1

Ectopic Fat Storage in the Pancreas using 1H-MRS: Importance of Diabetic status

and modulation with Bariatric Surgery induced Weight Loss

3
4

Running title: pancreatic fat deposition and modulation with weight loss

5
6

Corresponding author: Bndicte Gaborit, Inserm U1062, Inra U1260, Aix Marseille

Universit, Facult de Mdecine, 27 boulevard Jean Moulin, 13385 Marseille cedex 05,

France. Tel: +33-491324504; Fax: +33-491254336; E-mail: benedicte.gaborit@ap-hm.fr

9
10

Bndicte Gaborit*1,2,4, Ines Abdesselam*2,3,4, Frank Kober3,4, Alexis Jacquier3,4,6, Olivia

11

Ronsin1, Olivier Emungania7, Nathalie Lesavre8, Marie-Christine Alessi2,4, Jean Charles

12

Martin2, 4, Monique Bernard3,4, Anne Dutour1,2,4

13

*Authors contributed equally to the work

14
15

1Department

of Endocrinology, Metabolic Diseases and Nutrition, CHU Nord, Marseille,

16

France

17

2Inserm U1062,

18

19

Mdicale (CRMBM), UMR 7339, Marseille, France

20

4Aix

21

6Department

of Radiology, CHU Timone, Marseille, France

22

7Department

of digestive surgery, CHU Nord, Marseille, France

23

8Centre

24

9Gastroenterology

25

Conflict of Interest: none to declare

Inra U1260, Facult de Mdecine, Marseille, France

Aix Marseille Universit CNRS, Centre de Rsonance Magntique Biologique et

Marseille Universit, Marseille, France

dinvestigation Clinique (CIC), CHU Nord, Marseille, France

Department CHU Nord, Marseille, France

26

ABSTRACT

27

Objectives: recent literature suggests that ectopic fat deposition in the pancreas may

28

contribute to endocrine and exocrine organ dysfunction, such as type 2 diabetes,

29

pancreatitis or pancreatic cancer. The aim of this study was to determine factors

30

associated with pancreatic triglyceride content (PTGC), and to investigate the impact of

31

bariatric surgery on ectopic fat pads, pancreatic fat (PTGC) and hepatic fat (HTGC).

32

Subjects: 45 subjects (13 lean, 13 obese non diabetics and 19 type 2 diabetes, matched

33

for age and gender) underwent 1H-magnetic resonance spectroscopy, CT visceral

34

abdominal fat, metabolic and lipidomic analysis, including HOMA-IR, HOMA-B and

35

plasma fatty acids composition. Twenty obese subjects were reassessed 6 months after

36

bariatric surgery.

37

Results: PTGC was significantly higher in type 2 diabetic (23.83.2%), compared to

38

obese (14.03.3; p=0.03) and lean subjects (7.50.9%; p=0.0002). PTGC remained

39

significantly associated with type 2 diabetes after adjusting for age and sex (=0.47;

40

p=0.004), or even after adjusting for waist circumference, triglycerides and HOMA-IR

41

(=0.32; p=0.04). Type 2 diabetes, C18:1n-9 (oleic acid), uric acid, triglycerides, and PAI-

42

1 were the 5 more important parameters involved in PTGC prediction (explained 80% of

43

PTGC variance). Bariatric surgery induced a huge reduction of both HTGC (-51.27.9%)

44

and PTGC (-43.87.0%) reaching lean levels, while BMI remained greatly elevated. An

45

improvement of insulinresistance (HOMA-IR), and no change in HOMA-B was observed

46

after bariatric surgery. The PTGC or HTGC losses were not correlated, suggesting tissue

47

specific mobilization of these ectopic fat stores.

48

Conclusion: Pancreatic fat increased with type 2 Diabetes and drastically decreased

49

after bariatric surgery. This suggests that PTGC may contribute to beta cell improvement

50

seen after bariatric surgery. Further long-term interventional studies are warranted to

51

examine this hypothesis, and to determine the degree to which ectopic fat mobilization

52

may mediate the improvement in endocrine and exocrine pancreatic functions.

53

Key words: pancreatic fat, pancreatic triglyceride content, Proton magnetic resonance

54

spectroscopy, magnetic resonance imaging, obesity, type 2 diabetes, beta cell function,

55

bariatric surgery

56
57

58

INTRODUCTION

59
60

Ectopic lipid deposition is currently of growing interest, since body fat distribution has

61

been shown to be at least, as important as the degree of obesity in the development of its

62

comorbidities (1). The development of fat in ectopic undesirable sites (such as the liver,

63

the heart, the muscle, and the pancreas) is supposed to be due to a dysfunctional

64

subcutaneous adipose tissue not being able to appropriately store or oxidize the energy

65

excess surplus. This lipid overload, namely triglycerides overflow is redirected to cells of

66

internal organs, leading to steatosis and to organ dysfunction (2-4). Pancreatic fat is one

67

of the less studied ectopic fat. Pancreatic biopsies are difficult to obtain, necropsies are

68

rapidly degraded by digestive enzymes, and ultrasound is limited by digestive

69

interposition. Besides, its exact delimitation between triglycerides intra-betacells/acinar

70

cells and/or intrapancreatic adipocytes is unclear. Recent studies have suggested that

71

pancreatic fat accumulation is one of the mechanism leading to the impairment of cell

72

function, but conflicting results have been found in normal glucose tolerance, impaired

73

glucose tolerance/impaired fasting glycemia versus type 2 diabetes (5-9), even with the

74

hyperglycemic clamp method. Pancreatic fat deposition has been also associated with

75

lipotoxicity of the exocrine pancreas. An increase in pancreatic fat has been reported in

76

acute pancreatitis exacerbation, pancreatic fistula and pancreatic cancer (10-14).

77

Similarly, the association between pancreatic cancer and obesity or diabetes is well

78

known (15). Thus, evaluating pancreatic fat content noninvasively in obese subjects is of

79

major importance to precise the putative role of pancreatic fat deposition on endocrine

80

and exocrine pancreas diseases. Proton magnetic resonance spectroscopy (1H-MRS) is a

81

reliable noninvasive quantitative method to assess the pancreatic triglyceride content in

82

vivo, which has been validated in rodents and in humans (16). Remarkably, it has been

83

recently suggested that the modulation of ectopic fat stores with weight loss is tissue

84

specific. We have recently reported a significant decrease of epicardial fat after bariatric

85

surgery, without a significant change in myocardial triglyceride content (17). On the

86

contrary, other authors have demonstrated that hepatic triglyceride content is

87

extremely flexible with a drastic decrease only after two weeks of caloric restriction (18,

88

19). The effect of dietary interventions on pancreatic fat has been scarcely investigated

89

(20). Therefore, in this study we aimed at determining factors associated with an

90

increase in pancreatic fat content measured with 1H-MRS in 45 subjects (lean and obese)

91

and the modulation of pancreatic and hepatic triglyceride content 6 months after

92

bariatric surgery in a subgroup of 20 morbid obese patients.

93
94
95

SUBJECTS AND METHODS

Subjects

96

Forty five Caucasian patients, aged 43.3 1.8 years, with (n=13) healthy volunteers,

97

(n=13) obese non diabetics, and (n=19) type 2 diabetics, according to American

98

Diabetes Association (21) matched for age and gender were recruited after giving

99

written informed consent. The mean duration of diabetes was 4.8 5.1 years, and mean

100

HbA1c= 7.51 1.1%. Diabetic patients treated with insulin or thiazolinediones, agents

101

known to modulate body fat distribution were excluded from the study. Stable glucose

102

lowering therapy restricted to metformine (n=15) and/or sulfonylurea was required (n=

103

7). Exclusion criteria included: excess alcohol intake (>20units/week), history of

104

hepatitis, acute or chronic pancreatitis, malignant disease, recent (<3 months) change in

105

weight, and metal implant, claustrophobia or other MRI contraindications.

106

Among the obese patients, 20 (12 non diabetics and 8 diabetics, mean BMI=44.41.0

107

kg/m2) were included to undergo bariatric surgery for severe obesity (BMI>40 or 35

108

with at least one co-morbidity) and participated a second time, six months after surgery,

109

to the same exploration. The local ethics committee approved the study (Marseille,

110

Comit de Protection des Personnes, Sud-Mditerrane II), and the investigation

111

conformed to principles outlined in the Declaration of Helsinki.

112

Anthropometric measurements, abdominal CT scan

113

All subjects underwent weight, height, waist (WC) and hip circumference

114

measurements, blood pressure, and evaluation of adipose tissue (AT) distribution

115

(Subcutaneous (SAT) and visceral (VAT)) by abdominal computed tomography, as

116

described earlier (22, 23). A single breath-hold, 10-slice, centered on the fourth lumbar

117

vertebra was acquired. Quantification of SAT and VAT was performed by manual

118

delineation, and the areas were expressed in square centimeters. Percentage losses

119

were calculated according to the formula: percentage of parameter loss = (parameter

120

before surgery parameter after surgery/parameter before)*100.

121
122

Biochemical analyses

123

Patients underwent blood analysis after overnight fasting including lipid profile, glucose

124

profile, inflammation biomarkers and liver enzymes. Plasma adiponectin was measured

125

by ELISA (Quantikine Human Adiponectin, R&D Systems, Minneapolis, Minnesota), and

126

serum leptin levels were determined using a commercially available enzyme-linked

127

immunoassay kit (SPI-BIO, Bertin, France). The insulin-resistance homeostasis model

128

assessment (HOMA-IR) and the insulin-secretion homeostasis model assessment

129

(HOMA-B) indexes were calculated as follows: (Fasting Plasma Glucose (FPG)(mmol/L)

130

Fasting Plasma Insulin (FPI) (mUI/L)/22.5, and (20 x FPI (mUI/L) / (FPG-3.5),

131

respectively. The fatty acids of 250L of plasma, either in the free form or combined in

132

complexed lipids, were directly converted into volatile fatty acid methyl esters (FAME)

133

by adding an acetyl-chloride methanol solution, as already described (24). They were

134

then analyzed by fast gas chromatography, performed on a 0.5-l sample injected in split

135

mode with a hydrogen flow rate of 10 ml/min. The column was a capillary column

136

(BP70,10 m0.1 mm ID0.2m film thickness) (SGE International Pty. Ltd., Australia).

137

The temperature program was as follows: initial, 60C with a 0.5-min hold; ramp,

138

20C/min to 200C, 7C/min to 225C with a 1-min hold and then 160C/min to 250C

139

with a 1-min hold. The instrumental conditions were as follows: The carrier gas was H2

140

at a flow rate of 61.4 cm/s and a constant head pressure of 206.8 kPa; flameionization

141

detector was set at 280C; the air and nitrogen make-up gas flow rates were 450 ml/min

142

and 45 ml/min, respectively; the injector split ratio was 200:1; the detector sampling

143

frequency was 50 Hz; the autosampler injections had a volume of 0.5l; and the run time

144

for a single sample was 13.23 min with a sample injection-to-injection time of 16 min.

145
146

Pancreas and liver H-MRS:

147

All the 45 participants were scanned in a head first supine position, using a 3-T MRI-

148

scanner equipped with a 32-element phased array coil (Verio, Siemens Medical

149

Solutions, Erlangen, Germany). A spectroscopic volume of interest was positioned using

150

multi-slice FLASH image series with 3 different orientations (axial, coronal and sagittal),

151

and a point-resolved single-voxel proton spectroscopy sequence (PRESS) was used to

152

determine the molecular content of lipids and water (Figure 1). Special attention was

153

paid to voxel size and position to include only pancreatic and hepatic tissues, avoiding

154

any peri-visceral fat. Rectangular voxel (Volume of interest) dimensions for pancreatic

155

and hepatic acquisition were 17x15x15mm and 15x20x20mm, respectively. Pancreatic

156

fat content was measured in the corpus, which is the widest part of the pancreas. The

157

liver voxel was placed in the anterior part of liver, carefully avoiding major blood vessels

158

and biliary ducts (Figure 1). Scan parameters were as follows: 16 scans, repetition time

159

(TR)=1000 ms, echo time (TE)=35ms. Single scan spectra were also acquired as

160

reference for saturation correction. All acquisitions were performed during breath hold.

161

Spectra were analyzed using home-developed software running under an Interactive

162

Data Language environment (ITT Visual Solutions, Boulder, Colorado, USA). The

163

pancreatic triglyceride content (PTGC) and hepatic triglyceride content (HTGC) were

164

determined by integrating, for each spectrum, the frequency domain and expressing the

165

result as a percentage of the water signal (%TG (triglyceride) = TG/water 100). Final

166

amount of pancreatic and hepatic content of triglyceride were determined after

167

normalization with single-scan water peak.

168

The mean intra-subject coefficient of variation between five repeated pancreatic and

169

hepatic fat measurements was 1.65% and 1.67% respectively, with Pearson correlation

170

coefficients (r) of 0.99, and 0.99 respectively. The coefficient of variation inter-subject

171

between two observers was 4.55% with r=0.99 for pancreatic fat, and 5.87% with

172

r=0.99 for hepatic fat.

173

Pancreas and liver volumes:

174

Axial FLASH images series were acquired to cover the abdomen. Post processing liver

175

and pancreas contours were measured by manual delineation in each slice. Slices were

176

then summed and multiplied by slice thickness to obtain organs volumes.

177
178

Statistics:

179

For statistical analysis, data are presented as mean SEM, and level of significance set at

180

p=0.05. As values were not normally distributed, log transformation was performed.

181

Group differences (lean, obese and diabetics) were tested by ANOVAs, or Mann-

182

whitneys. Spearmans correlations were used to evaluate the relation between

183

pancreatic fat, liver fat and anthropometrical or biological parameters. Comparison of

184

baseline and 6-month parameters were performed using paired t-test or Wilcoxon

185

matched-pairs signed-ranks test when appropriate. Step by step multivariate regression

186

analysis was done to specify the link between diabetes and PTGC. Multivariate partial

187

least square (PLS) regression analyses were used to assess independent determinants of

188

pancreatic fat, and to select the best combination of variables able to predict pancreatic

189

fat variance; this method has been shown to be suitable for this kind of dataset (25, 26).

190

After performing the PLS regression using clinical and biochemical variables as X

191

explaining variables and pancreatic fat as Y dependent variables. The weight of each X

192

variable in predicting Y was determined using the variable importance in projection

193

(VIP) score of the SIMCA-P software (version 12, Umetrix, Umea, Sweden). For each X

194

variable, the VIP a cut-off value was set at 1, which corresponded to accepted values for

195

VIP (27). The validity of each VIP score was further statistically ascertained by Jack-

196

Knifing (a re-sampling method) (28) with 95% confidence intervals.

197
198

RESULTS

199

Factors associated with ectopic fat deposition

200

The general characteristics, anthropometrical and metabolic variables of the lean versus

201

obese non-diabetic, and type 2 diabetic patients (T2D) are shown in Table 1. Lean and

202

obese subjects were younger than T2D. All further analyses were thus age and gender

203

adjusted. As expected, obese and T2D subjects had significantly higher waist

204

circumference (WC), VAT, triglycerides (TG), low-grade inflammation, insulin-

205

resistance, liver enzymes and lower HDL cholesterol than lean subjects. Remarkably,

206

lean and non-diabetic obese were not significantly different for PTGC and HTGC

207

(p>0.05). By contrast, type 2 diabetic patients had more HTGC and PTGC than lean and

208

non-diabetic (ND) obese, HTGC: 26.13.9% (diabetics) vs 10.93.8% (obese ND),

209

p=0.004, 6.00.6% (lean), p=0.0002; PTGC: 23.83.2% (diabetics) vs 14.03.3% (obese

210

ND), p=0.03 or 7.50.9% (lean), p=0.0002 (Figure 2). PTGC increased with type 2

211

diabetes independently of BMI, (r=0.60, p=0.0002). A significant association was

212

observed between pancreatic triglyceride content (PTGC) and hepatic triglyceride

213

content (HTGC) (r=0.42, p=0.004). In univariate analysis, PTGC and HTGC were both

214

positively correlated with waist circumference, FPG, FPI, HOMA-R, HbA1c, visceral

215

abdominal fat, uric acid, triglycerides, FAME, PAI-1, ALT, GGT and negatively correlated

216

with HDL cholesterol (Table S1). HTGC was also negatively associated with adiponectin,

217

and PTGC positively associated with age and BMI. HTGC was associated with liver

218

volume (r=0.66, p=0.0008), while PTGC was not correlated to pancreas volume (Table

219

S2). PTGC remained significantly associated with diabetes (=0.47; p=0.004) after

220

adjusting for age, and sex, or after adjusting for WC, TG and HOMA-IR (=0.32; p=0.04).

221

All together these parameters explained 65% of PTGC variance (Table 2). To further

222

precise the parameters associated with pancreatic fat deposition, we analysed the

223

plasma lipodomic signature of our subjects. As shown in Table S3, PTGC was highly

224

linked to Oleic Acid C18:1n-9, (r=0.53,p=0.001) and to Palmitic acid C16: 0 (r=0.41,

225

p=0.013). To give a further insight into the biological variables that can predict PTGC, we

226

additionnally performed a PLS statistical analysis comprising all parameters. From that

227

analysis all parameters explained up to 80% of PTGC variance. Using the VIP scores of

228

the PLS procedure, we selected the 5 most relevant variables in this model, among

229

which diabetes was again involved in prediction of pancreatic fat along with C18: 1n-9,

230

uric acid, TG and PAI-1 (Figure S1).

231
232

Changes in ectopic fat stores after bariatric surgery

10

233

Twenty morbidly obese patients (6 men, 14 women, 8 with preoperative type 2

234

diabetes) underwent bariatric surgery (sleeve gastrectomy (n=13), or Roux-en-Y gastric

235

bypasses (n=7)) and lost significant weight at 6 months from 119.9.43.9 to 90.43.4 kg,

236

and BMI from 44.41.0 to 33.81.0 kg/m2 (p<0.0001) (Table 3). No difference was

237

observed regarding the percentage weight loss, neither for the type of surgery (p=0.4),

238

nor for the sex (p=0.5). As expected, bariatric surgery induced a significant reduction in

239

visceral and subcutaneous fat (-374, p=0.002; -314%, p<0.0001, respectively), and an

240

improvement of glucose tolerance, low-grade inflammation, liver enzymes and insulin-

241

resistance (Table 3). Among the 8 type 2 diabetic patients, 5 had a complete resolution

242

of their diabetes with HbA1c<6%. One patient could stop sulfonylurea and was glycemic

243

controlled with metformin only (HbA1c=6.5%). Lipid profile parameters did not change

244

but this was mainly associated with a drop in the use of lipid-lowering drugs. After

245

bariatric surgery, Docosahexaenoic Acid (DHA) increased significantly, and C18:3n-6,

246

C20:3n-6 decreased (Table S4). Interestingly, we observed a drastic decrease of HTGC (-

247

51.27.9%) and PTGC (-43.87.0%), higher than that of BMI (-24.51.2%), suggesting

248

an extreme rapid mobilization of these ectopic fat stores with weight loss (Figure 3).

249

Besides, PTGC and HTGC postoperatively reached lean levels.

250

The percentage of PTGC and HTGC losses were neither correlated, nor associated with

251

the percentage of weight loss (p=NS), suggesting a tissue specific mobilization of ectopic

252

fat stores. The percentage of PTGC and HTGC losses were not associated with the

253

plasmatic decrease in C18:3n-6, C20:3n-6. Whereas delta-BMI was not linked to baseline

254

BMI (p=0.22), delta-HTGC and delta-PTGC were highly linked to their initial levels

255

(Delta-HTGC with HTGC at baseline r=0.93, p<0.0001; Delta PTGC with PTGC at baseline

256

(r=0.88, p<0.0001), meaning that higher was the level at baseline, higher was the

257

decrease.

11

258

Regarding glucose metabolism, alongside with the decrease in ectopic fat stores,

259

bariatric surgery induced a huge decrease in insulin-resistance, as assessed by HOMA-IR

260

(from 5.70.9 to 1.30.3, p<0.0009), whereas HOMA-B remain stable (from 207.828.5

261

to 144.225.8), suggesting an improvement in cell function (Table 3).

262
263

DISCUSSION

264

The results of the present study show that inflammation, insulinresistance, liver

265

enzymes, C18:1n-9, which is a marker of lipogenesis, uric acid, and importantly type 2

266

diabetes are factors associated with ectopic fat accumulation in the pancreas. Moreover,

267

this is the first study to show the variation of pancreatic fat content with bariatric

268

surgery, and the tissue specific mobilization of these ectopic fat stores.

269

As rewieved by Smith et al, (13) many terms have been used to describe pancreatic

270

ectopic fat deposition: fatty pancreas, pancreatic fat, non alcoholic pancreatic steatosis,

271

non alcoholic fatty steatopancreatitis, pancreatic steatosis, the latter being often chosen.

272

While hepatic steatosis is characterized by the accumulation of lipid droplets (mainly

273

triglycerides and toxic intermediates, such as ceramides or diacylglycerol) in

274

hepatocytes, ectopic fat deposition in pancreas seems to be more complex. Indeed, it has

275

been described as the accumulation of triglycerides in islet and acinar cells, but also as

276

the infiltration of mature adipocytes in pancreatic tissue (13, 29). Lee et al, showed in

277

ZDF rats that the pancreatic profile of fat content in whole pancreas paralleled that of

278

islets (30). Moreover, they showed that intracellular lipid accumulation preceded

279

adipocyte infiltration. These results suggest that pancreatic 1H-MRS could be used as a

280

surrogate for evaluation of pancreatic triglyceride content both in acinar and beta cells.

281

This evaluation is of major importance to better precise the putative role of pancreatic

282

fat deposition on endocrine and exocrine pancreas diseases. Indeed, non alcoholic fatty

12

283

steatopancreatitis has been associated with type 2 diabetes, pancreatic cancer,

284

pancreatic fistula and to a more severe course of acute pancreatitis (11-13, 31-33). More

285

precisely, Sadr-Azodi et al demonstrated that waist circumference, but not BMI, is an

286

independent risk factor for the development of acute pancreatitis (34, 35). We confirm

287

in this study that pancreatic fat is highly linked to waist circumference and to visceral

288

abdominal fat deposition (36). Furthermore, one should expect that infiltrated

289

pancreatic adipocytes exert a deleterious paracrine effect on neighboring acinar cells.

290

Accordingly, Navina et al, showed that unsaturated fatty acids generated locally from

291

pancreatic fat were pro-inflammatory and induced a local toxic effect on pancreatic cells,

292

inducing necrotic cell death, thus explaining the increased severity of pancreatitis in

293

obese patients (10). Because of its potential local lipotoxicity both in exocrine and

294

endocrine pancreas, pancreatic fat level in obese and type 2 diabetic patients, and its

295

modulation with weight loss are crucial to analyze.

296

The association of peripheral insulin resistance, impaired regulation of hepatic glucose

297

production, with the declining in insulin secretion, progressing inexorably to cell

298

failure is the cornerstone of type 2 diabetes pathogenesis. In line with other studies, (8,

299

9, 16) we confirm that type 2 diabetic patients have more hepatic and pancreatic fat

300

than lean or obese individuals. Moreover, our results show that the increase in

301

pancreatic fat with diabetes was independent of BMI, and persisted after adjustment for

302

the index of insulin-resistance HOMA-IR. Furthermore, we did not find a significant

303

difference in PTGC between lean and obese subjects. This points out the specific link

304

between diabetes and pancreatic fat. Whether glucotoxicity accelerates pancreatic fat

305

deposition, or whether pancreatic fat is an important contributor to cell dysfunction

306

cannot be addressed in our study. Besides, the causal relationship between pancreatic

307

fat and cell function is very debated (6, 8, 9, 37). This may be due to the complex

13

308

kinetic of beta cell function, with an initial compensative increase in insulin secretion

309

due to insulin-resistance, and a second later decrease in insulin secretion. In obese ZDF

310

rats, an early increase in both pancreatic triglyceride content and insulin secretion has

311

been described; this increase in pancreatic fat was followed by a decline in insulin

312

secretion, 4 weeks after the increase in PTGC (30). In humans, literature has been

313

conflicting. Lim et al, in an interesting study, showed a significant reduction of

314

pancreatic fat in 11 type 2 diabetics submitted to an 8-week very low calorie diet. This

315

decrease of pancreatic fat was associated with a normalization of the first phase insulin

316

response (20). In healthy Chinese volunteers, Wong et al found no association between

317

fatty pancreas and HOMA-B after adjusting for HTGC and body mass index (38).

318

No chronological pattern of pancreatic fat content during time course of type 2 diabetes

319

development is available in humans. Probably, in a very early phase, pancreatic fat

320

content could be positively linked to the increased of insulin secretion, and later could

321

contribute to its decrease. The heterogeneity (age, BMI, duration of the disease,

322

ethnicity, genetic factors) of patients in most studies combines these two steps. This was

323

nicely observed by Szczepaniak et al (39), who demonstrated significant ethnic

324

differences in pancreatic steatosis, and striking differences in the relationship between

325

pancreatic steatosis and cell dysfunction among white, black and Hispanic non-

326

diabetic obese patients.

327

Interestingly, we found an association between PTGC and inflammatory, lipid markers

328

such as PAI-1, triglycerides, C18:1n-9 oleic acid, which is a marker of increase

329

lipogenesis and activity of delta 9 desaturase, and uric acid plasmatic level. With

330

lipidomic analysis we succeeded in explaining 80% of PTGC variance, exceeding

331

previous studies (36). Besides, uric acid has been shown to be associated with hepatic

332

fat deposition, (40, 41) and to stimulate inflammatory mediators and oxidative stress in

14

333

pancreatic islet cells in vitro (42). Moreover, Gotoh et al nicely showed the major role of

334

IL-10 in preventing obesity-induced fat accumulation in the pancreas of mice

335

splenectomized (43), confirming the important role of inflammation in pancreatic fat

336

deposition.

337

The effect of bariatric surgery on pancreatic fat content has never been reported to date.

338

Our results uniquely demonstrate a drastic decrease of pancreatic triglyceride content 6

339

months after bariatric surgery, in line with a major decrease in insulin resistance and

340

reversal of type 2 diabetes. Multiples studies have shown that diabetes remission after

341

bariatric surgery results from the joint improvement of both insulin resistance directly

342

linked to weight loss, and -cell function (44, 45). Changes in the pattern of

343

gastrointestinal hormones release have been linked to the early adaptation of -cell

344

function. Our results suggest that the decrease in pancreatic fat content could also

345

participate to the reversal of type 2 diabetes after bariatric surgery, or to the long-term

346

maintenance of the remission. Future prospective studies are needed to confirm this

347

hypothesis. The flexibility of hepatic fat and pancreatic fat that we report in this study is

348

in contrast with the lack of decrease of myocardial triglyceride content 6 months after

349

bariatric surgery (17). This suggests that endogenous fat stores do not respond equally

350

or to the same extent, or with the same kinetic to caloric restriction. Furthermore, the

351

percentage of PTGC or HTGC losses were not correlated, enhancing the tissue specific

352

mobilization of these ectopic fat stores, which has been suggested by others (46, 47).

353

The important differences in the modality of fat accumulation in those two tissues

354

(intracellular accumulation in liver, adipocytes development and intracellular

355

accumulation in pancreas) probably contribute to the differences in mobilization of

356

ectopic fat stores. Previous studies have found that bariatric surgery reduces the risk for

357

cancer in severe obese patients, particularly in women (48, 49). Whether the reduction

15

358

of pancreatic fat content could participate to a reduction in pancreatitis and pancreatic

359

cancer development deserves further investigation. The long-term effect of bariatric

360

surgery on endogenous fat stores is also needed: whether the decrease in pancreatic and

361

hepatic fat is maintained with the stabilization of weight, or whether there is a rapid

362

increase with weight regain has to be determined.

363
364

The limitations of this study must be considered. We did not assess insulin secretion

365

with the hyperglycemic clamp method, or insulin resistance with the euglycemic

366

hyperinsulinemic clamp, and used surrogate markers such as HOMA-B and HOMA-IR.

367

The cross sectional nature of the first part of the study limits our ability to draw

368

definitive conclusion and causal inferences from the relationships observed. The H-

369

MRS technique did not allow for the distinction between infiltration by adipocytes or

370

triglyceride deposition in non adipose cells.

371

In conclusion, our study shows that inflammation, insulin-resistance and importantly

372

type 2 diabetes are factors associated with ectopic fat accumulation in the liver and

373

pancreas. Our results show a significant drastic reduction of pancreatic and hepatic

374

triglyceride content 6 months after bariatric surgery, demonstrating fast adaptation and

375

tissue specific mobilization of these ectopic fat stores with marked weight reduction.

376

These improvements could be associated with the reduction in exocrine and endocrine

377

pancreas diseases with bariatric surgery. The effect of pancreatic fat on cell function

378

will have to be evaluated in long-term interventional studies.

379
380

ACKNOWLEDGMENTS

381

We thank Gabrielle Belpassi, Valrie Griset, Sylvie Bourrely, and Aurelia Mesnier, of the

382

Centre dInvestigation Clinique at Marseille North Hospital, for their help in choosing

16

383

patients for inclusion and for their biological technical assessment. We thank Patricia

384

Ancel, Virginie Tassistro, and Jacques Boukhalil for their technical help. We thank the

385

nurses and all the staff of the endocrinology and digestive surgery department for their

386

invaluable support.

387
388

CONFLICT OF INTEREST

389

None to declare

390
391

SUPPLEMENTARY INFORMATION

392

Supplementary information (tables and figures) is available at International Journal of

393

Obesity website.

394

17

395

REFERENCES

396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442

1.
Despres JP. Body fat distribution and risk of cardiovascular disease: an
update. Circulation. 2012 Sep 4;126(10):1301-13. PubMed PMID: 22949540. Epub
2012/09/06. eng.
2.
Szendroedi J, Roden M. Ectopic lipids and organ function. Current opinion in
lipidology. 2009 Feb;20(1):50-6. PubMed PMID: 19133412. Epub 2009/01/13.
eng.
3.
van Herpen NA, Schrauwen-Hinderling VB. Lipid accumulation in nonadipose tissue and lipotoxicity. Physiology & behavior. 2008 May 23;94(2):23141. PubMed PMID: 18222498. Epub 2008/01/29. eng.
4.
Wang MY, Grayburn P, Chen S, Ravazzola M, Orci L, Unger RH. Adipogenic
capacity and the susceptibility to type 2 diabetes and metabolic syndrome.
Proceedings of the National Academy of Sciences of the United States of America.
2008 Apr 22;105(16):6139-44. PubMed PMID: 18413598. Pubmed Central PMCID:
2329698. Epub 2008/04/17. eng.
5.
Hannukainen JC, Borra R, Linderborg K, Kallio H, Kiss J, Lepomaki V, et al.
Liver and pancreatic fat content and metabolism in healthy monozygotic twins
with discordant physical activity. Journal of hepatology. 2011 Mar;54(3):545-52.
PubMed PMID: 21112658. Epub 2010/11/30. eng.
6.
Heni M, Machann J, Staiger H, Schwenzer NF, Peter A, Schick F, et al.
Pancreatic fat is negatively associated with insulin secretion in individuals with
impaired fasting glucose and/or impaired glucose tolerance: a nuclear magnetic
resonance study. Diabetes/metabolism research and reviews. 2010
Mar;26(3):200-5. PubMed PMID: 20225188. Epub 2010/03/13. eng.
7.
Saisho Y, Butler AE, Meier JJ, Monchamp T, Allen-Auerbach M, Rizza RA, et
al. Pancreas volumes in humans from birth to age one hundred taking into
account sex, obesity, and presence of type-2 diabetes. Clin Anat. 2007
Nov;20(8):933-42. PubMed PMID: 17879305. Pubmed Central PMCID: 2680737.
Epub 2007/09/20. eng.
8.
Tushuizen ME, Bunck MC, Pouwels PJ, Bontemps S, van Waesberghe JH,
Schindhelm RK, et al. Pancreatic fat content and beta-cell function in men with and
without type 2 diabetes. Diabetes care. 2007 Nov;30(11):2916-21. PubMed PMID:
17666465. Epub 2007/08/02. eng.
9.
van der Zijl NJ, Goossens GH, Moors CC, van Raalte DH, Muskiet MH, Pouwels
PJ, et al. Ectopic fat storage in the pancreas, liver, and abdominal fat depots:
impact on beta-cell function in individuals with impaired glucose metabolism. The
Journal of clinical endocrinology and metabolism. 2011 Feb;96(2):459-67.
PubMed PMID: 21084401. Epub 2010/11/19. eng.
10.
Navina S, Acharya C, DeLany JP, Orlichenko LS, Baty CJ, Shiva SS, et al.
Lipotoxicity causes multisystem organ failure and exacerbates acute pancreatitis
in obesity. Science translational medicine. 2011 Nov 2;3(107):107ra10. PubMed
PMID: 22049070. Pubmed Central PMCID: 3321362. Epub 2011/11/04. eng.
11.
Mathur A, Hernandez J, Shaheen F, Shroff M, Dahal S, Morton C, et al.
Preoperative computed tomography measurements of pancreatic steatosis and
visceral fat: prognostic markers for dissemination and lethality of pancreatic
adenocarcinoma. HPB : the official journal of the International Hepato Pancreato
Biliary Association. 2011 Jun;13(6):404-10. PubMed PMID: 21609373. Pubmed
Central PMCID: 3103097. Epub 2011/05/26. eng.

18

443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491

12.
Pitt HA. Hepato-pancreato-biliary fat: the good, the bad and the ugly. HPB :
the official journal of the International Hepato Pancreato Biliary Association.
2007;9(2):92-7. PubMed PMID: 18333122. Pubmed Central PMCID: 2020794.
Epub 2008/03/12. eng.
13.
Smits MM, van Geenen EJ. The clinical significance of pancreatic steatosis.
Nature reviews Gastroenterology & hepatology. 2011 Mar;8(3):169-77. PubMed
PMID: 21304475. Epub 2011/02/10. eng.
14.
Rosso E, Casnedi S, Pessaux P, Oussoultzoglou E, Panaro F, Mahfud M, et al.
The role of "fatty pancreas" and of BMI in the occurrence of pancreatic fistula after
pancreaticoduodenectomy. Journal of gastrointestinal surgery : official journal of
the Society for Surgery of the Alimentary Tract. 2009 Oct;13(10):1845-51.
PubMed PMID: 19639369. Epub 2009/07/30. eng.
15.
Yeo TP, Lowenfels AB. Demographics and epidemiology of pancreatic
cancer. Cancer J. 2012 Nov-Dec;18(6):477-84. PubMed PMID: 23187833. Epub
2012/11/29. eng.
16.
Lingvay I, Esser V, Legendre JL, Price AL, Wertz KM, Adams-Huet B, et al.
Noninvasive quantification of pancreatic fat in humans. The Journal of clinical
endocrinology and metabolism. 2009 Oct;94(10):4070-6. PubMed PMID:
19773401. Pubmed Central PMCID: 2758729. Epub 2009/09/24. eng.
17.
Gaborit B, Jacquier A, Kober F, Abdesselam I, Cuisset T, Boullu-Ciocca S, et
al. Effects of bariatric surgery on cardiac ectopic fat: lesser decrease in epicardial
fat compared to visceral fat loss and no change in myocardial triglyceride content.
Journal of the American College of Cardiology. 2012 Oct 9;60(15):1381-9. PubMed
PMID: 22939560. Epub 2012/09/04. eng.
18.
Tamura Y, Tanaka Y, Sato F, Choi JB, Watada H, Niwa M, et al. Effects of diet
and exercise on muscle and liver intracellular lipid contents and insulin
sensitivity in type 2 diabetic patients. The Journal of clinical endocrinology and
metabolism. 2005 Jun;90(6):3191-6. PubMed PMID: 15769987. Epub 2005/03/17.
eng.
19.
Colles SL, Dixon JB, Marks P, Strauss BJ, O'Brien PE. Preoperative weight
loss with a very-low-energy diet: quantitation of changes in liver and abdominal
fat by serial imaging. The American journal of clinical nutrition. 2006
Aug;84(2):304-11. PubMed PMID: 16895876. Epub 2006/08/10. eng.
20.
Lim EL, Hollingsworth KG, Aribisala BS, Chen MJ, Mathers JC, Taylor R.
Reversal of type 2 diabetes: normalisation of beta cell function in association with
decreased pancreas and liver triacylglycerol. Diabetologia. 2011 Oct;54(10):250614. PubMed PMID: 21656330. Pubmed Central PMCID: 3168743. Epub
2011/06/10. eng.
21.
Diagnosis and classification of diabetes mellitus. Diabetes care. 2013 Jan;36
Suppl 1:S67-74. PubMed PMID: 23264425. Pubmed Central PMCID: 3537273. Epub
2013/01/04. eng.
22.
Gaborit B, Kober F, Jacquier A, Moro PJ, Cuisset T, Boullu S, et al. Assessment
of epicardial fat volume and myocardial triglyceride content in severely obese
subjects: relationship to metabolic profile, cardiac function and visceral fat. Int J
Obes (Lond). 2012 Mar;36(3):422-30. PubMed PMID: 21730964.
23.
Jensen MD, Kanaley JA, Reed JE, Sheedy PF. Measurement of abdominal and
visceral fat with computed tomography and dual-energy x-ray absorptiometry.
The American journal of clinical nutrition. 1995 Feb;61(2):274-8. PubMed PMID:
7840063.

19

492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540

24.
Du Q, Martin JC, Agnani G, Pages N, Leruyet P, Carayon P, et al. Dairy fat
blends high in alpha-linolenic acid are superior to n-3 fatty-acid-enriched palm oil
blends for increasing DHA levels in the brains of young rats. The Journal of
nutritional biochemistry. 2012 Dec;23(12):1573-82. PubMed PMID: 22445803.
Epub 2012/03/27. eng.
25.
Trygg J, Holmes E, Lundstedt T. Chemometrics in metabonomics. Journal of
proteome research. 2007 Feb;6(2):469-79. PubMed PMID: 17269704.
26.
Cramer RD. Partial Least Squares (PLS): Its strengths and limitations
Perspectives in Drug Discovery and Design ed1993 December 1993.
27.
Eriksson K JE, Kettaneh-Wold N, Wikstrm C, Wold S. Multi- and
megavariate analysis. Part I: Basic principles and applications. Umea, Umetrics
AB. 2006:1-419.
28.
Chavance M. [Jackknife and bootstrap]. Revue d'epidemiologie et de sante
publique. 1992;40(3):209-18. PubMed PMID: 1439063. Epub 1992/01/01. Le
canif et les bretelles. fre.
29.
Pinnick KE, Collins SC, Londos C, Gauguier D, Clark A, Fielding BA.
Pancreatic ectopic fat is characterized by adipocyte infiltration and altered lipid
composition. Obesity (Silver Spring). 2008 Mar;16(3):522-30. PubMed PMID:
18239594. Epub 2008/02/02. eng.
30.
Lee Y, Lingvay I, Szczepaniak LS, Ravazzola M, Orci L, Unger RH. Pancreatic
steatosis: harbinger of type 2 diabetes in obese rodents. Int J Obes (Lond). 2010
Feb;34(2):396-400. PubMed PMID: 20010902. Epub 2009/12/17. eng.
31.
Mathur A, Marine M, Lu D, Swartz-Basile DA, Saxena R, Zyromski NJ, et al.
Nonalcoholic fatty pancreas disease. HPB : the official journal of the International
Hepato Pancreato Biliary Association. 2007;9(4):312-8. PubMed PMID: 18345311.
Pubmed Central PMCID: 2215403. Epub 2008/03/18. eng.
32.
Patel AV, Rodriguez C, Bernstein L, Chao A, Thun MJ, Calle EE. Obesity,
recreational physical activity, and risk of pancreatic cancer in a large U.S. Cohort.
Cancer epidemiology, biomarkers & prevention : a publication of the American
Association for Cancer Research, cosponsored by the American Society of
Preventive Oncology. 2005 Feb;14(2):459-66. PubMed PMID: 15734973. Epub
2005/03/01. eng.
33.
Tranchart H, Gaujoux S, Rebours V, Vullierme MP, Dokmak S, Levy P, et al.
Preoperative CT scan helps to predict the occurrence of severe pancreatic fistula
after pancreaticoduodenectomy. Annals of surgery. 2012 Jul;256(1):139-45.
PubMed PMID: 22609844. Epub 2012/05/23. eng.
34.
Sadr-Azodi O, Orsini N, Andren-Sandberg A, Wolk A. Abdominal and total
adiposity and the risk of acute pancreatitis: a population-based prospective
cohort study. The American journal of gastroenterology. 2013 Jan;108(1):133-9.
PubMed PMID: 23147519. Epub 2012/11/14. eng.
35.
Petrov MS. Editorial: abdominal fat: a key player in metabolic acute
pancreatitis. The American journal of gastroenterology. 2013 Jan;108(1):140-2.
PubMed PMID: 23287945. Epub 2013/01/05. eng.
36.
Rossi AP, Fantin F, Zamboni GA, Mazzali G, Rinaldi CA, Del Giglio M, et al.
Predictors of ectopic fat accumulation in liver and pancreas in obese men and
women. Obesity (Silver Spring). 2011 Sep;19(9):1747-54. PubMed PMID:
21593811. Epub 2011/05/20. eng.
37.
van Raalte DH, van der Zijl NJ, Diamant M. Pancreatic steatosis in humans:
cause or marker of lipotoxicity? Current opinion in clinical nutrition and

20

541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588

metabolic care. 2010 Jul;13(4):478-85. PubMed PMID: 20489606. Epub

2010/05/22. eng.
38.
Wong VW, Wong GL, Yeung DK, Abrigo JM, Kong AP, Chan RS, et al. Fatty
Pancreas, Insulin Resistance, and beta-Cell Function: A Population Study Using
Fat-Water Magnetic Resonance Imaging. The American journal of
gastroenterology. 2014 Apr;109(4):589-97. PubMed PMID: 24492753.
39.
Szczepaniak LS, Victor RG, Mathur R, Nelson MD, Szczepaniak EW, Tyer N, et
al. Pancreatic steatosis and its relationship to beta-cell dysfunction in humans:
racial and ethnic variations. Diabetes care. 2012 Nov;35(11):2377-83. PubMed
PMID: 22968187. Pubmed Central PMCID: 3476895. Epub 2012/09/13. eng.
40.
Ryu S, Chang Y, Kim SG, Cho J, Guallar E. Serum uric acid levels predict
incident nonalcoholic fatty liver disease in healthy Korean men. Metabolism:
clinical and experimental. 2011 Jun;60(6):860-6. PubMed PMID: 20863537. Epub
2010/09/25. eng.
41.
Lonardo A, Loria P, Leonardi F, Borsatti A, Neri P, Pulvirenti M, et al. Fasting
insulin and uric acid levels but not indices of iron metabolism are independent
predictors of non-alcoholic fatty liver disease. A case-control study. Digestive and
liver disease : official journal of the Italian Society of Gastroenterology and the
Italian Association for the Study of the Liver. 2002 Mar;34(3):204-11. PubMed
PMID: 11990393. Epub 2002/05/07. eng.
42.
Roncal-Jimenez CA, Lanaspa MA, Rivard CJ, Nakagawa T, Sanchez-Lozada LG,
Jalal D, et al. Sucrose induces fatty liver and pancreatic inflammation in male
breeder rats independent of excess energy intake. Metabolism: clinical and
experimental. 2011 Sep;60(9):1259-70. PubMed PMID: 21489572. Pubmed
Central PMCID: 3137694. Epub 2011/04/15. eng.
43.
Gotoh K, Inoue M, Shiraishi K, Masaki T, Chiba S, Mitsutomi K, et al. Spleenderived interleukin-10 downregulates the severity of high-fat diet-induced nonalcoholic fatty pancreas disease. PloS one. 2012;7(12):e53154. PubMed PMID:
23285260. Pubmed Central PMCID: 3532347. Epub 2013/01/04. eng.
44.
Nannipieri M, Baldi S, Mari A, Colligiani D, Guarino D, Camastra S, et al.
Roux-en-Y gastric bypass and sleeve gastrectomy: mechanisms of diabetes
remission and role of gut hormones. The Journal of clinical endocrinology and
metabolism. 2013 Nov;98(11):4391-9. PubMed PMID: 24057293.
45.
Jackness C, Karmally W, Febres G, Conwell IM, Ahmed L, Bessler M, et al.
Very low-calorie diet mimics the early beneficial effect of Roux-en-Y gastric
bypass on insulin sensitivity and beta-cell Function in type 2 diabetic patients.
Diabetes. 2013 Sep;62(9):3027-32. PubMed PMID: 23610060. Pubmed Central
PMCID: 3749351.
46.
Snel M, Jonker JT, Schoones J, Lamb H, de Roos A, Pijl H, et al. Ectopic fat and
insulin resistance: pathophysiology and effect of diet and lifestyle interventions.
International journal of endocrinology. 2012;2012:983814. PubMed PMID:
22675355. Pubmed Central PMCID: 3366269. Epub 2012/06/08. eng.
47.
Khalaf KI, Taegtmeyer H. Slimming the heart with bariatric surgery. Journal
of the American College of Cardiology. 2013 Mar 5;61(9):990-1. PubMed PMID:
23449436. Epub 2013/03/02. eng.
48.
Adams TD, Hunt SC. Cancer and obesity: effect of bariatric surgery. World
journal of surgery. 2009 Oct;33(10):2028-33. PubMed PMID: 19672652. Epub
2009/08/13. eng.

21

589
590
591
592
593
594
595

49.
Sjostrom L, Gummesson A, Sjostrom CD, Narbro K, Peltonen M, Wedel H, et
al. Effects of bariatric surgery on cancer incidence in obese patients in Sweden
(Swedish Obese Subjects Study): a prospective, controlled intervention trial. The
lancet oncology. 2009 Jul;10(7):653-62. PubMed PMID: 19556163. Epub
2009/06/27. eng.

596

22

597

Figure legends

598
599

Figure 1. Assessment of pancreatic triglyceride content (PTGC) and hepatic

600

triglyceride content (PTGC) using proton magnetic resonance spectroscopy 1H-

601

MRS. Axial, coronal and sagittal images illustrate typical selection of the voxel in the

602

pancreas and in the liver; example of spectra obtained with water peak at 4.7 ppm and

603

triglyceride peak at 1.3 ppm.

604

Figure 2. Comparison of pancreatic fat and hepatic fat in lean, obese and type 2

605

diabetic patients.

606

Figure 3. Change in BMI, ectopic fat stores (HTGC and PTGC) after bariatric

607

surgery and comparison to lean subjects

608
609

Supplementary figures

610

Figure S1. Variable importance of projection (VIP) statistical analysis taking into

611

account all lipidomic, anthropometrical and metabolic parameters

23

Table 1. Comparison of baseline characteristics between lean (n=13), obese non

type 2 diabetics (n=13), and obese type 2 diabetic patients (n=19).
Lean

Obese non-T2D

Obese T2D

13

13

19

Age (years)

42 3

34 2

51 2 *

Gender (Men / Women)

5/8

6/7

10 / 9

BMI (kg/m)

21.7 0.6

44.1 1.5

36.6 1.7 *

Waist circumference (cm)

79 2

119 4

118 5 *

VAT (cm)

69 33

164 28

236 21 *

SAT (cm)

154 36

708 37

439 41 *

Fasting plasma glucose (mmol/L)

5.0 0.1

5.1 01

7.4 0.5 *

Fasting plasma insulin (mUI/L)

6.3 0.6

18.5 2.5

16.1 2.7 *

HDL cholesterol (mmol/L)

0.62 0.05

0.48 0.03

0.44 0.01 *

LDL cholesterol (mmol/L)

1.35 0.2

1.17 0.08

0.97 0.09 *

Triglycerides (mmol/L)

0.75 0.1

1.07 0.1

1.45 0.2 *

Uric acid (mol/L)

249 15

303 20

353 14 *

hs CRP (mg/L)

0.93 0.14

13.44 3.15

7.71 1.56 *

PAI-1 (UI/L)

82

30 7

26 5 *

HOMA IR

1.39 0.16

4.26 0.60

5.60 1.05 *$

ALT (UI/L)

16 1

28 3

38 3 *

GGT (UI/L)

16 3

25 2

46 7 *

Fasting adiponectin (g/L)

11.65 1.33

5.08 0.63

4.34 0.57 *

Fasting leptin (ng/L)

11.3 2.7

127.1 13.8

45.8 9.2 *

T2D type 2 diabetes ; BMI body mass index ; VAT visceral abdominal fat ; SAT
subcutaneous abdominal fat ; PAI-1 Plasminogen activator inhibitor-1 ; HOMA-IR
Homeostasis Model Assessment of Insulin Resistance
p<0.05 between lean and obese non T2D ; *p<0.05 between lean and obese T2D ;
p<0.05 between obese non T2D and obese T2D
$ HOMA-IR was calculated in T2D after exclusion of patients treated with
sulfonylureas

Table 2. Step by step multivariate regression analysis using pancreatic triglyceride content
(PTGC) as the dependent variable and anthropometrical or biological variables as
independent variables.
Variables entered in the

Total R2

coefficient

Sig.

0.55

0.12

0.80

0.42

Gender

0.01

0.10

0.92

Diabetes

0.47

3.01

0.004

0.23

1.45

0.16

Log triglycerides

0.29

1.82

0.08

Log HOMA-R

-0.02

-0.10

0.92

Diabetes

0.32

2.17

0.04

-0.08

-0.51

0.61

0.45

2.8

0.007

0.17

1.07

0.29

0.44

2.72

0.009

models
Model 1
Age

Model 2
Waist circumference

0.65

Model 3
Log adiponectin

0.24

Diabetes
Model 4
Log HTGC
Diabetes

0.56

Table 3. Characteristics of morbid obese subjects before and 6 months after bariatric
surgery (n=20)
Baseline

6 Months

BMI (kg/m)

44.4 1.0

33.8 1.0

<0.0001

WC (cm)

126 3

103 3

<0.0001

VAT (cm)

211 28

119 12

0.002

SAT (cm)

661 30

455 24

<0.0001

Total cholesterol (mmol/L)

4.73 0.17

4.4 0.32

NS

LDL-cholesterol (mmol/L)

3.00 0.18

2.66 0.22

NS

HDL-cholesterol (mmol/L)

1.24 0.06

1.17 0.09

NS

Triglycerides (mmol/L)

1.35 0.12

1.16 0.10

NS

FFA (mmol/L)

0.58 0.06

0.57 0.06

NS

Fasting plasma glucose (mmol/L)

5.9 0.5

4.6 0.2

0.004

Fasting plasma insulin (mUI/L)

20.64 2.7

6.7 1.24

0.0006

Fasting adiponectin (g/L)

4.6 0.6

6.48 0.88

0.0012

Fasting leptin (ng/L)

111.2 10.4

39.4 4.7

<0.0001

HbA1c (%)

6.4 0.3

5.6 0.1

0.0052

HOMA-IR

5.7 0.9

1.3 0.3

0.0009

HOMA-B

207.8 28.5

144.2 25.8

NS

Fat distribution

Lipids profile

Glucose tolerance

Inflammation biomarkers
PAI-1 (UI/L)

31.7 5.25

10.47 2.27

0.0029

hs CRP (mg/L)

12.2 2.2

5.8 1.8

0.0075

ASAT (UI/L)

25.5 1.8

21.1 1.2

NS

ALAT (UI/L)

35.2 3.0

26.6 1.7

0.01

GGT (UI/L)

45.5 6.7

17.5 2.1

0.0002

330.0 15.6

294.8 15.4

0.07

Liver enzymes

Cardiovascular biomarker
Uric Acid (mol/L)