High dietary carbohydrates decreasegallbladder volume and enhancecholesterol crystal formationAbhishek Mathur, MD, Marine Megan, BS, Hayder H. Al-Azzawi, MD, Debao Lu, MD,Deborah A. Swartz-Basile, PhD, Attila Nakeeb, MD, and Henry A. Pitt, MD, Indianapolis, Ind

Background. Animal and human data suggest that a diet high in refined carbohydrates leads togallstone formation. However, no data are available on the role of dietary carbohydrates on gallbladdervolume or on cholesterol crystal formation. Therefore, we tested the hypothesis that a high carbohydratediet would alter gallbladder volume and enhance cholesterol crystal formation.Methods. At 8 weeks of age, 60 lean and 36 obese leptin-deficient female mice were fed a 45%carbohydrate diet while an equal number of lean and obese mice were fed a 75% carbohydrate diet for4 weeks. All animals then underwent cholecystectomy, and gallbladder bile volume was recorded. Bilewas pooled, filtered, and maintained in a water bath at 37°C for 14 days. Birefringent cholesterolcrystals in bile were counted daily; crystal observation time and crystal mass were determined.Results. The crystal observation time was significantly shortened in both lean and obese mice on the75% diet compared with their counterparts on the 45% diet. The crystal mass was significantlyincreased in the lean mice on the 75% diet compared with the 45% diet. Gallbladder volumes weresignificantly reduced in both lean and obese mice on the 75% diet compared with their counterparts onthe 45% diet.Conclusions. These data suggest that a high carbohydrate diet decreases gallbladder volume, shortenscholesterol crystal observation time, and increases crystal mass. We conclude that dietary carbohydratesmay play a role in cholesterol gallstone formation by altering biliary motility and by enhancing crystalformation. (Surgery 2007;141:654-9.)

From the Department of Surgery, Indiana University School of Medicine, Indianapolis

Obesity has become an epidemic in the UnitedStates with 40% of Americans being overweight (bodymass index [BMI] �25) and/or obese (BMI �30).1

Obesity has been implicated as a risk factor ingallstone pathogenesis.2,3 Gallstone prevalence cor-relates positively with BMI and increases dramati-cally in the morbidly obese.3 The prevalence ofdiabetes also is on the rise; the number of Ameri-cans with diabetes has increased 61% since 1990,and diabetes has become the sixth leading cause of

Presented at the American Hepato-Pancreato-Biliary Associa-tion, March 9-12, 2006, Miami, Florida.

Supported in part by a research grant from the National Insti-tutes of Health (R-01 DK44279).

Accepted for publication November 3, 2006.

Reprint requests: Henry A. Pitt, MD, Department of Surgery,Indiana University School of Medicine, 535 Barnhill Drive, RT130D, Indianapolis, IN 46202. E-mail: hapitt@iupui.edu.

0039-6060/$ - see front matter

© 2007 Mosby, Inc. All rights reserved.

doi:10.1016/j.surg.2006.11.008

654 SURGERY

death in the United States.1,4 Diabetes is a riskfactor for the development of cholesterol gallstonedisease, and 30% of adults with diabetes developcholelithiasis.5 Insulin-resistant diabetes also corre-lates strongly with obesity; murine models of diabe-tes have demonstrated that resting gallbladdervolume is increased and gallbladder contraction isimpaired.6,7 Data gathered from human subjectshave corroborated these findings by showing thathumans with diabetes have enlarged resting gall-bladders with decreased gallbladder emptying.8

Cholesterol gallstone disease occurs as the resultof interactions between environmental and geneticfactors. Diet is an important environmental factorthat is a predisposition to cholesterol gallstone for-mation.9 Osler10 first postulated in 1892 that sweetfoods might play a role in gallstone development.More recently, a number of animal studies have dem-onstrated the role of dietary fats in gallstone forma-tion.11,12,13 Diets rich in cholesterol have beenshown to produce gallstones in various animalmodels.13,14 Moreover, the type of dietary fat plays

a role—saturated fats enhance and fish oils protect

Surgery Mathur et al 655Volume 141, Number 5

against gallstone formation in both animal modelsand human studies.11-13,15 The role of refined car-bohydrates in the diet were initially studied by Damand Christensen.16,17 They demonstrated that a re-fined carbohydrate diet produced gallstones in80% to 100% of the hamsters and mice used intheir study.16,17pve Research also has establishedthat refined sugars increase the biliary cholesterolsaturation index in humans.10 Further support fortheir role in gallstone formation includes the re-sults of a recent epidemiologic study of 70,000nurses, which documented that a high carbohy-drate diet is associated with an increased cholecys-tectomy rate.18 No data, however, have yet beengathered on the role of dietary carbohydrates ongallbladder volume or on cholesterol crystal forma-tion. Therefore, we tested the hypothesis that ahigh carbohydrate diet would alter gallbladder vol-ume and enhance cholesterol crystal formation.

MATERIAL AND METHODSAnimals and diets. A total of 120 lean control

(C57BL/6J) and 72 obese leptin-deficient (Lepob)female mice were obtained from Jackson Labora-tory (Bar Harbor, Me). The mice were housed 5per cage under controlled conditions (light, 6 am to6 pm; temperature, 22°C). During 1 week of envi-ronmental adjustment, the mice were fed a stan-dard low-cholesterol chow diet (Ralston Purina, StLouis, Mo). At 8 weeks of age, 60 lean C57BL/6Jmice and 36 Lepob female mice were fed a lowcarbohydrate diet (45% carbohydrate, 15% fat, and40% protein) (Dyets Inc, Bethlehem, Pa) while theremaining 60 lean and 36 obese mice were fed ahigh carbohydrate diet (75% carbohydrate, 15%fat, and 10% protein) (Dyets Inc) for 4 weeks. Thecarbohydrates in the 45% diet were 35% sucroseand 10% cornstarch. The 75% carbohydrate diethad 65% sucrose and 10% cornstarch. In bothdiets, the fat was anhydrous milk fat, and the pro-tein was exclusively casein. Both the animals andthe food were weighed weekly to determine growthand dietary intake. All protocols for these animalstudies were approved by the Indiana UniversityInstitutional Animal Care and Use Committee.

Bile and serum collection. At 12 weeks of age,after an overnight fast with water allowed ad libi-tum, the mice were anesthetized with an isoflu-rane–soaked gauze placed in a 2000 cm3 glass jar.They then received an intraperitoneal injection ofxylazine (15 mg/kg) and ketamine (50 mg/kg).The animals were weighed, and then underwentlaparotomy and cholecystectomy. Gallbladders wereplaced into individual microtubes and slit for bile

collection. Microtubes were centrifuged for 5 min-

utes at 15,000 rpm; gallbladders were removed; andbile was measured with a micropipette. Wholeblood was aspirated from the heart and centrifugedto isolate serum.

Crystal analysis. Bile was pooled into 100 �laliquots and microfiltered (Ultrafree sterile0.22-�m filter; Millipore Corp, Billerica, Mass) at15,000 rpm for 10 minutes. Three bile pools wereobtained for lean mice on the 45% and 75% car-bohydrate diets, as well as for obese mice on the75% carbohydrate diet. Five bile pools were ob-tained for obese mice on the 45% carbohydratediet. Pooled bile was incubated in a water bath at37°C for 14 days. Birefringent Maltese cross liquidand solid cholesterol monohydrate crystals werecounted in 10 high power fields by an investigatorwho was blinded as to group. Crystal observationtime (COT) in days, crystal growth rate (slope),and crystal mass (total number) were calculated aspreviously described.19

Serum analysis. Whole blood was spun at 15,000rpm for 5 minutes to separate serum. Serum waspooled to give 6 pools each for lean mice on the45% and 75% carbohydrate diets, as well as obesemice on the 75% carbohydrate diet. Five poolswere obtained for obese mice on the 45% carbohy-drate diet. Serum glucose was determined by thequantitative colorimetric method using the Glu-cose Liquicolor Kit (StanBio Lab, Boerne, Tex).Serum sodium was determined by indirect potenti-ometry using 2 glass sodium electrodes in the Syn-chron LX System (Beckman Coulter, Fullerton,Calif). Serum osmolality was measured by freezingpoint depression using the Advanced Micro Os-mometer, model 3300 (Advanced Instruments, Inc,Norwood, Mass).

Statistical analysis. Statistical analyses were per-formed using Sigma Stat statistical software (JandelCorp, San Rafael, Calif). All data are expressed asmean � standard error of the mean (SEM). Differ-ences in animal body weight, food intake, gallblad-der volume, biliary crystal data, and serum analysesdata were tested for statistical significance by anal-ysis of variance (ANOVA) and by the Studentunpaired t test. A P value of less than .05 wasconsidered statistically significant.

RESULTSAnimal weights and dietary intake. Data for

weekly animal weights and dietary intake are pre-sented in Figs 1, A and B. At 8 weeks of age, nosignificant difference in weight was observed be-tween the 2 groups of lean mice or the 2 groups ofobese mice. Subsequently, a significant decrease in

weight occurred in obese mice on the 75% carbo-

656 Mathur et al SurgeryMay 2007

hydrate diet compared with those on the 45% car-bohydrate diet in weeks 2, 3, and 4. No difference inweight existed between the 2 groups of lean mice atany time point. As expected, the obese mice weighedsignificantly more than the lean mice at all timepoints. The dietary intake was significantly lower inthe obese mice on the 75% compared to the 45%carbohydrate diet in weeks 2 and 3; there was asimilar trend in week 4. No difference in dietaryintake of the 2 groups of lean mice existed at anytime point. Dietary intake of the obese mice on the45% carbohydrate diet was significantly greaterthan the lean mice. Over the 4 weeks, the averagecarbohydrate consumption per 100 grams bodyweight for the lean and obese mice on the 75%CHO diet was 68 and 33, respectively. In compari-son, the mean carbohydrate consumptions were 40and 21, respectively, for the lean and obese mice on

Fig 1. A, Weekly animal weights for both lean and obesemice on the 45% and 75% carbohydrate diets. B, Weeklydietary intake for both lean and obese mice on the 45%and 75% carbohydrate diets.

the 45% CHO diet, which was significantly lower

(P � .01) than the corresponding mice on the 75%CHO diet.

In vitro biliary crystal observation. In vitro crys-tal results are presented in Figs 2, A and B. Forliquid crystals, the COT was significantly shortenedon the high carbohydrate diet in both lean andobese mice compared with their counterparts onthe low carbohydrate diet. The crystal mass wassignificantly greater in the lean mice on the highcarbohydrate versus the low carbohydrate diet. Nodifference in crystal growth was observed betweenthe 2 groups of lean or the 2 groups of obese mice.No strain differences were observed for any of the 3nucleation parameters, and no solid crystals wereseen in any of the 4 groups.

Gallbladder volume. Gallbladder volumes areshown in Fig 3. Gallbladder volumes were signifi-cantly reduced in both lean and obese mice onthe high carbohydrate diet compared with their

Fig 2. A, Mean crystal observation time for liquid crystalsin both lean and obese mice on the 2 diets. B, Crystalmass for liquid crystals over 14 days in both lean andobese mice on 2 two diets.

counterparts on the low carbohydrate diet. The

Surgery Mathur et al 657Volume 141, Number 5

gallbladder volumes were significantly increased inthe obese mice versus their lean counterparts onthe 45% carbohydrate diet. A similar strain differ-ence was not found on the 75% carbohydrate diet.

Serum data. Serum glucose, sodium, and osmo-larity are presented in the Table. Serum glucosewas decreased in the lean mice on the 75% carbo-hydrate diet compared with the 45% carbohydratediet (98 � 9 vs 129 � 14), but this difference wasnot statistically significant (P � .28). The glucosealso was decreased in the obese mice on the 75%versus the 45% carbohydrate diet (407 � 30 vs250 � 23), which was statistically significant (P �.01). As expected, a strain difference was also seen.The serum glucose was significantly higher in theobese mice on the 45% and 75% carbohydratediets compared with the corresponding groups oflean mice (P � .001). No strain or diet differencesin the serum sodium or osmolality were observed.The mean serum sodium was 150 mM/l, and themean serum osmolality was 341 mOsm/kg.

DISCUSSIONIn this study, 60 lean C57BL/6J mice and 36

Lepob female mice were fed a low (45%) carbohy-drate diet and an equal number of lean and obesemice were fed a high (75%) carbohydrate diet for4 weeks. The obese mice on the high carbohydratediet had a lower weekly dietary intake, gained lessweight, and had lower serum glucose levels thantheir counterparts on the 45% carbohydrate diet.The 75% carbohydrate diet enhanced the for-mation of liquid cholesterol crystal in both theobese and the lean mice by decreasing the COTand increasing the crystal mass. In addition, thegallbladder volumes were decreased in both leanand obese mice on the 75% carbohydrate diet. No

Fig 3. GB, Gallbladder. GB volume for both lean andobese mice on the 2 diets.

differences were observed in the serum sodium or

osmolality in either the lean or the obese mice onthe diets.

Leptin-deficient obese mice serve as a murinemodel for human obesity and insulin-resistant dia-betes. Previous studies from our laboratory havedemonstrated that Lepob mice have hyperglycemia,insulin-resistance, hyperinsulinemia, and hyperlip-idemia.20 Lepob mice also have been shown to havea shortened COT and an increased crystal mass andgrowth rate compared with lean controls on a non-lithogenic 60% carbohydrate chow diet.19 The factthat obese mice on a chow diet had increasedcholesterol crystal formation compared with theirlean counterparts despite similar cholesterol satu-ration index can be explained by the finding ofpronucleating proteins in the bile that have beenshown to enable rapid crystallization.21,22 Subse-quent data from our lab have shown that obesemice have an increased 46-kd nonmucin proteinand 61-kd and 84-kd nonmucin glycoproteins intheir bile.21 Further studies identified carboxyles-terase as the pronucleator in leptin-deficientmice.22 Because this present study showed no dif-ferences in crystal formation between obese miceand lean controls, one possibility is that the dietshigh in sucrose directly or indirectly altered thepronucleating factors in bile.

Diet is a well-established risk factor for gallstoneformation.18 Using data from human subjects, Tsaiet al9 suggested that a diet high in carbohydratesincreases the risk of symptomatic gallstone disease.Furthermore, their study demonstrated that su-crose, fructose, and starch have a significant posi-tive correlation with the increased gallstone risk.9

Animal data also support the role of dietary carbo-hydrates in gallstone formation.16,17 More compel-ling evidence that specific types of carbohydrateenhance gallstone formation comes from Thorn-ton et al10pve who demonstrated that human sub-jects eating a diet high in refined sugars had asignificantly higher biliary cholesterol saturationindex (CSI) compared with subjects on unrefined

Table. Serum data

Study groupGlucose(mg/dl)

Sodium(mM/l)

Osmolarity(mOsm/kg)

Lean 45% CHO 129 � 14 149 � 2 341 � 5Lean 75% CHO 98 � 10 152 � 1 332 � 1Obese 45% CHO 407 � 30* 149 � 1 341 � 3Obese 75% CHO 250 � 23*† 150 � 1 348 � 5

CHO, Carbohydrate.*P � .001 vs lean.†P � .01 vs obese 45%.

carbohydrates. For these reasons, we used diets

658 Mathur et al SurgeryMay 2007

with sucrose as the predominant sugar in thepresent study.

Our findings further elucidate the mechanismby which high dietary carbohydrates increase cho-lesterol crystal formation, namely by causing morerapid crystallization and also by increasing the crys-tal mass in the lean mice. The crystal mass in obesemice increased as well but did not reach statisticalsignificance, possibly because of the small numberof pools and a large standard error. The fact thatsolid crystals were not observed in this study isconsistent with the short time on the diet and thelow CSI observed in mice fed a low-cholesteroldiet.20 Further studies will be required to deter-mine which glycoproteins in bile are driving theenhanced liquid crystal formation. Dietary caseinalso has been thought to increase the cholesterolsaturation of bile and increase gallstone forma-tion.23 However, the decreased cholesterol crystalnucleation on the low 45% carbohydrate diet, de-spite the mice having increased casein contents intheir diet, reinforces the importance of dietary car-bohydrates in gallstone development.

Another interesting finding was that the obesemice ate less and weighed less on a high carbohy-drate diet, both of which contributed to lower theserum glucose compared with their counterpartson a low carbohydrate diet. This effect of diet onthe amount of weekly dietary intake can be ex-plained by dietary effects on neuropeptide Y (NPY)and leptin. NPY is a known potent stimulator offood intake and has been shown to preferentiallyenhance carbohydrate intake.24 Beck et al25 dem-onstrated that rats on a high carbohydrate diet hada decrease in NPY levels in the brain, lower bodyweight, and lower plasma glucose levels comparedwith rats on a well-balanced or high-fat diet.25 Thesefindings are similar to ours; and so we can postulatethat a decrease in NPY occurred on the high car-bohydrate diet in the present study.

Leptin is a hormone that regulates dietary intakeand body weight, among other effects.26 Robertset al27 have shown that a diet high in sucrose canincrease serum leptin and decrease food intake inrats. This possibility is a second explanation for ourresults. However, the low-carbohydrate diet that weused had more protein content than the high-carbohydrate diet, and an increase in protein con-tent in the diet has been shown to increase satietyand decrease weight in healthy humans.28 This ef-fect was accompanied by a decrease in leptin. In-terestingly, the high-carbohydrate, low-protein dietdid not affect food intake or weight in the leanmice with normal leptin metabolism but did so only

in the Lepob mice. This fact suggests that our ob-

servations are more likely due to alterations in NPYthan to leptin.

Another striking finding was that resting gallblad-der volumes significantly decreased in both lean andobese mice on the high-carbohydrate diet. The in-crease in gallbladder volume in obese diabetic miceon the low-carbohydrate diet was associated withvery high serum glucose levels. This finding sug-gests that the degree of hyperglycemia and insulinresistance in obese diabetic mice may affect hepaticbile flow, fasting gallbladder cyclic motor activity,gallbladder absorption, or sphincter of Oddi resis-tance, resulting in larger resting gallbladder vol-umes in diabetic mice with higher serum glucose.

Patients with diabetes have been shown to haveincreased resting gallbladder volumes. However,the degree of hyperglycemia in patients with diabeteshas not previously been shown to affect resting vol-ume. Spontaneous partial gallbladder emptying,which occurs during the fasting period, is known tobe under the control of the vagal-cholinergic sys-tem and is not affected by serum levels of cholecys-tokinin.29,30 Moreover, fasting, healthy humansubjects have been found to have fewer spontane-ous gallbladder-emptying periods and a greater de-gree of partitioning of liver bile to the gallbladderduring hyperglycemia.31

Both hepatic bile flow and gallbladder absorptionare affected by glucose levels.32 In isolated guinea piglivers, infusion of glucose decreased hepatic bile flowand increased bile sodium.32 This alteration of bilesodium might result in a decrease in gallbladderwater absorption, which would result in an increasein resting gallbladder volume. Additionally, datafrom our lab have documented that gallbladders ofLepob mice have decreased response to neurotrans-mitters such as acetylcholine, neuropeptide Y,and cholecystokinin.7 These obese, diabetic micehad increased resting gallbladder volumes com-pared with lean controls.33 Moreover, serum glu-cose and insulin correlated negatively with in vitrogallbladder motility.20 Thus, our observation withrespect to gallbladder volume may be due, in part,to the serum glucose levels. However, further stud-ies of hepatic bile flow, gallbladder absorption, andsphincter of Oddi and gallbladder motility will berequired to fully explain this observation.

In conclusion, in lean and obese mice a high-carbohydrate diet, containing primarily sucrose,decreases serum glucose and gallbladder volume,shortens cholesterol COT, and increases crystalmass. These alterations would be expected to en-hance gallstone formation, which is consistent withhuman data. Dietary carbohydrates thus may play a

role in cholesterol gallstone formation both by alter-

Surgery Mathur et al 659Volume 141, Number 5

ing biliary motility and by enhancing cholesterol crys-tal formation.

REFERENCES1. Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F,

Bales VS, et al. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA 2003;289:76-9.

2. Everhart JE, Khare M, Hill M, Maurer KR. Prevalence andethnic differences in gallbladder disease in the UnitedStates. Gastroenterology 1999;117:632-9.

3. Must A, Spadano J, Coakley EH, Field AE, Colditz G, DietzWH. The disease burden associated with overweight andobesity. JAMA 1999;282:1523-9.

4. Centers for Disease Control and Prevention. National dia-betes fact sheet: General information and national estimateson diabetes in the United States, 2003. Atlanta (GA): 2003,US Dept. of Health and Human Services, Centers for Dis-ease Control and Prevention; 2003.

5. Kayacetin E, Kisakol G, Kaya A, Akpinar Z. Real-time sonog-raphy for screening of gallbladder motility in diabetic pa-tients. Relation to autonomic and peripheral neuropathy.Neuroendocrinol Lett 2003;24:73-6.

6. Tran KQ, Swartz-Basile DA, Nakeeb A, Pitt HA. Gallbladdermotility in agouti-yellow and leptin-resistant obese mice.J Surg Res 2003;113:56-61.

7. Goldblatt MI, Swartz-Basile DA, Svatek CL, Nakeeb A, PittHA. Decreased gallbladder response in leptin-deficientobese mice. J Gastrointest Surg 2002;6:438-42.

8. Hahm JS, Park JY, Park KG, Ahn YH, Lee MH, Park KN.Gallbladder motility in diabetes mellitus using real timeultrasonography. Am J Gastroenterol 1996;91:2391-4.

9. Tsai CJ, Leitzmann MF, Willett WC, Giovannucci EL. Di-etary carbohydrates and glycaemic load and the incidenceof symptomatic gall stone disease in men. Gut 2005;54:823-8.

10. Thornton JR, Emmett PM, Heaton KW. Diet and gallstones.Effects of refined and unrefined carbohydrate diets on bilecholesterol saturation and bile acid metabolism. Gut 1983;24:2-6.

11. Dam H, Prange I, Christensen F. Alimentary production ofgallstones in hamsters. 17. Influence of butter fat and the fatof a dietetic margarine rich in linoleic acid on gallstone for-mation and composition of the bladder bile. Z Ernahrungswiss1965;6:97-106.

12. Dam H, Kruse I, Krogh Jensen M, Kallehauge HE. Studieson human bile. II. Influence of two different fats on thecomposition of human bile. Scand J Clin Lab Invest1967;19:367-78.

13. Jonnalagadda SS, Trautwein EA, Hayes KC. Dietary fats richin saturated fatty acids (12:0, 14:0, and 16:0) enhance gall-stone formation relative to monounstaturated fat (18:1) incholesterol-fed hamsters. Lipids 1995;30:415-24.

14. Tepperman J, Caldwell FT, Tepperman HM. Induction ofgallstones in mice by feeding a cholesterol-cholic acid con-taining diet. Am J Physiol 1964;206:628-34.

15. Magnuson TH, Lillemoe KD, High RC, Pitt HA. Fish oilprevents cholesterol crystal nucleation and gallstone forma-tion. Gastroenterology 1989;96:625.

16. Dam H, Christensen F. Alimentary production of gallstonesin hamsters: Influence of different carbohydrate sources ongallstone formation, diarrhea, and growth. Z Ernahrung-

swiss 1961;2:91-102.

17. Dam H, Christensen F, Prange I. The relationship betweendiet and composition of bladder bile in mice. Z Ernahrung-swiss 1969;9:200-8.

18. Tsai CJ, Leitzmann MF, Willett WC, Giovannucci EL. Glyce-mic load, glycemic index, and carbohydrate intake in rela-tion to risk of cholecystectomy in women. Gastroenterology2005;129:105-12.

19. Goldblatt MI, Swartz-Basile DA, Choi S-H, Nakeeb A, SarnaS, Pitt HA. Cholesterol crystal formation in congenitallyobese mice. Surg Forum 2000;51:1-2.

20. Tran KQ, Goldblatt MI, Swartz-Basile DA, Svatek C, NakeebA, Pitt HA. Diabetes and hyperlipidemia correlate with gall-bladder contractility in leptin-related murine obesity. J Gas-trointest Surg 2003;7:857-62.

21. Goldblatt MI, Swartz-Basile D, Svatek C, Nakeeb A, Pitt HA.Increased 46, 61, and 84 kDa gallbladder bile nonmucin pro-teins in genetically obese mice. Surgical Forum 2001;52:36-7.

22. Tran KQ, Goldblatt MI, Swartz-Basile DA, Svatek CL, Na-keeb A, et al. Carboxylesterase is a cholesterol pronucleatorin leptin deficient obese mice. J Amer Coll Surg 2002;195:13.

23. Ozben T. Biliary lipid composition and gallstone formationin rabbits fed on soy protein, cholesterol, casein and mod-ified caesin. Biochem J 1989;263:293-6.

24. Stanley BG, Daniel DR, Chin AS, Leibowitz SF. Paraventricu-lar nucleus injections of peptide YY and neuropeptide Ypreferentially enhance carbohydrate ingestion. Peptides1985;6:1205-11.

25. Beck B, Stricker-Krongrad A, Burlet A, Max JP, Musse N,Nicolas JP, Burlet C. Macronutrient type independently ofenergy intake modulates hypothalamic neuropeptide Y inLong-Evans rats. Brain Res Bull 1994;34:85-91.

26. Campfield LA, Smith FJ, Guisez Y, Devos R, Burn P. Recom-binant mouse OB protein: evidence for a peripheral signallinking adiposity and central neural networks. Science1995;269:546-9.

27. Roerts CK, Berger JJ, Barnard RJ. Long-term effects of dieton leptin, energy intake, and activity in a model of diet-induced obesity. J Appl Physiol 2002;93:887-93.

28. Weigle DS, Breen PA, Matthys CC, Callahan HS, MeeuwsKE, Burden VR, et al. A high-protein diet induces sustainedreductions in appetite, ad libitum caloric intake, and bodyweight despite compensatory changes in diurnal plasmaleptin and ghrelin concentrations. Am J Clin Nutr 2005;82:41-8.

29. Abiru H, Matsumoto T, Mochinaga N, Ura K, Tsunoda T,Tsuchiya R, Ozeki K, et al. Bilateral truncal vagotomy abol-ishes gallbladder cyclic motor activity. Gastroenterology1990;99:1221.

30. Qvist N, Oster-Jorgensen E, Rasmussen L, Pedersen SA,Olsen O, Cantor P, et al. Cholecystokinin, secretin, pancre-atic polypeptide in relation to gallbladder dynamics andgastrointestinal interdigestive motility. Digestion 1990;45:130-7.

31. Oster-Jorgensen E, Qvist N, Pedersen SA, Rasmussen L,Hovendal CP. The influence of induced hyperglycaemia onthe characteristics of intestinal motility and bile kinetics inhealthy men. Scand J Gastroenterol 1992;27:285-8.

32. Rutishauser SC, Millward SE. The effects of hypertonic glu-cose solutions on hepatic bile formation. J Hepatol 1990;11:22-8.

33. Graewin SJ, Lee KH, Kiely JM, Svatek CL, Nakeeb A, Pitt HA.Gallbladder myocytes are short and cholecystokinin-resis-

tant in obese diabetic mice. Surgery 2004;136:431-6.