Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanism of weight loss in the morbidly obese following ileogastrostomy and validation of reported energy… Su, Wanfang 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1995-060683.pdf [ 3.1MB ]
JSON: 831-1.0088354.json
JSON-LD: 831-1.0088354-ld.json
RDF/XML (Pretty): 831-1.0088354-rdf.xml
RDF/JSON: 831-1.0088354-rdf.json
Turtle: 831-1.0088354-turtle.txt
N-Triples: 831-1.0088354-rdf-ntriples.txt
Original Record: 831-1.0088354-source.json
Full Text

Full Text

MECHANISM OF WEIGHT LOSS IN THE MORBIDLY OBESE FOLLOWINGILEOGASTROSTOMY AND VALIDATION OF REPORTED ENERGYINTAKE IN NORMAL-WEIGHT AND MORBIDLY OBESE SUBJECTSbyWANFANG SUM.D., SHANXI MEDICAL COLLEGE, CHINA, 1983A THESIS SUBMITTED IN PARTIAL FULFILLMENTOF THE REQUIREMENTS FOR THE DEGREEOF DOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDIVISION OF HUMAN NUTRITIONSCHOOL OF FAMILY AND NUTRITIONAL SCIENCESWe accept this thesi nforming to the required standard______ITHE UNIVERSITY OF BRITISH COLUMBIAJUNE 1995(C)W. Su 1995In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of - -The University of British ColumbiaVancouver, CanadaDate OCb--( Ic?DE-6 (2/88)ABSTRACT ITo examine the mechanism ofweight loss following ileogastrostomy, 16 morbidly obesesubjects (36±2 years, 45.6±1.1 kg.m2’body mass index (BMI), 48.2±1.0% body fat(BF)(mean±SEM)) were selected and tested prior to and after this procedure. Due to variousreasons, complete data were not obtained from any one subject although a total of 16bypass patients participated in the study. Therefore, the number of subjects in each part ofthe study varied from six to ten. Body composition was determined using the isotopedilution space (IDS) method and bioelectrical impedanceanalysis (BIA), which werecompared with dual energy x-ray absorptiometry (DEXA) measurements in a subgroup ofthe participants. Gas exchange analysis was used to measure the changes in basal energyexpenditure (BEE) and thermic effect of food (TEF). Total energy expenditure (TEE) wasdetermined during 6-8 weeks after surgery using the doubly labeled water (DLW) method.Weighed food records were used to assess the changes in energy intake during the studyperiod. A group of normal-weight women (48±1 years, 23.4±0.5 kg.m2’BMI, 32.8±1.3%BF (mean±SEM)) was selected to supplement the overall research.Average body weight (111.1±3.2 kg, n=7) of the subjects completing the 3-monthmeasurement decreased by 5.2, 4.2 and 7.8 kg during each of the three months. There wasa significant decline (p<O.0001) in fat-free mass (FFM) and fat mass (FM) measured byboth IDS and BIA methods. The percentage of FFM and FM determined by IDS methodwas not significantly changed during the study period. Presurgical total body massII(125.6±4.5 kg) determined by DEXA was significantly different (p0.044) from that(130.3±6.3 kg) obtained by scale, but postsurgical data did not demonstrate thisdifference. The results raised an important question about the validity of DEXA in theassessment of body composition in the morbidly obese subjects. With the decrease of bodyweight induced by ileogastrostomy, body mass assessment by DEXA was not differentfrom that obtained from scale. There was a close agreement in fat mass (FM) andpercentage of body fat (%BF) obtained byIDS method and DEXA postsurgically,however, BIA showed a significant difference from the TBW method or DEXA (p<O.OS).Furthermore, the reduction of LBM and FM determined by IDS method and DEXA werefound to be smaller than those obtained by BIA.Ileogastrostomy did not significantly influence BEE levels but significantly affected TEF(p0.OO1). A very high percentage (44%) of energy expended for physical activity wasfound at the second month after surgery. Because TEE was not measured presurgery, wewere unable to assess the changes in TEE. However, our findings did show that TEE wasclosely correlated with the weight loss induced by ileogastrostomy (r0.719, p=O.0l9,n= 10). A almost significant relationship between weight loss and fecal energy (r0. 808,p=0.O52, n=6) but not urinary energy loss (r=0.01 1, pO.983, n=6) was observed in thisstudy. Surprisingly, energy intake as assessed by weighed food records was not related toweight loss during the short-term energy balance study. Energy intake was insignificantlycorrelated with energy expenditure (r=0 .628, p=O.l 68) and fecal energy (r=0. 732,p=O.O84).ifiResults of the validation of reported energy intake showed a large discrepancy betweenreported energy intake (El) and expenditure (EE) in both obese (1429±271 kcal.d’ El vs2933±239 kcal. d’ EE, respectively) and normal-weight groups (1653±76 kcal. d’ El vs2215±102 kcal. d’ EE, respectively). Underestimation, defined as {(EE-EI)IEEx 1001, was42.0% in the obese group and 20.5% in the normal-weight group after correcting for thechanges in body energy stores. The degree of underestimation was not associated withbody weight in the normal-weight group, however, a close relationship (r=0.868, p=O.O2S)between underestimation and body weight was observed in obese group.From these findings it is concluded that ileogastrostomy can significantly reduce bodyweight, reflected in the decline of LBM and FM. However, the percentage of LBM andFM during 3-month postsurgical measurements was not significantly different from thatpresurgery. Isotope dilution method and DEXA appeared to be accurate in assessing thereduction in body composition after intestinal bypass surgery although measurement oftotal LBM and FM compartments by the methods presently used did not agree very wellfor these morbidly obese subjects. Factorial energy expenditure results shQwed that BEEwas unchanged but TEF declined significantly. Total energy expenditure and fecal energyloss play very important roles in the weight loss following ileogastrostomy butenergy intake was not associated with this weight loss.TABLE OF CONTENTS IvABSTRACTTABLE OF CONTENTS IVLIST OF TABLES VIIILIST OF FIGURES XLIST OF ABBREVIATIONS XIACKNOWLEDGMENTS XIIISECTION1. INTRODUCTION 12. LITERATURE REVIEW 42.1. Implications, Prevalence and Definition of Obesity 42.2. Causes of Obesity 52.3. Treatment of Obesity 62.4. Mechanisms of Weight Loss Following Intestinal Bypass Surgery 92.4.1. Reduction in Food Intake 92.4.2. Malabsorption 132.4.3. Changes in Energy Expenditure 152.4.4. Changes in Hormonal Response and Nutrient Partitioning 172.5. Methodological Studies 192.5.1. Assessment of Body Composition Bioelectrical Impendance Analysis Isotope Dilution Space Method Dual Energy X-ray Absorptiometry 232.5.2. Measurement of Energy Expenditure 252.5.3. Food Intake Measurement 303. EXPERIMENTAL DESIGN AND METHODS 353.1. Examination of Mechanism of Weight Loss in Obese Subjects 38Following Ileogastrostomy3.1.1. Selection and Screening of Subjects 383.1.2. Measurement of Energy Intake 403.1.3. Measurement of Basal Energy Expenditure and Thermic Effect of Food 40V3.1.4. Body Composition Measurement 413.1.5. Measurement of Total Energy Expenditure 423.1.6. Measurement of Fecal and Urinary Energy Losses 433.2. Validation of Reported Energy Intake Using Doubly Labeled Water 44Technique in Normal-weight and Obese Subjects3.2.1. Selection and Screening of Subjects 443.2.2. Measurement of Energy Intake Using Weighed Food Records 443.2.3. Measurement of Total Daily Energy Expenditure Using Doubly 45Labeled Water Technique3.3. Analytical Procedures 473.3.1. Purification of Deuterium and Carbon Dioxide 473.3.2. Mass Spectrometric Determination 483.4. Data Calculation 493.4.1. Calculation of Total Body Water and Energy Expenditure 463.4.2. Estimation of Basal Energy Expenditure, Thermic Effect of 50Food and Substrate Utilization3.4.3. Evaluation of Postsurgical Energy Balance 513.5. Statistical Analyses 524. RESULTS 544.1. Examination of Mechanism of Weight Loss in Obese Subjects 54Following Ileogastrostomy4.1.1. Changes in Body Composition Following Ileogastrostomy 574.l.1.l.WeightLoss 574.1.1.2. Influence of Ileogastrostomy on Body Composition Measured 57by Isotope Dilution Method and Bioelectrical impedance Analysis4.1.1.3. Differences in Body Composition Assessed By Isotope 59Dilution and Bioelectrical Impedance Analysis4.1.2. Comparison of Isotope Dilution and Bioelectrical Impedance Analysis 61Methods with Dual Energy X-ray Absorptiometry Measurement4.1.2.1. Dual Energy X-ray Absorptiometry Measurement 614.1.2.2. Comparison of Isotope Dilution and Bioelectrical Impedance 61Analysis Methods with Dual Energy X-ray Absorptiometryin Assessing Body Composition4.1.2.3. Regional Changes in Body Composition Following 65Ileogastrostomy4.1.2.4. Changes in Body Composition Measured by Isotope Dilution 65Method, Bioelectrical Impedance Analysis and Dual EnergyX-ray AbsorptiometryVI4.1.3. Changes in Energy Expenditure Following Ileogastrostomy 674.1.3.1. Basal Energy Expenditure and Fasting Nutrient Utilization 674.1.3.2. Changes in Thermic Effect of Food 704.1.3.3. Changes in Fat and Carbohydrate Oxidation 734.1.3.4. Total Energy Expenditure in Bypassed Obese Subjects 754.1.3.5. Components of Total Energy Expenditure in Obese Subjects 78Following Ileogastrostomy4.1.4. Changes in Energy Intake Following Ileogastrostomy 784.1.5. Factors Associated with Weight Loss Following Ileogastrostomy 814.2. Validation of Reported Energy Intake Using Doubly Labeled Water 86Technique4.2.1. Physical Characteristics of Normal-weight and Obese Subjects 864.2.2. Total Energy Expenditure in Normal-weight Subjects 864.2.3. Accuracy of Reported Energy Intake in Normal-weight Subjects 894.2.4. Accuracy of Reported Energy Intake in Obese Subjects 934.2.5. Relationship between Underreporting of Energy Intake and 96Related Variables5. DISCUSSION 995.1. Influence of Ileogastrostomy on Body Composition and Methodology in 99Assessing the Changes in Body Composition5.2. Changes in Energy Expenditure 1065.3. Changes in Energy Intake 1145.4. Factors Associated with Weight Loss Following Ileogastrostomy 1165.5. Validation of Reported Energy Intake Using DLW Method 1206. SUMMARY AND CONCLUSIONS 129REFERENCES 131APPENDICES 1451. Invitation Letter for Obesity Study 1452. Sample Consent Form for Obesity Study 1463. Food Record Instruction 1484. Sample of Food Record 1495. Sample of Test Meal 150VII6. Sample of DEXA Measurement 1517. Preparation of Doubly Labeled Water and Sample Instruction Sheet 152for Total Energy Expenditure Measurement8. Postoperative Complications 1539. Relationship between BEE and BW, LBM, FM and El 154LIST OF TABLES VifiTable 1. Methods for Dietary Intake Assessment 31Table 2. Participation in Study Protocol in Obesity Study 55Table 3. Physical Characteristics of Presurgical Obese Subjects 56Table 4. Influence of Ileogastrostomy on Body Composition Determined by 58IDS and BIA MethodsTable 5. Changes in FFM and FM Measured by IDS and BIA Methods 60Table 6. Individual Data of DEXA Measurement Before and After Ileogastrostomy 62Table 7. Comparison of IDS and BIA methods with DEXA in Assessing Body 63Composition Before and After IleogastrostomyTable 8. Changes in Regional Body Composition Measured by DEXA 66Table 9. Changes in FFM and FM Determined by IDS, BIA and DEXA Methods 68Table 10. Basal Energy Expenditure and Thermic Effect of Food, Expressed as 71Absolute Amount and the Percentage of BEE and Ingested EnergyTable 11. Preprandial and Postprandial Fat and CHO Oxidation in the Obese 76Subjects Before and After IleogastrostomyTable 12. Individual Data of Total Body water, Elimination Rates and Total Energy 77Expenditure in Obese SubjectsTable 13. Changes of Energy Intake Following Ileogastrostomy 80Table 14. Individual Data for Energy Balance in the Morbidly Obese Subjects 82Following IleogastrostomyTable 15. Correlation Coefficients Between Energy Loss and Energy Expenditure 85Fecal and Urinary Energy and Energy IntakeTable 16. Physical Characteristics of Normal-weight and Obese Subjects in 87the Validation StudyTable 17. Individual Data of Total Body water, Elimination Rates and Total Energy 88Expenditure in Normal-weight SubjectsIxTable 18. Accuracy of Reported Energy Intake in Normal-weight Subjects 90Table 19. Accuracy of Reported Energy Intake in Obese Subjects 94Table 20. Summary of Energy Intake, Expenditure and the Representativeness of 97Reported Energy Intake in Normal-weight and Obese Subjects.Table 21. Multiple Correlation Coefficients Between Underreporting of Energy 98Intake and Some Physiological Variables in Obese and Normal-weightSubjectsTable 22. Reports of Decreased Food Consumption Following Intestinal Bypass 117Procedures in Humans.Table 23. Studies Comparing TEE in Normal-weight and Obese Subjects 124Using DLW MethodLIST OF FIGURES XFigure 1. Procedures of Ileogastrostomy 8Figure 2. Diagram of the Proposed Factors Associated with Weight Loss 10Following Intestinal Bypass SurgeryFigure 3. Principle of Doubly Labeled Water Method 28Figure 4. Outline of Overall Experimental Design 36Figure 5. Protocol of the Study 37Figure 6. Measurements of Each Metabolic Test 39Figure 7. Changes in Basal Energy Expenditure and Respiratory Quotients 69Figure 8. The Time Course of Thermic Effect of Food in Morbidly Obese Women. 72Figure 9. Changes in TEF Following Ileogastrostomy 74Figure 10. The Percentage of Energy Expenditure for BEE, TEF and Physical 79Activity at the Second Month Following IleogastrostomyFigure 11. Relationship Between Weight Loss and Energy Expenditure Following 83IleogastrostomyFigure 12. Correlation Between Energy Intake and Expenditure in Normal-weight 91Subjects.Figure 13. Difference Between Reported Energy Intakes and Total Energy 95Expenditure in Normal-weight and Obese Subjects, expressed as % TEELIST OF ABBREVIATIONS XIACOVA Analysis of covarianceANOVA Analysis of varianceAPE Atom percent excessBEE Basal energy expenditureBF Body fat%BFBIA Percent of body fat determined by BIA%BFDEXA Percent of body fat determined by DEXA method%BFIDS Percent of body fat determined by LDS methodBIA Bioeletrical impedance analysisBMC Bone mineral contentBIvil Body mass indexBW Body weightCCK CholecystokininCHO CarbohydratesD20 Deuterium oxideDEXA Dual energy x-ray absorptiometryDLW Doubly labeled waterDPA Dual photon absorptiometryEE Energy expenditureEl Energy intakeFM Fat massXIIFQ Food quotientFE Fecal energyFFM Fat-free massGLP Gastric inhibitory polypeptideHt Height113W Ideal body weightLDS Isotope dilution spaceJIB Jejunoileal bypassLBM Lean body massME. Metabolizable energyrCO2 Rate of carbon dioxide productionRQ Respiratory quotientSEE Sleeping energy expenditureSEM Standard error of meanSMOW Standard mean ocean waterTBW Total body waterTBMC Total bone mineral contentTEE Total energy expenditureTEF Thermic effect of foodTFM Total fat massTLM Total lean massACKNOWLEDGMENTS XfflI would like to thank my research supervisor, Dr. Peter Jones, for his kind help andknowledgeable guidance throughout my thesis research project. I am also deeply gratefulto Dr. Gwen Chapman, Dr. Laird Birmingham, Dr. lain Cleator and Dr. David Kitts fortheir constructive criticisms and inspiring discussion of work in progress. I appreciate andacknowledge the generous financial support of the BC Medical Services Foundation.Special thanks to Lisa Martin, Division of Epidemiology and Statistics, Ontario CancerInstitute, for her assistance throughout my thesis project. I appreciate IVIr. Brian Toy forhis help with computer consultation. I would also like to thank Mr. Lance Coombe forhelp in learning the drawing program.I am especially indebted to the 16 obese subjects who enthusiastically volunteered theirtime to participate in, collect samples and provide data for this research project.And finally, thanks to my family for their constant support and my son, Pengfei Hou, forthe lost love from his mother.11. INTRODUCTIONObesity has been clearly associated with adverse health consequences (CanadianGuidelines for Healthy Weights 1988; Reeder et al 1992). Most studies indicate that therelation between weight and total mortality is a J-shapecl curve, with those at the highestweights experiencing highest mortality rates (Canadian Guidelines for Healthy Weights1988; Canadian Health and Welfare 1991; Reeder et al 1992). The increased mortalityassociated with obesity is also significantly age-related with high mortality when obesitydevelops in the early ages. The secondary disorders of obesity include heart disease,diabetes, hypertension and certain forms of cancers which are collectively associated withapproximately sevenfold increase in mortality in the obese compared with normal-weightindividuals (Burton et al 1985; Reeder et al 1992).Weight reduction and maintenance are the chief goals in the treatment of obesity.However, at present there are few effective approaches to achieve weight loss, Surgicaltreatment of obesity is one of such means of producing and maintaining weight loss, withintestinal bypass surgery being one of the available techniques for those patients who meetthe criteria for obesity surgery. Although numerous studies have been conducted inhumans (Cleator et al 1991; Condon et al 1978; Pilkington et al 1976), the exactmechanism of weight loss after intestinal bypass is still unclear. The initial objective of thesurgery was to produce weight loss through malabsorption (Kremen et al 1954), but it isnow well-established in human patients that intestinal bypass causes a substantial reductionin food intake that is the major cause of weight loss (Bray et al 1976a, 1978,1980;2Pilkington et al 1976). Changes in energy expenditure have also been proposed as a factorin the weight loss (Condon et al 1978; Pilkington 1979) but these measurements have notbeen carried out in previous studies. Also, there is evidence that intestinal bypass surgerycan significantly change levels of gastrointestinal and systemic hormones (Besterman et al1978). These hormonal changes may be signals that result in alterations in satiety or inmetabolic disposition of macronutrients (Koopmans 1990).There have been no energy balance studies conducted to determine the exact mechanismsin the weight loss following intestinal bypass surgery. Therefore, we sought to betterexamine the role of energy intake, expenditure and malabsorption in relation to weight lossfollowing ileogastrostomy. The goal of this research was addressed by examination ofweight loss, changes in energy expenditure, and factors associated with this weight lossduring 90-day follow-up and 14-day balance studies after ileogastrostomy. Asupplementary study was added to validate the reported energy intake in normal-weightand obese subjects using the DLW method. Specifically, changes in body compositionwere measured by LDS, BIA and DEXA, and the methods were compared to validate theirapplicability for determining the changes in body composition during weight loss; changesin energy intake and expenditure including BEE and TEF were determined before andfollowing ileogastrostomy; a 14-day energy balance study was conducted to evaluate therole of energy intake, total energy expenditure and malabsorption in the weight lossfollowing ileogastrostomy.3Null Hypotheses:NH: Ileogastrostomy does not affect body composition in the morbidly obese.NFl2. Isotope dilution space method, bioelectrical impedance analysis and dual energyx-ray absorptiometry cannot accurately assess the changes in body compositionafter ileogastrostomy.NB3. Basal energy expenditure and thermic effect of food are not significantly changedfollowing ileogastrostomy.NFL1: The changes in basal energy expenditure and thermic effect of food are notcorrelated with the changes in body composition following ileogastrostomy.NET5: Ileogastrostomy does not reduce energy intake in the morbidly obese.NH6. Energy intake, malabsorption and total energy expenditure are not the primarydeterminants of weight loss following ileogastrostomy.NH7. Reported food records cannot accurately assess energy intake in the morbidlyobese after ileogastrostomy.NH8. Reported food records do not adequately measure energy intake in normalweight individuals.42. LITERATURE REVIEW2.1. Implications, Prevalence and Definition of ObesityObesity, which is a major health problem in North America, is directly or indirectlyassociated with a wide variety of diseases that collectively account for 15-20% of themortality rate (Burton et al 1985). Obesity complicates adult-onset diabetes mellitus,hypertension and cardiovascular diseases (Burton et al 1985; Reeder et al 1992). Morbidlyobese individuals also develop an array of diseases directly related to excess weight(Burton et al 1985). For these reasons, weight reduction and maintenance are the chiefpriority in the morbidly obese.The prevalence of obesity from a survey on well-being inCanada is approximately 27% in adults (Canada, Health and Welfare 1991). Theprevalence of men with a BMI above 26 and 28 was greater than that of women with aBIV11 above these values. However, the prevalence of women with a BIV11 above 30 and 35was greater than the prevalence of men (Canada, Health and Welfare 1988). Also, theprevalence of obesity increased with age and abdominal obesity was higher in men thanthat in women (Reeder et al 1992).Several criteria have been used in defining overweight and/or obesity. The commonly useddefinitions are percentage overweight and BMI range (Hunt and Groff 1990). In the firstcategory, mild, moderate and severe obesity are classified as 28-40, 40-100 and >100%over the desirable body weight. Using BIV11 range as the basis for obesity classification,5three grades are defined as 25-29.9, 30-40 and >40, respectively. Morbid obesity isdefined as BMI>40 or body weight more than 100 lb over ideal body weight (JEW) (Huntand Groff 1990). A body fat content >25% in men or >35% in women is also used todefine obesity (Weststrate 1993).2.2. Causes of ObesityA number of hypotheses have been proposed to explain the development and persistenceof obesity. Among these are genetic factors; metabolic defects; dietary indiscretions andphysical inactivity (Bray 1991; Burton et al 1985; Mayer 1953). Many investigatorsbelieve that obesity is caused by multiple factors in which unequivocal mechanisms forobesity have yet to be determined. Stability of body weight and body composition requiresthat over time, energy intake equals energy expenditure and also that the intakes ofprotein, carbohydrate and fat equal the oxidation of each (Flatt 1987,1988). Although it isunderstood that imbalance between energy intake and expenditure is the primary cause ofobesity, mechanisms through which this imbalance occurs remain to be fully defined. Theregulation of food intake is a complex interaction between special senses and action ofintake. The appearance of the food, its colour, its consistency, and its temperature areperceived by the sensory systems which recognize and translate the stimulus into anelectrochemical message to the brain. The hypothalamus is thought to be the mainintegrator of these signals. Both the brain and gastrointestinal tract release a variety ofhormones, of which insulin is a primary hormonal factor to regulate the food intake and6utilization (Bray et al 1980). Hyperinsulinemia are characteristics of obesity which mayreflect the high levels of nutrient intake and hypothalamic resistance to insulin action.Insulin resistance is frequently observed in obese patients which may result in a cluster ofmetabolic aberrations in obesity.2.3. Treatment of ObesityA large number of therapeutic approaches have been used in the treatment of obesepatients. These include behavioral modification (Foreyt and Goodrick 1991), exercise(Wilmore 1983), diets of various types (Bray 1991; Brownell 1987) and surgery(Halverson and Koehler 1981; MacLean et al 1981; Yale 1989). As a rule, losing andmaintaining body weight are extremely difficult. Dietary restriction is successful in alimited number of patients, but rarely helps obese patients maintain long-term weight loss.Other forms of behavioral modification have yielded similarly poor results (Brownell1987). Pharmacological preparations either do not work or have unacceptable side effects(Foreyt and Goodrick 1991). An approach for treatment of morbid obesity which hasshown promise in producing and maintaining weight loss is surgical intervention(Andersen et al 1984; Halverson and Koehler 1981; Kral 1992).Approximately 30 surgical techniques have been described for treating obesity. Of these,intestinal bypass has been described as one of the most acceptable procedures. Patientsboth lose weight and maintain reduced weight produced by intestinal bypass during longterm follow-up (Kral 1992). Unfortunately, the original jejunoileal bypass (JIB) procedure7was fraught with complication rates that exceeded 40%. These complications includedliver failure, nephrolithiasis, chronic electrolyte abnormalities, and persistent diarrhea(Bray et al 1977). In order to overcome some of these complications, alterations in theprocedure have been developed. The procedure of ileogastrostomy (Figure 1) wasdeveloped to produce weight loss and reduce complications created by JIB such ashepatitis and arthritis (Cleator et al 1988; Gourlay et al 1989). In this procedure, thestandard end-to-endjejunoileal bypass was performed. However, the ileal end of thebypassed segment was drained into the stomach in which hydrochloric acid suppressesbacterial overgrowth in the bypassed segment. Previous studies (Cleator et al 1988,1991)found that the procedure induced significant weight loss which was unaccounted for byreduced energy intake and malabsorption. Therefore, the exact causes of weight lossfollowing ileogastrostomy are still unknown. Further understanding the mechanism ofweight loss following this procedure will have potential applications for safer and moreeffective treatment of obesity.:A1I.ii•t •01••8Figure 1. Operative model for ileogastrostomy. In this procedure, thestandard end-to-end jejunoileal bypass was performed. However, the ilealend of the bypassed segment was drained into the stomach in order tosuppress bacterial overgrowth in the bypassed segment.92.4. Mechanisms of Weight loss Following Intestinal Bypass SurgeryIntestinal bypass surgery has been found effective in producing weight loss in the morbidlyobese since the procedure was introduced more than 20 years ago (Weisman 1973).However, presently there remains controversy regarding what fraction of the weight lossstems from shifts in energy intake and malabsorption as well as from other factors. Figure2 shows the proposed four factors which may be associated with weight loss followingintestinal bypass procedures.2.4.1. Reduction in Energy IntakeClinical and experimental studies have shown a significant reduction in food consumptionfollowing intestinal bypass surgery (Bray et al 1978,1979; Brewer et al 1974). Bray et alestimated energy intake in 14 female patients over a 4-day period in the hospital beforeJIB surgery, and 3 weeks and 6 months after surgery. They found a decline from 4766kcal.d’ preoperatively to 2965 kcal.d’ at 3 weeks, and 3389 kcal.d’ at 6 monthspostoperatively. The distribution of calories consumed as protein, fat, and carbohydratewas not different before and after surgery. Based on the daily reduction in energy intakeand daily weight loss during the study period, they concluded that reduced energy intakeaccounts for most of the weight loss produced by JIB surgery (Bray et al 1978,1979).Somewhat similar results were reported by Robinson et al (1979), who investigated therole of reduced food intake and fat absorption in weight loss in 31 bypass patients. Energyintake declined from 2425 kcal.cl1 preoperatively to 1115 kcal.d’ at 2FecalEnergytI?BEETotalEner________EnergyIntake,,Weight LossExpenditureTEE_____________________EEfor4Activity9MetabolicChangesFigure2.Adiagramaticmodelofweightlossfollowingintestinalbypasssurgery.Thecentralcomponentrepresentsweightloss.Fecalenergyloss(malabsorption),energyintake,energyexpenditureandmetabolicchangesarepresentedintheperipheralcomponents.Arrowsreturningtothecentralpartindicateeffectsontheweightlossafterintestinalbypassprocedures.Theinterrelationshipofeachfactorisalsodemonstratedinthefigure.BEE=basalenergyexpenditure,TEF=thermiceffectoffoodandEEforactivity=energyexpendedforphysicalactivities.11weeks, and 1904 kcal.d’ at 4 months after surgery. Also, the degree of reduction ofenergy intake was closely related to postoperative weight loss (r=0.95). Fat malabsorptionat 4 months was also correlated with weight loss (r=0. 89). This study accounted for theweight loss primarily on the basis of the reduction in energy intake with malabsorptionplaying a secondary role.Condon et al (1979) reported on the pre- and postoperative energy intakes of 65 bypasspatients. The mean energy intake decreased significantly from 3261 kcal to 2595 kcal aftersurgery. In agreement with Bray et a! (1978, 1979), energy consumed as carbohydrates,fat and protein decreased evenly compared with that presurgically. However, the authorsreported that there was no definite relationship between the changes in energy intake andweight loss in JIB patients. Of these 65 patients, 48 decreased whereas the remaining 17patients increased their food intake after surgery. The difference in weight loss betweenthe two groups was not significant. They concluded that alterations in energy intake aswell as malabsorption are important factors determining the rate of weight loss 1 to 9months after bypass surgery. The relative importance of these factors in the cause ofweight loss in the first operative month or two may be different. Marked decrease inenergy intake and striking steatorrhea often occur early, but both problems resolvepartially with the passage of time (Condon et al 1979).With ileogastrostomy, food intake decreases somewhat, and there is some malabsorption,however, these alterations have been found insufficient to account for the energy loss12associated with body weight loss. In a preliminary study of 12 subjects undergoingileogastrostomy, subjects lost a total of 23 kg body weight mostly as fat over the 90 dayperiod, with about 1300 calories per day unaccounted for using energy balancecalculations (Cleator et al 1991). It was suggested that other routes of energy loss mayplay a role in the weight loss following ileogastrostomy.In addition to energy intake changes, eating behavior was also reported to be altered afterintestinal bypass surgery (Mills and Stunkard 1976; Rodin et al 1976). Bypass patientsincrease food intake in the morning and decrease it at night (Bray et al 1978; Brewer et al1974). Pleasantness rating to highly concentrated sugar declined following surgery. Ingeneral, bypass surgery has been described as normalizing appetite behavior (Mills andStunkard 1976).There is a considerable variability in the degree of decreased energy intake and therelationship between postoperative undereating and weight loss (Benfield et al 1976;Condon et al 1979; Robinson et al 1979). This variability can be attributed, at least in part,to differences in energy intake measurement procedures and sampling periods used in thevarious studies. Shortcomings in the methodology of dietary intake assessment maytherefore have resulted in potentially misleading data. Recently, there has been a growingawareness that measuring food intake may be the most challenging problem faced in thesestudies, especially in studies of energy balance and obesity. The quality of dietary intakedata in both normal-weight and obese subjects has been questioned (Block and Hartman1989; Schoeller et al 1990).13Energy expenditure, as measured by DLW method, has been used to evaluate the accuracyof reported energy intake (Schoeller et al 1990). Energy is conserved, and thereforemetabolizable energy intake must equal expenditure plus the changes in body energystores. Energy expenditure and changes in body energy stores can be used to measuremetabolizable energy intake. Researchers have found that the obese tend to more greatlyunderestimate their energy intake compared with normal-weight subjects (Acheson et al1980; Schoeller et al 1990). The magnitude of this underestimation in bypass patients isstill unknown. Therefore, the role of reduced food intake in weight loss followingintestinal bypass needs to be further investigated.2.4.2. MalabsorptionThe original rationale for intestinal bypass was that a shorter intestine would producemalabsorption and thus facilitate the loss of body weight (Kremen et al 1954; Scott et al1971). Many authors have concluded that malabsorption accounts for most or all of theweight loss after JIB (Corso and Joseph 1974; Scott et al 1971; Weisman 1973). There isa decrease in the intestinal absorption of fat, nitrogen, carbohydrate, calcium, potassiumand vitamins (O’Leary et a! 1974; Scott et a! 1971). Scott et al (1971) reported pre- andpostoperative measurements of fat absorption in 7 patients. Postoperative fecal fat levelswere significantly higher in all patients than those preoperatively. Other studies havelikewise found steatorrhea after intestinal bypass surgery (Benfleld et al 1976; Bray et a!141976b). Preoperative fecal fat averaged 7.8±1.3 g.d’ in the stools. Four to six weeks afteroperation, fecal loss of fat rose to an average 44.4±6.6 g.d’. With the passage of time,there was a reduction in the quantity of fat appearing in the stools (Benfield et al 1976;Bray et al 1976b).The increased excretion of fat in the stools probably results from decreased ilealabsorption of bile acids (Wise and Stein 1976). The pancreatic exocrine function was alsoreported to decline after bypass surgery (Dano and Lenz 1974; Sorensen and Krag 1976).These changes reduce the intestinal digestion of triglyceride and thus absorption of fattyacids (Moore et al 1969). In response, the liver increases the production of bile acids fromcholesterol, and plasma cholesterol level decreases (Scott et al 1971). Scott et al (1971)found that plasma cholesterol levels fell rapidly within the first one to five months and thentended to stabilize. There was no tendency to rise with time, even though the absorptionof fecal fat increased.Malabsorption of carbohydrates has been documented (Bray et al 1 976b). Segmentalabsorption of glucose in the jejunum was reported to decline after surgery in one study(Barry et al 1977), but in another study was unchanged (Fogel et al 1976). The loss ofcalories in the stools rose from 131 kcal.d preoperatively to a maximum of 593 kcal.d’postoperatively, and this no doubt increases the rate of weight loss (Crisp et al 1977).Scott et a! (1971) also reported that the energy content of the stools rose from 100 kcal.d1 preoperatively to 500 kcal. d’ postoperatively.152.4.3. Changes in Energy ExpenditureIt is well established that energy expenditure falls in response to diminished energy intake(Apfethaum et a! 1971; Welle et al 1984), and it is generally recognized that this energyconserving phenomenon is counter-productive to the effectiveness of low energy diets intreating obesity (Bray et al 1969; Miller and Parsonage 1975). There are, however,conflicting views concerning whether this adaptive reduction in EE results from a loss ofFFM or an increased efficiency of energy utilization by cellular metabolic processes.Very few data are available on energy expenditure after intestinal bypass surgery(Pilkington 1980). The contribution of changes in energy expenditure to weight loss isunknown at the present time. Kopelman et al (1981) have found that a significant rise inserum 315131 triiodothyronine (T3) and a significant fall in 31315 triiodothyronine (rT3)concentration between 15 and 20 weeks after bypass surgery. They concluded that thisincrease in T3 after bypass may contribute to the substantial weight loss seen at this time.It is not known if there are any associated changes in metabolic rate, therefore, the role ofincreased energy output remains speculative.Total daily energy expenditure can be divided into three major components: BEE, TEF,and the energy cost of physical activity. Most studies have shown that BEE or sleepingenergy expenditure (SEE) in obese subjects is significantly higher than that of normal-16weight subjects (James et al 1978; Ravussin et al 1982). A decrease of SEE after weightloss was reported by Geissler et al (1987), where SEE was 10 percent lower in post-obesewomen compared with lean controls. Dale et al (1990) found a comparable decrease inSEE in subjects just after dietary induced weight loss and the decrease in SEE persistedover years.An impaired TEF has been suggested as a factor contributing to the development ofobesity (Jequier 1984). However, studies on postprandial thermogenesis in obesity haveshown conflicting results. Some studies demonstrated reduced postprandialthermogenesis in obese compared to lean subjects in response to a mixed meal (Segal et al1987a; Shetty et al 1981; Swaminathan et al 1985), whereas others could not find adifferent thermic response to a mixed meal (Cunningham et al 1981; Felig et al 1983).Bessard et a! (1983) found a significantly lower postprandial thermogenesis in obesesubjects after weight loss when compared to lean subjects in response to a liquid mixedmeal. There are no available data for the changes in TEF after intestinal bypass surgery.Total energy expenditure has also been shown to be elevated in the obese state (Ravussinet al 1982; Welle et al 1992). However, little is known about changes in TEE after weightloss. Controversies exist in changes in TEE after weight loss induced by various strategies.Some investigators (Bradfield and Jourdan 1972; Westerterp et a! 1990) reported nochanges or increase in TEE, whereas others (Bessard et a! 1983; Ravussin et al 1985)found that TEE declined after weight loss.172.4.4. Changes in Hormonal Response and Nutrient PartitioningThe gastrointestinal tract is an important component of the diffuse endocrine system.There are good reasons to believe that the anatomical changes induced by intestinal bypasssurgery alter the patterns of gut hormone release (Besterman et al 1978). Reductions ofthe upper small intestinal hormones such as gastric inhibitory polypeptide (GIP) werefound (Sarson et al 1981). Conversely, the ileal hormones such as neurotensin,enteroglucagon and cholecystokinin (CCK) were elevated following surgery (Buchan et al1993; Chan et al 1987; Sarson et al 1981). It is possible that other regulatory peptidessuch as somatostatin, enkephalins and pancreatic polypeptides, as well as unknownintestinal peptides also participate as hormones or neurotransmitters in the regulation ofsatiety and metabolism, and consequent weight loss. At present, changes in thesehormones and their physiological roles are not well understood, thus, any commentary ontheir effects on weight loss falls into the category of speculation.Morphological and functional alterations to a sub-group of regulatory peptides have beenfound after ileogastrostomy (Buchan et al 1993). Quantification of the endocrine cellpopulations in the jejunum in continuity three months after ileogastrostomy demonstrateda hyperplasia of cholecystokinin-, secretin-, gastric inhibitory polypeptide- motilin- andsomatostatin-containing cells. In samples of the ileum taken from within the bypass loopthe neurotensin- and somatostatin-containing cells were unaffected while the18enteroglucagon-containing endocrine cells were significantly increased in numbers. Themost significant alterations were the decreased circulating insulin and increased CCKlevels. The physiological roles of these hormonal changes were not addressed in this study.Whether the dramatic decline of insulin level following surgery influences the metabolismof glucose and other energy-containing nutrients needs to be addressed.In conclusion, reduced energy intake as the primary determinant inducing weight lossfollowing intestinal bypass procedures needs to be further clarified due to the limitations ofdietary intake assessment. The role of energy expenditure in weight loss following surgeryneeds to be investigated. The difficulties in clarification of the relationship between weightloss and factors associated with this weight loss may lie in the methods available. Fullyunderstanding the principles and limitations of each method is fundamental for researchersto interpret the results.192.5. Methodological Studies2.5.1 Assessment OfBody CompositionAssessment of body composition is important in order to describe metabolic consequencesof clinical interventions producing changes in body weight. The most commonly usedmethods for obtaining estimates of body composition are hydrodensitometry, IDS method,anthropometry, and 40K counting. When used individually, each of these methods can onlycrudely partition body weight into FM and FFM based on various assumptions.Hydrodensitometry and anthropometry are not applicable for these bypassed obesesubjects due to difficulty in measurement and inaccuracy of the method.40Potassiumcounting is based on the same priciple as the IDS method. Recent advances in bodycomposition methodology can expand body composition analysis from a two-compartmentmodel to four or more body weight fractions and validate old methods. Dual energy x-rayabsorptiometry is one of these advances in body composition research and was chosen tocompare with BIA and IDS methods on the basis of availability. This section reviews themethods we used in the study. Bioelectrical Impedance Analysis20The BIA method for determining body composition is based on the nature of theconduction of an applied electrical current through the organism under study. Electricalconduction in living organisms is related to water and electrolytes in the biologicalconductor. Because FFM contains virtually all the water and conducting electrolytes in thehuman body, conductivity is far greater in FFM than in FM of the body (Lukaski 1987).The electrical volume is inversely related to impedance (Z), resistance (R), and reactance(Xc) whereZ(R2+Xc)°5.Determination of resistance and reactance are made using afour terminal impedance plethysmograph. Estimation of FFM is obtained from anarithmetic calculation using a previously validated predictive equation.The first investigations to develop mathematical models to predict TBW and itsdistribution in humans using the impedance approach were performed by Thomasset et al(1962). Since then, investigators have demonstrated significant relationship betweenTBW, estimated as isotope dilution space, and Z, R, and Xc in 37 men (Lukaski et al1985) and in a group of 26 children and adolescents (Davies et al 1988). Thesepreliminary findings indicated the potential of the tetrapolar method to assesscompositional variables. Cross-validation studies were also undertaken to determine thevalidity of BIA method to assess TBW and FFM. Kushner and Schoeller (1986) tested thevalidity of an impedance model for the prediction of TBW derived in a sample of 40nonobese adults by applying it to 18 obese patients. The model successfully predictedTBW (r=0.95) with differences between measured and predicted values of only 0.6 to 1.021liters. In another study, it was shown that the prediction equation developed in the menwas capable of estimating FFM accurately in the women (Lukaski et al 1986). Thisapproach has also been used to demonstrate that impedance estimates of percentage bodyfat are similar to those determined using appropriate densitometric procedures.The application of BIA method to assessment of body composition in obese individualsindicates an obesity-dependant bias in predicting FFM determined densitometrically in alarge cross-validation study (Segal et al 1988). The controversies still exist in theapplication of BIA method to assessment of body composition during weight loss. Gray(1988) reported a significant correlation between TBW and Ht2/R before and after a 2-week fast in 6 obese women who lost 10 kg. However, Deurenberg et al (1989) reportedthat estimates of FFM were significantly less than those determined by densitometry in agroup of 13 obese women whose body weight decreased 10 kg after an 8-week weightreduction program. The apparent lack of consensus about validity of the BIA method toestimate the composition ofweight loss indicates the need to conduct controlled studies inwhich a multicompartmental assessment of body composition is employed.The BIA method offers a wide variety of potential applications for noninvasive assessmentof human body composition, because it is safe, convenient, and easy to use. Experimentalfindings from cross-validation studies demonstrate that the BIA method is valid andaccurate for estimation of FFM, TBW and BMC in healthy individuals. However, the22validation of BIA method in patients with abnormal water and electrolyte distributionsneeds to be evaluated. Isotope Dilution Space MethodThe finding that water occupies a relatively fixed fraction (73.2%) of fat-free mass (FFM)(Pace and Rathburn 1945) has stimulated the determination of total body water (TBW) asan index of human body composition. Some general assumptions of the isotope dilutiontechnique are that the isotope has the same distribution as water, it is exchanged by thebody in a manner similar to water, and it is nontoxic in the amounts used (Pinson 1952).Isotopes of hydrogen, deuterium and tritium, have been used to determine body watervolumes in healthy and diseased individuals (Culebras and Fitzpatrick 1977; Henry andPhyllis 1985; Schoeller and Jones 1987). The extensive use of the deuterium oxide dilutiontechnique for the estimation of TBW in mammals has demonstrated that the method isvalid and accurate (Culebras and Fitzpatrick 1977; Moore 1946). This isotope is rapidlyabsorbed in the gastrointestinal tract and equilibrates with body water in a few hours(Schoeller 1980, 1992). This method is comfortable for patients because it requires onlythe ingestion of the isotope and the collection of one or more urine or saliva samplesafterwards. It is the most precise method for the determination of pool sizes of body water(Schoeller 1992). This procedure has an analytical precision of 2.5% (Lukaski 1987). Thetechnique is generally advocated as the traditional method for body compositionmeasurement. It seems a particularly appropriate method to compare with BIA because23body fluids and electrolytes are responsible for electrical conductance (Henry and Phyllis1985; Schoeller 1989).Despite the high technical precision of the isotope dilution method, errors can be made incalculating FFM and FM because it is not known whether the water content in FFMremains constant in all subjects under all circumstances. Particularly, the assumption of73.2% of body water in FFM (Sin 1956) may be violated in the bypassed patients whomay experience dehydration following surgery. The validity of isotope dilution applicationto this group of patients has not been studied. Also, the technique provides no informationconcerning patterns of body fat distribution or changes in regional body composition afterweight loss. Dual Energy X-ray AbsorptiometrvDual-energy x-ray absorptiometry is a relatively new method for quantif,ring the skeletaland soft tissue components of body mass in vivo (Going et al 1993; Mazess et al 1990;Svendsen et al 1993). The fundamental principle of DEXA is based on the differentialattenuation by tissues at two energy levels. The composition of soft tissue is given by theratio of the soft-tissue attenuation (R) measured at the two energies. The attenuation ofpure fat (Rp) and of bone-free lean tissue (RL) are known from both theoreticalcalculations and human experiments. Given the subject’s and the known Rs for fat andlean, the proportion of fat and lean tissue in each pixel can be calculated. The method can24simultaneously measure bone mineral content and soft tissue for total and regional bodycompostion (FFM and FM).Preliminary results suggest that this new method can be used to accurately estimate softtissue composition with better precision (1-1.5%) than was possible with dual photonabsorptiometry (DPA) (Going et al 1993; Johansson et al 1993). However, the cross-validation of DEXA measurements has not been extensively researched, Comparisonbetween DEXA and hydrodensitometry estimates showed high intercorrelations betweenDEXA and hydrodensitometry methods (range 0.86-0.92). When DEXA was comparedwith skinfold anthropometry, bioelectrical impedance and total body potassium methodsfot the measurement of total body fat, significant differences in total body fat wereobserved between DEXA and the other methods (Oldroyd et a! 1993). More recently,Ryde et al (1993) compared DEXA with neutron activation method for the measurementof body fat. The results indicate that DEXA and neutron activation methods givecomparable measurements of body fat in a female population. On individual basis,however, there are clear differences between the methods. Measurement of abdominal andvisceral fat with computed tomography (CT) and DEXA showed that CT- and DEXAmeaused total abdominal fat were similar and highly correlated (r=0 .985, p<O.OO1) (Jensenet al 1995). The validation of DEXA measurement in obese subjects has not been studied.Because the DEXA method is theoretically independent of compartmental assumptions,the technique may prove useful for following changes in body composition over time.25Future studies are needed to expand the subject pooi by investigating obese individualsand patients with disturbed hydration. DEXA is independent of biological assumptionsabout the consistency of tissue densities and level of hydration. Also, DEXA isinexpensive and safe compared with the imaging techniques such as CT and nuclearmagnetic resonance imaging (IvIRI), because the radiation dose for a whole-body scan byDEXA is <5 mrem. DEXA is rapid and easy requiring only 15-20 mm and littlecooperation from the subject. Disadvantages of DEXA include the limited dimensions ofthe scanning table, which can exclude persons too large (>3 00 lb) or tall. Themanufacturers usually recommend an upper limit of body weight for DEXA instrumentsbecause body weight as well as tissue thickness may affect the accuracy of DEXAmeasurements.2.5.2. Measurement ofEnergy ExpenditureMeasurement of human energy expenditure is important in many areas of clinical andresearch investigation such as obesity, undernutrition, exercise, and a number of diseasessuch as trauma, infection and cancer. Both direct and indirect calorimetry have beenapplied to assess the metabolic rate and short- and long-term energy balance (Jequier1981; Jequier and Schutz 1985). The relative advantages and disadvantages of both directand indirect methods have been reviewed by Jequier (1981). A variety of measurementtechniques have been applied by investigators in the study of obesity with specificconsideration to the role of inadequate energy expenditure in the onset and persistence of26obesity. Conflicting results in this area of research have led to speculation about theimportance of methodological differences as a possible source of some of theseinconsistencies (Jequier 1981; Jequier and Schutz 1985). Several indirect methods havebeen used to study human energy expenditure on the basis of gas exchange analysis. It hasbeen suggested that the ventilated hood technique is the most comfortable indirect methodbecause subjects are able to breathe more naturally under the canopy than using a mask ora mouthpiece (Welle 1984). Furthermore, canopy systems require careful adjustment ofthe rate of the air flow into the canopy (Segal 1 987b). However, this technique cannotmeasure the energy expenditure of physical activity.The DLW method for measuring energy expenditure provides the capacity to measuretotal energy expenditure, including that of activity. Development of this method can betraced to a study performed by Lifson et al (1949) in the 1940s. They administered ‘SOlabeled water to animals and demonstrated that the oxygen in expired CO2 was derivedfrom body water. This is now known to result from the maintenance of isotopicequilibrium between the oxygen atoms of body water and CO2through the carbonicanhydrase reaction. On the basis of this observation, Lifson et al (1955) reasoned thatintegrated CO2 production could be measured from the differential elimination of waterlabeled with both isotopic hydrogen and oxygen. After a loading dose of DLW, the labeledhydrogen would be eliminated as water, whereas the oxygen isotope would be eliminatedas both water and CO2. Thus, it is theoretically possible to measure CO2 production bymeasuring the isotopic hydrogen and oxygen remaining in body water after administration27of the DLW. The DLW method was used to monitor the fluxes ofwater and CO2 throughthe body (Schoeller 1982, 1986). The principle of the DLW method is demonstrated inFigure 3.The DLW method has been extensively validated in animals (Nagy 1980) and humans(Coward et al 1988; Jones et al 1987, 1993a; Schoeller 1987,1988). The method has anaccuracy of 1% and a precision of 6% (Schoeller 1987). Moreover, replicate measures in3 subjects over a 2-year period and in 16 subjects in consecutive weeks have demonstrateda repeatability of 6% (Schoeller and Taylor 1987; DeLany et al 1989). Thus, it appearsthat the DLW method can accurately measure integrated energy expenditure over periodsof at least 2 weeks. Furthermore, the DLW method isnoninvasive, nonrestrictive and thus ideal for the measurement of total daily energyexpenditure in free-living subjects (Schoeller 1988).The applications of the DLW method are wide, but most uses to date have takenadvantage of the ability of the method to accurately measure energy expenditure in freeliving subjects. A number of investigators have applied the method to the study of obesity.These studies have measured the energy expenditure of obese individuals to determine ifthey are energy-efficient, as some intake studies have suggested. In addition to that, DLWmethod has been proposed as a method to validate energy intake assessment techniques.More recently, the majority of applications of the DLW method have been28I2112180 1/__________211 labels labels water Iwater pool and bicarbonate pooisj-4__________IjH218Ob-col8oj* IrHO r0 + r110* *k18— Ic2 rCOFigure 3. Principle of doubly labeled water method. k=experimentallydetermined rate constant. r=production rate.29aimed at determining energy requirements of healthy individuals. Currentrecommendations for energy requirements in adults with different levels of physicalactivities are based on fractional estimates of total energy expenditure TEE in which theprincipal components of TEE are either measured or estimated. In most cases, the valuesfor resting energy expenditure (REE) are predicted from age- and gender-specificequations. Application of an activity factor, derived from a crude assessment of thesubject’s physical activity level, to a measured or estimated level of REE are the basis ofcurrent energy requirement in healthy adults (Goran and Poehlman 1992). An activityfactor of 1.7 and 1.6 was used for male and female adult subjects, respectively. Becausethere have been no accurate methods to measure energy expended for physical activities,these factors are questionable. Many studies using the DLW method have been conductedin various age groups which include young adult men (Roberts et al 1991), adolescents(Bandini et al 1990), elderly persons (Goran and Poehlman 1992), lean and obese women(Prentice 1986), underweight adults (Riumallo et al 1989) and patients after surgery(Westerterp et al 1991). Token together, these studies provide evidence consistent withthe findings that current recommendations for energy requirements may underestimate theenergy needs in the healthy adults. Therefore, the activity factor should be increased to 1.8or more to cover the actual energy requirements in healthy adults.Clinical applications of the DLW method are currently in progress. Novick et al (1987)measured the energy expenditure of surgical patients and reported that energy expenditureincreased by 18% after surgery. In order to examine the role of energy expenditure in the30weight loss following ileogastrostomy and validate the reported energy intake, the DLWmethod was used to determine total energy expenditure in the postsurgical obese subjectsin the present study.2.5.3. Food Intake MeasurementVarious methods of dietary assessment have been used in studies of the role of diet anddiseases, and their strengths and deficiencies have been reviewed by Barrett-Connor(1991). The five main methods of diet assessment used in individuals are summarized inTable 1. At present, there are no dietary intake methods to assess energy intakewithout errors (Beaton 1994). The nature and magnitude of the error depends on both thedietary data collection methodology and the subjects studied. This section briefly reviewsthe weighed food record method, because it was the method most commonly used toinvestigate the role of reduced energy intake following surgical intervention in the past andwas chosen for energy intake assessment in the present study.The weighed food record approach has traditionally been perceived as the “gold standard”for dietary measurement and has been used in attempts to validate other methods such asfood frequency questionnaires. It is a method of diet assessment applicable to free-livingpopulations that assures quantitative and qualitative measurement of all nutrientsconsumed (Barrett-Connor 1991). The weighed food record method has both merits anddrawbacks. The main advantage is that the amounts31Table 1. Methods for Dietary Intake AssessmentMethod Expensive Behavior Quantitative RepresentativechangeDiet diary or record Yes Yes Yes(Condon et al 1978)Observation Yes No Yes Yes(Lansky and Brownell1982)Diet history Yes Yes Semiquantitative(Burke 1947)24-h diet recall Yes No Yes No(Balogh et al 1971)Food frequency No No Semiquantative Yesquestionnaire(Robinson et al 1979)32consumed can be recorded more accurately than by any other conventional methods(Pekkarinen 1970). The disadvantage of the method may be the problem of a trainingeffect and less representiveness of usual intake. The need to weigh and record intake maylead to a reduced intake or more monotonous diet and induce behavior change. The lengthof the recording period is also an important methodological concern (Tarasuk and Beaton1992). It is generally agreed that the recording period should be long enough to givereliable information on the normal food consumption (Pekkarinen 1970). However, toolong a survey may result in under-representing of the actual intake, since much work andtrouble is involved in the use of the method (Marr 1971). In conclusion, the collection offood records and the associated nutrient analysis comprise an extremely expensive andtime-consuming aspect of dietary studies and pose a considerable burden to studyvolunteers.The weighed food record method has been utilized over the decades; however, theaccuracy of reported intake remains largely unknown. The vast majority of evaluations ofthe weighed fOod record instrument have not included tests of accuracy or bias, becausesuch tests are extremely difficult to perform in a home environment. Lissner et al (1989)compared the reported intake of women with the actual intake fed these women duringsubsequent metabolic studies. The actual intakes were corrected for any change in bodyenergy stores during the metabolic period. The authors found that reported intakes tendedto underestimate maintenance energy intake. The degree of underestimation was related tosubject’s FFM. Because of this study and others (Bingham 1987; Schoeller 1990), there is33controversy about the accuracy of reported dietary intake. The obstacle to resolution ofthe controversy is the absence of methods to validate the accuracy of reported intake,especially for free-living subjects in which the goal is to determine habitual intake.The doubly labeled water technique has been proposed as a reference method to validatereported energy intake (Schoeller 1990). The method has an accuracy of 1% and aprecision (1 SD) of 6%. Thus, the DLW method is known to be accurate and preciseenough to serve as a reference method for the validation of reported dietary intake.Although some validation studies using the DLW method have found that reported intakeagrees well with energy expenditure and hence provides an unbiased estimate of habitualintake, the majority of the studies have detected bias in reported intake. Riumallo et al(1989) observed the highest level of accuracy for reported intake. Reported dietary intakeaveraged 2689±284 kcal.d’ (mean±SD) versus measured energy expenditure of 2724±3 03kcal.d’. Thus, in this study, reported intake was accurate. Another study in whichreported intake was found to be accurate was that of DeLany et a! (1989). Dietary intakeaveraged 2960±487 kcal.d’ versus an expenditure of 3230±520 kcal.d’. On the otherhand, more and more studies have noted a large bias in reported intake. Bandini et alcompared reported intake with expenditure in obese and nonobese adolescents. In thenonobese subjects, reported intake averaged 8 1±19% of measured expenditure (2 190±620kcal.d’ versus an expenditure of 2760±600 kcal.d’). In the obese, reported intakeaveraged only 59±24% of expenditure (1940±720 kcal.d’ versus an expenditure of 3390±34610 kcal.d’). This reported low intake could not be traced to undereating during thereporting period, because both obese and nonobese gained weight during the reportingperiod (0.1±0.7 and 0.4±1.0 kg, respectively). Therefore, the reported low intake in bothgroups probably reflects underreporting. In a similar study, Prentice et al (1986) comparedreported intake with expenditure measured by the DLW method. Reported dietary intake(1880±350 kcal. d’) compared quite well with measured energy expenditure (1910±240kcal.d’) in the lean group, but poorly (1610±430 kcal.d’ versus 2440±310 kcal.d’,respectively) in the obese group. Similar to the data reported by Bandini et al (1990), biaswas greater among the obese subjects. However, in contrast to the observations byBandini et al (1990), the lean control group reported quite accurately. A third study alsogrouped subjects as either normal or overweight (Bronstein and King 1988). Qualitatively,the results are similar to those of Prentice et al (1986); normal and overweight groupsreported very similar intakes, whereas energy expenditure was 550 kcal. d’ greater in theobese group than in the lean group.The findings presented above indicate great variation in the accuracy of reported intake.Perhaps the greatest limitation occurs in the use of dietary record for studies of energybalance, requirements, dietary intervention and obesity. The DLW technique provides aunique opportunity to identify the bias in reported energy intake and possibly improvetechniques for assessing dietary intake.353. EXPERIMENTAL DESIGN AND METHODSThe overall experimental design encompassed two parts (Figure 4). For the firstexperiment, 16 obese subjects were consecutively selected to participate in themeasurements of body composition, food intake and energy metabolism before and afterileogastrostomy. A 14-day energy balance study was conducted during the 90-dayexperiment (Figure 5). The strength of our design was the use of short-term measurementofweight loss and factors associated with this weight loss.Experiment 2 was intended to answer the question of validation of energy intakemeasurement in obese and normal weight subjects. Validation in obese subjects wasconducted during 6-8 weeks following ileogastrostomy using the DLW method. Subjectswho had completed the entire energy balance study were selected for the validation study.A supplementary study regarding validation of reported energy intake in normal-weightsubjects was carried out to assure the limitations of dietary record in estimation of energyintake. The reason for selection of normal-weight group in this study was that very fewobese subjects completed the energy balance study in experiment 1. Therefore, thevalidation of energy intake is limited due to small sample size.36Experiment 1: Examination of Mechanism of Weight LossFollowing IleogastrostomyObese subjectsIDS, BIA and DEXA Gas exchage analysis DLW method Weighed food recordBody composition BEE and TEE TEE Energy intakeExperiment 2: Validation of Reported Energy Intakein Normal-weight and Obese Subjects/Normal-weight group Obese groupTEE = Energy intake — Body energy stores(DLW) (weighed food record) (changes in composition)Degree of underestimationof energy intake in both groupsFigure 4. Experimental design: In experiment 1, obese subjectsparticipated in pre- and postsurgical corresponding measurements. Barsindicate subjects included in each experiment. Arrows indicatemeasurement by corresponding methods. In experiment 2, energybalance equation was used to evaluate the degree of underestimation ofenergy intake in both normal-weight and obese subjects.Fourtimesofmetabolictestduring12weeksIr1rDLWmeasurement,IrI 68I‘I,Collectionof5-dayfecesandurineFigure5.Protocol ofoverallstudyplan.TotalenergyexpenditureusingDLWmethodandcollectionoffecesandurinewerecarriedoutduringthe14-dayperiod.Fourmetabolictestswereconductedbeforeandat4,8and12weeksaftersurgery.02410l2wks0383.1. Examination of Mechanism of Weight Loss in Obese Subjects FollowingIleogastrostomy3.1.1. Selection and Screening ofSubjectsSixteen obese subjects who were scheduled for the ileogastrostomy were selected for a90-day study. Subjects who had been obese for over 5 years and failed at other strategieshad been selected at the Division of Internal Medicine at St. Paul’s Hospital. Subjects whohad previously diagnosed with diabetes or coronary heart disease were excluded. Eligiblesubjects were sent a description of the research project protocol (Appendix 1) andsubsequently contacted by telephone to further discuss the study. The consent form(Appendix 2) was signed at the presurgical test. The surgery performed in this study wasileogastrostomy (Figure 1) carried out by Dr. Cleator’ s group and described elsewhere(Cleator et al 1988). Anthropometric measurements were determined by the sameinvestigator at each of the 4 measurement points. Weight was determined to the nearest0.01 kg by the same scale with the patient dressed only in light underwear and withoutshoes. Body mass index was calculated dividing body weight (kg) by height squared (m2).Figure 6 outlines the protocol of each metabolic test. The protocol used at the other timepoints was nearly identical. The experimental protocol was approved by the EthicalCommittee of UBC and St. Paul’s Hospital.MeasurementofBEEAMeasurementofTEFh.TestmealD20BIA5hours 1’SalivaColiectionFigure6.Measurementsofeachmetabolictest.Deuteriumoxide(D20)wasadministeredat2hoursaftermealandBIAmeasurementwasperformedattheendofeachtest.3-dayfoodrecordwaskeptbyeachsubjects3daysbeforethetest.BEEmeasurementlastedfor30minutesandTEFcontinuedfor5hours.,Ir013-dayfoodrecords1234BaselinesalivaCollectionCo (0403.1.2. Measurement ofEnergy IntakeEnergy intake was determined by weighed food record method. Subjects were providedwith an instruction sheet (Appendix 3), food record (Appendix 4) and a dietary scale tomeasure food weight. The dietary food record was kept by each subject over 3 daysbefore each of the metabolic tests. On the presurgical assessment, each subject was trainedfor measuring food weight using a dietary scale and the instruction sheet was explained.The subjects were requested to specifj the brand names for commercial products or typeof ingredients used in preparing recipes. All items of food were weighed and recordedseparately. Leftover foods were also weighed and recorded. Some uncertainties about thefood records were clarified and corrected on the metabolic test days to ensure accuracy.The nutrient composition and energy intake of the records were analyzed using FoodProcessor II program (ESHA Research, Salem, OR, 1987). Consumption of energy,macronutrients and some micronutrients can be analyzed on the basis of food items andamount of food records.3.1.3. Basal Energy Expenditure and Thermic Effect ofFoodSubjects were tested prior to, and at 1, 2, and 3 months following surgery in the NutritionMetabolism Unit at St. Paul’s Hospital. All gas exchange measurements were performedby use of a ventilated-hood system (Deltatrak; Sensormedics, Anaheim, CA). Thismetabolic cart system was validated by directly measuring CO2 production and 0241consumption model in a previous study (Phang et al 1990). The overall errors for VCO2and V02 were 1.5% and 1.9%, respectively. The subjects were not hospitalized. After a30-minute rest, BEE was continuously measured between 7:00 and 7:30 a.m. After theBEE measurement, a test meal (Appendix 5) was served at about 7:35 a.m. and was eatenwithin 20 minutes. The energy level of the test meal was calculated to cover 30% of dailyenergy expenditure estimated from IBW, using Muffin’s equation of energy expenditure(Muffin et al 1990). This meal contained 15% kcal derived from protein, 35% kcal fromfat and 55% kcal from carbohydrates (CHO). The TEF was then continuously measuredfor 300 consecutive minutes while subjects remained at rest.3.1.4. Body Composition MeasurementOn the four tests of BEE and TEF, an oral dose of non-radioactive deuterium oxide (99%APE, 0.06’ estimated TBW) was consumed at 2 hours after the test meal. Salivasamples were collected predose and at 3 and 4 hours post dose. The deuterium enrichmentwas determined using mass spectrometry. Bioelectrical impedance analysis was alsoperformed as a comparative indicator of composition on each occasion. Bioelectricalimpedance was measured with the subject supine as described by Vazquez and Janosky(1991) with a body composition analyzer (RIL System, Detroit, MI). FFM was calculatedusing the equation of Segal et al (1988).FFM (kg) = 0.00091 186x(height)2-0.014 6x(R)+0.29990x(weight) (Eqn. 1)-0.070 12x(age)+9.3793842Dual-energy x-ray absorptiometry was performed in some of subjects presurgically and at6 weeks postsurgically. DEXA measurements were made with a total-body scanner(model DPX: Lunar Radiation Corp., Madison, WI) that uses a constant potential x-raysource to achieve a congruent beam of stable dual-energy radiation with effective energiesof 40 and 70 keV. The scanner was calibrated daily against standard calibration block tocontrol for possible baseline drift. Subjects lay supine on a comfortable table while thescintillation counter moved across the body from head to foot. A series of transverse scanswere made in 20 minutes at 1 cm intervals. DEXA directly measures three principalchemical components of the body: total fat mass (TFM), total lean mass (TLM) and totalbone mineral content (TBMC) (Appendix 6). Fat-free mass by DEXA is the sum of TLMand TBMC. The fat percentage of the body is calculated as TFM divided by the sum ofFFM and TFM. DEXA also provides measurements of these three components indifferent parts of the body (Appendix 6).3.1.5. Measurement of Total Energy ExpenditureTotal energy expenditure in obese subjects was measured between 6 and 8 weeksfollowing ileogastrostomy (Figure 5). On day 0 of the study, subjects reported to theMetabolic Lab in the fasting state. Subjects were weighed and the baseline urine and salivasamples were obtained. A single oral dose of 180 and deuterium labeled water (Appendix7) (0.25 gH2O’ estimated TBW from ideal body weight (IBW), 0.1’estimated TBW) was followed by about 100 ml tap water. Two urine samples (one in the43morning and the other in the afternoon) were collected on day 1 at the patient’s home. Onday 7 and 14, approximately at the same time, further urine samples were collected. Onthe morning of day 15, subjects reported to the Metabolic Clinic at St. Paul’s Hospital anda further dose of labeled water (0.06 g D20 kg’ estimated TBW) was administered todetrmine TBW at the end of the DLW period. All samples were frozen at -50°C in airtightparafllm wrapped plastic containers until use.3.1.6. Measurement ofFecal and Urinary Energy LossesAdditional routes of dietary and metabolic energy loss were investigated. During theperiod of TEE measurement, subjects were requested to comprehensively collect fecalmaterials and urine into containers provided for 5 days. The entire fecal and urinematerials for each subject were weighed, homogenized, pooled and freeze-dried. Aftergrinding and mixing of freeze-dried samples, feces and urine samples were subjected togross energy content determination by bomb calorimetry (Miller and Payne 1959).Measurements were performed in duplicate on approximately 1.0 g of the dried samplesusing benzoic acid as standard. Metabolizable energy was calculated for individualsubjects as the difference between energy consumed and energy loss in feces and urine.443.2. Validation of Recorded Energy Intake Using Doubly Labeled Water Techniquein Normal-weight and Obese Subjects3.2.1. Selection and Screening ofSubjectsValidation of reported energy intake in obese group was conducted during 6-8 weeksfollowing ileogastrostomy. The subjects who had completed the energy balance studywere selected for the validation study. Twenty-six normal-weight subjects were recruitedfrom all women who had participated in a low dietary fat trial for the prevention of breastcancer in the Toronto area. All women were sent a letter and information sheets describingthe validation study. Participants who reported current use of diuretics or who had ahistory of an eating disorder were excluded. Eligible subjects were contacted to furtherdiscuss the DLW project. The consent form was signed at the beginning of the first clinicvisit. The study protocol was approved by the University of Toronto Review Committeeon the Use of Human Subjects.3.2.2. Measurement ofEnergy Intake Using Weighed Food RecordsThe collection and analysis of food records in the obese and normal-weight groups wereconducted by different researchers using different computer programs in Toronto andVancouver areas. Each subject in the obese group was requested to complete weighedfood records as described above (Appendix 3,4) for 5 days during the 14-day DLW45period. The records were reviewed and analyzed using the Food Processor II (ESHAResearch, 1987). Normal-weight subjects were asked to record a 7-day weighed foodintake during the DLW period. Subjects were asked to weigh and record all food andbeverages consumed in the similar manner as the obese subjects had been instructed(Appendix 3). The nutrient composition and energy intake of the records were reviewedand analyzed using the nutrient data-analysis system (NDS, Nutritional CoordinatingCenter, Minnesota, 1993).3.2.3. Measurement of Total Energy Expenditure Using Doubly Labeled WaterTechniqueThe obese group included subjects who were participating in the examination of weightloss and energy metabolism following ileogastrostomy. The validation study in this groupwas conducted during 6-8 weeks following ileogastrostomy. The procedure of TEEmeasurement was described above.Normal-weight subjects were instructed to maintain their normal daily activities and tomake no conscious attempt to lose or gain weight during the study period. On day 0 of thestudy, subjects were weighed and baseline urine (10 ml) and saliva (5 ml) samples wereobtained. A single oral dose of 0.17 g H2O’8 and 0.07 g D20 per kg body weight wasadministered and followed by 100 ml of tap water. Saliva samples were collected at 3 and4 hours and a urine sample was collected at 4 hours post dose for measurement of TBW.46A urine sample was also collected by the subject at home on the morning of day 1. On themorning of day 14, subjects returned to the Clinic in the fasted state. A urine sample wasobtained and a further dose of labeled water (0.06 g D20 per kg body weight) wasadministered. Saliva samples were collected at 3 and 4 hours post dose for thedetermination of final body water volume.473.3. Analytical ProceduresUrine and saliva samples obtained for body composition and TEE measurements wereprepared for isotopic analyses using a vacuum system and analyzed by an isotope ratiomass spectrometer (VG Isomass, 903D, Cheshire, UK).3.3.1. Furfication ofDeuterium and Carbon DioxideA vacuum line was used for the preparation of hydrogen gas from the aqueous phase ofbiological fluids and working standards. Preannealed 6-mm-OD Pyrex tubes containing 60mg of zinc (Indiana University, Bloomington, IN) were attached to a vacuum inlet systemand dried of moisture for 10 minutes. A 2-pi microcapillary filled with urine or salivasamples was added to each tube before reattachment to the inlet system. Tubes wereimmersed in liquid nitrogen for 5 minutes, residual gases evacuated (<1 02 Torr) over 2minutes, and tubes flame-sealed. Reduction of samples was carried out at 540°C for 30minutes before mass spectrometric analysis.‘8Oxygen was purified as CO2. Urine was aliquoted (1.5 ml) into vacutainer tubes. Carbondioxide (1 ml) was added by injection, and then samples were agitated for 1 hour andincubated at 25°C for at least 48 hours. The CO2 collection line was evacuated to less than100 millitorr. The sample tube was immersed in liquid nitrogen for 5 minutes. The48contents were transferred toN2/methanol bath for 5 minutes. Then, CO2 was collected in atube placed in an N2 bath. The collecting tube was flame-sealed.3.3.2. Mass Spectrometric DeterminationIsotope ratio mass spectrometry permits an accurate isotopic enrichment analysis for anumber of low molecular weight compounds. The deuterium and 180 were measuredthrough the determination of2D/’H and46CO2t0.Sample tubes were manually inletedinto the mass spectrometer and analyzed against Vienna standard mean ocean water(SMOW). For deuterium measurements, the mass spectrometer was set up and calibrateddaily using SMOW standard and two working standards (GISP and V-Std). Regressionanalysis of observed enrichment values of standa.rds indicated good linearity of response atboth high and low enrichments (r>O.999). Appropriate standards, baseline and all samplesfor a given subject were analyzed in triplicate against an identical set of standards within a12-hour period. Isotopic data was expressed as per mil (ö%o) abundance relative toSMOW. Measurements were repeated in cases where replicate value differences exceededthe maximum acceptable tolerance for low and higher (>500 %o) enrichment samples of 2and 5%, respectively. As with 2D analyses, 180 was analyzed as CO2 against SMOW andcalibrated tank standards (-30%o). Maximum enrichment difference between replicateswas 0. 5%. Linearity of response was checked periodically with Vienna ‘O standardsenriched at 250 and 500%.493.4. Data Calculation3.4.1. Calculation of Total Body Water and Energy ExpenditureTotal body water was determined from deuterium dilution space (DS) calculated asfollows:DS (kg) = (dose x APE x 18.015)/(MW x R x A Enrichment) (Eqn. 2)where dose (g) is the amount of label given, APE is the atom percent excess of the dose,18.015 (g) is the molecular weight of water, MW (g) is the molecular weight of the dose,R is the known ratio of the heavy to light isotopes in a reference standard and Aenrichment is the observed isotopic enrichment over baseline enrichment. TBW is definedas DS of 2D divided by 1.04 (Schoeller et al 1980). Body water was assumed to comprise73.2% of FFM, so that FFM equals TBW divided by 0.732, and fat mass equals bodyweight minus FFM (Culebras and Fitzpatrick 1977; Moore 1946; Pace and Ruthburn1945).The calculation of energy expenditure was based on the assumptions of DLW techniquesusing Schoeller’ s equation ( Schoeller 1982,1986,1988). Mathematically,rCO2 (moled’)= 0.46 x TBW (1.01K8 - 1.04K2) (Eqn. 3)50where rCO2 is the rate of carbon dioxide production (moled’), TBW (mol) is the averageof TBW volume determined from deuterium dilution space during 14 days, and K18 and K2are the calculated elimination rates (pool,d’) of 180 and deuterium, respectively.Respiratory quotient (RQ) is estimated by food quotient (FQ) derived from 5-day foodrecords (Black et al 1986). Energy expenditure (kcal.d’) (Jones and Leitch 1993a) wascalculated from the CO2 production rate using the FQ (Black et al 1986).EE (kcal.d’) 3.9 x rCO2 / FQ + 1.1 x rCO2 (Eqn. 4)3.4.2. Estimation ofBasal Energy Expenditure, Thermic Effect ofFood and SubstrateUtilizationGas exchange data were corrected as computed values for oxygen and carbon dioxideexchange at standard temperature, pressure and humidity. The respiratory quotient(VCO2/ 0)was calculated with subtraction of gas exchange with protein oxidation andresults were converted to kilocalories by the Weir formula (Weir 1949). Similarly,carbohydrate and fat oxidation rates were determined each minute. A constant proteinoxidation rate of 0.7 g protein kg FFIVf’.d’ was assumed, and non-protein energyexpenditure and macronutrient utilization rates (per minute) were calculated (Jones andSchoeller 1 988a). Energy expenditure and macronutrient oxidation data were expressed as30-minute averages during the BEE period and as 60-minute averages during TEF51measurement. Postprandial thermogenesis was calculated as (i) the net increase over BEE;(ii) the percentage of increase relative to BEE and (iii) the percentage of increase relativeto ingested energy of the test meal. The latter approach (iii) used calculations as follows:TEF (%)=[(postprandial EE-preprandial EE)/EI] xl 00 (Eqn. 5)where postprandial EE and preprandial EE (kcal.h’) represent the mean energy expendedpostprandially and preprandially over 1 hour. The sum of the difference during 5 hourmeasurement divided by ingested energy was the percentage of meal energy production.3.4.3. Evaluation ofPostsurgical Energy BalanceThe following equation was used to evaluate the energy balance in each obese subject.EE = EI-[Body Stores+Fecal and Urinary Energy] (Eqn. 6)The data obtained from the DLW measurement was used to estimate energy expenditure.Energy intake was assessed using 5-day weighed food records. Body stores weredetermined by changes in body composition over a 2-week period divided by the numbersof days. The fecal and urinary energy levels were assessed by bomb calorimetry.523.5. Statistical AnalysesResults were analyzed using a statistical software package (Systat, Version 4.0, Evanston,IL, 1989). Results were expressed as means±SEM. Differences with pO.O5 wereconsidered to be significant.Repeated measures ANOVA with the subsequent Bonferroni multiple comparison wasused to compare energy intake determined by food records, BEE, TEF, and bodycomposition data over the presurgical and 3 postsurgical periods (Ott 1988). Comparisonof isotopic technique, BIA and DEXA measurement of body mass were performed usinglinear regression analysis and pairwise comparisons between corresponding variables fromthe three different methods.The evaluation of reported energy intake and adjusted energy intakes using changes inbody weight and composition was described by the following equation:(EI/EE)x 100 and (EE-EI)EEx 100 (Eqn. 7)The mean of reported energy intake from the food records was compared to the mean ofenergy expenditure measured by DLW method using a paired t-test and correlationcoefficient. Relationships between underestimation and body weight as well as otherrelated variables such as height and age were assessed using simple and multiple linearregression analysis.53544. RESULTS4.1. Examination of the Mechanism of Weight Loss in Obese Subjects FollowingIleogastrostomyTable 2 showed subjects’ overall completion of the study protocol in the obese subjects’study. Of the 16 obese subjects participating in the study, 10 subjects (1 M, 9F) completedthe presurgical metabolic tests and 7 subjects (7F) finished the tests over the 3postsurgical time points. The other 3 subjects did not undergo metabolic measurementsafter the first observation due to discomfort under the hood and time involved. From thenon, metabolic tests and 3-month body composition measurement were stopped to ensurethe completion of TEE measurement. The DLW method was perfonned in 10 subjectspostsurgically. Of these 10 subjects having completed TEE measurement, six finished thecollection of fecal and urine materials. Nine subjects took part in the presurgical total bodyscan using DEXA and eight completed the postsurgical measurement.Physical characteristics and body composition in the obese subjects before surgery areshown in Table 3. The average age was 36±2 years with a range of 20 to 49. We acceptedpatients with a BMJ above 35 kg.m2’and many were much heavier. Their mean bodyweight and BMI were 121.9±4.1 kg and 45.6±1.1 kg.m21,respectively. FFM and FMdetermined by IDS method were 62.9±2.2 kg and 59.0±2.8 kg, and FM representedabout 48% of body weight in this group of subjects.55Table 2. Participation in Study Protocol in Obese StudySubjects BEE and IDS for BIA for El TEE FE and UR DEXTEF body body measurement A1 yes yes yes yes yes yes2 yes yes yes yes yes yes3 yes yes yes yes yes yes4 yes yes yes yes yes yes5 yes yes yes yes6 yes yes yes yes yes yes7 yes yes yes yes yes yes8 yes* yes* yes*9 yes* yes* yes*10 yes* yes* yes yes*11 yes* yes* yes yes*12 yes* yes* yes yes*13 yes* yes* yes yes*14 yes* yes* yes*15 yes* yes* yes*16 yes* yes* yes*The first seven subjects completed corresponding measurements prior to and at 1, 2 and 3months following surgery.*Subjects only participated in the comparison of IDS and BIA methods with DEXA inassessing body composition before and at 6 weeks after surgery.“Yes” represents participation in the corresponding measurements. Symbols used areBEE, basal energy expenditure; TEF, thermic effect of food; El, energy intake; EE, energyexpenditure; FE, fecal energy; UE, urinary energy; DEXA, dual energy x-rayabsorptiometry.56Table 3. Physical Characteristics of Presurgical Obese SubjectsSubjects Sex Age HT BW BMI FFM FM FMJBW2-1(ys) (cm) (kg) (kg.m ) (kg) (kg) (/o)1 F 48 155 115.5 48.1 63.49 52.01 45.032 F 28 163 100.5 37.8 51.82 48.68 48.443 F 22 157 118.5 48.1 54.10 64.40 54.354 F 26 152 101.1 43.8 55.32 45.78 45.285 F 36 165 116.5 42.8 61.08 55.42 47.576 F 48 157 105.5 42.8 59.44 46.06 43.667 F 40 168 120.0 42.5 73.63 46.37 38.648 M 49 184 172.0 50.8 87.62 84.38 49.069 F 29 169 133.5 46.9 64.22 69.29 51.9010 F 38 157 108.5 44.2 54.31 54.19 49.9511 F 34 171 127.0 43.4 61.29 65.71 51.7412 F 40 166 125.5 45.7 64.55 60.95 48.5613 F 36 171 120.0 41.0 67.37 52.63 43.8614 F 40 156 119.0 48.9 56.59 62.41 52.4515 F 42 161 121.5 47.1 61.43 59.57 49.0316 F 20 161 146.0 56.3 69.54 76.46 52.37Mean 36 163 121.9 45.6 62.86 59.02 48.24SEM 2 2 4.4 1.1 2.22 2.80 1.03FFM and FM are the data derived from IDS method.57The postoperative complications occurred in this group of subjects are presented inAppendix 8. The most common complications were nausea, diarrhea and infection.Diarrhea was reported in virtually all patients in the early postoperative period, butthereafter usually subsided.4.1.1. Changes in Body Composition Following Ileogastrostomy4.1.1.1. Weight LossAll patients lost weight following surgery. Approximately 15% of preoperative bodyweight was lost during the 3-month period for the 7 subjects having completed the 3-month measurement (p<O.0001). Mean body weight decreased monthly by 5.2, 4.2 and7.8 kg at 1, 2 and 3 months postsurgically. However, the weight reduction in each monthwas not statistically significant. Influence of Ileogastrostomy on Body Composition Measured by Isotope DilutionMethod and Bioelectrical Impedance AnalysisBoth FFM and FM measured by IDS method and BIA showed a significant decline(p<O.0001, repeated measures ANOVA) across the 3 months (Table 4). However,multiple comparison did not show a significant difference in FFM among each month.When FFM and FM were expressed as the percentage of body weight during the 358Table 4. Influence of Ileogastrostomy on Body Composition Determinedby Isotope Dilution Method and Bioelectrical Impedance AnalysisPresurgery I mo. 2 mo. 3 mo.Body weight (kg) 111.1±3.2a 105.9±3.Oa lOl7±24ab 93.±3lbFFM-IDS (kg) 59.8±2.8 57.1±2.7 55.6±2.8 52.9±3.4FFM-BIA (kg) 55.4±1.7a 51.9±14th 5O.3±lth 48.2±1 6bFM-LDS (kg) 51.2±2.6a 48.8±2.2 46.l±23th 410±2bFM-BIA(kg) 55.2±1.5a 53.±l.7th 5O.7±1.3’° 45.9±1.5°FFM-IDS/BW(%) 53.9±1.8 53.9±1.8 54.6±2.1 56.2±2.5FFM-BIAJBW(%) 50.5±0.4 49.3±0.5 49.8±0.6 51.2±0.5FM-IDS/BW(%) 46.1±1.8 46.1±1.8 45.4±2.1 43.8±2.5FM-BIAIBW(%) 50.0±0.4 50.7±0.5 50.2±0.6 48.8±0.5Values are mean±SEM (n=7). Post hoc comparison was made between columns andmeans in each column not sharing a common superscript letter are significantly different(p<O. 05).Presurgery is the presurgical measurement and 1 mo., 2 mo. and 3 mo. are themeasurements at 1, 2 and 3 months following surgery, respectively.59months, these percentages of both FFM and FM measured by IDS method reachedborderline statistical significance (p=0. 08). Differences in Body Composition Assessment by Isotope Dilution Method andBioelectrical Impedance AnalysisThere was a significant difference (p<O.05) in the detection of FFM and FM using IDSmethod versus BIA (Table 5). The FFM was lower and FM higher measured by BIA thanthose by IDS method throughout the study period. The differences in FM were 3.98±1.76kg presurgically and 4.69±1.71 kg, 4.60±1.95 kg and 4.88±2.32 kg at 1, 2 and 3 monthspostsurgically, respectively. The differences in FFM were 4.48± 1.86 kg presurgically and5.24±1.79 kg, 5.25±2.07 kg and 4.67±2.32 kg at 1, 2 and 3 months postsurgically,respectively. Changes in FFM and FM following surgery and the percentage of thesereduced FFM and FM are also listed in Table 5. The monthly reduction of FFM and FMwere not significantly different between the two methods (Table 5), while the methodsdiffered significantly in assessing the absolute amount of FFM and FM. From IDS methodan average of 48.1, 64.6 and 71.0 % of weight loss was found to be body fat at 1, 2 and 3months postsurgically. A slightly lower but insignificant loss of body fat (32.7%) by BIAmethod was found at the first month compared with that obtained by the IDS method. Thepercentages of fat loss at the second and third months were quite similarto the percentage of fat loss determined by the IDS method.60Table 5. Changes in Fat-free Mass and Fat Mass Measured by IsotopeDilution and Bioelectrical Impedance Analysis MethodsIDS BIATime after 1 mo. 2 mo. 3 mo. 1 mo. 2 mo. 3 mo.surgexyFFM (kg) 57.1±2.7 55.6±2.8 52.9±3.4 51.9±1.4* 50.3±1.3* 48.2±1.6FM (kg) 48.8±2.2 46.1±2.3 41.0±2.4 53.5±1.7* 50.7±1.3 45.9±1.5Bodyfat(%) 46.1±1.8 45.4±2.1 43.8±2.5 507+05* 50.2±0.6* 48.8±0.5AFFM (kg) 2.7±0.6 1.5±0.5 2.7±1.0 3.5±0.7 1.5±0.3 2.1±0.5EFM (kg) 2.5±0.6 2.7±0.7 5.1±0.9 1.8±0.5 2.8±0.0.6 4.8±0.8AFFWABW 51.9±10.2 35.4±8.9 29.0±10.5 67.3±10.5 35.0±8.3 29.1±1.64(%)AFMJABW 48.1±10.2 64.6±8.9 71.0±10.5 32.7±10.5 65.0±8.3 70.9±1.6(%)Values are mean±SEM (n=7), compared between the two methods at correspondingmonths and statistical significance was symbolized by stars with *p<O.O51 mo., 2 mo. and 3 mo. are the measurements at 1, 2 and 3 months following surgery,respectively. ABW, AFFM, AFM represent the changes in body weight, fat-free mass, fatmass estimated by the difference of corresponding values with the previous measurement.614.1.2. Comparison of Isotope Dilution and Bioelectrical Impedance AnalysisMethods with Dual Energy X-ray Absorptiometry Measurement4.1.2.1. Dual Energy X-ray Absorptiometrv MeasurementThe individual data for the nine subjects who took part in the DEXA measurement arepresented in Table 6. All subjects completed the. pre- and postsurgical measurementsexcept subject 10, due to lack of interest in the study. On average the percentages of bodyfat before and after surgery were 51.2 and 50.8%, respectively. Both FFM (LM+BMC)and FM declined significantly with and without the male subject (p<O. 01) followingsurgery (Table 6). Comparison of Isotope Dilution Method and Bioelectrical Impedance Analysiswith Dual Energy X-ray Absorptiometrv in Assessing Body CompositionMeasurements of FFM and FM by the three techniques are shown in Table 7. Body weightmeasured by DEXA with and without the male subject correlated highly significantly withthat measured by scale (r=0. 989 and 0.991 with and without the male subjectspresurgically; r=0. 984 and 0.967 with and without the male subject postsurgically,p<O.0001). Significant differences in body weight estimates existed before surgery(p=O.O44 and 0.009 with and without the male subject, respectively). Inter-methodcomparisons showed that FFM measured by the three methods was significantly62Table 6. Individual Data of Dual Energy X-ray AbsorptiometryMeasurement Before and After IleogastrostomyPresurgery 6 wksSubjects BMC LM FM BMC LM FM(kg) (kg) (kg) (kg) (kg) (kg)8 3.534 68.406 79.167 3.266 65.300 72.7589 3.396 57.742 67.969 3.337 49.296 64.52510 2.853 47.652 56.46211 3.356 53.788 65.176 3.344 46.237 58.22612 3.352 56.024 62.897 3.465 51.999 60.76813 3.424 63.272 52.235 3.376 55.489 45.34214 3.313 52.263 61.717 3.246 52.798 55.39215 3.160 58.112 60.023 3.034 54.307 53.16216 3.161 63.989 72.349 3.060 59.416 68.280Females & MaleMean 3.283 57.916 64.405 3.266 54355** 59.807**SEM 0.067 2.153 2.852 0.053 2.097 3.100FemalesMean 3.252 56.605 62.354 3.266 52.792** 57.956**SEM 0.067 1.936 2.247 0.062 1.613 2.872Presurgery and 6 wks represent measurement time before and at 6 weeks after surgery.Pre- and postsurgical corresponding measurements were compared with **p<OO1Symbols are BMC, bone mineral content; LM, lean mass; FM, fat mass. FFMLM+BMC.63Table 7. Comparison of Isotope Dilution Method and BioelectricalImpedance Analysis With Dual Energy X-ray Absorptiometry inAssessing Body Composition Before and After IleogastrostomyIDS BL& DEXAPresurgeryFemales & MaleBW (kg)FFM (kg)FM (kg)FFMIBW (%)FMJBW (%)FemalesBW (kg)FFM (kg)FM (kg)FFMIBW (%)FMJI3W (%)PostsurgeryFemales & MaleBW (kg)FFM (kg)FM (kg)FFMII3W (%)FMJBW (%)FemalesBW (kg)FFM (kg)FM (kg)FFMIBW (%)FMIBW (%)130.3±6.3 a65.2±3.2a65. 1±3.450. 1±0.9a49.9±0.9”125. 1±3.9a62.4±1. 8a62.7±2.8”50.0±1 .Oa50.0±1.0”**1 17.2±5.3a*59524a57.7±3.751.0±1.4a49.0±1.4”* *1130±3 9a*579±2 1L* * 55.1±3. 1b51.3±1.6a48.7±1.6”130.3±6.3 a63.5±3.0’’66.8±3.5a48.8±0.6”51 .2±0.6a125. 1±3.9a61.0±1.9”64. 1±2.4a48.8±0.7°51 .2±0.7a1 17.2±5.3a**557±2 1°**615±34a47.6±O.7’52.4±0.7a**1 13.0±3.9L**54 1±1.6”* * 58.9±2. 5a47.9±O.7”°52. 1±0.7’125.6±4.51361.2±2.2064.4±2.9”48.8±1.0”51.2±1.Oa122.2±3.3’’59.9±2.0bc62.3±2.2l3049.0±1.51.0±1.1 ab**1 17.4±4.4a**5762b**598±3 1b49.2±1 3ab50.8±1 3b**1 14.0±3.3a**56016ab**58029ab49.1±1 5b50.9±1 5abValues are mean±SEM (n=9 with and n=8 without the male subject before surgery andn=8 with and n=7 without the male subject after surgery). Comparison was made betweenthe methods and symbolized by different superscript letters. Comparison betweenpresurgical and postsurgical measurements was symbolized by stars at the upper left sidewith *p<005, **p<001BW, FFM and FM are the body weight, fat-free mass and fat mass presurgically and at 6weeks postsurgically. Body weight for DEXA are the sum of LM, FM and BMC.64different (p<O.O5) before and after surgery, when the male subject was included. Thesignificant differences in FFM estimates before surgery were observed between the IDSand the other methods, when the male subject was excluded (Table 7). Estimates in FFMbetween the IDS and BIA methods still persisted after surgery, while the estimates in FFMby DEXA did not showed significant differences from those by the IDS and BIA methodspostsurgically.Presurgical %BF determined by IDS method was significantly lower than that obtained byDEXA (p=O.033) and by BIA (p=O.007). However, the postsurgical %BF measured byIDS method was similar to that obtained by DEXA, but significantly different from that byBIA method (p=O.029). The estimates in %BF described above were the results when themale subject was included. The significant differences in %BF estimates persisted betweenthe IDS and BIA methods before and after surgery (p=O.Ol7 and 0.0 16, respectively),when the male subject was excluded. The pre- and postsurgical estimates in %BF byDEXA were not significantly different from those by the BIA and IDS methods. Allmethods for determination of FFM and FM including the male subject correlatedsignificantly with each other before and after surgery (p<O.OO1 for FM and p<O.Ol forFFM. There were stronger correlations between body fat determined by the three methodsbefore surgery than that after surgery (DEXA vs IDS: presurgical r0.858, pO.OO3,postsurgical r=0.769, p=O.O43; DEXA vs BIA, presurgical r=0.889, p=O.OO1, postsurgicalr=0.822, p=O.O23; BIA vs IDS, presurgical r=0.963, p<O.0001, postsurgical r0.861,p=O.O06). The similar correlations in body fat estimates by the three methods were65observed, when the male subject was excluded (DEXA vs IDS: presurgical r=0.981,p<O.000l, postsurgical r0.891, p=O.0l7; DEXA vs BIA, presurgical r=0.965, p<O.000l,postsurgical r=0.948, p0.004; BIA vs IDS, presurgical r=0.996, p<O.0001, postsurgicalr=0.939, p=O.002). Regional Changes in Body Composition Following IleogastrostomyThe distribution of BMC, LM and FM prior to and after ileogastrostomy is shown inTable 8. There were no significant differences in BMC in other parts of the body beforeand after surgery except in the trunk, where BMC declined significantly (p=0. 015) aftersurgery. LM dropped significantly in the head (p=O.O22), the trunk (p=0.OO4) and the legs(0.001). There was a significant decline in FM in the head (pO.032), the trunk (p<O.0001)and the legs (0.016) but not in other parts of the body. The similar results were observedwhen the male subject was excluded (Table 8). Changes in Body Composition Measured by Isotope Dilution Method,Bioelectrical Impedance Analysis and Dual Energy X-ray AbsorptiometrvThe changes in body composition and the percentage of the reduced FFM and FMfollowing surgery were compared among the three methods. The changes in body weight (z\BW) assessed by DEXA were not significantly different from those determined by the66Table 8. Changes in Regional Body Composition Measuredby Dual Energy X-ray AbsorptiometryPresurgery PostsurgeiyBMC LM FM BMC LM FMFemales & MaleHead 0.54±0.02 3.80±0.12 1.13±0.07 0.55±0.02 3.57±0.14* 0.95±0.07*Trunk 1.34±0.03 28.92±1.16 37.14±2.06 1.22±0.04* 26.98±1.32** 34.05±2.31**Abdomen 0.61±0.02 12.92±0.67 18.88±1.39 0.58±0.02 12.48±0.79 17.45±1.31Arms 0.23±0.03 3.93±0.29 5.09±0.39 0.30±0.01 3.95±0.18 5.05±0.33Legs 1.19±0.04 21.27±0.94 20.87±1.38 1.20±0.03 19.86±0.66** 19.75±1.27*FemalesHead 0.54±0.02 3.73±0.11 1.09±0.07 0.56±0.02 3.47±0.11* 0.90±0.06*Trunk 1.33±0.03 28.24±1.07 35.86±1.83 1.23±0.04* 26.22±1.25** 32.79±2.24**Abdomen 0.60±0.02 12.36±0.40 17.85±1.07 0.57±0.03 11.92±0.64 16.59±1.14Arms 0.23±0.03 3.92±0.32 4.98±0.43 0.30±0.01 3.80±0.11 4.92±0.35Legs 1.16±0.04 20.72±0.87 20.41±1.48 1.19±0.03 19.31±0.43** 19.34±1.38*Values are mean±SEM (n=9 with and n=8 without the male subject before surgery andn=8 with and n=7 without the male subject after surgery). comparisons were madebetween columns for corresponding variables (paired t-test). *p<O 05, *67scale with and without the male subject (Table 9). Changes in FFM and FM as assessed byIDS and DEXA were more similar to one another than they were to changes as assessedby BIA. There were no significant differences in the percentage of reduced FFM and FMdetermined by DEXA and IDS method. However, BIA showed a higher loss in FFM thanDEXA and IDS method and a lower loss in FM than IDS method. The similar results werefound when the male subject was excluded (Table 9). Linear regressions for the reducedFFM and FM by each pair of the methods showed that high intercorrelations between BIAand DEXA estimates in the reduced FFM (r=0.938, p=O.006 without the male subject) butnot in FM. When the male subject was included, the relationship between BIA and DEXAestimates in the reduced FFM was not significant. Overall, poor correlations in theestimates of reduced FFM and FM by the three methods were observed in the presentstudy.4.1.3. Changes in Energy Expenditure Following Ileogastrostomy4.1.3.1. Basal Energy Expenditure and Fasting Nutrient OxidationBasal energy expenditure and RQ prior to and at 1, 2 and 3 months are shown in Figure 7.When expressed in absolute values, BEE were 1.158±0.065 kcal.min’ presurgically and1.097±0.055, 1.074±0.047, 1.040±0.039 kcal.min’ at 1, 2 and 3 months postsurgically,respectively (Table 10). As can be seen, at any time during the study, BEE tended to68Table 9. Changes in Fat-free Mass and Fat Mass Determined by IsotopeDilution, Bioelectrical Impedance Analysis and DualEnergy X-ray Absorptiometry MethodsIDS BIA DEXAFemales & MaleABW(kg) 12.7±2.1 12.7±2.1 10.5±1.2AFFM(kg) 5.8±l.9b 7.8±1.5a 4.9±1.1’°AFM (kg) 6.9±0.5’ 4.9±O.8b 5.6±O.7a1tFFMJBW(%) 38.9±8b0 60.0±4.9a 44.l±8.8’AFMIABW (%) 61. 1±8.8a 40.0±4.9°FemalesABW(kg) 10.8±1.0 10.8±1.0 10.4±1.4AFFM(kg) 4.3± 6.6±1.OaAFM(kg) 6.6±0.5a 43±06 5.2±0.8aAFFMIABW (%) 35. i±.2b 59.4±5.6aAFMJABW (%) 64.9±9.2a 40.6±5.6c 53.8±9.9w’Values are mean±SEM (n=9 with and n=8 without the male subject before surgery andn=8 with and n=7 without the male subject after surgery) and compared between columnsfor corresponding variables (paired t-test). Means not sharing a common superscript letterwere statistically significant.ABW, AFFM and zFM are the differences between pre- and postsurgical correspondingmeasurementsMonthsFigure7.Basalenergyexpenditureandrespiratoryquotientpresurgicallyandat1,2and3monthspostsurgically(n=7).Errorbarsare+SEM.Therewerenosignificantdifferencesbetweenpresurgicalandpostsurgicalmeasurements.BEERQ1.51.4ct C.)-a) D C a) 0 >< w1.2> a) C 1.1ctS C,)11 0.9C G).I-J D 0 > 0 :2 0.50123a) (070decrease following surgery. The BEE declined by 0.06 1, 0.022 and 0.034 kcal.min’ duringeach of the first, second and third months following surgery, respectively. The decrease inBEE was in parallel to the reduction of body weight and energy intake (Appendix 9).However, the changes of BEE were of borderline statistical significance p=O.O8) Becauseof the weight loss in this group of subjects, the changes in body weight and compositionmay affect the results. The analysis of covariance was applied to the BEE data with BW,FFM and FM as the covariates. The adjusted BEE was still not significantly different. Thepreprandial RQ level increased slightly but insignificantly following surgery (Figure 7).There were no changes in the utilization of carbohydrate (p=O. 838) and fat (pO. 628) in thepreprandial state (Table 10). Changes in Thermic Effect of FoodFigure 8 shows the time course of presurgical TEF expressed as a percentage over BEEand a percentage of ingested energy in nine obese women. The net increase in energyexpenditure after meal ingestion, expressed as a percentage above the BEE was 18.08±3.19, 19.93±3.70, 21.40±3.59, 19.41±3.19 and 14.40±3.33 % at 1, 2, 3, 4 and 5 hourspostprandially, respectively. The corresponding values for ingested energy were 2.08±0.37, 2.23±0.37, 2.45±0.35, 2.22±0.32 and 1.64±0.36 %, respectively (Figure 8). Thepattern of the thermic response curve showed that the morbidly obese subjects reachedtheir peak energy expenditure in the third hour after meal ingestion. At the end of themeasurement (t=3 00 minutes) energy expenditure was still higher than baseline values.71Table 10. Basal Energy Expenditure and Thermic Effect of Food,Expressed as Absolute Amount, Percentage of BasalEnergy Expenditure and Ingested EnergyPresurgery 1 mo. 2 mo. 3 mo.Testmealenergy 599±13.8a 3S9±63.Ob 424±55.O 599±13.8a(kcal)BEE (kcal.min’) 1.158±0.065 1.097±0.055 1.074±0.047 1.040±0.039BEE (kcal.d’) 1668±93.1 1579±79.9 1547±68.1 1498±55.9TEF (kcal.min’) 0.216±0.03 1’ 0.062±0.015” 0.054±0.01 1” 0.066±0.012”TEF (kcal.5h’) 64.9±9.2a l8.7±4b 16.1±3.3” 19.7±3.7”TEF(%mealenergy) 10.85±1.57a S.55±l.3&’ 3.96±0.90” 3.28±0.62”TEF (% BEE) 19.28±3.25a 5.65±1.30” 5.21±1.1113 6.22±1.23”PreprandialRQ 0.791±0.026 0.798±0.021 0.789±0.023 0.815±0.023PostprandialRQ 0.884±0.024a O.8ll±O.Ol8a 0.765±0.027” 08260016abValues are mean±SEM (n7). Comparison was made between months and the significancewas symbolized by different letters.Presurgery is presurgical measurement and 1 mo., 2mo. and 3 mo. are the measurementsat 1, 2 and 3 months following surgery, respectively.Symbols used are BEE, basal energy expenditure; TEF, thermic effect of food; PreprandialRQ is the average of RQ during 30 mm BEE measurement and postprandial RQ is theaverage of RQ during 5 h TEF measurement.30 15 10 010 0TEF/BEETEF/EIo- LU LU m LU F- U)-o ci) U) U) a) >< LU U- LU ‘—58LU U- LU6F— U)-D a) U)AU)‘+a) >< Lii LL1)LU I—12345Figure8.BEEandHoursThetimecourseofpresurgicalTEF(n=9),expressedasthepercentageofingestedcalories.Errorbarsindicate+SEM.I’.)73When TEF was expressed as absolute amount of energy over BEE, the presurgical TEFwas significantly higher than those postsurgically (p<O.0001). Posthoc comparisonsrevealed that the difference between the values was limited to the presurgicaldetermination, which was significantly greater (p<O.OOl) than each of the othermeasurements (Table 10). TEF values during the postsurgical study period were stabilizedand were not significantly different from each other. In order to calculate the postprandialthermogenesis, the cumulative energy expenditure increment above the premeal baselinewas calculated over 300 minutes. This integrated value divided by the energy content ofthe test meal were 10.85±1.57, 5.55±1.38, 3.96±0.90 and 3.28±0.62 % prior to and at 1,2 and 3 months, respectively (Figure 9). Because the subjects could not consume all fooditems of the test meal as they did on presurgical test, the percentage of TEF was analyzedusing ACOVA to adjust for the influence of meal composition. After this adjustment,presurgical TEF values were still significantly higher (pO.OO1) than those followingsurgery. The thermic responses, expressed as a percentage increase overbaseline energy expenditure dropped significantly following surgery (p<O.0001). Therewas no difference in preprandial RQ across the study period, however, postprandial RQfell significantly (p=O. 001). Changes in Fat and Carbohydrate OxidationFat and carbohydrate (CHO) oxidation rates (assuming constant oxidation for protein 0.7g. kg’ FFM.d’) during preprandial and postprandial measurements are presented in12o-. Cl)ci) 0 o8ci) Co ci)**w cD4 > (‘5 D E2D C)002MonthsFigure9.Cumulativethermiceffectoffoodfor5hours,expressedasthepercentageofingestedcalories(n=7).Errorbarsindicate+SEM.*1..movaluesignificantlydifferentfrompresurgicalvalue(p<O.05).**correspondingmcmeasurementsignificantlydifferentfrompresurgicalmeasurement(P<0.01).1375Table 11. There were no significant changes in preprandial oxidation of fat and CHOfollowing surgery. However, the postprandial oxidation of carbohydrate dropped offsignificantly (p=O.OO2) across the study period (Table 11). Because there was theconfounding effect of reduced macronutrient consumption and absorption, thepostprandial changes in fat and CHO oxidation could be indirect evidence of reducedintake and malabsorption of these nutrients. The relative oxidation of these two nutrientsexpressed as a percentage of the individual amounts ingested is also listed in Table 11.After surgery, obese women oxidized a greater proportion of the fat ingested than beforesurgery (p=O.O15). The percent of CHO oxidation was significantly lower than that beforesurgery (p=O.O5, repeated measures ANOVA). The difference was not observed whenmutiple comparisons were performed. Total Energy Expenditure in Bypassed Obese SubjectsThe isotopic and TEE data for obese subjects are summarized in Table 12. The means of180 and 2D elimination rates were 0.0927±0.0058 and 0.0708±0.0058 (pool.d’)respectively. Total body water volumes on day 0 and day 14, measured by deuteriumdilution method, were 41.21±1.45 and 40.44±1.47 kg, respectively. The body water pooldropped by 0.77 kg during the 14-day study period. Carbon dioxide production ratesvaried considerably, ranging from 17.41 to 28.14 (mol.d1) which resulted in largedifferences in TEE among subjects. Similarly, a considerable difference in energy76Table 11. Pre- and Postprandial Fat and Carbohydrate Oxidationin The Obese Subjects Before and After IleogastrostomyPresurgery 1 mo. 2 mo. 3 mo.PreprandialFat (g/min) 0.091±0.017 0.080±0.008 0.088±0.009 0.074±0.008CHO(g/min) 0.075±0.025 0.084±0.019 0.065±0.018 0.088±0.018Ingested MealFat(g) 23.30±0.54 13.97±2.45 16.51±2.14 23.30±0.54CHO (g) 82.38±1.89 49.41±8.67 58.36±7.57 82.38±1.89PostprandialFat(g/5h) 24.34±9.12 23.98±2.67 30.14±2.81 20.29±1.19CHO (g/5h) 60.66±8.28a 2980524b 2443±95b 32.44±5.67Oxidized/IngestedFat(%) lOS.56±40.SOb 208.91±46.36a 199.44±28.82a 8748588bCHO (%) 74.01±10.56 64.58±13.16 44.42±14.49 39.61±7.12‘Values are mean±SEM (n=7). Comparison was made between columns and means ineach column not sharing a common superscript letter are significantly different (p<O. 05).Presurgery is the presurgical measurement and 1 mo., 2 mo. and 3 mo. are themeasurements at 1, 2 and 3 months following surgery, respectively.Preprandial fat and CHO are the average of oxidized fat and carbohydrates during 30-minute BEE measurement and postprandial fat and CHO are the cumulative oxidized fatand CHO during 5-hour TEF measurements.77Table 12. Individual Data of Total Body Water, EliminationRates and Total Energy Expenditure in Obese SubjectsSubjects K2 K18 TBW-0 TBW-14 FQ rC021 TEEfFFyI TEEfB\(pool.d1) (pool.d1) (kg) (kg) (mole.d ) (kcal.d ) ( ) ( )1 0.0647 0.0903 42.50 42.86 0.842 28.17 3618 62.04 33.502 0.0525 0.0816 37.12 35.88 0.872 27.42 3423 68.65 35.553 0.1096 0.1325 39.50 38.32 0.861 23.02 2903 54.61 26.364 0.0494 0.0684 37.09 35.94 0.903 17.84 2165 43.41 22.716 0.0899 0.1081 39.69 38.66 0.858 18.44 2332 43.57 24.137 0.0709 0.0908 48.44 47.37 0.845 24.65 3154 48.20 28.8910 0.0811 0.1073 35.73 35.22 0.891 23.96 2939 60.66 30.2611 0.0599 0.0811 39.01 38.43 0.860 21.15 2669 50.46 23.5712 0.0609 0.0830 47.90 47.09 0.797 27.02 3626 55.88 30.5013 0.0687 0.0838 45.12 44.58 0.885 17.41 2148 35.06 20.15Mean 0.0708 0.0927 41.21 40.44 0.861 22.91 2898 52.27 27.56SEM 0.0058 0.0058 1.45 1.47 0.009 1.28 178 3.20 1.58Symbols used are K2 and K18, elimination rates of deuterium and 180 respectively; TBWo and TBW-14, total body water at the begining and end of the 14-day experimentalperiod; FQ, food quotient; rCO2, rate of CO2 production; TEE, total energy expenditure;FFM, fat-free mass; BW, body weight.78expenditure per FFM and BW existed in this group of subjects. Energy expended on thebasis of FFM and BWwere 52.27±3.20 and 27.56±1.58’, respectively. The Components of Total Energy Expenditure in Obese Subjects FollowingIleogastrostomyOf the 7 subjects completing metabolic tests, one could not conduct the TEEmeasurement using DLW method, due to the inconvenient transportation. The TEE wascarried out just before the second month metabolic test. The BEE and TEF at the secondmonth were used to estimate the percentage of energy cost for physical activity during thisperiod (Figure 10). The physical activity index calculated from the total energyexpenditure divided by resting energy expenditure (BEE+TEF) was 1.79±0.14 (mean±SEM) at the second month after ileogastrostomy.4.1.4. Changes in Energy IntakeEnergy intakes based on 3-day food records fell significantly (p<O .0001). Presurgically,the energy intake averaged 2022±295 kcal.d’ and there was a marked drop in reportedenergy intake by 629, 772 and 842 kcal.d’ during each of the first, second and thirdmonths following surgery, respectively. The postsurgical energy intakes at 1, 2 and 3months were relatively stable (Table 13). The decrease in food consumption after surgerywas also shown by the substantial reduction in the intake of different energy-containing79Percentage of Total Energy Expenditure for BasalEnergy Expenditure, Thermic Effect of Food andEnergy Cost of Physical Activity- 44% - ----.--Figure 10. Hourly energy expenditure for basal metabolism (BEE),food thermogenesis (TEF) and energy expended for physicalactivities (EE for activity) at the second month followingileogastrostomy (n=6). Numbers are the relative percentage of eachcomponent in total energy expenditure.BEEEE for Activity80Table 13. Changes in Energy Intake Following IleogastrostomyPresurgery 1 mo. 2 mo. 3 mo.Energy Intake (kcal.d1) 2022±295a 1393±2olb l251±257b 1180±74bFat(g) 85.1±20.7a 49.l±S.5b 51.9±13SbProtein(g) 77.7±8.7a 61.6±8.5 53.±9Ob 49.7±6.4”CHO (g) 236.3±27.9a 176 2±26 7ab l42.5±27.7’ l45.8±2O.bPercentage of Energy as 35.8±3.3 30.9±2.1 35.1±3.0 3 1.8±3.6Fat (%)Percentage ofEnergy as 16.1±1.3 18.1±1.3 17.9±1.8 17.4±1.6Protein (%)Percentage of Energy as 48.0±3.2 51.0±2.1 47.1±3.5 50.8±4.0CHO (%)‘Values are mean±SEM (n=7). Comparison was made between columns and means ineach column not sharing a common superscript letter are significantly different (p<O.O5).Presurgery is the presurgical measurement and 1 mo., 2 mo. and 3 mo. are themeasurements at 1, 2 and 3 months following surgery, respectively.81nutrients There were no significant changes in the distribution of energy consumed asprotein, fat and carbohydrate at the various times examined.4.1.5. Factors Associated With Weight Loss Following IleogastrostomyTable 14 presents individual data for energy balance study. Energy intakes were based on5-day food records reported by each subject who participated in the balance study. Fecaland urinary energy losses were derived from 6 subjects as described above. Total energyexpenditure was determined using DLW technique between 6 and 8 weeks followingileogastrostomy. On average energy expenditure was 2898±178 kcal.d’ and energy intakewas 1625±207 kcal.d1.Energy loss estimated by the changes in both FFM and FMmeasured by the IDS method was 929±157.2 kcal.d1.There was a close positivecorrelation between the energy loss and energy expenditure (r=0. 719, p=O.Ol 9, n= 10)during the DLW period (Figure 1 1A,B). On the contrary, energy intake showed a positivecorrelation with energy loss (r=0.582, p=0.O78, n=l0) and this relationship approachedsignificance. There was clear evidence of malabsorption as determined by 5-day fecalenergy measurement (Table 14). The amount of energy loss during the 14-day studyperiod may have been related to the energy content in feces (r=0.808, p’=O.O52)but wasnot correlated to the urinary energy loss (r0. 011). The energy lost in feces accounted for65.3% of variance in body energy loss obtained from the loss of FFM and FM during 14-day study period. However, the numbers were too small to draw justifiable conclusions.The correlation matrix for these variables in the 6 subjects who completed the energy82Table 14. Individual Data of Energy Balance in theObese Subjects Following IleogastrostomySubjects Weight loss Energy Loss RET TEE FE UE(kg. 14-d) (kcal.d’) (kcal.d’) ( (kcal.d’) (kcal.d’)1 -2.8 -1745.1 1386 3618 404.8 109.72 -2.4 -525.5 1601 3423 283.3 102.43 -3.3 -1044.1 1475 2903 448.3 90.04 -1.7 -199.7 535 2165 197.1 98.36 -1.7 -278.8 1042 2332 289.7 115.37 -3.2 -1068.1 2531 3154 436.5 105.010 -2.5 -1036.2 1363 293911 -2.5 -986.8 2305 266912 -3.2 -1552.2 2564 362613 -2.3 -853.7 1454 2148Mean -2.6 -929.0 1626 2899 343.3 103.5SEM 0.2 157.2 207 178 41.4 3.6RET, TEE, FE and UR represent reported energy intake, total energy expenditure, fecalenergy and urinary energy, respectively. Weight Loss was determined by scale.Energy loss was obtained from the loss of both FFM and FM during the 14-day period.Assuming that adipose tissue consists of 20% water and that each gram of fat stored has acaloric equivalent of 9.5 kcal.g’and each gram of protein has a caloric equivalent of 4.23kcal.g’, lg of FM equals 7.6 kcal amd lg of FFM equals 1.1 kcal because FFM contains26.8% of protein.830 I1500 2000 2500 3000 3500 4000 4500A •(‘)U) -.2 . .IS.5Energy expenditure (kcal,’d)0 I I1 C 1500 2000. 2500 3000 3500 4000 4500•0-4000SC,-800U)U) .0—-1200>0)-1600-2000Energy expenditure (kcal/d)Figure 11. Correlation between energy expenditure and weight loss (1 1A:r=O.675, p=O.032) as well as energy loss (1 1B: r=O.719, p=O.019) (n=lO).84balance study is presented in Table 15. There was an insignificant relationship betweenenergy intake and energy expenditure (r=O.628, p=O.l68, n=6) as well as fecal energy loss(r=O.732, p=O.O84, n6). Due to the large variability and inaccuracy of energy intake inthis study, the energy intake was not closely related to the weight loss.85Table 15. Correlation Coefficients Between Energy Loss and EnergyExpenditure, Fecal Energy, Urinary Energy and Energy IntakeEE FE UE ElFE 0.589(0.204)UE 0.054 -0.156(0.921) (0.763)El 0.628 0.732 0.023(0.168) (0.084) (0.966)E-loss 0.778 0.808 0.011 0.491(0.068) (0.052) (0.983) (0.323)Numbers are correlation coefficients for corresponding variables and p values are shown inbrackets (n=6).E-loss was energy lost in FM and FFM during the 14 -day period. EE, FE, UE and’ Elrepresent energy expenditure, fecal energy, urinary energy and energy intake, respectively.864.2. Validation of Reported Energy Intake Using Doubly Labeled Water in Normal-weight and Obese Subjects4.2.1. Physical Characteristics ofNormal-weight and Obese SubjectsThe physical characteristics of 26 normal-weight and 6 obese subjects are presented inTable 16, The normal-weight subjects were 48.3±1.0 years of age, mean body weight 61.7±1.3 kg, BMI 23 .4±0.5 kg.m2’, body fat 32.8±1.3%. The obese subjects were 36.0±2.8years of age, mean body weight 105.2±2.9 kg, BMI 40.3±1.0 kg.m2’, body fat 47.2±1.3%. Weight, percent 113W, and weight as fat in the obese group were all significantlygreater than those of normal-weight group. The mean body weight of the obese group was43.5 kg (70.5 %) greater than that of normal-weight group.4.2.2. Total Energy Expenditure in Normal-weight SubjectsTable 17 presents the isotopic and TEE data for normal-weight subjects. The means of180 and 2D elimination rates were 0.1158±0.0042 and 0.0903±0.004 1 (pool.d’). Totalbody water volumes on day 0 and day 14 were 30.34±0.87 and 30.34±0.88 kg,respectively. On average body weight declined by 0.138 kg and the body water pool wasmaintained during the 14-day study period in this group of subjects. Carbon dioxideproduction rates ranged from 8.64 to 24.69 (mo1.d’) with a mean value of 17.88±0.8487Table 16. Physical Characteristics of Normal-weight andObese Subjects in the Validation StudySubject Age Height Initial BW Final BW MEW BW/TBW BMI FM/MEWNormal ys cm kg kg kg kg.fn-weight1 41 166.5 71.0 69.4 70.2 110.4 25.3 38.42 45 162.6 61.4 61.2 61.3 98.6 23.2 29.63 55 163.0 54.0 54.2 54.1 87.0 20.4 36.14 54 149.7 61.0 61.5 61.3 113.0 27.3 41.15 52 170.0 61.9 61.0 61.5 92.8 21.3 31.86 45 165.0 66.2 65.4 65.8 103.5 24.2 35.87 45 166.7 61.2 60.7 61.0 96.0 21.9 30.88 45 160.5 57.1 56.7 56.9 93.5 22.1 35.19 45 166.2 56.8 58.0 57.4 90.3 20.8 26.910 57 166.5 64.5 64.2 64.4 101.3 23.2 45.311 52 164.0 57.5 57.4 57.5 92.4 21.4 38.212 50 177.0 70.0 69.0 69.5 100.7 22.2 16.213 45 155.0 62.5 62.5 62.5 107.6 26.0 28.914 47 165.5 70.0 69.5 69.8 109.8 25.5 31.515 45 158.5 59.5 59.1 59.3 99.7 23.6 19.116 54 162.5 59.5 60.0 59.8 96.1 22.6 25.017 48 162.5 52.9 53.1 53.0 85.2 20.1 26.418 46 157.3 71.4 71.1 71.3 122.7 28.8 40.719 55 170.0 75.0 74.9 75.0 113.2 25.9 38.520 54 155.3 48.7 48.7 48.7 83.8 20.2 32.721 37 161.8 58.2 58.0 58.1 95.5 22.2 29.822 45 169.0 57.4 57.8 57.6 88.7 20.2 29.823 45 153.3 58.5 57.8 58.2 102.6 24.7 42.224 54 153.2 52.0 52.9 52.5 92.5 22.3 37.625 49 159.3 63.9 63.1 63.5 106.8 25.0 34.426 45 165.5 75.0 75.0 75.0 118.0 27.4 30.5Mean 48.3 162.6 61.8 61.6 61.7 100.1 23.4 32.8SEM 1.0 1.2 1.4 1.3 1.3 2.0 0.5 1.3Obese1 48 155 109.4 106.6 108.0 195.0 45.0 46.02 28 163 97.5 95.1 96.3 161.2 36.3 48.23 22 157 111.8 108.5 110.2 194.1 44.7 51.74 26 152 96.2 94.5 95.4 175.8 41.3 47.86 48 157 97.5 95.8 96.7 170.2 39.2 44.67 40 168 110.8 107.6 109.2 165.9 38.7 40.1Mean 36.0 161.7 106.5 103.9 105.2 174.2 40.3 47.2SEM 2.8 2.2 2.7 2.5 2.6 4.4 1.0 1.388Table 17. Individual Data for Total Body Water, Elimination Ratesand Total Energy Expenditure in Normal-weight SubjectsSubjects K2-1 K181 TBW-0 TBW-14 FQ rCO21 TEEfFF1 ThEfB(pooLd ) (pool.d ) (kg) (kg) (mole.d ) (kcal.d ) ( ) (kcal.kg1 0.0932 0.1151 32.24 31.05 0.855 15.60 1962 45.38 27.942 0.0704 0.0969 31.60 31.55 0.853 19.86 2483 57.57 40.513 0.1054 0.1322 25.47 25.11 0.864 15.46 1871 54.15 34.594 0.0936 0.1230 26.66 26.14 0.870 18.18 2355 65.29 38.455 0.0857 0.1180 30.01 31.30 0.906 23.54 2869 68.49 46.486 0.0861 0.1101 31.56 30.28 0.833 17.08 2086 49.37 31.707 0.0857 0.1176 30.86 30.88 0.899 23.37 2850 67.60 46.778 0.1651 0.1960 27.10 26.99 0.897 18.14 2231 60.37 39.209 0.0954 0.1143 30.72 30.75 0.898 12.69 1620 38.58 28.2210 0.0810 0.1058 25.41 26.07 0.889 14.88 1895 53.91 29.4511 0.0820 0.1061 26.09 25.91 0.848 14.49 1808 50.90 31.4712 0.0860 0.1110 42.39 42.91 0.851 24.69 3070 52.69 44.1713 0.1134 0.1374 32.27 32.76 0.873 17.29 2108 47.45 33.7214 0.0998 0.1281 34.93 35.03 0.876 22.83 2807 58.69 40.2115 0.0777 0.1071 34.83 35.41 0.898 24.56 2999 62.50 50.5716 0.0732 0.1017 32.56 33.07 0.890 22.34 2738 61.08 45.8217 0.0923 0.1166 28.35 28.77 0.896 15.88 2005 51.38 37.8218 0.0955 0.1178 31.15 30.69 0.892 15.53 1937 45.86 27.1819 0.0717 0.0970 33.67 33.78 0.860 20.15 2496 54.21 33.3220 0.0971 0.1220 23.90 24.08 0.873 13.62 1631 49.76 33.4821 0.0769 0.0959 29.49 30.18 0.880 12.83 1557 38.20 26.8022 0.0709 0.0990 29.93 29.26 0.919 19.82 2428 60.06 42.1623 0.0551 0.0800 24.69 24.53 0.903 14.79 1817 54.05 31.2524 0.1158 0.1343 23.92 24.00 0.893 8.64 1095 33.44 20.8725 0.1021 0.1245 30.63 30.40 0.889 15.28 1960 47.03 30.8726 0.0766 0.1026 38.49 37.84 0.856 23.36 2913 55.88 38.84Mean 0.0903 0.1158 30.34 30.34 0.880 17.88 2215 53.23 35.85SEM 0.0041 0.0042 0.87 0.88 0.004 0.84 102 1.73 1.46Symbols used are K2 and K18, elimination rates of deuterium and 180 respectively; TBWo and TBW-14, total body water at the begining and end of the 14-day experimentalperiod; FQ, food quotient; rCO2, rate of CO2 production; TEE, total energy expenditure;FFM, fat-free mass; BW, body weight.894.2.3. Accuracy ofReported Energy Intake in Normal-weight SubjectsTable 18 lists the individual data of reported energy intake, changes in body weight andcomposition during the 2 week experimental period, energy intake after adjusting for bodyenergy stores and the intakes divided by expenditure in the normal-weight participants.Reported energy intake was significantly less than energy expenditure (-562 kcal,p<O.0001) and on average energy intake represented 76.8±3.4% of the measuredexpenditure for the whole group. The correlation between reported energy intake andexpenditure was 0.504 (p=O.009) (Figure 12A). The results described above were notadjusted for the changes in body energy stores during the study period. The changes inbody weight and composition were used to adjust for these body energy stores in thefollowing manners:Using the assumption of Black et al (1986) that the probable energy density of tissue lostor gained in adults under these normal conditions is 7000’ body weight, theadjusted energy intake for body weight changes was calculated in the following equation:Adjusted El (kcal.d’)=Reported EI-(ABWx7000 kcal.kg1)/number of days (Eqn. 8)On the basis of the assumptions that adipose tissue is 20% water and that each gram of fatstored has a caloric equivalent of 9.5 kcal.g’, the caloric equivalent of 1 g adipose tissue is7.6 kcal (Bandini et al 1990). The caloric equivalent of 1 g protein is 4.23 kcal. The90Table 18. Accuracy of Reported Energy Intakein Normal-weight SubjectsSubject LBW AFFM L\FM REI EIBW1 EI-1 REJJTEE EIBWITEE BIc /TEE(kg) (kg) (kg) (kcal.d ) (kcal.d ) (kcal.d ) xioo (%) xioo (%) xioo (%)1 -1.6 -1.63 0.03 1560 2360 1677 79.5 120.3 85.52 -0.2 -0.08 -0.12 1681 1781 1755 67.7 71.7 70.73 0.2 -0.49 0.69 1518 1418 1181 81.1 75.8 63.14 0.5 -0.72 1.22 1792 1542 1189 76.1 65.5 50.55 -0.9 1.75 -2.65 1938 2388 3237 67.6 83.3 112.86 -0.8 -1.75 0.95 2051 2489 1717 100.1 119.3 82.37 -0.5 0.03 -0.53 2405 2655 2688 84.4 93.1 94.38 -0.4 -0.14 -0.26 1790 1990 1940 80.3 89.2 87.09 1.2 0.04 1.16 1593 993 961 98.3 61.3 59.310 -0.3 0.90 -1.20 1321 1487 1898 69.7 77.6 100.211 -0.1 -0.25 0.15 1546 1596 1486 85.5 88.3 82.212 -1.0 0.71 -1.71 2355 2842 3213 76.3 92.6 104.713 0.0 0.67 -0.67 1409 1409 1719 66.8 66.8 81.514 -0.5 0.13 -0.63 1359 1609 1690 48.5 57.4 60.215 -0.4 0.79 -1.19 1144 1344 1728 38.2 44.8 57.616 0.5 0.70 -0.20 1432 1182 1483 52.3 43.2 54.217 0.2 0.58 -0.38 1389 1289 1547 69.3 64.3 77.218 -0.3 -0.64 0.34 989 1132 851 50.7 58.4 44.019 -0.1 0.15 -0.25 2201 2258 2330 88.4 90.4 93.320 0.0 0.24 -0.24 1716 1716 1829 105.3 105.3 112.221 -0.2 0.94 -1.14 1484 1554 1996 93.4 99.8 128.222 0.4 -0.92 1.32 1878 1678 1234 77.3 69.1 50.823 -0.7 -0.23 -0.47 1200 1553 1478 66.2 85.5 81.324 0.9 0.10 0.80 1115 665 647 101.9 60.8 61.525 -0.8 -0.31 -0.49 1826 2226 2116 93.1 113.5 107.926 0.0 -0.89 0.89 2280 2280 1868 78.3 78.3 64.1Mean -0.1 -0.01 -0.18 1653 1722 1749 76.8 79.0 79.5SEM 0.1 0.16 0.19 76 101 122 3.4 4.1 4.4Symbols used are the followings: zBW, AFFM and AFM, changes in body weight, FFMand FM during the 14-day experimental period; RET, EIBW and El c, reported energy intakeand intakes adjusted for changes in body weight and composition, respectively.91300002500 a2000• a ia I1500• •1000t.05000 I500 1000 1500 2000 2500 3000 3500Energy expenditure (kcalld)3000 Bg250OI •• 1500 • • II1000Ih. 5000LU500 1000 1500 2000 2500 3000 3500Energy expenditure (kcalld)Figure 12. Plots of energy expenditure against reported energy intake (1 2A:r=O.504, p=O.OO9) as well as energy intake after adjusting for the changes inbody energy stores (12B: r=0.503, pO.OO9) (n=26).92following equation were used to adjust energy intake for changes in body compositions:Adjusted El (kcal.d’) = Reported El - (AFM x 7600 kcal.kg1 + AFFM x 0.268 x4230’)/number of day (Eqn. 9)On average the subjects lost only 0.138±0.116 kg. This amount of weight loss couldaccount for a discrepancy of 69 kcal.d1.When the individual energy intakes were adjustedfor the weight lost or gained, the adjusted energy intakes were not statistically differentfrom the reported energy intakes (p=O.244) and the representiveness of adjusted energyintakes was also not different from that of reported energy intake divided by expenditure(p=O. 507). This adjusted energy intake represented 79.0±4.1% of energy expenditure(Table 18) and the correlation between energy intake and expenditure was 0.503(p=O. 009) (Figure 1 2B). Similarly, there were no significant differences between energyintake after adjusting for the changes in body composition and reported energy intake(Table 18).4.2.4. Accuracy ofReported Energy Intake in Obese SubjectsThe individual data for reported energy intake, changes in body weight and compositionand energy intake as a percent of TEE in the obese subjects are shown in Table 19.Reported energy intake was significantly less than expenditure (-1505 kcal. d’, p<O .0001).Reported energy intake represented 47.6±7.5 % of TEE measured by DLW93Table 19. Accuracy of Reported Energy Intake in Obese SubjectsSubjects AEW AFFM AFM REI MEBW MEc1 REIJTEE MEBW/TEE ME /TEE(kg) (kg) (kg) (kcal.d ) (kcal.d ) (kcal.d ) xioo (%) xioo (%) xioo (%)1 -2.8 0.49 -3.29 1386 2272 2617 38.3 62.8 72.32 -2.4 -1.69 -0.71 1601 2416 1741 46.8 70.6 50.93 -3.3 -1.62 -1.68 1475 2587 1981 50.8 89.1 68.24 -1.7 -1.57 -0.13 535 1090 440 24.7 50.3 20.36 -1.7 -1.40 -0.30 1042 1487 918 44.7 63.8 39.37 -3.2 -1.45 -1.75 2531 3590 3058 80.2 113.8 96.9Mean -2.5 -1.21 -1.31 1429 2240 1792 47.6 75.1 58.0SEM 0.3 0.34 0.48 271 359 405 7.5 9.3 11.0Symbols used are the followings: zBW, zFFM and z\FM, changes in body weight, LBMand FM determined by the IDS method during the 14-day experimental period; RET,IVIEBW and ME c, reported energy intake and metabolizable energy intakes adjusted forchanges in body weight and composition, respectively.94method. Because body weight declined dramatically in this group during the 2-weekperiod, part of the discrepancy between intake and expenditure may have been attributableto changes in body energy stores. Also, the obese subjects were patients followingileogastrostomy. There was some energy lost in feces and possibly in urine which mayoverestimate the metabolizable energy intake (ME). After correcting for the energy lost infeces and urine, ME fell to 982±242 kcal.d’ which was significantly less than reportedenergy intake (p<O.000l). The changes in body weight and composition were used toestimate the energy stores in the same manner as we did in normal-weight subjects. Theenergy loss estimated from the changes in body weight and composition using equations 8and 9 could account for 1258±145 and 810±240 kcal.d1 discrepancy,respectively. The adjusted energy intakes were 2240±3 59 and 1792±405 kcal,d’ obtainedfrom ME and the changes in body weight (1258±145 kcal.d’) and composition (810±240kcal.d’) (Table 19). These adjusted energy intakes represented 75.1±9.3 and 58.0±11.0% of energy expenditure. There was significant difference (pO.O29) in the degree ofunderestimation of energy intake using both means to adjust the changes in body stores inthe obese subjects, but not in the normal-weight subjects (Figure 13).Table 20 summarizes the results of the validation study. Obviously, there wereconsiderable differences in TEE and underestimation of energy intake, while these twogroups of subjects were not comparable because TEE in the obese group was measuredduring weight loss. In addtion, the two groups of subjects came from two geographicalareas, and the sample collection and data analyses were also slightly different between the95two groups. Obese subjects seem to have lower energy expenditure per body weight thannormal-weight subjects, however, energy expended per FFM was not different.4.2.5. The Relationship Between Underreporting ofEnergy Intake andBody WeightTo attempt to identifj factors associated with severity of underreporting, simplecorrelation analyses were performed between the under-reporting of energy intake withbody weight, age and height in normal-weight and obese groups. Results showed thatthere were no close correlations between these variables and the degree of under-reportingof energy intake in the normal-weight subjects. However, a significant negativerelationship was found between under-reporting of energy intake with body weightmeasured during the 14-day experimental period (r=O.868, p=O.O25) in the obese subjects.Multiple regression comparing degree of under-reporting of energy intake against bodyweight, age and height showed that there was a close relationship betweenthese variables and the degree of underreporting in obese group, but not in the normalweight group (Table 21).w w I 0 U) cT5 a) U) Cl) a) I ci x a) w Lu H C Cs Lu a: C a) ci) ci) £2 ci) C-) C U) I- a) ‘I Figure13.Differencebetweenreportedenergyintakes(REI)andtotalenergyexpenditureinnormal-weightandobesesubjectsexpressedasreportedenergyintake(UR1) andintakesadjustedforchangesinbodyweight(UR2)andcomposition(UR3) as%TEE.Statisticalsignificancewasindicatedbydifferentletters.•UR1EUR2•UR3100 80 60 40 20 0aZLNormal-weightObeseCo 0)97Table 20. Summary of Energy Intake, Expenditure and theRepresentativeness of Reported Energy Intake inObese and Normal-weight SubjectsObese Normal-weightRET (kcal.d’) 1429±271 1653±76TEE (kcal.d’) 2933±239 2215±102RET/TEE (%) 47.6±7.5 76.8±3.4TEE/BW (’) 28.5±2.1 34.8±1.6TEE/FFM (’) 53.4±4.2 53.2±1.7Values are mean±SEM. Symbols used are as follows: RET, reported energy intake; TEE,total energy expenditure; BW, body weight; FFM, fat-free mass.98Table 21. Multiple Correlation Coefficients Between Underreportingof Energy Intake and Some Physiological Variablesin Obese and Normal-weight SubjectsGroups Obese Normal-weightVariables Body weight, age and heightUnderreporting’ (%) 0.989 (0.032) 0.313 (0.508)Underreporting2(%) 0.962 (0.109) 0.254 (0.682)Underreporting3(%) 0.995 (0.015) 0.329 (0.463)Numbers are multiple correlation coefficients (r) and p values are shown in brackets.Underreporting”2’3represent the degree of underestimation of metabolizable energy intakeand and intakes adjusted for changes in body weight and composition, respectively.995. DISCUSSIONThis investigation expands previous studies of the energetics associated with weight lossfollowing intestinal bypass procedures. Based on the results obtained in the present study,there is evidence that TEE was related to weight loss after ileogastrostomy. Fecal energyloss was still an important determinant in the weight loss foloowing intestinal bypassprocedures. The difference between reported energy intake and expenditure was observedin both normal-weight and obese participants. This section will discuss the present findingsin the context of available data from similar studies.5.1. Influence of fleogastrostomy on Body Composition and Methodology inAssessing the Reduction in Body CompositionWe used the IDS method to measure the changes in body composition followingileogastrostomy in our first series of subjects. The results were compared with thosedetermined by BIA in which we used the equation of Segal et al (1988) to predict FFM.We found that FFM and FM measured by IDS method were significantly different fromthose obtained by BIA. There were no differences in the assessment of changes in FFMand FM between the two methods after ileogastrostomy.The difference in body composition measurement by IDS method and BIA in the presentstudy may be due to the influence of high %BF in obese subjects on the accuracy of BIA100measurements. It was reported that obesity affected the precision of BIA (Gray et al1989). Segal et a! (1985) reported a significant relationship (r= 0796) between residualFFM scores, calculated as the difference between observed and predicted values, and%BF. Our results showed higher FM and lower FFM determined by BIA than those byIDS method. Generally, underestimation of FFM was offset by overestimation of FM andvice versa. However, systematic errors occurring in both methods could alternativelyaccount for the present findings.Isotope dilution has been a traditionally used technique to investigate changes in bodycomposition. However, it is well known that this technique has limitations. Even with highprecision of TBW estimation, errors can be made in calculating FFM and FM because it isnot known whether the constant water content in FFM can be applied to all subjects.Particularly, the assumption of 73.2% of body water in FFM (Sin 1956) is violated inbypassed patients who may experience dehydration following these procedures. Even innormal sujects, the assumption appears flawed: Wellens et al (1994) showed the widerange of interindividual variation in the degree of hydration, where an average 74.0% inmen (range 65.8-86.2%) and 73.1% in women (range 60.8-84.5%) was observed. Otherstudies (Elia 1992; Fuller et al 1991) also reported an extensive range (67-78%) for thehydration fraction of FFM. This known wide range of interindividual variation in thedegree of hydration can affect the calculation of body composition estimates by theisotope dilution method.101Bioelectrical impendance analysis is a relatively new method for the assessment of bodycomposition. Because this approach is safe, noninvasive, rapid, portable, inexpensive, andeasy to use, it may be amenable for laboratory, clinical, and field assessment of humanbody composition. Many equations predicting FFM from weight, stature and resistancehave been reported (Lukaski and Bolonchuk 1988; Segal et al 1988; Vasquez and Janosky1991). Use of these equations to measure FFM in weight-stable subjects appears validbecause of the high correlation between BIA and independently measured FFM. However,controversy exists in the validity of BIA to measure changes in FFM (Deurenberg et al1 989b; Kushner et al 1990). Moreover, Vazquez and Janosky (1991) found that neitherresistance nor reactance changed significantly during reduction in body weight andprediction of FFM was based on factors other than resistance. Also, they showed that allequations used recently produced high prediction errors. Forbes et al (1992) analyzed thebasic equation used in the bioelectrical impedance methods. They raised doubt about theapplicability of the basic equation, which forms the foundation for this technique. Thevalidity of BIA method to estimate body composition among patients with altered fluidand electrolyte status is a critical and unsolved question. Among such individuals, theTBW/FFM and the intracellular to extracellular fluid volume may be altered. Our resultsshowed that significant differences in estimates of FFM and FM by the two methodspersisted, however, the changes in FFM and FM compartments were not different betweenthe two methods.102In this study a mean weight loss of 17.2 kg was found during the 3 month test period.Theoretically, 70-80% of this weight loss must be due to a loss of FM (Garrow 1978,1981). Our results showed less FM loss during the first month (48.1 and 32.7% of FMloss measured by IDS method and BIA, respectively) than the theoretical value (70-80%).The percentage of fat loss increased with the passage of time (64.6 and 71.0% of FM lossby IDS method at 2 and 3 months; 65.0 and 70.9% of FM loss by BIA at 2 and 3 months).The results indicate that ileogastrostomy induced higher loss of FFM compared with thetheoretical value of 20-30% of FFM loss, especially in the first month after surgery. Theloss of FFM and FM at 2 and 3 months are similar to those reported previously followingJIB (Brill et al 1972; Scott et a! 1975), however, a smaller loss of FM was found duringthe first month in this study. Regardless of the validity of IDS method and BIA, ourresults suggest that BIA overestimated FM and %BF compared with estimates obtainedby IDS method. However, there were no differences between the two methods inassessing the changes of FFM and FM following ileogastrostomy. It appears that IDS andBIA methods are capable of estimating changes in body composition, while there weresignificant differences in assessment of compartment size between the two methods.Resolution of the reliability of lBS and BIA methods in assessing body composition andchanges of FFM and FM compartments may require the use of advanced technology asreference method. Use of compositional models that account for altered fluid distributionare needed to avoid reliance upon the compartmental assumptions of a relatively constantTBW/FFM.103Dual energy x-ray absorptiometry has been developed to measure body composition and itis largely independent of compartmental assumptions. The error of DEXA was reported as1% in %BF and 0.8 kg in FFM (Wellens et al 1994). In addition to the high precision ofDEXA measurements, the procedure provides a direct measurement of both fatty and leanelements of the body, independent of dehydration state. Because of the characteristics ofDEXA, we hypothesized that DEXA would be useful for validation of the LDS and BIAmethods in which the assumption of water content in FFM is questionable. However, thevalidity of DEXA in obese subjects has not been investigated, thus, large errors inassessing body composition may occur in the morbidly obese subjects.Our results using DEXA showed a significant decline in both FFM and FM afterileogastrostomy. The percentage of FFM and FM after surgery was not different from thatpresurgically. The result was in agreement with the finding in our first series of subjectsusing the IDS method. As expected, the estimates of BMC were unaffected by surgery.Regional changes in body composition data showed that ileogastrostomy reduced LM andFM mainly in the trunk and legs. There was a significant difference between total bodymass by DEXA and BW by scale before surgery. This difference may be accounted for bythe large BW before surgery because the BW of a few subjects exceeded 300 lb that wasthe recommended upper level of DEXA measurement. With the reduction in BW afterileogastrostomy, there was no significant difference in assessing body mass by DEXA andBW by scale, which was an evidence that large BW before surgery can influence the104precision of DEXA measurements. It has been reported that the precision of DEXAmeasurements deteriorated with increasing depths of soft tissue (Laskey et al 1992).Significant effects of depths and adiposity on measurements of FFM and FM were foundin their study. It was suggested that DEXA might be least accurate for obese subjects.A significant difference in FFM and FM determined by BIA and IDS methods wasobserved pre- and postsurgically, however, the estimates in FFM and FM by DEXA werenot significantly different from those by BIA and IDS methods. Similarly, DEXA did notdiffer from BIA and IDS methods in the measurement of pre- and postsurgicalpercentages of BF. It appears that the estimates in FFM and FM by DEXA lay betweenthose by BIA and IDS methods. These results suggest that BIA, IDS and DEXA, asapplied in this study, gave poor measurement of body composition in the morbidly obesebefore and after ileogastrostomy. The differences may arise from biological as well asmethodological sources. For the DEXA method the potential sources include theconsiderable range of body thickness obseved in some of the subjects which is known toinfluence the ratio of the x-ray beam attenuation. For BIA and IDS methods, thelimitations have been discussed at the beginning of this section.The changes in FFM and FM obtained by IDS method was generally comparable to thatobtained by DEXA and different from that obtained by BIA. Furthermore, the percentageof these reduced FFM and FM likewise showed high agreement between 1DS method andDEXA. However, the reduced FFM and FM determined by the three methods were not105very well correlated with each other. The poor intercorrelations in the changes of PPM andPM between each pair of the methods may be explained by small sample size andphysiological changes induced by the surgical procedures. The different effects of thesechanges on individual method may lower the linear relationships between each pair of themethod in the assessment of changed body composition after surgery.Our results suggest that both DEXA and isotope dilution methods may accurately assessthe decrease of FPM and PM compartments induced by ileogastrostomy. There was noevidence indicating inability of the isotope dilution method to measure changes in bodycomposition after intestinal bypass surgery. It is unlikely that IDS method underestimatedPPM compartment due to dehydration in these subjects after surgery, because FFMcompartment determined by IDS method was higher than that by DEXA. Also,dehydration may not be obvious at 6 weeks after this procedure. In contrast, BIA may notbe a good method for assessing the changes in body composition during the study period(within 6 weeks after surgery) and its validity may improve with the passage of time aftersurgery.Acceptance of a method for assessing body composition is determined by the simplicity ofthe method as well as by its accuracy. Our results indicate that DEXA is a first choice forassessing changes in body composition in obese research if the subject’s body weight isnot excessive. DEXA has many advantages (Roubenoffet a! 1993) in that DEXA isnonivasive and the variance of the estimates is not affected by the subjects. It may be a106better means to estimate body composition than IDS method for individuals with alteredwater status.5.2. Changes in Energy ExpenditureThe level of BEE tended to decrease in obese individuals following ileogastrostomy,however, this reduction in BEE was of borderline significance. Previous studies revealedconflicting results concerning the changes of BEE in obese individuals after weightreduction. Some investigators reported that the relative BEE was unchanged with weightloss (Dore et al 1982; Warnold et al 1978). Warnold et al (1978) determined basalmetabolic rate (BMR) before and after dieting. They found that despite significant weightloss after dieting, BMR declined insignificantly which is accordant with our findings. Onthe other hand, Leibel and Hirsch (1984) reported that the BEE of obese individuals afterweight loss was less than lean subjects with comparable FFM. This finding suggests that aprolonged reduction in BEE may follow weight loss although sequential body compositionand metabolic measurements were not performed in their study. McFarland et al (1989)also reported that BMR declined following gastric partition and this decrease of BMR wasdue to the substantial reduction in energy intake.To our knowledge, there are very few data reported on the energy expenditure afterintestinal bypass surgery. Kopelman et al (1981) found a significant rise in serum 3’5’3triiodothyromne (T3) and a significant fall in 31315 triiodothyronine (rT3) concentration107between 15 and 20 weeks after bypass surgery. The increase of plasma T3 induced byintestinal bypass surgery is contrary to what has been described in the literature afterweight loss induced by dieting (Froidevaux et al 1993). It is generally accepted that low-energy diets induce a decrease of total T3 concentrations in plasma and the reduction in T3is due to a decreased conversion ofT4 into T3 in the peripheral tissues. In contrast to thefindings with dieting, an increase in T3 level after bypass was observed and this maycontribute to the unchanged BEE and substantial weight loss seen after these procedures.However, we did not measure the changes in plasma T3 levels.The current investigation, based on measurements of energy expenditure and bodycomposition, indicates that quantitative reduction in BEE was not associated with thechanges in BW and FFM. Previous studies (Bessard et al 1983; Geissler et al 1987)attributed the reduction in BEE to a loss of FFM. The following explanations may accountfor our findings. Firstly, BEE in our study subjects declined slightly but insignificantly aftersurgery. Secondly, the small sample size (n=7) and large variability in both BEE and bodycomposition data may contribute to the results in this study. Finally, the slight decrease inBEE may result from the reduced energy intake after surgery. It is possible that BEEgradually normalizes to presurgical level with the passage of time. Nevertheless, we failedto detect a significant decline in BEE following ileogastrostomy. Further studies areneeded, especially with a larger sample size, to confirm our findings.108The possibility that impaired thermogenesis, which is a blunted increase in energyexpenditure in response to certain stimuli, is associated with some types of human obesityhas received considerable investigative attention. The present study does not support theconcept of a reduced postprandial thermogenesis in obese subjects, although we do nothave TEF data for normal-weight subjects. The presurgical TEF expressed as thepercentage of either BEE or ingested meal was comparable to values reported in theliterature for the normal-weight subjects (Bukkens 1991; Cunningham et al 1981; Felig etal 1983), where reduced thermic responses to food were not detected. Cunningham et alcompared the thermogenesis after ingestion of an 800 kcal liquid meal (45% CHO, 40%fat and 15% prot) in 10 normal-weight and 10 obese subjects. They found that energyexpenditure was consequently 22-24% higher in obese than in normal weight subjectsthroughout the postmeal period (p<O.Ol). It appeared that obese subjects did not showimpaired thermogenesis. Yet a number of other studies (Bessard et al 1983; Schutz et al1984; Segal et al 1 987a, 1990) have shown that postprandial thermogenesis is significantlysmaller in obese than in lean humans.The controversy that exists in the relationship between thermogenesis and obesity mightpartly be due to heterogeneity of the obese and methodological differences concerning thetechniques of measuring energy expenditure. The energy content of the meal used in thisstudy was estimated from energy requirements to maintain IBW (Shetty 1981). Therewere two reasons to calculate the test meal based on IBW. Firstly, food given on the basisof LBW probably related better to the individual’s active mass of metabolizing tissue than109total body weight (James et a! 1978), therefore, the results could be more appropriate tocompare with lean controls (Shetty 1981). It has been reported that TEF increases in thesame subject with an increase in the meal’s energy content (Morgan et al 1982). In thisstudy, the energy load of test meal was lower than the energy requirement of the obesesubjects, because the obese subjects had more active tissue than they were at IBW.Therefore, it was not possible to overestimate the values in TEF due to the lower energyload of the test meal. Secondly, we tried to minimize the differences in meal’s energycontent before and after surgery, because most subjects could not complete the food weprovided after surgery.A substantial reduction of TEF corrected for the difference in energy content andmacronutrient composition was observed following ileogastrostomy. The reduced TEFafter surgery may be partly due to malabsorption of nutrients. However, the dramatic andcontinuous decline in TEF cannot be explained by malabsorption alone since the bowelrapidly adapts to the state of a shortened gut (Cleator et al 1991). With the improvementof malabsorption, TEF did not show a rise in the present study. The finding was inaccordance with dieting induced weight loss. Bessard et al (1983) found a significantlylower postprandial thermogenesis in obese subjects after weight loss when compared tolean subjects in response to a liquid mixed meal.The increase in heat production that occurs after a meal has been divided into two parts:obligatory and facultative. Obligatory thermogenesis is released in the processes of110transport, metabolism, and assimilation of metabolites absorbed after digestion. Theremainder of the thermogenesis include substrate cycles (Poehiman and Horton 1987).Quantitative significance of cycles in relation to heat production in humans cannot be fullyunderstood at the present time. The decline in TEF following ileogastrostomy may beaccounted for primarily on the basis of malabsorption of nutrients with the change insubstrate cycles playing a secondary role.What is the practical importance of the reduction in TEF after surgery? The extent towhich energy is saved by a reduction in TEF may contribute to slowing-down and possibledifficulty in achieving weight loss. The importance of the reduction in TEF on energybalance should be assessed together with the components of total energy expenditure.Due to the shortage of DLW, we did not conduct presurgical TEE measurements. TheTEE was measured during 6-8 weeks following ileogastrostomy. For theoretical andanalytical purposes, TEE can be broken down into BEE, TEF and the energy cost ofphysical activities. BEE is a measure of the energy expended for maintenance of normalbody function and homeostasis and usually, BEE equals 60-70% of TEE. Thermic effectof food is the increment in energy expenditure above BEE after food consumption andcomprises ‘-.1O% of TEE. Energy expended for physical activity is the most variablecomponent of TEE in humans. This component varies from <100 kcal.d’ for inactivepersons to >1000 kcal. d’ for those who are active. In ordinary life, physical activitycomprises 30% of TEE. Quantitatively, TEF, BEE and energy cost for physical activity in111patients after ileogastrostomy were determined in the present study. The value for TEFwas 3% at the second month after surgery, which is lower than the theoretical value(10%) and that (9.1%) obtained in normal-weight subjects (Weststrate et al 1989). Therelative percentage of presurgical TEF is unknown due to the lack of presurgical TEE inthis study. However, the findings that presurgical TEF, expressed as the percentage ofBEE or ingested energy, was comparable with that reported in normal-weight subjects(Bukkens 1991; Cunningham et al 1981; Felig et al 1983) suggest a higher percentage ofTEF presurgically than that postsurgically (3%). The value for physical activity (44%) inthis study was much higher than the average level and quite similar to those obtained in agroup of patients after gastroplasty (Westerterp et al 1991), where TEE was measured byDLW method. Before surgery, the level of energy expended for physical activity waslower than that after surgery judging from the physical activity index (1.52 vs 1.63). It wasconcluded that activity might rise after weight loss in their study (Westerterp et al 1991).Factorial measurements of energy expenditure showed that physical activity index forwomen was 1.56, 1.64 and 1.82 for light, moderate and heavy activity levels, respectively(FAO/WHO/IJNTJ report 1973). The physical activity index was 1.79 in the present study,which implied high physical activity levels in the subjects following ileogastrostomy.To our knowledge, there has been no work to investigate TEE after intestinal bypassprocedures that would be directly comparable to the present study. Controversies stillexist about the relationship between changes in TEE and weight loss. Bradfield andJourdan (1972) have studied six grossly obese women who lost 6 kg in one and half112months after dieting, and found no difference in TEE as predicted from heart rate/EEindividual regression lines. A study by Westerterp et al (1990) reported that TEE may riseafter weight loss because of increased physical activities. Most other studies found thatTEE declined during and after weight loss. Bessard et al (1983) reported a significantreduction in TEE after weight loss. They calculated an EE equivalent of weight loss,averaging 18’ weight loss per day. Ravussin et al (1985) also found a reduction inTEE and reported that approximately one half of the TEE reduction was accounted for bya decrease in BEE. Most of the remaining decline in TEE was explained by a decreasedTEF, and by the reduced cost of physical activity mainly due to lower body weight(Ravussin et al 1985). Although the changes in TEE after ileogastrostomy remainsunknown, the findings that BEE was unchanged after surgery and the level of energyexpenditure for physical activity was high in this study suggest that TEE may notsignificantly decline following ileogastrostomy.Preprandial fat and CHO oxidation rates were not affected by the surgical procedure andthe RQ during BEE measurement was essentially unchanged during weight loss, althoughthe energy balance was largely negative. These results are similar to those reported byFroidevaux et al (1993). There is some theoretical evidence that individuals who are closeto energy balance have an overall non-protein RQ largely influenced by the amount ofcarbohydrate and fat in the diet (Bessard et al 1983). If only fat was being oxidized, theRQ would be about 0.7 and if only carbohydrate was being oxidized, the RQ would be1.0. Most humans consuming a mixed diet will have an average RQ of about 0.87. When113the individual is in negative energy balance, endogenous fat is oxidized to provide energy.Therefore, significantly lower RQ may be observed. In this study, basal RQ level increasedslightly but insignificantly, however, postprandial RQ declined significantly. The resultsindicate that ileogastrostomy induced high postprandial fat oxidation. A significantincrease in postprandial fat oxidation and decrease in CHO oxidation was found in thisstudy. These metabolic changes may be related to hormonal alterations induced byileogastrostomy. Increased levels of insulin are characteristics of obesity. Hyperinsulinemiamay reflect insulin resistance in the obese individuals. A significant decline in plasmainsulin levels has been found in a previous study (Buchan et al 1993). It was suggestedthat the insulin sensitivity in postsurgical obese subjects was higher than that persurgery.However, the increased insulin sensitivity could not explain the changes in fat andcarbohydrate oxidation observed after ileogastrostomy. The changes in insulin level mayaffect the partitioning of carbohydrate and fatty acids as fuel in the body resulting inenhanced utilization of the latter for energy.5.3. Changes in Energy IntakeA significant reduction in food consumption has been recognized as a major cause forweight loss following intestinal bypass surgery (Bray et al 1978,1979; Brewer et al 1974).Our intake data indicate that energy intake declined significantly followingileogastrostomy. Nevertheless, postsurgical energy intake was not significantly differentacross the first 3 months. When intake was divided into protein, fat and carbohydrate, no114significant changes were observed in the percentage of energy obtained from thesemacronutrients. We did not demonstrate a significant association between reduced energyintake and weight loss due to large variability and small sample size in the present study.Similarly, the amount of energy consumed on the test days was not significantly correlatedwith the amount of weight loss over the preceding month.Bray et al (1978;1979) measured the energy intake of 14 female patients before JIBsurgery, and 3 weeks and 6 months after surgery. They found that energy intakes fell aftersurgery and there were no major changes in the distribution of calories consumed asprotein, fat, and carbohydrate. Our results are accordant with their findings. However,they and others (Robinson et al 1979) found a high inverse correlation between energyintake and weight loss. Therefore, it was concluded that weight loss induced by JIB wasaccounted for primarily on the basis of decrease in energy intake with malabsorptionplaying a secondary role.In agreement with Condon et al (1978), we did not show a significant relationshipbetween energy intake and weight loss during the 3-month and 14-day balance study. Wenoted that some subjects increased their intakes after surgery although their body weightdeclined constantly and continously. Condon et al also found an increase of energy intakein some of their subjects. They measured the energy intake of 65 bypass patients. Of the65 patients 48 decreased their food intake after surgery, whereas the remaining 17 patientsincreased their food intake. The difference in weight loss was not significant in the two115groups. There was also no close correlation between weight loss and energy intake in theirstudy.The study by Cleator et al (1991) demonstrated that reduced food intake andmalabsorption were insufficient to account for the energy loss calculated from the changesin body composition. In this study, 12 morbidly obese subjects were studied before andafter ileogastrostomy. Energy intake measured by 3-day food record and the decreasedenergy intake was calculated from preoperative 3-day energy intake minus postoperativevalue multiplied by 90 study days. Similarly, malabsorption was measured using 3-dayfecal collection for the determination of fecal protein, carbohydrate and fat content whichwere converted to energy loss. The energy loss was calculated by multipling 9.4 for FMand 4.0 for 26.2% protein in FFM compartment. Approximately 1300 calories per day wasunaccounted for using energy balance calculations. However, inaccuracy of energy intakedata and the indirect estimation of energy loss in feces were the main concern in theirstudy. Also, food intake and malabsorption changed with the passage of time aftersurgery. Food intake and fecal nutrient contents at the final 3 days may not represent theaverage values during the 90-day study period. Definite conclusion could not be drawnfrom simple calculation of energy balance.A considerable variability in the degree of decreased energy intake, and in the relationshipbetween postoperative undereating and weight loss existed in the previous studies(Benfield et al 1976; Condon et al 1979; Robinson et al 1979). This variability might be116due to the differences in energy intake measurement procedures and sampling periods usedin the various studies. Table 22 lists some studies describing reduced energy intakefollowing intestinal bypass surgery and the methods they used. The inaccuracy of dietarymethods is the main concern about the role of reduced energy intake in weight loss aftersurgery. We believe that reduced energy intake contributed to the negative energy balancefollowing ileogastrostomy, however, its magnitude in the weight loss could not bedetermined.5.4. Factors Associated to Weight Loss Following IleogastrostomyData concerning fecal energy loss in the present study provide evidence that malabsorptionplays an important role in weight loss although it alone probably does not account for allof the postsurgical weight loss. Moreover, the findings indicate that malabsorptiondepends on the amount of the food taken. Further studies are required to clarify thisrelationship. A justifiable conclusion cannot be drawn from these results due117Table 22. Reports of Decreased Food ConsumptionFollowing Intestinal Bypass in HumansStudy Subjects Methods TimeBray et al (1979) 14 women self-select preferred pre-, 3 wk andfoods 6 mo.Condon et al (1978) 48 women, 17 weighed food record pre- and 9 and dietary interviewCleator et al (1991) 12 women dietary record pre- and 3 mo.Pilkington et al 8 women and 8 prepared diets pre-, 4,12 and(1976) men 24 mo.Robinson et al 14 women, 17 questionnaire pre, 2 wk and(1979) men 4 mo.118to small sample size. Increasing the size of study sample may reduce some of the effects ofmisclassification. However, we believe that malabsorption existed and persisted at least 2months after surgery, and fecal energy loss certainly contributed to the weight loss afterileogastrostomy. Many authors have concluded that malabsorption accounted for all of theweight loss after JIB (Corso and Joseph 1974; Scott et al 1971; Weisman 1973). BothCorso and Scott et a! suggested that weight loss after JIB occurred in the absence of asignificant decrease in energy intake, although actual intake was not measured in eitherseries. A number of investigators have noted a distinctly inverse relationship between thelength of small bowel left in continuity and the degree of weight loss. Our fecal energydata were comparable with that reported in the literature (Crisp et a! 1977; Pilkington et al1976; Robinson et al 1979; Scott et a! 1971). Crisp et al (1977) reported that the loss ofenergy in the stools rose from 131 kca!. d’ preoperatively to a maximum of 593 kcal. d’postoperatively. Scott et al (1971) also reported that the energy content of the stools rosefrom 100 kcal. d’ preoperatively to 500 kcal. d’ postoperatively. In agreement with thesestudies, our result favors the concept that malabsorption may be the major contributingfactor to weight loss in intestinal bypass procedures.A significant correlation existed between weight loss and energy expenditure in the presentstudy. There has been no similar study to demonstrate the relationship between weightloss and energy expenditure as we did. Energy expenditure as a factor associated withweight loss following intestinal bypass procedures was proved in our study. A closenegative association between the reduction in energy intake and weight loss was not119demonstrated in this study due to the inaccuracy of reported energy intake. We found thatthere was a marginally positive relationship between energy intake and energy expenditureas well as fecal energy which may complicate the role of reduced intake in the weight lossfollowing intestinal bypass surgery. The simple regression analysis was performed in thisstudy due to the small sample size. Ideally, multiple regression analysis should beperformed as weight loss was a dependent variable and factors associated with the weightloss were independent variables. The relative contribution of these factors could beassessed by their coefficients in a multivariate model to correct the interclass correlations.The data available limited this statistical analysis. As to the mechanisms of weight lossafter ilegastrostomy, we believe that reduced energy intake plays some role in the weightloss, especially in the very early stage. This may be the reason for lower loss of FM at thefirst month observed in this studyUrinary energy loss arose from incomplete oxidation of the organic matter. Most of theorganic matter in urine are the nitrogenous compounds. It was proposed that measurementofurinary nitrogen by Micro-Kjeldahl procedure could be used to predict the energy lossin urine. A wide variation in the energy:nitrogen ratio was reported. Calculation of theurinary energy from energy:nitrogen ratio were always lower than the determined values(Southgate & Durnin 1969), although urinary energy was closely correlated with totalnitrogen. We measured the urinary energy determined by bomb calorimetry which wascomparable with that reported in normal subjects (Southgate & Durnin 1969).Furthermore, this urinary energy loss was not closely correlated with weight loss after120ileogastrostomy and we believed that the correlation cannot be improved with the increaseof sample size. The possibility of other routes of energy loss which was suggested by aprevious study (Cleator et al 1991) exists but is likely very small. Its contribution to theweight loss after ileogastrostomy is probably not important.Although this study has strengths, two limitations may have affected the conclusion wecan draw. First, the inaccuracy of energy intake may have introduced significant error inthe amount reported versus what they truly ate, which may underestimate the associationbetween intake and weight loss. Second, the small sample size limits the ability to reach ajustifiable conclusion.5.5. Validation of Reported Energy Intake Using DLW MethodWe found that the disagreement between TEE and energy intake after correcting for thechanges in body composition was 20.5% and 42.0% of TEE in normal-weight and obesesubjects, respectively. Our use of the DLW method as a reference test revealed seriousdiscrepancies between estimates of energy intake and expenditure. Such has been theresult of many other studies (Bandini et al 1990; Prentice et al 1986; Westerterp et al1986). The discrepancy raises several important questions. Firstly, did the error occur inthe estimation of energy intake or expenditure? Secondly, if it occurred in the estimationof energy intake were the results spurious or could they have general implications for121dietary research? Finally, might biased results be identifiable or related to some physicalcharacteristics in subjects studied?The DLW method is based on the differential elimination of 2D and 180 from body watersubsequent to a loading dose of these stable isotopes. The difference between the twoelimination rates is therefore a measure of carbon dioxide production, from which totaldaily energy expenditure is calculated according to the methods of indirect calorimetry.The method has been validated in small animals by comparing the method with measuredcarbon dioxide and the accuracy of DLW method has been reported to be 1-2%, with arelative standard deviation of 3-9% (Nagy et a! 1980). The method has been validated ininfants and young adults, healthy individuals and patients with gastrointestinal disorders,subjects under metabolic ward conditions, and free-living individuals under laboratory andnonlaboratory conditions (Jones et al 1978; Jones and Leitch 1993 a; Livingstone et al1990; Schoeller et al 1986, 1988). None of the studies indicated any significant bias. Whendata from all of these studies are combined the results suggest a small overestimation ofexpenditure by 2-3%.The accuracy of the method is not significantly affected by energy balance (Schoeller et al1986) or physical activity (Westerterp et al 1988). Schoeller et al (1986) investigated theerror in the DLW method as a function of energy balance to determine whether theaccuracy of the method is affected by the energy imbalance. Regression of percent errorcalculated from the difference with the reference method against energy balance status122(negative, zero and positive balance) was not statistically significant. The 95% confidencelimit about the slope suggests that the DLW error lies between -0.2 1% and +0.07%. Blacket al (1986) also covered most of the nutritional and physiological circumstances in whichthe DLW method is likely to be applied and demonstrated that errors arising fromassuming an RQ of 0.85 were very small. A mean FQ value to each community can beused without incurring significant error, although the precision can be improved stillfUrther by assessing each individual’s FQ. The influence of energy balance on theestimation of RQ using individual’s FQ values may be large and need to be considered inclinical studies involving rapid changes in body composition. Even if such changes in bodycomposition cannot be accurately assessed, the error for prediction of RQ from FQ shouldnever exceed ±2 percent. The total estimated error of the DLW method can be calculatedusing a root-mean-square summation of the errors arising from prediction of CO2production and the RQ assumption. The propagation of error analysis shows that theDLW method is robust and unlikely to be biased by more than 5% (Schoeller et al 1988).In subjects in energy imbalance errors in calculated energy expenditure will rarely exceed3-5% even if the imbalance is ignored. Therefore, DLW method is considered to be themost accurate method of assessing energy expenditure in free-living populations and validin obese subjects during weight loss.Previous work conducted in our laboratory entailed validation (Jones et al 1987; Jones andLeitch 1993a) and ongoing applications (Jones et al 1988,1993b; Su and Jones 1993)work in both humans and animals. Furthermore, our results of energy expenditure were123comparable to those reported in the literature (Banduni et al 1990; Welle 1992). Studiescomparing TEE between normal-weight and obese subjects using DLW method aresummarized in Table 23. Comparison of TEE between obese and normal-weight subjectsmay not be appropriate in this study since the TEE in obese group was measured duringthe dynamic phase of weight loss.It is therefore reasonable to conclude that the observed discrepancies arose largely frominaccurate estimates of habitual energy intake due to conscious or subconscious changes innormal dietary patterns or underreporting, or both. It is generally accepted that thearduous task of recording food intake may contribute to the unintentional underreportingof energy intake (Block and Hartman 1989). The major factors involved in generatingvalid nutrient estimation include the following: (1) selection of an appropriate datacollection methodology; (2) adequate level of food description; (3) appropriate techniquesfor quantifying amounts of food consumed; and (4) use of quality-controlled nutrientcalculation system that provides an adequate level of specificity. All the nutrientcalculation systems used recently are not complete and specific with the rapidly expandingnumber of foods available and increasing variety of foods in the diets. Difficulties inanalyzing food records may also result in large errors in the nutrient calculation. Tokentogether, bias in estimation of energy intake may underestimate or overestimate theassociation between energy intake and weight loss following intestinal bypass surgery.Also, this may be the reason for much of the controversies existed in the past.124Table 23. Studies Comparing Total Energy Expenditurein Normal-weight and Overweight Subjects,Using the Doubly Labeled Water MethodReference Group Mean weight (kg) Mean TEE (kcalld)Prentice et al (1986) Normal weight women 58±6 1914±287(n= 13)Overweight women 8 8±14 2440±33 5(n=9)Bandini et al (1990) Normal-weight boys 56±10 3110±502(n= 13)Overweight boys 94±26 . 36 14±646(n=18)Normal-weight girls 55±9 23 92±455(n= 12)Overweight girls 99±22 3278±431(n= 15)Welle et al (1992) Normal-weight women 60±4 2273±215(n=12)Overweight women 85±11 2679±431(n=26)Present study Normal-weight women 62±7 2215±518(n=26)Overweight women 104±10 2898±564(n=10)125As we compared energy intake data with objective measures of energy expenditure innormal-weight subjects, underestimation of food intake was also apparent in normal-weight group. Furthermore, the errors in estimating food intake are unlikely to be specificto the current study. Studies using the DLW method, conducted among diverse agegroups with a variety of health and/or disease states, confirm that self-report of energyintake tends to be lower than measured total energy expenditure (Johnson et al 1994;Lichtman et al 1992; Livingstone et al 1990; Schoeller 1990). Taken collectively, thesestudies found that reported food intakes underestimated habitual energy intakes.The magnitude of underreporting (20.5% in normal-weight group, 42.0% in obese group)in the present study is comparable with other published reports (Bandini et al 1990;Livingstone et al 1990; Mertz et al 1991). Bandini et a! (1990) compared reported intakewith expenditure determined by DLW method in obese and nonobese adolescents.Reported intake represented 8 1±19 and 59±24% of measured expenditure in nonobeseand obese, respectively. Livingstone et al (1990) compared energy intake as measured by7-day weighed records and total energy expenditure measure concurrently with the DLWmethod and found that on average the men underreported their intake by 19% and thewomen by 18%. In agreement with the literature, underreporting has been found to occurto a greater degree among obese than among normal-weight subjects. Lichtman et al(1992) reported that young obese subjects underreported their actual food intake by 47%and Lansky found reporting errors that averaged 53%. To date there is a paucity of work126done in bypassed obese subjects. This study provides new information on the degree ofunderreporting of energy intake that occurs in bypassed patients.The DLW method is too expensive and technically demanding to be used as a validator ofenergy intake measurements in large samples. It is possible that certain physiologicalcharacteristics may be predictors of the discrepancy between reported energy intake andtotal energy expenditure. Elucidating the relationship between these characteristics and themisreporting of energy intake could be a meaningful step toward the application ofcorrection factors to arrive at more accurate determinations of habitual energy intake.Thus, we tried to develop a prediction equation for understanding the bias that may existin reported energy intake data collected from both normal-weight and obese individualsusing independent variables, which are easily measured in a clinical settings. In our sample,both normal-weight and obese women were likely to underreport their energy intake. Wefound that there was no relationship between physical variables and underreporting ofenergy intake in normal-weight women. Thus, research is needed to examine other nonphysiological characteristics (income, marital, and educational status). However, bodyweight was a good predictor of underreporting of energy intake in obese women.Although the reason for this finding is unclear, it is possible that obese women purposelyreduce their recording of food, which made them appear to be ‘smaller eaters’.Johnson et al (1994) examined the relationship between physical characteristics and theunderreporting of energy intake in healthy older men and women. Reported energy intake127was obtained from a 3-day food diary and total energy expenditure was predicted by usinga published equation (Goran and Poehiman 1992). Predicted total energy expenditure wassignificantly higher than reported energy intake in both men and women. On average, menunderreported their intake by 12% and the women by 24%. Also, the over- andunderreporting of energy intake were not significantly correlated with any of the measuredphysical variables in the men. Among the women, underreporting of energy intakeincreased as FM and %BF increased. Percent body fat explained the most variation inunderreporting of energy intake (r= -0.42, p=O.OOl). The major findings were that olderwomen underreported energy intake to a greater degree than did older men and increasingadiposity was an independent predictor of underreporting in older women. Due to lack ofindependent measurement of FM in the present study, we did not show the relationshipbetween underestimation and %BF. However, the high correlation betweenunderestimation and body weight in obese group suggest that %BF may be a goodpredictor of underreporting of energy intake.We have defined body weight that was associated with underreporting of energy intake inobese subjects. It will be helpful to use this knowledge and begin to apply correctionfactors to reported energy intakes. Unfortunately, the small numbers and undefinedvariables in this study did not provide definitive markers and likely provided biasedestimates of intake.The fUture research should address the validity of the BIA, IDS and DEXA methods toestimate the changes in body composition during weight loss in patients with abnormal128water and electrolyte distributions using a multicompartmental assessment of bodycomposition model in controlled studies. Adequate cross-validation of DEXA should beperformed as has been done for BIA and IDS methods. Addition of presurgical TEEmeasurement would certainly strengthen the experimental design and provide valuableinformation about influence of ileogastrostomy on TEE. Improvement of energy intakemeasurement is necessary to clarify the role of reduced energy intake in the weight lossfollowing intestinal procedures. Alternatively, further studies are needed to identifycorrelates of underreporting and to correct the underreporting using conversion factors.1296. SUMMARY AND CONCLUSIONSThe results of this investigation indicate that ileogastrostomy induces a significant declinein FFM and FM measured by IDS method, BIA or DEXA, while the percentage of FFMand FM was not changed significantly. Bone mineral content determined by DEXA wasnot affected by ileogastrostomy. From the present study, it cannot be concluded whichmethod most accurately assesses body composition in morbidly obese subjects. Isotopedilution method and DEXA seem to be applicable to detect the changes in bodycomposition after intestinal bypass surgery. However, BIA was not a good choice forthese patients, especially during the very early stage.Basal energy expenditure declined slightly but insignificantly, however, a substantial andcontinuous decline in TEF was induced by ileogastrostomy. The direction of changes inTEE is still unknown because we did not measure presurgical TEE in this study. The TEEduring the short-term energy balance study demonstrated a significantly close correlationwith weight loss. Therefore, we concluded that energy expenditure may be an importantfactor in the weight loss after intestinal bypass procedures.Increased fecal energy loss was identified to be an important factor in the weight loss inthe present study, although its relative importance cannot be determined due to the smallsample size. We did not find a close relationship between urinary energy and weight lossafter ileogastrostomy.The urinary energy loss was comparable with normal individuals.130Therefore, we believed that the difference existed in the energy balance equation mayresult from misreporting of energy intake as reported in many recent publications. Thenotion was supported by the validation of reported energy intake in normal-weightsubjects. There were discrepancies between energy intake and expenditure observed inboth normal-weight and obese subjects. Misreporting of energy intake appears to occur inboth normal-weight and obese populations. The use of DLW method as an independentmarker of food intake has raised serious concern about the validity of much of the food-intake data published previously and the conclusion they have drawn. Therefore, reducedenergy intake as a major cause in the weight loss following intestinal bypass proceduresneed to be further evaluated. To a limited extent, this study has provided the data for thedirection and magnitude of misreporting of actual energy intake in normal-weight andobese subjects.These results have provided valuable information not only about weight loss andassessment of this weight loss but also about changes in energy metabolism. Energyexpenditure was first proved as a factor in the weight loss after intestinal bypassprocedures. The most challenging findings were to identif,’ the serious problems in energyintake data, although our study cannot provide definitive solution to correct this bias inpresent dietary assessment approaches.131REFERENCESAcheson KJ, Campbell IT, Edhoim OG, Miller DS, Stock MI. The measurement of foodand energy intake in man---an evaluation of some techniques. Am J Clin Nutr1980;33:1 147-1154.Albu I, Smolowitz J, Lichtman S, Heymsfield SB, Wang J, Pierson RN, Pi-Sunyer FX.Composition of weight loss in severely obese women: a new look at old methods.Metab 1992;41 :1068-1074.Andersen T, Backer OG, Stokholm KH, Quaade F. Randomized trial of diet andgastroplasty compared with diet alone in morbid obesity. N Engi J Med 1984;3 10:352-356.Apfelbaum M, Bostsarron J, Lacatis D. Effect of caloric restriction and excessive caloricintake on energy expenditure. Am J ClinNutr 1971;24:1405-1409.Balogh M, Kahn H, Medalie JH. Random repeat 24-hour dietary recalls. Am J Clin Nutr1971;24:304-3 10.Bandim LG, Schoeller DA, Cyr H, Dietz WH. Validity of reported energy intake inobese and nonobese adolescents. Am J Clin Nutr 1990;52:421-425.Barrett-Connor E. Nutrition epidemiology: how do we know what they ate? Am J ClinNutr 1991;54:S182-S187.Barry RE, Barisch J, Bray GA, Sperling MA, Morin U, Benfield J. Intestinal adaptationafter jejunoileal bypass in man. Am J Clin Nutr 1 977;30 :32-41.Beaton GH. Approaches to analysis of dietary data: Relationship between plannedanalyses and choice of methodology. Am J ClinNutr 1994;59:S253-S261.Benfield IR, Greenway FL, Bray GA, Barry RE, Lechago J, Mena I, Schedewie H.Experience with jejunoileal bypass for obesity. Surg Gynecol Obstet 1976;143 :40 1-410.Bessard T, Schutz Y, Jequier E. Energy expenditure and postprandial thermogenesis inobese women before and after weight loss. Am I Clin Nutr 1983;38:680-693.Besterman HS, Sarson DL, Blackburn AM. The gut hormone profile in morbid obesityand following jejunoileal bypass. Scand J Gastroenterol 1978; 13 : Si 5.Bingham SA. A dietary assessment of individuals, methods, accuracy, new techniques,and recommendations. Nutr Abstr Rev (Series A) 1987;57:705-742.132Black AE, Prentice AM, Coward WA. Use of food quotients to predict respiratoryquotients for the doubly labeled water method for measuring energy expenditure. HumNutr Clin Nutr I 986;40C:38 1-391.Block Q, Hartman AM. Issues in reproducibility and validity of dietary studies. Am JClinNutr 1989;50:1133-1138.Bradfield RB, Jourdan M. Energy expenditure of obese women during weight loss. Am JClin Nutr 1972;25:971-975.Bray GA. Effect of caloric restriction on energy expenditure in obese patients. Lancet1969;2:397-398.Bray GA, Barry RE, Castelnuova-Tedesco P. Intestinal bypass for obesity decreases foodintake and taste preference. Am J Clin Nutr 1976a;29:779-783.Bray GA, Dahms WT, Greenway FL, Marriott M, Molitch M, Atkinson R. Evaluation ofthe obese patient: 2. Clinical findings. JAmMed Ass 1976b;235:2008-2010.Bray GA, Greenway FL, Barry RE, Benfield JR, Fiser RL, Dahms WT, Atkinson RL,Schwartz AA. Surgical treatment of obesity: 1 Review of our experience and ananalysis of published reports. Tnt J Obes 1977;1 :331-367.Bray GA, Zachary B, Dahms WT. Eating patterns of massively obese individuals. J AmDiet Ass 1978;72:24-27.Bray GA, Dahms WT, Atkinson RL, Rodin J, Taylor I, Frame C, Schwartz A. Metabolicand behavioral differences between dieting and intestinal bypass. Horm Metab Res1979;1 1:648-654.Bray GA, Dahms WT, Atkinson RL, Mena I, Schwartz A. Factors controlling foodintake: A comparison of dieting and intestinal bypass. Am J Clin Nutr 1980;33:376-382.Bray GA. Obesity, a disorder of nutrient partitioning: the Mona Lisa Hypothesis. J Nutr1991;121:1 146-1162.Brewer C, White H, Baddeley H. Beneficial effects ofjejunoileostomy on compulsiveeating and associated psychiatric symptoms. Br Med J 1 974;4:3 14-315.Brill AB, Sandstead HE, Price R. Changes in body composition after jejunoileal bypassin morbidly obese patients. Am J Surg 1972;123:49-56.Bronstein MN, King JC. Energy expenditure of obese pregnant women. FASEB J1988;2:A1091.133Brownell KD. Obesity and weight control: The good and bad of dieting. Nutr Today1 987;22(3):4-9.Buchan AMJ, Pederson RA, Koop I, Gourlay RH, Cleator 1GM. Morphological andfunctional alterations to a sub-group of regulatory peptides in human pancreas andintestine after jejuno-ileal bypass. Tnt J Obes 1993;17: 109-1 13.Bukkens SGF, McNeill G, Smith JS, Morrison DC. Postprandial thermogenesis in post-obese women and weight-matched controls. Tnt J Obes 199 l;15: 147-154.Burke BS. The dietary history as a tool in research. JAm Diet Assoc 1947;23:1041-1046.Burton BT, Foster WR, Hirsch J, van Itallie TB. Health implications of obesity: An Nilconsensus development conference. Tnt J Obes 1985;9: 155-169.Canada, Health and Welfare.: Distribution of men and women at increasing levels of theBMI. Canadian Guidelines for healthy weights 1988;67-68.Canada, Health and Welfare. Prevalence of obesity. Report of the Task Force on theTreatment of Obesity 1991 ;6-7.Chan CB, Buchan AMJ, Green KA, Pederson RA. The effect ofjejunoileal bypass (JIB)in the obese zucker rat on a sub-group of enteroendocrine cells. Tnt J Obes987;1 1:285-293.Cleator 1GM, Dingee CK, Birmingham CL, MacFarlane 1K, Gourlay RH. Themechanism of weight loss after ileogastrostomy for morbid obesity. Obes Surg1991;1 :271-274.Cleator 1GM, Gourlay RH. Ileogastrostomy for morbid obesity. Can J Surg 1988;3 1:114-116.Condon SC, James NJ, Wise L, Alpers DII. Role of caloric intake in the weight loss afterjejunoileal bypass for obesity. Gastroenterol 1978;74:34-37.Corso PF, Joseph WL. Intestinal bypass in morbid obesity. Surg Gynecol Obstet1974;138:1-5.Crisp AH, Kalucy RS, Psych MRC, Pilkington TRE, Gazet J-C. Some psychosocialconsequences of ileojejunal bypass surgery. Am J Clin Nutr 1977;30:109-119.Culebras TM, Fitzpatrick GF. Total body water and the exchangeable hydrogen: TI. Areview of comparative data from animals with a report of new data from the rat. Am JPhysiol 1977;232:R60-R65.134Cunningham J, Levitt M, Hendler R, Nadel E, Felig P. Energy expenditure in obesity:Increased metabolic rate in the fasting and postprandial state. Clin Res 198 1;29:681A.Dale D, van Saris WHM, ten Hoor F. Weight maintenance and resting metabolic rate 18-40 months after a diet/exercise treatment. Tnt J Obesity 1990;14:347-359.Dano P, Lenz K. External pancreatic function following three types of intestinal shuntoperation for obesity. Scand J Gastroenterol 1974;9:451-455.Davies PSW, Preece MA, Hicks CJ, Halladay D. The prediction of total body water usingbioelectrical impedance in children and adolescents. Ann Hum Biol 1988;15:237-240.DeLany JP, Schoeller DA, Hoyt RW, Askew EW, Sharp MA. Field use ofD2180 tomeasure energy expenditure of soldiers at different energy intakes. J Appi Physiol1989;67: 1922-1929.Deurenberg P, Weststrate JA, van Der Kooy K. Body composition changes assessed bybioelectrical impedance measurements. Am J Clin Nutr 1989;49:401-403.Dore C, Hesp R, Wilkins D, Garrow JS. Prediction of energy requirements of obesepatients after massive weight loss. Hum Nutr Clin Nutr 1982;36C:41-48.Elia M. Body composition analysis: An evaluation of 2 component models,multicomponent models and bedside techniques. Clin Nutr 1992;ll :114-127.FAO/WHO/UNU. Energy and protein requirements: Report of a joint FAO/WHO/UNIJexpert consultation. WHO Tech Rep 1973; Ser 724.Felig P, Cunningham J, Levitt M, Hendler R, Nadel E. Energy expenditure in obesity infasting and postprandial state. Am J Physiol 1983;244:E45-E5 1.Flatt JP. Importance of nutrient balance in body weight regulation. Diabetes 1988;4:571-581.Flatt JP. The difference in the storage capacities for carbohydrate and for fat, and itsimplications in the regulation of body weight. Ann NY Acad Sci 1987;499: 104-123.Fogel MR, Ravitch Mlvi, Adibi SA. Absorptive and digestive function ofjejunum afterjejunoileal bypass for treatment of obesity. Gastroenterol 1976;71:729-733.Forbes GB, Simon W, Amatruda TM. Is bioimpedance a good predictor of bodycomposition change? Am J Clin Nutr 1992;56:4-6.Foreyt JP, Goodrick GK. Factors common to successful therapy for the obese patient.135Med Sci Sports Exerc 1991;23:292-297.Froidevaux F, Schutz Y, Christin L, Jequier E. Energy expenditure in obese womenbefore and during weight loss, after refeeding, and in the weight-relapse period. Am JClin Nutr 1993;57:35-42.Fuller NJ, Jenn SA, Goldberg GR. Inter-observer variability in the measurement of bodycomposition. Eur J ClinNutr 1991;45:43-49.Garrow JS, Durrant ML, Mann S, Stalley SF, Warwick PM. Factors determining weightloss in obese patients in a metabolic ward. mt j Obes 1978;2:441-447.Garrow JS. Thermogenesis and obesity in man. in Bjorntorp P, Cairella M, Howard AN,eds: Recent Advances in Obesity Research. London, John Libbey, 1981 ;3 :230-238.Geissier CA, Miller DS, Shah M. The daily metabolic rate of the post-obese and the lean.AmJ ClinNutr 1987;45:914-920.Going SB, Massett MP, Hall MC, Bare LA, Root PA, Williams DP, Lohman TG.Detection of small changes in body composition by dual-energy x-ray absorptiometryAm J Clin Nutr 1993;57:845-850Goran MI, Poehiman ET. Total energy expenditure and energy requirements in healthyelderly persons. Metab 1992;41:744-753.Gourlay RH, Cleator 1GM. Jejunoileal bypass with drainage of the bypassed small bowelinto the stomach (ileogastrostomy). in Deitel M eds: Surgery for the Morbidly ObesePatient. Philadelphia, Lea & Febiger, 1989;91-98.Gray DS, Bray GA, Gemayel N. Effect of obesity on bioelectrical impedance. Am J ClinNutr 1989;50:255-260.Halberg D. A survey of possible surgical methods of treating obesity. Acta Chir Scand1978;482:S3 1-S32.Halverson JD, Koehler RE. Gastric bypass: Analysis of weight loss and factorsdetermining success. Surgery 1981 ;90 :446-455.Henry CL, Phyllis EJ. A simple, inexpensive method of determining total body waterusing a tracer dose ofD2O and infrared absorption of biological fluids. Am J Clin Nutr1985;41:363-370.Hunt SM, Groff it. Energy Balance and weight control. in Advanced Nutrittion andHuman Metabolism. St. Paul, West publishing company, 1990; 405-444.136James WPT, Davies HL, Bailes J, Dauncey MJ. Elevated metabolic rates in obesity.Lancet 1978;l :1122-1125.Jensen MD, Kanaley JA, Reed JE, Sheedy PF. Measurement of abdominal and visceralfat with computed tomography and dual energy x-ray absorptiometry. Am J Clin Nutr1995;61 :274-278.Jequier E. Long-term measurement of energy expenditure in man. in Bjorntorp P, CairellaM, Howard AN, eds: Recent Advances in Obesity Research. London, John Libbey,1981;3:130-135.Jequier E. Energy expenditure in obesity. Clin Endocrinol Metab 1984;13:563-580.Jequier E, Schutz. New evidence for a thermogenic defect in human obesity. Tnt J Obes1985;9: 1-7.Johansson A, Forslund A, Sjodin A, Mailmin H, Hambraeus L, Ljunghall S.Determination of body composition-a comparison of dual-energy x-ray absorptiometryand hydrodensitometry. Am J Clin Nutr 1993;57:323-326.Johnson RK, Goran MI, Poehiman ET. Correlates of over- and underreporting of energyintake in healthy older men and women. Am J Clin Nutr 1994;59: 1286-1290.Jones PiN, Winthrop AL, Schoeller DA, Swyer PR, Smith J, Filler RM, Heim T.Validation of doubly labeled water for assessing energy expenditure in infants. PediatrRes 1987;21 :242-246.Jones PJH, Schoeller DA. Polyunsaturated: saturated ratio of dietary fat influences energysubstrate utilization in the human. Metab 1988a;37: 145-151.Jones PJH, Winthrop AL, Schoeller DA, Filler RIvI, Swyer PR, Smith J, Heim T.Evaluation of doubly labeled water for measuring energy expenditure during changingnutrition. Am J Clin Nutr 1988b;47:799-804.Jones PiN, Leitch CA. Validation of doubly labeled water for measurement of caloricexpenditure in collegiate swimmers. J Appl Physiol 1993a;74:2909-2914.Jones PJH, Jacobs I, Morris A, Ducharme MB. Adequacy of food rations in soldiersduring an arctic exercise measured using doubly labeled water. J Appl Physiol1993b;5: 1790-1797.Koopmans HS. Endogenous gut signals and metabolites control daily food intake. Tnt JObes 1990;14:S93-S104.Kopelman PG, Pilkington TRE, Gazet J-C, Lalloz MRA, Byfield PGH. Changes in137thyroid hormones in obese women following jejuno-ileal bypass surgery. Tnt J Obes1981;5:177-182.Kral JG. Overview of surgical techniques for treating obesity. Am J Clin Nutr1992;55:S552-S555.Kremen AJ, Linner JH, Nelson CH. An experimental evaluation of the nutritionalimportance of proximal and distal small intestine. Ann Surg 1954;140:439-448.Kushner RF, Schoeller DA. Estimation of total body water by bioelectrical impedanceanalysis. Am I Clin Nutr 1986;44:417-424.Kushner RF, Kunigk A, Alspaugh M, Andronis PT, Leitch CA, Schoeller DA. Validationof bioelectrical-impedance analysis as a measurement of changes in body compositionin obesity. Am J Clin Nutr 1990;52:219-223.Lansky D, Brownell KD. Estimates of food quantity and calories: Errors in self-reportamong obese patients. Am J Clin Nutr 1982;35:727-732.Laskey MA, Lyttle KD, Flaxman ME, Barber RW. The influence of tissue depth andcomposition on the performance of the Lunar dual-energy x-ray absorptiometer whole-body scanning mode. Eur I Clin Nutr 1992;46:39-45.Leibel RL, Hirsch J. Diminished energy requirements in reduced-obese patients, Metab1984;37: 164-170.Lichtman SW, Pisarska K, Berman ER, Pestone M, Dowling H, Offenbacher E, WeiselH, Heymsfield SB. Discrepancy between self-reported and actual caloric intake andexercise in obese subjects. N Eng I Med 1992;327:1893-1898.Lifson N, Gordon GB, Visscher MB, Nier AO. The rate of utilized molecular oxygen andthe source of heavy oxygen of respiratory carbon dioxide, studied with the aid ofheavy oxygen. I Biol Chem 1949;180:803-811.Lifson N, Gordon GB, McClintock R. Measurement of total carbon dioxide output bymeans ofD218O. I Appl Physiol 1955;7:704-710.Lissner L, Habicht JP, Strupp BI, Levitsky DA, Haas JD, Roe DA. Body compositionand energy intake: Do overweight women overeat and underreport? Am I Clin Nutr1989;49:320-325.Livingstone MBE, Prentice AM, Strain JJ, Coward WA, Black AE, Barker ME,McKenna PG, Whitehead RG. Accuracy of weighed dietary records in studies of dietand health. Br Med I 1990;300:708-712.138Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-free massusing bioelectrical impedance measurements of the human body. Am J Clin Nutr1985;41 :810-817.Lukaski HC, Bolonchuk WW, Hall CB, Siders WA. Validation of the tetrapolarbioelectrical impedance method to assess human body composition. J appl Physiol1986;60: 1327-1332.Lukaski HC. Methods for the assessment of human body composition: Traditional andnew. AmJClinNutr 1987;46:537-556.Lukaski HC, Bolonchuk WW. Estimation of body fluid volumes using tetrapolarbioeletrical impedance measurements. Aviat Space Environ Med 1988;59: 1163-1169.MacLean LD, Rhode BM, Shizgal FIM. Gastroplasty for obesity. Surg Gynecol Obstet1981;153:200-207.Marr JW. Individual dietary surveys: Purposes and methods. World Rev Nutr Dietet1971;13:105-164.Mayer J. Genetic, traumatic and environmental factors in the etiology of obesity. PhysiolRev 1953;33:472-508.Mazess RB, Barden HS, Bisek JP, Hanson J. Dual-energy x-ray absorptiometry for totalbody and regional bone-mineral and soft-tissue composition. Am J Clin Nutr1990;51:1 106-1112.McFarland RJ, Mg L, Parker W, Pilkington TRE, Gazet J-C. The dynamics ofweightloss after gastric partition for gross obesity. Tnt J Obes 1989; 13:81-88.Mertz W, Tsui JC, Judd JT, Reiser 5, Hallfrisch J, Morris ER, Steele PD, Lashley E.What are people really eating? the relation between energy intake derived fromestimated diet records and intake determined to maintain body weight. Am J Clin Nutr1991;54:291-295.Muffin MD, Jeor STS, Hill LA, Scott BJ, Daugherty SA, Koh YO. A new predictiveequation for resting energy expenditure in healthy individuals. Am J Clin Nutr1990;5 1:241-247.Miller DS, Parsonage S. Resistance to slimming. Lancet 1975; 1:733-779.Miller DS, PR Payne. A ballistic bomb calorimeter. Brit I Nutr 1959; 13:501-508.Mills MJ, Stunkard AJ. Behavioral changes following surgery for obesity. Am J Psychiat1976;133:527-531.139Morgan TB, York DA, Wasilewska A, Portman J. A study of the thermic response to ameal and to a sympathomimetic drug (ephedrine) in relation to energy balance in man.Br J Nutr 1982;47:21-32.Moore FD. Determination of total body water and solids with isotopes. Science1946;104: 157-160.Moore RB, Frantz ID Jr., Buchwald H. Changes in cholesterol pooi size, turnover rate,and fecal bile acid and sterol excretion after partial ileal bypass inhypercholesterolemic patients. Surgery 1 969;65 :98-100.Nagy KA. CO2 production in animals: Analysis of potential errors in the doubly-labelledwater methods. Am J Physiol 1980;238:R466-R473.Novick WM, Nusbaum N, Stein TP. The energy costs of elective surgery as measured bythe doubly labeled water(2H180) Fed Proc 1987;46:581.Oldroyd B, Bramley PN, Stewart SP, Simpson M, Truscott JG, Losowsky M, Smith MA.The measurement of total body fat by dual energy x-ray absorptiometry: Comparisonwith skinfold anthropometry, bioelectrical impedance and total body potassium. inEllis KJ & Eastman JD eds: Human body composition. New York, Plenum press1993 ;93-94.O’Leary JP, Thomas WC Jr, Woodward ER. Urinary tract stone after small bowel bypassfor morbid obesity. Am J Surg 1974;127: 142-147.Ott L. An Introduction to Statistical Methods and Data Analysis. Boston, PWS-KentPublishing Company, 1988.Pace N, Ruthburn EN. Studies on body composition: III. The body water and chemicallycombined nitrogen content in relation to fat content. J Biol Chem 1945; 158:685-691.Pekkarinen M. Major methodology in the collection of food consumption data. WorldRev Nutr Dietet 1970;12: 145-171.Phang PT, Rich T, Ronco J. A validation and comparison study of two metabolicmonitors. J Parent Enter Nutr l990;14(3):259-261.Pilkington TRE, Gazet JC, Ang L, Kalucy RS, Crisp AH, Day S. Explanations for weightloss after ileojejunal bypass in gross obesity. Bri Med J l976;1:1504-1505.Pilkington TRE. Weight loss following intestinal bypass. in Maxwell ID, Gazet J-C,Pilkington TR, eds: Surgical Management of Obesity. London, Academic PressInc., 1980;l71-179.140Pinson EA. Water exchanges and barriers as studied by the use of hydrogen isotopes.Physiol Rev 1952;32:123-134.Poehiman ET, Horton ES. The impact of food intake and exercise on energy expenditure.Nutr Rev 1989;47: 129-13 7.Prentice AM, Black AE, Coward WA, Davies HL, Goldberg GR, Murgatroyd PR,Ashford J, Sawyer M, Whiteheaci RG. High levels of energy expenditure in obesewomen. Br Med J 1986;292:983-987.Price RA. Genetics of human obesity. Ann Behav Med 1987;9:9-14.Ravussin E, Burnand B, Schutz Y, Jequier E. Twenty-four-hour energy expenditure andresting metabolic rate in obese, moderately obese, and control subjects. Am J ClinNutr 1982;35:566-573.Ravussin E, Burnanci B, Schutz Y, Jequier E. Energy expenditure before and duringenergy restriction in obese patients. Am J Clin Nutr 1985;41 :753-759.Reeder BA, Angel A, Ledoux M, Rabkin SW, Young TK, Sweet LE. Obesity and itsrelation to cardiovascular disease risk factors in Canadian adults. Can Med Assoc J1992;146(1 l):2009-2019.Riumallo JA, Schoeller DA, Barrera G, Uauy R. Energy expenditure in underweight free-living adults: impact of energy supplementation as determined by doubly labeledwater and indirect calorimetry. Am J Clin Nutr 1989;49:239-246.Roberts SB, Heyman MB, Evans WJ, Fuss P, Tsay R, Young VR. Dietary energyrequirements of young adult men, determined by using the doubly labeled watermethod. Am J Clin Nutr 1991;54:499-505.Robinson RG, Folstein IVIF, McHugh PR. Reduced caloric intake following small bowelbypass surgery: a systematic study of possible causes. Psychol Med 1979;9:37-53.Rodin J, Moskowitz HR, Bray GA. Relationship between obesity, weight loss, and tasteresponsiveness. Physiol Behav 1976; 17:591-597.RoubenoffR, Kehayias JJ, Dawson-Hughes B, Heymsfield SB. Use of dual-energy x-rayabsorptiometry in body composition studies: not yet a “gold standard”. Am J ClinNutr 1993;58:589-591.Ryde SJS, Laskey MA, Morgan WD, Compston JE. A comparison between neutronactivation and dual energy x-ray absorptiometry for the measurement of body fat. inEllis KJ & Eastman ID eds: Human body composition. New York, Plenum press1411993 ;3 67-370.Sarson DL, Scopinaro N, Bloom SR. Gut hormone changes after jejunoileal (JIB) orbiliopancreatic (BPB) bypass surgery for morbid obesity. Tnt J Obes 1981 ;5 :471-480.Schoeller DA, van Santen E, Peterson DA, Dietz W, Jaspan J, Klein PD. Total bodywater measurement in humans with 180 and 2H labeled water. Am J Clin Nutr1980;33 :2686-2693.Schoeller DA, van Santen E. Measurement of energy expenditure in humans by doublylabeled water method. J Appi Physiol 1982;53:955-959.Schoeller DA, Leitch CA, Taylor PB. Assessment of human energy expenditure withdoubly labled water method. in Gussler J eds: Appling New Technology to NutritionalAssessment: Report of the Ross Round Table on Medical Issues. Ohio, 1986;5-9.Schoeller DA, Jones Pifi. Measurement of total body water by isotope dilution: a unifiedapproach to calculation. in Ellis KJ, Yasumura S, Morgan WD, eds: In Vivo BodyComposition Analysis. London, 1987; 13 1-137.Schoeller DA, Taylor PB. Precision of the doubly labeled water method using the two-point calculation. Human Nutr Clin Nutr 1987;41C:21 5-223.Schoeller DA. Measurement of energy expenditure in free-living humans by using doublylabeledwater. JNutr 1988;l18:1278-1289.Schoeller DA, Bandim LG, Dietz WH. Inaccuracies in self-reported intake identified bycomparison with the doubly labeled water method. Can J Physiol Pharmacol1990;68:941-949.Schoeller DA. Isotope dilution methods. in Bjorntorp P, BrodoffBN eds: Obesity.Philadelphia, J.B. Lippincott Company, 1992; 80-88.Schols AMWJ, Wouter EFM, Soeters PB, Westerterp KR. Body composition bybioelectrical-impedance analysis compared with deuterium dilution and skinfoldanthropometry in patients with chronic obstructive pulmonary disease. Am J Clin Nutr1991;53:421-424.Schultink WJ, Lawrence M, van Raaij JM, Scott WM, Hautvast JGAJ. Body compositionof rural Beninese women in different seasons assessed by skinfold thickness andbioelectrical-impedance measurements and by a deuterium oxide dilution technique.Am I ClinNutr 1992;55:321-325.Schutz Y, Golay A, Felber J-P, Jequier E. Decreased glucose-induced thermogenesis afterweight loss in obese subjects: a predisposing factor for relapse of obesity? Am J Clin142Nutr 1984;39:380-387.Scott HW, Brill AB, Price RR. Body composition in morbidly obese patients before andafter jejunoileal bypass. Ann Surg 1975; 182:395-404.Scott WH, Sandstead HH, Brill AB, Burko H, Younger RK. Experience with a newtechnique of intestinal bypass in the treatment of morbid obesity. Ann Surg1971;174:560-572.Segal KR, Gutin B, Albu J. Thermic effects of food and exercise in lean and obese menof similar lean body mass. Am J Physiol 1987a;252:E1 10-El 17.Segal KR. Comparison of indirect calorimetric measurements of resting energyexpenditure with a ventilated hood, face mask, and mouthpiece. Am J Clin Nutr1987b;45: 1420-1423.Segal KR, van Loan M, Fitzgerald PT, Hodgdon JA, van Itallie TB. Lean body massestimation by bioelectrical impedance analysis: a four-site cross-validation study. AmJ ClinNutr 1988;47:7-14.Segal KR, Gutin B, Presta E, Wang J, and van Itallie TB. Estimation of human bodycomposition by electrical impedance methods: a comparative study. J Appi Physiol1985;58: 1565-1569.Segal KR, Edano Albert, Blando Lorna. Comparison of thermic effects of constantand relative caloric loads in lean and obese men. Am J Clin Nutr 1990;51 :14-21.Shetty PS, Jung RT, James WPT, Barrand MA, Callingham BA. Postthermogenesis inobesity. Clin Sci 1981;60:519-525.Sin WE. The gross composition of the body. in Lawrence iTT, Tobias CA, eds: Advancesin Biological and Medical Physics. New York, Academic Press Inc., 1956;4:239-280.Sorensen TIA, Krag E. Fat digestion after jejunoileal bypass operation for obesity. ScandJ Gastroenterol 1976;11:491-495.Southgate DAT, Durmn JVGA. Caloric conversion factors. An experimentalreassessment of the factors used in the calculation of the energy value of human diets.BrJNutr 1970;24:517-535.Su W, Jones PITT. Dietary fatty acid composition on energy accretion in rats. J Nutr1993:123:2109-2114.Svendsen OL, Haarbo J, Hassager C, Christiansen C. Accuracy of measurements of bodycomposition by dual-energy x-ray absorptiometny in vivo. Am J Clin Nutr1431993;57:605-608.Swaminathan R, King RFGJ, Holmfield J, Siwek RA, Baker M, Wales JK. Thermiceffect of feeding carbohydrate, fat, protein and mixed meal in lean and obese subjects.AmjClinNutr 1985;42:177-181.Tarasuk V, Beaton GH. Statistical estimation of dietary parameters: implications ofpatterns in within-subject variation---a case study of sampling strategies. Am J ClinNutr 1992;55:22-27.Thomasset A. Bioelectrical properties of tissue impedance measurements Lyon Med1962;207:107-1 18.Vazquez JA, Janosky JE. Validity of bioelectrical impendance analysis in measuringchanges in lean body mass during weight reduction. Am J Clin Nutr 1992;56:7-13.Warnold I, Carigren G, Krotkiewski M. Energy expenditure and body composition duringweight reduction in hyperplastic women. Am J Clin Nutr 1 978;3 1:750-763.Weir WE. New methods for calculating metabolic rate with special emphasis to proteinmetabolism. J Physiol 1949;109: 1-9.Welle L, Amatruda JM, Forbes GB. Resting metabolic rates of obese women after rapidweight loss. J Clin Endocrinol Metab 1984;59:41-44.Welle S, Forbes GB, Statt M. Energy expenditure under free living conditions in normal-weight and overweight women. Am J Clin Nutr 1992; 55: 14-21.Welle S. Metabolic responses to a meal during rest and low-intensity exercise. Am J ClinNutr 1984;40:990-994.Wellens R, Chumlea WC, Guo S, Roche AF, Reo NV, Siervogel RM. Body compositionin white adults by dual-energy x-ray absorptiometry, densitometry, and total bodywater. Am J Clin Nutr 1994;59:547-555.Weisman RE. Surgical palliation of massive and severe obesity. Am J Surg1973;125:437-446.Weststrate JA, Weys PJ, Poortvliet EJ, Deurenberg P, Hautvast GA. Diurnal variation inpostabsorptive resting metabolic rate and diet-induced thermogenesis. Am J Clin Nutr1989;50:908-914.Weststrate JA. Resting metabolic rate and diet-induced thermogenesis: a methodologicalreappraisal. Am J ClinNutr 1993;58:592-601.144Westerterp KR, Brouns F, Saris WTIM, Hoor FT. Comparison of doubly labeled waterwith respirometry at low- and high-activity levels. J App! Physiol 1988;65:53-56.Westerterp KR, Meijer GAL, Saris WHM, Hoor FT. Physical activity and sleepingmetabolic rate. Med Sci Sports Exerc 1990;23:166-170.Westerterp KR, Saris WHM, Soeters PB Hoor FT. Determinants of weight loss aftervertical banded gastroplasty. Tnt J Obes 1991;15:529-534.Wi!more iTT. Body composition in sport and exercise: directions for future research. MedSci Sports Exerc 1983;15:21-31.Wise L, Stein T. The pathogenesis of diarrhea after bypass of the small intestine. SurgGynecol Obstet 1976;142:686-691.Yale CE. Gastric surgery for morbid obesity. Arch Surg 1989;124:941-947.145Appendix 1. Invitation Letter for Obesity StudyDear_____________I am writing to invite you to participate in a study “Examination of weight loss andchanges in energy metabolism following ileogastrostomy” because you are scheduled forthe surgical procedure. Obesity can be effectively corrected by ileogastrostomy. However,the mechanism ofweight loss is still unknown. The purpose of this study is to helpdetermine factors associated with weight loss following ileogastrostomy.The study will last for 3 months after surgery. If you decide to take part in this study, youwill need to:1). have metabolic tests done at St. Paul’s Hospital at 4 different times (one beforesurgery, the others 1, 2 and 3 months following surgery;2). Write down everything you eat and drink for 3 days before each test;3). Have total body scan measured by DEXA twice; and4). Collect urine and feces completely for 5 days within 6-8 weeks following surgery.The total time commitment is about 25 hours. You will be compensated with $100 foryour participation in this study. Your participation is voluntary. You may decide not toparticipate or withdraw from the study at any time without affecting your normaltreatment. If you are interested in the study, please call Dr. Jones at 822-6253.Sincerely146Appendix 2. Sample Consent FormCONSENT FORMTitle: Examination of Weight Loss and Energy Metabolism FollowingIleogastrostomyInvestigators: Dr. Peter Jones #822-6253Dr. lain Cleator #681-1513Dr. Laird Birmingham # 631-5269You are being invited to participate in a study, “Examination of Weight Loss and EnergyMetabolism Following Ileogastrostomy”, because you are scheduled for the surgicalprocedure. You may decide not to participate or may withdraw from the study at any timewithout affecting your normal treatment.Purpose of the Study:Obesity can be effectively corrected by ileogastrostomy. However, the mechanism ofweight loss is still unknown. The purpose of this study is to help determine factorsassociated with weight loss following ileogastrostomy.Procedures:If you decide to take part in this study, you will need to go to St. Paul’s Hospital at 4different times, once before surgery and the others at 1, 2 and 3 months after surgery, formetabolic tests. During the tests, you will lie on the bed and wear a plastic hood. Yourbreath will go to a monitor which can analyze how much oxygen you inhale, how muchcarbon dioxide you exhale and how much energy your body is using. You will berequested to drink a small amount of stable isotope (no radioactivity). We will collect asmall amount of saliva sample to determine the fat mass and fat free mass in your body.Each metabolic test will require 6 hours of your time.You will need to write down everything you eat and drink for 3 days before each test.In addition, you need to collect urine and feces completely for 5 days within 6-8 weeksfollowing surgery. We will provide containers and deliver the freezer to your home.DEXA (to measure the changes of fat, muscles and bone density in your body aftersurgery) will be done twice, once before surgery and the other at 6 wks following surgeryon the same day of the metabolic test. The scan will be performed in the Department of147Nuclear Medicine at St. Paul’s. The scanner will take a series of pictures to estimate yourbody compartments.Total Time Commitment:The study will last for 3 months after surgery. The total time commitment is about 25hours.Risks and Significant Side Effects:There are no risks associated with the metabolic tests. During the DEXA scan, you will beexposed to a very small amount of radiation. The radiation exposure is equivalent to 10%of that a regular chest x-ray.Potential Benefits:There will be no direct benefit to you. However, improved understanding of themechanism of weight loss after surgery has potential application for effecting or achievingweight loss without surgery and for maintaining weight loss after such surgery.Monetary Compensation:You will be compensated with $100 for your participation in the study.Confidentiality:All data collected for this study will be kept confidential. Only Dr. Jones, Dr. Cleator andDr. Birmingham will have access to the data. The records will be identified by codenumbers, not by patient names.If you have any questions or concerns at any time during the study, you may contact Dr.Jones or Dr. Cleator at the numbers listed above.***********************I have read the above information. I freely consent to participate in the study andacknowledge receipt of a copy of the consent form.Signature of Participant DateSignature of Witness Date148Appendix 3. Food Record InstructionsPLEASE READ CAREFULLY:1. Write down EVERYTHING you eat and drink. Be sure to include all SNACKS andALCOHOL. Record immediately after each meal and snack to ensure accuracy.2. Write down HOW MUCH you eat and drink using the scale provided.A. Try to use GRAM measures.B. Be specific about the TYPE OF FOOD, BRAND NAME IF APPLICABLE, HOWTHE FOOD WAS PREPARED, AND CONTENT OF MIXED DISHES.C. For combination items, list each item separately, e.g. a cheeseburger would bedescribed as: bun, cooked ground beef; processed cheese, butter, relish, etc.D. IF THE FOOD IS PREPARED BY SOMEONE OTHER THAN YOURSELF:please try to describe the contents of the dish that is served to you and estimate theamount.E. If you leave some foods you have weighed, please write down HOW MUCH andthe TYPES.F. Don’t forget the EXTRAS? e.g. sugar on cereal or in coffee, dressing on salad,candy, soft drinks, alcohol.149Appendix 4. Sample of Food RecordName________________ Subject #_________________ Age years oldHeight cm Weight kgTest #_________________ Date_______________________CommentsMeals Types and Amounts (g) left-over FoodsBreakfastLunchSupperOthers150Appendix 5. Sample of Test MealEstimation of energy requirement for individual subject using Ivlifflin’ s equation:REE(male) 1 Oxbody weight (kg)+6 .25 xheight (cm)-5 xage+5REE(female)= 1 Oxbody weight (kg)+6 .25 xheight (cm)-5 xage- 161multiplied by 1.7 and 1.6 activity factor for male and female, respectively.Because this group of subjects was morbidly obese, the ideal body weight was used tocalculate the energy requirement for individual subjectThe energy content for test meal was derived from the estimated energy requirement forindividual subject times 30% for breakfast. This amount of energy was distributed to fat,CHO and protein, and served to the subjects as following foods.Food Item Quantitycereals gram2% milk mlorange juice mlsoft sunflower margarine gramomelette gramegg whitesegg yolksoft sunflower margarinewhole wheat bread grami51Appendix 6. Sample of Dual Energy X-ray AbsorptiometryLSt. Paul’. Bospital Wuclear Kedtc. Dept.1081 Rurrard St. Vaco.v.r, IC. Ph. 631.500*- It &ntc Cl.UCASIaJID J3451 2.tjht 1604g. 40 Sax Teasli V.4ht 119— ] H lody 04106194 1I—-lose tio, iot or dtagsoetsTotal (I ) I 1.354 —Total (g) I 3313Total s.(g)s 52263Total Pat Hiss (g)a 61717Total Pat 2 i 52.6Sin OW! Fat 2 i 45.2!ro:.k OW! Tat 2 I 43.0Soft Tissu. Pat 2 ; 54.12 ThHCIPN 6.0lTD CT Los Totat 14 See O.tde for etbir C’s.IS alSO —. RIO !s. Cl.7S eo be. I.S.O I I.S.O Celib. O4jO6I4cc**ii ROfl, mvs PoRzkP24s m NOTIN TEE FIELD OP VIEWSCAN If?OPMATIcJ)’. lodyScan P.wt.-S. 04106194 1 Z.solutlcc 6.5 x 13.0 —AzaJyc.fs Dat. 04106194 $..d 180 ..I.C,Jtbr,ttao1St. 04106194 b(de4 61.75f.chntc!aa 83 2o,t/Scann.r 2.3.0 I 1.3.0P4rstcta.o DR. CLEATOR 4.nal,a1. l.vls!aa 2.3.0DETAILED RESULTI81W !I4C AREA LSSIGTN WiDTH LEAN MASS TAT MASSglc.’ g caHead 2.408 380.0 240.9 3596 1037Trunk 1.243 1349 lOU 27326 36353Ab4oc. 1.379 640.7 403,7 11676 16921Ar.a 1.013 205.1 202.0 3103 3341L.1. 1.284 117* 917.4 18236 18766Total 1.354 3313 2447 32263 17l7L 52Appendix 7. Preparation of Doubly Labeled Water and Instruction Sheetfor Total Energy Expenditure Measurement.Original doubly labeled water was filtered and weighed to estimate exact amount of 180and 2D in the solution. The final ratio of 1 80:2D was 2.5:1 which was assumed as theoptimal dose (Schoeller 1988).PLEASE READ CAREFULLY:Total Energy Expenditure Instructions1. Fill out the information sheet in the blank.2. Collect 3h and 4h saliva samples after administration of DLW.3. Collect urine samples (10 ml) in the morning and afternoon at day 1.4. collect urine samples at day 7 and day 14 as you did in day 1.5. Select 5 days during this two-week period to COMPLETELY collect fecal and urinarymaterials into the containers provided. In the meantime, write down EVERYTHThTG youeat and drink in these 5 days using the scale as you did in presurgical test.5. If you have any comments, please write down in the information sheet in detail such asdiarrhea, vomiting.6. On day 15, you need to visit the Metabolic Lab. and a further dose of D20 will be givenand 3h and 4h saliva samples will be collected.Total Energy Expenditure Information SheetDate Initial Body Weight_______ Final Body Weight_______Date and time for sample collectionDay 1Day 3Day 5Day 7Day 9Day 11Day 13Comments__________ Day 2_________Day4______________Day 6Day 8Day 10Day 12Day 14Comments____________NameDose of DLW153Appendix 8. Postoperative ComplicationsType of complication No. ofpatients Interval (weeks)Wound infection 2, 6 2-3Severe nausea and vomiting 4 7-12Diarrhea almost all patients except 1 throughout the studyneed for antidiarrheal agent most patientsfoul-smelling flatus most patientsthirst 1,5,11,15 4-12electrolyte imbalancehypokalemia 2 8Anemia 2 8hypoproteinemia 2, 4 8stomal ulcer 4 12disturbances of liver functionscholelithiasis 3 7-8154Appendix 9. The Relationship Between BodyComposition and Energy ExpenditureValues are mean±SEM (n=7). BEE, BW, FFM, FM and El represent basal energyexpenditure, body weight, fat-free mass, fat mass and energy intake. rBEE, rBW, rLBM,rFM and rEl are the reduced BEE, BW, FFM, FM and El calculated from correspondingpresurgical data minus postsurgical data.R1, R2, P..3 and R4 are the correlation coefficients between BEE and BW, FFM, FM, andEl, respectively. P1, P2, P3 and P4 are the corresponding probabilities for these variables.BEE-1(kcal.min )1.16±0.071.10±0.061.07±0.051.04±0.04Omo.1 mo.2 mo.3 mo.BW(kg)111.1±3.2105.9±3.0101.7±2.493.9±3.1FFM(kg)59.8±2.857.1±2.75 5.6±2.852.9±3.4FM(kg)51.2±2.648.8±2.246.1±2.341.0±2.4El1(kcal.d )2022±29 51393±201998±1581164±201R1(P1)0.49(0.27)0.60(0.15)0.75(0.05)0.82(0.02)rBEE rBW rFFM rFM rEl rR1rPi(P2)0.12(0.80)0.16(0.73)0.27(0.55)0.74(0.06)rR2rP20.50(0.25)0.34(0.46)0.21(0.65)R3(P3)0.73(0.06)0.62(0.14)0.45(0.32)0.01(0.97)rR3rP30.21(0.66)0.13(0.78)0.31(0.50)R4(P4)0.04(0.93)0.89(0.01)0.74(0.06)0.790.03rR4rP40.09(0.85)0.30(0.51)0.22(0.63)imo. 0.06±0.06 5.2±0.6 2.7±0.6 2.5±0.6 628±2990.31(0.49)2mo. 0.02±0.02 4.2±0.7 1.5±0.5 2.7±0.7 1023±2310.34(0.46)3mo. 0.03±0.02 7.6±1.2 2.7±1.0 5.1±0.9 858±1410.04(0.93)


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items