Stability of cognitive performance in older patients with schizophrenia: an 8-week test-retest study

It is important to understand the stability of cognitive performance in schizophrenia in order to understand 1) the potential improvements in performance associated with the beneficial effects of cognitive-enhancing medications and 2) the potential variation in performance on cognitive measures that are components of neuroimaging studies. There are several factors that could lead to spurious improvements in cognitive test scores, including practice effects and random test-retest errors.

Forty-five older patients with schizophrenia who were receiving conventional antipsychotic medications participated in the study. All subjects completed a comprehensive neuropsychological test battery at baseline and again at an 8-week follow-up evaluation.

Performance on all of the cognitive measures was stable over time, as evidenced by significant test-retest correlations, and practice effects were generally absent or minimal. Tests administered with alternate forms were no more temporally stable than tests administered twice with the same form. Very few individual cases had substantial variation at retest across the 22 test scores. From these data, "norms for evaluating change" were developed and described using the reliable change index method.

The data suggest that the findings of modest cognitive improvements seen in some prior studies of schizophrenia patients when treated with second-generation antipsychotic medications were probably not due to simple retesting artifacts. At the same time, because of the variance in some of these test-retest performances, relatively substantial changes in performance on the part of individual patients would be required to be clearly interpretable.