Prevalence and antimicrobial resistance patterns of Campylobacter jejuni isolated from beef and bovine milk in Gondar town, Amhara region, Ethiopia

Abstract

Background

Campylobacter jejuni is a leading cause of foodborne diarrheal zoonotic diseases globally, often transmitted through contaminated food. Especially in Ethiopia, due to practices such as consuming raw milk and undercooked meat, as well as weak antibiotic regulations, the risk is higher. Therefore, this study aimed to assess the prevalence and antimicrobial resistance of Campylobacter jejuni in raw cow’s milk and beef carcass swabs in Gondar, Ethiopia.

Methods

A cross-sectional study was conducted between October 2022 and June 2023, collecting raw milk samples (n = 194) and beef carcass swabs (n = 196) using stratified multistage sampling. Raw milk was collected from the farm and cafeteria, and carcass swabs from 50 cm² areas of the neck, brisket, flank, and rump regions of beef. Campylobacter jejuni was identified by biochemical tests and amplification of the “hipO” gene, and antibiotic resistance was assessed using the Kirby-Bauer disk-diffusion method.

Results

A high prevalence of Campylobacter jejuni was found in beef (21.9%) and cow’s milk (19.6%), with older animals showing the highest rate (25%). Although not statistically significant, the prevalence was higher on farms with poor hygiene (23.53%) compared to those with better conditions (p > 0.05) (OR = 1.18, 95% CI: 0.42 to 3.32). Contamination of beef increased significantly after evisceration (p = 0.019) (OR = 3.24, 95% CI: 1.21 to 8.68). Regarding antimicrobial resistance, beef isolates showed higher resistance to Cefoxitin, Nalidixic acid, and Erythromycin than milk isolates. Notably, 44.4% of the total isolates demonstrated multidrug resistance to Tetracycline, Polymyxin B, and Erythromycin classes, with 28.4% originating from beef swabs.

Conclusion

The study outlined a concerning level of Campylobacter jejuni prevalence and significant levels of multidrug resistance to antibiotics commonly used in both humans and animals. This underscores the necessity for a One Health approach to tackle these challenges.

Supplementary Information

The online version contains supplementary material available at 10.1186/s42522-026-00208-5.

Introduction

Campylobacter is one of the leading food-borne zoonotic pathogens, considered the most common cause of bacterial gastroenteritis in both developed and developing countries, including Ethiopia [1–3]. Within the Campylobacter genus, thermophilic Campylobacter species, particularly C. jejuni and C. coli, have been recognized as a major cause of acute bacterial gastroenteritis in humans [4–6]. In severe cases, Campylobacter infection has been linked to post-infection complications, such as Guillain-Barré syndrome and Reiter’s syndrome, in addition to the typical diarrheal syndrome [7]. Campylobacter jejuni is a zoonotic pathogen, where poultry, wild birds, cattle, sheep, and pigs are recognized as sources of infection [8].

In Ethiopia, the incidence of foodborne diseases associated with Campylobacter is increasing due to various risk factors. A previous formative study indicated that close contact with carrier animals such as cattle and chickens, inadequate access to safe drinking water, socioeconomic challenges, and limited awareness of proper hygiene practices can increase the incidence [9]. The existing expanded mixed farming system in the country, which combines livestock production, chicken farm, and agriculture, also creates a high risk for cross-contamination and increases the prevalence of Campylobacter jejuni [10]. Previous studies in Ethiopia have highlighted a high prevalence of Campylobacter, with rates reaching 56.5% in animals [11], 50% among children in Eastern Ethiopia [12], 10.5% in water [13], and 9.3% in animal-sourced foods [14]. In the Amhara region, some researchers reported 21.5% in cattle and 28.9% in poultry fecal samples from backyard farms in Gondar town [15], 15.4% in children from the University of Gondar hospital [16], and 10.5% in diarrheic children at Dembia district [17].

Campylobacter can colonize most warm-blooded animals and poultry, and it is the leading cause of bacterial-derived foodborne diarrheal disease in humans, which may arise from contamination of meat and meat products [18–20]. It is frequently transmitted to humans through the consumption of contaminated raw or undercooked animal-origin food products [21]. Consumption of raw cow’s milk and undercooked beef is common in Ethiopia, which may play a vital role in the spread of infections.

The pathogen is often isolated from raw meat products due to a high likelihood of carcass contamination during slaughter, handling, and further processing [22]. A study at Jimma municipal abattoir, Ethiopia, indicated that not washing hands before meat processing, after using the toilet, or allowing meat to touch floors, walls, or become soiled during preparation were the primary factors linked to contamination by Campylobacter species [23].

The efficiency of foodborne disease treatments in Ethiopia is adversely affected by the development of drug resistance, particularly stemming from the unregulated accessibility and utilization of antibiotics for poultry production, as well as for human and animal treatment [24, 25]. Therefore, this study aimed to investigate the prevalence and antimicrobial resistance patterns of C. jejuni isolated from raw cow’s milk and beef carcasses swabs in Gondar town, Amhara region, Ethiopia, to support the development of effective strategies to prevent and control Campylobacter infections.

Materials and methods

Study area

The study was conducted from October 2022 to June 2023 in Gondar town, Amhara Region, Ethiopia, located in the Northwestern highlands of Ethiopia, at approximately 12°3′N latitude and 37°28′E (Fig. 1). The city has a total area of 192.3 km², with an average elevation of 2200 m above sea level. The annual temperature ranges from 22 °C to 30.7 °C, with an average of 26 °C. Relative humidity varies from 60% to 70% during the rainy season and from 30% to 40% during the dry season, with a mean annual rainfall of nearly 1025 mm [26, 27]. According to the district administration office report, the human population is 355,857, and there are 10,112 cattle, 24,690 goats, 3,830 sheep, 2,165 horses, 9,891 donkeys, and 75,834 poultry.

Fig. 1.

Study area, Gondar town, Ethiopia. (Map developed using QGIS-2.18.0 software)

Sample size determination

To determine the sample size, the expected prevalence was considered based on previously reported studies: 13.4% for raw milk and 7.4% for carcass swabs [28, 29]. The minimum required sample size was then calculated at 95% confidence interval (CI) and 5% desired precision using the formula: N = Z² * P * (1-P) / d² where: N = minimum required sample size; Z = desired level of confidence at 95% CI, which is 1.96; d = Margin of Error (MoE) set at 5%; and P = expected prevalence [30]. As a result, the minimum required sample size was 103 for milk and 57 for beef carcasses. To ensure the study is adequately powered and detects a meaningful difference, the sample size estimate was refined to 194 for the milk samples and 196 for the beef swab samples with a probability sampling method. This revised sample size reduces the desired margin of error (MoE) significantly and increases the precision of the estimates compared to the initial minimum requirements.

Study design and sampling method

A cross-sectional, multistage random sampling was applied to reduce sampling bias and investigate the relative prevalence and antibiotic resistance patterns of C. jejuni in raw cow’s milk and swabs from beef carcasses. The primary unit of sampling was “kebele”, which represents the lowest-level administrative division within the town. The secondary unit comprises dairy farms within kebeles, while the tertiary units include individual lactating cows in the farm, milk shop cafeterias, slaughtered cattle, and beef in butchers’ shops. Among the 21 kebeles in the town, 10 kebeles, which included 40 dairy farms that were randomly selected, 169 milk samples were collected (from 2 to 10 healthy lactating cows per farm based on the availability), and a total of 25 milk samples were collected from cafeterias that are distributed within these kebeles. Fifty-five beef carcass swabs from various butcher shops, and 141 carcass swabs from the municipal abattoir were collected randomly using the list obtained from local administrative offices and the municipal abattoir record sheets.

Inclusion and exclusion criteria

All randomly selected animals were included in the study; however, animals that had received antibiotic treatment in the last two weeks were excluded, and the next available animal was included instead.

Sample collection

Samples were collected from October 2022 to June 2023, following the guidelines provided by the World Organization for Animal Health (WOHA) [31]. Samples of raw cow’s milk were collected from dairy farms and cafeterias. On dairy farms, milk samples were collected individually from each animal’s milk, while in cafeterias, samples were taken from bulk milk tanks before heat treatment.

To assess potential risk factors, the body condition score (BCS) of the cattle was ranked on a scale of one to five, as described in a previous study [32]. The scoring categories were: poor (scores 1–2), medium (score 3), and good (scores 4–5). Additionally, the hygiene conditions of the farms were categorized as high (good), medium, and poor based on their adherence to basic hygiene practices. A high hygiene score is granted when the farm demonstrates excellent cleanliness and sanitation practices, including daily cleaning with disinfectants, restricted access to the farm, regular washing of udders with antiseptics before milking, and proper handwashing of milkers with antiseptics. Medium hygiene is given to farms that regularly clean floors with water, limit access to the farm, wash udders with water before milking, and ensure milkers wash their hands with antiseptics. A poor hygiene score indicates inadequate hygiene, characterized by infrequent cleaning and insufficient attention to the cleanliness of animals and milkers.

To identify potential differences between various breeds and age groups, cattle were classified into two categories: local breeds and crossbred. Additionally, animals were grouped by age as either young (5–8 years) or old (9–12 years), based on information provided by the farm owners.

During milk sampling, approximately 15 ml were aseptically collected using a sterile, screw-capped test tube and stored in an insulated box with ice packs before being transported to the University of Gondar (UoG) food safety laboratory for analysis. Beef carcass swab samples were collected from the municipal abattoir and butcher shops following previously established protocols [33], with some modifications. Briefly, a wet cotton swab was used to sample a 50 cm² area (10 cm x 5 cm) in four carcass regions: the neck, brisket, flank, and rump, which are more susceptible to contamination during slaughter due to their proximity to the animal’s hide and gastrointestinal tract [34]. These pulled swab samples from one carcass were placed in a 5 mL sterile buffered peptone water (BPW) (Oxoid Ltd., Hampshire, UK). After the swabbing was completed, the shaft of the swab handle was broken, leaving the cotton swab inside the collection test tube. The test tube was then sealed with a screw cap. Sampling at the abattoir was conducted on three separate operation lines, before evisceration, after evisceration, and after the final carcass wash. All samples were submitted to the University of Gondar food safety laboratory in a cold box and processed immediately.

Isolation of Campylobacter jejuni

The isolation and identification of C. jejuni were conducted according to the enrichment methods previously outlined [33, 35, 36]. Briefly, 15 ml of milk samples were centrifuged at 10,000 rpm for 10 min. After discarding the fat layer and supernatant, the pellets were resuspended in 1.0 ml buffered peptone water and vortexed. Subsequently, 9 ml of Bolton broth (Oxoid, Hants, UK) with a selective supplement and laked horse blood (HemoStat Laboratories, Dixon, USA) were added and pre-enriched by incubating the samples for 4 h at 37 °C, under microaerobic conditions, which was generated using airtight jars containing a gas-generating kit (CampyGen™ 3.5 L, Oxoid, Basingstoke, UK).

Enrichment was achieved by extending the incubation time for an additional 18 h at 42 °C under the same microaerobic conditions. After enrichment, 30 µl (three loopfuls) were spread onto Campylobacter blood-free modified charcoal-cefoperazone-deoxycholate selective agar (mCCDA-Preston; Oxoid, Basingstoke, Hants, UK) containing the Campylobacter CCDA selective supplement (Oxoid, Hants, UK). Following 16 to 18 h of incubation at 42 °C under microaerophilic conditions, as mentioned above, the plates were examined for the growth of colonies. When no colony growth was observed, the incubation was extended for an additional 24 h under the same conditions and reobserved.

All colonies that appeared grayish, flat, moist, often with a metallic sheen, and a tendency to spread were considered presumptive Campylobacter spp. and those that are Gram-negative, with spiral, “S” or a “V” shape roads, catalase and oxidase positive, and hydrolyze Hippurate were taken as presumptive C. jejuni.

Molecular detection of Campylobacter jejuni

DNA preparation

Template DNA was prepared by the thermal boiling method [37]. Briefly, a loopful of bacterial colonies from an agar plate was suspended in 1 mL of Tris-EDTA buffer, boiled for 10 min, and immediately chilled on ice for 5 min. Further, it was centrifuged for 10 min at 10,000 rpm, and 5 µL of supernatant was used as a PCR template DNA.

PCR-based identification of Campylobacter jejuni

For the detection of hippuricase gene (hipO) of C. jejuni by PCR, a forward primer (HIP400F): 5’-GAA GAG GGT TTG GGT GGT G-3’, and a reverse primer (HIP1134R): 5’-AGC TAG CTT CGC ATA ATA ACT TG3’ were used to amplify a 735 bp product size using the protocol reported by previous research with slight modification [38]. A 25 µl volume PCR protocol for the detection of hip0 gene of C. jejuni was: 10X Buffer (15mM Mg Cl₂; 2.5 µl); MgCl₂ (25 mMol; 1 µl); dNTP mix (25 mMol; 0.2 µl); HIP400F primer (with 10 pmol / µl; 1 µl); HIP1134R primer (with 10 pmol / µl;1 µl); Taq polymerase (3 Ս/ µl; 0.4 µl); Template DNA(2.5 µl) and nuclease-free water (16.4 µl).

The genes were amplified in a PCR thermocycler (Mastercycler, China), which was adjusted at different temperature and time combinations: - Initial denaturation at 94 °C for 5 min; followed by 30 cycles, each consisting of 94 °C for 1 min, 56 °C for 1 min, and 72 °C for 2 min. The final extension cycle was 7 min at 72 °C.

Before electrophoresis, 10 µl of each PCR product was mixed with 2 µl of a dye (EZ-Vision One™ DNA dye) and loaded into wells of a 1.5% agarose gel. For reference, a 100-bp DNA ladder (Quick-Load ^® Purple DNA Ladder; Biolabs, UK) was used, and electrophoresis was performed at 70 V. The image of the gel was captured using the Gel Documentation system (BioTop, gel documentation system, China).

Antimicrobial resistance test

Antimicrobial resistance testing was conducted using the Kirby–Bauer disk-diffusion method, following the standardized approach established by the European Committee on Antimicrobial Susceptibility Testing (EUCAST) for Campylobacter jejuni, as well as the reports from the American Society for Microbiology [39]. Nine different antimicrobial disks, including the broad-spectrum that acts on both gram-positive and negative bacteria, were tested: cefoxitin (CX)-30 µg, cefotaxime (CTX)-30 µg, polymyxin-B (PB)-30 µg, tetracycline (TE)-30 µg, nalidixic acid (NA)-30 µg, norfloxacin (NX)-10 µg, erythromycin (E)-15 µg, streptomycin (S)-10 µg, and co-trimoxazole (COT)-25 µg.

Confirmed Campylobacter jejuni isolates, from overnight subcultures, were tested for antibiotic resistance as follows: Colonies were suspended in 10 mL sterile normal saline solution (0.9% NaCl), and after vortex mixing, the turbidity was adjusted to the 0.5 McFarland standard. Muller-Hinton agar was streaked using a sterile cotton swab dipped into the inoculum tube, and the antimicrobial discs were placed on the surface of the agar using sterile forceps. Further overnight incubation was performed under microaerobic conditions at 42 °C as described earlier.

The resulting zone of inhibition was measured in mm with a caliper, recorded, and interpreted using the published zone diameter standards for Campylobacter and the standard zone diameter interpretive chart as susceptible (S), intermediate (I), or resistant (R) to the antimicrobial agent (Table 1) [28, 39–42]. The multidrug resistance (MDR) status was assessed based on the definition by Magiorakos et al. (2012), which defined multidrug resistance as the development of resistance to three or more different classes of antibiotics [43].

Table 1.

The guidelines used to determine the antimicrobial resistance breakpoints for the disk-diffusion methods

Antimicrobial agent	Disk Conc. (µg)	Zone diameter breakpoint (mm)
		S	I	R	References
Cefotaxime	30 µg	≥ 23	15–22	≤ 14	[28, 39–42]
Cefoxitin	30 µg	≥ 20	15–19	≤ 14
Polymyxin-B	30 µg	≥ 21	13–20	< 12
Norfloxacin	10 µg	≥ 17	13–16	≤ 12
Nalidixic acid	30 µg	≥ 19	14–18	≤ 13
Streptomycin	10 µg	> 15	12–14	< 11
Tetracycline	30 µg	≥ 15	12–14	< 11
Erythromycin	15 µg	≥ 23	14–22	≤ 13
Co-Trimoxazole	25 µg	≥ 17	13–16	≤ 12

S- Susceptible; I- Intermediate; R- Resistant; Conc - Concentrations

Data management and analysis methods

Raw data and laboratory results were recorded into a Microsoft Excel sheet and analyzed using STATA software version 14.0. Descriptive statistics were conducted to compute the number of samples positive for C. jejuni. A multivariate logistic regression test was used to assess the association between variables (breed, age, BCS, and hygienic system) and the occurrence of C. jejuni.

Results

Prevalence of Campylobacter jejuni

The overall prevalence of C. jejuni isolates, confirmed using PCR, from samples of raw cow’s milk and beef carcass swabs collected from Gondar town, was found to be 20.8% (81/390) (Table 2). Among the milk samples collected, 19.6% (38/194) were positive for C jejuni, and beef carcass swabs were 21.9% (43/196) positive. In raw milk samples collected, dairy farms showed a prevalence of 19.5% (33/169), whereas milk from the cafeteria had a prevalence of 20% (5/25) (Table 2).

Table 2.

Prevalence of Campylobacter jejuni from milk and beef carcass swabs in Gondar town, Ethiopia, 2022–2023

Sample type and source	Total examined	C. jejuni positive	Prevalence (%)	Odds ratio (OR)	95% CI	p. value (Fisher’s exact)	Symmetric measures
Beef swab in the abattoir:							Cramer’s V 0.201.
Before evisceration	47	7	14.9	Ref	-	-
After evisceration	47	17	36.2	3.24	1.21–8.68	0.019
After the final wash	47	11	23.4	1.74	0.61–4.96	0.298
Beef carcass swab:
In the abattoir	141	35	24.8	1.94	0.84–4.50	0.13	R-value 0.11
In the butcher’s shop	55	8	14.5	Ref	-	-
Raw cow’s milk:
From dairy farms	169	33	19.5	0.97	0.34–2.78	0.956	R-value -0.004
From cafeteria	25	5	20	Ref	-	-
Beef swab versus milk samples
Beef carcass swab	196	43	21.9	1.14	0.71–1.83	0.59	R-value 0.02
Raw cow’s milk	194	38	19.6	Ref	-	-
Total	390	81	20.8

Remark: - Ref= Reference

The highest prevalence of contamination, at 24.8% (35 out of 141 samples), was found in beef carcass swabs from municipal abattoirs. In contrast, butcher shops had a prevalence of 14.5% (8 out of 55 samples). Among the beef carcass swabs collected from the abattoir, the highest contamination level was detected in samples taken after evisceration, with a prevalence of 36.2%. and the lowest, at 14.9%, before evisceration (see Table 2). These results indicate a statistically significant association with C. jejuni contamination, particularly within the beef processing chain at the abattoir (p = 0.019). Notably, contamination increased significantly after the evisceration step, with samples collected after evisceration being over three times more than before evisceration (OR = 3.24, 95% CI: 1.21 to 8.68) (see Table 2). Similarly, the undifferentiated relationship between the sampling stages and the presence of C. jejuni for the abattoir beef swab samples suggests a weak-to-moderate association (Cramer’s V value of 0.201).

Additionally, the variations in the prevalence of C. jejuni across different locations (abattoir versus butcher shop, dairy farm versus cafeteria) and sample types (beef swabs versus milk samples) were not statistically significant (p > 0.05). The correlation metrics (R-values) for these comparisons were also weak, showing values of 0.11 for beef samples, -0.004 for milk samples, and 0.02 when comparing beef and milk samples, as illustrated in Table 2.

Milk samples collected from 40 dairy farms were assessed based on breed, age group, body condition score, and farm hygiene as described in the methodology section. Among C. jejuni-positive cows, a higher prevalence of C. jejuni was observed in local breeds (21.74%), in older age groups (25.0%), and in farms with poor hygiene practices (23.53%). Despite the differences in C. jejuni occurrence among potential risk factors, these differences were not statistically significant (p > 0.05) (Table 3).

Table 3.

Prevalence of Campylobacter jejuni and risk factors in milk samples in Gondar, Ethiopia, 2022–2023

Variables	Categories	Total	C. jejuni positive	prevalence (%)	Odds ratio (OR)	p. value (Fisher’s exact)	95% C.I.
Breed	Cross	100	18	18.00	0.79	0.56	0.37–1.70
	Local	69	15	21.74	Ref	-	-
Age	Young	93	14	15.05	0.53	0.121	0.246–1.148
	Old	76	19	25.00	Ref	-	--
BCS	Good	88	15	17.05	0.719	0.441	0.335–1.543
	Medium	81	18	22.22	Ref	-	-
Hygienic system	High	37	8	21.62	Ref	-	-
	Medium	81	13	16.05	0.76	0.58	0.28–2.05
	Poor	51	12	23.53	1.18	0.75	0.42–3.32

Remark: - Ref=Reference

Antimicrobial resistance patterns of isolated Campylobacter jejuni

Confirmed C. jejuni isolates (a total of 81) were tested for antimicrobial resistance against nine commonly used and reserve antibiotics in the study area. The highest resistance was observed for tetracycline (100%), followed by polymyxin B (74.1%) and erythromycin (61.7%). The majority of tetracycline-resistant C. jejuni strains were isolated from beef carcass swabs (53.1%), compared to 46.9% from milk samples. Additionally, a notable proportion of isolates, 24.7%, showed resistance to the third-generation antibiotic, cefotaxime, while the lowest resistance was for Norfloxacin (6.2%) and Co-Trimoxazole (13.6%) (see Table 4).

Table 4.

Antimicrobial-resistance patterns of Campylobacter jejuni isolates from milk and beef in Gondar town, Ethiopia, 2022–2023

Antibiotics	C. jejuni isolates (%)			Resistant C. jejuni from raw cow’s milk (%)			Resistant C. jejuni from carcass swab (%)			Association of Resistance between Milk & Beef isolates
	S*	I*	R*	Cafeteria	Farm	Sub Total	Abattoir	Butchers	Sub Total	Odds (95% CI)	p. value (2-tailed)
Cephalosporins:
Cefotaxime 30 µg (CTX)	44 (54.3)	17 (21)	20 (24.7)	0	6	6 (7.4)	14	0	14 (17.3)	0.388 (0.128–1.176)	0.1064
Cefoxitin 30 µg (CX)	65 (80.2)	2 (2.5)	14 (17.3)	0	2	2 (2.5)	10	2	12 (14.8)	0.144 (0.02–0.69)	0.0082
Polymyxin:
Polymyxin-B 30 µg (PB)	18 (22.2)	3 (3.7)	60 (74.1)	4	26	30 (37)	24	6	30 (37)	1.625 (0.588–4.487)	0.4480
Quinolone:
Norfloxacin 10 µg (NX)	66 (81.5)	10 (12.3)	5 (6.2)	0	0	0 (0)	5	0	5 (6.2)	0.091 # (0.0–1701)	0.057
Nalidixic acid 30 µg (NA)	28 (34.6)	18 (22.2)	35 (43.2)	1	9	10 (12.3)	17	8	25 (30.9)	0.257 (0.081–0.771)	0.016
Aminoglycoside:
Streptomycin 10 µg (S)	20 (24.7)	24 (29.6)	37 (45.7)	3	13	16 (19.8)	20	1	21 (25.9)	0.762 (0.316–1.835)	0.656
Tetracycline:
Tetracycline 30 µg	0	0	81 (100.0)	5	33	38 (46.9)	35	8	43 (53.1)	NaN	1.00
Macrolide:
Erythromycin 15 µg	19 (23.5)	12 (14.8)	50 (61.7)	3	15	18 (22.2)	29	3	32 (39.5)	0.31 (0.11–0.81)	0.0213
Sulfonamides:
Co-Trimoxazole 25 µg	67 (82.7)	3 (3.7)	11 (13.6)	0	4	4 (4.9)	7	0	7 (8.6)	0.605 (0.162–2.253)	0.528

Remarks: -C. jejuni positive results are presented numerically, with percentages in parentheses. S*- Susceptible; I*- Intermediate; R*- Resistant; Conc - Concentrations; NaN -unidentified; CI for OR are indicated in parenthesis; # - Haldane-Anscombe corrected Odds ration

Resistance patterns between milk and beef isolates showed statistically significant differences (P < 0.05) for Cefoxitin (OR = 0.144, 95% CI: 0.02 to 0.69), Nalidixic acid (OR = 0.257, 95% CI: 0.081 to 0.771), and Erythromycin (OR = 0.31, 95% CI: 0.11 to 0.81). In all three cases, resistance was significantly higher in beef isolates than in milk isolates. By contrast, for Polymyxin-B, Streptomycin, and Co-Trimoxazole, resistance levels of C. jejuni isolates were statistically comparable between the two sample sources (p > 0.05). For Tetracycline, all isolates were resistant, and there was no statistically significant difference in terms of drug resistance between the two sample group isolates (Table 4).

Multi-drug resistance (MDR) patterns of Campylobacter jejuni

Among a total of 81 tested isolates of C. jejuni, nearly half demonstrated resistance to three or more classes of antimicrobial drugs. Resistance levels were highest (44.4%) for tetracycline, polymyxin-B, and Erythromycin (TE/PB/E) classes, followed by tetracycline, polymyxin-B, and Nalidixic acid (TE/PB/NA) classes (35.8%).

In this study, majority of the multidrug-resistant isolates (28.4%) originated from beef carcasses, compared to those from milk samples (16.1%). Among the isolates obtained from milk samples, the majority demonstrating multidrug resistance were sourced from the farm (13.6%). Similar findings were observed in beef carcass swabs collected from the abattoir (25.9%). Moreover, one isolate (from a beef swab taken after evisceration) was resistant to all nine antibiotics tested (Table 5).

Table 5.

Multidrug-resistant patterns of 81 Campylobacter jejuni isolates from milk and beef in Gondar, Ethiopia, 2022–2023

Antibiotics	Resistant isolates (%)	Source and number of resistant isolates (%)
		Milk		Beef carcass swab
		Farm	Cafeteria	Abattoir	Butcher shops
TE/PB/E	36 (44.4)	11 (13.6)	2 (2.5)	21 (25.9)	2 (2.5)
TE/PB/NA	29 (35.8)	9 (11.1)	1 (1.2)	13 (16.0)	6 (7.4)
TE/PB/S	28 (34.6)	10 (12.3)	2 (2.5)	15 (18.5)	1 (1.2)
TE/PB/E/NA	18 (22.2)	4 (4.9)	1 (1.2)	11 (13.6)	2 (2.5)
TE/PB/NA/S	18 (22.2)	5 (6.2)	1 (1.2)	11 (13.6)	1 (1.2)
TE/PB/E/S	17 (21)	3 (3.7)	2 (2.5)	12 (14.8)	0
TE/E/S/NA	13 (16)	2 (2.5)	1 (1.2)	10 (12.3)	0
TE/PB/S/NA/CTX	13 (16)	4 (4.9)	0	9 (11.1)	0
TE/PB/E/S/NA	12 (14.8)	2 (2.5)	1 (1.2)	9 (11.1)	0
TE/PB/E/NA/CTX	11 (13.6)	3 (3.7)	0	8 (9.9)	0
TE/E/S/NA/CTX	10 (12.3)	2 (2.5)	0	8 (9.9)	0
TE/PB/E/S/CTX	9 (11.1)	2 (2.5)	0	7 (8.6)	0
TE/PB/E/S/NA/CTX	9 (11.1)	2 (2.5)	0	7 (8.6)	0
TE/PB/E/S/NA/COT	2 (2.5)	0	0	2 (2.5)	0
TE/PB/E/S/NA/CX/COT	2 (2.5)	0	0	2 (2.5)	0
TE/PB/E/S/NA/CTX/CX/COT/NX	1 (1.2)	0	0	1 (1.2)	0

Notes: Multidrug resistance patterns of the 81 C. jejuni isolates to a combination of different antibiotics, Cefoxitin (CX), Cefotaxime (CTX), Polymyxin-B (PB), Tetracycline (TE), Nalidixic acid (NA), Norfloxacin (NX), Erythromycin (E), Streptomycin (S), and Co-trimoxazole (COT), with percentages in parentheses

Discussion

This study focused on assessing the prevalence of lesser-studied and underreported pathogens that frequently cause foodborne infections through the consumption of raw milk and undercooked beef in Ethiopia. In this study, 19.6% of raw cow’s milk samples tested were positive for C. jejuni, which aligns with a study that reported 18% prevalence [44]. Cattle are typical carriers of C. jejuni; therefore, cross-contamination of milk can occur from the animal’s feces during milking [33, 45]. The study showed a slightly higher prevalence of C. jejuni in cafeteria milk samples (20%) compared to farm samples (19.5%). However, this difference was not statistically significant (p > 0.05) (OR = 0.97, 95% CI: 0.34 to 2.78) and Pearson’s R = -0.004). This variation could be due to a greater risk of contamination from the supply chain or the common practice of adulterating milk with water in local cafeterias, as reported by previous research [46].

Notable correlation between hygienic status and the occurrence of C. jejuni is also observed in this study. The group from poor-hygiene dairy farms had the highest prevalence at 23.53%, compared to 16.05% in the medium-hygiene group and 21.62% in the good-hygiene group, though not significantly impacted C. jejuni positivity (p > 0.05) (OR = 1.18, 95% CI: 0.42 to 3.32) (Table 3). Variations in breed, animal age, and body condition scores did not significantly affect the prevalence of C. jejuni (p > 0.05). This highlights the presence of additional confounding factors that may influence milk contamination beyond the aforementioned potential predictors. Factors such as the health and hygiene of the milker, poor post-harvest handling, and the cleanliness of the milk tank and utensils used for transporting and preserving the milk may significantly impact contamination level [47, 48].

Out of the 21.9% of beef swabs that tested positive for C. jejuni, 24.8% were from abattoirs, while 14.5% were from butcher shops. The positivity rates of beef swab samples did not show a statistically significant association between the two different sources (p > 0.05) (OR = 1.94, 95% CI: 0.84 to 4.50) (Table 2). When compared with previously reported findings, it is relatively higher: 5.3% in Debrezeyit and Addis Ababa abattoirs [19], 7.3% in Addis Ababa [49], and 11.8% in Arba Minch, Ethiopia [50]. Variations in disease prevalence can be attributed to several confounding factors, including differences in climate, seasonal changes, farm sanitation practices, and the level of farmers’ awareness regarding zoonotic diseases. Generally, areas with frequent training and awareness initiatives tend to have lower prevalence rates, particularly in central regions of the country compared to more remote areas [51, 52].

In many previous studies, it is well documented that during the slaughtering process, the main sources of contamination are the slaughtered animals themselves [53]. Findings from this study reinforce this fact. In three different operational systems of municipal abattoirs, the highest isolation rate was found in carcass swabs collected after the evisceration process, with a rate of 36.2% compared to before evisceration (14.9%) (p = 0.019), with an Odds Ratio (OR) of 3.24, indicating that carcasses were over three times more likely to be contaminated at this stage (Table 2). High C. jejuni prevalence after evisceration may result from contamination of carcasses with intestinal contents or improperly cleaned equipment during evisceration and subsequent processing in the slaughterhouse. Likewise, the relatively higher prevalence of C jejuni from samples after wash of the carcass compared to before evisceration may be associated with poor hygiene slaughtering practices in this abattoir, poor water quality used for washing and cleaning, and the usage of cold water for carcass washing, which did not significantly contribute to the removal of bacterial contamination [13, 54].

Antibiotic resistance is one of the major challenges in managing infectious diseases in Ethiopia. It is also an increasingly reported global public health concern in some Campylobacter spp [40, 51]. Following the Kirby–Bauer disk-diffusion method, this study demonstrates a high level of resistance to both commonly and rarely used antibiotics in animal treatment. Specifically, among the frequently used antibiotics in Ethiopian veterinary pharmacies, isolates of C. jejuni demonstrated 100% resistance to tetracycline, 61.7% to Erythromycin, and 45.7% to streptomycin (Table 4).

The 100% resistance to tetracycline observed in this study aligns closely with the 82.2% resistance reported in Jimma, Ethiopia [11]. However, it is significantly higher than the 13.5% resistance found in poultry and the 22.2% resistance in human isolates from Bahirdar in 2010 [58]. In Jimma, the rates of resistance reported were 39.5% in 2014 and 42.1% in 2022 [59, 60]. These increasing trends in resistance over the years suggest an ongoing development of tetracycline resistance, likely linked to prolonged and poorly regulated use of the drug.

Among the tetracycline-resistant isolates, 46.9% were from milk samples (mainly from dairy farms, accounting for 40.7%), while 53.1% were from carcass swabs (primarily from abattoir beef carcasses, at 43.2%). The overall findings of this work also support earlier research findings, demonstrating a significant association between the development of antimicrobial resistance (AMR), the prevalence of widely utilized antimicrobials, and the potential transfer of resistant genes between human and animal isolates [41, 55, 56, 57]. Moreover, the antimicrobial resistance patterns of C. jejuni isolates demonstrated statistically significant differences between milk and beef for Cefoxitin, Nalidixic acid, and Erythromycin (p < 0.05). In all three cases, the resistance was significantly higher in beef isolates than in milk isolates. This difference may be linked to the unregulated use of antibiotics in the study area, particularly in cattle intended for fattening, which may result in the transfer of resistant genes [25, 55, 56, 57].

In line with MDR, nearly half of the C. jejuni isolates of this study (44.4%) exhibited MDR to three or more different groups of tested antimicrobial agents (Table 5). The highest level of resistance observed among three classes of antibiotics was 44.4% to TE/PB/E (which included 28.4% from isolates from carcass swabs and 16.1% from milk samples). Resistance to four different classes of antibiotics was found in 22.2% isolates, specifically to TE/PB/E/NA and TE/PB/NA/S. Likewise, resistance to five different antibiotics (TE/PB/S/NA/CTX) was noted in 16% of the isolates, and one isolate from a carcass swab exhibited resistance to all nine tested antimicrobial drugs.

Overall, the isolates from carcass swabs showed a higher proportion of multidrug resistance to the tested antibiotics compared to those from milk samples. These high levels of resistance to various antimicrobial agents may be attributed to several factors, including the unrestricted use of these antimicrobial agents in treating animals, insufficient infection prevention and control measures, and a lack of surveillance and monitoring capabilities in Ethiopia [55, 61].

Limitations

Although the increased sample size and a multistage sampling strategy can significantly improve precision, selection bias may not be completely excluded because participation depended on the access to and willingness of slaughterhouses, butcher shops, and milk vendors.

This research adheres to the CLSI M02 standardized procedures for conducting antimicrobial disk susceptibility tests and determining breakpoints based on accepted scientific references. However, it has limitations due to not utilizing the gold standard broth microdilution method (MIC testing) recommended for C. jejuni by CLSI/EUCAST, which could impact the results. Moreover, molecular confirmation of antimicrobial resistance determinants such as gyrA, 23 S rRNA, and tet(O) was not performed. Therefore, the genetic basis of the observed phenotypic resistance could not be confirmed, and potential genotype–phenotype discordance cannot be excluded. Additionally, the coverage of the study area was limited, which may restrict the ability to generalize the findings to the national level; however, the similarities in drug administration rules, regulations, and farm management practices across different regions of the country remain consistent and reflect similar findings.

Conclusion and recommendation

This study indicates that animal products, especially raw cow’s milk and beef, are often contaminated with Campylobacter jejuni in the study area, which presents a significant risk of foodborne zoonotic infections for consumers. Furthermore, the study indicates the development of high-level drug resistance among C. jejuni isolates, including resistance to certain third-generation antibiotics that are not used for animal treatments in the country. This situation threatens human health by reducing available treatment options and contributing to the emergence and spread of this pathogen to human, animal, and environmental health. Therefore, to mitigate the risks associated with foodborne zoonotic diseases such as C. jejuni in the food chain, it is recommended to adopt a One-Health approach by enabling surveillance of Campylobacter throughout the food supply chain, implementing strict antimicrobial use programs in both human and veterinary sectors, and enhancing personnel hygiene during milking and slaughter.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

BCS

Body condition scores

BPW

Buffered Peptone Water

CSA

Central Statistical Agency

CI

Confidence interval

WOHA

World Organization for Animal Health

PCR

Polymerase Chain Reaction

UoG

University of Gondar

Author contributions

S.N.: conceptualization, data curation, formal analysis, investigation, methodology, visualization, writing original draft. D. B.: conceptualization, data curation, investigation, methodology, writing – reviewing & editing. A. B.: conceptualization, writing, reviewing & editing. B.H.: conceptualization, writing, reviewing & editing. S.M.: conceptualization, writing, reviewing & editing. A.G.A.: writing, reviewing & editing. A.E.Y.: supervision, writing, reviewing & editing. G. A.: supervision, writing, reviewing & editing. A. M.: supervision, writing, reviewing & editing. All authors read and approved the final manuscript.

Funding

This research was partially supported by the University of Gondar (UoG) laboratory facility in Ethiopia.

Data availability

All the data concerning this manuscript are contained within the text.

Declarations

Ethics approval and consent to participate

This research work was carried out after receiving ethical clearance from the College of Veterinary Medicine and Animal Sciences Research Ethics Review Committee (CVMASc-RERC), University of Gondar, Ethiopia, with a reference number CVMASc/UoG/RERC/2521.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.