Prevalence of antibiotic residues in commercial poultry feed and meat from industrial and backyard poultry in Iquitos, Perú

Abstract

Objectives:

This study aimed to determine the presence and prevalence of antibiotics in poultry feed and poultry meat in a region with high rates of highly antimicrobial resistant Campylobacter infections.

Methods:

Poultry feed was collected from local commercial feed stores over a period of 6 months. Standard open bag initial, growth, and fattening mixes were sampled at each commercial retailer. Meat was collected from local supermarkets, a central open-air market, and four open-air markets in small communities that raise their chicken in the community. Animal feed and meat were tested for the phenotypic presence of antibiotics using the Premi-Test, a bioassay that detects antimicrobials from multiple classes.

Results:

A total of 59 poultry feed samples were obtained, of which 49.2% were found to have detectable antimicrobial additives, and this did not differ between three classes of feed despite no labeling or claims of being a product that contains antimicrobials. Meat samples from 177 industrially produced broilers from supermarkets (n = 85) and open-air markets (n = 92) were evaluated for antibiotic residues. Antibiotics were detected in 27.7% of samples with no significant differences in prevalence between the market types. An additional 99 meat samples were obtained from regional chickens purchased from open markets in urban and rural open-air markets, which had antibiotic residues in 35.4% of samples. There was no significant difference in the prevalence of antibiotic residues in meat from broilers and regional chickens.

Conclusions:

Minimally processed poultry feed and poultry meat frequently had detectable antibiotics across different sectors of the poultry production chain. Antibiotic residues were frequently detected in chicken meat, whether it was purchased in supermarkets or through open-air markets. Backyard chickens reared in the community and thought to represent a more natural food source had the same prevalence of antibiotic residues as industrially produced chicken.

Introduction

Antibiotics are commonly used in agriculture, human medicine, and veterinary medicine to ensure the health of humans, animals, and crops. An adverse effect of antibiotic use is the increase of antimicrobial resistance (AMR). This poses a significant problem for global health because many antibiotics have become less effective [1]. In 2021, it was estimated that AMR caused 1.14 million global deaths and the disability-adjusted life years (DALY) adjudicated to AMR was a calculated 42.6 million; global deaths are forecasted to increase by 69% by 2050 [2]. In the United States (USA), it was estimated that 4.6 billion dollars was spent in 2017 to combat six types of multidrug-resistant infections [3].

The determinants of antibiotic resistance that arise due to the exposure of animals and their microbes can be shared cross-species; therefore, the use of antibiotics in veterinary medicine and agriculture can lead to the rise of antibiotic resistance in organisms that directly impact humans [4,5]. Antibiotic-resistant bacteria in livestock can be shared across species through livestock waste in the environment, contaminated water sources, contact between livestock and humans, or through consumption of livestock products. Furthermore, humans can share antibiotic-resistant bacteria with livestock due to their shared environment [6].

In poultry, antibiotics are used to fight infections and as growth promoters. In many high-income countries, the use of antibiotics for livestock is highly regulated, but antibiotics are still abused [7]. The USA established the National Antimicrobial Resistance Monitoring System (NARMS) to track antimicrobial drug use in animals, set guidelines for the allocation of antibiotics, and monitor retail meat [8]. However, in many low-middle income countries (LMIC), the use of antibiotics in livestock is widespread, and even when policy restrictions exist [9], limited to absent enforcement of restrictions of antibiotic use in the production of animals allows for widespread use [10].

In LMIC, small-scale commercial poultry counters food insecurity and helps alleviate poverty [11]. However, small-scale farms use antibiotics without proper knowledge or veterinary oversight [12,13]. In Perú, antibiotics can be obtained from veterinarians, drug representatives, and over the counter at feed stores [14]. Furthermore, surveillance of AMR within the livestock varies from country to country. In fact, South America has a low number of surveys about AMR in animal husbandry [15].

The objective of this study is to determine the prevalence of antibiotics in chicken meat and chicken feed in Iquitos, Perú. This was done as an exploratory response to the high prevalence of multidrug-resistant Campylobacter jejuni and Campylobacter coli strains isolated from children in the same region, as it is hypothesized that antibiotic exposure to poultry may affect human exposure directly through the ingestion of meat with residual antibiotics or indirectly through exposure to drug-resistant microbial populations of the gut flora of chickens. Furthermore, through conducting in-depth semi-structured interviews and surveys, this study seeks to better understand how, when, and why antibiotics are being used by small-scale poultry producers in Iquitos, Perú.

Methods

Sample collection

Feed

Iquitos is a city of approximately 500,000 individuals in the Eastern Amazon of Perú. An estimated 20 animal feed stores are clustered within a few blocks of each other in the city center. For this study, feed stores were randomly sampled on four different streets. On each block, every other feed store was sampled. In each sampling event, at least two different streets were sampled. Stores were randomly resampled to test if the results changed over a few months. Pre-wrapped feed was not sampled; samples from sacs of maize prepared for three stages of poultry growth (“inicio” (starter), “crecimiento” (growth), and “engorde” (fattening)) from the local mills were bought at market value and transported to the laboratory for processing within 3 h of purchase.

Meat

Chicken meat was bought at market value from supermarkets, urban open-air markets, rural open-air markets located in rural and peri-urban communities of Santo Tomas, La Union and Santa Clara de Nanay located 15 km from the city center of Iquitos. In urban Iquitos, chicken is available for commercial purchase at five supermarkets and four urban open-air markets. Chickens are divided into two types: broiler chickens and locally-reared backyard chickens (“gallina regional”/“gallinas amazonicas nativas”). Regional chickens are a locally bred mixed breed chicken that has a slower growth rate (3–6 months) and lower meat conversion rate in comparison to broiler chickens but are hardy and better breeders in the local climate and prized as a more natural food source. Samples were transported to the laboratory for processing within 3 h of purchase.

Laboratory procedures

Assay for antibiotic detection

Antibiotic presence or absence was determined using the Premi-Test bioassay, a phenotypic screening assay. This assay is a phenotypic test of antibiotic presence based on the inhibition of growth of Bacillus stearothermophilus. This test detects penicillin, tetracyclines, sulfonamides, and macrolides at or below the maximal residual levels [16]. It additionally detects cephalosporins, aminoglycosides, amphenicols, and quinolones, but does so above the maximal threshold level defined by the Food and Agricultural Organization and the World Health Organization [9]. Confirmatory quantitative assays with LC-MS/MS were not available to the remote setting and limited scope of this pilot project.

Feed

Thirty milliliters of distilled water was added to 10 g of feed. The mixture was shaken for 30 min in an orbital shaker. The mixture was centrifuged at 3000 g for 10 min. Then, 100 μl of supernatant was pipetted onto the Premi-Test ampoules (R Biopharm, Darmstadt, Germany). The ampoules were sealed with plastic foil and placed into the incubator at 80 °C for 10 min, then at 64 °C for 2 h and 30 min. After 2 h and 40 min, the ampoules were removed from the incubator every 5 min, the color was checked against the negative control, and a photo was taken. When the negative control changed from purple to translucent yellow, the results of all test vials were read and recorded. The negative controls were fresh yellow corn obtained from the market.

Meat

Lean chicken meat was squeezed with a meat presser over aluminum foil. Chicken juice (100 μL) was pipetted onto the Premi-Test Ampoules (R Biopharm, Darmstadt, Germany) and incubated for 20 min at room temperature. The juice was rinsed out of the ampoules twice using demineralized sterile water. The ampoules were inverted on a paper towel to remove excess water. The ampoules were sealed with plastic foil and placed into the incubator at 64.0 °C. After 2 h and 40 min, the ampoules were removed from the incubator every 5 min, the color was checked against the negative control, and a photo was taken. When the negative control changed from purple to translucent yellow, the results were read. The negative controls were obtained from the previous day’s negative samples. The positive control was 15 mcg erythromycin ethylsuccinate crushed into sample chicken juice.

Statistical analysis

Fisher’s exact test was done to compare the prevalence of antimicrobial contamination between different classes of chicken feed, and to compare the difference in the prevalence of antimicrobial contamination in meat from different sources (Python, 3.8.2).

Semi-structured interviews

Semi-structured interviews were conducted with small-scale poultry producers from Santa Clara de Nanay, Santo Tomas, and La Union. Small-scale poultry producers located in the communities of Santa Clara de Nanay, Santo Tomas, and La Union were identified by local field workers and invited to participate in the study. The purpose of the interviews was to better understand the small-scale commercial poultry farmer; in specific, poultry rearing practices, poultry illness, use of pharmaceuticals and or vaccines to treat or prevent illness, antibiotic usage, and the concept of antibiotic resistance. Interviews were guided by an interview guide containing open-ended questions. The interview was conducted in Spanish, recorded, transcribed verbatim and analyzed in the same language. Follow-up interviews were not conducted. During the analysis of transcribed material, a codebook was generated based on a-priori definitions. Codes were applied while analyzing transcribed material, after which codes were organized into categories.

Survey

A multiple-choice questionnaire was applied to feed store customers who agreed to participate in the study. A trained field worker stood outside two of the major feed stores in the city of Iquitos and requested customers to participate in the study. Customers were asked about their age, district of residence, type and number of animals being raised, use of antibiotics, veterinary guidance, and the concept of antibiotic resistance. This last topic was asked using an open-ended question. Data was digitized and proportions and percentages were calculated for each response.

Ethics statement

Chicken meat was purchased commercially according to market availability. Researchers did not participate or oversee slaughtering of poultry. The protocol to perform structured interviews to small-scale poultry producers and a survey to feed store customers was reviewed and approved by the ethics committee of Asociacion Benefica PRISMA. Written informed consent was obtained from all study participants by the field worker performing the interview.

Results

There were 59 feed samples tested: 21 samples of “starter”, 20 samples of “growth”, and 18 samples of “fattening” (Table 1). Of the 59 samples, 28 were positive for antibiotic residues (47%). The “starter” samples had the highest percent of positivity at 52% (11/21). The “growth” and “fattening” feed were 45% (9/20) and 44% (8/18) positive for antibiotics, respectively (Table 1). There were no significant differences in the prevalence of positivity to antibiotic residues between classes of feed.

Table 1.

Prevalence of antibiotic residuals in animal feed. There were three different types of feed sampled in the study: “Starter”, “Growth” and “Fattening”. The “Starter” feed had the highest positivity rate for antibiotic residue while “Growth” and “Fattening” had roughly the same positivity rate. N = 59. Fisher’s exact testing showed no difference in prevalence between feed classes (initial, growth, fattening) Starter vs Fattening (P-value = 0.76); Growth vs (P-value = 1.0), Starter vs P-value = 0.75.

Feed type	Percent positive	Total
Starter	52.4	21
Growth	45.0	20
Fattening	44.4	18
Total	49.2	59

A total of 276 chicken samples were obtained between December 2024 and March 2025 from four supermarkets and three urban open-air markets in the city of Iquitos, and from four rural open-air markets outside of Iquitos. Of these, 177 were commercially produced broiler chickens, 85 of which came from supermarkets and 92 of which came from open-air markets. Additionally, 99 meat samples from backyard regional chickens were obtained from open-air markets. Of the 276 samples of chicken meat, 84 (30%) samples were positive for antibiotic residues. There was antibiotic residue in 28% (24/85) samples of broiler chicken from the supermarket and in 27 % (25/92) samples from open-air markets (Table 2). Regarding meat from regional chickens, 35% (35/99) tested samples were positive for antibiotic residues (Table 2). There were no significant differences between the prevalence of antibiotic residues in poultry meat from broilers and regional chickens. Similarly, there was no significant difference between the prevalence of antibiotics in commercial supermarkets and open-air markets.

Table 2.

Prevalence of antibiotic residuals in chicken meat. Locally raised backyard chickens had a higher positivity rate for antibiotic residues than the broiler chickens. The supermarket and urban open-air market had comparable positivity rates for broiler chickens. The urban open-air market and rural open-air market had comparable positivity rates for regional chickens. N = 276. Fisher’s exact tests showed no significant differences between the prevalence of antibiotics in poultry meat across different market types or between broiler and regional chickens.

Market type	Broiler chickens		Locally-raised backyard chickens
	Percent positive	Total	Percent positive	Total
Supermarket	28.2	85	NA	0
Urban open-air market	29.1	86	35.6	87
Rural open-air marketa	0	6	33.3	12
Total	27.7	177	35.4	99

^aSample size is small and results should be interpreted with caution.

Semi-structured interviews

Demographics

In the communities outside of Iquitos, there were 10 poultry farmers interviewed. Nine interviewees were owners; one interviewee was a worker. The interviewees had a range of experience from new poultry owners, to raising chickens for 29 years. Eight of the owners had between 2 and 8 years of experience. Between 300 and 10,000 chickens were raised by the interviews; most raised broilers, but two interviewees raised regional chickens and one raised layer hens.

Medication

Most interviewees use vitamins, vaccines, and antibiotics to prevent and treat a range of chicken sicknesses. One interviewee used traditional medicine such as an onion and shoe polish to treat their chicken. Seven of the interviewees have a veterinarian prescribe medication—mostly only if the sickness “is bad”. One interviewee’s poultry raising program prescribes the medication, one interviewee has a family friend prescribe medication to the chickens, and one interviewee uses an agricultural advisor. One interviewee sends their chicken stool in for an antibiogram to determine the antibiotic resistance of the bacteria.

Antibiotics

Antibiotics were used for many different purposes. The interviewees used antibiotics differently, including always mixed in water, or only when chickens are sick, for growth promotion, at scheduled intervals, when the weather changes, or only for “coughing”. Four interviewees believe that antibiotics will stop working someday. Three interviewees believe that antibiotics will always work. Two interviewees understand antibiotic resistance—these interviewees were a veterinarian and doctor. Six interviewees did not know what antibiotic resistance was, and two did not respond.

Surveys to seed store customer

A total of 212 customers outside of the feed stores were surveyed. Of the 212 surveyed, 29% (52/212) of the customers were aged > 50 years, 23% (48/212) were between 18 and 25 years old, 18% (38/212) were between 26 and 30, 17% (35/212) were 30–39, and 14% (29/212) were between 40 and 49 years. There were 101 men surveyed, and 103 women surveyed. Only 35 of 211 (17%) customers responded that they had given antibiotics to their livestock before; however, 42% (87/212) customers did not respond to this question. When asked who recommended medicine to them, 21% (44/212) respondents answered veterinarians, 16% (34/212) respondents answered the retailers at the feed store owner, 11% (24/212) respondents said their friends, and 3% (6/212) respondents answered a colleague. Most respondents (104) did not answer this question. Of the 153 respondents that answered the question: “Have you heard of the word “antibiotic resistance” before?”, 90.8% (139/153) had not heard of the phrase (Table 3)

Table 3.

Survey results applied to feed store customers. Most feed store customers were aged > 50 years old or < 25 years old. Many customers chose not to answer if they gave antibiotics to their animals; only 17% of customers said “yes”. Veterinarians and Feed Store vendors recommended antibiotics to the respondents, but most respondents chose not to answer.

Question	Percent	Total
Age
18–25 years	23%	48
26–30 years	18%	38
30–39 years	17%	35
40–49 years	14%	29
> 50 years	29%	62
Sex
Male	48%	101
Female	49%	103
Prefer not to answer	4%	8
Have you given antibiotics to your animals before?
No	42%	89
Yes	17%	35
Prefer not to answer	42%	88
Who recommends the antibiotics to you?
Feed store vendor	16%	34
Neighbor/Friend	11%	24
Colleague	3%	6
Veterinarian	21%	44
Internet	0%	0
Other	0	0
Prefer not to answer	49%	104
Have you heard of the word “antibiotic resistance” before?
No	66%	139
Yes	8%	17
Prefer not to answer	26%	56
Total	-	212

Discussion

A One Health framework is necessary for the global control of AMR. Antimicrobial exposure of microbes in animal food production results in selective pressure on zoonotic pathogens to develop resistance to medically important antibiotics that then result in drug-resistant human infections. Additionally, antibiotics in human foodstuffs expose the human microbiome to antibiotics which broadly impacts all colonizing potential pathogens to this selective pressure. Control of microbial exposure of animals and the environment shared by humans and other animals is fundamental to maintaining the efficacy of the limited antimicrobials that we have to treat infections.

Because of a local study demonstrating high levels of macrolide resistance in Campylobacter isolates from children [17], we sought to characterize the exposure of chicken to antibiotics in the industrial and community context. The use of macrolides in poultry has been associated with macrolide resistance in Campylobacter in animal and human populations [5]. Broiler chicken samples had a positivity rate of 27.7%, whereas regional chicken, or Iquitos equivalent to an organic chicken, had a positivity rate of 35.4% of samples testing positive for antibiotic residues a prevalence that was not significantly different. The measured prevalence is credible given international reviews of the antibiotic residues in meat [18]. Limitations of this study include a limited sample size which did not permit a multivariable analysis and our lack of confirmation of antibiotic detection in samples with LC-MS/MS.

The prevalence of antibiotic residues in meat identified by this study was higher than those seen in Indonesia [19] and Bangladesh [20]. Widiasih et al. [19] found that 11.3% of chicken samples were positive when tested with Premi Tests. Chicken sold at the supermarket accounted for seven of the 13 positive chicken samples, while the central market had no positive chicken samples. In Bangladesh, 15.8% of samples of chicken feed tested positive for antibiotic additives and 23.5% of the poultry meat samples tested positive for antibiotic residues [20]. The prevalence of antibiotic residues was higher in Nigeria, where in a recent study 60% of the poultry (42 of 70) tested was positive for antibiotic residues [21]. In Lebanon, 62 out of 80 samples from poultry meat (77%) contained antibiotics [22]. There were only trace amounts of sulfonamides, quinolones, and tetracyclines, but the mean concentration levels of amoxicillin and penicillin G were above the maximum residue levels of the European Union [22]. In Tanzania, Nonga et al. [23] found that 14 of the 20 chicken meat samples (70%) were positive for antibiotic residues. Farmers surveyed used antibiotics frequently and slaughtered their chicken before the withdrawal period ended [23].

Although it is commonly believed that locally-reared backyard chickens are healthier and more wholesome given that they are raised at households in free-range settings, the levels of antibiotic residues in the meat of these animals were as high in these chickens as observed in industrially produced chickens. It appears that in this context, many small-scale farmers may be using antibiotics in local husbandry without their knowledge. Animal feed, not labeled or clearly stated as having antibiotics, was commonly found to have antibiotic additives from most local community commercial sources. Thus, although individuals raising animals did at times specifically seek out antibiotics to administer to their animals, 49.2% of the feed that is purchased in Iquitos has antibiotic additives and sold without any labeling or information regarding this ingredient.

We attempted to gain insight into the beliefs of local chicken producers regarding antibiotic use in animal husbandry. We experienced a high rate of non-response in questions specifically targeted toward personal use of antibiotics in their animals, which we feel was evidence of social desirability bias. Understanding both belief and practice at the local level is important, but this qualitative component of our study is of limited value and would need to be repeated with more advanced strategies to reduce desirability bias, such as randomized response technique or hypothetical scenario questionnaires [24].

Current methods for detecting antibiotic residues in foodstuffs in general and meat in particular have serious limitations which limit the screening, risk assessment, and enforcement of restrictions in use [25]. The most rigorous LC-MS/MS tests are sensitive, specific and can be calibrated to detect many antibiotics. However, the assays are demanding to standardize, expensive to perform (50–100 US$ a sample) and not able to be run outside of specialized reference laboratories. There are also simpler enzyme-linked immunosorbent assays, but they are less sensitive and are specific to individual antibiotics, which is problematic as antibiotic use is often occult and misreported and which antibiotic to test for is not clear as use is generally occult and misreported. Newer assays are under development [26], but not widely available to detect large range of antimicrobials in medical and veterinary use. We thus selected a broad phenotypic assay that would detect multiple classes of antimicrobials that are used in humans and their active metabolites. Confirmation by LC-MS/MS would have been desirable, particularly regarding the specific detection and quantification of macrolides (tylosin, erythromycin) but is not readily feasible in this and most contexts. This highlights the need for point of care tests with improved product performance characteristics to enable surveillance of antibiotic residuals in more diverse settings to enable knowledge of the true prevalence of exposure to risk and to enable regulatory oversite to control presence of these residues in the global food network.

Conclusion

Commercial poultry feed and approximately one-third of the chicken meat produced in the industrial sector and the community sector in the region of Iquitos Perú are contaminated by antibiotic residues. There is a need for increased judiciousness and monitoring of antimicrobial use in both the human and veterinary sectors.