A new anthropometric index for body fat estimation in patients with severe obesity

Body mass index (BMI) has been used to assess body adiposity, but it cannot adequately reflect body fat (BF) amount. The body adiposity index (BAI) has been shown a better performance than BMI for this purpose, but it can be inaccurate to estimate BF under extreme amounts of fat. Here, we propose a new anthropometric index, the Belarmino-Waitzberg (BeW) index, for specific estimation of BF in severely obese patients.

In 144 adult patients with severe obesity, BF was estimated by air displacement plethysmography (ADP), as the reference method, along with the follow anthropometric measurements: height, abdominal circumference (AC), hip circumference (HC), weight, BMI (weight/ height²) and BAI ([HC(cm) / height (m)^1.5) - 18] × 100). Patients were proportionately distributed into two distinct databases, the building model database (BMD) and the validation model database (VMD), which were applied to develop and validate the BeW index, respectively. The BeW index was tested for gender and ethnicity adjustment as independent variables. The agreement of BF% values obtained by the new index and by BAI with ADP was also assessed.

The BF% was 52.05 ± 5.42 for ADP and 59.11 ± 5.95 for the BeW index (all results are expressed as the mean ± standard deviation). A positive Pearson correlation (r = 0.74), a good accuracy (Cb = 0.94), and a positive Lin's concordance correlation (CCC = 0.70) were observed between the two groups. The 95% limits of individual agreement between the BeW index and ADP were 6.8 to 7.9%, compared to - 7.5 to 14.8% between the BAI and ADP. The new index, called the Belarmino-Waitzberg (BeW) index, showed an improvement of 2.1% for the R² value and a significant gender effect, therefore resulting in two different indexes for females and males, as follows: Female BeW = - 48.8 + 0.087 × AC(cm) + 1.147 × HC(cm) - 0.003 × HC(cm)2 and Male BeW = - 48.8 + 0.087 × AC(cm) + 1.147 × HC(cm) - 0.003 × HC(cm)2-7.195.

The new BeW index showed a good performance for BF estimation in patients with severe obesity and can be superior to the BAI for this purpose.