The participants included 28 adults (9 men, 19 women; ages 22–34) without known health conditions. They followed diets that were high in carbohydrate (65%-75% of total calories) and low in carbohydrate (15%-25% of total calories) for 6 days each. A 6-day washout period separated each diet regime, and the 24-day cycle was completed three times.

The different diets were calorie-matched, protein provided 15% of total energy, and the differences in carbohydrates and fat were mainly achieved through changes in staple foods like rice, bread, potatoes, and cooking oil.

The researchers looked at changes in gut microbiota composition and glycemic control within each individual participant^[1] and within the participant group. Certain gut microbiota were consistently associated with measures of glycemic control during the high-carbohydrate diet, but not the low-carbohydrate diet. The abundance of 10 species was used to develop a carbohydrate sensitivity score.

Among data from 609 cohort study participants who consumed a high-carbohydrate diet, the same 10-species score improved predictions for prediabetes or diabetes risk by 16% to 66% (random chance is 50%.) The microbiome score was also associated with poorer measures of glycemic control, like higher fasting glucose and HbA1c. Furthermore, 16 enzymes and 5 metabolites from the associated gut bacteria were similarly associated with markers of glycemic control.

Furthering personalized nutrition: A gut microbiota signature to predict blood sugar control Original paper