Skip to main content
. 2023 Oct 19;14:1257896. doi: 10.3389/fneur.2023.1257896

Table 3.

Multivariate analysis of the association between HA burden and CAA burden.

Variables Lobar CMBs ≥ 5* High CAA score*
Adjusted OR (95% CI) p Adjusted OR (95% CI) p
The presence of lacune, n (%) 3.909 (1.587–9.625) 0.003 3.297 (1.496–7.267) 0.003
The presence of deep CMB, n (%) 5.138 (0.955–27.631) 0.057 4.493 (1.110–18.182) 0.035
Deep CMB ≥ 5, n (%) 4.600 (1.811–11.687) 0.001 3.019 (1.285–7.094) 0.011
The number of deep CMBs, median (IQR) 1.057 (1.002–1.116) 0.043 1.052 (0.997–1.110) 0.066
Periventricular WMH scored 3, n (%) 3.002 (1.258–7.160) 0.013 5.306 (2.314–12.171) <0.001
Deep WMH scored ≥2, n (%) 5.075 (1.859–13.857) 0.002 5.595 (2.324–13.470) <0.001
The presence of WMH, n (%) 5.425 (1.911–15.398) 0.001 8.840 (3.355–23.291) <0.001
HA score, median (IQR) 2.317 (1.483–3.621) <0.001 2.241 (1.504–3.338) <0.001

HA, Hypertensive angiopathy; CAA, Cerebral amyloid angiopathy; CMB, Cerebral microbleed; WMH, White matter hyperintensities; and IQR, Interquartile range. *Multivariate binary regression was used. The presence of lobar CMBs ≥5 (vs. lobar CMBs <5) or high CAA score (vs. absence of high CAA score) was entered as dependent variable. Each of the HA related CSVD makers was entered into the regression analysis separately, by adjusting for age, male sex, hypertension, diabetes mellitus, hyperlipidemia, smoking, alcohol, and ICH etiology. The bold values in table represent that they were less than 0.05.