Refining the latent structure of neuropsychological performance in schizophrenia.


Journal Article

BACKGROUND: Elucidating the cognitive architecture of schizophrenia promises to advance understanding of the clinical and biological substrates of the illness. Traditional cross-sectional neuropsychological approaches differentiate impaired from normal cognitive abilities but are limited in their ability to determine latent substructure. The current study examined the latent architecture of abnormal cognition in schizophrenia via a systematic approach. METHOD: Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were carried out on a large neuropsychological dataset including the Brief Assessment of Cognition in Schizophrenia, Continuous Performance Test, Wisconsin Card Sorting Test, Benton Judgment of Line Orientation Test, and Wechsler Abbreviated Scale of Intelligence matrix reasoning derived from 1012 English-speaking ethnic Chinese healthy controls and 707 schizophrenia cases recruited from in- and out-patient clinics. RESULTS: An initial six-factor model fit cognitive data in healthy and schizophrenia subjects. Further modeling, which accounted for methodological variance between tests, resulted in a three-factor model of executive functioning, vigilance/speed of processing and memory that appeared to best discriminate schizophrenia cases from controls. Factor analytic-derived g estimands and conventionally calculated g showed similar case-control discrimination. However, agreement analysis suggested systematic differences between both g indices. CONCLUSIONS: Factor structures derived in the current study were broadly similar to those reported previously. However, factor structures between schizophrenia subjects and healthy controls were different. Roles of factor analytic-derived g estimands and conventional composite score g were further discussed. Cognitive structures underlying cognitive deficits in schizophrenia may prove useful for interrogating biological substrates and enriching effect sizes for subsequent work.

Full Text

Duke Authors

Cited Authors

  • Lam, M; Collinson, SL; Eng, GK; Rapisarda, A; Kraus, M; Lee, J; Chong, SA; Keefe, RSE

Published Date

  • December 2014

Published In

Volume / Issue

  • 44 / 16

Start / End Page

  • 3557 - 3570

PubMed ID

  • 25066336

Pubmed Central ID

  • 25066336

Electronic International Standard Serial Number (EISSN)

  • 1469-8978

Digital Object Identifier (DOI)

  • 10.1017/S0033291714001020


  • eng

Conference Location

  • England