Gastroesophageal Reflux and Gastrointestinal Symptoms After Metabolic and Bariatric Surgery in Adolescents: An 8-Year Follow-Up Analysis

Abstract

Importance:

Vertical sleeve gastrectomy (VSG) and Roux-en-Y gastric bypass (RYGB) are the most commonly performed metabolic and bariatric surgery (MBS) procedures in adolescents and adults. Despite their safety and effectiveness, there is concern over postoperative gastrointestinal symptoms (GIS), especially gastroesophageal reflux symptoms (GERS), in those undergoing VSG.

Objective:

To evaluate the long-term prevalence of GIS in adolescents who underwent RYGB or VSG.

Design, Setting, and Participants:

This is a prospective, multicenter, observational cohort study at five academic referral centers in the United States. Patients were enrolled from February 28, 2007, through December 30, 2011. The analysis included 228 adolescents: 161 RYGB and 67 VSG followed prospectively for 8 years.

Main Outcomes and Measures:

Patient-reported GIS before surgery and across 8 years of postoperative follow-up were assessed. We dichotomized postoperative symptom severity and analyzed the data using general linear mixed models.

Results:

Adolescents undergoing either VSG or RYGB demonstrated significant increases in abdominal pain (10% vs. 17%), bloating (8% vs. 20%), and constipation (3% vs. 9%) between baseline and 8 years (p<0.05). Following RYGB, the prevalence of GERS was not statistically significantly different between baseline (12%) and 8 years (13%)(p>0.05). Following VSG, however, GERS increased from 9% preoperatively to 27% at 8 years (p<0.05). In adjusted analyses, VSG was associated with higher odds of GERS at 8 years (adjusted odds ratio 2.67 [1.57-4.55, 95%CI]).

Conclusions and Relevance:

GERS represents a considerable concern pre- and post-MBS in adolescents, especially after VSG. Appropriate patient selection along with counseling and objective monitoring for pathologic consequences of gastroesophageal reflux after MBS are warranted.

Trial Registration:

clinicaltrials.gov Identifier: NCT00474318

Introduction

Obesity is a global pandemic with 39% of the world population either being overweight or having obesity[1]. In the past several decades, this pandemic has expanded to younger individuals with one-fifth of U.S. children having obesity (BMI ≥ 95^th percentile); with approximately 8% being further categorized as having class 2 obesity or higher (BMI ≥ 120% of the 95^th percentile or greater or BMI ≥ 35 kg/m²)[1-3]. Obesity in adults and children is associated with numerous comorbidities (e.g. hypertension, diabetes mellitus, and obstructive sleep apnea)[4] that are becoming more prevalent in affected youth[5-9]. Weight loss improves the management and remission of obesity-associated comorbidities[10]. Metabolic and bariatric surgery (MBS) has been proven to offer superior outcomes compared to intensive lifestyle treatment alone or in combination with medical therapy in the management of obesity and its sequalae[11]. However, recently developed pharmacologic approaches using endocrine-based therapies are emerging and may markedly improve results over lifestyle modification therapy alone and/or in combination with earlier-generation pharmacological obesity therapies[12]. The long-term outcomes of these obesity medications in pediatrics and adolescence are, however, still largely unknown.

Roux-en-Y gastric bypass (RYGB) and vertical sleeve gastrectomy (VSG) are the two most commonly performed MBS operations in adolescents and adults[13, 14]. Both procedures have been shown to be safe and effective[15, 16]. During the past decade, VSG has surpassed RYGB as the most common procedure and has been shown to have similar clinical outcomes in terms of weight loss and resolution of related comorbidities when compared to RYGB[13, 17, 18]. Both procedures are associated with distinct adverse event profiles related to the alteration of gastrointestinal anatomy, and VSG is thought to have an overall favorable risk profile compared to RYGB[18, 19]. However, conflicting evidence exists about the association of VSG with gastroesophageal reflux (GER)[20-24].

Our group previously demonstrated that participants enrolled in the Teen-Longitudinal Assessment of Bariatric Surgery (Teen-LABS) prospective observational study who underwent VSG reported higher levels of gastroesophageal reflux symptoms (GERS) compared to those who underwent RYGB at 5 years postoperatively[24]. In this current analysis, we aimed to compare overall GERS between VSG and RYGB up to 8 years postoperatively.

Methods

Study Design, Subjects, and Data Collection

Data from the Teen-LABS study were used for this analysis. The methodology and design have been well-described previously[25]. The study protocol, data and safety monitoring plans, and assent and/or consent forms were approved by all institutional review boards and an independent data and safety monitoring board convened by the NIH.

Adolescents (age 13-19) with a BMI of 35 kg/m² or greater who met eligibility criteria for MBS were enrolled at five participating sites in the U.S. Participants were recruited between February 2007 and December 2011 and completed study visits at 6 months, 12 months, then annually to assess outcomes up until year 6, when data collection transitioned from annually to every other year.

After excluding data for participants who underwent adjustable gastric band insertion (n=14), 228 individuals underwent baseline data collection (161 RYGB and 67 VSG). At 8 years, data were available for 175/228 (77%); 126/161 (78%) in the RYGB group, and 49/67 (73%) in the VSG group.

Data points and Definitions

Demographic data were collected and included the participant’s age at the time of surgery, race (White, Black, Asian, American Indian, or Alaskan Native and mixed race), ethnicity (Hispanic or non-Hispanic), and biological sex (male or female). BMI and percent BMI change were calculated at each visit. Full data on participant baseline characteristics as well as full data-point definitions can be found in our previous publication[24].

Gastrointestinal symptoms (GIS) were surveyed at baseline and each follow-up visit using a 15-item GIS scale[26]. Of the seven possible responses of increasing symptom severity (none to very severe [1 to 7]), patients who reported their symptoms as moderate or higher (≥4) were considered positive, whereas those with mild to no symptoms (<4) were considered negative for that symptom. Responses to 1) heartburn and 2) acid reflux were combined into one symptom defined as GERS and, if either heartburn or acid reflux were positive, they were considered GERS-positive. Those with mild to no symptoms of both heartburn and acid reflux were considered GERS-negative. Incidence (symptom reported at a follow-up visit but not present at baseline), prevalence (presence of the symptom at any of the study visits), and remission (symptoms reported at baseline but not at a follow-up visit) were calculated.

Statistical Methods and Analysis

Descriptive statistics were reported as mean ± standard deviation (SD) for continuous variables or frequency and percentage for categorical variables. Sex, race (Black vs. non-Black), and BMI change from baseline were selected a priori as covariates for multivariable regression models and assessed for multicollinearity. General linear mixed models with logit link were used to examine the adjusted odds ratios (aOR) for the prevalence of GIS at baseline, 6 months, and yearly at 1 year through 8 years post-surgery, except year 7. Models were created for each symptom of interest. BMI change from baseline was included as a time-dependent covariate, and years since surgery and surgery type were the independent variables of interest. A visit by surgery type interaction was used to examine the difference in GIS rates at baseline to 8 years post-surgery between RYGB and VSG. If the interaction was not significant, the main effect of surgery type was examined to test the overall difference between RYGB and VSG. An additional subgroup analysis examining the aOR for the prevalence of severe/very severe GERS was also performed to evaluate whether findings were similar when evaluating severe GERS only. For this analysis, we elected to adjust for medication use, considering that we could not extract the severity of symptoms from reports of medication use only. SAS^®, version 9.4 (SAS Institute, Cary, NC) was used for analysis. A p-value of <0.05 was considered statistically significant.

Results

Participant Characteristics at Baseline

Selected patient characteristics are available in Table 1. At baseline, the mean age of participants at surgery was 17 ± 2 years and 75% of the total cohort was female. The most prevalent symptoms in the combined cohort included hunger pain (20%), rumbling (13%), flatulence (11%), and GERS (11%). In addition, 18% also reported the use of antacid or acid reflux medications.

Table 1:

Patient Demographic, Anthropomorphic Characteristics and Gastrointestinal Symptoms at Baseline

Variable	Overall (n=228)	RYGB (n=161)	VSG (n=67)
Age at surgery (years)	16.5 (±1.6)	16.5 (±1.5)	16.4 (±1.7)
Sex (Female)	171 (75%)	126 (78%)	45 (67%)
Race:
White	164 (72%)	119 (73%)	45 (67%)
Black	50 (22%)	35 (22%)	15 (22%)
Asian	1 (0.4%)	1 (1%)	0 (0%)
American Indian or Alaskan Native	1 (0.4%)	0 (0%)	1 (2%)
More than one race	12 (5%)	6 (4%)	6 (9%)
BMI (kg/m2)	52.6 (±9.4)	53.7 (±9.6)	50.1 (±8.3)
Gastrointestinal symptoms (score ≥ 4)
Abdominal Pain	22 (10%)	15 (9%)	7 (10%)
Hunger pain	46 (20%)	27 (17%)	19 (28%)
Rumbling	29 (13%)	18 (11%)	11 (16%)
Bloating	18 (8%)	12 (7%)	6 (9%)
Flatulence	26 (11%)	18 (11%)	8 (12%)
Constipation	6 (3%)	5 (3%)	1 (1%)
Diarrhea	18 (8%)	13 (8%)	5 (7%)
Loose stools	11 (5%)	7 (4%)	4 (6%)
Hard stools	12 (5%)	9 (6%)	3 (4%)
Urgent bowel	17 (7%)	13 (8%)	4 (6%)
Belching	8 (4%)	7 (4%)	1 (1%)
Nausea	23 (10%)	17 (11%)	6 (9%)
Heartburn	19 (8%)	17 (11%)	2 (3%)
Acid reflux	18 (9%)	12 (7%)	6 (9%)
Other Related Variables
GERS*	25 (11%)	19 (12%)	6 (9%)
Antacid	35 (15%)	24 (15%)	11 (16%)
Acid reflux medication	30/227 (13%)	19 (12%)	11/66 (17%)
Medication**	42 (18%)	28 (17%)	14 (21%)

Data presented as mean (standard deviation) or n (%)

RYGB is Roux-en-Y gastric bypass and VSG is vertical sleeve gastrectomy

^*GERS – Gastroesophageal reflux symptoms (defined as the presence of heartburn or acid reflux)

^**Reported Antacid use or Acid reflux medication

Gastrointestinal Symptoms Post-Surgery

Between baseline and 8 years post MBS, participants reported an increase in the unadjusted prevalence of abdominal pain, bloating, and constipation with no significant change in the unadjusted prevalence of diarrhea (Table 2). Hunger pain decreased between baseline and 8 years (20% vs 6%, p<0.001). GERS prevalence showed a trend toward an increase (11% vs 17%, p=0.07) in the combined cohort, though this result was not statistically significant. The adjusted odds of reporting GERS increased by 13% between baseline and 8 years (OR 1.13 [1.07-1.19, 95%CI]).

Table 2:

Baseline and 8-year Crude Prevalence Rates and Adjusted Odds Ratios (aOR) for Change (per year: baseline to 8 years postoperatively) in the Presence of Moderate to Very Severe Gastrointestinal Symptoms and for the Type of Surgery

	Prevalence N (%)		Change in prevalence p- value	aOR (95% CI)
	Baseline N=228	8 Years N=175	Baseline to 8y	Change per year Baseline to 8y	Surgery (RYGB reference)
Abdominal Pain	22 (10)	30 (17)	0.03	1.08 (1.02, 1.15)	1.21 (0.74, 1.99)
Hunger pain	46 (20)	11 (6)	<.0001	0.91 (0.84, 0.98)	1.60 (0.93, 2.75)
Rumbling	29 (13)	14 (8)	0.13	0.94 (0.87, 1.02)	1.16 (0.70, 1.92)
Bloating	18 (8)	35 (20)	0.0004	1.18 (1.11, 1.25)	1.26 (0.78, 2.04)
Flatulence	26 (11)	22 (13)	0.72	1.01 (0.96, 1.08)	1.03 (0.59, 1.80)
Constipation	6 (3)	15 (9)	0.008	1.09 (1.01, 1.18)	1.29 (0.68, 2.45)
Diarrhea	18 (8)	14 (8)	0.96	1.09 (1.02, 1.17)	0.95 (0.53, 1.71)
Loose stools	11 (5)	15 (9)	0.13	1.14 (1.06, 1.23)	1.03 (0.57, 1.87)
Hard stools	12 (5)	11 (6)	0.65	1.03 (0.94, 1.13)	1.22 (0.60, 2.49)
Urgent bowel	17 (7)	19 (11)	0.23	1.09 (1.02, 1.17)	1.20 (0.66, 2.17)
Belching	8 (4)	10 (6)	0.29	0.97 (0.90, 1.05)	1.64 (0.85, 3.17)
Heartburn	19 (8)	22 (13)	0.16	1.12 (1.05, 1.19)	2.47 (1.39, 4.38)
Acid reflux	18 (8)	24 (14)	0.06	1.13 (1.06, 1.20)	3.13 (1.70, 5.76)
GERS*	25 (11)	30 (17)	0.07	1.13 (1.07, 1.19)	2.67 (1.57, 4.55)
Nausea	23 (10)	28 (16)	0.08	1.07 (1.02, 1.14)	1.39 (0.88, 2.18)
Reported Antacid use	35 (15)	10 (6)	0.002	0.75 (0.68, 0.83)	0.80 (0.45, 1.42)
Acid reflux medication	30 (13)	26 (16)	0.39	0.94 (0.87, 1.01)	1.45 (0.89, 2.35)
Medication**	42 (18)	27 (17)	0.70	0.86 (0.80, 0.92)	1.20 (0.75, 1.93)
Severe GERS	7 (3)	11 (6)	0.12	1.14 (1.04, 1.25)	2.87 (1.26, 6.55)

Adjusted odds ratios (aOR) and 95% confidence intervals for change per year and surgery type were adjusted for sex, race, and percent change in BMI from baseline. This analysis focused on postoperative symptoms from baseline to 8 years, thus the aOR was calculated based on data from baseline to 8 years after surgery. All year-by-surgery interactions were non-significant. Surgery OR is for VSG vs RYGB (reference).

^*GERS – Gastroesophageal reflux symptoms (defined as the presence of heartburn or acid reflux)

^**Reported Antacid use or Acid reflux medication

Gastrointestinal Symptoms by Surgery Type

The unadjusted prevalence for selected gastrointestinal symptoms at baseline and every follow-up point is presented in Figure 1. Higher adjusted odds of GERS were found in the VSG cohort compared to RYGB over the 8 years (aOR 2.67 [1.57-4.55, 95%CI]; Table 3). The 8-year prevalence of GERS was 27% in the group, representing a significant increase over baseline (p=0.04) while the prevalence of GERS was 13% at 8 years in the RYGB group, which was not different from baseline for the RYGB group (p>0.05).

Figure 1. Prevalence of Selected Moderate to Very Severe Gastrointestinal Symptoms at Baseline to 8 Years after Gastric Bypass and Sleeve Gastrectomy.

*GERS - Gastroesophageal reflux (defined as the presence of heartburn or acid reflux)

Table 3:

Adjusted Odds Ratio (aOR) for Models Used to Examine Moderate to Very Severe Gastrointestinal Symptoms at Baseline to 8 years Postoperatively

	Change per year	Surgery (RYGB reference)	Female	Black	Percent decrease from baseline in BMI at visit (each 5%)
Abdominal Pain	1.08 (1.02, 1.15)	1.21 (0.74, 1.99)	2.76 (1.59, 4.82)	1.42 (0.88, 2.31)	0.98 (0.93, 1.04)
Bloating	1.18 (1.11, 1.25)	1.26 (0.78, 2.04)	2.78 (1.44, 5.38)	1.59 (0.96, 2.62)	0.98 (0.92, 1.06)
Constipation	1.09 (1.01, 1.18)	1.29 (0.68, 2.45)	1.75 (0.85, 3.62)	1.12 (0.59, 2.15)	0.98 (0.92, 1.05)
Diarrhea	1.09 (1.02, 1.17)	0.95 (0.53, 1.71)	1.88 (0.92, 3.83)	0.71 (0.40, 1.29)	1.11 (1.03, 1.20)
Belching	0.97 (0.90, 1.05)	1.64 (0.85, 3.17)	4.14 (1.90, 9.02)	1.08 (0.56, 2.06)	0.92 (0.86, 0.98)
GERS*	1.13 (1.07, 1.19)	2.67 (1.57, 4.55)	1.79 (0.98, 3.28)	0.99 (0.54, 1.82)	1.03 (0.97, 1.09)
Nausea	1.07 (1.02, 1.14)	1.39 (0.88, 2.18)	2.00 (1.24, 3.21)	1.34 (0.85, 2.13)	0.99 (0.93, 1.05)

Adjusted odds ratio (aOR) 95% confidence intervals for models examining the effect of year and surgery type adjusted for sex, race, and percent change in BMI from baseline. This analysis focused on postoperative symptoms from baseline to 8 years, thus aOR was calculated based on data from baseline to 8 years after surgery. Surgery aOR is for VSG vs RYGB (reference).

^*GERS – Gastroesophageal reflux symptoms (defined as the presence of heartburn or acid reflux)

^**Reported Antacid use or Acid reflux medication

When only evaluating acid reflux symptoms, the postoperative prevalence of GERS was significantly higher in the VSG group (11–24%) compared to the RYGB group (2–13%). Whereas remission of GERS was 100% in the first 6 months for RYGB and one year for VSG, remission decreased to a nadir at 8 years of 69% for RYGB and 40% for VSG (Figure 2).

Figure 2: Incidence and Remission of Moderate to Very Severe Gastroesophageal Reflux Symptoms (GERS) over time (6 months to 8 years) by Surgery Type.

^aIncidence was defined as GERS not reported at baseline but reported at the subsequent visit, 6 months through 8 years

^bRemission was defined as GERS reported at baseline but not reported at the subsequent visit 6 months through 8 years

RYGB is Roux-en-Y gastric bypass and VSG is vertical sleeve gastrectomy

GERS – Gastroesophageal reflux symptoms (defined as the presence of heartburn or acid reflux)

Severe/Very Severe GERS Subgroup Analysis

The unadjusted prevalence of severe GERS in the combined cohort did not change significantly between baseline and 8 years postoperatively (3% vs 6%, p=0.12). However, when separating the groups by surgery type, participants who underwent VSG demonstrated a significant increase from 1% at baseline to 14% at 8 years whereas the prevalence after RYGB remained stable between baseline (4%) and 8 years (3%). The adjusted odds ratio for severe GERS was higher in the VSG group (aOR 2.87 [1.26-6.55, 95%CI]) and persisted after adjusting for medication use (aOR 2.69 [1.21-5.99, 95%CI]).

Discussion

This analysis prospectively evaluated long-term gastrointestinal symptoms after MBS in adolescents. Building on our previously reported 6-month and 5-year postoperative outcomes[24], we found a higher prevalence of abdominal pain, bloating, and constipation at 8 years compared to baseline in the combined surgical cohort. We also found that those who underwent VSG had approximately three-fold higher odds of severe reflux symptoms at 8 years compared to those who underwent RYGB. These findings have important implications that should serve to inform patient selection and provide insight during preoperative clinical decision-making and counseling with patients regarding postoperative expectations. Additionally, the results of this analysis highlight the need for long-term patient follow-up and further research on postoperative esophageal inflammation and injury, including the development of clear clinical monitoring pathways and novel treatment approaches to reduce GIS occurrence.

The long-term outcomes of MBS among adolescents beyond 5 years remain an ongoing area of research. Our study is the first to evaluate GIS in post-MBS adolescents beyond 5 years, with the most closely associated research on this topic pertaining to post-MBS adults. Regarding abdominal pain, retrospective studies have demonstrated a prevalence of approximately 20% in post-MBS adults, similar to the 17% we observed[27]. Given abdominal pain has been identified as one of the most common reasons for emergency department (ED) visits and hospital readmissions following MBS in adults, several studies have evaluated risk factors for its development[28]. In particular, preoperative use of strong analgesics, depression, and postoperative complications have been implicated[29, 30]. Given the high prevalence of long-term abdominal pain following MBS in adolescents, further research into specific risk factors and possible underlying mechanisms is warranted to improve patient selection and assist in designing targeted preventive interventions.

While our study demonstrated an increased prevalence of bloating and constipation, similar studies in adults are heterogenous and report a wide range of prevalences. This is likely due to the broad range of etiologies that contribute to these symptoms in the non-MBS and MBS populations. Similar to abdominal pain, a direct relationship between these symptoms and the surgical procedure would be challenging to elucidate on a long-term basis postoperatively. However, when comparing the prevalence of these symptoms in our cohort to the general U.S. population, little difference is found. For instance, 20% of our cohort reported bloating at 8 years. This is similar to that of the average U.S. population with a prevalence quoted as one in seven individuals[31, 32]. In terms of bowel function, our data demonstrated an increase in the prevalence of constipation from 3% at baseline to 9% at 8 years, whereas unlike our 5-year results, the prevalence of diarrhea remained largely unchanged at long-term follow-up compared to baseline. Even so, the estimated prevalence of constipation in U.S. children and adolescents is approximately 10%,[33] and therefore not significantly different than our cohort. Ultimately, our results pertaining to bloating and constipation demonstrate the need for targeted postoperative monitoring of these symptoms with a patient-centered, individualized approach to the diagnosis and treatment of the underlying etiology.

Our findings pertaining to the persistence, relapse, and increased prevalence of GERS in VSG compared to RYGB are not only consistent with those in our 5-year report, but also with the adult literature[34, 35]. Several systematic reviews, and retrospective and meta-analyses in the adult MBS population found a significantly higher prevalence of de novo or worsening GERS following VSG compared to RYGB[34-37]. In one meta-analysis, the odds ratio for GERS following VSG in adults was found to be five times higher than in RYGB[38], greater than the three-fold risk observed in our study. The proposed mechanisms behind this phenomenon include increased gastric intraluminal pressure, decreased gastric compliance, and altered emptying mechanics due to the final shape of the gastric sleeve[24, 39-41].

A commonly employed method to achieve symptom improvement or remission in adult candidates for MBS with concomitant GERS is up-front RYGB, as this is considered an effective antireflux procedure[42]. Likewise, conversion from VSG to RYGB has proven safety and efficacy in adults for the treatment of medically refractory GERS with no clear anatomic abnormalities, such as hiatal hernia[43-45]. The superior improvement in GERS with RYGB compared to VSG is hypothesized to occur due to multiple factors, including the greater degree of weight loss typically achieved with RYGB along with physiological differences, including a greater reduction in the mass of gastric acid secreting parietal cells, downward traction of the gastric pouch and gastrojejunal anastomosis created by the roux limb, the removal of the gastric pylorus limiting upward fluid pressure, and surgical technique requiring little to no dissection at the esophageal hiatus[45-48].

In terms of surgical selection, the 2018 American Society for Metabolic and Bariatric Surgery (ASMBS) pediatric metabolic and bariatric surgery guidelines recommend RYGB in patients with severe GERS and severe obesity unless contraindicated, acknowledging that it is unclear if VSG may also be a reasonable approach to patients with GERD and obesity as long as there is no evidence of Barrett’s esophagus[49]. Similarly, the ASMBS released a position statement in 2021 advocating for RYGB over VSG in adults with severe GERS, while also acknowledging that GERS is not the only outcome of importance and that the surgeon, multidisciplinary team, and patient may appropriately decide to pursue VSG once the risks of persistent GERS are discussed[50]. Our data supports up-front RYGB for adolescent MBS candidates with preexisting severe GERS, though more data is needed to determine if adolescents benefit from conversion from VSG to RYGB with the same efficacy and safety profile described in the adult literature.

A useful clinical adjunct to further inform the suitability of performing RYGB or VSG is preoperative esophagogastroduodenoscopy (EGD). The primary goal of EGD in this setting is to evaluate for mucosal or anatomic abnormalities that may alter medical and/or surgical management, including esophagitis (e.g. GERD or eosinophilic disease), Barrett’s esophagus, peptic ulcer disease, hiatal hernia, celiac disease, malignancy, and the presence of H. Pylori[51]. The necessity of preoperative EGD in patients undergoing MBS is a topic of ongoing debate, with recommendations varying amongst professional societies, often based on the presence or absence of preoperative GIS. Interestingly, pooled studies including a total of 729 asymptomatic adults undergoing EGD prior to MBS demonstrated a prevalence of hiatal hernia, esophagitis, and positive pH testing in 17%, 16.9%, and 39.3% of patients, respectively[50]. In the pediatric MBS population, a retrospective review of 134 patients, 70% of whom underwent preoperative EGD, demonstrated that 46% had either an anatomic abnormality or mucosal disease requiring medical treatment; however, abnormal EGD findings were not associated with changes in surgical management or any postoperative complication[52]. These studies suggest that the incidence and prevalence of GER without symptoms is perhaps higher than one might infer from our study, though the clinical significance of this has yet to be determined, particularly if there is no impact on the choice or outcome of surgery.

The most recent societal guidelines pertaining to the use of preoperative EGD in adolescents undergoing MBS were published in 2021 by the Pediatric Committee of the ASMBS. The committee recommended routine preoperative EGD with biopsy be performed in all pediatric patients with significant upper GIS (high-quality evidence)[53]. For asymptomatic patients undergoing VSG, the committee concluded that there is insufficient evidence to recommend for or against routine preoperative EGD and advised that this decision should be left to the surgeon’s discretion (moderate-quality evidence). For patients undergoing RYGB, the committee stated that preoperative EGD can be considered due to the difficulty in assessing the GI tract postoperatively (moderate-quality evidence)[53]. The findings of our study, along with the 2018 ASMBS Pediatric Committee guidelines, substantiate the use of preoperative EGD in patients with severe GERS regardless of surgery type, particularly in those who prefer to undergo VSG, as the findings could significantly inform clinical and shared decision-making. However, further study of the utility of preoperative EGD in patients undergoing MBS with asymptomatic GERS is necessary, particularly in light of studies demonstrating abnormal EGD findings of uncertain clinical significance.

While a clear correlation between VSG and increased GERS has been established in the adult population, the data pertaining to the association between VSG, GERS, and the subsequent development of Barrett esophagus is discordant. For instance, a 2021 systematic review and meta-analysis of 10 studies reported an 11.6% prevalence of Barrett esophagus after VSG in adults over 6 months to 10 years follow-up. Interestingly, the prevalence of Barrett esophagus did not correlate with that of GERS[54]. Conversely, long-term outcomes from 240 adults included in the Sleeve vs. Bypass (SLEEVEPASS) randomized-control trial demonstrated that while the risk of GERS was higher in patients undergoing VSG compared to RYGB, there was no statistically significant difference in the cumulative incidence of Barrett esophagus[35]. Additionally, a large retrospective bariatric surgery registry trial including long-term follow-up data showed no difference in the postoperative incidence of Barrett esophagus between VSG and RYGB in adults[34].

Postoperative endoscopic surveillance is addressed by the aforementioned 2018 ASMBS Pediatric Committee guidelines and 2021 ASMBS position statement pertaining to adults. The 2018 guidelines recommend considering a baseline post-VSG screening EGD in all pediatric patients when transitioning to adult care, due to the risk of poor follow-up and asymptomatic reflux (low-quality evidence). The guidelines also recommend annual screening for GIS by the patient’s pediatrician or primary care provider and if present, timely endoscopic evaluation (high-quality evidence). The 2021 position statement conditionally recommends clinicians consider a screening EGD for all adults who have undergone MBS and have GIS, including GERS, and in all adults 3 or more years following VSG, regardless of symptoms[50]. When compared to adults, there is a paucity of data regarding monitoring and management of Barrett esophagus in children, regardless of MBS status, and therefore, further research is warranted to obtain objective longitudinal data on GERS and its consequences following VSG in this patient population. Nonetheless, our data further validates the 2018 ASMBS Pediatric Committee guidelines and supports screening for GIS with a low threshold for EGD in patients who have undergone MBS.

Limitations of this study include a risk of recall bias due to the subjective nature of the symptoms which could also be influenced by the relative infrequency of data collection, occurring predominantly once per year. Additionally, while our data is heterogeneous due to the observational design of the study, adjustment for potential confounders was performed in the logistic regression analysis. Despite these limitations, the prospective design, length of follow-up, uniformity of data definitions and collection methods (particularly the use of a standardized GIS rating scale), and high subject retention and visit completion from multiple distinct sites are major strengths of our study.

In conclusion, adolescents who have undergone MBS experience an increased risk of abdominal pain, bloating, constipation, and GERS at 8-years postoperatively, with patients undergoing VSG encountering a disproportionately higher risk of GERS compared to RYGB. Our results reinforce our 5-year study findings and highlight the importance of appropriate patient selection and counseling regarding postoperative risks. This includes preferential consideration of RYGB in select patients with preexisting severe GERS where this concern has been identified as a priority during the preoperative shared decision-making process. While the data regarding Barrett esophagus are unclear, given its theoretically increased risk following MBS, routine monitoring for pathologic consequences of reflux after MBS is warranted and should serve as an imperative for further research to aid in the establishment of clear, evidence-based clinical guidelines for pre- and postoperative management of pediatric patients undergoing MBS.