All-Cause Acute Illness Hospitalisations in the Preceding Two Years Are Associated With Cognitive Decline in Older Adults: The Sydney Memory and Ageing Study.
Level 3 - non-randomized controlled study
Prospective longitudinal cohort study with linked administrative health records
PubMed 40312128 · doi:10.1002/gps.70077
What was done
This was a secondary analysis of four biennial assessment waves over 10 years from the Sydney Memory and Ageing Study, linked probabilistically to hospitalization data. The study evaluated 1026 functionally independent older adults aged 70-90 years at baseline. The primary outcomes were baseline global cognition intercept and cognitive change slope, analyzed against all-cause acute hospitalization episodes and cumulative length of stay across five consecutive 2-year time intervals using time-lagged models adjusted for age, sex, and education.
What was found
The cohort had a mean age of 78.8 years, baseline mean Mini-Mental State Examination score of 28.7, and averaged 3.3 hospitalizations and 18.9 days in hospital over 10 years. Overall adjusted cognitive change was -0.133 z-scores per year. In the final time interval, cognitive decline accelerated by -0.012 z-scores per hospitalization per year (SE = 0.005, p = 0.014) and -0.002 z-scores per day in hospital per year (SE = 0.001, p < 0.001). Time-lagged models revealed that pooled recent hospitalizations drove this decline (-0.036 z-scores per year per episode, SE = 0.012, p = 0.004; -0.008 z-scores per year per day, SE = 0.002, p < 0.001), significantly exceeding non-recent hospitalizations (Wald test p = 0.011 for episodes, p < 0.001 for days).
Why it matters
This study shows that all-cause acute hospitalizations, rather than only intensive care or surgery, are dose-dependently and recency-dependently linked with accelerated cognitive decline in older adults. It identifies recent hospital admissions as a critical window for cognitive monitoring and intervention.
Limits
The observational design cannot establish causality, as hospitalization may serve as a proxy for underlying acute or worsening chronic medical illness. Specific inpatient drivers such as in-hospital delirium, medication changes, or specific primary diagnoses were not reported in the abstract, and results from this community-dwelling cohort may not generalize to other settings.