by Steven D. Targum, MD, and Richard S.E. Keefe, PhD

Dr. Targum is an executive-in-residence at Oxford BioScience Partners and on the faculty of the Department of Psychiatry at the Massachusetts General Hospital. Dr. Targum is chief medical officer at BrainCells Inc., and chief medical advisor to Prana Biotechnology Ltd. Dr. Targum is on the editorial board of Psychiatry 2008. Dr. Keefe is Professor of Psychiatry and Behavioral Sciences, Duke University Medical Center Durham, North Carolina. Dr. Keefe is also on the editorial board of Psychiatry 2008.

Psychiatry (Edgemont) 2008;5(12):55–59


Dr. Targum has stock or stock options in BrainCells Inc. and Prana Biotechnology Ltd. In the past year, Dr. Targum has been a consultant to United BioSource Corporation, Dynogen, Epix, DOV Pharmaceuticals, Sepracor, NuPathe, and Memory Pharmaceuticals. In the past year, Dr. Keefe has received grant/research support from Eli Lilly and Pfizer. He has received honoraria or served as a consultant or advisory board member for Abbott Pharmaceuticals, Acadia, BiolineRx, Bristol Myers Squibb, Cephalon, Cortex, Dainippon Sumitomo Pharma., Eli Lilly, Johnson & Johnson, Lundbeck, Memory Pharmaceuticals, Merck, Neurosearch, Orexigen, Pfizer, Sanofi/Aventis, Shering-Plough, Wyeth, and Xenoport. In addition, Dr. Keefe receives royalties from the Brief Assessment of Cognition in Schizophrenia (BACS) testing battery and the MATRICS Battery (BACS Symbol Coding).


Cognitive problems have long been considered a core component of schizophrenia but have only recently been considered as possible diagnostic features of the illness or as potential treatment targets. Although the cognitive impairment can range from mild to severe compared to healthy control subjects, almost all patients with schizophrenia demonstrate some cognitive deficits. These cognitive deficits are distinguished from the hallucinations and delusions, which generally characterize the positive symptoms of schizophrenic illness. In fact, the cognitive impairment tends to be more independent of the typical clinical symptoms of schizophrenia than similar problems in the affective disorders. In recent years, it has become apparent that there are some aspects of impaired neurocognition that are consistently found in schizophrenia and may have both prognostic and treatment implications for clinical practice. In this column, I interviewed the psychologist Richard S.E. Keefe, PhD, Professor in the Department of Psychiatry and Behavioral Sciences at Duke University Medical School in Durham, North Carolina. Dr. Keefe has focused his research career on the neurocognitive deficits of schizophrenia and offers his insights into the diagnostic, prognostic, and treatment implications for cognitive assessments in schizophrenia.

Most clinicians have always considered schizophrenia to be a form of “cognitive” disorder. Are there specific aspects of cognition that are more important for people with schizophrenia?

Dr. Keefe: Neurocognitive impairment is clinically relevant and profound for all patients with schizophrenia. As a rule, schizophrenic patients perform an average of 1.5 to 2 standard deviations below healthy controls on many neurocognitive tests. The severity of this impairment is greatest in the domains of memory, attention, working memory, problem solving, processing speed, and social cognition.

Do psychotic symptoms or antipsychotic treatments influence these findings?

Dr. Keefe: In fact, these neurocognitive deficits are present prior to the initiation of antipsychotic treatment and are not caused by psychotic symptoms in patients who are able to complete cognitive testing, which includes the overwhelming majority.

How do you differentiate specific aspects of cognitive dysfunction from hallucinations, delusions, and paranoia?

Dr. Keefe: They are very different. In fact, they are largely unrelated. In the CATIE study, we assessed cognition as well as symptoms at baseline before randomized treatment, and none of the cognitive domains were significantly correlated with hallucinations or delusions, despite a sample size exceeding 1,000 patients (n=1,331).

Are there neuroanatomical or neurophysiological correlates of the cognitive dysfunction you describe

Dr. Keefe: Certainly. Functional magnetic resonance imaging (fMRI), positron emission tomography (PET), electroencephalogram (EEG), and even structural MRI studies all show relations between neuroanatomical measures and cognitive deficits in schizophrenia. These relations are strongest in frontal regions, temporal cortex, and hippocampus.

Do the cognitive deficits identified in schizophrenia correlate with behavior or social functioning?

Dr. Keefe: Absolutely. This is the most consistent and relevant aspect of cognition in schizophrenia. Many of the various cognitive deficits in schizophrenia have been shown to be associated with functional outcomes, such as difficulty with community functioning, difficulty with instrumental and problem-solving skills, reduced success in psychosocial rehabilitation programs, and the inability to maintain successful employment. Several studies have found that cognitive deficits are better able to explain important functional outcomes, such as work performance and independent living, than the positive or negative symptoms we generally associate with schizophrenia. In the CATIE patient sample, cognition was found to be significantly related to quality of life although residual symptoms were also important.

Can cognitive testing assist in the diagnosis of schizophrenia?

Dr. Keefe: This is controversial, but I think so. In fact, we have recently proposed that the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V) and International Statistical Classification of Diseases and Related Health Problems, Eleventh Revision (ICD-11) include cognition as part of the diagnosis.[1] It is interesting that while most schizophrenia experts consider cognition to be a core component of the disorder, and it is mentioned several times in the DSM-IV-TR (Fourth Edition, Text Revision) description of schizophrenia, it has not been part of the diagnostic criteria. We have proposed the following criterion for consideration in the diagnostic criteria for DSM-V and ICD-11 schizophrenia: “A level of cognitive functioning suggesting a consistent severe impairment and/or a significant decline from premorbid levels considering the patient’s educational, familial, and socioeconomic background.” If these diagnostic systems focus less on specific criteria in favor of a completely dimensional approach, which seems likely at this point, the above recommendation could be easily revised to include cognitive impairment as one of the key dimensions.

Do you think that there are different forms of schizophrenia based upon the extent of cognitive dysfunction?

Dr. Keefe: No. I don’t see evidence of that. You can always cut up a pie in different ways, but it is still a pie. I think that almost everyone who has schizophrenia has some cognitive impairment. It is the core of the disorder.

Would the inclusion of cognitive assessments in the evaluation of schizophrenic patients facilitate treatment planning?

Dr. Keefe: I think so. In my experience, psychiatrists rarely consider cognitive function in their evaluation of patients with schizophrenia. Including cognitive impairment in the criteria for schizophrenia may increase psychiatrists’ attention toward a core component of the disorder that is the largest determinant of long-term functioning. Since cognitive impairment is also rarely considered as an important treatment target, its inclusion in the diagnosis may help to educate clinicians about the importance of cognition in their treatment options. Furthermore, representatives from the US Food and Drug Administration (FDA) have indicated that the recognition of cognitive impairment in the diagnostic nomenclature would be an important step in approving a drug for a cognitive improvement indication for patients with schizophrenia. A large number of pharmaceutical companies and government agencies are involved in intense work to develop compounds that may improve cognition with schizophrenia. If successful, these compounds have the potential to alter the treatment of schizophrenia. However, if clinicians are not trained to recognize cognitive deficits distinct from psychotic symptoms, a potential great benefit to patients would be missed. I think that the inclusion of cognitive impairment in the diagnostic criteria for schizophrenia may force educational systems to teach clinicians how to recognize cognitive impairment, measure improvement, and direct treatments accordingly.

Is the extent of cognitive deficit a prognostic as well as diagnostic marker?

Dr. Keefe: Cognitive impairment has clear prognostic value. Patients with the most extreme deficits will tend to have the greatest difficulty functioning in the long run, while patients who can perform in the “normal range” will tend to do better.

What specific tests can be used to evaluate cognitive dysfunction in schizophrenia?

Dr. Keefe: This is a tough issue. There are no specific tests that will identify schizophrenia, and it appears as though many, many aspects of cognition are affected. It really depends upon the purpose of the cognitive tests. We need to ask what the objective of the assessment is. A comprehensive identification of neuropsychological strengths and weaknesses across multiple domains can take several hours of testing, but a brief assessment of overall cognitive skill can probably be completed in about a half-hour.

What is MATRICS? Can it be used in clinical practice?

Dr. Keefe: In response to increased interest in developing new pharmacological compounds to treat cognitive impairment in schizophrenia, the National Institute of Mental Health (NIMH) funded the Measurement and Treatment Research to Improve Cognition (MATRICS) project.[2] This project engaged over 120 leaders from industry, government, and academia to devise methods for conducting clinical trials for cognitive-enhancing drugs in schizophrenia and also developed an intensively vetted battery of tests for use in clinical trials. This set of tests, the MATRICS Consensus Cognitive Battery (MCCB), was approved by the FDA and is currently in use for over a dozen multisite clinical trials of new compounds for schizophrenia. The MCCB can be used in clinical practice, but it is used mainly to determine a single overall composite outcome measure. While domain scores can be calculated, a comprehensive assessment of treatment effects on different cognitive domains, like executive function, may require additional testing.

What is executive function?

Dr. Keefe: The term measures of executive functioning refers to a set of processes involved in complex, goal-directed thought and behavior. Common executive abilities include judgment, problem solving, decision making, planning, and social conduct, and depend upon many of the other cognitive abilities, such as attention, perception, memory, and language.

How do you measure executive function?

Dr. Keefe: Executive functions may be among the most interesting cognitive tests, as they tend to be the aspects of cognition that are the most complex and most associated with what we think of as being human. There are various brain regions involved in executive functions, but the prefrontal cortex is the most frequently engaged. In the clinic, executive function is tested in a number of ways and may involve problem solving (e.g., Wisconsin Card Sorting Test or maze tests), suppression of a salient response with development of an alternate response (e.g., Stroop test), or response conflict or task switching (e.g., Trailmaking B). Given the challenges associated with assessing executive function, some researchers and clinicians have recommended delineating specific component processes instead of treating it as a global domain.

Can the identified cognitive deficits be treated?

Dr. Keefe: To date, treatment benefits appear to be minimal. Psychosocial, cognitive, and psychotropic treatments can have some impact on cognition, but almost all of the measurable cognitive impairment remains in most patients despite these treatments. Regarding psychotropics, we had some early hope that the newer antipsychotics improved cognition. But my interpretation of the recent studies is that the early optimism was largely due to the fact that the comparator in many of these studies was a large dose of one of the older drugs, usually haloperidol. Some of the benefit we were seeing with the newer antipsychotics was probably due to practice effects and placebo effects. There are currently a number of large efforts underway to find new medications that might be added as “cotreatments” to antipsychotics in order to improve cognition. While nothing is on the market yet, clinicians should prepare themselves to begin assessing cognition so that they will be able to determine if a patient is benefitting. Besides, antipsychotic dose, adjunctive treatment, and other factors can have an impact on cognition, so clinicians really should be aware of cognitive changes in their patients.

Can cognitive therapy abet the cognitive deficits in schizophrenia?

Dr. Keefe: Some recent work on cognitive remediation suggests that some of these behavioral strategies can be very effective, especially if patients are motivated to get better because they want to work or improve in some other way. It is a fascinating area of research, but the results are still preliminary and require long-term studies to evaluate sustained benefits.

Can neurocognitive testing be applied in clinical practice where time is a factor?

Dr. Keefe: This is a very important question. Although formal cognitive testing is sensitive to the cognitive impairments found in schizophrenia, the resources required to complete full neuropsychological evaluations are often prohibitive. However, resource requirements have not kept cognitive impairment out of the diagnostic criteria for other disorders of cognition, such as Alzheimer’s dementia and attention deficit hyperactivity disorder (ADHD), which do not require formal cognitive testing. Brief assessment may help reduce the burden of collecting cognitive performance data, but psychiatrists frequently are pressed to find enough time even to complete standard clinical evaluations. The methods for establishing the presence of cognitive impairment in schizophrenia for diagnostic purposes will need to be established. A few possible methods are brief assessment, such as the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) or the Brief Assessment of Cognition for Schizophrenia (BACS), which was developed in our lab at Duke; interview-based assessments of cognition, such as the Schizophrenia Cognition Rating Scale (SCoRS) or the Cognitive Assessment Interview (CAI); and measures of functional capacity, such as the UCSD Performance-based Skills Assessment. Each of these measures can be completed in 30 to 45 minutes or less per assessment. It is of course also important to assess a patient’s longitudinal course of cognitive functioning in the context of his or her personal background, which may require clinicians to gather substantial amounts of historical information. This may be fairly burdensome to some clinicians, but will help frame a patient’s current cognitive ability: Is it a lot worse than expected? Or is it about where it should be given the person’s family, personal, and educational history?

Dr. Keefe, what do you see on the horizon for neurocognitive studies in schizophrenia?

Dr. Keefe: Cognition will increasingly be involved as a crucial assessment in research and clinical practice. Researchers have known for a long time that cognition is at the core of the disorder, and it is likely that it will be represented in DSM-V and ICD-11. It has become very clear that good functioning depends upon cognitive ability and that treatments that can improve cognition are needed. Astute clinicians have begun to incorporate cognitive measures into their practices so that they can adjust doses of medications or refer patients for rehabilitation as necessary. If ongoing studies on new compounds to improve cognition are fruitful, clinicians will thus be in a good position to use these medications and assess their efficacy in the clinic.

1. Keefe RS, Fenton WS. How should DSM-V criteria for schizophrenia include cognitive impairment? Schizophr Bull. 2007;33(4):912–920.
2. Nuechterlein KH, Green MF, Kern RS, et al. The MATRICS consensus cognitive battery: Part 1. Test selection, reliability, and validity. Am J Psychiatry. 2008;165:203–213.