Structural brain alterations in anorexia nervosa: a global brain volume and anatomical likelihood estimation (ALE) meta-analysis combined with a functional decoding approach

TL;DR

This study finds that anorexia nervosa causes significant brain volume loss, which partially recovers with weight gain but remains reduced for up to 1.5 years.

Contribution

The study combines global brain volume analysis and ALE meta-analysis to reveal structural brain changes and preserved regions in anorexia nervosa.

Findings

Acute anorexia nervosa is associated with significant gray and white matter volume reductions, especially in adolescents.

Brain volume improvements occur with weight restoration but do not fully normalize within 1.5 years of recovery.

Key brain regions like the cingulate gyrus and precuneus show consistent structural reductions in anorexia nervosa.

Abstract

•Acute AN shows a reduction of –4.79% gray matter and –2.48% white matter with numerically larger reductions in adolescents.•Brain volumes improve with weight restoration but remain reduced up to 1.5 years post‑recovery.•Anatomical likelihood estimation (ALE) indicates widespread gray matter and cortical thickness reductions.•Key areas affected in AN include the cingulate gyrus, precentral gyrus, and precuneus.•Preserved regions colocalize with brain areas for eating, food cues, threat, and reinforcement (FDR-corrected).

Acute AN shows a reduction of –4.79% gray matter and –2.48% white matter with numerically larger reductions in adolescents.

Brain volumes improve with weight restoration but remain reduced up to 1.5 years post‑recovery.

Anatomical likelihood estimation (ALE) indicates widespread gray matter and cortical thickness reductions.

Key areas affected in AN include the cingulate gyrus, precentral gyrus, and precuneus.

Preserved regions colocalize with brain areas for eating, food cues, threat, and reinforcement (FDR-corrected).

Substantial brain volume loss is well-documented during acute anorexia nervosa (AN); however, longitudinal outcomes are unclear. Our comprehensive meta-analysis investigated global and regional structural brain alterations in adult and adolescent individuals with AN by extracting reported brain volume scores and neuroimaging coordinates from the literature. Results showed significant global brain volume reductions in gray matter (GM), white matter (WM), and increases in cerebrospinal fluid (CSF) in acute AN (N = 1130 patients; N = 40 papers), gradually improving upon weight rehabilitation. However, even after 1.5 years of recovery, significantly lower global GM volume compared to healthy controls was found (N = 232 patients; N = 12 papers). Regarding potential regional changes, our search identified 35 eligible papers with neuroimaging coordinates for 412 foci as input for our anatomical likelihood estimation (ALE) analyses. The results revealed widespread reductions of GM volume and cortical thickness, but notably also identified consistently affected brain regions including the cingulate gyrus, precentral gyrus, and precuneus. Spatial colocalization analyses using the Neurosynth data base indicated brain areas associated with eating, food, threat, and reinforcement to be relatively preserved. The findings of our meta-analysis contribute to a better understanding of the underlying pathophysiology of AN, the time course and residuals of brain structural alterations during recovery and clinical implications potentially relevant for more-targeted treatment options.

Figures