Exploring the relative contribution of genetic and external exposomic risk scores to allergies in elderly women

Allergies have been linked to immune dysfunction, genetics, and environmental factors. However, environmental exposures are often highly correlated similar to genetic predictors making the cumulative assessment of exposures difficult. Here, we aim to investigate the relative contribution of genetic variants as well as different individual and environmental factors on the presence or absence of allergies in 450 elderly German women enrolled in the SALIA cohort study living in the Ruhr area by using genetic risk score (GRS) and exposomal risk scores (ERS). We used the novel cross leverage scores (CLS) to select genetic variants to be included in the GRS. The weights of the risk scores were obtained through bootstrapped and cross-validated ridge regression. We characterized the relative contributions of the risk scores to presence of allergies such as atopic dermatitis, asthma, or allergic rhinitis using McFadden's Pseudo R-squared. Overall, our model was able to explain 11.13% of the variance of allergy diagnosis. The modest variance explained is consistent with prior work on complex polygenic and environmentally influenced traits, reflecting that no single exposure or domain is likely to fully capture individual risk. The GRS had the highest relative contribution at 3.80%, followed by the meteorological risk score with 1.13%. This method can easily be adapted to other diseases and can facilitate health risk assessments of exposomal factors. In addition, the results may aid policy-making, for example, by regulating specific sources of exposure.