This interview took place on April 16, 2007, and was conducted by Norman Sussman, MD.


This interview is also available as an audio PsychCastTM at

Disclosure: Dr. Lieberman is a consultant to Eli Lilly and Pfizer; is on the advisory boards of AstraZeneca, Eli Lilly, GlaxoSmithKline, Lundbeck, Organon, and Pfizer; has a patent from Repligen; and receives research support from Acadia, Bristol-Myers Squibb, GlaxoSmithKline, Janssen, Merck, Organon, and Pfizer.



Dr. Lieberman is the Lawrence E. Kolb Chairman of Psychiatry at the Columbia University College of Physicians and Surgeons and director of the New York State Psychiatric Institute in New York City. He also holds the Lieber Chair and directs the Lieber Center for Schizophrenia Research in the Department of Psychiatry at Columbia, and serves as the principal investigator of the Clinical Antipsychotic Trials of Intervention Effectiveness Research Program sponsored by the National Institute of Mental Health. Dr. Lieberman’s research focuses on the neurobiology, pharmacology, and treatment of schizophrenia and related psychotic disorders.


What is tardive dyskinesia?

Tardive dyskinesia is a neurologic side-effect syndrome associated with chronic antipsychotic administration affecting neuromuscular function. It consists of an involuntary movement disorder wherein the chronic effects of treatment induce a change in neuromuscular function in certain muscle groups, which in turn produce involuntary movements. Such movements usually occur in the form of choreoathetoid-like movements (ie, writhing or flicking movements), occasionally in the form of dystonic movements (ie, sustained muscle contractions), and more occasionally in the form of tic-like movements (ie, rapid contractions of small muscle groups). These involuntary movements may occur on a persistent basis and may even be irreversible, meaning that they do not stop even if the medication that has induced them is discontinued.

The movements generally affect small muscle groups, usually in the extremities such as fingers, toes, and hands, as well as in the oral facial region on the tongue or lips. Rarely, but occasionally, they can affect larger muscle groups such as the neck and torso. When this occurs, it is usually in the context of these dystonic-like manifestations. Very occasionally, it can also involve smooth muscle groups such as the diaphragm, which can affect respiration and speech. This is fairly rare, but is one of the more severe variants of the condition.

Tardive dyskinesia is usually mild and not disabling or life-threatening, albeit mildly or moderately disfiguring and appearing unusual to others. However, in its extreme form it can be distressing and in some very rare cases potentially life threatening.


What are the early indicators of tardive dyskinesia?

An early indicator of vulnerability to tardive dyskinesia is when a patient is prone to developing acute extrapyramidal symptoms (EPS) with antipsychotic administration. For example, a predictor of increased vulnerability to tardive dyskinesia would be if rigidity, tremor, and bradykinesia develop following initial administration of an antipsychotic. However, the earliest manifestations of tardive dyskinesia itself are the involuntary movements, which are best elicited by examining the patient and conducting a brief neurologic exam. Such an exam consists of observing the patient’s body parts, extremities, tongue, and mouth under resting conditions as well as in response to a motor or cognitive task which acts as a stressor. A clinician can conduct a task activation wherein patients are asked to perform a cognitive test like counting backwards from 100 by sevens, or a manual stressor or task like touching his or her thumb and fingers together in sequential fashion. A clinician can then see if those activities activate other regions of the body to produce involuntary movements. In addition, the clinician should observe the patient ambulate and check muscle flexibility and resistance or rigidity.


Should primary care physicians (PCPs) conduct an Abnormal Involuntary Movement Scale (AIMS) exam on all patients taking antipsychotics?

PCPs should be aware of and monitor patients for these side effects. Whether they should conduct a full AIMS assessment and score it on a rating scale is a matter of individual preference. There is no standard of care that requires it; however, monitoring patients for side effects is certainly required and is part of the current standard of care. When prescribing a first-generation antipsychotic (FGA) or a second-generation antipsychotic (SGA), PCPs should indicate that there is the potential for developing neurologic side effects such as EPS or tardive dyskinesia. At follow-up visits, patients should be examined for such side effects, particularly during the drug initiation and dose escalation period. PCPs should examine patients for EPS and at some frequency for tardive dyskinesia. EPS should be evaluated at every clinic visit.

However, once patients have been on a medication at a stable dose for a period of time and have not exhibited symptoms, the frequency of examining for EPS can be reduced. Tardive dyskinesia—being a chronic neurologic side effect—usually does not occur until the patient has been treated for months if not years; thus, the frequency of examining for it can be less than for acute EPS.


Are SGAs associated with lower risk or incidence of tardive dyskinesia compared to low doses of FGAs?

SGAs are associated with a lower risk for EPS and tardive dyskinesia than FGAs. The incidence rates are related to the potency of the antipsychotic (meaning its affinity for the dopamine [D]2 receptor) and the dose and duration at which the drug is administered. The precise estimates of the incidence of SGAs and the individual FGAs are not well determined, so there are only rough estimates of tardive dyskinesia risk by class of antipsychotics. The difficulty is that not all FGAs or SGAs are similar or comparable to one another. There is variation within those groups which is associated with differential risk for tardive dyskinesia. Thus, current estimates apply for classes of drugs but not necessarily for individual drugs in a precise way.

Longitudinal incidence data with FGAs as a class suggest that the annual incidence of tardive dyskinesia risk is approximately 4% per year of exposure.1-3 Thus, of 100 patients started on haloperidol or chlorpromazine, four would have some manifestations of tardive dyskinesia at the end of 1 year. That rate increases after the second or third year. It is unknown for how long this rate remains linear and if it would extend to 100%. It seems not to, since not all patients are vulnerable to tardive dyskinesia, even given substantial exposure.

The annual incidence rate of SGAs is certainly less, but it is not zero nor is it equal for all SGAs. Although there are no precise estimates, there is evidence on which to base probabilistic estimates of tardive dyskinesia liability for the SGAs.

Certain inferences between FGAs and SGAs can also be made based on the pharmacology of the different drugs and their capacity to induce EPS. That is, the level of affinity that an antipsychotic has for the D2 receptor, where it acts as a full antagonist, is associated with its risk for both EPS and tardive dyskinesia.

This seems to be the case for the FGAs; hence, the higher-potency drugs like haloperidol and fluphenazine are the more tardive dyskinesia-inducing drugs. If such thinking is applied to the SGAs, one might think that risperidone may have a somewhat higher risk than drugs that have a very low D2 affinity such as clozapine or quetiapine. Based on this inferential and speculative reasoning, one can deduce that if FGAs have an approximately 4% annual incidence in patients treated with a variety of both high- and low-potency FGAs, then the SGAs have a lower rate. If the rate is somewhere between 0% and 4%, the drugs that are most likely to produce rates of ≤1% are probably clozapine and quetiapine. The drugs that produce the rates that are >1%, but not as high as 4% are probably drugs like risperidone and possibly ziprasidone, which have a higher D2 affinity. Aripiprazole has very high affinity for the D2 receptor but acts as a partial agonist so this reasoning may not apply.


Do any medications treat tardive dyskinesia?

Unfortunately, there has not been any real success in developing therapies for tardive dyskinesia. A range of remedies have been tried. One of the most promising and practical of these has been vitamin E. However, in a multicenter Veteran’s Administration study the results were negative regarding the use of vitamin E for treatment of tardive dyskinesia.4

In addition, all the cholinergic agents that have been tried have only been inconsistently effective. The best treatment for tardive dyskinesia is simply reducing dosage or reducing the D2 antagonist load on the D2 receptor by switching patients to a very low-potency D2 antagonist or withdrawing them from antipsychotics. However, withdrawal is not practical in the case of a patient with schizophrenia. The one exception to that has been the use of clozapine as a treatment for the severe variant of tardive dyskinesia called tardive dystonia. Patients with tardive dystonia who have been treated with or switched to clozapine have had in some cases a dramatic amelioration of their symptoms. However, beyond this there has been no successful treatment developed for tardive dyskinesia.


Are there other factors aside from sensitivity to EPS that may predispose a patient to developing tardive dyskinesia?

Age is a big factor. There is abundant evidence that the older the patient, the more vulnerable he or she is to develop tardive dyskinesia. Patients ≥60 years of age have a four- to ten-fold greater risk of developing tardive dyskinesia than young or middle-aged adults. Gender is also a factor, as females tend to be more prone postmenopausally. There is also some evidence that suggests diagnosis is a risk factor. The risk factor of bipolar disorder is particularly worrisome given the increasing use of antipsychotics as treatment. Patients with bipolar disorder seem to be more susceptible to multiple side effects of antipsychotics, not just EPS. They seem to be potentially more vulnerable to weight gain and metabolic effects of SGAs, as well.


Do any of the newer drugs carry a lowered risk of acute EPS and tardive dyskinesia?

Paliperidone, which has been put on the market recently, does not seem to carry any particular advantage with respect to EPS or tardive dyskinesia. It is the 9-hydroxy metabolite of risperidone and is a similar drug in that regard. Bifeprunox is a partial D2 agonist which offers some potential incremental benefit in that it is another D2 agonist with a higher intrinsic activity than aripiprazole. Aripiprazole offers high D2 affinity but with partial agonism, which mitigates its tardive dyskinesia-inducing effect. However, experience with aripiprazole is still too limited to know what its real tardive dyskinesia liability is. Bifeprunox, because it seems to have a higher  intrinsic agonist activity at the D2 receptor, would be expected to have a somewhat lower ability to cause tardive dyskinesia than does aripiprazole. Again, this is only a hypothetical estimate of the incidence rate with aripiprazole, which is the prototype partial D2 agonist.


Is there a trade-off between using FGAs or SGAs?

SGAs certainly have reduced tardive dyskinesia liability, albeit by varying degrees relative to particularly the high-potency FGAs. However, with the emergence of concern about weight gain and metabolic effects—such as hyperglycemia, hyperlipidemia, and the effects on cardiovascular disease risk—the question of trade-off must be considered. Tardive dyskinesia is clearly a very devastating and serious condition because it reflects an effect on the nervous system that is potentially irreversible. The condition, however, is by and large mildy disfiguring and has, only in rare instances, had any appreciable morbidity or been life threatening. On the other hand, weight and metabolic effects influence cardiovascular morbidity and mortality. Thus, the variable risk of these drug classes poses a complicated question. It is like being stuck between a rock and a hard place of adverse effects. One is apparent immediately or fairly soon in terms of EPS, and even in the case of tardive dyskinesia is certainly worrisome and serious but is not necessarily life threatening. The other (ie, obesity, hyperglycemia, hyperlipidemia) is slower to develop, more benign in its initial manifestations, and does not cause any significant subjective distress but is potentially associated with the development of the life-shortening conditions of cardiovascular disease and diabetes. Clinicians in the field are wrestling with determining the risk-benefit analysis of these different classes of agents.


Do the newer studies of FGAs use lower doses than were used in previous studies?

Newer studies with FGAs are using lower doses, which is probably producing a more accurate and valid comparison of the two drug classes. In some of these studies, the advantages of the SGAs to FGAs are appearing less dramatic and consistently, while in other studies the FGAs appear comparable or slightly advantageous.5,6

An interesting set of first-episode studies have been conducted Europe. These included a multicenter trial in first-episode patients where risperidone and haloperidol were compared using very low 2-mg doses of both.7 No real substantial differences on therapeutic or adverse effects were found.

A recent report of the European First Episode Study8 is comparable to a first-episode Clinical Antipsychotic Trials of Intervention Effectiveness study. Multiple SGAs have been compared to low-dose haloperidol in first-episode patients. This will be an interesting study in determining the comparative effectiveness of the different SGAs as compared to a representative FGA. In some ways, the first-episode patient sample is a purer assay of drug activity because these patients are in an earlier stage of the illness and had less prior treatment exposure. PP



1. Glazer WM. Expected incidence of tardive dyskinesia associated with atypical antipsychotics. J Clin Psychiatry. 2000;61(suppl 4):21-26.
2. Jeste DV. Tardive dyskinesia rates with atypical antipsychotics in older adults. J Clin Psychiatry. 2004;65(suppl 9):21-24.
3. Kane JM. Tardive dyskinesia rates with atypical antipsychotics in adults: prevalence and incidence. J Clin Psychiatry. 2004;65(suppl 9):16-20.
4. Adler LA. Vitamin E treatment for tardive dyskinesia. Veterans Affairs Cooperative Study #394 Study Group. Arch Gen Psychiatry. 1999;56(9):836-841.
5. Lieberman JA, Stroup S, McEvoy J, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia: primary efficacy and safety outcomes of the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia trial. N Engl J Med. 2005;353 (12):1209-1223.
6. Jones PB, Barnes TR, Davies L, et al. Randomized controlled trial of the effect on Quality of Life of second- vs first-generation antipsychotic drugs in schizophrenia: Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS 1). Arch Gen Psychiatry. 2006;63(10):1079-1087.
7. Gaebel W, Reisbeck M, Wölwer W, et al. Maintenance treatment with risperidone or low-dose haloperidol in first-episode schizophrenia. One-year results of a randomized controlled trial within the German Research Network on schizophrenia. J Clin Psychiatry. In press.
8. Fleischhacker WW, Kahn R. Paper presented at: Winter Workshop of the American College of Neuropsychopharmacology. February 2-8, 2006; Davos, Switzerland.