A Network Approach to Bipolar Symptomatology in Patients with Different Course Types.

Area of Science:

• Psychiatry
• Network Science
• Bipolar Disorder Research

Background:

• Bipolar disorder (BD) presents a highly variable longitudinal mood course.
• Past disease course is a strong predictor of future BD trajectory, suggesting consistent symptom patterns.
• Novel network analysis approaches can elucidate underlying structures of BD symptomology.

Purpose of the Study:

• To investigate if distinct longitudinal course types in BD patients are associated with unique symptom correlation network characteristics.
• To apply network analysis to understand symptom interdependencies in different BD patient subgroups.

Main Methods:

• Categorized 125 BD patients into minimally impaired, predominantly depressed, and cycling groups based on two-year monthly life charts.
• Calculated groupwise Spearman's rank correlation coefficients between Young Mania Rating Scale (YMRS) and Quick Inventory of Depressive Symptomatology (QIDS) items.
• Compared weighted symptom networks and centrality measures across the three patient groups.

Main Results:

• Significant differences were observed in weighted symptom networks among the three BD course groups.
• The cycling group showed the strongest interconnections between manic and depressed symptoms.
• Central symptoms varied by group: elevated mood/increased speech (stable), loss of self-esteem/psychomotor slowness (depressed), and concentration loss/suicidality (cycling).

Conclusions:

• Symptom networks differ significantly among bipolar disorder patients with distinct longitudinal course types.
• These findings highlight the utility of network analysis in understanding BD heterogeneity.
• Further clinical interpretation and implications of these network differences are discussed.