COVID‐19 Infection and Dementia: Analyses of time‐varying risk, subtypes, and subpopulations from the UK Biobank

Background Although COVID‐19 patients were suggested to experience worse cognitive outcomes, there is a paucity of evidence on time‐varying risk of dementia, especially the subtypes, as well as among critical subpopulations. Method Out of over 50,000 individuals from a general population in the UK Biobank, SARS‐COV‐2 infected patients between March 1, 2020, and July 31, 2021 and maximally 5:1 propensity score matched contemporary non‐infected individuals were selected, with baseline dementia cases excluded. Matching was done on demographic characteristics, lifestyle, and comorbidities. Dementia was captured according to primary care, inpatient records, and death registry, with the follow‐up ending at the earliest of outcome occurrence, death, or August 31, 2021. Associations were evaluated using time‐varying hazard ratios (HRs) and odds ratios (ORs). Result With a mean age of 65.0 years for 18,032 COVID‐19 patients and 83,008 controls, participants were followed for a median of 247 (IQR: 204‐305) days and 255 dementia cases occurred, including 90 Alzheimer’s disease (AD) cases and 42 vascular dementia cases. Compared with matched controls, COVID‐19 infection was associated with increased risks of dementia during the acute phase (first 30 days), with HRs (95% CIs) being 12.77 (6.77, 24.08) for all‐cause dementia, 9.21 (2.77, 30.59) for AD, 5.53 (1.69, 18.11) for vascular dementia, and 25.35 (8.74, 73.56) for other dementia. Dementia risk declined drastically after COVID‐19 infection and sustained for all‐cause dementia, vascular dementia, and other dementia, while became non‐significant for AD. Among those not hospitalized during the acute phase, elevated dementia risk in the follow‐up remained for all‐cause dementia, vascular dementia, and other dementia, with ORs being 1.82, 4.55, and 1.64, respectively. Among most of the subpopulations classified by demographic characteristics, APOE genotype, and comorbidities (except for those with baseline chronic obstructive pulmonary diseases), COVID‐19 infection was associated with an elevated all‐cause dementia risk and no modification effect was detected. Conclusion Declined yet sustained elevated dementia risk since COVID‐19 infection was found, with differential trends for dementia subtypes. Increased dementia risk from COVID‐19 infection also applied for the non‐hospitalized during the acute phase and most subpopulations. The potential dementia risk associated with Omicron variants warrants further evaluation.