Inferring Multi-Organ Genetic Causal Connections using Imaging and Clinical Data through Mendelian Randomization

Abstract

Functional and morphological architectures of major human organs have been well characterized using imaging biomarkers. Nevertheless, deciphering the causal relationships between imaging biomarkers and major clinical outcomes, as well as understanding the causal interplay across multiple organs, remains a formidable challenge. Mendelian randomization (MR) presents a framework for inferring causality by using genetic variants as instrumental variables. Here we report a systematic multi-organ MR analysis between 402 imaging biomarkers and 88 clinical outcomes. We identified 488 genetic causal links for 62 diseases and 130 imaging biomarkers from 9 organs, tissue, or systems, including the brain, heart, liver, kidney, lung, pancreas, spleen, adipose tissue, and skeleton system. We prioritized crucial intra-organ causal connections, such as the bidirectional genetic links between Alzheimer's disease and brain function, as well as inter-organ causal effects, such as the adverse impact of heart diseases on brain health. Our findings uncover the genetic causal links spanning multiple organs, offering a more profound understanding of the intricate relationships between organ imaging biomarkers and clinical outcomes.