Skip to main content
. 2011 Apr 15;66A(7):801–808. doi: 10.1093/gerona/glr059

Table 3.

Results of Linear Regression Models of Body Composition Measures (Year 1) as Independent Variables and FEV1 (mL) (Year 1) as the Dependent Variable

N = 1893 Model 1 Model 2 Model 3
Beta (SE) p Value Beta (SE) p Value Beta (SE) p Value
Total hip bone mineral density (gm/cm2) 1246.12 (83.60) <.001 210.75 (78.44) .007
Fat mass (kg) −9.29 (1.72) <.001 −2.12 (1.34) .113
Subcutaneous adipose tissue (cm2) −1.50 (0.13) <.001 −0.05 (0.11) .669
Visceral adipose tissue (cm2) 0.73 (0.23) .001 −1.14 (0.17) <.001 −1.66 (0.18) <.001
Total thigh muscle area (cm2) 5.37 (0.23) <.001 1.43 (0.31) <.001 1.87 (0.35) <.001
Total thigh intermuscular fat area (cm2) −3.58 (1.23) .004 −2.17 (0.92) .018
Height (cm) 43.05 (1.24) <.001 29.00 (1.68) <.001 27.77 (1.75) <.001

Notes: Model 1: Body composition measures were entered in separate unadjusted models. Model 2: Model 1 + clinic site, age, sex, race, FEV1 or FVC quality score, and height (CT and DEXA measures only). Model 3: Model 2 + remaining body composition measures, smoking, prevalent cardiovascular disease, osteoporosis, emphysema, and chronic bronchitis. Y1 FEV1 (N = 88) or Y1 FVC (N = 196) of quality score <2 excluded. Model 3 results provided for variables retained in forward selection procedure (p-entry = .20).