Slow potential correlates of attention dysfunction in senile dementia: II.

The CNV was recorded from Fz and Cz placements in five seniles and five normal elderly subjects over two sessions under five ISI levels. In the first session 32 trials each of five foreperiods with 2, 4, 6, 8 and 10 sec ISIs were recorded. The second session involved presentation of 32 trials of five foreperiods with 1.4, 1.8, 2.2, 2.6 and 3.0 sec ISIs. The use of a repeated measures within subjects design was to avoid the confounding effect of between subject variance. The results suggested that difference between groups in preparatory strategies as measured by amplitude and latency of the CNV were exacerbated for longer ISIs. Though the seniles seemed able to 'keep time' with small changes in ISI the effort invested in preparation of response was far lower than that of the normals. CNV amplitude and latency and RT measures showed that for both groups the optimal preparatory interval was around 2 sec, the seniles requiring a slightly shorter interval than the normal to synchronise preparation and response. Cortical topography differences suggested that the seniles' preparatory strategies were predominantly influenced by the extent of their initial frontal orienting response and that the seniles showed less differentiation between frontal and central recordings. The seniles' dependence on frontal orienting suggests that their performance does not benefit from preparatory set and results in a more stimulus bound divisive set. Implications of these qualitative differences in attentional strategy for behavioural management of senile patients are discussed.