4if é Measures of Body Fat and Related Factors in Normal Adults —I1 1299 | | | i Table 5 presents the regression equations using two, three and four variables for each category and combinations of categories. Weight is in pounds; skinfold thickness is in millimeters; other measurements are in centimeters. Incorporation of additional variables resulted in little further increase in the multiple correlation coefficients. Although the linearity of the relationships has been questioned [10], LTIPLE CORRELATIONS the size of the multiple correlation coefficients indicates that the non-linearity. gra| ol ay Multiple correlation with measured : — 39.408 0.920 x — 41.448 0.909 ngth 0.913 0.930 +23.173 0.972 0.977 0.980 0.946 0.963 +39.358 0.970 0.917 0.933 th ~ 37.294 0,937 — 29.836 0.914 0.922 0.927 0.914 0.924 — 20.832 0.929 0.905 0.913 — 36.165 0.918 Sey Je athe oo 0.937 wnateatinak a nedlern SRYee‘yale. SPERroad wpe Loyleh cag eyes 8 aie Ca. gth — 26.189 Sfatebn teste. — It - 0.930 eet 35.689 . 0.918 though perhapsstatistically significant, is substantively not important. TABLE 6. STANDARD ERROR OF ESTIMATE AND 95°5 CONFIDENCE INTERVALS FOR KILOGRAMS BODY FAT OF ‘AVERAGE’ INDIVIDUALS, BY CATEGORY AND NUMBER OF PREDICTOR VARIABLES Category I Il No. of variables 2 3 4 2 1V Vv S.E. Mean 95°% confidence interval . Individual Population piean 2.635 2.402 2840 i7141128-2300 17.14 17.14 11.69-22.59 12.17-22.11 16.16-18.12_ 2.761 20.67 14.96-26.38 19.70—21.64 2.683 2.557 20.67 15.37-25.97 19.77-21.57 2 3.037 19.00 12.70-25.30 17.92-20.08 3 4 2 3 2.792 2.690 2.337 2.151 20.67 19,00 19,00 23,30 23.30 15.11-26.23 16.23-18.05 16.31-17.97 3 4 III Body fat in kg 19.73-21.61 13.20-24.80 13.41-24.59 18.43-28. 17 18.81-27.79 18.00-20.00 18.04—19.96 22.44-24.16 22,.51—-24,09 4 2.044 23.30 19,02—27.458 22.54-24.06 2 3.776 18.58 10.76—26.40 £7.26-19,90 4 3.665 18.58 10.98-26.18 17.29-19.87 3 3.747 18.58 10.8 1-26.35 17.27-19.89 tes ngeth — 31.268 0.959 A ala SF. ~ 45.464 Pady whe eet REEeb. +ni, » Pad,2 2a ax SFE. ie ayh yates ‘ea!aieSSL SRSe fatin ke Table 6 illustrates the precision of estimation of the equations by considering, in each category, a hypothetical ‘average’ individual, each of whose anthropometric measurements is exactly equal to the mean for his category. In each case the 95°. confidence interval [11] for the estimate of kilograms of body fat of an individual is given, as well as the 95% confidence interval for the estimate of the mean value for kilograms of body fat for a population in which all individuals had measure- ments equal to the mean forhis category. ESTIMATION OF LEAN BODY WEIGHT CONTENT FROM K-40 Means, standard deviations and ranges for K*® measurements in counts/min and the measurements derived therefrom appear in Table 7. LBW**° and LBM for subjects studied with both the whole-body counting and TBW-body density techniques are also presented. (The body weight as measured at the time of laboratory study was used for the calculation of LBM for this table. For Table 3, LBM was calculated from the body weight taken for anthropometric measurements. The largest discrepancy between these two measurements for any one individual was 0.7 kg). | if ' i \ oeat t necessarily those able after the first

Select target paragraph3