C-reactive protein and cognitive function in bipolar disorder: clinical associations and exploratory transcriptomic profiling

Abstract Background Cognitive dysfunction is common in bipolar disorder (BD), and systemic inflammation has been implicated in cognitive abnormalities. However, the association between peripheral C-reactive protein (CRP) levels and cognitive performance in newly diagnosed, drug-naïve patients with BD remains incompletely characterized. This study examined CRP–cognition associations and additionally explored peripheral transcriptomic alterations in BD. Methods A total of 135 newly diagnosed, drug-naïve patients with BD and 85 healthy controls (HCs) were recruited. Cognitive performance was assessed using the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) and the Stroop Color-Word Test. Between-group cognitive comparisons were corrected for multiple testing using the Benjamini–Hochberg false discovery rate (FDR) procedure, and CRP–cognition associations were further examined using multivariable linear regression adjusting for demographic and clinical covariates. Exploratory RNA sequencing was performed using peripheral blood mononuclear cells from 6 patients with BD and 6 HCs, followed by bioinformatic analyses and RT-qPCR evaluation of selected candidate genes. Results Serum CRP levels were significantly higher in patients with BD than in HCs ( P = 0.020), and patients with BD showed poorer performance across multiple cognitive measures. In unadjusted analyses, higher CRP was associated with lower RBANS total, List Learning, List Recognition, Stroop total, Word-reading, and Color-word scores. In multivariable analyses adjusting for age, sex, education, BMI, smoking status, and HAMD-17 score, higher log-transformed CRP remained associated with lower RBANS total scores (standardized β = −0.214, P = 0.032) and poorer Stroop Word-reading performance (β = −0.207, P = 0.029), whereas the association with Stroop total score was attenuated (β = −0.168, P = 0.082). Exploratory RNA-seq identified 617 candidate differentially expressed genes; only one gene remained significant after genome-wide FDR correction. RT-qPCR identified significant differences in several candidate genes after FDR correction, although concordance with RNA-seq directionality was incomplete. Conclusions Higher peripheral CRP levels were associated with selected aspects of cognitive performance in newly diagnosed, drug-naïve patients with BD. After adjustment for demographic and clinical covariates, associations remained for RBANS total score and Stroop Word-reading performance. The exploratory transcriptomic findings identified candidate peripheral molecular alterations in BD but should be considered hypothesis-generating given the small RNA-seq sample and limited findings after multiple-testing correction.