Esra Emekli, Burak Can Demirel, Emre Emekli
Whole-hypothalamus volume was lower in AN (p = 0.006). The multivariate subregional effect was borderline (p = 0.049), with five subunits showing lower volumes after FDR correction. After global brain-volume adjustment, the whole-hypothalamus effect remained significant (p = 0.006), whereas the subregional effect was attenuated (p = 0.056). Among underweight participants with AN, whole-hypothalamus (p < 0.001) and multivariate subregional differences (p = 0.002) remained significant after global brain-volume adjustment, with the same five subunits surviving FDR correction. The group effect was not significant after BMI adjustment, and the initial group-by-BMI interaction was not robust to sensitivity analyses. Eating-disorder symptom scores were not independently associated with hypothalamic volume.
OBJECTIVE: The hypothalamus plays a central role in energy homeostasis and neuroendocrine regulation, processes altered in anorexia nervosa (AN), yet hypothalamic morphology remains insufficiently characterized. We examined whole and subregional hypothalamic volumes in adult women with AN and healthy controls.
METHOD: This secondary analysis included 47 women with AN and 51 healthy controls. T1-weighted MRI data were processed with FastSurfer and FreeSurfer. Primary comparisons adjusted for age and estimated total intracranial volume. Sensitivity analyses additionally adjusted for segmented brain volume excluding ventricular volume, restricted AN participants to BMI < 18.5 kg/m2, and examined BMI effects. Eating Disorder Examination Questionnaire associations were tested. Longitudinal weight, treatment, and endocrine data were unavailable.
RESULTS: Whole-hypothalamus volume was lower in AN (p = 0.006). The multivariate subregional effect was borderline (p = 0.049), with five subunits showing lower volumes after FDR correction. After global brain-volume adjustment, the whole-hypothalamus effect remained significant (p = 0.006), whereas the subregional effect was attenuated (p = 0.056). Among underweight participants with AN, whole-hypothalamus (p < 0.001) and multivariate subregional differences (p = 0.002) remained significant after global brain-volume adjustment, with the same five subunits surviving FDR correction. The group effect was not significant after BMI adjustment, and the initial group-by-BMI interaction was not robust to sensitivity analyses. Eating-disorder symptom scores were not independently associated with hypothalamic volume.
DISCUSSION: Lower hypothalamic volume in AN was not explained by global brain-volume variation, although evidence for a specific subregional pattern in the full sample was less robust. Findings were strongest among underweight participants and support a close association between nutritional status and hypothalamic morphology.