Delirium in hospital was followed by a five-fold higher dementia risk in older adults

Design

Prospective population-based cohort with time-dependent exposure

Population

3501 adults, median age 74, hospitalised, without prior dementia

Primary outcome

Incident mild cognitive impairment and dementia

Effect

MCI aHR 3.67 (2.19 to 6.15); dementia aHR 5.39 (2.94 to 9.85)

This cohort drew on the Mayo Clinic Study of Aging, a population-based study with neuropsychological testing every 15 months. It included 3501 adults, median age 74, with at least one hospital admission and no dementia beforehand. Delirium was identified from clinical notes using a validated language-processing algorithm. Median follow-up was 7.2 years.

Delirium was documented in 75 people (2.1%). It was associated with a higher risk of mild cognitive impairment (adjusted hazard ratio 3.67, 95% CI 2.19 to 6.15) and of dementia (5.39, 2.94 to 9.85), and with faster decline on cognitive testing, including in those cognitively normal at baseline.

The association is strong, but the design cannot separate cause from marker: delirium may unmask brain vulnerability rather than create it. Either way, an episode of delirium identifies someone who deserves cognitive follow-up. The low detection rate, 2%, almost certainly undercounts delirium, which is often missed in notes.

• Arrange cognitive follow-up after any hospital admission complicated by delirium.
• Record delirium episodes clearly in the discharge summary so they are not lost.
• Use proven non-drug measures to prevent delirium: orientation, sleep, mobility, hearing and vision aids.
• Review sedative and anticholinergic drugs in older inpatients.

Why it matters

Delirium is often treated as a passing inpatient problem, not a warning about the years ahead.

Don't overread it

Observational; it cannot show that preventing delirium prevents dementia.

The statistics, in plain English

A hazard ratio of 5.39 for dementia means about five times the rate. The wide interval (2.94 to 9.85) reflects only 75 people with delirium. Adjustment covered age, sex, education and APOE status, but not every illness that might cause both delirium and decline, so this is an association.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.