The term second generation antidepressants is used clinically, rather than to denote a single pharmacological class, for a group of drugs introduced after tricyclic antidepressants and monoamine oxidase inhibitors and characterized by more selective or multimodal mechanisms. They include mainly selective serotonin reuptake inhibitors, serotonin and norepinephrine reuptake inhibitors and several atypical antidepressants. In adults with major depression, they are now a central component of pharmacotherapy because, at similar average class efficacy, they are generally easier to manage and less toxic in overdose than tricyclic antidepressants.
Response to an individual drug nevertheless remains heterogeneous. Comparative meta analyses show differences among agents in average efficacy, acceptability and adverse effects, but no antidepressant is universally superior. Selection must therefore integrate episode characteristics, previous individual or family responses, comorbidities, suicide risk, concomitant medications, interactions, adverse effects the patient considers most relevant, possible pregnancy, age, preferences and availability of suitable formulations. Before starting an antidepressant, it is also essential to assess for features suggestive of bipolar disorder, substance use or medical conditions that could alter the diagnosis and treatment strategy.
Contemporary guidelines do not regard pharmacotherapy as the only option for every depressive presentation. In less severe disorders, psychological strategies or other interventions may be appropriate depending on preferences and context, whereas in more severe forms an antidepressant is often one of the first line options, either alone or combined with psychotherapy. The goal is not merely to reduce symptom scores, but to achieve clinical remission, functional recovery and relapse prevention while minimizing adverse effects and the risk of premature discontinuation.
Second generation antidepressants are pharmacologically heterogeneous. SSRIs primarily inhibit the serotonin transporter, or SERT, increasing synaptic availability of 5 hydroxytryptamine. SNRIs inhibit SERT and the norepinephrine transporter, or NET, with a ratio between the two effects that varies among drugs and, for some agents, with a more evident noradrenergic component at higher doses. This pharmacodynamic distinction does not, however, reliably predict which patient will respond to one class rather than another.
Among atypical antidepressants, mirtazapine antagonizes presynaptic alpha2 adrenergic receptors and several serotonin receptors, with marked H1 antagonism that contributes to sedation and increased appetite. Bupropion inhibits norepinephrine and dopamine reuptake without a clinically dominant serotonergic action. Trazodone combines 5 HT2 antagonism with serotonin reuptake inhibition, whereas vortioxetine combines SERT inhibition with modulation of several serotonin receptor subtypes. Agomelatine, by contrast, acts as an MT1/MT2 melatonergic agonist and 5 HT2C antagonist.
Classification does not coincide with regulatory indication. A drug may be approved for major depression and, separately, for anxiety disorders or neuropathic pain, while the same indication may not be authorized in every jurisdiction. In the Italian context, scientific evidence must therefore be distinguished from AIFA/EMA authorization and reimbursement status. For example, desvenlafaxine is authorized in Italy for major depressive disorder in adults, whereas the European profile of duloxetine includes major depression, generalized anxiety disorder and diabetic peripheral neuropathic pain.
Pharmacokinetic differences are clinically important. Fluoxetine and its metabolite norfluoxetine have long half lives and therefore concentrations decline more gradually after discontinuation, but they also require a longer washout before incompatible drugs. Paroxetine and venlafaxine, by contrast, are among the agents in which rapid dose reduction can more readily produce discontinuation symptoms. Fluvoxamine and paroxetine have clinically relevant potential for enzyme mediated interactions; bupropion inhibits CYP2D6; duloxetine is metabolized mainly by CYP1A2 and CYP2D6; agomelatine requires particular attention to interactions that increase its exposure.
The concepts of “serotonergic”, “noradrenergic” or “dopaminergic” action should not be converted into a simplistic correspondence between symptoms and neurotransmitters. There is no clinical test demonstrating a serotonin, norepinephrine or dopamine deficiency that can guide individual prescribing. Symptoms such as anergia, psychomotor slowing, insomnia, hypersomnia, anxiety or loss of appetite may help guide selection mainly through the profile of desirable or undesirable effects, not because they identify a definite neurochemical mechanism.
The time to response must also be interpreted correctly. Biological changes begin after the first doses, but clinically meaningful benefit often requires several weeks. Early improvement during the first two weeks may be prognostically favorable, whereas failure to achieve remission after an adequate period at a therapeutic dose requires reassessment of diagnosis, adherence, dose, duration, comorbidities and concomitant treatment before concluding that there is a true pharmacological failure.
Treatment duration depends on recurrence risk. After remission from a first episode, the antidepressant is generally continued for months to consolidate benefit; a history of recurrent episodes, severe depression, chronicity, comorbidities or relapse after discontinuation may justify longer maintenance. The decision should be individualized and periodically reassessed, avoiding both premature discontinuation and automatic continuation without review.
SSRIs include fluoxetine, sertraline, escitalopram, citalopram, paroxetine and fluvoxamine. They are among the most widely used drugs for major depression because they offer a favorable balance among efficacy, tolerability and safety in overdose compared with tricyclic antidepressants. Average efficacy differences among individual SSRIs are generally modest; in clinical practice, half life, interactions, adverse effect profile, concomitant indications and the patient's previous response are more important.
The most common early adverse effects include nausea, diarrhea or dyspepsia, headache, insomnia or somnolence, transient worsening of anxiety, tremor and sexual dysfunction. Serotonergic inhibition can lead to reduced libido, delayed orgasm or anorgasmia and ejaculatory disorders. The true frequency of sexual dysfunction is probably underestimated in studies relying on spontaneous reporting, so it is useful to assess it actively before and during treatment.
Fluoxetine has a particularly long half life and an active metabolite; this reduces the risk of abrupt plasma fluctuations after an occasional missed dose and generally makes an immediate discontinuation syndrome less likely, but also causes persistent interactions and requires a longer interval before starting an MAOI. It can be relatively activating and, particularly early in treatment, may promote insomnia, restlessness or gastrointestinal symptoms in some patients.
Sertraline is widely used in depressive and anxiety disorders. Its cardiovascular profile is generally favorable and the drug has been extensively studied in patients with heart disease, but this does not mean that it is free of interactions or that it should automatically be preferred in every cardiovascular condition. Gastrointestinal effects, particularly diarrhea, can be more prominent in some patients.
Escitalopram is the S enantiomer of citalopram and has high selectivity for SERT. It is often well tolerated and has relatively limited interaction potential. Citalopram and escitalopram nevertheless require attention to QT prolongation, especially at high doses, in older age, heart disease, bradycardia, electrolyte abnormalities or concomitant use of other QT prolonging drugs; dose restrictions in the respective product information should be observed.
Paroxetine has a shorter half life, greater anticholinergic activity than other SSRIs and potent CYP2D6 inhibition. It is more frequently associated with discontinuation symptoms when stopped rapidly and may promote weight gain and sexual dysfunction. In older adults, the anticholinergic burden is an unfavorable feature, and in people taking drugs metabolized by CYP2D6 the interaction may be clinically important.
Fluvoxamine is widely used for obsessive compulsive disorder and has a clinically relevant enzyme inhibition profile, particularly involving CYP1A2 and CYP2C19. Interactions with other drugs must therefore be assessed carefully. There is insufficient evidence to describe it as inherently more anxiolytic than other SSRIs in major depression.
An uncommon but important complication is hyponatremia, often mediated by SIADH and more likely in older adults, people with low body weight, previous hyponatremia or those receiving diuretics. Bruising and gastrointestinal bleeding may increase because of interference with platelet function, particularly when SSRIs are combined with NSAIDs, antiplatelet agents or anticoagulants. Serotonin syndrome is rare with appropriately dosed monotherapy but becomes a genuine risk when other strongly serotonergic drugs or MAOIs are added.
SNRIs inhibit serotonin and norepinephrine reuptake. Venlafaxine has a marked serotonergic effect at lower doses, while noradrenergic inhibition becomes more relevant as exposure increases; desvenlafaxine is the main active metabolite of venlafaxine and has simpler pharmacokinetics; duloxetine produces relatively balanced inhibition of SERT and NET within the therapeutic range. These pharmacological differences still do not allow reliable matching of a specific agent to an individual symptom subtype.
Venlafaxine is effective in major depressive disorder and several anxiety disorders. Dose increases may be associated with rises in blood pressure and heart rate in a proportion of patients, so blood pressure should be monitored, particularly at higher doses or in patients with preexisting hypertension. Nausea, sweating, insomnia, sexual dysfunction and discontinuation symptoms are clinically relevant adverse effects.
Abrupt discontinuation of venlafaxine can produce dizziness, disequilibrium, paresthesias or electric shock like sensations, nausea, insomnia, irritability and flu like symptoms. Frequency and severity vary and should not automatically be mistaken for depressive relapse. Gradual dose reduction, adapted to dose, treatment duration and the patient's previous experience, is generally preferable.
Desvenlafaxine is authorized in Italy for major depressive disorder in adults; the recommended dose in the product information is 50 mg once daily. Higher doses have not consistently shown a proportional efficacy advantage and may increase adverse effects. Because renal elimination makes an important contribution to its pharmacokinetics, renal impairment requires dose adjustments according to the product information.
Duloxetine is authorized in Europe for major depressive episodes, generalized anxiety disorder and diabetic peripheral neuropathic pain. The presence of chronic pain may make it clinically useful to consider a drug with documented efficacy for specific pain syndromes, but does not justify automatically extending this rationale to every form of pain. The European indication does not include all conditions for which duloxetine is approved in other countries.
With duloxetine, hepatic function, alcohol use and drug interactions must be considered. Nausea, dry mouth, constipation, sweating, insomnia or somnolence, sexual dysfunction and increased blood pressure can occur. Discontinuation should be gradual. In patients with severe renal or hepatic impairment, use may be contraindicated or subject to specific restrictions according to the product information.
Milnacipran is pharmacologically an SNRI and is used in some jurisdictions for different indications; availability and approved indications are not the same across countries. On a page intended for Italian clinical practice it should therefore not be presented as an automatic prescribing equivalent of venlafaxine, desvenlafaxine or duloxetine without verifying the actual product available and its authorization.
SNRIs share with SSRIs the risk of serotonin syndrome when combined with other strongly serotonergic drugs and should not be combined with MAOIs without the required washout periods. They can also cause hyponatremia and increase bleeding risk when combined with drugs that interfere with hemostasis. The choice over an SSRI should therefore be based on the overall clinical picture, not on an assumption that SNRIs are generally superior.
Mirtazapine antagonizes presynaptic alpha2 adrenergic receptors and modulates serotonergic transmission through 5 HT2 and 5 HT3 antagonism, in addition to potent H1 antagonism. It is effective in major depression, and its profile can be useful when insomnia, reduced appetite or weight loss are important clinical problems. This orientation is based mainly on pharmacological effects and tolerability, not on the existence of a biological subtype of depression that is “mirtazapine responsive”.
Sedation and increased appetite are especially common during the first weeks and may promote weight gain. It is widely believed that higher doses are always less sedating because the noradrenergic component increases, but this relationship has not been sufficiently demonstrated to be used as a clinical rule. Titration should follow response, tolerability and product information, without increasing the dose for the purpose of reducing sedation.
Mirtazapine causes less sexual dysfunction on average than strongly serotonergic antidepressants, but it is not entirely free of this effect. It may cause dry mouth, constipation, edema and lipid abnormalities; neutropenia or agranulocytosis are very rare but warrant urgent assessment in the presence of fever, sore throat or unexplained signs of infection, in accordance with product warnings.
Bupropion inhibits norepinephrine and dopamine reuptake and has little direct activity on the serotonergic system. In Italy, the modified release formulation for depression is authorized for the treatment of major depressive episodes. A different bupropion product is used for smoking cessation: indications, formulations and dosing regimens should not be interchanged without reference to the specific product information.
Bupropion has a lower propensity for sexual dysfunction and weight gain than many serotonergic antidepressants, but it can promote insomnia, agitation, anxiety, headache, dry mouth and increased blood pressure. The risk of seizures is dose dependent; according to the product information, the drug is contraindicated in conditions that significantly lower the seizure threshold, including epilepsy and certain eating disorders. It is also a CYP2D6 inhibitor and can increase exposure to several other drugs.
Trazodone combines 5 HT2A/5 HT2C antagonism, SERT inhibition and alpha1 and H1 antagonism in a dose dependent manner. Low doses are predominantly sedating and are not equivalent to doses used for antidepressant treatment. Its widespread use for insomnia should therefore not lead to the assumption that a small evening dose constitutes adequate treatment for major depression.
Adverse effects of trazodone include sedation, dizziness and orthostatic hypotension; the risk of falls is especially relevant in older adults or when it is combined with other sedatives and antihypertensive drugs. Priapism is rare but is a medical emergency. Its interaction profile also depends on CYP3A4 metabolism and on concomitant use of other drugs that prolong the QT interval or increase sedation.
The choice between these drugs and SSRIs or SNRIs should not be framed as a rigid hierarchy. Mirtazapine may be preferable when insomnia and weight loss are problematic; bupropion may be considered when sexual dysfunction or somnolence makes a serotonergic drug less desirable; trazodone may be appropriate in selected patients. In every case, the entire risk profile must be assessed, rather than only one potentially favorable effect.
Vortioxetine is a multimodal antidepressant: it inhibits SERT and interacts with several serotonin receptors, including 5 HT1A, 5 HT1B, 5 HT3 and 5 HT7. It is authorized in the European Union for the treatment of major depressive episodes in adults. The recommended dose for adults younger than 65 years is generally 10 mg once daily, with adjustment between 5 and 20 mg according to response and tolerability; in older adults, the product information recommends greater caution and a lower starting dose.
Nausea is the most frequent adverse effect and occurs mainly during the first weeks. Headache, dizziness and sexual dysfunction may occur, although its sexual adverse effect profile may be more favorable than that of some SSRIs in certain comparisons. It is incorrect to present vortioxetine as free of sexual dysfunction or as specifically indicated for “cognitive depression”.
In clinical studies, vortioxetine has shown effects on cognitive measures in patients with depression, but part of the improvement may depend on the overall reduction in depressive symptoms. Cognitive difficulties may contribute to treatment selection, but there is no diagnostic criterion or biomarker that reliably identifies patients who will obtain a specific cognitive benefit.
Vortioxetine is metabolized mainly through CYP2D6, with contributions from other pathways. Strong CYP2D6 inhibitors can increase its exposure and strong enzyme inducers can reduce it. Like other serotonergic antidepressants, it should not be combined with MAOIs and requires attention to serotonergic drugs, anticoagulants, antiplatelet agents and conditions predisposing to hyponatremia.
Agomelatine is an MT1/MT2 agonist and 5 HT2C antagonist authorized in the European Union for major depressive episodes in adults. It is taken in the evening; the authorized starting dose is 25 mg and can be increased to 50 mg in patients without adequate improvement after assessment of the benefit risk balance and in accordance with hepatic monitoring requirements.
The main safety concern with agomelatine is potential hepatotoxicity. Liver function tests should be performed before treatment and approximately after 3, 6, 12 and 24 weeks, after a dose increase and thereafter when clinically indicated. The drug should not be started in the presence of contraindications specified in the product information and should be discontinued according to the hepatic thresholds and criteria provided there.
Agomelatine has a distinctive profile compared with antidepressants that directly inhibit SERT and may have a relatively limited impact on sexual function, but it should not be described as risk free or generally superior. Its average efficacy is that of an antidepressant, and the requirement for liver monitoring is an important factor in selection and adherence.
Vilazodone is sometimes included internationally among multimodal antidepressants because it combines SERT inhibition with partial 5 HT1A agonism, but it is not a routine regulatory option in Europe. To avoid confusion between United States literature and Italian practice, it is preferable to discuss separately the drugs that are actually authorized and available in the European context.
Initial selection begins with the diagnosis. Before prescribing an antidepressant, clinicians should investigate episodes of pathological mood elevation, decreased need for sleep, increased activity or other features suggestive of bipolarity, because bipolar depression requires a different strategy from unipolar depression. Alcohol and substance use, concomitant medications, thyroid disorders, anemia, sleep disorders and other conditions that can cause or worsen depressive symptoms should also be assessed.
Titration should be specific to the drug. Some agents can be started directly at a standard therapeutic dose, while others require gradual increases to reduce nausea, activation, hypotension or sedation. In older or frail patients, those with comorbidities or those who are highly sensitive to adverse effects, it is often appropriate to start at lower doses and titrate more slowly. The aim remains to reach an effective dose, not to maintain a subtherapeutic dose indefinitely because of concern about initial adverse effects.
Response should be assessed in a structured way. After treatment begins, tolerability, adherence, clinical worsening, agitation, insomnia, emergence of manic or hypomanic symptoms and suicide risk should be reviewed early. Initial follow up should be particularly careful in younger patients and people at high risk. The emergence of suicidal ideation should neither be automatically attributed to the drug nor dismissed as part of depression: it requires immediate clinical assessment.
If benefit is partial or absent after an adequate period at a therapeutic dose, the first question is whether treatment has actually been adequate. Adherence, absorption, interactions, substance use, diagnosis, comorbidities and persistent psychosocial factors should be reviewed. Subsequently, dose escalation when supported, switching to another antidepressant, adding psychotherapy or augmentation strategies may be considered. There is no single sequence that is valid for every patient.
Laboratory monitoring is not the same for all agents. SSRIs do not routinely require serial blood panels in every healthy adult, but sodium measurement may be appropriate in older adults or people predisposed to hyponatremia. Blood pressure and heart rate are more important with some SNRIs and bupropion. ECG and electrolytes are indicated when there is a risk of QT prolongation. With agomelatine, liver monitoring is an integral part of treatment safety.
During pregnancy and breastfeeding, the decision should not be reduced to the risk category of an individual drug. The risks of exposure must be weighed against those of untreated depression, taking into account severity, previous relapses, response to earlier treatments and drug specific characteristics. Abrupt discontinuation of an effective antidepressant solely because pregnancy is discovered can be harmful and requires specialist assessment.
In older adults, polypharmacy, renal and hepatic function, risk of falls, hyponatremia, bleeding and QT prolongation must be considered. Dose modifications in renal or hepatic disease are drug specific. In patients with epilepsy or factors that lower the seizure threshold, bupropion requires particular caution or is contraindicated; in patients with obesity or metabolic syndrome, mirtazapine may be less desirable if weight gain represents an important risk.
Pharmacogenetics can be useful in selected contexts. CPIC guidelines provide recommendations for some drug genotype combinations, particularly CYP2D6 and CYP2C19, but a pharmacogenetic test does not fully predict efficacy and tolerability and does not replace clinical assessment. Proprietary tests that combine many genes into commercial scores should not be interpreted as infallible diagnostic tools.
Adverse effects should be discussed with the patient in concrete terms, distinguishing transient effects from those that require reassessment. Nausea, headache, gastrointestinal disturbances, sleep changes and mild activation may diminish during the first weeks; sexual dysfunction, weight gain or sweating may persist. Management may include changing the timing of administration, adjusting the dose, treating the symptom or switching to a drug with a different profile, while avoiding unsupported empirical strategies.
Serotonin syndrome results from excessive serotonergic activity and is characterized by a variable combination of altered mental status, autonomic hyperactivity and neuromuscular signs, particularly clonus and hyperreflexia. The risk is greater when strongly serotonergic drugs are combined, especially MAOIs with SSRIs or SNRIs. Clinical suspicion requires withdrawal of the causative drugs and urgent assessment.
Interactions with anticoagulants, antiplatelet agents and NSAIDs can increase bleeding risk with serotonergic antidepressants. CYP inhibitors may increase concentrations of other psychotropic drugs, antiarrhythmics, beta blockers or analgesics, whereas inducers may reduce exposure. Checking only the package leaflet is not sufficient when a patient takes many medications: pharmacokinetics, pharmacodynamics and concomitant clinical conditions need to be assessed systematically.
QT deserves attention particularly with citalopram and escitalopram and when several risk factors are present. The absolute risk of torsades de pointes remains low in the general population, but increases with heart disease, bradycardia, hypokalemia or hypomagnesemia, high doses and use of multiple drugs that prolong repolarization. An ECG is therefore reasonable when the individual risk profile warrants it, but not necessarily in every healthy young person.
Discontinuation symptoms can occur with all classes, but their likelihood varies among drugs. The most recent meta analysis estimates that a clinically relevant proportion of patients develop symptoms after stopping treatment; paroxetine and venlafaxine are among the drugs commonly associated with greater risk, whereas the long half life of fluoxetine tends to make symptoms less immediate. Symptoms can include dizziness, nausea, insomnia, anxiety, irritability, disequilibrium and paresthesias.
Dose reduction should be individualized. After short courses at low doses, a relatively simple reduction may be sufficient, whereas after prolonged use or previous discontinuation symptoms, slower tapering with proportionally smaller reductions at low doses may be necessary. Worsening immediately associated with a dose reduction should not automatically be labeled as relapse: timing, symptom characteristics and response to reintroduction help with the differential diagnosis.
When switching from one antidepressant to another, the strategy may involve a direct switch, cross taper or washout. Temporary coadministration may be inappropriate when there is a risk of serotonergic or metabolic interaction. Switching to or from an MAOI requires specific intervals; after fluoxetine the washout is longer because of persistence of the active metabolite. Timing should be checked against the product information for the specific drugs being used.
Effective antidepressant management does not end with the prescription. Education about response times, adverse effects, adherence, alcohol and substance use, discontinuation risk, recognition of manic symptoms and the follow up plan reduces errors and inappropriate treatment interruption. The final decision should be shared, documented and updated according to the clinical course.
| Class or drug | Predominant mechanism | Factors useful in selection | Relevant risks and monitoring |
| SSRIs | SERT inhibition | Extensive experience in depression and anxiety disorders; generally favorable profile compared with TCAs | Sexual dysfunction, gastrointestinal disturbances, hyponatremia, bleeding; QT especially with citalopram/escitalopram |
| Venlafaxine | Dose dependent SERT and NET inhibition | Major depression and anxiety disorders | Blood pressure, sexual dysfunction, discontinuation symptoms |
| Desvenlafaxine | SERT and NET inhibition | Authorized in Italy for adult MDD | Renal function, blood pressure, discontinuation symptoms |
| Duloxetine | SERT and NET inhibition | Depression; possible usefulness when diabetic neuropathic pain coexists | Blood pressure, hepatic function, CYP1A2/CYP2D6 interactions |
| Mirtazapine | Alpha2, 5 HT2/5 HT3 and H1 antagonism | Insomnia, reduced appetite or weight loss may guide selection | Sedation, weight gain, metabolic changes |
| Bupropion | Norepinephrine/dopamine reuptake inhibition | Lower propensity for sexual dysfunction and weight gain | Insomnia, blood pressure, CYP2D6 interactions, seizure risk |
| Trazodone | 5 HT2 antagonism and SERT inhibition | Sedating profile; antidepressant dose differs from low hypnotic doses | Orthostatic hypotension, sedation, rare priapism |
| Vortioxetine | SERT inhibition and serotonergic receptor modulation | Adult MDD; multimodal profile | Nausea, CYP2D6 interactions, possible sexual dysfunction |
| Agomelatine | MT1/MT2 agonism and 5 HT2C antagonism | Alternative with a nonclassical serotonergic profile | Mandatory hepatic monitoring according to product information |
Informational notice: the information contained on this page is provided solely for informational and educational purposes and does not replace the advice, diagnosis or treatment provided by a physician. If needed, always consult a qualified healthcare professional.
Artificial intelligence transparency: this page was created with the support of artificial intelligence tools, used to assist in the production and processing of its content.