What is the best predictor of body fat percentage for older brazilian women?

This study aimed to investigate the validity of several equations and predictive indices for estimating body fat percentage anthropometric measurements including height, body weight, circumferences (waist and hip), and a dual energy X-ray absorptiometry (DXA) scan were performed. All measurements were performed by trained researchers following specific protocols. The results were compared to the dual energy X-ray absorptiometry (DXA) technique, which is considered the reference method. The analyzed equations showed moderate to good correlation coefficients with DXA, with particular emphasis on Visser's equation 6, which showed the best correlation (r = 0.752, p < 0.001). However, the agreement between the equations and DXA, as assessed by the Lin concordance coefficient, was classified as poor (rho c < 0.90). This indicates that although the equations have a positive correlation with body composition, they tend to deviate from the identity line when compared to the reference method. Additionally, the equations showed high sensitivity for detecting obesity when the cut-off point of 30% body fat was adopted, indicating a good ability to identify the presence of the condition. However, the equations, with the exception of equation 4, showed low specificity, meaning they had limited ability to detect normal individuals, resulting in a low negative predictive value. The results suggest that BF% equations and indices are dependent on the populations in which they were developed. The specificity and sensitivity of these equations may vary, and it is important to carefully select the most appropriate equation for estimating BF% in older Brazilian women.