Skip to main content
. Author manuscript; available in PMC: 2009 Mar 31.
Published in final edited form as: Diabetes Care. 2001 Oct;24(10):1711–1721. doi: 10.2337/diacare.24.10.1711

Table 4.

Body composition characteristics associated with major weight gain

Characteristic Major weight gain Treatment group n First quartile Median Third quartile Adjusted mean* P value for treatment group P value for weight gain
Percent body fat
   Men Yes INT 38 24.5 27.1 31.7 25.9 <0.0001
CON 5 23.0 24.7 28.7 24.3 0.3733
No INT 260 15.1 19.2 22.7 17.8 0.1479
CON 332 14.7 18.4 23.2 17.2 0.0076
   Women§ Yes INT 67 36.8 38.9 41.6 39.8 <0.0001
CON 9 36.3 38.1 40.8 40.1 0.5958
No INT 220 27.0 30.7 34.0 32.3 0.2342
CON 249 26.8 30.3 33.9 31.9 <0.0001
Fat-free mass (kg)
   Men Yes INT 38 65.8 69.4 78.5 68.8 <0.0001
CON 5 67.6 67.9 73.7 66.9 0.7189
No INT 260 60.4 65.3 70.6 63.2 <0.0001
CON 332 60.3 64.8 70.9 61.6 0.0131
   Women Yes INT 67 47.8 53.3 56.5 54.9 <0.0001
CON 9 47.4 50.9 54.8 53.8 0.2537
No INT 220 43.1 45.7 49.5 51.0 0.0019
CON 249 42.4 45.3 48.7 50.1 0.0005
*

Least-squares estimates of cell means after adjusting for covariance with weight at baseline

from the Wilcoxon’s test of differences between treatment groups within weight-gain categories after adjusting for covariance with weight at baseline

from the Wilcoxon’s test of differences between patients with and without major weight gain within treatment groups after adjusting for covariance with weight at baseline

§

includes all women randomized as adults. Excluding women who became pregnant during the trial yields substantially identical results. INT, intensive; CON, conventional.