A Retrospective Study to Examine the Correlation of Bioelectrical Impedance Analysis with Shear-wave Elastography in Indian Patients with Non-alcoholic Fatty Liver Disease and Diabetes on Background Sodium-glucose Cotransporter-2 Inhibitor Therapy

Background Patients with non-alcoholic fatty liver disease (NAFLD) are often insulin resistant. Several recent studies show NAFLD to be associated with cardiovascular risk. Bioimpedance analysis (BIA) is a common approach for body composition measurements and is a noninvasive, low-cost modality. Shear-wave elastography (SWE) works using an acoustic radiation force pulse sequence that generates shear-waves that estimates the liver stiffness. Objectives The primary objective was to assess the correlation between SWE values and BIA values in an Indian population. The hypothesis is that with the increase in BIA value measuring visceral fat percentage, the SWE value measuring liver stiffness should increase. Materials and methods We conducted a retrospective analysis of clinic data of 36 patients properly screened from July 2018 to December 2018, who matched our prespecified inclusion criteria. Statistical analysis was performed using GraphPad Insta Version 3.0(®) using regression analysis. Visceral fat percentage and skeletal muscle percentage of lower limbs were calculated using an Omron HBF 375(®) analyzer. SWE values for liver fat were measured using a Philips Affinity 70(®) using two-dimensional imaging and expressed in kilopascal (kPa) units. Results We found that 88.88% of the patients with diabetes had above normal SWE values (2.0 to 4.5 kPa), and a corresponding 83.33% of patients had above the high cut-off for BIA values (>10%) but without any positive correlation between the two parameters as evident from the p-value of 0.079. Conclusions This study found a high prevalence of fat burden amongst our patients with type 2 diabetes and NAFLD. This is the first of its kind of study where we searched for a correlation between the two commonly used parameters in assessing the fat burden and liver stiffness of an individual but found there was no significant correlation between the two parameters used.