|
androstenediol, point, gadgets, l glutamine, the zone diet, burning, cardiopathy, scale, tribex 500, accu measurefat loss design n/a, multi vitamin, nutritional supplements, bodymass index, ami accu measure fat tester by accu-fitness, used, bodyfat calculators, measure body fat, body fat monitors, bmi, fattrackgold digital body fat caliper and myotape combo,
|
between the general characteristics of the validation and cross-validation groups. To identify anthropometric measurements that best predicted BFM estimated by DXA, multiple linear regression analyses with backward elimination of independent variables were performed using a tolerance level of 0.3 as exclusion criteria to avoid co-linearity. Removal criteria were a significance badbreath level of >0.15 and a tolerance level of <0.3. Development of equations was based on different steps. 1) Preliminary hypotheses were built on existing skinfold equations and our own observations. 2) A dimension reduction was conducted based on preliminary regression models. badbreath SFs and other anthropometric variables were used in two separate regression models, and those not contributing significantly to the variance were excluded. SFs were then combined as sums without using the information obtained by linear regression models, and different sums of log-SFs were included in the final regression model with backward elimination.
|
