Uncovering covariance patterns across energy balance traits enables the discovery of new obesity-related genes

Objective

Effective solutions to obesity remain elusive, partly owing to its root in a positive energy balance (EB), which stems from the interplay of numerous traits spanning body size and composition, diet, physical activity, and metabolic profile. Nevertheless, EB-contributing traits are typically studied in isolation. We integrate numerous EB-related traits measured in the UK Biobank to uncover the underlying patterns of EB and associated genes in study participants.

Methods

We used sparse factor analysis to integrate traits and performed genome-wide association analyses on the integrated phenotypes to elucidate EB-related genes and metabolic pathways. We performed pleiotropy analyses on candidate single-nucleotide polymorphisms to uncover the genetic basis of EB.

Results

We identified multiple genes and genomic regions associated with EB, including many that have previously not been directly associated with obesity measures (e.g., MIR5591, FNDC3B, ANAPC10, SULT1A1, AXIN1, SKIDA1, ERLIN1, DOCK7), which we validated using an independent subset of the UK Biobank dataset along with data from the Atherosclerosis Risk in Communities cohort. We found that the covariances in EB traits are primarily driven by genome-wide pleiotropic associations.

Conclusions

We offer new insight into EB patterns and the genetic basis of EB.