Verbal memory and processing speed are two of the proposed neurocognitive predictors in schizophrenia. the objective is to determinate neurocognitive predictors of functioning in one five years follow-up period on ambulatory schizophrenia patients.

We conduct one cohort study on 30 schizophrenia out-patients (19 male; age mean: 32.8 years; SD:7.2). at the moment of inclusion it was applied one neuropsychological battery sensitive to neuropsychological deficit in schizophrenia: WAIS-III, BADS, WCST, Colour Trails, Trail Making A and B, BVRT, California Verbal Learning Test (Spanish version: TAVEC). Variables was summarized determining Z values and principal components. the cohort was prospectively studied for up to five years. as result variables it were considered: time to the first antipsychotic-drug change, time to the first psychiatric admission, and average five-years annual score in the Strauss-Carpenter Outcome Scale. It was constructed Cox and Linear Regression Models to determinate the better predicting neurocognitive components.

The better global outcome was directly related with WAIS-III processing speed index (corrected R square: 0.19; p = 0.02; beta constant= 0.469; beta WAIS-III processing speed= 0.125). None predictor was selected for the outcome variable change of antipsychotic. Psychiatric admission was predicted (p=0.006) by implicit learning (beta= -2.19), executive functioning (beta= 1.02), WAIS-III Total IQ (beta= 0.45) and WAIS-III Perceptual Organization Index (beta= -0.20).

Higher processing speed index predicts one better functioning outcome during five-years follow-up. the risk for psychiatric admission was heterogeneously related with neurocognitive predictors. Verbal memory did not predict functional outcome.