Abstract
BACKGROUND: Polypharmacy and use of anticholinergics are associated with adverse health outcomes. Because these 2 exposures covary, it is not known to what extent the detrimental health effects attributed to anticholinergic burden measured by anticholinergic burden scales (ABS) may be due to polypharmacy. METHODS: Our aims were to evaluate the added effect beyond polypharmacy of (1) anticholinergic use and (2) scores according to existing ABS. We used linked data from the UK Biobank prospective cohort study to investigate drugs prescribed in primary care in the year 2015 (n = ∼200 000; median age = 65 years). We randomly sampled among drugs to create 2000 pseudoscales-drug burden scales designed to reflect the strength of the background effect of polypharmacy. We differentiated pseudoscales constructed to capture either general polypharmacy or putative anticholinergic polypharmacy. For each scale, we fitted logistic regression adjusted for confounders to assess associations between pseudoscales from each set and risk of death, dementia, or delirium. We also assessed 23 existing ABS by comparing them to the effects of pseudoscales that included the same numbers of drugs as each ABS. RESULTS: Odds ratios for death, dementia, and delirium using anticholinergic-polypharmacy pseudoscales (median = 1.09-1.10) were greater than those of general-polypharmacy pseudoscales (1.05-1.06). The added effect of anticholinergic use beyond polypharmacy was greater in adjusted models and when prescribed to participants when they were older. 35%-90% of ABS exhibited stronger effects than most pseudoscales. CONCLUSIONS: The models show an anticholinergic effect of drugs beyond polypharmacy or drug count, but it is small, ABS-dependent, and varies between outcomes.