Impacts of risk thresholds and age on clinical high risk for psychosis: a comparative network analysis

Christophe Gauld, Pierre Fourneret, Ben Alderson-Day, Emma Palmer-Cooper, Clément Dondé
Eur Arch Psychiatry Clin Neurosci. 2024-08-21; :
DOI: 10.1007/s00406-024-01876-y

PubMed

Lire sur PubMed

Gauld C(1)(2), Fourneret P(1)(2), Alderson-Day B(3), Palmer-Cooper E(4), Dondé C(5)(6)(7).

Author information:
(1)Department of Child and Adolescent Psychopathology, CHU de Lyon, 69000, Lyon,
France.
(2)Institut des Sciences Cognitives Marc Jeannerod, UMR 5229 CNRS & Université
Claude Bernard Lyon 1, 69000, Lyon, France.
(3)Department of Psychology, Durham University, Durham, UK.
(4)School of Psychology, Centre for Innovation in Mental Health, University of
Southampton, Southampton, UK.
(5)Univ. Grenoble Alpes, 38000, Grenoble, France.
.
(6)INSERM, U1216, 38000, Grenoble, France. .
(7)Psychiatry Department, CHU Grenoble Alpes, 38000, Grenoble, France.
.

One of the main goals for supporting people with a psychotic disorder is early
detection and intervention, and the detection of Clinical High Risk (CHR) is a
major challenge in this respect. This study sought to compare core symptoms of
CHR for psychosis networks based on two CHR self-assessment tools, across
different risk thresholds and age groups. This cross-sectional online
investigation analyzed 936 individuals for CHR, in France and the UK, with the
Prodromal Questionnaire-16 (PQ-16) and the Perceptual and Cognitive Aberrations
(PCA). Twelve different symptom networks were constructed, assessing
relationships, compactness, centrality, predictability, and comparisons between
them, based on different thresholds and age groups. In the above-threshold PQ-16
network, the most central symptom was « Voices or whispers »; in the PCA network,
the most central symptom was « Non-relevant thoughts distract or bother ». They
presented low overall predictability. No significant difference was found
between them. This study makes three key contributions. First, this
cross-network analyses highlight the relative importance of some central
symptoms. Secondly, comparisons between networks demonstrate the unity of the
CHR construct across scales, thresholds, and ages, affirming its phenotypic
homogeneity, an essential issue for patient care pathways. Thirdly, the low
average network predictability suggests the existence of unconsidered symptoms
within these CHR networks. These results shed light on the organization of CHR
symptoms using routine clinical questionnaires, offering insights for preventive
targets in a logic of precision semiology.