Antipsychotics: A Comprehensive Overview

A high-level examination of the typical and atypical antipsychotics — their receptors, likely mechanisms, real-world value, and problems

What These Drugs Are, and What They Actually Treat

Antipsychotics (historically "neuroleptics," "major tranquilizers") are the drugs that made deinstitutionalization possible: their arrival in the 1950s transformed schizophrenia from a condition warehoused in asylums into one substantially treatable in the community. That is their genuine and enormous achievement, and it should anchor any balanced account. They are also among the most burdensome drugs in psychiatry — carrying movement disorders, metabolic toxicity, and a long-term outcome literature contested enough that serious clinicians and researchers disagree about how, and how long, they should be used.

A precise framing of what they treat matters, because it is widely misunderstood. Antipsychotics are genuinely effective against the positive symptoms of psychosis — hallucinations, delusions, thought disorder — and against the agitation and risk of acute psychotic states, and they reliably reduce relapse in schizophrenia. They are far less effective, often nearly inert, against the negative symptoms (avolition, flattening, social withdrawal) and cognitive symptoms that drive most of the long-term disability in schizophrenia — the great unmet need the entire class has failed to meet. So "antipsychotic" overstates the case: these drugs dampen the florid, productive symptoms of psychosis well and leave its most disabling dimensions largely untouched.

Their use has also expanded far beyond psychosis: across bipolar disorder (covered in the mood-stabilizers document), as antidepressant adjuncts, for agitation and aggression, in delirium, and — most contestably — for insomnia, anxiety, and behavioral control in dementia and nursing homes, much of it off-label and on thin evidence. This expansion is central to the problems section.

This overview covers the first-generation (typical) antipsychotics, the second-generation (atypical) antipsychotics, the uniquely important agent clozapine, and the newest mechanistic departures. The honest framing: a class that genuinely controls psychosis and enables lives that would otherwise be lost to it — and that is seriously toxic, ineffective against the symptoms that most disable, widely overextended into off-label uses, and entangled in real and unresolved questions about its long-term effects.

The Dopamine Hypothesis and Its Evolution

The theory underpinning these drugs has itself evolved instructively.

The original dopamine hypothesis. Two observations converged: every effective early antipsychotic blocked dopamine D2 receptors, and the clinical potency of these drugs correlated tightly with their D2-blocking affinity; meanwhile, dopamine-increasing drugs (amphetamine, L-DOPA) could induce psychosis. This yielded the dopamine hypothesis: psychosis reflects dopamine overactivity, and antipsychotics work by blocking it. The model was refined into a regional account: mesolimbic hyperdopaminergia drives positive symptoms (and D2 blockade there is therapeutic), while mesocortical hypodopaminergia (too little prefrontal dopamine) may underlie negative and cognitive symptoms — explaining why D2 blockade helps the positive symptoms but cannot fix, and may even worsen, the negative ones.

The salience reframing. The more sophisticated modern account, developed by Shitij Kapur, reframes the dopamine story psychologically. Dopamine mediates salience — the process by which the brain tags stimuli and thoughts as important, attention-worthy, motivationally relevant. The hypothesis: psychosis is a state of aberrant salience, in which dysregulated dopamine assigns significance to stimuli and internal events that do not warrant it — the neutral becomes portentous, coincidence becomes conspiracy, an inner thought becomes an external voice — and delusions are the mind's reasoned attempt to explain this storm of misattributed significance. On this view antipsychotics work not by directly deleting hallucinations or delusions but by dampening the aberrant salience — turning down the abnormal significance so that the symptoms lose their grip. Patients often describe exactly this: not that the voices vanished overnight, but that they stopped mattering, receded, became ignorable. Kapur's phrase — the drugs provide "a platform for psychological resolution" — captures that the chemistry creates the conditions for recovery rather than constituting it, and explains the time-course (below): salience re-attribution and the psychological reworking it permits take weeks, even though D2 receptors are blocked in hours.

The limits of the dopamine story. The hypothesis remains incomplete: it explains positive symptoms and antipsychotic action reasonably well but negative/cognitive symptoms poorly, and the broader neurobiology of schizophrenia clearly involves glutamate (NMDA-receptor hypofunction — the basis of the ketamine/PCP psychosis model and of glutamatergic drug efforts that have largely failed), GABA, and developmental and genetic factors that dopamine alone cannot carry. Dopamine is the lever the drugs pull, not the whole disease.

The Two Generations and the Mechanism of "Atypicality"

First-generation (typical) antipsychotics

Chlorpromazine (Delay and Deniker, 1952) opened the era; haloperidol (Janssen) became the high-potency standard. Mechanistically these are predominantly D2 antagonists, distinguished mainly by potency (high-potency agents like haloperidol cause more movement side effects and less sedation/hypotension; low-potency agents like chlorpromazine the reverse, owing to additional histaminergic, cholinergic, and adrenergic blockade). Their defining liability is extrapyramidal symptoms (EPS) — the motor consequences of striatal D2 blockade — culminating in tardive dyskinesia, the potentially irreversible movement disorder that became the typicals' signature harm and the central motivation for developing something better.

The D2 occupancy window

PET studies (Kapur and others) established a clarifying quantitative framework: antipsychotic efficacy requires roughly 65–70% striatal D2 occupancy, while EPS emerge above roughly 78–80%. There is, in other words, a therapeutic window of D2 occupancy — enough to treat psychosis, not so much as to cause movement disorders — and much of the typicals' problem was routine dosing that blew past 80% occupancy. This single insight reframed antipsychotic dosing (lower is often better) and helped explain atypicality.

Second-generation (atypical) antipsychotics

The atypicals — risperidone, olanzapine, quetiapine, ziprasidone, aripiprazole, paliperidone, asenapine, lurasidone, cariprazine, brexpiprazole, and others — were defined clinically by lower EPS/tardive-dyskinesia risk at effective doses. Two mechanistic theories explain how:

• The 5-HT2A/D2 ratio (Meltzer). Atypicals combine D2 antagonism with strong 5-HT2A antagonism; because serotonin (via 5-HT2A) modulates dopamine release in motor pathways, 5-HT2A blockade preferentially preserves dopamine in the nigrostriatal (motor) and mesocortical pathways while still blocking it in the mesolimbic (psychosis) pathway — yielding antipsychotic effect with less EPS and, in theory, less worsening of negative symptoms.
• Fast-off / loose binding (Kapur & Seeman). An alternative (or complementary) account: atypicals bind D2 more loosely and dissociate faster, blocking dopamine enough to treat psychosis but releasing the receptor readily enough that normal phasic dopamine signaling (and motor function) is less disrupted. On this view, "atypicality" is partly a kinetic property, not just a receptor-ratio one.

The dopamine partial agonists

A genuinely distinct mechanism: aripiprazole, cariprazine, and brexpiprazole are D2 partial agonists — they occupy the receptor but produce submaximal activation, acting as functional antagonists where dopamine is high (dampening mesolimbic overactivity) and functional agonists where dopamine is low (supporting mesocortical signaling). This "dopamine system stabilizer" concept aims to treat positive symptoms without the EPS, prolactin elevation, and (in part) the negative-symptom worsening of pure antagonism — and these agents are indeed cleaner on EPS and prolactin, though akathisia is notably common with aripiprazole.

Clozapine: the exception that matters most

Clozapine stands alone and deserves emphasis. It is the only antipsychotic proven superior to the others for treatment-resistant schizophrenia (the ~30% of patients who fail two adequate trials of other agents), the only one with a demonstrated antisuicide effect in schizophrenia (the InterSePT trial), and it produces essentially no EPS or tardive dyskinesia — a remarkable profile. Its mechanism is famously promiscuous (relatively weak, loose D2 binding plus action across a wide range of serotonergic, muscarinic, histaminergic, and adrenergic receptors), and its uniqueness has never been fully explained or replicated. And yet it is tragically underused, for the same structural reason as lithium: its burdens are severe and demanding — agranulocytosis (a potentially fatal loss of white cells, requiring mandatory regular blood-count monitoring under a REMS program), myocarditis, seizures, severe metabolic effects, constipation (which can be fatal — ileus), and sedation. The result is that the drug uniquely able to help the most refractory patients is withheld from many who would benefit, because the monitoring and risk are onerous — a genuine quality-of-care failure that parallels lithium's neglect.

The Time-Course Paradox and What "Working" Means

As with the antidepressants, there is a delay paradox: D2 receptors are blocked within hours, but the full antipsychotic effect builds over days to weeks. Proposed explanations include depolarization block (chronic D2 blockade eventually silences dopamine neurons through sustained depolarization, developing over weeks) and — more satisfyingly, on the salience model — the time required for aberrant salience to subside and for the patient to psychologically rework the delusional framework built during psychosis. The salience account fits the clinical phenomenology: improvement often begins with the symptoms losing emotional grip and salience ("the voices are still there but I don't care about them") before, and sometimes without, their disappearance — which is why "antipsychotic response" is better understood as the loosening of psychosis's hold than its erasure, and why the drugs create conditions for recovery rather than constituting recovery.

What They Actually Do: The Evidence and Its Hard Questions

Acute psychosis and positive symptoms: robustly effective, the evidence base solid (Leucht's network meta-analyses establishing efficacy over placebo and the relatively modest differences among agents, with clozapine, olanzapine, and a few others ranking somewhat higher on efficacy and the agents differing more in side-effect profile than in efficacy).

Relapse prevention: maintenance antipsychotic treatment substantially reduces relapse versus discontinuation — one of the clearer benefits, and the basis for long-term treatment and for long-acting injectable formulations (which address the enormous adherence problem in schizophrenia and improve outcomes in many patients).

Negative and cognitive symptoms: the persistent failure — antipsychotics do little for the avolition, flattening, and cognitive impairment that drive most long-term disability, and may worsen some of it. This is the field's central unmet need.

The long-term outcome controversy: here genuine, unresolved disagreement exists, and an honest account must present it. The standard view — maintenance antipsychotics improve long-term outcomes by preventing relapses, each of which may be neurotoxic and worsen the course — is challenged by several lines of evidence. Wunderink's randomized trial found that patients assigned to early dose reduction/discontinuation had better long-term functional recovery at seven years than those maintained on standard doses, despite more early relapses — suggesting that for some patients, lower long-term exposure yields better functioning. Long-term naturalistic data (and the critical-psychiatry analyses of Whitaker, Harrow's long-term follow-up showing some unmedicated patients faring well, and Moncrieff's work) raise the possibility that continuous antipsychotic treatment, while preventing relapse, may impair long-term functional recovery in a subset and may induce dopamine supersensitivity (upregulated, hypersensitive D2 receptors from chronic blockade) that paradoxically increases relapse vulnerability and makes discontinuation harder. The mainstream rejoinder: relapse is genuinely harmful and often catastrophic, the dose-reduction studies are small and their generalizability uncertain, and discontinuation carries serious risk. The honest synthesis: maintenance treatment clearly prevents relapse, the optimal long-term dose and duration are genuinely uncertain and probably patient-specific, the field has moved toward using the lowest effective dose and considering carefully-monitored dose reduction in stable patients, and the question of whether everyone with a psychotic episode needs indefinite antipsychotic treatment is more open than older teaching allowed.

Problems With Their Use

Extrapyramidal symptoms and tardive dyskinesia. Acute dystonia (frightening, treatable), drug-induced parkinsonism, akathisia (a tormenting inner restlessness, under-recognized, linked to distress and suicidality, prominent with aripiprazole and high-potency typicals), and — the gravest — tardive dyskinesia, the involuntary, often irreversible movement disorder from chronic D2 blockade. TD risk is lower with atypicals and lowest with clozapine and the partial agonists, but real with all D2 antagonists over time. The VMAT2 inhibitors (valbenazine, deutetrabenazine) are now available treatments for TD — a genuine advance, though they do not undo the underlying problem.

Other serious effects. Sedation; hyperprolactinemia (risperidone, paliperidone, typicals — causing sexual dysfunction, galactorrhea, menstrual disruption, bone-density loss); QT prolongation (ziprasidone, others); neuroleptic malignant syndrome (rare, life-threatening — hyperthermia, rigidity, autonomic instability); anticholinergic effects (low-potency typicals, clozapine, olanzapine); and clozapine's specific dangers (agranulocytosis, myocarditis, seizures, ileus).

The dementia black box. Antipsychotics increase mortality in elderly patients with dementia-related psychosis and agitation — an FDA boxed warning — yet are widely used for exactly this (behavioral control in dementia and nursing homes), often without adequate justification. This off-label use in the elderly is one of psychiatry's clearest overuse scandals: vulnerable patients given drugs that increase their risk of death for the convenience of behavioral management, frequently without the non-pharmacological approaches that should come first.

Off-label overuse generally. Beyond dementia: quetiapine for insomnia (low-dose, extremely common, exposing patients to metabolic and other risks for a sleep aid), antipsychotics for anxiety, for non-bipolar mood instability, and for behavioral control across settings — much of it poorly evidenced and committing patients to seriously toxic drugs for indications that do not warrant them.

The brain-volume controversy. Longitudinal imaging (Ho, Andreasen and others) found associations between cumulative antipsychotic exposure and loss of brain volume — a genuinely alarming finding that remains contested and confounded (schizophrenia itself is associated with volume changes, and disentangling drug from illness is difficult). It is not established that antipsychotics cause clinically meaningful brain atrophy, but the signal is real enough to take seriously and feeds the long-term-outcome debate.

The critical-psychiatry critique. A coherent dissenting position (Moncrieff, Whitaker, others) argues for a "drug-centered" rather than "disease-centered" model: antipsychotics do not correct a specific disease process but produce a state of dopaminergic dampening/neurological suppression that reduces the intensity of psychosis (and of much else), useful in acute crisis but potentially harmful as indefinite treatment given supersensitivity, functional, and volume concerns. The mainstream regards this as overstated and dangerous if it deters treatment of a serious illness; the critics regard the mainstream as minimizing real long-term harms. A balanced account holds that both the genuine, often life-saving acute and relapse-prevention benefits and the serious long-term burdens and unresolved outcome questions are real, and that the truth is neither "take your medicine indefinitely" nor "the drugs are the problem."

A Theoretical Synthesis

The antipsychotics are the clearest case in this series of a drug class that works through a mechanism we partly understand for symptoms we can partly treat, while missing the heart of the disorder. The dopamine/salience account explains the positive symptoms and the antipsychotic effect with real elegance — psychosis as aberrant salience, the drugs as dampening that aberrant significance so the symptoms lose their grip and psychological recovery becomes possible — and the salience framing dignifies what patients actually report and explains the time-course better than any "deleting hallucinations" model. But dopamine blockade does nothing for the negative and cognitive symptoms that disable, because those likely arise from glutamatergic, developmental, and mesocortical-hypodopaminergic processes the drugs do not address (and partly worsen). The class has thus been stuck for seventy years: every agent since chlorpromazine works through the same essential D2 lever, the atypicals improved tolerability (EPS) more than efficacy, clozapine's unique superiority remains unexplained and unreplicated, and the glutamatergic and other novel mechanisms that might reach the unmet symptoms have repeatedly failed in trials. The newest agents (the muscarinic-receptor approach reaching the market, working through cholinergic rather than direct dopaminergic mechanisms) represent the first genuinely new antipsychotic mechanism in decades and a test of whether the D2 monopoly can finally be broken.

The deepest unresolved tension is temporal: these drugs are genuinely, sometimes dramatically beneficial in acute psychosis and for relapse prevention, and their long-term effects — on functional recovery, on the brain, on the trajectory of the illness — are genuinely uncertain, with credible evidence that the optimal long-term strategy is lower-dose, more individualized, and more open to careful reduction than the historical "indefinite full-dose maintenance" default. Holding the acute benefit and the long-term uncertainty together, without collapsing into either uncritical maintenance or anti-medication dismissal, is the honest clinical stance.

The Clinical Bottom Line

For acute psychosis and schizophrenia, antipsychotics are genuinely effective, often essential, and enabling of lives otherwise lost to the illness — first-line for positive symptoms and acute stabilization, with agent choice driven mainly by side-effect profile (favoring metabolically cleaner agents and, increasingly, partial agonists and lowest effective doses) since efficacy differences among non-clozapine agents are modest.

Clozapine is uniquely effective for treatment-resistant schizophrenia and uniquely antisuicidal, and is tragically underused; it deserves consideration far earlier than it typically receives, with its serious monitoring burden accepted as the price of reaching patients no other drug helps.

Maintenance treatment prevents relapse and is appropriate for established schizophrenia, but the optimal long-term dose and duration are genuinely uncertain; modern practice favors the lowest effective dose, long-acting injectables where adherence is a problem, and individualized consideration of careful dose reduction in stable patients — holding the relapse-prevention benefit against the long-term burden and outcome questions.

The off-label expansion — antipsychotics for insomnia, anxiety, non-bipolar mood instability, and especially behavioral control in dementia and the elderly (where they increase mortality) — is largely an overuse problem warranting real restraint and non-pharmacological alternatives first.

Across all of them: serious movement and metabolic toxicities that contribute to premature death demand monitoring, metabolically conscious agent selection, the lowest effective dose, and honest informed consent. These drugs control psychosis and prevent relapse genuinely and often life-savingly; they are seriously toxic, ineffective against the most disabling symptoms, widely overextended, and shadowed by real long-term-outcome uncertainty — and good practice treats genuine psychosis vigorously while taking every one of those caveats seriously.

Selected References and Further Reading

1. Kapur, S. (2003). Psychosis as a state of aberrant salience: A framework linking biology, phenomenology, and pharmacology in schizophrenia. American Journal of Psychiatry, 160(1), 13–23.
2. Leucht, S., et al. (2013). Comparative efficacy and tolerability of 15 antipsychotic drugs in schizophrenia: A multiple-treatments meta-analysis. The Lancet, 382(9896), 951–962.
3. Howes, O.D., & Kapur, S. (2009). The dopamine hypothesis of schizophrenia: Version III — the final common pathway. Schizophrenia Bulletin, 35(3), 549–562.
4. Kapur, S., & Seeman, P. (2001). Does fast dissociation from the dopamine D2 receptor explain the action of atypical antipsychotics? American Journal of Psychiatry, 158(3), 360–369.
5. Meltzer, H.Y. (1999). The role of serotonin in antipsychotic drug action. Neuropsychopharmacology, 21(2 Suppl), 106S–115S.
6. Kane, J., et al. (1988). Clozapine for the treatment-resistant schizophrenic. Archives of General Psychiatry, 45(9), 789–796.
7. Meltzer, H.Y., et al. (2003). Clozapine treatment for suicidality in schizophrenia (InterSePT). Archives of General Psychiatry, 60(1), 82–91.
8. Wunderink, L., et al. (2013). Recovery in remitted first-episode psychosis at 7 years of follow-up of an early dose reduction/discontinuation or maintenance treatment strategy. JAMA Psychiatry, 70(9), 913–920.
9. Harrow, M., & Jobe, T.H. (2007/2012). Factors involved in outcome and recovery in schizophrenia patients not on antipsychotic medications. Journal of Nervous and Mental Disease / Psychological Medicine.
10. Leucht, S., et al. (2012). Antipsychotic drugs versus placebo for relapse prevention in schizophrenia: A systematic review and meta-analysis. The Lancet, 379(9831), 2063–2071.
11. Ho, B.C., Andreasen, N.C., et al. (2011). Long-term antipsychotic treatment and brain volumes. Archives of General Psychiatry, 68(2), 128–137.
12. Correll, C.U., et al. (2015). Cardiometabolic risk in patients with first-episode schizophrenia. JAMA Psychiatry, 72(12), 1227–1238.
13. Schneider, L.S., Dagerman, K.S., & Insel, P. (2005). Risk of death with atypical antipsychotic drug treatment for dementia. JAMA, 294(15), 1934–1943.
14. Moncrieff, J. (2013). The Bitterest Pills: The Troubling Story of Antipsychotic Drugs. Palgrave Macmillan.
15. Chouinard, G., et al. (2017). Antipsychotic-induced dopamine supersensitivity psychosis. Psychotherapy and Psychosomatics, 86(4), 189–219.
16. Kishi, T., et al. (2021/2023). VMAT2 inhibitors (valbenazine, deutetrabenazine) for tardive dyskinesia: Meta-analyses. Journal of clinical psychopharmacology / movement disorders literature.
17. Howes, O.D., et al. (2017). Treatment-resistant schizophrenia: TRRIP working group consensus. American Journal of Psychiatry, 174(3), 216–229.
18. Tiihonen, J., et al. (2017). Real-world effectiveness of antipsychotic treatments in schizophrenia (and 20-year mortality cohort studies). JAMA Psychiatry / World Psychiatry.
19. Correll, C.U., et al. (2022/2024). Trial of xanomeline–trospium (muscarinic agonist, KarXT) in schizophrenia. New England Journal of Medicine / The Lancet.
20. Goff, D.C., et al. (2017). The long-term effects of antipsychotic medication on clinical course in schizophrenia. American Journal of Psychiatry, 174(9), 840–849.