. Author manuscript; available in PMC: 2021 Oct 12.

Published in final edited form as: J Alzheimers Dis. 2021;83(2):861–877. doi: 10.3233/JAD-210251

Table 7.

Odds ratio with 95% confidence intervals (CI) from conditional logistic regression models for specified unit of decrement in cross-sectional or longitudinal One Card Learning performance with corresponding increases in likelihood of progression to MCI.

	Decrement	Odds Ratios (95% CI)	p-value
sOBJ
baseline¹	1 z	1.70 (1.20, 2.42)	0.003
30 months¹	1 z	2.09 (1.44, 3.03)	<0.001
ΔOBJ
WSD¹	1 z	1.20 (0.88, 1.64)	0.25
LR^1,2	0.046 SD	1.06 (0.80, 1.41)	0.67
LME^1,2	0.005 SD	2.02 (1.43, 2.86)	<0.001

Note. Odds ratios represent the likelihood of progressing from cognitively unimpaired to MCI given a 1 SD lower performance for each individual method. sOBJ = subtle objective cognitive impairment at a single assessment. ΔOBJ = subtle objective cognitive decline (longitudinal). WSD = within subjects’ standard deviation. LR = annualized change from linear regression model. LME = annualized change from linear mixed effects model.

Reverse scaling applied so that the odds ratios are greater than one.

ΔOBJ-LR and ΔOBJ-LME were scaled by dividing by their standard deviations.