Multi-omics study points to glycolytic genes as suggestive links to insulin resistance in obesity

Researchers have used an integrative multi-omics Mendelian randomization approach to identify suggestive glycolytic candidates associated with insulin resistance in metabolically unhealthy obesity, in a discovery-based study published in Diabetology & Metabolic Syndrome. The work, led by Fengyu Zhong, Xue Qin, Sufen Zhang, Yi Wang, Lin Lu and Dan Zhang, was published on 14 October 2026 as an open-access research article.

Metabolically unhealthy obesity is characterized by adipose tissue dysfunction and insulin resistance, but the causal genetic drivers remain elusive, according to the study’s authors. To probe these drivers, the team applied multi-omics Mendelian randomization, a framework that uses genetic instruments to test whether molecular signals are associated with a disease-relevant trait, while explicitly framing the results as suggestive rather than definitive.

The methods combined several layers of biological information. The researchers first integrated transcriptomic data from subcutaneous adipose tissue comparing metabolically unhealthy obesity with metabolically healthy obesity, drawing on the GSE156906 dataset to screen for glycolysis genes related to the unhealthy obesity phenotype. They then integrated insulin resistance genome-wide association study data with DNA methylation quantitative trait loci, expression quantitative trait loci and protein quantitative trait loci, and performed summary-data-based Mendelian randomization to investigate genetic associations between the obesity-related glycolysis genes and insulin resistance.

The screening step revealed a marked asymmetry in the adipose tissue transcriptome. In metabolically unhealthy obesity, 205 glycolysis genes were downregulated and 89 were upregulated, at a significance threshold of p < 0.05. This pattern of predominantly reduced glycolytic gene expression in the adipose tissue of metabolically unhealthy individuals formed the starting point for the downstream genetic analyses.

Built on those obesity-related glycolysis genes, the discovery dataset analysis identified 64 methylation quantitative trait loci, 11 expression quantitative trait loci and 7 protein quantitative trait loci associated with insulin resistance risk. Colocalization analysis was performed to determine whether the identified signals shared a causal genetic variant, and the authors report that it provided suggestive evidence in support of the associations.

One candidate stood out from the methylation-to-expression analysis. The gene HKDC1, tagged by the methylation probes cg05532178 and cg11762346, showed convergent summary-data-based Mendelian randomization and colocalization evidence suggestive of an association with insulin resistance in the discovery cohort. Tissue-specific analyses using GTEx v8 expression quantitative trait loci data showed directionally consistent negative associations between HKDC1 expression and insulin resistance across metabolically relevant tissues, reinforcing the signal across multiple tissue contexts.

A second candidate emerged from the combined expression and protein analysis. ITGAM showed directionally consistent associations at both the transcriptomic and proteomic levels with insulin resistance risk. The authors note, however, that these associations reached only nominal significance and that colocalization support was modest, marking ITGAM as a weaker, more tentative lead than HKDC1.

The study’s conclusion is deliberately cautious. The authors describe the results as exploratory multi-omics evidence for candidate genes potentially linking glycolytic dysregulation to insulin resistance in obesity, and they state that further validation in independent cohorts and experimental systems is warranted. As a discovery-based study, the work identifies hypotheses about which glycolysis-related genes may be genetically associated with insulin resistance, rather than establishing causal mechanisms in patients, and the suggestive nature of the colocalization evidence underscores the need for replication.

Subject of Research: Genetic associations between glycolysis-related genes and insulin resistance in metabolically unhealthy obesity

Article Title: Integrative multi-omics mendelian randomization identifies suggestive glycolytic candidates associated with insulin resistance in obesity: a discovery-based study

Article References: Zhong, F., Qin, X., Zhang, S., Wang, Y., Lu, L., & Zhang, D. (2026). Integrative multi-omics mendelian randomization identifies suggestive glycolytic candidates associated with insulin resistance in obesity: a discovery-based study. Diabetology & Metabolic Syndrome. https://doi.org/10.1186/s13098-026-02253-6

Image Credits: AI Generated

DOI: 10.1186/s13098-026-02253-6

Keywords: Mendelian randomization, multi-omics, glycolysis, insulin resistance, obesity, adipose tissue, HKDC1, ITGAM, colocalization, eQTL, mQTL, pQTL