This study aimed at using latent variable modelling to explore the significantly contributing variables to functioning in schizophrenia patients.

The study cohort comprised 296 schizophrenia patients evaluated once for demographic characteristics, functioning (FROGS, SWN-K, QLS) and symptomatology (Positive and Negative Syndrome Scale [PANSS]). First exploratory multivariate analyses were conducted and then a model with functioning as a latent variable was proposed and tested with the data.

Symptomatology as negative, cognitive and excitation factor are significant predictors of functioning assessed through FROGS (P<0.0001), SWN-K and QLS (P<0.001). The model was constructed with functioning defined as a latent variable, indicators are subscores on FROGS, SWN-K, QLS and exogenous variable included symptomatology, Duration of Untreated Psychosis (DUP) and educational level.

Using the five clinical dimensions of the PANSS, (Positive, Negative, Cognitive, Anxiety/Depression and Excitation) the negative and cognitive dimensions are highly correlated via the latent variable to the three dimensions of functioning evaluated by the FROGS: “daily life”, “social functioning” and “treatment” and the QLS subscores (interpersonnal, common object, instrumental role). Educationnal level is positively linked to functioning but not DUP. The model emphasizes the need for treatment strategies that have an effect on cognitive-factors, to improve functioning in schizophrenia.