White matter microstructural changes in adolescent anorexia nervosa including an exploratory longitudinal study

Abstract

Background: Anorexia nervosa (AN) often begins in adolescence, however, the understanding of the underlying pathophysiology at this developmentally important age is scarce, impeding early interventions. We used diffusion tensor imaging (DTI) to investigate microstructural white matter (WM) brain changes including an experimental longitudinal follow-up.

Methods: We acquired whole brain diffusion-weighted brain scans of 22 adolescent female hospitalized patients with AN at admission and nine patients longitudinally at discharge after weight rehabilitation. Patients (10-18 years) were compared to 21 typically developing controls (TD). Tract-based spatial statistics (TBSS) were applied to compare fractional anisotropy (FA) across groups and time points. Associations between average FA values of the global WM skeleton and weight as well as illness duration parameters were analyzed by multiple linear regression.

Results: We observed increased FA in bilateral frontal, parietal and temporal areas in AN patients at admission compared to TD. Higher FA of the global WM skeleton at admission was associated with faster weight loss prior to admission. Exploratory longitudinal analysis showed this FA increase to be partially normalized after weight rehabilitation.

Conclusions: Our findings reveal a markedly different pattern of WM microstructural changes in adolescent AN compared to most previous results in adult AN. This could signify a different susceptibility and reaction to semi-starvation in the still developing brain of adolescents or a time-dependent pathomechanism differing with extend of chronicity. Higher FA at admission in adolescents with AN could point to WM fibers being packed together more closely.

Keywords: Adolescence; Anorexia nervosa; DTI; Fractional anisotropy; TBSS.

Supplementary Fig. 1

TBSS results comparing FA for adolescent AN at admission versus TD corrected for age. Only patients measured with the identical protocol one (iso-voxel) were entered in the analysis. Significant areas with AN > TD are marked in orange, AN < TD are marked in blue. Underlying WM skeleton marked in green. AN — anorexia nervosa FA — fractional anisotropy TD — typically developing controls.

Supplementary Fig. 2

TBSS results comparing FA, MD, AD and RD for adolescent AN at admission versus TD corrected for age. Only patients measured with the identical protocol one (iso-voxel) were entered in the analysis. Significant areas with AN > TD are marked in orange, AN < TD are marked in blue. Underlying WM skeleton marked in green. AD — axial diffusivity AN_admission — anorexia nervosa patients at admission FA — fractional anisotropy MD — mean diffusivity RD — radial diffusivity TD — typically developing controls z — brain slice coordinate.

Supplementary Fig. 3

TBSS results using a different alignment strategy (DTI-TK) comparing FA for adolescent AN at admission versus TD corrected for age and protocol. Significant areas with AN > TD are marked in orange, AN < TD are marked in blue. Underlying WM skeleton marked in green. AN — anorexia nervosa FA — fractional anisotropy TD — typically developing controls

Supplementary Fig. 4

TBSS results excluding four patients with comorbidities and medication comparing FA for adolescent AN at admission versus TD corrected for age and protocol. Significant areas with AN > TD are marked in orange, AN < TD are marked in blue. Underlying WM skeleton marked in green. AN — anorexia nervosa FA — fractional anisotropy TD — typically developing controls.

Supplementary Fig. 5

Alternative method (SPM) results comparing FA for adolescent AN at admission versus TD corrected for age and protocol. Top: Significant areas with TD > AN are marked in grey. Bottom: AN > TD are marked in grey. Uncorrected for multiple comparison. AN — anorexia nervosa AN_admission — anorexia nervosa patients at admission FA — fractional anisotropy TD — typically developing controls.

Supplementary Fig. 6

ANCOVA results comparing the global WM skeleton for all diffusivity values corrected for age and protocol. AD — axial diffusivity AN_admission — anorexia nervosa patients at admission AN_discharge — anorexia nervosa patients at discharge FA — fractional anisotropy MD — mean diffusivity.

Fig. 1

TBSS results comparing FA for adolescent AN at admission versus TD corrected for age and protocol. Significant areas with AN > TD are marked in orange. Underlying WM skeleton marked in green. AN — anorexia nervosa FA — fractional anisotropy TD — typically developing controls WM — white matter.

Fig. 2

TBSS results comparing FA, MD, AD and RD for adolescent AN at admission versus TD. Significant areas with AN > TD are marked in orange, AN < TD are marked in blue. Underlying WM skeleton marked in green. AD — axial diffusivityAN — anorexia nervosaFA — fractional anisotropyMD — mean diffusivityRD — radial diffusivityTD — typically developing controlsz — brain slice coordinate.

Fig. 3

ANCOVA results comparing the region of interest differing in FA between AN_admission, AN_discharge and TD for all diffusivity values corrected for age and protocol. AD — axial diffusivity AN_admission — anorexia nervosa patients at admission AN_discharge — anorexia nervosa patients at discharge FA — fractional anisotropy MD — mean diffusivity RD — radial diffusivity roi — region of interest TD — typically developing controls # — original contrast for defining roi * — p < 0.05.