Startle Latency improves Positive Predictive Power for Psychotic Conversion in Clinical High Risk Subjects: Findings from the North American Prodromal Longitudinal Study (NAPLS)

Presentation First Author: 
Kristin Cadenhead

Background: The NAPLS Consortium investigates risk, prediction and mechanism of psychosis in individuals at clinical high risk (CHR) for psychosis. Methods: 506 CHR subjects were assessed with the Structured Interview for Prodromal Syndromes (SIPS) and the startle paradigm, a candidate biomarker for psychosis risk. All subjects received clinical follow-up for up to 3.5 years and were classified as having progressed to psychosis (N=47) or not (N=459). Sensitivity and specificity of startle latency values and psychotic conversion were assessed using a Receiver Operating Characteristic (ROC) curve and then used in conjunction with the SIPS as a means to improve the positive predictive power (PPV) for psychosis. Results: The conversion rate to psychosis was 27% over 3.5 years. The AUC for startle latency was 0.60 (p69 ms was 0.37, while that of P1>3 was 0.38. If CHR subjects were above the cutoffs for both latency and P1, the PPV increased to 0.51. Psychosis prediction analyses using Cox Regression for latency, P1 and the two combined were all significant (p's

Conference Name: 
Presentation Date: 
January, 2015
Additional Authors: 
Jean Addington - Carrie Bearden - Tyrone Cannon - Barbara Cornblatt - Daniel Mathalon - Thomas McGlashan - Diana Perkins - Larry Seidman - Ming Tsuang - Elaine Walker - Scott Woods
See other presentations in this session: