May Its the Diet, Not the Dieter? Interview with:
Dr. Eran Elinav PhD
Immunology Department and
Prof. Eran Segal PhD
Department of Computer Science and Applied Mathematics,
Weizmann Institute of Science,
Rehovot , Israel

Medical Research: What is the background for this study?

Response: Obesity and diabetes are rising epidemics, affecting a significant portion of the world’s population. Blood glucose levels are a major contributor to these epidemics, as they are associated with obesity and with risk to develop diabetes. Normalizing blood glucose can have highly beneficial effects on the health of individuals. Therefore, we decided to study the factors that affect blood glucose. Among these factors are the macro- and micronutrient content of the meal, a person’s lifestyle, medical state, and their microbiome. We then started collecting from participants these exact parameters, and studied their effect on blood glucose. We managed to successfully predict the blood glucose response of many people to many meals and to assign for people beneficial diets based on this prediction.

Medical Research: What are the main findings?

Response: Briefly, that people’s response to meals, measured by the postprandial glycemic response (PPGR) are associated with disease risk; that these responses are highly personal, meaning two people eating the same food may have radically different responses; and that we suggest a solution. Using a huge 800 people cohort we devised a predictor for a person’s PPGR to untested food by taking into account their health state, medical history, lifestyle and microbiome. We validated this predictor on 100 unseen people with very good results, and then used it to construct diets that lowered the PPGR to food.

Medical Research: What should clinicians and patients take away from your report?

Response: For many years, our thinking has been that people develop obesity, diabetes and other diet-related diseases because they are not compliant with our dietary advice. However, based on our study, another possibility is that people are in fact compliant but that the dietary advice that we are giving them is inappropriate. Thus, while the extensive profiling that each of our study participants underwent is not yet available to the general public, we believe that a take-home message for people from our work is that if a diet did not work for you, it may be the diet’s fault and not your fault.

We believe our research paves the way to a publicly available personalized nutrition solution. In our research settings, we showed that the comprehensive profile that we measured can be utilized to achieve and design personally tailored diets. We are now working on identifying the minimal set of inputs that would be needed to achieve accurate predictions of post-meal blood glucose responses. Our vision is to be able to derive predictions and personalized diets using a small set of inputs that people could fill out in questionnaires and a single microbiome sample, and we believe that this is both achievable in the near future and that that would be cost effective and financially viable.

Medical Research: What recommendations do you have for future research as a result of this study?

Response: We are currently pursuing multiple different paths out of the many research directions that our results highlighted. First, after realizing the great potential that lies within the microbiome with regards to glycemic response predictions, we believe that there is further work to be done. Our group has been and continues to develop new methods for microbiome analysis, that will provide us with better features to base our predictions on, as well as methodological advances in the machine-learning aspects of our algorithm, that would allow us to better utilize the features that our currently available to us. Our vision is that if we are successful, microbiome-based features would be able to represent or replace all other personal features, and thus, in the future, instead of having to extensively measure, test, and profile each and every individual with trained clinical personnel and expensive various tests, all we would need is to have them mail us a stool sample.

Second, in this work we performed a short-term dietary intervention trial that showed great promise. We are currently pursuing more challenging long-term dietary interventions, with the hope of successfully improving glycemic control and/or insulin sensitivity, weight, and other metabolic parameters.

Additionally, our study presented a unique methodological framework, and after its success in the very hard setting of healthy, not perturbed individuals, we are applying the same design to other, diseased populations. We are currently working on diabetic populations (both type I and II), and hope to show that prediction algorithms specifically tailored to these populations could improve quality of life and reduce comorbidities.


Personalized Nutrition by Prediction of Glycemic Responses

Zeevi, David et al. Cell , Volume 163 , Issue 5 , 1079 – 1094

[wysija_form id=”5″]

Dr. Eran Elinav PhD, & Prof. Eran Segal PhD (2015). May Its the Diet, Not the Dieter?