Torgeir T. Søvik, MD; Erlend T. Aasheim, MD, PhD; Osama Taha, MD; My Engström, RN; Morten W. Fagerland, MSc, PhD; Sofia Björkman, RD; Jon Kristinsson, MD, PhD; Kåre I. Birkeland, MD, PhD; Tom Mala, MD, PhD; Torsten Olbers, MD, PhD
Acknowledgment: The authors thank Tomm Bernklev, PhD, for support with statistical analysis, and Diana Lustgarten, RN, and Niclas Johansson, RN, for data collection at Sahlgrenska University Hospital.
Grant Support: By research fellowship grants from the South-Eastern Norway Regional Health Authority (Drs. Søvik and Aasheim) and a postdoctoral research grant from Sahlgrenska University Hospital (Dr. Olbers).
Potential Conflicts of Interest: Disclosures can be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M10-2877.
Reproducible Research Statement:Study protocol: Available from Dr. Søvik (e-mail, firstname.lastname@example.org). Statistical code: Available from Dr. Fagerland (e-mail, email@example.com). Data set: Not available.
Requests for Single Reprints: Torgeir T. Søvik, MD, Department of Gastrointestinal Surgery, Oslo University Hospital Aker, PO Box 4959 Nydalen, N-0424 Oslo, Norway; e-mail, firstname.lastname@example.org.
Current Author Addresses: Dr. Søvik: Department of Gastrointestinal Surgery, Oslo University Hospital Aker, PO Box 4959 Nydalen, N-0424 Oslo, Norway.
Dr. Aasheim: Department of Endocrinology, Oslo University Hospital Aker, PO Box 4959 Nydalen, N-0424 Oslo, Norway.
Dr. Taha: Department of Plastic Surgery, Bariatric Surgery Unit, Assiut University Hospital, 71526 Assiut, Egypt.
Ms. Engström: Department of Surgery, Sahlgrenska University Hospital, S-41345 Gothenburg, Sweden.
Dr. Fagerland: Unit of Biostatistics and Epidemiology, Oslo University Hospital Ullevål, PO Box 4956 Nydalen, N-0424 Oslo, Norway.
Ms. Björkman: Department of Clinical Nutrition, Sahlgrenska University Hospital, S-41345 Gothenburg, Sweden.
Dr. Kristinsson: Department of Morbid Obesity and Bariatric Surgery, Oslo University Hospital Aker, PO Box 4959 Nydalen, N-0424 Oslo, Norway.
Dr. Birkeland: Department of Endocrinology, Oslo University Hospital Aker, PO Box 4959 Nydalen, N-0424 Oslo, Norway.
Dr. Mala: Department of Gastrointestinal Surgery and Department of Morbid Obesity and Bariatric Surgery, Oslo University Hospital Aker, PO Box 4959 Nydalen, N-0424 Oslo, Norway.
Dr. Olbers: Department of Gastrosurgical Research & Education, Sahlgrenska University Hospital, S-41345 Gothenburg, Sweden.
Author Contributions: Conception and design: T.T. Søvik, E.T. Aasheim, M. Engström, S. Björkman, J. Kristinsson, K.I. Birkeland, T. Mala, T. Olbers.
Analysis and interpretation of the data: T.T. Søvik, E.T. Aasheim, O. Taha, M. Engström, M.W. Fagerland, S. Björkman, K.I. Birkeland, T. Mala, T. Olbers.
Drafting of the article: T.T. Søvik, E.T. Aasheim.
Critical revision of the article for important intellectual content: T.T. Søvik, E.T. Aasheim, O. Taha, M. Engström, M.W. Fagerland, J. Kristinsson, K.I. Birkeland, T. Mala, T. Olbers.
Final approval of the article: T.T. Søvik, E.T. Aasheim, O. Taha, M. Engström, M.W. Fagerland, J. Kristinsson, K.I. Birkeland, T. Mala, T. Olbers.
Provision of study materials or patients: T.T. Søvik, E.T. Aasheim, O. Taha, M. Engström, J. Kristinsson, T. Olbers.
Statistical expertise: T.T. Søvik, M.W. Fagerland.
Obtaining of funding: E.T. Aasheim, J. Kristinsson, T. Mala, T. Olbers.
Administrative, technical, or logistic support: T.T. Søvik, E.T. Aasheim, M. Engström, J. Kristinsson, K.I. Birkeland, T. Mala.
Collection and assembly of data: T.T. Søvik, E.T. Aasheim, O. Taha, M. Engström, S. Björkman, T. Olbers.
Søvik TT, Aasheim ET, Taha O, Engström M, Fagerland MW, Björkman S, et al. Weight Loss, Cardiovascular Risk Factors, and Quality of Life After Gastric Bypass and Duodenal Switch: A Randomized Trial. Ann Intern Med. 2011;155:281-291. doi: 10.7326/0003-4819-155-5-201109060-00005
Download citation file:
Published: Ann Intern Med. 2011;155(5):281-291.
Gastric bypass and duodenal switch are currently performed bariatric surgical procedures. Uncontrolled studies suggest that duodenal switch induces greater weight loss than gastric bypass.
To determine whether duodenal switch leads to greater weight loss and more favorable improvements in cardiovascular risk factors and quality of life than gastric bypass.
Randomized, parallel-group trial. (ClinicalTrials.gov registration number: NCT00327912)
2 academic medical centers (1 in Norway and 1 in Sweden).
60 participants with a body mass index (BMI) between 50 and 60 kg/m2.
Gastric bypass (n = 31) or duodenal switch (n = 29).
The primary outcome was the change in BMI after 2 years. Secondary outcomes included anthropometric measures; concentrations of blood lipids, glucose, insulin, C-reactive protein, and vitamins; and health-related quality of life and adverse events.
Fifty-eight of 60 participants (97%) completed the study. The mean reductions in BMI were 17.3 kg/m2 (95% CI, 15.7 to 19.0 kg/m2) after gastric bypass and 24.8 kg/m2 (CI, 23.0 to 26.5 kg/m2) after duodenal switch (mean between-group difference, 7.44 kg/m2 [CI, 5.24 to 9.64 kg/m2]; P < 0.001). Total cholesterol concentration decreased by 0.24 mmol/L (CI, −0.03 to 0.50 mmol/L) (9.27 mg/dL [CI, −1.16 to 19.3 mg/dL]) after gastric bypass and 1.07 mmol/L (CI, 0.79 to 1.35 mmol/L) (41.3 mg/dL [CI, 30.5 to 52.1 mg/dL]) after duodenal switch (mean between-group difference, 0.83 mmol/L [CI, 0.48 to 1.18 mmol/L]; 32.0 mg/dL [CI, 18.5 to 45.6 mg/dL]; P < 0.001). Reductions in low-density lipoprotein cholesterol concentration, anthropometric measures, fat mass, and fat-free mass were also greater after duodenal switch (P ≤ 0.010 for each between-group comparison). Both groups had reductions in blood pressure and mean concentrations of glucose, insulin, and C-reactive protein, with no between-group differences. The duodenal switch group, but not the gastric bypass group, had reductions in concentrations of vitamin A and 25-hydroxyvitamin D. Most Short Form-36 Health Survey dimensional scores improved in both groups, with greater improvement in 1 of 8 domains (bodily pain) after gastric bypass. From surgery until 2 years, 10 participants (32%) had adverse events after gastric bypass and 18 (62%) after duodenal switch (P = 0.021). Adverse events related to malnutrition occurred only after duodenal switch.
Clinical experience was greater with gastric bypass than with duodenal switch at the study centers.
Duodenal switch surgery was associated with greater weight loss, greater reductions of total and low-density lipoprotein cholesterol concentrations, and more adverse events. Improvements in other cardiovascular risk factors and quality of life were similar after both procedures.
South-Eastern Norway Regional Health Authority.
Two types of bariatric surgery are commonly used to treat severe obesity: gastric bypass and duodenal switch. Few controlled clinical trials of these procedures have been reported.
Patients with a body mass index between 50 and 60 kg/m2 were randomly assigned to undergo gastric bypass or duodenal switch. After 2 years, duodenal switch resulted in significantly greater reduction in body mass index than gastric bypass but was more commonly associated with adverse events. Dietary factors, quality of life, and cardiovascular markers varied with each procedure.
The effect of surgical experience on outcomes was not assessed.
Patients benefited from both types of bariatric surgery. The choice of procedure should be individualized.
The risk for obesity-related comorbid conditions increases progressively with weight (1, 2), and body mass index (BMI) is associated with increased mortality, primarily from cardiovascular disease and several types of cancer (3–5). Weight loss can reduce the risk for health complications (6, 7). Although severely obese persons may benefit from lifestyle intervention in the short term (8), few data on the long-term efficacy of nonsurgical interventions are available (9, 10). Recent studies have demonstrated that bariatric surgery can induce a sustained weight loss that is associated with favorable effects on obesity-related comorbid conditions (11), cancer incidence (12), quality of life (13), and death (14).
Recent survey data from the United States indicate that at least 5% of the population meets the criteria for bariatric surgery (15). The prevalence of severe obesity has increased relatively more than that of overweight and moderate obesity (16). About one third of patients who have bariatric surgery in the United States has a BMI of more than 50 kg/m2 (called superobesity) (17, 18). After Roux-en-Y gastric bypass surgery and achieving peak weight loss, these patients often still have a BMI above 35 kg/m2(19).
Evidence from nonrandomized studies suggests that biliopancreatic diversion with duodenal switch may lead to greater weight loss than gastric bypass (20, 21). To our knowledge, these procedures have not been compared in a randomized clinical trial. We therefore conducted a randomized trial of gastric bypass and duodenal switch in superobese patients to address the hypothesis that greater weight loss after duodenal switch is associated with improvements in cardiovascular risk factors and health-related quality of life. Previously published 1-year follow-up data from the same trial demonstrated a greater weight loss after duodenal switch (22, 23), and the present study contributes longer-term follow-up and supplementary data. We sought to compare gastric bypass with 1-stage duodenal switch surgery. We did not include patients with a BMI greater than 60 kg/m2; in such patients, duodenal switch is often performed in 2 stages to reduce surgery-related morbidity (24).
The study was an unblinded, prospective, randomized clinical trial. Participants were recruited from referrals for treatment of morbid obesity between March 2006 and August 2007. Two-year follow-up was completed in November 2009. Signed, informed consent was obtained from all participants. The trial protocol was approved by appropriate local ethics committees.
Recruitment, interventions, and data collection were performed at Oslo University Hospital Aker, Oslo, Norway, and Sahlgrenska University Hospital, Gothenburg, Sweden.
Participants were eligible for inclusion if they had a BMI of 50 to 60 kg/m2, were aged 20 to 50 years, and had not sustained previous weight loss. Exclusion criteria were a history of major abdominal or bariatric surgery, disabling cardiopulmonary diseases, cancer, long-term treatment with oral corticosteroids, and conditions associated with poor adherence. Before inclusion, participants were informed about the expected outcomes of both operations. The baseline assessment included recording of medical history and current medications and a physical examination. Blood samples were obtained after an overnight fast.
Randomization was computer-derived at Sahlgrenska University Hospital by using LabView software, version 7.1 (National Instruments, Austin, Texas), with stratification according to study center, sex, age (<35 years or ≥35 years), and BMI (<55 kg/m2 or ≥55 kg/m2) by using the minimization method. Blocking was not used. Participants and surgeons were masked to treatment allocation until 1 week before surgery.
Participants followed a low-calorie diet (1000 kcal/d) for 3 weeks before surgery. Both operations were performed at both study centers by using standardized laparoscopic techniques (Figure 1) (22, 25). In gastric bypass, the stomach was divided proximally to create a pouch of 25 mL and a gastric remnant. The jejunum was measured 50 cm from the ligament of Treitz and anastomosed to the pouch with the linear stapling technique. To create the alimentary limb, the jejunum was measured 150 cm distally from the gastrojejunal anastomosis and anastomosed to the 50-cm biliopancreatic limb. The small bowel was then transected between the 2 anastomoses to create a Roux-en-Y configuration with a long common channel, allowing food from the alimentary limb to be mixed with bile and pancreatic juices transported in the biliopancreatic limb. Duodenal switch consisted of longitudinal (“sleeve”) gastrectomy along a nasogastric tube of 30 to 32 Fr. The small bowel was measured from the cecum to create a common channel of 100 cm and an alimentary limb of 200 cm. The procedure then involved a hand-sewn anastomosis between the duodenum and the ileum, an anastomosis between the biliopancreatic limb and the ileum, and transection of the bowel between the anastomoses. Mesenteric defects were not closed in either procedure. Participants consumed a liquid diet for the first week after surgery and a semiliquid diet for 1 week, then gradually returned to normal food intake.
Bowel limb lengths for gastric bypass: alimentary limb, 150 cm; biliopancreatic limb, 50 cm; and common channel, variable length. Bowel limb lengths for duodenal switch: alimentary limb, 200 cm; biliopancreatic limb, variable length; and common channel, 100 cm. Figures by Ole-Jacob Berge, MD, and reproduced from reference 25 with permission.
The primary end point of the study was the change in BMI from baseline until 2 years after surgery. Secondary end points included changes in anthropometric measures, cardiovascular risk factors, health-related quality of life, body composition, vitamin concentrations, and adverse events. Weight was measured with participants wearing light clothing and no shoes, with 1 kg subtracted to account for clothing. Weight change was reported as the reduction in BMI units (kg/m2), total weight loss in kilograms, and percentage of body weight loss. In addition, the number of participants who had a BMI of 40 kg/m2 or higher at 2 years after surgery was evaluated. Waist circumference was measured halfway between the caudal point of the costal arch and the iliac crest (26). Hip circumference was measured at the level of the femoral great trochanters. With the participant in the supine position, sagittal abdominal diameter was measured as the distance from the back to the highest point of the abdomen during normal expiration, as a proxy for visceral adipose tissue mass (27).
Cardiovascular risk factors were evaluated by changes in fasting lipid concentrations (total cholesterol, high-density lipoprotein [HDL] cholesterol, low-density lipoprotein [LDL] cholesterol, and triglycerides), glucose, insulin, and C-reactive protein and changes in blood pressure. Blood pressure was measured 3 times with an appropriately sized cuff after the participant had rested for 5 minutes, and the last 2 measurements were averaged. Comorbid conditions were diagnosed on the basis of referrals and clinical assessment. We previously reported methods for measuring vitamin concentrations and the participants' vitamin levels and dietary supplement intakes within 1 year after surgery (23, 28). We now report vitamin concentrations until 2 years after surgery.
Generic health-related quality of life was assessed by using validated Norwegian and Swedish versions of the Short Form-36 Health Survey (SF-36) (4-week recall, version 2.0). This self-administered questionnaire consists of 8 domains, of which 4 reflect physical health and 4 reflect mental health. A score was calculated for each domain on a scale from 0 (worst possible health state) to 100 (best possible health state) (29).
Body composition was evaluated by changes in fat mass and fat-free mass, as measured by bioelectrical impedance analysis in Norwegian participants (n = 30; 2 participants exceeded the 200-kg weight capacity of the apparatus at baseline) by using the BC-418 (Tanita, Tokyo, Japan). In Swedish participants (n = 30), total body potassium was estimated as a proxy for fat-free mass by measuring the amount of radioactive K-40 in a whole-body counter (Nuclear Enterprises, Edinburgh, United Kingdom) at the Department of Radiation Physics at Sahlgrenska University Hospital. The ratio of total body potassium to fat-free mass was set at 64.8 mmol/kg for men and 59.6 mmol/kg for women (30). The percentage of weight lost as fat-free mass ([Δfat-free mass/Δbody weight] × 100%) was calculated after both body composition studies.
At each postsurgery visit, participants were asked open-ended questions about the occurrence of any adverse events, including readmissions, and diagnostic or therapeutic procedures. We report the overall numbers of adverse events and events occurring more than 30 days after surgery. Thirty-day outcomes have been reported before (22).
All participants were prescribed daily supplements of multivitamins, 100 mg of iron sulfate, 1000 mg of calcium carbonate, and 20 mcg of vitamin D3. In addition, participants in the gastric bypass group received a vitamin B12 substitute. Ursodeoxycholic acid was prescribed for 6 months, except in participants who had had cholecystectomy. Participants were seen by a dietitian and surgeon at inclusion and at 6 weeks, 6 months, 1 year, and 2 years after surgery. The dietitian visits included a dietary recall and evaluation of nutritional status and adherence to vitamin supplementation. Participants received top-up supplementation according to predefined cutoff levels of vitamin concentrations (23).
To estimate sample size, we performed a retrospective analysis of superobese patients who had undergone gastric bypass (n = 19) or duodenal switch (n = 13) at Sahlgrenska University Hospital. These patients had mean reductions in BMI of 18.0 kg/m2 (SD, 6.7) and 24.9 kg/m2 (SD, 5.0) after a mean follow-up of 34 and 31 months, respectively. Applying these data, we calculated that a total of 26 study participants would give 80% power to detect a significant (P < 0.050) difference between the groups. To allow for possible dropouts and add power for analysis of secondary outcomes, we decided to enroll 60 participants. Sample size was calculated in SamplePower 2.0 (SPSS, Chicago, Illinois).
Linear mixed models were fitted to all outcome variables to account for the repeated measures by patient. Competing models were compared by using the Akaike information criterion and the Bayesian information criterion. We modeled time by using a piecewise linear spline with a knot at 12 months. Two variables (glucose and insulin) had a knot at 6 months instead of 12 months, and 2 variables (systolic and diastolic blood pressures) were modeled with only 1 slope. Time and time-by-treatment interaction were fixed effects in all models. All models included a random intercept. For some variables, models also included fixed effects for treatment and site and random slopes by patient.
We compared binomial proportions between treatment groups with the Pearson chi-square test and the Suissa–Shuster exact unconditional test, depending on whether all expected counts were above 5 (31).
Linear mixed model analyses were performed with Stata, version 11.1 (StataCorp, College Station, Texas), and categorical analyses were performed with StatXact 9 (Cytel, Cambridge, Massachusetts). All P values are 2-tailed.
The trial was supported by the South-Eastern Norway Regional Health Authority. The funding source had no role in the design of the study; collection, analysis, and interpretation of the data; or drafting of the manuscript.
Figure 2 shows the participant flow through the study. One participant withdrew after randomization but before being informed about the allocated intervention. Two participants in the duodenal switch group were lost to follow-up. Baseline characteristics did not significantly differ between groups (Table 1).
BMI = body mass index.
After 2 years, the reduction in BMI was 17.3 kg/m2 (95% CI, 15.7 to 19.0 kg/m2) after gastric bypass and 24.8 kg/m2 (CI, 23.0 to 26.5 kg/m2) after duodenal switch (mean between-group difference, 7.44 kg/m2 [CI, 5.24 to 9.64 kg/m2]; P < 0.001) (Figure 3 and Table 2). The corresponding weight loss was 50.6 kg (CI, 45.4 to 55.8 kg) after gastric bypass and 73.5 kg (CI, 68.1 to 79.0 kg) after duodenal switch (mean between-group difference, 23.0 kg [CI, 16.2 to 29.7 kg]; P < 0.001). Mean percentage of body weight loss was 31.2% (CI, 29.2% to 33.2%) after gastric bypass and 44.8% (CI, 42.8% to 46.8%) after duodenal switch. Two years after surgery, 8 of 31 participants (26%) in the gastric bypass group and 0 of 27 in the duodenal switch group had a BMI of 40 kg/m2 or higher (P = 0.006).
Estimated postsurgery values are expected means from the linear mixed-effects models. Error bars indicate 95% CIs. P < 0.001 for between-group changes. BMI = body mass index.
Both procedures were associated with significant reductions in waist circumference, hip circumference, and sagittal diameter at 2 years (Table 2). Reductions in all of these measures were significantly greater in the duodenal switch group.
Fat mass and fat-free mass were significantly reduced in both groups, as measured by bioelectrical impedance analysis (Appendix Table 1). Participants in the duodenal switch group lost significantly more fat mass and fat-free mass than those in the gastric bypass group. However, the mean percentage of weight lost as fat-free mass did not differ between groups (23.6% [CI, 21.2% to 25.9%] for gastric bypass and 24.6% [CI, 21.9% to 27.2%] for duodenal switch; mean between-group difference, 1.03 percentage points [CI, −2.36 to 4.42 percentage points]; P = 0.54). Total body potassium measurements showed a greater reduction of fat-free mass after duodenal switch (Appendix Table 1). Total body potassium measurements also showed that the percentage of weight lost as fat-free mass did not significantly differ between groups (data not shown).
Appendix Table 1.
Two years after surgery, total cholesterol concentration decreased by 0.24 mmol/L (CI, −0.03 to 0.50 mmol/L) (9.27 mg/dL [CI, −1.16 to 19.3 mg/dL]) after gastric bypass and 1.07 mmol/L (CI, 0.79 to 1.35 mmol/L) (41.3 mg/dL [CI, 30.5 to 52.1 mg/dL]) after duodenal switch (mean between-group difference, 0.83 mmol/L [CI, 0.48 to 1.18 mmol/L]; 32.0 mg/dL [CI, 18.5 to 45.6 mg/dL]; P < 0.001). Low-density lipoprotein cholesterol concentration decreased by 0.26 mmol/L (CI, 0.09 to 0.43 mmol/L) (10.0 mg/dL [CI, 3.47 to 16.6 mg/dL]) after gastric bypass and 0.78 mmol/L (CI, 0.61 to 0.96 mmol/L) (30.1 mg/dL [CI. 23.6 to 37.1 mg/dL]) after duodenal switch (mean between-group difference, 0.53 mmol/L [CI, 0.29 to 0.76 mmol/L]; 20.5 mg/dL [CI, 11.2 to 29.3 mg/dL]; P < 0.001). Participants in both groups had significant reductions in mean triglyceride concentrations and significant increases in mean HDL cholesterol concentrations (Figure 4 and Table 2). Mean systolic and diastolic blood pressures, fasting levels of glucose and insulin, and mean levels of C-reactive protein were all significantly reduced in both study groups with no between-group differences in the changes at 2 years (Table 2 and Appendix Figure 1).
Estimated postsurgery values are expected means from the linear mixed-effects models. Error bars indicate 95% CIs. To convert cholesterol values to mg/dL, divide by 0.0259. To convert triglyceride values to mg/dL, divide by 0.0113. HDL = high-density lipoprotein; LDL = low-density lipoprotein.
* P < 0.001 for between-group changes.
† P = 0.78 for between-group changes.
Estimated postsurgery values are expected means from the linear mixed-effects models. Error bars indicate 95% CIs. Values for participants who used insulin were excluded (1 in the gastric bypass group at all time points and 1 in the duodenal switch group at baseline). To convert glucose values to mg/dL, divide by 0.0555. To convert insulin values to pmol/L, multiply by 6. To convert CRP values to nmol/L, multiply by 9.524. CRP = C-reactive protein.
* P = 0.077 for between-group changes.
† P = 0.123 for between-group changes.
‡ P = 0.94 for between-group changes.
Participants at Sahlgrenska University Hospital had significantly greater reductions in hip circumference and blood pressure than those at Oslo University Hospital Aker. Changes in any other outcome variable did not significantly differ by hospital (data not shown). At study entry, 14 participants used antihypertensive drugs, whereas 2 used such drugs at 2 years. Similarly, 4 participants used oral hypoglycemic drugs and 2 used insulin before surgery, whereas 1 received insulin after surgery. Two participants received lipid-lowering therapy with statins at inclusion and none at 2-year follow-up.
At 2-year follow-up, the duodenal switch group had lower mean concentrations of vitamin A and 25-hydroxyvitamin D than did the gastric bypass group (Appendix Table 2). Participants in the gastric bypass group received additional cobalamine supplements and had higher vitamin B12 concentrations during follow-up. Other vitamins (vitamin B6, folate, vitamin C, and vitamin E adjusted for serum lipids) either were stable or increased at 2 years compared with baseline.
Appendix Table 2.
One participant in each group did not complete SF-36 forms after surgery and were excluded from analysis. After 2 years, 7 of 8 subscores significantly improved in gastric bypass participants (Appendix Figure 2 and Appendix Table 3). In the duodenal switch group, 5 of 8 subscores significantly improved after 2 years. Bodily pain was the only domain in which improvements differed between groups, with a mean 20.2-point (CI, 6.71 to 33.7 points) greater improvement in the gastric bypass group (P = 0.003).
Estimated postsurgery values are expected means from the linear mixed-effects models; a score of 0 represents worst possible health and 100 represents best possible health. Estimated postsurgery values are expected means from the linear mixed-effects models. BP = bodily pain; GH = general health perceptions; MH = general mental health; PF = physical functioning; RE = role limitations due to emotional problems; RP = role limitations due to physical health problems; SF = social functioning; SF-36 = Short Form-36 Health Survey; VT = vitality.
Appendix Table 3.
Overall, significantly more participants in the duodenal switch group had adverse events occurring from surgery and up until 2 years: 10 of 31 (32%) after gastric bypass compared with 18 of 29 (62%) after duodenal switch (P = 0.021). Long-term (>30-day) adverse events occurred in 9 participants (29%) after gastric bypass (1 to 2 events per participant) and in 12 participants (41%) after duodenal switch (1 to 6 events per participant) (P = 0.32) (Table 3).
In the gastric bypass group, 6 participants had consultations for vomiting, abdominal pain, or diarrhea, of whom 1 had laparoscopy and 2 had upper endoscopies (all negative). Three participants developed symptomatic cholelithiasis and 1 had cholecystectomy. One participant had internal herniation requiring laparoscopic small-bowel reduction. Another was treated with antibiotics for bacterial enterocolitis. One participant was hospitalized for psychiatric disease.
In the duodenal switch group, 1 participant without excessive alcohol consumption developed acute hepatic failure (liver biopsy revealed steatohepatitis with fibrosis and cirrhosis), pneumonia, and protein-calorie malnutrition after laparoscopic surgery for bile duct stones (32). Liver function improved from Child–Pugh score class C to A after nutritional support and management of liver failure. Two other participants required parenteral nutrition and a protein-enriched diet for malnutrition. Two participants had low serum concentrations of vitamin A and night blindness revealed by dark adaptometry, which resolved after adequate supplementation of vitamin A. One of these participants was later hospitalized for traumatic subarachnoid hemorrhage and Haemophilus influenzae meningitis, which was complicated by gangrene of fingers and toes that required partial amputations. One participant with iron deficiency received iron infusions. Laparotomy was performed in 3 participants because of peritonitis, inflammation of the transverse mesocolon, and small-bowel obstruction after an intra-abdominal abscess. The latter participant subsequently had another laparotomy for small-bowel obstruction. Laparoscopic cholecystectomy was performed in 1 participant. Four participants had upper endoscopies performed because of vomiting or abdominal pain (1 had a prepyloric ulcer with Helicobacter pylori infection). One participant experienced recurrent severe diarrhea that was managed with metronidazole, 1 participant developed urolithiasis, and 1 participant had a traumatic humerus fracture after a car accident.
In this randomized trial comparing duodenal switch with another bariatric surgical procedure, which we believe to be the first to do so, we show that duodenal switch induced a significantly and substantially greater weight loss than gastric bypass in superobese patients. Two years after surgery, 26% of participants who had gastric bypass still had a BMI more than 40 kg/m2 compared with no participants after duodenal switch. These weight-loss data extend our previously published findings after shorter-term follow-up (22, 23) and those of uncontrolled studies conducted by others (20, 33), which have been limited by the inherent bias in treatment allocation.
Bariatric surgery is associated with long-term weight loss and decreased overall mortality (14). However, whether a dose–response relationship exists between the magnitude of weight loss after surgery and reduction in total or cardiovascular mortality is unknown. An important finding in our study was that despite a large difference in weight loss, both procedures induced substantial and, for most variables studied, similar improvements in markers of risk for cardiovascular disease. For blood lipids, we found significantly greater reductions in total and LDL cholesterol concentrations after duodenal switch, extending the shorter-term changes reported previously (23). Other investigators have found a reduction in cardiovascular mortality in patients with hyperlipidemia and previous myocardial infarction who were randomly assigned to undergo partial ileal bypass (which is no longer done) compared with conservative treatment (34).
Previous studies have shown beneficial effects of both gastric bypass and duodenal switch on cardiovascular risk factors and reduced mortality from heart disease after gastric bypass (21, 35–37). Our finding of a 0.53-mmol/L (20.5-mg/dL) greater reduction in LDL cholesterol concentration after duodenal switch could mean that patients having this procedure have a greater reduction in long-term cardiovascular risk. In a meta-analysis of more than 90 000 participants in cholesterol-lowering trials, a 1-mmol/L (38.6-mg/dL) reduction in LDL cholesterol concentration translated into a 12% reduction in total mortality and a 19% reduction in coronary mortality over 5 years (38). The reduction in LDL cholesterol may be explained by both weight loss and procedure-specific effects. A decreased absorption of cholesterol has been demonstrated after gastric bypass (39). After duodenal switch, mixing of food with pancreatic enzymes and bile acids is more delayed, which may lead to more pronounced malabsorption of fat than with gastric bypass (40).
The available evidence suggests that duodenal switch is associated with higher rates of resolution of hypertension, diabetes, and dyslipidemia, including in superobese patients (11, 41). Although we observed some evidence of lower blood pressures and lower fasting serum glucose levels after duodenal switch than after gastric bypass, these differences were not statistically significant.
Patients who had duodenal switch had significantly lower levels of vitamin A and 25-hydroxyvitamin D than did patients who had gastric bypass. Additional supplementation of these fat-soluble vitamins in patients having duodenal switch has been proposed (21). After gastric bypass, concentrations of vitamins were stable or increased. This indicates that the dietary supplementation regimen described in our trial may be adequate after gastric bypass surgery (23), as also shown in a study of vitamin status after gastric bypass in a less obese cohort taking similar supplements (42).
The substantially greater weight loss after duodenal switch was not associated with significantly greater improvements in health-related quality of life. The higher overall morbidity after duodenal switch may have had a negative effect on these participants' quality of life, although our study did not have statistical power to formally address this. The SF-36 subscores 2 years after duodenal switch are similar to those reported by other studies (43). Weight regain in the long term after bariatric surgery is associated with deteriorations in health-related quality of life (13). This suggests that if longer-term maintenance of weight loss differs between gastric bypass and duodenal switch, this may also influence quality of life.
Meta-analyses of current bariatric procedures suggest that duodenal switch is associated with the largest weight loss and greatest improvements of obesity-related comorbid conditions (11). This operation is, however, difficult to perform by laparoscopy and has been associated with a higher mortality rate than laparoscopic gastric bypass (44). Although some series have reported low morbidity rates after bariatric surgery in superobese patients (45, 46), others have identified very high BMI as a risk factor for surgical complications (17). Because duodenal switch is often reserved for patients with a BMI greater than 50 kg/m2, balancing the health benefits and safety of this operation to those of other procedures is important. Data from our study suggest that superobese patients opting for gastric bypass should be informed that they are likely to experience clinically important health benefits even though they may expect to still be severely obese after surgery. In contrast, duodenal switch provides greater weight loss and greater improvements in total and LDL cholesterol concentrations but at a potentially greater risk. Patients should also be informed that health-related quality of life seems to be relatively similar after both surgeries.
We suggest that duodenal switch surgery be restricted to well-informed superobese patients who are likely to adhere to clinical follow-up. Furthermore, we recommend more closely monitoring patients after duodenal switch than after gastric bypass because of the greater risk for micronutrient deficiencies after duodenal switch. In women of fertile age, malabsorptive procedures, such as duodenal switch, may be associated with an increased risk for pregnancy-related nutritional deficiencies and adverse neonatal outcomes (47, 48). In future studies, the potential benefits of performing the duodenal switch as a 2-stage procedure need to be clarified (49).
Strengths of our study include the prospective, randomized design and a high rate of completion. Our study also has several limitations. The study sample was predominantly female, which is common in studies of bariatric surgery. The study did not include participants older than 50 years, which may help explain why relatively few participants had type 2 diabetes. The results are limited by the small study size and the 2-year follow-up. Surgeons and the multidisciplinary treatment teams at both study centers had accumulated more clinical experience with gastric bypass than with duodenal switch, which may influence operating times and morbidity rates (22). Although all known adverse events were recorded, we did not use a formal classification system for surgical complications. The findings may be generalizable only according to the surgical techniques used, such as the lengths of the intestinal limbs. Evaluation of body composition by bioelectrical impedance analysis may be useful for group comparisons but has not been thoroughly validated in obese persons (50). Finally, the study setting should be considered when interpreting the data; for example, most of the participants were of white ethnic origin (51).
In summary, both gastric bypass and duodenal switch induced significant and durable weight loss in superobese patients. Weight loss and reductions of total and LDL cholesterol concentrations were greater after duodenal switch, and HDL cholesterol concentration increased after both procedures. Duodenal switch was also associated with lower concentrations of vitamin A and 25-hydroxyvitamin D and more adverse events during the 2-year follow-up. Both interventions were associated with significant improvements in other cardiovascular risk factors and health-related quality of life.
The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.
John G Kral
SUNY Downstate Medical Center
October 9, 2011
Ledtter to the Editor: Bypassing Bariatric Surgery and Editorial Evidence
The journal published a rarity: an ethical, scientific, randomized, trial of bariatric surgery (1) exhibiting equipoise: the two operations are similar, in contrast with randomizations between operations with different mechanisms and effects (2) or between any operation and "medical" care (3), although long-term data demonstrate superiority of the one operation. One might question equality of the two hospitals, equivalency of the teams' learning curves, follow-up schedules lacking postoperative "dietitian and surgeon" visits during the important 12-24 months after bypass operations in superobese patients and drawing conclusions after only 2 years when slopes of the mean BMI diverge in groups with different follow-up rates, and particularly too early to evaluate the most valuable benefits of these operations. However, the accompanying Editorial (4) raises more serious questions:
Is it accurate or "evidence-based" to portray superobese patients as "metabolically normal", "healthy" or "near -normal"?
Should all malabsorptive operations be branded as "inherently dangerous"? Gastrointestinal surgery is "dangerous" yet meets risk-benefit and cost-benefit criteria or else would not be allowed?
Can outcomes of treatments of superobese patients be evaluated after only 2 years? No 2-year evidence supports stating: "greater weight loss does not seem to change the anticipated long-term outcomes for bariatric surgery". Predictions before weight and compensatory adaptive mechanisms have stabilized are unfounded. There is no evidence that "Many patients are well-adapted to their obesity"; neither superobese patients in general nor 'bariatric surgery' patients. The editorial criticizes the "low methodological quality" of studies of bariatric surgery and requests "good evidence", yet describes a rare complication as "particularly worrisome for these young women" based on one case-report (5). Not to trivialize this, but the complication is rare and easily prevented by standard of care following authoritative guidelines (editorial refs 3 and 4).
"Complications are serious" and "may take years to manifest" but it is not accurate that "the true complication rate over the lifetime of these young patients can be expected to be formidable". Both operations have been performed for 35 years with large series exceeding 15 years, including two with offspring of mothers with the potentially more "dangerous" of the operations followed for 2-24 years. The complications are known, germane to gastrointestinal surgery, but most important: They are easily preventable by competent follow-up care and are substantially easier to treat effectively than superobesity or less extreme forms of obesity. In the final analysis patients must decide whether effects of these operations, beneficial or adverse, outweigh their own suffering with their disease. Responsible physicians must provide the best available understandable evidence to lay patients.
1. Sovik TT et al. Weight loss, cardiovascular risk factors and quality of life after gastric bypass and duodenal switch. Ann Intern Med 2011;155: 281-91
2. Kral JG. Psychosurgery for obesity. Obes Facts 2009;2:339-41
3. Kral JG et al. Flaws in methods of evidence-based medicine may adversely affect public health directives. Surgery 2005;137:279-84
4. Livingston EH. Primum Non Nocere. Ann Intern Med 2011;155:329-330.
5. Huerta S, Rogers LM, Li Z, Heber D, Liu C, Livingston EH. Vitamin A deficiency an a newborn resulting from maternal hypovitaminosis A after biliopancreatic diversion for the treatment of morbid obesity. Am J Clin Nutr. 2002:76:426-9.
October 12, 2011
Unidentified flaws in both original paper and editorial
Sovik and colleagues (1) are to be congratulated on performing one of the few RCTs comparing bariatric surgical procedures, but their paper has serious flaws which were not identified in the accompanying editorial (2) nor the subsequent reports in other journals(3) and specialist websites.
What Sovik's data actually illustrates is that if duodenal switch (DS) is performed on randomly (instead of carefully) selected patients by relatively inexperienced surgeons and if patients are then given sub- optimal vitamin replacement (more suited to a short-limb Roux-en-Y gastric bypass) and infrequent follow-up, post-operative problems are not uncommon. In fact with this study design it was remarkable that so few adverse events were seen.
Our experience of 125 DS patients followed for up to 5 years(4) confirmed Sovik's report that the DS is indeed associated with reversible nutritional deficiencies (particularly vitamin D). However, we and others recognize that the DS is not a universally applicable operation because it requires an exceptional level of post-operative patient compliance. Few patients are capable of this and without careful selection problems can occur, particularly if, as in this study, their nutrition is only being monitored once every 6-12 months during the crucial mid-late post- operative period. The very notion of randomizing potentially non-compliant patients to a DS is flawed.
Sovik previously reported (5) that most of the difference in complication rates between the two operations occurred in the first 30 days after surgery and was related to technical complications (including an unusually high leak/abscess rate of 10.3%) or complications attributable to the DS's prolonged operative time. Their current paper confirms this to be the case as only an additional 4 gastric bypass and 7 DS patients developed complications between 30 days and 1 year post-op. Although not statistically significant, the most common adverse event in the DS group was vomiting (3/7), which is understandable given the extremely tight (30-32F) sleeve gastrectomy the authors performed. In year two, a further 5 gastric bypass patients developed adverse events, compared to just 3 DS patients (two due to trauma rather than surgery).
The DS is a technically challenging procedure and mandates a strict post-operative regime with intensive support from the bariatric team. It is clear that to keep the frequency of adverse events to a minimum in the early and late post-operative periods the DS should only be offered to carefully selected patients treated in specialist centres. Sovik's paper simply confirms this and adds little more.
DD Kerrigan CJ Magee AI Mitchell
1. Sovik TT, Aasheim ET, Taha O, Engstrom M, Fagerland MW, Bjorkman S, et al. Weight loss, cardiovascular risk factors, and quality of life after gastric bypass and duodenal switch: a randomized trial. Ann Intern Med. 2011;155(5):281-91.
2. Livingston EH. Primum non nocere. Ann Intern Med. 2011;155(5):329-30.
3. BMJ. Short Cuts: Duodenal Switch is a poor choice in super obese adults. British Medical Journal. 2011.
4. Magee CJ, Barry J, Brocklehurst J, Javed S, Macadam R, Kerrigan DD. Outcome of laparoscopic duodenal switch for morbid obesity. Br J Surg. 2011;98(1):79-84.
5. Sovik TT, Taha O, Aasheim ET, Engstrom M, Kristinsson J, Bjorkman S, et al. Randomized clinical trial of laparoscopic gastric bypass versus laparoscopic duodenal switch for superobesity. Br J Surg. 2010;97(2):160- 6.
Fernando B Bonanni MD FACS FASMBS
Abington Memorial Hospital
October 16, 2011
Duodenal Switch: Experience counts
Letter to the Editor,
Those of us who have dedicated ourselves to offering the entire arsenal in weight loss surgery have come to understand the important role the Duodenal Switch has in the Super Obese patient and in the future of Bariatric Surgery.
The data provided reveals greater weight loss in DS patients resulting in reduction of 25 BMI units compared to 17 in the GBP. (p value of < .001) Improvement in cholesterol in DS patients was greater. (p value of < .0001). 26% of GBP group were still morbidly obese as compared to none in the DS group. Early and late post-operative morbidity was not statistically significant. The authors comment that there is a tendency toward protein malnutrition in the DS patient.
The authors offer an unusually aggressive construction of the DS. The Sleeve component is more aggressive than a standard VSG. The average bougie size in a VSG is 34 to 38 F Bougie. The Sleeve component of a standard DS procedure is usually created larger than a standard sleeve anticipating significant contribution to weight loss by the diversionary component of the procedure. A Common Channel of 100 cm plus an aggressive sleeve is a set up for excessive weight loss and serious sequale of protein malnutrition. Intense nutritional follow up with specific instruction on adequate protein supplementation, hydration, and vitamin supplement is paramount. Most programs require that Vitamins A D E K Fe Ca++, Zinc, and protein are supplemented and followed closely.
Persistent morbid obesity in 26% of GBP patients underscores the shortcomings of the GBP procedure in the super obese. A predictable recidivism rate in GBP of 8 - 23 % is well known to start at post op two years. It is probable that the majority of the GBP patients in this study will regain weight and comorbid disease will return. The construction of the DS in this study will lead to predictable and serious nutritional consequences in the DS patients in this study.
The DS is superior in the Super Obese. We cannot continue to ignore inadequate weight loss and high rate of recidivism in the super obese patient after GBP. Experience and a comprehensive nutrition program achieve safe results. Standardization of procedures and nutritional protocol are necessary to study procedures accurately.
Fernando B. Bonanni Jr. MD FACS, FASMBS Director, Institute of Metabolic and Bariatric Surgery Abington Memorial Hospital
Department of Experimental Medicine, Sapienza University of Rome
October 21, 2011
Weight loss, cardiovascular risk factors, and quality of life after gastric bypass and duodenal switch
TO THE EDITOR We read with great interest the recent article by Sovik and colleagues (1) regarding the change of weight, cardiovascular risk factors and quality of life 2 years after gastric bypass and duodenal switch. However, we have some comments on the paper. The selection criteria for different bariatric techniques are not well defined and are usually based on psychiatric statement, body weight and relative surgical risk. The authors didn't clarify if the patients underwent psychiatric assessment, which were the criteria used for the eligibility of the bariatric surgery options. As clearly discussed by Sovik and colleagues, duodenal switch surgery should be restricted to well -informed super obese patients who are likely to adhere to clinical follow -up. As a consequence, the patients should be carefully selected instead of being randomized. In this case the randomization procedure raises important ethical concerns that may be overcome with case-control study design (2). Since fat distribution more than the total amount of fat is critical in defining cardiovascular risk and insulin resistance, an evaluation of fat distribution in both visceral and peripheral sites through either a surrogate marker like waist to hip ratio or direct measurements of the different body compartments by DXA analysis of body composition is advisable in order to assess the quality of weight loss (3). Furthermore, the calculation of the HOMA-IR index would have more likely better characterized the changes of insulin resistance rather than fasting glucose and insulin serum concentrations values. Faecal excretion of cholesterol is one of the main mechanisms through which malabsorptive techniques can reduce circulating total, LDL and HDL cholesterol (4). The authors show a greater LDL reduction after duodenal switch, but the reported increase of HDL cholesterol appears not significant; on the contrary, the HDL cholesterol values are clearly increased in the gastric bypass group. Thus, calculated indices such as total cholesterol/HDL cholesterol ratio would have been appropriate. It is difficult to explain why both gastric bypass and duodenal switch patients showed a superimposable increase of PTH whilst only duodenal switch patients were vitamin D deficient. Whether increased PTH levels are relevant to future skeletal health is still an open question. Recent epidemiologic and clinical studies showed that obesity may not be protective for osteoporosis and weight reduction per se causes bone loss (5). Therefore, an assessment of bone mineral density modifications after surgery-induced weight loss is advisable. Finally, the authors did not account for confounding like sex and age including menopause and exposure to oral contraceptives in the female subgroup.
1. Sovik TT, Aasheim ET, Taha O, Engstrom M, Fagerland MW, Bjorkman S, Kristinsson J, Birkeland KI, Mala T, Olbers T. Weight loss, cardiovascular risk factors, and quality of life after gastric bypass and duodenal switch : a randomized trial. Ann Intern Med. 2011 Sep 6;155(5):281-91
2. Lubrano C, Mariani S, Badiali M, Cuzzolaro M, Barbaro G, Migliaccio S, Genovesi G, Rossi F, Celanetti M, Fiore D, Pandolfo MM, Specchia P, Spera G. Metabolic or bariatric surgery? Long-term effects of malabsorptive vs restrictive bariatric techniques on body composition and cardiometabolic risk factors. Int J Obes (Lond). 2010 Sep;34(9):1404-14
3. Bays HE, Laferr?re B, Dixon J, Aronne L, Gonzalez-Campoy JM, Apovian C, Wolfe BM; Adiposopathy and Bariatric Surgery Working Group.Adiposopathy and bariatric surgery: is 'sick fat' a surgical disease? Int J Clin Pract. 2009 Sep; 63(9):1285-300.
4. Corradini SG, Eramo A, Lubrano C, Spera G, Cornoldi A, Grossi A et al. Comparison of changes in lipid profile after bilio-intestinal bypass and gastric banding in patients with morbid obesity. Obes Surg 2005; 15: 367- 377.
5. Greco EA, Fornari R, Rossi F, Santiemma V, Prossomariti G, Annoscia C, Aversa A, Brama M, Marini M, Donini LM, Spera G, Lenzi A, Lubrano C, Migliaccio S. Is obesity protective for osteoporosis? Evaluation of bone mineral density in individuals with high body mass index. Int J Clin Pract. 2010 May; 64(6):817-20.
Tor I., Karlsen, PhD-student, Joran Hjelmesaeth
Morbid Obesity Center, Vestfold Hopital Trust, Norway
November 11, 2011
Quality of life assessment
We read with interest Sovik et al's report on the well performed randomized controlled trial comparing the effects of gastric bypass and duodenal switch . From the observed changes in the 8 sub-scales of health reated quality of life (HRQL)-questionnaire SF-36, the authors conclude that the two procedures had a similar effect on quality of life. We wish to suggest that this conclusion is questionable.
Cardiology, Hospital Medicine, Obesity, Coronary Risk Factors, Prevention/Screening.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only