Skip to main content
. 2019 Apr 22;76(7):818–826. doi: 10.1001/jamaneurol.2019.0834

Table 2. Associations of β-Amyloid Protein With Cognitive Decline in the Absence of β-Amyloid Depositsa.

Cognitive Outcome Age Male Education Level β-Amyloid
Estimate (SE) P Value Estimate (SE) P Value Estimate (SE) P Value Estimate (SE) P Value
Global cognition 0.0001 (0.001) .96 0.003 (0.018) .87 0.001 (0.002) .59 −0.014 (0.005) .008
Episodic memory 0.0004 (0.001) .77 0.001 (0.018) .95 0.002 (0.002) .51 −0.006 (0.005) .26
Semantic memory 0.0005 (0.001) .66 −0.007 (0.016) .65 0.002 (0.002) .44 −0.004 (0.005) .40
Working memory −0.0004 (0.001) .73 0.0001 (0.017) >.99 0.002 (0.002) .38 −0.007 (0.005) .16
Processing speed −0.002 (0.001) .13 0.007 (0.018) .69 0.001 (0.002) .53 −0.014 (0.005) .008
Visuospatial ability 0.0008 (0.001) .48 0.001 (0.016) .93 −0.001 (0.002) .75 −0.013 (0.005) .006
a

The results were obtained using linear mixed models in which annual cognitive scores are longitudinal outcomes. Each row represents a separate model for a different cognitive outcome. The model includes a term for time in years prior to death, which estimates the mean rate of cognitive decline, terms for independent variables (age [per year], sex, education level [for each year of education received], and β-amyloid protein), and their interactions with time. The interaction terms were used to examine the associations of independent variables with the rate of cognitive decline. Numbers presented are estimates (SEs) and P values for the interaction terms.