Gut bacteria signature identified as early predictor of type 2 diabetes risk

Researchers examining the gut bacteria of more than 200,000 people have identified a distinct microbial signature that appears years before type 2 diabetes develops—a finding that could reshape how doctors identify and intervene with people at risk. The work, presented this week at the European Association for the Study of Diabetes conference in Milan, suggests that changes in the composition of gut bacteria can be detected long before blood sugar levels begin to climb, opening a window for prevention that current diagnostic methods simply do not capture.

The study analyzed genetic data from 229,025 adults in the UK and US, of whom 3,627 had type 2 diabetes and 14,022 had prediabetes. The remaining 211,376 had normal blood glucose levels. When researchers accounted for age, sex, and body mass index, they identified 789 bacterial species associated with diabetes. Some 168 of these species were more abundant in people with the condition, while 621 were less abundant. One example is Enterocloster bolteae, a spore-forming bacterium that was more common in diabetics and has been linked to poor cardiovascular and metabolic health. Another, Romboutsia timonensis, appeared less frequently in diabetic patients. Notably, some of the 789 species had never been isolated or studied in a laboratory before—they were identified for the first time through this analysis.

The bacterial changes were most pronounced in people taking medication for diabetes, but they also appeared in those with untreated diabetes. The researchers then asked a crucial question: could this microbial pattern predict risk in people who had not yet developed prediabetes or diabetes? They examined 135,093 individuals with normal blood glucose and measured how their blood sugar responded two hours after eating. Those whose blood sugar remained elevated longer than normal—a sign of insulin resistance, an early step toward diabetes—were significantly more likely to carry the bacterial signature associated with diabetes. As insulin resistance increased across the study population, the microbiome changes became more pronounced.

Current diabetes diagnosis relies on blood tests that measure glucose levels: the HbA1c test, fasting plasma glucose, and oral glucose tolerance tests. These tests identify people who already have abnormal blood sugar. They do not flag people with normal glucose levels who are on a trajectory toward diabetes. Risk assessment today depends on demographic and family factors—age, sex, body mass index, and family history. A microbiome signature could add another layer of information, potentially identifying people at risk before any metabolic changes show up in blood work.

Professor Tim Spector, a scientific co-founder of ZOE and one of the study's authors, described the findings as a major advance in understanding how the gut microbiome connects to metabolic disease. Earlier detection of these microbial shifts could enable intervention through diet and lifestyle changes, or through treatment, before diabetes takes hold. Dr. Gabriel Baldanzi, who led the work at the University of Trento in Italy, emphasized that lifestyle interventions—diet and physical activity—can prevent or delay diabetes onset, but people often struggle to see results beyond weight loss. If a microbiome test could show someone that their gut bacteria are shifting in the right direction, it might provide the motivation to sustain those changes.

The researchers were careful to note that microbiome analysis should not replace standard diagnostic tests. Instead, it could complement existing risk factors, offering additional insight into who might benefit most from preventative action. The study was made possible by the size of the cohort—more than 200,000 people—and by advances in computational methods for analyzing metagenomic data, the DNA sequences of bacteria in the gut. Many of the bacterial species identified as part of the diabetes signature are entirely new to science, discovered through this analysis rather than through traditional laboratory cultivation.

Previous research has suggested links between the gut microbiome and type 2 diabetes, but those studies often disagreed on which bacteria mattered most across large populations. This work, by examining hundreds of thousands of individuals, provides a more consistent picture. The findings raise the possibility of a future in which people at risk of diabetes could receive microbiome testing alongside traditional risk assessment, and could monitor changes in their bacterial composition over time as they make lifestyle changes. Whether this translates into clinical practice, and whether microbiome testing becomes part of routine diabetes prevention, remains to be seen—but the research suggests the gut may hold clues to diabetes risk that blood tests alone cannot reveal.