



* Department of Pediatrics, Clinique de Pédiatrie Saint Antoine, Hôpital Saint Vincent de Paul, Catholic University, Lille, France
Department of Pediatric Gastroenterology, Hepatology, and Nutrition, Hôpital Jeanne de Flandre, Faculty of Medicine, Lille, France
| ABSTRACT |
|---|
|
|
|---|
Study Design. A prospective double-blind study was conducted between March 2001 and April 2002 in children at least 6 years old with NUD who had been referred for upper gastrointestinal endoscopy for epigastric pain. A standardized questionnaire was administered blindly by a pediatric gastroenterologist. This questionnaire characterized epigastric pain and associated factors. Infection was confirmed by positive culture and histologic examination of the gastric mucosa.
Results. From 100 children enrolled, 26 proved infected (12 female, 14 male; mean age: 11.4 ± 2.6 years), and 74 were noninfected (44 female, 30 male; mean age: 10.4 ± 3.1 years). There were no differences in age or symptom characteristics between groups except for epigastric pain during meals that was more frequent in noninfected than in infected children (25.6% vs 3.8%).
Conclusion. There were no specific characteristics of symptoms in nonulcer-dyspeptic H pyloriinfected children as compared with noninfected children.
Key Words: Helicobacter pylori nonulcer dyspepsia recurrent abdominal pain children
Abbreviations: RAP, recurrent abdominal pain NUD, nonulcer dyspepsia GI, gastrointestinal
The role of Helicobacter pylori in the colonization of the stomach in adults and children with chronic gastritis, peptic ulcer, and possibly gastric carcinomas is now well documented,1 and eradication of the bacteria is very effective in preventing peptic-ulcer relapses in both adults2 and children.3
Recurrent abdominal pain (RAP), according to Apley's criteria (ie, at least 3 discrete episodes of abdominal pain of sufficient severity to interrupt normal daily activities or performance, occurring over a period of
3 months4), and nonulcer dyspepsia (NUD), which refers to pain or discomfort centered in the upper abdomen,5 are difficult conditions to define in children, because many have imprecise symptoms. Moreover, the role and clinical manifestations of H pylori remain unclear in such children. Vomiting and acute abdominal pain related to ulcer disease may be associated with H pylori infection, whereas the role of this bacterium in children with RAP and NUD is the subject of conflicting reports. Localized epigastric pain6 and nocturnal awakening7,8 have been reported occasionally in children with this infection. However, a high incidence of H pylori infection in the course of RAP does not prove a causal relationship between this infection and abdominal pain.9
Recently, a European pediatric consensus recommended a search for H pylori infection using upper-gastrointestinal (GI) endoscopy with gastric biopsy in children suffering from upper-digestive symptoms suggestive of organic disease without any additional clear information on the nature of these symptoms.10
There have been several proposals to define criteria for functional GI dyspepsia in children mature enough to provide an accurate history of pain.5 Functional dyspepsia is subdivided into 3 forms: ulcer-like dyspepsia, characterized by a centered pain in the upper abdomen (epigastric pain); dysmotility-like dyspepsia, characterized by unpleasant discomfort centered in the upper abdomen or associated with upper abdominal fullness, early satiety, bloating, or nausea; and unspecified dyspepsia.5 The use of refined clinical characteristics of RAP could be of help in identifying a subgroup of patients with RAP, in whom H pylori infection might need investigation and treatment.11 The aim of this study was therefore to characterize the symptoms in H pyloriinfected children with NUD.
| PATIENTS AND METHODS |
|---|
|
|
|---|
6 years old were enrolled, with the assumption that they were mature enough to provide an accurate history of their pain characteristics. A figure showing the localization of the epigastric region was first shown and explained to all the enrolled children. Epigastric pain and tenderness were identified by a pediatric gastroenterologist after careful clinical assessment. Then a questionnaire was administered that asked the children to describe the characteristics of their epigastric pain, namely its intensity, using a visual analog scale (from 0- to 100-mm extreme ends; 0 indicated the absence of pain, and 100 indicated the highest degree of pain), and frequency (days per week). Other items included any occurrence of daytime or nocturnal awakening; epigastric tenderness; relationship of pain episodes with eating (before, during, or after); nausea, episodes of vomiting, or weight loss (>5% of the last body weight). The questionnaire also recorded any episodes of hematemesis, any drugs taken during the last month (antispasmodics or analgesics), any school absences, and any family history of peptic-ulcer disease.
The socioeconomic levels of the enrolled children and the numbers of people in their household were recorded. The ethnic background of each child was defined as European ("white"), North African, Middle Eastern, African, Asian, or Far Eastern, according to the mother's place of birth. The mother's education level was defined as unschooled, primary plus secondary school (high school), or tertiary. The father's socioeconomic level was defined as unemployed, low, intermediate, or high according to the French national social classification system.
We excluded children <6 years old; those who had already suffered gastric H pylori infections; institutionalized encephalopathic children; and children who had received antibiotics, acid-suppressing medications, or a nonsteroidal antiinflammatory drug during the month preceding evaluation.
H pylori infection was confirmed by positive culture as described12 and by histologic examination (Sydney classification) of gastric antral and fundic biopsy specimens. Noninfected children were defined as those who exhibited negative H pylori cultures and negative histologic findings from their gastric mucosa. Cultures and histologic examinations of biopsy samples were conducted blindly. Informed consent was obtained from the parents of all the enrolled patients.
All statistical tests were performed by using StatView software (Abacus, CA). Means, SDs, medians, and extremes were calculated for all quantitative parameters. Differences between qualitative parameters were analyzed by using
2 tests. A P value of < .05 was taken as statistically significant.
| RESULTS |
|---|
|
|
|---|
There were no differences between centers in the prevalence of H pylori infection or the distributions of age, gender ratio, symptom characteristics, education outcome, or socioeconomic levels. There were no children presenting with ulcer disease during the study period.
There were no differences for most of the symptom characteristics when comparing infected with noninfected children. These characteristics included histories of nausea (12 of 26 [46.1%] vs 35 of 74 [47.2%]), vomiting (10 of 26 [38.1%] vs 27 of 74 [36.1%]), weight loss (1 of 26 [3.8%] vs 6 of 74 [8.1%]), hematemesis (1 of 26 [3.8%] vs 2 of 74 [2.7%]), no drugs being taken (10 [38.1%] vs 36 [48.6%]), a positive family history of peptic-ulcer disease (16 of 26 [61.5%] vs 36 of 74 [48.6%]), or durations of school absence (2.6 ± 4.2 vs 4.5 ± 10.1 days) (Table 1).
|
As expected, H pylori infection was more frequent in nonwhite than in white children (15 of 26 [57.6%] vs 9 of 74 [12.2%]; P < .0003) and among those with low educational backgrounds and socioeconomic levels (26 of 26 [100%] vs 62 of 74 [83.7%] and 19 of 26 [73%] vs 32 of 74 [43.2%], respectively; P < .05 for both).
Finally, H pyloriinfected children demonstrated any erosive esophagitis compared with noninfected children (7 of 74 [9.4%]; P = not significant), and H pyloriinfected children more frequently demonstrated nodular gastritis (19 of 26 [73%]) compared with noninfected children (7 of 74 [9.4%]; P < .001). Surprisingly, the noninfected children more frequently exhibited macroscopic aspects of gastritis at endoscopy (55 of 74 [74.4%]) rather than mild histologic gastritis (Table 2). H pyloriinfected children demonstrated nodular bulbitis and duodenitis (5 of 26 [19.2%]) compared with noninfected children (3 of 74 [4.1%]; P = not significant).
|
| DISCUSSION |
|---|
|
|
|---|
It is unknown whether H pylori gastritis causes symptoms in children lacking gastric or duodenal ulcers.9,1315 Two types of studies have sought to address the association of H pylori infection with RAP and dyspepsia in children and adults.16 The first approach investigates the possible association between H pylori infection and RAP; the second approach studies whether the eradication of H pylori infection results in the resolution of symptoms.16
Dyspepsia is poorly characterized in children. Hyams et al,17 in a prospective study on subjects with dyspepsia and dyspepsia subtypes (ulcer-like and dysmotility-like), showed that H pylori infection was unusual. They found only 5 cases out of the 127 subjects fulfilling their criteria for dyspepsia and concluded that most children with dyspepsia do not have serious disease.17 Similar results were found by another study, in which organic abnormalities in the course of RAP were detected in only 45% of children, and H pylori infection was found in only 1 of 44.18 However, several authors have suggested that nighttime pain associated with nocturnal awakening, fasting pain relieved by food, pain associated with meals, postprandial pain, bitter taste, and heartburn are the clinical signs that help to distinguish ulcer-positive children from those who are ulcer-negative yet positive for H pylori infection.19
We found here that the noninfected children more frequently exhibited epigastric pain during meals than did the infected ones. However, this clinical feature was not clinically significant in the noninfected group, because only 26% exhibited epigastric pain during meals and 74% did not. The absence of specific characteristics of RAP between these 2 groups in our study could be related either to the difficulty of children in describing their symptoms precisely or to an inappropriate selection of our population based on their symptoms, which could have caused a bias toward the reflux-like group. However, the prevalence of endoscopic esophagitis was very weak in the studied children and varied from none in the infected children to only 9.4% in the noninfected ones. Moreover, considering results, we observed that calculation of the required statistical power for the number of enrolled children to demonstrate a significant difference between both groups did not show any difference concerning most clinical dyspeptic manifestations.
On the other hand, in this study we have an extremely high prevalence of macroscopic gastritis in the noninfected children. As a matter of fact, our results in this highly selected population of children with epigastric pain are in agreement with the results of Snyder et al,20 in which primary antral gastritis was found in
20% of the 4 different age groups of 408 children studied, in contrast with only 4 of 39 children <10 years old with primary antral gastritis who had evidence of H pylori infection.
The main feature of this study is that we only selected children with epigastric-like NUD and that we detailed the characteristics of their abdominal pain. Thus, this study differs from other published studies in which the characteristics of RAP were not fully described12,21 or were limited to an imprecise definition of RAP.22 Our hypothesis was that the use of refined clinical characteristics for describing RAP could be of help in identifying a subgroup of patients with this condition in whom H pylori infection might need to be investigated and treated.11 Our study also used a rigorous double-blind approach, taking into account the usual factors associated with infection: socioeconomic level, the number of subjects in the same household, and ethnic background.
Children with H pyloriassociated gastritis are often asymptomatic, but a small minority of infected subjects will experience complications of peptic ulcer and gastric cancers including adenocarcinoma and lymphoma.2325 Fiedorek et al26 studied 245 healthy children for evidence of H pylori colonization and found that 30% of them were colonized. Blecker et al24 investigated 466 asymptomatic children for evidence of H pylori infection by using specific serum containing H pylori antibodies and found the prevalence of H pylori infection to increase from 5.4% to 13.4% with increasing age. Moreover, Bode et al,27 in a large epidemiologic study of healthy children, also showed that H pylori infection was not a cause of GI symptoms. In addition, a Dutch study failed to detect any difference in the prevalence of H pylori antibodies between children with RAP by using Apley's criteria and asymptomatic controls.28 Hardikar et al29 conducted a prospective case-control study in children with RAP, testing for serum antiH pylori IgG antibodies: 5 subjects (5%) and 14 controls (14%) had raised serum antibody titers, indicating a negative association between this infection and RAP. Controversially, Chong et al30 found that the incidence of positive antiH pylori IgG antibodies was significantly higher in children with RAP (17.4%) than in those without (10.5%).
Glassman et al21 assessed the presence of abdominal pain and vomiting in children undergoing upper-GI endoscopy; no difference was detected between H pyloriinfected children and those without infection in regard to their clinical manifestations. Mahony et al31 also found that the presence of epigastric pain did not discriminate between children with H pylori gastritis and those with a normal mucosa. Another retrospective study reviewed the symptomatology of 143 children referred for upper-GI endoscopy because of RAP for
6 weeks; H pylori infection was diagnosed in only 25.2% of children, and no statistically significant differences could be detected between the symptoms experienced by H pyloriinfected children as compared with those not infected.22 Our previous experience is in agreement with those reports: we found that RAP was not significantly present in 63% of patients with H pylori versus 49% of a control group of 74 age-matched children negative for H pylori.12,32 A meta-analysis of 45 studies has shown that the reported prevalence rates of H pylori infection in children undergoing upper endoscopy for RAP are inconsistent (median: 22%; range: 0-81%), with lower rates in children meeting Apley's criteria (median: 6%; range: 0-9%).33 Our study is in agreement with these studies: we found no significant difference between H pyloriinfected and noninfected children either in the intensity and frequency of RAP or in the other symptoms. These results suggest that there is no evidence for an association of this infection with RAP even in a subgroup of children presenting with epigastric-type NUD.33
The evidence supporting an association between H pylori gastritis and RAP is based on studies in which the authors have reported an improvement in symptoms after treatment.3442 On the other hand, controversial results were obtained by a Scandinavian study conducted in nonulcer-dyspeptic H pyloriinfected children in which a long-term follow-up study after treatment with colloidal bismuth subcitrate and tinidazole did not reveal any symptom relief for children in whom infection was eradicated.43 Most of those studies suffered from methodological problems because they were retrospective, open, nonrandomized versus placebo trials. Wever et al44 conducted a double-blind treatment study to elucidate whether symptoms disappeared in children with H pylori infection and RAP after bacterial eradication. The eradication rates were 81% (30 of 37) in the overall group and 74% (17 of 23) in the blind group; however, no correlation was seen between the eradication of H pylori and the disappearance of RAP after 3 or 6 months of observation in the total group or in the blind group of studied children. This study was thus unable to prove any causal relationship between RAP and H pylori infection.
Nevertheless, the interpretation of 1-arm treatment trials, even if well conducted, must be cautious, given the lack of a comparison group receiving standard therapy, unbiased investigators, and patients' subjective symptom scores. In addition, in most published studies, large numbers of children have been lost to follow-up, making the interpretation of long-term treatment effects difficult.32
This question thus cannot be resolved. An appropriate answer will only be achieved through prospective, well-structured, randomized, double-blind, placebo-controlled therapeutic clinical trials in nonulcer-dyspeptic H pyloriinfected children in comparable groups matched for age and socioeconomic class who undergo specific treatment designed to relieve dyspeptic symptoms. Only such studies will lead to standardized common guidelines for treatment.
| CONCLUSIONS |
|---|
|
|
|---|
| FOOTNOTES |
|---|
Reprint requests to (N.K.) Department of Pediatrics, Clinique de Pédiatrie Saint Antoine, Hôpital Saint Vincent de Paul, Catholic University, Boulevard Belfort, BP 387, 59020 Lille, France. E-mail: kalach.nicolas{at}ghicl.net
No conflict of interest declared.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
B. S. Collins and D. W. Thomas Chronic Abdominal Pain Pediatr. Rev., September 1, 2007; 28(9): 323 - 331. [Full Text] [PDF] |
||||
![]() |
L. Boyanova, E. Lazarova, C. Jelev, G. Gergova, and I. Mitov Helicobacter pylori and Helicobacter heilmannii in untreated Bulgarian children over a period of 10 years J. Med. Microbiol., August 1, 2007; 56(8): 1081 - 1085. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Falsafi, N. Valizadeh, S. Sepehr, and M. Najafi Application of a Stool Antigen Test To Evaluate the Incidence of Helicobacter pylori Infection in Children and Adolescents from Tehran, Iran Clin. Vaccine Immunol., September 1, 2005; 12(9): 1094 - 1097. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. pylori Infection and Nonulcer Dyspepsia in Children Journal Watch (General), February 11, 2005; 2005(211): 5 - 5. [Full Text] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||