Dynamics of verotoxin-producing Escherichia coli isolated from German beef cattle between birth and slaughter.

A total of 85 cattle from three German beef farms were followed between birth and slaughter during a period of 2 years and monthly faecal samples were submitted for bacterial culture. Verotoxin-producing Escherichia coli (EC) were detected using a standard diagnostic cascade. Potentially pathogenic VTEC ((p)VTEC) were defined as: positive for (1) verotoxin 1 (vt1) and eae, (2) positive for verotoxin 2 (vt2) and eae, (3) positive for both verotoxins 1 and 2 and eae, while verotoxinogenic EC (EC(vt1,2)) were defined as: (1) positive for vt1, (2) positive for vt2 or (3) positive for both vt1 and vt2. There were 1587 observations (1462 valid) available for the statistical analysis including 6 (0.4%) samples from 6 (7.1%) different animals positive for VTEC O157, 78 (5.3%) pVTEC isolates and 389 (26.6%) EC(vt1,2) isolates. The median day of the study at first detection was 280 days for EC(vt1,2) and 315 days for pVTEC. The median age at first detection was: 121 days for EC(vt1,2) and 215 days for pVTEC. Time series analysis, survival analysis, and stochastic SI models were used to find differences in the population dynamics of EC(vt1,2) and pVTEC. There was a strong farm and age effect for the first detection of EC(vt1,2) and for pVTEC while the seasonal effect was significant for the first EC(vt1,2) detections only. With increasing age at first and all consecutive detections, EC(vt1,2) and pVTEC were detected less frequently. The serotype O157 was found more frequently together with detection of other serotypes of pVTEC in the same sample. The EC(vt1,2) were found more often together with pVTEC. The first EC(vt1,2) were on average found before the first pVTEC's and positive cross-correlations existed between EC(vt1,2) and pVTEC. The critical duration for the shedding period above which the VTEC could propagate themselves on the farms by f.e. transmission between animals was found to be between 8 and 18 sampling intervals of 28 days (224-504 days) for EC(vt1,2), and between 5 and 6 sampling periods of 28 days each (140-168 days) for the pVTEC which is smaller than all critical shedding periods for EC(vt1,2). The reasons for EC(vt1,2) being isolated from faeces earlier than pVTEC are discussed.