Ancestry-Stratified Transcriptome-Wide Association Study Reveals Tissue-Specific Mechanisms of Type 2 Diabetes Heterogeneity

Type 2 diabetes (T2D) is a heterogeneous disease with clear differences in prevalence, pathophysiology, and treatment response across ancestral groups. To investigate transcriptomic mechanisms that may underlie this variation, ancestry-stratified transcriptome-wide association studies were performed with S-PrediXcan using summary statistics from the multi-ancestry T2D genome-wide association study reported by Suzuki et al [1]. Genetically predicted gene expression was evaluated with GTEx v8 models in six metabolically relevant tissues: subcutaneous adipose, visceral omentum, skeletal muscle, pancreas, liver, and whole blood. Analyses were carried out separately in African (AFA), East Asian (EAS), European (EUR), Hispanic (HIS), and South Asian (SAS) ancestry groups. Gene-set enrichment testing was then applied to TWAS Z-scores, and results were combined across tissues with Stouffer meta-analysis.

The strongest prioritized signals were ancestry-specific. In African-ancestry individuals, insulin and GLP-1 secretion pathways were consistently downregulated, with the strongest signals seen in pancreas and visceral adipose tissue and driven largely by GLP1R, ADRA2A, ABCC8, ADCY5, and GNG5. In Hispanic-ancestry individuals, mitochondrial and oxidative phosphorylation programs were upregulated, while NAD+ metabolism and sirtuin-related pathways were downregulated; key driver genes included CPOX and CLN3. East Asian and Hispanic groups showed increased HLA and antigen-presentation signaling, whereas European and South Asian groups showed the strongest downregulation of glutamatergic signaling. These patterns were tissue-dependent and broadly consistent with the mechanistic clusters described by Suzuki et al [1].

Because the enrichment results were not corrected for multiple testing, these findings should be viewed as prioritized hypotheses rather than definitive associations. Even so, they offer a biologically coherent picture of ancestry-related transcriptomic differences in T2D and point to possible directions for more tailored therapeutic strategies. Replication in independent cohorts and ancestry-specific functional studies will be necessary.