Personalised nutrition is the provision of dietary advice which is tailored to a person based on their individual characteristics and lifestyle. As analytical techniques become faster and cheaper, and AI tools more advanced, businesses are seeking to capitalize on heightened consumer awareness of the impact of diet on health as well as their desire for a personalized approach.
Many companies rely on genetic testing, such as UK-based BioSynergy, which offers personalized dietary advice based on DNA analysis. Hong Kong start-up GUTolution sequences the gut microbiome to provide diet recommendations, while US-based company Levels provides personalized advice based on blood glucose levels and biomarkers which assess metabolic health.
Other companies circumvent biological testing altogether, relying entirely on an AI-based approach which tailors advice to users’ reported diets. One such platform is the Birdie app, which takes a user’s answers to dietary questions then searches the scientific literature and medical resources to provide tailored nutritional advice.
There is vast growth potential for businesses operating in the rapidly expanding personalized nutrition sector. However, to succeed, companies will need to convince consumers that they can deliver substantially more than just standard nutritional advice, and protect their IP to ensure a competitive advantage.
Challenges in patenting personalized nutrition
Generally, technological innovations are protected using patents, and the personalized nutrition field is no exception. However, as with personalized medicine, there are certain challenges to obtaining patents in this field. For example, European law prohibits the patenting of therapeutic and diagnostic methods practiced on the body. In the US, methods which are solely based on a natural phenomenon, such as an increase in a particular biomarker in response to food, are also not eligible for patent protection. Therefore, patent applications for methods which involve testing patient samples to determine nutritional or metabolic health, or administering certain foods to improve health, must be drafted with these exclusions in mind.
Another challenge is establishing that a method of delivering personalized nutritional advice is novel and non-obvious over what is already known in the field. In order to provide commercially useful protection, patent claims are often intentionally “broad” or generic. This may also be necessary to capture the way a method is applied across a whole population. However, this can seem at odds with the concept of providing nutritional advice which is specific to the individual.
Patenting computer-implemented methods
A common way to protect personalized nutrition, which overcomes some of the above patent eligibility challenges, is to claim a computer-implemented method and a system for carrying out the method. Such methods typically include steps such as providing test data (e.g. from glucose monitoring or sequencing), using the data to generate a score, and outputting the score, or guidance based on the score, to a user.
In order to pass the inventive step hurdle, the claimed method must go beyond delivering standard dietary advice which is known to be linked to certain physiological states, such as lowering fat consumption to counter high cholesterol. Experimental data, for example data showing a previously-unknown link between diet and a physiological effect, can help to overcome this hurdle.
Case study
One example of a successful method patent in this field is European patent EP4203712B1 owned by Nestlé. The patent claims a computer-based method for providing a personalised fibre recommendation based on an individual’s carbohydrate-active enzyme (“CAZyme”) profile. Based on sequence analysis of faecal samples from 60 subjects, together with the subjects’ answers to a food questionnaire, a machine learning algorithm was developed to determine the likelihood of an individual having a specific CAZyme profile based on the questionnaire alone, thereby enabling the profile of subsequent subjects to be predicted without sampling. The novelty and inventiveness in the methodology, therefore, lies in the relationship identified between a subject’s food intake and their enzyme profile.
What this shows is that reasonably broad method claims can be obtained in this field when they are based on data which identifies a previously-unknown correlation between food and a particular characteristic.
Composition patents
Companies can also seek composition patents as part of their IP strategy. Claims to new compositions or other products are often considered desirable since it is generally easier to detect infringement of competitor products than methods.
Companies pursuing claims to new food compositions include ZOE Ltd, which is seeking patent protection for its “Daily30+” plant-based supplement (European patent application EP4585050), and Myota GmbH, which has filed a family of patent applications (based on international application WO2024/115691) directed to a dietary fibre composition.
Food compositions, while not subject to the patent eligibility hurdles that methods are, face an inventive step challenge in that a combination of known ingredients for their known functions is likely to be considered obvious. Indeed, such objections have been raised against ZOE’s and Myota’s applications by the European Patent Office. Applicants who can demonstrate, ideally with comparative data, that their formulation provides an unexpected benefit over existing products are therefore more likely to succeed in reaching grant.
Top take-aways for personalized nutrition companies
Personalized nutrition is a complex field encompassing genetics, microbiology, biochemistry, nutrition, formulation chemistry and AI. The IP issues involved are equally complex, requiring a strategy that is tailored to the company just as its products or services are tailored to the individual. Things to consider include:
- Ensuring that patent claims are carefully drafted with the patentability exclusions of key markets in mind. Consider directing claims to computer-implemented methods to avoid some of these exclusions;
- Generating experimental data to support new methods and compositions prior to filing. For more on data in IP, see our earlier post here;
- Taking steps to protect unregistered IP that may reside in databases, AI algorithms, and proprietary software, for example by implementing a trade secrets policy, as discussed further here;
- Using registered trade marks to protect branding; this may be especially important for companies who are not carrying out original research.
This article was written by Partner and Patent Attorney Jennifer Bailey