Journal of the International Neuropsychohgical Society (1996), 2, 486-493. Copyright © 1996 INS. Published by Cambridge University Press. Printed in the USA. Proactive inhibition and semantic organization Relationship with verbal memory in patients with schizophrenia DAVID A. KAREKEN, 1 PAUL J. MOBERG, 1 AND RUBEN C. GUR 1 'Department of Psychiatry, University of Pennsylvania, Philadelphia, Pennsylvania 19104 (RECEIVED July 10, 1995; REVISED January 12, 1996; ACCEPTED January 26, 1996) Abstract Compared to other cognitive functions in schizophrenia, evidence suggests that verbal memory is particularly impaired. This study used the California Verbal Learning Test (CVLT) to examine proactive inhibition (PI) and semantic processing in verbal memory in 29 patients with schizophrenia and 29 healthy controls. Patients showed significantly less PI, but also did not organize (cluster) their recall according to semantic category. Controls and patients demonstrated small retroactive inhibition (RI) effects regardless of semantic content. Although both groups made similar types and numbers of free recall intrusion errors, patients committed more phonemic and nonshared recognition errors. Results suggest that reduced semantic processing prevented build of PI, and contributes to defective memory in schizophrenia. The anatomic-physiologic abnormalities that underlie these findings may be particularly pronounced in prefrontal and temporal-parietal cortical areas. (JINS, 1996, 2, 486-493.) Keywords: Proactive inhibition, Verbal memory, Schizophrenia INTRODUCTION Research suggests that patients with schizophrenia have greater impairments in memory than in other cognitive func- tions, (e.g., Saykin et a!., 1991, 1994; Karekenetal., 1995). Further, attention and abstraction do not appear to account for differences between patients and healthy controls in mem- ory test performance (Saykin et al., 1991, 1994; Gold etal., 1992). The character of this memory deficit appears more closely related to inadequate initial encoding, rather than to the time dependent loss of trace seen in conditions such as AlzheimerDisease(Heatonetal., 1994; Paulsen etal., 1995). Some studies indicate that schizophrenics' verbal mem- ory is more impaired than their visual memory (Saykin et al., 1994; Karekcn et al., 1995). Failure to adequately en- code semantic information may be one reason. Koh et al. (1973) observed that when patients with schizophrenia re- called a list of words, the order in which they recited the words was unrelated to word meaning. This contrasted with healthy controls who "clustered" semantically similar words together. More recently, Paulsen et al. (1995) administered Reprint requests lo: Ruben C. Gur, Ncuropsychiatry Section (10 Gates), University of Pennsylvania, Philadelphia, PA 19104. a list learning test (CVLT, Delis et al., 1987) and also found that schizophrenic patients had less categorical organiza- tion (clustering) in their recall than controls. Greater semantic structure appears to facilitate memory in schizophrenics. Koh et al. (1976) reported that inducing patients to rate the pleasantness of words, forcing greater use of word meaning, rendered the memory performance of controls and patients equal. Similarly, Koh et al. (1980) found that patients' word recall was made equal to that of controls when the patients were required to unscramble sentences. However, latent semantic structure is not sufficient to facil- itate word recall in patients with schizophrenia (i.e., an un- organized list of words that can be grouped into different categories). Rather, semantic characteristics must be highly salient (i.e., preclustered) in order to facilitate recall signif- icantly (Gold ct al., 1992). Insofar as schizophrenics do im- pose semantic organization on material that they encode and recall, the degree of this organization predicts their mem- ory performance, as it does in healthy controls (Harvey etal., 1986). The phenomenon of proactive inhibition (PI) provides an- other measure of how semantic organization affects verbal memory. PI is operationally defined by a declining slope of word recall with the presentation of successive lists of words 486