Anticholinergic burden on admission assessed by the Japanese Anticholinergic Risk Scale and incident in-hospital delirium in older adults: A multicenter prospective cohort study.

OBJECTIVES

To examine whether admission anticholinergic burden, quantified by the Japanese Anticholinergic Risk Scale (JARS), is associated with incident in-hospital delirium and whether baseline cognition modifies this association.

METHODS

In this prospective multicenter cohort (Japan; October 2019-March 2025), 987 adults aged ≥60 years admitted to acute geriatric wards at 3 university hospitals and 1 national medical center were assessed. JARS from admission medication reconciliation was modeled as 0, 1-2, and ≥3, and per 1-point increase. Delirium was diagnosed on ward rounds using DSM-5 criteria. Cognition was assessed by Mini-Mental State Examination (MMSE) and stratified as ≤23 vs ≥24. Logistic regression adjusted for age, sex, MMSE, emergency admission, Clinical Frailty Scale, Charlson Comorbidity Index, and C-reactive protein; effect modification was tested with a JARS×MMSE interaction and stratified models.

RESULTS

Delirium occurred in 156 participants (15.8%). JARS was 0 in 48.1%, 1-2 in 38.8%, and ≥3 in 13.1%. High burden (JARS ≥3) was associated with delirium (adjusted OR 1.76, 95% CI 1.09-2.85). Per 1-point higher JARS, the overall association was borderline (OR 1.13, 95% CI 1.00-1.27) but differed by cognition: OR 1.80 (1.27-2.56) for MMSE ≥24 and 1.06 (0.93-1.21) for MMSE ≤23 (interaction P=0.011-0.029).

CONCLUSIONS

This prospective multicenter study provides the first evidence supporting JARS-based delirium risk stratification, with a clearer dose-response among cognitively preserved patients. JARS-based medication review at admission may help target potentially preventable delirium risk in acute geriatric wards.