Epidemiology of functional dyspepsia and gastroparesis as diagnosed in Flemish‐Belgian primary care: A registry‐based study from the Intego database

Dyspepsia is a prevalent condition in the general population. Besides organic causes, the differential diagnosis of dyspepsia includes functional dyspepsia (FD) and gastroparesis (GP) which share similar pathophysiological mechanisms and clinical presentation. So far, no study investigated the prevalence of FD and GP in a primary care in Belgium.


| INTRODUC TI ON
3][4] The symptoms of FD include postprandial fullness, early satiation, epigastric pain, and/or epigastric burning based on the ROME IV diagnostic criteria. 5On the contrary, GP is a clinical syndrome characterized by symptoms of nausea and vomiting, postprandial fullness, early satiety, and bloating with delayed gastric emptying in the absence of a mechanical obstruction. 6,7The etiology of FD is likely multifactorial, including disordered communication between the gut and the brain, motility dysfunction, visceral hypersensitivity, and low-grade mucosal inflammation. 8Conversely, the major causes of GP are diabetic and idiopathic. 91][12][13] In the UK, the prevalence of FD was 8% in the general population, and the prevalence of GP in general practice was 13.8 per 100,000 persons. 9,14The incidence of FD was 2.8% per year, based on dyspepsia questionnaires, and the incidence of GP ranged between 1.5 and 1.9 per 100,000 personyears, according to general practice records. 9,15In Belgium, previous studies from an Internet survey showed a 5% prevalence of uninvestigated dyspepsia in the general population and 0.5% of the population was affected by symptoms suggestive of GP. 2,16 However, the impact of dyspeptic symptoms on health care seeking behavior and its role in determining the prevalence of FD and GP diagnoses in clinical practice in Belgium remain unknown.
Therefore, the current study assessed the epidemiology of FD and GP in Flemish speaking Belgian general practice which will help to define future approaches to target general practitioners for education, screening, and specific intervention in patients with FD and GP.

| Study design and data source
This retrospective, cross-sectional study utilized the Intego (Integrated Computer Network) database, a Flemish-Belgian general practice-based morbidity registration network managed at the Academic Centre for General Practice of the KU Leuven, which was previously described. 17The Intego database captures longitudinal patient-level data from registration by more than 100 general practitioners. 17

| Data analysis
Due to the categorization of patients diagnosed with FD/GP as healthy if diagnosed prior to registration, any subsequent registration for FD/GP is mistakenly considered as an incident case.
This phenomenon can lead to an overestimation of incidence while concurrently underestimating the prevalence of the disease, especially in the initial years of registration.To avoid this, a cross-sectional analysis was conducted for the prevalence and incidence over last 10 years from 2011 to 2021.We determined prevalence, demographics (age and sex), and relevant comorbidities (diabetes, gastroesophageal reflux disease (GERD), irritable bowel syndrome (IBS), and chronic constipation) from this dataset.
The prevalence of FD/GP was calculated as a percentage with 95%

Key points
• Functional dyspepsia and gastroparesis are not commonly diagnosed in primary care in Flemish-Belgium.
• The presence of somatization, anxiety, and depressive disorders significantly increase the risk of diagnosed functional dyspepsia and gastroparesis.
• The lower prevalence of functional dyspepsia and gastroparesis reported in primary care, comparing to general population, implicates infrequently health care seeking of patients with dyspeptic symptoms or, more likely, a lack of awareness of both entities among physicians in Flemish-Belgium.
confidence interval (CI).The group of patients that consulted their general practitioner at least once in the given year, was used as denominator of prevalence, called the Yearly Contact Group (YCG).
Prevalent cases of FD/GP were patients who had a diagnosis on or before 31 December of the given year and incident cases were those who had a new diagnosis in the YCG.For incidence, the YCG who did not have FD/GP were considered at risk.Case-control comparison was conducted to investigate the impact of diabetes, somatization, anxiety, and depressive disorders on the diagnosis of FD/GP.Patients were matched 1:1 on birth year, gender, and the identification of general practice.The date of registration FD/ GP was as close as possible during the 1:1 matching, and the difference was less than 2 years.Optimal matching algorithm as previous studies was used to increase the balance of the matched sample. 18The effects of preexisting comorbidities (somatization, anxiety, and depressive disorders) were determined by using odds ratios (ORs).Categorical variables were analyzed using the chisquare test and continuous variables were compared with the Student's t-test.Multivariate logistic regression analysis was used to yield an adjusted odds ratio.Throughout, the statistical significance level used was 0.05.Statistical analyses were performed in R v.4.0.1 (R Foundation for Statistical Computing, Vienna, Austria).
The ccoptimalmatch R package was used as the analytic tool for case-control matching. 18

| Prevalence and incidence of FD/GP
From 2011 to 2021, the YCG varied between 83,405 and 224,726 people.The FD/GP prevalence remained stable over the study period (1.03%-1.21%, Figure 1).In 2021, among 224,970 subjects in the YCG, 2488 patients received a diagnosis of FD or GP.The prevalence of FD/GP in 2021 was 1.11% (95% CI 1.06-1.15).There were 190 patients with FD or GP among 13,888 diabetic patients.The prevalence of FD/GP in diabetic subjects was 1.37% (95% CI 1.18-1.58),which was significantly higher than nondiabetic patients (p < 0.01).
The prevalence of FD/GP did not differ significantly among patients with type 2 diabetes and type 1 diabetes (1.42% (95% CI 1.22-1.63)versus 0.86% (95% CI 0.43-1.53),p = 0.10).Figure 2 shows the incidence of FD/GP over the course of years.The crude incidence rate of FD/GP was stable, from 109 to 142 per 100,000 adults between 2011 and 2021.

| FD/GP and other gastrointestinal motility disorders
Among 9136 observations (9110 with FD, 26 GP), a total of 7526 patients received a diagnosis of FD or GP.After screening for organic gastrointestinal disease diagnostic codes, 5242 eligible patients (20 GP, mean age = 45.6 ± 18.9 years, 53.7% female) with FD/GP were identified (Figure 3).When considering overlapping with other gastrointestinal motility disorders, patients with FD/ GP showed 18.8% overlap with GERD, 17.1% overlap with IBS, and 18.7% overlap with chronic constipation.Up to 41.6% patients had at least one comorbidity (31.7% occurring after the diagnosis of FD/gastroparesis) and 11.6% of patients had two or more of these comorbidities.

| DISCUSS ION
To our knowledge, this is the first epidemiology study on FD/GP in Belgium based on the data collected from primary care setting.To enable patients to receive early treatment for dyspeptic symptoms, both patient and doctor must be aware of the presence of this problem.We found stable prevalence and incidence over the 11 years with 1.0%-1.2%and 109 to 142 per 100,000 adults, respectively, even though the Rome Diagnostic Criteria of FD was changed from Rome III to Rome IV during the period studied (2011-2021).Compared to the previous worldwide study via an online survey, the prevalence of diagnosed FD/GP in primary care setting, according to our result, is substantially lower than the prevalence in the general population. 2 The incidence of FD/GP in the primary care was also much lower than the 2.8% from a longitudinal study of FD with 10-year follow-up of subjects originally enrolled in a population-screening program for Helicobacter pylori in the UK. 15 It is essential to note that both studies had differences in the study design and were not primary care practice-based research.
One of the possible explanations for this finding is that a majority of patients with mild severity of dyspepsia symptoms in general population do not necessitate a visit to their general practitioner. 19ditionally, many mildly to moderately symptomatic people may use self-medication to control their digestive symptoms instead of seeking medical care. 20One population-based study in another European country demonstrated that only about one fourth of individuals with dyspeptic symptoms seek medical consultation. 21Our finding may implicate infrequent health care seeking behavior of individuals with dyspeptic symptoms in Flemish-Belgium.However, this is in contrast with an earlier Belgian survey where over 61% of subjects with dyspeptic symptoms reported health care contacting. 22Furthermore, among individuals attending an outpatient clinic with primary care physicians, a study conducted in Japan reported the prevalence of FD was 2.9%, which was comparable to the prevalence of 2.5% found in the general population of Japan by a recent multinational study. 2,23r finding also highlights the issue of the presumably high rates of undetected or not formally diagnosed FD/GP in primary care and a significant number of patients with FD would be labeled with gastroesophageal reflux disease or other disorders of gut-brain interaction such as IBS. 24Although technically a normal endoscopy is required to diagnose FD, the recent consensus on FD suggested that patients suspected of having FD without alarm symptoms or risk factors can be managed without endoscopy in primary care. 25,26It has been reported that the vast majority of subjects with dyspepsia in the general population do not have organic disease explaining their symptoms. 27Since access to upper gastrointestinal endoscopy and gastric emptying tests is less readily available in a general practice setting, symptom pattern recognition and an appropriate clinical referral for diagnostic testing are the critical steps for early diagnosis of GP at that level.
Symptoms of FD and GP are a frequent finding in patients with diabetes. 28In the present study, a higher prevalence of FD symptoms was present in diabetics compared to nondiabetics.The prevalence of FD/GP symptoms did not differ between type 1 and type 2 diabetes mellitus, based on a limited identified case number.Currently, there are no studies that compare the prevalence of FD in type 1 and type 2 diabetes in primary care.The diagnosis of FD in diabetic patients may be challenging because diabetic patients may have gastroparesis, enteropathy, and autonomic neuropathy.However, a study in the US, based on medical records, found a higher prevalence of GP in type 1 diabetes, compared to type 2 diabetes. 29A possible explanation for the absence of such finding in the current paper is that there is only a limited number of GP cases in our database and the result represents the prevalence of FD alone.
In Europe, GERD, IBS, and chronic constipation are common disorders of gut-brain interaction affecting 8.8%-25.9%,3.3%-5.9%,and 8.6%-14.5% of the population of European countries, respectively. 2,30The frequency of overlap between FD and GERD or IBS in our study is lower than in earlier literature reports.
Previous general population studies show that patients with FD frequently have overlapping IBS and GERD. 31Up to 33% of subjects with dyspepsia symptoms had GERD and 34% of FD individuals also met criteria for IBS in population-based surveys in Europe. 22,32A lower prevalence of overlap probably reflects the possibility of undetected IBS or GERD in primary care.Despite the fact that chronic constipation in combination with various gastrointestinal and extraintestinal symptoms can also be due to other diseases, we found that 18.7% FD or GP patients overlap with chronic constipation which is similar to the result from online survey in the United States reporting chronic constipation in 17% of FD patients. 33It also worth mentioning that distinguishing between chronic constipation and IBS with constipation can be challenging as these two disorders often overlap. 34,35 the literature, both FD and GP have been associated with psychological comorbidities, such as anxiety, depression, and somatization. 36,379][40][41] Herein, we found that patients with psychological comorbidities had a higher risk of FD/GP.Pre-existing chronic diseases might affect patients in help-seeking for symptoms and doctors' clinical diagnosis.Our finding emphasizes the importance of considering a diagnosis of FD or GP in patients with preexisting psychological comorbidities in primary care.On the contrary, anxiety, depression, and somatization were not associated with altered risk of FD/GP in the diabetic background population.
Diabetes mellitus probably has an overriding effect on developing GP/FD and reduces the effects of psychological factors.
The present study has potential clinical implications but also has several limitations.First, in this study, the diagnosis of GP and FD was based on diagnostic codes, instead of combining symptom pattern with appropriate diagnostic investigation as described by the Rome guidelines, and this approach may lack optimal specificity.Second, our findings likely underestimate the prevalence and incidence of FD/GP since under-reporting bias by general practitioners cannot be excluded.However, utilizing the same database, epidemiological reports on different chronic diseases have yielded comparable results to those found in the literature. 42,43nally, inherent to a retrospective study, the analysis is limited to the available data cohort and sets.There was a limited number of cases with diagnosis of GP.We are unable to estimate prevalence or incidence of GP alone and investigate the impact.Future prospective epidemiological research needs to be done in the primary care population.
In summary, the incidence and prevalence rates of FD/GP remained stable over the past decade in primary care and were less prevalent than in the general population.The available data suggest that a number of patients experiencing dyspeptic symptoms may not seek medical care and highlight a potential lack of awareness of FD and GP among physicians in Belgium during the studied period.
Between 2000 and 2021, the database included 586,164 patients (age 18-104 years old) with a total of 2,663,412 observations.General practitioners with optimal registration performance were included in the registry.The registered data, including patient information, new diagnoses made by general practitioners or specialists entered by the general practitioners after hospitalization, drug prescriptions, and laboratory test results, were updated continuously and accumulated for each patient.New diagnoses were classified according to a detailed thesaurus (Medidoc codes) automatically linked to the International Classification of Primary Care (ICPC-2) and International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10).All patients registered in the Intego for at least 1 day from 2000 to 2021 were included for the current analysis.Cases of FD and GP were identified using ICD-10 codes (K30 for FD, K31.0, K31.84 for GP).E10 and E11 diabetes ICD-10 codes were selected for Type 1 diabetes mellitus and Type 2 diabetes mellitus, respectively.According to ICD-10 code, comorbidities of gastrointestinal disorders were extracted as well.For cases with missing ICD-10 code, diagnosis in medical records was used to detect a diagnosis of FD or GP.Patients with a relevant organic gastrointestinal disease diagnostic code, such as peptic ulcer disease, inflammatory bowel disease, and cancer, were excluded.The study was approved by the ethics committee of the Leuven University Hospital in Belgium (reference study number S63825) and was performed in accordance with Good Clinical Practice guidelines.
After the matched pairing of subjects with and without FD/GP (mean age 56.6 ± 20.0, 2813 female), 5236 FD/GP (204 diabetes and 5032 nondiabetes) patients were compared to 5236 subjects without dyspepsia symptoms (222 diabetes and 5014 nondiabetes).Patients with somatization/anxiety/depression had a significantly higher risk of developing FD/GP, compared to controls (OR 1.38, 95% CI 1.18-1.61,p < 0.01).The odds ratio remained increased after adjusting for diabetes with conditional logistic regression (OR 1.39, 95% CI 1.19-1.61,p < 0.01).After the matched pairing of diabetic patients with and without FD/GP, 340 diabetic patients with FD/ GP (mean age 60.7 ± 14.3 years, 183 female) and 340 diabetic patients without FD/GP (mean age 60.9 ± 14.3 years, 183 female) were analyzed.Diabetic patients with somatization/anxiety/depression F I G U R E 1 Trends in the prevalence of functional dyspepsia and gastroparesis between 2011 and 2021 in primary care.did not have an increased risk of FD/GP, compared to patients with diabetes alone (OR = 1.15 (95% CI 0.72-1.85,p = 0.56)).

F I G U R E 2
Trends in the incidence of functional dyspepsia and gastroparesis between 2011 and 2021 in primary care.F I G U R E 3The flowchart of patient inclusion and exclusion.