Journal of Gastroenterology

, Volume 44, Issue 6, pp 518–534

Epidemiology and clinical characteristics of GERD in the Japanese population

Authors

    • Department of GastroenterologyOsaka City University Graduate School of Medicine
  • Tetsuo Arakawa
    • Department of GastroenterologyOsaka City University Graduate School of Medicine
Review

DOI: 10.1007/s00535-009-0047-5

Cite this article as:
Fujiwara, Y. & Arakawa, T. J Gastroenterol (2009) 44: 518. doi:10.1007/s00535-009-0047-5

Abstract

We reviewed articles on the epidemiology and clinical characteristics of gastroesophageal reflux disease (GERD) in Japan to clarify these features of GERD in this country. Although the definition of GERD depends on the individual study, the prevalence of GERD has been increasing since the end of the 1990s. The reasons for the increase in the prevalence of GERD may be due to increases in gastric acid secretion, a decrease in the Helicobacter pylori infection rate, more attention being paid to GERD, and advances in the concept of GERD. More than half of GERD patients had non-erosive reflux disease, and the majority (87%) of erosive esophagitis was mild type, such as Los Angeles classification grade A and grade B. There were several identified risk factors, such as older age, obesity, and hiatal hernia. In particular, mild gastric atrophy and absence of H. pylori infection influence the characteristics of GERD in the Japanese population. We also discuss GERD in the elderly; asymptomatic GERD; the natural history of GERD; and associations between GERD and peptic ulcer disease and H. pylori eradication. We examined the prevalence of GERD in patients with specific diseases, and found a higher prevalence of GERD, compared with that in the general population, in patients with diabetes mellitus, those with obstructive sleep apnea syndrome, and those with bronchial asthma. We provide a comprehensive review of GERD in the Japanese population and raise several clinical issues.

Keywords

Gastroesophageal reflux disease Epidemiology Non-erosive reflux disease Hiatal hernia Gastric atrophy

Introduction

Gastroesophageal reflux disease (GERD) is caused by the abnormal reflux of the gastric contents into the esophagus and is now recognized as a common gastrointestinal (GI) disease in Japan. GERD has a multifactorial pathogenesis, including the presence of gastric acid [1] or bile [2], hiatal hernia [3], lower esophageal sphincter (LES) dysfunction [3, 4], esophageal motility dysfunction [4], impairment of esophageal epithelial resistance [5], and hypersensitivity [6]. Gastric acid in the reflux contents has major harmful effects on the esophageal mucosa and acid-suppressive drugs are effective in most cases of GERD [1]. Because several factors, including Helicobacter pylori infection and gastric atrophy, influence gastric acid secretion, these factors are inversely related to GERD [7]. Impairment of gastroesophageal barrier functions leads to reflux of the gastric contents into the esophagus. Hiatal hernia and transient LES relaxation play a major role in abnormal gastroesophageal reflux (GER) [3, 4]. Defensive factors such as esophageal clearance and epithelial resistance may be involved in the severity of GERD [4, 6].

GERD is diagnosed by esophago-gastro-duodenoscopy (EGD) and/or the presence of specific symptoms such as heartburn and acid regurgitation. There are several excellent reviews [811] of the epidemiology of GERD by Japanese researchers, but most of these studies are reviewed by limited publications, mainly in English. To clarify the situation in Japan, we reviewed articles on the epidemiology of GERD in the Japanese population published in both English and Japanese.

We examined epidemiological studies of GERD in the Japanese population by searching PubMed or Ichushi web in August 2008. Relevant articles were identified using the terms “Japanese” or “Japan” and “heartburn”, “esophagitis”, “GERD”, “non-erosive reflux disease”, or “esophageal ulcer”. We selected the articles (excluding abstracts) on the epidemiology and clinical characteristics of GERD in Japan or quoted other older papers in Japanese as needed. We collected and reviewed a total of 120 papers.

We reviewed the epidemiology of reflux esophagitis (erosive esophagitis) and symptomatic GERD in Japan, and the literature showed that the prevalence of GERD has been increasing from the end of the 1990s.

We also discuss GERD in the elderly; asymptomatic GERD; non-erosive reflux disease (NERD) and erosive reflux disease, the natural history of GERD; and finally, the prevalence of GERD among patients with specific diseases. However, we excluded postoperative GERD and infant GERD from this review.

Terminology used in GERD diagnosis

Several diagnostic terms used for GERD may cause some confusion. We define these diagnostic terms (Fig. 1) so that the present review can be better understood. When EGD is used as a diagnostic tool, any positive findings of esophagitis by any classification are defined as “reflux esophagitis” Patients with grade A–D but not grade M by the modified Los Angeles (LA) classification [12] are diagnosed as having “erosive esophagitis” (Fig. 1a). Patients in whom GERD symptoms, such as heartburn, are present, or those who have positive scores in questionnaires specific for GERD are diagnosed as having “symptomatic GERD” (Fig. 1b). When both EGD and GERD symptoms (or questionnaires) are used as diagnostic tools, persons in whom GERD symptoms are present but in whom esophageal mucosal injury is absent are diagnosed as having “non-erosive reflux disease (NERD).” In these cases, grade M by the modified LA classification is included. Persons in whom GERD symptoms are present and who have erosive esophagitis (LA classification grade A or more) are diagnosed as having “erosive reflux disease (ERD)”, and persons with positive findings of esophagitis by EGD but without symptoms are diagnosed has having “asymptomatic ERD” (Fig. 1c).
https://static-content.springer.com/image/art%3A10.1007%2Fs00535-009-0047-5/MediaObjects/535_2009_47_Fig1_HTML.gif
Fig. 1

Definition of diagnostic terminology of gastroesophageal reflux (GERD). a Cases diagnosed by esophago-gastro-duodenoscopy (EGD). b Cases diagnosed by GERD symptoms or questionnaires. c Cases diagnosed by both EGD and GERD symptoms. LA Los Angeles, ERD erosive reflux disease, NERD non-erosive reflux disease

Prevalence and clinical characteristics of reflux (or erosive) esophagitis in Japan

First, we analyzed GERD in the Japanese population as diagnosed by EGD. There are several classifications of reflux esophagitis but most studies have used the LA classification. Although it was reported that observer variations, dependent on the level of endoscopic experience, occurred during the early period of the introduction of the LA classification to Japan [13], the modified LA classification including grade M [12] has been commonly used more recently for the diagnosis of reflux esophagitis. Grade M is defined as minimal changes to the esophageal mucosa, such as reddish erythema and/or whitish turbidity [12]. Although the definition of reflux esophagitis depends on the individual study, GERD defined as LA classification grade A or more, termed “erosive esophagitis,” could exclude interobserver bias because of the extremely poor agreement in the endoscopic diagnosis of grade M [14, 15].

There were 30 studies [1645] on the prevalence of reflux esophagitis in outpatients and 12 studies [4657] on subjects who underwent regular health check-ups (Table 1). The prevalence of reflux esophagitis ranged from 1.4 to 52.1%. This wide range persists even when the studies that included erosive esophagitis (LA classification grade A or more) in their analysis were selected. The reasons for the wide range of reported prevalence for reflux esophagitis might be due to the study subjects (outpatients versus healthy subjects, or their ages), the period when the subjects were enrolled, and the area where the study was conducted. As expected, the prevalence of reflux esophagitis among outpatients was relatively higher than that among subjects who underwent regular health check-ups (mean prevalence for the entire data, 10.6% in outpatients and 7.6% in persons who underwent regular health check-ups).
Table 1

Prevalence of reflux esophagitis in Japan

Subjects

First author

Study period

Diagnosis

Prevalence

Outpatients

Endo et al. [16]

1966–1971

RE

145/4169 (3.5%)

Makuuchi [17]

1967–1976

JSED grade II

10/127b (7.9%)

Makuuchi [17]

1967–1977

JSED grade II

88/441c (20.0%)

Yoshimori et al. [18]

–1974

RE

159/2400 (6.6%)

Furuya et al. [19]

1975–1989

RE or Ulcers

512/31186 (1.6%)

Tsuchiya et al. [20]

1976–1985

RE or Ulcers

482/30609 (1.6%)

Sakurai [21]

1978–1997

LA (including grade M)

2341/119887 (2.0%)

Kuma et al. [22]

1983–1989

SM

115/7714 (1.5%)

Yamaguchi et al. [23]

1986–1995

LA

1661/35745 (4.6%)

Keida et al. [24]

1989–2002

LA

1889/24858 (7.6%)

Aizawa et al. [25]

1989–1990

Ulcer or erosion

19/1368 (1.4%)

Yamaguchi et al. [26]

1990–2004

LA

23/1621d (1.4%)

Maekawa et al. [27]

1993–1996

LA

119/2278 (5.2%)

Tomiyama et al. [28]

1995–1997

LA

100/3721 (2.7%)

Furukawa et al. [29]

1996–1998

LA

977/6010e (16.3%)

Okamoto et al. [30]

1996–1998

LA

1199/8031 (14.9%)

Sakamoto et al. [31]

1998

LA

302/14367 (2.1%)

Kawai et al. [32]

1998

LA

103/2690 (3.8%)

Iwakiri et al. [33]

1998–2004

LA

214/1897 (11.3%)

Inamori et al. [34]

1999

LA

54/392 (13.8%)

Inaba et al. [35]

1999

LA

115/825 (13.9%)

Ohta et al. [36]

1999

LA (grade B or more)

26/232 (11.2%)

Morichika et al. [37]

1999–2004

LA

414/794 (52.1%)

Sekiguchi et al. [38]

2000

LA (including grade M)

298/1496 (19.9%)

Fujiwara et al. [39]

2000–2001

LA

42/548 (7.7%)

Nagoshi et al. [40]

2000–2001

LA (including grade M)

725/8239 (8.8%)

Amano et al. [41]

2000–2002

LA (including grade M)

750/3219 (23.3%)

Shimazu et al. [42]

2000–2003

LA

72/1234 (5.8%)

Ohara et al. [43]

2003

LA

602/3608 (16.7%)

Keida et al. [24]

2003–2004

LA

556/4031 (13.8%)

Nagoshi et al. [44]

2005

LA (including grade M)

580/2383 (24.3%)

Sakaguchi et al. [45]

2005–2006

LA

202/1813 (11.1%)

Regular health check-up

Amano et al. [46]

1995–1998

LA

274/2788 (9.8%)

Yamaji et al. [47]

1996–1997

SM

108/5732 (1.9%)

Fujishiro et al. [48]

1996–2000

LA

69/781 (8.8%)

Uetake et al. [49]

1997

RE

132/1986 (6.5%)

Sekiguchi et al. [50]

(1997)a

RE

415/31952 (1.3%)

Moki et al. [51]

1998–2002

LA

191/5159 (3.7%)

Nakamura et al. [52]

1999

LA

74/539 (13.7%)

Furuta et al. [53]

2002–2003

LA

143/1683 (8.5%)

Yagi et al. [54]

2003–2004

LA

490/3818 (12.8%)

Kobayashi et al. [55]

2004–2005

LA

84/691 (12.2%)

Mishima et al. [56]

(2005)a

LA

195/2760 (7.1%)

Funatsu et al. [57]

2005–2006

LA

32/659 (4.9%)

RE reflux esophagitis (classification used was not available), JSED classification of Japanese Society for Esophageal Diseases (1973), LA Los Angeles classification, SM Savary-Miller classification

aPublication year

bSubjects without hiatal hernia

cSubjects with hiatal hernia

dSubjects who underwent emergency endoscopy

eIncluded 1616 subjects who underwent regular health check-up

Twenty-five studies [21, 23, 2635, 3741, 43, 44, 48, 52, 55, 56, 58, 59] described the severity of reflux esophagitis according to the LA classification. Because interobserver agreement on grade M is extremely poor, [14, 15] and because the two large studies[29, 30] combined cases of grade C and grade D, and because the LA classification for grade C and grade D was revised in 1999, [60] we analyzed the severity of erosive esophagitis as grade A, grade B, and grade C + D. A total of 9782 cases with erosive esophagitis were analyzed. Of these cases, the numbers with grade A, grade B, and grade C + D were 5338 (54.6%), 3169 (32.4%), and 1275 (13.0%), respectively (Fig. 2). The majority of erosive esophagitis cases in Japan were mild types such as grade A or grade B, accounting for 87%. In the studies which separated grade C and grade D, the prevalence of grade C was more than double that of grade D (grade C, 7.4% and grade D, 3.0%).
https://static-content.springer.com/image/art%3A10.1007%2Fs00535-009-0047-5/MediaObjects/535_2009_47_Fig2_HTML.gif
Fig. 2

Grades of severity of erosive esophagitis in Japan according to the LA classification

Several risk factors for reflux esophagitis have been identified. Most studies agree that the presence of hiatal hernia [17, 18, 22, 29, 34, 46, 48, 52, 55, 56], higher body mass index (BMI) or obesity [29, 37, 45, 51, 55], older age [19, 23, 29, 33, 34, 37, 46], or mild gastric atrophy [33, 34, 3739, 46, 47, 52, 55, 56], assessed by either higher serum pepsinogen level or closed-type atrophic gastritis by the Kimura-Takemoto classification [61] during EGD, and possibly the absence of H. pylori infection [38, 47, 48] are risk factors for reflux esophagitis. Some studies identified male gender as a risk factor for reflux esophagitis [22, 42, 46, 51, 56, 57] but others failed to find a significant gender difference [19, 52]. Recent interesting studies have shown an association between reflux esophagitis and metabolic syndrome [51, 57], which is recognized as being associated with several GI diseases [62]. Because the H. pylori infection rate is low in younger and middle-aged persons [63], and because most cases of atrophic gastritis are due to H. pylori infection [64], and because the Japanese lifestyle has become increasingly westernized, the risk factors for reflux esophagitis are also changing.

Prevalence and clinical characteristics of symptomatic GERD in Japan

Heartburn is a specific symptom of GERD; however, there are several dilemmas regarding the symptom-based diagnosis of GERD. First, there are differences in the recognition of heartburn among GERD patients. Manabe et al. [65]. showed that, in Japan, recognition of GERD did not differ between patients with erosive esophagitis and physicians, whereas NERD patients did not recognize a “burning sensation in the chest” as heartburn as often as physicians, while confusing “stomach ache” with heartburn. Second, it is uncertain how the frequency and severity of the experienced heartburn influences the diagnosis of GERD. Most researchers agree that GERD can be diagnosed when a person experiences heartburn at least twice weekly and these symptoms disturb their daily life. Disturbance of daily life is usually assessed by administering a questionnaire on health-related quality of life (HR-QOL). Three studies using the Medical Outcomes Study Short Form-36 (SF-36) demonstrated that HR-QOL in GERD patients was lower than that in the general Japanese population. [6668] A recent study by Hongo et al. [69] using a new questionnaire (QOLRAD) which has been developed as a specific QOL assessment of GERD, showed impairment of HR-QOL in GERD patients.

To achieve an accurate symptom-based diagnosis of GERD, several questionnaires have been established. The Carlsson-Dent self-administered questionnaire (QUEST) [70] has been translated to create a Japanese version. A study by the Osaka GERD Society showed a sensitivity of 72 and 65% and a specificity of 54 and 74% when the cutoff was set at 4 or 6 points, respectively [71]. A frequency scale for symptoms of GERD (FSSG) was recently developed by Kusano et al. [72]. It includes 12 symptoms and uses a 6-point Likert scale. The response scale is designed to measure the frequency of symptoms a patient has experienced (never, occasionally, sometimes, often, and always). When the cutoff was set at 8 points, FSSG showed a sensitivity of 62% and a specificity of 59% [72, 73].

Table 2 presents a summary of the prevalence of symptomatic GERD in Japan [30, 43, 45, 5457, 7481]. The diagnosis of symptomatic GERD in ten studies [30, 43, 45, 54, 7479] was based on the presence of heartburn, while five other studies [5557, 80, 81] used questionnaires (3 QUEST and 2 FSSG). There were several differences in the diagnostic criteria among the studies, including the frequency, severity, and duration of heartburn; the presence or absence of a detailed explanation of heartburn; face-to-face interview versus self-report questionnaire; and the setting of the cutoff score. The subjects of ten studies [54, 56, 57, 7480] were an unselected population, mainly persons who underwent regular health check-ups, while in five studies [30, 43, 45, 55, 81] the subjects were outpatients. Although the prevalence of symptomatic GERD ranged from 6.6 to 37.6%, the mean prevalence of GERD was 11.5% (3216/27 870) when GERD was defined as the presence of heartburn at least twice weekly [43, 45, 54, 7678]. A relatively higher prevalence of GERD was found in a study using a questionnaire. Watanabe et al. [81] showed that patients rarely visited general physicians with heartburn as the chief complaint, although about 37% of patients were positive for FSSG.
Table 2

Prevalence of symptomatic GERD in Japan

Authors

Study period

Diagnostic criteria

Prevalence

Stanghellini [74]

1996

Heartburn ≥1/W, ≥moderate severity, during 3 monthsf

49/500 (9.8%)

Kato et al. [75]

1996

Heartburn ≥1/W

173/1662 (10.4%)

Okamoto et al. [30]

1996–1998

Heartburn presentf

2223/8031a (27.7%)

Fujiwara et al. [76]

2001

Heartburng ≥2/W during 1 year

399/6035 (6.6%)

Watanabe et al. [77]

2001

Heartburng ≥2/W during 1 year

276/4095 (6.7%)

Kubota et al. [78]

2002

Heartburn ≥2/W

618/7386 (8.4%)

Yamagishi et al. [79]

2003

Heartburn present (usually or sometimes) during 1 month

30345/160983 (18.8%)

Ohara et al. [43]

2003

Heartburng ≥2/W

725/4723b (15.4%)

Yagi et al. [54]

2003–2004

Heartburn ≥2/W

929/3818 (24.3%)

Kobayashi et al. [55]

2004–2005

QUEST

128/691c (18.5%)

Mishima et al. [56]

(2005)h

QUEST

351/2760 (12.7%)

Sudou et al. [80]

2005

QUEST

148/869 (17.0%)

Watanabe et al. [81]

2005–2006

FSSG

1554/4139d (37.6%)

Sakaguchi et al. [45]

2005–2006

Heartburn ≥2/W

269/1813e (14.8%)

Funatsu et al. [57]

2005–2006

FSSG

178/659 (27%)

aOutpatients who underwent regular health check-up or had GI symptoms

bFirst-visit outpatients

cOutpatients

dPatients who visited a GP

eOutpatients with GI symptoms

fFace-to-face interview

gDetailed explanation of heartburn was included in the questionnaire given to patients

hPublication year

Large studies by Yamagishi et al. [79] showed a higher prevalence of symptomatic GERD in elderly women (≥60 years) than men but other studies failed to find a gender difference in terms of the prevalence of symptomatic GERD [75, 76]. Several studies identified that obesity (or weight gain) was a risk for symptomatic GERD [43, 55, 57, 75, 78], and diets containing such foods as sweet cake, rice cake, and fatty or spicy food exacerbated heartburn [43, 56, 75]. Other factors, such as the use of nonsteroidal anti-inflammatory drugs (NSAIDs) or anti-asthmatic drugs [43], irregular lifestyle, overeating, and the presence of stress [54], alcohol intake and smoking habit in male workers [77], and H. pylori infection in younger persons [78] were associated with symptomatic GERD in all the reports.

Recently, functional heartburn was proposed as one of the functional GI disorders [82]. Although the definition of functional heartburn is not well established, the Rome III criteria suggest that functional heartburn is defined as the presence of heartburn but no evidence of abnormal acid reflux and a positive-symptom index by ambulatory 24-h pH monitoring, and no response to treatment with double doses of proton-pump inhibitors (PPIs) [82]. Therefore, symptomatic GERD might include some cases of functional heartburn. However, it is difficult to distinguish functional heartburn from symptomatic GERD, especially in large studies where it is almost impossible to carry out pH monitoring. Novel biomarkers or diagnostic tools specific for functional heartburn will help us to distinguish these two diseases in the near future.

Increasing prevalence of GERD in Japan

All studies included in Tables 1 and 2 are summarized in Fig. 3 in relation to the study period. The relatively higher prevalence of GERD before 1980 might be due to selection bias and the definition of GERD, but the prevalence of GERD began to increase from the end of the 1990s. Five studies [20, 21, 23, 24, 49] on the prevalence of GERD with study periods of at least 10 years in the same hospitals or area showed a pattern similar to that of the entire data. The precise reasons for the increasing prevalence of GERD in Japan, especially from the end of the 1990s, are unknown. Possible reasons or mechanisms are described below.
https://static-content.springer.com/image/art%3A10.1007%2Fs00535-009-0047-5/MediaObjects/535_2009_47_Fig3_HTML.gif
Fig. 3

Prevalence of GERD in relation to the study period. Fifty-seven studies on the prevalence of GERD were included. Bars show the range of the periods when the subjects were enrolled. Colors represent numbers of study subjects, with blue color indicating less than 100; black 100–5000; and red, more than 5000

Increase in gastric acid secretion

Kinoshita et al. [83] examined gastric acid secretion in Japanese subjects in both the 1970s and 1990s. They found that basal and stimulated gastric acid secretion had increased over the past 20 years in both the elderly and non-elderly, irrespective of H. pylori infection. Since gastric acid plays a major role in the pathogenesis of GERD, an increase in gastric acid secretion in the Japanese affects the epidemiology of GERD.

Decrease in H. pylori infection rates

The prevalence of H. pylori infection is higher in Japan compared with western countries but the rate of infection is decreasing. Sugiyama et al. [63] showed that H. pylori seropositivity was 74.9% in asymptomatic subjects born before 1950 and 20.7% in those born after 1950 in the general population. Because H. pylori infection is inversely related to GERD, a decrease in the infection rate is associated with an increasing prevalence of GERD in the Japanese population, especially in younger and middle-aged persons.

More attention being paid to GERD

There are three PPIs available in Japan. Omeprazole, lansoprazole, and rabeprazole appeared on the market in 1991, 1992, and 1997, respectively. The increasing use of PPIs in the clinical field has led to a greater recognition of GERD. The number of manuscripts, reviews, and themes of symposia related to GERD began increasing in the 1990s. Especially, the first large epidemiological study, published in 1999, [29] which is the most cited Japanese paper in this field, had an influence on Japanese gastroenterologists. Finally, the Japanese GERD Society was established in 1996 and the number of members is increasing. These factors have led to Japanese doctors paying more attention to GERD.

Advance in the concept of GERD

The modified LA classification has been commonly used for the diagnosis of GERD. Although a consensus about the diagnosis of grade M has yet to be reached, an expanded concept of GERD including NERD could affect the reported increase in GERD prevalence. This is further supported by advances in videoendoscopy, enabling the detection of minimal esophageal changes or tiny superficial esophageal mucosal injuries.

Proportions and clinical characteristics of NERD and ERD in Japan

GERD is divided into two subtypes, NERD and ERD, characterized by the presence (ERD) or absence of breaks in the mucosa (NERD). Ten studies [40, 4345, 5458, 84] show the proportions of NERD and ERD in GERD. The subjects of three studies [54, 56, 57] were persons who underwent regular health check-ups, while the subjects of the other studies [40, 4345, 55, 58, 84] were outpatients. Figure 4 shows the proportions of NERD and ERD in Japan. Among a total of 5022 subjects, NERD was identified in 2944 (58.6%) and ERD in 2078 (41.4%). A higher proportion of NERD among GERD patients was more prominent in the unselected subjects who underwent routine health check-ups, with 1205 (80.4%) of 1490 subjects with GERD. Several studies showed that NERD was found more frequently in females [5458, 84] and in younger persons [55], and was associated with the absence of hiatal hernia [54, 56, 58, 84], and with lower BMI [54, 5658, 84] compared with ERD among subjects with GERD. These clinical characteristics of NERD are also found in western countries; however, more severe gastric atrophy [54, 56, 84] and the higher rate of H. pylori infection [84] compared with ERD are specific characteristics of Japanese NERD. Joh et al. [68] demonstrated that the frequency of abnormal acid reflux with NERD was higher in patients with minimal changes than in patients without such changes. Thus, if gastric acid secretion lead to the development of NERD (grade M or grade N) or ERD, these clinical features found in Japanese NERD could agree with the theoretically advanced features. Other factors, such as smoking habit [55, 58, 84] or a low incidence of metabolic syndrome [57] are also suggested. The clinical characteristics of ERD are virtually the same as those of reflux esophagitis. Finally, as we mentioned above, the possibility that functional heartburn is included in NERD should be noted.
https://static-content.springer.com/image/art%3A10.1007%2Fs00535-009-0047-5/MediaObjects/535_2009_47_Fig4_HTML.gif
Fig. 4

Proportions of NERD and ERD among Japanese GERD patients

GERD in elderly Japanese

We should discuss the epidemiology and clinical features of GERD in the elderly because Japan is known for the longevity of its people. Several studies showed a higher prevalence of GERD in the elderly, and the clinical characteristics in this group. First, GERD is predominant in females [8588]. Women in Japan live longer than men (mean lifespan in 2006 was 85.8 years for women and 79.0 years for men according to a report from the Ministry of Health, Labour and Welfare [89]) and osteoporosis and kyphosis commonly found in elderly women may be a factor in the higher prevalence of GERD in elderly women [90]. In addition, Kusano et al. [91] demonstrated that the presence of kyphosis correlated with hernia size in elderly women. Second, the proportion of severe esophagitis is high in the elderly [27, 29, 59, 92]. Furukawa et al. [29] showed that severe esophagitis was found in women more than 70 years old and in men more than 80 years old. The mechanisms responsible for the higher proportion of severe esophagitis in the elderly are unknown, but impairment of esophageal motility in the elderly [92, 93] might be a factor. Third, the presence of hiatal hernia is a risk factor for GERD in both the elderly and non-elderly, but the incidence (46–72%) of hiatal hernia in the elderly is extremely high [27, 29, 87, 9294], especially in women [29]. In addition, the prevalence of hiatal hernia increases with age [95]. Fourth, Ohara et al. [96] demonstrated that the rate of H. pylori infection in elderly GERD subjects was lower (39%) than that in elderly non-GERD subjects (63%).They found that the serum pepsinogen I/II ratio in elderly GERD subjects was significantly higher than that in elderly non-GERD subjects, revealing mild gastric atrophy, [96] although Urita et al. [85, 86] showed that the incidence of open-type atrophic gastritis was higher in elderly GERD subjects compared with non-elderly GERD subjects. Because the H. pylori infection rate and the presence or severity of gastric atrophy increases with age, clearer associations between GERD and H. pylori-negative or mild gastric atrophy appear in the elderly [85, 86, 96]. Finally, it is a clinically important issue that 24–54% of elderly GERD patients are asymptomatic [27, 86, 87], especially those with mild esophagitis [27]. Sometimes GI bleeding such as hematemesis or tarry stool is the first episode seen in the elderly, accounting for 13.7–33.3% [8688]. Although GERD is rarely (1.2%) seen as a cause of upper GI bleeding in emergency endoscopy units [26], clinicians should consider that GERD is a potential differential diagnosis in cases of acute GI bleeding in elderly Japanese.

Asymptomatic ERD in Japan

As described above, asymptomatic ERD is one of the clinical features in elderly GERD subjects; thus, we reviewed asymptomatic ERD in the general adult population, with results reported in this section. Individuals often experience asymptomatic erosive esophagitis, which is frequently found incidentally during regular health check-ups or upon further examination for other reasons during an EGD examination. Two large studies showed that 144 (73.8%) out of 195 subjects with erosive esophagitis (LA grade ≥ A) were asymptomatic when assessed by QUEST (cutoff score 6) [56], and 288 (58.8%) of 490 subjects with erosive esophagitis (LA grade ≥ A) were asymptomatic when this was defined as less than two episodes of heartburn per week [54]. The high prevalence of asymptomatic ERD might be due to the definition of asymptomatic (high cutoff score on QUEST and high frequency of heartburn) or the study subjects (persons undergoing a regular health check-up). Nozu and Komiyama [59] demonstrated that 23 (26.4%) of 87 outpatients with erosive esophagitis were asymptomatic when this was defined as the absence of either typical symptoms such as heartburn and acid regurgitation, or atypical symptoms such as epigastric pain, discomfort, dysphagia, cough, or globus. They also found that smoking habit, male gender, and lower BMI were independent risk factors for asymptomatic ERD [59]. Further study of whether these factors are actually associated with asymptomatic ERD is required because of the small sample size of the Nozu study. In addition, no detailed natural history of asymptomatic ERD exists.

Natural history of GERD in Japan

Three questions about the natural history of GERD are suggested: (1) How often do non-GERD subjects develop GERD? (2) Is there a natural sequence of NERD-mild esophagitis-severe esophagitis among GERD patients? (3) How often do GERD patients develop (or progress) to Barrett’s esophagus? A few studies have provided information on the natural history of GERD in the Japanese population.

Three studies demonstrated the incidence of GERD development among non-GERD subjects. Miyamoto et al. [97] showed 37 (15.4%) of 241 elderly subjects developed GERD as diagnosed by QUEST during a 6-year follow-up period, and they identified absence of H. pylori infection, constipation, and the use of calcium-channel antagonists as risk factors for the development of GERD. Similarly Azumi et al. [98] reported 35 (8%) of 451 subjects (mean age, 47 years) who developed GERD during a 5-year follow-up period. Kawanishi [99] demonstrated that 51 (11.3%) of 450 subjects (mean age, 48.0 years) developed erosive esophagitis (47 with grade A and 4 with grade B) during a 5-year follow-up period. These studies suggest that about 10% of non-GERD adults go on to develop GERD, with this being more frequent in the elderly. Azumi et al. [98] demonstrated that 65 (74%) out of 88 subjects who were QUEST-positive at initial examination became QUEST-negative after 5 years, suggesting that, there are some GERD populations who improve spontaneously. Because GERD goes through a cycle of remission and recurrence, there are some difficulties in defining precisely what constitutes a new development of GERD.

Kawanishi [99] elucidated the natural history of NERD as diagnosed by the occurrence of heartburn at least twice weekly without breaks in the esophageal mucosa. He found that 17 (36.2%) of 47 NERD patients developed erosive esophagitis, including 15 with grade A and 2 with grade B esophagitis during a 5-year follow-up period [99]. The incidence of esophagitis was higher compared with that in non-NERD subjects (11.3%), and the absence of H. pylori infection, the absence of gastric atrophy, an increase in BMI, and elevated triglycerides were risk factors for the development of erosive esophagitis among NERD subjects. An impressive study by Manabe et al. [100] elucidated the natural history of mild esophagitis including grade A and grade B. Using EGD, they followed 105 patients with mild esophagitis for 5 years. Only 11 (10.5%) of 105 progressed to severe esophagitis, including 9 with grade C and 2 with grade D, while esophagitis resolved in 31, and 63 showed no progression of esophagitis. They identified increased age, female gender, GERD symptoms at initial examination, the presence of hiatal hernia, the absence of gastric atrophy, and the absence of H. pylori infection as risk factors for progression of esophagitis.

Limited available data about the natural history of GERD in the Japanese population suggest that a subpopulation, especially those who are H. pylori-negative or those in whom gastric atrophy is absent, progress to more severe types of GERD, but the rate of such progression may be low. The rate of progression to Barrett’s esophagus in Japanese GERD patients is still unknown.

Association between peptic ulcer disease and reflux esophagitis

A strong association between peptic ulcer disease and reflux esophagitis had been reported in Japan before the H. pylori era. Table 3 shows the co-incidence of peptic ulcer diseases among patients with reflux esophagitis [1823, 25, 101, 102]. Co-incidence of duodenal ulcer, gastric ulcer, and duodenogastric ulcer was found in 9.6–31.6%, 6.0–20.1%, and 3.5–5.4% of patients with reflux esophagitis, respectively. The high co-incidence of peptic ulcer diseases might be due to the high prevalence of peptic ulcer disease in Japan before the H. pylori era. The presence of duodenal ulcer might be related to the hypersecretion of gastric acid and/or delay in gastric emptying because of stenosis due to edema or deformity of the duodenal bulb. Interestingly, Amano et al. [41] demonstrated that elderly patients (65 years or older) with duodenal ulcer or distal gastric ulcer had a significantly higher prevalence of reflux esophagitis, including LA grade M (33.3 and 31.7%, respectively), than those with proximal gastric ulcer (18.0%), but these differences were not observed among younger patients. This finding might be related to the role of hypersecretion of gastric acid in elderly GERD patients. Because H. pylori eradication therapy is the first-line treatment for peptic ulcer disease and prevents ulcer recurrence, the rate of co-incidence of peptic ulcer diseases in patients with reflux esophagitis has been decreasing. Associations between reflux esophagitis and NSAID- or aspirin-induced duodenogastric ulcer are unknown. The effects of H pylori eradication therapy on pre-existing GERD are discussed below.
Table 3

Co-incidence of peptic ulcer disease among GERD patients

 

Study period

Number of GERD cases

Duodenal ulcer

Gastric ulcer

Duodenogastric ulcer

Yoshimori et al. [18]

–1974

159

19 (11.9%)

33 (20.1%)

 

Furuya et al. [19]

1975–1989

512

59 (11.5%)

75 (14.6%)

 

Yoshida et al. [101]

1976–1980

50

6 (12.0%)

3 (6.0%)

 

Tsuchiya et al. [20]

1976–1985

482

89 (18.5%)

65 (13.5%)

17 (3.5%)

Sakurai [21]

1978–1997

2439

9.6%

7.7%

 

Takaaki et al. [102]

1980–1990

112

12 (10.7%)

11 (9.8%)

6 (5.4%)

Kuma et al. [22]

1983–1989

115

16 (13.9%)

11 (9.6%)

 

Yamaguchi et al. [23]

1986–1995

1661

224 (13.5%)

224 (13.5%)

 

Aizawa et al. [24]

1989–1990

19

6 (31.6%)

3 (15.8%)

 

GERD and Helicobacter pylori eradication in Japan

An elegant systematic review by Nakajima and Hattori [103] is useful in understanding how GERD is related to H. pylori therapy. There are two proposed issues including recurrence of GERD after cure of infection and the effect of eradication therapy on pre-existing GERD.

The development of reflux esophagitis was found in 4.8–20.5% after cure of infection (Table 4) [104110]. Most of the newly developed reflux esophagitis was mild type. Two studies [104, 109] revealed a higher incidence of reflux esophagitis in patients with successful eradication compared with those with persistent infection (failure of eradication or matched control) but one study showed no difference [105]. The relatively wide range of incidence might be due to the study subjects (patients with duodenal ulcer, gastric ulcer, or other diseases) or the observation periods established to detect the development of GERD. Several reports suggested that the presence of hiatal hernia, corpus gastritis, and an increase in gastric acidity were risk factors for development of reflux esophagitis after cure of infection [104, 109]. Several mechanisms for the increased gastric acidity after cure of infection, such as a decrease in ammonia production, cytokines or hormones, and recovery of gastric inflammation of the corpus have been suggested. Whether an increase in gastric acidity alone is responsible for the development of GERD is unclear because of the lack of detailed associations between H. pylori infection and gastroesophageal barriers such as LES. Sasaki et al. [111] reported on the long-term observation of 45 patients with reflux esophagitis (all mild esophagitis such as grade A or grade B) that developed after H. pylori eradication. They found improvement in 78.8%, worsening in 8.9%, and no progression to severe type (grade C or grade D) 3 years after eradication. These findings suggest that most cases of newly developed GERD after H. pylori eradication are transient phenomena and eradication therapy rarely becomes a long-term clinical problem.
Table 4

Development of GERD after cure of H. pylori infection

 

Follow-up (months)

Incidence (%)

 

Successful eradication (%)

Failure of eradication

Erosive esophagitis

 Hamada et al. [104]

3 years

18a

(0.3%b)

 Murai et al. [105]

>6

5.4

5.2%

 Yachida et al. [106]

24

4.8a

 

 Koike et al. [107]

7

10.5

 

 Fukuchi et al. [108]

6

5.0

 

 Inoue et al. [109]

12

20.5

3.8%

GERD symptoms

 Murai et al. [105]

>6

4.6

0.0%

 Yamamori et al. [110]

12

9.7

 

aEstimated incidence within 3 years

bAge- and disease-matched H. pylori-positive controls

Murai et al. [105] showed that 4.6% of patients with successful H. pylori eradication had a recurrence of GERD symptoms after eradication. Yamamori et al. [110] demonstrated that approximately 10% of patients with peptic ulcer disease redeveloped GERD symptoms after cure of H. pylori infection, and age more than 70 years and gastric ulcer were associated with the recurrence. However, the development of GERD symptoms is complex because these symptoms were masked during anti-ulcer treatment such as by the administration of acid-suppressive drugs [103]. In addition, heartburn is seen in patients with other diseases, especially in those with functional dyspepsia. Thus, which disease (GERD or functional dyspepsia) appears after the cure of H. pylori infection in patients with peptic ulcer disease could not be distinguished.

The effect of H. pylori eradication therapy on pre-existing GERD in Japanese patients is summarized in Table 5 [105, 106, 112, 113]. Four studies of the effect of H. pylori eradication therapy on pre-existing GERD have been conducted in Japan. These studies suggest improvement of reflux esophagitis and GERD symptoms after cure of H. pylori infection. In particular, two studies, by Miwa et al. [112]. and Ishiki et al. [113], demonstrated a statistically significant improvement in GERD in a group cured of infection compared with a group with persistent infection. Ishiki et al. [113] found that duodenal ulcer, cure of H. pylori infection, the absence of hiatal hernia, and lower BMI were independently associated with improvement of reflux esophagitis after H. pylori eradication. How does cure of H. pylori infection improve GERD? There are several proposed mechanisms. First, normalization of gastric acidity in patients with duodenal ulcer [114] might play an important role. Second, as Nakajima and Hattori [103] suggested, symptoms such as heartburn and/or acid regurgitation are directly related to peptic ulcer disease; thus, symptoms are improved after the cure of H. pylori infection. Third, peptic ulcer disease, especially duodenal ulcer, is known to be co-incident with GERD (as described in the section above). If peptic ulcer itself directly or indirectly induces reflux esophagitis, healing of the ulcer results in improvement of GERD.
Table 5

Effect of H. pylori eradication on pre-existing GERD

 

Follow-up (months)

Cured infection

Persistent infection

 

n

Improvement

No change

Got worse

n

Improvement

No change

Got worse

Murai et al. [105]

>6

11

3 (27.3%)

6 (54.5%)

2 (18.2%)

    

Murai et al. [105]a

>6

36

34 (94.4%)

      

Yachida et al. [106]

24

27

5 (18.5%)

20 (74.1%)

2 (7.4%)

    

Miwa et al. [112]a

3 years

237

155 (65.4%)

70 (29.5%)

12 (5.1%)

237

72 (30.4%)

147 (62.0%)

18 (7.6)

Ishiki et al. [113]

22

120

73 (60.8%)

42 (35.0%)

5 (4.2%)

36

14 (38.9%)

18 (50.0%)

4 (11.1)

Ishiki et al. [113]a

22

38

25 (65.8%)

13 (34.2%)

 

11

5 (45.5%)

6 (54.5%)

 

aStudy of effect of H. pylori eradication on pre-existing GERD symptoms

Prevalence of GERD in patients with specific diseases

We reviewed the prevalence of GERD in Japanese patients with specific diseases, including diabetes mellitus, chronic liver disease, obstructive sleep apnea syndrome (OSAS), and bronchial asthma. We selected these diseases because there was a sufficient number of publications to review. Other diseases such as collagen disease and osteoporosis are well known as having a high co-incidence with GERD but were excluded in this review because few such Japanese epidemiological studies have been conducted. Non-cardiac chest pain, chronic cough, laryngitis, and dental disease, which are strongly associated with GERD [115], were excluded for the same reason.

Diabetes mellitus

Diabetes mellitus (DM) is common in Japanese adults, and diabetic patients often complain of GI symptoms. Because diabetes mellitus and GERD share similar risk factors such as obesity, and because diabetes mellitus affects autonomic nerve functions, a higher prevalence of GERD is to be expected in diabetic patients. Table 6 summarizes the prevalence of GERD in patients with diabetes mellitus [116121]. Nishida et al. [117] demonstrated a significantly higher prevalence of GERD in diabetic patients compared with controls, but Ariizumi et al. [121] showed no difference in the prevalence of reflux esophagitis between diabetic patients and controls. When QUEST was used for the diagnosis of GERD [116121], about 30% of diabetic patients were positive for GERD, showing a higher rate compared with the general adult population (see Table 2). Several studies showed that disease duration of diabetes [117, 120] and the presence of diabetic neuropathy [119, 120] were associated with GERD. Because esophageal motility disorder and abnormal acid reflux in diabetic patients are associated with diabetic motor neuropathy [122] and such esophageal dysfunction is worsened with long disease duration [123], esophageal dysfunction may result in a higher prevalence of GERD in diabetic patients. On the other hand, Kirizuka et al. [116] reported that the presence of hiatal hernia or the use of calcium-channel antagonists was more closely related to the incidence of GERD in diabetic patients than the duration or control of diabetes. Additional associated factors such as an increase in BMI or HbA1c level [117], the use of oral hypoglycemic agents [117], or constipation [119] were suggested as factors related to GERD incidence.
Table 6

Prevalence of GERD in patients with diabetes mellitus

 

GERD

Controls

Diagnosis

Prevalence

Kirizuka et al. [116]

 QUEST

51/163 (31.3%)

 

 LA

20/163 (12.3%)

 

Nishida et al. [117]

QUEST

61/241 (25.3%)

4/42b (9.5%)

Akiyama et al. [118]

QUEST

17/32 (53.1%)

 

Hisano et al. [119]

QUEST

22/77 (28.6%)

 

Kase et al. [120]

QUEST

71/156a (45.5%)

 

Ariizumi et al. [121]

EGD

15/85 (17.6%)

97/944c (10.3%)

QUEST

28/85 (32.9%)

 

a156 patients with GI symptoms among 531 diabetic patients

bPatients with chronic hepatitis due to HCV

cPatients without diabetes mellitus

There is another important issue concerning GERD in diabetic patients. Kinekawa et al. [124] examined 53 diabetic patients by QUEST immediately before 24-h pH monitoring. They found that diabetic patients had fewer symptoms and extremely low scores, and there was no difference in scores between patients with and without GER. Similarly, Hisano et al. [119] showed that diabetic patients with GERD had fewer symptoms than nondiabetic GERD patients. If GERD symptoms are masked in diabetic patients, the precise prevalence of GERD might actually be higher.

Chronic liver disease

Six studies [117, 125129] have reported an association between GERD and chronic liver disease (Table 7). First, Akatsu et al. [125] demonstrated that 8 (27.6%) of 29 patients developed reflux esophagitis after living-donor liver transplantation, although only 1 patients had reflux esophagitis before transplantation. The mechanisms of GERD development after living-donor liver transplantation are still unknown. The other five studies [117, 126129] showed the prevalence of GERD in patients with chronic liver disease using specific questionnaires such as QUEST or FSSG. Two studies[126, 128] showed a significantly higher prevalence of GERD (approximately 19%) in liver disease patients compared with controls, but the prevalence of GERD in controls was extremely low compared with that in the general adult population (see Tables 1, 2). Only one report, by Suzuki et al. [129] using QUEST, showed that more than 30% of patients with chronic liver disease had GERD. Thus, we could not conclude that the prevalence of GERD in Japanese patients with chronic liver disease was high. Most authors showed no differences in GERD prevalence according to the etiology or stage (chronic hepatitis or cirrhosis) of liver disease [126, 128, 129], but Ueda et al. [127] showed a relatively higher incidence of GERD in patients with alcoholic liver disease. Because drinking alcohol worsens or induces GERD symptoms, and because non-alcoholic steatohepatitis and GERD share common risk factors, such as metabolic syndrome [62], their associations might be examined in future. Because interferon affects gastric emptying [130], such antiviral therapy might affect the pathogenesis of GERD. Researchers should examine in more detail the associations between GERD and chronic liver disease. Clinically, erosive esophagitis is believed to be a risk factor for the rupture of esophageal varices. A recent study by Okamoto et al. [131]. showed that the positive predictor for bleeding from esophageal varices was the presence of a red color sign in the right anterior wall of the esophagus, where mucosal breaks induced by GERD were more frequently found, and the administration of a PPI was a negative predictor. These findings might explain an association between GERD and rupture of esophageal varices, but further study of this is needed.
Table 7

Prevalence of GERD in patients with chronic liver disease

 

GERD

Controls

Diagnosis

Prevalence

Akatsu et al. [125]

EGD

8/29a (27.6%)

 

Nishida et al. [117]

QUEST

4/42 (9.5%)

 

Kakizaki et al. [126]

FSSG

28/145 (19.3%)

3/40b (7.5%)

Ueda et al. [127]

FSSG

24/153 (15.3%)

 

Abe et al. [128]

FSSG

66/338 (19.5%)

1/37c (2.6%)

Suzuki et al. [129]

QUEST

80/238 (33.6%)

 

aPatients with living-donor liver transplantation

bVolunteers

cPatients with colon polyps

Obstructive sleep apnea syndrome (OSAS)

Suganuma et al. [132] first found a significant increase in the incidence of sleep disturbance in Japanese OSAS patients with GERD compared with OSAS patients without GERD. Subsequently, six studies [133138] on the associations between GERD and OSAS were reported (Table 8). The prevalence of GERD in OSAS patients was reported to be 19.2–42.1%. Uchiyama et al. [133], using 24-h pH monitoring, demonstrated no significant difference in GERD prevalence between sleep disturbance patients with OSAS, defined as a score on the apnea-hypopnea index (AHI) of 5 or more, and non-OSAS, defined as AHI scores of less than 5 (19.2% in OSAS patients and 15.4% in non-OSAS patients), while the other studies [134138] showed a higher prevalence, compared with the findings of Uchiyama et al. [133]. Tanaka et al. [137] showed a significant association between the prevalence of GERD and the severity of OSAS as assessed by AHI score, but Tanimura et al. [136] showed no such association. Three studies [133, 135, 138] demonstrated that nasal continuous positive airway pressure improved OSAS as well as GERD, but the sample size was small and the effect of GERD therapy, such as a PPI, on the improvement of OSAS was not reported in Japan. As shown in a review by Mizuta et al. [139], the association between GERD and OSAS remains controversial in Japan, as well as worldwide, because of failure to establish a causal link between the two diseases in a large study, inconsistencies in definition, similar risk factors for the two diseases (such as obesity), and several biases. However, the high prevalence of GERD in OSAS patients in Japan should be noted.
Table 8

Prevalence of GERD in patients with obstructive sleep apnea syndrome

Authors

GERD

Diagnosis

Prevalence

Uchiyama et al. [133]

pH monitoring

5/26 (19.2%)

Sugai et al. [134]

QUEST

133/320 (41.6%)

pH monitoring

12/50 (24.0%)

Sato et al. [135]

QUEST

25/73 (34.2%)

Tanimura et al. [136]

QUEST

53/126 (42.1%)

Tanaka et al. [137]

FSSG

45/143 (31.5%)

Sato et al. [138]

FSSG

42138 (30.4%)

Bronchial asthma

Bronchial asthma is one of the four extra-esophageal syndromes strongly associated with GERD in the Montreal definition and classification [115]. There are several proposed mechanisms of association between bronchial asthma and GERD. Vagally mediated reflex and microaspiration worsens asthma, while autonomic nerve disturbance and the use of bronchodilators affect GERD. Ten papers [43, 140148] identified an association between GERD and bronchial asthma, and most studies focused on the prevalence of GERD in asthmatics. Although the diagnostic criteria for GERD were different among these studies, the prevalence of GERD in asthmatics was high, ranging from 22.1 to 75.6% (Table 9). Several studies showed the efficacy of PPIs in the treatment of GERD and improvement of pulmonary function or asthma symptoms in asthmatic patients. PPIs attenuated GERD symptoms as well as asthma symptoms, but the efficacy of PPIs in improving pulmonary function remains controversial; improvement of peak expiratory flow was shown in three studies [142, 143, 145], but no change was observed in two studies [146, 147]. A further large study is needed. Although data are limited concerning the prevalence of asthma in GERD patients, the use of bronchodilators was reported to be significantly higher in persons with heartburn compared with those without heartburn [43].
Table 9

Prevalence of GERD in patients with bronchial asthma

Authors

GERD

Diagnosis

Prevalence

Suzuki et al. [140]

pH monitoring

42/58 (72.4%)

Tomioka et al. [141]

Heartburn

41/106 (38.7%)

LA

23/104 (22.1%)

Nakase et al. [142]

LA

20/72 (27.8%)

Tsugeno et al. [143]

LA or QUEST

25/94 (26.6%)

Shimizu et al. [144], [145]

LA

37/78 (47.6%)a

QUEST

59/78 (75.6%)

Sato et al. [146]

Heartburn

48/290 (32.0%)

Nogami et al. [147]

QUEST

25/65 (38.5%)

Takezawa [148]

QUEST

48/88 (54.5%)

a54/78 (69.2%) when grade M was included

In conclusion, we have provided a comprehensive review of the epidemiology and clinical characteristics of GERD in the Japanese population. Several factors are associated with the increase in the prevalence of GERD in Japan. Especially, H. pylori infection, gastric atrophy, and long life affect the epidemiology and clinical characteristics of GERD in the Japanese population at the present time.

Acknowledgments

This study was supported, in part, by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture in Japan.

Copyright information

© Springer 2009