Aging is a universal process of life, but its mechanisms are minimally understood. In particular, there is a need to identify biomarkers, and associated genetic features, that modulate lifespan variation across individuals. Companion dogs are a natural model of aging due to their shorter lifespans and sophisticated medical system, facilitating biomarker collection. Dog breed population structure enables genetic association studies with smaller sample sizes. Here, we use dogs as a model to discover aging-related biomarkers and relevant genomic variation.

We developed a reproducible genome-wide association study (GWAS) pipeline that removes human-specific assumptions. Our scalable pipeline enables association studies of hundreds of phenotypes across thousands of companion animals. We created the first dog GWAS catalog of 7,627 companion dogs enrolled in the Dog Aging Project. In 976 dogs, we investigated the genetic basis of 159 blood traits, including routine clinical analytes and metabolites. We find that blood traits are highly heritable in dogs, and many of these genetic associations in dogs replicate findings from human GWAS, supporting shared underlying biology.

We find that breed strongly influences variation in blood trait levels. Violating the common assumption that body size explains most of the lifespan variation across breeds, we find that potassium is associated with decreased breed median lifespan independent of breed average weight. High potassium levels are also associated with increased mortality risk in individual dogs. Our work illustrates the power of dogs as a model to identify blood-based predictors of lifespan and potentially develop interventions targeting these biomarkers.