The Role of Cognition and Social Functioning as Predictors in the Transition to Psychosis for Youth With Attenuated Psychotic Symptoms

Abstract

In the literature, there have been several attempts to develop prediction models for youth who are at clinical high risk (CHR) of developing psychosis. Although there are no specific clinical or demographic variables that seem to consistently predict the later transition to psychosis in those CHR youth, in addition to attenuated psychotic symptoms, the most commonly occuring predictors tend to be poor social functioning and certain cognitive tasks. Unfortunately, there has been little attempt to replicate alogorithms. A recently published article by Cornblatt et al suggested that, for individuals with attentuated psychotic symptoms (APS), disorganized communication, suspiciousness, verbal memory, and a decline in social functioning were the best predictors of later transition to psychosis (the RAP model). The purpose of this article was to first test the prediction model of Cornblatt et al with a new sample of individuals with APS from the PREDICT study. The RAP model was not the best fit for the PREDICT data. However, using other variables from PREDICT, it was demonstrated that unusual thought content, disorganized communication, baseline social functioning, verbal fluency, and memory, processing speed and age were predictors of later transition to psychosis in the PREDICT sample. Although the predictors were different in these 2 models, both supported that disorganized communication, poor social functioning, and verbal memory, were good candidates as predictors for later conversion to psychosis.

Keywords: clinical high risk; prediction; prodromal; schizophrenia.

Fig. 1.

Kaplan–Meier survival estimates in 3 risk groups. Kaplan–Meier curves for 3 risk groups are presented. To give a reasonable spread of risk, we chose 3 prognostic groups which are good, fairly good, and poor. The required cut-points are the 27th and 73th centiles of the continuous variable, the PI in our dataset. Absolute values are 0.49 and 1.19.