Article Text

Download PDFPDF

Tardive movement disorders: a practical approach
  1. Tabish A Saifee,
  2. Mark J Edwards
  1. Sobell Department of Motor Neuroscience and Movement Disorders, UCL Institute of Neurology, London, UK
  1. Correspondence to Dr M J Edwards, Sobell Department of Motor Neuroscience and Movement Disorders, UCL Institute of Neurology, Queen Square, London WC1N 3BG, UK; m.edwards{at}

Statistics from

Tardive movement disorders are a pervasive problem, particularly in a mental health setting, and the cause of significant disability. They typically present as an orolingual chorea but also can manifest as retrocollis, other forms of dystonia, tics and tremor. They are induced exclusively by dopamine receptor blocking drugs and often do not remit, even when the drug is stopped, and can paradoxically worsen. Here we discuss their phenomenology, identify important features for diagnosis and differential diagnoses, and discuss practical management steps and potential prognostic features.


There are three main groups of drug induced movement disorders. Firstly, there are acute onset movement disorders, typically occurring on drug initiation and less commonly with changes in dose—for example, acute dystonic reactions and neuroleptic malignant syndrome.1 Secondly, there are movement disorders that emerge during drug treatment but stop when the drug is withdrawn2—for example, parkinsonism induced by dopamine receptor blocking drugs (DRBs) or tremor induced by sodium valproate. Thirdly, and the subject of this article, there are the ‘tardive’ movement disorders (TMDs), exclusively induced by DRBs, emerging typically after chronic treatment and often not remitting, even when the drug is stopped. Here we discuss their phenomenology, diagnosis and practical management.

The term ‘tardive’ derives from the French word for slow—tardif—signifying the typical onset of symptoms after chronic drug treatment. These disorders are best thought of as drug induced movement disorders that persist: this is preferable to specifying a length of time of treatment before onset, as rarely this can be within a few weeks or months.3 TMDs are caused exclusively by DRBs such as antipsychotics (neuroleptics) but also by DRBs prescribed for nausea (metoclopramide)4 and dizziness (prochlorperazine). As so often the case with movement disorders, there is a rather confusing terminology used to describe these disorders. The catch-all term ‘tardive dyskinesia’ is used to describe both the whole range of TMDs and also the specific case of choreic movements: brief, semi-purposeful, irregular movements that are not repetitive or rhythmic but appear to flow from one muscle to the next that typically affect the mouth. Other terms describe particular patterns of tardive movement disorder: tardive dystonia, and the less certain categories of tardive myoclonus, tardive tics, tardive tremor, tardive akathisia and tardive parkinsonism. This review aims to cover all types of TMDs but concentrates on the two commonest manifestations: tardive dyskinesia (chorea) and tardive dystonia.


Chlorpromazine, first introduced in 1952, revolutionised the treatment of psychosis. The frequent admission of patients to ‘correctional institutions’ was needed less often and outpatient management became the norm. However, within a few years, there were reports of chlorpromazine induced movement disorders.5 While many such patients had acute drug reactions or parkinsonism on the drug, a number developed a chronic movement disorder related to chlorpromazine exposure which did not remit on drug withdrawal. The entity of TMD was born, and over subsequent years clinicians have increasingly recognised the disability associated with this movement disorder (box 1).

Box 1 How do tardive movement disorders affect patients?

  • They can be disabling and disfiguring

  • Some dyskinetic movements create social stigma

  • Remission is rare, and therefore a lifelong movement disorder is typical

  • The symptoms may adversely affect the ability to carry out activities of daily living

  • They can make a patient reluctant to take any further medications

  • They can rarely be life threatening, particularly when affecting respiratory or bulbar muscles

Why does it occur?

Although the mechanism is not clearly understood, TMD most likely relates to dopamine receptor hypersensitivity caused by chronic receptor blockade by the DRBs, particularly in the nigrostriatal pathway. The hypothesis is that DRBs block D2 receptors while endogenous dopamine chronically stimulates D1 receptors thus creating an imbalance between the direct and indirect pathways in the basal ganglia, and leading to the movement disorder. Older or ‘typical’ DRBs have a greater affinity for D2 receptors and have a higher risk of causing TMDs than atypical DRBs (table 1). Other neurotransmitter systems such as γ-aminobutyric acid and acetylcholine are probably also involved.6 Genetic polymorphisms seem to contribute to the risk of developing a tardive syndrome7 which may explain the large variations in likelihood of occurrence, severity of movement disorder and in the duration of drug exposure between individual patients (see ‘Who is at risk’, below).

Table 1

Typical and atypical dopamine receptor blocking drugs

How common are TMDs?

The incidence of TMD following typical DRB use is around 5% per year for the first few years. Over 10 years though, the cumulative risk is nearly 50%. Indeed, just under a third of all patients on chronic typical DRB treatment have a TMD. This is lower for atypical DRBs although the reported prevalence varies widely.8

Who is at risk?

Although TMDs are more likely on typical DRBs, they also occur with atypical DRB treatment. Although atypical DRBs are favoured partly because they are thought less likely to cause TMDs, it may be that the total dose of DRBs is more important than whether the patient takes an atypical or typical DRB.9 It is hard to separate the contribution of the underlying pathophysiology of the mental health disorders being treated from the risk of developing a TMD. However, from cases of TMD occurring in patients with normal mental health treated with DRBs for nausea and dizziness, it is clear that the disorder can occur in anyone. Patients with affective disorders rather than psychotic disorders appear more likely to develop TMDs but this may simply be a chance association. Other risk factors (possibly also chance associations related to DRB type, duration and dose) include: non-white patients, older patients, frequent stopping and starting of DRBs, structural brain lesions and previous DRB associated acute extrapyramidal symptoms.10

How does it start?

TMD rarely begins before 6 weeks of DRB treatment and takes an average of 5 years to begin. Rarely, it occurs after just a few doses and, at the other end of the spectrum, after two decades of treatment. The nature of the disorder in those few cases with persistent movement disorder after very brief exposure is unclear, and clinicians should be very cautious in diagnosing a TMD in this group. Patients often do not notice or report the initial mild abnormal movements, so careful monitoring and examination is useful here. Bear in mind that dyskinesias are widely reported to occur as part of schizophrenia, independently of drugs.11 Also, it can take many months to eliminate typical DRBs completely from the brain after long term use, so your patient still has a potential risk of developing a TMD after stopping DRB treatment. In fact, the paradoxical onset of symptoms on stopping or reducing treatment points strongly towards a TMD. This contrasts with the acute effects of DRBs, which develop on treatment initiation, dose increase or swapping to a higher potency agent.

What are the typical features?

Most patients with TMD are affected by oro-facio-lingual movements, usually called tardive dyskinesia, but best described as chorea. The movements include lip pursing, lip smacking, tongue protrusion, writhing tongue movements and chewing. Protrusion of the tongue into the cheek is a classical feature. These can be disfiguring and, at worst, can impact on eating, adequate nutrition, swallowing and speaking. Piano playing movements of the fingers commonly accompany the facial movements. Raising the eyebrows, frowning and head nodding may also occur. TMD also manifests as hip rocking or to and fro movements of the torso. It can give the impression of a patient walking on the spot, with rhythmic shifting of weight between feet, often attributed to akathisia.12 TMD may affect the diaphragmatic muscles, giving speech an arrhythmic quality, which can become interrupted by grunting and groaning noises. TMD can also present with respiratory symptoms such as irregular breathing, tachypnoea and, rarely, when severe, aspiration, pneumonia and even death. Some authors have proposed describing tardive dyskinesia movements as ‘stereotypies’.13 Although an academic point, the increase in tardive dyskinesia movements occurring with distraction (eg, asking the patient to close their eyes and mentally to count backwards) is typical of chorea but atypical for stereotypies.

The next commonest presentation of TMD is with dystonia: abnormal involuntary muscle contraction leading to posturing of the affected body parts. As with idiopathic dystonia, the postures are not usually fixed and there may be slow writhing movements. Abnormal postures commonly affect the neck, and may spread to become segmental or generalised. It is uncommon for tardive dystonia to remain focal and also uncommon for it to start in the legs or arms. Task specificity is not typical for tardive dystonia and sensory tricks (or geste) may be absent3—that is, the ability of patients to correct abnormal posture with a light touch to the face/neck. Often the most striking clinical feature in tardive dystonia is extreme retrocollis (neck hyperextension). This contrasts with idiopathic cervical dystonia where retrocollis is very rare compared with the much more common turn or tilt of the head (torticollis/laterocollis). An axial dystonia, often with hyperextension of the back and scoliosis, also occurs in tardive dystonia.

Other movement disorders may occur as tardive phenomena but these are much rarer and indeed some doubt their existence. Thus there are reports of patients with myoclonus, tics, tremor and even parkinsonism developing after chronic DRB exposure, persisting even after drug withdrawal. It is possible that these are coincident to DRB. Akathisia, motor restlessness accompanied by a subjective urge to move, can also occur as a tardive phenomenon. Akathisia is a well known acute DRB side effect but in some patients it emerges only after years of treatment, leading some to suspect it may be a tardive syndrome too. This may be so, but supposed tardive akathisia often remits with drug withdrawal, unlike typical tardive dyskinesia or dystonia.

Mixed movement disorders are common in TMD (often combined chorea and dystonia), in which case it is useful to identify the dominant hyperkinetic movement and to use this to guide treatment (table 2).

Table 2

Classical features of tardive movement disorders

Differential diagnosis and investigation

Clinicians must consider a TMD in patients presenting with movement disorders, especially orofacial chorea and neck dystonia. It is not enough simply to ask, “Have you taken any antipsychotic drugs?” and to dismiss the possibility of a TMD if the patient answers, “No”. Patients may not be aware of the nature of the drugs they are on, and may be being treated with DRBs that are not antipsychotics. Box 2 covers important questions to ask patients with possible drug induced movement disorders.

Box 2 Important questions to ask your patient

  • What was the onset of the movement disorder in relation to starting any new medications?

  • Was there a change in the dose of an existing medication? Which drug, what dose and for how long was it taken prior to tardive movement disorder onset?

  • Were there factors that may have altered plasma levels of the drug (renal/hepatic failure, liver enzyme modulators)?

  • Were dopamine receptor blocking drugs used to treat psychiatric symptoms, nausea, anxiety or dizziness?

  • Has there been exposure to: contraceptive pill, illicit drugs (eg, cocaine), alcohol, nicotine, caffeine, ‘over the counter’ medications, herbal remedies/nutritional supplements?

  • Are there any previously used medications that have since been stopped?

Most cases with a typical presentation and a clear history of chronic DRB exposure require only screening with copper studies (particularly in younger patients) and brain imaging. There are several degenerative disorders, mostly very rare, which can present with the combination of psychiatric disturbance and the later appearance of a movement disorder, for example Wilson's disease or Huntington's disease (table 3). If such patients are treated with DRBs for their psychiatric disturbance, the later occurrence of a movement disorder can be incorrectly diagnosed as a TMD. This situation is uncommon compared with the overall incidence of TMDs. However, among referrals to a neurology clinic, the prevalence may be considerably higher, as the symptoms may have been refractory to treatment or the clinical course unusual, prompting the referral in the first place. It is important to identify these conditions as they have significant implications for prognosis, treatment and also genetic counselling and testing, particularly for Huntington's disease and Wilson's disease. Red flags to neurodegenerative diagnoses rather than TMD include a rapidly deteriorating course, persisting unilateral symptoms, additional neurological signs such as hyperreflexia, and other medical symptoms or signs, such as jaundice.14 Table 3 gives a summary of the important diseases that can masquerade as TMD. Remember also that DRB exposure may be incidental to the emergence of a primary movement disorder, such as primary dystonia, and this can cause some diagnostic confusion.

Table 3

Differential diagnoses of tardive movement disorders

It is of course possible for patients on DRBs to develop idiopathic craniocervical/segmental dystonia. There are no tests that can distinguish tardive dystonia from idiopathic dystonia; the clinical picture is all that the clinician can work with. There is some evidence that patients with idiopathic dystonia are susceptible to worsening on DRBs, and such patients should avoid these drugs. Late onset hyperkinetic movement disorders are common in Parkinson's disease after chronic levodopa therapy and also sometimes after chronic treatment with tricyclic antidepressants, selective serotonin reuptake inhibitors and serotonin and norepinephrine reuptake inhibitors.15 These can confuse the clinical picture if patients are also receiving DRBs, but critically, movement disorders due to these other drugs will stop if the drug is withdrawn.

Clinicians should be aware of withdrawal emergent syndrome and be able to distinguish this from typical TMD, of which it may be a forme fruste. Withdrawal emergent syndrome occurs in children and mirrors some of the clinical features of TMD.16 However, it develops within days of abrupt DRB withdrawal after chronic administration. Usually, dyskinetic movements occur in the neck, torso and arms but not the face. It is generally short lived and usually subsides after 3 months.


Treating TMD is often unsuccessful, making prevention particularly important. Prevention can be achieved by limiting DRB use to only when it is necessary and using atypical DRBs where possible. Try to use a minimally effective dose and avoid frequent ‘drug holidays’ after starting DRB. Also, be sure to address possible coexisting substance abuse—for example, cocaine—as this may worsen symptoms.

When patients present with a TMD, an important first step is to weigh up whether DRB treatment can be stopped. This decision must be taken with the treating psychiatrist. Those patients who stop the DRB quickly after the movement disorder emerges are much more likely to go into remission than those who stay on the drug. Failure to diagnose the movement disorder in the early stages and to stop the offending drug may be one reason why remission rates are only about 14%. Be wary that sudden DRB withdrawal can exacerbate TMD for the first few weeks but subsequently there is usually improvement.

For many psychiatric indications, DRB withdrawal is not possible. In this situation, a reasonable first line treatment is to switch a typical to an atypical DRB, in particular clozapine (figure 1). This drug requires blood monitoring due to potential side effects of agranulocytosis, and can also cause seizures.17 In view of this, other atypical DRBs may be helpful, such as quetiapine or reserpine, but these come with their own risks, such as diabetes mellitus, weight gain and sedation.

Figure 1

Management algorithm for tardive disorders.

For patients with predominant dystonia, the next step is a trial of an anticholinergic, particularly in young patients (trihexyphenidyl—table 4), However, this should not be done if dyskinesia is predominant, as this can worsen with anticholinergics. The dose titration should be gradual but some younger patients can reach high doses with benefit and limited side effects. If chorea dominates, catecholamine depleting drugs, particularly tetrabenazine, are useful,18 significantly alleviating symptoms in up to half of all patients. Depression, being a common side effect of tetrabenazine, is an important relative contraindication in the patient group who commonly develop TMDs. Parkinsonism as a side effect can also limit its use.

Table 4

Doses of common treatments in tardive dyskinesia and tardive dystonia

A combination of other drugs (each with fairly limited evidence for effectiveness) can supplement these firstline measures. Benzodiazepines such as diazepam and clonazepam may be useful19; tolerance can be a problem but a brief withdrawal period may help this. Baclofen can help in some patients (perhaps particularly those with axial spasms). Several other drugs have been beneficial in small case series: L-dopa,20 amantadine,21 levetiracetam,22 β-blockers, baclofen23 and ceruletide,24 a cholecystokinin analogue. Vitamin E is of variable benefit but probably has a role in prevention rather than treatment25; it could be tried in patients with mild symptoms not requiring more aggressive forms of treatment.

If these measures fail, and the patient cannot withdraw their DRB, the next step is to start an atypical DRB, in particular clozapine. The rationale for using certain DRBs to treat a movement disorder caused by DRBs may seem rather twisted but relates both to the different mode of action, certainly of clozapine compared with other DRBs, and the clinical finding that TMDs sometimes first occur and certainly may worsen during withdrawal of DRBs.

Specific circumstances

If symptoms are focal—for example, blepharospasm, oro-facio-lingual-masticatory dyskinesia or cervical dystonia—then botulinum toxin injections are beneficial.26 This benefit lasts 3–4 months before wearing off and so requires regular repeated treatment. Botulinum toxin can help even when other treatments have failed.26 Botulinum toxin is useful as it is easily administered and has relatively few common significant side effects such as dysphagia when muscles near the pharynx are injected. However, the benefits of botulinum toxin are ultimately short lived.

Patients with severe axial dystonia who have not responded to standard treatment measures can consider continuous intrathecal baclofen, although there is only very limited evidence for its benefit. It is a last resort, reserved for patients whose dystonia is functionally impairing their walking or other activity of daily living.27 As with other patients with dystonia who receive intrathecal baclofen, catheter displacement or breakage is a common adverse event.

For tardive akathisia, try β-blockers first and if these do not help, then add a benzodiazepine such as clonazepam; finally, try amantadine or clonidine if the first steps were unsuccessful.28

No response to medical intervention?

Previously, patients with tardive dystonia were assumed to be poor candidates for deep brain stimulation (DBS) and were excluded from trials. Recently, there have been several reports of tardive dystonia treated with DBS of the internal segment of the globus pallidus. Contrary to previous assumptions, most instances of tardive dystonia may show motor improvement from DBS approaching that seen in treated patients with primary dystonia29; there does not, as yet, appear to be an increase in psychiatric adverse events. However, in view of the invasiveness and lack of long term follow-up in treated patients, this surgical option should be reserved for those with severe symptoms refractory to other simpler measures.

Practice points

  • Tardive movement disorder (TMD) most commonly presents as an orolingual dyskinesia but can also manifest as retrocollis, other forms of dystonia, tics and tremor.

  • Prevention is most important: use DRBs only when required and then use the smallest effective doses for the briefest necessary time.

  • Consider alternative diagnoses such as a neurodegenerative disease.

  • Aim to reduce and stop the causative drugs. Explore the history of drugs of abuse and alternative therapies used by patient.

  • Avoid abrupt cessation of the causative drug, as this may worsen matters.

  • Stopping the causative drug is often not feasible and should be balanced with the need for treatment of the initial problem, eg, psychosis.

  • Try swapping causative drug for quetiapine or clozapine if continuation of DRB therapy is required.

  • For tardive dyskinesia, use tetrabenazine. Look out for depression and parkinsonism as side-effects. Use a high-dose anticholinergic for tardive dystonia.

  • For alternative treatments, use baclofen, benzodiazepines, beta-blockers or levetiracetam.

  • Try botulinum toxin injections for treatment of focal symptoms.

  • With refractory symptoms, consider re-starting DRB or deep brain stimulation

  • Short duration of DRB use and withdrawal of offending agent predict more favourable outcomes.


Tardive symptoms tend to come on insidiously and plateau. Once they reach a steady state, they do not usually worsen unless there is drug withdrawal or concomitant use of antidepressants or anticholinergic drugs. The commonest outcome, in half of all patients, is mild to moderate symptoms with some waxing and waning over several years without extended remission or significant progression. A quarter improves, and about one in seven fully remit. A quarter of patients worsen over time. Patients who have a higher risk of progression or persistence of symptoms are highlighted in box 3. It is reasonable to aim to withdraw DRB therapy in these patients who are particularly at risk of progression or persistence. Further, the likelihood of persisting TMD is much smaller in patients who have their DRB withdrawn. Also, there seems to be a favourable remission rate for patients who have been on DRB treatment for a shorter period of time. Patients with tardive dystonia tend to have a poor prognosis with improvement in only one in seven; remission occurs a mean of 2.5 years after stopping DRB treatment.3

Box 3 Risk factors for progression or persistence of symptoms of tardive dyskinesia

  • Older age

  • Non-white ethnicity

  • Women aged over 65 years

  • Frequent on–off manipulations

  • Long history of tardive movement disorder

  • Long history of exposure to dopamine receptor blocking drugs

  • Organic brain disease or affective disorder

Follow-up of patients with TMD is important to assess treatment efficacy and also to monitor complications of treatment. The abnormal involuntary movement scale30 is a validated and useful method for documenting serial scores to assess long term trends in a patient's symptoms. Alternatively, a good clinical description of the type of movement disorder, its distribution and any disability is useful for monitoring patients. Ideally, patients should be reviewed at least every 6 months. Some patients need more frequent follow-up if there are medication changes, complications or the need for repeat botulinum toxin injections. Joint management of the patient with a psychiatrist is helpful as decisions on modulating DRB treatment need expert psychiatric input.


This paper was reviewed by Dr Nick Fletcher, Liverpool, UK.


View Abstract


  • Funding TAS is funded by a doctoral research fellowship from the National Institute for Health Research. MJE is funded by a clinician scientist fellowship from the National Institute for Health Research.

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.