Schizophrenia can bring terrible hallucinations as well as impair a variety of cognitive functions. A large number of antipsychotic medications purport to combat both psychotic symptoms and cognitive impairments, and they are widely considered successful. However, this "success" is suspicious, because in many trials with multiple medications, all of the medications achieved the same level of improvement for cognition. How could there be so little difference between medications? An eye-opening new study in the October issue of the Archives of General Psychiatry suggests that the effects are the same because they are non-existent. In other words, the study offers evidence that all patients, even those receiving no treatment, would show "improved" cognition after taking the same cognitive test multiple times.
Researchers are warned of the "Practice Effect" during training. Subjects score better with repetition on many tests, including standard IQ tests, because they either become more familiar with the style of the test or with specific tasks. Even though researchers have been warned to look out for the practice effect, it is easy to miss it when tests are administered weeks apart and patients give no indication that they are becoming too familiar with the tasks. Dr. Terry Goldberg did not forget the warning, so he set up an experiment to determine what, if any, effect could be attributed to practice.
Dr. Goldberg subjected three groups to a large battery of tests that were repeated with a frequency used by many previous researchers. The three groups were: schizophrenics given olanzapine, schizophrenics given risperidone, and healthy controls. There were no significant differences between the olanzapine and risperidone groups, but this mysterious similarity has already been observed. The astonishing finding was that the healthy group, with no medication and no impairments to start with, showed significant improvements on almost all of the cognitive tests.
The control groups improvement was .33, whereas the treated groups showed only a .36 improvement. The .36 seems impressive when viewed in isolation, so we cannot strongly fault previous researchers for being enthusiastic about this ostensible benefit of medication, but the number appears negligible when viewed alongside the improvement that automatically occurs after repeated testing.
Subjects were given 14 different tests to assure that results are not the result of the peculiarities of a few tests: Verbal fluency, MMSE, Trail Making, WAIS-R digit symbol, WCST, CPT-IP, CVLT, WMS-R logical memory, WMS-R visual reproduction, Judgement line orienatation, DMS, WMS-R digit span, Finger rapping, and Grooved pegboard. Only five tests did not show improvement: fluency, digit span, CPT-IP, DMS, and the motor. Of the nine remaining tests, treated subjects did not improve more than healthy controls on the majority (5).
The results of this experiment are consistent with previous observations of the practice effect, leading the authors to conclude that there is a high probability that antipsychotics do little to improve cognition. This casts doubt on an extensive body of research, and necessitates more careful exploration before patients are assured of the benefits of antipsychotics. While antipsychotics have other benefits besides improved cognition, such as a lessening of extreme symptoms like hallucinations, they also have severe side-effects. A large portion of patients develop distressing motor problems after taking antipsychotics, raising the bar of evidence needed to justify a prescription. Although this study has identified an issue in testing, deciding how to solve it is much more difficult. The authors of this study do though offer some help. They noticed that the tests showing the least practice effect were those with "simple directions, large numbers of trials, and a restricted set of stimuli." If researchers design tests along these specifications, and remain aware of the problem that can creep into other tests, then they should be able to accurately gauge the effectiveness of antipsychotics on cognitive function.
|