Methods: Children 6-35 months old with acute non-bloody diarrhea (<72 hours) were enrolled in an acute diarrhea clinical trial between 3/2015 and 1/2016 at two sites (1 rural, 1 urban) in Guatemala. Stool samples collected at baseline and 30 days later were analyzed by multiplex PCR (FilmArrayTM GI-Panel, BioFire, USA) that identifies 22 viral, parasitic and bacterial diarrheal pathogens including C.difficile. Subjects were characterized by combination of baseline and 30 day C.difficile sample results: -/+ (new acquisition), +/- (clearance), and +/+ (persistence). Associations between these categorizations and demographic, epidemiologic and co-infecting pathogenic organisms were assessed using multivariable generalized linear models.
Results: CDI was present in 26 of 298 subjects at baseline; 13 (50%) had persistence at 30 days and 13 (50%) cleared. New acquisition at day 30 occurred in 23 subjects. In multivariable analysis adjusted for age, recent hospitalization was marginally significantly associated with C.difficile presence in stool at baseline (Prevalence Ratio (PR) 2.65, p = 0.07). In subjects with either new C.difficile acquisition or persistence between baseline and day 30, residence in the rural site (PR 0.33, p = 0.003)) and presence of E.coli pathotypes: enteropathogenic (EPEC), enteroaggregative (EAEC) and enterotoxigenic (ETEC) (PR 0.43, p = 0.01)) were associated with reduced risk of CDI.
Conclusion: In a LMIC pediatric population, the presence of E.coli pathotypes appeared protective against C.difficile persistence/new acquisition. These findings add to our current understanding that CDI occurs in part as a result of competition within the intestinal microbiota, which may be independent of the potential pathogenicity of competing microbes. We hypothesize that this phenomenon could be suppressing the C.difficile burden among children in LMICs.
M. Melgar, None
D. M. Calvimontes, None
I. Contreras-Roldan, None
E. Asturias, None
S. Dominguez, None