Skip to main content
letter
. 2020 Aug 28;21(11):1568–1569. doi: 10.1016/j.jamda.2020.08.032

Table 1.

Multilevel Model Coefficients for Predictors of Weight Loss among Nursing Home Residents

Parameters Estimate SE 95% CI Test Statistic P
Fixed effects
 Intercept 151.68 3.52 144.72, 158.64 43.06 .000
 Age −1.92 0.33 ‒2.58, −1.26 −5.74 .000
 Sex 23.74 6.90 10.11, 37.38 3.44 .001
 Mo −0.24 0.26 ‒0.76, 0.29 −0.89 .37
 NH restrictions −3.68 1.00 ‒5.65, −1.71 −3.68 .00
Variance components
 Intercept (1) 1393.94 164.18 1106.60, 1755.89 8.49 .000
 NH restrictions (2) 27.32 6.81 16.76, 44.54 4.01 .000
 Covariance (1, 2) −58.57 23.90 ‒106.41, −11.73 −2.45 .014
 Residual 20.91 1.72 17.80, 24.56 12.19 .000

CI, confidence interval; SE, standard error.

Age was centered around the mean of 86.91 years. Sex was coded as 0 (women) and 1 (men). Month was coded as 0 (December), 1 (January), 2 (February), and 4 (April). NH restrictions were coded as 0 (December/January/February) or 1 (April). Test statistics for fixed effects and variance components were t-tests and Wald z, respectively.

NH restrictions are the time periods where nursing home restrictions were implemented.