Baseline neurocognitive deficits in the CATIE schizophrenia trial
Keefe, R. S. E., Bilder, R. M., Harvey, P. D., Davis, S. M., Palmer, B. W., Gold, J. M., ... Lieberman, J. A. (2006). Baseline neurocognitive deficits in the CATIE schizophrenia trial. Neuropsychopharmacology, 31(9), 2033-2046. DOI: 10.1038/sj.npp.1301072
Neurocognition is moderately to severely impaired in patients with schizophrenia. However, the factor structure of the various neurocognitive deficits, the relationship with symptoms and other variables, and the minimum amount of testing required to determine an adequate composite score has not been determined in typical patients with schizophrenia. An 'all-comer' approach to cognition is needed, as provided by the baseline assessment of an unprecedented number of patients in the CATIE (Clinical Antipsychotic Trials of Intervention Effectiveness) schizophrenia trial. From academic sites and treatment providers representative of the community, 1493 patients with chronic schizophrenia were entered into the study, including those with medical comorbidity and substance abuse. Eleven neurocognitive tests were administered, resulting in 24 individual scores reduced to nine neurocognitive outcome measures, five domain scores and a composite score. Despite minimal screening procedures, 91.2% of patients provided meaningful neurocognitive data. Exploratory principal components analysis yielded one factor accounting for 45% of the test variance. Confirmatory factor analysis showed that a single-factor model comprised of five domain scores was the best fit. The correlations among the factors were medium to high, and scores on individual factors were very highly correlated with the single composite score. Neurocognitive deficits were modestly correlated with negative symptom severity (r=0.13-0.27), but correlations with positive symptom severity were near zero (r<0.08). Even in an 'all-comer' clinical trial, neurocognitive deficits can be assessed in the overwhelming majority of patients, and the severity of impairment is similar to meta-analytic estimates. Multiple analyses suggested that a broad cognitive deficit characterizes this sample. These deficits are modestly related to negative symptoms and essentially independent of positive symptom severity.