A longitudinal study of the prevalence and super-shedding of Escherichia coli O157 in dairy heifers

•We studied shedding in a cohort of 52 dairy heifers weekly for 6 months.•Point prevalence varied greatly (9.6-94.3%).•Super-shedding was detected in 32/893 (3.6%) of samples tested.•Shedding was not predictable from prior shedding beyond two weeks.•Prevalence of shedding was not associated with super-shedding.

The fecal shedding and super-shedding of the human pathogen Escherichia coli O157 by cattle has been the focus of many previous studies with varied results observed. The heterogeneity of shedding is becoming more accepted, both in the numbers of animals shedding and the levels at which animals shed. To clarify patterns in shedding and super-shedding we undertook a longitudinal study to investigate shedding within a cohort of replacement dairy heifers. The cohort of 52 heifers was sampled 18 times at approximately weekly intervals with no significant changes in management during the sampling period. An overall prevalence of 44.3% (412/930 samples) was detected with prevalence ranging from 9.6 to 94.3% at individual sampling points. Each of the 52 heifers yielded at least one sample which was detected positive for E. coli O157. Super-shedding was detected at a sample level of 3.6% (32/893) and ranged between 0 and 9.6% at each sampling point. Of the 52 heifers, 24 (46.2%) were detected to be super-shedding at some point during the study, 19 of which were detected as super-shedding at only one point. From our findings we conclude that super-shedding is not associated with a small subset of animals that shed at high levels continually as had been proposed by earlier studies. We propose that the term 'super-shedding event' as opposed to 'super-shedding animal' better describes the nature of shedding.