Abstract
<title>Abstract</title> <p>Background: Obesity is metabolically heterogeneous, and conventional adiposity measures mainly quantify fat amount rather than adipose tissue quality. This study evaluated computed tomography (CT)-derived relative fat density (RFD), an individualized ratio of visceral to subcutaneous fat attenuation, for metabolic risk stratification in adults with obesity. Methods: This retrospective cross-sectional study included adults who underwent routine health examinations between June 2024 and July 2025. Visceral fat area (VFA), subcutaneous fat area (SAT), visceral fat attenuation density (FAD), and subcutaneous adipose density (SAD) were measured on non-contrast abdominal CT images at the L3 level. Sex-specific VFA cutoffs for defining obesity and sex-specific RFD cutoffs were determined using receiver operating characteristic (ROC) curve analysis with the maximum Youden index. Restricted cubic spline (RCS), DeLong tests, and multivariable logistic regression were used to evaluate the association between RFD and metabolic abnormality (MA). Results: Among 1,426 participants, 698 (48.9%) were classified as having obesity based on sex-specific VFA cutoffs (123.173 cm² for men and 64.87 cm² for women). In adults with obesity, the optimal RFD cutoffs were 99.599 in men and 95.356 in women. RFD showed better discriminative performance for MA than FAD in the overall obese population (AUC 0.657 vs. 0.574; DeLong P < 0.001) and in obese men (AUC 0.609 vs. 0.509; P = 0.034). High RFD was independently associated with MA after adjustment for age, VFA, and BMI (OR = 2.153, 95% CI: 1.565–2.963; P < 0.001). MA was more frequent in the high-RFD group than in the low-RFD group (64.53% vs. 37.57%), with a stronger stratification effect in women. Conclusions: RFD is independently associated with metabolic abnormality in adults with obesity and may provide individualized, sex-specific risk stratification beyond conventional fat attenuation measures.</p>