Genomewide Association Study of Severe Covid-19 with Respiratory Failure

Prognosis InfectiologyVirology
Ellinghaus D et al

Main result

Detection of significant cross-replicating associations with rs11385942 at locus 3p21.31 and with rs657152 at locus 9q34.2 in the meta-analysis of two case-control panels: OR = 1.77 (95%CI: 1.48-2.11), and OR = 1.32 (95%CI: 1.20-1.47), respectively.

At locus 3p21.31, the association signal covered the genes SLC6A20, LZTFL1, CCR9, FYCO1, CXCR6 and XCR1.

At locus 9q34.2, the association signal coincided with the A/B/O blood group locus; in this cohort, a specific analysis of blood groups showed a higher risk in blood group A than in other blood groups (OR = 1.45 [95% CI: 1.20-1.75]), and a protective effect in blood group O compared with other blood groups (OR= 0.65 [95% CI: 0.53-0.79]).


This study identified the 3p21.31 gene group as a locus of genetic susceptibility among COVID-19 patients with respiratory failure.

It also reported a potential involvement of the A/B/O blood types.

Strength of evidence Moderate

- The genetic study was well-made and used reliable detection techniques
- The statistical analysis conforms with current practices for genetic analyses, taking into account many potential confounding factors
- Lack of information on patient characteristics and lack of correlations with clinical data - notably their respiratory status


To identify the genetic factors potentially involved in COVID-19 development and evolution.


Pan-genomic association study.

Included 1,980 patients with a severe form of COVID-19 (defined as a respiratory failure) in seven hospitals within the Italian and Spanish epicentres of the SARS-CoV-2 pandemic in Europe.

Quality control and exclusion of population outliers lead to the inclusion of 835 patients and 1,255 controls in Italy as well as 775 patients and 950 controls in Spain.

In total, 8,582,958 single-nucleotide polymorphisms were analyzed in a fixed-effect meta-analysis (phenotypic associations with alleles were assessed separately for the Italian and Spanish case-control panels, using PLINK logistic regression with adjustment and stratification based on sex and age). The threshold for statistical significance was set at 5x10^-8.

bibliovid.org and its content are bibliovid property.

Legal Notice