Longitudinal beta regression models for analyzing health-related quality of life scores over time

Table 3 Parameter estimates of LMM, beta GLMM and beta GEE in the ICF stroke data (N = 517)

Parameter coefficients	LMM		Beta GLMM		Beta GEE
	Estimate	p value	Estimate	p value	Estimate	p value
Intercept	0.4955	<0.0001	−0.0106	0.9569	−0.0766	0.6828
Age (centered)	−0.0031	0.0060	−0.0226	0.0015	−0.0162	0.0137
Male sex	0.0528	0.0599	0.2931	0.0991	0.3200	0.0574
Time 2	0.0801	0.0003	0.4637	0.0005	0.3343	0.0010
Time 3	0.1597	<0.0001	0.9254	<0.0001	0.6792	<0.0001
Phase D	0.2941	<0.0001	1.6160	<0.0001	1.4316	<0.0001
Phase D*Time2	−0.0463	0.0807	−0.2005	0.2240	−0.0832	0.5154
Phase D*Time 3	−0.1178	<0.0001	−0.5465	0.0017	−0.3573	0.0477
σ²	0.0126
ϕ			10.80
Variance of random effects	Estimate	SE	Estimate	SE	Estimate	SE
Variance	0.0325	0.0038	1.2782	0.1632
Covariance estimates
Variance					0.0764	0.0061
Compound symmetry					0.1928	0.0230
Scale					0.0514
Fit statistics
−2LogL	−396.4		−924.1		-
AIC	−376.4		−904.1		-
BIC	−343.6		−871.3		-
Pseudo-R²†	0.2277		0.7217		-

LMM Linear mixed model, GLMM Generalized linear mixed model, GEE Generalized estimating equations, SE Standard error, CS Compound Symmetry, AIC Akaike information criterion, BIC Bayesian information criterion.
†Compared to linear random-intercept model with -2LogL = −262.8.

ISSN: 1471-2288