Skip to main content
. Author manuscript; available in PMC: 2011 Feb 22.
Published in final edited form as: Arch Gen Psychiatry. 2008 May;65(5):542–550. doi: 10.1001/archpsyc.65.5.542

Table 5.

Multivariate-Adjusted Correlates of Each Cognitive Domain as an Outcome in Subjects Without Depression Combined With Those With Either Amyloid-Associated Depression or Nonamyloid Depressiona

Subjects With Amyloid-Associated
Depression + Without Depression
 Subjects With Nonamyloid
Depression + Without Depression
LM Delayed Recall
(n=735)
 Block Design
(n=710)
 LM Delayed Recall
(n=730)
 Block Design
(n=704)
β Estimate
(SE)		 P
Value		 β Estimate
(SE)		 P
Value		 β Estimate
(SE)		 P
Value		 β Estimate
(SE)		 P
Value
Age, y −0.3 (0.04) <.001 −0.2 (0.04) <.001 −0.3 (0.04) <.001 −0.2 (0.04) <.001
School, y 0.9 (0.1) <.001 0.8 (0.1) <.001 0.9 (0.1) <.001 0.9 (0.1) <.001
APOE4 −1.5 (0.7) .04 −2.0 (0.7) .005 −1.1 (0.8) .16 −1.1 (0.7) .14
Antidepressant use 1.3 (0.8) .09 0.1 (0.7) .87 0.4 (0.8) .65 0.5 (0.7) .46
Amyloid-associated depression −1.9 (0.4) <.001 −1.4 (0.4) <.001
Nonamyloid depression −0.9 (0.8) .30 .30 <.001

Abbreviation: LM, Logical Memory.

a

All of the listed variables were included in the regression model after adjusting for sex, race, creatinine level, diabetes mellitus, stroke, and cardiovascular disease. Because Bonferroni correction was applied, P<.0167 is considered significant.