Is slice thickness from computed tomography relevant for body composition analysis?

Computed tomography (CT) is a reference method to assess body composition at the tissue level. Despite its accuracy, CT's technical parameters are less discussed. Therefore, this study aimed to evaluate the impact of different CT slice thicknesses on body composition measurements.

A cross-sectional study including healthy and young subjects (20-45 years). CT images were evaluated under a standardized protocol at five different slice thicknesses for each participant (0.6, 1.25, 2.5, 5, and 10 mm). Tissue analysis was performed using the semi-automatic mode of SliceOmatic® software (TomoVision). Paired-measures ANOVA with Tukey's post-hoc test evaluated differences between slice thicknesses.

A total of 550 slices (from n = 110 individuals) were analyzed (56.4% females). Radiodensity differences were observed between slice thicknesses of the intermuscular and visceral adipose tissue (DIMAT and DVAT in HU) and for the area of intermuscular adipose tissue (IMAT in cm²). Most relevant differences occurred between the thinner slices (0.6 and 1.25 mm) vs. the thickest slice (10 mm), all _adjustedp-values < 0.001. Differences in skeletal muscle area were not significant in any slice thickness. These results were consistent across sex- and BMI-stratified analyses.

Slice thickness may influence the radiodensity of adipose compartments, potentially affecting the analysis of tissues' composition. Although it remains unclear whether these differences are clinically relevant, our findings reinforce the need for operational standardization for CT image acquisition parameters.