Concordance between dual indirect methods for assessing fat percentage.

Abstract

To evaluate the concordance between two low-cost and easily accessible double indirect methods, which have been used indistinctly in different studies where access to more accurate methods is not available, and to determine fat percentage and its relationship with age, sex, body mass index (BMI), waist circumference, level of physical activity and sedentary hours. Twenty-four persons between 18 and 38 years and 28 between 39 and 59 years from a university community were evaluated. Calculations were made: BMI, fat % was estimated by anthropometry with a digital adipometer (Skyndex System I USA) and by Electrical Bioimpedance Analysis - BIA (Biody Expert ZM II FRA), physical activity level and sedentary hours were determined with the short IPAQ questionnaire. Pearson's correlation coefficient, Bland & Altman's graphical method and Lin's concordance correlation index were calculated. The significance level p<0.05 was estimated. The fat percentage by anthropometry was: 30.5% ±8.5 (18-38 years) 35.0% ±6.7 (39-59 years); by BIA 27.0% ±7.3 (18-38 years) and 29.2% ±7.0 (39-59 years). Both techniques showed good correlation, but low degree of concordance (Lin index less than 0.9) except for the group of young persons with moderate level of physical activity (0.95). The doubly indirect methods used in the study showed strong correlation, but low concordance, so their use is not recommended indistinctly for the follow-up of a particular case. According to the study data for this specific population in young people with moderate physical activity, follow-up could be performed with either of the two methods.