Table 3.

Relationships between nEV Aβ₄₂ and AV45 SUVR

	β	Standard error	Standard β	t	p
The plasma nEV Aβ42 term alone explained 41.1% variation in average AV45 uptake
Intercept	1.109	0.011		104.429	< 0.001
nEV Aβ42	0.028	0.003	0.641	10.921	< 0.001
The plasma nEV Aβ42 plus clinical features explained 46.4% variation in average AV45 uptake
Intercept	0.939	0.071		13.225	< 0.001
nEV Aβ42	0.025	0.003	0.573	9.695	< 0.001
Age	0.003	0.001	0.154	2.691	0.008
Female	− 0.011	0.015	− 0.043	− 0.763	0.447
APOE ε4 status	0.041	0.015	0.165	2.819	0.005
The plasma nEV Aβ42 plus clinical features and the interaction term explained 46.6% variation in average AV45 uptake
Intercept	0.915	0.077		11.986	< 0.001
nEV Aβ42	0.029	0.005	0.651	5.986	< 0.001
Age	0.003	0.001	0.166	2.815	0.005
Female	− 0.01	0.015	− 0.039	− 0.691	0.491
APOE ε4 status	0.055	0.022	0.221	2.517	0.013
nEV Aβ42 * APOE ε4 status	− 0.005	0.006	− 0.121	− 0.858	0.392

The analysis was performed in total individuals including Aβ− NCs, Aβ+ NCs, and patients with aMCI and ADD. In the first model, nEV Aβ₄₂ was used as a predictor of AV45 SUVR; in the second model, nEV Aβ₄₂ plus age, sex, and APOE ε4 status were used as predictors of AV45 SUVR; in the third model, the interaction term between nEV Aβ₄₂ and APOE ε4 status was additionally included

Abbreviations: Aβ β-amyloid, NCs cognitively normal controls, aMCI amnestic mild cognitive impairment, ADD Alzheimer’s disease dementia, APOE apolipoprotein E, nEV neuronal-derived extracellular vesicle, AV45 [¹⁸F]florbetapir