Serotonin syndrome is a potentially life-threatening drug toxidrome caused by excess serotonergic activity at central and peripheral 5-HT receptors. It is predominantly iatrogenic, arising from drug combinations that simultaneously enhance serotonin availability through different mechanisms, and its recognition requires distinguishing it from other hyperadrenergic and hypermetabolic syndromes with which it shares clinical features.
Serotonin syndrome results from excessive stimulation of postsynaptic 5-HT1A and 5-HT2A receptors in the central nervous system (CNS) and peripheral nervous system. The most dangerous combinations are those that simultaneously block serotonin reuptake and inhibit its enzymatic degradation, since these two mechanisms are normally redundant and each alone provides partial protection against serotonergic excess.1 The highest-risk combination in clinical practice is a selective serotonin reuptake inhibitor (SSRI) or serotonin-norepinephrine reuptake inhibitor (SNRI) combined with a monoamine oxidase inhibitor (MAOI), which blocks both the transporter-mediated removal of serotonin from the synapse and the enzyme responsible for its subsequent intraneuronal degradation. Other combinations that generate sufficient serotonergic excess to produce the syndrome include SSRIs or SNRIs with tramadol (a weak SERT inhibitor), SSRIs or SNRIs with triptans, SSRIs or SNRIs with dextromethorphan, SSRIs with linezolid (an antibiotic with reversible MAO-A inhibitory activity), and SSRIs with the antifungal agent methylene blue, which also inhibits MAO. The 5-HT2A receptor is particularly implicated in the neuromuscular manifestations of the syndrome, specifically the clonus and hyperthermia, whereas 5-HT1A excess mediates the autonomic features including diaphoresis, tachycardia, and mydriasis.1
The clinical presentation of serotonin syndrome is characterized by a triad of altered mental status, autonomic instability, and neuromuscular abnormalities, though the full triad is not invariably present in mild cases.2 Altered mental status ranges from mild agitation and anxiety to frank confusion and delirium. Autonomic features include hyperthermia, diaphoresis, tachycardia, hypertension, and tachypnea. The neuromuscular findings are the most diagnostically specific: clonus (both spontaneous and inducible), hyperreflexia, and muscle rigidity, which in severe cases produces a temperature-driven rhabdomyolysis and hyperthermia that can exceed 41.1℃ (106℉). The Hunter Toxicity Criteria Decision Rules, the most sensitive and specific validated instrument for diagnosing serotonin syndrome, require the presence of a serotonergic agent and one of the following: spontaneous clonus; inducible clonus with agitation or diaphoresis; ocular clonus with agitation or diaphoresis; tremor with hyperreflexia; or hypertonic rigidity with temperature above 38℃ and either ocular or inducible clonus.2 The Hunter Criteria have sensitivity of 84% and specificity of 97% for serotonin syndrome diagnosed by a clinical toxicologist, substantially outperforming the earlier Sternbach criteria.
Neuroleptic malignant syndrome (NMS) and serotonin syndrome are the two principal hypermetabolic syndromes encountered in psychiatric pharmacology, and their distinction is clinically consequential because their management differs. Both can present with hyperthermia, altered mental status, and rigidity. The distinguishing features favor serotonin syndrome when: onset is rapid (hours rather than days to weeks as in NMS); neuromuscular abnormality is characterized by clonus and hyperreflexia rather than the lead-pipe rigidity and hyporeflexia of NMS; bowel sounds are hyperactive rather than absent; and the history includes a serotonergic drug combination rather than antipsychotic initiation or dose escalation.2 In NMS, the dopamine (DA)-blocking mechanism produces bradykinesia and the extrapyramidal rigidity that reduces heat dissipation; in serotonin syndrome, the 5-HT2A receptor-mediated rigidity is typically more severe in the lower extremities and accompanied by clonus. Creatine kinase (CK) elevation occurs in both but tends to be more pronounced in NMS. Leukocytosis is common in NMS, less so in serotonin syndrome.
Management of serotonin syndrome is stratified by severity. Mild cases require only discontinuation of the causative serotonergic agent and supportive care with benzodiazepines for agitation and autonomic control. Moderate-to-severe cases require hospital admission, continuous monitoring, active cooling for hyperthermia, intravenous fluid resuscitation, and aggressive benzodiazepine sedation. The specific serotonin antagonist cyproheptadine, a first-generation antihistamine with potent 5-HT2A and 5-HT1A antagonist properties, is administered orally or via nasogastric tube at an initial dose of 12 mg followed by 2 mg every two hours until symptom control, not to exceed 32 mg in 24 hours.3 Cyproheptadine is not supported by randomized controlled trials given the difficulty of conducting such trials in a toxicological emergency, but its receptor pharmacology is mechanistically appropriate and its use is supported by case series and expert consensus. Life-threatening serotonin syndrome with temperatures above 41.1℃ requires endotracheal intubation, neuromuscular paralysis to terminate thermogenesis from muscle rigidity, and intensive care unit management. Dantrolene, which is used in malignant hyperthermia and occasionally in NMS, is not recommended in serotonin syndrome because the primary driver of hyperthermia is 5-HT receptor-mediated rigidity rather than a defect in calcium channel regulation in skeletal muscle.
SSRI or SNRI + any MAOI (including linezolid or methylene blue) — absolute contraindication. SSRI + tramadol — underrecognized risk, particularly in postoperative pain management. SSRI + dextromethorphan — relevant in patients self-medicating with over-the-counter cough preparations. Fluoxetine has a five-week washout period before MAOI initiation due to norfluoxetine's extended half-life; all other SSRIs require two weeks.
Antidepressant adverse effects are not random — they follow directly from the receptor binding profiles of each class. Understanding the mechanistic basis of each adverse effect allows the clinician to predict which symptoms a given patient is likely to experience, to select agents with more favorable profiles for patients with specific comorbidities, and to manage adverse effects rationally without empirical trial and error.
Sexual dysfunction is the most common reason for antidepressant discontinuation in patients who have achieved an adequate therapeutic response, occurring in 40% to 65% of patients taking SSRIs and SNRIs on prospective assessment with validated instruments, substantially higher than the rates reported in passive clinical trial adverse event collection.4 The mechanism involves sustained 5-HT2 receptor activation in spinal reflex arcs mediating ejaculation and orgasm, inhibition of nitric oxide synthase in genital vasculature, and prolactin elevation via tuberoinfundibular dopamine pathway suppression. Management options include dose reduction if clinically feasible; switching to an agent with a lower sexual dysfunction burden such as bupropion, mirtazapine, or vortioxetine; augmenting with bupropion (which enhances dopaminergic and noradrenergic tone in circuits mediating sexual response); or adding a phosphodiesterase type 5 inhibitor for erectile dysfunction in male patients, which addresses the nitric oxide-mediated vascular component.9 Drug holidays, a strategy sometimes employed by patients who reduce or omit doses on weekends, have limited utility because the relevant receptor adaptations that suppress sexual function persist over days, and the risk of breakthrough depression from intermittent dosing outweighs the modest symptomatic benefit.
Weight gain with antidepressants occurs through multiple mechanisms and varies considerably across classes. Paroxetine produces the most consistent and clinically significant weight gain among SSRIs, attributable to its antihistaminic activity and anticholinergic effects that reduce satiety signaling and increase appetite. Mirtazapine causes substantial weight gain in most patients through potent histamine H1 receptor blockade, which reduces hypothalamic satiety signaling, combined with 5-HT2C antagonism, which disinhibits appetite-promoting orexinergic and neuropeptide Y circuits.5 TCAs produce weight gain through a combination of H1 blockade and muscarinic antagonism. In contrast, bupropion is weight-neutral or modestly weight-reducing in most patients and is the only antidepressant approved as an adjunct for weight management when combined with naltrexone. SSRIs other than paroxetine, SNRIs, and vortioxetine are generally weight-neutral over the first three to six months, though long-term treatment beyond one year may be associated with a modest increase in body weight even with these agents, which is thought to reflect the resolution of depression-associated weight loss and normalization of appetite rather than a direct pharmacological effect on energy balance.
The most clinically significant cardiovascular effect specific to modern antidepressants is QTc prolongation. Citalopram produces dose-dependent QTc prolongation through blockade of the cardiac hERG potassium channel (IKr current), a mechanism identified post-marketing; the FDA issued a safety communication in 2011 recommending a maximum dose of 40 mg/day in most patients and 20 mg/day in patients over 60 years, those with hepatic impairment, poor CYP2C19 metabolizers, or those taking concomitant CYP2C19 inhibitors.6 Escitalopram, the S-enantiomer of citalopram, also produces dose-dependent QTc prolongation and carries the same 20 mg/day maximum dose in the high-risk populations listed. No other SSRI or SNRI produces clinically meaningful QTc prolongation at therapeutic doses. TCAs carry a substantially greater cardiovascular burden through combined cardiac sodium channel blockade (Nav1.5 — covered in Module 04), alpha-1 adrenergic blockade producing orthostatic hypotension, and QTc prolongation. Venlafaxine at higher doses produces a dose-dependent increase in blood pressure through NET inhibition, requiring blood pressure monitoring when doses exceed 150 mg/day.
Nausea is the most common early adverse effect of SSRIs and SNRIs, occurring in 20% to 30% of patients and typically resolving within one to two weeks as 5-HT3 receptors in the gastrointestinal tract and area postrema desensitize to sustained serotonergic stimulation. Initiating at the lowest available dose and titrating slowly, administering the dose with food, or starting with extended-release formulations reduces early nausea without compromising eventual therapeutic efficacy. Diarrhea occurs more frequently with sertraline than other SSRIs and is typically dose-dependent and manageable. GI bleeding risk is elevated with SSRIs because serotonin in platelets, derived from uptake from plasma rather than endogenous synthesis, is required for platelet aggregation; SERT blockade depletes platelet serotonin and impairs platelet plug formation.7 This effect is additive with nonsteroidal anti-inflammatory drugs (NSAIDs) and anticoagulants, and concomitant SSRI plus NSAID use substantially increases upper GI bleeding risk. Proton pump inhibitor co-prescription attenuates, but does not eliminate, this excess risk.
H1 blockade predicts sedation and weight gain (mirtazapine, doxepin, paroxetine, TCAs). Alpha-1 blockade predicts orthostatic hypotension and falls (TCAs, trazodone). mAChR blockade predicts anticholinergic burden (tertiary TCAs, paroxetine). SERT inhibition predicts nausea, sexual dysfunction, platelet effect (all SSRIs, SNRIs). Nav1.5 blockade predicts cardiac conduction toxicity in overdose (TCAs). hERG block predicts QTc prolongation (citalopram, escitalopram, TCAs).
Antidepressants generate drug interactions through two distinct CYP mechanisms: as substrates, their plasma concentrations are altered when co-administered drugs inhibit or induce the enzymes that metabolize them; as inhibitors, they alter the plasma concentrations of co-administered drugs that share the same metabolic pathway. The second mechanism is of greater clinical consequence because antidepressants are among the most potent CYP inhibitors in common clinical use.
Fluoxetine and paroxetine are both potent inhibitors of CYP2D6 and can convert an extensive metabolizer into a phenotypic poor metabolizer during treatment, a phenomenon termed phenocopying.8 The clinical consequences are extensive because CYP2D6 metabolizes a large proportion of commonly prescribed drugs. Tamoxifen requires CYP2D6-mediated conversion to endoxifen, its active metabolite responsible for the majority of the anti-estrogenic efficacy in hormone receptor-positive breast cancer; fluoxetine or paroxetine co-prescription substantially reduces endoxifen plasma concentrations and is associated with reduced tamoxifen efficacy and increased breast cancer recurrence risk in observational studies.8 This interaction has prompted oncology guidelines recommending against fluoxetine and paroxetine in patients on tamoxifen; sertraline, citalopram, escitalopram, or venlafaxine are preferred alternatives. Other CYP2D6 substrates affected include TCAs (elevated concentrations and toxicity risk), codeine (reduced conversion to morphine, therapeutic failure in extensive metabolizers and opioid toxicity in ultra-rapid metabolizers), metoprolol (increased beta-blockade), and risperidone and haloperidol (elevated antipsychotic concentrations).
Fluvoxamine, an SSRI primarily used for obsessive-compulsive disorder (OCD), is a uniquely broad CYP inhibitor with potent inhibitory activity at CYP1A2, CYP2C19, and CYP3A4, making it the SSRI with the highest drug interaction burden.9 Its CYP1A2 inhibition substantially elevates concentrations of clozapine, olanzapine, theophylline, and caffeine. Its CYP2C19 inhibition increases concentrations of omeprazole, diazepam, phenytoin, and warfarin. Clozapine co-administration with fluvoxamine is particularly dangerous, as CYP1A2 is the primary metabolic route for clozapine, and fluvoxamine can increase clozapine concentrations three-fold or more, substantially raising the risk of clozapine-associated seizures, agranulocytosis threshold lowering, and cardiotoxicity. Theophylline toxicity, with its narrow therapeutic window, is another interaction requiring active management. In practice, fluvoxamine should be prescribed with a comprehensive medication review; patients on clozapine should use an alternative SSRI for OCD.
Several antidepressants are themselves significantly affected when co-administered with CYP inducers or inhibitors. Carbamazepine, a potent inducer of CYP3A4, CYP2C9, and CYP2C19, substantially reduces the plasma concentrations of sertraline, citalopram, escitalopram, mirtazapine, and TCAs. Rifampin, another broad inducer, can reduce antidepressant concentrations to subtherapeutic levels within days of co-initiation.9 St. John's wort (Hypericum perforatum) is a clinically important inducer of CYP3A4 and P-glycoprotein, reducing concentrations of sertraline and other SSRI substrates while simultaneously contributing its own mild SERT inhibitory activity, creating a dual mechanism for serotonin toxicity when combined with prescription antidepressants. Conversely, azole antifungals such as fluconazole and ketoconazole are potent CYP2C19 and CYP3A4 inhibitors that can substantially increase concentrations of citalopram, escitalopram, and TCAs; this interaction is particularly relevant in immunocompromised patients requiring antifungal therapy while maintained on antidepressants.
Antidepressants interact with warfarin through two distinct mechanisms that are pharmacokinetically additive. SSRIs and SNRIs deplete platelet serotonin through SERT inhibition, impairing platelet aggregation and potentiating warfarin's anticoagulant effect on the clotting cascade without changing warfarin pharmacokinetics. Fluvoxamine, fluoxetine, and to a lesser degree paroxetine inhibit CYP2C9, which metabolizes the more potent S-warfarin enantiomer, pharmacokinetically increasing warfarin concentrations and international normalized ratio (INR).9 The combined pharmacodynamic and pharmacokinetic interaction requires INR monitoring within one to two weeks of antidepressant initiation or dose change in patients on warfarin. Among SSRIs, citalopram and escitalopram have the least CYP2C9 inhibitory activity and are the preferred choices in patients on warfarin when SSRI therapy is indicated.
Fluoxetine or paroxetine + tamoxifen: use an alternative SSRI. Fluvoxamine + clozapine: contraindicated in practice. Any SSRI + warfarin: INR check within 2 weeks. St. John's wort + any prescription antidepressant: avoid combination. Carbamazepine + antidepressant: expect reduced antidepressant concentrations; consider dose adjustment and concentration monitoring.
Antidepressant discontinuation syndrome (ADS) is a predictable pharmacological consequence of abrupt or rapid cessation of antidepressants in patients who have been taking them for four weeks or longer. It is not equivalent to addiction or physiological dependence in the classic sense — it does not involve craving, compulsive drug-seeking, or dose escalation — but it produces genuine physical and psychological symptoms that can be severe enough to prevent discontinuation and require clinical management.10 Recognition is important both to avoid unnecessary alarm when symptoms appear and to distinguish them from relapse, which requires a different management response.
The mechanism of ADS reflects the neuroadaptations that develop during sustained serotonergic stimulation. During chronic SSRI or SNRI treatment, presynaptic 5-HT1A autoreceptors downregulate and postsynaptic 5-HT2A receptors undergo compensatory changes in expression and sensitivity. The termination of SERT blockade produces a rapid fall in synaptic serotonin that these receptor adaptations are temporarily unable to compensate for, resulting in a transient functional serotonin deficiency state.10 The speed of onset is governed by the elimination half-life of the agent and its active metabolites: agents with short half-lives produce symptoms within one to two days of cessation, whereas fluoxetine, with an effective half-life of seven to fifteen days when norfluoxetine is included, produces symptoms only after one to two weeks of cessation if at all, because norfluoxetine self-tapers automatically. SNRIs with short half-lives, particularly venlafaxine immediate-release and desvenlafaxine, are associated with some of the most severe and rapidly appearing discontinuation syndromes among all antidepressants.
The symptoms of ADS are usefully organized by the FINISH mnemonic: Flu-like symptoms (myalgia, fatigue, sweating, chills, nausea); Insomnia (vivid dreams, nightmares, disrupted sleep architecture); Nausea (and vomiting, particularly with paroxetine and venlafaxine); Imbalance (dizziness, gait unsteadiness); Sensory disturbances (electric shock sensations or "brain zaps" — brief, painful paresthesias spreading from the head, highly characteristic and pathognomonic when present); and Hyperarousal (anxiety, agitation, irritability).11 The sensory disturbances, particularly the "brain zap" phenomenon, are the most distinctive feature and are highly useful for distinguishing ADS from relapse, since they do not occur as a manifestation of depression itself. Relapse, in contrast, presents with the gradual re-emergence of depressive symptoms over days to weeks without the acute neurological sensory features. The temporal pattern also helps: ADS symptoms begin within days of cessation and resolve within one to four weeks without retreatment, whereas relapse symptoms intensify progressively and do not self-resolve.
The risk of ADS is a direct function of elimination half-life and receptor binding selectivity. Paroxetine carries the highest risk among SSRIs because of its combination of the shortest half-life in the class and potent anticholinergic activity, which adds a cholinergic rebound component to the serotonergic withdrawal symptoms. Venlafaxine immediate-release produces severe ADS given its two-hour half-life, and clinicians frequently misattribute the rapidly appearing sensory symptoms to an intercurrent neurological condition. Sertraline, citalopram, and escitalopram carry intermediate ADS risk. Fluoxetine has the lowest ADS risk of any SSRI or SNRI due to its extended effective half-life through norfluoxetine, and can occasionally be used as a bridging agent to facilitate discontinuation of other antidepressants in patients with intractable discontinuation syndromes.11 TCAs and MAOIs also produce discontinuation syndromes, with the TCA syndrome including prominent cholinergic rebound symptoms and the MAOI syndrome including vivid dreams, agitation, and myoclonic jerks.
The primary management strategy for planned antidepressant discontinuation is gradual dose tapering rather than abrupt cessation. There is no universal tapering schedule applicable to all patients, but a general principle is that the taper should extend over at least two to four weeks for most agents and potentially months in patients who have been on high doses for prolonged periods or who have previously experienced severe ADS.11 A hyperbolic tapering strategy, in which dose reductions become proportionally smaller as the total dose decreases, is supported by receptor occupancy modeling: because the relationship between dose and receptor occupancy is non-linear (following a hyperbolic curve), equal absolute dose reductions produce proportionally larger reductions in receptor occupancy at the lower end of the dose range, where the risk of ADS is greatest. Switching to a longer-acting agent such as fluoxetine before tapering is a validated strategy for patients in whom direct tapering of paroxetine or venlafaxine has failed. If symptoms develop during or after tapering, reinstatement at the prior dose followed by a slower taper is preferable to prolonged symptom exposure.
ADS: onset within days of cessation; includes brain zaps, dizziness, flu-like symptoms; self-limited within 1–4 weeks; responds to dose reinstatement within 24 hours. Relapse: gradual re-emergence of depressive symptoms over weeks; no sensory disturbances; does not self-resolve; requires treatment decision (restart, switch, or augment).
Antidepressant prescribing in pregnancy and in older adults requires integrating pharmacokinetic changes, altered risk-benefit ratios, and population-specific safety data that often differ substantially from the general adult data on which most prescribing decisions are based. These are not edge cases — major depressive disorder (MDD) affects approximately 10% to 15% of pregnant women, and depression is among the most common psychiatric diagnoses in patients over 65.
The framing of antidepressant use in pregnancy as a binary choice between drug exposure and no exposure misrepresents the clinical reality. Untreated depression in pregnancy carries its own set of fetal and maternal risks: poor prenatal care adherence, inadequate gestational weight gain, increased rates of preterm birth, neonatal complications related to maternal HPA axis dysregulation, and elevated rates of postpartum depression and impaired mother-infant bonding.12 The decision to continue, discontinue, or initiate antidepressant therapy in pregnancy requires weighing the risk of the medication against the risk of undertreated or untreated maternal depression, not against the risk of no pharmacological exposure in a healthy patient. This risk-benefit framework should be made explicit in clinical conversations with pregnant patients.
SSRIs are the most extensively studied antidepressants in pregnancy and remain first-line for pharmacological management of MDD during gestation when drug therapy is indicated. The overall congenital malformation risk with SSRI exposure does not appear to differ significantly from background rates in adequately controlled studies that account for confounding by indication and severity of maternal illness.12 Paroxetine carries the most cautious labeling, with earlier data suggesting a possible association with ventricular septal defects at higher doses, though subsequent larger studies have not consistently replicated this finding and the absolute risk, if real, is small. The neonatal adaptation syndrome, occurring in approximately 30% of neonates exposed to SSRIs in the third trimester, consists of transient jitteriness, hypoglycemia, respiratory distress, and feeding difficulties, and typically resolves within two weeks without specific intervention.12
Persistent pulmonary hypertension of the newborn (PPHN) has been reported with late gestational SSRI exposure in observational studies; the absolute risk is low (estimated at 2 to 3 per 1000 exposed compared with 1 to 2 per 1000 unexposed), but it is a severe outcome requiring neonatal intensive care, and the finding has contributed to ongoing clinical uncertainty. Sertraline and escitalopram are the SSRIs most commonly recommended in pregnancy on the basis of the most favorable available safety data and lowest placental transfer rates relative to other agents in the class.
Most antidepressants are excreted in breast milk, but relative infant dose (RID) — the infant's weight-adjusted dose as a percentage of the maternal weight-adjusted dose — is low for the preferred agents. Sertraline and paroxetine have the lowest RID values among SSRIs, generally below 1% to 2%, and are considered compatible with breastfeeding by most expert bodies including the American Academy of Pediatrics.13 Fluoxetine has a higher RID and longer neonatal half-life due to norfluoxetine, and is generally avoided during breastfeeding when an alternative is feasible. Citalopram and escitalopram have intermediate RID values and are considered acceptable. TCAs and nortriptyline specifically have low RID values and a long safety record in lactation. All decisions regarding antidepressant use during lactation require case-by-case assessment integrating the severity of maternal illness, the availability of alternatives, and the gestational age and health status of the infant.
Depression in patients over 65 years requires modified prescribing approaches driven by four pharmacokinetic and pharmacodynamic changes of aging: reduced hepatic cytochrome P450 activity, reduced renal clearance, reduced albumin binding capacity (increasing free drug fraction for highly protein-bound agents), and increased CNS sensitivity to adverse drug effects at equivalent plasma concentrations.14 The consequence is that equivalent doses produce higher effective exposures, and adverse effects that are manageable in younger adults become clinically important safety risks in older patients. Anticholinergic burden is particularly dangerous in the elderly, producing cognitive impairment, urinary retention, constipation, and increased fall risk; tertiary amine TCAs are strongly discouraged in patients over 65 and appear on the Beers Criteria list of potentially inappropriate medications for older adults. Alpha-1 adrenergic blockade producing orthostatic hypotension is a leading cause of falls and hip fractures in elderly patients on TCAs and trazodone; if trazodone is used for insomnia in this population, starting at 25 mg and titrating cautiously is essential.
QTc prolongation risk from citalopram is specifically amplified in older adults, underpinning the 20 mg/day maximum dose recommendation. Among SSRIs, sertraline and escitalopram are most widely recommended in the elderly based on pharmacokinetic data, tolerability profiles, and minimal anticholinergic activity; escitalopram has evidence supporting efficacy in late-life depression from randomized controlled trials in patients over 65.14 Hyponatremia through the syndrome of inappropriate antidiuretic hormone secretion (SIADH) is a disproportionate risk in elderly patients taking SSRIs and SNRIs, with incidence rates several-fold higher than in younger patients, and baseline sodium levels should be checked before initiation and rechecked within four weeks in patients over 65.
Pregnancy: weigh drug risk against risk of untreated depression; prefer sertraline or escitalopram; counsel on neonatal adaptation syndrome. Lactation: sertraline and paroxetine lowest RID; avoid fluoxetine when alternatives exist. Elderly: start at half the usual adult dose and titrate slowly; avoid tertiary TCAs; monitor sodium at baseline and 4 weeks post-initiation; check orthostatic blood pressure; use escitalopram or sertraline as first choice; apply Beers Criteria.
Boyer EW, Shannon M. The serotonin syndrome. N Engl J Med. 2005;352(11):1112–1120.
doi:10.1056/NEJMra041867Dunkley EJ, Isbister GK, Sibbritt D, Dawson AH, Whyte IM. The Hunter Serotonin Toxicity Criteria: simple and accurate diagnostic decision rules for serotonin toxicity. QJM. 2003;96(9):635–642.
doi:10.1093/qjmed/hcg109Graudins A, Stearman A, Chan B. Treatment of the serotonin syndrome with cyproheptadine. J Emerg Med. 1998;16(4):615–619.
doi:10.1016/s0736-4679(98)00057-2Montejo AL, Llorca G, Izquierdo JA, Rico-Villademoros F. Incidence of sexual dysfunction associated with antidepressant agents: a prospective multicenter study of 1022 outpatients. J Clin Psychiatry. 2001;62(suppl 3):10–21.
Fava M. Weight gain and antidepressants. J Clin Psychiatry. 2000;61(suppl 11):37–41.
US Food and Drug Administration. Celexa (citalopram hydrobromide) drug safety communication: abnormal heart rhythms associated with use. Silver Spring, MD: FDA; 2011. Available at: https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-revised-recommendations-celexa-citalopram-hydrobromide-related
Anglin R, Yuan Y, Moayyedi P, Tse F, Armstrong D, Leontiadis GI. Risk of upper gastrointestinal bleeding with selective serotonin reuptake inhibitors with or without concurrent nonsteroidal anti-inflammatory use: a systematic review and meta-analysis. Am J Gastroenterol. 2014;109(6):811–819.
doi:10.1038/ajg.2014.87Borges S, Desta Z, Li L, et al. Quantitative effect of CYP2D6 genotype and inhibitors on tamoxifen metabolism: implication for optimizing analgesic and antidepressant therapy. Clin Pharmacol Ther. 2006;80(1):61–74.
doi:10.1016/j.clpt.2006.03.013Spina E, Trifirò G, Caraci F. Clinically significant drug interactions with newer antidepressants. CNS Drugs. 2012;26(1):39–67.
doi:10.2165/11594710-000000000-00000Fava GA, Gatti A, Belaise C, Guidi J, Offidani E. Withdrawal symptoms after selective serotonin reuptake inhibitor discontinuation: a systematic review. Psychother Psychosom. 2015;84(2):72–81.
doi:10.1159/000370338Davies J, Read J. A systematic review into the incidence, severity and duration of antidepressant withdrawal effects: are guidelines evidence-based? Addict Behav. 2019;97:111–121.
doi:10.1016/j.addbeh.2018.08.027Yonkers KA, Wisner KL, Stewart DE, et al. The management of depression during pregnancy: a report from the American Psychiatric Association and the American College of Obstetricians and Gynecologists. Gen Hosp Psychiatry. 2009;31(5):403–413.
doi:10.1016/j.genhosppsych.2009.04.003Hale TW. Medications and Mothers' Milk: A Manual of Lactational Pharmacology. 19th ed. New York: Springer Publishing; 2021. Antidepressants chapter.
Roose SP, Schatzberg AF. The efficacy of antidepressants in the treatment of late-life depression. J Clin Psychopharmacol. 2005;25(suppl 1):S1–S7.
doi:10.1097/01.jcp.0000162807.84570.4e