Pancreas volumes in humans from birth to age one hundred taking into account sex, obesity, and presence of type-2 diabetes

Abstract

Our aims were (1) by computed tomography (CT) to establish a population database for pancreas volume (parenchyma and fat) from birth to age 100 years, (2) in adults, to establish the impact of gender, obesity, and the presence or absence of type-2 diabetes on pancreatic volume (parenchyma and fat), and (3) to confirm the latter histologically from pancreatic tissue obtained at autopsy with a particular emphasis on whether pancreatic fat is increased in type-2 diabetes. We measured pancreas volume in 135 children and 1,886 adults (1,721 nondiabetic and 165 with type-2 diabetes) with no history of pancreas disease who had undergone abdominal CT scan between 2003 and 2006. Pancreas volume was computed from the contour of the pancreas on each CT image. In addition to total pancreas volume, parenchymal volume, fat volume, and fat/parenchyma ratio (F/P ratio) were determined by CT density. We also quantified pancreatic fat in autopsy tissue of 47 adults (24 nondiabetic and 23 with type-2 diabetes). During childhood and adolescence, the volumes of total pancreas, pancreatic parenchyma, and fat increase linearly with age. From age 20-60 years, pancreas volume reaches a plateau (72.4 +/- 25.8 cm(3) total; 44.5 +/- 16.5 cm(3) parenchyma) and then declines thereafter. In adults, total ( approximately 32%), parenchymal ( approximately 13%), and fat ( approximately 68%) volumes increase with obesity. Pancreatic fat content also increases with aging but is not further increased in type-2 diabetes. We provide lifelong population data for total pancreatic, parenchymal, and fat volumes in humans. Although pancreatic fat increases with aging and obesity, it is not increased in type-2 diabetes.

Fig. 1

Phantom control and representative CT images demonstrating the pancreas in normal, obese, and elderly individuals. a: Phantom study of volume measurement (SD at 50 = 0.8, at 100 = 0.5, and at 150 = 1.2). b: Representative CT image of a normal pancreas (33-year-old female, BMI 24.1). Pancreas is indicated by the broken line. c: Representative CT image of an obese subject (53-year-old male, BMI 35.3) showing an enlarged pancreas. CT density of the pancreas is low and interlobular fissures are seen on the surface edge. d: Representative CT image of an elderly subject (83-year-old male, BMI 25.8). The pancreas is atrophic. CT density is decreased and inhomogeneous with marked interlobular fissures present.

Fig. 2

Total, parenchymal, and pancreatic fat volumes with age calculated from CT images. a: Total pancreas volume. b: Parenchymal volume. c: Fat volume. d: Fat/parenchyma ratio (F/P ratio). Total pancreas volume increases with age until ∼30 years and then reaches a plateau. After age 60, total pancreas volume declines due to a decline in parenchymal pancreas volume. Pancreatic fat remains proportionate to pancreas parenchyma until age ∼50 years, after which the F/P ratio increases because of loss of pancreas parenchyma, but not fat. For n value for each age group, see Supplementary Table 2.

Fig. 3

Effect of BMI on pancreatic volume. Total pancreatic volume (a), parenchymal volume (b), fat volume (c), and fat/parenchyma ratios (F/P ratio) (d) were measured from CT scan images in lean (n = 460), overweight (n = 460), and obese groups (n = 230). All components of pancreatic volume are significantly increased in overweight and obese subjects compared with lean subjects. **P < 0.01.

Fig. 4

Interaction of aging and BMI on pancreatic volumes. Total pancreas volume (a), parenchymal volume (b), and fat volume (c) are shown for each decile for lean, overweight, and obese groups. *P < 0.05, **P < 0.01 vs. lean. §P < 0.05, §§P < 0.01 vs. overweight.

Fig. 5

Relationship between BMI and pancreatic volumes. Despite considerable individual variance there is a significant relationship between BMI and total (a), parenchymal (b), and fat (c) volumes in nondiabetic subjects (gray dots, broken line) and subjects with type-2 diabetes (solid dots and solid line). The relationships are comparable in individuals with type-2 diabetes (T2DM) and nondiabetics (ND). (a) ND r = 0.3, P < 0.0001; T2DM r = 0.4, P < 0.0001, (b) ND r = 0.2, P < 0.0001; T2DM r = 0.2, P = 0.01, (c) ND r = 0.4, P < 0.0001: T2DM r = 0.5, P < 0.0001.

Fig. 6

Comparison of pancreatic fat percentages from pancreata obtained at autopsy in lean and obese subjects with or without type-2 diabetes. a: Histological comparison of pancreas fat (%) between lean and obese subjects with or without type-2 diabetes. LND, lean nondiabetic; LD, lean diabetic; OND, obese nondiabetic; OD, obese diabetic. *P < 0.05 vs. LND. #P < 0.05 vs. LD. b: Correlation analysis between BMI and histological pancreas fat (%). Circle, LND; triangle, LD; reverse triangle, OND; rhombus, OD.