A PREDICTION EQUATION FOR ESTIMATING BODY FAT PERCENTAGE IN WOMEN USING SIMPLE MEASURES 1056

To develop a regression equation to predict body fat percentage (BF%) using simple measures, 200 white women, 20-65 yrs, were equally distributed across four age categories, and within each age category, three BMI categories. All were hydrostatically weighed and residual volumes, circumferences, and lifestyles were assessed. Estimation and cross-validation groups were formed randomly. A regression model was created using the estimation group and the equation was then cross-validated using the second group with excellent results: R=0.91, SEE=3.5%. Equations from the two groups were statistically equivalent, so the groups were combined and a final equation was developed using the total sample: BF%=0.001637 (Age2) + 1.087765 (A) + 0.582779(BMI) + 2.041608 (Hip) - 0.741206 (B) - 0.612696 (Neck) - 0.007792(Hip2) - 0.500325 (C) - 93.236853. Question A was “Compared to other women your age, how physically active are you?” with responses[1-7]: extremely (ex.) active, very (v.) active, active, in between, inactive, v. inactive, ex. inactive. Question B was “How careful are you concerning what you eat?” with responses [1-6]: ex. careful, v. careful, somewhat careful, not v. careful, rather laxed, v. laxed. Question C was“How difficult would it be for you to rapidly jog 2 miles without stopping?” with responses [1-7]: ex. difficult, v. difficult, difficult, in between, easy, v. easy, ex. easy. BMI: wt in kg. ÷ (ht in m.)2. Hip: circumference in cm. around the buttocks at the largest point. Neck: circumference in cm. taken at a right angle to the long axis of the neck just inferior to the laryngeal prominence. A simpler model was also created without the neck measure with virtually the same SEE. The findings indicate that BF% can be accurately estimated across a variety of white women using simple measures.