Small vessel cerebrovascular disease is associated with cognition in prospective Alzheimer’s clinical trial participants

Table 2 The association of regional and total WMH volumes with PACC scores

	Model 1 (frontal lobe)			Model 2 (temporal lobe)			Model 3 (parietal lobe)			Model 4 (occipital lobe)			Model 5 (total volume)
	B	95% CI	p-value	B	95% CI	p-value	B	95% CI	p-value	B	95% CI	p-value	B	95% CI	p-value
Age	− 0.17	[− 0.2, − 0.15]	p < 0.001	− 0.18	[− 0.20, − 0.16]	p < 0.001	− 0.18	[− 0.20, − 0.02]	p < 0.001	− 0.18	[− 0.20, − 0.16]	p < 0.001	− 0.18	[− 0.20, − 0.15]	p < 0.001
Education	0.12	[0.08, 0.17]	p < 0.001	0.13	[0.09, 0.17]	p < 0.001	0.13	[0.08, 0.17]	p < 0.001	0.13	[0.09, 0.17]	p < 0.001	0.13	[0.09, 0.17]	p < 0.001
WMH^a	− 0.03	[− 0.06, − 0.01]	p = 0.009	− 0.07	[− 0.17, 0.03]	p = 0.15	− 0.04	[− 0.08, − 0.000008]	p = 0.04	− 0.05	[− 0.13, 0.03]	p = 0.20	− 0.01	[− 0.03, − 0.001]	p = 0.03
Amyloid	− 0.38	[− 0.62, − 0.13]	p = 0.003	− 0.38	[− 0.63, − 0.13]	p = 0.002	− 0.38	[− 0.62, − 0.13]	p = 0.003	− 0.37	[− 0.62, − 0.12]	p = 0.003	− 0.38	[− 0.62, − 0.13]	p = 0.003
Adjusted R²	0.141			0.138			0.139			0.138			0.140
ΔR² with inclusion of interaction term	0.0005			0.0005			0.0005			0.0003			0.0005
ΔAIC with inclusion of interaction term	3.54			1.44			3.35			1.09			2.01

Five separate models were run, which included regional (frontal, temporal, parietal, and occipital) and total WMH volumes, respectively, as the primary WMH predictor of PACC, together with age, education, and amyloid status, which were invariant across models. Included statistics reflect unstandardized B estimate, 95% confidence intervals, and p-values of the association of each factor (age, education, regional or total WMH, and amyloid SUVR) with PACC scores. The bottom two rows show change in adjusted R² and AIC when comparing models with and without a WMH by amyloid interaction term. A positive delta AIC indicates that there is less parsimony with interaction term in the model
^aThe “WMH” variable refers to frontal, temporal, parietal, occipital, and total WMH volume for each of the five models, respectively

ISSN: 1758-9193