Reflections on the inclusion of cognitive impairment in the diagnostic criteria for schizophrenia

Abstract

Richard Keefe outlines the reasons for including cognitive impairment as one of the criteria for the diagnosis of schizophrenia in future diagnostic manuals (DSM-V and ICD-11). First, he convincingly argues that cognitive impairment has a major impact on the lives of almost all individuals who meet DSM-IV and ICD-10 criteria for schizophrenia. He reminds the reader that even patients who obtain normal or almost normal psychometric scores might still be cognitively impaired, compared to their potential as predicted by their parents’ intelligence and their socioeconomic context. Than he presents the advantages and stumbling blocks associated with adding cognitive impairment to future diagnostic criteria, and suggests solutions. The first advantage Keefe identifies is help in defining a point of rarity between schizophrenia and other diagnostic entities, to reduce diagnostic overlap. The second advantage is the encouragement of the clinical, academic, regulatory, and pharmaceutical communities to develop treatments targeting cognitive impairment. The main impediment he identifies is lack of assessment instruments which on the one hand are brief and simple and on the other are sufficiently detailed and comprehensive to capture developmental and current information in a valid and reliable manner. He voices the concern that non-psychologist clinicians, who are not sensitized and trained in the assessment of cognition, might be unwilling to carve out the necessary time from the already-limited patient-clinician encounter. Since, even for the most comprehensive scales, much of the variance is accounted for by only a few items, the solution he advocates is the use of brief scales which can be quickly learned and applied by clinicians. Keefe is correct to point out that cognitive impairment in schizophrenia differs from cognitive impairment in mood disorders, which is the main overlapping diagnostic group, in terms of prevalence, severity and course. However, the differences are neither large nor consistent, and it is unclear by how much the inclusion of cognitive impairment as a criterion would reduce the diagnostic overlap. Furthermore, cognitive impairment is present in almost all psychiatric diagnostic categories, some of which overlap with schizophrenia in manifestations other than cognitive impairment (e.g., severe personality disorders, drug abuse, obsessive-compulsive disorder). Moreover, there is considerable overlap between schizophrenia with severe cognitive impairment and mild mental retardation with unusual ideation, and the inclusion of cognitive impairment as a criterion for schizophrenia might only enhance the overlap. There is little doubt that the inclusion of cognitive impairment as a diagnostic criterion would increase awareness and focus on it the attention of all stakeholders. Whether this would hasten the development of treatments targeting cognitive impairment is less certain. Increased awareness might help divert budgets and other resources to education and rehabilitation programs focused on cognitive enchantment. Since education and rehabilitation activities have a incremental and cumulative effect, more budgets and resources might benefit cognition. However, the development of a pharmacological intervention targeting cognitive impairment is an endeavor waiting for a conceptual breakthrough, rather than the incremental assigning of more funds which would result from increased awareness. Cognitive impairment in schizophrenia may have more in common with cognitive impairment occurring in most other mental disorders than with schizophrenic psychosis per se 1,2. Cognitive attributes and predisposition to psychosis are partially inherited traits in schizophrenia patients. However, whether these traits co-segregate and follow common pathophysiologic pathways is far from certain. The lack of correlation between the severity of cognitive impairment and psychosis, and the fact that cognitive impairment appears before and persists after psychosis, are consistent with the idea that they are independent. Conceptually, therefore, it would take the same breakthrough to treat cognitive impairment in schizophrenia as it would take to improve cognitive functioning in any individuals, mentally ill or not, whose performance is 1.5 to 2.5 SDs below norms or expected performance. This major scientific endeavor could hardly benefit from the inclusion of cognitive impairment as a criterion for schizophrenia. I would not be too concerned with the willingness of those who treat schizophrenia (psychiatrists, nurses, social workers, and others) to learn and devote time to the assessment of cognition. The relentless efforts of researchers like Keefe and others over the past 15 years have already sensitized clinicians to the issue of cognitive impairment in schizophrenia patients. Once they are convinced of the benefit of diagnosing it and the potential of treating it, they are likely to adjust their priorities and fit cognitive assessment into the clinical encounter. In spite of the above reservations, I hope that Richard Keefe’s suggestion to include cognitive impairment as a criterion for schizophrenia is accepted. A diagnostic classification that leaves out an aspect which is so prevalent among the affected individuals and so central to their daily life would be deficient.