Abstract
Background: Cognitive decline in Huntington's disease (HD) remains an area of inconsistencies, especially far from disease onset. Objective: To clarify the course of cognition in premanifest HD. Methods: Twenty-six premanifest HD, 19 manifest HD, and 87 control subjects were followed for ten years, using an extensive cognitive battery. Differences in baseline levels and change over time, on four factors (motor speed, global cognition, executive functioning (EF), and memory) were examined, using multilevel regression analyses. Converters were additionally analysed as a separate group. Also, the influence of motor speed and predicted years to disease onset on the cognitive factors was studied. Results: Manifest HD subjects showed lower baseline scores compared to controls on the motor speed (p = 0.002), memory (p < 0.001) and EF (p < 0.001). They additionally deteriorated over the ten-year follow-up on memory (p = 0.01). Converters deteriorated on EF (p = 0.04). Further analyses of premanifest subjects ‘far from and close to predicted onset’ revealed lower baseline scores for the ‘close’ group on EF, as compared to controls (p = 0.001). They also deteriorated on memory (p = 0.01). Motor speed substantially mediated the results of the three cognitive factors; when added as covariate to the model several baseline and slope differences for the cognitive factors ceased to be significant. Conclusions: Memory and EF are highly sensitive for ascertaining deterioration in premanifest HD gene carriers, especially in subjects close to onset. Lack of deterioration for the subjects further away from onset suggests that both domains are largely unaffected in those far from onset. Also, motor influence on cognition is substantial and should be taken into account in cognitive HD research.