Attentional bias in paranoia: systematic review and meta-analysis

TY - JOUR

T1 - Attentional bias in paranoia

T2 - systematic review and meta-analysis

AU - Eid, Laura

AU - Hampshire, Chloe

AU - Vamvakas, George

AU - Jacobsen, Pamela

AU - Stahl, Daniel

AU - Yiend, Jenny

N1 - Publisher Copyright: © The Author(s), 2026. Published by Cambridge University Press.

PY - 2026/4/6

Y1 - 2026/4/6

N2 - BACKGROUND: Paranoia is a transdiagnostic symptom and is associated with cognitive and social impairments. Attentional bias toward threat is thought to maintain paranoia.AIMS: Despite many studies, attentional biases in paranoia have not been systematically summarised, which was the aim of the current work.METHOD: We conducted a systematic review and meta-analysis, identifying 10 964 studies, of which 35 met inclusion criteria for review and 15 for meta-analysis.RESULTS: Findings showed a significant negative attentional bias (average standardised effect size 0.26; 95% CI 0.01-0.52; p = 0.046). Preliminary indications suggested bias was strongest for paranoia-related stimuli (average effect size 0.30; 95% CI 0.03-0.57; p = 0.027) and stronger for words than faces (average effect size 0.41; 95% CI 0.05-0.77; p = 0.027), but more data is needed to confirm these effects. Limitations were primarily statistical and included likely underestimation of the overall effect size of the association between negative attentional bias and paranoia and a lack of sufficient studies to robustly examine moderators. CONCLUSIONS: Summarising this literature provides a rationale for existing and new interventions for paranoia that target biased attentional mechanisms.

AB - BACKGROUND: Paranoia is a transdiagnostic symptom and is associated with cognitive and social impairments. Attentional bias toward threat is thought to maintain paranoia.AIMS: Despite many studies, attentional biases in paranoia have not been systematically summarised, which was the aim of the current work.METHOD: We conducted a systematic review and meta-analysis, identifying 10 964 studies, of which 35 met inclusion criteria for review and 15 for meta-analysis.RESULTS: Findings showed a significant negative attentional bias (average standardised effect size 0.26; 95% CI 0.01-0.52; p = 0.046). Preliminary indications suggested bias was strongest for paranoia-related stimuli (average effect size 0.30; 95% CI 0.03-0.57; p = 0.027) and stronger for words than faces (average effect size 0.41; 95% CI 0.05-0.77; p = 0.027), but more data is needed to confirm these effects. Limitations were primarily statistical and included likely underestimation of the overall effect size of the association between negative attentional bias and paranoia and a lack of sufficient studies to robustly examine moderators. CONCLUSIONS: Summarising this literature provides a rationale for existing and new interventions for paranoia that target biased attentional mechanisms.

UR - https://www.scopus.com/pages/publications/105035170976

U2 - 10.1192/bjo.2026.10993

DO - 10.1192/bjo.2026.10993

M3 - Article

C2 - 41936630

SN - 2056-4724

VL - 12

JO - BJPsych Open

JF - BJPsych Open

IS - 3

M1 - e100

ER -