Answers are grounded in this paper and cite their sources.
Background The role of human leukocyte antigen (HLA) variation in COVID-19 disease severity remains uncertain, with large studies based on hospitalization endpoints often reporting weak or inconsistent associations. We hypothesized that these inconsistencies reflect limited phenotypic resolution, whereby broad clinical outcomes obscure HLA effects on specific acute symptom domains. Methods We performed an association analysis of imputed classical HLA alleles in SARS-CoV-2-positive participants from the Danish Blood Donor Study (DBDS; n = 15,666). Using logistic regression under a dominant carrier model, we identified HLA alleles associated with acute debilitation (being bedridden), mapped significant signals across 12 acute symptoms, tested for sex and vaccination interactions, and evaluated replication in the EFTER-COVID symptom subset of the Copenhagen Hospital Biobank (CHB; n = 13,322). Hospitalization outcomes were evaluated in the CHB-COVID19 cohort (n = 70,400; 3,449 hospitalizations). Results Three highly correlated alleles were associated with reduced odds of being bedridden, consistent with the AH8.1 ancestral haplotype. Conditional analyses could not resolve independent allele-specific associations. Symptom-level analyses identified a reproducible association with fever in both study cohorts (DBDS; OR = 0.87, p adj = 0.036: EFTER-COVID; OR = 0.88, p adj = 0.036). These associations remained significant after adjusting for pandemic period of infection and vaccination status at time of infection. A nominal sex interaction was observed in DBDS but did not survive multiple-testing correction or replicate independently in EFTER-COVID. No significant interactions were observed for vaccination status. No associations were detected with respiratory symptoms, post-COVID syndrome, or COVID-19 hospitalization (HLA-DRB1*03:01P: OR = 1.06, P = 0.38). Interpretation HLA variation linked to AH8.1 was associated with lower odds of acute debilitation and reproducibly associated with lower odds of acute fever. These findings demonstrate that phenotype-resolved analyses are critical for identifying reproducible immunogenetic associations in acute COVID-19.
No takes yet. Share an insight, caveat, or question.
Quinn et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: