Table 1.

Characteristics of study participants at baseline

Variable	Controls, n = 164	MCI, n = 258	AD, n = 103	p§	post hoc
Age, mean (SD)	76.03 (5.22)	74.48 (7.25)	75.11 (7.35)	.069	C > M*
Female, %	48.8	35.7	45.6	.019	C A > M**
Caucasian, %	92.7	92.6	94.2	.149	----
Education, mean (SD)	16.02 (2.88)	15.67 (2.94)	14.98 (3.29)	.022	C M > A
On anti-dementia medication, %	0.0	48.1	92.2	.001	A > M > C**
MMSE, mean (SD)	29.16 (0.96)	27.08 (1.79)	23.54 (1.99)	.001	C > M > A
CDR-global, %
0.0	100.0	0.0	0.0	.001	A > M > C†
0.5	0.0	100.0	48.5
1.0	0.0	0.0	51.5
APOE ε4, %	26.8	56.2	66.0	.001	C < M A **
‡ VBR, mean (SD)	3.65 (1.91)	4.55 (1.91)	5.15 (1.90)	.001	A > M > C
‡ FAQ, mean (SD)	0.15 (4.32)	3.95 (4.30)	12.82 (4.28)	.001	A > M > C

MCI = mild cognitive impairment; AD = Alzheimer disease; MMSE = Mini-Mental State Examination; CDR = Clinical Dementia Rating scale; APOE = apolipoprotein E; VBR = ventricle-to-brain ratio; FAQ = Pfeffer Functional Activities Questionnaire.

^§p value for omnibus test of group differences.

^*Though the omnibus test for group differences in age was only marginally significant, we nonetheless conducted post hoc tests to further probe the omnibus finding because of the centrality of age in studies of dementia.

^**“Female,” “on anti-dementia medication,” and “APOE ε4” are dichotomous variables. Therefore, these are subsequent 2×2 chi square or Fisher’s exact tests, not pairwise comparisons.

^†“CDR–global” is a categorical variable. Therefore, these are subsequent 2×3 Fisher’s exact tests, not pairwise comparisons.

^‡Adjusted for age, gender, and education.

C > M = control mean greater than MCI mean; C A > M = control and AD counts greater than MCI count; C M > A = control and MCI mean greater than AD mean; A > M > C = AD mean (count) greater than MCI and control means (counts), and MCI mean (count) greater than control mean (count); C < M A = control count less than MCI and AD counts; C > M > A = control mean greater than MCI and AD means, and MCI mean greater than AD mean.