Atlas‐Based Mapping of Hepatic Fat by Sex: Results From the Lifestyle Intervention ( LION ) Study
Year: 2026
Type: article
Abstract: ABSTRACT Background Hepatic proton density fat fraction (PDFF) is often reported as a whole‐liver mean, without a population reference for individual fat distribution. Purpose To construct longitudinal, sex‐specific three‐dimensional PDFF atlases and assess regional fat loss, sex differences in heterogeneity, and metabolic associations. Study Type Prospective; secondary analysis of a randomized lifestyle intervention trial. Population A total of 127 adults with obesity imaged at baseline (, 73 female [F], 54 male [M]), after an 8‐week low‐caloric diet (, 49F/32M), and after 12‐month weight maintenance (, 40F/32M). Field Strength/Sequence 3T; six‐echo chemical shift‐encoded gradient‐echo. Assessment Sex‐ and visit‐specific atlases were built by deformable registration; z ‐maps quantified voxel‐wise deviation from the matched atlas. Six heterogeneity features (standard deviation [ σ x ], coefficient of variation [CV], entropy []; z ‐ skewness, z ‐kurtosis, volume fraction z > 2) were compared between sexes and correlated, cross‐sectionally and longitudinally, with hepatocellular, glycemic, adiposity markers and triglycerides, and longitudinally with steatosis remission (PDFF < 5%). Statistical Tests Welch t‐test; Spearman correlation ( ρ ); Benjamini–Hochberg correction within each test set. Significance at adjusted p < 0.05. Results Liver PDFF decreased significantly from to in males (13.6% ± 7.0% to 6.6% ± 2.5%) and females (9.1% ± 6.2% to 5.9% ± 3.3%) with regionally non‐uniform, sex‐specific patterns. At , σ x and H were significantly higher in males and CV, z ‐skewness and z ‐kurtosis in females; at , σ x (males) and z‐skewness and z‐kurtosis (females) remained significant; no feature differed at (adjusted p = [0.583–0.953]). Cross‐sectionally, features correlated with hepatocellular, glycemic and adiposity markers, more extensively in females. Longitudinally, significant correlations were almost limited to adiposity ( V 1 to V 3 : CV × visceral fat volume in females, ρ = −0.65; CV × visceral PDFF in males, ρ = −0.67); apart from × gamma‐glutamyl transferase in females, no feature tracked hepatocellular, glycemic or steatosis remission outcomes (adjusted p = [0.051–0.999], [0.092–0.999], and [0.073–0.887]). Data Conclusion Hepatic fat loss showed spatially heterogeneous and sex‐specific patterns. Heterogeneity reflected metabolic state cross‐sectionally but mainly adiposity change longitudinally. Evidence Level 2. Technical Efficacy Stage 1.
DOI: 10.1002/jmri.70624
Language: English
Subfield: Hepatology
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
Estimated APC paid: $3,680
OpenAlex ID: W7226039342
