Tardive dyskinesia and antipsychotics: a 5-year longitudinal study of frequency, correlates and course.

The present study comprised a naturalistic, multicentre, 5-year study of course and correlates of tardive dyskinesia (TD). One hundred and sixty-six patients treated with risperidone were included during 1995/96 and followed once a year for 5 years. Mean age at inclusion was 38 years, and mean illness duration was 12 years. Tardive dyskinesia was assessed by the Abnormal Involuntary Movement Scale, and each patient's cognitive function was tested with a comprehensive computerised test battery. At study entry, 14% had TD according to a criterion index. Fifty percent were aware of it, but few reported distress. Age and sex did not correlate with TD, but schizophrenia and bipolar diagnoses did. The presence and intensity of TD correlated with all Positive and Negative Syndrome Scale for Schizophrenia symptom dimensions except the affective factor, but not with type of medication or chlorpromazine-equivalent levels. Tardive dyskinesia patients were cognitively impaired in tests reflecting mental speed, but not in other cognitive modalities. Over the 453 patient years of exposure, five patients developed TD and 14 became free of it. Our findings support the view that TD: (i) is a dynamic phenomenon; (ii) is only partly drug-induced; (iii) has a mild course during treatment with modern neuroleptics; and (iv) appears to have some correlation with mental slowness.