Background
Pancreatic cancer has the lowest survival of all common cancers. Obesity is a major modifiable risk factor for pancreatic cancer, yet the causal molecular pathways linking obesity to disease risk remain poorly understood.

Objectives
To investigate molecular intermediates underlying the relationship between obesity and pancreatic cancer using genetic and molecular epidemiological approaches.

Methods
We systematically evaluated the associations of 21 previously reported molecular risk factors for pancreatic cancer – including lipids, glycemic traits, plasma hormones, and cytokines – as well as 5,810 unique plasma proteins from the deCODE consortium and UK Biobank, using Mendelian randomization (MR) and colocalization analyses. We used genome-wide association studies (GWAS) with up to 697,734 participants, to proxy various obesity traits, namely body mass index (BMI), waist to hip ratio (WHR) Abdominal subcutaneous adipose tissue (ASAT), Gluteofemoral adipose tissue (GFAT), Visceral adipose tissue (VAT), as well as Magnetic resonance imaging defined Pancreas and Liver fat. Genetic association estimates for pancreatic cancer were obtained from a genome-wide association study (GWAS) meta-analysis conducted by the Pancreatic Cancer Cohort Consortium (PanScan) and the Pancreatic Cancer Case-Control Consortium (PanC4), comprising 21,536 participants, including 9,040 cases. For molecular intermediates showing evidence of an association with pancreatic cancer, we further quantified their mediating effects in the relationship between adiposity-related traits and pancreatic cancer using MR mediation analyses. Finally, we explored tissues relevant to adiposity and/or pancreatic cancer, including pancreatic single-cell–level data, to investigate adiposity-driven mechanisms underlying pancreatic cancer risk using colocalization analyses.

Results
A one–standard deviation (SD; 4.7 kg/m²) increase in genetically proxied BMI was associated with a 40% higher risk of pancreatic cancer (OR = 1.40, 95% CI: 1.24–1.58). Positive associations were also observed for WHR (OR = 1.31, 95% CI: 1.12–1.53) and ASAT (OR = 1.96, 95% CI: 1.31–2.93), while GFAT, adjusted for BMI, was inversely associated with pancreatic cancer risk (OR = 0.82, 95% CI: 0.69–0.91). Consistent with previous reports, fasting insulin and plasma triglycerides showed positive associations with pancreatic cancer risk. In proteome-wide analyses, 39 unique plasma proteins associated with pancreatic cancer were supported in colocalization analyses. Across obesity-related traits, BMI was linked to CTRB1, MST1, B4GALT1, LGR4, GRP, and CCL3; WHR to LGR4, MST1, CCL3, and PACS2; and GFAT, adjusted for BMI, to CTRB1, ACP5, SERPINH1, GRP, LRCH4, ENO3, KRT18, and PNLIP. At the pancreatic tissue level, colocalization analyses integrating pancreatic gene expression and pancreatic cancer data supported the relevance of CTRB1 and KRT18 in bulk pancreatic tissue, and LGR4 and PNLIPRP2 in pancreatic islets, in mediating pancreatic cancer risk.

Conclusions/Implications
Our findings provide novel evidence for specific molecular intermediates linking obesity to pancreatic cancer risk, consistent with a potential etiological role and offering insights that may inform the development of targeted prevention strategies.