Modelling reaction time distribution of fast decision tasks in schizophrenia: Evidence for novel candidate endophenotypes. (November 2018)

• Record Type: Journal Article
• Authors: Fish, Simon; Toumaian, Maida; Pappa, Eleni; Davies, Timothy J.; Tanti, Ruth; Saville, Christopher W.N.; Theleritis, Christos; Economou, Marina; Klein, Christoph; Smyrnis, Nikolaos
• Is Part Of: Psychiatry research. Volume 269(2018)
• Page Start: 212
• Page End: 220
• Publication Date: 2018-11
• Subjects: Ex-Gaussian model -- Drift Diffusion Model -- Heritable characteristic -- Working memory -- Cognitive deficit -- Psychosis; Psychiatry -- Periodicals; Psychiatry -- periodicals; Psychiatrie -- Périodiques; 616.89
• DOI: 10.1016/j.psychres.2018.08.067
• Languages: English
• ISSNs: 0165-1781

Abstract

Highlights: Processing speed measured by Reaction Time (RT) is slower in schizophrenia patients. Patients, healthy patient siblings and controls performed two-choice RT tasks. Ex-Gaussian RT model and Drift Diffusion Model (DDM) was applied on RT distributions. Ex-Gaussian RT parameters dissociated patients and siblings from controls. The rising decision signal in DDM dissociated patients and siblings from controls. RT distribution modeling reveals novel candidate endophenotypes of schizophrenia. Abstract: Increased reaction time (RT) and variability of RT in fast decision tasks is observed in patients with schizophrenia and their first degree relatives. This study used modelling of the RT distribution with the aim of identifying novel candidate endophenotypes for schizophrenia. 20 patients with schizophrenia, 15 siblings of patients and 25 healthy controls performed an oddball task of varying working memory load. Increases in mean and standard deviation (SD) of RT were observed for both patients and siblings compared to controls and they were again independent of working memory load. Ex-Gaussian modelling of the RT distribution confirmed that parameters μ, σ and τ increased significantly in patients and siblings compared to controls. The Drift Diffusion Model was applied on RT distributions. A decrease in the diffusion drift rate ( v ) modeling the accumulation of evidence for reaching the decision to choose one stimulus over the other, was observed in patients and siblings compared to controls. The mean time of the non-decisional sensorimotor processes ( t0 ) and it's variance ( st0 ) was also increased in patients and siblings compared to controls. In conclusion modeling of the RT distribution revealed novel potential cognitive endophenotypes in the quest of heritable risk factors for schizophrenia.