A model using three patient characteristics was a significant predictor of general neurocognitive impairment in patients 3 years after their first episode of a nonaffective psychosis was diagnosed, in a study conducted in Spain.

Of the different clinical, neuropsychological, premorbid, and sociodemographic variables, three – premorbid IQ, baseline performances on verbal memory, and motor dexterity tasks – were the only statistically significant predictors of global neurocognitive function in the patients at the 3-year follow-up, reported Dr. Rosa Ayesa-Arriola, of the department of psychiatry at Marqués de Valdecilla University Hospital, Santander, Spain, and her associates. (Prog. Neuropsychopharmacol. Biol. Psychiatry 2013;43:23-8).

The model using these three variables was almost 80% accurate in predicting neurocognitive functioning overall, correctly classifying 75% of the patients who did not have neurocognitive impairment and 83.6% of those who had impairment at 3 years, they said. These results indicate that a lower premorbid IQ and worse performances on verbal memory and motor dexterity tasks “robustly predicted” general neurocognitive impairment, and that it might be possible to correctly identify long-term neurocognitive impairment “by detecting baseline deficits in verbal memory, fine motor skills, and low IQ,” they wrote.

The study is the first to compare neurocognitive outcomes in patients with a first episode of psychosis “within the context of a multifactorial model,” they noted.

The study evaluated data on about 150 patients, aged 15-60 years, who were treated as outpatients and inpatients at the hospital for their first episode of psychosis (most had schizophrenia) and were randomly assigned to antipsychotic treatments. (The patients are part of a large epidemiologic and longitudinal intervention program of patients with first-episode psychosis.) The data collected included neurocognitive assessments at baseline, and at the 3-year follow-up and information on premorbid and sociodemographic variables for patients were collected from patients, relatives, and medical records.

Although more studies are needed to replicate and evaluate the effectiveness of the model, they concluded, since people with “schizophrenia have cognitive deficits that impair [fundamental] aspects of their daily lives, the suggested predictive model could be an important step in approving a drug for a cognitive improvement indication or to alter the way that schizophrenia is currently treated.”

The authors cited the large homogenous sample and the 3-year longitudinal design as strengths of the study but acknowledged several limitations. For instance, some patients were not included in the 3-year analysis because neuropsychological evaluations were incomplete, and other factors could influence results, such as medication adherence or hospitalizations during the 3-year period. Dr. Ayesha-Arriola had no conflicts of interest.