This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.Functional dyspepsia (FD) is characterized by postprandial fullness, early satiation, epigastric pain, or epigastric burning despite the absence of organic digestive or metabolic disorders.1 FD is one of the most common gastrointestinal disorders and has a high prevalence worldwide. FD can limit affected individuals’ social life and affect their quality of life.2 The reported risk factors for FD are female, current smoker, nonsteroidal anti-inflammatory drug (NSAID) use, and
In epidemiologic studies from Iceland, Taiwan, and 3 other countries (United Kingdom, Canada, and United States), the prevalence of FD in a young population was higher than that in middle-aged and older people.5-7 FD can impair patient work productivity and increase their medical costs. In the overall United States population, the estimated medical costs (including indirect costs) related to FD were reported to be 18.4 billion dollars.8 There is little information on FD in young people, however. Previous cohorts mainly consist of middle-aged subjects and/or patients with gastrointestinal symptoms.
Body mass index (BMI) is a well-known risk factor for gastroesophageal reflux disease (GERD) and erosive esophagitis.9,10 The association between BMI and FD in Western countries remains inconsistent, however. Several studies reported that BMI including obesity positively associated with dyspepsia symptoms mainly in Europe, the United States, and Latin American countries.11-16 In a large 10-year population-based study in Iceland, weight loss positively associated with the development of FD.5 In other studies, no association between BMI and FD was found.17-21
In the Asian population, 5 studies showed an inverse association between BMI and FD,22-26 and visceral adiposity was associated with FD in a Korean study.27 In an Iranian study of the general population, no association between BMI and FD was found.28 The association between BMI and FD in the Asian population remains unclear. Further, to date, no study has evaluated this issue in young people in Asian and Western populations. The primary aim of the present study is to investigate the association between BMI and FD based on the Rome III criteria in young Japanese people.
We enrolled 10 104 university students who had no missing health check-up examination data at university between April 2015 and April 2017 in this study. A specific questionnaire pertaining to FD based on the Rome III criteria classification29,30 was sent to all subjects who underwent a health check-up. Information that was consistent with the Rome III criteria was collected from each subject’s digestive medical history, which focused on the diagnosis of organic conditions and concerning symptoms. The exclusion criteria were as follows: medication for digestive disorders within 6 months, use of NSAIDs and steroids, and concerning signs (recurrent vomiting, weight loss, bloody stool/melena, and dysphagia). Subjects were also excluded if they reported organic diseases such as GERD, gastritis, peptic ulcers,
Figure 1. Study population.
Using a self-administered questionnaire, information was collected and the following definitions were used for smoking, drinking, and regular exercise habit, and medical history. Current smoking was defined as smoking 1 cigarette or more per day. Current drinking was defined as drinking alcohol, regardless of the amount or frequency. Regular exercise was defined as exercising 1 or more times per week.
FD was defined as the subject reporting 1 or more symptoms, such as postprandial fullness, early satiation, or epigastric pain or burning. The definition of FD in this study was based on the Rome III criteria.29,30 Subjects with digestive symptom onset within 6 months prior to this study’s survey were excluded from this cohort.
BMI was classified into 4 categories (quartiles) on the basis participant distribution, as follows: (1) lowest BMI, < 19.36 kg/m2; (2) low BMI, 19.36 kg/m2 to 20.90 kg/m2; (3) moderate BMI, 20.90 kg/m2 to 22.74 kg/m2; and (4) high BMI, > 22.74 kg/m2 (reference). On the basis of the men’s and women’s BMI distribution, men’s and women’s BMI were classified into 4 categories (quartiles): men’s BMI: (1) lowest BMI, < 19.56 kg/m2; (2) low BMI, 19.56 kg/m2 to 21.12 kg/m2; (3) moderate BMI, 21.12 kg/m2 to 23.12 kg/m2; and (4) high BMI, > 23.12 kg/m2 (reference), and women’s BMI: (1) lowest BMI, < 19.10 kg/m2; (2) low BMI, 19.10 kg/m2 to 20.53 kg/m2; (3) moderate BMI, 20.53 kg/m2 to 22.18 kg/m2; and (4) high BMI, > 22.18 kg/m2 (reference). The definitions of lean, normal, and overweight and obese were < 18.5 kg/m2, 18.5 kg/m2 ≤ BMI < 25 kg/m2 (reference), and 25 kg/m2 ≤ BMI < 30 kg/m2, and 30 ≤ BMI kg/m2, respectively. Estimations of crude odds ratios (ORs) and their 95% confidence intervals (CIs) for FD in relation to BMI, leanness, and obesity were performed using a logistic regression analysis. We selected the following potential confounding factors: age, drinking, smoking, exercise habits, anemia, and sports injury. Trend of an association was assessed using a logistic regression model assigning consecutive integers to the categories of BMI variables. For tests of quadratic trend, we including linear and quadratic terms in the model. SAS software package version 9.4 (SAS Institute Inc, Cary, NC, USA) was used to perform the statistical analyses.
Table 1 shows the characteristics of the 8923 study participants. The percentage of men was 61.4% in this cohort. The mean age and BMI were 20.1 years and 21.35 kg/m2, respectively. The frequency of smoking, drinking, exercise habit, and sports injury in men was higher than that in women (
Table 1 . Clinical Characteristics of 8923 Study Participants
| Variable | Total (N = 8923) | Men (n = 5478) | Women (n = 3445) | |
|---|---|---|---|---|
| Age (yr) | 20.10 ± 2.80 | 20.20 ± 2.59 | 20.00 ± 3.10 | 0.003 |
| BMI (kg/m2) | 21.35 ± 3.05 | 21.68 ± 3.24 | 20.84 ± 2.63 | 0.001 |
| BMI < 18.5 | 1241 (13.9) | 656 (12.0) | 585 (17.0) | 0.001 |
| 18.5 ≤ BMI < 25 | 6778 (76.0) | 4131 (75.4) | 2647 (76.8) | |
| 25 ≤ BMI< 30 | 747 (8.3) | 558 (10.2) | 189 (5.5) | |
| 30 ≤ BMI | 157 (1.8) | 133 (2.4) | 24 (0.7) | |
| Smoking | 527 (5.9) | 492 (9.0) | 35 (1.0) | 0.001 |
| Drinking | 973 (10.9) | 736 (13.4) | 237 (6.9) | 0.001 |
| Exercise habit | 3408 (39.3) | 2415 (44.1) | 1093 (31.7) | 0.001 |
| Medical history | ||||
| Irregular pulse | 83 (0.9) | 52 (1.0) | 31 (0.9) | 0.810 |
| Heart murmur | 47 (0.5) | 30 (0.6) | 17 (0.5) | 0.730 |
| ECG abnormality | 63 (0.7) | 33 (0.6) | 30 (0.9) | 0.140 |
| Kidney disease | 7 (0.1) | 4 (0.1) | 3 (0.1) | 0.810 |
| Anemia | 239 (2.7) | 52 (1.0) | 187 (5.4) | 0.001 |
| Traffic accident | 115 (1.3) | 75 (1.4) | 40 (1.2) | 0.390 |
| Sports injury | 273 (3.1) | 219 (4.0) | 54 (1.6) | 0.001 |
| Frequency of symptoms | ||||
| Postprandial fullness | 74 (0.8) | 39 (0.7) | 35 (1.0) | 0.120 |
| Early satiation | 111 (1.2) | 51 (0.9) | 60 (1.7) | 0.001 |
| Epigastric pain or epigastric burning | 36 (0.4) | 12 (0.2) | 24 (0.7) | 0.001 |
| Functional dyspepsia, % | 168 (1.9) | 81 (1.5) | 87 (2.5) | 0.001 |
| EPS | 33 (0.4) | 12 (0.2) | 21 (0.6) | 0.003 |
| PDS | 146 (1.6) | 74 (1.4) | 72 (2.1) | 0.007 |
| Overlap | 14 (0.2) | 5 (0.1) | 9 (0.3) | 0.048 |
BMI, body mass index; ECG, electrocardiogram; EPS, epigastric pain syndrome; PDS, postprandial distress syndrome.
Data are presented as mean ± SD or n (%).
Table 2 shows the crude and adjusted ORs and 95% CIs for FD compared to BMI. The prevalence of FD among the lowest, low, moderate, and in high BMI groups was 3.1%, 1.9%, 1.3%, and 1.2%, respectively. After adjustment for age, sex, drinking, smoking, exercise habits, anemia, and sports injury, the lowest BMI was independently associated with FD (adjusted OR, 2.28; 95% CI, 1.46-3.67;
Table 2 . Crude and Adjusted Odds Ratios and 95% Confidence Intervals for Functional Dyspepsia in Relation to Body Mass Index
| Variable | Prevalence (%) | Crude OR (95% CI) | Adjusted OR (95% CI) |
|---|---|---|---|
| FD | |||
| Total | |||
| BMI ≤ 19.36 kg/m2 | 70/2231 (3.1) | 2.75 (1.77-4.40) | 2.28 (1.46-3.67) |
| 19.36 kg/m2 < BMI ≤ 20.90 kg/m2 | 42/2230 (1.9) | 1.63 (1.001-2.70) | 1.44 (0.88-2.39) |
| 20.90 kg/m2 < BMI ≤ 22.74 kg/m2 | 30/2232 (1.3) | 1.16 (0.68-1.97) | 0.98 (0.48-1.19) |
| 22.74 kg/m2 ≤ BMI | 26/2230 (1.2) | 1.00 | 1.00 |
| 0.001 | |||
| Men | |||
| BMI ≤ 19.56 kg/m2 | 28/1369 (2.1) | 1.66 (0.91-3.11) | 1.48 (0.81-2.78) |
| 19.56 kg/m2 < BMI ≤ 21.12 kg/m2 | 22/1370 (1.6) | 1.30 (0.69-2.49) | 1.19 (0.63-2.28) |
| 21.12 kg/m2 < BMI ≤ 23.12 kg/m2 | 14/1370 (1.0) | 0.82 (0.40-1.67) | 0.77 (0.37-1.57) |
| 23.12 kg/m2 ≤ BMI | 17/1369 (1.2) | 1.00 | 1.00 |
| 0.100 | |||
| Women | |||
| BMI ≤ 19.10 kg/m2 | 39/861 (4.5) | 3.10 (1.69-6.07) | 2.94 (1.59-5.77) |
| 19.10 kg/m2 < BMI ≤ 20.53 kg/m2 | 20/861 (2.3) | 1.55 (0.77-3.22) | 1.53 (0.76-3.17) |
| 20.53 kg/m2 < BMI ≤ 22.18 kg/m2 | 15/847 (1.7) | 1.16 (0.55-2.48) | 1.18 (0.55-2.53) |
| 22.18 kg/m2 ≤ BMI | 13/861 (1.5) | 1.00 | 1.00 |
| 0.001 |
Odds ratios adjusted for age, drinking, smoking, exercise habit, anemia, and sports injury.
BMI, body mass index; FD, functional dyspepsia.
The association between leanness (BMI < 18.5 kg/m2), overweight (25 ≤ BMI < 30 kg/m2), obesity (30 ≤ BMI kg/m2), and FD is shown in Table 3. In crude analysis, leanness was associated with FD in men and women (OR, 1.85; 95% CI, 1.02-3.18 and OR, 2.28; 95% CI, 1.42-3.59, respectively). After adjustment, leanness was independently associated with FD in women (OR, 2.19; 95% CI, 1.35-3.45;
Table 3 . Crude and Adjusted Odds Ratios and 95% Confidence Intervals for Functional Dyspepsia in Relation to Lean, Overweight, and Obese
| Variable | Prevalence (%) | Crude OR (95% CI) | Adjusted OR (95% CI) |
|---|---|---|---|
| FD | |||
| Total | |||
| BMI < 18.5 kg/m2 | 44/1241 (3.6) | 2.19 (1.52-3.09) | 2.01 (1.40-2.86) |
| 18.5 kg/m2 = BMI < 25 kg/m2 | 112/6778 (1.7) | 1.00 | 1.00 |
| 25 kg/m2 = BMI < 30 kg/m2 | 10/747 (1.3) | 0.81 (0.40-1.47) | 0.87 (0.43-1.60) |
| 30 kg/m2 = BMI | 2/157 (1.2) | 0.77 (0.13-2.45) | 0.95 (0.16-3.06) |
| 0.001 | |||
| Men | |||
| BMI < 18.5 kg/m2 | 16/656 (2.4) | 1.85 (1.02-3.18) | 1.74 (0.96-3.01) |
| 18.5 kg/m2 = BMI < 25 kg/m2 | 55/4131 (1.3) | 1.00 | 1.00 |
| 25 kg/m2 = BMI < 30 kg/m2 | 8/550 (1.4) | 1.08 (0.47-2.15) | 1.15 (0.51-2.30) |
| 30 kg/m2 = BMI | 2/133 (1.5) | 1.13 (0.18-3.69) | 1.35 (0.22-4.45) |
| 0.35 | |||
| Women | |||
| BMI < 18.5 kg/m2 | 28/585 (4.8) | 2.28 (1.42-3.59) | 2.19 (1.36-3.45) |
| 18.5 kg/m2 = BMI < 25 kg/m2 | 57/2647 (2.2) | 1.00 | 1.00 |
| 25 kg/m2 = BMI < 30 kg/m2 | 2/189 (1.1) | 0.49 (0.08-1.57) | 0.50 (0.08-1.63) |
| 30 kg/m2 = BMI | 0/24 (0.0) | N/A | N/A |
| 0.001 |
Odds ratios adjusted for age, drinking, smoking, exercise habit, anemia, and sports injury.
BMI, body mass index; FD, functional dyspepsia; N/A, not applicable.
This study evaluated the association between BMI and the prevalence of FD in young Japanese people. In the present study, BMI was independently inversely associated with FD based on the Rome III criteria. This is the first study to show the inverse association between BMI and FD in a young population.
There is limited evidence on the relationship between BMI and the prevalence of FD in Asian people. In a Japanese web-based study of 8038 participants who had no history of severe illness, FD based on the Rome III criteria with constipation was associated with a lower BMI among women.22 In another Japanese study of 7112 participants who underwent upper endoscopy examination for health screening, PDS, which is an FD subgroup, was significantly inversely associated with BMI.23 In a Taiwanese study of patients with GERD, BMI was independently inversely associated with FD.24 An Iranian cross-sectional study of 18 180 participants showed that the prevalence of FD without GERD was higher in subjects with a lower BMI.25 In a Malaysian cross-sectional study of 1002 young adults, leanness (BMI < 18.5 kg/m2) was significantly positively associated with FD.26 The findings in the present study were consistent with the results in previous studies that found an inverse association between BMI and FD.
In contrast, in the Asian population, some studies did not show the inverse association between BMI and FD. In a Korean case control study of 363 subjects that included 90 subjects with FD, higher visceral adiposity, not BMI, was associated with an increased risk of FD.27 In an Iranian cross-sectional study of 790 patients with gastrointestinal symptoms (mean age, 49.9 years; mean BMI, 25.4 kg/m2), BMI was not associated with FD.28 The Domestic/International Gastroenterology Surveillance Study showed that BMI positively associated with the prevalence of dyspepsia symptoms using an original questionnaire of the general population, including Japan.31 Discrepancies for BMI and FD in Asian people may be partially explained by differences in age, sex, distributions of BMI, prevalence of obesity, definition of FD, sample size, and confounding factors.
The underlying mechanism that links low BMI and FD remains unclear. In a study of patients with FD, weight loss was associated with visceral hypersensitivity and delayed gastric emptying.32 Similarly, delayed gastric emptying was found in participants with low BMI.33 In several animal models, activation of corticotrophin releasing factor (CRF) receptors may inhibit gastric emptying.34 Peripheral injection of CRF and related peptides inhibit gastric emptying in lean mice.35 A low BMI may cause FD via visceral hypersensitivity and delayed gastric emptying due to CRF.
The strengths of the present study were the sample size and the definition of FD based on the Rome III criteria. Our study had several limitations, however. First, this was a cross-sectional analysis. Second, we did not have access to medical records, including those related to medication and endoscopy. Therefore, an unknown digestive disease such as GERD, cancer, or ulcer may have caused FD. The exclusion of any reported organic disorder or concerning signs likely limited this bias, however. Third, due to several exclusion criteria, the prevalence of FD in this cohort may be lower than that in previous studies. FD in this cohort was symptomatic but untreated. Therefore, the severity of FD may be milder than that in previous studies. Fourth, nutritional evaluation data was not available in this cohort. As diet varies among cultures, further research is needed to examine the influence of nutrition on the association between BMI and FD. Fifth, the association between psychological disorders and FD has been reported.36,37 Data on psychological disorders, including depression and anxiety, were not collected in this cohort. Finally, the subjects in this study were not representative of Japanese young people. Notably, the prevalence of FD was far lower in this study than in previous population-based studies that have found inverse associations between exercise habits and FD. Similarly, the rates of smoking and drinking were lower in this cohort than in previous cohorts. The percentage of obese subjects was low in this cohort, and the distribution of BMI may affect the association between BMI and FD. Given that the present cohort consisted solely of university students, it is possible that the relatively high educational status of our population affected health behaviors, and that this was partly responsible for the low prevalence of FD in this study.
In conclusion, among Japanese young people, BMI may be independently inversely associated with FD. Leanness may be an independent associated factor for FD in young Japanese women.
The authors would like to acknowledge Katsutoshi Okada, Syuichi Saheki, Mikage Oiwa, Hiromi Miyauchi, Yuko Matsumoto, Takako Yamamoto, Hiroko Suzuki, Masumi Hino, Tomo Kogama, and all of the Health Services Center staff for their support.
None.
None.
Conception and design: Yasunori Yamamoto and Shinya Furukawa; material preparation and data collection: Aki Kato and Katsunori Kusumoto; data analysis: Shinya Furukawa and Yasunori Yamamoto; interpretation of data: Yasunori Yamamoto, Shinya Furukawa, Junichi Watanabe, Eiji Takeshita, Yoshio Ikeda, Naofumi Yamamoto, Katsuhiko Kohara, Yuka Saeki, and Yoichi Hiasa; the first draft of the manuscript was written by Yasunori Yamamoto and Shinya Furukawa; and supervision: Yuka Saeki, Katsunori Kusumoto, and Yoichi Hiasa. All authors read and approved the final manuscript.
 |
 |
