Trazodone and SNRIs (Venlafaxine, Duloxetine): Interaction Guide

What the labels actually warn about

The trazodone label is explicit. Serotonin syndrome risk is increased when trazodone is co-administered with other serotonergic drugs, and the examples it gives include SNRIs, SSRIs and triptans. It also states that serotonin syndrome can occur with trazodone alone [1].

The Effexor XR label lists other serotonergic drugs the same way: other SNRIs, SSRIs, triptans, tricyclic antidepressants, opioids, lithium, buspirone, amphetamines, tryptophan and St. John's wort. It warns that these combinations raise serotonin syndrome risk and says to monitor for symptoms [2].

The signs both labels describe are the same set: mental status changes such as agitation, hallucinations, delirium or coma; autonomic instability such as fast heart rate, labile blood pressure, dizziness, sweating, flushing and fever; neuromuscular signs such as tremor, rigidity, muscle twitching, overactive reflexes and incoordination; seizures; and nausea, vomiting or diarrhea [1][2].

The trazodone label tells clinicians to watch for these signs, particularly when trazodone is started, and to consider stopping trazodone and the other serotonergic drug if serotonin syndrome occurs [1]. The venlafaxine label says to discontinue and start supportive treatment if it occurs [2].

The pharmacokinetic question: does an SNRI raise trazodone exposure?

Trazodone's active metabolite, mCPP, is made by CYP3A4. In 16 human liver microsome preparations and 8 expressed human CYP enzymes, mCPP formation tracked CYP3A4 activity alone, and ketoconazole blocked it [4]. A 2025 in vitro study found that CYP2D6, CYP2C19 and flavin monooxygenases also contribute to trazodone metabolism, but to other, inactive metabolites; CYP3A4 remained the primary enzyme [6].

Duloxetine is a moderate CYP2D6 inhibitor: at 60 mg twice daily it raised desipramine exposure about threefold [3]. Venlafaxine's label says coadministration raises the peak concentration and total exposure of a CYP2D6 substrate and suggests considering a lower dose of that substrate [2]. The trazodone label's examples of strong CYP3A4 inhibitors are itraconazole, ketoconazole, clarithromycin and indinavir; neither duloxetine nor venlafaxine is among them [1].

The trazodone label does not identify trazodone as a CYP inhibitor, and trazodone does not appear in the FDA's tables of CYP substrates, inhibitors and inducers [1][5]. The label separately notes that trazodone can raise digoxin or phenytoin levels, with no mechanism stated [1].

What clearly moves trazodone is CYP3A4. In 10 healthy volunteers, low-dose ritonavir cut trazodone's oral clearance from 155 to 75 mL/min and lengthened half-life from 6.7 to 14.9 hours after a single 50 mg dose; peak levels rose from 842 to 1,125 ng/mL. Sedation, fatigue and side effects increased, three volunteers had nausea, dizziness or hypotension, and one fainted [7]. The label's instruction is to consider a lower trazodone dose with a strong CYP3A4 inhibitor, and a higher dose with a strong inducer such as rifampin, carbamazepine, phenytoin or St. John's wort [1].

Blood pressure effects differ by SNRI

Trazodone blocks alpha-1 adrenergic receptors, which its label links to postural hypotension [1]. Orthostatic hypotension and syncope are listed warnings, and in outpatient MDD trials hypotension was reported by 4% of patients on trazodone versus 0% on placebo, and syncope by 5% versus 1% [1]. The label notes that patients on blood pressure medicines may need a lower dose of the blood pressure drug [1].

Venlafaxine's label warns of dose-related increases in systolic and diastolic pressure, with sustained hypertension in 19 of 705 MDD patients (3%) at doses of 75 to 375 mg a day [2]. A rise of at least 15 mm Hg in supine diastolic pressure to 105 or higher occurred in 1.4% versus 0.9% on placebo, and a rise of at least 20 mm Hg in systolic pressure to 180 or higher in 1% versus 0.3% [2]. A meta-analysis of 3,744 depressed patients found the venlafaxine effect highly dose dependent and clinically significant only above 300 mg a day of the immediate-release form [8].

Duloxetine's average effect in adult trials was small: mean systolic 0.5 and diastolic 0.8 mm Hg, against falls of 0.6 and 0.3 on placebo, with no significant difference in sustained elevated blood pressure [3]. That label still says to check blood pressure before and periodically, and carries warnings for orthostatic hypotension, falls and fainting [3].

Bleeding risk adds up

The trazodone label warns that aspirin, NSAIDs, other antiplatelet drugs, warfarin and other anticoagulants add to bleeding risk. For patients on warfarin it says to monitor INR when trazodone is started or stopped, and notes that prothrombin time may rise or fall [1]. The Effexor XR label carries the same antiplatelet and anticoagulant warning and advises close monitoring when the antidepressant is started or stopped [2]. Duloxetine 60 or 120 mg did not change warfarin INR or warfarin pharmacokinetics in 15 healthy subjects, though monitoring at start and stop is still advised [3].

QT prolongation is a trazodone problem here

Trazodone prolongs the QT/QTc interval. The label says to avoid it in known QT prolongation, a history of arrhythmia, low potassium or magnesium, congenital long QT and the early recovery phase after a heart attack, and to avoid combining it with other QT-prolonging drugs [1]. Torsades de pointes and ventricular tachycardia have been reported after marketing at doses of 100 mg a day or less [1]. Venlafaxine looks different: a thorough QT study in 54 healthy adults found no significant QT prolongation at 450 mg a day, twice the maximum labeled dose [2].

MAOIs and the washout gaps

Trazodone is contraindicated in anyone taking an MAOI or within 14 days of stopping one, including linezolid and intravenous methylene blue, and 14 days must pass after stopping trazodone before an MAOI antidepressant is started [1]. Effexor XR is contraindicated in patients taking an MAOI or within 14 days of stopping one [2]. For duloxetine the label gives a 5-day gap after stopping duloxetine and 14 days after stopping an MAOI [3]; desvenlafaxine and levomilnacipran labels give 7 days after stopping them [9][10].

Sleep use sits outside the label

Trazodone's only labeled use is major depressive disorder in adults, and the label gives no insomnia indication or insomnia dose [1]. Drowsiness is the most common label effect in outpatients, reported by 41% versus 20% on placebo [1].