Papers

Helicobacter pylori and childhood recurrent abdominal pain: community based case-control study

BMJ 1999; 319 doi: https://doi.org/10.1136/bmj.319.7213.822 (Published 25 September 1999) Cite this as: BMJ 1999;319:822
  1. Colin Macarthur, epidemiologist (colinmac{at}ucalgary.ca)a,
  2. Norman Saunders, paediatricianb,
  3. William Feldman, paediatricianb,
  4. Moshe Ipp, paediatricianb,
  5. Patricia Winders-Lee, research nurseb,
  6. Susan Roberts, graduate studentb,
  7. Linda Best, research assistant, department of microbiologyc,
  8. Philip Sherman, paediatric gastroenterologistb,
  9. Paul Pencharz, paediatric gastroenterologistb,
  10. Sander Veldhuyzen van Zanten, paediatric gastroenterologistc
  1. a Department of Community Health Sciences, Health Sciences Center, Calgary, Alberta, Canada T2N 4N1
  2. b Department of Paediatrics, Hospital for Sick Children, Toronto, Ontario, Canada M5G 1X8
  3. c Victoria General Hospital, Halifax, Nova Scotia, Canada B3H 2Y9
  1. Correspondence to: C Macarthur
  • Accepted 20 May 1999

Recurrent abdominal pain (at least three discrete episodes of abdominal pain over a period of three or more months, and of sufficient severity to interrupt normal activities) is a common childhood complaint. We set out to determine the association, if any, between Helicobacter pylori infection and childhood recurrent abdominal pain.

Participants, methods, and results

Cases and controls were drawn consecutively from the practice populations of six primary care paediatricians in Toronto. (Convenience sampling was used to select paediatricians; they were chosen because of their interest in the study.) Cases were children aged 5-15 years presenting with recurrent abdominal pain; controls were healthy children undergoing a routine check-up or vaccination. Excluded were children with concurrent disease, suspected organic disease, aged under five years, or who had used bismuth in the previous month. All families approached consented to participate.

Serum IgG antibodies to H pylori were measured by using a flow microsphere immunofluorescent assay (FMIA) and a commercial immunoassay kit (Bio-Rad Laboratories, Hercules, CA). The FMIA method has been validated in children (100% sensitivity, 97% specificity) against a gold standard of culture and histology.1 Infection was also diagnosed by using a 13C-urea breath test (99% sensitivity, 98% specificity).2 A standardised questionnaire was used to gather social, demographic, and clinical information on each participant.

One hundred children with recurrent abdominal pain and 100 healthy controls participated in the study. The groups did not differ in age, height, weight, history of colic, previous admissions to hospital, seasonal allergies, or antibiotic use in the previous month (table). The two groups were also similar with respect to socioeconomic status, country of birth of the parents, number of siblings, smokers in the home, household pets, and family history of recurrent abdominal pain, peptic ulcer, and migraine. Children with recurrent abdominal pain, however, were more likely to be girls, to have symptoms of chronic headache or chronic limb pain, and to have more frequent school absences than control children.

Comparison of children with recurrent abdominal pain (cases) and healthy children (controls) on social, demographic, and clinical factors. Values are numbers unless otherwise specified

View this table:

Results of serology were available for 174 children (87%). Only five had positive results (3/93 cases v 2/81 controls). Breath test results were available for 193 children (97%). Nine were positive (4/97 (4%) cases v 5/96 (5%) controls; crude odds ratio 0.78 (95% confidence interval 0.20 to 3.01); odds ratio adjusted by logistic regression 0.65 (0.08 to 2.56)).

Comment

This community based case-control study found no association between H pylori infection and recurrent abdominal pain in childhood. Strengths of the study included the primary care setting, the use of incident cases of recurrent abdominal pain, and the use of healthy children as controls. Information bias was minimised by using a standardised questionnaire, with the research nurse blind to the serology and breath test results. H pylori infection was measured by two independent serology tests and a urea breath test. As the families involved were of a high socioeconomic status (which could explain the low prevalence of infection among control children), the findings should be generalised with caution.

One hospital based case-control study found that H pylori infection could be a risk factor for recurrent abdominal pain.3 The case children, however, had a higher prevalence of organic disease than in previous reports. Also, the control group included children with concurrent gastrointestinal disease. Other community based case-control studies have found no association between H pylori infection and recurrent abdominal pain.4 5

On the basis of our findings and other published evidence, H pylori is not a causal factor for childhood recurrent abdominal pain Therefore, primary care doctors should not routinely investigate for H pylori infection in children who present with classic symptoms of recurrent abdominal pain.

Acknowledgments

We thank Drs M Gans, M Goldbach, and J Hilton for their contributions.

Contributors: CM discussed core ideas, designed the study protocol, analysed the data, and wrote the manuscript; he is guarantor for the paper. NS initiated the primary study hypothesis, discussed core ideas, helped design the study protocol, participated in data collection, and edited the paper. WF helped formulate the primary study hypothesis, discussed core ideas, helped design and write the study protocol, and edited the paper. MI participated in data collection, discussed core ideas, and edited the paper. PW-L participated in the design and execution of the study, particularly data collection and quality control, and edited the paper. SR participated in the design and execution of the study, particularly outcome measurement (urea breath test) and quality control, and edited the paper. LB participated in the design and execution of the study, particularly outcome measurement (serological tests) and quality control, and edited the paper. PS discussed core ideas, helped design the study protocol, helped interpret the data, and edited the paper. PP participated in the design and execution of the study, particularly outcome measurement (urea breath test), helped interpret the data, and edited the paper. SVZ participated in the design and execution of the study, particularly outcome measurement (serological tests), helped interpret the data, and edited the paper.

Footnotes

  • Funding Health and Welfare Canada, National Health and Research Program grant No 6606-5282.

  • Competing interests None declared.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract