Obesity Surgery

, Volume 18, Issue 11, pp 1479–1484

Gastroesophageal Reflux in Patients with Morbid Obesity: A Role of Obstructive Sleep Apnea Syndrome?


  • J. M. Sabaté
    • Assistance Publique - Hôpitaux de Paris (AP-HP), Hôpital Louis Mourier, Service d’Hépato-GastroentérologieUniversité Paris 7
  • P. Jouët
    • Assistance Publique - Hôpitaux de Paris (AP-HP), Hôpital Louis Mourier, Service d’Hépato-GastroentérologieUniversité Paris 7
  • M. Merrouche
    • Assistance Publique - Hôpitaux de Paris (AP-HP), Hôpital Louis Mourier, Service d’Hépato-GastroentérologieUniversité Paris 7
  • J. Pouzoulet
    • Assistance Publique - Hôpitaux de Paris (AP-HP), Hôpital Louis Mourier, Service d’Hépato-GastroentérologieUniversité Paris 7
  • D. Maillard
    • Assistance Publique - Hôpitaux de Paris (AP-HP), Hôpital Louis Mourier, Service des Explorations FonctionnellesUniversité Paris 7
  • F. Harnois
    • Assistance Publique - Hôpitaux de Paris (AP-HP), Hôpital Louis Mourier, Service d’Hépato-GastroentérologieUniversité Paris 7
  • S. Msika
    • Assistance Publique - Hôpitaux de Paris (AP-HP), Hôpital Louis Mourier, Service de Chirurgie viscéraleUniversité Paris 7
    • Assistance Publique - Hôpitaux de Paris (AP-HP), Hôpital Louis Mourier, Service d’Hépato-GastroentérologieUniversité Paris 7
Research Article

DOI: 10.1007/s11695-008-9508-9

Cite this article as:
Sabaté, J.M., Jouët, P., Merrouche, M. et al. OBES SURG (2008) 18: 1479. doi:10.1007/s11695-008-9508-9



Obesity is a risk factor for gastroesophageal reflux disease (GERD) and for obstructive sleep apnea (OSA). Our aim was to evaluate in morbidly obese patients the prevalence of OSA and GERD and their possible relationship.


Morbidly obese patients [body mass index (BMI) >40 or >35 kg/m2 in association with comorbidities] selected for bariatric surgery were prospectively included. Every patient underwent a 24-h pH monitoring, esophageal manometry, and nocturnal polysomnographic recording.


Sixty-eight patients [59 women and 9 men, age 39.1 ± 11.1 years; BMI 46.5 ± 6.4 kg/m2 (mean ± SD)] were included. Fifty-six percent of patients had an abnormal Demester score, 44% had abnormal time spent at pH <4, and 80.9% had OSA [apnea hypopnea index (AHI) >10] and 39.7% had both conditions. The lower esophageal sphincter (LES) pressure was lower in patients with GERD (11.6 ± 3.4 vs 13.4 ± 3.6 mm Hg, respectively; P = 0.039). There was a relationship between AHI and BMI (r = 0.337; P = 0.005). Patients with OSA were older (40.5 ± 10.9 vs 33.5 ± 10.4 years; P = 0.039). GERD tended to be more frequent in patients with OSA (49.1% vs 23.1%, respectively; P = 0.089). There was no significant relationship between pH-metric data and AHI in either the 24-h total recording time or the nocturnal recording time. In multivariate analysis, GERD was significantly associated with a low LES pressure (P = 0.031) and with OSA (P = 0.045) but not with gender, age, and BMI.


In this population of morbidly obese patients, OSA and GERD were frequent, associated in about 40% of patients. GERD was significantly associated with LES hypotonia and OSA independently of BMI.


Morbid obesityGERDOSAEsophagusManometry24-h pH monitoring



body mass index


obstructive sleep apnea


gastroesophageal reflux disease


apnea hypopnea index


lower esophageal sphincter


waist-to-hip ratio


Gastroesophageal reflux disease (GERD) is an important and increasingly common condition, with a prevalence as high as 20% in the USA and 8% to 30% in France depending on the frequency of symptoms used for the definition [1, 2]. In these countries, the prevalence of obesity has been rising to epidemic proportions and is, in France, about 8–10% of the adult population [3]. Obesity is considered as a risk factor for GERD: a direct relationship has been found between increasing body mass index (BMI) and reflux symptoms of GERD [4] and between high BMI and the severity of GERD evaluated by the DeMeester score [5]. In a previous study performed in morbidly obese patients, we reported that about half of the patients had GERD defined by an abnormal pH-metric result [6].

Obstructive sleep apnea (OSA) is a common disorder characterized by repetitive pharyngeal collapse during sleep. The most common features of OSA are frequent snoring, periods of apnea during sleep, and excessive daytime somnolence. Its prevalence in a middle-aged North American population is up to 5% [7]. Obesity is the most important reversible risk factor for OSA, as 40% to 90% of obese subjects are affected by OSA [8, 9]. Other risk factors for OSA include older age and male gender. In the literature, different studies performed in non-morbidly obese populations suggested a link between GERD and OSA [10, 11]. While GERD and OSA are more frequent in morbidly obese patients, only few studies have assessed both conditions in this population [12].

The aim of the present study was to determine, in morbidly obese patients selected for bariatric surgery, the respective prevalence of GERD defined by an abnormal pH-metric result and of OSA and to search for a possible relationship.

Patients and Methods

Study Design

Consecutive patients selected for bariatric surgery between April 2003 and January 2005 at our institution were prospectively included in this study. Patients were considered for surgery if they had a BMI ≥40 kg/m2 or ≥35 kg/m2 and at least one severe obesity-related medical complication (e.g., type 2 diabetes mellitus, hypertension, cardiovascular disease) and after an unsuccessful hypocaloric regimen. Every obese patient of this study had a prospective evaluation of symptoms suggesting GERD, such as heartburn, epigastric pain, regurgitations, dysphagia, and extra esophageal manifestations (chronic pharyngitis or asthma), and had an upper gastrointestinal endoscopy, a 24-h pH-metry, and an esophageal manometry performed before surgery. Polygraphic recording was also performed systematically before surgery to determine if OSA was present.

Patients and Preoperative Assessment

The following data were recorded before bariatric surgery: age, gender, anthropometrical measurements (weight, height, waist and hip circumferences). Increased waist-to-hip ratio (WHR) reflecting central obesity was defined as ≥0.90 in men and ≥0.85 in women [13]. Exclusion criteria were as follows: history of compulsive eating disorder, previous bariatric surgery, or a severe cardiopulmonary chronic disease. Informed written consent was obtained from every patient and the study was conducted in accordance with the Helsinki Declaration.


Esophagogastroduodenoscopy was performed using a standard videogastroscope to exclude esophageal or gastric lesions and also to describe the cardial region which could be normal or with the presence of hiatus hernia. Severity of esophagitis was scored according to the Los Angeles classification [14].

Twenty-four-hour pH-metry

Ambulatory 24-h pH testing was performed with the patients off all peptic medication for at least 7 days. The pH electrode (internal diameter 2.3 mm, Medtronic, Paris, France) was positioned 5 cm above the upper border of the lower esophageal sphincter (LES). Data were recorded in a portable digital data logger (Digitrapper, Mark III Gold, Synectics, Paris, France) for a period of 24 h. We measured the eosophageal acid exposure defined by percent of time spent at pH <4 (abnormal if superior to 4.2% of the total recording). The duration of longest reflux episode, the total number of reflux, and the number of refluxes longer than 5 min were also recorded. A Johnson–DeMeester score [15] was calculated using a standard software program (Synectics) (abnormal if score >14.72).

Esophageal Manometry

Esophageal manometry was performed with a four-channel water-perfused catheter (5 mm diameter, Marquat SA, Boissy, France), with the distance between the transducers at 5 cm. The LES was located using the stationary pull-through technique and the resting LES pressure (normal range: 18 ± 5 mm Hg), LES relaxation, and esophageal body contractility were determined for a minimum of 10 wet swallows. Normal peristaltic waves were considered if their morphology, amplitude (normal range: 116 ± 25 mm Hg), duration (normal range: 3.4 ± 0.5 s), and propagation were normal.

Polygraphic Recordings

Ambulatory polygraphic monitoring was performed by a CID 102 device (CIDELEC, Angers, France). Measurements included nasal airflow (nasal cannulae), rib cage and abdominal wall motions (respiratory inductance plethysmography), and pulse oxygen saturation via a finger probe. All data were analyzed visually. Apnea was defined as cessation of airflow for more than 10 s. Hypopnea was defined as a reduction of airflow, regardless of its amplitude, with a more than 3% fall in pulse oxygen saturation. The number of apnea and hypopnea episodes per hour of monitoring, or apnea hypopnea index (AHI), was calculated. OSA was defined as absent when AHI was less than 10/h [16], moderate when it was 10 to 50/h, and severe when it was above 50/h.

Statistical Analysis

Qualitative variables were expressed as a percentage and were compared using the Chi-square and the Fisher exact test as appropriate. Continuous variables were expressed as the mean ± SD and were compared using the Student t test. Correlations between variables were assessed using Spearman coefficient correlations. Variables with a P value less than 0.2 in the univariate analysis were included in the model for multivariate analysis. Multivariate analysis was performed using binary logistic regression. The risk factors age, sex ratio, BMI, LES pressure, and the presence of OSA were taken into account in multivariate analysis for abnormal esophageal acid exposure. Differences were considered significant if the P value was less than 0.05. The analysis was performed with SPSS computer software (version 14.0; SPSS, Chicago, IL, USA).


Characteristics of Patients

Sixty-eight consecutive morbidly obese patients (59 females/9 males) selected for bariatric surgery were included in the study. All but two patients were of Caucasian origin. The mean age was 39.1 ± 11.1 years and mean BMI was 46.5 ± 6.4 kg/m2 (range: 38–65 kg/m2). For females, the mean waist circumference was 122.5 ± 12.6 cm and the mean WHR was 0.87 ± 0.10; for males, the mean waist circumference was 136.5 ± 12.9 cm and the mean WHR was 1.01 ± 0.08. Central obesity was present in 89% of males and 53% of females, 13.2% (9/68) had type 2 diabetes mellitus.


Heartburn was present in 44.1% (30/68), regurgitations in 26.5% (18/68), and epigastric pain in 26.5% (18/68) of patients. Clinical GERD, defined by the presence of either heartburn or regurgitation, was present in 50% (34/68) of patients. Extraesophageal manifestations were found in 20.6% (14/68) of obese patients: 11.8% (8/68) had chronic pharyngitis and 11.8% (8/68) had asthma.


Endoscopy was performed in all patients and showed a hiatus hernia in 36.8% (25/68) patients and esophagitis in 10.3% (7/68) patients (four patients had grade A esophagitis and three had grade B; no patient had grade C or D esophagitis) [14].


A 24-h pH ambulatory recording was completed in every patient. The time spent at pH <4 was pathological in 30 patients (44%). In these patients, the mean percentage of time with pH below 4 was 7.7 ± 3.0% and the mean longest reflux was 30.9 ± 21.3 min; 42.6% (29/68) of the patients had more than 50 reflux episodes per day, and 22% (15/68) had more than 3 refluxes longer than 5 min. The percentage of patients with an abnormal DeMeester score was 55.9% (38/68), and the global mean score was 18.5 ± 13.1. The mean DeMeester score was higher in patients with than in those without GERD symptoms (22.8 ± 15.7 vs 14.1 ± 7.9; P = 0.006).

Esophageal Manometry

Esophageal manometry was performed in every patient. The mean resting LES pressure was 12.6 ± 3.7 mm Hg (range: 7–21 mm Hg); 67.6% of the patients (46/68) had incompetent LES defined by a LES pressure under the normal range (LES tone <15 mm Hg). All patients had normal LES relaxation. The amplitude of esophageal contraction waves was normal (superior to 70 mm Hg) in 25.6% of the patients. Three patients (4.4%) had esophageal dyskinesia. The LES pressure was significantly lower in patients with abnormal time spent at pH <4 compared with those with normal time spent at pH <4 (11.6 ± 3.4 vs 13.4 ± 3.6 mm Hg, respectively; P = 0.039).

Characteristics of Patients According to Polygraphic Results

Every patient had polygraphic recordings. The mean AHI was 32.7 ± 26.3/h (range: 0–131/h). OSA (AHI >10/h) was found in 55 (80.9%) patients, moderate in 43 (63.3%) patients and severe (AHI >50/h) in 12 (17.6%). The characteristics of patients according to polygraphic results are shown in Table 1. Patients with OSA were significantly older than patients without OSA. BMI was significantly correlated with AHI (r = 0.337; P = 0.005), whether patients had abnormal (P = 0.019) or normal (P = 0.015) pH-metric recording. No difference was found for the other variables. No correlation was found between AHI and WHR (P > 0.5). Twenty-seven patients (39.7% of the whole group) had both OSA and GERD. Mean AHI was not different in patients with or without clinical GERD (33.0 ± 25.1 vs 32.4 ± 28.0, respectively, P = 0.92) or in patients with or without esophagitis (34.9 ± 15.2 vs 32.5 ± 27.5, respectively, P = 0.82).
Table 1

Clinical characteristics of patients according to the presence or absence of OSA defined by AHI < or ≥10


OSA (n = 55)

No OSA (n = 13)

P value

Mean age (years)

40.5 ± 10.9

33.5 ± 10.4


Gender (% of women)




Mean BMI (kg/m2)

47.1 ± 6.6

44.4 ± 5.1


Presence of hiatal hernia (% of patients)




Mean DeMeester score (n < 14.72)

19.1 ± 14.1

15.6 ± 7.5


Mean percent of time spent at pH <4

4.9 ± 3.7

3.8 ± 1.9


Mean number of refluxes

50.4 ± 33.2

49.1 ± 17.9


Mean longest reflux (min)

16.1 ± 19.1

9.9 ± 6.5


Mean number of refluxes >5 min

2.2 ± 2.4

2.0 ± 1.6


Mean percent of nocturnal time spent at pH <4

3.5 ± 4.7

2.9 ± 2.8


Mean number of nocturnal refluxes

8.6 ± 9.4

11.0 ± 5.6


Mean longest nocturnal reflux (min)

11.3 ± 16.6

7.9 ± 8.1


Mean number of nocturnal refluxes >5 min

0.8 ± 1.7

0.9 ± 1.1


Mean LES pressure (mm Hg)

12.6 ± 3.4

12.9 ± 4.8


Normal contraction amplitude (% of patients)




Normal esophageal propagation (% of patients)




at Test

bChi-square test with Yates’ correction

Risk Factors for Abnormal pH-metric Recording

In univariate analysis, the percentage of abnormal pH-metric recording tended to be higher in patients with OSA compared with patients without OSA (49.1% vs 23.1%, respectively; P = 0.089). AHI was not correlated with the percentage of time spent at pH <4 (r = 0.173; P = 0.476), the total number of refluxes (r = −0.179; P = 0.144), the percentage of nighttime spent at pH <4 (r = −0.088; P = 0.476), or with the number of nocturnal refluxes (r = −0.32; P = 0.287). Results of multivariate analyses are presented in Table 2. The presence of abnormal pH-metric recording was associated with a lower LES pressure (P = 0.031) and with the presence of OSA (P = 0.045), but not with gender, age, and BMI.
Table 2

Multivariate analysis (binary logistic regression) for pathological time spent at pH <4


Odds ratio

CI 95%

P value









Male gender




LES pressure




Presence of OSA





In this study of morbidly obese patients waiting for bariatric surgery, GERD and OSA were highly frequent, with 40% of patients having both conditions. We found that OSA was significantly associated with abnormal esophageal acid exposure.

GERD defined by clinical symptoms and/or abnormal 24-h pH-metric data was present in this series in about half of the patients. This frequency is similar to that reported in literature [17] and in our previous study [6]. As was described [6, 17], there was a relationship between low LES pressure and abnormal pH-metric data.

In our study, 80% of the patients had OSA, a prevalence similar to that reported in the literature ranging between 70% and 80% of morbidly obese patients [9, 18, 19]. OSA was severe in 17% and moderate in 63% of the patients. As described before [18], we found that patients with OSA were older and had a higher BMI.

In univariate analysis, the percentage of patients with pH-metric GERD was twice as high as in presence of OSA (from 23 to 49%), although it did not reach the level of significance possibly because of a small number of patients included in this study. This association between GERD and OSA was confirmed in multivariate analysis independently of BMI, the only other factor associated with GERD being low LES pressure. The coexistence of OSA and GERD was initially described in children [20] and in the nonobese adult population [10]. However, in patients referred for sleep disorders, conflicting results about this association were found when reflux was defined only on the basis of symptom questionnaires [21, 22]. Small studies based on polygraphic recording and pH-metric recording, as in our study, showed a high prevalence of GERD in patients with OSA, ranging from 64% to 100% [23, 24].

Because obesity is a risk factor for both OSA and GERD, some authors argued for a simple coincidence [24, 25]. We found that about 40% of our morbidly obese patients had both conditions. Others argued for a causal relationship because the treatment of one condition improves the other condition, but the nature and the direction of the relationship between the two conditions are not yet fully established. Different mechanisms have been proposed to explain the association between GERD and OSA in obese patients. GERD can lead to worsening of airway reactivity by directly inducing airway inflammation or indirectly by vasovagal reflex arc from the esophagus to the lung, resulting in bronchoconstriction. This is supported by the fact that antireflux medications with proton pump inhibitors decreased weekly frequency of apnea attacks [11, 26]. On the other hand, the generation of a negative intrathoracic pressure by OSA during respiratory efforts could favor the passage of gastric content into the esophagus by decreasing mean lower esophageal resting pressure and LES pressure gradient [27]. This hypothesis is supported by the improvement of GERD after the administration of continuous positive airway pressure [27, 28]. However, such improvement was also found in patients without OSA [29], and in our study, there were no differences in mean LES pressure or percentage of patients with normal relaxation whether OSA was present or not. Some authors looked for a temporal link between OSA and GERD. In a controlled study with simultaneous pH-metric and polygraphic recording, Ing et al. found abnormal nocturnal pH-metric data with an increase in the percent of time spent at pH <4 in patients with OSA [29]. However, in this study, only about half of all refluxes were temporally related to apneas [29]. In our study, although pH-metric and polygraphic recordings were not performed simultaneously, nocturnal pH-metric data were not different in patients with or without OSA. Kerr et al. found that both nocturnal reflux and reflux in supine position were increased, suggesting that factors other than OSA could play a role to produce GERD [27].

In the absence of a clear temporal relationship between OSA and GERD, some authors looked for an association between severity of OSA and GERD. Kim et al. [22] did not find a correlation between GERD symptoms and OSA. Demeter et al. found a positive correlation between the occurrence of erosive reflux disease and AHI [30]. In our study, we did not find any relationship between OSA severity and GERD symptoms or endoscopic findings or between AHI and the percent of time spent at pH <4. However, a higher score of AHI does not necessarily reflect the magnitude of respiratory effort during obstruction [31].

Our results confirm high frequencies of GERD and OSA in morbidly obese patients, with both conditions being present simultaneously in 40% of them. For these reasons, morbidly obese patients should be systematically explored for GERD and OSA. The risk of long-term complications of these comorbidities is especially important in nonoperated patients, as bariatric surgery may improve both GERD and OSA [6, 32, 33]. In this study, we found that OSA was associated with abnormal esophageal acid exposure independently of BMI, but mechanisms linking these two conditions remain to be elucidated. To better understand the pathophysiology, we suggest that future studies should be performed with simultaneous pH-metric, manometric, and polygraphic recordings.

Copyright information

© Springer Science + Business Media, LLC 2008