. Author manuscript; available in PMC: 2010 Jul 14.

Published in final edited form as: J Int Neuropsychol Soc. 2009 Dec 4;16(2):233–243. doi: 10.1017/S1355617709991160

Table 5. Further Multinomial Logistic Regression Models: Additional Variables Significant in Univariate Testing (p<0.05) after Adjustment for Minimum Model Covariates.

		Odds Ratios (ORs)
	Analysis Population (N)	Class 1: Mild	Class 2: Attention/Construction	Class 3: Memory	Class 4: Severe	p-value
Male gender	590	.	2.09	0.80	0.70	0.01
Frontal gait^*	578	.	2.66	0.95	3.00	0.09
Frontal release signs - grasp	580	.	0.76	1.05	3.88	0.01
Early/ prominent visuospatial impairment^*	510	.	2.13	0.92	2.02	0.00
White matter changes	438	.	0.72	0.39	1.16	0.02
AD with psychotic features	590	.	1.90	2.16	3.26	0.14
Current use of medications for AD	590	.	1.29	0.69	0.62	0.14

Due to data sparseness, a data-derived prior was applied to stabilize logistic regression model

^†

Odds ratio is odds of class membership relative to Class 1 associated with a one unit increase (continuous variables) or relative to reference value (categorical variables), adjusted for Minimum Model covariates