2026 ECNP Congress

PS03-3184

Differential contributions of cognitive reserve components to cognition and functioning in early affective and non-affective psychosis

Ainara Espluga — M.F. Forte, B. Sole, M. Garriga, D. Clougher, A. Meseguer, M. Mur, R. Lopez, D. Rey, N. Miguel, V. Llorca, I. Forcada, G. Parramon, M.A. Sandonis, J. Jesús-Suarez, S. Mas, P. Gassó, A. González-Segura, E. Vieta, S. Amoretti, M. Serra-Navarro · Bipolar and Depressive Disorders Unit- Hospital Clínic de Barcelona- Departament de Medicina- Institut de Neurociències UBNeuro- Universitat de Barcelona UB, Fundació Clínic-Institut d’Investigacions Biomèdiques August Pi I Sunyer IDIBAPS- CIBERSAM- ISCIII- Barcelona- Spain, Barcelona, Spain

2026 ECNP Congress

Introduction:

Cognitive Reserve (CR) has been associated with better cognitive and functional outcomes in early psychosis. However, it is unclear whether different components of CR contribute differentially to these outcomes. This study aimed to investigate associations between CR components and cognition and functioning in early psychosis, and whether these relationships differ by diagnostic group.

Materials and Methods:

A total of 103 patients with early psychosis were included and classified into affective and non-affective psychosis. CR was estimated using four proxies, based on Cognitive Reserve Assessment Scale in Health (CRASH): years of education, occupational attainment, engagement in social/intellectual activities. We also used estimated premorbid IQ from the vocabulary subtest from the Wechsler Adult Intelligence Scale IV.

Cognitive performance was assessed using a neuropsychological battery (CVLT, WCST, FAS, CPT-III). Composite scores for verbal memory, verbal fluency, and executive function were derived using PCA. In addition, standardized measures for attention, processing speed, and working memory were included. An overall cognitive score was derived using a second PCA, providing a comprehensive index of cognitive performance.

Psychosocial functioning was assessed using the Functioning Assessment Short Test (FAST) total score, with higher scores indicating poorer functioning.

General linear models were fitted separately for cognitive and functional outcomes, including diagnostic group and the four CR indicators. Interactions between diagnostic group and each CR indicator were tested to examine whether associations differed by group; as these interactions were not significant, final models retained only main effects.

Results:

The model for global cognition was significant, R² = .368, adjusted R² = .325, F(5,74) = 8.618, p < .001. Diagnostic group was not significantly associated with global cognition, p = .567. Among the CR indicators, education and estimated IQ were significant independent predictors. Higher education was associated with better global cognition, B = 0.107, p = .016, ηp² = .076. Estimated IQ showed the strongest association with global cognition, B = 0.028, p = .001, ηp² = .135. Leisure activities and occupational attainment were not significantly associated with global cognition.

The model for FAST total score was significant, R² = .208, adjusted R² = .168, F(5,97) = 5.105, p < .001. Diagnostic group showed a trend-level association, B = −4.621, p = .056, with the affective group showing lower FAST scores than the non-affective psychosis group. Among the CR indicators, occupational attainment was the only significant predictor of FAST total score, B = −1.517, p = .028, ηp² = .049. This result suggests that higher occupational attainment was associated with better psychosocial functioning. Leisure activities, education, and estimated IQ were not significantly associated with FAST total score.

Conclusions:

CR components in early psychosis are differentially associated with cognitive and functional outcomes, with premorbid IQ and educational attainment being the strongest predictors of cognition, and occupational attainment the main predictor of psychosocial functioning. These associations are consistent across affective and non-affective psychosis, with no evidence of group moderation. This supports the relevance of CR as a transdiagnostic protective factor in early psychosis.

The authors declare no conflict of interest.

References

No references