Inferring multi-organ genetic connections using imaging and clinical data through Mendelian randomization

Nat Biomed Eng. 2025 Nov 19:10.1038/s41551-025-01554-x. doi: 10.1038/s41551-025-01554-x. Online ahead of print.

Abstract

Understanding the complex relationships among major clinical outcomes and the interplay among multiple organs remains a considerable challenge. By using imaging phenotypes, we can characterize the functional and structural architecture of major human organs. Mendelian randomization (MR) provides a valuable framework for uncovering robust relationships between phenotypes by leveraging genetic variants as instrumental variables. Here we conduct a systematic multi-organ MR analysis involving 402 imaging traits and 372 clinical outcomes. Our analysis reveals 184 MR associations for 58 diseases and 56 imaging traits across various organs, tissues and systems, including the brain, heart, liver, kidney, lung, pancreas, spleen, adipose tissue and skeletal system. We identify intra-organ MR connections, such as the putative bidirectional genetic links between Alzheimer's disease and brain function, and interorgan associations, such as heart diseases and brain health. Metabolic disorders, such as diabetes, show genetically rooted putative MR effects across multiple organs. These findings shed light on the genetic links spanning multiple organs, providing targets for future mechanistic follow-up for clinical disease research.