Host Genetics versus Environment (Thrifty Genes) in COVID-19 severity and progression by PCA-based approaches

Background COVID-19 studies identified age, sex and selected comorbidities as key factors for the worst prognosis (PMID32737124). Host genetics also play a crucial role in determining clinical phenotypes among different populations. Actually, hundreds SNPs and haplotypes, characterizing distinct ethnicities (PMID26690364), have been identified as risk factors for COVID-19 progression (PMID36531241). Aim Investigating the potential role of candidate thrifty genes of main metabolism pathways in COVID-19 severity and prognosis. Methods We recruited 361 COVID-19 hospitalised Italian patients, before vaccine availability, together with clinical dataset. We genotyped several thrifty gene variants (APOE, ACE1, ACE2, ABO, OAS3, LZTFL1, TP53) previously identified by GWAS (PMID32558485). Alleles associated with severe COVID-19 have been identified by single and multiple analyses. Then, correlations among clinical data and gene variants were verified by logistic regressions based on Principal Component Analysis (PCA) coefficients. Results PCA showed 14 of the 32 PCs describe about 70% of the sample variance. Two PCs resulted significantly associated in a logistic regression model for intra-hospital death due to SARS-CoV-2 infection (PC1: OR=3.91; 2.56-5.96, p<0.001; PC2: OR=0.47; 0.33-0.68, p<0.001, respectively) indicating risk (PC1) and protective (PC2) role of the respective variable combinations. Additionally, PC7 correlated with MEWS (Modified Early Warning Score) (OR=1.47; 1.03-2.09; p=0.03). Conclusions Identifying host-specific genetic predictive markers for complex disease as COVID-19 may help in revealing molecular mechanisms with therapeutic relevance clarifying the mutual interactions of additional environmental risk factors. Accordingly, the genetic information obtained could boost preventive screening programmes and personalised clinical approaches. Since thrifty concept describes metabolic adaptation in which gene variants, leading to favourable phenotype, become more common among different populations in different environments, we could trace back information to design novel drugs, targeting the metabolic paths and contrast novel emerging diseases recognizing in advance the most susceptible population.