Anticholinergic burden and cognition in schizophrenia

A cognitive association with practical limits

Difficulty with learning, attention, or processing information can remain important even when psychotic symptoms are controlled. Anticholinergic medication exposure is one potentially relevant part of that picture. The central question is how much the available studies establish: an association between classified exposure and test performance is informative, but it does not identify the cause of one person’s difficulty or predict the result of a medication change.

This article separates a large observational study, a broader evidence synthesis, and the interpretation needed when using UnifiedACB. It is a narrative reading of named publications, not an additional meta-analysis. The original scales and their versions remain the units of evidence; no new UnifiedACB cognitive risk score is calculated.

What a large outpatient study contributed

Joshi and colleagues analyzed cross-sectional data from 1,120 people with schizophrenia or schizoaffective disorder. They used a modified ACB scale and assessed cognition with a computerized battery. Higher burden groups performed worse across the tested domains, including after adjustment for measured clinical and treatment variables. The scale modification and cross-sectional design are essential parts of interpreting that finding. Joshi et al., 2021

An adjusted association is stronger evidence than an unqualified impression that a long regimen “must” explain cognition, but adjustment is not randomization. The analysis cannot guarantee that all reasons for receiving a particular medication have been measured. Nor does its modified scale automatically become identical to a public source version bearing the same ACB family name.

What the 2025 synthesis added

Mancini and colleagues combined studies of clinical burden scales, serological anticholinergic activity, and medication tapering in psychosis. Their meta-analysis reported negative associations with global cognition and several domains. The authors also found improvement in selected domains in tapering studies, while calling for randomized trials to quantify benefit with less bias. These are related lines of evidence, not interchangeable measurements. Mancini et al., 2025

The existence of a global association does not mean every cognitive domain changes equally in every person. A study that measures verbal learning answers a narrower question than one measuring social cognition. Likewise, a short follow-up after an intervention cannot by itself establish sustained improvement in education, independent living, or employment. The outcome, interval, and study design should travel with the conclusion.

Why three evidence layers should stay separate

Clinical burden scales categorize medicines according to their own rules. A serum or laboratory measure reports a different kind of signal. A treatment study asks what happens when a regimen is altered under specified conditions. Agreement across these layers can strengthen a hypothesis, but none can simply be substituted for another in an individual calculation.

For example, a source-specific category does not supply a blood concentration, receptor occupancy, or expected number of points lost on a cognitive test. Moving from the category to any of those quantities requires an explicit, validated relationship. UnifiedACB therefore shows the underlying classification rather than converting it into an estimated cognitive deficit.

Read the comparison group carefully

“Higher” and “lower” burden are study-defined categories. Their meaning depends on which medicines were covered, how as-needed treatment was handled, whether dose entered the method, and how the study selected participants. A higher group may differ from a lower group in more than medication classification. Treatment complexity and illness history can influence both exposure and performance.

When reading a paper, ask whether its lower-exposure group had similar clinical needs and whether the model accounted for the factors most relevant to the question. If a paper reports a correlation, retain that language. Calling the result a medication effect can imply a counterfactual comparison the study did not perform: how the same person would have performed under an otherwise identical regimen with different exposure.

Bring the finding into a cognitive assessment

A useful assessment records the test, cognitive domain, testing conditions, and medication history at the relevant time. The history should distinguish prescriptions from actual use and capture recent changes. A result obtained after an unusually disrupted night or an acute episode should not be interpreted as though it were a stable measure of long-term performance.

The source lookup then helps identify which medicines merit discussion in that context. It can show that the evidence differs by scale or that a particular item was not publicly listed. It cannot decide whether a test result reflects sedation, a motor limitation, illness-related difficulty, medication exposure, or several of these together.

If a clinical team changes treatment for an independently justified reason, follow-up should use a clearly defined outcome rather than assuming that a lower source total is itself proof of cognitive benefit. A numerical exposure change and a meaningful improvement in daily experience should be documented separately. Repeating a test also raises questions about practice effects and comparability of conditions.

A research summary should also keep the study’s exposure definition visible when it is translated into practice. If an analysis used a modified scale, its findings concern that modified method and the population in which it was applied. Looking up a medicine in another version can help locate relevant evidence, but it does not reproduce the original study’s classification or adjusted model. The distinction matters when two apparently similar burden descriptions are being compared across visits, services, or publications.

What the evidence does not justify

These studies do not validate a universal burden cutoff for schizophrenia, a ranking of all antipsychotics by cognitive safety, or a formula for subtracting anticholinergic influence from a test score. They also cannot promise that tapering an adjunct will be appropriate or successful for every patient. The indication for the adjunct and the outcome it was intended to prevent still matter.

The clinical implication is to make anticholinergic exposure visible when evaluating cognition and considering a regimen review. The evidence is most useful when it improves the questions asked about treatment, while medication changes remain tied to individual assessment and follow-up.

References

1. Mancini V, et al. Anticholinergic Burden and Cognitive Function in Psychosis: A Systematic Review and Meta-Analysis. Am J Psychiatry. 2025;182:349-359. Source
2. 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. Source
3. Georgiou R, et al. Anticholinergic Burden and Cognitive Performance in Patients With Schizophrenia: A Systematic Literature Review. Front Psychiatry. 2021;12:779607. Source
4. Joshi YB, et al. Anticholinergic Medication Burden-Associated Cognitive Impairment in Schizophrenia. Am J Psychiatry. 2021. doi:10.1176/appi.ajp.2020.20081212. Source

Disclosure

Funding: None. Conflict of interest: None declared for the UnifiedACB educational project.

Educational evidence summary. This content does not provide patient-specific medical advice and does not replace clinical judgment, medication reconciliation, or prescribing guidance.