Examining Associations of Anticholinergic Burden with Cognition and Mobility in Older Adults

: (Alternative Format Thesis)

Abstract

Background
Anticholinergic burden (AChB) refers to the cumulative effect of medications with anticholinergic activity and is increasingly recognised as an important determinant of adverse outcomes in older adults. Elevated AChB has been associated with cognitive decline, delirium, falls, functional impairment, and mortality. However, existing measurement tools are inconsistent, largely expert-based, and show limited concordance. Addressing these limitations is essential for improving pharmacoepidemiological research and guiding safer prescribing in geriatric care.

Methods
This thesis, presented in alternative format, adopts a multi-method pharmacoepidemiological approach. Four interlinked studies were conducted: (1) a systematic review and meta-analysis examining the association between AChB and mobility outcomes; (2) a drug property evaluation comparing pharmacological characteristics with classifications in the Anticholinergic Cognitive Burden (ACB) scale and exploring machine learning methods for classification; (3) a narrative review of innovations in AChB detection and management, including digital health and deprescribing strategies; and (4) the development and clinical validation of a novel machine learning–derived AChB scale using data from 14,067 community-dwelling older adults.

Results
The systematic review demonstrated consistent associations between higher AChB and impaired mobility, extending the literature beyond cognition. The drug property analysis revealed discrepancies between expert-opinion classifications and mechanistic evidence, while exploratory machine learning analyses indicated that property-based models could better align with pharmacological plausibility. The narrative review placed these methodological advances within the wider translational context, identifying opportunities and barriers to clinical application. Finally, the new machine learning–derived scale was optimised and validated, showing improved predictive accuracy for cognitive and functional outcomes compared with legacy measures.

Conclusions
This thesis advances the measurement of anticholinergic burden by integrating evidence synthesis, mechanistic evaluation, conceptual innovation, and clinical validation. The findings demonstrate that expert-based classifications may not adequately reflect pharmacological reality, while machine learning–based approaches show promise in producing more robust and clinically relevant tools. Continued validation, integration into electronic prescribing systems, and evaluation within deprescribing interventions are recommended to translate these advances into improved outcomes for older adults.

Date of Award	25 Mar 2026
Original language	English
Awarding Institution	University of Bath
Supervisor	Prasad Nishtala (Supervisor) & Nikoletta Fotaki (Supervisor)