James · Neurology 2019 · prospective cohort study · n=777

Cognitive decline after elective and nonelective hospitalizations in older adults.

Cited 36 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective longitudinal cohort study with linked claims data

PubMed 30635482 · doi:10.1212/WNL.0000000000006918 · record verified 2026-08-26

What was done

Data from 777 community-dwelling older adults without baseline dementia in the Rush Memory and Aging Project were linked to 1999–2010 Medicare claims records. Participants completed annual cognitive testing using a battery of 19 tests over a mean follow-up of 5.0 years (SD = 2.6). Mixed-effects regression models compared global cognitive decline before and after elective versus nonelective (emergent/urgent) hospitalizations relative to no hospitalization, adjusting for age, sex, education, medical conditions, length of stay, surgery, intensive care unit admission, and comorbidities.

What was found

A total of 460 participants (59.2%) were hospitalized during follow-up, with 222 (28.6%) having at least one elective and 418 (53.8%) at least one nonelective admission. Non-hospitalized participants had a mean cognitive decline of 0.051 units per year. Elective hospitalizations showed no significant difference in decline rates either before (0.044 units/year) or after (0.048 units/year) admission. In contrast, participants with nonelective hospitalizations showed faster prehospital decline (0.076 units/year; estimate = -0.024, SE = 0.011, p = 0.032) and accelerated by 0.036 units/year (SE = 0.005, p < 0.001) to a mean loss of 0.112 units per year post-hospitalization, more than doubling the rate of non-hospitalized individuals.

Why it matters

This study shows that post-hospitalization cognitive acceleration is driven by unplanned, nonelective admissions rather than hospital stays in general. These findings identify nonelective admissions as critical events warranting targeted cognitive monitoring and follow-up in older adults.

Limits

As an observational cohort study, residual confounding from acute illness severity, unmeasured clinical factors, or in-hospital delirium cannot be excluded. The study is based on a single volunteer cohort of community-dwelling older adults, potentially limiting generalizability to broader demographic and socioeconomic groups.

Cited by