Trazodone FAERS Safety Signals: What Post-Market Surveillance Data Shows

Trazodone (originally marketed as Desyrel, now available only as a generic) is a serotonin antagonist and reuptake inhibitor (SARI) that received FDA approval in 1981 for major depressive disorder. Most of its current use is off-label, at low doses, as a sleep aid. That mismatch between the population studied in the original approval trials and the population taking the drug today is the central reason its post-market safety record deserves its own look, separate from the label.

The direct answer

FAERS surveillance of trazodone has not overturned its basic risk-benefit profile: it remains an FDA-approved treatment for depression and a widely used, if off-label, sleep aid. What the surveillance record has done is sharpen where the real risk concentrates. Priapism is a rare but mechanistically well-understood event tied to trazodone's alpha-1 adrenergic blockade and occurs across the dose range, including low insomnia doses. Falls and orthostatic hypotension are the more common practical hazard in patients over 65 taking trazodone for sleep. QT prolongation and serotonin syndrome are real but context-dependent signals that concentrate in patients with cardiac risk factors or multi-drug serotonergic regimens, not in the average low-dose user. None of these signals were derived from randomized trial precision; they come from voluntary, passive reporting, so they establish plausibility and direction, not incidence.

What FAERS can and cannot tell you

The FDA Adverse Event Reporting System is a voluntary, passive surveillance database. Reports come from clinicians, patients, and manufacturers, and can be submitted whenever someone believes a drug may be linked to an adverse event. The FDA's own FAERS Public Dashboard lets anyone query aggregate counts by drug and reaction term, but the FDA is explicit that a report in the system does not establish that the drug caused the event.

Three limitations matter specifically for trazodone. First, reporting volume tracks media attention and labeling changes as much as it tracks true event rates, so raw counts over time are not a clean incidence curve. Second, voluntary systems are widely understood to capture only a fraction of actual adverse events, so FAERS cannot be used to calculate a population-level incidence rate for any of the signals below. Third, and specific to this drug, the population using trazodone today (mostly adults taking 25 to 100 mg for insomnia, often older adults) differs substantially from the population in the original 1970s-80s registration trials for depression at doses up to 400 mg. A disproportionality signal in FAERS can reflect who is taking the drug and at what dose as much as it reflects a fixed property of the molecule.

Priapism: the signal that shaped the label

Priapism is the adverse event most specifically associated with trazodone among psychiatric medications, and it is one of the reasons the drug's label carries a distinct warning. The mechanism is plausible and well described: trazodone's active metabolite blocks alpha-1 adrenergic receptors in penile smooth muscle, which can interfere with detumescence.

Older case-series literature has put the incidence in a range of roughly 1 in 6,000 to 1 in 8,000 male patients, but this figure comes from decades-old case reports rather than a systematic modern registry, and it should be treated as an approximate order of magnitude rather than a precise current rate. Verification against the primary literature is warranted before quoting a specific number to a patient. What is more consistently reported across cases is that priapism is not confined to high antidepressant doses; it has been reported at doses as low as 50 mg, the range commonly used for insomnia. A prolonged erection lasting more than four hours is a urologic emergency regardless of the dose that preceded it, because ischemic damage and permanent erectile dysfunction can result from delay.

Cardiac signals: QT prolongation and arrhythmia

Trazodone blocks the cardiac hERG potassium channel in vitro, which is the same channel implicated in drug-induced QT prolongation for many psychotropic medications. This gives the cardiac signal a plausible mechanistic basis, and post-market reports of QT prolongation and arrhythmia involving trazodone exist in FAERS.

The clinical qualifier matters more than the raw signal count. Reports concentrate in patients who already have structural heart disease, electrolyte abnormalities, or who are taking other QT-prolonging drugs, and the effect is dose-related, so it is more relevant at antidepressant-range doses (300 to 400 mg/day) than at typical low-dose insomnia use. The prescribing information advises caution in patients with cardiac disease and supports ECG monitoring when other risk factors are present; the exact current wording should be checked against the label on Drugs@FDA rather than assumed, since label language is periodically revised. In overdose, trazodone has been linked to cardiac arrhythmia less often than tricyclic antidepressants, which is one reason it is generally considered to have a wider (though not unlimited) margin of safety in overdose compared with older antidepressant classes.

Serotonin syndrome when trazodone is added to an existing regimen

Serotonin syndrome reports involving trazodone rise in parallel with polypharmacy: the drug is frequently layered onto an existing SSRI or SNRI regimen specifically because it is being added as a sleep aid rather than as the primary antidepressant. The combination is common in practice and most patients tolerate it, but the interaction is pharmacologically real, not theoretical, since both drug classes increase synaptic serotonin activity through different mechanisms.

A decision framework for reading a trazodone FAERS signal

Not every signal in this article changes what a clinician or patient should do. The table below is organized around the question that actually matters at the point of care: does this signal change the decision in front of you, or is it background noise from a drug used by millions of people for over four decades?

The pattern across every row is the same: the FAERS signal is a starting point for asking who is at risk, not a reason to treat every patient as if they carry the same risk.

Falls and orthostatic hypotension in older adults

Orthostatic hypotension, driven by the same alpha-1 blockade responsible for priapism, is one of the most frequently reported adverse events for trazodone, and it has outsized consequences in older adults because a fall in this population can lead to a hip fracture and a cascade of further decline. An observational cohort study of nursing home residents has reported a meaningfully elevated fracture risk associated with trazodone use; the precise magnitude of that risk should be checked against the primary study before it is quoted to a patient, since exact figures vary by cohort and methodology.

The American Geriatrics Society's Beers Criteria, the standard reference for potentially inappropriate medications in older adults, lists trazodone among drugs to use with caution because of fall risk, orthostasis, and sedation. This is a guideline-level recommendation, one step up the evidence hierarchy from a single observational study, and it is the more durable reason to start low and monitor closely in this age group. Case narratives consistently describe falls clustering at night or on first standing, consistent with peak sedative and vasodilatory effect after a bedtime dose; blood pressure can drop meaningfully after the first dose, before tolerance develops over subsequent days.

Suicidality: a class warning, not a trazodone-specific one

In 2004 the FDA required a black box warning across antidepressants, including trazodone, for increased suicidal thinking and behavior in patients under 25, later extended to include young adults through age 24. The FDA's black box warning on antidepressants and suicidality describes the basis for this warning, which came from pooled analysis of short-term antidepressant trials showing greater risk of suicidal ideation compared with placebo, concentrated in the first one to two months of treatment.

This warning applies to trazodone as a class member, but its practical relevance depends heavily on how the drug is being used. Most current trazodone prescribing is low-dose, off-label, and in adults well past the age range the warning targets. The warning is most directly relevant when trazodone is used at antidepressant doses in adolescents or young adults, a smaller slice of current prescribing than the insomnia use case.

Hepatotoxicity: rare, but real enough to name

A small cluster of hepatotoxicity reports, including elevated liver enzymes, cholestatic hepatitis, and rare acute liver failure, appears in the post-market record for trazodone. The NIH LiverTox entry for trazodone describes this as a rare, idiosyncratic reaction, typically emerging one to six weeks after starting therapy and usually reversible after the drug is stopped. Recurrence on re-challenge has been described in some reports, which is one of the stronger forms of pharmacovigilance evidence for causation even at low overall event counts. Routine liver monitoring is not standard practice for trazodone given how infrequent this signal is; the practical implication is to check liver function promptly if a patient develops unexplained jaundice, nausea, or right-upper-quadrant pain during treatment, rather than to screen everyone in advance.

What changed on the label, and what that process actually shows

Trazodone's FDA label has been revised multiple times since 1981, and the pattern each time has followed the same sequence: a signal accumulates in post-market reporting, the FDA reviews it, and the label is updated. The priapism warning, the 2004 suicidality black box, and cardiac and serotonergic drug-interaction language all followed this path. The current, authoritative label text should always be checked directly on Drugs@FDA rather than assumed from a secondary summary, since exact wording is revised periodically and quoting outdated language as current is a common error in secondary sources.

This is the intended function of post-market surveillance, and it is also its limit. Pre-approval trials for trazodone enrolled a few hundred patients over weeks to months; FAERS aggregates reports across millions of patient-years and every combination of comorbidity and co-medication a trial could never enroll for. That scale is why rare events like priapism or idiosyncratic hepatotoxicity surfaced after approval rather than during trials. But FAERS cannot substitute for a controlled study when the question is how large an effect is, only whether a signal exists and roughly who it concentrates in. Confirming an exact incidence, relative risk, or reporting odds ratio for any of the signals above requires going to the specific epidemiological study or FDA safety review behind it, not to the aggregate FAERS count alone.

What this means at the point of prescribing

None of this changes the basic calculus that trazodone is an effective, generally well-tolerated option for depression and, off-label, for insomnia. What the surveillance record supports is a more targeted monitoring conversation: ask male patients about priapism risk before the first prescription regardless of the intended dose; check for cardiac risk factors and other QT-prolonging medications before adding trazodone in a patient with heart disease; reconcile the full serotonergic medication list before adding trazodone to an existing SSRI or SNRI regimen; start at the lowest effective dose in patients 65 and older and counsel on rising slowly, especially at night; and treat unexplained jaundice or right-upper-quadrant pain during treatment as a reason to check liver function.

When to seek urgent care

An erection lasting more than four hours, chest pain, fainting, or an irregular heartbeat, and the combination of agitation, muscle twitching or rigidity, fever, and a racing heart (possible serotonin syndrome) are all reasons to seek emergency evaluation rather than wait for a scheduled follow-up. New jaundice, dark urine, or persistent nausea during treatment warrants prompt medical evaluation as well.

Evidence boundary

Established: trazodone can cause priapism, orthostatic hypotension with fall risk in older adults, and, in combination with other serotonergic drugs, serotonin syndrome; these mechanisms are pharmacologically well understood and consistently reported. Plausible but not precisely quantified: exact incidence rates for priapism, hepatotoxicity, and QT-related events, since FAERS and older case series cannot generate reliable population-level rates. Not established from this evidence base: that trazodone at typical low insomnia doses carries suicidality risk comparable to antidepressant-dose use in younger patients, or that any single FAERS signal count reflects a true incidence rather than reporting behavior. Readers and clinicians should treat every specific number in this article as a starting estimate to be checked against current primary sources, not a fixed clinical fact.