Physiology Based Digestion and Absorption Modelling
Physiology Based Digestion and Absorption Modelling predicts the interaction of food with the body based on a large amount of information available from the physiological literature on gastroenterology and in vitro enzymatic digestion studies. The set of physiological mechanisms and variations between subjects that are included in the model is constantly updated in the one model fits all program.
The potential advantage of such a model is that it combines a vast amount of knowlegde available form the literature of diverse scientific research disciplines, including enzymology, gastroenerology, food digestion, food and health, pharmaceutics, medical science, sensory science and neurology. As a result, each simulation includes aspects that are difficult to oversee from the perspective of a single discipline. This allows interdisciplinary collaborations. the model also allows to predict and evaluate the effect of variations between individuals or groups within the population (infants, elderly, obese, the effect of diseases and surgical interventions) by variations in the model parameters. Because it is a one-model-fits-all-application, the model is far beyond giving a best fit to experimental data, but progressively builds on validated scientific knowledge, allowing to predict and evaluate the effect of variations in model parameters such as consumption rate, gastric emptying regualation, food composition, structure and buffering capacity, time-of-day effects on outcomes such as glycemic and insulinemic excurions, fullness and protein absoprtion.
The model is therefore very effective for predicting physiological outcomes for for example product development of healthy foods and pharmaceutical absorption in the fed state during or after consumption of different foods, and to assist in setting up new human trials and interpretation of trial resuls.
The model has been applied successfully in several applications, modelling the effect of food structure (solid versus liquid) on gastric processing and emptying, intestinal absorption kinetics, glycemic effects of realistic food products in a meal composed of different food matrials, e.g. a drink of water or a beverage together with a pasta meal with grated cheese and sauce, predicting aspects such as oral transit, gastric emptying rate, gastric tone, sensations of fullness and hunger, gastric solids content, gut transit speed, addition of digestive fluids such as saliva, gastric fluid, bile, pancreartic fluid materials, release of gut hormones such as CCK, GIP, GLP-1, PYY, Motilin, Gastrin and Secretin, nutrient absorption, blood glucose and insulin excursions. Some examples have been demonstrated recently at the Infogest conference in Granada 2019, the slides of this presentation "In silico modelling of the physiology of the digestive system, digestion and absorption" can be found below.
Opportunities
The existing model can be extended by adding new modules describing new aspects of the gastro-intestinal tract. Obviously, any further development of the model will be greatly supported by collaboration on these topics with experts in pharmaceutical and medical application and companies that are interested in the application for food or feed product development. I am therefore highly interested in collaboration with the academia and commercial companies to further develop knowledge and research tools, share insight and bring these to application through contract research.
Examples of possible developments are:
- put in a more detailed description of gastric acidification,
- improve the time-of-day effects already included in the model,
- improve the migrating motor complex already included in the model,
- include functionalities of the large intestine such as the dynamics of the intestinal microbiota,
- add physiological functionalities more remotely related to the gastrointestinal tract (for example the lymphatic transport of absorbed fat, liver function and skeletal muscle protein accretion).
- adapt the physiological parameters towards subgroups of the population (infants, elderly, ethnic variations), diseased (obese, insulin resistance, medical situations) and animal physiology (ruminants, poultry, pig, dog),
- predict effects of digestion and absorption in satiation from food and sugar craving,
- modelling gastro-instestinal absorption of nutrients and bioavalability of pharmaceuticals by cellular ttansport mechanisms.
- applying the in silico model for further development and alignment with in vitro digestion models, for example developing physiologically more realistic in vitro stomach and intestinal models by including feedback regulations from the in silico model.
THE CONCEPT
The body is a highly functional biosystem consisting of several integrated functional units (mouth, stomach, intestine, liver, brain, etc.) which has developed to optimally select, digest, absorb and metabolize food. If we want to optimize the composition, structure and timing of meals for optimal nutritional performance regarding for example nutrient delivery, glycemic effects and hunger suppression, we need to adapt the food to the conditions exposed by the digestive system. However, the difficulty is that the digestive system adjusts its digestive settings in response to the food. For that reason, in vitro digestion setups, which are non-reactive in general, fail to mimic the actual adapting conditions in the gastrointestinal tract. To overcome this problem, a literate-based in silico model for the gastrointestinal tract been developed, which calculates the digestive processes and transport of the food through the gastrointestinal, which include the adaptations of the digestive parameters (transit times, digestive fluid secretion, pH control by acidification and bicarbonate addition) and absorption.
The model is structured as a lineup of well-mixed compartments with different functionalities (plate, mouth, stomach (corpus, fundus, antrum), proximal duodemum, distal duodenum, jejunum, ileum, colon) representing the lumen of the gastrointestinal tract (see figure below). Each time step calculates the digestive processes within each compartment and the material transport between subsequent compartments. The small intestinal compartments are lined with compartments representing the epithelial surface that contain the receptor and absorptive cells. In between each luminal and epithelial compartment is a mucus compartment, which is programmed to behave as a selective filter, only allowing small molecular species such as small sugars, peptides, amino-acids and fatty acid bound to bile micelles to pass. Absorption is programmed to be competitive between the macronutrients and limited to a specific maximum per unit length of the small intestine in accordance to the experimental findings in pigs reported in literature. The epithelial compartments also release gut hormones (CCK, PYY, GLP-1 and GIP) in correspondence to the physiological sensitivity of the receptor I, K and L-cells and their typical distribution along the alimentary tract. The released hormones enter a simulated blood compartment, where they have a residence half time corresponding to their typical physiological values. Various settings of the model, such as the gastric tone, gastric emptying time and small intestinal transit times are adjusted by either the blood content of these hormones or their rate of release by the epithelial compartments (and a simulated neural coupling tot the target compartment). Also the release of digestive fluids (gastric acid, digestive enzymes, bile and water) are adjusted according to the detection of nutrients in the epithelial compartments (hence specific small molecular species that were able to pass the mucosal lining).
The stomach
As an example of the type of information that has been included in the model, the stomach will be discussed. The stomach has multiple physiological functions of which the most important ones are that it prepares the food for optimal digestion by the small intestine (e.g. solid chunks are broken down into a fluidic system of sufficiently small pieces), it is acidified to low pH to reduce microbial survival, and stored and released from the stomach at a rate that does not exceed the digestive and absorptive capacities of the small intestine.
Secretion of gastric fluids occurs mainly in the proximal region of the stomach (corpus + fundus). Gastric acid secretion is modelled according to experimental in vivo secretion rates described in the literature, by setting a basal secretion rate and activated release is modelled to be proportional to the deviation from a target pH in the antral region (where most pH receptors reside), in such a way that these together lead to a steady state pH of 2 in the antrum. Hence, gastric acid release dynamically adapts to the deviation from steady state, and will temporarily increase when food enters the stomach, similar to in vivo experimental observations.
Because of the relative slowness of this process of pH adjustment (among other caused by the distance between pH receptors and acid secretory glands), this system can become unstable and lead to temporarily excessive acid secretion, leading to a pH in the antral region far below 1, which would be harmful for the duodenum. Likely for that reason, physiology provides a secretin-mediated brake mechanism that reacts in response to a low duodenal pH, resulting an inhibition of gastric acidification and gastric emptying. This mechanism is also been included in the model.
Exchange of material between the proximal (fundus, corpus) and distal (antrum) regions of the stomach results in a redistribution of the gastric contents, which is modelled such that the gastric handling of any floating material (fat) and solid food material (e.g. curdled milk, cheese, meat) corresponds to the in vivo behavior described in the literature.
Based on observation from x-ray imaging movies of gastric emptying in dogs, it seems likely that the antral volume adjusts “as a balloon” to the gastric tone, which in turn is a linear function of the gastric volume, with a proportionality constant determined by the gut hormone CCK. Gastric tone is modelled such that it describes experimental gastric tone data reported in literature. The proportionality constant between antral volume ant gastric tone is adjusted in such a way that the gastric emptying rate of low-caloric viscous fluids (fast emptying) corresponds to in vivo experimental findings, giving a roughly exponential emptying with a half-emptying time. This adjustment of the antral volume depends on the transfer rate from the proximal to the antral compartments, which results in the visosity dependence of fast gastric emptying in accordance to the literature.
Solid material is actively transported from the antral to the proximal compartment by peristalsis at a rate that is roughly estimated from videos on this process. The model allows this solid material to be introduced in the form of solid food entering through the mouth, or be formed in the gastric compartment as a result of the interaction with gastric secretion (acid and pepsin), which is of particular relevance for modelling the effect of curdling of milk proteins.
Retropulsion of food from the antrum to the corpus and fundus occurs during contraction of the antrum, which in the extremes can lead to a complete emptying of the distal lumen into the small intestine (fast emptying), or a complete retropulsion into the proximal compartment. The ratio between these limiting processes determines to what extent the antral content is mixed with the content retained in the proximal stomach. During this retropulsion, the content that is retropelled is subject to considerable shear, which supports the break up any solid material (“gastric grinder”). The extent to which this breakup break up occurs is dependent on the properties of the solid material, and must be estimated for example by in vitro experiments.
Sensation of gastric fullness is calculated on the basis of blood hormone levels, gastric volume and gastric tone. A gastric tone that exceeds a critical value is associated with gastric discomfort.
Try it out for yourself
If you are intersterested in exploring a curring version of the model, please feel free to contact me at info@insightfoodinside.com. By doing so you can get access to an executable version of the programme in a dropbox share map. In the current version data input and output are delivered and produced as tekst files (extension .txt) in the same map as the executable program. The contents of these text files can be conveniently modified and read by using the meal and results excell documents. The dropbox shared map contains a few example cases of input and output exel files.
An example of a of the input and output repersenting daily diet of a normal individual
The following two excel files show the examples of the input data and simulated outcomes for a normal healthy adult, culculated with the trial program avaialble through thee dropbox share folder mentioned above. As can be seen, an indefinate number of meals or meal components can be included in the diet at any time from 5 different dishes or beakers, each with a freely defined maximum consumption rate, and each food material can be decomposed with adjustable half times of decomposition for the diffent units (mouth, stomach, small intestin) into other food materials or any of the nutrients. For each nutrient the model calculates the digestion and absorption speed and physiologically adjusts digestive fluids release and transit times between the diffrent compartments of the alimentary tract. Some post absorptive features, such as glycemic control, are also included in the demo program. In a more advanced version of the program, most of the modeling parameters that are now set for a normal healthy adult can be adjusted to personalized physiology (e.g. individual differentiation in saliva relaese, gastric emptying time, absorption capacity and insulin sensitivity, and instead of first order decay times for the desintegration of food materials into coumpunds and nutrients, also more advanced models can be included in the form of subunits.
Because the simulation program is still in development, the demonstration version has a expiration date.
Demonstration video about the application of the in silico digerstion model.
Selected Publications
Relating food emulsion structure and composition to the way it is processed in the Gastro Intestinal tract and physiological responses. What are the opportunities?
Food Biophysics 5(4) (2010) 258–283. http://link.springer.com/article/10.1007%2Fs11483-010-9160-5
Specific food structures suppress appetite through reduced gastric emptying rate.
American Journal of Physiology- Gastrointestinal and Liver Physiology, http://ajpgi.physiology.org/content/304/11/G1038
Differences in in vitro gastric behaviour between homogenized milk and emulsions stabilised by Tween 80, whey protein, or whey protein and caseinate.
Food Hydrocolloids 25(4) (2011) 781–788 http://www.sciencedirect.com/science/article/pii/S0268005X10002316