Amnesia and recognition memory: a re-analysis of psychometric data.

The present study compared the recognition memory deficit in different groups of amnesics using scores from a standard test. The data, taken from a literature search, came from 33 studies reporting the performance of amnesic subjects on the recognition memory test (RMT) [77]. A total of 112 amnesic subjects were grouped according to their pathology. In addition, the analysis included subjects with schizophrenia, amygdala damage, or frontal lobe damage. Of these three nonamnesic groups, only the frontal lobe subjects were impaired on both RMT subtests, while the schizophrenics showed a disproportionate impairment for the recognition of faces. The amygdala subjects were also poor at face recognition. Among the amnesic groups, those subjects likely to have multiple sites of pathology (e.g. Korsakoff amnesics, post-encephalitics) were found to be the most impaired on the RMT. In contrast, those amnesics with more focal, limbic lesions in the hippocampus, fornix, or mamillary body region showed much milder deficits on the RMT task, some performing at normal levels. Despite their apparent sparing of recognition, the overall severity of amnesia in those subjects with limbic lesions appeared comparable to that in the remaining amnesics. These findings indicate that deficits on both subtests of the RMT are a frequent but not inevitable component of anterograde amnesia. They also point to a distinct subgroup of amnesias associated with selective damage in the hippocampus or its diencephalic targets, in which there is a relative sparing of recognition under certain test conditions.