Effect of treatment status on whole brain fractional anisotropy in patients with anorexia nervosa

Holly Carrington, BA

Brigham and Women’s Hospital
Effect of treatment status on whole brain fractional anisotropy in patients with anorexia nervosa

Scientific Abstract

Background: Anorexia nervosa (AN) and atypical-AN (at-AN) are restrictive eating disorders characterized by restricted energy intake, an intense fear of weight gain and a distorted view of one’s body. Restriction in both disorders leads to weight loss, with less severe low- weight in at-AN. Prior work using diffusion magnetic resonance imaging (dMRI) have reported reductions in fractional anisotropy (FA), thought to be indicative of impaired white matter (WM) health. Longitudinal studies of FA have shown changes in WM in young underweight patients with AN can reverse following brief weight restoration therapy in higher level of care. Cross-sectional studies of WM microstructure in AN/at-AN rarely account for level of treatment on WM health which may explain conflicting results of WM integrity in patients with AN/at-AN. The purpose of this study was to understand the impact of level of treatment on WM microstructure in AN/at-AN. Given the greater severity of symptoms in AN/at-AN individuals in higher level of care (HLOC), we hypothesized that HLOC individuals would show the greatest deficits in WM microstructure compared to participants in outpatient treatment (OUT) or not in treatment (NT).

Methods: dMRI data were acquired from adolescent female healthy controls (HC[n=31]; 18.2±3.0 years old) and participants with AN in one of the following treatment levels: HLOC (n=18; 19.0±2.8 years old), OUT (n=25; 19.4±2.6 years old) or NT (n=12; 20.1±1.3 years old). FA maps were calculated for each individual and a WM skeleton was generated using an automated analytic pipeline (tract-based spatial statistics-TBSS). Group comparisons were carried out on the average whole brain FA values using analysis of variance (ANOVA) with an a=0.05

Results: Groups did not differ for age. BMI did not differ among the HLOC, OUT and NT groups, and was lower in patients with AN than in HCs (p<0.001). Differences in whole-brain FA between groups were not statistically significant (F(3,82)=0.762, p=0.519).

Conclusions: No statistically significant differences in whole brain FA were observed between groups. Future plans include using FW imaging to assess treatment effect on whole brain FA.

Live Zoom Session – April 21st

research Areas


Holly W. Carrington, BA, Danielle L. Kahn, BA, Laura M. Holsen, MS, PhD, Franziska Plessow, PhD, Kendra R. Becker, PhD, Jennifer J. Thomas, PhD, Martha E. Shenton, PhD, Kamryn T. Eddy, PhD, Madhusmita Misra, MD, MPH, Elizabeth A. Lawson, MD, MMSc, Lauren Breithaupt, PhD, Amanda E. Lyall, PhD

Principal Investigator

Amanda E. Lyall, PhD

Affiliated Website