.1%, covering 601 of 12,603 participants) used a systematic and relevant assessment of withdrawal symptoms. [reanalysis-2025]

The reanalysis lists further problems with the earlier data: 46 of 62 studies (74%) had definite or probable funding from a drug company; the average time on the drug was under six months (23.4 weeks); and the most common observation period was two weeks, which could miss symptoms that start later. It also argues that subtracting placebo-withdrawal symptoms is not like-for-like, because the placebo and drug figures came from different groups of studies. [reanalysis-2025]

In the five studies that did measure withdrawal systematically, the pooled share was 55%, with high heterogeneity. Excluding one trial where not everyone stopped the drug gave 61%, and also excluding a trial of agomelatine, a drug with a different mechanism, gave 69%. [reanalysis-2025]

Who is more likely to have withdrawal symptoms?

The meta-analysis found that risk factors included being female, being younger, having adverse effects early in treatment, taking higher doses or taking the drug for longer, stopping abruptly, and having lower drug clearance or a variation in the serotonin 1A receptor gene. [zhang-2025] By class, the incidence was 29.7% for serotonin-noradrenaline reuptake inhibitors, 45.6% for selective serotonin reuptake inhibitors (SSRIs) and 59.7% for tricyclic antidepressants, but the authors report no significant differences between classes, so the ranking should not be treated as established. [zhang-2025]

Duration showed a dose-response pattern in the pooled data: 35.1% at 6 to 12 weeks, 42.7% at 12 to 24 weeks and 51.4% beyond 24 weeks. [zhang-2025] The reanalysis makes a related point from earlier research, that withdrawal effects are greater after long-term use, which is why trials of short courses may understate what happens after years. [reanalysis-2025]

Does tapering the dose prevent withdrawal?

Not demonstrably. The meta-analysis found withdrawal symptoms in 34.5% of people who tapered and 42.5% of people who stopped abruptly, a gap that was not statistically significant. [zhang-2025] That is not evidence that tapering does not help; it means the pooled studies could not show it, and many were short and not designed around tapering. Whether to taper, and how fast, is a decision for you and your prescriber, and nothing here replaces that conversation.

Is it withdrawal or is my depression coming back?

The symptoms alone cannot reliably tell you, and that overlap is part of why withdrawal has been hard to measure. The reanalysis argues that because the common withdrawal symptoms (anxiety, fatigue, poor concentration, low mood) overlap with the conditions antidepressants treat, they can be overlooked or misclassified as symptoms of the underlying condition. [reanalysis-2025]

What can help is timing and novelty, and these are our suggestions rather than something the studies tested: note whether symptoms started within days of a dose change, whether they include things you did not have before (dizziness, electric-shock sensations) and how they change from week to week. A record like that gives a prescriber something to work with. If you have been through a depressive episode and your mood is dropping, our depression screener can structure what you are noticing, though it cannot tell withdrawal from relapse.

What can this research not show?

It cannot tell you what will happen to you, how long symptoms last, or what is typical after years on a drug. Five limits matter most:

  • Most data come from short treatment and short follow-up. The meta-analysis notes that treatment mostly lasted 8 to 12 weeks and symptoms were generally measured for under 4 weeks, and its authors call for studies of long-term users with long follow-up. [zhang-2025]
  • Raw percentages include symptoms the drug did not cause. Some symptoms follow stopping any pill, or stopping a placebo. The reanalysis says its own figures are likely to include such incidental symptoms, and for illustration deducted a 30% placebo rate, which cut its 55% to 69% range to 25% to 39%. [reanalysis-2025]
  • Severity is poorly measured. The reanalysis reports that for 11 of 19 studies, the earlier review’s severity figures were based on adverse events that led to study discontinuation or on serious adverse events, which it argues are not valid indicators of how severe withdrawal feels. [reanalysis-2025]
  • The critique is a partial view. Several authors of the reanalysis declare interests in tapering research and in books about psychiatric drugs, and the page we read lists a 2025 Psychological Medicine paper by the earlier review’s authors, titled “Double counting, double standards”, which we have not read. Weigh it as one side of an active dispute. [reanalysis-2025]
  • Neither paper tests what to do. They count symptoms; neither is a trial of tapering schedules, so neither can say how to come off a drug safely.

When to seek help

Speak to a doctor before stopping or reducing an antidepressant, and sooner if symptoms after a change are distressing, last more than a few days or interfere with work, sleep or relationships. Do not restart, stop or change a dose on your own on the strength of anything on this page. If you have thoughts of harming yourself or feel unable to stay safe, contact your local emergency services or a crisis helpline.

How MyFreud can help

A record shows a prescriber what changed and when. Daily mood tracking in MyFreud shows how your days run across a dose change, and the notepad is somewhere to note sleep, dizziness, anxiety and anything new. It does not replace a doctor or a pharmacist.

Download MyFreud and start today: App Store or Google Play.

Frequently asked questions

What is antidepressant discontinuation syndrome?

It is the group of physical and emotional symptoms some people get after stopping or cutting down an antidepressant. Research papers also call it antidepressant withdrawal syndrome. How often it happens is debated, with estimates in the studies we checked ranging from roughly 15% to over 50%, depending on how symptoms were measured.

How common is antidepressant withdrawal?

A 2025 meta-analysis of 35 studies found a pooled incidence of 42.9%, and 44.4% in the 11 randomised trials. A reanalysis of an earlier review that estimated about 15% argues that figure rests on unreliable data, and found 55% in the five studies that measured withdrawal systematically. Both estimates mostly come from short-term studies.

How long does antidepressant discontinuation syndrome last?

The 2025 meta-analysis reports that symptoms usually appear within 2 weeks of stopping, but most of its studies measured symptoms for less than 4 weeks, so it cannot say how long they last. The reanalysis notes that short observation periods are likely to miss effects that start later or continue.

Does tapering the dose prevent withdrawal symptoms?

Not clearly. In the 2025 meta-analysis, withdrawal symptoms occurred in 34.5% of people who tapered and 42.5% of people who stopped abruptly, but that difference was not statistically significant. Tapering is still what prescribers generally advise, and the right pace for you is a decision to make with them.

Is it withdrawal or is my depression coming back?

You cannot reliably tell from the symptoms alone. The reanalysis notes that common withdrawal symptoms, such as anxiety, fatigue, poor concentration and low mood, overlap with the conditions antidepressants treat, so withdrawal can be mistaken for relapse. Timing and whether the symptoms are new can help, and a prescriber can assess them.

References

  1. 1.Zhang MM, Tan X, Zheng YB, et al. ( 2025). Incidence and risk factors of antidepressant withdrawal symptoms: a meta-analysis and systematic review. Molecular Psychiatry. doi.org . doi:10.1038/s41380-024-02782-4
  2. 2.Authors named on the journal article page (first author listed as JM in the article's contribution statement) ( 2025). Evidence on antidepressant withdrawal: an appraisal and reanalysis of a recent systematic review. Psychological Medicine. doi.org . doi:10.1017/S0033291725100652