An exposure-to-affinity research model for anticholinergic burden

Introduction

Drug-category scales do not always incorporate a particular dose, route, exposure time, or laboratory assay. A systematic review found that anticholinergic burden scales vary substantially and identified no gold standard. UnifiedACB preserves those categories and separately explores an exposure-to-affinity index. The research question is whether a transparent quantitative construct might later add information; no clinical risk interpretation is established. ^{1, 2, 4}

The proposed calculation is deliberately tied to a specified reference regimen and an identified laboratory measurement. Its output cannot be substituted for a source's 0–3 drug category. The website calls an incomplete result ‘UnifiedACB Score not assigned’ because a numerical answer from mismatched or absent inputs would create more apparent precision than the research supports. ^{1, 2}

Methods

For a specified reference regimen, the archived calculation estimates unbound plasma concentration as total peak concentration multiplied by the unbound fraction. R is that concentration divided by a named muscarinic affinity or inhibitory-concentration value after unit alignment. The proposed bounded index is U = 10R/(1 + R). Assay species, tissue, receptor subtype, analyte, regimen, and timing must accompany each estimate. Ki and IC50 are not automatically interchangeable; conversion requires specific assay assumptions. ^{1, 3}

Eligibility review considers medication identity and route, exposure inputs, active moieties, and assay comparability. Published 0–3 scale values are displayed separately; they cannot substitute for missing model inputs or be converted to U. ¹

The transformation is applied after the exposure-to-affinity ratio is calculated: when R equals 1, U equals 5; as R increases, U approaches 10 without becoming a probability. For a proposed list calculation, only eligible and comparable raw R contributions would be summed before one transformation. Summing already transformed U estimates, or adding unlike publication categories, would be a different calculation. Unresolved medicines must remain explicitly unresolved rather than contribute zero. ^{1, 2}

The release audited historical estimates against displayed input ratios. That check tests arithmetic and traceability of selected inputs; it does not validate the exposure measurement, comparability of assays, central nervous system penetration, or a relationship to a patient outcome. A single-dose peak exposure cannot silently be transferred to steady state, another route, or an active-metabolite mixture. ¹

Results

The frozen v1.0 audit retains 20 historical calculations. Their arithmetic was reconciled for the stated inputs, but they are not validated whole-drug scores for arbitrary doses, formulations, or patients. The site therefore displays ‘UnifiedACB Score not assigned’ for a general medication or medication list when eligibility is incomplete. ¹

As an arithmetic illustration only, R = 1 yields U = 5. This is neither a 50% probability of harm nor a validated moderate-risk category. ¹

Eighteen archived calculations used rat-brain binding inputs and two involved human M1 cases. The audit reproduced the displayed 20 calculations to within 5 × 10⁻⁸ using the rounded R values; three retained small rounding differences that require original exposure inputs for exact reproduction. Chlorpromazine and clomipramine were moved from an older Final status to provisional after active-moiety review, leaving the current 20-estimate research set. ¹

Discussion

The model is a bounded transformation of selected pharmacologic inputs. It is not a direct measure of central receptor occupancy, adverse effects, delirium, or dementia. Parent-drug exposure may not cover active metabolites, and an oral reference regimen cannot be silently transferred to a patch, eye drop, or injection. A larger numerical estimate within this framework does not create a treatment threshold. ^{1, 2}

Before a medication-list total can be interpreted, eligible drug contributions, missing exposure, route and time alignment, and external outcome validation must be addressed. Unknown components remain unknown; they are not zero. ^{1, 2}

The assay issue is substantive. A rat-brain radioligand IC50 and a human M1 affinity value may reflect different species, tissues, receptor mixtures, and experimental conditions even when both are expressed in molar units. The denominator therefore stays attached to each estimate. Laboratory binding is not itself proof of drug exposure at central receptors or a clinical effect at a specific dose. ¹

The independent MEPS primary analysis did not confirm a cognitive-limitation association for the frozen research index. This outcome result does not invalidate the arithmetic, but it sharply limits any claim that the present U values predict clinical outcomes. Further work must address input completeness and prospective dose- and time-aware validation before a score could be considered for patient-level guidance. ^{1, 2}

References (AMA style)

1. Chaar M. Research v1.0 release notes: scope, provenance, compatibility and open issues. Unified ACB Tool; 2026. Published September 28, 2026. Source ↗
2. Chaar M. Public claims and safety specification: website copy, display rules, privacy and review. Unified ACB Tool; 2026. Published September 28, 2026. Source ↗
3. Cheng Y, Prusoff WH. Relationship between the inhibition constant (Ki) and the concentration of inhibitor which causes 50 percent inhibition (IC50) of an enzymatic reaction. Biochem Pharmacol. 1973;22(23):3099-3108. doi:10.1016/0006-2952(73)90196-2 Source ↗
4. Vennard O, Stewart C, Tolia M, Soiza RL, Myint PK. Anticholinergic medication burden scales: a systematic review. J Am Geriatr Soc. 2026;74(6):1771-1784. doi:10.1111/jgs.70352 Source ↗