Comparative Adiposity Study of a Body Shape Index and Body Adiposity Index with BMI using DEXA as Gold Standard

Abstract

Introduction: In today's world, rising obesity epidemic is one of the biggest healthcare challenges which subsequently contributes towards other significant health issues like diabetes, musculoskeletal diseases certain cancers and cardiovascular diseases. Various anthropometric measures like BMI, body adiposity index(BAI), ABSI has been devised for estimating the body fat. Objective: This current study was conducted to assess the agreement of a body shape index (ABSI) and body adiposity index (BAI) measurement methods with body mass index (BMI) using DEXA as gold standard for estimating the adiposity of the body. Methodology: A cohort study was conducted on a total of 40 subjects belonging to the age group of 25-45 years. The subjects underwent a 24-hour dietary recall. Data pertaining to frequency of food intake, general health, social habits and lifestyle were collected through a questionnaire. Routine demographic details were collected and measurements of height, weight, waist circumference and hip circumference were performed using the SECA scale. Biochemical parameters evaluated includes: lipid profile and blood sugar levels. The subjects were screened for the risk of cardiovascular disease and diabetes. Body adiposity measures like BMI, BAI, ABSI and DEXA were calculated to estimate the body's adiposity. The statistical analysis of the data was performed with Medcalc software. Results: Correlation studies revealed gender (r-0.727, P <0.001, CI-0.847 to-0.537), BMI (r 0.76, P <0.001, CI 0.592 to 0.868), BAI (r 0.69, P <0.001, CI 0.482 to 0.824), ABSI (r-0.37, P 0.019, CI-0.611 to-0.066) and region fat percentile (r 0.99, P <0.001, CI 0.99 to 0.99) as significantly correlated with DEXA tissue fat percentile. A strong positive correlation was also observed between DEXA tissue percentile and measurements like BMI, BAI and region fat percentile, whereas the correlation with ABSI was moderate inverse. Insignificant correlation was attained with variables like age and bone mineral content. However, the inter-rater agreement statistics performed on classified variables showed that BMI possessed significant moderate agreement (K = 0.54) and BAI had poor agreement (K = 0.14) with DEXA tissue fat percentile. Additionally, mountain plot analysis revealed a positive bias of 9.35 between BMI and DEXA tissue fat (%) (limits of agreement-5.89 to 25.72), and a bias of 4.44 for BAI, with more variability (limits of agreement between-13.58 and 22.48) compared to BMI. Conclusion: From the current study we can conclude that, BMI, BAI and ABSI underestimate the body adiposity when compared to DEXA tissue fat %. Additionally, BAI also exhibited a poor agreement with reference when assessed as a classified measure, compared to BMI.