Validity of body adiposity index in predicting body fat in Brazilians adults

OBJECTIVES: The aim of this study was to compare various methods of assessing body compositions with body adiposity index (BAI) and to identify the validity of BAI as a predictor of body fat in Brazilian adults. METHODS: This study included 706 individuals (average age 37.3 years, SD = 12.1). Anthropometric data included percent body fat obtained by skinfold thicknesses, bioelectrical impedance analysis and DXA. Body mass index (BMI), waist/hip ratio, and BAI were calculated. The correlation between variables was assessed by Pearson's correlation coefficient, and the Bland-Altman and Kaplan Meier graphic approaches were used to verify the agreement between BAI and DXA. RESULTS: There was a strong correlation between BAI and BMI (r = 0.84 in men and r = 0.86 in women, P < .001), waist circumference (r = 0.77 in men and r = 0.75 in women, P < .001) percent fat by skinfold thicknesses (r = 0.71 in men and r = 0.71 in women, P < .001) and by DXA (r = 0.72 in men and r = 0.78 in women, P < .001). The Bland-Altman approach showed an overestimation of BAI in males and an underestimation in women using DXA as the reference method. The agreement between BAI and DXA through the Kaplan-Meier analysis was 41%. CONCLUSIONS: It was found that BAI does not replace other measurements of body fat, but compared with more complex methods can be an alternative for estimating the body fat in the absence of these methods.