Table 2.

Linear mixed-effect modelling for CSO-PVS and BG-PVS in cognitively unimpaired subjects. Models show different trajectories over time, effects of age, sex and years of education. Models with correlated slope and random intercept: PVS ~ age + age² + time + sex + years of education + total intracranial volume + (1 + time | subject)

	CSO-PVS				BG-PVS
	B	SE	CI	p	B	SE	CI	p
(Intercept)	0.03	0.09	-0.14 – 0.20	0.733	0.09	0.09	-0.08 – 0.26	0.307
age (linear)	0.12	0.04	0.02 – 0.22	0.017	0.19	0.05	0.09 – 0.28	< 0.001
age (quadratic)	-0.13	0.04	-0.22 – -0.04	0.006	-0.09	0.05	-0.18 – -0.00	0.047
time	0.04	0.01	0.02 – 0.05	< 0.001	0.05	0.01	0.03 – 0.07	< 0.001
sex	0.18	0.08	-0.07 – 0.43	0.166	0.12	0.12	-0.25 – 0.24	0.959
years of education	-0.01	0.04	-0.11 – 0.08	0.830	-0.01	0.05	-0.10 – 0.08	0.831
sbTIV	0.23	0.06	0.10 – 0.35	< 0.001	0.09	0.06	-0.03 – 0.21	0.150
σ2	0.09				0.18
τ00	0.74 Subject				0.69 Subject
τ11	0.01 Subject.time				0.00 Subject.time
ρ01	0.01Subject				-0.25Subject
ICC	0.89				0.79
N	383 Subject				387 Subject
Observations	1364				1381
Marginal R2 / Conditional R2	0.066 / 0.901				0.058 / 0.804

Annotations. σ² = residual variance; τ₀₀ = random intercept variance; τ₁₁ = random slope variance; ρ₀₁ = covariance between random slope and intercept