Pancreatic dysfunction causally impacts region-specific cortical thinning: a mendelian randomization study revealing...

Alterations in cerebral cortex structure among patients with pancreatic diseases have been reported, yet the underlying causes remain unclear. This study employs Mendelian randomization (MR) to elucidate the causal effects of four pancreatic diseases (benign and malignant pancreatic cancer, acute and chronic pancreatitis) on cerebral cortex structure.

This study utilized genome-wide association study (GWAS) summary data from 437,418 participants in the Finnish consortium to identify the predictive role of genetic factors in acute pancreatitis. Similarly, data from these 437,418 individuals were used to evaluate the determining role of genetic factors in chronic pancreatitis. Data from 378,749 participants in the Finnish database were employed to assess the predictive effect of genetic factors on malignant pancreatic cancer, while data from 499,507 participants were used to examine the influence of genetic factors on benign pancreatic tumor. Additionally, this study incorporated data from 51,665 patients in the ENIGMA consortium to explore the associations between genetic susceptibility and both benign/malignant pancreatic cancers, acute/chronic pancreatitis, as well as the relationships between these diseases and changes in cerebral cortical thickness (TH) or surface area (SA). Magnetic resonance imaging (MRI) was used to conduct global measurements of SA and TH across 34 functional regions. The primary analytical approach employed the inverse-variance weighted (IVW) method for estimation, supplemented by pleiotropy residual sum and outlier test (MR-PRESSO), MR-Egger regression, and weighted median methods to detect heterogeneity and pleiotropy.

At the overall level, no significant effects on the cerebral cortex were observed for any of the four pancreatic diseases. At the functional level, malignant pancreatic cancer reduced cortical thickness in the bankssts region under global weighted conditions (β = - 0.12 mm, P = 1.17*10-7). For benign pancreatic tumor, negative associations were observed with cuneus (β = - 4.54, p = 0.028) and rostralmiddlefrontal (b = - 11.15, p = 0.063) surface areas. Acute pancreatitis was linked to middletemporal surface area (β = 14.59, p = 0.0042) and precentral thickness (β = - 0.006, p = 0.0046). Chronic pancreatitis exhibited a negative association with bankssts surface area (b = - 3.45, p = 0.0419). No pleiotropy was detected.

These findings provide MR evidence supporting a potential causal link between pancreatic diseases and cerebral cortex morphology. Our findings suggest that malignant pancreatic cancer has a causal effect on the structure of the cerebral cortex, suggesting the existence of a "pancreas-brain axis" mechanism.