Prevalence and infection risks of zoonotic enteropathogenic bacteria in Swiss cow-calf farms

Veterinary Microbiology 69 (1999) 251±263

Prevalence and infection risks of zoonotic enteropathogenic bacteria in Swiss cow-calf farms A. Busatoa,*, D. Hofera, T. Lentzea, C. Gaillarda, A. Burnensb b

a Institute of Animal Breeding, University of Berne, Bremgartenstrasse 109a, 3012 Berne, Switzerland Institute of Veterinary Bacteriology, Faculty of Veterinary Medicine, University of Bern, Bern, Switzerland

Received 25 January 1999; accepted 25 June 1999

Abstract A longitudinal study was performed in 67 larger Swiss cow-calf farms from September 1996 through November 1997. The objectives of the study were to estimate prevalence and risk factors for colonization with potentially zoonotic enteropathogenic bacteria in younger calves and in calves at weaning age. The study included data from 395 calves with three to four fecal samples each. Fecal samples were analyzed for Campylobacter spp., verotoxin producing E. coli (VTEC), Yersinia spp. and Salmonella sp. Possible environmental and individual factors associated with colonization of these agents were examined. The calves were housed indoor during the first 3 months of life (winter 1996/1997). The prevalences within this time period were: C. coli 3.4%, C. fetus 15.5%, C. hyointestinalis 9.6%, C. jejuni 38.5%, VTEC 44.3% and Yersinia spp. 2%. At the end of the grazing season the prevalences at weaning (8±10 months of age) were: C. coli 1.7%, C. fetus 4.0%, C. hyointestinalis 25.9%, C. jejuni 13.3%, VTEC 38.2% and Yersinia spp. 0%. No salmonellae were present at any time of the study. The prevalences of C. jejuni and VTEC increased significantly within the first 3 months of life, whereas C. hyointestinalis decreased. None of the environmental factors such as housing or feeding had any consistent influences on colonization by the bacteria studied. VTEC, Campylobacter spp. and Yersinia spp. should probably be considered as normal inhabitants of the bovine intestinal tract. However, as they represent a source of gastrointestinal infections in humans, management factors limiting intestinal colonization of these bacteria should be considered in cow-calf operations. # 1999 Elsevier Science B.V. All rights reserved. Keywords: Cattle-bacteria; Enteropathogens; Zoonosis; Epidemiolgy; Switzerland

*

Corresponding author. Tel.: +41-31-631-23-25; fax: +41-31-631-26-40 E-mail address: [email protected] (A. Busato) 0378-1135/99/$ ± see front matter # 1999 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 1 3 5 ( 9 9 ) 0 0 1 1 9 - 4

252

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

1. Introduction The introduction of milk quotas in the late 1970's and the structural changes in the Swiss agriculture in the last few years increased a shift from intensive dairy production towards extensive beef production methods. Consequently the numbers of cow-calf farmers have increased considerably in the last years and in spring 1998 approximately 1900 farms were registered in the Swiss association of cow-calf farmers, representing approximately 4% of Swiss cattle breeders. Cow-calf herds are raised in Switzerland usually in loose stalls constructed either as deep litter barns or as converted dairy barns with boxes for individual cows and specifically designed deep litter areas accessible for calves only. The animals have unlimited access to outdoor paddocks in winter and in summer the herds are grazed, often on remote mountain pastures. Calves are slaughtered either at weaning at 8±10 months or after a finishing period at age 12±14 months. Marketing is performed through the usual retail channels partially labeled as ecological products. On larger farms beef is usually processed on site and sold directly to customers. An earlier study in Swiss cow-calf farms (Steiner et al., 1997) has shown that zoononotic foodborne bacteria Campylobacter spp., VTEC, and Yersinia spp. are more prevalent in cow-calf herds than previously assumed. The objectives of this study were to describe the prevalence of zoonotic enteropathogenic bacteria in beef calves raised in Swiss cow-calf farms at different ages, to characterize the temporal shedding of these bacteria under field conditions and to investigate possible associations between intestinal colonialization with VTEC and C. jejuni and environmental and individual calf factors during the first 3 months of life and at weaning. 2. Materials and methods 2.1. Study design This project was designed as a longitudinal study with multiple samples per animal over a period of 8±10 months. The target population was defined as calves born during the winter season and raised in Swiss cow-calf herds with at least 20 breeding cows. It was planned to collect four fecal samples from each calf, one sample each month during the first 3 months of age and one sample at weaning (8±10 months). Sample size calculations were appropriate for a two-stage cluster sample according to Levy and Lemeshow (1991). Estimates of frequency of infections were derived from a previous project (Steiner et al., 1997). Expenses for laboratory analyses for each sample and mileage costs to reach each farm were included in these calculations. Based on these assumptions and requirements it was decided to collect 400 calves from 80 farms (approximately 15% of all the calves in each farm). Two hundred and eight cow-calf farmers located in western Switzerland with a minimal herd size of 20 adult cattle were informed about the study in summer 1996 by letter and asked to take voluntarily part. Seventy-six farmers agreed to participate, 96 gave no answer and 36 refused. Finally, 67 farms were selected for the study and nine farms were excluded because of logistic constraints.

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

253

The purpose of a first visit at the farm was the collection of farm and calf-management data. At least four additional farm visits were subsequently performed in order to collect fecal samples and individual calf data such as the date of birth, sex, breed, birthweight, birthplace, age of the dam, health status, therapeutic/prophylactic measures etc. Since the number of farms selected was below the number of farms originally planned it was decided to increase the fraction of sampled calves per farm from 15 to 20%. 2.2. Bacteriology Fecal swabs (Transwab1, Medical Wire & Equipment Co. Ltd., UK) and 50 ml plastic tubes were used to collect fecal samples. The bacteriological assays included screening for enterotoxigenic E. coli (ETEC), verotoxin-producing E. coli (VTEC), Salmonella spp., Campylobacter spp. and Yersinia spp. All bacteriological tests were performed at the Institute of Veterinary Bacteriology, Faculty of Veterinary Medicine, University of Berne. Fecal samples were stored at 48C and processed within 16 h. For the isolation of Yersinia spp. fecal specimens were plated on CIN-Agar (Catalogue no. 43203, bioMeÂrieux, Marcy l'Etoile, France) and the plates read after incubation for 48 h at 228C. Enterotoxigenic E. coli (ETEC) were recovered by plating fecal specimens on MINCA agar and identified by slide agglutination of five colonies with K99 antiserum (Boss et al., 1992). For recovery of Salmonella spp., fecal matter was directly plated on Bromthymolblue-Lactose-Agar (Brolac, cat.-no. 1639, Merck, Darmstadt, Germany) and brilliant green agar. The presence of Verotoxin-producing E. coli (VTEC) was ascertained by immunological detection of toxin in overnight broth cultures in a commercially available enzyme immunoassay (Premier EHEC1, cat.-no. 608 096, Meridian Diagnostics, Cincinnati, OH, USA). Briefly, about 0.1 g of feces were inoculated into MacConkey broth (cat.-no. 0020-01-5, Difco, Detroit, USA), and after overnight incubation at 378C, 50 ml of the broth culture was mixed with sample diluent as recommended by the manufacturer. The results were read spectrophotometrically at 450 nm and a cutoff value of 0.180 was used. The sensitivity of the enzyme immunoassay (EIA) for the detection of fecal verotoxin has been shown to be comparable to the cytotoxicity assay on Vero cells, and its sensitivity for the detection of VTEC in overnight broth cultures was similar or superior to other established methods including PCR (Kehl et al., 1997). For the detection of campylobacters, the fecal specimens were plated on a commercially available selective medium (Campylosel, cat.-no. 43253, bioMeÂrieux, Marcy l'Etoile, France), and incubated for 48 hours at 378C in an atmosphere of 6% O2, 7% CO2, and 7% H2 in nitrogen. Typical colonies of Gram negative, oxidase positive, motile organisms with spiral morphology were restreaked onto blood agar and identified by dot-blot DNA hybridization (Burnens et al., 1993). 2.3. Data collection, statistical analyses The data were stored in a relational database and all diagnostic data were coded as positive, negative or missing. The final dataset available for statistical analyses comprised data of 67 farms and 1521 samples derived from 395 calves.

254

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

Since multiple measurements per animal cannot be regarded as independent units of observation and since calves within the same barn tend to respond similarly, data were analyzed considering two levels of clustering. It was not always possible to select exactly the same proportion of calves per farm. In order to avoid biased point estimates of prevalences as a consequence of unequal sampling probabilities (Brogan, 1998) each observation was weighted to represent a target population of 6697 calves. This population corresponded to the number of calves born from September 1996 through March 1997 in Swiss cow-calf farms with at least 20 cows. Explanatory variables were derived from questionnaires and from the protocols from individual calves and samples. The analyses of possible environmental and individual effects of an infection up to 3 months of age were achieved in several steps. In a first step, descriptive methods were applied in order to establish estimates of prevalence with corresponding 95% confidence intervals (CI). A second, analytical step included logistic models as screening procedures in order to set the basis of multivariate models. Most of the bacteria that were analyzed showed age dependent excretion patterns. Age in days at sampling was, therefore, always included as a confounding variable in this screening process. Multivariate logistic models were developed based on the results of the screening procedure, and finally the following model with a repeated measurement design over three samples was used for all the bacteria under study Y ˆ AGE ‡ N COW ‡ BARN ‡ TEMP ‡ BREED ‡ TREATMNT ‡ FEEDS ‡ FEED TYP ‡ ERROR The presence of an infection at each sampling date was defined as binary target variable (Y). Age in days at each sampling date (AGE) and the number of breeding cows in each herd (N_COW) were included as continuous explanatory covariables. In order to test the hypothesis that housing systems have an influence on the prevalence of bacteria two variables were included: the type of barn (BARN: deep litter barn, loose stall with boxes) and as a rough measure for the environmental temperature, barns were classified into closed barns or open barns with large outside openings (TEMP) (Table 1). Open barns were considered to be colder and dryer than closed barns. All the calves were sired by beef bulls; the dams, however, were classified into dairy or dual-purpose cows or beef cows. Since the potentially lower milk production of beef breed dams may influence the bacterial infections of their calves, a binary categorical variable BREED was included in the model (Table 1). Several animals were treated for various diseases by the herdsman or by a veterinarian prior to the collection of fecal samples. These treatments included antibiotics, antiparasitics or other therapeutic measures. In order to account for the effect of these therapeutic treatments a corresponding four level and time varying categorical variable was included in the model (TREATMNT) (Table 2). Each treatment level was Table 1 Descriptive statistics for binary farm-level variables used in the multivariate model Binary variables

Code = 0

Code = 1

Proportion code = 0 (%)

Housing system (BARN) Temperature in barn (TEMP) Breed of dam (BREED) Feed type (FEED_TYP)

Boxes Closed Beef breeds Not restricted

Deep litter Open front Others Restricted

34.4 56.3 81.0 74.6

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

255

Table 2 Descriptive statistics for nominal farm-level variables used for the multivariate model Variable

Levels

Relative proportion (%)

Therapy/ treatment (TREATMNT)

Antibiotics Antiparasitics Therapy without AB None >=50% DM grass silagea >=50% DM corn silage >=50% DM other silage >=50% DM straw >=50% DM hay Balanced

7.1 1.4 1.0 90.5 19.6 8.2 3.1 6.6 12.2 50.3

Feed (FEEDS)

a

DM = dry matter.

compared to samples without prior treatments. Based on the hypothesis that particular diets and feeding strategies at farm level may influence the intestinal colonialization of bacteria in calves, two additional time varying variables were included in the final model. A six level categorical variable FEEDS was used to assess possible effects of specific diet components and rations (Table 2). They were categorized into balanced rations as the reference and into rations with more than 50% dry matter (DM) of grass silage, corn silage, other silage, hay or straw. Possible effects of unrestricted or restricted feeding were evaluated with a two level variable (FEED_TYP) (Table 1). Descriptive statistical procedures and statistical tests of unclustered data were performed with SAS 6.12. (SAS1, Institute, Cary, NC, USA). Confidence intervals and statistical tests of clustered data were analyzed with SUDAAN Software (Research Triangle Institute, Triangle Park NC, Release 7.00). Agreement between observed data and logistic models were evaluated with likelihood ratio tests. Continuous results are given as means and standard deviations (SD), binomial results and estimates of relative infection-risks as proportions and odds ratios (OR) with 95% confidence intervals (CI). Throughout the paper the level of significance was set at p  0.05. 3. Results Sixty-six percent of the herds were kept in deep litter free stalls that were designed specifically as cow-calf stalls (Table 1). The remaining herds were housed in free stalls with individual boxes for cows and with specifically designed areas accessible for calves only. Angus crossbred animals were most frequent (61%). The population of calves initially selected comprised 403 animals; on average six calves were selected on each farm (2±20). The dams of the 403 calves were multiparous cows in 334 cases, the other were heifers. Sixty-one percent of the farmers reported problems of calf diarrhea 1 or 2 years earlier than the study. The calf mortality in the first 3 months of life was 1.2%, two calves died of omphalitis, two of unknown reasons and one calf was killed prematurely because of poor

256

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

performance. At weaning, data from 348 calves were available, two animals were culled because of accidents, one died of unknown disease and the remaining animals were slaughtered without notifying the project team before fecal samples could be collected. Forty-nine (12%) calves were diagnosed with diarrhea at the age of 1 month, 45 (11%) at 2, 35 (9%) at 3 months and only 1% at the weaning age. Calves were treated for various diseases in 114 cases prior to sample collection during the first 3 months (1173 samples). Antibiotics were applied in 82 cases and in 16 cases antiparasitics or other treatments were applied. 4. Prevalences Salmonella were not detected in the study. Verotoxin producing E. coli (VTEC) were detected in 43% of all the samples or in 78% of all the farms. Thirty-five percent in the first, 45% in the second, 53% in the third and 39% of the samples in the fourth collecting period were positive (Table 3). The analysis of possible risk factors of an colonization with VTEC in calves up to 3 months revealed significant effects promoting a colonization (OR > 1) for age, dairy crossbreed dams (other breeds) and treatments with antiparasitics prior to sample collection. Protective effects (OR < 1) were obtained for open barns, antibiotics and feeding more than 50% DM grass silage (Table 4). The overall prevalence of C. jejuni was 32%. In two farms (3%) no C. jejuni were found, whereas in eight farms more than 50% of all the samples were positive. Prevalence of C. jejuni are shown in Table 3. Effects significantly promoting a colonization were observed for age, open barns, more than 50% DM grass or cornsilage. Factors decreasing the risk of a colonization were number of cows, crossbreed animals, antiparasitics and feeding more than 50% DM other silage (Table 4). C. coli were in found in 3% of all the samples and in 19% of all the farms and prevalences for C. coli varied between 2 and 4% (Table 3). C. fetus were detected between 13 and 18% in the first three samples and only 4% at weaning (Table 3). C. hyointestinalis were found in 14% of the samples or in 82% of the farms (Table 3).

Table 3 Prevalences of bacterial enteropathogens depending on age at sampling Month 1

Month 2

Number of specimen 395 392 Age (days  SD) 21.2  5.4 48.4  5.9 Prevalences (%) with 95% confidence intervals VTEC 34.8a  4.2 44.7b  4.9 C. jejuni 35.3a  4.3 35.0a  4.8 a C. coli 2.6  0.1 3.7ab  1.4 17.9b  3.6 C. fetus 13.0a  3.1 C. hyoint. 14.4a  3.3 11.0a  3.0 Yersinia 1.9  1.3 2.1  1.2 a

Month 3

Months 8±10

386 77.8  6.6

348 298.5  35.4

53.4c  4.9 45.2b  5.1 3.8a  1.8 15.7ab  3.1 3.5b  1.6 2.6  1.7

38.7a  4.7 13.3c  3.6 1.7ac  1.2 4.3c  0.2 25.6c  4.3 0.0  0.0

Different superscripts denote significant differences between sampling times.

Average

42.9  2.4 3.22  1.2 2.9  0.7 12.7  1.4 13.6  1.5 16.7  0.6

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

257

Table 4 Values of odds ratios for variables of the multivariate logistic model Variable/variable name Herdsize Age Type of barn Deep litter Boxes Construction Open front Closed Breeds Other breeds Beef-breeds Treatment Antibiotics Antiparasitics Others None Diet composition 50%DM grass silage 50%DM corn silage 50%DM other silage 50%DM straw 50%DM hay Balanced ration Feeding procedure Not restricted Restricted

VTEC N_COW AGE BARN TEMP BREED TREATMNT

FEED

FEED_TYP

*

Significant OR-values.

a

Categories with less than 5% of observations.

C. jejuni

1.00 1.01*

0.99* 1.01*

1.08 1.00

0.97 1.00

0.62* 1.00

1.43* 1.00

1.46* 1.00

0.68* 1.00

0.63* (22.91a) (0.75) 1.00

0.74 (0.01*) (1.65) 1.00

0 .71* 0.77 (0.75) 1.07 0.79 1.00

1.35* 2.72* (0.44*) 1.36 0.70 1.00

0.90 1.00

0.94 1.00

Yersinia sp. were found in 24 cases (1.7%) or in 22% of the farms. Prevalences in the first three samples were: 2, 2 and 3%, at weaning no positive samples were diagnosed (Table 3 ). The differentiation of species revealed Y. pseudotubeculosis in 23 cases and Y. frederiksenii in the remaining specimen. 5. Discussion The sampled calves represented 6% of all calves raised in Swiss cow-calf farms with at least 20 cows in 1996. Since all the farmers participated voluntarily in the study, inferences from our data may, therefore, be biased towards more motivated producers. It is however, difficult to establish a direction of bias. The motivation of each farmer to participate ranged from basic agreement to gain additional knowledge on animal health to the expectation of solving existing problems of calf diseases on individual farms.

258

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

5.1. E. coli Infections with verotoxin producing E. coli (VTEC) in cattle gained increasing significance in public health after outbreaks of VTEC related diseases in humans were linked to foods of bovine origin (Griffin and Tauxe, 1991). Human diseases associated to VTEC include diarrhea, hemorrhagic colitis, hemolytic uremic syndrome and thrombotic thrombocytopenic purpura (Karmali, 1989). VTEC of the O157 : H7 serotype were identified as the major etiologic strain of these diseases. The role of VTEC as cause of diarrhea in cattle is not fully established and infections seem to be limited to the intestinal tract (Gyles, 1994). However, it appears that a direct relationship between infection of VTEC and diarrhea exists (Busato et al., 1998). In the present study an average 43% of samples tested positive for VTEC without further determination of specific serotypes. Generally lower prevalences in cattle have been reported in other work (Wilson et al., 1992; Blanco et al., 1996; Rahn et al., 1997), possibly due to different diagnostic methods. The prevalences in this study increased significantly from 0.35 to 0.53 within the first 3 months and at weaning the prevalence of VTEC was similar to the value obtained in the first sample. The relatively high prevalence in the first month of life can be associated to the epidemiological situation in cow-calf herds where calves are exposed immediately after birth to the microbiological environment of an entire herd. Increasing prevalences of VTEC with age were also observed specifically for E. coli O157 in dairy calves (Garber et al., 1995; Roy, 1980) and may be the consequence of a decreasing maternal immunity. The risk of an infection relative to the result of the previous sample did not indicate any specific pattern and supports the hypothesis of the transient nature of VTEC infections in cattle (Besser et al., 1997; Pell, 1997). Based on the prevalence at weaning of 39% it appears that weaned calves raised in cow-calf farms are probably still more affected than adult animals (Wilson et al., 1992; Griffin and Tauxe, 1994; Burnens et al., 1995). Possible reasons for these findings are not known but may be related to different immunity status, effects of certain feeding strategies and different levels of ruminal development. (Mechie et al., 1997). In contrast with reports specific for E. coli O157, prevalences in summer were lower than in winter (Hancock et al., 1997). These observations and the fact that animals raised in open barns were significantly less at risk of an infection are consistent with experimental data indicating that E. coli O157 : H7 survives better in colder environments (Kearney et al., 1993). However, the samples were always collected from the same animals and effects of age and season can, therefore, not be discriminated. With respect to effects of breed of the dam it appears that calves from breeds with a relatively high potential of milk production such as Simmental and Brown Swiss, are significantly more likely to shed VTEC than calves from beef breed cattle. A detrimental effect of higher milk intake on the intestinal flora and/or differences of colostral immunoglobulin concentrations between breeds (Penhale and Christie, 1969; Logan, 1977; Muller and Ellinger, 1981) can be used as a possible explanation of these findings. Similar to most of the other bacteria diagnosed in this study, treatments with antibiotics prior to the collection of the samples had significant negative effects on the amount of shedding VTEC. However, the strong positive effect of antiparasitic treatments on shedding VTEC remains inexplicable. Relatively few animals were treated with antiparasitics or non-

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

259

antibiotic medicaments and some feedstuff categories occurred rarely in the sample population (Table 2). Therefore, the corresponding estimates of relative risks have to be interpreted cautiously. However, it seemed that influences of feedstuff and feeding strategy on the prevalence of an infection were generally less pronounced for VTEC than for Campylobacter spp. 5.2. Campylobacter Several species of Campylobacter, particularly C. jejuni are recognized as major causes of human diarrheal disease. Campylobacter spp. are widespread and a large variety of wild and domestic animals are considered to be reservoirs for human diseases. Depending on geographical location, the prevalence of Campylobacter spp. in cattle varies from very low to 100% (Rosef et al., 1983; Warner et al., 1986). The most important sources of an infection in humans are the consumption of contaminated foods of animal origin or polluted water. Fresh poultry, raw milk and raw or undercooked beef represent special hazards of human illness. (Butzler and Oosterom, 1991; Doyle and Jones, 1992). The resistance of Campylobacter spp. to physical factors is relatively low. The germs cannot grow at environmental temperatures and survive in feces, in water and on foods for only a few days (Butzler and Oosterom, 1991; Doyle and Jones, 1992; Norcross et al., 1992). It has to be considered however, that cooling and freezing can lead to a long survival of Campylobacter spp. in foods and that an infective dose of 500 cells can be sufficient to produce human illness. Therefore, Campylobacter spp. have to be regarded as an important hazard of food borne diseases in cow-calf farms particularly on farms where meat is processed and commercialized on site. It is evident that C. jejuni are a responsible factor of diarrhea in several animal species. However, the role of C. jejuni in calf diarrhea is not entirely assured. Experimental data (Al-Mashat and Taylor, 1980, 1983; Firehammer and Myers, 1981) indicate that infections with C. jejuni can induce diarrhea in calves, whereas in other studies no differences in prevalence between healthy and diseased calves were observed (Weber et al., 1985; Snodgrass et al., 1986; Busato et al., 1998). In the present investigation the average prevalence of C. jejuni over all the four fecal samples was 32% which is slightly above other observations made in several types of cattle populations (HaÈnninen and Raevuori, 1981; Prescott and Bruin-Mosch, 1981; Weber et al., 1984, 1985; Giacoboni et al., 1993). A possible explanation of these high prevalences may lie in the epidemiological situation of cow-calf herds where animal to animal contacts occur more frequently than in traditional dairy barns with cows attached and calves raised separately in boxes. Other possible reasons could be different diagnostic procedures between studies, geographical and seasonal variations. The average prevalence of C. jejuni increased significantly within the first three samples but at weaning the prevalence was three times lower than the initial value at 1 month of age. It can be hypothesized that this time trend of infection is the result of an increasing exposition due to the progressing time of confinement to barns during the winter months and a much lower probability of exposition in summer when calves were on pasture. Consequently the assessment of infections risks from sample to sample indicated no signs of development of immunity against an intestinal

260

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

colonization with C. jejuni during the three first months of life. The finding of higher prevalences in winter are supported by similar observations made in German cattle populations (Weber et al., 1985). In contrast to the other species of campylobacters, calves raised in colder and dryer open barns seemed to be significantly more at risk to be infected than calves raised in closed barns. These results are not consistent with evidence from in vitro experiments, indicating that C. jejuni survives in feces better when kept in a colder environment. (Blaser et al., 1980; Nachamkin, 1997). The risk of an C. jejuni infection was significantly lower in larger herds, a finding which is particular for C. jejuni. For a more extended interpretation it would be necessary to perform further studies with measuring the exact area available per animal and determining the specific housing conditions of each herd. For all Campylobacter spp. detected in this study no common and partially contradicting patterns of odds ratios were observed with respect to treatments prior to sample collection and to effects of feedstuff compositions and feeding strategies. The epidemiological features of C. coli and C. jejuni appear to be very similar and the two species are sometimes not further differentiated in the literature. However, C. coli is less frequently involved in human illness than C. jejuni (Rosef et al., 1983; Karmali et al., 1983). With respect to human disease it appears that swine populations are an important natural reservoir of C. coli whereas cattle populations seem to play only a minor role in this context. The average prevalence of C. coli in this study was 3% which reflects the results of other work (Prescott and Munroe, 1982; Munroe et al., 1983; Rosef et al., 1983; Weber et al., 1985). C. fetus have long been recognized as a cause of sporadic abortion in cattle and sheep and it appears that C. fetus are normal inhabitants of bovine intestinal tract (Garcia et al., 1983). Our data indicate a relatively constant presence of 13±18% of C. fetus in calves up to 3 months of life. At weaning the prevalences were considerable lower. Similar infection patterns although higher in prevalence, were observed in Japanese cattle populations, where 12% prevalence in adult animals and 27% in calves are reported (Giacoboni et al., 1993). Reports of infections with C. hyointestinalis in cattle are rare and there is no clear evidence of a pathogenic role (Snodgrass et al., 1986; Diker et al., 1990). Comparable to other data the average prevalence of C. hyointestinalis in this study was 14% (samples 1±4). Thus C. hyointestinalis were second to C. jejuni as the most commonly diagnosed campylobacter infection. 5.3. Yersinia The fact that 1.7% of all samples tested positive for Yersinia spp. confirms other findings indicating that infections with Yersinia spp. may be occasionally found in cattle populations (Slee et al., 1988; Steiner et al., 1997). Similar to other enteropathogenic bacteria, yersiniae have reservoirs in livestock and can, therefore, contaminate foods derived from these animals. The results of the present study indicate a constant but low prevalence of Yersinia pseudotubeculosis in Swiss cow-calf herds. In contrast to other observations (Slee et al., 1988) the infection appears to be limited to animals younger than 8 months. However, there is evidence that infections with Yersinia spp. are more

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

261

prevalent in winter (Hodges and Carman, 1985) and it is possible that the observed variation of prevalence is not related to age but to seasonal influence. 6. Conclusions The study showed relatively high prevalences of food borne enteric pathogens in calves raised in cow-calf herds. In most cases the bacteria were isolated in fecal specimen taken from clinically normal animals. No common risk profiles between VTEC and C. jejuni were observed with respect to individual as well as environmental effects and management factors. The study provided important knowledge that cow-calf herds have to be recognized as significant reservoirs of zoonotic enteropathogenic bacteria. Hygienic aspects and measures to prevent the microbiological contamination of carcasses and finished products have to be ensured at each level in the process of production and distribution of beef, particularly on farms with own processing facilities and direct commercialization. Acknowledgements The authors acknowledge the excellent technical assistance provided by M. Krawinkler. The project was funded by the Swiss Federal Veterinary Office, the Swiss Association of Artificial Insemination and the Swiss Beef Cattle Breeding Association. References Al-Mashat, R.R., Taylor, D.J., 1980. Production of diarrhoea and dysentery in experimental calves by feeding pure cultures of Campylobacter fetus subspecies jejuni. Vet. Rec. 107, 459±464. Al-Mashat, R.R., Taylor, D.J., 1983. Production of enteritis in calves by the oral inoculation of pure cultures of Campylobacter fetus subspecies intestinalis. Vet. Rec. 112, 54±58. Besser, T.E., Hancock, D.D., Pritchett, L.C., McRae, E.M., Rice, D.H., Tarr, P.I., 1997. Duration of detection of fecal excretion of Escherichia coli O157:H7 in cattle. J. Infect. Dis. 175, 726±729. Blanco, M., Blanco, J.E., Blanco, J., Gonzales, E.A., . Alonso, M.P., Maas, H., Jansen, W.H., 1996. Prevalence and characteristics of human and bovine verotoxgenic Escherichia coli strains isolated in Galicia (northwestern Spain). Euro. J. Epidemiol. 12, 13±19. Blaser, M.J., Hardesty, H.L., Powers, B., Wang, W.L., 1980. Survival of Campylobacter fetus subsp. jejuni in biological milieus. J. Clin. Microbiol. 11, 309±313. Boss, P., Monckton, R.P., Nicolet, J., Burnens, A.P., 1992. Nachweis von Toxigenen Verschiedener E. coli beim Schwein mit nichtradioaktiv markierten. Sonden. Schweiz. Arch. Tierheilk. 134±137. Brogan, D.J., 1998. Pitfalls of using standard statistical software packages for sample survey data. Encyclopedia of Biostatistics. Wiley, New York (in press). Burnens, A.P., Stanley, J., Schaad, U.B., Nicolet, J., 1993. Novel Campylobacter-like organism resembling Helicobacter fennelliae isolated from a boy with gastroenteritis and from dogs. J. Clin. Microbiol. 31, 1916± 1917. Burnens, A.P., Frey, A., Lior, H., Nicolet, J., 1995. Prevalence and clinical significance of vero-cytotoxinproducing Escherichia coli (VTEC) isolated from cattle in herds with and without calf diarrhoea. J. Vet. Med. B 42, 311±318.

262

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

Busato, A., Lentze, T., Hofer, D., Burnens, A., Hentrich, B., Gaillard, C., 1998. A case control study of potential enteric pathogens for calves raised in cow-calf herds. J. Vet. Med. B 45, 519±528. Butzler, J.P., Oosterom, J., 1991. Campylobacter: pathogenicity and significance in foods. Int. J. F. Mic. 12, 1±8. Diker, K.S., Diker, S., Ozlem, M.B., 1990. Bovine diarrhea associated with Campylobacter hyointestinalis. J. Vet. Med. B. 37, 158±160. Doyle, M.P., Jones, D.M., 1992. Food- borne transmission and antibiotic resistance of Campylobacter jejuni. In: Nachamkin, I., Blaser, M.J. (Eds.), Campylobacter jejuni: Current Status and Future Trends, pp. 45±48. Firehammer, B.D., Myers, L.L., 1981. Campylobacter fetus subsp jejuni: Its possible significance in enteric disease of calves and lambs. Am. J. Vet. Res. 42, 918±922. Garber, L.P., Wells, S.J., Hancock, D.D., Doyle, M.P., Tuttle, J., Shere, J.A., Zhao, T., 1995. Risk factors for fecal shedding of Escherichia coli O157:H7 in dairy calves. J. Am. Vet. Med. Assoc. 207, 46±49. Garcia, M.M., Eaglesome, M.D., Rirby, C., 1983. Campylobacter important in veterinary medicine. Vet. Bull. 53, 793±818. Giacoboni, G.I., Itoh, K., Hirayama, K., Takahashi, E., Mitsuoka, T., 1993. Comparison of fecal campylobacter in calves and cattle of different ages and areas in Japan. J. Vet. Med. Sci. 55, 555±559. Griffin, P.M., Tauxe, R.V., 1991. The epidemiology of infections caused by Escherichia coli O157:H7, other enterohemorrhagic E. coli, and the associated hemolytic syndrome. Epidemiol. Rev. 13, 60±98. Gyles, C.L., 1994: Escherichia coliverotoxins and other cytotoxins. In: Gyles, C.L. (Ed.), E. coliInfection In Man and Domesticated Animals. Wallingford, UK, pp. 365±398. Hancock, D.D., Besser, T.E., Rice, D.H., Herriot, D.E., Tarr, P.I., 1997. A longitudinal study of Escherichia coli O157 in fourteen cattle herds. Epidemiol. Infect. 118, 193±195. HaÈnninen, M.L., Raevuori, M., 1981. Occurence of Campylobacter fetus subsp. jejuni and Yersinia enterocolitica in domestic animals and in some foods of animal origin in Finland. . Nord. Vet. Med. 33, 441±445. Hodges, R.T., Carman, M.G., 1985. Recovery of Yersinia pseudotuberculosis from the faeces of healthy cattle. NZ Vet. J. 33, 175±176. Karmali, M.A., Penner, J.L., Fleming, P.C., Williams, A., Hennessy, J.N., 1983. The serotype and biotype distribution of clinical isolates of Campylobacter jejuni and Campylobacter coli over a three-year period. J. Infect. Dis. 147, 243±246. Karmali, M.A., 1989. Infection by verocytotoxin producing Escherichia coli. Clin. Microbiol. Rev. 2, 15±38. Kearney, T.E., Larkin, M.J., Levett, P.N., 1993. The effect of slurry storage and anaerobic digestion on survival of pathogenic bacteria. J. Appl. Bacteriol. 74, 86±93. Kehl, K.S., Havens, P., Behnke, C.E., Acheson, D.W.K., 1997. Evaluation of the premier EHEC assay for the detection of Shiga toxin-producing Escherichia coli. J. Clin. Microbiol. 35, 2051±2054. Levy, P.S., Lemeshow, S., 1991. Sampling of Populations. Wiley, New York. Logan, E.F., 1977. The influence of husbandry on colostrum yield and immunoglobulin concentration in beef cows. Br. Vet. J. 133, 120±125. Mechie, S.C., Chapman, P.A., Siddons, C.A., 1997. A fifteen month study of Escherichia coli O157:H7 in a dairy herd. Epidemiol. Infect. 118, 17±25. Muller, L.D., Ellinger, D.K., 1981. Colostral immunoglobulin concentrations among breeds of dairy cattle. J. Dairy Sci. 64, 1727±1730. Munroe, D.L., Prescott, J.F., Penner, J., 1983. Campylobacter jejuni and Campylobacter coli serotypes isolated from chickens, cattle and pigs. J. Clin. Microbiol. 18, 877±881. Nachamkin, I., 1997. Campylobacter jejuni. In: Doyle, M.P., Beuchat, L.R., Montville T.J. (Eds.), Food Microbiology: Fundamentals and Frontiers. Washington DC, USA, pp. 159±170. Norcross, M.A., Johnston, R.W., Brown, J.L., 1992. Importance of Campylobacter spp. to the food industry. In: Nachamkin, I., Blaser, M.J. (Eds.), Campylobacter jejuni: Current Status and Future Trends. Washington DC, USA pp. 61±65. Pell, A.N., 1997. Manure and microbes: Public and animal health problem? J. Dairy Sci. 80, 2673±2681. Penhale, W.J., Christie, G., 1969. Quantitative studies on bovine immunglobulins I. Adult plasma and colostrum levels. Res. Vet. Sci. 10, 493±501. Prescott, J.F., Bruin-Mosch, C.W., 1981. Carriage of Campylobacter jejuni in healthy and diarrheic animals. Am. J. Vet. Res. 42, 164±165.

A. Busato et al. / Veterinary Microbiology 69 (1999) 251±263

263

Prescott, J.F., Munroe, D.L., 1982. Campylobacter jejuni enteritis in man and domestic animals. J. Am. Vet. Med. Assoc. 181, 1524±1530. Rahn, K., Renwick, S.A., Johnson, R.P., Wilson, J.B., Clarke, R.C., Alves, D., Mcewen, S., Lior, H., Spika, J., 1997. Persistance of Escherichia coli O157:H7 in dairy cattle and the dairy farm environment. Epidemiol. Infect. 119, 251±259. Rosef, O., Gondrosen, B., Kapperud, G., Underdal, B., 1983. Isolation and characterization of Campylobacter jejuni and Campylobacter coli from domestic and wild mammals in Norway. Appl. Environ. 46, 855±859. Roy, J.H.B., 1980. Factors affecting susceptibility of calves to disease. J. Dairy Sci. 63, 650±664. Slee, K.J., Brightling, P., Seiler, R.J., 1988. Enteritis in cattle due to Yersinia pseudotuberculosis infection. Aust. Vet. J. 65, 271±275. Snodgrass, D.R., Terzolo, H.R., Sherwood, D., Campbell, I., Menzies, J.D., Synge, B.A., 1986. Aetiology of diarrhoea in young calves. Vet. Rec. 119, 31±34. Steiner, L., Busato, A., Burnens, A., Gaillard, C., 1997. HaÈufigkeiten und Ursachen von KaÈlberverlusten und KaÈlberkrankheiten in Mutterkuhbetrieben II Mikrobiologische und parasitologische Diagnosen bei KaÈlbern mit Durchfall. Dtsch. TieraÈrztl. Wschr. 104, 169±173. Warner, D.P., Bryner, J.H., Beran, G.W., 1986. Epidemiologic study of campylobacteriosis in Iowa cattle and the possible role of unpasteurized milk as a vehicle of infection. Am. J. Vet. Res. 47, 254±258. Weber, A., Bergmann, I., Bauer, K., 1984. Nachweis von Campylobacter jejuni in Kotproben von KaÈlbern mit und ohne Enteritiden. Berl. MuÈnch. TieraÈrztl. Wschr. 97, 10±13. Weber, A., Lembke, C., SchaÈfer, R., 1985. Ueber die jahreszeitliche HaÈufigkeit des Nachweises von Campylobacter jejuni und Campylobacter coli in Kotproben gesunder Schlachtrinder. Zbl. Vet. Med. B. 32, 197±201. Wilson, J.B., McEwen, S.A., Clarke, R.C., Leslie, K.E., Waltner-Toews, D., Gyles, C.L., 1992. A case-control study of selected pathogens including verocytotoxigenic Escherichia coli in calf diarrhea on an Ontario veal farm. Can. J. Vet. Res. 56, 184±188.