By Evgeny Legedin

The recent publication in The Lancet Psychiatry (2026; 13: 24-36)¹ of a systematic review and network meta-analysis by Debora Zaccolotti and colleagues has sparked interest in the ongoing debate about depression pills use and discontinuation. Titled “Comparison of antidepressant deprescribing strategies in individuals with clinically remitted depression: a systematic review and network meta-analysis,” this study synthesises evidence from 76 randomised controlled trials (RCTs) involving over 17,000 participants. As a platform dedicated to critical reflection on psychiatric practices, we examine this work through a sceptical lens, highlighting its contributions while questioning its implications for the medicalisation of distress.

Who Organised the Study?

The study was led by researchers from the WHO Collaborating Centre for Research and Training in Mental Health and Service Evaluation at the University of Verona, Italy. Corresponding author Dr Giovanni Ostuzzi, along with a team including Debora Zaccolotti, Carlotta Mosconi, Chiara Gastaldon, and others from Verona, collaborated with international experts such as Florian Naudet from Université de Rennes, France, and Ioana Alina Cristea from the University of Padova, Italy. The authors explicitly declare no competing interests in the paper, and the research was unfunded, which is a positive note in an area often influenced by pharmaceutical interests. Input from individuals with lived experience was incorporated post-analysis to aid interpretation, though they were not involved in design or conduct.

What Were They Looking For?

The authors aimed to compare the effectiveness of various depression pills deprescribing strategies in adults with fully or partially remitted depressive or anxiety disorders. Concerns about overprescribing, long-term use, and the lack of evidence-based discontinuation methods prompted this network meta-analysis (NMA). They searched databases up to April 2025 for RCTs comparing strategies like abrupt discontinuation, fast tapering (≤4 weeks), slow tapering (>4 weeks), reduced-dose continuation (≤50% minimal effective dose), and standard-dose continuation, with or without adjunctive psychological support (e.g., mindfulness-based cognitive therapy or CBT). The primary outcome was relapse rate at trial end (mean follow-up ~46 weeks), with secondary measures including psychopathology scores, quality of life, dropouts, adverse events, and withdrawal symptoms.

By pooling data across depression (79% of studies) and anxiety (21%), the team sought to identify which approaches best prevent “relapse” while minimising harm, hypothesising that slow tapering might outperform others.

What Did They Find?

The analysis included 17,379 participants (mean age 45 years, 67% female, mostly White). Key findings:

• Strategies outperforming abrupt discontinuation for relapse prevention (with relative risks [RRs] and moderate certainty unless noted):
• Standard-dose continuation + psychological support: RR 0.40 (NNT 4.3).
• Standard-dose continuation: RR 0.51 (NNT 5.3).
• Slow tapering + psychological support: RR 0.52 (NNT 5.4).
• Reduced-dose continuation: RR 0.62 (low certainty, NNT 6.8).

No significant difference from abrupt discontinuation for fast tapering + support (low certainty), abrupt + support (very low certainty), or slow tapering alone (low certainty).

Head-to-head: Slow tapering + psychological support was equivalent to continuation but superior to abrupt or fast tapering.

Tolerability was similar across groups, though data on withdrawal symptoms, quality of life, and long-term outcomes were sparse.

The authors conclude that guidelines should promote individualised, gradual tapering with psychological support, especially for depression, while cautioning generalisation to anxiety due to limited data.

Limitations

While ambitious, the study has notable shortcomings:

Methodological Choices: Tapering categories (≤4 vs. >4 weeks) were arbitrary and pragmatic, potentially oversimplifying real-world practices. An alternative analysis separating “very slow” tapering (>12 weeks) showed similar trends but underscored the under-representation of hyperbolic (non-linear) tapering schedules.

Data Quality and Gaps: Moderate heterogeneity (I²=60.7%) persisted despite sensitivity analyses. Many comparisons had low or very low certainty due to imprecision, risk of bias (28% of RCTs high risk), and indirect evidence. Withdrawal symptoms were poorly reported and possibly misclassified as adverse events or relapse. Follow-up was short (~46 weeks), limiting insights into long-term effects. Anxiety-specific data were scarce, and partial remission was included without full stratification.

Generalisability: Predominantly White participants and assumptions of similar effects across antidepressants (validated but not exhaustive) restrict applicability. The pooling of depression and anxiety overlooks potential differences in underlying experiences.

Bias Risks: While unfunded, some included RCTs had industry ties, though sensitivities mitigated this. Transitivity (comparability across comparisons) was checked but not fully violated.

These limitations align with broader critiques in the field, where RCTs often favour short-term, controlled settings over real-world complexities – as if the messiness of human experience can be neatly boxed into trial endpoints.

A Critical Perspective

This meta-analysis represents a quiet step toward acknowledging deprescribing challenges, yet it falls short by perpetuating some flawed assumptions about depression pills and “relapse.” It risks conflating withdrawal symptoms with true relapse, a common pitfall in discontinuation research.² “Relapses” often peak early (within 6-12 weeks) in abrupt or fast tapering arms, aligning with withdrawal timelines rather than disease recurrence.³ The high relapse rates in abrupt/fast groups here likely reflect this, yet the authors downplay sparse withdrawal data, which may inflate antidepressants’ perceived prophylactic value. The arbitrary “slow” category (>4 weeks) is insufficient; even current NICE guidance recommends stepwise reduction taking “at least 1 month and potentially much longer” for many patients,⁴ and recent evidence strongly supports hyperbolic (non-linear) tapering schedules down to tiny doses to minimise severe, prolonged withdrawal in long-term users – a group still grossly underrepresented in RCTs.⁵

Moreover, the biomedical framing treats depression/anxiety as chronic illnesses requiring “maintenance,” ignoring views that depression pills induce dependence rather than correct imbalances. Psychological support’s benefits may stem from addressing life stressors, not augmenting drugs – a subtle shift that challenges the pill-first paradigm. The short follow-up misses how benefits wane over time, and the neglect of long-term harms (e.g., emotional blunting) perpetuates overprescribing. Ultimately, while praising gradual approaches, the evidence calls for trials prioritising withdrawal differentiation, hyperbolic protocols, and non-medical alternatives – reframing “deprescribing” as liberation from iatrogenic dependence.

This review underscores the need for humanism in psychiatry: beyond pills, toward understanding distress in context. Readers are encouraged to explore related posts on our site, such as deprescribing editorials. What are your thoughts? Share in the comments.

1. Zaccolotti D, et al. (2026). Comparison of antidepressant deprescribing strategies in individuals with clinically remitted depression: a systematic review and network meta-analysis. The Lancet Psychiatry.
2. Moncrieff, J., et al. (2022). The serotonin theory of depression: a systematic umbrella review of the evidence. Molecular Psychiatry.
3. Hengartner, M. P. (2021). Prophylactic effects or withdrawal reactions? An analysis of time-to-event data from antidepressant discontinuation trials. Therapeutic Advances in Psychopharmacology.
4. National Institute for Health and Care Excellence (NICE). Depression in adults: treatment and management. NICE guideline [NG222]. Published 29 June 2022. (Section 1.11.6–1.11.9 on stopping antidepressants).
5. Horowitz, M. A., & Taylor, D. (2019). Tapering of SSRI treatment to mitigate withdrawal symptoms. The Lancet Psychiatry.

About the author

Evgeny Legedin is a trainee psychiatrist working in Glasgow, Scotland. He has a long-standing interest in critical psychiatry, evidence-based practice and the honest communication of uncertainty to patients.