Metabolic traits affecting the relationship between liver fat and intrapancreatic fat: a mediation analysis

Abstract

Aims/hypothesis The clinical importance of fat deposition in the liver and pancreas is increasingly recognised. However, to what extent deposition of fat in these two depots is affected by intermediate variables is unknown. The aim of this work was to conduct a mediation analysis with a view to uncovering the metabolic traits that underlie the relationship between liver fat and intrapancreatic fat deposition (IPFD) and quantifying their effect. Methods All participants underwent MRI/magnetic resonance spectroscopy on the same 3.0 T scanner to determine liver fat and IPFD. IPFD of all participants was quantified manually by two independent raters in duplicate. A total of 16 metabolic traits (representing markers of glucose metabolism, incretins, lipid panel, liver enzymes, pancreatic hormones and their derivatives) were measured in blood. Mediation analysis was conducted, taking into account age, sex, ethnicity and BMI. Significance of mediation was tested by computing bias-corrected bootstrap CIs with 5000 repetitions. Results A total of 353 individuals were studied. Plasma glucose, HDL-cholesterol and triacylglycerol mediated 6.8%, 17.9% and 24.3%, respectively, of the association between liver fat and IPFD. Total cholesterol, LDL-cholesterol, alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, γ-glutamyl transpeptidase, insulin, glucagon, amylin, C-peptide, HbA1c, glucagon-like peptide-1 and gastric inhibitory peptide did not mediate the association between liver fat and IPFD. Conclusions/interpretation At least one-quarter of the association between liver fat and IPFD is mediated by specific blood biomarkers (triacylglycerol, HDL-cholesterol and glucose), after accounting for potential confounding by age, sex, ethnicity and BMI. This unveils the complexity of the association between the two fat depots and presents specific targets for intervention. Graphical abstract

Remnant cholesterol, but not low‐density lipoprotein cholesterol, is associated with intra‐pancreatic fat deposition

“…Prior to blood sample collection, participants were required to fast for 8 to 10 hours 22,25 . Venous blood samples were then collected and centrifuged at 4°C.…”

“…This was a cross-sectional study conducted at the University of Auckland, whose study population was described in detail elsewhere. 22,25 In brief, participants were from the general population of Auckland (New Zealand), aged at least 18 years, and had no history of acute infectious or inflammatory disorders requiring medical care in the previous 6 months. All participants provided informed written consent for undergoing magnetic resonance imaging (MRI) of the abdomen.…”

“…29 For the quantification of liver fat deposition, magnetic resonance spectroscopy was used (which involved the placement of a region of interest in the right lobe of the liver). 22…”

“…Prior to blood sample collection, participants were required to fast for 8 to 10 hours. 22,25 Venous blood samples were then collected and centrifuged at 4 C. Blood lipids analysed included triglycerides, HDL cholesterol and total cholesterol from the standard lipid panel. LDL cholesterol (mg/dL) was calculated using the 2020 Sampson formula: total cholesterol/0.948 -HDL cholesterol/ 0.971 -(triglycerides/8.56 + [triglycerides x non-HDL cholesterol]/2140triglycerides 2 /16100) -9.44.…”

“…Fourth, the 17 studies included in the 2017 meta‐analysis of biomarkers of IPFD did not adjust for covariates 12 . However, it is appreciated that factors such as sex, age, ethnicity, insulin resistance, and liver fat deposition may affect IPFD 7,21,22 . Fifth, it has been suggested that there may be differences in fat deposition among the different regions of the pancreas, 23,24 and it is therefore possible that differences in regional IPFD could confound associations with components of the lipid panel.…”