A major discovery has been made in the field of nutrition and healthcare: the same foodstuffs affect different people in different ways, including the influence on their blood glucose levels.
To reach this conclusion, two researchers from the Weizmann Institute of Science, Eran Segal and Eran Elinav, gathered data from 800 people. During the second phase of the project, they incorporated this data into a machine learning algorithm, and found that it was able to predict the way sugar levels vary from one person to another after a given meal. This experiment is further proof that data science is not the stuff of fantasy but a very real innovation that is transforming the way researchers work.
YOU ARE WHAT YOU EAT? WELL, IT’S NOT QUITE THAT SIMPLE…
Do universal diets work? Not according to a report published by Eran Elinav and Eran Segal on 19 November 2015 in scientific journal Cell. By monitoring the glucose levels of 800 people over a week using a mobile app and a blood glucose monitor which regularly reported their glycaemic responses, they were able to glean a wealth of information such as meal content, exercise and sleep times. They also took daily stool samples which enabled them to monitor their microbiome i.e. the microorganisms in their intestine.
Eran Segal’s conclusion was surprising:
” The huge differences that we found in the rise of blood sugar levels among different people who consumed identical meals highlights why personalized eating choices are more likely to help people stay healthy than universal dietary advice.”
Blood glucose levels after the same meals do indeed vary from one person to another, due to the composition and function of their gut microbes or microbiome.
USING machine learning TO DEVISE PERSONALISED SOLUTIONS
So does this report mean the end of “one-size-fits-all” diets? The scientists have come up with a personalised approach to enable patients to monitor their glucose levels and thus help prevent and treat obesity and diabetes. Using the various data collected, they have developed an algorithm that can predict individualised responses to foods, based on the person’s diet, lifestyle, etc. Dr Elinav and Professor Segal then conducted follow-up study of another 100 volunteers to confirm the algorithm’s ability to predict changes in blood sugar levels as a response to different foods. The final stage of the study, involving around thirty volunteers, tested the scientists’ ability to use the algorithm to prescribe personal dietary recommendations for lowering blood-glucose-level responses to food. Half the volunteers followed the predictions of a nutritionist assisted by a team of researchers using a glucose monitor, whilst the others followed the algorithm’s predictions. The results showed significant reductions in blood glucose levels of all the volunteers, proving that the algorithm’s ability to predict based on the data provided.