A new longitudinal study from Switzerland suggests that emotional disturbances experienced after reducing or stopping antidepressants are more accurately described as withdrawal reactions than relapse, a finding the researchers say could reshape how clinicians approach tapering. The work was led by Michael P. Hengartner of the Kalaidos University of Applied Sciences and tracked 32 adults in primary care as they began to discontinue their medications.
The participants had been on antidepressants for an average of 42 months and were in stable remission when they planned to taper. Researchers assessed them just before tapering and again at two, four, six, eight, sixteen and twenty-six weeks, using standardized measures for depression and anxiety alongside a tailored scale to capture withdrawal symptoms.
Specifically, the team calculated two withdrawal scores: affective withdrawal symptoms such as nervousness and mood swings, and discriminatory physical symptoms including incoordination, ringing in the ears and brain zaps. These were tracked alongside established questionnaires—the Patient Health Questionnaire-9 and the Generalized Anxiety Disorder-7—as asked about experiences over the preceding two weeks.
Across the 26 weeks, researchers observed a clear pattern: when participants reduced their dose, depression and anxiety rose in parallel with both emotional and physical withdrawal symptoms. Conversely, during periods when the dose was kept stable, both anxiety and emotional withdrawal gradually eased. The researchers describe this wave-like pattern as characteristic of pharmacological withdrawal rather than relapse.
In a key test, the team examined whether removing physical withdrawal symptoms from the analysis would change the link between dose reductions and emotional symptoms. They found that accounting for physical withdrawal substantially weakened the association; for depression, the link disappeared entirely when these physical signs were considered. The researchers concluded that the emotional symptoms and physical withdrawal signs were part of the same withdrawal process rather than separate phenomena.
“Rapid onset of depressive and anxiety symptoms and abatement within a few weeks after sequential dose reductions and stabilizations reflect the typical wave-like pattern of repeated withdrawal reactions, especially when these emotional problems are accompanied by discriminatory physical withdrawal symptoms such as dizziness, incoordination, visual or auditory sensory changes, muscle cramps, or nausea,” Hengartner said. “In such cases, patients are advised to reduce their antidepressant more slowly and in smaller steps to mitigate withdrawal effects.”
The study also highlighted that the magnitude of dose reductions mattered. Reductions of more than 25 percent of the minimum therapeutic dose were more likely to be followed by acute spikes in withdrawal-related distress, whereas smaller reductions tended to lessen these effects.
Hengartner emphasised that there is substantial inter-individual variability in withdrawal risk. “Effect sizes are variable, because they depend on the medication taken, the duration of use, the tapering steps, and the individual propensity to pharmacological withdrawal effects,” he explained. “That is, some people will barely experience withdrawal effects, some have milder and tolerable symptoms, while still others suffer severe and impairing symptoms. For the latter group, the effects have strong clinical relevance and important practical consequences.”
He also noted that this variability was present even among those who had been on long-term medication with considerable withdrawal risk: “We already knew from previous research that there is substantial inter-individual variability in the propensity to experience clinically relevant withdrawal reactions, but we were still surprised to see that, in some cases, even people who had been long-term on a medication with considerable withdrawal risk could discontinue treatment quite rapidly without problems, while others experienced substantial withdrawal effects.”
However, the researchers cautioned that the study’s naturalistic design and small sample size limit the strength of the conclusions. “This was a naturalistic study, that is, people tapered at their own pace,” Hengartner noted. “Moreover, since this was not a randomized controlled study we cannot draw firm causal conclusions. Our findings show a pattern that indicates withdrawal effects, but it is not a definite proof.”
Other limitations include reliance on self-reported measures without independent clinical evaluations, and longer gaps between some assessment points later in the study, which may have missed day-to-day fluctuations in withdrawal symptoms. The researchers therefore called for larger, more frequent studies to confirm timelines and to compare tapering strategies.
Looking ahead, Hengartner argued that replication in a larger cohort and, ideally, a randomised controlled tapering trial are needed to validate causality and to identify risk factors for antidepressant withdrawal more clearly. “Our findings need to be replicated in a larger cohort study,” he concluded. “Moreover, a randomized controlled tapering trial is required to validate the causality of the suggested effects. Finally, we also need more research into inter-individual differences and risk factors of antidepressant withdrawal.”
The study, titled Acute Affective Symptoms during Antidepressant Tapering Indicate Withdrawal Reactions Rather than Relapse: Results from a Prospective Longitudinal Cohort Study, lists Michael P. Hengartner, Andri Rennwald, Oliver Senn, Stefan Neuner-Jehle and Mark A. Horowitz as authors.
