Introduction

Exposure to anticholinergic medications may be associated with cognitive decline. Understanding which cognitive domains are involved may provide insights into the mechanisms.

Research question

What is the association between anticholinergic burden and specific domains of cognitive function in older adults who are initially without major cognitive impairment?

Study design

Post-hoc analysis of longitudinal observational data from the ASPirin in Reducing Events in the Elderly (ASPREE) study.

Methods

19,114 participants from Australia and the United States aged 70 years and older (65 years and older for US minorities) were recruited and followed for a median of 4.7 years. At enrollment, participants were free of known cardiovascular disease, major physical disability, or dementia. Cognitive assessments administered at baseline and biennially at follow-up visits included the Modified Mini-Mental State examination (3MS), Hopkins Verbal Learning Test–Revised (HVLT-R) delayed recall, Controlled Oral Word Association Test (COWAT), and Symbol Digit Modalities Test (SDMT). Anticholinergic burden was calculated at baseline using the Anticholinergic Cognitive Burden (ACB) scale and grouped as scores of 0 (no burden), 1-2 (low to moderate), or 3+ (high). Linear mixed effects models were used to assess the relationship between ACB score and cognition over time.

Results

After adjusting for sex, age, education, minority status, smoking status, hypertension, diabetes, depression, chronic kidney disease, country, and frailty, participants with a high ACB score had worse performance over time for 3MS (Adjusted [Adj] B=-0.092, P=0.034), HVLT-R delayed recall (Adj B=-0.104, P<0.001), COWAT (Adj B=-0.151, P<0.001), and SDMT (Adj B=-0.129, P=0.026), than participants with an ACB score of 0. A low to moderate ACB score was also associated with worse performance over time for HVLT-R delayed recall (Adj B=-0.037, P=0.007) and COWAT (Adj B=-0.065, P=0.003), compared to those with no ACB.

Conclusions

Anticholinergic burden predicts worse cognitive function over time in initially dementia-free older adults, particularly for executive function (COWAT) and episodic memory (HVLT-R).