King's College London

Research portal

Meta-Analysis of longitudinal neurocognitive performance in people at clinical high-risk for psychosis

Research output: Contribution to journalReview articlepeer-review

Original languageEnglish
Pages (from-to)2009-2016
Number of pages8
JournalPsychological Medicine
Issue number11
Published13 Aug 2022

Bibliographical note

Funding Information: The authors thank S. Morris for his advice on the statistical methods and J. Addington, L. Liu, K. Allott, M. Fujioka, J. Lee, K. Lim, T. Lee and C. Liu for providing the additional data necessary to complete the meta-analysis. C. See was supported by the UK Medical Research Council (MR/N013700/1). S. Si was supported by the China Scholarship Council. H. Dickson is affiliated with the National Institute for Health Research (NIHR) Specialist Biomedical Research Centre for Mental Health at the South London and Maudsley NHS Foundation Trust and Institute of Psychiatry, Psychology and Neuroscience, King's College London, United Kingdom. M. Kempton was supported by a Medical Research Council Fellowship (grant MR/J008915/1). Publisher Copyright: © The Author(s), 2022. Published by Cambridge University Press.

King's Authors


Persons at clinical high-risk for psychosis (CHR) are characterised by specific neurocognitive deficits. However, the course of neurocognitive performance during the prodromal period and over the onset of psychosis remains unclear. The aim of this meta-Analysis was to synthesise results from follow-up studies of CHR individuals to examine longitudinal changes in neurocognitive performance. Three electronic databases were systematically searched to identify articles published up to 31 December 2021. Thirteen studies met inclusion criteria. Study effect sizes (Hedges' g) were calculated and pooled for each neurocognitive task using random-effects meta-Analyses. We examined whether changes in performance between baseline and follow-up assessments differed between: (1) CHR and healthy control (HC) individuals, and (2) CHR who did (CHR-T) and did not transition to psychosis (CHR-NT). Meta-Analyses found that HC individuals had greater improvements in performance over time compared to CHR for letter fluency (g =-0.32, p = 0.029) and digit span (g =-0.30, p = 0.011) tasks. Second, there were differences in longitudinal performance of CHR-T and CHR-NT in trail making test A (TMT-A) (g = 0.24, p = 0.014) and symbol coding (g =-0.51, p = 0.011). Whilst CHR-NT improved in performance on both tasks, CHR-T improved to a lesser extent in TMT-A and had worsened performance in symbol coding over time. Together, neurocognitive performance generally improved in all groups at follow-up. Yet, evidence suggested that improvements were less pronounced for an overall CHR group, and specifically for CHR-T, in processing speed tasks which may be a relevant domain for interventions aimed to enhance neurocognition in CHR populations.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454