You are on page 1of 6

Surgery for Obesity and Related Diseases ] (2013) 00–00

Original article
Improvement in gastroesophageal reflux disease symptoms after various
bariatric procedures: review of the Bariatric Outcomes Longitudinal Database
Pradeep K. Pallati, M.D.c, Abhijit Shaligram, M.D.a, Valerie K. Shostrom, M.S.b,
Dmitry Oleynikov, M.D., F.A.C.S.a, Corrigan L McBride, M.D., F.A.C.S.a,
Matthew R. Goede, M.D.a,*
a
Department of Surgery, University of Nebraska Medical Center, Omaha, Nebraska
b
Department of Biostatistics, University of Nebraska Medical Center, Omaha, Nebraska
c
Department of Surgery, Creighton University Medical Center, Omaha, Nebraska
Received May 14, 2013; accepted July 23, 2013

Abstract Background: The prevalence of gastroesophageal reflux disease (GERD) in the morbidly obese
population is as high as 45%. The objective of this study was to compare the efficacy of various
bariatric procedures in the improvement of GERD.
Methods: The Bariatric Outcomes Longitudinal Database is a prospective database of patients who
undergo bariatric surgery by a participant in the American Society of Metabolic and Bariatric
Surgery Center of Excellence program. GERD is graded on a 6-point scale, from 0 (no history of
GERD) to 5 (prior surgery for GERD). Patients with GERD severe enough to require medications
(grades 2, 3, and 4) from June 2007 to December 2009 are identified; the resolution of GERD is
noted based on 6-month follow-up.
Results: Of a total of 116,136 patients, 36,938 patients had evidence of GERD preoperatively.
After excluding patients undergoing concomitant hiatal hernia repair or fundoplication, there were
22,870 patients with 6-month follow-up. Mean age was 47.6 11.1 years, with an 82% female
population. Mean BMI was 46.3 8.0 kg/m2. Mean preoperative GERD score for patients with
Roux-en-Y gastric bypass (RYGB) was 2.80 .56, and mean postoperative score was 1.33 1.41
(P o .0001). Similarly, adjustable gastric banding (AGB, 2.77 .57 to 1.63 1.37, P o .0001)
and sleeve gastrectomy (SG, 2.82 .57 to 1.85 1.40, P o .0001) had significant improvement
in GERD score. GERD score improvement was best in RYGB patients (56.5%; 7955 of 14,078)
followed by AGB (46%; 3773 of 8207) and SG patients (41%; 240 of 585).
Conclusion: All common bariatric procedures improve GERD. Roux-en-Y gastric bypass is
superior to adjustable gastric banding and sleeve gastrectomy in improving GERD. Also, the greater
the loss in excess weight, the greater the improvement in GERD score. (Surg Obes Relat Dis
2013;]:00-00.) r 2013 Published by Elsevier Inc. on behalf of American Society for Metabolic and
Bariatric Surgery.
Keywords: Gastroesophageal reflux disease (GERD); Bariatric Outcomes Longitudinal Database (BOLD); Roux-en-Y
gastric bypass; Adjustable gastric banding; Sleeve gastrectomy

The prevalence of obesity has been increasing at an


Presented as a poster of distinction at the 29th American Society of alarming rate during the past decades in the United States,
Metabolic and Bariatric Surgery Meeting, San Diego, California, as shown by the recent reports from Centers for Disease
June 2012. Control and Prevention [1]. According to this report, in
*
Correspondence: Matthew Goede, M.D., Department of Surgery,
University of Nebraska Medical Center, 983280 Nebraska Medical Center,
2009–2010, 35.7% of U.S. adults are obese. Obesity, defined
Omaha, NE 68198-3280. as body mass index (BMI) Z30 kg/m2, increases the risk of
E-mail: mgoede@unmc.edu cardiovascular disease, hypertension, dyslipidemia, and type

1550-7289/13/$ – see front matter r 2013 Published by Elsevier Inc. on behalf of American Society for Metabolic and Bariatric Surgery.
http://dx.doi.org/10.1016/j.soard.2013.07.018
2 P.K. Pallati et al. / Surgery for Obesity and Related Diseases ] (2013) 00–00
2 diabetes, among others, and decreases the life expectancy SRC. An Institutional Review Board approval was also
significantly [2]. In addition, obesity increases the risk of obtained from our institution to conduct the research. The
gastroesophageal reflux (GERD), with increased prevalence data were presented to us in a Microsoft Access format with
of Barrett’s esophagus and esophageal cancer. The preva- a unique identification for the individual patient. Several
lence of GERD in the general population has been shown to variables were identified, but the data were approved for
be in the range of 8% to 26%. In obese individuals, this is research on effect of various procedures on GERD only.
considerably higher than in the nonobese population [3].
The treatment of GERD involves dietary and lifestyle Study data
intervention along with pharmacotherapy and antireflux
procedures. As noted in a review article, dietary and lifestyle All patients operated on between June 2007 and Decem-
intervention may improve GERD in obese patients; how- ber 2009, with at least 6 months of follow-up are identified.
ever, the most favorable effect is likely to be found after All the variables entered into the BOLD database have been
bariatric surgery [4]. The fundoplication for treatment of previously reported in the literature [10]. The data used in
GERD has been associated with increasing failures in the this study included age, sex, race, BMI, excess weight and
severely obese patients, although results are conflicting, and GERD co-morbidity. GERD is identified as a co-morbidity,
it has been suggested that bariatric surgery rather than and it is graded as shown in Table 1. Because grade 5
fundoplication should be strongly considered in these includes patients who meet criteria for both antireflux
patients [5]. Several studies have focused specifically on surgery and prior surgery for GERD, these patients were
the improvement of GERD after various bariatric procedures excluded. The study population included patients with
[6–9]. Roux-en-Y gastric bypass (RYGB) has been associ- GERD severe enough to require medications (grades 2, 3,
ated with the most improvement in GERD symptoms, but and 4). Patients undergoing concomitant hiatal hernia repair
adjustable gastric banding and sleeve gastrectomy have also or fundoplication at the time of the bariatric surgery were
shown some improvement; the results are conflicting [8,9]. excluded from the results.
The reports so far involve only small cohort of patients in an
individual center study or a review of the current literature. Statistics
Bariatric Outcomes Longitudinal database (BOLD) is a
nationwide database that collects information on various PC SAS version 9.2 (SAS Institute, Cary, USA) was used
procedures, outcomes, and other parameters as entered by for all summaries and analyses. Gender and race are
American Society for Metabolic and Bariatric Surgery summarized using frequencies and percentages. Descriptive
Centers of Excellence (BSCOE) participants. The present statistics, including the mean and standard deviation are
study reviews the outcome of GERD with various proce- used to describe age and BMI. The difference in GERD
dures in the BOLD database. score at 6 months is calculated by subtracting the pre-
GERD score for each subject. This difference is classified as
Methods improvement (a decrease in GERD score), no improvement
(no change in GERD score), or worsening (an increase in
Database GERD score).
Percentage of excess weight loss (%EWL) was catego-
BOLD is a proprietary, Internet-based software product
rized as 4 levels. The first level was no change or a gain in
developed by the Surgical Review Corporation (SRC) to
excess weight. The remaining 3 levels were based on the
collect prospective data for all bariatric surgery patients treated
tertiles for %EWL. These levels were low change in %EWL
by American Society for Metabolic and Bariatric Surgery
BSCOE participants for the purpose of assessing outcomes
and quality of care. Data entry started in June 2007, and all Table 1
BSCOE participants have been required to enter patient data Symptom grading based on medication use
into BOLD since January 2008. The participants enter
Grade Symptoms Number of patients
preoperative, intraoperative, and postoperative data on all (Total number ¼ 116,136)
research-consented patients, as well as long-term follow-up
0 No symptoms of GERD 62,881
information for a minimum of 5 years. Recently, SRC has
1 Intermittent or variable symptoms, 15,908
transferred administration of the BSCOE program to the no medication
American Society for Metabolic and Bariatric Surgery; how- 2 Intermittent medication 10,969
ever, the data was obtained before transfer of the database. 3 H2 blockers or low dose PPI 23,241
4 High dose PPI 2729
5 Meet criteria for antireflux surgery, 408
Data acquisition
or prior surgery for GERD

A formal BOLD Data Access Request along with a GERD ¼ gastroesophageal reflux disease; H2 ¼ histamine 2; PPI ¼
BOLD Data Dissemination Request was submitted to the proton pump inhibitor.
Gastroesophageal Reflux After Bariatric Surgery / Surgery for Obesity and Related Diseases ] (2013) 00–00 3
(change between 0% and 34%), moderate change in %EWL
(change between 35% and 54%), and high change in
%EWL (change 454%). A general linear model was used
to compare the difference in GERD values among %EWL
categories. Pairwise comparisons were made when the
overall P value was o.05. The Tukey adjustment was used
for multiple comparisons.
A χ2 test was used to compare the change in GERD score
among the surgery types and %EWL categories. Because
both χ2 tests were statistically significant, a multivariate
logistic regression model with terms for %EWL category,
surgery type, and the interaction of these 2 terms was fit.
The interaction was not significant and was then removed
from the model leaving the fixed effect terms for %EWL
category and surgery type. The statistical level of signifi-
cance was set at .05 for all comparisons. Fig. 1. Change in gastroesophageal reflux disease composite score based
on procedure type.

Results by RYGB (2%) and AGB (1.2%). The remainder of the


patients had no change in their GERD status.
A total of 116,136 patients underwent bariatric surgery The change in the composite score is shown in Fig. 1.
from June 2007 to December 2009. Of these, 36,938 Pairwise comparisons showed that RYBG is better than
patients had preoperative evidence of GERD severe enough both AGB and SG (P o .0001) and that AGB is better than
to require medications. After excluding patients undergoing SG in improving GERD (P ¼ .02).
concomitant hiatal hernia repair or fundoplication, there The change in the composite score based on %EWL is
were 31,642 patients. Of these, 22,870 (72.3%) patients had shown in Fig. 2. All levels showed significant improvement
a 6-month follow-up recording the outcome of GERD. in GERD composite score. Tukey adjusted pairwise compar-
Demographic characteristics of these patients are shown ison showed that there were statistically significant differ-
in Table 2. Predominantly more patients had RYGB ences in all of the %EWL category comparisons except for
(61.6%) than adjustable gastric banding (AGB, 35.9%) low %EWL versus no change/gain in %EWL. The greater
and sleeve gastrectomy (SG, 2.5%). There were signifi- the %EWL, the greater the improvement in GERD score.
cantly more men in the SG group compared with AGB and Table 3 presents the results of multivariate logistic
RYGB (P o .05). Mean age of the patients was regression for improvement in GERD score. When surgery
47.6 11.1 years and involves predominantly Caucasians type is held constant, the odds of GERD improvement
(83.7%). Weight loss was significantly greater in RYGB increase with greater change in %EWL compared with no
patients, followed by SG and AGB. change/gain in %EWL. And when %EWL category is held
GERD score improvement was significantly highest in constant, the odds of GERD improvement are much greater
RYGB patients (56.5%), followed by AGB (46%) and SG for RYGB compared with the other 2 surgery types.
patients (41%) (P o .05). Worsening of GERD was seen in We also reviewed the effect of surgery type on patients
a small number of patients, mostly in SG (4.6%), followed who had GERD score of 0 or 1 preoperatively. Of 69,542

Table 2
Demographic data
Variable RYGB N ¼ 14,078 AGB N ¼ 8207 SG N ¼ 585 Total N ¼ 22,870 P value

Gender .1162
Male 17.9% 17.8% 21.2% 18%
Female 82.1% 82.2% 78.8% 82%
Age, years (mean SD) 47.5 10.9 47.7 11.5 48.0 11.0 47.6 11.1 .1189
Preoperative BMI (mean SD) 47.1 8.18 44.8 7.17 48.5 10.3 46.3 7.98 o.0001
Postoperative BMI (mean SD) 33.2 6.58 38.6 6.48 36.1 8.19 35.2 7.07 o.0001
Race
Caucasian 83.9% 83.4% 83.4% 83.7% .5214
Black 7.5% 8.7% 8.2% 8% .0109
Hispanic 4.8% 4% 4.8% 4.5% .0088
Other race 4.2% 4.5% 3.6% 4.3% .3826

AGB ¼ adjustable gastric banding; RYGB ¼ Roux-en-Y gastric bypass; SD ¼ standard deviation; SG ¼ sleeve gastrectomy.
4 P.K. Pallati et al. / Surgery for Obesity and Related Diseases ] (2013) 00–00

Fig. 2. Change in gastroesophageal reflux disease composite score based on percentage of excess weight loss (%EWL).
patients, only 48,101 (69.2%) patients had 6 month follow- literature and found that “AGB has antireflux properties
up score. Worsening of GERD was significantly more in resulting in resolution or improvement of reflux symptoms,
SG (9.2%) followed by RYGB (4.6%) and AGB (2.7%). The normalized pH monitoring results, and a decrease of
remainder of the patients had no change or had esophagitis on short term. However, worsening or newly
improvement in their GERD status. developed reflux symptoms and esophagitis are found in a
subset of patients during longer follow-up.” The present
study was a 6-month follow-up that showed improvement
Discussion
in GERD in 46.5% patients, worsening in only 1.3%, and
Roux-en-Y gastric bypass has been shown to be asso- no change in the remainder of the patients. In patients with
ciated with significant improvement in GERD, which is minimal or no symptoms before the surgery, only 2.7%
corroborated by our study. AGB is also associated with developed worsening GERD.
improvement in GERD in 46.5% patients. Although this is SG has been implicated with worsening and new onset
significantly less than RYGB, it does represent improve- GERD in previously nonreflux patients. Howard et al. [14]
ment in up to half the patients undergoing AGB. Similarly, reported an 82% worsening of GERD symptoms after SG.
sleeve gastrectomy (SG) is associated with 41.7% improve- Tai et al. [15] also noted a significant increase in the
ment in GERD. This represents a fairly good number of prevalence of GERD symptoms (12.1% versus 47%) and
patients overall. erosive esophagitis (16.7% versus 66.7%) after SG in 66
AGB has been implicated with both improvement and patients with 1 year follow-up. However, Daes et al. [16]
worsening of GERD in the literature [9,11–13]. Dixon et al. reported that with careful attention to surgical technique
[11] and Woodman et al. [12] reported significant improve- along with concomitant hiatal hernia repair, incidence of
ment in GERD of up to 91% in a small number of patients. GERD decreased from 49.2% to 1.5% at 6–12 months
On the contrary, Ovrebø et al. [13] reported markedly postoperatively. Also, Soricelli et al. [17] noted GERD
increasing gastroesophageal reflux after AGB. De Jong remission in 73.3% of the patients who had preoperative
et al. [9] performed a systematic review of the available GERD when combined hiatal hernia repair was performed.

Table 3
Odds of gastroesophageal reflux disease improvement at Z6 months
Effect level Reference level Odds ratio 95% confidence interval Odds ratio P value

%EWL category
Low %EWL No change/gain 1.107 (0.865, 1.416) .4206
Moderate %EWL No change/gain 1.364 (1.063, 1.749) .0146
High %EWL No change/gain 1.579 (1.228, 2.030) .0004
Surgery type
AGB RYGB 0.809 (0.749, .874) o.0001
SG RYGB 0.567 (0.479, .671) o.0001

%EWL ¼ percentage of excess weight loss; AGB ¼ adjustable gastric banding; RYGB ¼ Roux-en-Y gastric bypass; SG ¼ sleeve gastrectomy.
Gastroesophageal Reflux After Bariatric Surgery / Surgery for Obesity and Related Diseases ] (2013) 00–00 5
“De novo” GERD symptoms developed in 22.9% of the the disease based on the use of acid suppressive medica-
patients undergoing SG alone compared with 0% of patients tions and does not have any objective data, such as 24-hour
undergoing SG plus hiatal hernia repair. Finally, a recent pH monitoring, to confirm gastroesophageal reflux. The
review of literature on the effect of SG on GERD by Karmali variation in the number of procedures is inherent to the data,
et al. [8] in 2011 noted that of a total of 11 studies that had especially the low number of sleeve gastrectomies noted.
both preoperative and postoperative data, 4 studies reported Also, these procedures are performed with minor modifica-
an increase in GERD after SG and 7 studies found reduced tions across the nation with no standardized technique
GERD prevalence after SG. Similarly, the present study across various institutions. As with sleeve gastrectomy,
found GERD score improvement in 41% of patients with there has been variation in the distance from pylorus, size of
preexisting GERD. Worsening of GERD was seen in the sleeve, and hiatal dissection, all of which could affect
4.6%, and “de novo” GERD was seen in 9.2% of the the GERD itself. We did remove the procedures involving
patients, which was significantly higher than other groups. hiatal hernia repair to remove the bias, however this
RYGB has been the procedure of choice for patients with limitation needs to be noted.
GERD. In the present study, it was shown to result in
improvement of GERD in 56.5% of patients. This is in Conclusions
concordance with the available literature, which shows
similar results. Murr et al. [7] noted 94% improvement in All common bariatric procedures improve GERD. Roux-
typical reflux symptoms after gastric bypass for morbid en-Y gastric bypass is superior to adjustable gastric banding
obesity by 9 months with 4% showing worsening symp- and sleeve gastrectomy in improving GERD. Improvement
toms. Schauer et al. [18] noted similar significant decrease in GERD is also a function of weight loss, because patients
in GERD-related symptoms, including heartburn (from 87% with higher weight loss experience significantly higher
to 22%, P o.001), in their series after gastric bypass. improvement in GERD.
RYGB combines a low-pressure small gastric pouch that
enters directly into the jejunum. It reduces exposure to acid Disclosures
and prevents bile reflux, thus resulting in maximum
improvement in GERD. Prachand et al. [19] compared the The authors have no commercial associations that might
resolution of co-morbidities between patients undergoing be a conflict of interest in relation to this article.
either RYGB or duodenal switch (DS). Although DS
offered better weight loss, greater risk of remission of References
diabetes, greater improvement in hypertension, and much
better control of hyperlipidemia, RYGB was more effective [1] Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity in
for GERD. the United States, 2009–2010. NCHS data brief, no 82. Hyattsville,
MD: National Center for Health Statistics; 2012.
The %EWL categories noted in this study take into [2] Peeters A, Barendregt JJ, Willekens F, Mackenbach JP, Al Mamun A,
account the tertiles based on the %EWL noted. We Bonneux L. NEDCOM, the Netherlands Epidemiology and Demog-
observed that, of the patients losing weight, one third raphy Compression of Morbidity Research Group. Obesity in adult-
lose o35%, one third lose 455%, and the remainder lose hood and its consequences for life expectancy: a life-table analysis.
between 35%–55%. Surprisingly, the group with no change Ann Intern Med 2003;138:24–32.
[3] Ogden CL, Yanovski SZ, Carroll MD, Flegal KM. The epidemiology
or gain in weight within 6 months after surgery also noted of obesity. Gastroenterology 2007;132:2087–102.
significant improvement in postoperative GERD scoring. [4] De Groot NL, Burgerhart JS, Van De Meeberg PC, de Vries DR,
On further review, we noted that all the patients with Smout AJ, Siersema PD. Systematic review: the effects of conserva-
improvement were in the AGB or RYGB group only. We tive and surgical treatment for obesity on gastro-oesophageal reflux
hypothesize that AGB and RYGB inherently decrease disease. Aliment Pharmacol Ther 2009;30:1091–102.
[5] Prachand VN, Alverdy JC. Gastroesophageal reflux disease and
GERD as a result of the decrease in amount of gastric severe obesity: Fundoplication or bariatric surgery? World J Gastro-
juice present in the proximal pouch. We also noted that, if enterol 2010;16:3757–61.
you have significantly higher %EWL with whatever the [6] Klaus A, Gruber I, Wetscher G, et al. Prevalent esophageal body
surgery type, you have significantly greater improvement in motility disorders underlie aggravation of GERD symptoms in
GERD. This is expected from the loss of high intra- morbidly obese patients following adjustable gastric banding. Arch
Surg 2006;141:247–51.
abdominal pressure resulting in decreased pressure gra- [7] Nelson LG, Gonzalez R, Haines K, Gallagher SF, Murr MM. Ameli-
dients between stomach and esophagus. oration of gastroesophageal reflux symptoms following Roux-en-Y
This study is based on the largest registry of bariatric gastric bypass for clinically significant obesity. Am Surg 2005;71:950–3.
surgery patients available today with good follow-up. [8] Chiu S, Birch DW, Shi X, Sharma AM, Karmali S. Effect of sleeve
However, there are a few limitations. This is a retrospective gastrectomy on gastroesophageal reflux disease: a systematic review.
Surg Obes Relat Dis 2011;7:510–5.
database research that is based on the quality of database, [9] De Jong JR, Besselink MGH, van Ramshorst B, Gooszen HG, Smout AJ.
which is subject to continuous quality improvement within Effects of adjustable gastric banding on gastroesophageal reflux and
BOLD. GERD, as documented in the BOLD, only identifies esophageal motility: a systematic review. Obes Rev 2010;11:297–305.
6 P.K. Pallati et al. / Surgery for Obesity and Related Diseases ] (2013) 00–00
[10] DeMaria EJ, Pate V, Warthen M, et al. Baseline data from American 1 year after laparoscopic sleeve gastrectomy among obese adults.
Society for Metabolic and Bariatric Surgery-designated Bariatric Surg Endosc 2013;27:1260–6.
Surgery Centers of Excellence using Bariatric Outcomes Longitudinal [16] Daes J, Jimenez ME, Said N, Daza JC, Dennis R. Laparoscopic
Database. Surg Obes Relat Dis 2010;6:347–55. sleeve gastrectomy: symptoms of gastroesophageal reflux can be
[11] Dixon JB, O’Brien PE. Gastroesophageal reflux in obesity: the effect reduced by changes in surgical technique. Obes Surg 2012;22:
of lap band placement. Obes Surg 1999;9:527–31. 1874–9.
[12] Woodman G, Cywes R, Billy H, Montgomery K, Cornell C, Okerson
[17] Soricelli E, Iossa A, Casella G, Abbatini F, Calì B, Basso N. Sleeve
T, APEX Study Group. Effect of adjustable gastric banding on
gastrectomy and crural repair in obese patients with gastroesophageal
changes in gastroesophageal reflux disease (GERD) and quality of
reflux disease and/or hiatal hernia. Surg Obes Relat Dis. Epub 2012
life. Curr Med Res Opin 2012;28:581–9.
Jun 19.
[13] Ovrebø KK, Hatlebakk JG, Viste A, et al. Gastroesophageal reflux in
morbidly obese patients treated with gastric banding or vertical banded [18] Frezza EE, Ikramuddin S, Gourash W, et al. Symptomatic improve-
gastroplasty. Ann Surg 1998;228:51–8. ment in gastroesophageal reflux disease (GERD) following laparo-
[14] Howard DD, Caban AM, Cendan JC, Ben-David K. Gastroesopha- geal scopic Roux-en-Y gastric bypass. Surg Endosc 2002;16:1027–31.
reflux after sleeve gastrectomy in morbidly obese patients. Surg Obes [19] Prachand VN, Ward M, Alverdy JC. Duodenal switch provides
Relat Dis 2011;7:709–13. superior resolution of metabolic comorbidities independent of weight
[15] Tai CM, Huang CK, Lee YC, Chang CY, Lee CT, Lin JT. Increase in loss in the super-obese (BMI Z50 kg/m2) compared with gastric
gastroesophageal reflux disease symptoms and erosive esophagitis bypass. J Gastrointest Surg 2010;14:211–20.

You might also like