Changes in anticholinergic cognitive burden and risk of single and recurrent falls: population-based cohort study

Background Medications with high anticholinergic cognitive burden (ACB) are associated with increased fall risk in older adults. However, the potential alteration of risk with changes in ACB over time has yet to be established. Objective To estimate the association between the changes in ACB with single and recurrent falls. Methods Data from European Investigation of Cancer-Norfolk (EPIC-Norfolk) study participants, aged 40 years and above, who attended the first (1HC:1993–98), second (2HC:1998–2000) and third (3HC: 2004–11) health checks were utilised. The main outcome was a single fall event or recurrent ($\ge 2$) falls occurring during the 12 months preceding the time point of the 3HC. Results Data from 10 717 participants with a median, Interquartile range (IQR) age of 55.6 (13.1) years were included. Three thousand four hundred forty-five (32.2%) participants had an ACB of one or greater at baseline. Participants were classified into four groups: no (67.8%), late (21.1%), transient (6.8%) and continuous (4.3%). Late (OR 1.49, 95% CI 1.25–1.79), transient (1.66, 1.28–2.14) and continuous (1.67, 1.22–2.29) exposure were significantly associated with increased recurrent falls compared with no exposure. Mediation analysis revealed that gait speed (GS) contributed to 16.9% (CI: 9.4%–27.8%) of the increase in risk of recurrent falls associated with ACB. Discussion Anticholinergic medication use, in adults aged 40 years and above, was linked to recurrent falls at 14-year follow-up, regardless of whether introduction or cessation occurred during the follow-up. Future research should determine effective strategies for minimising the long-term risk of falls when starting anticholinergic medications, which could include GS as a risk-detection and monitoring tool.