Table 2: Multivariable linear regression for interlimb volume difference outcome.

Coefficient estimateStandard error

Race (black ref.)−0.261.790.89
Body mass index
Age in years at study entry0.180.120.13
Education: some college versus HS grad or less−4.892.240.03
Education: college graduate versus high grad or less−2.492.310.28
Hours/week of paid outside work−
Number of people living in home−0.760.660.25
Number of dependent children at home1.341.990.50
Cancer stage at diagnosis1.120.920.23
Years since cancer treatment0.550.16<0.001
Axillary lymph node dissection4.751.890.01
Radiation therapy0.042.370.99
Hormone therapy−0.441.690.79
Breast reconstruction−0.231.890.91
>25% lymphedema care adherence4.101.620.01

Italicized items in table represent statistically significant covariates at .
Coefficient estimates represent the change in interlimb difference percentage expected with a “yes” response for categorical variables or per unit change for continuous variables.