Biome-based individualized diets are most effective for managing blood-sugar levels, study confirms

If you and your partner or best friend eat the same foods but react differently, you might be able to blame differences in your gut microbiomes—the mix of bacteria in your digestive system—as well as your individual genes and physiologies.

The researchers’ goal for this study was to be able to predict individuals’ glycemic responses to foods. Analyzing the composition of gut microbiomes, which varies between people, and other factors is more accurate than predicting the reactions based on different foods.

Summary: Personalized diets based on one’s physical characteristics and gut biome may be the most effective way for people to manage their blood sugar, according to the findings of Assessment of a Personalized Approach to Predicting Postprandial Glycemic Responses to Food Among Individuals Without Diabetes.

The study: A growing body of evidence shows that individuals’ glycemic responses to the same foods differ greatly, depending on each person’s physiological and genetic characteristics, as well as the makeup of their gut microbiome. The microbiome might affect the body’s energy metabolism and its regulation of insulin.

Using a regression analysis that involves thousands of decision trees, researchers predicted participants’ post-meal glycemic responses to meals with a pool of 72 features:

The findings: The results showed “a wide variation in individual glycemic responses to the foods consumed,” the study reported. Each individuals’ results were consistent, but results between participants ranged from 6 milligrams of glucose per deciliter of blood to 94 mg/dL, with a mean of 30.7 mg/dL.

Study conclusions: The researchers’ model, based on each participant’s gut microbiome, correctly predicted the body’s blood-sugar response 62 percent of the time. When predictions about blood-sugar response are based on only carbohydrates, they are correct 40 percent of the time; when based on calories alone, the predictions are correct only 32 percent of the time.

This study confirmed the 2015 findings of an Israeli study, Personalized Nutrition by Prediction of Glycemic Responses that included 800 people. 

Why the research is interesting: High glucose levels are related to diabetes, heart disease, obesity, vision loss and heart disease. The results of this study explains, in part, why some people are energized after they eat fruit but others suffer a blood sugar spike that causes them to feel tired later.

“This study is the first critical step in defining and proving the value of a personalized diet. As a clinician, I have seen that my patients do not respond to the same foods the same way — just like not all weight-loss diets work for all people the same,” Dr. Heidi Nelson, a co-author on the study, said in a prepared statement. “For people who want to manage their blood glucose levels, we have a new model that predicts their unique response to foods.”

With our individualized model, people no longer have to give up all foods within a certain category,” said Purna Kashyap, M.B.B.S., co-director of the Mayo Clinic Center for Individualized Medicine Microbiome Program and an author on the study. “It allows them to choose specific foods within certain categories that fit well with their microbiome.”

Current recommendations: Recommendations for patients who need to lose weight or control their blood sugar are often based on foods’ calories or carbohydrate content. Eating fewer calories often results in weight loss, and eating low-carbohydrate or high-protein foods can be effective in improving blood-sugar levels among diabetics.

Who and when: All but nine of the 327 participants came from Olmsted and Hennepin counties in Minnesota; the other nine lived in Florida. (Valid data was available for 293 participants.)

Participants had to be 18 years old or older, with access to a mobile device and a web browser. None of them had diabetes, were pregnant or were substance abusers. Others excluded from the study included patients who had chronic gastrointestinal or metabolic disorders; had used antibiotics in the previous three months; had undergone bariatric weight-loss surgery; had taken fertility treatments in the three months before the study; had undergone radiation or chemotherapy treatment for cancer in the previous two year; had cancer at the time of the study; and more.

The mean age of the 327 participants was 45, and 78 percent were women. More than three-quarters of the participants were white and non-Hispanic; 3.7 percent were Asian; 1.5 percent were black or African-American. About 14 percent did not report their race.

Participants were considered to be in good health:

How it was done: Patients were recruited from Oct. 11, 2016, to Dec. 13, 2017. During a meeting at the beginning of each study week, staff measured participants’ height, weight, waist and hip circumference, blood pressure and pulse. Staff also drew blood to measure participants’ long-term blood-sugar, or A1c, levels.

What was measured: Each participant’s stool sample was analyzed for microbiome features prior to the week they took part of the study. Based on those analyses, researchers predicted how each participant’s blood sugar would react to specific foods.

For six days, the participants ate a specified breakfast—a plain bagel with cream cheese or one of three cereals, with or without milk— but maintained their normal eating habits for other meals. The participants kept track of what they ate, how much and when, using a mobile application that included a food catalog.

With a continuous glucose monitor, participants’ blood sugar levels were tracked and compared to each participant’s food consumption.

Authors: Helena Mendes-Soares, who has a doctorate in microbial ecology from Indiana University Bloomington, is a research associate at Mayo Clinic in Rochester, Minnesota; Tali Raveh-Sadka, Ph.D. in computational biology and is director of research at DayTwo, an Israeli company that analyzed the participants’ microbiomes (the Mayo Clinic is invested in Day Two); Heidi Nelson, M.D., is a professor of surgery whose research focuses on colon and rectal cancer; and more.

Published: Assessment of a Personalized Approach to Predicting Postprandial Glycemic Responses to Food Among Individuals Without Diabetes was published Feb. 8 at JAMA Network Open.

Related reading:

Big companies and investment firms discover personalized nutrition

Accountant's personal crisis inspires microbiome breakthrough

Experts reveal advancements in microbiome research at Natural Products Expo West 2017