Cognitive reserve among potentially inappropriate medication users with and without preclinical Alzheimer’s disease

Abstract

AbstractBackgroundAugmenting cognitive reserve (CR) may delay the progression from preclinical Alzheimer’s disease (pAD) to symptomatic AD. Because studies have shown that potentially inappropriate medication (PIM) use may lower CR,1 optimizing individual medication regimens may protect against AD dementia symptomatic onset. We investigated the association between baseline measures of CR and PIM use among participants enrolled in a randomized medication therapy management trial.2 MethodsParticipants completed cognitive tests –Trail Making Tests A and B (TMT) and California Verbal Learning Test Trials 1‐5 (CVLT) – while wearing a 1.5mg scopolamine patch (“challenged”) and again four weeks later under normal conditions (“unchallenged”). Change scores between the challenged and unchallenged tests (standardized for age, sex, and education) were used as a surrogate for CR. The Medication Appropriateness Index for medications considered potentially inappropriate for older adults based on the 2015 Beers Criteria3 served as a measure of PIM use (MAI‐PIM). pAD was defined as total brain relative standardized uptake values > 1.4 using amyloid PET.4Spearman correlation coefficients (rs) assessed the relationship between MAI‐PIM and CR.ResultsIncluded participants (N=79) had mean age 73.9 years [SD 5.9] and were 66% female; 31.6% met pAD criteria. The median (IQR) PIMs reported at baseline was 2 (2‐3). Participants with and without pAD had similar mean (SD) baseline change scores for TMT B minus A (‐0.17 [1.37] and ‐0.13 [1.10] respectively; p=0.90), but lower total CVLT change scores (‐0.85 [0.84] and ‐0.29 [0.83]; p=0.01). Higher MAI‐PIM was also correlated with lower baseline CVLT change scores (rs=‐0.16), driven by the participants with pAD (rs=‐0.33). TMT B minus A change scores and MAI‐PIM were uncorrelated (rs=0.02).ConclusionAmong community‐dwelling non‐demented older adults, executive function and total learning CR was operationalized via participants’ abilities to compensate for anticholinergic challenge. There was a moderate negative correlation between MAI‐PIM and low total learning CR among participants with pAD, but not for those without. These results support the hypothesized relationship between CR and medication appropriateness.