Background: Older people exposed to medicines with anticholinergic properties are at an increased risk of experiencing adverse events. Anticholinergic burden refers to exposure to 1 or more medicines with anticholinergic activity. Studies have reported anticholinergic burden to be a strong predictor of cognitive and physical impairment in older people1. Background work in this area has provided a model for quantifying anticholinergic burden2. In this work we explored a competitive multiple-ligand model to estimate relative anticholinergic potencies and anticholinergic burden.
Aim: To quantify the anticholinergic burden using a dose-response based pharmacological model that takes into account the muscarinic receptor binding affinity.
Methods: The anticholinergic activity of commonly prescribed medicines was derived from the literature and categorised into high, moderate, and low anticholinergic activity. An a priori model was developed using the principles of competitive multiple-ligand binding to calibrate the relative potencies of anticholinergic medicines. The binding was generalised to be a function of defined daily dose, bioavailable fraction, and relative potency. A data set was extracted from prescription and diagnostic data that included the events delirium, constipation and urinary retention and drug treatments the patient was taking. These events were chosen as they represent typical central, local and peripheral anticholinergic effects. The extracted data was rearranged into a NONMEM readable format using MATLAB. Repeated measures modelling were conducted using NONMEM v7.2. Parameters estimated from the model will be relative potency (ED50) for different categories of anticholinergic medicines.
Results: A literature search revealed 102 unique drug entities that were described as having anticholinergic activity. A search of the database provided a total of 3461 patients who were exposed to 1 or more anticholinergic agents. Of those patients, 1.4%, 3.0%, 0.7% experienced delirium, constipation or urinary retention, respectively. Patients were taking 1 to 15 medicines. An initial analysis in NONMEM identified drugs associated with these events and their relative potencies have been estimated.
Discussion: There are large number of drugs that have been reported to have anticholinergic activity and the reported activity differs significantly for some drugs. The multivariate model developed in this work will be used as the basis for estimating anticholinergic burden in future patients.
References:
1. Nishtala PS et al., (2009), J Clin Pharmacol, 49, 1176-84.
2. Hilmer SN et al., (2007), Arch Intern Med, 23, 781-87.
