Blood-based dietary signatures accurately predict both diet and risk of diseases

When it comes to food and diet research, it’s difficult to know what people are eating. Not to mention the risk of illness caused by what they eat.

Doctors and researchers usually ask people to complete a food frequency questionnaire that estimates calorie intake, food groups and nutrients. there is.

However, a research team led by a cardiologist at the University of Michigan School of Medicine is using molecular profiling and machine learning to identify blood-based diet characteristics that more accurately predict both diet and risk of cardiovascular disease and type 2 diabetes. I found a way to developResult is European Heart Journal.

Dieting is not one dimensional. It is constantly changing and traditional methods of evaluation are not perfect. ”

Venkatesh Murthy, MD, Ph.D., Senior Author, Cardiologist, University of Michigan Health Frankel Cardiovascular Center, Associate Professor of Cardiology, UM Medical School

“We need tools that are easier for everyone to use, more reliable and accurate. We can develop a better understanding of what is at risk, a disease that affects millions of Americans,” Mercy said.

Researchers used blood samples and food surveys in more than 2,200 white and black adults to identify dietary metabolite signatures and subsequent disease risk over a 25-year period. I conducted a study on development. Through a machine learning model, the researchers were able to create a blood-based diet signature that predicted a person’s overall diet across 19 food groups more accurately than her by 10-20%.

Additionally, to identify who is more likely to develop both diabetes and cardiovascular disease based on each food group, blood-based signatures were used to determine the standard measure of diet quality, a healthy diet. For example, if the food frequency questionnaire showed that red meat eaters had an 18% increased risk of diabetes, blood-based signatures We found a 55% increase.

“The use of metabolites to understand food intake and nutrition is an expanding area of ​​nutritional science,” said co-author Ravi Shah, M.D., Ph.D., cardiologist and associate professor of medicine at Vanderbilt University Medical Center. I’m here. “In addition to understanding which types of nutrition are better or worse for our health, the methods here allow those working in food science to take a metabolic snapshot of nutrition and diets and use them to improve our health. We may be able to better understand the impact of

The findings follow a $170 million award awarded to clinics and centers nationwide by the National Institutes of Health for “precision health nutritional research for development.” and[ing] An algorithm that predicts an individual’s response to food and diet,” according to the release.

Researchers say blood-based signature techniques should be tested in prospective, controlled studies of different diets. Knowing exactly what you’re adhering to gives even stronger results.

“Diet and nutrition research is really hard,” Marty said. Ultimately, studies like this may help us better understand the optimal diet for our patients.”