Canonical representation of the human energy metabolism of lean mass, fat mass, and insulin resistance

Cyber-Physical System for Management and Self-Management of Cardiometabolic Health

“…Building on the above observations and reviewing energy perturbation studies from the international literature, we observed also a high level of correlation between weight, fat weight, and HOMA-IR [26]. We found also that our newly defined R-ratio and Rw-ratio showed highly significant correlation with HOMA-IR, and we proposed these measures as metrics for insulin resistance [2,3,26]. We recognized that monitoring R-ratio and Rw-ratio may give an important tool for monitoring changes of insulin resistance; we developed CPS for this purpose.…”

“…For demonstration purposes, I plugged the mandatory input variables ∆W k , ∆F k , and EB k as well as the known ingested macronutrient calories CI k , FI k , and PI k into SAM-HEM algorithm with Kalman filter [1][2][3]. As a measure of goodness of fit of the metabolic model SAM-HEM, I calculated the predicted mean value and standard deviation of the modeling error, i.e., model-predicted value minus the known trajectory of weight, fat weight, and lean mass.…”

“…Components of CPS are (A) a management software tool (MST); (B) a metabolic health monitor (MHM) app; (C) software on MHM capturing biometric signals from sensors of heart rate, physical activity, and pulse oximetry from a smart watch; and (D) software on MHM capturing biometric signals related to body composition and hydration status from Ori Diagnostic Instruments' (ODI) patented apparatus for impedance spectroscopy. MHM is running ODI's proprietary self-adaptive individualized stochastic mathematical model of the human energy metabolism (SAM-HEM) [1][2][3] via cloud computing. Based on our published simulation studies, SAM-HEM is a suitable concept to capture daily changes of the following SVs: weight; fat mass; lean mass; protein mass; intracellular water mass; extracellular water mass; utilized macronutrient intake and substrate oxidation of carbohydrate, fat, protein; and the R-ratio (ratio of the daily lean mass change velocity divided by the daily fat mass change velocity) which could be used as a surrogate marker for insulin resistance.…”

“…However, mathematical models can be created with input variables with easier realization in daily life such as weight and fat weight measurement with bioimpedance fat scale. Further, appropriately built mathematical models can provide indirect measurements of difficult-to-measure variables of the human energy metabolism like fat versus carbohydrate oxidation rate or changes of insulin resistance [1][2][3] and provide a solution to gain a special quantified insight into the fat and the entire energy metabolism. Currently, the computational model of the human energy metabolism (CM-HEM) [17] and its improved version [18] is considered the most complete.…”

“…CM-HEM is neither linear nor recursive nor individualized to a particular subject, and therefore, it is not suited to performing recursive parameter identification of the human energy metabolism, nor is it able to perform inverse calculation of utilized energy intake. Further, insulin resistance is not considered in CM-HEM when in fact insulin resistance plays a crucial role influencing fat and carbohydrate oxidation rates and the entire dynamic of body composition change [1][2][3]. Obviously, individualized models are needed which can be tied to easily measurable input variables such as weight and fat weight and provide insight into the fat and nonfat energy balance and change of insulin resistance.…”