Interobserver reproducibility of single-slice computed tomography for quantifying visceral and subcutaneous abdominal fat in obese adults

Abstract Background Abdominal fat distribution plays a critical role in metabolic and surgical risk assessment. Although volumetric computed tomography (CT) is the gold standard for fat quantification, single-slice techniques are increasingly used in clinical practice due to their simplicity and efficiency. However, reproducibility varies across anatomical slice levels and remains understudied in obese adults. Objective To evaluate the interobserver reproducibility of single-slice CT measurements of visceral, subcutaneous, and total abdominal fat at the L3–L4 and L4–L5 levels in obese adults. Methods This cross-sectional study included 40 consecutive obese adults who underwent non-contrast abdominal CT. Two independent radiologists performed manual segmentation of visceral fat area (VFA), subcutaneous fat area (SFA), and total fat area (TFA) using standard Hounsfield unit thresholds. Reproducibility was assessed using intraclass correlation coefficients (ICC), concordance correlation coefficients (CCC), and Bland–Altman analysis. Results Interobserver agreement was excellent for VFA and SFA at both slice levels (ICC range: 0.982–0.997; CCC range: 0.981–0.998). TFA demonstrated excellent reproducibility at L4–L5 but only moderate-to-good reproducibility at L3–L4. Bland–Altman plots confirmed narrow limits of agreement for VFA and SFA, with wider dispersion for TFA at L3–L4. Conclusion Single-slice CT quantification of visceral and subcutaneous fat demonstrates excellent interobserver reproducibility in obese adults, particularly at the L4–L5 level. The L4–L5 slice may represent the most reliable anatomical site for standardized fat assessment in clinical and research settings.