NUTRITION-AI GAP: HOW LACK OF BIG DATA IS SLOWING INNOVATION IN NUTRITION SCIENCE
- Maryam Adegboyega
- 2 days ago
- 3 min read
The nutrition AI gap big data challenge highlights how the lack of harmonized food datasets is limiting the use of artificial intelligence in nutrition research and precision health.
Artificial intelligence (AI) has transformed how most industries operate; however, the field of nutrition is still struggling to catch up. A recent article published in Frontiers in Nutrition describes the nutrition–AI gap as being driven largely by a lack of robust big data, particularly harmonised databases that capture the precise chemical composition of foods as a whole. The article acknowledges the breakthrough of nutrition research in the 20th century with the discovery of key food components such as macronutrients and micronutrients. This gave rise to a reductionist approach, where scientists have spent decades trying to link individual nutrients or isolated food compounds to disease outcomes. The authors argue that this approach has had limited success and suggest that it may be time to rethink the framework used to study diet–disease relationships.
With the ability to detect complex dietary patterns that are not easily apparent to humans, AI presents a promising opportunity to uncover hidden relationships between diet and health. However, realising this potential depends on the availability of comprehensive, well-annotated datasets that describe foods in their entirety. Such datasets could transform nutrition into a true big data science, enabling more precise and actionable insights.
In emerging areas such as precision nutrition, a more data-driven approach can support the design of dietary recommendations tailored to age, metabolic status, health conditions, goals, ethnicity, gender, and environmental context. While such initiatives already exist, their adoption is still limited. The United States NIH “All of Us” Research Program, for example, uses large-scale data and AI-driven methods to explore individual variability in health and disease, including responses to lifestyle and dietary patterns. Similarly, the UK-based PREDICT studies integrate mobile applications, wearable sensors, and AI analytics to examine how genetics, gut microbiome composition, lifestyle, and metabolic differences influence postprandial responses.
Another area where AI shows strong potential is real-time dietary assessment. One of the major challenges in nutrition research and intervention design is accurately assessing dietary intake. Food recall is often unreliable and subject to bias, while traditional dietary assessment methods are time-consuming. Modern computer vision models, powered by deep learning, now enable automated food recognition and instant estimation of calories and nutrient content, with steadily improving accuracy.
Eradicating malnutrition in Africa, including Cameroon, remains a high priority for governments and development partners. With robust and representative data, countries could benefit significantly from AI models capable of identifying risk patterns and enabling early, targeted interventions. Beyond malnutrition, policymakers could also use AI-driven insights to better prioritise populations for disease prevention and resource allocation based on exposure and vulnerability profiles.
In food formulation, AI also offers opportunities to explore and optimise novel or underutilised local ingredients. Through clustering and optimisation techniques, it is possible to design nutrient-dense foods with improved organoleptic properties, making them both nutritious and acceptable to consumers.
There is an urgent need for collaboration between AI engineers, nutrition scientists, and clinicians to build the future of food and nutrition. This multidisciplinary synergy will support the development of tools that are transparent, ethical, validated, and practically useful in addressing real-world nutritional challenges.
Sources
How multi-omics and AI are revolutionizing nutrition research for sustainable precision diets. https://www.nutritioninsight.com/news/multi-omics-ai-nutritional-epidemiology-precision-diets.html
Personalised REsponses to DIetary Composition Trial (PREDICT): an intervention study to determine inter-individual differences in postprandial response to foods. https://www.researchgate.net/publication/345349897_Personalised_REsponses_to_DIetary_Composition_Trial_PREDICT_an_intervention_study_to_determine_inter-individual_differences_in_postprandial_response_to_foods
NIH AllofUs Program. https://allofus.nih.gov/
Comments