Comparison of Esophageal Dysmotility and Reflux Burden in Patients with Different Metabolic Obesity Phenotypes Based on High-Resolution Impedance Manometry and 24-h Impedance-pH

Gastroesophageal reflux disease (GERD) is associated with regurgitation and persistent, distressing heartburn resulting from reflux of the contents of the stomach into the esophagus. In addition, the distinct complications of GERD include reflux esophagitis (RE), peptic stricture, and Barrett’s esophagus (BE) [1]. GERD can reduce quality of life and increase healthcare costs [2]. Thus, it is crucial to understand the causes of GERD and to adopt precautionary measures to lower the chance of developing this condition.

Obesity, which poses a myriad of challenges to public health throughout the world, is a major contributing factor to GERD. Moreover, earlier research showed a link between increased body mass index (BMI) and the incidence of GERD symptoms, RE, nonerosive reflux disease (NERD), and hiatal hernia (HH) [3]. This is most likely because acid exposure is linearly and strongly correlated with BMI, with an increased acid exposure time (AET) per unit rise in BMI. Furthermore, a higher percentage of esophageal motility disorders have been linked to patients with obesity. These disorders are associated with a greater possibility of abnormal morphology of the esophagogastric junction (EGJ), an increased gastroesophageal pressure gradient, abnormal esophageal function and contractility, and other high-resolution manometry (HRM) findings that create the ideal conditions for reflux to develop [4]. Burgerhart et al. [5] found that the link between BMI and esophageal acid is largely due to pressure; the combined effect contributes to the incidence of GERD. Obesity is inherently linked to a cluster of metabolic abnormalities, including dyslipidemia, hypertension, and hyperglycemia [6]. GERD, reflux hypersensitivity (RH), and functional heartburn (FH) are among the gastrointestinal disorders for which metabolic syndrome has been found to be a risk factor [7]. However, patients with obesity have distinct metabolic phenotypes, and the effect of obesity on GERD is not yet fully understood. Individuals with obesity but who do not show signs of metabolic abnormalities have metabolically healthy obesity (MHO) [8].

Although the link between obesity and GERD has received much research attention, there are limited data on esophageal motility based on HRM and esophageal reflux indicators based on 24-h multichannel intraluminal impedance-pH (24-h MII-pH) monitoring in individuals with MHO. In the present study, we utilized endoscopy, ambulatory 24-h MII-pH monitoring, and esophageal function testing to gain valuable insights for clinical prevention and intervention strategies regarding GERD and ineffective esophageal motility (IEM) in individuals with different metabolic obesity phenotypes.

From January 1, 2018, to December 31, 2022, data were collected from 712 patients at the First Affiliated Hospital of Dalian Medical University, who had typical reflux symptoms for at least 3 months and underwent 24-h MII-pH monitoring, HRM, and upper endoscopy. The exclusion criteria were underweight (BMI <18.5 kg/m²; n = 38); missing information on metabolic syndrome, weight, and height (n = 72); severe cardiopulmonary diseases and coagulation disorders (n = 10); and a history of gastric surgery (n = 25). In the final stage, 567 individuals were selected for the examination. They were categorized into four groups: 176 patients with metabolically healthy non-obesity (MHNO), 137 patients with the MHO phenotype, 122 patients with the metabolically unhealthy non-obesity (MUNO) phenotype, and 132 patients with the metabolically unhealthy obesity (MUO) phenotype (Fig. 1).

Data on demographic characteristics, biochemical parameters, personal medical history, and medication use were recorded by trained nurses according to standardized methods. Demographic characteristics included age, body weight, and height. Biochemical parameters included high-density lipoprotein-cholesterol, triglycerides, and fasting blood glucose. In addition, a history of diabetes mellitus, hypertension, and surgery were recorded.

An upper endoscopy examination was performed on the patients before esophageal HRM and 24-h MII-pH monitoring. It was performed with a GIF-H260 or GIF-HQ260 endoscope (Olympus, Tokyo, Japan) following international standards. HH was detected with upper endoscopy. BE necessitated the presence of salmon-colored mucosa based on endoscopy as well as goblet cells and intestinal metaplasia based on histology [9]. The reliability of severity was classified using the Los Angeles categorization [10].

An HRM system (Medtronic Inc, Minneapolis, MN, USA) was used to record manometry parameters. Two separate, skilled physicians used proprietary software (ManoView, Medtronic) to manually assess the findings. The Chicago Classification Version 4.0 (CCv4.0) diagnostic criteria were used to evaluate the metrics [11]. Among the parameters acquired were the lower esophageal sphincter pressure (LESP), the mean distal contractile integral (DCI), the median integrated relaxation pressure (IRP), the number of ineffective swallows, and the type of esophageal peristalsis. When the EGJ contractile integral (EGJ-CI) was <25 mm Hg/cm, it was deemed abnormally low [11, 12]. The DCI was employed to measure the strength of esophageal body contractions. Ineffective swallows included: weak, with a DCI between 100 and <450 mm Hg/cm/s; failed, with a DCI <100 mm Hg/cm/s; and fragmented, defined as a peristaltic transition zone defect of >5 cm with an isobaric contour of 20 mm Hg, in the presence of a DCI ≥450 mm Hg/cm/s. For the identification of IEM according to the CCv4.0 criteria, at least 70% of swallows need to be ineffective, or a minimum of 50% of swallows fail with a DCI measurement of <100 mm Hg/cm/s [11].

A 24-h MII-pH monitoring device (Sandhill Scientific Inc., Highland Ranch, CO, USA) was used to record pH-impedance parameters, namely (1) the total, upright, and supine AET; (2) the total number of reflux episode (TRs); (3) the symptom-association probability (SAP; positive if ≥95%) and the symptom index (SI) (positive if ≥50%); (4) the mean nocturnal baseline impedance (MNBI); and (5) the post-reflux swallow-induced peristaltic wave index (PSPWI). An antegrade 50% impedance drop that starts in the most proximal impedance channel, travels to the most distal impedance channel, and of which at least 50% returns to the baseline within 30 s of a reflux event was deemed to be a PSPW [13]. The PSPWI was determined as the TRs divided by the number of PSPWs [14]. During the night, the MNBI was measured through the most distal impedance channel. To determine the MNBI, three 10-min intervals (at approximately 01:00, 02:00, and 03:00 h) were selected and the mean was computed; times involving swallows, refluxes, and pH drops were eliminated [14]. The cut-off values for the PSPWI, the MNBI, and the TRs were 50%, 1,500, and 40, respectively [15, 16].

NERD was characterized by endoscopy showing no mucosal pathology but an AET >6%. For RH, endoscopy revealed no mucosal pathology, and patients had AET ≤6% but a positive SI/SAP. FH was characterized by endoscopy showing no mucosal pathology, AET <4%, and the absence of positive SI/SAP [17]. The pH-impedance investigations provided the recorded data, which were subsequently extracted and analyzed according to the Lyon Consensus thresholds [15]. Pathologic cases were defined as having a total AET >6%, while physiologic cases had values <4%. Values falling between 4% and 6% were considered indeterminate. Additionally, for upright and supine AET, values >6% and 2%, respectively, were categorized as pathologic.

BMI was computed as kg/m². Obesity was assessed according to the World Health Organization (WHO) criteria for East Asians: a BMI of ≥25 kg/m² is considered obese [18]. The metabolic status was evaluated using the Adult Treatment Panel III criteria [19]. Being metabolically healthy was characterized as having fewer than two of the following criteria: (1) systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥85 mm Hg; (2) fasting blood glucose ≥100 mg/dL; (3) high-density lipoprotein-cholesterol <40 mg/dL for men and <50 mg/dL for women; and (4) triglyceride ≥150 mg/dL. Finally, the patients were categorized into four metabolic obesity phenotypes: (1) MHNO, BMI <25 kg/m², and fewer than two components of metabolic syndrome; (2) MHO, BMI >25 kg/m² and fewer than two components of metabolic syndrome; (3) MUNO, BMI <25 kg/m², and two or more components of metabolic syndrome; (4) MUO, BMI ≥25 kg/m², and two or more components of metabolic syndrome.

SPSS Statistics 26.0 (IBM, Armonk, NY, USA) was used for all data analysis. The Shapiro-Wilk test was employed to determine whether the continuous variables met the normality assumption; they are presented as the mean ± standard deviation or median (interquartile range). Categorical variables are presented as frequencies and percentages. One-way analysis of variance followed by the Tukey or the Kruskal-Wallis H test. Categorical variables were assessed by χ² tests. Logistic regression was employed to investigate the relationships between the different phenotypes and pathologic AET and the prevalence of IEM. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for the MHO, MUNO, and MUO groups, with the MHNO group as the reference category. A p value <0.05 (two-tailed p value) was deemed statistically significant.

We included a total of 567 patients, of whom 176 had the MHNO phenotype, 137 had the MHO phenotype, 122 had the MUNO phenotype, and 132 had the MUO phenotype. In the endoscopy-positive group, there were more patients with the MHO, MUO, or MUO phenotype and RE, BE, or HH compared with patients with the MHNO phenotype and RE, BE, or HH (p < 0.05). Similarly, the severity of RE differed considerably across the four phenotypes (p < 0.05). In the endoscopy-negative group, there were more patients with the MUO or MHO phenotype and NERD than patients with the MHNO or MUNO phenotype and NERD (p < 0.001). Furthermore, there were more patients with the MHNO phenotype and RH or FH compared with patients with the MHO, MUNO, or MUO phenotype and RH or FH (p < 0.05) (Table 1).

The mean DCI and the number of failed, weak, and fragmented swallows were measured similarly for all patients. Compared to people with the MUNO and MHNO groups, the MHO and MUO groups had a substantially higher median IRP, number of ineffective swallows, and prevalence of IEM, and lower ECJ-CI and LESP (Table 2; p < 0.05).

Compared with the MHNO and MUNO groups, the MHO and MUO groups showed substantially higher supine, upright, and total AET; more TRs; and a lower MNBI and PSPWI (Table 3; p < 0.05). SI >50% and SAP >95% were similar among the four groups.

Figure 2a illustrates the distribution of impedance parameters across 567 subjects, categorized by AET levels: pathologic (>6%), indeterminate (4%–6%), and physiologic (<4%). There were significant differences among the metabolic obesity phenotypes across the three AET levels (p < 0.05). Figure 2b categorizes the same subjects based on their supine (>2%), upright (>6%), and total AET (>6%). There were significant differences across all phenotypes (all p_s < 0.001).

Next, we evaluated the adjunctive 24-h MII-pH evidence (Fig. 3a). The patients with the MUO and MHO phenotypes demonstrated substantially elevated TRs and a decreased MNBI and PSPWI compared with the patients with the MUNO and MHNO phenotypes (p < 0.001). Figure 3b shows the adjunctive HRM evidence. The patients with the MUO and MHO phenotypes displayed a substantial reduction in ECJ-CI as well as a higher prevalence of IEM and HH compared with patients with the MUNO and MHNO phenotypes (p < 0.05).

Logistic regression for AET risk showed that the MUO, MUNO, and MHO phenotypes were linked to a higher possibility of developing pathologic AET (Table 4; p < 0.001). After adjusting for sex, age, and BMI (model 3), the adjusted OR (95% CI) for the prevalence of pathologic AET compared with the MHNO phenotype was 15.614 (5.314–45.879) for the MHO phenotype, 6.402 (2.093–19.583) for the MUNO phenotype, and 19.086 (6.170–59.044) for the MUO phenotype.

Logistic regression for IEM risk indicated that the MUO, MUNO, and MHO phenotypes had a greater risk of IEM (Table 5; p < 0.05). After adjusting for BMI, age, and sex (model 3), the adjusted OR (95% CI) for IEM risk compared with the MHNO phenotype was 2.761 (1.373–5.551) for the MHO phenotype, 2.077 (1.038–4.156) for the MUNO phenotype, and 3.659 (1.647–8.130) for the MUO phenotype.

Our study represents the first known evaluation of a substantial group of individuals with various metabolic obesity phenotypes that has utilized HRM and 24-h MII-pH monitoring. These findings confirm the need to preserve a healthy weight to prevent the development of GERD and IEM: Independent of the metabolic health status, people with obesity have a higher risk of esophageal dysmotility and reflux burden than those without obesity.

Previous research has shown a significant association between obesity and GERD. Many processes have been suggested to explain this relationship, although the precise mechanisms that relate fat and GERD are not yet known. Individuals with obesity experience a reduction in LESP, which impairs the anti-reflux barrier and subsequently leads to the development of gastroesophageal reflux. This phenomenon may be associated with heightened intra-abdominal pressure [20]. Saliva secretion, gravity, and esophageal motility collectively determine the esophageal clearance rate. Obesity often results in reduced saliva secretion and impaired esophageal motility, compromising esophageal clearance [21]. Ortiz et al. [22] found that individuals with obesity demonstrate reduced esophageal sensitivity to acid perfusion, potentially affecting esophageal clearance. Cytokine-mediated esophageal inflammation has recently been suggested as the process underlying the etiology of GERD [23]. Visceral adipose tissue functions as a significant depot of adipocyte-derived factors, releasing cytokines such as tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), IL-1, leptin, adiponectin, and other molecules. These mediators can induce systemic effects, influencing and amplifying systemic inflammatory responses [24]. These findings indicate that obesity alone may serve as a significant risk factor for GERD. We found that compared with patients with the MHNO or MUNO phenotype, patients with the MUO or MHO phenotype had a higher incidence of RE, BE, or NERD, and that various metabolic obesity phenotypes were associated with the severity of RE. However, there were more patients with the MUNO or MHNO phenotype and RH or FH compared with patients with the MHO or MUO phenotype and RH or FH. These results could be attributed to the abnormal acid exposure experienced by individuals with obesity. In contrast, the pathogenesis of RE, BE, and NERD primarily revolves around exposure to abnormal acid levels. Nevertheless, RH and FH may not be directly associated with abnormal acid exposure; they are more likely associated with psychiatric factors [25].

The high prevalence of manometric anomalies in patients with the MHO phenotype included changes in baseline LESP, difficulties with LES relaxation, and abnormalities in esophageal body contractility. Furthermore, as BMI increases, so do the chances of esophageal reflux symptoms and aberrant reflux measurements [26]. Similarly, we found an increased percentage of pathologic AET and IEM in patients with the MHO or MUO phenotype. These findings indicate that aberrant AET and esophageal dysmotility may be the main causes of the higher burden of esophageal symptoms in individuals with obesity. Thus, the determination of objective esophageal indicators could enhance the diagnostic workup of patients with symptomatic MHO. Consistent with our findings, researchers have found that metabolic disorders are significantly associated with GERD, although the mechanism underlying this association is unclear [7, 27]. Hypertension leads to visceral nerve sensitivity, which in turn leads to increased esophageal visceral nerve sensitivity, which in turn leads to reflux symptoms [28]. GERD seems to emerge from peripheral neuronal abnormalities due to hyperglycemia, which exacerbate high gastroesophageal pressure gradients and enhance transitory sphincter relaxation [29]. Moreover, elevated lipid levels can impair esophageal clearance and weaken the LES, ultimately contributing to the development of RE [30]. As shown previously, the risk of GERD is increased by dyslipidemia, hypertension, and hyperglycemia. Thus, our findings emphasize the necessity of considering metabolic abnormalities independent of the obesity status. There has been a lack of research to confirm esophageal motility and esophageal reflux function employing objective testing in individuals with metabolic disorders.

HRM indicated various alterations in esophageal function among the four metabolic obesity phenotypes. According to Braghetto et al. [31], a lower LESP was more typical in patients with obesity and was linked, based on 24-h pH monitoring, to increased erosive esophagitis and pathogenic gastroesophageal reflux. Our findings align with this observation: Patients with the MUO or MHO phenotype demonstrated substantially and significantly lower baseline LESP compared with those with the MUNO or MHNO phenotype. The cause could be attributed to the increased occurrence of temporary lower esophageal sphincter relaxation in individuals with the MHO phenotype. The novel EGJ-CI parameter provides a thorough representation of the barrier function. The EGJ-CI is negatively linked to exposure to acid and the overall reflux episode [32]. We demonstrated that patients with the MUO phenotype had lower EGJ-CI compared with those with the MHNO phenotype. Moreover, the incidence of hypotensive ECJ-CI was greater for those with the MUO or MHO phenotype compared with those the MHNO or MUNO phenotypes. The results of our study confirm previous studies that obesity has a notable effect on the functioning of the esophagus. Specifically, it causes changes in the morphology of the EGJ, leading to a higher burden of acid reflux [33]. In addition, the MUO phenotype is the most common in individuals with HH [15].

According to the CCv4.0 criteria, the median IRP has a significant impact on defining esophageal motility disorders. We found that patients with the MUO or MHO phenotype demonstrated a considerably greater median IRP compared with those with the MUNO or MHNO phenotype. According to Yen et al. [34], patients with obesity have much higher LES IRP-4s than healthy participants. More cases of esophageal dysmotility disorders should arise from the greater LES IRP-4s in people with obesity, which should offset the higher DCI. In our cohort, the mean DCI was not appreciably greater, which makes sense given the small number of patients with morbid obesity in our study. We showed that the prevalence of IEM varied between patients with the MHO and MUNO phenotypes; it was more common in patients with the MHO phenotype. Nonetheless, the IEM prevalence was markedly higher in patients with the MHO or MUNO phenotype compared with patients with the MHNO phenotype, indicating that obesity, apart from metabolically unfavorable phenotypes, was the primary risk factor for IEM. The risk of IEM was increased by both obesity and metabolic abnormalities, and the incidence of IEM was greatest in patients with the MUO phenotype. Hypercontractility has been linked to an elevated esophageal acid dwell duration and, thus, to a higher risk of esophageal mucosal damage [35]. Comparably, IEM increases the risk of an elevated AET by causing a lower contractile vigor and a larger acid burden [36]. Thus, excessive weight could lead to esophageal dysmotility.

Patients with the MHO or MUO phenotype showed substantially higher levels of all reflux-related markers based on 24-h MII-pH compared with patients with the MHNO or MUNO phenotype. In contrast, pathologic reflux was more likely to coexist with obesity. We did not find an association between obesity and metabolic diseases: More than half of our patients were SAP/SI positive, and the rate of positivity did not vary across the metabolic obesity phenotypes. Usually, being SI or SAP positive indicates the strongest correlation between reflux symptoms [37]. Nevertheless, SAP and the SI are excessively reliant on the accuracy of patient recordings, and in many instances, symptoms are not documented during reflux monitoring. Further evaluation of relevant and objective parameters is necessary to develop a standard diagnosis of GERD in patient with various metabolic obesity phenotypes. The diagnostic efficacy of TRs in patients with typical reflux symptoms and the MHO phenotype has not been exhaustively examined. The TR physiologic threshold is the new benchmark [15]. TR was positive and found in more patients with the MHO or MUO phenotypes than in patients with the MHNO or MUNO phenotype, indicating that obesity can lead to more reflux episodes. Therefore, typical reflux symptoms should be noted in patients with the MHO phenotype to avoid the development of GERD.

Esophageal chemical clearing, or the PSPWI, is a primary defense mechanism against reflux [38], which is triggered by the salivary reflex and is directly linked to the harmfulness of refluxate [39]. In research using MII-pH monitoring to identify inconsistent GERD individuals with AET 4%–6%, PSPWI had an area under the receiver operating characteristic curve of 0.839 to detect reflux [40]. The PSPWI demonstrated overall accuracy and greater sensitivity for the diagnosis of GERD compared with the use of traditional MII-pH parameters, including AET, TRs, and the longest reflux event [14]. In our cohort, patients with the MUO or MHO phenotype had a PSPWI that was considerably lower than patients with the MHNO or MUNO phenotype. Patients with obesity often exhibit reduced esophageal clearance compared with individuals without obesity. This may explain the susceptibility to GERD in patients with obesity.

Baseline impedance indicates the integrity of the esophagus mucosa. A reduced MNBI can reflect increasing GERD severity [14]. We found a lower MNBI in patients with the MHO, MUNO, or MUO phenotype compared with patients with the MHNO phenotype. Furthermore, Blevin et al. [41] suggested that obesity modifies the function of the esophageal barrier and is linked it to an aberrant esophageal MNBI. According to the Lyon Consensus 2.0, an MNBI threshold of <1,500 Ω is abnormal. In our study, 78.1% of patients with the MHO phenotype had an abnormal MNBI, which suggests that these patients had more pronounced esophageal mucosal impairment. The precise mechanism underlying this impairment in the esophageal mucosal in the context of obesity remains unclear. According to Gibbens et al. [42], central obesity increases intercellular space, decreases desmosome density, and increases fluorescein leakage, all of which compromise the anatomical and functional integrity of the esophageal barrier.

There are several limitations to our study. Initially, it was conducted at a single medical facility with analysis of previous data, making it an observational study. It is important to note that such studies may be susceptible to inherent bias. Subsequent investigations with increased sample sizes in each subgroup could assist in confirming the validity of these findings. Second, we did not measure some potentially confounding factors, such as dietary patterns, psychosocial stress, medication usage (opiates and other analgesics), and the socioeconomic status, which may induce a potential bias. Finally, we evaluated heartburn or regurgitation as it is the primary symptom that confirms the presence of GERD. However, it is still necessary to determine the relevance of these results in patients experiencing extraesophageal reflux symptom.

We found that patients with the MHO phenotype are at higher risk of GERD and IEM than patients with the MUNO phenotype, suggesting that obesity plays a crucial role in developing GERD and IEM. Given these considerations, we must pay attention to symptoms to avoid the development of GERD and IEM in patients with the MHO phenotype.

We thank the study participants for their contribution.

This study was approved by the Ethics Committee of Institutional Review Board of the First Affiliated Hospital of Dalian Medical University (PJ-KS-KY-2020-04). Written informed consent was obtained from all participants.

The authors declare that there is no conflict of interest.

This study was not supported by any sponsor or funder.

T.H. was primarily responsible for the data analysis and writing of the manuscript. T.H. and Z.J.D. significantly revised the draft, interpreted the data, and involved in data analyses. M.J.Z. and M.H.T. collected the information and participated in data interpretation. All authors read and approved the final manuscript.

The data that support the findings of this study are not publicly available due to reason why their containing information that could compromise the privacy of research participants but are available from the corresponding author (Z.J.D.) or Data Sharing Committee (email address: cathydoctor@sina.com) upon reasonable request.