Cognitive Trajectories After Hospitalization for COVID-19: A 36-Month Longitudinal Study.

OBJECTIVE

Neurological symptoms are commonly reported among individuals with postacute sequelae of COVID-19 (PASC), with cognitive impairment being the most common feature, but cognitive trajectories in PASC have not been clearly defined. The authors determined long-term cognitive changes in a cohort of individuals hospitalized with COVID-19, characterized their distinct cognitive trajectories, and identified factors associated with each trajectory at 36 months posthospitalization.

METHODS

The sample comprised 630 patients who were hospitalized with severe COVID-19 symptoms; 214 patients returned for at least one follow-up assessment over a 36-month period. Cognitive function, including attention, cognitive flexibility, processing speed, and memory, was evaluated by using BrainCheck, a validated digital platform, at 3-6, 12, 24, and 36 months posthospitalization. Longitudinal changes in cognitive performance and their associations with demographic factors, medical history, and neuropsychiatric symptoms were analyzed with linear mixed-effects models.

RESULTS

Four distinct cognitive trajectories over 36 months were identified and characterized. Group 1 (N=103, 48%) exhibited consistent normal cognitive function, group 2 (N=14, 7%) transitioned from unimpaired to impaired, group 3 (N=29, 14%) changed from impaired to unimpaired, and group 4 (N=68, 32%) showed persistent impairment across all domains. Poorer cognitive outcomes were associated with Hispanic ethnicity, although effects varied across domains.

CONCLUSIONS

As with many viral encephalitides, some patients showed stable normal or abnormal cognition or improved cognition over time. Surprisingly, however, a fourth subset exhibited delayed cognitive decline. This observation suggests that PASC-associated mechanisms, perhaps including chronic cerebral inflammation, may cause progressive cognitive impairment well after the initial infection clears.