SSRIs and SNRIs in FAERS: 968,183 adverse-event reports by the numbers

The FDA Adverse Event Reporting System (FAERS) holds 968,183 individual safety reports that name at least one of the ten marketed selective serotonin reuptake inhibitor (SSRI) and serotonin-norepinephrine reuptake inhibitor (SNRI) antidepressants — sertraline (Zoloft), duloxetine (Cymbalta), citalopram (Celexa), escitalopram (Lexapro), fluoxetine (Prozac), venlafaxine (Effexor), paroxetine (Paxil), desvenlafaxine (Pristiq), fluvoxamine (Luvox), and levomilnacipran (Fetzima) — in any role, filed between 2004 and early 2026. Sertraline and duloxetine anchor the class with 186,315 and 183,705 reports respectively, each about 19 percent of the class total; citalopram (139,220), escitalopram (136,169), fluoxetine (128,758), and venlafaxine (120,038) form a dense middle band, and paroxetine, the oldest large-volume SSRI, contributes 89,453. Of the class-level reports, 662,310 — 68.4 percent — carry an FDA seriousness flag, and 87,261 (9.0 percent) record a fatal outcome, a higher death-flag rate than the cardiometabolic and oncology classes this publication has analyzed and a direct reflection of overdose-toxicity reporting in a population at elevated baseline suicide risk. The class signature is the suicidality footprint: 55,342 reports in the suicidal-ideation-and-behaviour family, with completed suicide coded as a reaction term on 22,180 reports and suicidal ideation on 19,985. The serotonergic and electrolyte signals that shaped the labels are visible too — 11,850 serotonin-syndrome reports, 10,584 hyponatraemia/SIADH reports, and 7,355 QT-prolongation reports that concentrate in citalopram (1,818), the agent the FDA dose-capped in 2011.

This article is a class-level descriptive read of SSRI and SNRI reporting in FAERS, written for pharmacovigilance, medical-affairs, behavioral-health policy, and pharmacy-benefit teams who encounter antidepressant safety numbers in labeling reviews, prior-authorization design, formulary-tiering, and overdose-surveillance briefings. Every figure here is computed from the public openFDA FAERS extract (20,328,575 total reports, export dated 2026-06-08). It is not a disproportionality signal analysis, it is not a safety claim, and it is not incidence data. FAERS is a spontaneous-reporting system: it captures what was reported, never what was caused, it has no exposure denominator, and antidepressant reporting is dominated by a population with high baseline suicide and overdose risk in which fatal outcomes are common regardless of attribution. Two reporting artifacts dominate this class and must be read into every number below: the 2004 boxed warning on suicidality generated sustained reporting attention to suicidal-ideation terms, and overdose-toxicity reports (intentional and accidental) inflate the death and serious counts. For the class-wide FAERS methodology and the database aggregate picture, see the companion analysis of 20 million FAERS reports; for the other class cuts, see GLP-1 receptor agonists, SGLT2 inhibitors, JAK inhibitors, DOACs, checkpoint inhibitors, and statins in FAERS. This article stays on the ten SSRI and SNRI antidepressants.

Methodology, in one paragraph

A report is counted for an antidepressant when the drug's generic or brand name appears as an exact drug-field token in that report — as a suspect, concomitant, or interacting drug — using an exact-token match across the drug_substances, drug_generics, and drug_brands fields. Because a single report can name several antidepressants (intentional-overdose polypharmacy, or sequential-trial documentation), the per-substance totals sum to more than the 968,183 class-level (union, deduplicated) reports. A "serious" report is any report the FDA coded as serious. A report counts toward death when either the seriousness field carries a death flag or the outcome field carries a Fatal outcome; for this class in particular, that death measure is heavily weighted by overdose-toxicity reporting, not by therapeutic-dose adverse effects. Reactions are coded to MedDRA Preferred Terms; reaction counts are case-normalized. The adverse-event families aggregate related Preferred Terms (for example, "suicidal ideation and behaviour" combines suicidal ideation, suicidal behaviour, suicide attempt, completed suicide, suicidality, and self-injurious ideation; "hyponatraemia / SIADH" combines hyponatraemia, syndrome of inappropriate antidiuretic hormone secretion, and blood sodium decreased), and a report contributes once to a family if any term in the family appears. The "any role" counts reported here are intentionally inclusive.

The reporting-volume picture, by drug

Three patterns matter for access and medical-affairs teams. First, volume tracks market penetration: sertraline and duloxetine lead because Zoloft and Cymbalta are the two highest-volume agents across depression and anxiety (and duloxetine's pain indications), and fluoxetine's share is lower than its historical primacy would suggest because generic fluoxetine is now a small share of new starts. Second, citalopram stands out for the wrong reason: its 14.2 percent death-flag rate is the highest in the class, a function not of therapeutic-dose toxicity but of citalopram's comparatively narrow therapeutic index in overdose and its dose-dependent QT prolongation — the same pharmacology that drove the 2011 FDA dose restriction. Third, the SNRIs with the highest overdose-toxicity and discontinuation profiles (venlafaxine at 10.1 percent death) sit above the SSRIs with cleaner overdose profiles (escitalopram at 7.8 percent), a ranking consistent with the published toxicology literature. The newer, lower-volume SNRIs (desvenlafaxine, levomilnacipran) show lower serious and death shares largely because of denominator and reporting-maturity effects, not because they are intrinsically safer.

The class-level outcome profile

A report can carry more than one outcome, so these shares sum to more than 100 percent. The Fatal outcome count (61,509) sits below the death-flag measure (87,261) because many death-flagged reports carry unresolved outcome fields. As with every FAERS outcome table, none of these percentages are incidence — and for this class the fatal count is especially not comparable across drugs, because overdose-toxicity reporting concentrates in the agents most commonly involved in intentional self-harm.

The suicidality, serotonergic, and electrolyte footprint

Stripping out the non-specific terms (drug ineffective, fatigue, nausea, dizziness, headache, anxiety) that dominate the raw reaction list for any high-volume psychiatric class, the label-relevant adverse-event families leave a clear footprint:

The suicidality family (55,342 reports) is the class-defining reporting cluster, and it must be read against the regulatory history, not as a count of drug-caused events. Completed suicide is coded on 22,180 reports and suicidal ideation on 19,985, but FAERS cannot separate a drug-attributable emergent ideation from the underlying disorder it was prescribed to treat, and the reports are inflated by the very warning that generated them.

The boxed warning — and its unintended downstream effect

The defining regulatory event for the class is the October 2004 boxed warning on antidepressants and suicidality in pediatric patients, extended in May 2007 to adults aged 18–24. The FDA's meta-analysis placed the excess risk of suicidal thoughts and behaviors at roughly 4 percent on drug versus 2 percent on placebo in pooled pediatric trials, with no completed suicides in the analyzed trials and a reduction in suicidality in patients aged 65 and older. The warning reshaped prescribing — peer-reviewed analyses have found antidepressant prescribing fell on the order of 20 to 50 percent in adolescents and young adults after the warnings — and a 2024 systematic review in Health Affairs concluded that the warnings were associated with declines in depression-related visits and increases in psychotropic-drug poisonings (a proxy for suicide attempts) and in suicide deaths among youth, an unintended consequence the surveillance and access community continues to debate.

The FAERS suicidality counts here are part of that reporting ecosystem: the warnings produced durable reporter attention to suicidal-ideation terms, so the volume is as much a measure of the warning's surveillance footprint as of the underlying events. For pharmacovigilance and medical-affairs teams, the operational implication is that suicidality signal detection in this class requires a denominator-adjusted, age-stratified disproportionality design — the raw counts are not actionable on their own.

Serotonin syndrome, hyponatraemia, and the citalopram QT cap

The serotonergic and electrolyte signals are the second tier of the class profile. Serotonin syndrome (11,850 reports) is the headline pharmacodynamic risk and sits in the Warnings and Precautions section of every SSRI and SNRI label; it clusters toward the highest-serotonergic agents (fluoxetine and sertraline carry the most reports in this read) and is amplified by serotonergic-drug combinations — a 2025 FAERS analysis found SSRI–SNRI combinations generated the largest serotonin-syndrome signal (reporting odds ratio 25.42) and that fluvoxamine carried the highest per-report serotonin-syndrome risk of any SSRI (ROR 2.66), with SSRI–opioid and SSRI–antidepressant pairings the most frequently reported combinations. Hyponatraemia and SIADH (10,584 reports) is the underappreciated risk in older patients — SSRIs and SNRIs stimulate antidiuretic-hormone release via hypothalamic serotonergic pathways, the effect is not dose-dependent, and the SNRIs carry a dual noradrenergic mechanism that the comparative literature links to higher clinically relevant hyponatraemia risk.

The QT signal (7,355 reports) is the one that produced a hard dose cap. In August 2011, the FDA restricted citalopram to a maximum of 40 mg per day (20 mg in patients over 60, in hepatic impairment, or who are CYP2C19 poor metabolizers) after a thorough-QT study showed dose-dependent QTc prolongation of 8.5 ms at 20 mg, an estimated 12.6 ms at 40 mg, and 18.5 ms at 60 mg, with post-marketing torsades-de-pointes reports. A 2025 FAERS Bayesian disproportionality analysis of antidepressant arrhythmias confirmed citalopram carries the highest QT-prolongation constituent ratio of any antidepressant (versus duloxetine OR 10.10, sertraline 2.31, fluoxetine 1.46), with escitalopram second. The descriptive count here — citalopram leading the QT family at 1,818 reports — is the reporting echo of that dose restriction, and it is why citalopram also carries the class's highest death-flag rate: the same narrow-overdose pharmacology that caps its dose drives its serious-outcome reporting.

The reporting trajectory, by year

The trajectory tracks the reporting ecosystem more than a safety trend. The 2010–2012 step-up coincides with the post-generic-escitalopram and post-generic-duloxetine volume expansion and the citalopram dose-restriction communication; the sharp 2014–2015 jump to a 73,779-report peak coincides with a broadening of direct-to-consumer and consumer-reporting channels; and the post-2019 decline to a ~48,000-report steady state reflects reporting normalization and the maturation of the FAERS intake. The 2026 figure (11,773) is a partial year through the extract cut-off.

What this means for pharmacovigilance, medical-affairs, and access teams

• For pharmacovigilance teams, the 9.0 percent death-flag rate is an overdose-toxicity artifact, not a therapeutic-dose signal. Read the fatal counts against the class's intentional-self-harm reporting base; the signals that matter are denominator-adjusted suicidality monitoring and the serotonergic, electrolyte, and QT families.
• For medical-affairs and label teams, the 2011 citalopram QT cap remains the controlling precedent. Any QT data on a new serotonergic agent will be read against the citalopram 40 mg framework, and the dose-limit logic is now standard regulatory expectation for the class.
• For behavioral-health market-access and pharmacy-benefit teams, citalopram's overdose-toxicity and QT profile and the SNRIs' discontinuation burden inform quantity-limit and risk-management design. The discontinuation-syndrome family (23,466 reports, concentrated in duloxetine, paroxetine, and venlafaxine — the short-half-life agents) supports taper and quantity-limit policies on those specific drugs.
• For dossier and evidence teams, the suicidality counts cannot be quoted without the warning-history and exposure context. A sertraline or fluoxetine suicidality number is partly a measure of the 2004–2007 boxed warning's surveillance footprint; the comparable figure for a newer agent reflects a different reporting baseline.
• For policy teams, the unintended-consequences literature on the boxed warning is now part of the access conversation. The trade-off between suicidality-risk communication and under-treatment of depression is a live policy question that payer and formulary decisions increasingly touch.

Sources

• US FDA, openFDA FAERS (Adverse Event Reporting System) extract, export dated 2026-06-08 (20,328,575 total reports). Per-drug, class-union, outcome, AE-family, and trajectory aggregates computed by PharmaDossier from the public FAERS extract. https://open.fda.gov/data/faers/
• US FDA. "FDA Drug Safety Communication: Abnormal heart rhythms associated with high doses of Celexa (citalopram hydrobromide)." August 2011 — citalopram dose restriction to 40 mg/day (20 mg in patients over 60, hepatic impairment, or CYP2C19 poor metabolizers); thorough-QT study showing 8.5 ms (20 mg), ~12.6 ms (40 mg), and 18.5 ms (60 mg) QTc prolongation. https://www.fda.gov/Drugs/DrugSafety/ucm297391.htm
• US FDA. "Suicidality in Children and Adolescents Being Treated With Antidepressant Medications." October 2004 boxed warning; expanded 2007 to adults aged 18–24 — ~4 percent on-drug versus ~2 percent on-placebo suicidal thoughts and behaviors in pooled pediatric trials. https://www.fda.gov/Drugs/DrugSafety/InformationbyDrugClass/UCM096273
• Lu CY, et al. "Changes in antidepressant use by young people and suicidal behavior after FDA warnings and media coverage: quasi-experimental study." BMJ 2014;348:g3596 — 20–50 percent prescribing declines in adolescents and young adults after the warnings. https://www.bmj.com/content/348/bmj.g3596
• Glober T, et al. "The FDA Antidepressant Warnings and Suicide Risk Among Youth: A Systematic Review." Health Affairs 2024 — unintended declines in depression care and increases in psychotropic-drug poisonings and suicide deaths. https://www.healthaffairs.org
• Frontiers in Psychiatry 2025. "Arrhythmic events pertinent with antidepressants: a Bayesian disproportionality analysis mining the FDA Adverse Event Reporting System database" — citalopram highest QT-prolongation constituent ratio (vs duloxetine OR 10.10, sertraline 2.31, fluoxetine 1.46); escitalopram second. https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2025.1637471/full
• Liu B, et al. "The risk of hyponatremia induced by SSRIs and SNRIs: a systematic review and meta-analysis." 2025 — SIADH-mediated hyponatraemia mechanism via hypothalamic 5-HT2C stimulation; higher clinically relevant risk with SNRIs. https://pmc.ncbi.nlm.nih.gov/articles/PMC12323117
• Zhou Q, et al. "Selective Serotonin Reuptake Inhibitors and Risk of Serotonin Syndrome as Consequence of Drug-Drug Interactions: Analysis of the FDA Adverse Event Reporting System." 2025 — fluvoxamine highest SSRI serotonin-syndrome ROR (2.66); SSRI–SNRI combination signal (ROR 25.42); SSRI–opioid and SSRI–antidepressant pairings most reported. https://pmc.ncbi.nlm.nih.gov/articles/PMC12140641
• US FDA, fluoxetine prescribing information (Boxed Warning on suicidality; Warnings on serotonin syndrome, hyponatremia, and QT prolongation). https://nctr-crs.fda.gov/fdalabel/services/spl/set-ids/887fc670-db67-4cfe-967b-46b38375dae5/spl-doc
• US FDA, Drugs@FDA approval records — fluoxetine (Prozac, 1987), sertraline (Zoloft, 1991), paroxetine (Paxil, 1992), fluvoxamine (Luvox, 1994), venlafaxine (Effexor, 1993), citalopram (Celexa, 1998), escitalopram (Lexapro, 2002), duloxetine (Cymbalta, 2004), desvenlafaxine (Pristiq, 2008), levomilnacipran (Fetzima, 2013). https://www.accessdata.fda.gov/scripts/cder/daf/