Comparison of Anticholinergic Burden Scales and Their Association with Cognitive and Functional Impairment in Older Adults: A Cross-Sectional Study Using the REPOSI Database

Background: The increasing use of anticholinergic medications in older adults with multiple chronic conditions raises significant concerns regarding their cumulative anticholinergic burden, which is linked to several adverse outcomes. This study aimed to compare existing anticholinergic burden scales to identify those most effective at correlating drug-induced anticholinergic load with cognitive and functional impairment. In addition, we proposed a new classification system on the basis of published scales to optimally correlate total anticholinergic burden with observed clinical deficits. Methods: This cross-sectional study analyzed data from the REPOSI registry, which collects clinical and therapeutic information on patients aged 65 years and older admitted to internal medicine and geriatric wards across Italy. Anticholinergic exposure was assessed using 20 established anticholinergic burden scales from literature. In addition, seven experimental scales were developed on the basis of published scales and various mathematical functions (maximum, mode, median, and mean) to evaluate potential differences in measuring anticholinergic load. Outcomes included cognitive impairment, assessed using the Short Blessed Test (SBT), and functional independence, measured by the Barthel Index (BI). A zero-inflated negative binomial model was applied to analyze associations between anticholinergic burden and each outcome. Given the variability in drug scoring across published scales, we developed seven experimental scales using different mathematical functions (maximum, mode, median, and mean) to standardize scoring. Three versions included a null-score adjustment to account for drugs omitted in some scales, ensuring a more comprehensive measure of anticholinergic burden. Results: Among 7735 patients, higher anticholinergic burden was consistently associated with increased cognitive impairment (SBT) and physical dependency (BI) across all existing and proposed scales. The modified Anticholinergic Risk Scale (mARS) scale showed the strongest associations with cognitive (rate ratio [RR] 1.10, 95% confidence interval [CI] 1.06, 1.13; P < 0.0001) and physical impairment (RR 1.18, 95% CI 1.11, 1.24; P < 0.0001), indicating an 18% higher risk of dependency per unit increase, while the CRIDECO Anticholinergic Load Scale (CALS) scale exhibited the best model fit. Our newly developed scales confirmed these associations, with the median with null score and the mean with null score scale showing the strongest link to cognitive impairment (RR 1.07, 95% CI 1.05, 1.09; P < 0.0001) and the strongest association with physical dependency (RR 1.11, 95% CI 1.08, 1.15; P < 0.0001). Conclusions: Scales that identify a greater proportion of patients with at least one anticholinergic drug may provide a more comprehensive assessment of anticholinergic burden in clinical practice. While no single scale demonstrated a definitive advantage across all outcomes, these scales may identify patients at risk. Prioritizing the use of scales with broader coverage could enhance clinical decision-making and optimize management of polypharmacy in older adults and recognize its potential impact on cognitive and functional outcomes.