Neuroanatomical subtypes of bulimia nervosa show divergent gray matter patterns and distinct clinical profiles.
Bulimia nervosa (BN) is a clinically heterogeneous eating disorder. Variability in clinical presentation and treatment response suggests differences in underlying neurobiology. Neuroimaging studies have reported inconsistent findings of widespread brain structural alterations in BN, which may reflect unrecognized neuroanatomical subtypes. Identifying such subtypes could inform targeted treatment approaches.
81 treatment-naïve female Han Chinese patients with BN and 72 healthy controls from a single center underwent T1-weighted MRI. Computational Anatomy Toolbox 12 was used to quantify regional gray matter volumes (GMVs). The heterogeneity through discriminative analysis (HYDRA) algorithm was applied to identify neuroanatomical subtypes. Neuroanatomical characteristics of the subtypes were compared, and clinical characteristics were compared in 55 patients with available clinical assessments.
Two distinct BN subtypes with divergent GMV aberrations and clinical characteristics were identified. Subtype 1 exhibited widespread GMV reductions mainly involving the cerebellum, prefrontal, temporal, parietal, and occipital cortices, insula, and limbic regions. It was characterized by higher depressive symptoms, younger age, and lower estimated total intracranial volume (eTIV), with eTIV negatively correlated with eating disorder psychopathology, particularly weight and shape concerns. Subtype 2 showed relatively preserved GMV, with alterations restricted to the right cerebellum. In this subtype, emotion dysregulation showed a subtype-specific association with eating disorder psychopathology.
Two neuroanatomical subtypes of BN are characterized by distinct structural patterns and clinical profiles. Subtype 2 patients may benefit from emotion regulation-focused interventions, while those with Subtype 1 may benefit from more comprehensive treatment. These implications remain hypothetical and require validation in longitudinal interventional studies.
81 treatment-naïve female Han Chinese patients with BN and 72 healthy controls from a single center underwent T1-weighted MRI. Computational Anatomy Toolbox 12 was used to quantify regional gray matter volumes (GMVs). The heterogeneity through discriminative analysis (HYDRA) algorithm was applied to identify neuroanatomical subtypes. Neuroanatomical characteristics of the subtypes were compared, and clinical characteristics were compared in 55 patients with available clinical assessments.
Two distinct BN subtypes with divergent GMV aberrations and clinical characteristics were identified. Subtype 1 exhibited widespread GMV reductions mainly involving the cerebellum, prefrontal, temporal, parietal, and occipital cortices, insula, and limbic regions. It was characterized by higher depressive symptoms, younger age, and lower estimated total intracranial volume (eTIV), with eTIV negatively correlated with eating disorder psychopathology, particularly weight and shape concerns. Subtype 2 showed relatively preserved GMV, with alterations restricted to the right cerebellum. In this subtype, emotion dysregulation showed a subtype-specific association with eating disorder psychopathology.
Two neuroanatomical subtypes of BN are characterized by distinct structural patterns and clinical profiles. Subtype 2 patients may benefit from emotion regulation-focused interventions, while those with Subtype 1 may benefit from more comprehensive treatment. These implications remain hypothetical and require validation in longitudinal interventional studies.