A large global genetics study shows that many key drivers of Type 2 diabetes operate outside the bloodstream.

Scientists are getting a clearer picture of why Type 2 diabetes is so hard to pin down. In a major international project led in part by the University of Massachusetts Amherst and Helmholtz Munich in Germany, researchers linked hundreds of genes and proteins to the disease in ways that suggest cause, not just correlation. Many of these signals would likely stay hidden in studies that look only at blood.

The work, published in Nature Metabolism, points to a key challenge in diabetes research: the biology behind rising blood sugar does not play out the same way in every part of the body. It also shows why including people from many backgrounds matters, since genetic clues that stand out in one population may be faint or invisible in another.

Moving Beyond Blood-Based Studies

To sort likely drivers from bystanders, the team analyzed genetic data from more than 2.5 million people worldwide and treated it as a “natural experiment.” Because genetic variants are assigned at conception, they can help researchers trace which biological changes are more likely to sit upstream of disease. The study compared patterns across seven tissues tied to diabetes and four global ancestry groups, then asked a simple question: what do you miss if you only measure blood?

“We’ve known for some time now that tissue context is important to consider when trying to understand the mechanisms underlying the development of Type 2 diabetes,” says Cassandra Spracklen, associate professor of epidemiology at UMass Amherst and co-senior author of the paper. “But this work demonstrates just how important that context truly is.”

Blood is convenient, but Type 2 diabetes involves a coordinated breakdown across multiple organs, including adipose (fatty) tissue, the liver, skeletal muscle, and the insulin-producing cells of the pancreas. Gene activity and protein levels can look very different depending on where you measure them, so a blood sample can miss changes happening at the source.

Tissue-Specific Genetic Signals

Across the seven tissues, the researchers found causal evidence pointing to 676 genes. Yet overlap with blood was limited: only 18% of genes with a causal effect in a primary diabetes tissue, such as the pancreas, showed a matching signal in blood. At the same time, 85% of genetic effects observed in diabetes-relevant tissues were completely absent from blood-based analyses.

“By revealing both shared and tissue-specific mechanisms, our findings move us closer to improving strategies for Type 2 diabetes prevention and treatments that may be more effective across global populations,” adds Chi “Josh” Zhao, a doctoral student in epidemiology at UMass Amherst and co-first author of the study.

The Role of Global Genetic Diversity

The work builds on data from the Type 2 Diabetes Global Genomics Initiative, an international consortium that includes substantial representation from non-European populations. Using genome-wide association studies, the initiative has identified thousands of DNA variants linked to diabetes risk.

In the new analysis, researchers examined how these variants influence gene activity and protein levels across populations from Europe, Africa, the Americas, and East Asia. The study focused on genetic variants located near genes that affect gene expression or protein abundance and tested more than 20,000 genes and 1,630 proteins.

The results provide strong evidence that genetically predicted levels of 335 genes and 46 proteins influence Type 2 diabetes risk. Some associations were consistent across ancestry groups, while others emerged only when data from historically underrepresented populations were included.

The findings lay out a roadmap for future research aimed at understanding the biological pathways underlying Type 2 diabetes and developing more effective treatments.