Globally, diabetes has become one of the defining health crises of the modern era. According to the World Health Organization, 14% of adults had diabetes in 2022, up from 7% in 1990. In the United States, more than 40 million people have diabetes, but about 11 million people remain undiagnosed. It is estimated that more than 115 million Americans have prediabetes, and about 80% of them don’t know it. In the UK, around 5.8 million people live with diabetes, and it is thought that up to 1.3 million people are undiagnosed.
“We’re talking about a pandemic that, in my opinion, is much worse than the Covid pandemic,” says Michael Snyder, a professor of genetics at Stanford University. “We need new ways to deal with this.”
The danger is not limited only to diabetes itself, but also the damage that accumulates silently for years before diagnosis. Persistently high blood sugar increases the risk of heart disease, stroke, kidney failure, blindness, and nerve damage. The earlier the disease is recognized, the greater the chance of preventing these complications or avoiding diabetes altogether.
Diagnosis still relies heavily on measuring blood glucose levels, most commonly using the HbA1c test, which estimates the average blood sugar over the past few months. Although widely used and generally reliable, they are not infallible. Results cannot reflect certain medical conditions or physiological factors that can affect blood sugar levels.
Researchers are increasingly concerned that current diagnostic tools are also less effective in some populations. Recent studies suggest that HbA1c can be falsely low in some black and South Asian people, delaying diagnosis until the disease becomes more advanced.
This disparity has sparked growing interest in more personalized, data-rich approaches to diabetes detection: those that combine biomarkers, wearable devices, and artificial intelligence to identify risk earlier and understand the disease in greater detail.
At Stanford University, Snyder and his colleagues have been exploring whether continuous glucose monitors (CGMs) – wearable sensors that track glucose levels in real time – can detect hidden metabolic patterns long before the traditional diagnosis of type 2 diabetes, which accounts for about 95% of cases. Although it is often associated with obesity – an important risk factor – thin people can also develop type 2. Snyder himself developed type 2 diabetes despite not fitting the stereotype of the disease.
“Glucose regulation involves many organ systems: the liver, muscles, intestines, pancreas, and even the brain,” Snyder says. “There are a lot of biochemical pathways, and it makes sense that glucose dysregulation might not be just one bucket.”
The Stanford team has developed an AI-powered algorithm that analyzes patterns in continuous glucose monitoring data to identify different forms of type 2 diabetes. In tests, the system identified some of these patterns with about 90 percent accuracy.
The researchers believe the findings could help identify people who already have metabolic problems long before a traditional diabetes diagnosis is made. “It’s a tool people can use to take preventative measures,” Snyder says. “If levels trigger a diabetes warning, dietary or exercise habits can be modified, for example.”
CGMs have also become cheaper and more accessible, with many now available without a prescription in the United States. Snyder believes these medications could eventually become part of routine preventive health care. “In an ideal world, people would wear them once a year,” he says. “The goal from our perspective is to keep people healthy rather than trying to fix them later.”
