Skip to main content

Table 2 Trajectory of functional limitation prior to versus after cancer onset among participants with cancer (n = 139)

From: Functional limitations before and after cancer diagnosis and contributing factors: findings from the China health and retirement longitudinal study

 

Model 2

Model 3.1

Model 3.2

 

Estimate (se)

Estimate (se)

Estimate (se)

Fixed Effects

 Intercept, β00

0.37 (0.97)***

2.26 (0.35)***

2.21 (0.35)***

 Time to/from cancer diagnosis (CTF), β10

0.43 (0.22)*

0.48 (0.22)*

0.45 (0.22)*

 Time-varying cancer diagnosis, β20

-0.39 (0.36)

-0.48 (0.36)

-0.43 (0.36)

 CTF × time-varying cancer diagnosis, β30

-0.48 (0.24)*

-0.50 (0.24)*

-0.48 (0.24)*

Demographic characteristics at baseline

 Age, β01

0.03 (0.01)***

0.04 (0.01)***

0.04 (0.01)***

 Male, β02a

-0.90 (0.25)*

-0.63 (0.22)***

-0.63 (0.22)***

 Education, β03

0.20 (0.43)

-

-

 Marital status, β04

-0.01 (0.05)

-

-

Contributing factors for disablement-

 Pain, β05

-

0.46 (0.08)***

0.93 (0.28)***

 Self-reported memory problems, β06

-

0.47 (0.17)**

0.49 (0.17)**

 Have fallen down in the last two years, β07

-

0.71 (0.28)**

0.72 (0.28)**

 CTF × Pain, β11

-

-

0.30 (0.18)

 Time-varying cancer diagnosis × Pain, β21

-

-

-0.42 (0.29)

 CTF × Time-varying cancer diagnosis × Pain, β31

-

-

-0.39 (0.19)*

Random Effects

 Intercept Variance,\({\sigma }_{\upsilon }^{2}\)

1.21 (0.22)***

0.81 (0.16)***

0.80 (0.16)***

 Residual Variance,\({\sigma }_{e}^{2}\)

1.48 (0.13)***

1.39 (0.12)***

1.39 (0.12)***

Model Fit Index

 AIC

1498.1

1406.1

1408.7

 BIC

1504.0

1411.8

1414.5

 -2 Res Loglikelihood

1494.1

1402.1

1404.7

  1. Notes. ***p < 0.001, **p < .0.01, *p < 0.05
  2. Analysis was based on N = 417 observations from n = 139 participants with cancer diagnosis over 3 waves. Demographic characteristics were measured at baseline and time-invariant; all contributing factors for disablement were measured over time and time-varying except for pain which was measured at wave 2
  3. aModel 2 included demographic covariates. Specifically, the gender variable was dummy coded as male = 1 and female = 0 and used as is for Model 2, however, it was centered at the sample mean for Models 3.1–2
  4. Models 3.1–2 included demographic and disablement covariates, where non-significant main effects and interaction terms were trimmed off for model parsimony. The outcome was measured in the original units. All the estimates are unstandardized regression coefficients that correspond to the change in Y relative to a one-unit increase in X (independent variable)