Optimising personalised nutrition products based on individual micronutrient utilisation rate

Abstract

The invention provides a method of optimising the micronutrient formulation of a personalised nutrition product intended for use by an individual subject. the method comprises establishing an initial baseline micronutrient profile for the individual subject by performing an initial baseline assay on a sample obtained from the individual subject; establishing an individualised optimal micronutrient level, wherein the individualised optimal micronutrient level is calculated from a personal health dataset that relates to the individual subject; administering a baseline nutrition product to the individual subject for a first period of time; performing a first follow up analysis that comprises performing a first follow up assay on a first follow up sample obtained from the individual subject to establish a first follow up micronutrient profile for the individual subject, and determining a micronutrient utilisation rate (MUR) for one or more micronutrients within the individual subject. Systems utilizing the methods are also provided.

Description

[0001] OPTIMISING PERSONALISED NUTRITION PRODUCTS BASED ON INDIVIDUAL MICRONUTRIENT UTILISATION RATE

[0002] FIELD OF THE INVENTION

[0003] This invention relates to the technical field of human nutrition, and nutritional supplement, or nutritional guidance tailored to individual characteristics and needs. In particular, the invention relates to methods and systems for measuring an individual’s micronutrient utilisation rates (MUR) and optimising a micronutrient formulation of a personalised nutrition product based on this assessment.

[0004] BACKGROUND OF THE INVENTION

[0005] The present invention relates to the field of personalised nutrition products and optimisation of the personal nutritional products, specifically micronutrient formulation.

[0006] Micronutrients encompassing essential vitamins and minerals play a pivotal role in human health by facilitating enzymatic reactions as cofactors and coenzymes, contributing to energy metabolism by assisting in the catabolic process of macronutrients, and maintaining enzyme structure and function. These nutrients are indispensable for immune system function, bone health, antioxidant defence, neurological activities, and hormone regulation. While macronutrients provide the necessary energy and building blocks for bodily functions, micronutrients are critical to maintaining protein and cellular processes, as well as metabolic processes.

[0007] To date, formulating personalised nutrition products has generally been limited to identifying individual variability in nutritional needs, through measured factors such as genetics, lifestyle, and health conditions. However, the type and amount of personal health data acquired is not sufficient to capture the complexity and nuances of an individual's micronutrient health profile and characterise their micronutrient-metabolism type. Specifically, the collected health data does not account for characterising or predicting an individual need for micronutrients or an individual’s micronutrient utilisation rate (MUR) for each micronutrient. An individual’s MUR predicts the cumulative impact of a regularly consumed quantity of a micronutrient on their baseline micronutrient status. It is necessary to calculate MUR in order to precisely formulate a personalised nutrition product which can predictably influence an individual’s micronutrient status, and thereby address and prevent deficiencies, while avoiding potential over-dosing. US patent No. 5,542,420 discloses a system for prescribing a suitable diet to users based on their personalised health characteristics. Information about food consumed and medical data is entered at terminals and communicated to a ‘health computer’ which then determines the dietary prescription. However, the health characteristics used to prescribe are limited to a database of messages indicating food consumption. W02022094670 discloses a method for providing a personalised caloric and / or macronutrient intake profile for a user. However, the types of data collected, including bioimpedance measurements or anthropometric measurements, and the types of data generated, such as basal metabolic rate, do not address individual requirements for micronutrients. US2023038011 provides a method of creating a genetically personalised intravenous or intramuscular nutrition therapy, US2019145988 provides a method for providing nutritional supplement information for a subject, and WO2023119203 discloses a method for implementing a nutritional supplement plan for a user, however, these processes do not address optimisation of a micronutrient supplement over a period of time, such as by prediction of micronutrient utilisation rates. Similarly, US2017156386 discusses preparing a personalised micronutrient composition, and US2010266723 discloses the methods for formulating and customising the nutritional supplements. However, the current state of the art lacks a systematic model or a method that addresses the dynamic nature of the optimal micronutrient levels and micronutrient utilisation rates in individuals or population sub-groups in order to optimise the personalised nutrition product compositions.

[0008] Therefore, it would be highly desirable to have an alternative method of optimising the micronutrient formulation by incorporating the measured metabolic properties of an individual subject, including micronutrient utilisation rates (MUR), which are required to be able to precisely formulate a personalised nutrition product which predictably influences an individual’s micronutrient status, and thereby address and prevent deficiencies, while avoiding potential over-dosing.

[0009] The present invention seeks to overcome or at least alleviate one or more of the problems found in the prior art.

[0010] SUMMARY OF THE INVENTION

[0011] The present invention provides a new method of identifying an individual’s micronutrient-metabolism type, by calculating micronutrient utilisation rates (MUR) and then optimising the micronutrient formulation of a personalised nutrition product accordingly. An individual’s MUR predicts the cumulative impact of a regularly consumed quantity of a micronutrient upon their baseline micronutrient status. It is necessary to calculate MUR in order to precisely formulate a personalised nutrition product which can predictably influence an individual’s micronutrient status, and thereby address and prevent deficiencies, while avoiding potential over-dosing. The invention may be particularly useful for an individual subject that is administering the micronutrient formulation to reach an individualised optimal micronutrient level.

[0012] Embodiments of the invention involve a method of optimising a micronutrient formulation, comprising: establishing an initial baseline micronutrient profile from a sample obtained from an individual subject; establishing an individualised optimal micronutrient level calculated from a personal health dataset; administering a baseline nutrition product to the individual subject for a first period of time; performing a first follow up analysis to establish a first follow up micronutrient profile for the individual subject; determining the MUR for one or more micronutrients within the individual subject; identifying a requirement for increasing or decreasing certain micronutrients, factoring in MUR, for the next follow up analysis to reach an individualised optimal micronutrient level; and formulating the personalised nutrition product so that it satisfies the requirements for reaching the individualised optimal micronutrient level when administered to the individual subject. Accordingly, a first aspect of the invention provides a method of optimising the micronutrient formulation of a personalised nutrition product intended for use by an individual subject, the method comprising: establishing an initial baseline micronutrient profile for the individual subject by performing an initial baseline assay on a sample obtained from the individual subject; establishing an individualised optimal micronutrient level, wherein the individualised optimal micronutrient level is calculated from a personal health dataset that relates to the individual subject; administering a baseline nutrition product to the individual subject for a first period of time; performing a first follow up analysis that comprises performing a first follow up assay on a first follow up sample obtained from the individual subject to establish a first follow up micronutrient profile for the individual subject, and determining a micronutrient utilisation rate (MUR) for one or more micronutrients within the individual subject by:

[0013] (i) identifying any change in one or more micronutrient levels in the first follow up micronutrient profile in comparison to that of the initial baseline micronutrient profile;

[0014] (ii) identifying an amount of the micronutrient consumed from the baseline nutrition product during the first period of time;

[0015] (iii) identifying an amount of the micronutrient consumed from a diet excluding a baseline nutrition product during the first period of time;

[0016] (iv) identifying an amount of the micronutrient used by an individual during the over a first period of time;

[0017] (v) identifying a requirement for an increase or decrease in micronutrients for the first follow up micronutrient profile to reach an individualised optimal micronutrient level; and

[0018] (vi) formulating the personalised nutrition product so that it satisfies the requirements for reaching the individualised optimal micronutrient level when administered to the individual subject.

[0019] In some embodiments, a subsequent follow up analysis can be performed, followed by identifying any change in micronutrient level, and reformulation of the personalised nutrition product which satisfies the requirements for additional micronutrients in the individual subject.

[0020] A second aspect of the invention provides for a computer-implemented method for calculating a Micronutrient Utilisation Rate (MUR) for an individual subject, comprising: receiving input data comprising an individual subject's baseline micronutrient profile, and one or more additional factors selected from: a first follow-up micronutrient profile; an amount of a micronutrient consumed from a diet in a first period of time; an amount of a micronutrient consumed from a baseline nutrition product in a first period of time; an amount of a micronutrient used over a first period of time; and a personal health dataset; constructing a predictive model using a machine learning model comprising (i) an input layer configured to receive the baseline micronutrient profile and the one or more additional factors,

[0021] (ii) one or more hidden layers, wherein a value for the MUR can be inferred from the latent variables of the one or more hidden layers that transform baseline micronutrient profiles into predicted follow-up micronutrient levels, and

[0022] (iii) an output layer configured to provide a predicted follow-up micronutrient profile; training the predictive model using the input data; applying the trained predictive model to generate the predicted follow-up micronutrient profile; estimating MUR values by inverse inference using the trained predictive model, by identifying latent variables of the hidden layers that minimises the error between predicted and actual micronutrient levels at follow-up.

[0023] A third aspect of the invention provides a system for identifying a micronutrient utilisation rate (MUR) of a plurality of micronutrients in an individual subject, comprising: a database that is adapted to store a record of micronutrient levels identified as present within in a sample of a body fluid obtained from the individual subject at a specific timepoint; a device that collects data indicating metabolic status of an individual; a server; and a processor in communication with the server and the database, the processor being configured to calculate the MUR.

[0024] A fourth aspect of the invention provides a method of optimising the micronutrient formulation of a personalised nutrition product intended for use by an individual subject comprised within a metabolic group, comprising: establishing one or more metabolic groups by clustering individual subjects with similar rates of uptake for one or more micronutrients; identifying an average optimal micronutrient level for the one or more metabolic groups by averaging an optimal micronutrient level for individuals comprised within the one or more metabolic groups; identifying an average first follow up micronutrient profile for the one or more metabolic groups by averaging first follow up micronutrient profile for the one or more metabolic groups; establishing a metabolism type dataset for the one or more metabolic groups that incorporates the average optimal micronutrient level and the average first follow up micronutrient profile; classifying an individual subject to the one or more metabolic groups if an individual metabolism type dataset for a given micronutrient lies within three standard deviations of the metabolism type dataset of the metabolic group for the same micronutrient; identifying a requirement for additional micronutrients for the average first follow up micronutrient profile to reach the average optimal micronutrient level; and formulating the personalised nutrition product so that it satisfies the requirements for reaching the average optimal micronutrient level when administered to an individual subject of the metabolic group.

[0025] Accordingly, in embodiments the invention relates to a method for defining a metabolic group comprising of individuals with similar metabolism dataset, in terms of the rate of utilisation of different micronutrients. The invention further relates to a method of assigning an individual subject to the metabolic group, and formulating a nutritional product that will satisfy the average optimal micronutrient level for the metabolic group that an individual is associated with.

[0026] In aspects and embodiments, the rate of utilisation of micronutrients is calculated by: identifying a change in one or more micronutrient levels in the first follow up baseline micronutrient profile in comparison to that of the initial baseline micronutrient profile; identifying a total amount of micronutrients consumed from the baseline nutrition product during the first period of time and from the individuals diet in comparison to the amount of micronutrient analytes present in the individual samples at the followup analysis; and predicting the amount of micronutrients used by an individual during the first period of time, typically factoring in variables including age, medical conditions, medications, diet, lifestyle, BMI, waist and hip circumference, body composition (fat, lean, and water mass), blood pressure, hydration levels, physical activity, and strength.

[0027] In some embodiments, micronutrient utilisation rates (MUR) are calculated with further information concerning dietary intake and / or lifestyle, for example from fitness and health trackers, and this further information is stored in a database that characterises a micronutrient-metabolic status for the individual. The data characterising the micronutrient-metabolic status of the individual includes at least one indicator of demographic status and / or medical status.

[0028] In some embodiments, the sample obtained from the individual subject can be selected from one or more of the following: whole blood; serum; plasma; stool; urine; sweat; cerebrospinal fluid; tears; saliva; solid tissue biopsy; or exhaled breath.

[0029] Further, a computer implemented method can be used to identify various parts of the invention, including a rate of utilisation of micronutrients, an amount of micronutrient consumed from a baseline I individualised nutritional product, and / or an amount of micronutrient consumed from a diet excluding a generic nutrition product.

[0030] It will be appreciated that any features of one aspect or embodiment of the invention may be combined with any combination of features in any other aspect or embodiment of the invention, unless otherwise stated, and such combinations are envisaged and are intended to be directly and unambiguously disclosed herein, and to fall within the scope of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The invention is further illustrated by the accompanying drawings in which:

[0032] Figure 1 Shows a schematic diagram of an embodiment of a system of the invention. Micronutrient Utilisation Rate (MUR) Predictor, uses few-shot supervised artificial intelligence (Al) models to convert patient-reported variables into a prediction of the expected MUR value and its components, with the possible inclusion of error bars (confidence intervals, variance around the expected variables) or similar. Multiple model families may be considered, including probabilistic regularised parametric models, non-parametric models, neural networks (shallow or deep, with or without skip connections, with or without attention layers), as well as resampling-based methods such as ensembles of tree-based classifiers adapted for prediction of ordinal target variables (ordinal regression) or continuous target variables (regression).

[0033] Figure 2 Shows a schematic diagram of an embodiment for an MUR recommender that employs approximate inverse inference methods to recommend personalised nutrition products. These recommendations are based on current and previous MUR variables, as well as patient-reported variables, including blood tests and questionnaires collected at baseline and / or one month. The specific methods involve variational Bayesian inference, Markov Chain Monte Carlo (including Gibbs, Metropolis-Hastings, Hamiltonian Monte Carlo), loopy belief propagation, and expectation propagation for inferring optimal personalised nutritional interventions from MUR and patient-reported variables.

[0034] Figure 3 Shows a schematic diagram of an example of an embodiment of the invention outlining the timepoints of initial baseline assay and the follow up assay after 28 days. In the example, to obtain the initial baseline micronutrient profile, the blood sample was taken following fasting, along with additional anthropometric measurements, such as waist: hip ratio, to establish a personal health data dataset. The participants then orally consumed a sachet of organic or synthetic baseline nutritional product. Another blood sample was taken at 2 hours point to measure the acute blood vitamin and mineral concentration response to the consumption of the product. The participants then orally consumed an organic or synthetic generic nutrition product one sachet per day, for 28 days. Following 28 days, the follow up assay was performed to establish the follow-up micronutrient profile.

[0035] DETAILED DESCRIPTION OF THE INVENTION

[0036] All references cited herein are incorporated by reference in their entirety. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0037] Prior to setting forth the invention, a number of definitions are provided that will assist in the understanding of the invention

[0038] As used herein, the term ‘comprising’ means any of the recited elements are necessarily included and other elements may optionally be included as well. ‘Consisting essentially of’ means any recited elements are necessarily included, elements that would materially affect the basic and novel characteristics of the listed elements are excluded, and other elements may optionally be included. ‘Consisting of’ means that all elements other than those listed are excluded. Embodiments defined by each of these terms are within the scope of this invention.

[0039] The term ‘micronutrients’ in the context of the present invention means a chemical element or substance that is essential in minute or trace amounts to the human body. Micronutrients are essential for growth, development, in normal homeostasis and for regulating immune function amongst other key functions within the body. Micronutrients may include vitamins and minerals, such as elemental substances. Some non-limiting examples of vitamin micronutrients include the vitamins A, B1 , B2, B3, B5, B6, B7, B9, B12, C, D, E and K. Some non-limiting examples of minerals include iron, boron, calcium, zinc, magnesium, potassium, sodium, iodine, molybdenum, selenium, copper and folate. It will be appreciated that micronutrients may exist in elemental, compound, salt or chelated forms. In addition, certain micronutrients may exist in precursor metabolite form within food or drink that is subject to one or more biochemical conversions in the cell prior to attaining the final active form. Vitamins can occur in a variety of related forms known as vitamers. A vitamer of a particular vitamin is one of several related compounds that performs the functions of said vitamin and prevents the symptoms of deficiency of said vitamin. Micronutrients may, thus, include vitamers and bioavailable precursors of specified vitamins and dietary minerals.

[0040] The term ‘nutrition product’ as used herein refers to products that either supplement nutrition or provide part or all the daily nutritional requirements. Nutritional products can contain micronutrients, fatty acids, or other nutrients, or a combination of those. Administering nutrition products can result in compensating for deficiencies of a certain nutrient, enhancing athletic performance, or promoting specific health goals such as supporting bone health. Nutritional products may also comprise pre- or probiotics, or other complex formulations of macromolecules that may contribute to improving the nutritional balance of the recipient individual. A personalised nutrition product will comprise a personalised and optimised micronutrient composition. The personalised nutrition product may be in the form of a pill, tablet, capsule, syrup, liquid composition, powder, food product, food additive or shake. In a specific embodiment of the invention the personalised nutrition product may further include one or more ingredients including, protein powder, flavourings, emulsifiers, stabilisers, bulking agent, preservatives, and colourings. In one embodiment of the invention, the personalised nutrition product is in the form of a powdered protein shake composition that may be reconstituted by the addition of a liquid, such as water.

[0041] ‘Organic micronutrients’ refers to micronutrient compounds sourced directly from non-synthetic organic sources, such as from extracts or derivatives of plants and fungi. In certain instances, the organic sources are grown without exposure to synthetic pesticides, growth hormones, antibiotics, modern genetic engineering techniques (including genetically modified crops), chemical fertilisers, or sewage sludge. Organic micronutrients may be comprised within formulations and compositions that preserve their original structure as found in the original sources, thereby leading to improved bioavailability and stability. Where micronutrients comprise minerals (e.g. metal ions such as iron, sodium, or non-metals such as iodine, sulphur or phosphorous) an organic micronutrient may comprise them in a chelated or biologically complexed form that makes them more readily available for utilisation in the body when consumed.

[0042] The term ‘synthetic micronutrients’ refers to micronutrients typically synthesised in a laboratory or via an industrial process. Laboratory or industrially produced synthetic micronutrients may include products of chemical or biochemical synthetic routes, chemical extracts of minerals from mined ores, or biotechnological extracts from cells or microorganisms grown in culture. Although synthetic micronutrients often share a similar chemical composition with their natural counterparts, they can differ in structure, charge or morphology from the organic micronutrients.

[0043] An ‘individual subject’ refers to an individual who is a participant in the activity of receiving administration of a nutrition product. In embodiments of the invention, the individual is a mammal, suitably selected from a human, monkey, bear, rat, mouse, rabbit, guinea pig, pig, dog, cat, goat, sheep, horse or cow. Typically, the individual subject is a human and may be a human patient in need of therapeutic intervention. In a specific embodiment the individual subject is a from one or more specific populations or groups, such as elderly (i.e. geriatric), infant or neonate, child (i.e. paediatric), adult, a male, a female, a pregnant female, lactating female, immunocompromised, bariatric, prediabetic, and / or diabetic. In a specific embodiment of the invention, the individual subject has a cancer.

[0044] The term ‘sample’ refers to a sample obtained from the individual subject, which comprises material of biological origin that is sufficient to allow for a determination of micronutrient status to be made. In one embodiment the sample may comprise at least one of the following: whole blood; serum; plasma; stool; urine; sweat; cerebrospinal fluid; tears; saliva; solid tissue biopsy; or exhaled breath.

[0045] The term ‘first follow up sample’ refers to the sample obtained from the individual subject following a first period of time. Likewise, second or further follow up samples refer to samples obtained from the individual subject following a second or further periods of time respectively.

[0046] The term ‘micronutrient profile’ refers to the specific composition and level of one or more micronutrients within an individual subject. In a specific embodiment, the micronutrient profile comprises a quantitative measure of a plurality of micronutrients in a sample obtained from an individual subject, and / or an in silico prediction of the levels of one or more micronutrients within an individual subject derived from data comprised from one or more samples.

[0047] The term ’level’ or ’levels’ refers to a quantitative or relative measurement of a specific micronutrient. Levels may be determined in relation to objective concentrations, absolute quantities or by reference to understood standards such as functional units of activity, as appropriate to the specific micronutrient under consideration.

[0048] ‘Baseline micronutrient profile’ means a measurement or a value of a micronutrient profile that is maintained by an individual for a period of time. This might be the profile at which an individual has as a starting point (initial baseline), or where the profile consistently stays at another point in time (e.g. after nutrition intervention). Baseline profiles are determined from baseline assays which are taken when an individual is fasted, to avoid direct short-term impacts from food / drink that is consumed.

[0049] ‘Initial baseline assay’ refers to the first assessment or measurement taken from a sample from an individual that indicates the initial baseline micronutrient profile. Baseline profiles are determined from baseline assays which are taken when an individual is fasted, to avoid direct short-term impacts from food / drink that is consumed.

[0050] ‘First follow up assay’ refers to the follow up assessment or measurement taken from a first sample from an individual following the first period of time. Likewise, second or further follow up assays refer to assays performed following a second or further periods of time respectively.

[0051] ‘Baseline nutrition product’ means a reference nutritional product that comprises a defined composition of nutrients, including non-personalised or non-optimised micronutrients. The baseline nutrition product may be described as a generic, or non-optimised, nutritional product. Hence, the formulation of a baseline nutrition product is not individually optimised to a person or metabolic grouping but instead comprises generic levels of micronutrients, usually set at general adult population recommended daily rates or ranges. A baseline generic nutrition product may consist essentially of one or more micronutrients derived from organic sources and / or synthetic sources. These generic levels may be informed by standard metrics of recommended daily consumption. An exemplary baseline generic nutrition product may include one or more ingredients including, protein powder, flavourings, emulsifiers, stabilisers, bulking agent, preservatives, and colourings. In one embodiment of the invention, the baseline generic nutrition product is in the form of a powdered protein shake composition that may be reconstituted by the addition of a liquid, such as water. In other embodiments, the baseline nutrition product is in the form of an oral enteral feeding composition. In a further embodiment, the baseline nutrition product is in the form of a parenteral feeding composition.

[0052] ‘Individualised optimal micronutrient level’ refers to an ideal or recommended amount of micronutrients that an individual should consume for optimal health and well-being.

[0053] ‘Metabolism’ is defined as the chemical reactions that occur within the body's cells that control energy utilisation as well as a range of biochemical processes that underpin the physiology of an individual subject. Metabolism consists of two different processes: anabolism and catabolism. Anabolism refers to the set of metabolic processes that build complex molecules from simpler ones. Anabolic processes require energy and are involved in the synthesis of proteins, nucleic acids, and other essential molecules needed for growth and repair. Catabolism is the set of metabolic processes that break down complex molecules into simpler ones, sometimes releasing energy in the process. Catabolic processes are responsible for the breakdown of nutrients, such as carbohydrates, fats, and proteins, to generate energy.

[0054] ‘Metabolism type dataset’ is defined as a dataset that represents one or more metabolic characteristics of an individual subject. This dataset is typically used to determine a micronutrient-metabolism type. The metabolism type dataset may incorporate an average optimal micronutrient level, and an average first follow up micronutrient profile. Further, the metabolism type dataset may comprise a level of a specialised protein, such as ferritin and / or transferrin, that is critical to maintaining micronutrient levels in the body of an individual subject, by regulating iron distribution.

[0055] The metabolism type dataset may comprise readouts of various markers and parameters relating to metabolic level such as HbA1 c, fasting glucose, haemoglobin, blood pressure, bioelectrical impedance analysis (BIA), lipids (High Density Lipoprotein, Low Density Lipoprotein, triglycerides, cholesterol, Fat mass percentage), and Homeostasis Model Assessment of Insulin Resistance (HOMA-IR) provided detailed profiles of metabolic health. Further, inflammatory markers, including C-reactive protein (CRP), N-terminal pro B-type natriuretic peptide (NT-pro-BNP), and various cytokines such as Granulocyte- Macrophage Colony-Stimulating Factor (GM-CSF), Interferon-alpha (IFN-y), Interleukin 6 (IL-6), and Tumour Necrosis Factor (TNF-a), may be tracked to assess systemic inflammation and its correlation with metabolic changes. It will be appreciated that levels of other inflammatory cytokines may be monitored, including various interleukins.

[0056] The term ‘bioavailability’ or ‘nutrient bioavailability’ refers to the proportion of a nutrient, typically a micronutrient, that is digested, absorbed, and metabolised through physiological pathways, thereby rendering it available for utilisation by the body. While the bioavailability of macronutrients is generally stable, that of micronutrients is subject to significant variability, influenced by a range of intrinsic and extrinsic factors. Intrinsic factors may include biological characteristics such as sex, age, body composition, life stage (e.g., pregnancy or childhood), and an individual’s pre-existing nutrient stores. Extrinsic factors encompass the preparation and processing methods of the nutrient source, the chemical form in which the nutrient is delivered, and the presence of substances that may act as inhibitors, diminishing bioavailability, or as enhancers, promoting absorption.

[0057] Micronutrient Utilisation Rate or “MUR” as used herein refers to the prediction of the cumulative impact of a regularly consumed quantity of a micronutrient upon their baseline micronutrient status. A micronutrient is considered to have been “utilised” when it’s either used in the functioning of the body or contributes to baseline concentrations of the micronutrient within an individual. Micronutrients which are not utilised are excreted by the body without either being used in its functioning or impacting baseline micronutrient status. It is necessary to calculate MUR in order to precisely formulate a personalised nutrition product which can predictably influence an individual’s micronutrient status, and thereby address and prevent deficiencies, while avoiding potential over-dosing.

[0058] This concept describes the kinetics of the absorption and assimilation of one or more micronutrients into the body of an individual subject. In the context of this application, ‘micronutrient utilisation’ for one or more micronutrients’ is an indicator of a metabolic parameter of an individual subject. The rate of utilisation typically relates to an individual in a fasted state. MUR uses baseline micronutrient levels, ensuring no immediate short-term impacts from food / drink that is consumed and instead represents the true baseline levels of micronutrients in an individual. This differs to micronutrient levels used in common bioavailability measures, which are post-prandial. Unlike bioavailability, which refers to the immediate proportion of a nutrient available for absorption, the rate of utilisation represents the individualised amount of a micronutrient that is required to achieve and maintain individual optimum micronutrient levels, factoring in variables like age, medical conditions, medications, diet, lifestyle, BMI, waist and hip circumference, body composition (fat, lean, and water mass), blood pressure, hydration levels, physical activity, and strength. The rate of utilisation provides a prediction of micronutrients utilised by cells within the body, while “bioavailability” only indicates the potential for absorption.

[0059] ‘Micronutrient Utilisation Rate (MUR); as used herein refers to a prediction of the individualised rate at which a specific micronutrient is utilised by cells within the body. MUR considers long-term effects, factoring in nutrient bioavailability, retention, dietary intake, the cumulative impact of sustained micronutrient intake, nutrient form, lifestyle, medical conditions (previous and current), and medication use. MUR provides characteristic metabolic data rather than just a snapshot which may be influenced by acute or isolated instances. MUR calculations typically require data collected over a month or longer, or periodic retesting, to provide a comprehensive understanding. MUR calculation will involve taking account of personal health datasets, including age, medical conditions, medications, changes in diet and lifestyle, changes in pregnancy or lactation status, BMI, waist and hip circumference, systolic and diastolic blood pressure, fat mass, lean mass, dry lean mass, water, volume of water consumed, and changes in physical activity levels and physical strength.

[0060] Micronutrient Utilisation Rate vector (MUR vector) as used herein is defined as a quantitative prediction of an individual’s capacity to metabolise and utilise specific micronutrients. According to an embodiment of the invention a MUR vector comprises at least one micronutrient utilisation rate (MUR) corresponding to the utilisation rate for at least one micronutrient. In embodiments of the invention the MUR may be expressed as a fraction or as a percentage.

[0061] ‘Metabolic group’ as used herein refers to a group formed by clustering individual subjects with similar rates of utilisation for one or more micronutrients so as to form populations or sub-populations. Typically, an individual subject can be classified into a metabolic group if an individual metabolism type dataset lies within three standard deviations of the metabolism type dataset of the metabolic group. A metabolic group may define a patient group, if the individuals within that group are suffering from a disease or condition that alters, biases or constrains their rates of utilisation for one or more micronutrients. For example, lactating mammals may fall within a metabolic group that has higher levels of utilisation for certain micronutrients, whereas patients suffering from cancer may form a metabolic group that have lower levels of utilisation for certain micronutrients.

[0062] ‘Computer implemented model’ refers to a computational system or method that is implemented using computer hardware and software to perform a specific function, process, or calculation. The model may involve the processing and manipulation of data, the execution of algorithms, or the generation of outputs based on inputs received by the computer system.

[0063] Importance of an individualised micronutrient utilisation rate profile and optimal micronutrient level

[0064] Micronutrients are required for facilitating enzymatic reactions as cofactors and coenzymes, contributing to energy metabolism by assisting in the catabolic process of macronutrients, maintaining enzyme structure and function, and the maintenance and repair of cells, tissues, and organs that make up the human body. Formulation and production of micronutrient supplements have been mainly focussed on addressing common deficiencies in micronutrients such as iron, vitamin A and iodine. Although it is known that certain demographics, such as children and pregnant women, are known to be vulnerable to micronutrient deficiency, a systematic method for analysing an individual optimal micronutrient level, as well as the required intake for additional micronutrients to reach an individualised optimal micronutrient level, is less well researched.

[0065] Understanding an optimal micronutrient level is necessary as the amount of micronutrient required for each individual significantly varies. Some cases of micronutrient deficiencies are due to poor diet or lifestyle, while others are due to conditions involving genetic mutation that prevents synthesis or metabolism of micronutrients correctly. Approximately 50 human diseases have been identified that are caused by enzyme polymorphisms leading to micronutrient deficiency. The type of micronutrient deficiency and the degree of deficiency depend not only on the genetic variation but on overall metabolic characteristics of an individual.

[0066] Micronutrient utilisation rates (MURs) will vary significantly between individuals, due to variables such as age, sex, genetics, epigenetics and lifestyle. This is why it is valuable to calculate an individual's MUR, which predicts the cumulative impact of a regularly consumed quantity of a micronutrient upon their baseline micronutrient status. Without factoring in MUR, it is not possible to precisely formulate a personalised nutrition product which can predictably influence an individual’s micronutrient status, and thereby address and prevent deficiencies, while avoiding potential over-dosing. The MUR is needed to determine the levels of micronutrients required for an individual to achieve their optimal micronutrient profile.

[0067] In healthy individuals without apparent symptomatic micronutrient deficiencies, identifying an individual micronutrient profile and optimal micronutrient levels can provide health benefits while preventing side effects. For instance, while it is known that high intake of multiple micronutrients, such as calcium, folate (vitamin B9), nicotinic acid (vitamin B3), vitamin E and A can affect genome stability, which is one of the hallmarks of cancer, it is not clear which combination or dosage of the micronutrients would be beneficial for an individual. Further, there are known negative effects of some dietary supplements, such as an increased risk of colon polyps and kidney stones from calcium supplements.

[0068] Embodiments of the present invention provide for optimising the micronutrient formulation of personalised nutrition products, identifying the type and dosage of micronutrients that require supplementation, and prevention of unexpected side effects.

[0069] Establishing an initial baseline micronutrient profile

[0070] In aspects and embodiments of the invention, an initial baseline micronutrient profile of an individual is established by performing an initial baseline assay on a sample obtained from the individual subject. Baseline profiles are determined from baseline assays which are taken when an individual is fasted, to avoid direct short-term impacts from food / drink that is consumed. A micronutrient initial baseline assay refers to an assessment or analysis conducted to determine the baseline levels of various micronutrients in an individual's body from one or more sample obtained from the individual. The sample obtained from the individual subject comprises at least one of the following: whole blood; serum; plasma; stool; urine; sweat; cerebrospinal fluid; tears; saliva; solid tissue biopsy; or exhaled breath. The specific components of a micronutrient baseline assay can vary depending on the context and the goals of the assessment, but generally, it involves measuring the concentrations of (e.g. establishing a level of) one or more target micronutrients in the biological sample.

[0071] Establishing an initial baseline measurement

[0072] In the initial baseline measurement, a comprehensive set of parameters is collected to establish a comprehensive health profile of the patient. The initial baseline measurement may involve obtaining a personal health dataset, which may include data such as medical history, current medications, and any pre-existing conditions. Body measurements, including weight, height, waist to hip ratio, and body mass index (BMI), are taken to assess overall physical health (as shown in Example 1 , see below). Cognitive assessments are conducted to evaluate mental function, memory, and attention span, using standardised testing protocols. Cognitive assessment involves a range of standardised tests designed to measure various cognitive abilities such as attention, memory and executive function. The digit span test measures working memory by asking individuals to repeat sequences of digits in forward and reverse order. The Stroop test requires participants to name the colour of ink used to print a word, rather than reading the word itself. The Visual Working Memory Test evaluates the ability to retain and manipulate visual information by presenting a series of visual stimuli and testing recall after a short delay. Mental State Examination (MMSE) is a widely used 30-point screening test that assesses orientation, memory, attention, language, and visual-spatial skills. The Montreal Cognitive Assessment (MoCA) provides a more comprehensive 30-point screening, with greater sensitivity to mild cognitive impairment. The Wechsler Adult Intelligence Scale (WAIS) offers a detailed assessment of intellectual capabilities, measuring verbal comprehension, perceptual reasoning, working memory, and processing speed. Neuropsychological batteries like the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) provide a standardised method for evaluating multiple cognitive domains, including immediate and delayed memory, visuospatial / constructional abilities, language, and attention. For specific populations, specialised protocols such as the Cambridge Neuropsychological Test Automated Battery (CANTAB) offer computerised cognitive assessments with precise measurement of cognitive function across various neurological and psychiatric conditions.

[0073] Metabolism data is gathered through measurements of metabolic rate, insulin sensitivity, and lipid profiles, alongside markers such as homeostatic model assessment for insulin resistance (HOMA-IR) and fasting glucose levels. This multifaceted baseline measurement provides an insight into an individuals’ health, metabolism, and nutritional status.

[0074] Establishing an individualised optimal micronutrient level

[0075] The optimal level of the micronutrient for an individual subject may be derived from combining population recommendations with a personal health dataset that relates to the individual subject. In embodiments, the personal health dataset may be acquired from the initial baseline measurement process. The personal health dataset comprises data relating to one or more of: lifestyle; demographics; bioinformatics data and medical data. Lifestyle data is related to an individual's habits, behaviours, and choices that influence health. It might involve information about diet, physical activity, sleep patterns, substance use (such as smoking, drug or alcohol consumption), and other factors. Demographic data provides information about the characteristics of individuals, such as age, gender, ethnicity, socioeconomic status, education level, and geographic location. Bioinformatics data can involve genetic data (genomics), proteomic data, metabolomic data, microbiome data or other molecular-level information. Medical data encompasses information, amongst other things, regarding an individual's health history, diagnoses, treatments, medications, laboratory test results, and other healthcare-related details.

[0076] The information required for the personal health dataset can be collected in a variety of ways. Lifestyle data can be collected from surveys, questionnaires or mobile apps which allow individuals or personal devices, such as smart watches, to log and track lifestyle factors. Demographic data can be collected from surveys or from electronic health records. Bioinformatics data can be collected from genetic testing, and medical data can be collected from medical records or health information system with the consent of an individual subject.

[0077] The optimal level of the micronutrients can be found by a variety of methods such as statistical methods or computer implemented methods (this is exemplified in Example 2, see below). Some examples of statistical methods include regression analysis, Bayesian optimisation, and gradient descent. Some examples of computer implemented methods include both supervised and unsupervised machine learning models that can be trained to predict optimal values. Regression models or classifiers can be used for this purpose.

[0078] Establishing a rate of utilisation for a micronutrient within the individual subject (this is exemplified further in Example 5, see below)

[0079] It is known that some micronutrients such as boron, molybdenum, and iodine can be absorbed at over 90 percent of a supplied bioavailable composition, while the average absorption rates of zinc, copper, and selenium can range from 30 to 80 percent. It is known that population level differences in absorption of micronutrients also exist. For instance, children, pregnant or lactating women are known to have significantly enhanced absorption rates for specific nutrients and dietary minerals.

[0080] However, measuring or determining a rate of utilisation for a micronutrient within the individual subject remains a challenge. On animal subjects, invasive studies using inverted jejunal sacs have accurately calculated the level of thiamine absorption. It is particularly desirable in humans to calculate the rate of micronutrient utilisation from the amount of micronutrient consumed from various sources in the diet and the amount of micronutrient used over a defined period of time.

[0081] For each micronutrient, a rate of utilisation for a micronutrient within the individual subject over a first period of time is defined as:

[0082] Wherein

[0083] MUR = Micronutrient Utilisation Rate for the individual subject - the cumulative impact of a regularly consumed quantity of a micronutrient upon an individual’s baseline micronutrient status (when an individual is fasted).

[0084] U = Amount of the micronutrient used over a first period of time - predicted by Al / machine learning models, based on a range of factors (including, but not limited to, physical activity, lifestyle, BMI, health conditions, and age), where a relationship between each variable and the usage of each micronutrient is predicted, and then refined using individual data. Over time, as more data is collected, these predictions will become increasingly precise and reliable.

[0085] T = Amount of the micronutrient from the first follow up micronutrient profile (see Example 4).

[0086] E = Amount of the micronutrient from the baseline micronutrient profile.

[0087] C = Amount of micronutrient consumed from a diet during the first period of time.

[0088] B = Amount of the micronutrient consumed from the generic nutrition product during the first period of time.

[0089] Prior to measuring the rate of utilisation, a baseline nutrition product is administered to the individuals. The baseline nutrition product is a generic nutrition product that may comprise one or more organically or synthetically sourced micronutrients. The organic and synthetic micronutrients baseline products have been produced and administered separately to determine if differences in their origins and, in some cases, their composition, impacts their bioavailability and rate of utilisation.

[0090] The “amount of the micronutrient used over a first period of time” is defined as the amount of a micronutrient utilised by the body, over a period of time, given an individual's level of activity and profiling data. As such, increased physical activity could lead to greater use of certain micronutrients. The “amount of micronutrient consumed from a diet” can be estimated from diet diaries that are text and / or photography based. It is also possible to use estimates from data sets, and for the system to sensecheck as individual subjects' diet diary based on existing data sets to identify outlier information which may need to be disregarded because it is probably the result of input error. The “amount of the micronutrient consumed from the baseline nutrition product” are directly measured from the composition of the baseline nutrition product.

[0091] Determining amount of micronutrient required fora portion of the nutrition personalised product (This is exemplified further in Example 6, see below): G = Gap between an individual’s optimal baseline micronutrient level and an estimated level of micronutrients in the individual subject in the absence of consumption of the personalised nutrition product.

[0092] E = Amount of the micronutrient from the baseline micronutrient profile (when an individual is fasted).

[0093] S = Amount of micronutrient synthesised within the body.

[0094] C = Amount of micronutrient consumed from a diet during the first period of time.

[0095] U = Amount ofthe micronutrient used over a first period of time - predicted by Al, based on a range of factors (including, but not limited to, physical activity, lifestyle, BMI, health conditions, and age), where a relationship between each variable and the usage of each micronutrient is predicted, and then refined using individual feedback. Over time, as more data is collected, these predictions will become increasingly precise and reliable.

[0096] P = Number of portions.

[0097] A = Amount of micronutrient required per serving.

[0098] The “amount of micronutrient synthesised within the body” is estimated by sensors and / or using digital devices, for instance, a sensor measuring light exposure to approximate the synthesised amount of vitamin D. The value can be estimated by the computer implemented method, with the inputs including geographic, seasonal and lifestyle factors.

[0099] Using inverse inference methods to estimate variables

[0100] Inverse inference methods are used on quantities, variables, distributions, and functions that are not directly observable. Inferring on a hidden variable via the observation of another variable g is the main objective of inverse inference methods. Various inverse inference methods can be used to estimate variables that cannot be directly measured.

[0101] Embodiments of the present invention make use of inverse inference methods to estimate several variables, including an individualised optimal micronutrient level, the amount of micronutrient consumed from a diet during the first period of time, the amount of micronutrient used over the first period of time, and the estimated level of micronutrients in the individual subject in the absence of consumption of the personalised nutrition product. Estimates for these variables are used as inputs to calculate the rate of utilisation for a micronutrient within the individual subject, and the amount of micronutrient required per portion.

[0102] Estimating the above variables involves sampling from complex probability distributions of multiple data points. For example, estimating an individual optimal micronutrient level involves taking into account lifestyle, demographics, bioinformatics data and medical data. Calculating the amount of micronutrient consumed from a diet, and the amount of micronutrient used over the first period of time, involves readout data from one or more samples obtained from the individual subject. In some embodiments, variables such as the level of micronutrients in the individual subject in the absence of consumption of the personalised nutrition product can be estimated, from an input comprising at least one of: initial baseline micronutrient profile, follow up micronutrient profile, data from follow up assays on a follow up sample.

[0103] In some embodiments, a cross-reference dataset generated from the rate of utilisation calculated using different samples obtained from the same individual subjects are taken as an input.

[0104] In an alternative embodiment, variational Bayesian inference approximates the variable value by estimating the posterior distribution, which indicates a probability describing the uncertainty of the statistical model after taking into account of the observed data. Variational Bayesian inference involves optimising the divergence between the true posterior distribution and the variational distribution to produce the estimated variable.

[0105] In an alternative embodiment of the invention, Markov Chain Monte Carlo is used to optimise the target variable distribution by exploring a variety of parameter spaces from multiple datasets. Markov Chain Monte Carlo involves constructing a Markov chain that has the desired probability distribution as its equilibrium distribution. Running the Markov chain for a sufficiently long time will generate the samples that resemble draws from the target distribution. Other Markov Chain Monte Carlo algorithms, such as Gibbs sampling and the Hamiltonian Monte Carlo have been developed to improve the efficiency of sampling.

[0106] In another embodiment, loopy belief propagation is used to approximate inferences on the variables. In loopy belief propagation, the algorithm iteratively passes messages between nodes in the graph to update beliefs about the variables until the variables reach the optimised value. Similarly, expectation propagation is used to iteratively refine an approximate distribution provided by complex datasets by updating messages between variables in a probabilistic graphical model.

[0107] Types of computer implemented methods

[0108] In some embodiments, a computer implemented method is used to identify at least one or more of the following: any change in micronutrient levels in the micronutrient profile, optimal micronutrient level, amount of micronutrient required per portion, rate of utilisation absorption of micronutrients, micronutrient utilisation rate (MUR), or MUR vectors for one or more micronutrients and any other variable of the invention that can be estimated from an inverse inference method.

[0109] In some embodiments, the computer implemented method involves a machine learning model that is trained on a diverse dataset including a personal health dataset. A machine learning approach learns complex patterns within the data to approximate variables, distributions, and functions that are not directly observable for optimising the micronutrient formulation of a personalised nutrition product. Further, the model can calculate and incorporate error bars, offering insights into the confidence intervals and variance associated with each prediction. A machine learning approach may comprise (mathematical) functions, algorithms or models that can be used to perform a task related to data input to the model. Models may be taught to perform a wide variety of tasks on input data, examples including but not limited to: determining a label for the input data, performing a transformation of the input data, making a prediction or estimation of one or more output parameter values based the input data: or producing any other type of information that might be determined from the input data.

[0110] In a specific embodiment, a widely used class of machine learning algorithms involves simple linear models. Linear models are some of the most straightforward to use in machine learning approaches and make a prediction by using a linear function of the input features. Known linear models may include linear regression, linear regression (ordinary least squares), ridge regression, Lasso and polynomial regression, or ordinal regression. In common with all linear regression models is the need to consider the given data points (the training data) and plot a best fit line to fit the model in the best way possible and to thereby allow predictions to be made with a high level of accuracy. Regression techniques, such as those described above and that are more widely known in the art, may be used to generate a range of in silico rules and models based upon training data sets comprising multiple metric values. Linear regression models are particularly useful for extrapolation, where there is a need to estimate values beyond the observational range provided within a training data set.

[0111] In an alternative embodiment of the invention, the machine learning models utilise a neural network approach. A neural network is a model containing an interconnected group of processing elements or "neurons" that process information using a connectionist approach to computation. Neurons are organised into layers which typically include an input layer, one or more hidden layers, and an output layer. Neural networks are often used to model complex relationships between inputs and outputs or to find patterns within data. Typically, neural networks process data in a non-linear, distributed, parallel fashion. Often a neural network is an adaptive system that changes its structure during a learning phase. Functions are performed collectively and in parallel by the processing elements, rather than there being a clear delineation of subtasks to which various units are assigned. Generally, a neural network involves a network of simple processing elements that exhibit complex global behaviour determined by the connections between the processing elements and element parameters. Neural networks may be used with algorithms designed to alter the strength of the connections in the network to produce a desired signal flow. The strength, also known as a weighting, is altered during the training or learning phase. A neural network employed can be shallow, meaning the network has a small number of hidden layers, or deep, with multiple hidden layers. Shallow networks are simpler and computationally less intensive, making them suitable for certain tasks where the complexity of deep neural networks may not be necessary. A neural network used can have a skip connection, meaning bypassing one or more intermediate layers in deep neural networks. A neural network used can involve a few-shot supervised model which is trained on a very small dataset, typically much smaller than what is traditionally required for training robust models.

[0112] In some embodiments, supervised machine learning models may be implemented. A supervised machine learning model process a set of training data comprising example inputs and corresponding ground-truth outputs. The ground truth output refers to a verified outcome that represents the correct output, against which the model’s predicted output is assessed. Examples of the supervised machine learning include feedforward neural networks, convolutional neural networks, recurrent neural networks, and transformer networks. The training process typically involves iteratively optimising the values of weights or bias values of the model to tune the model to reproduce the ground truth outputs for the input data. The optimisation process, implemented via as backpropagation and gradient-based algorithms, enables the model to improve its accuracy through iteration. In some embodiments, the optimisation of the parameters through iteration can infer one or more latent variables, representing underlying factors that are not directly observable but significantly influence the predicted output such as first follow-up micronutrient profiles.

[0113] In a particular embodiment, the training data may comprise the measured baseline micronutrient profile, along with personal health dataset that includes one or more anthropometric measurements, such as but not limited to: age, medical conditions, medications, diet and lifestyle, BMI, waist and hip circumference, systolic and diastolic blood pressure, fat mass, and physical strength. In this embodiment, micronutrient utilisation rate (MUR) or one or more MUR vectors may be represent as a latent variable of the machine learning model. The first follow-up micronutrient profile measured at the first follow-up assay may serve as the ground truth output to be compared with the outputs, and the values of weightings or bias optimised to infer the latent variable of the model.

[0114] In an alternative embodiment of the invention, supervised or unsupervised clustering techniques may be used. In one embodiment, unsupervised clustering methods, including but not limited to hierarchical clustering and spectral clustering, may be employed to identify metabolic groupings within the micronutrient utilisation rate (MUR) or MUR vector of each individual, wherein each resulting cluster is indicative of a distinct metabolism type. In a different embodiment, a supervised clustering method, including but not limited to mixtures-of-experts, may be used to map MUR or MUR vectors to metabolism types in conjunction with specific outcomes. In this context, cluster allocations are determined based on both the individual-level MUR variables and the target outcomes, such as micronutrient profiles observed at follow-up. The cluster allocations may inform the predictive relationship between MURs and the metabolic outcomes.

[0115] A further embodiment of the invention provides for the use of decision tree based classifiers to predict both original and continuous target variables. In particular, a random forest approach comprises a supervised machine learning algorithm that is constructed from decision tree algorithms. This algorithm is suitably used to predict behaviour and outcomes in a given set of circumstances. The term ‘random forest’ refers to the use of a combination of classification tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the so-called ‘forest’. A random forest is a learning ensemble consisting of a bagging of un-pruned decision tree learners with a randomised selection of features at each split of the decision tree. Where ‘bagging’ is an ensemble meta-algorithm that improves the accuracy of the machine learning algorithm. A random forest grows a large number of classification trees, each of which votes for the most popular class. The random forest algorithm establishes the prediction outcome based on the predictions of the decision trees. It predicts by taking the average or mean of the output from various trees in the forest. Increasing the number of trees in the forest, thus, increases the predictive power of the algorithm. Random forest algorithms are useful for predictive accuracy within a rich dataset and are suitable for both regression and classification tasks.

[0116] Data architecture

[0117] Embodiments of the present invention are described herein with reference to block diagrams. It will be understood that each block of the diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams are implemented by computer implemented methods. The data architecture represented in blocks, and data processing flow represented in arrows are executed via the processor of the computer or other programmable data processing apparatus. The data architecture can be stored in a computer readable storage medium, a programmable data processing apparatus, and / or multiple other devices to function in a particular manner, such that the computer readable storage medium implements aspects of the function / act specified in the block diagram or blocks.

[0118] Figure 2 illustrates a data structure represented in a block diagram that sets out how different data structures of the invention interact. A dedicated private server can host the knowledge base of micronutrient utilisation rates (MUR). The server can also host user data, comprising of one or more personal health data, and variables required for inverse inference methods such as individualised optimal micronutrient level, amount of micronutrient consumed from a diet during the first period of time, an amount of micronutrient used over the first period of time, and an estimated level of micronutrients in the individual subject in the absence of consumption of the personalised nutrition product.

[0119] In one embodiment, a server is adapted to make use of inverse inference methods to estimate variables. In a specific embodiment, a high-level, general-purpose programming language, such as Python, is used to facilitate running of predictions of rate of utilisation of a micronutrient for an individual subject.

[0120] In some embodiments, an application is linked to external devices and the knowledge base private server, by collecting user data and variables.

[0121] Hence, in one embodiment the invention comprises an apparatus configured to perform some or all of the methods described. In embodiments of the invention the apparatus may include a (computer) system. The system can be configured for engineering compliant communications. The system can comprise one or more processors and one or more non-transient computer-readable storage media. The computer readable storage media can have stored thereon computer-executable instructions that are executable by the one or more processors to cause the computer system to perform some or all of the methods and procedures described herein. Hence, it will be appreciated that the present invention may be a system, an apparatus, a method, and / or a computer program product. The computer program product may include a computer readable storage medium (or media) having computer readable program instructions thereon for causing a processor to carry out aspects of the present invention.

[0122] The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device. The computer readable storage medium may be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of the computer readable storage medium includes the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, and any suitable combination of the foregoing. Computer readable storage media may be accessible within a local area network in the form of one or more linked servers or located remotely in cloud based virtual machines or servers. Cloud based services may be accessed via wired or wireless (wi-fi) telecommunications, such as over the internet. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fibre-optic cable), or electrical signals transmitted through a wire.

[0123] The steps of sample acquisition or sample processing may be fully or semi-automated. By way of nonlimiting example, a breath analyser may be integrated into a system of an embodiment of the invention. Breath analysers (also called breathalysers) operate on the principle that the subject’s breath contains a wealth of valuable biomarker data, offering insights into an individual's health and well-being. These devices utilise sensor technologies to detect and quantify the presence of specific compounds in exhaled breath, known as biomarkers, which can be informative about micronutrient status. The breath sample is introduced to the analyser, and sophisticated sensors interact with the breath molecules, producing measurable signals. Key biomarkers commonly targeted include ethanol for alcohol detection, as well as various volatile organic compounds (VOCs) associated with specific health conditions. Advanced breath analysers leverage technologies such as infrared spectroscopy, electrochemical sensors, or metal oxide semiconductors to selectively identify and measure these biomarkers.

[0124] In an alternative embodiment of the invention a wearable, non-intrusive microneedle sensor device may be used for sample data acquisition. Typically, such a device includes a microneedle sensor unit coupled to an electronics unit, where the microneedle sensor unit comprises a substrate, an array of spiked microneedle structures that include sensor electrodes, and electrical interconnections that electrically couple the sensor electrodes to the electronics unit for processing of detectable signals associated with one or multiple biomarkers within a biofluid. Microneedle sensor patches may be integrated into the systems and apparatus of the present invention to provide suitable bioassay information.

[0125] In a second aspect, the invention relates to a method for defining a metabolic group comprising of individuals with similar metabolic characteristics defined in terms of the rate of utilisation of different micronutrients. The metabolic groups are defined by clustering individual subjects with similar rates of utilisation for one or more nutrients. Then an optimal micronutrient level is defined for each metabolic group, by averaging an optimal micronutrient level for individuals in each group. Average first follow up nutrient profile is defined following the first follow up tests on the individual subjects of the metabolic group. The metabolic group can be represented by a metabolism type dataset, which incorporates the average optimal micronutrient level and the average first follow up micronutrient profile. The invention further relates to a method of assigning an individual subject to the metabolic group, by identifying a metabolism type dataset of an individual subject and comparing this value to the metabolism type dataset of the metabolic group. If the average value of at least one component of an individual metabolism type dataset falls within at least three standard deviations of the metabolism type dataset of the metabolic group, an individual subject is classified to the metabolic group. The amount of additional micronutrients required to reach the average optimal micronutrient level for the metabolic group is then calculated. The personalised nutrition product of the metabolic group is formulated so that it satisfies the average optimal micronutrient level for the metabolic group an individual is associated with.

[0126] In further aspects, the invention provides systems for the manufacture of micronutrient formulations that are optimised to the micronutrient requirements of an individual subject or to that of a metabolic group, or to an individual identified as being within a metabolic group. The systems may comprise computer apparatus that operates according to the methods and architecture described by the embodiments disclosed herein. The systems may further comprise integrated hardware and software designed to formulate compositions comprising optimised levels of micronutrients sourced from organic and / or synthetic origin. The systems may comprise manufacturing execution functionality and software programs in order to operate a manufacturing process that comprises one or more of liquid mixing, powder blending, pasteurisation, homogenization, standardisation, packaging, and sterilisation processes. Further embodiments of the system of the invention may comprise product tracking, individual barcoding, logistics and distribution aspects necessary to ensure that individualised nutritional products may be placed in the hands of the correct end user. The invention is further demonstrated by reference to the following non-limiting examples.

[0127] EXAMPLES

[0128] The following examples provide an exemplification of an embodiment of the present invention.

[0129] EXAMPLE 1 - Performing an initial baseline measurement of candidate subjects

[0130] At a first visit, participant subjects underwent a series of basic health screen measurements. All participants in the study had fasted overnight and had not engaged in any physical exercise prior to the assessment.

[0131] The list of parameters that were measured is outlined in Table 1 below. A venous blood sample was taken before the fasted participant consumed one sachet of predetermined amounts of organically sourced vitamins and minerals (generic nutrition products). These measurements are all referred to in this document as ‘Baseline’ measurements or Timepoint 1 (T1) (see Figure 3).

[0132] Table 1. Parameters of the baseline measurements

[0133] Example 2 - Establishing an individualised optimal micronutrient level

[0134] Table 2. Example of an optimised micronutrient level

[0135] In an example, establishing an individualised optimal micronutrient level may start from defining baseline recommendation provided by General Reference Nutrient Intakes (RNI) or Recommended Dietary Allowances (RDA).

[0136] Based on individual measurement parameters, an optimal micronutrient profile was established by evaluating the levels of certain micronutrients such as iron, vitamin D, vitamin B12, folate, calcium, magnesium, and zinc. Demonstration of levels of vitamin D or iron below the optimum indicated a deficiency that required dietary adjustments or supplementation. Example 3 - Administering a generic nutrition product

[0137] The following are exemplary compositions that can be utilised as the baseline composition and also as a base recipe for a personalised nutrition product.

[0138] Table 3: Generic nutrition product in the form of a powder for reconstitution as a nutritional shake: The above formulation does not include any additional added micronutrients beyond those included in the listed ingredients. Optionally, depending upon the target individual up to 5%m (mass percent) of micronutrient formulation may be included to meet non-personalised requirements based upon, for example, average requirements for an adult (e.g. recommended daily doses). Baseline nutritional products are available in two distinct compositions: an organic baseline nutritional product and a synthetic baseline nutritional product, each containing micronutrient composition paired with maltodextrin as the carrier. The organic baseline nutritional product comprises at least one micronutrient extracted from an organic source, whereas the synthetic blend incorporates synthetically manufactured micronutrients. The synthetic blend mirrors the composition of the organic blend, with equivalent synthetic amounts of each micronutrient. Both blends are paired with maltodextrin and are designed to deliver a consistent 60% of the recommended intake per serving, providing a standardised basis for evaluating micronutrient bioavailability and efficacy.

[0139] Table 4. Composition of an example of an organically sourced baseline nutritional product

[0140] To establish the micronutrient level immediately after the supplementation formula, the participant subjects consumed a sachet of organic or synthetic baseline nutritional product orally, having been assigned in a randomised, single-blind fashion. Another blood sample was taken at 2 hours point to measure the acute blood vitamin and mineral concentration response to the consumption of the product. This sample was referred to as ‘After 2 hours’ or Timepoint 2 (T2) (Figure 3).

[0141] The participant subjects then consumed an organic or synthetic generic nutrition product orally one sachet per day, for 28 days.

[0142] The second visit was scheduled 28 days after the first visit. At the second visit, all measurements were repeated as in the first visit including a venous blood sample, but without the acute 2-h supplement test. All measurements taken in the second visit were referred to as ‘After 28 days’ or Timepoint 3 (T3) (Figure 3). The above shake recipes may be combined with a liquid, such as water, as required and to taste to enable consumption by an individual subject.

[0143] Example 4 - Identifying any change in one or more micronutrient levels in the first follow up profile

[0144] Table 5. Summary of micronutrients, metabolites, body measurement, and cognitive tests

[0145] The results of the first follow-up metabolic dataset or micronutrient profile are summarised in Table 5. Both organic and synthetic baseline product administered groups experienced increases in iron, calcium, ferritin, folate, and vitamin D levels from baseline to 2 hours post-intervention. Over 28 days, both groups showed a decrease in systolic blood pressure, waist ratio, Stroop test time, and fat mass, with folate and vitamin B12 also increasing. Notably, the synthetic group had higher total cholesterol levels throughout the study. Additionally, the groups differed in how certain parameters changed over the 28-day period; triglycerides remained the same in the organic group but increased in the synthetic group, and glucose levels stayed the same in the organic group but increased in the synthetic group.

[0146] Example 5 - Determining a rate of utilisation for one or more micronutrients within the individual subject

[0147] In this example, a micronutrient utilisation rate (MUR) within the individual subject was calculated by identifying amount of the micronutrient used over a first period of time (U), amount of the micronutrient from the first follow up micronutrient profile (T), Amount of the micronutrient from the baseline micronutrient profile (E), amount of micronutrient consumed from a diet during the first period of time (C), and amount of the micronutrient consumed from the generic nutrition product during the first period of time (B).

[0148] U + T — E

[0149] MUR)=

[0150] C + B

[0151] The amount of micronutrient consumed from a diet during the first period of time (C) was calculated by the individual subjects tracking their micronutrient intake by photographing each meal they consumed. The images were processed using image recognition and natural language processing (NLP) techniques, analysing the types and portions of food in each image. By cross-referencing visual image data with a nutritional database it was possible to estimate the nutrient content for each food item consumed. Patterns of any outliers or inconsistencies — such as unusual portion sizes or unlikely food combinations were identified.

[0152] Example 6 - Identifying a requirement for increasing or decreasing micronutrients for the first follow up micronutrient profile to reach an individualised optimal micronutrient level

[0153] The first follow up micronutrient levels are compared to the individualised optimal micronutrient level. The required amount of an additional micronutrient for supplementation is calculated based on how much was needed to achieve the individualised optimal micronutrient levels as well as forecast dietary intake and micronutrient utilisation rates (MUR).

[0154] Example 7- Determining a metabolism type dataset or Micronutrient Utilisation Rate (MUR) for one or more micronutrients within the individual subject

[0155] Step 1 : Data collection and identification of input data

[0156] To calculate Micronutrient Utilisation Rate (MUR), an input dataset of the initial baseline micronutrient profile is considered (see Table 6). Additional factors that account for individual differences in utilisation and metabolism, including age, medical conditions, medications, diet, lifestyle, BMI, waist and hip circumference, body composition (fat, lean, and water mass), blood pressure, hydration levels, physical activity, and strength, may be considered. The MUR value for each micronutrient is expressed in % of the utilisation rate. To determine MUR of an individual, supervised machine learning, such as a feedforward neural network is used.

[0157] Table 6. Example of an input data (baseline micronutrient profile) for MUR calculation. Table 7. Example of true follow-up level after approximately 1 month (T3) to be compared with the output layer.

[0158] Step 2: Predictive modelling:

[0159] A feed-forward neural network for predicting follow-up micronutrient levels was constructed with an input layer that takes input data. One or more hidden layers were incorporated to model the complex utilisation rates of micronutrients (referred to as Micronutrient Utilisation Rates, or MUR). The MURs are the latent variables of the network that transforms baseline micronutrient profile to the predicted first follow-up micronutrient levels, effectively capturing individual variability in nutrient metabolism and utilisation. The MUR values in the hidden layers were constrained to a physiologically feasible nonnegative range. The output layer provided the predicted first follow up micronutrient profile, representing the model’s final prediction based on the learned utilisation rates and input data. Standard optimisation techniques such as backpropagation, using the true follow-up micronutrient profile were employed during training to iteratively adjust the network parameters, minimising the error between predicted and observed follow-up levels and improving the network’s accuracy in modelling changes in nutrient status over time.

[0160] Step 3: Estimating utilisation rates by inverse inference:

[0161] The structure and parameters (weights) of the trained neural network, as well as the observed baseline levels, follow-up levels, and dietary micronutrient intake, were fixed. The objective function — such as the mean squared error between predicted and actual micronutrient levels at follow-up — was optimised with respect to the hidden units of the trained network corresponding to the utilisation rates.

[0162] To reduce overfitting, additional constraints were applied to the model, such as regularisation techniques. Standard methods, including cross-validation, were used to select hyperparameters and estimate performance. The model was then tested on an external dataset to evaluate its predictive accuracy.

[0163] The resulting MUR vector for each individual represented their unique micronutrient utilisation profile. Combined with other health parameters, this vector provided insights into individual metabolic health and nutritional needs.

[0164] Example 8- Forming the personalised nutrition product so that it satisfies the requirements for reaching the individualised optimal micronutrient level

[0165] The personalised nutrition product is configured to meet the individualised optimal micronutrient levels of the user and can be delivered in a variety of forms. The product may be provided in the form of capsules, which contain only the specified micronutrient extract ingredients, or as a powdered nutrition shake, wherein the micronutrient extracts are combined with a generic base mix. The base mix is formulated to include a protein blend, fruit powder, tapioca fibre, natural sweetener, and natural flavouring. A base mix composition example of a personalised nutrition product is in Table 8. Nutritional values of a base mix composition of a personalised nutrition product is in Table 9.

[0166] Table 8: Base mix composition of a personalised nutrition product in the form of a powder for reconstitution as a nutritional shake. Table 9. Nutrient values of a base mix composition for a personalised nutrition product.

[0167] Example 9 - Establishing a metabolism type dataset / profile and assigning an individual to one or more metabolic groups

[0168] Following identification of an individual’s unique MUR vector, individuals were assigned to distinct metabolic groups based on similarities in their MUR vectors. The metabolic grouping can be performed using both unsupervised clustering methods, such as hierarchical or spectral clustering, and supervised clustering approaches, including mixtures-of-experts. In the case of supervised clustering, the cluster allocation provided valuable insights into how an individual’s MUR vector mapped to labelled outcomes, such as the presence or absence of micronutrient deficiencies or predicted levels of micronutrients at follow-up. For example, individual subjects with MUR vectors characterised by low utilisation rates for vitamin D and calcium might consistently cluster into a group associated with a higher risk of bone-related deficiencies, such as osteoporosis or osteopenia. Similarly, a cluster dominated by individuals with high utilisation rates for iron but low for vitamin B12 might correspond to a subgroup prone to anaemia caused by vitamin B12 deficiency rather than iron deficiency.

[0169] Further, a MUR vector of an individual, measured over time, can be redefined dynamically to represent evolving metabolic types with increasing precision. Rather than being static, an individual’s MUR set - or “MUR vector” - will adapt and change in response to a wide range of factors, including age, health status, lifestyle modifications, dietary habits, and environmental influences. These temporal shifts reflect the complex and dynamic nature of nutrient metabolism, enabling the model to capture more accurate and personalised insights into an individual’s metabolic health as their circumstances evolve.

[0170] Example 10 - Linking a metabolism type dataset / profile to health outcome categories

[0171] Establishing metabolic groups of Example 9 allowed the model to associate distinct metabolism type dataset / profile to health outcomes. For example, representative MUR vectors of each metabolic group, provided predicted nutrient levels for further follow-up points, such as identifying individuals with MUR profiles that suggested a likelihood of dropping below adequate levels of folate or magnesium at followup, enabling earlier dietary or supplementation interventions. The metabolic groupings provided further insights such as common nutrient absorption inefficiencies or metabolic inefficiencies across the group.

[0172] MUR vectors of an identified metabolic group, combined with additional health markers, may relate to specific health outcome categories, including obesity and broader metabolic disorders. For example, weight reduction, side effects of weight-loss treatments, such as nausea and diarrhoea quantified using gastrointestinal symptom rating scale (GSRS) scores or visual analogue scales (VAS) were combined with MUR vectors to define a specific health outcome category of an individual.

[0173] Patient retention and medication adherence were also evaluated where appropriate, offering insights into long-term treatment sustainability. Inflammatory markers, including CRP, NT-pro-BNP, and various cytokines such as GM-CSF, IFN-y, IL-6, and TNF-a, were tracked to assess systemic inflammation and its correlation with metabolic changes. Metabolic markers such as HbA1 c, fasting glucose, haemoglobin, blood pressure, bioelectrical impedance analysis (BIA), lipids (HDL, LDL, triglycerides), and HOMA-IR provided detailed profiles of metabolic health.

[0174] In this way the methods and systems of the invention allow for the remediation of micronutrient deficiencies in both healthy subjects as well as in subjects suffering from chronic pathological conditions or diseases.

[0175] Although particular embodiments of the invention have been disclosed herein in detail, this has been done by way of example and for the purposes of illustration only. The aforementioned embodiments are not intended to be limiting with respect to the scope of the appended claims, which follow. It is contemplated by the inventors that various substitutions, alterations, and modifications may be made to the invention without departing from the spirit and scope of the invention as defined by the claims.

Claims

CLAIMS1. A method of optimising the micronutrient formulation of a personalised nutrition product intended for use by an individual subject, the method comprising: establishing an initial baseline micronutrient profile for the individual subject by performing an initial baseline assay on a sample obtained from the individual subject; establishing an individualised optimal micronutrient level, wherein the individualised optimal micronutrient level is calculated from a personal health dataset that relates to the individual subject; administering a baseline nutrition product to the individual subject for a first period of time; performing a first follow up analysis that comprises performing a first follow up assay on a first follow up sample obtained from the individual subject to establish a first follow up micronutrient profile for the individual subject, and determining a micronutrient utilisation rate (MUR) for one or more micronutrients within the individual subject by:(i) identifying any change in one or more micronutrient levels in the first follow up micronutrient profile in comparison to that of the initial baseline micronutrient profile;(ii) identifying an amount of the micronutrient consumed from the baseline nutrition product during the first period of time;(iii) identifying an amount of the micronutrient consumed from a diet excluding a baseline nutrition product during the first period of time;(iv) identifying an amount of the micronutrient used by an individual during the over a first period of time;(v) identifying a requirement for an increase or decrease in micronutrients for the first follow up micronutrient profile to reach an individualised optimal micronutrient level; and(vi) formulating the personalised nutrition product so that it satisfies the requirements for reaching the individualised optimal micronutrient level when administered to the individual subject.

2. The method of Claim 1 , wherein the personal health dataset comprises data relating to one or more of: lifestyle; demographics; bioinformatics data and medical data.

3. The method of Claim 1 or 2, wherein the first period of time is at least 1 week, suitably at least 4 weeks, typically at least six weeks.

4. The method of any previous Claim, further comprising: providing the personalised nutrition product to the individual subject, with an indication to consume the personalised nutrition product for a second period of time;performing a second follow up analysis that comprises performing a second follow up assay on a second follow up sample obtained from the individual subject to establish a second follow up micronutrient profile for the individual subject, and identifying any change in micronutrient levels in the second follow up micronutrient profile in comparison to that of the initial baseline micronutrient profile and / or the first follow up micronutrient profile; identifying requirements for increases or decreases in micronutrients so as to optimise micronutrients in the individual subject; and reformulating the personalised nutrition product so that satisfies the requirements for changes to micronutrients in the individual subject.

5. The method of any previous Claim, wherein formulating the personalised nutrition product involves calculating the amount of micronutrient required per administered dose.

6. The method of any previous Claim, wherein the second period oftime is at least 1 week, suitably at least 2 weeks, optionally at least 4 weeks, typically at least 12 weeks.

7. The method of Claim 6, wherein calculating the amount of micronutrient required per dose involves predicting a gap between the individualised optimal micronutrient profile and an estimated level of micronutrients in the individual subject in the absence of consumption of the personalised nutrition product.

8. The method of any previous Claim, wherein the micronutrient utilisation rate (MUR) is calculated with further information concerning dietary intake and / or lifestyle, and wherein the further information is stored in the form of a database that additionally includes data that characterises the metabolic status of the individual.

9. The method of Claim 8, wherein the further information includes at least one population-based dataset.

10. The method of any previous Claim, wherein the sample obtained from the individual subject comprises at least one of the following: whole blood; serum; plasma; stool; urine; sweat; cerebrospinal fluid; tears; saliva; solid tissue biopsy; or exhaled breath.

11. The method of any previous Claim, wherein the micronutrient utilisation rate (MUR) is calculated using different samples obtained from the same individual subject is compared to generate a cross-reference dataset.

12. The method of Claim 8, wherein the data characterising the micronutrient-metabolic status of an individual includes at least one of indicator of demographic status and / or medical status.

13. The method of any previous Claim, wherein a computer implemented method is used to identify any change in micronutrient levels in the first or a second follow up micronutrient profile, in comparison to that of an earlier micronutrient profile.

14. The method of any previous Claim, wherein a computer implemented method is used to calculate the optimal micronutrient level.

15. The method of any previous Claim, wherein a computer implemented method is used to calculate the amount of micronutrient required per portion of the personalised nutrition product.

16. The method of any previous Claim, wherein a computer implemented method is used to determine the micronutrient utilisation rate (MUR) of micronutrients in the individual subject.

17. The method of any previous Claim, wherein a computer implemented method is used to determine the amount of a micronutrient consumed from a diet excluding a baseline nutrition product.

18. The method of any one of Claims 13 to 17, wherein the computer implemented method utilises one or more of a probabilistic regularised parametric model, a non-parametric model, or a neural network.

19. The method of any previous Claim, wherein the micronutrients comprise vitamins and / or dietary minerals.

20. The method of any previous Claim, wherein the personalised nutrition product is in a form of a dry powder.

21. The method of any previous Claim, wherein the personalised nutrition product comprises one or more organically sourced micronutrients.

22. The method of any previous Claim, wherein the personalised nutrition product comprises one or more synthetic micronutrients.

23. The method of any previous Claim, wherein the individual subject is suffering from a disease or condition that impairs nutritional uptake of micronutrients from the diet.

24. A computer-implemented method for calculating a Micronutrient Utilisation Rate (MUR) for an individual subject, comprising: receiving input data comprising an individual subject's baseline micronutrient profile, and one or more additional factors selected from: a first follow-up micronutrient profile; an amount of a micronutrient consumed from a diet in a first period of time; an amount of a micronutrient consumed from a baseline nutrition product in a first period of time; an amount of a micronutrient used over a first period of time; and a personal health dataset; constructing a predictive model using a machine learning model comprising(i) an input layer configured to receive the baseline micronutrient profile and the one or more additional factors,(ii) one or more hidden layers, wherein a value for the MUR can be inferred from the latent variables of the one or more hidden layers that transform baseline micronutrient profiles into predicted follow-up micronutrient levels, and(iii) an output layer configured to provide a predicted follow-up micronutrient profile; training the predictive model using the input data; applying the trained predictive model to generate the predicted follow-up micronutrient profile; estimating MUR values by inverse inference using the trained predictive model, by identifying latent variables of the hidden layers that minimises the error between predicted and actual micronutrient levels at follow-up.

25. The method of claim 24, wherein the machine learning model is a feed-forward neural network.

26. The method of claim 24, wherein the training process involves iterative optimisation of the weights associated with one or more nodes comprised within the machine learning model.

27. A system for identifying a micronutrient utilisation rate (MUR) of a plurality of micronutrients in an individual subject, comprising: a database that is adapted to store a record of micronutrient levels identified as present within in a sample of a body fluid obtained from the individual subject at a specific timepoint; a device that collects data indicating metabolic status of an individual; a server; and a processor in communication with the server and the database, the processor being configured to calculate the MUR.

28. The system of Claim 27, wherein the system is further configured to calculate a projected optimal baseline micronutrient level for the individual subject.

29. The system of Claim 28 wherein the system is configured to manufacture a composition for a personalised nutrition product that satisfies the requirements for an optimal baseline micronutrient level for the individual subject.

30. The system of any one of Claims 27 to 29, wherein the device comprises a breath analyser and / or a microneedle patch sensor.

31. A method of optimising the micronutrient formulation of a personalised nutrition product intended for use by an individual subject comprised within a metabolic group, comprising: establishing one or more metabolic groups by clustering individual subjects with similar rates of uptake for one or more micronutrients; identifying an average optimal micronutrient level for the one or more metabolic groups by averaging an optimal micronutrient level for individuals comprised within the one or more metabolic groups; identifying an average first follow up micronutrient profile for the one or more metabolic groups by averaging first follow up micronutrient profile for the one or more metabolic groups;establishing a metabolism type dataset for the one or more metabolic groups that incorporates the average optimal micronutrient level and the average first follow up micronutrient profile; classifying an individual subject to the one or more metabolic groups if an individual metabolism type dataset for a given micronutrient lies within three standard deviations of the metabolism type dataset of the metabolic group for the same micronutrient; identifying a requirement for additional micronutrients for the average first follow up micronutrient profile to reach the average optimal micronutrient level; and formulating the personalised nutrition product so that it satisfies the requirements for reaching the average optimal micronutrient level when administered to an individual subject of the metabolic group.

32. The method of Claim 31 , wherein classifying an individual subject to the metabolic group involves supervised clustering or unsupervised clustering.

33. The method of Claim 32, wherein the supervised clustering comprises use of either a mixtures- of-experts model or of a random forest model.

34. The method of Claim 32, wherein the unsupervised clustering comprises use of hierarchical or spectral clustering methods.

35. The method of any one of Claims 31 to 34, wherein the individual subjects in the metabolic group may share one or more health markers indicative of health outcome categories.

36. The method of Claim 35, wherein the one or more health markers comprise at least one measure selected from: weight change; a gastrointestinal symptom rating scale (GSRS) score; a visual analogue scales (VAS); one or more inflammatory markers; or one or more metabolic markers.

37. The method of Claim 36, wherein the one or more inflammatory markers comprise at least one of: C-reactive protein (CRP) level; N-terminal prohormone of brain natriuretic peptide (NT-pro-BNP) level; or inflammatory cytokine levels.

38. The method of Claim 37, wherein the inflammatory cytokine is selected from: GM-CSF; IFN-y; IL-6; or TNF-a.

39. The method of claim 36, wherein the metabolic markers comprise at least one of: HbA1 c level; fasting glucose level; haemoglobin level; diastolic and / or systolic blood pressure; bioelectrical impedance analysis (BIA); lipid levels (HDL, LDL, and / or triglycerides); and homeostatic model assessment for insulin resistance (HOMA-IR).

Images