《生命科学》 2026, 38(9): 1607-1619
人工智能技术赋能间歇性禁食的健康效应解析及其转化应用
摘 要：
间歇性禁食（intermittent fasting，IF）作为一种非药物干预策略，通过调控代谢重编程、增强细胞应激抵抗和重塑免疫微环境等机制，在促进健康及防治慢性非传染性疾病中展现出显著潜力。人工智能（artificial intelligence，AI）技术的迅速发展，为多模态数据整合、复杂机制解析及个体化干预优化提供了创新工具。本文系统综述了AI技术在揭示IF健康效应中的应用进展，重点探讨其在代谢调控、细胞应激反应与免疫调节机制解析中的作用，并评估其在个体化干预方案制定与动态监测中的潜在价值。通过结合多组学数据与机器学习（machine learning，ML）模型，AI技术为识别在IF中发挥作用的关键分子通路及预测健康获益提供了新的分析范式。最后，本文讨论了当前AI在饮食干预机制研究中面临的数据标准化、模型可解释性与跨场景泛化等挑战，并展望了AI驱动的精准营养与主动健康管理的未来方向。总之，AI技术正推动IF研究从经验性探索迈向机制性与个体化的新阶段，为慢病防控和健康促进提供了更科学、更精准的策略支撑。
通讯作者：杨阳 , Email:yyang93@shsmu.edu.cn 王慧 , Email:huiwang@shsmu.edu.cn
Abstract:
Chronic non-communicable diseases (NCDs) constitute a severe threat to the health of residents in China, impairing patients’ quality of life and imposing a heavy burden on individuals and society due to the rising costs of prevention and treatment. Currently, the focus of NCD prevention has shifted toward dietary intervention, a kind of non-invasive strategy: Formulate personalized dietary regimens based on patients’ disease-specific physiological characteristics and integrate them into daily life to effectively slow disease progression and reduce related burdens to individuals and society. Among dietary strategies, intermittent fasting (IF) has demonstrated significant potential in promoting health and preventing NCDs through mechanisms including metabolic reprogramming regulation, enhanced cellular stress resistance, and immune microenvironment remodeling. As an emerging dietary pattern, IF exerts metabolic effects through complex multi-level pathways. Diverse fasting protocols further expand the associated pathways and mechanisms. Comprehensive multi-omics analysis is therefore crucial to elucidate the specific characteristic mechanisms of each fasting mode and integrate core pathways of IF, thereby fully clarifying its health impacts. The rapid advancement of artificial intelligence (AI) technology shows promises in early diagnosis, dynamic monitoring, multi-dimensional data interpretation, and key information extraction. It provides innovative tools for multi-omics data integration, complex mechanism decipherment, and individualized intervention optimization. This review systematically summarizes the progress of AI applications in uncovering IF’s health effects. It focuses on AI’s roles in decoding metabolic regulation, cellular stress response, and immune modulation mechanisms, and evaluates its potential value in formulating personalized interventions and conducting dynamic diagnostic monitoring. By integrating multi-omics data with machine learning (ML) models, AI offers a novel analytical paradigm for identifying key molecular pathways of IF and predicting its health benefits. Finally, the review discusses current challenges of AI in dietary intervention research, such as data standardization, model interpretability, and cross-scenario generalization. It also prospects the future of AI-driven precision nutrition and proactive health management. Overall, AI is propelling IF research from empirical exploration to a new stage of mechanistic understanding and individualization, providing more scientific and precise support for NCD prevention, control, and health promotion.
Communication Author：YANG Yang , Email:yyang93@shsmu.edu.cn WANG Hui , Email:huiwang@shsmu.edu.cn
SciCloud投稿系统