Obesity Surgery

, Volume 20, Issue 3, pp 283–289

Long-Term Outcomes of Laparoscopic Roux-en-Y Gastric Bypass in US Veterans


  • Debra L. Hauser
    • Veterans Administration Pittsburgh Healthcare System
  • Rebecca L. Titchner
    • Veterans Administration Pittsburgh Healthcare System
  • Mark A. Wilson
    • Veterans Administration Pittsburgh Healthcare System
    • Veterans Administration Pittsburgh Healthcare System
    • Division of Minimally Invasive SurgeryThe University of Pittsburgh Medical Center
Clinical Research

DOI: 10.1007/s11695-009-0042-1

Cite this article as:
Hauser, D.L., Titchner, R.L., Wilson, M.A. et al. OBES SURG (2010) 20: 283. doi:10.1007/s11695-009-0042-1



The objective of this study is to evaluate the long-term outcomes following laparoscopic Roux-en-Y gastric bypass (LREYGB) in veteran patients. The VA bariatric population differs from its counterpart in the private sector by the predominance of a male population, a higher percentage of patients from a lower socioeconomic background, a higher mean age, and a higher rate of obesity-related comorbidities.


A retrospective review with prospectively collected data was used to analyze postoperative changes of comorbidities and percent of excess weight loss (% EWL) in consecutive patients who underwent LREYGB between August 2003 and September 2006.


Among 70 patients, 73% were men with a mean age of 52 years (29–66 years). Average preoperative weight and body mass index were 310 lbs (224–397 lbs) and 46 kg/m2 (36–60 kg/m2), respectively. The incidence of major and minor complications was 1.4% and 15.7%, respectively. There were no mortalities. Follow-up (f/u) was possible in all patients. At a mean f/u rate of 39 months, % EWL was 56%. At 1, 3, and 5 years, % EWL was 61%, 53%, and 59%, respectively. Thirty-five patients (50%) had type 2 diabetes mellitus (T2DM). Glycosylated hemoglobin concentrations returned to normal levels in 91% of patients and improved in an additional 6% of T2DM cases. Only 7% of patients are still maintained on antidiabetic medications. In patients with more than 1 year f/u, most other comorbidities were improved or resolved.


Long-term f/u of LREYGB in veteran patients demonstrated significant and durable weight loss (56% EWL) with marked improvements in comorbidities especially T2DM.


Morbid obesityVeteransBariatric surgeryType 2 diabetes mellitusGastric bypassHigh riskLong-term


Morbid obesity in the US has more than doubled in the past 20 years [1]. The most recent prevalence surveys report 32.7% of US adult citizens are overweight, defined as body mass index (BMI) of 25 to 30, 34.3% are obese (BMI ≥30) and 5.9% are extremely obese (BMI ≥40) [2]. The Veterans Affairs (VA) patient population is reported to have a higher incidence of obesity [3]. Laparoscopic Roux-en-Y gastric bypass (LREYGB) has been shown to be effective at achieving significant long-term weight loss in morbidly obese patients with acceptable morbidity and mortality rates [47]. In addition, LREYGB has been shown to improve and/or resolve the multiple comorbidities typically associated with obesity [810]. Given the high incidence of morbid obesity in the VA population, LREYGB has become an important treatment modality for VA patients with these life-threatening conditions. However, the VA population presents unique challenges in patient management. Advanced age, predominately male gender, and lower socioeconomic background are typical characteristics of this population [11] and are known factors for increasing morbidity and mortality rates following bariatric surgery [1215]. The purpose of this study is to report on the long-term outcomes following LREYGB in veteran patients. We will discuss surgical techniques, our protocol for preoperative evaluation, and postoperative management in this unique patient population.


This is a retrospective review with prospectively collected data obtained from consecutive VA patients at a tertiary care center who underwent LREYGB surgery between August 2003 and September 2006. The computerized record system was used to extract all data. This study was approved by the Research and Development Review Board under number 02758 at the VA Pittsburgh Healthcare System.

Data collection included demographics, preoperative weight and BMI, gender, age, biochemical data, preoperative comorbidities, surgical technique, surgical time, length of stay, postoperative complications, weight loss, and changes in comorbidities. Weight loss was reported as the mean percentage of excess weight loss (% EWL). Ideal body weight was identified using the midpoint of a medium frame from the 1999 Metropolitan Life Height and Weight Tables. Patients were considered for bariatric surgery if they met the criteria established by the NIH Consensus Development Panel Report of 1991 including a BMI of 35 to 39 with one or more severe comorbidities or BMI ≥40 with or without any comorbidities. Patients were excluded from surgery if they had severe inoperable coronary artery disease (CAD), severe chronic obstructive pulmonary disease (COPD) with a forced expiratory volume (FEV1) <1.0, history of noncompliance with treatment, psychosis, tobacco use, and alcohol and/or illicit drug abuse within the year prior to the referral.

All patients received extensive preoperative education by a multidisciplinary team including emphasis on diet regimen and exercise. Patients received a comprehensive preoperative evaluation including the following:

Psychology consults were obtained in order to evaluate patients for any issues which might hinder compliance with postoperative care.

Biochemical testing included complete blood count (CBC), glycosylated hemoglobin (HgA1C), glucose, electrolytes, blood urea nitrogen, creatinine, liver profile tests, albumin, calcium, thyroid-stimulating hormone, iron studies, vitamin B12/folate, homocysteine, parathyroid hormone, and 25-hydroxyvitamin D levels. Preoperative evaluation of vitamin D levels was only initiated in 2006 following the high incidence of postoperative vitamin D deficiencies.

Cardiovascular evaluations consisted of an adenosine stress test on all patients aged 40 and above or under the age of 40 with a strong family history of CAD and/or with one or more risk factors. An echocardiogram was obtained on patients with a history of congestive heart failure, phentermine–fenfluramine use, and history of rheumatic fever or heart murmur. Patients determined to have a high risk for cardiovascular disease but who could not have an adenosine stress test due to their weight were placed on an 800 calorie liquid diet with the goal to lose enough weight to qualify for the study. Patients with poor mobility, history of deep vein thrombosis (DVT), pulmonary embolus, and/or a hypercoagulable state had an inferior vena cava (IVC) recovery filter placed 2 days prior to surgery.

Pulmonary assessment included a chest X-ray, a sleep study, and pulmonary function tests. Appropriate positive pressure treatment was initiated if a patient was diagnosed with obstructive sleep apnea (OSA). Asthma or COPD patients were initiated on optimal bronchodilator agents. Patients were deemed inoperable if they had an FEV1 < 1.0 that did not improve with medical treatment or with preoperative weight loss.

Gastrointestinal work-up consisted of an upper endoscopy. Appropriate treatment was initiated if significant gastritis, esophagitis, ulcers, or Helicobacter pylori were found prior to surgery. Patients with a history of symptomatic gallstones underwent a cholecystectomy concomitantly with the gastric bypass. A 4 hour liquid and solid phase gastric emptying study was done on all patients with a history of diabetes-related complications such as retinopathy, nephropathy, or neuropathy. A screening colonoscopy was performed on patients aged >50 and on those younger than 50 years of age with a strong family medical history of colon cancer. The guidelines established by the American Society for Gastrointestinal Endoscopy were followed for recommendations on repeat colonoscopies in patients with a history of colon polyps.

Preoperative weight loss was required from all patients. Patients with a BMI of 35 to 49 were placed on a low-calorie diet (LCD) consisting of a 4-week 1,000-cal diet and a 2-week liquid diet. Progression of weight loss and compliance with diet were monitored regularly by a registered dietitian (RD). Patients with a BMI > 50 or a BMI < 50 with a history of diabetes or severe comorbidities were placed on a very-low-calorie diet (VLCD) consisting of an 800 calorie liquid diet. The goal for weight loss was 10% of body weight. Most patients achieved an average weight loss of 4 to 5 lbs/week. All patients were started on a daily multivitamin, and ursodiol was initiated to prevent gallstone formation. Patients were monitored with weekly visits which included vital signs, blood work, behavioral modification, and support group sessions with an RD. Patients with hypertension (HTN) and/or type 2 diabetes mellitus (T2DM) were required to participate in self-monitoring of blood pressure and/or blood glucose. Failure to comply with the requirements involved with the weight loss program resulted in either delay or denial of surgery.

Perioperative Management

All LREYGB were performed by a single team of surgeons. This technique was previously described [16]. Briefly, it involves the creation of a 15-mL gastric pouch and an intra-abdominal end-to-side gastrojejunal anastomosis constructed using a linear stapler technique. A segment of the jejunum is bypassed with the creation of an antecolic, antegastric Roux limb, which measures 150 cm. All patients were maintained on a cardiac monitor postoperatively. An upper gastrointestinal study of gastrografin followed by thin barium was performed on the first postoperative day to rule out anastomotic leak and/or obstruction. Patients were subsequently started on a low-sugar clear-liquid diet. DVT prophylaxis consisted of subcutaneous low-molecular-weight heparin and sequential compression devices. Physical therapy and occupational therapy assisted patients with ambulation and evaluated patients for any aides needed for activities of daily living.

Postoperative management involved follow-up (f/u) visits scheduled 1 week following surgery, then 1, 3, 6, 12, 18, and 24 months, and then yearly. Patients were seen by the bariatric surgeon, a nurse practitioner, and RD on every f/u visit. All patients were started on a vitamin supplement regimen of a daily multivitamin, cyanocobalamin, ferrous sulfate, ascorbic acid, and calcium citrate with vitamin D. Vitamin levels, electrolytes, liver function tests, and CBCs were evaluated every 6 months for the first 2 years following surgery and then yearly thereafter. Diet and vitamin compliance, weight loss, laboratory values, and changes of comorbidities were monitored at each f/u visit.


Seventy consecutive patients were included in the study. Patient demographics are summarized in Table 1. Table 2 reflects early (within 30 days) and late (beyond 30 days) complications. Most early complications were minor in nature. The one major complication was due to a bowel obstruction distal to the jejunojejunostomy due to thick barium impaction from inadvertent radiology administration. The majority of late complications involved nutritional deficiencies. Fifty-two percent of patients had vitamin D deficiencies; however, the baseline levels are unknown. All patients were treated appropriately with resolution of vitamin deficiencies. Three patients developed marginal ulcers, each several months following surgery. Frequent doses of nonsteroidal anti-inflammatory agents were thought to have caused the ulcers. Four patients developed symptomatic cholelithiasis and were treated successfully with laparoscopic cholecystectomy. One patient developed a small bowel obstruction due to adhesions that was treated surgically.
Table 1

Demographics and Peri-Operative Characteristics (n = 70)

Mean age in years (range)

52 (29–66)


51/19, 73% male

African American/Caucasian

8/62, 89% Caucasian

Mean no. of co-morbidities/patient (range)

6.8 (1–14)

Percent with history of prior abdominal surgery


Mean BMI at initial visit (range)

50 (38–68)

Mean BMI post-LCD/VLCD (range)

46 (36–60)

% Super-morbid obese patients (BMI > 50)


No. (%) patients maintained on Pre-Operative Weight Loss Programs

VLCD 25 (36%)

LCD 31 (44%)

Neither 14 (20%)

Mean operative time in minutes (range)

162 (108–313)

No. (%) patients with additional procedures

27 (39%)

IVC filter placed, n (%)

3 (4.3%)

Mean Length of stay in days (range)

4 (2–16)

Table 2

Early and Late Complications

Early Complications


Late Complications




Symptomatic marginal ulcer

1 (1.4%)



Perforated marginal ulcer

2 (2.8%)

Minor wound infection


Symptomatic cholelithiasis

4 (5.6%)



Small bowel obstruction due to adhesions

1 (1.4%)



Abdominal pain, negative exploratory laparoscopy

1 (1.4%)





Bleeding intra-abdominal, asymptomatic


Vitamin B12 deficiency

3 (4.2%)

Questionable pulmonary embolism


Vitamin D deficiency

37 (51.8%)


11 (15.7%)

Folate deficiency

1 (1.4%)


Iron deficiency

15 (21.0%)



Bowel obstruction






GI bleeding













1 (1.4%)


Weight Loss

Fig. 1 depicts weight loss at different time intervals. At a mean f/u of 39 months, the average % EWL achieved was 56%. The % EWL increases to 63% when preoperative weight loss was considered. The nadir mean % EWL of 64% was achieved at 18 months postoperatively. While weight regain was observed on some patients, this weight gain stabilized as demonstrated by the fact that 48 patients with 4-year f/u maintained a mean % EWL of 55%.
Fig. 1

Mean percent excess weight loss with standard deviation according to duration of f/u


Following surgically induced weight loss, significant improvement or complete remission was noted with all comorbidities. These comorbidities included T2DM, HTN, OSA, and gastroesophageal reflux disease (GERD). Table 3 summarizes the changes in comorbidities with categorization as aggravated, unchanged, improved, or in remission. The most significant improvement was observed with T2DM. At a mean f/u of 36 months, the mean HgA1C level was 6.0%. T2DM was considered improved if a patient had improved HgA1C levels and/or was maintained on less T2DM medications. T2DM was considered in remission if a patient had a normal HgA1C level and was not maintained on any T2DM medications.
Table 3

Change in obesity-related comorbidities


Improvement in select comorbidities


Percent aggravated

Percent unchanged

Percent improved

Percent remission

Percent unknown





























Fig. 2 illustrates changes in T2DM over the f/u period. As the % EWL increases, the percentage of patients in remission also increases. At the same time, there is a significant decrease in the usage of T2DM medication. However, some patients required reinitiation of DM medications around 36 months following surgery, coinciding with some weight regain.
Fig. 2

Change in diabetes according to duration of f/u

HTN was considered improved if a patient had improved blood pressure control and/or was taking less blood pressure medications. HTN was considered to be in remission if the patient had a normal blood pressure and was not taking any medications. OSA was assessed both objectively and subjectively and was defined as in remission if a repeat sleep study was negative. However, the majority of patients did not have a repeat sleep study following surgery. These patients were assessed subjectively with results reported as unchanged or improved. GERD was evaluated subjectively, and over 50% of patients reported no GERD symptoms.


The incidence of obesity has increased substantially in the US and among veterans [1, 3]. This rise in obesity is associated with increases in the development of obesity-related comorbidities such as T2DM, dyslipidemia, HTN, OSA, and CAD [17], as well as an increased mortality risk [18]. LREYGB has been shown to be an effective treatment for morbid obesity that results in significant and sustained weight loss with acceptable morbidity and mortality rates [4, 7, 19]. A meta-analysis of 136 studies on bariatric surgery by Buchwald et. al. revealed a mean % EWL of 61.6% with a 30-day mortality rate of 0.5% for gastric bypass patients [9]. In addition, weight loss following LREYGB has resulted in improvement and/or resolution of comorbidities commonly associated with morbid obesity [4, 5, 7]. Given the high incidence of morbid obesity in the VA patient population, LREYGB has become an important treatment modality for these patients. However, the VA patient population presents unique challenges in patient management.

Large studies of open and LREYGB in the private sector have shown that the majority of patients are female with an average age in the early 40s [12, 20, 21]. In contrast, the VA patient population has a very different profile. Advanced age, predominately male gender, and lower socioeconomic status are typical characteristics of this population [11]. In addition, published studies of VA patients undergoing bariatric surgery report a higher prevalence of comorbidities than patients in other health care systems [5, 15, 19, 2224]. All of these characteristics are known risk factors for increasing morbidity and mortality rates following bariatric surgery [1215, 25]. Livingston et. al. summarized 1,067 consecutive patients undergoing gastric bypass surgery to assess individual risk factors predictive for adverse outcomes. These results demonstrate that age over 55 years has a threefold higher mortality from gastric bypass surgery than younger patients. In addition, male gender and higher weight are predictive for the development of severe life-threatening complications [12]. The patients in our series fit the profile of a typical VA patient population. Male gender, advanced age, BMI > 50, and a high percentage of comorbidities were common characteristics. Our cohort of patients had a higher percentage of comorbidities than reported in most published series on LREYGB [5, 15, 19] with preoperative incidences of T2DM, HTN, dyslipidemia, and OSA of 50%, 77%, 70%, and 80%, respectively. In comparison, a meta-analysis by Maggard et. al. on surgical treatment of obesity revealed corresponding prevalence of T2DM, HTN, dyslipidemia, and OSA of 11%, 38%, 32%, and 15%, respectively [26]. In addition to the above risk factors, around half of our patients had prior abdominal surgeries.

In spite of our high-risk patient population, our mortality rate was zero with early major and minor complication rates of 1.4% and 15.7%, respectively. Our late complication rate, excluding nutritional deficiencies, was 11.4%. Most of these late complications involved nutritional deficiencies with half of the patient population having vitamin D deficiency. However, in the majority of these patients, we did not have a baseline level. These complication rates are below the 30-day mortality rate of 1.4% and the overall complication rate of 19.7% reported in other published studies of VA patients [11].Our low complication rate may be attributed to our extensive preoperative evaluation, which was summarized earlier, and by our strict requirement for preoperative weight loss. Previous studies have shown that preoperative weight loss reduces operative time [27, 28] and morbidity rates [29] following gastric bypass surgery. Another study reports that preoperative weight loss in patients undergoing gastric bypass results in decreased length of hospital stay and more rapid postoperative weight loss [30]. Our unpublished data (presented at the American College of Surgeons Annual Meetings, October 2008) indicates that preoperative weight loss reduced the technical difficulty of the surgery and resulted in major physiological improvements. Abdominal wall depth and liver volume were significantly reduced. Physiological changes were evidenced by the improvements of DM, OSA, and HTN. The mean BMI of 56 at baseline was reduced to a mean BMI of 49 following completion of the VLCD.

Improvement and/or resolution of comorbidities following bariatric surgery have been well documented [9, 26]. Our results were comparable and revealed substantial improvement or remission in all obesity-related comorbidities following surgery. Importantly, these improvements were largely sustained over the long term. The most prominent of these improvements occurred with T2DM as evidenced by 97% of patients with either improvement or complete remission. However, some patients required reinitiation of T2DM medications coincident with weight regain in a subset. While the number of patients is too small to draw any significant conclusion, it remains an important observation and stresses the importance of long-term f/u to maintain good outcomes following bariatric surgery. HTN improved or was in remission in over 60% of patients. At the same time, 50% of patients required less or no HTN medication. Our study demonstrated significant and sustained excess weight loss following LREYGB with a mean 56% EWL achieved at a mean f/u of 39 months. However, when preoperative weight loss was accounted for, the mean % EWL increased to 63%. This weight loss was sustained over the long term with 48 patients demonstrating 55% EWL at 48 months following surgery. Our long-term % EWL was similar to that reported in other long-term studies [8, 20].

It should be noted that a 100% f/u rate was possible in our study due to availability of the Veterans Affairs Computerized Patient Record System. This system adds to the validity of our data, as we are not aware of any long-term outcome paper that demonstrates a similar f/u rate. On the other hand, our study is limited due to the small number of patients, the inherent weaknesses of a retrospective review, and the bias of a predominately male population.

In conclusion, careful approach in this high-risk population can lead to a desired outcome with minimal morbidity and mortality. Optimization of organ systems and preoperative weight loss is thought to have played a significant role in our favorable outcomes. Long-term f/u in our patients has shown a sustained % EWL of >50%. This weight loss resulted in significant improvement or remission of most obesity-related comorbidities.

Conflict of interest

The authors claim no commercial associations that might be a conflict of interest in relation to this article.

Copyright information

© Springer Science + Business Media, LLC 2009