The Longitudinal Course of Cognitive Impairment in Schizophrenia: An Examination of Data From Premorbid Through Posttreatment Phases of Illness

Cognitive impairment is a core feature of schizophrenia that is present across the course of the illness. However, due to complexities of studying cognitive decline in patients prior to the onset of illness, the longitudinal course is not fully understood. The cognitive effects in patients with schizophrenia are robust, with a 1.5 to 2.5 standard deviation gap between patients and healthy controls on composite scores. People with schizophrenia manifest a prior history of cognitive impairment in the premorbid phases of the illness. Examination of school records suggests that children who will eventually develop schizophrenia begin school at a level of functioning that is a full grade behind their peers, with the gap increasing by the time they finish high school. Epidemiologic work suggests that there are both static cognitive impairments and developmental lags in these patients during childhood, well before the illness is fully manifest. Although there was initial promise of improved cognitive function with second-generation antipsychotic treatment, more recent studies have suggested no differences among antipsychotics, with the initial appearance of improvement very likely attributable to practice effects, inappropriate medication dosing, and poor study design. Two large, prominent studies evaluating first- and second-generation antipsychotics suggested that, although there was slight to modest improvement in cognitive function for all treatments, there were no differences among medications, regardless of the generation of the agents. In summary, patients who develop schizophrenia, on average, demonstrate cognitive impairment beginning as early as the first grade, with deterioration seen across school years. Further, these patients had substantial cognitive deficits after the initiation of psychosis. Finally, while antipsychotic treatment improves symptoms, antipsychotics have little impact on cognition, and there appear to be no differences in the degree of cognitive improvement between first- and second-generation agents.