Skip to main content

Table 5 Marginal model and random-effects models analysing the influence of time and age

From: Choosing marginal or random-effects models for longitudinal binary responses: application to self-reported disability among older persons

 

Marginal model GEE

Random model likelihood integration

 

Estimate

SE

p value

Estimate

SE

p value

Intercept

-1.83

0.12

< 0.0001

-3.61

0.25

< 0.0001

Time (years)

0.12

0.03

< 0.0001

0.17

0.06

0.002

Age at baseline*

0.09

0.02

< 0.0001

0.16

0.03

< 0.0001

Age* × time

0.01

0.004

0.003

0.04

0.009

< 0.0001

Baseline disability

1.54

0.10

< 0.0001

3.18

0.22

< 0.0001

Death during the study

0.68

0.12

< 0.0001

1.36

0.26

< 0.0001

Random intercept variance

-

-

-

7.25

0.68

< 0.0001

Random slope variance

-

-

-

0.18

0.04

< 0.0001