What is the impact of metabolic syndrome and its components on reflux esophagitis? A cross-sectional study

Reflux esophagitis (RE) has become a major public health issue in Asia. A low prevalence rate of 3.3% was reported in an epidemiological report in Singapore by Kang et al. in 1993 [1], however, subsequent studies have reported a rapid increase in Asian populations, ranging from 4.3 to 37.8% [2, 3].

RE is defined as visible esophageal mucosal damage on endoscopy [4, 5]. RE results from the abnormal retrograde flow of gastric contents into the esophagus [5]. RE has been hypothesized to be a multifactorial process, including an abnormal frequency of gastroesophageal reflux, delayed esophageal clearance, delayed gastric emptying, and transient lower esophageal sphincter tone relaxation [6–8]. RE can cause serious complications if untreated for a prolonged period, including esophageal stricture, Barrett’s esophagus, and esophageal adenocarcinoma [4, 5].

Previous studies have reported positive associations between RE and the following associated factors: older age, male gender, smoking, alcohol consumption, and hiatal hernia, and a negative association between RE and Helicobacter pylori (H. pylori) infection [9–13]. Obesity is also known to play an important role in RE [9, 10, 14–16]. Further evidences have demonstrated that abdominal obesity, which is the core component of metabolic syndrome (MetS), may be a stronger predictor of RE than obesity [10, 11, 16–20]. MetS is generally recognized to be a combination of metabolic abnormalities, including abdominal obesity, hypertension, hyperglycemia, reduced high-density lipoprotein cholesterol (HDL-C), and elevated triglycerides (TG) [21, 22]. Visceral fat accumulation and insulin resistance are currently thought to be the predominant causes of MetS [6, 21, 23]. Patients with MetS have a high risk of cardiovascular disease, diabetes mellitus, and other atherosclerotic diseases [16, 23, 24]. A retrospective case-control study in China in 2010 found that a high waist-hip ratio, hyperglycemia, hypertriglyceridemia, and MetS were the associated factors for RE, and that HDL-C was associated with a reduced risk of RE in men [25]. Another study in Taiwan in 2017 also found a positive association between MetS and RE [26]. These findings suggest that there may be pathogenetic links between RE and MetS [10, 12, 27].

With the increasing levels of obesity and Western diet in Taiwan, it is important to understand the relationship between MetS and RE and further explore the risk factors for RE. Identification of modifiable risk factors and further prevent RE and MetS earlier in healthy people in Taiwan is an important task. However, there was lack of study related to the correlation between the two diseases in Taiwanese. Therefore, the primary aim of this study was to assess the relationship of MetS and its components and RE. The secondary aim was to investigate the association between the number of MetS components and the severity of RE. We hypothesized that there is a positive association between the two diseases in a Taiwanese population.

All variables in the model 2 (including age, gender, BMI, uric acid, smoking, alcohol drinking, hiatal hernia, and MetS) were controlled by each other. The variables that showed strong collinearity with others were excluded from model 2. The results of model 2 showed that MetS remained a significantly positive association with RE (adjusted OR^ß = 1.251; 95% CI = 1.071–1.462; p = 0.005).

Furthermore, we conducted model 3 to assess the association between RE and the individual components of MetS. All variables in the model 3 (including age, gender, uric acid, smoking, alcohol drinking, hiatal hernia, elevated BP, abdominal obesity and hyperglycemia) were controlled by each other. The variables that showed strong collinearity with others were excluded from model 3. Among the five MetS components, elevated BP (adjusted OR^γ = 1.163; 95% CI = 1.023–1.323; p = 0.021), abdominal obesity (adjusted OR^γ = 1.173; 95% CI = 1.020–1.349; p = 0.026) and hyperglycemia (adjusted OR^γ = 1.306; 95% CI = 1.142–1.495; p < 0.001) remained to be positively associated with the presence of RE.

Take a comprehensive look at the results of model 2 and model 3, older age (≥65 years), male gender, smoking, alcohol consumption, higher BMI, higher uric acid, and hiatal hernia were all found significantly related to RE. Particularly, the presence of hiatal hernia was found to be a strong associated factor for RE.

Same as our initial hypothesis, this cross-sectional population-based study demonstrated that MetS was positively associated with RE. Among the five MetS components, abdominal obesity, elevated BP, hyperglycemia, and elevated TG were independent predictors for RE. Most importantly, this study found a linear trend of association between the number of MetS components and the severity of RE.

Abdominal obesity has been hypothesized to induce gastroesophageal reflux through both mechanical and humoral causes. Abdominal obesity, also called central obesity, is defined as an elevated waist circumference. The significance of an elevated waist circumference is the accumulation of visceral fat. Compared with simple obesity (only defined by BMI), visceral fat accumulation is considered to be a stronger risk factor for the development of hypertension, dyslipidemia, diabetes mellitus, coronary artery disease, and cerebrovascular disease [9]. Visceral fat accumulation has also been reported to increase intra-abdominal pressure and transient lower esophageal sphincter relaxation, which can promote gastroesophageal reflux [9, 15, 19]. In addition to mechanical causes, adipose tissues have been shown to overproduce pro-inflammatory cytokines (interleukin-6, tumor necrosis factor-α, etc.), which can cause esophageal circular muscle relaxation and insulin resistance [9, 19, 24]. These mechanisms can then lead to chronic inflammation of the gastroesophageal junction and result in the development of RE.

Hypertension has been reported to be significantly associated with RE in several previous studies, which is consistent with the our study [9, 30, 31], although the mechanism for this relationship remains unclear. Calcium channel blockers are widely used to treat hypertension. Some studies have reported that calcium channel blockers are important associated factors for RE as they inhibit esophageal muscle contraction and reduce lower esophageal sphincter pressure leading to RE [9, 12, 30]. However, in addition to the side effects of antihypertensive treatment, the influence of hypertension itself must be taken into account. Several studies have suggested that genetic factors may play a role and that environmental factors known to increase blood pressure, such as high salt intake and life stress, may also have an impact on this issue [31]. Since this study is a retrospective cross-sectional study, we did not include environmental factors or exclude the patients using calcium channel blockers in our analysis.

The association between hyperglycemia and RE is controversial. Most previous studies have reported no significant correlation between hyperglycemia and RE [18, 19, 24, 32]. However, an increasing number of studies support a positive relationship between the two diseases [9, 19]. In our study, hyperglycemia (FPG ≥100 mg/dL) was an independent associated factor for RE. Current studies hypothesized that hyperglycemia causes RE due to diabetic autonomic neuropathy and insulin resistance. Vagus nerve injury, a common complication of diabetes autonomic neuropathy, affects esophageal motor and sensory function. Motor dysfunction of the esophagus includes slower gastric emptying, abnormal esophageal motility, ineffective bolus transport, and more frequent esophageal transient lower esophageal sphincter relaxation [9, 30, 33]. The perceptional abnormalities such as nausea and abdominal fullness, which result from esophageal sensory dysfunction, are important in clinical reflux symptoms [33]. Few studies have discussed the occurrence of RE in patients with impaired fasting glucose or glucose intolerance. Some studies have suggested that a high serum glucose level influences the motor function of the esophagus before the diagnosis of diabetes mellitus was made [9, 34]. Our study results strengthen the hypothesis that hyperglycemia is associated with the development of RE.

A weak association was found between elevated TG and RE after adjusting for age and gender (shown in model 1, Table 3). It has been reported that hypertriglyceridemia may be related to RE in several previous studies [9, 19, 26, 30, 35]. Hypertriglyceridemia is known to be associated with high dietary fat. An excessively high fat diet can result in delayed gastric emptying, and further increases the risk of RE [19, 32]. In accordance with our study, most previous studies have not reported a significant association between HDL-C and RE [10, 24, 35].

The most important result of this study is the positive association between MetS and RE. We further identified a linear trend of correlation between the number of MetS components and the severity of RE. According to multivariate analysis, we also found a dose-response relation between the number of MetS components and the presence of RE (shown in model 1, Table 3).

An increase in the number of MetS components was associated with the presence and a more severe grade of RE. This dose-response relation means that the two diseases are not only related to each other, but also have a degree of influence on each other. As described above, most MetS components have been proven associated with the occurrence of RE. A case-control study in Korea in 2008 identified insulin resistance and MetS as risk factors for RE [24]. Moreover, a cross-sectional study in 2011 reported that insulin resistance was an independent predictor for the prevalence and severity of RE [35]. Insulin resistance is currently known as a core factor in the development of MetS. Therefore, it is possible that MetS is an important predictor for the development of RE. From an another point of view, persistent gastric acid stimulation can result in chronic inflammation of the esophagus, and chronic inflammation has been proven to be an important leading cause of MetS [32]. Therefore, it is also possible that RE leads to the occurrence of MetS. Further prospective longitudinal studies are needed to elucidate which MetS or RE occurs first.

We also found different distributions of several baseline characteristics between the groups with and without RE. The individuals with RE were predominant male and exhibited older age, higher BMI, more smoking, and more alcohol consumption. A longitudinal study in 2008 reported possible spontaneous regression in patients with low-grade RE (LA grade A or B) without pharmacological treatment [15]. Accordingly, patients with low grade RE should be encourage to make lifestyle changes to reverse the disease, such as weight reduction, stop smoking and stop drinking alcohol.

There are several limitations to this study. First, previous studies have reported the prevalence of RE ranged from 9.0 to 24.6% in a screening health examination in Taiwan [36]. However, there was a particularly high prevalence rate of RE (59.6%) in our study population. All participants were enrolled from the database of our Health Examination Center, and most of them were able to afford self-paid health examinations, including upper gastrointestinal endoscopy. Although a high socioeconomic status had not been proven to be a causative factor for RE [18], it possibly caused selection bias in our study. Due to the upper gastrointestinal endoscopy was optional, relative invasive and high price, those who already had gastrointestinal symptoms tended to receive it. However, this hypothesis was not confirmed due to lack of asking gastrointestinal symptoms in the initial questionnaire. Overdiagnosis is an another possible explanation of the high prevalence rate of RE in our study population. Most people diagnosed with RE belong to LA grade A (87%). The diagnostic criteria for LA grade A by each gastroenterologist vary widely and may lead to overdiagnosis. Second, this study did not include the effects of several confounding factors, such as lifestyle and dietary habits, co-medications (including aspirin or non-steroid anti-inflammatory drugs, proton-pump inhibitor or histamine-2 receptor antagonists), and socioeconomic status. Excessive fatty sweat food consumption can also weaken lower esophageal sphincter and cause gastroesophageal reflux [8]. Recent aspirin or non-steroid anti-inflammatory drugs use may increase the risk of reflux esophagitis. Otherwise, taking proton-pump inhibitor or histamine-2 receptor antagonists may improve reflux esophagitis [12]. Third, self-reported medical and pharmacology history (such as anti-dyslipidemia medicine use) may result in misclassification bias. Fourth, not everyone received a rapid urease test, so there would be misclassification bias.

There are several strengths to this study. To the best of our knowledge, this is the first study to identify the linear trend of an association between the number of MetS components and the severity of RE. In addition, we enrolled a large number of participants. We also focused on a healthy population so we could suggest early prevention measures for the known risk factors.

In summary, the present study suggests that MetS is significantly associated with the presence and the severity of RE, although the study could not clarify whether MetS or RE occurred first. Therefore, further prospective longitudinal studies are needed to evaluate the association between RE and MetS in the future.