Risk Factors for Dysmotility, Acid Reflux Symptoms, and Overlap Using FSSG in Japan

1 Department of General Medicine, Oita University Faculty of Medicine, 1-1 Idaigaoka, Hasama-machi, Yufu-City, Oita 879-5593, Japan 2 Department of Environmental and Preventive Medicine, Oita University Faculty of Medicine, 1-1 Idaigaoka, Hasama-machi, Yufu-City, Oita 879-5593, Japan 3 Department of Gastroenterology, Oita University Faculty of Medicine, 1-1 Idaigaoka, Hasama-machi, Yufu-City, Oita 879-5593, Japan


Introduction
Dyspepsia such as dysmotility and epigastric pain and acid reflux symptoms including heartburn are common gastrointestinal symptoms and major symptoms for functional dyspepsia (FD) and gastroesophageal reflux disease (GERD) in general population [1,2].Several studies showed that there is a significant overlap between dyspepsia and acid reflux symptoms [1,3,4].As disturbances of gastrointestinal sensory and motor function are commonly accepted as the underlying pathogenesis of both dyspepsia and acid reflux symptoms, it is possible that the overlap induces by common risk factors.However, the risk factors for overlap of dyspepsia and acid reflux symptoms have not been enough elucidated.
Regarding dyspepsia, dysmotility symptoms are the major type in Japan [5,6].In 2004, Kusano et al. developed a screening tool for dysmotility symptoms and acid reflux symptoms, named frequency scale for the symptoms of GERD (FSSG) [7].Although several studies reported the usefulness of FSSG regarding GERD [7,8], there is no epidemiological study examined the prevalence of dysmotility symptoms by using FSSG.
Recently, many reports have shown that metabolic risk factors including obesity were related with gastrointestinal diseases including GERD [9][10][11][12][13][14]. Several mechanisms such as increased intraabdominal pressure resulted from obesity are proposed [15].Although obesity is also related with dyspepsia in some reports [16][17][18], the relation between Epidemiology Research International metabolic risk factors and dysmotility symptoms remains unclear.
In this study, we examined the prevalence and risk factors including metabolic risk factors for dysmotility symptoms, acid reflux symptoms, and overlap of the two symptoms by using FSSG in Japanese population.

Study Population.
A cross-sectional survey was performed in Japanese underwent the routine medical examination at Oita occupational health service center between April 2008 and May 2009.The information for height, weight, waist circumference, blood pressure, serum lipid, and fasting blood sugar was obtained from each subject.Abdominal obesity was defined as a waist circumference ≥85 cm in males and ≥90 cm in females.Measurements were performed at the World Health Organization (WHO) recommended site (midpoint between the lower border of the rib cage and the iliac crest) by a trained personnel [19].Blood testing was done for all persons after more than 12 hours of fasting.The information about smoking, alcohol intake, use of aspirin, nonsteroidal anti-inflammatory drugs (NSAIDs), antiacid secretion, antihypertensive drugs, antidyslipidemic drugs, and/or antidiabetic drugs was also obtained from each subject.Participants were diagnosed with metabolic syndrome if they had an elevated waist circumference and two of the following components: (1) high blood pressure (≥130 mm Hg systolic or ≥85 mm Hg diastolic), (2) hypertriglyceridemia (≥150 mg/dL) or low levels of high density lipoprotein-cholesterol (HDL-C) (<40 mg/dL), or (3) DM/hyperglycemia (≥110 mg/dL) [20][21][22].The subjects taking antihypertensive drugs, antilipidemic drugs, and/or antidiabetic drugs were also considered as positive of the corresponding components.Participants were requested to answer the questionnaire before the endoscopic examination.Informed consent was obtained by the ethical committee of Oita University Faculty of Medicine.

Questionnaire for Survey.
Participants completed a FSSG questionnaire, which is a self-administered validated questionnaire [7,8].Questions of FSSG consist of acid reflux symptoms and dysmotility symptoms.In this study, typical acid reflux symptoms were evaluated by 3 items as follows: "do you get heartburn?","do you get heartburn after meals?", and "do you get bitter liquid (acid) coming up into your throat?"Dysmotility symptoms were evaluated by 4 items as follows: "does your stomach get bloated?","does your stomach ever feel heavy after meals?", "do you ever feel sick after meals?", and "do you feel full while eating meals?"In FSSG, each question was validated on five points (range from 0 to 4): "never", "occasionally", "sometimes", "often", or "always".The subjects with more than two points at any items of acid reflux symptoms without significant endoscopic findings were considered as acid reflux symptoms group.The subjects with more than two points at any items of dysmotility symptoms without significant endoscopic findings were considered as dysmotility symptoms group.Subjects met both criteria were considered as overlap group.Persons other than acid reflux symptoms group, dysmotility symptoms group, and overlap group were considered as control group.

Endoscopy.
Upper endoscopy was offered to participants who responded to the survey.Skilled endoscopists, who was blinded to the results of the questionnaire performed endoscopy.The severity of erosive esophagitis was graded from A-D according to the LA classification [23].We considered LA-A to be the cut-off for erosive esophagitis.Peptic ulcer including scar phase was reported.The Makuuchi's classification was used in the assessment of hiatus hernia as follows: grade 1 indicated no columnar mucosa in the lower esophagus; grade 2, <1 cm of columnar mucosa; grade 3, 1 to 3 cm of columnar mucosa; grade 4, >3 cm of columnar mucosa [24].Reflux esophagitis, peptic ulcer including scar phase, gastric tumor, and postgastrectomy were considered as significant findings.

Statistical Analysis.
All statistical analyses were performed by SPSS version 18 (SPSS inc., Chicago, IL, USA).Statistical analysis was done using χ 2 test for comparison of discrete variables and t-test for comparison of continuous variables.Multiple backward stepwise logistic regression analyses were used to examine association with each disease.Predictor variables consisted of age (continuous variable), sex (dichotomous variable), waist circumference (defined as categorical variables with five levels: ≥74.9 cm, 75.0-84.9cm, 85.0-94.9cm, or ≤95.0 cm), hypertension (dichotomous variable), dyslipidemia (dichotomous variable), diabetes mellitus (dichotomous variable), smoking (defined as categorical variables with three levels: none, former smoking, currently smoking), alcohol intake (defined as categorical variables with three levels: none, former drinking, currently drinking), and hiatus hernia (defined as categorical variables with four levels: 1-4).For each variable, odds ratio (OR) and 95% confidence interval (CI) were given.A two-tailed P value of <0.05 was considered statistically significant.

Prevalence of Dysmotility Symptoms, Acid
Reflux Symptoms, and Overlap.Total of 778 Japanese underwent the routine medical examination.Among them, 267 subjects were excluded due to insufficient data of symptom questionnaire, and/or the lacking of information for the metabolic syndrome, thus 511 received upper gastrointestinal endoscopy with symptom questionnaire, and blood test for the metabolic syndrome (including blood sugar and serum lipid).Total of 116 persons were further excluded because of using antiacid secretion drugs and/or NSAIDs.Persons with significant endoscopic findings (reflux esophagitis, peptic ulcer including scar phase, gastric tumor, postgastrectomy) were also excluded to avoid the bias from these endoscopic findings.Therefore 395 persons (259 males and 136 females, mean age 47.9 ± 8.4 years) were finally included in this study.Figure 1 showed the prevalence of dysmotility, acid reflux symptoms, and overlap.Dysmotility symptoms were found in 129 (32.6%) and acid reflux symptoms were in 81 (20.5%).Overlap was found in 55 (13.9%) of all 395 subjects, which in 42.6% of dysmotility symptoms group and 67.9% of acid reflux symptoms group.Thus 74 persons were considered as dysmotility symptoms alone group and 26 were considered as acid reflux symptoms alone group.The remaining 240 subjects who did not have any of dysmotility symptoms nor acid reflux symptoms were considered as controls.

Characteristics of Dysmotility Symptoms Group, Acid
Reflux Symptoms Group, and Overlap.Table 1 shows the characteristics of dysmotility symptoms alone group, acid reflux symptoms alone group, and overlap.Comparing with controls, the female gender, the absence of diabetes, and lower alcohol use were significantly associated with dysmotility symptoms alone group in univariate analyses.The absence of abdominal obesity diagnosed by waist circumference was significantly associated with dysmotility symptoms alone group.In addition, when waist circumference was categorized into four levels, waist circumference was significantly smaller in dysmotility symptoms alone group than controls.The presence of abdominal obesity, hypertension, and smoking were significantly associated with acid reflux symptoms alone group compared with controls.Smoking and hiatus hernia were significantly higher in overlap than in controls.
Results from multiple logistic regression analysis examining associations of metabolic risk factors with each disease compared with controls are shown in Table 2.After adjusting for each factor, female gender was significantly associated with dysmotility symptoms.Increased waist circumference was inversely associated with dysmotility symptoms.No significant differences were observed between acid reflux symptoms and controls.Female gender, smoking, and hiatus hernia were significantly associated with overlap compared with controls.

Risk Factors for Overlap between Dysmotility and Acid
Reflux Symptoms.In univariate analyses, male gender, smoking, and hiatus hernia were significantly associated with overlap compared with dysmotility symptoms alone.In multivariate analyses adjusted by each factor, smoking was remained to be significantly associated factors for overlap (OR = 1.68, 95% CI; 1.12-2.53)(Table 3).Regarding acid reflux symptoms, abdominal obesity, and hypertension were higher in acid reflux symptoms alone than in overlap, on the other hand, smoking was higher in overlap than in acid reflux symptoms alone.In multivariate analyses adjusted by each factor, smoking was also remained to be significantly associated factors for overlap (OR = 2.70, 95% CI; 1.23-5.93).Increased waist circumference was inversely associated with overlap compared with acid reflux symptoms alone (OR = 0.40, 95% CI; 0.18-0.88).

Discussion
This is the first study evaluated the prevalence of dysmotility symptoms by using FSSG.In the present study, the prevalence of dysmotility symptoms, acid reflux symptoms, and overlap were 32.6, 20.5 and 13.9%, respectively.Although it is difficult to compare this result with previous reports due to the different diagnostic criteria, high prevalence of overlap between dysmotility symptoms, and acid reflux symptoms were consistent with previous reports [25,26], which suggesting that the common pathophysiological mechanism such as diffuse motor disturbances, altered visceral sensitivity, and brain-gut dysfunction should be involved in the formation of both dysmotility symptoms and acid reflux symptoms [27,28].
Recently, many reports have shown that metabolic risk factors, especially visceral obesity, appear to be involved in several gastrointestinal diseases, such as GERD [29].Some reports showed that obesity also had higher risk for dyspepsia [16][17][18]; however there are still few reports analyzing the relationship between metabolic risk factors and dyspepsia.Several mechanisms such as increased intraabdominal pressure and increased gastroesophageal pressure gradient inducing LES relaxation by which obesity may cause are proposed [15].In this study, any of metabolic risk factors except for smoking were not related with dysmotility symptoms, overlap, and even acid reflux symptoms.No relation of acid reflux symptoms with metabolic risk factors may be due to the small number of acid reflux symptoms alone subjects in this study.Interestingly, the presence of metabolic risk factors was rather tended to be inversely associated with dysmotility symptoms, although it can be the result of dyspeptic symptoms such as early satiety and fullness.In our knowledge, this is the first report investigating the association of dysmotility symptoms with metabolic risk factors.Consistent with our results, dysmotility symptoms were reported to be higher in underweight than desirable weight and obese subjects [5,30].The presence of hiatus hernia was the highest in overlap type.Although hiatus hernia induces gastroesophageal reflux disease by reflux of gastric acid, subjects with reflux esophagitis diagnosed by endoscopy were excluded in this study.Acid reflux induced by hiatus hernia may contribute to not only the acid reflux symptoms but also the dysmotility symptoms as consistent with previous reports [31,32].In addition, smoking was more associated with overlap than dysmotility symptoms or acid reflux symptoms alone.Lee et al. reported that the smoking was not an independent risk factor for overlap between GERD and dyspepsia [26].However, since smoking evokes antral hypomotility [33], dysmotility type dyspepsia of some smokers might be affected by smoking.Kim et al. reported that smoking was an independent risk factor for GERD [34].Common pathophysiological mechanism induced by smoking may be involved in subjects with overlap.Dose of smoking might be associated with the severity of symptoms.However, unfortunately, we do not have detailed information for the dose of smoking in this study.Further study is necessary to clarify the relationship between the dose of smoking and severity of symptoms.Dysmotility symptoms were more frequent in females, which was consistent with previous reports [35][36][37].Studies of somatic pain have shown that women tend to have lower thresholds for certain stimuli such as pressure and electrical stimuli [38].Moreover, delayed gastric emptying has been noted in women compared with men in most other studies [39,40].Further study is needed to clarify the pathogenesis.
As the other risk factors, for dyspepsia, Helicobacter pylori (H.pylori), psychological factors, and dietary factors are also reported [37,41].Recently, it is reported that anxiety was significantly associated with overlap between GERD and dyspepsia [26].It is necessary to investigate the relationship between H. pylori, dietary factors and overlap between dysmotility symptoms and acid reflux symptoms.
Our study has some limitations.First, it is possible that there was a selection bias.Not all subjects included for final analysis, overall response rate was 50.7% (395 of 778).Subjects with symptoms might be easy to answer the questionnaire.Second, we did not consider the epigastric pain symptoms, because FSSG questionnaire did not include the items.Some additional questionnaires, such as Rome III Integrative Questionnaire, might be needed in the future studies [42].Third, the question for frequency of symptom was different from other studies as described above.In FSSG, each question is validated on five points (range from 0 to 4): "never", "occasionally", "sometimes", "often", or "always".This ambiguous question in FSSG might contribute the difference of prevalence.Specific question such as "at least one day a week" will be reasonable.Finally, the number of subjects examined in this study was not so large.Metabolic risk factors except for smoking were not related with even acid reflux symptoms.This may be beta-error due to the small number of acid reflux symptoms alone subjects.Further large-scale study is necessary to confirm our findings.

Conclusions
We first used FSSG to examine the prevalence of dysmotility symptoms and acid reflux symptoms and their overlap.Overlap between dysmotility symptoms and acid reflux symptoms was common in Japanese.Furthermore, smoking was an independent risk factor for overlap group.Common pathophysiological mechanism induced by smoking may be involved in subjects with overlap.

Figure 1 :
Figure 1: The prevalence of dysmotility symptoms, acid reflux symptoms, and their overlap in the 395 subjects.

Table 1 :
Characteristics of four groups.
b P < 0.05 compared with the control group, using chi-square tests.

Table 2 :
Multivariate analyses of the risk for dysmotility, acid reflux symptoms, and overlap comparing with control.

Table 3 :
Risk factors for overlaps compared with a single disorder.
a Statistically significant by multivariate analyses.