Quantitative metabolism: data and mechanism as dual engines for a continuous, computable, and adaptive health framework

Abstract Metabolism is a dynamic, multiscale system, yet metabolic medicine still relies largely on static, threshold-based measurements that poorly capture longitudinal heterogeneity and early state transitions. We propose an operational framework for quantitative metabolism that integrates high-resolution measurements with mechanistically constrained computational models to represent metabolic states, predict trajectories, and guide intervention. The framework comprises four interconnected layers: multi-omics, defining individual molecular fingerprints; physiological and behavioral dynamics, capturing time-resolved exposures and proximal responses; phenotypic function, representing integrated organ- and system-level consequences; and adaptive metabolic regulation, extending toward digital twins and model-guided interventions such as digital therapeutics (DtX). Four sequential validation stages—quantifiable representation, predictive trajectory, operable intervention, and adaptive regulation—define the evidence required to advance from measurement to closed-loop clinical application. By replacing threshold-defined categories with continuous, time-resolved state representations, quantitative metabolism provides a structured research agenda for longitudinal metabolic assessment, mechanistically interpretable prediction, and prospectively validated personalized intervention across the life course.