Skip to main content
v2026.11,986 entries · CC-BY 4.0
LAC HealthSponsored

Shop medical supplies

  • FDA registered
  • ISO 13485
  • BBB Accredited A+

Shop at LAC Health

Editorial · CASRAI · Clinical research and clinical trials

Lancet Meta-Analysis of 115 Clinical Trials Compares Long-Term Antidepressant Maintenance Outcomes

A Lancet Psychiatry meta-analysis of 115 clinical trials finds escitalopram and duloxetine maintain depressive response longer than benchmark fluoxetine.

Medically reviewed byDr. Diana Nieves Castro, MD
Published 21 Nov 2026· Last updated 20 Nov 2026· 6 minute read
Pharmaceutical blister pack designed for solid oral medication dosage forms
Photo: Plastic Ingenuity / Wikimedia Commons, CC BY-SA 4.0 (cropped). Source · Licence

An international research consortium led by Radboud University Medical Centre and the University of Oxford published a systematic review and network meta-analysis of 115 randomized controlled trials in The Lancet Psychiatry on November 10, 2026, comparing the long-term maintenance efficacy and discontinuation rates of 21 antidepressant medications in over 30,000 adults. The findings offer vital comparative benchmarks for clinical trialists, research offices, and academic medical centers designing extended psychiatric follow-up protocols.

This page describes research findings and general evidence synthesis; it is not medical advice.

Evidence Synthesis Evaluates Long-Term Maintenance Across 21 Antidepressants

The study, published in The Lancet Psychiatry, evaluates which pharmacotherapies sustain depressive symptom remission over extended horizons. While acute efficacy over six to eight weeks is established in registrational trials, comparative maintenance evidence lasting six months or longer has historically remained fragmented across heterogeneous designs.

Lead investigator Dr. Paul C. F. van Haaren and senior author Dr. Henricus G. Ruhé of Radboud University Medical Centre, working with Dr. Andrea Cipriani and Dr. Edoardo G. Ostinelli at the University of Oxford alongside teams at Amsterdam UMC, Kyoto University, and the University of Bern, searched ten databases and regulatory repositories through March 27, 2026. Following the PRISMA guidelines, the authors screened 54,231 records to identify 115 double-blind randomized controlled trials enrolling 30,428 adult participants with unipolar major depressive disorder.

The study cohort had a mean age of 45.3 years, and 66% (20,066 participants) were female. The authors conducted random-effects pairwise and network meta-analyses to evaluate two primary endpoints: efficacy (sustained response rate) and acceptability (all-cause discontinuation). Fluoxetine served as the reference comparator across ranked network forest plots.

Comparative Odds Ratios and Risk-of-Bias Distribution

The primary comparative analysis synthesized 35 head-to-head continuation and continuation-extension trials comprising 9,888 patients. In this dataset, three antidepressant compounds demonstrated statistically higher sustained response rates than fluoxetine:

  • Escitalopram: Demonstrated an odds ratio (OR) of 1.62 (95% CI 1.21–2.18) for sustained response alongside lower all-cause discontinuation (OR 0.60, 95% CI 0.43–0.83) compared with fluoxetine.
  • Duloxetine: Showed higher sustained response (OR 1.57, 95% CI 1.08–2.26) and lower all-cause discontinuation (OR 0.64, 95% CI 0.44–0.95) relative to fluoxetine.
  • Agomelatine: Showed higher sustained response (OR 1.40, 95% CI 1.06–1.85) than fluoxetine, while all-cause discontinuation was comparable.

Conversely, pairwise league table comparisons identified agents that performed less favorably over time. Bupropion and reboxetine showed lower sustained response rates against four other active antidepressants in pairwise rankings. In trial acceptability, reboxetine and milnacipran exhibited higher all-cause discontinuation against six and four other antidepressants, respectively.

Antidepressant Agent Class Sustained Response vs Fluoxetine (OR, 95% CI) All-Cause Discontinuation vs Fluoxetine (OR, 95% CI) Comparative Performance Summary
Escitalopram SSRI 1.62 (1.21–2.18) 0.60 (0.43–0.83) Higher sustained response rate and significantly lower discontinuation than benchmark
Duloxetine SNRI 1.57 (1.08–2.26) 0.64 (0.44–0.95) Higher sustained response rate and significantly lower discontinuation than benchmark
Agomelatine Melatonergic / 5-HT2c 1.40 (1.06–1.85) 0.94 (0.69–1.28) Higher sustained response rate; discontinuation comparable to benchmark
Fluoxetine SSRI 1.00 (Reference) 1.00 (Reference) Established historical benchmark comparator for network meta-analysis models
Bupropion NDRI 0.89 (0.58–1.38) 1.12 (0.74–1.70) Lower sustained response rates against four comparator agents in league table
Reboxetine NRI 0.74 (0.46–1.18) 1.71 (1.11–2.63) Lower sustained response and higher discontinuation against multiple active comparators

Methodological quality assessment using the Cochrane RoB 2 tool identified substantial vulnerability across the evidence base. Of the 35 head-to-head continuation trials, 19 studies (54%) were judged to carry a high risk of bias, driven by participant attrition during maintenance phases, allocation concealment limitations, and incomplete outcome reporting.

Background and Context in Maintenance Depression Research

Major depressive disorder is a recurrent disorder, and clinical guidelines recommend continuing pharmacotherapy for six to twelve months following remission. However, observational data reported in The Conversation and summarized by ScienceDaily indicate that prescribing durations frequently extend to two to four years, far beyond clinical trial observation periods.

The broad protective effect of maintenance treatment was confirmed by the review: approximately 62% of patients continuing active antidepressants maintained symptom improvement compared with 45% of those switched to placebo. This aligns with findings from the landmark ANTLER trial published in the New England Journal of Medicine. Nevertheless, historical trials frequently used enrichment designs where only acute responders are randomized to maintenance, introducing potential selection artifacts that require rigorous randomization controls and sensitivity modeling.

Independent Psychiatrists and Clinical Trialists Urge Caution

In expert commentary gathered by the Science Media Centre and Science Media Centre España, independent clinicians emphasized that population rankings require careful clinical interpretation. Dr. Sameer Jauhar, Clinical Associate Professor in Affective Disorders and Psychosis at Imperial College London and a peer reviewer for the study, highlighted that the primary comparative dataset rested on approximately 10,000 participants across 35 trials, with over half at high risk of bias.

Dr. Jauhar noted that typical follow-up was around six months, with only six trials tracking patients beyond that duration. Consequently, the evidence cannot establish comparative effectiveness over several years. Dr. Jauhar stressed that population averages cannot predict individual patient response, warning that patients currently stable on an antidepressant should not switch medications based on meta-analytic effect sizes.

Commentators also emphasized safety and regulatory differences. Agomelatine is not approved by the US FDA and requires routine hepatic monitoring in Europe due to hepatotoxicity risks. Furthermore, bupropion was evaluated in relatively small participant cohorts, underscoring why statistical rankings cannot replace individualized clinical decisions.

What It Means for Researchers, Universities, and Research Offices

The findings carry operational and methodological implications for academic medical centers and clinical research administrative offices:

  • Longitudinal Protocol Horizons: Academic sponsors must design trial protocols extending beyond six months to match real-world maintenance timelines and monitor emergent long-term adverse events.
  • Managing Attrition Bias: Because 54% of maintenance trials showed high risk of bias, trial offices and principal investigators must strengthen retention frameworks and longitudinal missing-data strategies.
  • Distinguishing Withdrawal from Relapse: Clinical trial designs must differentiate between depressive relapse and antidepressant discontinuation symptoms, embedding structured tapering rather than abrupt placebo substitution.
  • Data Harmonization for Evidence Synthesis: Institutions managing clinical data should adopt standardized electronic outcome reporting to enable future individual participant data meta-analyses.

What Happens Next in Long-Term Treatment Evidence

The study authors emphasized that further head-to-head pragmatic trials across licensed antidepressants are needed to resolve uncertainties from indirect network comparisons. Funder programs in the United Kingdom, European Union, and North America are prioritizing comparative effectiveness studies that incorporate biomarker and pharmacogenomic profiling.

Clinical guideline committees, including the UK National Institute for Health and Care Excellence (NICE), will review this network meta-analysis during their upcoming cyclic updates to depression management guidance. In parallel, academic psychiatric networks are establishing longitudinal registries to track multi-year outcomes, patient-reported quality of life, and structured discontinuation pathways.

Sources

Ask CASRAI · full access

Ask about this story

Full access: large monthly usage, up to 10 source passages behind each answer, and MCP and API access for Claude Code, Codex and other CLIs. Cancel any time.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →