Abstract

Abnormalities in saccadic function are known features of Huntington's disease (HD) in presymptomatic and symptomatic patients. The study sought to evaluate implicit prediction of rhythmical stimuli by saccades in manifest HD. In evaluating the process, 22 symptomatic HD patients and 22 matched controls were recruited and had their saccadic and cognitive function assessed. Saccades were measured using a high-speed video eye tracking system (iViewXTM, SMI, Berlin) and cognition was assessed with a series of domain-specific neuropsychological tests. The saccadic tasks began with the stimulus appearing at non-predictable locations and times for 17 trials, followed by 40 trials of the stimulus appearing at fixed positions of ±10° from the centre on the horizontal axis at different inter-stimulus intervals (ISI:750 ms, 1000 ms, 1400 ms, 2050 ms). The mean difference in latency (ms) of the last 10 trials and the trials with non predictable stimuli for each task was used as measures of prediction. Multi-level regression modelling showed significant group differences in the ability to predict in ISI: 750 ms task [HD:−15 ms (σ=132), Controls: −213 ms (114)] but such differences progressively diminished with increasing ISI; ISI: 2050 ms task [HD:−47 ms (114), Controls: −51 ms (114)]. Overall results showed that shorter ISI encouraged prediction in controls but not in HD group. A weak but non significant correlation was observed between poorer global cognitive scores and the failure to predict in HD group only. Rhythmic movement is thought to be regulated by basal ganglia and the degeneration of basal ganglia in HD may result in the impairment of implicit prediction in saccadic function.