Prevalence of antimicrobial resistance in fecal generic Escherichia coli isolated in western Canadian beef herds. Part II — Cows and cow-calf pairs

Abstract

The objective of this study was to describe antimicrobial resistance in fecal generic Escherichia coli isolated from cows and cow-calf pairs from western Canadian beef herds. Susceptibility testing was completed on 1555 E. coli isolates (n = 533 cows from 69 herds) harvested from fresh fecal samples in the spring of 2002, and 630 isolates (n = 105 cow-calf pairs from 10 herds) collected in the spring of 2003. Only 1 cow isolate was resistant to an antimicrobial classified by Health Canada as being of very high importance to human medicine. Resistance to at least 2 antimicrobials was detected in 7.1% of the 2002 cow isolates, in 3.4% of the 2003 cow isolates, and 23.2% of the 2003 calf isolates. In the cows, resistance to at least 1 antimicrobial was not associated with cow breed (P = 0.16), cow age (P = 0.14), or previous cow treatment (P = 0.56). In the calves, resistance to at least 1 antimicrobial was not predicted by whether or not its dam was resistant to at least 1 antimicrobial (P = 0.36).

Introduction

Cow-calf herds in western Canada are subjected to less intensive management practices and different antimicrobial exposures than livestock species raised more intensively, such as poultry, swine, or cattle in dairies or feedlots (1). The potential difference in selective pressure within cow-calf herds could, therefore, result in micro-organisms that have a different profile of antimicrobial resistance (AMR) compared with other types of livestock. There is currently no information on AMR in the cow-calf industry in western Canada. The objective of this study was to describe the prevalence and patterns of AMR in fecal Escherichia coli isolated from cows and cow-calf pairs from western Canadian beef herds; the associations between individual cow risk factors and AMR; and the associations between AMR in cows and in calves. This study was undertaken to complement a study of AMR in beef calves (2), presented in Part I, as a part of research investigating AMR and AMU in cow-calf herds (1,2) (Figure 1).

Figure 1.

Sampling structure for study of AMR in western Canadian cow-calf herds.

Materials and methods

Herd and animal selection

Herd selection is described in Part I (2), and Figure 1 herein. For this study, veterinarians were asked to collect up to 10 cow fecal samples per herd from a convenience sample of 69 herds enrolled in the larger productivity study. The number of samples collected, however, was limited by the accessibility of the animals, and the time available for sampling during the herd visit. Samples were obtained from those cows that could be individually identified and defecated while being observed. A separate disposable glove and container were used for each animal. In 2002, all cow fecal samples were collected during a single visit to each herd between April and June. In 2003, fecal samples were collected from a convenience sample of 10 cow-calf pairs on each of 10 farms during a single visit to each herd between March and June. Cow-calf pairs were handled through a chute system; fecal samples were obtained directly per rectum. A separate disposable glove and container were used for each sample. Health status was evaluated by the study veterinarian at the time of sample collection, and recorded as healthy or unhealthy with appropriate comments.

Laboratory methods

The materials and methods used in this study are described in Part I (2). Briefly, fecal samples were cultured for generic E. coli (Prairie Diagnostic Services, Saskatoon, Saskatchewan). A minimum of 3 isolates per sample identified as E. coli were selected and stored at −80°C. Isolates were tested for susceptibility using broth micro-dilution (Sensititre; TREK Diagnostic Systems, Cleveland, Ohio, USA) and the standard 2002 National Antimicrobial Resistance Monitoring System (NARMS) panel (3). All testing was done in accordance with Clinical Laboratory Standards Institute (CLSI) guidelines (4,5). The significance of the MIC results for each antimicrobial tested were classified according to the Veterinary Drug Directorate, Health Canada (Tables I–III) (6). Multiple resistance was defined as resistance to ≥ 2 antimicrobials.

Table I.

Minimum inhibitory concentrations (MICs) for fecal generic Escherichia coli isolates recovered from cows in the spring of 2002 arranged by the Veterinary Drug Directorate, Health Canada, classification of drugs (n = 1555)

Table III.

Minimum inhibitory concentrations (MICs) for generic fecal Escherichia coli isolates collected from the calves of the cow-calf pair samples in the spring of 2003, the Veterinary Drug Directorate, Health Canada, classification of drugs (n = 318)

Statistical analysis

The approach used for data manipulation and population averaged prevalence estimates are described in detail in Part I (2). The cow data collected in 2002 were analyzed and are presented separately from the cow-calf pair data collected in 2003. All models were developed using generalized estimating equations (GEE) to account for clustering within herd [SAS v.8.2 for Windows (PROC GENMOD); SAS Institute, Cary, North Carolina, USA]. Model specifications included a binomial distribution, logit link function, repeated statement with subject equal to herd, and an exchangeable correlation structure.

Samples collected from cows in 2002 were examined for unconditional associations between AMR and animal level risk factors of interest. The AMR outcomes (yes/no) included resistance to any antimicrobial or ≥ 2 antimicrobials. Resistance to specific antimicrobials (tetracycline and sulphamethoxazole) was also investigated; these antimicrobials being selected because they are the most commonly observed resistances, and resistance to other drugs is very rare. The risk factor variables examined included: cow breed (British, continental, or crossbred), cow age (2-year-old heifers, 3-year-old cows, 4- to 10-year-old cows, and cows > 10 y of age), and whether the cow was ever treated with any oral or injectable antimicrobial, excluding ionophores, between January 1, 2002 and the date of sample collection.

The methods for calf fecal sample collection and data collected in 2002 are described in Part I (2). In the spring of 2002, herds with both cow and calf fecal samples (Figure 1) (N = 37), were examined to determine if the proportion of cows in the herd with resistance to sulphamethoxazole and tetracycline was associated with the proportion of calves in the herd (number of AMR positive calves/number of calves sampled) with resistance to the same antimicrobials. The relative difference between the proportion of cows and calves sampled in each herd that had any AMR positive isolates was also determined. The total number of samples with any type of AMR (numerator) as a proportion of the total number of animals sampled (denominator) was determined, first between cows and calves in the spring of 2002, and then in a separate model between cows in the spring of 2002 and calves in the fall of 2002 (Figure 1). The regression coefficient for the fixed effect for age group was used to estimate the odds ratio (OR) and 95% confidence interval (95% CI) describing how much more or less likely fecal samples from calves were to be positive for AMR than those from cows.

For the cow-calf pair data collected in 2003, the proportion of animals with any AMR positive isolates over the total number of animals sampled per herd was compared between calves and cows. Also, the association between any resistance in the dam’s isolates (yes/no) and any resistance in the calf’s isolates (yes/no) was examined.

Results

Cows (2002 sampling)

Cow age ranged from 2 to 14 y [median, 5 y; interquartile range (IQR), 3 to 8]. Median herd size was 154 (range, 71 to 437) breeding females. The median number of samples collected per herd was 8 (range, 2 to 10; IQR, 6 to 10). Before sample collection, producers reported that 4.1% (22/533) of cows had been treated with antimicrobials between January 1, 2002 and the time of sample collection. Time between last treatment and sample collection ranged from 6 d to 147 d (median, 37 d).

Resistance to at least 1 antimicrobial was identified in 9.8% of the 1555 isolates, in 15.1% of cows, and in 61% of herds (Tables IV–VI). Tetracycline and sulphamethoxazole were the antimicrobials to which resistance was most commonly identified (Tables IV–VI). For all other drugs, the frequency of resistance was < 2% (Tables IV and V). No resistance was identified to ceftriaxone, ciprofloxacin, and nalidixic acid. The maximum number of antimicrobials to which an isolate demonstrated resistance was 11; 0.5% (7/1555) of isolates were resistant to at least 5 antimicrobials. The most common multiresistance pattern was ampicillin-chloramphenicol-streptomycin-sulphamethoxazole-tetracycline. Three of the 7 isolates with resistance to 5 or more antimicrobials exhibited this pattern. No resistance was detected to the Category I antimicrobials except for 1 isolate resistant to ceftiofur (Table I). The median minimum inhibitory concentration (MIC) for all antimicrobials, with the exception of streptomycin, was several dilutions below the breakpoint.

Table IV.

Prevalence (%) of AMR in Escherichia coli isolates recovered from cows (n = 1555) in the spring of 2002 and for cows (n = 312) and calves (n = 318) in the spring of 2003

	Cows 2002			Pair cows 2003			Pair calves 2003
		95% Confidence interval			95% Confidence interval			95% Confidence interval
Antimicrobial	Prevalence (%)	Lower CI	Upper CI	Prevalence (%)	Lower CI	Upper CI	Prevalence (%)	Lower CI	Upper CI
Amikacin	0.0	0.0	0.3	0.0	0.0	1.5	0.0	0.0	1.5
Amoxicillin-clavulanic acid	0.3	0.1	0.8	0.0	0.0	1.5	3.9	1.1	13.0
Ampicillin	0.7	0.2	2.3	2.0	0.3	12.0	8.7	4.1	17.7
Cefoxtin	0.2	0.0	0.8	0.0	0.0	1.5	3.2	0.7	13.1
Ceftiofur	0.1	0.0	0.4	0.0	0.0	1.5	2.9	0.6	13.8
Ceftriaxone	0.0	0.0	0.3	0.0	0.0	1.5	0.0	0.0	1.5
Cephalothin	0.3	0.1	0.8	0.7	0.1	4.1	6.0	2.7	12.7
Chloramphenicol	0.8	0.2	3.0	0.0	0.0	1.5	7.9	3.4	17.1
Ciprofloxacin	0.0	0.0	0.3	0.0	0.0	1.5	0.0	0.0	1.5
Gentamicin	0.2	0.1	0.6	0.0	0.0	1.5	0.3	0.0	2.0
Kanamycin	0.7	0.2	2.9	2.7	0.6	10.4	5.9	1.8	17.3
Nalidixic acid	0.0	0.0	0.3	0.0	0.0	1.5	0.0	0.0	1.5
Streptomycin	3.4	1.2	9.3	3.4	1.1	9.7	13.0	7.2	22.3
Sulphamethoxazole	7.1	3.4	14.2	4.4	1.8	10.2	24.3	15.2	36.5
Tetracylcine	8.7	4.4	16.5	5.1	2.2	11.2	20.6	12.5	32.1
Trimethoprim-sulphamethoxazole	0.5	0.2	1.7	1.7	0.3	10.1	7.3	2.7	18.2
AMR (≥ 1 antimicrobial)	9.8	5.1	18.2	6.1	3.1	11.7	25.8	16.6	37.8
Multi AMR (≥ 2 antimicrobials)	7.1	3.2	15.0	3.4	1.1	9.7	23.2	14.3	35.3
A3Ca	0.1	0.0	0.4	0.0	0.0	1.5	2.9	0.6	13.8
ACSSuTb	0.1	0.0	0.4	0.0	0.0	1.5	0.9	0.3	3.5
AKSSuTc	0.1	0.0	0.4	2.0	0.3	12.0	1.2	0.3	5.0
ACKSSuTd	0.2	0.0	1.3	0.0	0.0	1.5	2.9	0.6	13.8

CI — confidence interval.

^aA3C — ampicillin, cefoxitin, ceftiofur, cephalothin.

^bACSSuT — ampicillin, chloramphenicol, streptomycin, sulphamethoxazole, and tetracycline.

^cAKSSuT — ampicillin, kanamycin, streptomycin, sulphamethoxazole, and tetracycline.

^dACKSSuT — ampicillin, chloramphenicol, kanamycin, streptomycin, sulphamethoxazole, and tetracycline.

Table VI.

Herd prevalence (%) of AMR for cows (N = 69 herds) sampled in the spring of 2002 and for cows (N = 10 herds) and calves (N = 10 herds) of the cow-calf pairs sampled in the spring of 2003

	Cows 2002			Pair cows 2003			Pair calves 2003
		95% Confidence interval			95% Confidence interval			95% Confidence interval
Antimicrobial	Prevalence (%)	Lower CI	Upper CI	Prevalence (%)	Lower CI	Upper CI	Prevalence (%)	Lower CI	Upper CI
Amikacin	0.0	0.0	6.5	0.0	0.0	34.5	0.0	0.0	34.5
Amoxicillin-clavulanic acid	5.8	2.2	14.5	0.0	0.0	34.5	30.0	10.0	62.4
Ampicillin	11.6	5.9	21.5	10.0	1.4	46.7	50.0	22.5	77.5
Cefoxtin	4.4	1.4	12.6	0.0	0.0	34.5	30.0	10.0	62.4
Ceftiofur	1.5	0.2	9.6	0.0	0.0	34.5	20.0	5.0	54.1
Ceftriaxone	0.0	0.0	6.5	0.0	0.0	34.5	0.0	0.0	34.5
Cephalothin	5.8	2.2	14.5	10.0	1.4	46.7	70.0	37.6	90.0
Chloramphenicol	10.2	4.9	19.8	0.0	0.0	34.5	50.0	22.5	77.5
Ciprofloxacin	0.0	0.0	6.5	0.0	0.0	34.5	0.0	0.0	34.5
Gentamicin	2.9	0.7	10.9	0.0	0.0	34.5	10.0	1.4	46.7
Kanamycin	10.2	4.9	19.8	20.0	5.0	54.1	40.0	15.8	70.3
Nalidixic acid	0.0	0.0	6.5	0.0	0.0	34.5	0.0	0.0	34.5
Streptomycin	34.8	24.5	46.7	40.0	15.8	70.3	80.0	45.9	95.0
Sulphamethoxazole	49.3	37.7	60.9	50.0	22.5	77.5	100.0	50.0	50.0
Tetracylcine	53.6	41.9	65.0	50.0	22.5	77.5	100.0	50.0	50.0
Trimethoprim-sulphamethoxazole	8.7	4.0	18.0	10.0	1.4	46.7	40.0	15.8	70.3
AMR (≥ 1 antimicrobial)	60.9	49.0	71.6	60.0	29.7	84.2	100.0	N/A	N/A
Multi AMR (≥ 2 antimicrobials)	53.0	41.9	65.0	40.0	15.8	70.3	100.0	N/A	N/A
A3Ca	1.5	0.2	9.6	0.0	0.0	34.5	20.0	5.0	54.1
ACSSuTb	1.5	0.2	9.6	0.0	0.0	34.5	20.0	5.0	54.1
AKSSuTc	1.5	0.2	9.6	10.0	1.4	46.7	20.0	5.0	54.1
ACKSSuTd	2.9	0.7	10.9	0.0	0.0	34.5	20.0	5.0	54.1

CI — confidence interval.

^aA3C — ampicillin, cefoxitin, ceftiofur, cephalothin.

^bACSSuT — ampicillin, chloramphenicol, streptomycin, sulphamethoxazole, and tetracycline.

^cAKSSuT — ampicillin, kanamycin, streptomycin, sulphamethoxazole, and tetracycline.

^dACKSSuT — ampicillin, chloramphenicol, kanamycin, streptomycin, sulphamethoxazole, and tetracycline.

Table V.

Prevalence (%) of AMR in cows (n = 533) sampled in the spring of 2002 and for cows (n = 105) and calves (n = 105) of the cow-calf pairs sampled in the spring of 2003

	Cows 2002			Pair cows 2003			Pair calves 2003
		95% Confidence interval			95% Confidence interval			95% Confidence interval
Antimicrobial	Prevalence (%)	Lower CI	Upper CI	Prevalence (%)	Lower CI	Upper CI	Prevalence (%)	Lower CI	Upper CI
Amikacin	0.0	0.0	0.9	0.0	0.0	4.4	0.0	0.0	4.4
Amoxicillin-clavulanic acid	0.8	0.3	1.9	0.0	0.0	0.0	4.9	1.6	14.2
Ampicillin	1.5	0.8	2.9	2.0	0.3	11.8	11.7	5.7	22.3
Cefoxtin	0.6	0.2	1.7	0.0	0.0	4.4	4.9	1.6	14.2
Ceftiofur	0.2	0.0	1.3	0.0	0.0	4.4	3.8	0.9	14.3
Ceftriaxone	0.0	0.0	0.9	0.0	0.0	4.4	0.0	0.0	4.4
Cephalothin	0.8	0.3	1.9	0.9	0.1	6.0	10.5	5.9	17.9
Chloramphenicol	1.7	0.8	3.6	0.0	0.0	4.4	10.2	4.3	22.1
Ciprofloxacin	0.0	0.0	0.9	0.0	0.0	4.4	0.0	0.0	4.4
Gentamicin	0.4	0.1	1.5	0.0	0.0	4.4	0.9	0.1	6.0
Kanamycin	1.3	0.7	2.6	2.8	0.7	10.1	7.7	2.5	21.4
Nalidixic acid	0.0	0.0	0.9	0.0	0.0	4.4	0.0	0.0	4.4
Streptomycin	6.7	4.5	9.9	4.8	2.1	10.5	20.7	12.4	32.6
Sulphamethoxazole	11.1	8.0	15.1	5.7	2.9	10.8	33.8	21.3	49.0
Tetracylcine	13.4	10.2	17.5	7.7	3.5	16.2	31.6	20.6	45.2
Trimethoprim-sulphamethoxazole	1.7	0.7	4.3	2.0	0.3	11.8	9.2	3.6	21.2
AMR (≥ 1 antimicrobial)	15.1	11.7	19.3	8.6	4.3	16.3	37.9	25.3	52.4
Multi AMR (≥ 2 antimicrobials)	11.6	8.6	15.6	4.8	2.1	10.5	32.7	21.5	46.2
A3Ca	0.2	0.0	1.3	0.0	0.0	4.4	3.8	0.9	14.3
ACSSuTb	0.2	0.0	1.3	0.0	0.0	4.4	1.9	0.5	6.5
AKSSuTc	0.2	0.0	1.3	2.0	0.3	11.8	2.1	0.7	6.3
ACKSSuTd	0.4	0.1	1.4	0.0	0.0	4.4	3.8	0.9	14.3

CI — confidence interval.

^aA3C — ampicillin, cefoxitin, ceftiofur, cephalothin.

^bACSSuT — ampicillin, chloramphenicol, streptomycin, sulphamethoxazole, and tetracycline.

^cAKSSuT — ampicillin, kanamycin, streptomycin, sulphamethoxazole, and tetracycline.

^dACKSSuT — ampicillin, chloramphenicol, kanamycin, streptomycin, sulphamethoxazole, and tetracycline.

The detection of any AMR or multiple AMR was not associated with cow breed (P = 0.16; P = 0.11), cow age (P = 0.14, P = 0.42), or previous cow antimicrobial treatment (P = 0.56, P = 0.32). Tetracycline and sulphamethoxazole resistance were not associated with cow breed (P = 0.09, P = 0.45), cow age (P = 0.20, P = 0.22), or previous cow antimicrobial treatment (P = 0.44, P = 0.28).

Cow-calf pairs (2003 sampling)

The median number of samples collected per herd was 10 (range, 9 to 16; IQR, 10 to 10), and the median herd size was 130 (range, 86 to 382) breeding females. Cow age ranged from 2 to 19 y of age (median, 5 y; IQR, 3 to 8). Ninety-two percent of the cows were classified as healthy at the time of sample collection. Fifty-eight percent (61/105) of the calves sampled were male, and 91.4% (96/105) of the calves were classified as healthy at sample collection. Median calf age was 47 d (range, 1 to 129 d; IQR; 28 to 60).

Of the 312 isolates recovered from the cow fecal samples in 2003, 6.1% were resistant to at least 1 antimicrobial (Table IV); 8.6% of cows had at least 1 resistant isolate as did 60% of the herds (Tables V and VI). Most of the resistance was to tetracycline and sulphamethoxazole. The maximum number of antimicrobials that an isolate was resistant to was 7; 1.9% (6/312) of isolates were resistant to at least 5 antimicrobials. The most common multiresistance pattern was ampicillin-kanamycin-streptomycin-sulphamethoxazole- tetracycline. No resistance was identified to the Category I antimicrobials. The median MIC for all antimicrobials was several dilutions below the breakpoint (Table II), except for streptomycin, which had a median MIC in the dilution immediately below the breakpoint.

Table II.

Minimum inhibitory concentrations for generic fecal Escherichia coli isolates recovered from the cows of the cow-calf pair samples in the spring of 2003, arranged by the Veterinary Drug Directorate, Health Canada, classification of drugs (n = 312)

Of the 318 calf isolates, 25.8% were resistant to at least 1 antimicrobial (Table IV). The prevalences of calves and herds with at least 1 resistant isolate were 37.9% and 100%, respectively (Tables V and VI). Most of the resistance was to tetracycline and sulphamethoxazole; no resistance was identified to ceftriaxone, nalidixic acid, and ciprofloxacin. The maximum number of antimicrobials that an isolate was resistant to was 12; 9.1% (29/318) of isolates and 12.3% (13/105) of calves demonstrated resistance to at least 5 antimicrobials. The most common multiresistance pattern was ampicillin-choramphenicol-kanamycin-streptomycin-sulphamethoxazole-tetracycline.

Ceftiofur was the only Category I antimicrobial to which isolates demonstrated resistance (Table III). As for the cows, the median MIC for all antimicrobials, except streptomycin, was several dilutions below the breakpoint (Table III).

Resistant E. coli were identified in both the cow and the calf for only 4.8% (5/105) of the pairs and 3 of 10 farms. Three of the 5 resistant pairs had isolates resistant to ≥ 2 antimicrobials. Tetracycline was the most common resistance identified in 4 of 5 pairs. There was no association between the detection of any resistant isolates in the calf and the detection of any resistant isolates in its dam (P = 0.36).

Comparison of AMR prevalence between cows and calves

The prevalence of AMR was lower in isolates collected from beef cows in the spring than in spring calves from the same herds. In the 10 herds described previously with cow-calf pair data, the calves were 7.1 (95% CI, 2.8 to 18.3; P < 0.0001) times more likely to have had at least 1 resistant isolate detected than the cows (Figure 1). A similar trend was detected for 37 herds that had samples collected in the spring of 2002 from both calves (2) and cows (Figure 1). The calves examined were 10.0 (95% CI, 5.8 to 17.0; P < 0.0001) times more likely to have had at least 1 resistant isolate detected than the cows from the same herds. The median number of samples collected from the 37 herds (212 calves and 299 cows) was 13 (IQR, 12 to 17; range, 7 to 20).

There was no difference, however, in the proportion of cows with resistance when compared with calves sampled in the fall from the same herds [odds ratio (OR), 1.1; 95% CI, 0.3 to 3.7; P = 0.91] (Figure 1). The median number of samples from these 9 herds (81 calves and 74 cows) was 19 (IQR, 15 to 20; range, 9 to 20).

For the 37 herds that had both cow and calf samples collected in the spring of 2002, the odds that calves would have at least isolate resistant to sulphamethoxazole increased with the proportion of cows that had at least 1 isolate resistant to sulphamethoxazole (OR, 7.5; 95% CI, 1.3 to 41.7; P = 0.02). A similar increase in the odds of tetracycline resistance in the calves was seen with an increasing proportion of cows being positive for tetracycline resistance (OR, 6.1; 95% CI, 1.5 to 25.3; P = 0.01).

Discussion

This study, together with the companion study of calves in cow-calf herds described in Part I (2), provides a baseline overview of the occurrence of AMR in cow-calf herds in western Canada. The prevalence of AMR was relatively low in the study cows, particularly to drugs classified by Health Canada as being of very high importance to human medicine (6). Cows were much less likely to shed resistant E. coli than very young (spring) calves; however, cows and calves had similar AMR patterns in the fall. Other key findings were that the AMR status of the individual calf was not associated with its dam, and that the frequency of common types of resistance in the calves was associated with exposure from the cow herd.

Escherichia coli isolates from both the cows and calves were most commonly resistant to tetracycline and sulphamethoxazole. This was further explored by considering whether the proportion of cows in the herd with either tetracycline or sulphamethoxazole resistance was predictive of these same resistances in young calves sampled at the same time (spring). The association between resistance in the cow herd and the occurrence of the same types of resistance in the calves indicates that young calves might acquire resistant strains through contact with the cow herd or from fecal contamination of the environment by the cow herd. The data from the cow-calf pair study suggests that the AMR status of the calf’s dam is less important than the herd environment in determining the calf’s AMR status.

The most common resistances observed were consistent with previous reports for E. coli from a variety of different animal species (2,7,8,9). The difference in AMR prevalence among species (9–11) may result from variation in selection pressure among different animal industries. The frequency of resistance was low in generic fecal E. coli isolates harvested from beef cows, especially to antimicrobials classified as important to human health. No ciprofloxacin or ceftriaxone resistance was present, and only 1 ceftiofur resistant isolate was identified from the cow samples. The median MICs for the majority of the antimicrobials tested were well below the breakpoint for resistance, indicating that most of the E. coli in this study population was highly susceptible the drugs on the panel.

There are slight differences in the AMR prevalence estimates between the 2002 and 2003 cow samples, but the distribution of drugs to which resistance was detected is very similar. The confidence intervals for the 2 prevalence estimates overlap, suggesting that the difference was not significant. In the cow-calf pair study, only 105 animals from 10 herds were sampled, while over 500 animals from 69 herds were sampled for the cow study. There were also differences in the prevalence estimates between the calves from the cow-calf pair study in 2003 (25.8%) and a larger sample of beef calves from the companion study conducted in the spring of 2002 (48.8%) (2), but again, the pattern of drugs to which resistance was detected was very similar in the 2 populations. The 2002 studies involving larger populations of cows and calves were probably more representative of western Canadian cow-calf herds. The calves sampled in 2003 pairs study were older on average (median, 47 d) than the calves sampled in 2002 (median, 6 d).

While the prevalence of AMR in the cows was significantly lower than that observed in young calves in the spring, the prevalence estimates from cows were similar to those of older calves from the same herds sampled in the fall (2). These studies were not specifically designed to study the effect of age on the prevalence of resistant E. coli in beef cattle; however, it appears that there may be an age-related difference between the pre-ruminant calves and older animals in each study [see Part I (2)]. With the exception of a lower risk of AMR in calves < 3 d of age, there were no individual animal risk factors identified for AMR in either calves or cows in either this or the previous study (2).

The intention of the current study was to assess cattle in cow-calf herds at a time when animals were potentially under the highest stress. The study was not designed to assess the effect of seasonal variation. The calving season is often the period when antimicrobial treatment is most common because animals are most susceptible to disease as a result of crowding, confinement, stress associated with calving, and potentially adverse weather conditions.

One limitation of this and the associated calf study (2) was the use of convenience sampling rather than having a formal random sampling for selecting herds and animals within herds. Access to the herds and the necessary calving and treatment records were provided through a larger study looking at factors affecting productivity in beef herds (1). Because herds were enrolled in the larger productivity study based on their ability to provide the required data, these herds probably represent some of the more progressive, commercially viable, and intensively managed herds in western Canada (1,2). The study herds enrolled in this project were larger than the national reported averages (2). The data from the herds in this study, therefore, best reflect AMR in this segment of the industry.

The findings of this study indicate that on-farm exposure to beef cows probably poses a relatively low risk as a source of AMR for human health. However, additional molecular studies would provide more insight into what AMR genes are being carried in these populations. Follow-up monitoring is needed to detect emerging resistance issues of public health concern in this commodity.