A comparative study of disorders of gut–brain interaction in Western Europe and Asia based on the Rome foundation global epidemiology study

Many studies have been published on disorders of the gut–brain interaction (DGBI) in Asia and Western Europe, but no previous study has directly assessed the difference between the two regions. The aim was to compare the prevalence of DGBI in Asia and Western Europe.


| INTRODUC TI ON
Disorders of the gut-brain interaction (DGBI) are prevalent disorders that affect the global population. DGBI diagnoses are based on the Rome criteria, which define the frequency and duration of symptoms or symptom combinations that are required for diagnosis of each disorder, as well as the necessary exclusion of organic diseases that might explain the symptoms. 1 DGBI in the Rome criteria are divided into anatomical regions based on the presumed origin of the symptoms, that is, esophageal, gastroduodenal, gallbladder and sphincter of Oddi, bowel and anorectal, and an additional category with centrally mediated disorders of gastrointestinal pain.
There are a wide range of studies available on individual DGBI in Western Europe and Asia, especially on the more common DGBI such as irritable bowel syndrome (IBS) and functional constipation. These studies have been summarized in meta-analyses [2][3][4][5][6] where the prevalence of DGBI has been found to be higher in Europe, although with some exceptions. 3 However, these meta-analyses have been severely limited by study heterogeneity. Furthermore, studies on the less common DGBI are scarce, and there is a lack of studies that collect data on both subjects from Western Europe and Asia with the same methodology, limiting the possibility to perform direct comparisons. Similarly, data on the differences and similarities regarding characteristics of subjects with DGBI in the two geographical regions are lacking. In addition, studies using factor analyses have identified different symptom clusters in Asia compared with Western countries, potentially indicating differences in gastrointestinal symptom patterns in Europe and Asia. 7,8 The prevalence of IBS has been found to be lower when defined by the Rome IV compared to the Rome III, 2,9 but it is unclear whether this is true in both Western Europe and Asia, and more specifically, if the change in IBS prevalence between Rome III and IV criteria is similar in the two geographical regions. In addition, infectious diarrhea has been associated with the onset of IBS and the frequency of postinfection IBS has been shown to be similar in studies performed in Asia and Europe. 10 However, the proportion of IBS with onset after infectious diarrhea has not been directly compared in the two geographical regions using uniform methodology.
Hence, the primary aims of the study were to compare the prevalence and overlap of DGBI in Western Europe and Asia, and to compare the characteristics of individuals with DGBI in the two geographical regions. Secondary aims were to compare the occurrence France, Germany, Netherlands, Italy, Spain, Sweden, and the United Kingdom) and Asia (China, Japan, South Korea, and Singapore). We assessed DGBI diagnoses (Rome IV Adult Diagnostic Questionnaire), anxiety/depression (Patient Health Questionnaire-4, PHQ-4), non-GI somatic symptoms (PHQ-12), and access to and personal costs of doctor visits. Some of the data in this paper have already been reported in previous Global Study papers. This is inevitable since the original paper 9 included a broad range of descriptive statistics for all countries (33) and all disorders (22). Other papers, including the present one, which use the same database, are reporting in-depth analyses for countries, disorders, and methods, and these include a brief overview of some specific data previously reported.

Results
The Internet survey in the RFGES was reviewed by the Institutional Review Board (IRB) of the University of North Carolina at Chapel Hill before data collection started and was deemed exempt from IRB oversight due to the anonymity of the participants.

| Variables and definitions
DGBI were defined according to the Rome IV criteria. 1 "Any" DGBI was defined as the presence of at least one DGBI. DGBI were categorized by anatomical regions as defined by the Rome IV criteria, as displayed in Table 1. Two other DGBI categories, centrally mediated abdominal pain and biliary pain, were not included due to the low number of diagnosed individuals meeting criteria for these diagnoses (less than 0.1% of the population surveyed). When estimating the prevalence of esophageal and gastroduodenal DGBI, subjects who reported celiac disease, GI cancer, inflammatory bowel disease, or peptic ulcer were excluded from being classified as DGBI but were retained as non-DGBI cases. For bowel and anorectal DGBI, subjects who reported celiac disease, GI cancer, inflammatory bowel disease, diverticulitis, or bowel resection were excluded from DGBI and retained as non-DGBI cases. Otherwise, no subjects were excluded. In the comparison of IBS prevalence by Rome III and Rome IV criteria, South Korea, Italy, Spain, Sweden, and the UK, where Rome III questions were not included in the survey, were excluded.
The presence of psychological co-morbidity was evaluated with the Patient Health Questionnaire-4 (PHQ-4). The PHQ-4 is a fouritem questionnaire used to screen for anxiety and depression. Based on the PHQ-4, anxiety was defined as a total score of >2 on the first two questions of the questionnaire, and depression as a total score of >2 on the last two questions. 11 The PHQ -12 12

| Statistical analysis
Categorical variables are summarized as proportion in percentages and their 95% confidence intervals (95% CI). Continuous variables are summarized as means and 95% CIs. We planned to test differences in the proportion of subjects with at least one DGBI, esophageal, gastroduodenal, bowel, or anorectal DGBI in Western Europe and Asia, whereas other comparisons are descriptive. The magnitude of differences (effect sizes) is described with odds ratios TA B L E 1 All DGBI included in the study and their classification to anatomical regions.

| Prevalence of DGBI and the impact of associated factors
In unadjusted comparison, the proportion of patients with any DGBI was higher in Western Europe when compared to Asia (OR 1.14; 95% CI 1.08-1.20), and the same was true when comparing DGBI prevalence by anatomical regions (esophageal, gastroduodenal, bowel, and anorectal), OR 1.08-1.45, p < 0.05 (Figure 1). When assessing factors to be adjusted for when comparing prevalence rates between regions, the following variables were all associated with DGBI; younger age, female sex, and higher scores on PHQ-4, PHQ-12, access to a doctor and personal costs of a doctor visit (Table 2). In an adjusted comparison, correcting for the abovemen-   (Figure 3). Certain DGBI were so rare that reliable ORs and CIs could not be estimated and therefore not included Figure 2; these were functional biliary pain (prevalence 0.098% vs 0.021% in Western Europe and Asia, respectively), centrally mediated TA B L E 2 Demographic variables of subjects with and without DGBI in the total sample.  Tables S1 and S2.
When analyzing prevalence rates in individual countries within regions, in Western Europe the countries with the highest prevalence of at any DGBI were France, Spain, and Italy (47%, 44% and 43%, respectively), the Netherlands had the lowest (31%). In Asia, the prevalence of at any DGBI in Japan and South Korea was the highest (39% in both); it was lowest in Singapore (31%) (Figure 4).

| Comparison of subjects with DGBI in Western Europe and Asia
With regard to demographic factors, subjects with DGBI in Western Europe compared to Asia were older, more likely to be female, and had higher BMI but lower education level (Table 4). Furthermore, psychological factors were more prominent among DGBI subjects in Western Europe, and they were more likely to report no/small personal cost of medical care, but no difference was observed in the subjects' ability to visit a doctor if needed (Table 4). Additional details are provided in Table S3.
Subanalyses were performed in the same manner as above for subjects with IBS, functional dyspepsia, functional constipation, and functional bloating in Western Europe and Asia. The most no- Europe and in Asia. The same was true for age and sex for subjects with functional constipation in the two regions. Apart from that, the same general pattern of differences and similarities was found when comparing these subanalyses to the main analysis on all DGBI (Tables S4-S19).

| DISCUSS ION
In this study, we found that the prevalence of DGBI was generally higher in Western Europe when compared to Asia. However, when taking psychological and non-GI somatic symptoms into account, the difference in DGBI prevalence between the two regions was significantly diminished. Overlap of DGBI across anatomical regions was greater in Western Europe, further substantiating its greater DGBI burden when compared to Asia. The characteristics of subjects with DGBI in the two geographical regions varied considerably, and how they varied was different for individual DGBI, for example, subjects with IBS were quite similar in the two geographical regions.
Post-infection onset of IBS was more common in Asia, but IBS was more common in Western Europe, as was the case for several other DGBI. Lastly, the difference in IBS prevalence between these world regions seemed to be greater when using the Rome IV compared to Rome III criteria.
To our knowledge, this is the first time that prevalence rates of all DGBI in Western Europe and Asia are compared in a single study where all data were obtained and defined with the same methodology. Most previous studies have reported the prevalence of a single or few DGBI in only one of these geographical regions. There are systematic reviews and meta-analyses available that have gathered and compared prevalences of IBS, 2-4 functional constipation, 5 and uninvestigated dyspepsia. 6 These have generally shown that the prevalance of these DGBI are higher in Europe compared to Asia.
In a meta-analysis from 2012, the prevalence of IBS according to Manning and Rome I-III criteria in Northern and Southern Europe was shown to be 12%-15%, higher than the 7% prevalence found in Southeast Asia. 3 In contrast, one meta-analysis showed higher prevalence of IBS in Asia, 9.6%, compared to Europe pooled with North America, Australia, and New Zealand, 8.1%. 4 In a meta-analysis on functional constipation, the prevalence was often higher in Western European countries compared to Asia, but the pooled prevalence across geographical regions was not reported and it is important to note that the study included data from the RFGES data set. 5 With regard to uninvestigated dyspepsia, a meta-analysis showed a prevalence of 14.6% in Southeast Asia and 21.7%-24.3% in North and South Europe. 6 All of these meta-analyses have been severely limited by study heterogeneity, and few studies exist that directly compare Asia and European countries using the same methodology.
In the current study, psychological factors and non-GI somatic symptoms were shown to have an important role in DGBI prevalence differences between Western Europe and Asia. The association F I G U R E 3 Prevalence of individual DGBI in Western Europe (blue) and Asia (red). Odds ratios derived from a simple logistic regression model with geographical region as dependent variable and each DGBI as an independent variable are displayed on the right side of the graph. Error bars and numbers in parentheses represent 95% confidence intervals.
between these factors and DGBI is well known and is thought to be mediated through the brain-gut axis. 15 Importantly, the association has been shown to be bidirectional. 16 The current study was crosssectional, so data on DGBI, psychological factors, and non-GI somatic symptoms were collected at a single time point; therefore, we cannot conclude on the direction of the association. Furthermore, it is likely that these factors only constitute a part of the explanation for different prevalence between the two geographical regions, as our analysis did not include all known risk factors for DGBI and it is likely that complex residual confounders exist between the two regions that we were not able to account for.
The general pattern was that most DGBI were more common in Western Europe than in Asia. However, there were four DGBI that were marginally more common in Asia, that is, functional constipation, functional diarrhea, cyclic vomiting syndrome, and excessive belching. The reason why only these DGBI were found more commonly in Asia is unclear and we have no unifying explanatory factor or theory for these findings. However, it should be noted that for three out of these four DGBI the differences in prevalence rates were small. The characteristics of DGBI subjects in Western Europe and Asia were quite different, which may be related to the fact that certain DGBI were more common in Western Europe and some were less common, which may lead to an imbalance in the characteristics of the two groups. Additionally, a part of the explanation may be underlying differences in certain characteristics that may be more prominent in either region without it being related to DGBI in any way.
It is well known that prevalence of IBS is lower when defined by the Rome IV compared to the Rome III 2,9,17  Gastroenteritis is a known risk factor for the development of IBS 10 and in the current study a considerable proportion of subjects with IBS reported a post-infection onset, which was more commonly seen in Asia. This difference is not easily explained, but one possible explanation could be that infectious diarrhea is more common in Asia F I G U R E 4 Prevalence of any DGBI and DGBI by anatomical region in all countries included in the study. Error bars represent 95% confidence intervals.
compared to Europe, but reliable estimates comparing the frequency of infectious diarrhea in these regions are scarce. Previously, the prevalence of IBS after infectious diarrhea has been shown to be similar in Asia and Europe, 10 so different susceptibility to post-infection IBS in the two geographical regions does not seem to be an explanation.
The strengths of the current study include its large sample size and electronic data gathering that ensured completeness of data.
Furthermore, the data gathering was conducted in the same way in each country and efforts were made to translate questionnaires as accurately as possible. With regard to limitations, there are factors important for the context of DGBI about which data were not collected in the current study. These factors may differ between Western Europe and Asia, for example diet differences and H.pylori status. Furthermore, there are additional differences between the regions that are less definable, such as cultural, environmental, and language differences. However, our inability to correct for all of these factors does not change the fact that DGBI were found to be more common in Western Europe. There were some intra-regional variations observed in the study analyses. To account for this, we considered using mixed models with country as a nested random effect within region, but unfortunately the number of countries in Asia was too low to carry out this analysis. Of course, the results of the study have to be interpreted with these intra-regional variations in mind, but ultimately the primary aim of the study was to compare geographical regions and not countries.
To conclude, in this large, multinational survey using uniform methodology across countries, DGBI were found to be more common and have greater overlap in Western Europe compared to Asia. However, the difference was less apparent when correcting for differences in the severity of psychological and non-GI somatic symptoms, underlining the importance of these factors for DGBI.
The characteristics of subjects with DGBI varied by geographical region, especially with regard to psychological symptoms and non-GI somatic symptoms. Postinfectious onset of IBS was more common in Asia, and the difference in IBS prevalence between these two regions tended to be greater when using the Rome IV rather than the Rome III criteria.

AUTH O R S CO NTR I B UTI O N S
JPH conceptualized the study, analyzed and interpreted data, wrote the first and revised manuscripts and approved of the final paper.