Table 2.

Linear mixed models of plasma and neuroimaging biomarkers predicting MMSE score change in MCI. The first columns represent the univariate model with each biomarker modelled individually, and the last columns represent the model comprising all the biomarkers. All the models included age, sex, and years of education as fixed factors

Predictors	Univariate model with one biomarker				Multivariate model with all biomarkers (AIC = 621)			Multivariate model with significant predictors (AIC = 715)
	Estimate	Std. Error	p	AIC	Estimate	Std. Error	p	Estimate	Std. Error	p
Plasma biomarkers
p-Tau181	-0.15	0.17	0.38	944	0.59	0.32	0.06
p-Tau231	-0.19	0.29	0.28	957	-0.15	0.38	0.69
Aβ42/Aβ40	0.34	0.18	0.05	1002	-0.12	0.22	0.59
GFAP	-0.38	0.16	0.01	996	0.18	0.29	0.53
NfL	-0.43	0.16	0.006	1000	-0.8	0.31	0.009	-0.42	0.17	0.02
Neuroimaging biomarkers
Hippocampal volume	0.44	0.18	0.01	921	0.17	0.25	0.49
Centiloid	-0.38	0.19	0.04	874	-0.06	0.3	0.84
Tau-SUVR	-0.66	0.19	0.001	715	-0.71	0.31	0.02	-0.55	0.17	0.002