Learning and Brain Function in Schizophrenia

To the Editor: The positron emission tomography (PET) imaging study of deficit and nondeficit patients with schizophrenia and healthy volunteers conducted by Adrienne C. Lahti, M.D., and colleagues (1) revealed statistically significant mean differences among these groups in certain brain areas. However, two important points were not discussed by the authors, and these issues challenge the view that the authors’ work supports the hypotheses regarding putative brain areas relevant to the distinction between deficit and nondeficit schizophrenia.

1. Given their use of categorical differentiation of the three groups by diagnostic and negative symptom assessments and the statistically different mean differences among these groups, how do the authors explain the striking and graphic overlapping of the distributions of both pretask and posttask differences and the direction of changes, as displayed so clearly in their Figure 1?

2. The authors found no differences in the learning of the tone-discrimination task among the three conditions; all three groups reached high levels of accuracy in the training procedure. Therefore, to what compensatory brain mechanisms do the authors attribute these similarities in performance across groups in the face of the differences found in the hypothesized—but few—brain areas? If the learning and performance of these three groups were equally good, there would appear to be either 1) some compensatory processes or brain regions that function well to compensate for the hypothesis-driven impairments found in a few brain areas or 2) the hypothesis-driven brain areas that were found to be deficient in the schizophrenia groups were unrelated to the performance on the auditory recognition and discrimination tasks used by the authors.

For me, the most important findings from the study were those noted in these two points. A thoughtful exchange with the authors on the interpretation of their results would help to elucidate further the purported relations between learning and brain function, as measured by the particular PET imaging technique used by Dr. Lahti and her colleagues.