Hospitalisation with infection linked to later psychiatric and neurological diagnoses

Design

Retrospective propensity-matched multicohort study

Population

1,062,722 matched pairs hospitalised in 62 US health organisations, 2014 to 2018

Primary outcome

14 psychiatric and neurological outcomes at 1 month to 2 years

Effect

Raised risk in 116 of 140 comparisons; cognitive deficits 19.08% vs 6.71% after cardiac infection

This US multicohort study matched 1,062,722 pairs of patients from 62 health organisations, comparing those hospitalised with infection across 10 body systems against those hospitalised for other reasons and the general population, over two years of follow-up.

Infection was associated with higher risk in 116 of 140 infection-disorder comparisons against other hospitalisations. The largest relative increase was for encephalitis; the largest absolute difference was for cognitive deficits (for cardiac infections, 19.08% vs 6.71%, a difference of 12.37 percentage points). Children also had raised risks, but the absolute differences were significantly smaller than in adults. Infectious encephalitides were the leading risk factor for 7 of 14 disorders.

The breadth is the point: the association is not confined to COVID-19 or to infections of the nervous system. Because this is observational data from diagnostic codes, it cannot show that infection causes the disorders, and sicker patients are more likely to be seen and coded. It does justify asking about recent serious infection when a new psychiatric or cognitive presentation arrives.

• Ask about hospitalisation for infection in the past two years when assessing a new psychiatric or cognitive presentation in an adult.
• Look for cognitive complaints after serious infection; the absolute excess was largest for cognitive deficits.
• Keep encephalitis in mind as a cause of new-onset psychiatric symptoms, especially with fever, seizures or confusion.
• Expect smaller absolute risks in children than in adults after the same infections.

Why it matters

Post-infection brain health risk looks like a general feature of serious infection, not a COVID-19 peculiarity.

Don't overread it

These are associations from coded records; they do not show that infections cause the diagnoses.

The statistics, in plain English

The restricted mean time lost ratio compares how much disorder-free time each group loses over two years; above 1 means the infection group lost more. With over a million pairs, even small differences reach significance, so the absolute risk differences are the more useful numbers.

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.