White matter microstructural and inflammation-based subgroups in bipolar disorder II depression differentiate in depressive and psychotic symptoms.

BACKGROUND

Elevated inflammation and impaired white matter (WM) microstructure have been observed in bipolar disorder (BD). The link between inflammation, WM integrity, and psychiatric symptoms in BD-II depression (BDII-D) remains unknown. We aimed to define BDII-D subgroups through the interplay of inflammation and WM microstructure, and to explore differences in psychiatric symptoms between subgroups, thus offering insight into elucidating the explanatory measures linked to BDII-D.

METHODS

WM differences were compared between 146 BDII-D individuals and 151 health controls (HCs) by Tract-Based Spatial Statistics. Partial correlation with multiple comparison corrections was used to explore associations between WM, inflammation, and psychiatric symptoms. The canonical correlation analysis metrics of WM and inflammation followed by k-means clustering were used to define WM microstructural-inflammation subgroups of BDII-D. The differences in clinical profiles were compared between the subgroups.

RESULTS

Compared with HCs, BDII-D showed significant WM alterations in the anterior thalamic radiation (ATR), cingulum, forceps, and inferior fronto-occipital fasciculus. In BDII-D, lower fraction anisotropy (FA) within the right ATR and cingulum were significantly associated with higher interleukin-6, while lower FA in the cingulum and lower axial diffusivity in the forceps major exhibited significant links with higher C-reactive protein. Among the subgroups identified, subgroup II characterized by elevated inflammation and impaired WM integrity displayed greater psychiatric symptoms.

CONCLUSIONS

WM alterations are concentrated in emotional neurocircuits and are linked to inflammation in BDII-D. WM-inflammation subgroups exhibit distinct variations in psychiatric symptoms. Thus, WM alterations and inflammation might be an explanatory process in the pathophysiology of BDII-D.