Background: Classical conditioning provides a means of addressing mechanisms of learning and can therefore help understand the pathophysiology of memory alteration in schizophrenia.
Methods: Single cue delay and trace eyeblink conditioning were used in patients with schizophrenia and matched normal control subjects to explore, respectively, cerebellar and hippocampal integrity during learning. We measured percent of conditioned (CRs) and unconditioned responses (URs), their amplitude, and onset and peak latencies. We also accounted for spontaneous blink rates and stimulus-induced responses before learning.
Results: During delay conditioning, patients showed CRs with longer onset and peak latencies and improved efficiency compared to normal volunteers without there being differences between patients and normal control subjects in the percentage of CRs. During trace conditioning, neither group showed an increase in CRs as a function of conditioned stimulus-unconditioned stimulus pairings, in part because the level of spontaneous blink rates exceeded the level of CRs; however, patients with schizophrenia showed increased responding 150-400 msec after the conditioned stimulus and in the last 100-150 msec before the unconditioned stimulus, whereas normal control subjects showed only the latter type of responses. The former type of response was more frequent in patients with schizophrenia even before either trace or delay conditioning.
Conclusions: These results suggest integrity of cerebellar mechanisms underlying conditioning, although the altered timing of CRs in patients may indicate differences in the modulation of such responses. Both the greater CR onset latency during delay and the presence of early nonadaptive responses during trace are compatible with the pattern of responding seen in animals with hippocampal damage.