RESUMO
BACKGROUND: The body adiposity index (BAI) comprises a simple method for estimating body fat (BF) that needs to be validated in patients with severe obesity. The present study aimed to determine BAI accuracy with respect to the determination BF in patients with severe obesity. METHODS: A cross-sectional prospective study comparing two methods for BF estimation was conducted in 433 patients with severe obesity between August 2012 to December 2014. BF was estimated by bioelectrical impedance analysis (BIA) with specific equations developed for BF estimation in patients with severe obesity and BAI. The BF estimation in 240 patients with severe obesity (Group 1: G1) was used to evaluate BAI limitations and to develop a specific equation in this population. The new equation proposed was validated in another 158 patients with severe obesity (Group 2: G2). RESULTS: There was a significant difference between BF determination by BIA and BAI (P = 0.039). The mean (SD) BF in G1 was 52.3% (6.1%) determined by BIA and 51.6% (8.1%) determined by BAI. Sex, waist-hip ratio (WHR) and obesity grade determined significant errors on BF estimation by BAI. A new equation (modified body adiposity index; MBAI) was developed by linear regression to minimise these errors [MBAI% = 23.6 + 0.5 × (BAI); add 2.2 if body mass index ≥ 50 kg m-2 and 2.4 if WHR ≥ 1.05]. The new equation reduced the difference [1.2% (5.9%), P < 0.001 to 0.4% (4.12%), P = 0.315] and improved the correlation (0.6-0.7) between methods. CONCLUSIONS: BAI present significant limitations in severe obesity and MBAI was effective for BF estimation in this population.
