Pharmacoepidemiologic studies have demonstrated that the prevalence of TD is approximately 30% in patients receiving first-generation antipsychotic agents and 20% in patients receiving second-generation agents who may also have previously been on first-generation agents. When second-generation antipsychotics became available, the hope was that we had done away with drug-induced parkinsonism and, possibly, TD. We believed that we had potentially solved the TD problem with these agents, but that was proven to be untrue.

The rate of TD among patients who have only been on second-generation antipsychotics is not zero and has been estimated to be approximately 7%. Moreover, this lower rate may be misleading. If we consider those who have only been exposed to second-generation antipsychotics and have not used first-generation agents, many are relatively young and have been on treatment for relatively briefer periods of time compared with previous cohorts on first-generation medications. We do not yet have second-generation–only exposed patients who have been on treatment for 30 years or longer and are now advanced in age. Therefore, we do not yet really know the cumulative exposure–associated risk of second-generation agents on TD. In my view, 7% is likely an underestimate of what we will eventually see.

Cumulative exposure to dopamine receptor–blocking agents and advanced age are among the strongest risk factors for TD. As one gets older, the risk of developing TD is higher. If a patient begins treatment with an antipsychotic at 60 years of age, their risk of developing TD would be much higher than that of a 20-year-old person who is initiating the same treatment. It will be interesting to know the true prevalence of TD in people who have only been on second-generation agents and have been on them for 30 years or longer.

Anyone on a dopamine receptor–blocking agent must be regularly screened for TD. Guidelines have suggested that the less-frequent use of formal, structured assessments such as the Abnormal Involuntary Movement Scale might, perhaps, be appropriate for those patients on second-generation agents who have no other risk factors for TD, such as advanced age or a history of movement disorders. However, these assessments must still be done.