Hippocampal cell loss was induced by the four-vessel occlusion (4VO) method
, a model of global ischaemia. Global ischaemia for 15 min induced a select
ive damage to the CA1 subfield. Occlusion for 25 min produced a larger cell
loss within the CA1 and more variably the CA2, CA3, the striatum and corte
x. Ischaemic and sham control groups were assessed on two conditional discr
imination tasks (presenting the conditional cues either in the choice arms
or the start arm) and two spatial tasks (water maze and a simple spatial di
scrimination task). No significant effects were found on either of the spat
ial tasks (apart from the speed measure on the water maze). However, on the
conditional discrimination task with the cues in the choice arms, animals
with 25 min ischaemia learned the task significantly more slowly than the 1
5 min ischaemic and control groups. Results for the task with cues presente
d in the start arm differed according to choice of criterion for learning.
With a standard criterion of 90% accuracy on one session controls were sign
ificantly superior to both ischaemic groups. However, in this task rats wit
h 15 min occlusion showed the greatest impairment, and were significantly w
orse than both the controls and the 25 min occlusion group. These results s
uggest that hippocampal ischaemic damage disrupts the learning of condition
al discrimination but not simple spatial tasks. No clear relationship betwe
en the extent of hippocampal cell loss and behavioural impairment was evide
nt. These results highlight the critical importance of procedural factors i
n the assessment of cognitive impairment. (C) 2000 Elsevier Science B.V. Al
l rights reserved.