Taylor-Rowan · The Cochrane database of systematic reviews 2021 · Systematic review and meta-analysis · n=25 studies (968,428 participants)

Anticholinergic burden (prognostic factor) for prediction of dementia or cognitive decline in older adults with no known cognitive syndrome.

Cited 87 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review and meta-analysis of observational prognostic studies

PubMed 34097766 · doi:10.1002/14651858.CD013540.pub2 · record verified 2026-08-28

What was done

A Cochrane systematic review and random-effects meta-analysis searched five databases up to March 2021. Authors included prospective and retrospective longitudinal cohort and case-control observational studies (minimum 1 year follow-up) assessing anticholinergic burden scales as prognostic factors for future cognitive decline or dementia in cognitively unimpaired older adults. Twenty-five studies (968,428 participants) met inclusion criteria, and 8 studies (320,906 participants) provided odds ratio (OR) data suitable for meta-analysis.

What was found

The Anticholinergic Cognitive Burden (ACB) scale was the only tool with sufficient data for scale-specific pooling: - Unadjusted ACB pooled OR: 1.47 (95% CI 1.09 to 1.96). - Adjusted ACB pooled OR: 2.63 (95% CI 1.09 to 6.29). - Exploratory pooled adjusted OR across all scales: 2.16 (95% CI 1.38 to 3.38). - Dose-response gradient across ACB scores: ACB 1 (OR 2.18, 95% CI 1.11 to 4.29), ACB 2 (OR 2.71, 95% CI 2.01 to 3.56), and ACB 3 (OR 3.27, 95% CI 1.41 to 7.61). - Overall certainty of evidence was evaluated as low by GRADE.

Why it matters

This review establishes that cumulative anticholinergic exposure is associated with a dose-dependent increase in the odds of cognitive decline or dementia in unimpaired older adults, supporting clinical caution with anticholinergic prescribing.

Limits

The certainty of evidence was graded as low, relying entirely on observational studies where residual confounding (e.g., confounding by indication) remains possible. Only 8 of 25 identified studies could be quantitatively pooled because hazard ratios and continuous metrics were unsuitable for meta-analysis. Only one scale (ACB) had enough data for scale-specific meta-analysis.

Cited by