Tardive dyskinesia (TD) is a persistent, often irreversible movement disorder caused by chronic exposure to dopamine receptor blocking agents, most commonly antipsychotics, but also metoclopramide and prochlorperazine. It is classified in DSM-5-TR under Medication-Induced Movement Disorders and Other Adverse Effects of Medication, distinct from acute dystonia, akathisia, and drug-induced parkinsonism. The classic presentation is choreoathetoid movements of the orofacial region (lip smacking, tongue protrusion, chewing), though limb and truncal involvement is common and disabling. Two VMAT2 inhibitors, valbenazine and deutetrabenazine, are FDA-approved and represent the only evidence-based pharmacologic treatments; clozapine switch and offending-agent dose reduction are adjunctive strategies. Bottom line: screen every patient on a dopamine blocker with the AIMS at baseline and at least every six months, because early recognition is the single most important determinant of reversibility.

TD remains common despite the shift to second-generation antipsychotics; the prevalence reduction with newer agents is real but modest.

Prevalence and incidence

• Pooled prevalence among patients on chronic antipsychotic therapy is approximately 25%, with first-generation antipsychotic (FGA) exposure yielding around 30% and second-generation antipsychotic (SGA) exposure around 20%.^[1]
• Annual incidence with FGAs is roughly 5% per year of exposure in mixed-age adults, with cumulative risk continuing to rise across years of treatment.^[1-2]
• Annual incidence with SGAs is approximately 3% per year of exposure, lower than FGAs but not negligible.^[1]

Demographic and treatment-related risk factors

• Older age is the strongest single risk factor; cumulative incidence in patients over 55 exceeds 25% within the first year of antipsychotic exposure.^[2]
• Female sex, particularly postmenopausal, confers higher risk.^[2]
• African ancestry is associated with elevated TD risk in several cohorts.^[2]
• Mood disorders carry higher TD risk than schizophrenia at equivalent antipsychotic doses.^[2]
• Diabetes mellitus, intellectual disability, traumatic brain injury, and HIV are all independently associated with TD.^[2]
• Early extrapyramidal symptoms (drug-induced parkinsonism, acute dystonia) within the first months of therapy predict later TD.^[2]
• Cumulative antipsychotic dose and duration of exposure show a dose-response relationship.^[1-2]
• Intermittent dosing and treatment interruptions increase risk relative to continuous exposure.^[2]
• Concurrent anticholinergic use is associated with worsening of orofacial TD, though it may relieve coexisting dystonia.^[3]

The dominant model is post-synaptic dopamine D2 receptor supersensitivity in the nigrostriatal pathway, though this framework is incomplete and several competing mechanisms remain plausible.

Dopamine receptor supersensitivity hypothesis

• Chronic D2 blockade upregulates and sensitizes striatal D2 receptors, producing exaggerated dopaminergic signaling when receptor occupancy fluctuates.^[3-4]
• This explains the paradoxical worsening of dyskinesias on dose reduction and transient suppression on dose escalation, the so-called masking effect.^[3]
• The hypothesis is supported by animal models and the efficacy of presynaptic dopamine depletion via VMAT2 inhibition.^[4]
• Limitations: not all chronically exposed patients develop TD, and the time course does not always match receptor changes.^[3-4]

Alternative and complementary mechanisms

• GABAergic dysfunction in the striatum and globus pallidus, with reduced inhibitory output through the indirect pathway, is implicated by animal and post-mortem data.^[3]
• Oxidative stress and mitochondrial dysfunction from chronic antipsychotic exposure may damage striatal neurons, consistent with the partial benefit of antioxidants in some trials.^[3,5]
• Maladaptive synaptic plasticity at corticostriatal synapses is an emerging framework that integrates dopaminergic and glutamatergic contributions.^[3]

Genetic contributions

• Polymorphisms in DRD2, DRD3 (notably Ser9Gly), CYP2D6, COMT, and MnSOD have been associated with TD risk, though no single variant has clinical utility for prediction.^[2-3]
• Heritability estimates from family studies suggest a meaningful genetic contribution but most variance is unexplained.^[2]

Causative agents

• All dopamine D2 receptor blocking agents carry TD risk, including non-psychiatric drugs.^[3]
• Higher-risk: haloperidol, fluphenazine, and other high-potency FGAs.^[1,3]
• Intermediate-risk: risperidone, paliperidone, olanzapine.^[1]
• Lower-risk among antipsychotics: quetiapine, aripiprazole; clozapine has the lowest risk and may even improve established TD.^[1,6]
• Non-antipsychotic culprits: metoclopramide (often unrecognized; carries an FDA boxed warning), prochlorperazine, promethazine.^[7]

DSM-5-TR places TD in the Medication-Induced Movement Disorders chapter. The diagnostic framework hinges on three elements: characteristic movements, sufficient exposure, and persistence beyond drug withdrawal. DSM-5-TR criteria (ICD-10-CM code G24.01):

• Involuntary athetoid or choreiform movements lasting at least a few weeks, developing during or within four weeks of withdrawal from an oral neuroleptic, or within eight weeks of withdrawal of a depot preparation.^[8]
• Cumulative neuroleptic exposure of at least a few months: three months for adults under 60, and one month for adults 60 and older.^[8]
• Symptoms are not better explained by another medical condition (e.g., Huntington disease, Wilson disease, Sydenham chorea) or by another medication.^[8]

Specifiers and qualifiers

• DSM-5-TR does not formally subtype TD by topography, but clinical practice distinguishes orofacial (lingual-facial-buccal), limb-truncal, and generalized variants.^[3]
• Tardive dystonia, tardive akathisia, tardive tics, and tardive tremor are recognized as related tardive syndromes with overlapping mechanisms but distinct phenomenology.^[3]
• Withdrawal-emergent dyskinesia is a transient self-limited dyskinesia appearing after abrupt discontinuation, typically resolving within several weeks; it is most common in children and overlaps but is not synonymous with TD.^[3]

ICD-11 placement

• ICD-11 codes neuroleptic-induced TD under 6E40, in the chapter on disorders due to substances or medications, with subcategories paralleling the tardive phenomenology recognized in DSM-5-TR.^[9]
• ICD-11 does not impose the DSM-5-TR minimum exposure thresholds; clinicians outside the United States may diagnose at shorter durations when phenomenology and exposure clearly correspond.^[9]

TD is phenomenologically heterogeneous; the classic orofacial pattern is the most recognized but limb, truncal, and respiratory involvement frequently coexist or dominate.

Orofacial (buccolingual-masticatory) features

• Repetitive tongue movements including protrusion, lateral darting, and the bon-bon sign (tongue pushing into the cheek).^[3]
• Lip smacking, puckering, and pursing.^[3]
• Chewing or grinding movements without food in the mouth.^[3]
• Grimacing and excessive blinking or blepharospasm.^[3]

Limb and truncal features

• Choreoathetoid movements of the fingers, producing the classic piano-playing or guitar-playing appearance.^[3]
• Foot tapping, toe writhing, and ankle inversion.^[3]
• Truncal rocking, pelvic thrusting, and respiratory dyskinesia presenting as grunting, irregular breathing, or unexplained dyspnea.^[3,10]

Tardive dystonia

• Sustained, often painful posturing (retrocollis, opisthotonos, blepharospasm) that responds less reliably to VMAT2 inhibitors and may benefit from botulinum toxin or anticholinergics.^[3,11]
• More common in younger patients and in men, in contrast to the orofacial form.^[3]

Course-related features

• Movements are typically suppressed by voluntary motor activity in the same body part and worsened by distraction or emotional arousal.^[3]
• Symptoms cease during sleep, distinguishing TD from many primary movement disorders that persist.^[3]
• Patients are often unaware of mild orofacial movements; family members or clinicians are the first to notice.^[3]

The differential includes other movement disorders triggered by the same drugs, primary neurologic disease, and medical mimics. A careful temporal history of drug exposure and movement onset is decisive.

Other medication-induced movement disorders

• Acute dystonia: sustained muscle contractions, typically within hours to days of initiation, responsive to anticholinergics, distinguished by acute onset.^[3,12]
• Akathisia: subjective restlessness with stereotyped pacing or rocking, distinct from the choreoathetoid quality of TD, though tardive akathisia complicates the picture.^[3,12]
• Drug-induced parkinsonism: bradykinesia, rigidity, and pill-rolling tremor, the opposite phenomenology to TD, though both may coexist in the same patient.^[3,12]
• Neuroleptic malignant syndrome: hyperthermia, rigidity, autonomic instability, presenting as an acute emergency rather than a chronic dyskinesia.^[12]

Primary movement disorders

• Huntington disease: family history, cognitive decline, psychiatric features, and CAG repeat expansion in HTT.^[13]
• Wilson disease: hepatic dysfunction, Kayser-Fleischer rings, low ceruloplasmin. Essential to exclude in any patient under 40 with a new movement disorder.^[13]
• Sydenham chorea: post-streptococcal, typically children and adolescents, often resolves spontaneously.^[13]
• Idiopathic orofacial dyskinesia of the elderly (edentulous dyskinesia): occurs in older adults without antipsychotic exposure, often related to ill-fitting dentures or tooth loss.^[3]
• Essential tremor and dystonic tremor: rhythmic rather than choreoathetoid; absence of neuroleptic exposure history.^[13]

Medical and metabolic mimics

• Hyperthyroidism: tremor, chorea in severe cases; check TSH.^[13]
• Systemic lupus erythematosus and antiphospholipid syndrome: chorea may be a presenting feature.^[13]
• HIV-associated movement disorders.^[13]
• Stimulant intoxication or chronic methamphetamine use: stereotypies and chorea.^[13]
• Anti-NMDA receptor encephalitis: orofacial dyskinesias are a hallmark; consider in younger patients with psychiatric symptoms, seizures, or autonomic instability.^[13]

Assessment combines a structured movement examination, a thorough exposure history, and targeted workup to exclude mimics. The AIMS is the standard instrument and should be documented in the chart.

History elements

• Complete antipsychotic exposure history including cumulative dose, duration, dose changes, and treatment interruptions.^[2]
• Exposure to non-antipsychotic dopamine blockers (metoclopramide, prochlorperazine, promethazine) that patients and prescribers often overlook.^[7]
• Onset and progression of movements, awareness of movements, and functional impact (eating, speech, social interaction, respiration).^[3]
• Family history of movement disorders or early-onset psychiatric illness.^[13]
• Substance use, particularly stimulants.^[13]

Examination using the AIMS protocol

• The Abnormal Involuntary Movement Scale (AIMS) is a 12-item examination assessing seven body regions on a 0-4 severity scale plus global judgments.^[15]
• Conduct in three phases: observation while the patient is unaware, activation maneuvers (tongue protrusion, finger tapping, walking), and a brief structured rating.^[15]
• A score of 2 (mild) in two or more regions, or 3 (moderate) in one region, with at least three months of antipsychotic exposure, supports a TD diagnosis under Schooler-Kane research criteria.^[15]
• Document AIMS at antipsychotic initiation, every six months on FGAs, and at least annually on SGAs; in older adults, every three to six months regardless of class.^[14]

Laboratory and imaging workup

• Targeted to exclude mimics rather than confirm TD, which is a clinical diagnosis.^[3]
• Reasonable initial screen in atypical or new presentations: TSH, CBC, comprehensive metabolic panel, ceruloplasmin, HIV testing if indicated.^[13]
• Neuroimaging (MRI) when the presentation is unilateral, rapidly progressive, or accompanied by cognitive decline.^[13]
• Genetic testing (HTT for Huntington disease) when family history or course suggests primary neurodegeneration.^[13]

Always exclude Wilson disease in any patient under 40 with a new movement disorder, regardless of neuroleptic exposure.^[13]

Management proceeds along three parallel lines: address the offending agent when feasible, initiate a VMAT2 inhibitor for clinically significant movements, and treat associated dystonia or psychiatric instability. The American Academy of Neurology guideline and the APA practice guideline converge on VMAT2 inhibitors as the only treatments with strong supporting evidence.^[14]

Pharmacotherapy

• VMAT2 inhibitors are first-line. Strong evidence supports both valbenazine and deutetrabenazine for symptom reduction in adults with TD.^[14,16-17]
• valbenazine 40 mg PO QD initial dose for one week, then increase to 80 mg daily; some patients benefit from 60 mg daily when 80 mg is not tolerated.^[16]
• deutetrabenazine 6 mg PO BID initial total daily dose 12 mg, titrated weekly by 6 mg increments to a maximum of 48 mg per day (36 mg with strong CYP2D6 inhibitors or in poor metabolizers).^[17]
• Tetrabenazine is used off-label for TD where valbenazine and deutetrabenazine are unavailable; it is less tolerable due to depression, sedation, and akathisia and is not FDA-approved for TD.^[14]
• Evidence suggests benefit from clonazepam 1-4.5 mg PO daily for symptomatic relief, particularly in dystonic forms, though tolerance develops within months.^[14]
• Limited evidence suggests Ginkgo biloba extract EGb 761 reduces TD severity in patients with schizophrenia at 240 mg per day.^[14]
• Amantadine has been used adjunctively with modest evidence and is sometimes recommended when VMAT2 inhibitors are contraindicated or unavailable.^[14]
• Anticholinergic agents should be discontinued or reduced; they typically worsen orofacial TD even though they relieve coexisting drug-induced parkinsonism or acute dystonia.^[3,14]
• Vitamin E and vitamin B6 have shown small benefits in some trials but evidence is inconsistent and they are not first-line.^[14]

Offending-agent management

• Reassess whether continued dopamine receptor blockade is clinically necessary; for non-psychiatric indications (metoclopramide for nausea), discontinue when feasible.^[7,14]
• For schizophrenia and bipolar disorder, abrupt discontinuation risks relapse and may transiently worsen TD via unmasking; do not stop antipsychotics solely to treat TD without psychiatric assessment.^[14]
• Evidence suggests switching to clozapine reduces TD severity and is the preferred substitution when antipsychotic continuation is required.^[6,14]
• Dose reduction of the offending antipsychotic may be considered; expect transient worsening before improvement.^[14]
• Switching from an FGA to an SGA other than clozapine has not been consistently shown to improve established TD.^[14]

Psychotherapy

• No psychotherapy modifies TD directly.^[14]
• Supportive psychotherapy and psychoeducation help patients adapt to functional impact and cope with stigma; this is a routine adjunct, not a disease-modifying intervention.^[14]

Neuromodulation

• Botulinum toxin injection is recommended for focal tardive dystonia (blepharospasm, cervical dystonia, oromandibular dystonia) with moderate supporting evidence.^[11,14]
• Deep brain stimulation (DBS) of the globus pallidus internus is an option for severe, medication-refractory tardive dystonia or generalized TD; evidence is limited to case series and small cohorts but effect sizes are substantial.^[14,18]
• Electroconvulsive therapy does not treat TD; it may be used for coexisting psychiatric indications without worsening TD.^[14]

Adjunctive

• Treat coexisting depression, anxiety, and psychotic relapse aggressively; psychiatric decompensation makes TD harder to manage and worsens functional outcomes.^[14]
• Optimize dental care; mucosal injury from chronic orofacial movements requires attention.^[3]
• Speech and swallow evaluation in patients with significant orofacial or pharyngeal TD; aspiration risk is real in advanced disease.^[10]

VMAT2 inhibitors are tolerable but not benign, and the broader TD evidence base is constrained by short trial durations and selected populations. Clinicians should weigh the adverse-effect profiles against the disability of the dyskinesia itself.

• Common adverse effects of VMAT2 inhibitors include somnolence, fatigue, akathisia, headache, and dry mouth; rates are dose-dependent.^[16-17]
• Serious or rare adverse effects include QT prolongation, neuroleptic malignant syndrome (rare), parkinsonism, and mood lowering with suicidal ideation in vulnerable patients.^[16-17]
• CYP2D6 interactions are clinically important for deutetrabenazine; dose reduction is required with strong CYP2D6 inhibitors such as paroxetine, fluoxetine, bupropion, and quinidine.^[17]
• Long-term safety beyond two years is incompletely characterized for both valbenazine and deutetrabenazine.^[16-17]
• Cost remains a major access barrier in many health systems; both agents are branded and substantially more expensive than tetrabenazine or supportive care.^[14]
• Most pivotal trials excluded patients with severe psychiatric instability, recent suicidality, or substance use disorders, limiting generalizability to the broader population most affected by TD.^[16-17]
• Trial endpoints rely on AIMS reduction, which correlates imperfectly with patient-reported function and quality of life.^[14]
• There is no evidence that any intervention reverses fully established, longstanding TD; the goal is symptom reduction and prevention of progression rather than full resolution.^[14]

Risk and management considerations vary meaningfully by age, sex, and comorbid condition. Tailoring is essential.

Older adults

• Incidence after one year of antipsychotic exposure exceeds 25% in adults over 55.^[2]
• DSM-5-TR shortens the minimum exposure threshold to one month for adults 60 and older.^[8]
• Consider AIMS screening every three to six months in older adults on any antipsychotic.^[14]
• Both VMAT2 inhibitors are approved for older adults; start at the lowest dose given orthostasis and fall risk.^[16-17]

Pediatric and adolescent populations

• Withdrawal-emergent dyskinesia after antipsychotic discontinuation is more common in children than adults and usually resolves within weeks to months.^[3]
• True persistent TD does occur in pediatric patients on long-term antipsychotics; AIMS monitoring is appropriate when continued treatment is necessary.^[14]
• Valbenazine and deutetrabenazine are not FDA-approved in pediatric populations; off-label use requires specialist involvement.^[16-17]

Pregnancy and postpartum

• Pregnancy data on VMAT2 inhibitors are limited; both have animal reproductive toxicity signals and should be avoided in pregnancy unless benefits clearly outweigh risk.^[16-17]
• Postmenopausal women carry elevated TD risk; estrogen status appears protective.^[2]

Comorbid medical illness

• Hepatic impairment requires dose adjustment for valbenazine; deutetrabenazine is contraindicated in hepatic impairment.^[16-17]
• Cardiovascular disease: QT-prolonging comedications should be reviewed before VMAT2 inhibitor initiation.^[16-17]
• HIV, traumatic brain injury, and diabetes mellitus all elevate baseline TD risk; threshold for monitoring should be correspondingly lower.^[2]

Comorbid substance use

• Stimulant use independently produces stereotypies and chorea and complicates assessment; obtain a urine drug screen when the diagnosis is unclear.^[13]
• Alcohol use disorder, common in serious mental illness, worsens functional outcomes and may worsen movement-disorder phenomenology.^[14]

Cultural considerations

• Underrecognition of TD in patients with limited English proficiency or whose movements are normalized by family or clinicians delays treatment; structured AIMS screening helps reduce this disparity.^[14]

TD is often persistent. Some patients improve substantially after offending-agent withdrawal, but a meaningful fraction develop chronic, partially disabling movements that require long-term management.

Natural history

• Approximately one-third of patients experience some spontaneous improvement after offending-agent discontinuation, though improvement is slow and often incomplete.^[3,14]
• A subset shows progression despite intervention, particularly older patients and those with prolonged exposure before recognition.^[2-3]
• Tardive dystonia tends to be more persistent than orofacial dyskinesia and less responsive to VMAT2 inhibitors.^[11]

Treatment response

• Approximately 30-40% of patients achieve at least 50% AIMS reduction on a VMAT2 inhibitor at the recommended dose over 6-12 weeks.^[16-17]
• Response is typically maintained on continued treatment; discontinuation often leads to symptom return.^[16-17]
• Earlier initiation, before the disorder is well established, is associated with better outcomes.^[14]

Functional and mortality outcomes

• Severe TD compromises eating, speech, social participation, and ambulation; functional disability is meaningful and underrecognized.^[10,14]
• Respiratory dyskinesia is associated with aspiration and, rarely, respiratory compromise.^[10]
• Excess mortality data are confounded by underlying psychiatric illness, but severe TD likely contributes to morbidity and accelerated decline in older patients.^[10]

TD is rarely an acute emergency, but acute presentations and safety-relevant decompensations do occur.

Acute presentations requiring urgent evaluation

• New-onset severe orofacial dyskinesia with autonomic instability or altered mental status: consider anti-NMDA receptor encephalitis, drug intoxication, or NMS.^[12-13]
• Respiratory dyskinesia producing hypoxia or aspiration: airway assessment, oxygen, swallow evaluation.^[10]
• Acute worsening of dystonic posturing causing pain or musculoskeletal injury: consider benztropine or botulinum toxin referral.^[11-12]

Safety in psychiatric decompensation

• Patients with TD who experience psychotic relapse often require antipsychotic continuation or escalation; do not withhold necessary antipsychotic treatment to preserve TD response.^[14]
• Clozapine is the preferred substitute when antipsychotic continuation is required in established TD.^[6,14]
• Monitor mood during VMAT2 inhibitor treatment; suicidal ideation is a class warning.^[16-17]

Several important questions remain unsettled despite the regulatory milestone of two FDA-approved VMAT2 inhibitors.

• Whether SGAs meaningfully reduce TD risk compared to FGAs: meta-analyses suggest a modest reduction, but methodological heterogeneity and confounding by indication leave the magnitude uncertain.^[1,3]
• Whether clozapine is disease-modifying for established TD or simply has lower D2 affinity and therefore less ongoing receptor blockade: mechanistic studies are inconclusive.^[6]
• Optimal duration of VMAT2 inhibitor therapy: trials extend to one year, but most patients require indefinite treatment to maintain response, with limited long-term safety data.^[16-17]
• Comparative effectiveness of valbenazine versus deutetrabenazine: no head-to-head trial exists; indirect comparisons suggest broadly similar efficacy.^[14]
• Role of VMAT2 inhibitors in mild TD identified on screening: trials enrolled patients with at least moderate severity, and the threshold for treating mild movements is debated.^[14,16-17]
• Generic tetrabenazine as a lower-cost alternative: less tolerable but considerably cheaper, raising access and equity questions where branded VMAT2 inhibitors are unaffordable.^[14]
• Genetic prediction of TD risk: candidate-gene findings have not yielded a clinically useful risk panel, and polygenic scores remain investigational.^[2-3]
• Whether anticholinergic withdrawal alone improves orofacial TD in modern practice, where many patients are no longer maintained on benztropine: evidence is older and may not generalize.^[3,14]

• TD is a hyperkinetic movement disorder caused by chronic dopamine D2 receptor blockade, characterized by choreoathetoid movements of the orofacial region, limbs, or trunk.^[3]
• DSM-5-TR requires at least three months of cumulative neuroleptic exposure for adults under 60, and at least one month for adults 60 and older.^[8]
• The AIMS is the standard screening instrument; document at antipsychotic initiation and every six months thereafter.^[15]
• Valbenazine and deutetrabenazine are the only FDA-approved treatments for TD.^[14,16-17]
• Clozapine is the antipsychotic with the lowest TD risk and is the preferred switch when continued antipsychotic treatment is required.^[6,14]
• Strongest risk factors are older age, female sex (particularly postmenopausal), African ancestry, mood disorder diagnosis, and early extrapyramidal symptoms.^[2]
• Metoclopramide is a frequently overlooked cause of TD and carries an FDA boxed warning; cumulative exposure beyond 12 weeks should prompt reassessment.^[7]
• Anticholinergic medications typically worsen orofacial TD even though they relieve coexisting drug-induced parkinsonism and acute dystonia.^[3,14]
• Tardive movements characteristically suppress during sleep and worsen with distraction.^[3]
• Dose reduction of the offending antipsychotic transiently worsens TD through unmasking before any improvement.^[3,14]
• Tardive dystonia is more common in younger patients and men, in contrast to the orofacial form which predominates in older women.^[3]
• Botulinum toxin is recommended for focal tardive dystonia; deep brain stimulation of the globus pallidus internus is an option for refractory severe cases.^[11,14,18]
• Always exclude Huntington disease, Wilson disease, and anti-NMDA receptor encephalitis when the presentation is atypical or progressive.^[13]
• VMAT2 inhibitors carry a class warning for depression and suicidal ideation derived from the tetrabenazine experience in Huntington disease.^[16-17]
• Approximately one-third of patients experience some improvement after offending-agent discontinuation; full reversal is uncommon once TD is well established.^[3,14]

No external funding. No conflicts of interest declared. Peer-review status: pending.