How Machine Learning and Data Science Can Advance Nutrition Research

Print Friendly, PDF & Email

In this special guest feature, Kyle Dardashti, CEO & Founder of Heali, discusses how machine learning and data science bring exciting potential to the world of personalized nutrition. Heali is a personalized nutrition company focused on supporting people with medical nutrition therapy. Formerly Kyle led a marketing agency and product incubator which he co-founded. He was responsible for innovating global brands with new-age marketing strategies. Today at Heali he leads a diversified and experienced staff combining world-renowned leaders from technology, nutrition and medicine sectors.

Nutrition plays a pivotal role in health and wellness. Diet-related diseases are the most common cause of death in the United States and diet alone is the leading risk factor for premature death worldwide.1 Fortunately, health conditions linked to poor diet are almost always preventable. However, individuals must understand the link between diet and disease in order to make appropriate lifestyle and behavior changes.

Even though food and nutrition has been studied for centuries, the field of nutrition has only recently accepted the idea of personalized nutrition: providing individuals with specific, actionable dietary insights based on genetics, metabolism, disease, and environment.

Advancing Personalized Nutrition Recommendations 

Machine learning and data science bring exciting potential to the world of personalized nutrition. Combining these two technologies together on a cohesive platform that supports continuous tracking would allow for real-time validated nutrition recommendations tailored to an individual’s lifestyle. In other words, collecting data on an individual’s meal habits, symptom patterns, physical activity, and lab values can be compiled and analyzed to produce personalized suggestions on what to eat, when to eat, and why. For example, this platform could identify individual foods or ingredients that trigger specific symptoms and suggest a dietary pattern that is supported to reduce these symptoms.

Advancing Nutrition Research Efforts

A platform such as the one described above has strong potential to improve nutrition on an individual level; it also brings unprecedented opportunity to advance the field of nutrition research. Currently, nutrition research efforts are costly and have a high participant burden causing them to be inaccessible and/or unfeasible to use long-term. Experimental studies often require participants to adhere to an exact dietary pattern that is different from a standard diet, leading to high participant drop-off rates. In observational studies, diet evaluation is often based on self-reported food intake and can be impacted by interviewer bias and/or participant bias and often leads to inaccurate summary data. And, because of ethical considerations, nutrition research efforts are almost always conducted on the healthy adult population, leading to undiverse sample sizes and outcomes that fail to represent the population at large or over time.

Current nutrition research efforts fail to recognize all of the many factors that affect individual dietary needs. Leveraging machine learning and data science on a common platform would allow complex links between age, disease, lifestyle, and diet to be recognized on both an individual and community level. Researchers can extrapolate this data to further understand individual nutrition needs and  validate dietary recommendations.

In-Silico Research Models

On a wider scale, this technology can be used to develop in silico clinical trials, which involve using statistical modeling to create a synthetic patient pool with previously-collected user data. Because in silico trials are conducted entirely within computer models, they can help to explain cause and effect relationships in rapid time with zero participant burden. In silico trials would allow researchers to evaluate the effectiveness of proposed diet therapies and understand how nutrition needs change throughout the lifecycle, even within vulnerable populations such as the acutely ill who are often left out of current nutrition research efforts. These insights can also be used to guide and improve the success of future human subject research trials. Long-term, these insights can advance clinical care by providing medical professionals with validated nutrition interventions specific to their patient population.

The Future

Technological advancements such as machine learning and data science bring vast potential to the future of nutrition and can help us to better understand the link between diet and health across all ages, genders, ethnicities, disease states, lifestyles, and more. Combining these insights with current advancements in microbiome testing, genetic testing, and continuous lab monitoring technologies, personalized nutrition will be at the forefront of this next era of integrated healthcare. Long-term, these advancements have the power to solve some of today’s biggest health concerns.

[1] National Institutes of Health. “2020–2030 Strategic Plan for NIH Nutrition Research.” National Institutes of Health, May 2020.

Sign up for the free insideBIGDATA newsletter.

Join us on Twitter: @InsideBigData1 –

Speak Your Mind