Density of ectopic fat depots predict distinct biomarkers of glycemic and insulinemic status in persons with HIV

Abstract

Background: HIV and obesity are conditions of impaired lipid storage where ectopic lipid accumulates in organs and tissues, promoting glucose intolerance and insulin resistance. Persons with HIV (PWH) are at high risk for diabetes, and one indicator of risk is the density of organs and tissues involved in glucose metabolism, which reflects ectopic lipid content and can be quantified using CT-tissue attenuation. We investigated relationships between subcutaneous adipose (SAT), visceral adipose (VAT), liver, pancreas, and skeletal muscle densities with biomarkers of glycemic/insulinemic status.

Methods: Demographic, anthropometric, and clinical data were utilized with automated segmentation of CT morphometric data from images acquired at the 3rd lumbar vertebra level in PWH who had normoglycemia, prediabetes, and T2DM.

Results: Of 217 PWH, 29.0% had prediabetes and 30.4% had T2DM. Liver, pancreas, and skeletal muscle densities were lower, and SAT density was higher, in PWH with T2DM. No differences were observed for VAT density. Receiver operating curves adjusted for age, sex and BMI showed tissue densities had similar ability to discriminate glycemic/insulinemic status. Adjusted multivariable logistic regression showed higher SAT density associated with higher glucose (p = 0.002), HbA1c (p < 0.001), and diabetes status (p < 0.001). Lower liver density is associated with diabetes status (p = 0.007) and higher HbA1c (p = 0.03), whereas lower skeletal muscle density is associated with higher glucose (p = 0.03) and insulin (p = 0.04).

Conclusions: Tissue densities, which differed significantly among the three groups, were robustly associated with various biomarkers of glycemic/insulinemic status. CT-morphometrics may enhance the detection of metabolic perturbations and diabetes risk, possibly earlier than some clinical biomarkers.

Fig. 1. Heatmap displaying univariate relationships between demographic, clinical, and morphometric factors.

Associations are presented as Spearman’s correlations with shades of red indicating positive correlations and shades of blue indicating negative correlations.

Fig. 2. Comparisons of receiver operating curves for CT-quantified densities for abdominal subcutaneous adipose tissue, visceral adipose tissue, abdominal skeletal muscle, liver and pancreas.

Each figure (a–e) presents a different glycemic/insulinemic outcome.