Objectives. Antimicrobial treatment may disturb the colonization resistance of gastrointestinal microflora, which may induce clinical symptoms, most commonly diarrhea. The severity of antibiotic-associated diarrhea may range from a brief, self-limiting disease to devastating diarrhea with electrolyte disturbances, dehydration, crampy abdominal pain, pseudomembranous colitis, toxic megacolon, or even death. The incidence of diarrhea in children receiving a single antimicrobial treatment is unclear. In addition to more critical use of antimicrobials, adjunctive preventive measures to antibiotic-associated diarrhea are needed. The objective of this study was to evaluate the incidence of diarrhea after antimicrobial treatment in children with no history of antimicrobial use during the previous 3 months. Another aim of this study was to assess the preventive potential of Lactobacillus rhamnosus GG (Lactobacillus GG; American Type Culture Collection 53103), a probiotic strain with a documented safety record and a therapeutic effect in viral gastroenteritis on antibiotic-associated diarrhea. Methods. Oral antimicrobial agents were prescribed for the treatment of acute respiratory infections at the clinics of the Health Care Center of the City of Tampere or Tampere University Hospital, Finland, to 167 patients who were invited to participate in the study. Of the patients, 48 were lost to follow-up; therefore, the final study population consisted of 119 children from 2 weeks to 12.8 years of age (mean: 4.5 years). All study subjects met the inclusion criteria: they had not received any antimicrobial medication during the previous 3 months, they did not suffer from gastrointestinal disorders, and they did not need intravenous antimicrobial treatment. The patients were randomized to receive placebo or 2 x 10(10) colony-forming units of Lactobacillus GG in capsules given twice daily during the antimicrobial treatment. Lactobacillus GG and placebo capsules were indistinguishable in appearance and taste. The parents kept a daily symptom diary and recorded stool frequency and consistency at home for 3 months. Diarrhea was defined as at least three watery or loose stools per day for a minimum of 2 consecutive days. In the case of diarrhea, viral (adenovirus, rotavirus, calicivirus and astrovirus) and bacterial (Salmonella, Shigella, Yersinia, Campylobacter, Clostridium difficile, Staphylococcus aureus, and yeasts) analyses were studied in fecal samples. The metabolic activity of the gut microflora was assessed by analysis of fecal urease, beta-glucosidase, and beta-glucuronidase activities. The primary outcome measure was diarrhea during the first 2 weeks after the beginning of the antimicrobial treatment, because this period most likely reflects the effects of antimicrobial use. Secondary outcome measures were the activities of fecal urease, beta-glucuronidase, and beta-glucosidase. Results. On the entire follow-up, 80% of any gastrointestinal symptoms were reported during the first 2 weeks after the beginning of the antimicrobial treatment. The incidence of diarrhea was 5% in the Lactobacillus GG group and 16% in the placebo group within 2 weeks of antimicrobial therapy (chi(2) = 3.82). The treatment effect (95% confidence interval) of Lactobacillus GG was -11% (-21%-0%). In diarrheal episodes, the viral and bacterial analyses were positive for Clostridium difficile in 2 cases and for Norwalk-like calicivirus in 3 cases. The age of the patients with diarrhea was between 3 months and 5 years in 75% of cases in both groups. The severity of diarrhea was comparable in the study groups, as evidenced by similar stool frequency (mean: 5 per day; range: 3-6) and the duration of diarrhea (mean: 4 days; range: 2-8). The activities of fecal urease and beta-glucuronidase, but not beta-glucosidase, changed significantly after the beginning of the antimicrobial treatment in the Lactobacillus GG group and in the placebo group alike. The decrease in urease and beta-glucuronidase activities was reversible in patients with no diarrhea, but in patients with diarrhea, the modifications in gut microflora were more profound and prolonged. The activities of the three enzymes were normalized within 3 weeks, evidenced by stable enzyme activities in samples collected 3 weeks, 1 month, and 3 months after the beginning of the antimicrobial treatment, compared with those obtained before treatment. Discussion. In the present study, after a single antimicrobial treatment, the incidence of diarrhea was 16%. The higher incidence of antibiotic-associated diarrhea in previous reports may be attributable to a recent antimicrobial therapy that disturbs intestinal flora and exposes to complications. Also, in the present study, changes in the metabolic activity of the intestinal flora were observed, evidenced by a transient decline in fecal enzyme activities. Different probiotic preparations, including lactobacilli, are recommended frequently to prevent antibiotic-associated diarrhea. Although probiotics have been shown to be efficient in the prevention and the treatment of viral gastroenteritis, their usefulness during antimicrobial therapy in children has not been elucidated before. We observed that the administration of Lactobacillus GG to children receiving antimicrobial therapy for respiratory infection reduced the incidence of antibiotic-associated diarrhea to one third. The beneficial effect may be mediated by a number of functions of probiotics, ie, production of antimicrobial substances, local competition of adhesion receptors and nutrients, and stimulation of intestinal antigen specific and nonspecific immune responses. Conclusion. A probiotic strain, Lactobacillus GG, is effective in the prevention of diarrhea in children receiving antimicrobial treatment to respiratory infections.
