Pancreatic Steatosis Demonstrated at MR Imaging in the General Population: Clinical Relevance

Abstract

Purpose: To determine the relationship between pancreatic fat content and type 2 diabetes and prediabetes.

Materials and methods: From the prospective population-based Study of Health in Pomerania (SHIP), 1367 volunteers (563 men, 678 women; median age, 50 years) underwent whole-body magnetic resonance (MR) imaging at 1.5 T, which included multiecho chemical shift-encoded acquisition of the abdomen. SHIP was approved by the institutional review board, and written informed consent was obtained from all participants. The proton density fat fraction (PDFF) was calculated after correction for T1 bias, T2* bias, multipeak spectral complexity of fat, and noise bias. On the basis of oral glucose tolerance test results, participants were grouped into those with normal glucose tolerance (n = 740), those with prediabetes (n = 431), and those with confirmed type 2 diabetes but without medication (n = 70). PDFF was assessed in the pancreatic head, body, and tail. Multivariable regression analysis was conducted to investigate possible relationships of PDFF with demographic factors, behavioral factors, and laboratory data associated with the metabolic syndrome.

Results: In all subjects, the mean unadjusted pancreatic fat content was 4.4% (head, 4.6%; body, 4.9%; tail, 3.9%; being unequally distributed, P < .001). There was no significant difference in pancreatic PDFF among subjects with normal glucose tolerance, prediabetes, and type 2 diabetes (P = .980). Pancreatic PDFF showed a positive association with age and body mass index and a negative association with serum lipase activity (P < .001).

Conclusion: The presence of pancreatic fat is not related to prediabetes or diabetes, which suggests that it has little clinical relevance for an individual's glycemic status.

Figure 1:

Phantom data. Graph shows excellent agreement between PDFF and phantom fat content, with no differences in slope and intercept. Correlation is also near perfect (r² = 0.989). Results demonstrate high reliability and robustness of chemical shift–encoded PDFF for assessing tissue fat.

Figure 2:

PDFF maps illustrate different amounts of pancreatic fat in pancreatic head (arrow) in subjects without diabetes. (a) Image shows a very small amount of fat in pancreatic head. (b) Image shows fatty replacement of pancreatic head with PDFF of 19%. (c) Image shows complete fatty replacement as a maximum variant of pancreatic steatosis.