Exploring Novel Biomarkers for Rosacea Through Cohort Study and Mendelian Randomisation

The current therapeutic landscape for rosacea is notably deficient in targeted medications, underscoring an urgent need for the identification of novel biomarkers. Utilising a longitudinal cohort of 54 306 individuals from the UK Biobank (UKB), we conducted a comprehensive assessment of the associations between 2923 serum proteins and the risk for rosacea. Our cohort analysis identified 18 proteins significantly associated with rosacea risk. Next, we complemented the two-sample Mendelian randomisation (TSMR) and Mendelian randomisation (SMR) analysis based on pooled data to identify genetic links between protein targets and rosacea. TSMR analysis refined this list to nine proteins demonstrating significant causal relationships with at least one form of rosacea. Further refinement through SMR and differential expression analysis reduced this to five key proteins, including four (ABHD14B, CHMP6, DBNL and MCFD2) that inhibit rosacea onset and one (MSR1) that promotes it. The differential expression of these five biomarkers was validated by multiple omics datasets as well as in vitro experiments. We calculated the protein score based on the expression levels of these proteins, noting that participants with higher scores demonstrated an increased incidence of rosacea. The integrative examination of proteomic and genetic data from a European adult cohort provides robust causal evidence for several proteins as promising new biomarkers for the development of rosacea treatments.