Antimicrobial Resistance in Fecal Escherichia coli and Salmonella enterica from Dairy Calves: A Systematic Review

Abstract

The discovery of antibiotics brought with it many advances in the health and well-being of humans and animals; however, in recent years development of antimicrobial resistance (AMR) has increasingly become a concern. Much of the antibiotic use on dairy farms is for disease management in mature cattle, and AMR in fecal organisms is relatively rare in this group. However, young dairy calves often carry high levels of AMR in their fecal Escherichia coli and Salmonella enterica, which could provide a potential reservoir of AMR genes on dairy farms. To develop practical and effective antibiotic stewardship policies for dairy calf rearing, it is vital to have a solid understanding of the current state of knowledge regarding AMR in these animals. A systematic review process was used to summarize the current scientific literature regarding AMR in fecal S. enterica and E. coli and associations between management practices and AMR prevalence in dairy calves in the United States and Canada. Seven online databases were searched for literature published from 1997 to 2018. Multiple studies indicated an association between preweaned calves and increased risk of fecal shedding of resistant bacteria, compared to other animal groups on dairy farms. There also was evidence, although less consistent, of an impact of antibiotic treatment, antibiotic-containing milk replacer feeding, and feeding nonsalable or waste milk (WM) on the presence of AMR bacteria. Overall, the research summarized in this systematic review highlights the need for continued research on the impact of management practices, including antibiotic use, WM feeding, and disease prevention practices in reducing AMR in E. coli and S. enterica in dairy calves. In addition, few data were available on physiological and microbiological factors that may contribute to the high relative populations of resistant bacteria in young calves, suggesting another valuable area of future research.

Introduction

As in humans, the use of antibiotics in dairy cattle has greatly benefited health, welfare, and productivity; however, their use is not without the possibility of adverse consequences. Antimicrobial resistance (AMR) is a concern in both human and animal populations. Despite uncertainty in the evidence linking AMR development in animals to AMR in humans (Muloi et al., 2018), the use of antibiotics in animal agriculture has come under scrutiny. A World Health Organization-funded meta-analysis concluded that practices restricting antibiotic use in animal agriculture are associated with reduced AMR in animals and humans, suggesting that improved antibiotic stewardship could curb AMR risks in animal agriculture (Tang et al., 2017).

Antibiotic stewardship on dairy farms often focuses on mastitis management, the cause for the majority of antibiotic use (USDA-APHIS, 2008). However, despite a greater frequency and volume of antibiotic use in adult cows, there is growing concern because dairy calves carry proportionally higher levels of resistance in commensal Escherichia coli and Salmonella enterica (Edrington et al., 2008; Berge et al., 2010; Cao, 2015). It is vital to have a solid understanding of the current knowledge of AMR in dairy calves to develop practical and effective antibiotic stewardship practices. A systematic review process was chosen to allow an in-depth rigorous assessment of the currently published research. The objective of this review was to identify areas of consensus and to summarize data to aid in the development of evidence-based antibiotic stewardship programs, as well as to identify areas in need of additional research relating to AMR in dairy calves in the United States and Canada.

Materials and Methods

Literature search

A systematic literature search of seven electronic databases (Fig. 1) was conducted using Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (Moher et al., 2009). The search terms included descriptors of dairy cattle, drug resistance, and Salmonella or E. coli (Table 1). Searches were limited to publications from 1997 to the date searches were conducted (February 2018).

FIG. 1.

Flow diagram depicting selection of studies for inclusion.

Table 1.

Systematic Search Parameters

Search terms

(((((calf or cattle or heifer or bovine or calves)) AND dairy) AND (antibiotic resistance or antimicrobial resistance or drug resistance)) AND (E. coli or Escherichia coli or salmonella))

Inclusion criteria

Fecal or environmental manure samples from dairy-breed heifer or bull calves were collected.

Calves were raised in the United States or Canada.

Data for calves aged birth to 12 weeks of age were reported, and these data were presented distinctly from animals of other ages.

A qualitative or quantitative measure of AMR to at least one class of antibiotics was included.

AMR was measured in S. enterica and/or E. coli.

Exclusion criteria

Only diagnostic samples were analyzed.

The study was published in a language other than English.

The study did not report original research.

Quality criteria

Animal housing and management represented commercial herd conditions.

AMR was measured and reported in a manner that allowed adequate evaluation of the population or intervention.

The type of statistical analysis was appropriate for the study design.

AMR, antimicrobial resistance.

Study selection and data extraction

Following the removal of duplicate records, two of the authors assessed each article for eligibility. Studies were eligible if two authors agreed that all inclusion and exclusion criteria were satisfied (Table 1). For each eligible study, two authors evaluated the quality of the research based on predetermined criteria, which, when met, allowed inclusion in the final review (Table 1). Dissertations and theses were assessed on a chapter-by-chapter basis. Chapters published in peer-reviewed journals were excluded in favor of the peer-reviewed article. Data extraction was performed by one author with a second checking for accuracy and completeness. Due to the diversity of data collection and reporting among the studies, a meta-analysis was not performed.

Results

Search

The results of the systematic literature search are summarized in Figure 1. The 36 articles included in this review comprised 23 observational and 13 experimental studies. AMR was reported for E. coli in 19 studies, for S. enterica in 8, and for both in 9. Methods of measuring AMR varied, with 22 studies using broth dilution minimum inhibitory concentration, 15 utilizing disc diffusion, four reporting AMR genes, and one using break point agar. Multiple methods were used in four articles. Multidrug resistance (MDR) was defined differently across studies, ranging from resistance to more than one antibiotic, to resistance to five or more antibiotics or three or more classes of antibiotics.

Age of calf

Age frequently was reported as a significant risk factor associated with AMR in E. coli, with preweaned calves exhibiting greater risk of resistance compared to other age groups (Table 2) (Berge et al., 2005, 2010; Galvao et al., 2005; Villarroel, 2007; Awosile et al., 2018). Donaldson et al. (2006) reported decreasing counts of ceftiofur-resistant Gram-negative enteric bacteria (90% of which were E. coli) as calves aged (p < 0.001). Cao (2015) reported the greatest farm-level prevalence of resistant, MDR (≥3 classes of antimicrobials), and bla _CMY-2-carrying E. coli in preweaned calves, but the prevalence of E. coli carrying the bla _CTX-M gene was low and more evenly distributed across age groups (Table 3). Among young calves, maximal risk of resistance typically was reported at around 2 weeks of age in calves not fed antibiotic-containing medicated milk replacer (MMR) (Berge et al., 2005, 2006a; Galvao et al., 2005; Awosile et al., 2018).

Table 2.

Summary of Multivariable Logistic Regression from Studies with Outcomes Related to Age of Calf

Reference	Sampling	Risk factors	Statistics
Awosile et al. (2018)	Cohort study. Four hundred eighty-eight Holstein calves on eight dairy farms. Rectal swab collected at 2–15 and 42–56 d of age	Frequency of recovery of suspected extended-spectrum cephalosporin-resistant E. coli	OR	SE	p-Value
Awosile et al. (2018)		2–15 d of age (compared to 45–56 d of age)	2.42	0.31	<0.001
Berge et al. (2005)	Cohort study. One hundred forty-nine calves in a total of five cohorts from three dairy farms. Rectal swab collected at 1–2 d of age, then every 2 weeks through 6 weeks of age	E. coli belonging to a cluster with higher level of antibiotic resistance	OR	95% CI	p-Value
		Calf age 2 weeks (compared to 0 weeks)	53.6	38.0–75.7	<0.0001
		Calf age 4 weeks (compared to 0 weeks)	29.8	21.6–41.0	<0.0001
		Calf age 6 weeks (compared to 0 weeks)	16.4	11.9–22.5	<0.0001
Berge et al. (2010)	Cross-sectional study. Rectal swab collected from up to 22 calves (2–4 weeks), and other groups at multiple types of operations, each visited once or twice	Increasing multiple resistance in fecal E. coli
		Dairy calf (compared to dairy cow)	23.62	16.48–33.85	p < 0.01

Berge et al. (2006a)	Experimental study. One hundred twenty dairy-source bull calves enrolled into one of four treatment groups, two fed MMR, the other two fed non-MMR. Escherichia coli isolated from rectal swabs. Calves followed through 28 d of age	Increasing multiple resistance in fecal E. coli
		Non-MMR-fed calves (age 14 d compared to 1 d)	10.18	5.69–18.3	<0.0001
		Non-MMR-fed calves (age 28 d compared to 1 d)	5.16	2.82–9.41	<0.0001
		MMR-fed calves (age 14 d compared to 1 d)	608.5	186.27–1989.03	<0.0001
		MMR-fed calves (age 28 d compared to 1 d)	645.87	196.25–2125.79	<0.0001
Galvao et al. (2005)	Experimental study. Fifty-two Holstein calves with failure of passive transfer (serum IgG <0.73 g/dL). Experimental treatments: Live yeast cultures in milk, starter, both, or none. Rectal swabs analyzed. Data from all calves presented here	Isolation of commensal E. coli with increasing multidrug resistance
		Calf age 25 d (compared to 13 d)	0.21	0.12–0.37	0.0001
		Calf age 38 d (compared to 13 d)	0.09	0.05–0.15	0.0001
		Calf age 59 d (compared to 13 d)	0.18	0.11–0.31	0.0001
		Calf age 74 d (compared to 13 d)	0.22	0.13–0.38	0.0001
Villarroel (2007)	Longitudinal study. Calves from a single 1200 cow dairy. Fifty calves that had been treated and 50 age-matched calves that had not been treated were sampled (50 g feces) at ∼8-week intervals. At each visit, an additional 5–10 calves per pen were randomly sampled	Isolation of resistant E. coli
		Calf age 0–60 d (compared to >180 d)	3.39	2.58–4.46	<0.05
		Calf age 61–120 d (compared to >180 d)	1.45	1.18–1.77	<0.05
		Calf age 121–180 d (compared to >180 d)	1.42	1.16–1.73	<0.05
		Isolation of MDR (≥3 antimicrobials) E. coli
		Calf age 0–60 d (compared to >180 d)	9.66	7.01–13.30	<0.05
		Calf age 61–120 d (compared to >180 d)	1.96	1.53–2.50	<0.05
		Calf age 121–180 d (compared to >180 d)	1.43	1.13–1.79	<0.05
		Isolation of resistant S. enterica
		0–45 d Individually housed (compared to 75–300 d group housed)	26.58	8.69–81.29	<0.05

MMR, medicated milk replacer; MDR, multidrug resistance; IgG, immunoglobulin G; OR, odds ratio; CI, confidence interval; SE, standard error.

Table 3.

Summary of Prevalence Data for Studies with Outcomes Related to Age of Calf

Reference	Sampling	Variable	Preweaned calves	Postweaned calves	Lactating cows	Dry cows	Sick cows
Cao (2015)	Cross-sectional study. Manure composite sample from 80 Pennsylvania dairy farms. For each farm: three lactating cow samples and one sample per group for other groups collected	Recovery of resistant E. coli (% farms)	85%	70%	44%	28%
		Recovery of MDR (≥3 classes of antimicrobials) E. coli (% farms)	70%	39%	20%	6%
		Prevalence of E. coli carrying bla _CMY-2 gene (% farms)	31%	13%	4%	0%
		Prevalence of E. coli carrying bla _CTX-M gene (% farms)	1.3%	1.3%	1.4%	0.0%
Cummings et al. (2009)	Cross-sectional study. Fecal samples from suspect Salmonella cases on 831 farms in NY, PA, VT, MA, and CT	Prevalence of MDR (≥5 antimicrobials) S. enterica isolates	77.0%		65.2%
Edrington et al. (2008)	Cross-sectional study. Four commercial dairies (≥2000 head) in Southwestern United States Calves from all four dairies housed at centralized calf-rearing facility. Fecal samples collected on two visits from the various groups	Proportion of S. enterica isolates resistant to ≥4 antibiotics	82%		65%		83%
Habing et al. (2012)	Retro-prospective study. Thirty-one MI dairy farms sampled in 2000 and 2009. Fecal and environmental samples were collected every other month	Proportion of S. enterica isolates resistant to at least one antibiotic	30.5% (40/131)		2.3% (7/309)
Ray et al. (2007)	Cohort study. One hundred ten herds in MI, MN, NY, and WI sampled at 2-month intervals from August 2000 to October 2001	Proportion of S. enterica isolates resistant to ≥5 antibiotics	25.30%		4.20%		13.3%

MDR, multidrug resistance; NY, New York; PA, Pennsylvania; VT, Vermont; MA, Massachusetts; CT, Connecticut; MI, Michigan; MN, Minnesota; WI, Wisconsin.

Several studies demonstrated a greater proportion of resistant or MDR S. enterica isolated from calves compared to lactating cows (Table 3) (Ray et al., 2007; Edrington et al., 2008; Habing et al., 2012). In suspected salmonellosis cases, prevalence of MDR (≥5 antimicrobials) S. enterica isolates was also higher in calves than lactating cows (Cummings et al., 2009). Edrington et al. (2011) reported no difference between pre- and postweaned dairy calves; however, this study was sampled at 2 d before and 2 d after weaning, which may have influenced the ability to detect a difference between these groups. Cao (2015) reported low levels of resistance in S. enterica across all age groups sampled.

Antibiotic treatment

The impact of antibiotic treatment on resistance in E. coli was reported in several studies (Berge et al., 2005, 2006a; Villarroel, 2007; Pereira et al., 2014c; Liu et al., 2016; Awosile et al., 2018). Increasing risk of AMR in E. coli, by various measures, was reported in some, but not all instances for studies reporting impact of treatment before sampling (Table 4) (Berge et al., 2005, 2006a; Villarroel, 2007; Pereira et al., 2014c). Berge et al. (2006a) reported an increased risk of isolating more highly resistant E. coli from non-MMR-fed calves treated within 5 days of sampling, but not in treated calves fed MMR. The latter group had a highly resistant E. coli population, possibly masking any impact of parenteral treatment. One study reported treatment as a risk factor for isolation of resistant S. enterica; however, this was significant at 3 weeks and 2 months post-treatment, but not at 2 or 4 weeks post-treatment (Villarroel, 2007).

Table 4.

Summary of Multivariable Logistic Regression from Studies with Outcomes Related to Antibiotic Treatment

Reference	Sampling	Risk factors	Statistics
Awosile et al. (2018)	Cohort study. Four hundred eighty-eight Holstein calves on eight dairy farms. Rectal swab collected at 2–15 and 42–56 d of age	Frequency of recovery of suspected extended-spectrum cephalosporin-resistant E. coli	OR	SE	p-Value
		Regular use of ceftiofur (compared to occasional)	3.83	1.00	<0.001
		Florfenicol use (yes compared to no)	2.02	0.36	<0.001
		Use of ceftiofur for respiratory disease (yes compared to no)	0.57	0.09	0.001
Berge et al. (2005)	Cohort study. One hundred forty-nine calves in a total of five cohorts from three dairy farms. Rectal swab collected at 1–2 d of age, then every 2–6 weeks of age	E. coli belonging to a cluster with higher level of antibiotic resistance	OR	95% CI	p-Value
Berge et al. (2005)		Antibiotic treatment within 5 days of sampling (yes compared to no)	2.0	1.4–3.0	0.0003
Berge et al. (2006a)	Experimental study. One hundred twenty dairy-source bull calves enrolled into one of four treatment groups, two fed MMR, the other two fed non-MMR. E. coli isolated from rectal swabs. Calves followed through 28 d of age	Increasing multiple resistance in fecal E. coli
		Non-MMR-fed calves treated with an antibiotic w/in 5 d of sampling (yes compared to no)	3.03	1.15–7.98	0.02
		MMR-fed calves treated with an antibiotic w/in 5 d of sampling (yes compared to no)	0.51	0.21–1.24	ns
Pereira et al. (2014c)	Cohort study. Rectal swabs from 479 calves from eight NY dairy farms were sampled at a single visit to each farm. Calves were grouped by age (3–14 d of age, 15–35 d of age, 36–65 d of age). All OR ns for ceftiofur-treated calves	Calf-level isolation of 3 compared to 0 nonsusceptible E. coli isolates (enrofloxacin treated compared to no)
		Nalidixic acid	41	10–172	<0.0001
		Ciprofloxacin	92	14–613	<0.0001
		Isolate level isolation of nonsusceptible E. coli (enrofloxacin treated compared to no)
		Ciprofloxacin	2.0	1–2	0.005
		Nalidixic acid	3.0	2–6	<0.0001
		Ceftriaxone	0.8	0.5–1	ns
		Neomycin	0.9	0.6–1	ns
Villarroel (2007)	Longitudinal study. Calves from a single 1200 cow dairy. Fifty calves that had been treated and 50 age-matched calves that had not been treated were sampled (50 g feces) at ∼8-week intervals. At each visit, an additional 5–10 calves per pen were randomly sampled	Isolation of resistant E. coli
		Tx 1 week ago vs. no tx	4.56	1.57–13.24	≤0.05
		Tx 2 week ago vs. no tx	8.24	2.9–23.41	≤0.05
		Tx 3 week ago vs. no tx	3.68	1.70–7.94	≤0.05
		Tx 4 week ago vs. no tx	1.83	1.06–3.15	≤0.05
		Isolation of MDR (≥3 antimicrobials) E. coli
		Tx 1 week ago vs. no tx	4.67	2.00–10.89	≤0.05
		Tx 2 week ago vs. no tx	3.05	1.68–5.52	≤0.05
		Tx 3 week ago vs. no tx	1.84	1.04–3.24	≤0.05
		Isolation of resistant S. enterica
		Tx 3 week ago vs. no tx	2.51	1.07–5.92	≤0.05
		Tx longer than 2 months vs. no tx	4.23	1.54–11.62	≤0.05

MMR, medicated milk replacer; MDR, multidrug resistance; ns, non-significant; NY, New York; OR, odds ratio; CI, confidence interval; SE, standard error, Tx, treatment.

Liu et al. (2016) demonstrated that calves treated with therapeutic doses of ceftiofur or florfenicol had elevated numbers of fecal E. coli resistant to the treatment medication (p = 0.005, p = 0.001), but this was not observed in oxytetracycline-treated calves. Florfenicol-treated calves also had increased isolation of ceftiofur-resistant fecal E. coli (r = 0.5; p < 0.05), and untreated calves in a neighboring pen had increased prevalence of florfenicol-resistant fecal E. coli compared to more remotely housed calves (p < 0.001).

Medicated (antibiotic-containing) milk replacer

Several studies evaluated the impact of feeding MMR on AMR in fecal bacteria (Table 5) (Khachatryan et al., 2004, 2006; Berge et al., 2006a; Kaneene et al., 2008, 2009; Averill, 2009; Pereira et al., 2011). All evaluated the use of a tetracycline-class antibiotic alone or in combination with either neomycin or sulfamethazine and amprolium. Studies that only evaluated changes in tetracycline resistance had varied results, with reports of elevated resistance (Khachatryan et al., 2004), reduced resistance (Kaneene et al., 2008), and no difference (Averill, 2009) in AMR E. coli in MMR-fed calves compared to non-MMR-fed calves. In contrast, most studies measuring resistance to a wider variety of antibiotics demonstrated greater resistance to at least some classes of antibiotics and/or higher prevalence of MDR bacteria in MMR-fed calves compared to non-MMR-fed calves (Berge et al., 2006a; Khachatryan et al., 2006; Kaneene et al., 2009; Pereira et al., 2011). An exception demonstrated higher prevalence of resistance to chloramphenicol in calves fed milk replacer (MR) alone, compared to calves fed MR mixed with an added dietary supplement, regardless of whether it contained antibiotics or not (Khachatryan et al., 2006).

Table 5.

Summary of Results from Studies with Outcomes Related to Feeding of Antibiotic-Containing Milk Replacer

Reference	Medication	Study description	Statistics
Averill (2009)	Neomycin, tetracycline	Experimental study, 15 calves per group. One group fed medicated MR, control group fed same MR w/o antibiotics. Fecal samples from 30 calves collected at single time point (2 weeks of age). Single E. coli from each sample assessed for resistance	No significant difference in tetracycline resistance (all but one isolate had MIC >8 μg/mL). No other measures of resistance in E. coli reported
Berge et al. (2006a)	Neomycin, tetracycline	Experimental study, 30 calves/group. Groups 1–3 were not fed medicated MR, w/varying housing and treatment regimens. Group 4 fed medicated MR. No difference in resistance between groups 1–3, so reported together here. Fecal samples collected at 1, 14, and 28 d	Calves receiving antibiotics in milk had greater proportion of E. coli isolates belonging to the three most resistant phenotypes reported (48%) compared to calves receiving nonmedicated MR (4%; p < 0.0001)
Kaneene et al. (2008)	Neomycin, oxytetracycline	Intervention study, in eight dairy herds in NY and MI. Preintervention, all herds fed medicated MR, for 3 months. Half of farms (intervention herds) then switched to same MR w/o antibiotics during postintervention period. Data reported only for calf fecal samples, not environmental	Change in % of isolates susceptible to tetracycline between pre- and postintervention phases
			Intervention group had significant increase in E. coli tetracycline susceptibility between pre- and postintervention periods (2.0% and 9.3%, respectively, p = 0.0004)
			No significant difference in control group
			No significant difference in S. enterica isolates
			Tetracycline MIC50 during intervention phase
			No significant difference in E. coli tetracycline MIC50
			S. enterica from intervention group had significantly lower MIC50 than controls (36.7 μg/mL, 64.0 μg/mL, respectively; p < 0.0001)
Kaneene et al. (2009)	Neomycin, oxytetracycline	Same as Kaneene et al. (2008)	Proportion of MDR isolates (resistant to ≥5 antimicrobials)
			67.4% of E. coli isolates in control herds, 50.5% in intervention herds (p < 0.01; OR = 0.5; 95% CI = 0.4–0.6)
			94% of S. enterica isolates in control herds, 45.4% in intervention herds (p < 0.0001; OR = 0.05; 95% CI = 0.03–0.1)
Khachatryan et al. (2004)	Oxytetracycline	Experimental study. Calves fed bulk milk with or without TM-50 (terramycin) supplement. Calves were followed to 12 weeks of age, with weaning at 4–6 weeks of age	Prevalence of tetracycline resistance among E. coli isolates was higher than expected in the calves not receiving antibiotics (0.55 observed, 0.533 expected, p = 0.039)
Khachatryan et al. (2006)	Oxytetracycline	Retrospective study comparing fecal E. coli isolates from calves on the same WA dairy from 2001 and 2004. The addition of a dietary supplement containing oxytetracycline, dry milk, and vitamins to bulk tank milk for feeding calves was discontinued in 2003	Prevalence of E. coli isolates resistant to streptomycin (p < 0.001), sulfadiazine (p < 0.001), and tetracycline (p = 0.019) declined. There was no significant difference in prevalence of chloramphenicol-resistant isolates and a rise in ampicillin-resistant isolates (p < 0.001)
Khachatryan et al. (2006)	Oxytetracycline	Experimental study with three groups (n = 9), one fed milk with the oxytetracycline supplement described above, one fed milk with the supplement, but without antibiotics, and a control group fed no supplement in milk	No significant difference between any group for prevalence of resistance to ampicillin, streptomycin, sulfadiazine, and tetracycline in E. coli. Prevalence of chloramphenicol-resistant E. coli was higher in the nonsupplemented group than either of the groups receiving the supplement, either with or without antibiotics (p = 0.03)
Pereira et al. (2011)	Chlortetracycline, sulfamethazine, amprolium	Calves from three NY dairy farms studied. Two farms did not feed growth promoting antibiotics, the third did feed growth promoting antibiotics	E. coli isolates from the farm feeding antibiotics had smaller inhibition zone diameter for tetracycline, ampicillin, streptomycin, sulfamethoxazole–trimethoprim, amoxicillin–clavulanic acid, neomycin, ceftiofur, gentamycin, and cefepime (p < 0.05). There were no differences in inhibition zone diameter for nalidixic acid, enrofloxacin, or ciprofloxacin

MR, milk replacer; MIC, minimum inhibitory concentration; MDR, multidrug resistance; OR, odds ratio; CI, confidence interval.

Nonsalable or waste milk feeding

Antibiotic treatment of lactating cows generally requires that milk is withheld from sale due to the possibility of it containing violative antibiotic residues. On some dairies, this nonsalable, or waste milk (WM), is used to feed calves, which might present an opportunity for selection of AMR bacteria in the gut. Three studies assessed the impact of WM feeding on the presence of AMR E. coli in calf feces (Table 6) (Pereira et al., 2014b; Maynou et al., 2017; Awosile et al., 2018). In the two experimental studies, feeding regimens differed, but both found higher prevalence of resistant E. coli to some, but not all assessed antibiotics in the treatment group compared to non-WM-fed calves (Pereira et al., 2014b; Maynou et al., 2017); however, the study by Maynou et al. (2017) had large numbers of calves removed from analysis due to the need for therapeutic antibiotic treatment.

Table 6.

Summary of Results from Studies with Outcomes Related to Feeding of Nonsalable Milk or Similar

Reference	Diet	Description	Statistics
Awosile et al. (2018)	Feeding of unpasteurized nonsalable milk	Cohort study. Four hundred eighty-eight Holstein calves on eight dairy farms. Rectal swab collected at 2–15 and 42–56 d of age.	Predictors of recovery of suspected extended-spectrum cephalosporin-resistant E. coli (multivariable logistic regression). Feeding of unpasteurized, nonsalable milk (yes, compared to no) OR = 1.61, SE = 0.25 (p = 0002)
Maynou et al. (2017)	MR or pasteurized WM (pWM)	Experimental study. Following colostrum feeding, calves were fed either MR or pWM until weaning at d 49. Rectal swabs collected at d 1 (10 per group), d 35, and d 56. A total of 27 of 52 calves were removed from the study due to therapeutic antimicrobial administration, leaving 10 MR-fed calves and 16 pWM-fed calves for assessment.	Erythromycin (98.9%), penicillin (100%), and pirlimycin (99.4%) resistance high across all treatments.
			Prevalence of resistance to ampicillin and cephalothin rose between d 1 and 35, then fell between d 35 and 56 in pWM fed calves (p < 0.01), but not in MR fed calves.
			Prevalence of resistance to ceftiofur was greater in pWM-fed calves (p < 0.05)
			No difference between treatments for prevalence of florfenicol, tetracycline, streptomycin, or sulfamethoxazole–trimethoprim resistance
Pereira et al. (2014b)	BT milk or BT milk with ceftiofur, penicillin, oxytetracycline and ampicillin added at expected waste milk concentrations (DR, drug residue).	Experimental study. Following colostrum feeding, calves were fed either BT milk or DR milk. Three feeding trials with 5 calves per treatment in each trial were conducted. Rectal fecal samples collected before treatments (week 0) and weekly through 6 weeks of age.	Larger proportion of E. coli from DR-fed calves resistant to ampicillin, cefoxitin, and tetracycline for week 1–6 (p < 0.05) and streptomycin on weeks 3 and 5 (p < 0.05).
			Larger proportion of fecal E. coli from DR-fed calves nonsusceptible to ceftiofur (weeks 1–6) and ceftriaxone (weeks 2–6) (p < 0.05)
			Larger proportion of E. coli from DR-fed calves was resistant to 3 or more antimicrobial drugs for weeks 1–6 (p < 0.05).
			No difference in prevalence of resistance between treatments for neomycin and trimethoprim–sulfamethoxazole.

MR, milk replacer; WM, waste milk; BT, salable bulk tank; OR, odds ratio; SE, standard error.

Other (uncategorized)

Several studies did not fit in the previous categories, but yielded valuable information for this review. A comparison of AMR E. coli between farms housing calves in individual versus group pens demonstrated a greater risk of resistance to some antibiotics in group-housed calves (ciprofloxacin, nalidixic acid; p ≤ 0.001) and others (ampicillin, ceftiofur, gentamycin, streptomycin, tetracycline; p ≤ 0.04) in individually-housed calves (Pereira et al., 2014d). In a comparison between calves on calf ranches and dairy farms, E. coli isolated from the former had a greater likelihood of being more highly resistant (odds ratio = 2.4, 95% confidence interval = 1.6–3.6; p < 0.0001) (Berge et al., 2005). E. coli isolates from calf ranch calves were also shown to be more strongly associated with phenotypes that could not be explained by common resistance genes (Davis et al., 2011). A study of the emergence of E. coli carrying the bla _CTX-M gene in Washington State demonstrated a higher prevalence of bla _CTX-M on farms with >3000 adult cows, that moved animals onto the premises, where fresh bedding was added less than weekly, or where residual fly sprays were not used (Davis et al., 2015). An assessment of farm management factors on California dairies and calf ranches indicated that none was significantly associated with the presence of MDR S. enterica (Berge et al., 2006b).

Descriptive studies were included to capture a full understanding of the current knowledge of AMR in dairy calves. Several studies reported data suggesting that patterns of AMR could vary widely from farm to farm during a salmonellosis outbreak, but often remained consistent within the farm (Bischoff et al., 2004; Alexander et al., 2009; Kaneene et al., 2010). During an outbreak in calves on a single dairy, 86% of S. enterica isolates were serovar Oranienburg, of which 97% were resistant to nine or more antimicrobials (Kaneene et al., 2010). In addition, a report from two farms experiencing an outbreak in calves detailed that S. enterica isolates from one farm were all serovar Typhimurium var. Copenhagen with a core pattern of resistance to nine antibiotics, while on the other farm, all isolates were serovar Typhimurium, with 36/39 isolates considered pan-susceptible (Alexander et al., 2009). In contrast, 65% of isolates collected from calves on six different operations displayed a core resistance pattern to four antibiotics, with or without resistance to other antimicrobials (Berge et al., 2003). In a case–control study comparing case farms with a history of salmonellosis caused by highly resistant S. enterica to farms with no salmonellosis history, calves from case farms shed E. coli isolates with a slightly higher prevalence of resistance to some antibiotics, but not others (DeFrancesco et al., 2004). A longitudinal study of calves on five Michigan dairy farms found the highest prevalence of resistance among E. coli strains to spectinomycin, penicillin, and clindamycin (Ferguson, 2000).

Two studies reported in vitro transfer of resistance from calf-derived bacteria (Bischoff et al., 2004; Jiang et al., 2006). Resistance was transferred to E. coli from calf-derived ampicillin-resistant Salmonella Kinshasa, with evidence of cotransfer of resistance to other antibiotics (Bischoff et al., 2004). In vitro transfer of both bla _CMY2 and the class 1 integron to various Salmonella serovars from calf-derived S. enterica and E. coli isolates was also reported (Jiang et al., 2006). However, an in vivo transfer study did not indicate the transfer of resistance between naturally occurring MDR E. coli in dairy calves and cattle-origin Salmonella Newport or Reading (Edrington et al., 2013). In a study of in vivo fitness, an E. coli strain carrying the predominant resistance pattern among calves on the study farm was demonstrated to have an in vivo fitness advantage in neonatal calves, but not older heifers. (Khachatryan, 2005).

Discussion

The observation that preweaned dairy calves carried a relatively high prevalence of AMR in fecal bacteria was a driving factor in conducting this review. The evidence supporting a difference in AMR prevalence between preweaned calves and other cattle on dairies was, for the most part, consistent across studies and was demonstrated in both E. coli and S. enterica. The breadth of these studies indicates that this phenomenon has been observed geographically across the United States and Canada and over time. However, the demonstration of an in vivo fitness advantage of a resistant strain of E. coli in neonatal calves that could not be demonstrated in older heifers, as well as the digestive differences between a pre- and postweaned ruminant, suggests that other physiological or microbiological factors may also be at play (Khachatryan, 2005). Further research into the factors involved in increased AMR presence in preweaned calves may bring valuable new insight into how to best manage AMR on dairy operations.

The full implementation of the Veterinary Feed Directive (VFD) in January of 2017 brought major changes to how medically-important in-feed antibiotics, including those used in milk or MR, are used in the United States by placing these drugs under the order of a veterinarian (FDA-CVM, 2013). The evidence summarized in this study may suggest that the use of MMR could select for resistant organisms. The VFD regulations may reduce the quantity of antibiotics used in milk-fed calves, but further research will be needed to determine their full impact on populations of AMR bacteria.

Parenteral administration of antibiotics also may select for AMR. In studies specifying the treatment antibiotic, it was demonstrated that different antibiotic treatment choices may impact resistant populations differently (Pereira et al., 2014c; Liu et al., 2016). In addition, Liu et al. (2016) demonstrated an increase in both florfenicol- and ceftiofur-resistant E. coli populations after florfenicol treatment, suggesting possible coselection of resistance. Although these studies show that the antibiotic chosen for treatment of disease in calves may play a role in selection pressure on E. coli populations, neither provides sufficient evidence to truly inform antibiotic selection. Further research on the impact of specific treatment antibiotics and the role of coselection on AMR organisms could inform prudent antibiotic selection to minimize the development of resistance, particularly to antimicrobials vital to human health.

Although treatment protocols are the hallmark of on-farm antimicrobial stewardship, practices to prevent and limit the spread of disease also play a role in reducing antimicrobial use. Davis et al. (2015) demonstrated that one of several management factors associated with an increased prevalence of the bla _CTX-M gene in E. coli was the movement of animals onto the premises, suggesting a potential role for between-farm biosecurity practices in reducing the spread of resistance. In addition, calf ranch calves, which are typically comingled from multiple sources, were shown to be at risk of shedding more highly resistant E. coli than calves raised on their farm of origin (Berge et al., 2005). Although there are many differences in management between these calf rearing systems, the comingling of calves from multiple sources may influence the prevalence of AMR E. coli on calf ranches. Within-farm animal contact may also impact resistant populations. Liu et al. (2016) demonstrated that untreated calves housed adjacent to florfenicol-treated calves shed E. coli with a higher prevalence of florfenicol resistance compared to more remotely housed calves. Improving our understanding of the impact of biosecurity, as well as other disease prevention practices, may help to strengthen antimicrobial stewardship programs on dairy farms.

Feed-based and parenteral antibiotics may be the primary routes for administering antibiotics to calves, but antibiotic residues in WM represent another potential risk for selection of resistance. The NAHMS 2007 Dairy Study found that over 30% of U.S. dairy farms fed WM to heifer calves (USDA-APHIS, 2010). The concentration of antibiotics in WM may be low (Pereira et al., 2014a), but even very low concentrations of antibiotics have been shown to select for resistant organisms (Gullberg et al., 2011). A cross-sectional evaluation of WM on 34 New York State dairies demonstrated that over 80% of samples were positive using a commercial antibiotic screening test, with ceftiofur determined to be the most common drug residue (Pereira et al., 2014a). Although an economic advantage of feeding pasteurized WM to dairy calves has been demonstrated (Godden et al., 2005), it is also important to understand the potential undesired impacts of this practice on AMR. Although few studies reported here assess the impact of WM on resistant organisms, these suggest that WM feeding may exert some selection pressure on fecal E. coli (Pereira et al., 2014b; Maynou et al., 2017). Further research into the impact of WM feeding on AMR in calves and the persistence of AMR beyond weaning will be valuable for further assessing the impact of this feeding practice.

Conclusion

The purpose of this systematic review was to aid in the development of antibiotic stewardship programs and to guide future research by summarizing evidence relating to AMR populations and associations between management factors and AMR in dairy calves on United States and Canadian farms. The heterogeneity in the methods used to both measure and report AMR made interpretation and comparison across studies challenging. The majority of studies reported phenotypic resistance; however, with the rapid evolution of genotypic methods, interpretation will continue to be complex, but will bring the potential to provide new insights as well. To develop effective antimicrobial stewardship programs, it is vital to have a strong understanding of factors that influence development, spread, and persistence of AMR in dairy calves. Future research to improve understanding of calf physiology and the impact of management factors on AMR will help inform evidence-based biosecurity, treatment, and management decisions to minimize AMR on dairy operations.

Footnotes

Acknowledgments

This work was supported, in part, by funding from the Pennsylvania Department of Agriculture (Agreement OSP# 191748) and the U.S. Department of Agriculture (Agreement OSP# 176619).

Disclosure Statement

No competing financial interests exist.

References

Alexander

, Warnick

, Cripps

, McDonough

, Grohn

, Wiedmann

, Reed

, James

, Soyer

, Ivanek

. Fecal shedding of, antimicrobial resistance in, and serologic response to SalmonellaTyphimurium in dairy calves. J Am Vet Med Assoc, 2009; 235:739–748.

Averill

. Use of Antimicrobial Agents in Dairy Calves and the Public Health Concern. Michigan State University, Ann Arbor, MI: ProQuest Dissertations Publishing, 2009.

Awosile

, McClure

, Sanchez

, Rodriguez-Lecompte

, Keefe

, Heider

. Salmonella enterica and extended-spectrum cephalosporin-resistant Escherichia coli recovered from Holstein dairy calves from 8 farms in New Brunswick, Canada. J Dairy Sci, 2018; 101:3271–3284.

Berge

ACB

, Atwill

, Sischo

. Assessing antibiotic resistance in fecal Escherichia coli in young calves using cluster analysis techniques. Prev Vet Med, 2003; 61:91–102.

Berge

ACB

, Atwill

, Sischo

. Animal and farm influences on the dynamics of antibiotic resistance in faecal Escherichia coli in young dairy calves. Prev Vet Med, 2005; 69:25–38.

Berge

ACB

, Hancock

, Sischo

, Besser

. Geographic, farm, and animal factors associated with multiple antimicrobial resistance in fecal Escherichia coli isolates from cattle in the western United States. J Am Vet Med Assoc, 2010; 236:1338–1344.

Berge

ACB

, Moore

, Sischo

. Field trial evaluating the influence of prophylactic and therapeutic antimicrobial administration on antimicrobial resistance of fecal Escherichia coli in dairy calves. Appl Environ Microbiol, 2006a;72:3872–3878.

Berge

ACB

, Moore

, Sischo

. Prevalence and antimicrobial resistance patterns of Salmonella enterica in preweaned calves from dairies and calf ranches. Am J Vet Res, 2006b;67:1580–1588.

Bischoff

, Edrington

, Callaway

, Genovese

, Nisbet

. Characterization of antimicrobial resistant Salmonella Kinshasa from dairy calves in Texas. Lett Appl Microbiol, 2004; 38:140–145.

10.

Cao

. Antimicrobial Resistance of Salmonella and E. coli from Pennsylvania Dairy Herds. University of Maryland, College Park, MD: ProQuest Dissertations Publishing, 2015.

11.

Cummings

, Warnick

, Alexander

, Cripps

, Grohn

, McDonough

, Nydam

, Reed

. The incidence of salmonellosis among dairy herds in the northeastern United States. J Dairy Sci, 2009; 92:3766–3774.

12.

Davis

, Besser

, Orfe

, Baker

KNK

, Lanier

, Broschat

, New

, Call

. Genotypic-phenotypic discrepancies between antibiotic resistance characteristics of Escherichia coli isolates from calves in management settings with high and low antibiotic use. Appl Environ Microbiol, 2011; 77:3293–3299.

13.

Davis

, Sischo

, Jones

, Moore

, Ahmed

, Short

, Besser

. Recent emergence of Escherichia coli with cephalosporin resistance conferred by bla _CTX-M on Washington state dairy farms. Appl Environ Microbiol, 2015; 81:4403–4410.

14.

DeFrancesco

, Cobbold

, Rice

, Besser

, Hancock

. Antimicrobial resistance of commensal Escherichia coli from dairy cattle associated with recent multi-resistant salmonellosis outbreaks. Vet Microbiol, 2004; 98:55–61.

15.

Donaldson

, Straley

, Hegde

, Sawant

, DebRoy

, Jayarao

. Molecular epidemiology of ceftiofur-resistant Escherichia coli isolates from dairy calves. Appl Environ Microbiol, 2006; 72:3940–3948.

16.

Edrington

, Callaway

, Anderson

, Nisbet

. Prevalence of multidrug-resistant Salmonella on commercial dairies utilizing a single heifer raising facility. J Food Prot, 2008; 71:27–34.

17.

Edrington

, Carter

, Farrow

, Islas

, Hagevoort

, Friend

, Callaway

, Anderson

, Nisbet

. Influence of weaning on fecal shedding of pathogenic bacteria in dairy calves. Foodborne Pathog Dis, 2011; 8:395–401.

18.

Edrington

, Farrow

, Hume

, Anderson

, Hagevoort

, Caldwell

, Callaway

, Anderson

, Nisbet

. Evaluation of the potential antimicrobial resistance transfer from a multi-drug resistant Escherichia coli to Salmonella in dairy calves. Curr Microbiol, 2013; 66:132–137.

19.

FDA-CVM. Guidance for Industry #213: New Animal Drugs and New Animal Drug Combination Products Administered in or on Medicated Feed or Drinking Water of Food-Producing Animals: Recommendations for Drug Sponsors for Voluntarily Aligning Product Use Conditions with GFI #209. U.S. Department of Health and Human Services, Rockville, MD: Center for Veterinary Medicine, Food and Drug Administration, 2013.

20.

Ferguson

. A Longitudinal Study of Select Bovine Pathogens and Other Factors which Affect Calf Health on 5 Michigan Dairy Farms. Michigan State University, Ann Arbor, MI: ProQuest Dissertations Publishing, 2000.

21.

Galvao

, Santos

JEP

, Coscioni

, Villasenor

, Sischo

, Berge

ACB

. Effect of feeding live yeast products to calves with failure of passive transfer on performance and patterns of antibiotic resistance in fecal Escherichia coli . Reprod Nutr Dev, 2005; 45:427–440.

22.

Godden

, Fetrow

, Feirtag

, Green

, Wells

. Economic analysis of feeding pasteurized nonsaleable milk versus conventional milk replacer to dairy calves. J Am Vet Med Assoc, 2005; 226:1547–1554.

23.

Gullberg

, Cao

, Berg

, Ilback

, Sandegren

, Huges

, Andersson

. Selection of resistant bacteria at very low antibiotic concentrations. PLoS Pathog, 2011; 7:e01918-14.

24.

Habing

, Lo

Y-J

, Kaneene

. Changes in the antimicrobial resistance profiles of Salmonella isolated from the same Michigan dairy farms in 2000 and 2009. Food Res Int, 2012; 45:919–924.

25.

Jiang

, Yang

, Dettman

, Doyle

. Analysis of fecal microbial flora for antibiotic resistance in ceftiofur-treated calves. Foodborne Pathog Dis, 2006; 3:355–365.

26.

Kaneene

, Miller

, May

, Hattey

. An outbreak of multidrug-resistant Salmonella enterica serotype Oranienburg in Michigan dairy calves. Foodborne Pathog Dis, 2010; 7:1193–1201.

27.

Kaneene

, Warnick

, Bolin

, Erskin

, May

, Miller

. Changes in tetracycline susceptibility of enteric bacteria following switching to nonmedicated milk replacer for dairy calves. J Clin Microbiol, 2008; 46:1968–1977.

28.

Kaneene

, Warnick

, Bolin

, Erskin

, May

, Miller

. Changes in multidrug resistance of enteric bacteria following an intervention to reduce antimicrobial resistance in dairy calves. J Clin Microbiol, 2009; 47:4109–4112.

29.

Khachatryan

. Mechanisms Responsible for the Maintenance of High Prevalence of Antimicrobial Drug-Resistant Escherichia coli in Dairy Calves. Washington State University, Ann Arbor, MT: ProQuest Dissertations Publishing, 2005.

30.

Khachatryan

, Besser

, Hancock

, Call

. Use of a nonmedicated dietary supplement correlates with increased prevalence of streptomycin-sulfa-tetracycline-resistant Escherichia coli on a dairy farm. Appl Environ Microbiol, 2006; 72:4583–4588.

31.

Khachatryan

, Hancock

, Besser

, Call

. Role of calf-adapted Escherichia coli in maintenance of antimicrobial drug resistance in dairy calves. Appl Environ Microbiol, 2004; 70:752–757.

32.

Liu

, Zhao

, Orfe

, Subbiah

, Call

. Soil-borne reservoirs of antibiotic-resistant bacteria are established following therapeutic treatment of dairy calves. Environ Microbiol, 2016; 18:557–564.

33.

Maynou

, Migura-Garcia

, Chester-Jones

, Ziegler

, Bach

, Terre

. Effects of feeding pasteurized waste milk to dairy calves on phenotypes and genotypes of antimicrobial resistance in fecal Escherichia coli isolates before and after weaning. J Dairy Sci, 2017; 100:7967–7979.

34.

Moher

, Liberati

, Tetzlaff

, Altman

, Prisma

Group

. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med, 2009; 6:e1000097.

35.

Muloi

, Ward

, Pedersen

, Fevre

, Woolhouse

MEJ

, van Bunnik

BAD

. Are food animals responsible for transfer of antimicrobial-resistant Escherichia coli or their resistance determinants to human populations? A systematic review. Foodborne Pathog Dis, 2018; 15:467–474.

36.

Pereira

, Santos

TMA

, Bicalho

, Caixeta

, Machado

, Bicalho

. Antimicrobial resistance and prevalence of virulence factor genes in fecal Escherichia coli of Holstein calves fed milk with and without antimicrobials. J Dairy Sci, 2011; 94:4556–4565.

37.

Pereira

, Siler

, Bicalho

, Warnick

. Multiresidue screening of milk withheld for sale at dairy farms in central New York State. J Dairy Sci, 2014a;97:1513–1519.

38.

Pereira

, Siler

, Bicalho

, Warnick

. In vivo selection of resistant E. coli after ingestion of milk with added drug residues. PLoS One, 2014b;9:e115223.

39.

Pereira

, Siler

, Ng

, Davis

, Grohn

, Warnick

. Effect of on-farm use of antimicrobial drugs on resistance in fecal Escherichia coli of preweaned dairy calves. J Dairy Sci, 2014c;97:7644–7654.

40.

Pereira

, Siler

, Ng

, Davis

, Warnick

. Effect of preweaned dairy calf housing system on antimicrobial resistance in commensal Escherichia coli. J Dairy Sci 2014d;97:7633–7643.

41.

Ray

, Warnick

, Mitchell

, Kaneene

, Ruegg

, Wells

, Fossler

, Halbert

, May

. Prevalence of antimicrobial resistance among Salmonella on midwest and northeast USA dairy farms. Prev Vet Med, 2007; 79:204–223.

42.

Tang

, Caffrey

, Nóbrega

, Cork

, Ronksley

, Barkema

, Polachek

, Gangshorn

, Sharma

, Kellner

, Ghali

. Restricting the use of antibiotics in food-producing animals and its associations with antibiotic resistance in food-producing animals and human beings: A systematic review and meta-analysis. Lancet Planet Health, 2017; 1:e316–e327.

43.

USDA-APHIS. Highlights of Dairy 2007 Part III: Reference of Dairy Cattle Health and Management Practices in the United States, 2007. Fort Collins, CO: National Animal Health Monitoring System, Animal and Plant Health Inspection Service, United States Department of Agriculture, 2008.

44.

USDA-APHIS. Heifer Calf Health and Management Practices on U.S. Dairy Operations, 2007. Fort Collins, CO: National Animal Health Monitoring System, Animal and Plant Health Inspection Service, United States Department of Agriculture, 2010.

45.

Villarroel

. Antimicrobial Use in Dairy Cattle and its Potential Impact on Antimicrobial Resistance in Enteric Bacteria: Sampling, Data Collection and Analytical Methods. Colorado State University, Ann Arbor, MI: ProQuest Dissertations Publishing, 2007.