The relationship between neuropsychological and overall performance in people with schizophrenia is known. Smoking and stimulant drugs use can improve neuropsychological outcomes, however the existence of drugs use may be a more severe illness marker.

The purpose of this study is to investigate which clinical and epidemiological variables, including stimulant drugs use and smoking, influence on neuropsychological performance in patients with psychosis.

92 patients with different psychosis were assessed with a battery that included SCIP, to assess neuropsychological performance, PANSS, to evaluate psychotalogy, GAF and SIX as global performance measures. We also explore clinic and sociodemographic data. A binary logistic regresion model was applied on scores on the task ‘words’ of the SCIP (memory and learning), dichotomized at the median. The model included: sex, age, onset age, family history, negative scale, positive scale, global psychopatology and PANSS subtypes, estimated premorbid IQ using Barona Index, stimulant drugs use, smoking and funcionality.

The absence of stimulating drugs use and smoking (trials 2 and 3 of SCIP), was associated with better memory and learning in patients with psychosis. Younger age (trials 1-4 and total of words), higher functionality (trials 1,2 and total of words) and premorbid IQ (trials 2 and 4) were also positively associated with better neuropsychological performance.

Stimulant drugs use and smoking could be markers of poorer previous neuropsychological function in psychotic patients. It is necessary to do longitudinal studies evaluating these variables as markers, risk or protective factors of cognitive performance