TABLE 2.

Models showing significant associations between plasma and neuroimaging biomarkers before and after covariate adjustment and multiple‐comparisons correction.

	NfL				GFAP				Aβ42/40				p‐tau181				PLASMA SUM
	Base	Adj	F‐Adj	F‐Adj‐Dx	Base	Adj	F‐Adj	F‐Adj‐Dx	Base	Adj	F‐Adj	F‐Adj‐Dx	Base	Adj	F‐Adj	F‐Adj‐Dx	Base	Adj	F‐Adj	F‐Adj‐Dx
Aβ‐PET	+	+			+	+	+	+	–				+	+	+	+	+	+	+	+
BVOL	–	–	–	–	–	–	–	–	+	+			–	–	–	–	–	–	–	–
WMH	+	+			+	+	+	+					+	+			+	+	+	+
FA	–	–			–	–	–	–					–				–	–
FW	+	+			+	+	+	+					+	+			+	+

Note: Base (unadjusted) and adjusted models associating plasma and neuroimaging‐derived biomarkers are presented. ±, plasma biomarker coefficient significant at p < 0.05, (+) positive association and (−) negative association.

Abbreviations: Adj, linear models adjusted for comorbidities (eGFR, BMI); Aβ, amyloid beta; base, baseline unadjusted models; BVOL, brain volume (adjusted for intracranial volume); CU, cognitively unimpaired; DEM, dementia; DWI, diffusion weighted imaging; Dx, diagnosis; FA, fractional anisotropy; F‐Adj, adjusted models plus additional covariates (age, sex, race, education, APOE ε4 status, medications). F‐Adj‐Dx., fully adjusted models plus clinical diagnosis; FW, free water; GFAP, glial fibrillary acidic protein; logWMH, log‐transformed global white matter hyperintensity values; MCI, mild cognitive impairment; NfL, neurofilament light chain; NODDI, neurite orientation dispersion and density imaging; SUVr, standardized uptake value ratio.