Impaired reproduction of second but not millisecond time intervals in Parkinson's disease

The basal ganglia have been associated with temporal processing in ranges of milliseconds and seconds. However, results from PD patient studies are elusive. Time perception in these patients has been tested with different approaches including repetitive movement tasks (i.e. finger tapping) and cognitive tasks (i.e. time reproduction), and both abnormal and normal performances have been reported for different time intervals. Furthermore, when PD patients were required to learn two target durations in the same session when they were off medication, they overestimated the short duration and underestimated the long duration in the seconds range. This pattern of temporal accuracy was described as a "migration effect" and was interpreted as a dysfunctional representation of memory for time (Malapani, C., Rakitin, B. C., Levy, R., Meek, W. H., Deweer, B., Dubois, B., et al. (1998). Coupled temporal memories in Parkinson's disease: A dopamine-related dysfunction. Journal of Cognitive Neuroscience, 10, 316-331). Here, we controlled the emergence of similar behaviour also during millisecond time processing in PD patients. A time reproduction task was employed in which subjects were required to estimate intervals in millisecond (500 ms) and few second (2000 ms) ranges. In the first experiment, these intervals were tested in the same session to verify whether the migration effect was present also between time intervals in different millisecond and few second ranges. In a second experiment, they were not intermingled but were tested in two separate sessions to verify whether abnormalities depended on a selective perceptual deficit of the time intervals tested (i.e. millisecond or second ranges). All experiments were performed in both off and on therapy conditions. Our results demonstrated that PD patients showed no deficits in time estimation for time intervals in either the millisecond or few second range when the different time intervals were tested in separate sessions. This negative finding was obtained in both on and off conditions. However, when the different ranges were tested in the same session, we found that PD patients were impaired selectively for time intervals in the seconds range. Our data seem to indicate that time processing in PD patients for time intervals spanning up to 2s is unimpaired and that abnormalities in such temporal scale may emerge only when patients have to deal with different durations, when timing involves further cognitive processes such as memory and attention. (C) 2007 Elsevier Ltd. All rights reserved.