Clinical impact of an integrated e-health system for diabetes self-management support and shared decision making (POWER2DM): a randomised controlled trial.

AIMS/HYPOTHESIS: There is a lack of e-health systems that integrate the complex variety of aspects relevant for diabetes self-management. We developed and field-tested an e-health system (POWER2DM) that integrates medical, psychological and behavioural aspects and connected wearables to support patients and healthcare professionals in shared decision making and diabetes self-management. METHODS: Participants with type 1 or type 2 diabetes (aged >18 years) from hospital outpatient diabetes clinics in the Netherlands and Spain were randomised using randomisation software to POWER2DM or usual care for 37 weeks. This RCT assessed the change in HbA1c between the POWER2DM and usual care groups at the end of the study (37 weeks) as a primary outcome measure. Participants and clinicians were not blinded to the intervention. Changes in quality of life (QoL) (WHO-5 Well-Being Index [WHO-5]), diabetes self-management (Diabetes Self-Management Questionnaire - Revised [DSMQ-R]), glycaemic profiles from continuous glucose monitoring devices, awareness of hypoglycaemia (Clarke hypoglycaemia unawareness instrument), incidence of hypoglycaemic episodes and technology acceptance were secondary outcome measures. Additionally, sub-analyses were performed for participants with type 1 and type 2 diabetes separately. RESULTS: A total of 226 participants participated in the trial (108 with type 1 diabetes; 118 with type 2 diabetes). In the POWER2DM group (n=111), HbA1c decreased from 60.6±14.7 mmol/mol (7.7±1.3%) to 56.7±12.1 mmol/mol (7.3±1.1%) (means ± SD, p<0.001), compared with no change in the usual care group (n=115) (baseline: 61.7±13.7 mmol/mol, 7.8±1.3%; end of study: 61.0±12.4 mmol/mol, 7.7±1.1%; p=0.19) (between-group difference 0.24%, p=0.008). In the sub-analyses in the POWER2DM group, HbA1c in participants with type 2 diabetes decreased from 62.3±17.3 mmol/mol (7.9±1.6%) to 54.3±11.1 mmol/mol (7.1±1.0%) (p<0.001) compared with no change in HbA1c in participants with type 1 diabetes (baseline: 58.8±11.2 mmol/mol [7.5±1.0%]; end of study: 59.2±12.7 mmol/mol [7.6±1.2%]; p=0.84). There was an increase in the time during which interstitial glucose levels were between 3.0 and 3.9 mmol/l in the POWER2DM group, but no increase in clinically relevant hypoglycaemia (interstitial glucose level below 3.0 mmol/l). QoL improved in participants with type 1 diabetes in the POWER2DM group compared with the usual care group (baseline: 15.7±3.8; end of study: 16.3±3.5; p=0.047 for between-group difference). Diabetes self-management improved in both participants with type 1 diabetes (from 7.3±1.2 to 7.7±1.2; p=0.002) and those with type 2 diabetes (from 6.5±1.3 to 6.7±1.3; p=0.003) within the POWER2DM group. The POWER2DM integrated e-health support was well accepted in daily life and no important adverse (or unexpected) effects or side effects were observed. CONCLUSIONS/INTERPRETATION: POWER2DM improves HbA1c levels compared with usual care in those with type 2 diabetes, improves QoL in those with type 1 diabetes, improves diabetes self-management in those with type 1 and type 2 diabetes, and is well accepted in daily life. TRIAL REGISTRATION: ClinicalTrials.gov NCT03588104. FUNDING: This study was funded by the European Union's Horizon 2020 Research and Innovation Programme (grant agreement number 689444).

A computational model of the effects of macronutrients absorption and physical exercise on hormonal regulation and metabolic homeostasis.

Regular physical exercise and appropriate nutrition affect metabolic and hormonal responses and may reduce the risk of developing chronic non-communicable diseases such as high blood pressure, ischemic stroke, coronary heart disease, some types of cancer, and type 2 diabetes mellitus. Computational models describing the metabolic and hormonal changes due to the synergistic action of exercise and meal intake are, to date, scarce and mostly focussed on glucose absorption, ignoring the contribution of the other macronutrients. We here describe a model of nutrient intake, stomach emptying, and absorption of macronutrients in the gastrointestinal tract during and after the ingestion of a mixed meal, including the contribution of proteins and fats. We integrated this effort to our previous work in which we modeled the effects of a bout of physical exercise on metabolic homeostasis. We validated the computational model with reliable data from the literature. The simulations are overall physiologically consistent and helpful in describing the metabolic changes due to everyday life stimuli such as multiple mixed meals and variable periods of physical exercise over prolonged periods of time. This computational model may be used to design virtual cohorts of subjects differing in sex, age, height, weight, and fitness status, for specialized in silico challenge studies aimed at designing exercise and nutrition schemes to support health.

Exploring dynamic metabolomics data with multiway data analysis: a simulation study.

BACKGROUND: Analysis of dynamic metabolomics data holds the promise to improve our understanding of underlying mechanisms in metabolism. For example, it may detect changes in metabolism due to the onset of a disease. Dynamic or time-resolved metabolomics data can be arranged as a three-way array with entries organized according to a subjects mode, a metabolites mode and a time mode. While such time-evolving multiway data sets are increasingly collected, revealing the underlying mechanisms and their dynamics from such data remains challenging. For such data, one of the complexities is the presence of a superposition of several sources of variation: induced variation (due to experimental conditions or inborn errors), individual variation, and measurement error. Multiway data analysis (also known as tensor factorizations) has been successfully used in data mining to find the underlying patterns in multiway data. To explore the performance of multiway data analysis methods in terms of revealing the underlying mechanisms in dynamic metabolomics data, simulated data with known ground truth can be studied. RESULTS: We focus on simulated data arising from different dynamic models of increasing complexity, i.e., a simple linear system, a yeast glycolysis model, and a human cholesterol model. We generate data with induced variation as well as individual variation. Systematic experiments are performed to demonstrate the advantages and limitations of multiway data analysis in analyzing such dynamic metabolomics data and their capacity to disentangle the different sources of variations. We choose to use simulations since we want to understand the capability of multiway data analysis methods which is facilitated by knowing the ground truth. CONCLUSION: Our numerical experiments demonstrate that despite the increasing complexity of the studied dynamic metabolic models, tensor factorization methods CANDECOMP/PARAFAC(CP) and Parallel Profiles with Linear Dependences (Paralind) can disentangle the sources of variations and thereby reveal the underlying mechanisms and their dynamics.

Sleep duration and lipid metabolism in patients with diabetes mellitus: from the POWER2DM study.

This study assesses the association between sleep duration and plasma lipid profiles in people with diabetes mellitus (DM). Sleep duration data were obtained in 91 patients from the POWER2DM study (NCT03588104). The patients were divided in tertiles, based on their sleep duration, and blood samples were obtained at the beginning and after 9 months. Significant differences were found, specifically, patients in Tertile 3 (≥ 7.51 h) showed lower plasma levels of high-density lipoprotein cholesterol HDL-c (p < 0.05), apolipoprotein A1 (apo-A1; p < 0.05) and low HDL-c/apo-A1 ratio (p < 0.05). This study shows that sleep duration is associated with plasma lipid profiles in people with DM.

A Predictive Model-Based Decision Support System for Diabetes Patient Empowerment.

The main objective of POWER2DM is to develop and validate a personalized self-management support system (SMSS) for T1 and T2 diabetes patients that combines and integrates i) a decision support system (DSS) based on leading European predictive personalized models for diabetes interlinked with predictive computer models, ii) automated e-coaching functionalities based on Behavioral Change Theories, and iii) real-time Personal Data processing and interpretation. The SMSS offers a guided workflow based on treatment goals and activities where a periodic review evaluates the patients progress and provides detailed feedback on how to improve towards a healthier, diabetes appropriate lifestyle.

Determination of physiological parameters for endogenous glucose production in individuals using diurnal data.

BACKGROUND: Triple tracer meal experiments used to investigate organ glucose-insulin dynamics, such as endogenous glucose production (EGP) of the liver are labor intensive and expensive. A procedure was developed to obtain individual liver related parameters to describe EGP dynamics without the need for tracers. RESULTS: The development used an existing formula describing the EGP dynamics comprising 4 parameters defined from glucose, insulin and C-peptide dynamics arising from triple meal studies. The method employs a set of partial differential equations in order to estimate the parameters for EGP dynamics. Tracer-derived and simulated data sets were used to develop and test the procedure. The predicted EGP dynamics showed an overall mean R 2 of 0.91. CONCLUSIONS: In summary, a method was developed for predicting the hepatic EGP dynamics for healthy, pre-diabetic, and type 2 diabetic individuals without applying tracer experiments.

Actions of metformin and statins on lipid and glucose metabolism and possible benefit of combination therapy.

Patients with diabetes type 2 have an increased risk for cardiovascular disease and commonly use combination therapy consisting of the anti-diabetic drug metformin and a cholesterol-lowering statin. However, both drugs act on glucose and lipid metabolism which could lead to adverse effects when used in combination as compared to monotherapy. In this review, the proposed molecular mechanisms of action of statin and metformin therapy in patients with diabetes and dyslipidemia are critically assessed, and a hypothesis for mechanisms underlying interactions between these drugs in combination therapy is developed.

Feed-forward neural network model for hunger and satiety related VAS score prediction.

BACKGROUND: An artificial neural network approach was chosen to model the outcome of the complex signaling pathways in the gastro-intestinal tract and other peripheral organs that eventually produce the satiety feeling in the brain upon feeding. METHODS: A multilayer feed-forward neural network was trained with sets of experimental data relating concentration-time courses of plasma satiety hormones to Visual Analog Scales (VAS) scores. The network successfully predicted VAS responses from sets of satiety hormone data obtained in experiments using different food compositions. RESULTS: The correlation coefficients for the predicted VAS responses for test sets having i) a full set of three satiety hormones, ii) a set of only two satiety hormones, and iii) a set of only one satiety hormone were 0.96, 0.96, and 0.89, respectively. The predicted VAS responses discriminated the satiety effects of high satiating food types from less satiating food types both in orally fed and ileal infused forms. CONCLUSIONS: From this application of artificial neural networks, one may conclude that neural network models are very suitable to describe situations where behavior is complex and incompletely understood. However, training data sets that fit the experimental conditions need to be available.

The effect of plant sterols and different low doses of omega-3 fatty acids from fish oil on lipoprotein subclasses.

SCOPE: Consumption of a low-fat spread enriched with plant sterols (PS) and different low doses (<2 g/day) of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) from fish oil reduces serum triglycerides (TGs) and low-density lipoprotein-cholesterol (LDL-Chol) and thus beneficially affects two blood lipid risk factors. Yet, their combined effects on TG and Chol in various lipoprotein subclasses have been investigated to a limited extent. METHODS AND RESULTS: In a randomized, double-blind, placebo-controlled, parallel study, we determined TG and Chol in 13 LP subclasses in fasting serum of 282 hypercholesterolemic subjects, who consumed either a placebo spread or one of the four spreads containing PS (2.5 g/day) and EPA+DHA (0.0, 0.9, 1.3, and 1.8 g/day) for 4 weeks. After PS treatment, total LDL-Chol was reduced, which was not further changed by EPA+DHA. No shift in the LDL-Chol particle distribution was observed. The addition of EPA+DHA to PS dose-dependently reduced VLDL-Chol and VLDL-TG mainly in larger particles. Furthermore, the two highest doses of EPA+DHA increased Chol and TG in the larger HDL particles, while these concentrations were decreased in the smallest HDL particles. CONCLUSION: The consumption of a low-fat spread enriched with both PS and EPA+DHA induced shifts in the lipoprotein distribution that may provide additional cardiovascular benefits over PS consumption alone.

Predicting individual responses to pravastatin using a physiologically based kinetic model for plasma cholesterol concentrations.

We used a previously developed physiologically based kinetic (PBK) model to analyze the effect of individual variations in metabolism and transport of cholesterol on pravastatin response. The PBK model is based on kinetic expressions for 21 reactions that interconnect eight different body cholesterol pools including plasma HDL and non-HDL cholesterol. A pravastatin pharmacokinetic model was constructed and the simulated hepatic pravastatin concentration was used to modulate the reaction rate constant of hepatic free cholesterol synthesis in the PBK model. The integrated model was then used to predict plasma cholesterol concentrations as a function of pravastatin dose. Predicted versus observed values at 40 mg/d pravastatin were 15 versus 22 % reduction of total plasma cholesterol, and 10 versus 5.6 % increase of HDL cholesterol. A population of 7,609 virtual subjects was generated using a Monte Carlo approach, and the response to a 40 mg/d pravastatin dose was simulated for each subject. Linear regression analysis of the pravastatin response in this virtual population showed that hepatic and peripheral cholesterol synthesis had the largest regression coefficients for the non-HDL-C response. However, the modeling also showed that these processes alone did not suffice to predict non-HDL-C response to pravastatin, contradicting the hypothesis that people with high cholesterol synthesis rates are good statin responders. In conclusion, we have developed a PBK model that is able to accurately describe the effect of pravastatin treatment on plasma cholesterol concentrations and can be used to provide insight in the mechanisms behind individual variation in statin response.

Dietary medium chain fatty acid supplementation leads to reduced VLDL lipolysis and uptake rates in comparison to linoleic acid supplementation.

Dietary medium chain fatty acids (MCFA) and linoleic acid follow different metabolic routes, and linoleic acid activates PPAR receptors. Both these mechanisms may modify lipoprotein and fatty acid metabolism after dietary intervention. Our objective was to investigate how dietary MCFA and linoleic acid supplementation and body fat distribution affect the fasting lipoprotein subclass profile, lipoprotein kinetics, and postprandial fatty acid kinetics. In a randomized double blind cross-over trial, 12 male subjects (age 51±7 years; BMI 28.5±0.8 kg/m2), were divided into 2 groups according to waist-hip ratio. They were supplemented with 60 grams/day MCFA (mainly C8:0, C10:0) or linoleic acid for three weeks, with a wash-out period of six weeks in between. Lipoprotein subclasses were measured using HPLC. Lipoprotein and fatty acid metabolism were studied using a combination of several stable isotope tracers. Lipoprotein and tracer data were analyzed using computational modeling. Lipoprotein subclass concentrations in the VLDL and LDL range were significantly higher after MCFA than after linoleic acid intervention. In addition, LDL subclass concentrations were higher in lower body obese individuals. Differences in VLDL metabolism were found to occur in lipoprotein lipolysis and uptake, not production; MCFAs were elongated intensively, in contrast to linoleic acid. Dietary MCFA supplementation led to a less favorable lipoprotein profile than linoleic acid supplementation. These differences were not due to elevated VLDL production, but rather to lower lipolysis and uptake rates.

A new flexible plug and play scheme for modeling, simulating, and predicting gastric emptying.

BACKGROUND: In-silico models that attempt to capture and describe the physiological behavior of biological organisms, including humans, are intrinsically complex and time consuming to build and simulate in a computing environment. The level of detail of description incorporated in the model depends on the knowledge of the system's behavior at that level. This knowledge is gathered from the literature and/or improved by knowledge obtained from new experiments. Thus model development is an iterative developmental procedure. The objective of this paper is to describe a new plug and play scheme that offers increased flexibility and ease-of-use for modeling and simulating physiological behavior of biological organisms. METHODS: This scheme requires the modeler (user) first to supply the structure of the interacting components and experimental data in a tabular format. The behavior of the components described in a mathematical form, also provided by the modeler, is externally linked during simulation. The advantage of the plug and play scheme for modeling is that it requires less programming effort and can be quickly adapted to newer modeling requirements while also paving the way for dynamic model building. RESULTS: As an illustration, the paper models the dynamics of gastric emptying behavior experienced by humans. The flexibility to adapt the model to predict the gastric emptying behavior under varying types of nutrient infusion in the intestine (ileum) is demonstrated. The predictions were verified with a human intervention study. The error in predicting the half emptying time was found to be less than 6%. CONCLUSIONS: A new plug-and-play scheme for biological systems modeling was developed that allows changes to the modeled structure and behavior with reduced programming effort, by abstracting the biological system into a network of smaller sub-systems with independent behavior. In the new scheme, the modeling and simulation becomes an automatic machine readable and executable task.

Lipoprotein metabolism indicators improve cardiovascular risk prediction.

BACKGROUND: Cardiovascular disease risk increases when lipoprotein metabolism is dysfunctional. We have developed a computational model able to derive indicators of lipoprotein production, lipolysis, and uptake processes from a single lipoprotein profile measurement. This is the first study to investigate whether lipoprotein metabolism indicators can improve cardiovascular risk prediction and therapy management. METHODS AND RESULTS: We calculated lipoprotein metabolism indicators for 1981 subjects (145 cases, 1836 controls) from the Framingham Heart Study offspring cohort in which NMR lipoprotein profiles were measured. We applied a statistical learning algorithm using a support vector machine to select conventional risk factors and lipoprotein metabolism indicators that contributed to predicting risk for general cardiovascular disease. Risk prediction was quantified by the change in the Area-Under-the-ROC-Curve (ΔAUC) and by risk reclassification (Net Reclassification Improvement (NRI) and Integrated Discrimination Improvement (IDI)). Two VLDL lipoprotein metabolism indicators (VLDLE and VLDLH) improved cardiovascular risk prediction. We added these indicators to a multivariate model with the best performing conventional risk markers. Our method significantly improved both CVD prediction and risk reclassification. CONCLUSIONS: Two calculated VLDL metabolism indicators significantly improved cardiovascular risk prediction. These indicators may help to reduce prescription of unnecessary cholesterol-lowering medication, reducing costs and possible side-effects. For clinical application, further validation is required.

A systematic approach to obtain validated partial least square models for predicting lipoprotein subclasses from serum NMR spectra.

A systematic approach is described for building validated PLS models that predict cholesterol and triglyceride concentrations in lipoprotein subclasses in fasting serum from a normolipidemic, healthy population. The PLS models were built on diffusion-edited (1)H NMR spectra and calibrated on HPLC-derived lipoprotein subclasses. The PLS models were validated using an independent test set. In addition to total VLDL, LDL, and HDL lipoproteins, statistically significant PLS models were obtained for 13 subclasses, including 5 VLDLs (particle size 64-31.3 nm), 4 LDLs (particle size 28.6-20.7 nm) and 4 HDLs (particle size 13.5-9.8 nm). The best models were obtained for triglycerides in VLDL (0.82 < Q(2) <0.92) and HDL (0.69 < Q(2) <0.79) subclasses and for cholesterol in HDL subclasses (0.68 < Q(2) <0.96). Larger variations in the model performance were observed for triglycerides in LDL subclasses and cholesterol in VLDL and LDL subclasses. The potential of the NMR-PLS model was assessed by comparing the LPD of 52 subjects before and after a 4-week treatment with dietary supplements that were hypothesized to change blood lipids. The supplements induced significant (p < 0.001) changes on multiple subclasses, all of which clearly exceeded the prediction errors.

A physiologically based in silico kinetic model predicting plasma cholesterol concentrations in humans.

Increased plasma cholesterol concentration is associated with increased risk of cardiovascular disease. This study describes the development, validation, and analysis of a physiologically based kinetic (PBK) model for the prediction of plasma cholesterol concentrations in humans. This model was directly adapted from a PBK model for mice by incorporation of the reaction catalyzed by cholesterol ester transfer protein and contained 21 biochemical reactions and eight different cholesterol pools. The model was calibrated using published data for humans and validated by comparing model predictions on plasma cholesterol levels of subjects with 10 different genetic mutations (including familial hypercholesterolemia and Smith-Lemli-Opitz syndrome) with experimental data. Average model predictions on total cholesterol were accurate within 36% of the experimental data, which was within the experimental margin. Sensitivity analysis of the model indicated that the HDL cholesterol (HDL-C) concentration was mainly dependent on hepatic transport of cholesterol to HDL, cholesterol ester transfer from HDL to non-HDL, and hepatic uptake of cholesterol from non-HDL-C. Thus, the presented PBK model is a valid tool to predict the effect of genetic mutations on cholesterol concentrations, opening the way for future studies on the effect of different drugs on cholesterol levels in various subpopulations in silico.

Clustering by plasma lipoprotein profile reveals two distinct subgroups with positive lipid response to fenofibrate therapy.

Fibrates lower triglycerides and raise HDL cholesterol in dyslipidemic patients, but show heterogeneous treatment response. We used k-means clustering to identify three representative NMR lipoprotein profiles for 775 subjects from the GOLDN population, and study the response to fenofibrate in corresponding subgroups. The subjects in each subgroup showed differences in conventional lipid characteristics and in presence/absence of cardiovascular risk factors at baseline; there were subgroups with a low, medium and high degree of dyslipidemia. Modeling analysis suggests that the difference between the subgroups with low and medium dyslipidemia is influenced mainly by hepatic uptake dysfunction, while the difference between subgroups with medium and high dyslipidemia is influenced mainly by extrahepatic lipolysis disfunction. The medium and high dyslipidemia subgroups showed a positive, yet distinct lipid response to fenofibrate treatment. When comparing our subgroups to known subgrouping methods, we identified an additional 33% of the population with favorable lipid response to fenofibrate compared to a standard baseline triglyceride cutoff method. Compared to a standard HDL cholesterol cutoff method, the addition was 18%. In conclusion, by using constructing subgroups based on representative lipoprotein profiles, we have identified two subgroups of subjects with positive lipid response to fenofibrate therapy and with different underlying disturbances in lipoprotein metabolism. The total subgroup with positive lipid response to fenofibrate is larger than subgroups identified with baseline triglyceride and HDL cholesterol cutoffs.

Diagnostic markers based on a computational model of lipoprotein metabolism.

BACKGROUND: Dyslipidemia is an important risk factor for cardiovascular disease and type II diabetes. Lipoprotein diagnostics, such as LDL cholesterol and HDL cholesterol, help to diagnose these diseases. Lipoprotein profile measurements could improve lipoprotein diagnostics, but interpretational complexity has limited their clinical application to date. We have previously developed a computational model called Particle Profiler to interpret lipoprotein profiles. In the current study we further developed and calibrated Particle Profiler using subjects with specific genetic conditions. We subsequently performed technical validation and worked at an initial indication of clinical usefulness starting from available data on lipoprotein concentrations and metabolic fluxes. Since the model outcomes cannot be measured directly, the only available technical validation was corroboration. For an initial indication of clinical usefulness, pooled lipoprotein metabolic flux data was available from subjects with various types of dyslipidemia. Therefore we investigated how well lipoprotein metabolic ratios derived from Particle Profiler distinguished reported dyslipidemic from normolipidemic subjects. RESULTS: We found that the model could fit a range of normolipidemic and dyslipidemic subjects from fifteen out of sixteen studies equally well, with an average 8.8% ± 5.0% fit error; only one study showed a larger fit error. As initial indication of clinical usefulness, we showed that one diagnostic marker based on VLDL metabolic ratios better distinguished dyslipidemic from normolipidemic subjects than triglycerides, HDL cholesterol, or LDL cholesterol. The VLDL metabolic ratios outperformed each of the classical diagnostics separately; they also added power of distinction when included in a multivariate logistic regression model on top of the classical diagnostics. CONCLUSIONS: In this study we further developed, calibrated, and corroborated the Particle Profiler computational model using pooled lipoprotein metabolic flux data. From pooled lipoprotein metabolic flux data on dyslipidemic patients, we derived VLDL metabolic ratios that better distinguished normolipidemic from dyslipidemic subjects than standard diagnostics, including HDL cholesterol, triglycerides and LDL cholesterol. Since dyslipidemias are closely linked to cardiovascular disease and diabetes type II development, lipoprotein metabolic ratios are candidate risk markers for these diseases. These ratios can in principle be obtained by applying Particle Profiler to a single lipoprotein profile measurement, which makes clinical application feasible.

Simulating the physiology of athletes during endurance sports events: modelling human energy conversion and metabolism.

The human physiological system is stressed to its limits during endurance sports competition events. We describe a whole body computational model for energy conversion during bicycle racing. About 23 per cent of the metabolic energy is used for muscle work, the rest is converted to heat. We calculated heat transfer by conduction and blood flow inside the body, and heat transfer from the skin by radiation, convection and sweat evaporation, resulting in temperature changes in 25 body compartments. We simulated a mountain time trial to Alpe d'Huez during the Tour de France. To approach the time realized by Lance Armstrong in 2004, very high oxygen uptake must be sustained by the simulated cyclist. Temperature was predicted to reach 39°C in the brain, and 39.7°C in leg muscle. In addition to the macroscopic simulation, we analysed the buffering of bursts of high adenosine triphosphate hydrolysis by creatine kinase during cyclical muscle activity at the biochemical pathway level. To investigate the low oxygen to carbohydrate ratio for the brain, which takes up lactate during exercise, we calculated the flux distribution in cerebral energy metabolism. Computational modelling of the human body, describing heat exchange and energy metabolism, makes simulation of endurance sports events feasible.

A physiologically-based kinetic model for the prediction of plasma cholesterol concentrations in the mouse.

The LDL cholesterol (LDL-C) and HDL cholesterol (HDL-C) concentrations are determined by the activity of a complex network of reactions in several organs. Physiologically-based kinetic (PBK) computational models can be used to describe these different reactions in an integrated, quantitative manner. A PBK model to predict plasma cholesterol levels in the mouse was developed, validated, and analyzed. Kinetic parameters required for defining the model were obtained using data from published experiments. To construct the model, a set of appropriate submodels was selected from a set of 65,536 submodels differing in the kinetic expressions of the reactions. A submodel was considered appropriate if it had the ability to correctly predict an increased or decreased plasma cholesterol level for a training set of 5 knockout mouse strains. The model thus defined consisted of 8 appropriate submodels and was validated using data from an independent set of 9 knockout mouse strains. The model prediction is the average prediction of 8 appropriate submodels. Remarkably, these submodels had in common that the rate of cholesterol transport from the liver to HDL was not dependent on hepatic cholesterol concentrations. The model appeared able to accurately predict in a quantitative way the plasma cholesterol concentrations of all 14 knockout strains considered, including the frequently used Ldlr-/- and Apoe-/- mouse strains. The model presented is a useful tool to predict the effect of knocking out genes that act in important steps in cholesterol metabolism on total plasma cholesterol, HDL-C and LDL-C in the mouse.