Neurocognitive functioning of subjects with putative pre-psychotic states and early psychosis.

BACKGROUND The neurocognitive functioning of patients with schizophrenia is likely to decline at the early stage of the illness. More evidence is needed to determine whether deficits in certain domains of neurocognition precede the onset of illness and can predict the onset of psychosis. METHODS Subjects were recruited from the SOPRES study in Taiwan. A neuropsychological battery including the continuous performance test, Wisconsin Card Sorting Test, Wechsler Adult Intelligence Scale-Third Edition, Trail Making Tests, Mandarin version of the Verbal Fluency Test, and Wechsler Memory Scale-Third Edition, was applied at baseline and 1-year follow-up. Neurocognitive profiles derived from these tests were categorized into 9 domains for comparisons among subjects with different levels of clinical severity. RESULTS A total of 324 participants, including 49 with first episode psychosis (FEP), 53 with ultra-high risk (UHR), 42 with intermediate risk (IR), 43 with marginal risk (MR), and 137 normal controls completed a baseline assessment and 71% of the participants completed a 1-year follow-up assessment. The profiles of the UHR and IR groups were identical at baseline. Those who converted to FEP later on (UHR+) showed relatively poorer performance than non-converters (UHR-) at baseline. At follow-up the performance of UHR+ was compatible to that of FEP, while UHR- generally improved. CONCLUSIONS By including subjects with early putative pre-psychotic states, our study clarifies some inconsistencies about the timing and stability of changes in neurocognitive functioning that occur at the start of psychosis; it also raises questions regarding the feasibility of using neurocognitive deficits to predict the risks of transition to psychosis.