Ectopic fat accumulation in the pancreas and its biomarkers: A systematic review and meta-analysis

Presence of fat in the pancreas increases the risk of metabolic co-morbidities. Detection and quantification of pancreatic fat is not a routine clinical practice, at least in part because of need to use expensive imaging techniques. We aimed to systematically review common markers of pancreatic fat in blood and to investigate differences in these markers associated with fatty pancreas. The search was conducted in 3 databases (EMBASE, Scopus, and MEDLINE). Studies in humans were eligible for inclusion if they reported on biological markers and percentage of pancreatic fat or fatty pancreas prevalence. Data were pooled for correlation and effect size meta-analysis. A total of 17 studies including 11 967 individuals were eligible for meta-analysis. Markers of lipid metabolism, including circulating triglycerides (r = 0.38 [95% confidence interval (CI) 0.31, 0.46]) and high-density lipoprotein cholesterol (r = -0.33 [95% CI -0.35, -0.31]), and markers of glucose metabolism, including glycated haemoglobin (r = 0.39 [95% CI 0.30, 0.48], insulin (r = 0.38 [95% CI 0.33, 0.43]), and homeostasis model assessment-insulin resistance (r = 0.37 [95% CI 0.30, 0.44], yielded the best correlations with percentage of pancreatic fat. Further, effect size analysis showed large and medium effects for the above markers of lipid and glucose metabolism. Circulating levels of triglycerides and glycated haemoglobin appear to be the best currently available markers of pancreatic fat. The approach of non-invasive and accurate detection of pancreatic fat by blood analysis should be further explored in the future, by investigating other potential biological markers of pancreatic fat.