Complete blood count-based inflammation indexes and symptom severity in people with bipolar disorder: an analysis based on structural equation modelling

Aims: Bipolar disorder (BD) may be linked to immune-inflammatory dysregulation. Recently, complete blood count (CBC)-based inflammation indexes—neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), and platelet-to-lymphocyte ratio (PLR)—have emerged as potential, reproducible, and cost-effective markers for mental disorders. This study thus aimed to investigate the relationship of NLR, MLR, and PLR with manic and depressive symptom severity in people with BD, jointly testing the interactions with relevant clinical variables. Methods: We included inpatients with BD aged ≥ 18 consecutively hospitalized from May 2020 to March 2025. CBC-based ratios were calculated from fasting blood samples. Structural equation modelling (SEM) was performed to test the relationships of CBC-based ratios with manic and depressive symptom severity—assessed by the Young Mania Rating Scale (YMRS) and the Montgomery–Åsberg Depression Rating Scale (MADRS), respectively—accounting for age, sex, body mass index, alcohol/substance use disorders, and psychotropic medication doses. Results: We included 175 participants (mean age 46.8 ± 16.1 years; 48.6% males), 126 with a manic episode and 49 with a depressive episode. The MLR was higher in mania than in depression (p = 0.019), while no significant differences emerged for NLR and PLR. The SEM showed that greater YMRS scores were associated with higher NLR (coeff. = 0.077, p < 0.001) and MLR (coeff. = 0.096, p < 0.001), and lower PLR (coeff. = –0.088, p < 0.001). Moreover, higher MADRS scores were associated with lower MLR (coeff. = –0.189, p < 0.001) and higher PLR (coeff. = 0.123, p < 0.001). Conclusion: This study provides novel insights into the differential associations of NLR, MLR, and PLR with symptom severity across manic and depressive episodes, underscoring a complex immune-inflammatory dysregulation in BD. Notwithstanding generally small coefficients, our findings suggest that CBC-based ratios may represent accessible indexes to monitor mood state severity. Further investigation into their clinical utility is needed.