A study led by researchers at Stanford Medicine shows that differences in blood sugar responses to certain carbohydrates depend on details of an individual’s metabolic health status
The differences in blood sugar response patterns among individuals were associated with specific metabolic conditions such as insulin resistance or beta cell dysfunction, both of which can lead to diabetes. The study findings suggest that this variability in blood sugar response could lead to personalized prevention and treatment strategies for prediabetes and diabetes
“Right now, the American Diabetes Association dietary guidelines do not work that well because they lump everyone together. This study suggests that not only are there subtypes within prediabetes, but also that your subtype could determine the foods you should and should not eat,” said Michael Snyder, PhD, the Stanford W. Ascherman, MD, FACS Professor in Genetics
A paper explaining the research will be published in Nature Medicine on June 4. Tracey McLaughlin, MD, a professor of endocrinology, co-led the study with Snyder. Yue Wu, PhD, postdoctoral scholar in genetics; Ben Ehlert, a graduate student in biomedical data science; and Ahmed Metwally, PhD, a former postdoctoral scholar at Stanford Medicine who is now a research scientist at Google, are joint first authors.
Michael Snyder
Blood sugar responses are a window to metabolic health
There is more than one pathway to diabetes, which is currently diagnosed based on elevated blood sugar levels, called hyperglycemia. Beta cells in the pancreas make the hormone insulin, which is then distributed to cells throughout the body to help convert glucose, or sugar, in the blood into energy. Beta cell dysfunction occurs when the pancreas fails to make or to release enough insulin, and insulin resistance occurs when cells in the body do not respond fully to insulin. Both beta cell dysfunction and insulin resistance can contribute to the high blood sugar levels that define prediabetes and Type 2 diabetes.
In the study, 55 participants without a history of Type 2 diabetes underwent metabolic testing for insulin resistance and beta cell dysfunction in addition to multi-omics profiling, which included tests for triglyceride levels, metabolites in plasma of the blood, measures of liver function and gut microbiome data. Just under half of the participants, 26 in total, had prediabetes, which was not surprising given that one in three adults in the U.S. has prediabetes.
Starchy foods were not equal; there was a lot of individual variability in which foods produced the highest glucose spike.”
The study participants wore continuous glucose monitors and ate same-sized portions of different carbohydrates that were delivered to their homes. There were seven foods tested: jasmine rice; buttermilk bread; shredded potato; pasta; canned black beans; grapes; and a berry mix containing blackberries, strawberries and blueberries. The participants consumed the food first thing in the morning, after fasting for 10 to 12 hours. Each participant ate each food type twice, and the research team tracked their blood sugar response during the three hours after their meal.
Many participants had a blood glucose spike after eating rice or grapes, regardless of their metabolic health status. The blood glucose responses to foods containing the highest amounts of resistant starch — potatoes and pasta — varied depending on the participants’ metabolic dysfunction
“Starchy foods were not equal; there was a lot of individual variability in which foods produced the highest glucose spike,” Wu said
The highest blood sugar spikes after eating pasta occurred in participants who had insulin resistance, and the highest spikes after eating potatoes occurred in participants who were either insulin resistant or had beta cell dysfunction. The multi-omics profiling showed that the potato-spiking participants also had high levels of triglycerides, fatty acids and other metabolites commonly seen in people with insulin resistance.
Tracey McLaughlin
Glucose spikes to beans were associated with histidine and keto metabolism, a state in which the body primarily uses fat for energy. Participants whose blood sugar spiked after eating bread were more likely to have hypertension, or high blood pressure
The highest blood glucose spikes after eating potatoes occurred in the participants who were the most insulin resistant and had the lowest beta cell function. Everyone spiked to some extent after eating grapes. The comparison of the blood glucose responses to potatoes versus grapes was associated with having insulin resistance, suggesting that this ratio could serve as a real-world biomarker for insulin resistance in the future.
“Such a biomarker would be useful because insulin resistance is amenable to lifestyle and medication interventions that can reduce risk for diabetes in high-risk individuals. At present there is no easy way to diagnose it in the clinic,” McLaughlin said
Eating to prevent blood glucose spiking
The researchers also examined whether eating a portion of fiber, protein or fat before carbohydrates reduced blood sugar spikes. The participants ate pea fiber powder, protein from boiled egg whites or fat in the form of crème fraîche 10 minutes before eating rice
Eating fiber or protein before the rice lowered the glucose spike, and eating fat before the rice delayed the peak of the spike. But these changes in blood glucose response occurred only in the metabolically healthy participants who were insulin sensitive or had normal beta cell function. Though eating fat, protein or fiber before carbohydrates had minimal impact on the blood glucose response patterns in participants with insulin resistance or beta cell dysfunction, McLaughlin and Snyder both think this type of mitigation is worth further research.
“Eating carbohydrates later in a meal is still a good idea even though it has not yet been sorted out whether it is best to eat protein, fat or fiber before carbohydrates. Eat your salad or hamburger before your French fries,” Snyder recommended
Researchers from Johns Hopkins University, Ultima Genomics, Amrita Vishwa Vidyapeetham University, University of Bergen and Cairo University contributed to this study
This study was funded by the National Institutes of Health (grants R01 DK110186-01, 2T32 HL09804911, UM1TR004921 and R01 DK114183), the National Library of Medicine (grant 2T15LM007033), the Stanford Precision Health and Integrated Diagnostics Center, Stanford Diabetes Center, American Diabetes Association (grant 11-23-PDF-76), and Nutrition for Precision Health
