P- 52 CORRELATION AMONG DIFFERENT METHODS TO ESTIMATE BODY AND LIVER FAT IN HEALTHY ADULTS

An increase in body fat is a risk factor to develop fatty liver disease. Until now, few studies have related different methods to estimate body fat related to liver fat. We aimed to determinate the correlation among different methods to estimate body and liver fat in healthy adults. In a cross- sectional study registered in ethics and research committees GAS3794 evaluating healthy adults; chronic illness, uncontrolled diabetes, hypertension or thyroid, bariatric surgery or pacemaker, were not included. The estimation of body fat (BF) was by 3 methods, two with bioelectric impedance of 19 frequencies (19F), 4 frequencies (4F) and the third by the Durnin- Womersley (DW) skinfold thickness formulae; also, were measured body mass index (BMI), waist circumference (WC) and visceral fat (VF). The liver fat was estimated by controlled attenuation parameter (CAP) using transitory elastography. Correlations were calculated with Pearson coefficient among body composition methods and CAP; each anthropometric isolated parameter was associated with hepatic steatosis grades by a logistic regression analysis using SPSS v21.0. In 231 participants, mean age was 41.8 years (SD 11.3), WC 91 cm (SD 12), BMI 27.8 kg/m2 (SD 4.6). 112 had some grade of steatosis (S3 n=72, S2 n=18). The correlation among 3 methods was on average r= 0.853 (p=0.000), and between CAP and BF was 0.290 (p= 0.000). BMI, WC, VF and suprailiac skinfold thickness showed correlations of r=0.570 (p=0.000), r=0.477 (p=0.000), r=0.393 (p= 0.000) y r= 0.471 (p=0.000) respectively. Regression analysis demonstrated that BMI (OR=1.32, p=0.000), WC > 80 women and > 90 cm men (OR=14.7, p=0.010), VF (OR=1.8, p=0.008) and suprailiac skinfold thickness (OR=1.14, p=0.000) showed association with steatosis. Three methods were like to estimate body fat although they were not able to represent the liver fat; waist circumference was the best indicator related with steatosis