Visual dysfunction and psychosis: A systematic review and meta-analysis

Visual impairment and broader visual-related abnormalities have been implicated in psychosis, but the consistency, specificity, and direction of this association remain unclear. We systematically reviewed evidence linking clinically defined visual impairment, visual-processing abnormalities, oculomotor/visuomotor abnormalities, higher-order visual dysfunction, and psychosis-related outcomes.

PubMed/MEDLINE, EMBASE, and PsycINFO were searched according to PRISMA guidelines. Studies were synthesized narratively by design and visual-related domain. An exploratory random-effects meta-analysis was conducted when sufficiently comparable cross-sectional estimates were available for clinically defined or functional visual impairment.

Fifty-six unique studies were included and synthesized across longitudinal or genetically informed, case-control, cross-sectional, and case-report evidence. Longitudinal findings were heterogeneous, with positive, inverse, null, attenuated, and reverse-direction associations. Case-control studies reported abnormalities across retinal/ocular, central visual-system, electrophysiological, oculomotor, subjective, and higher-order visual-cognitive domains, whereas visual-acuity findings were mixed. Cross-sectional studies provided the most consistent evidence for an association between clinically defined or functional visual impairment and psychosis-related outcomes. In the primary meta-analysis of 12 cross-sectional estimates, visual impairment was associated with higher odds of psychosis-related outcomes (OR = 1.98, 95% CI 1.35-2.90). Sensitivity analyses remained positive under broader eligibility criteria and after exclusion of the largest effect estimate. Direction-stratified and hallucination-related analyses also supported a positive association, although smaller subgroups were less precise.

Visual dysfunction is associated with psychosis-related outcomes, but interpretation depends on visual domain and study design. Current evidence supports a cautious, bidirectional, domain-based framework rather than a single causal pathway from visual impairment to psychosis.