Skip to main content
. 2011 Oct 18;26(6):443–449. doi: 10.1177/1533317511424892

Table 2.

Factor Structure of the New ADAS-cog 2-Factor Analysis for the AD, PDD, and Pooled Patient Populations (Randomized Population a )

Item IDEAL and ADENA (AD) EXPRESS (PDD) IDEAL, ADENA, and EXPRESS (AD and PDD) a
N = 3959 N = 513 N = 4472
Factor 1 (4.71) b Factor 2 (0.80) b Factor 1 (3.90) b Factor 2 (0.72) b Factor 1 (4.56) b Factor 2 (0.79) b
Word recall 0.67 0.23 0.58 0.19 0.68 0.22
Following commands 0.56 0.43 0.52 0.29 0.54 0.43
Constructional praxis 0.47 0.19 0.52 0.18 0.45 0.21
Naming objects/fingers 0.53 0.36 0.43 0.08 0.53 0.33
Ideational praxis 0.57 0.31 0.50 0.35 0.54 0.34
Orientation 0.67 0.17 0.62 0.28 0.67 0.16
Word recognition 0.65 0.18 0.49 0.12 0.64 0.16
Remembering test instructions 0.54 0.43 0.58 0.44 0.55 0.42
Spoken language ability 0.22 0.80 0.20 0.73 0.21 0.79
Word-finding difficulty 0.27 0.77 0.17 0.76 0.27 0.76
Comprehension of spoken language 0.32 0.73 0.37 0.69 0.32 0.73

Abbreviations: IDEAL, Investigation of transDermal Exelon in ALzheimer’s disease; EXPRESS, EXelon in PaRkinson’s disEaSe dementia Study; ADENA, Alzheimer's Disease with ENA 713; AD, Alzheimer’s disease; ADAS-cog, Alzheimer’s Disease Assessment Scale–cognitive subscale; PDD, Parkinson’s disease dementia.

aOf the randomized population (n = 4540), baseline ADAS-cog data were provided by 4472 patients.

bEigenvalues. Two domains were derived from a factor analysis on the covariance matrix of 11 baseline ADAS-cog items using a varimax rotation. Factors retained had eigenvalues >0.5. Domains took the sum of each item that loaded >0.4; items that loaded on multiple factors were included in the factor with the highest loading.

Items that loaded most highly on the factor are shown in bold.