All statistical analyses used SAS 9.4 25 and/or R version 4.4.0. 26 To model CDR‐SB progression over time we used the hlme function of the R lccm package, 27 estimating a linear mixed model (without intercept) on the change in CDR‐SB from baseline (time = 0) as a function of time, with random effects on slopes for individual participants. In a second model, baseline CDR global score (0.5 or 1) was included as a categorical variable interacting with time.
