Microbiome Model Predicts Glucose Spikes Better Than Carbs, Calories

Marlene Busko

February 08, 2019

An algorithm based on a person's gut microbiome more accurately predicts an individual's spike in blood glucose after eating than just knowing a food's calories or carbohydrates, new data show.

The study, conducted in adults who did not have diabetes, "really shows that we can use models to predict glycemic responses to food better than just looking at calorie or carbohydrate content," senior author Heidi Nelson, MD, from the Mayo Clinic in Rochester, Minnesota, told Medscape Medical News.

This is preliminary research, she cautioned. "This doesn't make a claim about preventing a disease" such as prediabetes or diabetes, she emphasized.

Rather, "this is a first step in proving that we can maybe personalize nutrition in those who seek that option."

For the study, Nelson's team from the Mayo Clinic collaborated with researchers from the Israeli company DayTwo, which markets a test kit based on earlier work in Israeli cohorts.

The new research extends the previous findings to an American cohort and suggests that this model "may allow individuals to better manage their glycemic responses to food consumed," first author Helena Mendes-Soares, PhD, the Mayo Clinic, and colleagues report in an article published online today in JAMA Network Open.

"It's exciting on many levels," Nelson said. "First, this is a real tool...that can be made accessible to people on an app; by submitting [stool] samples and [using] that tool...people can predict which foods are going to give the rise in the sugar value.

"It also suggests that in the future, similar types of models might become available for other types of patients," such as those with high cholesterol, she added.

"It's a nice, exciting study," Michael P. Snyder, PhD, professor and chair of genetics, Stanford University, California, told Medscape Medical News.

As reported previously, Snyder's team recently found that some people with normal blood glucose levels experience postprandial glycemic spikes that are in the diabetic range.

"It's still early days," he echoed, and "we need more...data and more...people doing this sort of thing, so that we can get more precise" models to predict glycemic responses.

"They had people eating a standard meal of a bagel and cream cheese," he noted. "We need a lot of people eating different foods to see if they compare."

Gut Bacteria and Postprandial Glycemic Response

Growing evidence suggests that individuals experience significantly different changes in blood glucose level (which can be estimated using a continuous glucose monitor that measures glucose in interstitial fluid) after eating the same foods, the authors write.

In 2015, Israeli researchers led by Professors Eran Segal and Eran Elinav from the Weizmann Institute of Science, Rehovot, Israel, developed an algorithm, described in an article in Cell, that used microbiome data as well as other factors (eg, age, sex, height, hip circumference, and physical activity) to predict postprandial glycemic responses in people who did not have diabetes.

The model was developed and validated in Israeli cohorts of 800 and 100 individuals, respectively.

To test the algorithm in an American cohort, Nelson's team enrolled 327 individuals who did not have diabetes (318 individuals in Minnesota and nine individuals in Florida) in the 6-day study.

The mean age of the participants was 45 years, and 78% were women.

Two days before the study began, participants provided stool samples, which were sent to the DayTwo laboratory in Israel, where the gut microbiome composition was determined.

During the study, participants wore a continuous glucose monitor. To calibrate the monitor, participants used a manual glucose monitor at least four times a day. They used an app — the DayTwo Food & Activity Logger, which contains the MyNetDiary food catalog — to log their food intake and physical activity.

Participants were instructed to eat their usual foods, except that for breakfast on 4 days, they were asked to eat a standardized meal consisting of a bagel and cream cheese.

The R coefficient for a correlation between predicted and actual postprandial glycemic spike was 0.62 for the model that used microbiome data (a perfect positive correlation would be 1.0), but, as determined on the basis of a food's calorie content, it was only 0.34, and it was 0.40 when determined using carbohydrate content.

"When we talk about diet, we tend to make it 'one size fits all,' " Nelson noted. This study shows "that everybody responds differently according to a number of variables, including their microbiome."

"We really need to understand how people are different" and how this affects health, Snyder agrees.

"I think this is the future, and each of these studies is an important step in that direction," he said.

The study was funded by DayTwo, Inc. The Mayo Clinic has investments in the company. Mendes-Soares and Nelson received support from DayTwo, and several coauthors are employees of the company. Full disclosures of relevant financial relationships are listed in the journal article. Snyder has disclosed no relevant financial relationships.

JAMA Netw Open. Published online February 8, 2019. Full text

Follow Medscape on Facebook, Twitter, Instagram, and YouTube

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....