2026 ECNP Congress

PS03-3077

Immune-related correlates of cognitive variability in bipolar disorder: a multiblock partial least squares analysis

Federico Bottaro — P. Enrico, C. Cabasino, E. Tassi, S. Peri, G. De Franceschi, E. Maggioni, G. Delvecchio, C. Molent, F. Papaleo, P. Brambilla · University of Milan, Department of Pathophysiology and Transplantation, Milan, Italy

2026 ECNP Congress

Introduction

Cognitive impairment is a relevant dimension of bipolar disorder (BD), contributing to poor functional outcome and inter-individual variability [1]. Dysbindin-1, encoded by the dystrobrevin-binding protein 1 gene (DTNBP1), is involved in synaptic trafficking and dopamine D2 receptor-related mechanisms [2]. The DysBray haplotype has been associated with reduced DTNBP1 expression, while dysbindin-1 variation has been linked to cognitive response to antipsychotics [3,4]. Immune dysregulation, including cytokine alterations, is implicated in BD, although its relationship with cognitive variability remains unclear [5].

Aims

We examined whether predefined haplotype, demographic, environmental, immunological and clinical blocks were associated with cognitive performance in BD patients, focusing on DysBray haplotype status and immune markers.

Methods

We analyzed baseline data from 117 BD patients (mean age 45.29 years) receiving aripiprazole, quetiapine or olanzapine combined with a mood stabilizer. Baseline assessment included (i) demographic variables (age, sex) (ii) clinical variables (BPRS, GAF, YMRS,⁠ HAM-D), (iii) environmental variables (age at onset, alcohol use, drug use, medication dose equivalents), (iv) DysBray haplotype genotyping and (v) immune gene-expression profiling. Immune markers were quantified from whole blood by quantitative real-time polymerase chain reaction using OpenArray technology. These sets of variables were used as candidate predictors of cognitive outcomes. Cognitive performance was assessed with the Brief Assessment of Cognition in Affective Disorders. Multi-block partial least square (MBPLS) was applied to extract latent variables (LVs) capturing multivariate associations among predictive blocks and cognitive outcome variables. Specifically, we evaluated MBPLS models considering: (a) all five predictors; (b) all predictors excluding the immunological block, and (c) haplotype, demographic and immunological blocks, with or without additional environmental and clinical blocks. For each LV, the block scores and loadings were extracted. Significance of each LV was quantified by computing the correlation among predictor and response scores, including permutation testing.

Results

In the full MBPLS model including haplotype, demographic, environmental, immunological and clinical blocks, only the immunological block survived permutation testing, showing significant associations between immunological and cognitive outcome (permutation p=0.013, 0.016, 0.021 and 0.012; corresponding nominal p values=0.002, 0.051, 0.027 and 0.0009). The haplotype block resulted not significantly associated with cognition (permutation p=0.314 and 0.314; nominal p=0.147 and 0.941). Demographic components showed associations in the full model (nominal p=0.0002 and 0.019) but did not survive permutation testing. Environmental and clinical blocks were nominally significant, but did not pass permutation testing. In reduced models excluding immunological variables, the demographic block was significantly associated with cognition (haplotype plus demographic model: permutation p=0.010 and 0.030; after adding environmental variables: permutation p=0.007 and 0.008; after adding clinical variables: permutation p=0.003 and 0.006). In models including immune markers together with the haplotype block, immunological components remained significant, whereas haplotype components remained non-significant.

Conclusions

These findings suggest a significant multivariate association among cognitive variability and immune markers in BD, while DysBray haplotype status did not. Demographic factors contributed mainly to the absence of immune markers, and clinical-environmental associations, though nominally significant, did not survive permutation testing. Consistent with evidence of BD immune dysregulation [5], these results support further investigation of immune-cognitive pathways for biological stratification.

References

  1. MacQueen GM, Memedovich KA. Cognitive dysfunction in major depression and bipolar disorder: assessment and treatment options. Psychiatry and Clinical Neurosciences. 2017;71(1):18–27.
  2. Ji Y, Yang F, Papaleo F, Wang HX, Gao WJ, Weinberger DR, Lu B. Role of dysbindin in dopamine receptor trafficking and cortical GABA function. Proceedings of the National Academy of Sciences of the United States of America. 2009;106(46):19593–19598.
  3. Bray NJ, Preece A, Williams NM, et al. Haplotypes at the dystrobrevin binding protein 1 (DTNBP1) gene locus mediate risk for schizophrenia through reduced DTNBP1 expression. Human Molecular Genetics. 2005;14(14):1947–1954.
  4. Scheggia D, Mastrogiacomo R, Mereu M, et al. Variations in Dysbindin-1 are associated with cognitive response to antipsychotic drug treatment. Nature Communications. 2018;9:2265.
  5. Munkholm K, Vinberg M, Vedel Kessing L. Cytokines in bipolar disorder: a systematic review and meta-analysis. Journal of Affective Disorders. 2013;144(1–2):16–27.