Tardive dyskinesia in the aged. Duration of treatment relationships.

Although tardive dyskinesia (TD) is recognized to result from neuroleptic drug exposure, data conflict about the importance of the quantity of that exposure in producing TD. The relationship between duration of neuroleptic treatment (one to 301 months) and TD was studied in 57 elderly psychiatric inpatients. Examinations for TD and parkinsonism were quantified on the Abnormal Involuntary Movement Scale (AIMS) and on a parkinsonism severity scale. Prevalence of presumed TD was 49% and of parkinsonism 51%. Prevalence of TD increased with longer treatment, but parkinsonism was independent of treatment duration. Linear multiple regression analysis showed that the AIMS score was correlated positively with treatment duration and negatively with parkinsonism. Logistic multiple regression analysis verified these relationships and was more successful at predicting TD. The length of neuroleptic treatment necessary to produce TD was calculated from the logistic model at 10.8 months (95% confidence interval, zero to 25.6 months). These analyses showed the greatest rise in risk of TD occurred within the first two years of drug therapy.