Health is influenced by interplay of molecular, physiological and environmental factors. To effectively maintain health and prevent disease, health-relevant relations need to be understood at multiple levels of biological complexity. Network-based methods provide a powerful platform for integration and mining of data and knowledge characterizing different aspects of health. Previously, we have reported physiological and gene expression changes associated with adaptation of murine epididymal white adipose tissue (eWAT) to 5 days and 12 weeks of high-fat diet (HFD) and low-fat diet feeding (Voigt et al. in Mol Nutr Food Res 57:1423-1434, 2013. doi: 10.1002/mnfr.201200671 ). In the current study, we apply network analysis on this dataset to comprehensively characterize mechanisms driving the short- and long-term adaptation of eWAT to HFD across multiple levels of complexity. We built a three-layered interaction network comprising enriched biological processes, their transcriptional regulators and associated changes in physiological parameters. The multi-layered network model reveals that early eWAT adaptation to HFD feeding involves major changes at a molecular level, including activation of TGF-β signalling pathway, immune and stress response and downregulation of mitochondrial functioning. Upon prolonged HFD intake, initial transcriptional response tails off, mitochondrial functioning is even further diminished, and in turn the relation between eWAT gene expression and physiological changes becomes more prominent. In particular, eWAT weight and total energy intake negatively correlate with cellular respiration process, revealing mitochondrial dysfunction as a hallmark of late eWAT adaptation to HFD. Apart from global understanding of the time-resolved adaptation to HFD, the multi-layered network model allows several novel mechanistic hypotheses to emerge: (1) early activation of TGF-β signalling as a trigger for structural and morphological changes in mitochondrial organization in eWAT, (2) modulation of cellular respiration as an intervention strategy to effectively deal with excess dietary fat and (3) discovery of putative intervention targets, such those in pathways related to appetite control. In conclusion, the generated network model comprehensively characterizes eWAT adaptation to high-fat diet, spanning from global aspects to mechanistic details. Being open to further exploration by the research community, it provides a resource of health-relevant interactions ready to be used in a broad range of research applications.