Microbiome holds clues to success with personalized nutrition in diabetes
BALTIMORE — Postprandial glucose responses to specific foods vary from person to person depending on the makeup of his or her gut microbiome, underscoring a need for personalized nutrition to help better manage diabetes, according to a speaker here.
Humans carry within them thousands of different species of bacteria, mostly located inside the large intestine, with each species holding thousands of different genes, Eran Segal, PhD, professor in the department of computer science and applied mathematics at the Weizmann Institute of Science, said during a presentation at the American Association of Diabetes Educators annual meeting.
Microbiome researchers, Segal said, are only beginning to understand the influence of those millions of bacterial genes on the human body, affecting everything from obesity, asthma and cardiovascular disease to glucose excursions after consuming a meal. Additionally, research conducted by Segal and colleagues has suggested that the foods humans consume can induce metabolic changes by directly altering the gut microbiota, whereas microbiota, much like humans, exhibit diurnal rhythmicity that can further effect host epigenetics.
“In our human genome, we have only about 25,000 genes,” Segal said. “That means that by focusing only on human genetics, we have been studying only 1% of the genetic material that we carry among us.”
Predicting glucose responses
A quick glance at the popular press shows there is still a great deal of confusion surrounding a very basic question, Segal said: What is a healthy diet that we should all follow?
“Being scientists, we wanted to take an unbiased, data-driven approach to the study of nutrition,” Segal said. “We thought a lot about what marker of nutrition we should consider as a healthy marker.”
The researchers decided to focus on postprandial glucose response to food, Segal said, in part because changes in glucose in the 2 hours after consuming a meal are related to weight management and the management and progression of diabetes. Postprandial hyperglycemia is also associated with several other comorbid conditions, Segal said, including changes in drug metabolism, depression and increased risk for dementia.
As part of the Personalized Nutrition Project, Segal and colleagues used blinded continuous glucose monitoring (CGM) to analyze glucose levels in 800 adults without diabetes who consumed 46,898 meals during 1 week while logging food intake, exercise and sleep (54% with overweight; 22% with obesity; 60% women) in an effort to measure individual postprandial glucose responses and variability across people and identify factors associated with any variability. Over the week, participants followed normal daily routines and dietary habits, apart from the first meal of every day, which was one of four standardized meals with 50 g carbohydrates provided by researchers. Prior to the study, researchers collected blood samples and a single stool sample for microbiota profiling.
After 7 days, the researchers observed high variability in postprandial glucose response among people who consumed identical meals and noted multiple associations between the standardized meal postprandial glucose responses of participants and both their clinical and gut microbiome data.
“Any food, as long as it has some amount of carbohydrates to digest, resulting in glucose, the response to that food will be very different across different people,” Segal said. “This underscores and highlights the needs for personalized nutrition, at least as far as balancing blood sugars go.”
The researchers then created a machine-learning algorithm integrating blood parameters, dietary habits, anthropometrics, physical activity and gut microbiota, which was able to accurately predict personalized, postprandial glucose responses to real-life meals. Dietary interventions based on this algorithm successfully normalized blood glucose levels, Segal said. The results were published in Cell in November 2015, and the findings were translated into a product that provides personalized nutrition recommendations, Segal said.
Tailored nutrition interventions
A large variability in postprandial glucose responses can have several public health implications, Segal said. Any universal nutrition recommendations, he said, will have limited efficacy in balancing glucose levels, as each person is different. Similarly, concepts like the glycemic index will be useful for some, but not for others. It is microbiota composition and function, Segal said, that explains much of the variability in post-meal glucose responses to food, and dietary interventions targeting postprandial glucoses responses, he said, may induce consistent changes in microbiota, although that needs to be seen in longer randomized controlled trials.
“We saw that following a diet that balances your blood sugar levels, there are some changes in the microbiome composition that are consistent across people,” Segal said. “These changes were also beneficial, in the sense, when there is a certain bacterium where low levels are associated with greater weight loss, this bacterium tended, on average, to go down in participants when following the low [postprandial glucose response] diet. We think this is very intriguing, because it may suggest that if you follow such diets, then, on the long-term, it will have an effect not just while you’re on the diet, but perhaps improve [microbiota] composition for some time after you finish the diet.
“Different people have very different blood glucose responses to food,” Segal said. “This tells you that there is a need to tailor our nutrition to the individual if we want to balance blood sugar levels.” – by Regina Schaffer
Segal Evan. Personalized medicine approaches for treatment of diabetes based on gut microbiome. Presented at: American Association of Diabetes Educators; Aug. 17-20, 2018; Baltimore.
Zeevi D, et al. Cell. 2015; doi:10.1016/j.cell.2015.11.001.
Disclosure: Segal reports he is a consultant for Day Two.