Diarrheagenic Potential of Escherichia coli in Children in a Developed Country

Escherichia coli is an underestimated diarrheal pathogen in developed countries. This organism may cause gastroenteritis by many different mechanisms, and on the basis of this, diarrheagenic E. coli has been subdivided into different groups, of which the most important ones are Shiga-toxigenic E. coli, enteropathogenic E. coli (EPEC), other attaching and effacing E. coli (A/EEC), enteroaggregative E. coli (EAEC), enterotoxigenic E. coli, and enteroinvasive E. coli. The latter two are associated with travel to less developed countries. Shiga-toxigenic E. coli is a well-established pathogen, whereas the pathogenic role of EPEC, A/EEC, and EAEC is less well established.

We read the paper by Pabst et al. in the June issue of the Journal of Clinical Microbiology (1) with great interest. The authors try to shed light on the role of diarrheagenic E. coli in Switzerland by using a case-control setup and molecular diagnostic methods for the identification of the different E. coli groups. However, the study has a couple of shortcomings that makes us wonder if all of the conclusions presented are correct.

Firstly, the authors do not take into account the potential bias of the travel history of the patients when comparing cases and controls. Travel may be an important confounder because none of the controls reported traveling, whereas travel was quite common among the patients; i.e., if a pathogen is more common in travelers than in patients with domestically acquired cases, it will not be possible to ascertain the relative importance of travel if the patients are compared with a group of controls who have not traveled in a univariate analysis. A way of circumventing the problem would be to exclude the travel-associated cases from the analysis. In the case of EAEC, the number of patients who had traveled was 9 or 47%; the total number of diarrheal patients who had traveled was 31. It is thus possible to compare the occurrence of EAEC in patients and controls with no travel history: The prevalence in patients was 9 of 156, and that in controls was 3 of 137. This difference is not statistically significant (P = 0.21; Fisher's two-tailed exact test).

Secondly, the authors did not find any difference in the prevalence of EPEC between cases and controls. This conclusion may be questioned too, because the EPEC definition used is based solely on detection of the intimin gene (eae). In the definition of EPEC, the serotype of the strains should also be taken into account (2). We performed a study similar to that of Pabst et al. and found a strong correlation between disease in infants less than 2 years of age and EPEC, defined as eae-positive strains belonging to the classical EPEC O groups, whereas other A/EEC showed no correlation to disease (B. Olesen, P. Schiellerup, M. Helms, J. Neimann, F. Scheutz, and P. Gerner-Smidt, Abstr. 12th Eur. Congr. Clin. Microbiol. Infect. Dis., abstr. P607, 2002). Classical EPEC is the most common bacterial cause of diarrhea in children less than 2 years old in Denmark. We wonder if Pabst et al. would have reached the same conclusion had they serotyped their eae-positive isolates.