Animal husbandry and environmental conditions are associated with cefotaxime-resistant <i>Escherichia coli</i> in yard soil in peri-urban Malawi

Figures
Abstract
Soil is an important reservoir for antimicrobial resistance (AMR) and increasingly recognized as a pathogen transmission pathway, especially for young children. However, drivers of domestic AMR soil contamination in low-income countries remain unidentified. We conducted a cross-sectional study with 237 peri-urban households in southern Malawi to identify household and environmental factors associated with cefotaxime-resistant E. coli in yard soil. Enumerators employed structured surveys and sampled 900 cm2 of yard soil per household. We enumerated cefotaxime-resistant E. coli in soil using IDEXX Quanti-Tray/2000 with Colilert-18 and cefotaxime supplement, and assessed associations with household sanitation, animal ownership and management, child health and antibiotic use, and weather. Among children <5 years, 25–90% played, ate, slept, or crawled on the ground outside. Of 233 soil samples, 69% harbored cefotaxime-resistant E. coli at a mean of 0.90 log10 most probable number (MPN) per dry gram. Compared to households without animals, household soil had approximately 0.50-log lower mean cefotaxime-resistant E. coli concentration if animals were enclosed at night and 0.40-log higher concentration if they were not (p-values<0.05). Mean cefotaxime-resistant E. coli concentrations were approximately 0.90-log lower if soil was dry at the time of collection, 0.70-log lower if the household was in the top wealth quintile (p-values<0.005), and 0.30-log lower if any child in the household used antibiotics in the last four weeks (p-value = 0.05). There were no associations with daytime animal confinement, household sanitation, temperature, and ambient humidity. Findings suggest that animal husbandry and soil moisture had stronger associations with cefotaxime-resistant E. coli in soil compared to sanitation or antibiotic use, underscoring the importance of a One Health approach to AMR that incorporates domestic animals and environmental factors. Given children’s frequent soil contact, our findings also highlight potential AMR acquisition from soilborne pathways. Studies should quantify soilborne AMR exposure and evaluate associations with animal management/enclosure practices.
Citation: Budden E, Niven CG, Clark B, Floess E, Chirwa B, Matekenya M, et al. (2026) Animal husbandry and environmental conditions are associated with cefotaxime-resistant Escherichia coli in yard soil in peri-urban Malawi. PLOS Glob Public Health 6(7): e0006264. https://doi.org/10.1371/journal.pgph.0006264
Editor: Ben Pascoe, University of Oxford, UNITED KINGDOM OF GREAT BRITAIN AND NORTHERN IRELAND
Received: March 20, 2026; Accepted: June 22, 2026; Published: July 13, 2026
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: The minimal data set and accompanying code, where generated is available at The Open Science Framework via https://doi.org/10.17605/OSF.IO/5RG6E.
Funding: Grants that supported this work include the following: National Science Foundation (1951006 to CLW), National Science Foundation (1950212 to ARH), National Science Foundation (2236372 FDLR), and United States Department of Education GAANN (P200A240029 to EB). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Antimicrobial resistance (AMR) caused 1.27 million deaths in 2019 [1]. By 2050, AMR is projected to result in 10 million deaths per year [2], while others predict 2 million annual deaths directly from AMR and 8 million deaths related to AMR [3]. Infections caused by antibiotic-resistant organisms (AROs) are increasingly difficult to treat with available antibiotics and are associated with increased risk of death and complications, duration of illness, and healthcare costs [4–7]. Community colonization with AMR is particularly high in low-income countries. In sub-Saharan Africa, the prevalence of intestinal colonization with extended-spectrum beta-lactamase (ESBL) producing E. coli ranged from 5 to 59%, where children <5 years in the Central African Republic had the highest prevalence of colonization [8]. Further, a study conducted in 2023 found that 89% of infants in a Kenya hospital were colonized with ESBL-producing bacteria [9]. In many low-income countries, regulations on antibiotic use are limited [8], and clinically-relevant antibiotics are often available without restriction, both for human use and animal husbandry [10]. Frequent antibiotic use is a known driver of AMR, including self-prescribing or over-prescribing by healthcare professionals [11] and therapeutic or sub-therapeutic use for animals [12].
Poor water and sanitation infrastructure, and contaminated environments further facilitate the spread of AROs and antimicrobial resistance genes (ARGs) between reservoirs and hosts via horizontal gene transfer [13–15]. Antibiotic residues, AROs, and ARGs can be shed in human feces and disseminated into the environment in settings with open defecation or inadequate sanitation [16,17]. Similarly, when domestic animals defecate near the dwelling, antibiotic residues, AROs and ARGs from their feces can spread into soil and other environmental compartments [18,19], as domestic animals are frequently administered antibiotics both therapeutically to treat illness and sub-therapeutically to support growth [20,21]. Within environmental compartments, mobile genetic elements can be exchanged between native communities and pathogens deposited via human or animal fecal waste [22,23]. Environmental conditions, such as temperature and precipitation, may affect the proliferation of ARGs [24]. Additionally, while ARGs can spread naturally in the environment, anthropogenic disturbances can add selective pressures [16,25]. For example, ARG exchange between organisms can be triggered by co-selectors, such as heavy metals, pesticides, and microplastics [25].
As one of the most diverse matrices for microbial activity, soil is a particularly important environmental reservoir for AMR [26] because it can offer an efficient environment for ARGs to transfer between soil-inhabited bacteria and clinically-relevant pathogens, such as E. coli, Klebsiella pneumoniae, and Salmonella enterica [27,28] (Fig 1). In rural Bangladesh, a study detected cefotaxime-resistant E. coli on 89% of soil floors [29]. Resistance against at least one antibiotic was detected in 42% of E. coli isolates from yard soil samples in Bangladesh [30] and 33% of isolates in Peru [31]. Further, a recent study in Nigeria found ESBL-producing E. coli in 68% of farm soil samples [32], and a study conducted along a river basin in Tanzania found that 14% of samples positive for E. coli harbored ESBL-producing organisms [33].
Soil contaminated with fecal waste can be directly ingested by young children via mouthing behavior or indirectly via contaminated food, water, fomites, and/or hands, which can transfer fecal pathogens to hosts [16,34,35] (Fig 1). Direct soil consumption among children has been correlated with increased diarrheal diseases [36,37]. Early childhood behaviors, such as mouthing soiled hands or objects, further elevates exposure to indirect soilborne routes [38]. In a Monte Carlo simulation utilizing data from rural Bangladesh, 40% of children between six months and two years of age were found to ingest soil [39]. While no studies have assessed whether soil exposure is associated with increased AMR colonization, it is plausible that AROs and ARGs can be transmitted to hosts via soilborne pathways. For example, in a 2022 study in Malawi, having a toilet with a soil floor was associated with higher odds of having ESBL-producing E. coli in household members’ stool samples [40].
Sanitation practices have been explored as a determinant of both fecal and AMR contamination in domestic soils, but the evidence is mixed for fecal indicators and scarce for AMR. A study in Tanzania found that having an improved versus unimproved pit latrine was not associated with fecal indicator bacteria in yard soil; soil at the entrance to the home had some of the highest concentrations of fecal indicator bacteria of all sampled locations, including the latrine entrance [41]. It was hypothesized that latrine superstructure may influence this finding because most latrines had open roofs, which may result in bacterial inactivation by sunlight [41]. Similarly, a 2018 study in rural Bangladesh found that having an improved versus unimproved pit latrine was not associated with E. coli or ESBL-producing E. coli in yard soil [30]. A randomized sanitation intervention providing improved latrines to rural Bangladeshi households did not reduce E. coli levels in yard soil; though this study did not evaluate resistance [42]. Sanitation improvements are recommended in AMR control programs [43], and a better understanding of how sanitation infrastructure/practices influence AMR in soil can inform AMR mitigation strategies.
Animal husbandry is a major source of fecal contamination in the environment and plays an important role in disseminating AMR [13]. Hog feces, soil from hog farms, and samples from poultry and cattle farms in various settings have been shown to have high rates of ARG detection, pinpointing animal feces as a source of AMR in animal production facilities [44–48]. Other studies have investigated fecal indicator bacteria and AMR in informal backyard animal husbandry settings, a common practice in low- and middle-income countries where animals are raised within the premises. In Bangladesh, the presence of chickens, cows, goats, and sheep in the household compound increased E. coli concentration in yard soil, and chicken ownership in particular was correlated with significantly higher concentration [34]. Also in Bangladesh, cefotaxime-resistant E. coli on indoor soil floors had a strong correlation with owning domestic animals and unsafe animal management practices, such as letting animals roam free in the home and keeping them inside the home at night [29]. A recent review found that, in low- and middle-income countries, animals roaming in or near living spaces allows for the exchange of ARGs through environmental compartments, and soil in particular plays a critical role as a mediator in AMR transmission between humans and domestic animals [49]. Given that backyard animal husbandry is a common form of livelihood in low-income countries, the influence of specific animal management practices (e.g., roaming, daytime and nighttime confinement) on the environmental dissemination of AMR needs further study.
Prior research has documented that microbial growth is influenced by environmental and weather factors. Higher soil moisture content promotes proliferation of E. coli and ESBL-producing E. coli in soil [30]. A 2024 study found a higher prevalence of beta-lactamase genes in soils where manure was applied under higher moisture conditions [50]. Temperature and sunlight also influence E. coli concentration in soil [41,51]. A study in rural Tanzania found lower concentrations of fecal indicator bacteria in soil that was in direct sunlight compared to in shade or partial sunlight [41]. Additionally, E. coli inoculated into soils from Lake Superior (bordered by the United States and Canada) were found to grow more successfully at temperatures ≥30 °C, with the most growth occurring at 37 °C [51]. Further evidence is needed on the influence of environmental conditions (e.g., sunlight, rainfall, ambient temperature, ambient humidity, soil moisture) specifically on antimicrobial-resistant organisms.
In this study, we aimed to quantify cefotaxime-resistant E. coli in yard soil among peri-urban Malawian households. Cefotaxime is a broad-spectrum cephalosporin antibiotic that is commonly used in clinical settings [52] and a known indicator for ESBL-producing bacteria, which are of particular concern because they confer multi-drug resistance against clinically relevant antimicrobials [52]. Specifically, ESBL-producing E. coli forms the basis of World Health Organization’s global AMR surveillance in environmental compartments [53]. Here, we used a One Health approach to assess associations between the concentration of cefotaxime-resistant E. coli in yard soil and a range of human, animal and environmental factors, including sanitation, animal management practices, antibiotic use by humans and animals, and weather-related variables.
Methods
Study design
Ethical considerations
Data collection
Enumerators from Mzuzu University and the Malawi University of Science and Technology were trained to administer a structured questionnaire in Chichewa using Open Data Kit (ODK). The questionnaire included questions and spot check observations of household water and sanitation indicators drawn from the Joint Monitoring Programme’s (JMP) Core Questions on Drinking Water, Sanitation and Hygiene for Household Surveys (water source, latrine type and access, child defecation) [55], animal ownership and management (types of domestic animals, daytime and nighttime confinement, antibiotic use), caregiver-reported child health for children <5 years (enteric/respiratory infection symptoms, antibiotic use), environmental factors (sampled soil in sunlight or wet), socio-demographic indicators (assets, expenditures, education, household size, floor material), and caregiver-reported child soil exposure behaviors. If antibiotic use was reported, enumerators asked caregivers for the packaging of any reported antibiotics to confirm. We installed five EXTECH Instruments Humidity/Temperature Dataloggers RHT10 in various locations around Bangwe to record ambient temperature and relative humidity every ten minutes throughout the data collection period. Additionally, we collected publicly available weather data from Weather API, which recorded ambient temperature and humidity every three hours in Blantyre [56], the closest city to Bangwe.
Soil sample collection
Enumerators trained in aseptic technique identified a courtyard area immediately adjacent to the entrance of the household, avoiding stepping into the sampling area. Two stainless steel corner stencils and a stainless steel spoon were sterilized with 10% bleach and 70% ethanol. The stencils were placed on the soil to measure a 30 x 30 cm area. The spoon was used to scrape the topsoil 10 times vertically and 10 times horizontally and transfer the collected soil into a sterile Whirlpak bag. The Whirlpak bag was placed in a secondary plastic bag to avoid cross-contamination. Samples were transported to the Biological Sciences Laboratory at the Malawi University of Science and Technology in coolers with ice. A field blank was administered every other day. At the first household visited, enumerators opened a Whirlpak bag containing sterile DI water, stirred the sanitized sampling spoon ten times, closed the bag, and returned it to the lab with the soil samples. Due to resource constraints, replicate samples were not collected.
Enumeration of cefotaxime-resistant E. coli
Samples were processed within five hours of collection. The soil sample was homogenized by shaking the Whirlpak bag. Then, 4 g of soil and 40 mL of sterile DI water were added to a 50 mL falcon tube and vortexed for 2 minutes. 10 mL of the homogenized sample was retrieved without letting the soil settle and added to a sterile 100 mL Whirlpak bag containing 90 mL of sterile DI water, creating a 100 mL aliquot that contained 1 g of soil. 80 µL of 5 mg/mL filter-sterilized cefotaxime was added to the 100 mL solution, for a final cefotaxime concentration of 0.004 mg/mL [57], followed by adding Colilert-18 media. Whirlpak bags were gently inverted to mix and, once the media dissolved, the contents were added into a Quanti-Tray/2000, sealed, incubated at 35 °C for 20–22 hours, and visually enumerated for the most probable number (MPN) of cefotaxime-resistant E. coli using a UV lamp. To assess moisture content, an approximately 5 g aliquot of soil was weighed to record the wet mass [34], placed into a drying oven at 110 °C, and weighed again after 24 hours to record the dry mass. The moisture content was used to calculate the MPN of cefotaxime-resistant E. coli per dry gram of soil.
Field blanks and 10% lab blanks were analyzed using the same procedure as actual samples, using 100 mL of field blank sample and 100 mL of sterile DI water, respectively. Two aliquots were analyzed for each blank, one with and one without cefotaxime. We did not assess samples for generic E. coli nor process duplicate samples or positive controls due to resource constraints. However, other studies conducted in low- and middle-income country settings have found a high concentration of generic E. coli in soil [29,34,41]. Our enumeration approach was validated in a previous study, which found that the positive control (E. coli NC11) used yielded no statistically significant differences in cefotaxime resistant E. coli concentration between the IDEXX method and MacConkey plates [57].
Statistical analysis
We generated a binary variable for the prevalence of cefotaxime-resistant E. coli and log10-transformed the MPN values per dry gram. Non-detect samples were assigned 0.5 MPN/wet gram of soil (half the lower detection limit) [58], and samples exceeding the upper limit of quantification of 2419 MPN were assigned 2420 MPN/wet gram prior to calculating the MPN/dry gram and taking the logarithm. We tabulated the mean log10-transformed MPN of cefotaxime-resistant E. coli in yard soil across cross-categories of animal ownership and husbandry practices and cross-categories of latrine type and user load. We used bivariate and multivariable regression to assess associations between household and environmental variables and the log10-transformed MPN and prevalence of cefotaxime-resistant E. coli. We considered the following independent variables in bivariate models.
Sanitation.
We generated binary variables for whether: any feces observed within 2x2 meters of soil sampling area; the household had an improved latrine according to the JMP definition [59]; the household had a flush/pour flush latrine; the latrine was utilized by a single household or shared by multiple households; and children in the household open defecated.
Animal ownership and management.
We generated binary variables for whether: the household owned animals; animals were kept in the compound (any animals combined, as well as specific animal types); and any animal feces were observed in the compound. Among households that owned animals, we generated binary variables for whether animals were enclosed during the day and at night, and whether the household gave them antibiotics (to any animals, as well as to specific animal types). Animals were considered enclosed if they were kept in a basket or cage, an animal-only enclosure that they could not leave, or an animal-only enclosure that they could leave freely. Additionally, we generated a categorical variable for animal cohabitation intensity: household owns animals that are not enclosed, household owns animals that are enclosed, household does not own animals.
Environmental factors.
We generated binary variables for whether the soil sampling area was in sunlight and whether it was wet at the time of collection as observed by the enumerator. We performed a two-tailed t-test to determine if the enumerator-recorded observation of soil wetness correlated with the sample’s measured moisture content. We tabulated the daily average temperature and humidity using data from the loggers. For study dates where only a subset of loggers was online for 24 hours because of installation/de-installation partway through the day (June 25), we only used data from the loggers that had data for the full 24-hour period. For study dates with no available logger data (June 20, 21, 24), we calculated the daily average temperature and humidity using data from Weather API (weatherapi.com/history/q/blantyre-1598090). We created tertiles of daily average temperature and humidity across the study period and generated binary variables for the top vs. bottom two tertiles.
Child illness and antibiotic use.
We generated binary variables for whether any child <5 years in the household had symptoms of caregiver-reported diarrhea, acute respiratory infection (ARI), ARI with fever, or fever in the last seven days, and if any child <5 years in the household used antibiotics in the last four weeks. Diarrhea was defined as 3 or more loose stools in a 24-hour period. ARI was defined as persistent coughing accompanied with shortness of breath, panting, or wheezing [60].
Bivariate models were run for each independent variable. Variables with a p-value of <0.20 in bivariate models were shortlisted for inclusion in multivariable regression models. This broad inclusion criterion was used to ensure that potentially relevant variables were not disregarded. Variables with a p-value <0.05 in multivariable models were considered to have a statistically significant association with the outcome. We estimated variance inflation factors to assess collinearity between weather-related variables (sampling area in sunlight, sampling area wet, average ambient temperature, average ambient humidity) before including them in multivariable models.
Multivariable models also controlled for household’s primary water source, highest education level by any household member, household size, weekly household expenditure, indoor floor material and asset-based household wealth quintile. The water source was categorized according to the JMP definition of improved vs. unimproved [59]. The floor material was categorized as improved (cement/concrete, tile) vs. unimproved (earth/mud). The wealth quintile was calculated using principle component analysis of reported assets [61]. These additional variables were selected for inclusion in multivariable models as potential confounders because household-level water service and socio-demographic factors are expected to influence AMR in the domestic environment [62,63]. Models only included confounders with sufficient variation in the sample (> 5% prevalence in each stratum).
We used generalized linear models to estimate associations. Models with continuous outcomes (log10-MPN) used a Gaussian error distribution and identity link to estimate differences in log10-transformed MPN, and models with binary outcomes (prevalence) used a Poisson distribution and log link to estimate prevalence ratios. Models accounted for geographic clustering of enrolled households using robust standard errors. All analyses were conducted in R version 4.3.3 through Rstudio version 2024.12.0 + 467. De-identified data and analysis scripts are publicly available at Open Science Framework (https://doi.org/10.17605/OSF.IO/5RG6E).
Results
Participant characteristics
We enrolled 237 households and grouped them into 38 independent spatial clusters based on their GPS coordinates [54]. Respondents were an average age of 31.1 years old, and households included an average of 1.2 children <5 years, 1.4 children aged 5–16 years, and 2.7 adults >16 years (S1 Table). The most common education level was secondary (partial or complete), and 87% (205) of respondents reported being able to read and write (S1 Table). Average weekly household expenditure was 17.0 USD (S1 Table). Additionally, 69% (164) of households had electricity in the previous 7 days, 24% (56) owned a refrigerator, 70% (166) owned a mosquito net, and 87% (205) owned a mobile phone (S1 Table).
Water and sanitation practices.
84% (200) of households had an improved primary water source, and the most common source was piped water outside the compound (29%, n = 69) (S1 Table). Most households had a latrine, while 19% (46) had an improved latrine, and the most common latrine type was a pit latrine with slab (65%, n = 154) (S1 Table). 71% (168) of respondents reported sharing their latrine with other households, with an average of 11 people per shared latrine (S1 Table). The primary place of defecation for children <5 years was in latrines (40%, n = 95) and nappy/diapers (39%, n = 92) (S1 Table). However, children <5 years were also reported to openly defecate in the compound in 25% (59) of households and outside of the compound in 5% (11) of households (S1 Table). The majority (65%, n = 153) of respondents reported putting/rinsing child feces into the latrine (S1 Table). In 8% (20) of households, fecal waste was observed within 2x2 meters of the soil sampling area in the yard (S1 Table).
Animal ownership and management.
Animals were owned by 24% (57) of households; the most common animal was chickens/poultry (16%, n = 37) while other animals included dogs (9%, n = 22), cats (5%, n = 12) and sheep/goats (0.4%, n = 1) (Table 1). Almost three quarters (75%, n = 177) of households reported that animals lived in their compound, where a compound was defined as the land area where the household lives, including extended family members. The most common animal living in the compounds was chickens/poultry (61%, n = 144) while dogs, cats, sheep/goats, and pigs were also reported (Table 1). Households owned an average of 1.8 animals, and an average of 7.1 animals lived in the compound (Table 1). Animals were observed within 2x2 meters of the soil sampling area in the yard in 7% (16) of households, and animal feces were observed in the compound in 37% (21) of households (Table 1).
Among the 57 households that owned animals, during the day animals roamed free outdoors in 49% (28), roamed free indoors in 47% (27), and were enclosed in 12% (7) of households (Table 1). At night, animals roamed free outdoors in 26% (15), roamed free indoors in 67% (38), and were enclosed in 21% (12) of households (Table 1). Among households that used a daytime enclosure, 57% (4) had only chickens/poultry, 29% (2) had chickens/poultry and other animals, and 14% (1) had only other animals, while among households that used a nighttime enclosure, 50% (6) had only chickens/poultry, 42% (5) had chickens/poultry and other animals, and 8% (1) had only other animals (Table 1). In the last four weeks, 28% (16) of households gave antibiotics to any animal in the household, most commonly to chickens/poultry (Table 1).
Child illness, antibiotic use and soil exposure.
Among 292 children <5 years in enrolled households, caregivers reported that 16% (48) of children had diarrhea, 46% (134) had an ARI, 24% (70) had an ARI with fever, and 33% (97) had a fever in the last 7 days. In the last four weeks, 33% (97) of children used antibiotics, with an average of 1.2 courses and 1.8 doses of antibiotics taken (Table 2). Of the 97 children who had used antibiotics, the packaging was available for confirming the antibiotic for 17 children. Among these, the packaging matched the reported antibiotic for 76% (13) of children, while for 4 children, the packaging indicated that the reported medication was not an antibiotic. The most common antibiotics were amoxicillin (part of penicillin class) and trimethoprim/sulfamethoxazole (part of sulfonamide class) (Table 2). Antibiotic recommendations most commonly came from doctors (54%, n = 52), followed by self-prescription (29%, n = 28) and pharmacists (17%, n = 16) (Table 2). The most common reason why the child stopped taking antibiotics was that the child felt better (47%, n = 46) (Table 2). In the last three months, caregivers reported an instance of antibiotics being ineffective for 25% (24) of children, and of these, 71% (n = 17) took an additional antibiotic to improve (Table 2). Additionally, among children <5 years, 90% (263) played, 74% (217) ate, 49% (143) slept, and 25% (74) crawled on the ground outside; 9% (27) of children spent no time on the ground outside (Table 2).
Cefotaxime-resistant E. coli in yard soil
Yard soil samples were collected from 233 of 237 enrolled households; 4 households declined sample collection. Additionally, in 2 samples, we were able to assess the binary prevalence of cefotaxime-resistant E. coli but could not quantify it because the IDEXX tray did not seal correctly. 69% (160) of samples harbored cefotaxime-resistant E. coli at a mean concentration of 0.90 log10-MPN/dry gram (standard deviation [SD]=1.1), corresponding to a geometric mean of 8.0 MPN/dry gram (Table 3).
Soil and weather conditions
At the time of collection, 39% (90) of soil samples were in sunlight and 23% (53) were dry (Table 3). When the sample was visually wet as reported by enumerators, the measured soil moisture content was significantly higher (20% for wet soil vs. 4% for dry soil, t-test p-value<0.05) (Table 3 and S1 Fig). The daily average temperature was 18.0°C, and the daily average relative humidity was 67% over the study period (Table 3).
Household and environmental factors associated with cefotaxime-resistant E. coli in yard soil
Among cross-categories of animal type and animal enclosure practices, the concentration of cefotaxime-resistant E. coli in yard soil was highest among households that owned poultry and did not enclose their animals at night (mean log10-MPN: 1.42) and generally appeared higher when animals were not enclosed at night, regardless of the types of animals owned (S2 Fig). Among cross-categories of number of latrine users and latrine type, cefotaxime-resistant E. coli concentration in soil was high among households whose latrine was unimproved and used by an above-median (>9) number of people (mean log10-MPN: 1.05), as well as among households whose latrine was not flush/pour flush and used by an above-median number of people (mean log10-MPN: 1.07). Cefotaxime-resistant E. coli concentration in soil generally appeared higher among households where the latrine was used by multiple households. Unless the latrine was used by a single household, households with an unimproved latrine appeared to have higher cefotaxime-resistant E. coli levels compared to households with an improved latrine.
In bivariate regression, households had a statistically significant lower concentration of cefotaxime-resistant E. coli per dry gram of yard soil if owned animals were reported to be enclosed at night (log10: -1.12 (-1.56, -0.67); p-value<0.0005), if the family’s latrine was used by a single household rather than shared with other households (log10: -0.42 (-0.72, -0.12); p-value = 0.01), if the soil was dry at time of collection (log10: -0.92 (-1.29, -0.56); p-value<0.0005), or if the ambient temperature was in the top tertile (log10: -0.31 (-0.59, -0.03); p-value = 0.03) (S2 Table). While not statistically significant, households also appeared to have lower concentration of cefotaxime-resistant E. coli if they had an improved latrine, soil was in sunlight at time of collection, or if children used antibiotics in the last four weeks (S2 Table). On the other hand, households appeared to have a higher concentration of cefotaxime-resistant E. coli in yard soil if the household owned chickens/poultry (S2 Table). The following variables were not associated with cefotaxime-resistant E. coli concentration in yard soil: utilizing a flush/pour flush latrine, children in household openly defecating, feces or animals observed near sampling area, owning animals, owning dogs/cats, compound keeping animals, animals enclosed during the day, household giving antibiotics to any animal, ambient humidity, and children having diarrhea, ARI, fever, or ARI with fever (S2 Table).
In multivariable models that adjusted for potential confounders, animal cohabitation intensity had a strong association with the concentration of cefotaxime-resistant E. coli in yard soil. Specifically, compared to households that did not own animals, households that owned animals and kept them enclosed at night had a significantly lower concentration of cefotaxime-resistant E. coli in yard soil (log10: -0.50 (-0.83, -0.16); p-value<0.005), whereas households that owned animals and did not enclose them at night had a higher concentration (log10: 0.39 (0.11, 0.66); p-value<0.05) (Fig 2 and S3 Table). Households also had a significantly lower concentration of cefotaxime-resistant E. coli in yard soil if the soil was dry at the time of collection (log10: -0.87 (-1.35, -0.39); p-value<0.005) and if the household was in the top wealth quintile (log10: -0.70 (-1.10, -0.30); p-value<0.005) (Fig 2 and S3 Table). Any child in the household using antibiotics in the last four weeks was borderline associated with a lower concentration of cefotaxime-resistant E. coli in yard soil (log10: -0.26 (-0.51, 0.0); p-value = 0.05) (Fig 2 and S3 Table). In a separate multivariable model that considered poultry ownership rather than overall animal ownership, owning poultry was not associated with cefotaxime-resistant E. coli concentration, while the soil being dry and greater household wealth remained associated and any child in the household using antibiotics in the last four weeks remained borderline associated with lower cefotaxime-resistant E. coli concentration (S4 Table). Sanitation and ambient temperature did not have an association with cefotaxime-resistant E. coli concentration in yard soil in multivariable models (Fig 2 and S3 and S4 Tables).
Models include variables that were associated with the outcome with a p-value of <0.20 in bivariate analyses and also control for improved primary water source, highest education in the household, asset-based household wealth quintile, weekly household expenditure, number of people in the household and improved floor material. Circles indicate point estimates and horizontal lines indicate 95% confidence intervals.
Findings for cefotaxime-resistant E. coli prevalence in yard soil were broadly similar to those for cefotaxime-resistant E. coli concentration. In bivariate models, households had significantly lower prevalence of cefotaxime-resistant E. coli if animals were enclosed at night, any child in the household used antibiotics in the last four weeks, if the soil was in sunlight or dry at time of collection, or if the ambient temperature was in the top tertile (S5 Table). While not statistically significant, households appeared to have lower prevalence of cefotaxime-resistant E. coli if they had a flush/pour flush latrine or the latrine was used by a single household (S5 Table). In contrast, while not significant, households appeared to have higher prevalence of cefotaxime-resistant E. coli if the household owned poultry, if children openly defecated, or if the ambient humidity was in the top tertile (S5 Table).
In multivariable models, lower prevalence of cefotaxime-resistant E. coli in yard soil was associated with any child in the household using antibiotics in the last four weeks (PR: 0.84 (0.70, 0.99); p-value = 0.04), soil being dry at time of collection (PR: 0.36 (0.26, 0.49); p-value<0.0005), ambient humidity in the top tertile (PR: 0.84 (0.72, 0.97); p-value = 0.02), and the household being in higher wealth quintiles (PR: 0.64 (0.46, 0.89); p-value = 0.01) (S6 and S7 Tables).
Discussion
In our study in a peri-urban setting in Malawi, approximately two thirds of enrolled households with a child <5 years had cefotaxime-resistant E. coli in yard soil. Malawi is a low-income country and experiences high rates of AMR colonization both among humans and animals, consistent with prior evidence from other low-income countries [64–67]. A previous study in southern Malawi detected ESBL-producing E. coli or Klebsiella pneumoniae in 30% of animal feces (n = 290 samples) and 42% of human stools (n = 1190 samples) [68]. Strengthening AMR surveillance and regulations has been identified as a priority [69], and it is critical to investigate local drivers of AMR transmission. In our study, unenclosed domestic animals were associated with higher concentration of cefotaxime-resistant E. coli in yard soil. Enclosed domestic animals, soil being dry and higher household wealth were associated with lower concentration, and any child in the household using antibiotics in the last four weeks was borderline associated with lower concentration of cefotaxime-resistant E. coli in yard soil. Household sanitation was not associated with the prevalence or concentration of cefotaxime-resistant E. coli in yard soil.
Previous studies have identified soil as an important reservoir for AMR [23,30,31,70] and reported a range of prevalence values for antimicrobial-resistant E. coli in domestic, environmental and agricultural soils, including 14% of soil samples from a river basin in Tanzania [33], 42% of yard soil samples in Bangladesh [30], 68% of farm soil samples in Nigeria [32] and 89% of indoor soil floors in Bangladesh [29]. In Malawi, 77% of farm soil samples positive for E. coli showed cefotaxime resistance [71], and ESBL-producing bacteria (E. coli, K. pneumoniae) were found in 60% of soil samples from play and waste deposition areas [70]. Our finding of 69% prevalence in yard soil among peri-urban Malawian households is consistent with prior evidence. Recent studies in Malawi have also detected AMR in other environmental matrices such as farmland and household soil, vegetables, drains, standing water, and household drinking water [54,70,71]. However, in additional data from our study, the prevalence of cefotaxime-resistant E. coli in drinking water (8.4%) [54] was substantially lower than what we detected in yard soil, pointing to soil as a notable reservoir. Children <5 years in our study were reported to play, eat, sleep and crawl on the ground outside, presenting opportunities for soil exposure. Given previous evidence on the contribution of direct and indirect soil ingestion to child exposure to fecal contamination [39], our findings suggest that domestic soils can be a source of AMR exposure for young children.
As environmental pathways are increasingly recognized to facilitate AMR transmission, understanding drivers of AROs in environmental matrices is critical. Here, animal cohabitation, specifically nighttime enclosure of domestic animals, had a strong association with cefotaxime-resistant E. coli concentration in yard soil. Domestic animals often share living space with household members in low-income countries [29,72,73]. Animal presence has previously been associated with soil contamination with fecal indicators and AROs [29,74], and living in close proximity to animals has been shown to increase the odds of being colonized with AROs [75]. However, as 3.5 million people in Malawi experience chronic food insecurity [76], animal-source foods are critical for early childhood development [77]. While meat is infrequently sourced due to expense [77,78], the consumption of fish and eggs is common in Malawi [77]. Therefore, while owning domestic animals can present AMR risks, it can also support nutrition for household members, especially children.
There is also growing evidence on the contribution of specific animal management practices to soil contamination with AROs. Our findings align with a recent study in rural Bangladesh, where households who kept their animals inside the home at night or let them roam free inside the home or in the compound had higher prevalence and concentration of cefotaxime-resistant E. coli on indoor soil floors [29]. This could be due to antibiotic use in backyard animal husbandry and resulting dissemination of AMR via fecal waste from unenclosed or freely roaming domestic animals. A study in Kenya found that 80% of survey respondents utilized antibiotics in domestic livestock practices [12], while in our study 28% of households that owned animals gave them antibiotics (primarily to chickens). A recent study examining antimicrobial usage polices in Malawi found that animal husbandry is an area where further policies are needed, as clinically-relevant antimicrobials are not regulated for animal usage, and animal feces disposal is unregulated [10]. However, we found no association between animal antibiotic use and cefotaxime-resistant E. coli in yard soil. We did not record the specific antibiotics given to animals; it is possible that the antibiotics used for animals were not beta-lactams (e.g., tetracycline) and do not trigger ESBL production and/or cefotaxime resistance. Nonetheless, while antibiotic stewardship for domestic animals is recognized as an AMR control strategy [79], our findings highlight animal enclosure practices as an additional One Health approach to addressing AMR.
Poultry ownership was specifically associated with elevated cefotaxime-resistant E. coli in bivariate models in our study while this association was attenuated after controlling for potential confounders. Our findings are consistent with a 2017 study in Bangladesh that demonstrated domestic animals, especially chickens, contribute to fecal contamination in soil [34], and with a recent study that found associations between the number of chickens owned and cefotaxime-resistant E. coli on indoor soil floors [29]. In a study in Burkina Faso, children in households that owned poultry had greater odds of tetracycline resistance genes present in the gut compared to households that did not own poultry [80]. The specific contribution of poultry to fecal and AMR contamination in the domestic environment may be due to the difficulty in maintaining poultry enclosures [81]. Backyard poultry owners in low-income countries often prefer not to use enclosures because enclosed poultry require feed while free-range poultry can forage for food. Backyard poultry are also more likely to be allowed to roam freely inside living areas than larger animals, which are often tied up or kept in dedicated areas; roaming birds can contaminate the domestic environment as they scratch soil, forage and defecate. On the other hand, a study in Peru found that chicken corrals require an intensive process for maintaining hygiene [82], and corralling chickens increased Campylobacter infections among young children, compared to letting them range free [83]. Our survey collected enclosure information broadly for all animals in the compound and did not record whether poultry specifically were confined in an enclosure. However, among households that used nighttime enclosures, 50% owned poultry while 42% owned poultry and at least one other type of animal. Analyzing enclosure types and practices that households use specifically for poultry may provide further insight on how poultry management may affect cefotaxime-resistant E. coli contamination in yard soil.
Vulnerable water and sanitation infrastructure has been identified as a key barrier in addressing AMR in Malawi, as weak infrastructure perpetuates fecal-oral transmission and AMR dissemination [20,84]. In our study, cefotaxime-resistant E. coli concentration in soil was higher among households that had an unimproved latrine shared by a higher number of users, and in unadjusted regression models, multiple indicators of higher-quality sanitation, such as having an improved latrine or latrine used by a single household, were borderline associated with lower concentration of cefotaxime-resistant E. coli in yard soil. However, these associations were attenuated after controlling for potential confounders. Similarly, existing evidence on the association between sanitation and AMR is mixed and vulnerable to confounding by wealth. In a global analysis using human gut metagenome data (i.e., genetic material within the human gut) from 26 countries in Africa, North America, Europe, South America, and Asia, access to sanitation infrastructure that effectively isolates waste from human contact was associated with lower concentration of ARGs in the gut metagenome [14]. A 2018 modelling study found that lacking WaSH infrastructure was associated with higher AMR colonization in humans [19]. A study in Malawi found that having inadequate sanitation facilities was associated with higher AMR levels in drains and soil [70]. However, much of this evidence comes from observational studies, and it is possible that the findings are confounded by wealth (i.e., lower AMR in settings with better sanitation reflecting overall better health from higher wealth). In a randomized controlled trial in Bangladesh, on-site sanitation improvements did not decrease fecal contamination of courtyard soil [42], and there is mixed evidence on how effectively on-site sanitation systems isolate fecal waste from the environment in low-income countries [85]. Taken together with this prior evidence, our findings suggest that household-level sanitation practices may not fully capture drivers of cefotaxime resistance in environmental compartments such as soil; integrating other variables, such as community-level infrastructure and practices, environmental factors and animal husbandry, may identify critical additional drivers [86]. Future randomized sanitation trials that measure AMR in the environment or among residents can help generate unconfounded evidence on the relationship between sanitation status and AMR.
In our analysis, child health variables such as diarrheal and respiratory infections, which are common drivers of pediatric antibiotic use, were not associated with cefotaxime-resistant E. coli in yard soil. In bivariate models, child antibiotic use in the last month was associated with lower cefotaxime-resistant E. coli prevalence and borderline associated with lower cefotaxime-resistant E. coli concentration in soil. In multivariable models, child antibiotic use remained borderline associated with lower concentration and significantly associated with lower prevalence of cefotaxime-resistant E. coli in yard soil. Notably, approximately a third of children in our study used antibiotics in the last month. The most commonly reported antibiotics were amoxicillin and trimethoprim/sulfamethoxazole. While trimethoprim/sulfamethoxazole belong to the sulfonamide class of antibiotics [87], amoxicillin belongs to the penicillin class and is a beta-lactam antibiotic like cefotaxime [88]. Therefore, frequent amoxicillin use by children would be expected to be associated with ESBL-production and/or cefotaxime resistance in children’s guts and consequently in domestic soils, given that 25% of children in our study open defecated within the compound. Our finding of borderline association between child antibiotic use and lower cefotaxime resistant E. coli in yard soil conflicts with prior research that suggests that environmental AMR may result from high antibiotic use through the excretion of antibiotic residues [89], while it also supports existing evidence that environmental transmission may contribute more to human colonization with AMR than the direct consumption of antimicrobials [19]. We note that self-reported antibiotic use can be inaccurate. One study assessing antibiotics sold in Malawi found high numbers of unregistered antibiotics and inconsistent dosage between pills [90], while another study in Malawi that analyzed 56 antibiotic samples sold by drug retailers found six sub-standard (i.e., does not meet drug standards) and one falsified (i.e., distorts the drug’s true identity) antibiotics/antimalarials [91]. We therefore asked to see the reported antibiotic administered to children. While we were able to confirm the reported antibiotic only for a small subset of children whose caregivers had the packaging available, in more than 75% of these instances, the packaging matched the reported antibiotic. Similarly, a previous study in eight low- and middle-income countries found that caregiver-reported pediatric antibiotic use matched the children’s medical records [92]. Our findings are consistent with prior studies showing high antibiotic use among young children in sub-Saharan Africa. In a recent study in Kenya, caregivers reported that 63% of children <5 years had taken antibiotics in the previous 90 days; caregivers were asked to name the specific antibiotic used, and data were excluded if the named medication was not an antibiotic [93]. While we did not record antibiotic consumption by household members other than children <5 years, our findings suggest that factors outside of human consumption of antibiotics may be dominant drivers of cefotaxime resistance in the home environment in peri-urban Malawi.
Some environmental factors, such as soil being dry, were associated with lower cefotaxime-resistant E. coli concentration in yard soil in our study. In contrast, while high ambient temperature was associated with lower cefotaxime-resistant E. coli concentration in bivariate models, this relationship was attenuated in multivariable models, and ambient humidity was not associated with cefotaxime-resistant E. coli concentration in bivariate or multivariate models. Our findings are consistent with previous evidence on the effect of moisture on E. coli and antimicrobial behavior. E. coli proliferates in soil under favorable environmental conditions, including higher moisture content [30,94]. Sunlight can inactivate microorganisms [95] and may break down antibiotic residues in soil; when exposed to sunlight, oxytetracycline and tetracycline, common broad-spectrum antibiotics used in humans and animals [96], lose antibiotic potential [97] and therefore may not trigger resistance. On the other hand, our findings differ from previous evidence on the relationship between ambient temperature/humidity and AMR. In a global study utilizing AMR databases, AMR colonization in inpatients was more common in areas with warmer temperatures [19]. An ecological study in China utilizing human clinical data found that, although ambient humidity confounds with ambient temperature, the median relative humidity of the area may influence the proliferation of AMR alongside ambient temperatures [62]. Our findings of no association with these weather parameters may reflect our data collection period during the cool, dry winter season in Malawi. In contrast, associations between ambient temperature and AMR in human clinical data have been derived from multi-year simulation studies, which better capture temperature variations [98]. We also did not analyze soil temperature specifically, which may influence the survival and proliferation of ARGs in soil [99]. Understanding how weather affects AROs is critical, given increasing temperatures and rainfall intensity in many world regions where AMR is common.
Our study had some limitations. First, we solely measured cefotaxime-resistant E. coli, which may not capture other AMR classes or specific resistant pathogens. Cefotaxime is a third generation cephalosporin and a broad-spectrum antibiotic [52,100]. Cefotaxime resistance is conferred by ESBL production; ESBL-producing organisms have multi-drug resistance against penicillins and cephalosporins, including cefotaxime [101]. Therefore, our study used a validated protocol [57] to quantify cefotaxime-resistant E. coli in yard soil as a proxy for ESBL-producing E. coli [52], which is widely monitored as a sentinel organism for AMR in environmental matrices, such as in the World Health Organization’s Tricycle global AMR surveillance protocol [53]. We also did not measure cefotaxime-resistant E. coli or other AROs in human or animal stool to corroborate the associations with sanitation and animal management practices. Additionally, our cross-sectional data collection only captured the dry/cold season in Malawi; the concentration of cefotaxime-resistant E. coli in yard soil and factors associated with it may differ during warm/rainy seasons. Further, due to our observational cross-sectional study design, we could only establish association, not causation, between household and environmental variables and cefotaxime-resistant E. coli in yard soil. While we accounted for a range of potential confounders (water source, education, socio-demographics), the possibility of confounding remains. Finally, we collected self-reported data on household practices using a structured questionnaire. Respondents may remember events and answer survey questions inaccurately. However, we have no reason to expect that any such misreporting would be differential with respect to our study exposure and outcomes variables; any non-differential misreporting would be expected to attenuate the observed associations.
Conclusion
Findings from this study suggest that while sanitation and child health did not influence cefotaxime-resistant E. coli in yard soil, environmental and animal husbandry characteristics were significantly associated and child antibiotic use was borderline associated with the concentration and/or prevalence of cefotaxime-resistant E. coli in this peri-urban study setting. Results highlight the importance of taking a One Health approach to incorporate factors related to animals and the environment in global AMR mitigation strategies. Future studies should investigate the influence of specific animal husbandry practices, including animal cohabitation intensity and animal types, on cefotaxime-resistant E. coli in yard soil, other matrices in the domestic environment and among household members.
Supporting information
S1 Checklist. PLOS Global Public Health inclusivity in global research questionnaire.
https://doi.org/10.1371/journal.pgph.0006264.s001
(DOCX)
S1 Table. Socio-demographics and water, sanitation and hygiene conditions of enrolled households.
https://doi.org/10.1371/journal.pgph.0006264.s002
(DOCX)
S1 Fig. Moisture content (%) of soil, stratified by enumerator observation if the soil was wet/dry at time of collection.
The X-axis reflects whether the soil was wet or dry at time of collection, and the Y-axis reflects the average moisture content (%) of the sample. The bottommost and topmost horizontal lines represent the lower and upper limits (excluding outliers), the bottom and top borders of the box represent the first and third quartiles, and the bold horizontal line represents the median of moisture content (%).
https://doi.org/10.1371/journal.pgph.0006264.s003
(TIF)
S2 Fig. Mean log10-transformed most probable number (MPN) of cefotaxime-resistant Escherichia coli among cross-categories of animal ownership and husbandry characteristics.
Y-axis variables include animals the household or compound owns, including any animals, dogs/cats, and poultry. X-axis variables include specific animal husbandry characteristics, including if animals are enclosed in the daytime/nighttime, household administers antibiotics to animals, and animal feces observed in the compound during sampling. The heatmap gradient reflects the mean log10-transformed MPN of cefotaxime-resistant E. coli in yard soil, where darker colors represent higher MPN values. Boundaries indicate complementary variables.
https://doi.org/10.1371/journal.pgph.0006264.s004
(TIF)
S2 Table. Bivariate associations between household environmental characteristics and concentration of cefotaxime-resistant E. coli in yard soil.
Bolded values indicate associations with p-value <0.20.
https://doi.org/10.1371/journal.pgph.0006264.s005
(DOCX)
S3 Table. Adjusted associations between household environmental characteristics, animal ownership and concentration of cefotaxime-resistant E. coli in yard soil.
Models include variables that were associated with the outcome with a p-value of <0.20 in bivariate analyses and also control for household’s primary water source, education, socioeconomics, household size, and indoor floor material. Bolded values indicate associations with p-value <0.05.
https://doi.org/10.1371/journal.pgph.0006264.s006
(DOCX)
S4 Table. Adjusted associations between household environmental characteristics, poultry ownership and concentration of cefotaxime-resistant E. coli in yard soil.
Models include variables that were associated with the outcome with a p-value of <0.20 in bivariate analyses and also control for household’s primary water source, education, socioeconomics, household size, and indoor floor material. Bolded values indicate associations with p-value <0.05.
https://doi.org/10.1371/journal.pgph.0006264.s007
(DOCX)
S5 Table. Bivariate associations between household environmental characteristics and prevalence of cefotaxime-resistant E. coli in yard soil.
Bolded values indicate associations with p-value <0.05.
https://doi.org/10.1371/journal.pgph.0006264.s008
(DOCX)
S6 Table. Adjusted associations between household environmental characteristics, animal ownership and prevalence of cefotaxime-resistant E. coli in yard soil.
Models include variables that were associated with the outcome with a p-value of <0.20 in bivariate analyses and also control for household’s primary water source, education, socioeconomics, household size, and indoor floor material. Bolded values indicate associations with p-value <0.05.
https://doi.org/10.1371/journal.pgph.0006264.s009
(DOCX)
S7 Table. Adjusted associations between household environmental characteristics, poultry ownership and prevalence of cefotaxime-resistant E. coli in yard soil.
Models include variables that were associated with the outcome with a p-value of <0.20 in bivariate analyses and also control for household’s primary water source, education, socioeconomics, household size, and indoor floor material. Bolded values indicate associations with p-value <0.05.
https://doi.org/10.1371/journal.pgph.0006264.s010
(DOCX)
References
- 1. Antimicrobial Resistance Collaborators. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet. 2022;399(10325):629–55. pmid:35065702
- 2. O’Neill J. Tackling drug-resistant infections globally: final report and recommendations; 2016. Available from: https://amr-review.org/sites/default/files/160525_Final%20paper_with%20cover.pdf
- 3. GBD 2021 Antimicrobial Resistance Collaborators. Global burden of bacterial antimicrobial resistance 1990-2021: a systematic analysis with forecasts to 2050. Lancet. 2024;404(10459):1199–226. pmid:39299261
- 4. Founou RC, Founou LL, Essack SY. Clinical and economic impact of antibiotic resistance in developing countries: a systematic review and meta-analysis. PLoS One. 2017;12(12):e0189621. pmid:29267306
- 5. Koukoubani T, Makris D, Daniil Z, Paraforou T, Tsolaki V, Zakynthinos E, et al. The role of antimicrobial resistance on long-term mortality and quality of life in critically ill patients: a prospective longitudinal 2-year study. Health Qual Life Outcomes. 2021;19(1):72. pmid:33658021
- 6. Matos ECOD, Andriolo RB, Rodrigues YC, Lima PDLD, Carneiro ICDRS, Lima KVB. Mortality in patients with multidrug-resistant Pseudomonas aeruginosa infections: a meta-analysis. Rev Soc Bras Med Trop. 2018;51(4):415–20. pmid:30133622
- 7. Shrestha P, Cooper BS, Coast J, Oppong R, Do Thi Thuy N, Phodha T, et al. Enumerating the economic cost of antimicrobial resistance per antibiotic consumed to inform the evaluation of interventions affecting their use. Antimicrob Resist Infect Control. 2018;7:98. pmid:30116525
- 8. Tompkins K, Juliano JJ, van Duin D. Antimicrobial resistance in enterobacterales and its contribution to sepsis in sub-Saharan Africa. Front Med (Lausanne). 2021;8:615649. pmid:33575265
- 9. Edwards T, Williams CT, Olwala M, Andang’o P, Otieno W, Nalwa GN, et al. Molecular surveillance reveals widespread colonisation by carbapenemase and extended spectrum beta-lactamase producing organisms in neonatal units in Kenya and Nigeria. Antimicrob Resist Infect Control. 2023;12(1):14. pmid:36814315
- 10. Mhone AL, Muloi DM, Moodley A. One Health: governance and regulatory framework for antimicrobial use in Malawi. Sci One Health. 2025;4:100119. pmid:40838097
- 11. Limwado GD, Aron MB, Mpinga K, Phiri H, Chibvunde S, Banda C, et al. Prevalence of antibiotic self-medication and knowledge of antimicrobial resistance among community members in Neno District rural Malawi: a cross-sectional study. IJID Reg. 2024;13:100444. pmid:39435378
- 12. Rware H, Monica KK, Idah M, Fernadis M, Davis I, Buke W, et al. Correction: Examining antibiotic use in Kenya: farmers’ knowledge and practices in addressing antibiotic resistance. CABI Agric Biosci. 2024.
- 13.
World Health Organization. Antimicrobial resistance. World Health Organization; 2023. Available from: https://www.who.int/news-room/fact-sheets/detail/antimicrobial-resistance
- 14. Fuhrmeister ER, Harvey AP, Nadimpalli ML, Gallandat K, Ambelu A, Arnold BF, et al. Evaluating the relationship between community water and sanitation access and the global burden of antibiotic resistance: an ecological study. Lancet Microbe. 2023;4(8):e591–600. pmid:37399829
- 15. Zhou Z, Chen H. Evaluating human exposure to antibiotic resistance genes. Biosaf Health. 2024;6(2):98–100. pmid:40078948
- 16.
Centers for Disease Control and Prevention. The environment and food production – causes of antimicrobial resistance. Centers for Disease Control and Prevention; 2024. Available from: https://www.cdc.gov/antimicrobial-resistance/causes/environmental-food.html
- 17. Lienert J, Güdel K, Escher BI. Screening method for ecotoxicological hazard assessment of 42 pharmaceuticals considering human metabolism and excretory routes. Environ Sci Technol. 2007;41(12):4471–8. pmid:17626454
- 18. Peng S, Zhang H, Song D, Chen H, Lin X, Wang Y, et al. Distribution of antibiotic, heavy metals and antibiotic resistance genes in livestock and poultry feces from different scale of farms in Ningxia, China. J Hazard Mater. 2022;440:129719. pmid:35985212
- 19. Collignon P, Beggs JJ, Walsh TR, Gandra S, Laxminarayan R. Anthropological and socioeconomic factors contributing to global antimicrobial resistance: a univariate and multivariable analysis. Lancet Planet Health. 2018;2(9):e398–405. pmid:30177008
- 20.
Kalitsilo L, Morse T, Feasy N, Malenga T, Kamthunzi V, Chidziwosano K. Water, sanitation and hygiene (WASH): a critical barrier against antimicrobial resistance (AMR) in Malawi. 2021.
- 21. Mankhomwa J, Tolhurst R, M’biya E, Chikowe I, Banda P, Mussa J, et al. A qualitative study of antibiotic use practices in intensive small-scale farming in urban and peri-urban Blantyre, Malawi: implications for antimicrobial resistance. Front Vet Sci. 2022;9:876513. pmid:35685344
- 22. Blair JMA, Webber MA, Baylay AJ, Ogbolu DO, Piddock LJV. Molecular mechanisms of antibiotic resistance. Nat Rev Microbiol. 2015;13(1):42–51. pmid:25435309
- 23. Zhu Y-G, Zhao Y, Zhu D, Gillings M, Penuelas J, Ok YS, et al. Soil biota, antimicrobial resistance and planetary health. Environ Int. 2019;131:105059. pmid:31374443
- 24. Delgado-Baquerizo M, Hu H-W, Maestre FT, Guerra CA, Eisenhauer N, Eldridge DJ, et al. The global distribution and environmental drivers of the soil antibiotic resistome. Microbiome. 2022;10(1):219. pmid:36503688
- 25. Imran M, Das KR, Naik MM. Co-selection of multi-antibiotic resistance in bacterial pathogens in metal and microplastic contaminated environments: an emerging health threat. Chemosphere. 2019;215:846–57. pmid:30359954
- 26. Pagaling E, Hough R, Avery L, Robinson L, Freitag T, Coull M, et al. Antibiotic resistance patterns in soils across the Scottish landscape. Commun Earth Environ. 2023;4(1):403.
- 27. Forsberg KJ, Reyes A, Wang B, Selleck EM, Sommer MOA, Dantas G. The shared antibiotic resistome of soil bacteria and human pathogens. Science. 2012;337(6098):1107–11. pmid:22936781
- 28. Nguyen AT, Ratnasiri K, Barratt Heitmann G, Tazin S, Anderson C, Hanif S, et al. Potential pathogens and antimicrobial resistance genes in household environments: a study of soil floors and cow dung in rural Bangladesh. Appl Environ Microbiol. 2025;91(6):e0066925. pmid:40422288
- 29. Ercumen A, Hossain MS, Tabassum T, Haque A, Rahman A, Rahman MH, et al. Dirt floors and domestic animals are associated with soilborne exposure to antimicrobial-resistant E. coli in rural Bangladeshi households. Environ Sci Technol. 2025;59(47):25173–85. pmid:41221823
- 30. Montealegre MC, Roy S, Böni F, Hossain MI, Navab-Daneshmand T, Caduff L, et al. Risk factors for detection, survival, and growth of antibiotic-resistant and pathogenic Escherichia coli in household soils in rural Bangladesh. Appl Environ Microbiol. 2018;84(24):e01978-18. pmid:30315075
- 31. Hartinger SM, Medina-Pizzali ML, Salmon-Mulanovich G, Larson AJ, Pinedo-Bardales M, Verastegui H, et al. Antimicrobial resistance in humans, animals, water and household environs in rural Andean Peru: exploring dissemination pathways through the one health lens. Int J Environ Res Public Health. 2021;18(9):4604. pmid:33925280
- 32. Igbinosa EO, Beshiru A, Igbinosa IH, Cho G-S, Franz CMAP. Multidrug-resistant extended spectrum β-lactamase (ESBL)-producing Escherichia coli from farm produce and agricultural environments in Edo State, Nigeria. Aworh MK, editor. PLOS ONE. 2023;18:e0282835.
- 33. Kimera ZI, Mgaya FX, Mshana SE, Karimuribo ED, Matee MIN. Occurrence of Extended Spectrum Beta Lactamase (ESBL) producers, quinolone and carbapenem resistant enterobacteriaceae isolated from environmental samples along Msimbazi River Basin Ecosystem in Tanzania. Int J Environ Res Public Health. 2021;18(16):8264. pmid:34444027
- 34. Ercumen A, Pickering AJ, Kwong LH, Arnold BF, Parvez SM, Alam M, et al. Animal feces contribute to domestic fecal contamination: evidence from E. coli measured in water, hands, food, flies, and soil in Bangladesh. Environ Sci Technol. 2017;51:8725–34.
- 35. Wang Y, Moe CL, Null C, Raj SJ, Baker KK, Robb KA, et al. Multipathway quantitative assessment of exposure to fecal contamination for young children in low-income urban environments in Accra, Ghana: the SaniPath analytical approach. Am J Trop Med Hyg. 2017;97(4):1009–19. pmid:29031283
- 36. Bauza V, Ocharo RM, Nguyen TH, Guest JS. Soil ingestion is associated with child diarrhea in an urban slum of Nairobi, Kenya. Am J Trop Med Hyg. 2017;96(3):569–75. pmid:28093532
- 37. Shivoga WA, Moturi WN. Geophagia as a risk factor for diarrhoea. J Infect Dev Ctries. 2009;3(2):94–8. pmid:19755737
- 38. Kwong L, Ercumen A, Pickering A, Unicomb L, Davis J, Luby S. Hand- and object-mouthing of rural Bangladeshi children 3–18 months old. Int J Environ Res Public Health. 2016;13:563.
- 39. Kwong LH, Ercumen A, Pickering AJ, Unicomb L, Davis J, Leckie JO, et al. Soil ingestion among young children in rural Bangladesh. J Expo Sci Environ Epidemiol. 2021;31(1):82–93. pmid:31673039
- 40. Sammarro M, Rowlingson B, Cocker D, Chidziwisano K, Jacob ST, Kajumbula H, et al. Risk factors, temporal dependence, and seasonality of human extended-spectrum β-lactamases-producing Escherichia coli and Klebsiella pneumoniae colonization in Malawi: a longitudinal model-based approach. Clin Infect Dis. 2023;77(1):1–8. pmid:36869813
- 41. Pickering AJ, Julian TR, Marks SJ, Mattioli MC, Boehm AB, Schwab KJ, et al. Fecal contamination and diarrheal pathogens on surfaces and in soils among Tanzanian households with and without improved sanitation. Environ Sci Technol. 2012;46(11):5736–43. pmid:22545817
- 42. Ercumen A, Pickering AJ, Kwong LH, Mertens A, Arnold BF, Benjamin-Chung J, et al. Do sanitation improvements reduce fecal contamination of water, hands, food, soil, and flies? Evidence from a cluster-randomized controlled trial in rural Bangladesh. Environ Sci Technol. 2018;52:12089–97.
- 43.
World Health Organization. Technical brief on water, sanitation, hygiene (WASH) and wastewater management to prevent infections and reduce the spread of antimicrobial resistance (AMR). World Health Organization; 2020. Available from: https://www.who.int/publications/i/item/9789240006416
- 44. Peng J-J, Balasubramanian B, Ming Y-Y, Niu J-L, Yi C-M, Ma Y, et al. Identification of antimicrobial resistance genes and drug resistance analysis of Escherichia coli in the animal farm environment. J Infect Public Health. 2021;14(12):1788–95. pmid:34785168
- 45. Scicchitano D, Leuzzi D, Babbi G, Palladino G, Turroni S, Laczny CC, et al. Dispersion of antimicrobial resistant bacteria in pig farms and in the surrounding environment. Anim Microbiome. 2024;6(1):17. pmid:38555432
- 46. Shen C, He M, Zhang J, Liu J, Su J, Dai J. Effects of the coexistence of antibiotics and heavy metals on the fate of antibiotic resistance genes in chicken manure and surrounding soils. Ecotoxicol Environ Saf. 2023;263:115367. pmid:37586197
- 47. Shen D, Li C, Guo Z. Dynamics of antibiotic resistance in poultry farms via multivector analysis. Poult Sci. 2025;104(2):104673. pmid:39693773
- 48. Zhou B, Wang C, Zhao Q, Wang Y, Huo M, Wang J, et al. Prevalence and dissemination of antibiotic resistance genes and coselection of heavy metals in Chinese dairy farms. J Hazard Mater. 2016;320:10–7. pmid:27505289
- 49. Swarthout JM, Chan EMG, Garcia D, Nadimpalli ML, Pickering AJ. Human colonization with antibiotic-resistant bacteria from nonoccupational exposure to domesticated animals in low- and middle-income countries: a critical review. Environ Sci Technol. 2022;56:14875–90.
- 50. Shawver S, Ishii S, Strickland MS, Badgley B. Soil type and moisture content alter soil microbial responses to manure from cattle administered antibiotics. Environ Sci Pollut Res Int. 2024;31(18):27259–72. pmid:38507165
- 51. Ishii S, Ksoll WB, Hicks RE, Sadowsky MJ. Presence and growth of naturalized Escherichia coli in temperate soils from Lake Superior watersheds. Appl Environ Microbiol. 2006;72(1):612–21. pmid:16391098
- 52. Carmine AA, Brogden RN, Heel RC, Speight TM, Avery GS. Cefotaxime: a review of its antibacterial activity, pharmacological properties and therapeutic use. Drugs. 1983;25:223–89.
- 53.
World Health Organization. WHO integrated global surveillance on ESBL-producing E. coli using a “One Health” approach: implementation and opportunities. World Health Organization; 2021. Available from: https://www.who.int/publications/i/item/9789240021402
- 54. Niven CG, Clark B, Floess E, Chirwa B, Matekenya M, Budden E. Associations between water supply intermittencies and drinking water quality, child health, and caregiver emotional stress in peri-urban Malawi. 2026.
- 55.
JMP. Core questions on drinking water, sanitation and hygiene for household surveys: 2018 update. United Nations Children’s Fund (UNICEF) and World Health Organization; 2018.
- 56.
Blantyre Weather History. Weather API; Available from: https://www.weatherapi.com/history/q/blantyre-1598090?loc=1598090
- 57. Hornsby G, Ibitoye TD, Keelara S, Harris A. Validation of a modified IDEXX defined-substrate assay for detection of antimicrobial resistant E. coli in environmental reservoirs. Environ Sci Process Impacts. 2023;25(1):37–43. pmid:36562251
- 58.
Weldelberger J, Campbell K. Non-detect data in environmental investigations. Toronto, Canada: Los Alamos National Laboratory; 1994. Available from: https://www.osti.gov/servlets/purl/10156972
- 59.
World Health Organization (WHO), United Nations Children’s Fund (UNICEF). Progress on household drinking water, sanitation and hygiene 2000–2024: special focus on inequalities. Geneva; 2025.
- 60. Ashraf S, Islam M, Unicomb L, Rahman M, Winch PJ, Arnold BF, et al. Effect of improved water quality, sanitation, hygiene and nutrition interventions on respiratory illness in young children in rural Bangladesh: a multi-arm cluster-randomized controlled trial. Am J Trop Med Hyg. 2020;102(5):1124–30. pmid:32100681
- 61. Martel P. Review of options for reporting water, sanitation and hygiene coverage by wealth quintile. U N Child Fund UNICEF Data Anal Sect Div Data Res Policy UNICEF N Y. 2016;No. 4. https://mics.unicef.org/sites/mics/files/2024-05/MICS%20Methodological%20Paper%204.pdf
- 62. Li W, Liu C, Ho HC, Shi L, Zeng Y, Yang X, et al. Estimating the effect of increasing ambient temperature on antimicrobial resistance in China: a nationwide ecological study with the difference-in-differences approach. Sci Total Environ. 2023;882:163518. pmid:37080321
- 63. Muloi DM, Hassell JM, Wee BA, Ward MJ, Bettridge JM, Kivali V, et al. Genomic epidemiology of Escherichia coli: antimicrobial resistance through a One Health lens in sympatric humans, livestock and peri-domestic wildlife in Nairobi, Kenya. BMC Med. 2022;20(1):471. pmid:36482440
- 64. Bizimungu O, Crook P, Babane JF, Bitunguhari L. The prevalence and clinical context of antimicrobial resistance amongst medical inpatients at a referral hospital in Rwanda: a cohort study. Antimicrob Resist Infect Control. 2024;13(1):22. pmid:38389102
- 65. Cooper LN, Beauchamp AM, Ingle TA, Diaz MI, Wakene AD, Katterpalli C, et al. Socioeconomic disparities and the prevalence of antimicrobial resistance. Clin Infect Dis. 2024;79:1346–53.
- 66. Jajoo M, Manchanda V, Chaurasia S, Sankar MJ, Gautam H, Agarwal R, et al. Alarming rates of antimicrobial resistance and fungal sepsis in outborn neonates in North India. PLoS One. 2018;13(6):e0180705. pmid:29953451
- 67. Johnson J, Robinson ML, Rajput UC, Valvi C, Kinikar A, Parikh TB, et al. High burden of bloodstream infections associated with antimicrobial resistance and mortality in the neonatal intensive care unit in Pune, India. Clin Infect Dis. 2021;73(2):271–80. pmid:32421763
- 68. Cocker D, Chidziwisano K, Mphasa M, Mwapasa T, Lewis JM, Rowlingson B, et al. Investigating One Health risks for human colonisation with extended spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae in Malawian households: a longitudinal cohort study. Lancet Microbe. 2023;4(7):e534–43. pmid:37207684
- 69.
World Health Organization. Malawi national action plan on antimicrobial resistance: review of progress in the human health sector. Geneva: World Health Organization; 2022.
- 70. Mwapasa T, Chidziwisano K, Mphasa M, Cocker D, Rimella L, Amos S, et al. Key environmental exposure pathways to antimicrobial resistant bacteria in southern Malawi: a SaniPath approach. Sci Total Environ. 2024;945:174142. pmid:38906299
- 71. Abraham A, Mtewa AG, Chiutula C, Mvula RLS, Maluwa A, Eregno FE, et al. Prevalence of antibiotic resistance bacteria in manure, soil, and vegetables in Urban Blantyre, Malawi, from a farm-to-fork perspective. Int J Environ Res Public Health. 2025;22(8):1273. pmid:40869858
- 72. Barnes AN, Mumma J, Cumming O. Role, ownership and presence of domestic animals in peri-urban households of Kisumu, Kenya. Zoonoses Public Health. 2018;65(1):202–14. pmid:29205887
- 73. Passarelli S, Ambikapathi R, Gunaratna NS, Madzorera I, Canavan CR, Noor RA, et al. The role of chicken management practices in children’s exposure to environmental contamination: a mixed-methods analysis. BMC Public Health. 2021;21(1):1097. pmid:34103022
- 74. Ercumen A, Pickering AJ, Kwong LH, Arnold BF, Parvez SM, Alam M, et al. Animal feces contribute to domestic fecal contamination: evidence from E. coli measured in water, hands, food, flies, and soil in Bangladesh. Environ Sci Technol. 2017;51:8725–34.
- 75. Kurowski KM, Marusinec R, Amato HK, Saraiva-Garcia C, Loayza F, Salinas L, et al. Social and environmental determinants of community-acquired antimicrobial-resistant Escherichia coli in children living in semirural communities of Quito, Ecuador. Am J Trop Med Hyg. 2021;105(3):600–10. pmid:34280150
- 76.
World Food Program. World Food Programme Malawi. Malawi [Internet]; 2026. Available from: https://www.wfp.org/countries/malawi
- 77. Kaimila Y, Divala O, Agapova SE, Stephenson KB, Thakwalakwa C, Trehan I, et al. Consumption of animal-source protein is associated with improved height-for-age z scores in rural Malawian children aged 12-36 months. Nutrients. 2019;11(2):480. pmid:30823563
- 78. Zemene MA, Kebede N, Anteneh RM, Moges N, Tsega SS, Dessie AM, et al. Determinants of animal source food consumption among children aged 6-23 months in sub-Saharan Africa: multilevel mixed effect model. Sci Rep. 2024;14(1):26294. pmid:39487171
- 79. Van Boeckel TP, Pires J, Silvester R, Zhao C, Song J, Criscuolo NG, et al. Global trends in antimicrobial resistance in animals in low- and middle-income countries. Science. 2019;365(6459):eaaw1944. pmid:31604207
- 80. Brogdon JM, Sié A, Dah C, Ouermi L, Coulibaly B, Lebas E, et al. Poultry ownership and genetic antibiotic resistance determinants in the gut of preschool children. Am J Trop Med Hyg. 2021;104(5):1768–70. pmid:33755581
- 81. Thomas ED, Kwong LH, Rimi NA, Shanta IS, Uddin MR, Khan S, et al. Who has sheds? Exploring practices and determinants of overnight housing for backyard poultry in rural Bangladesh to inform an intervention to limit exposure to poultry and poultry feces. PLOS Glob Public Health. 2025;5(7):e0004929. pmid:40694593
- 82. Harvey SA, Winch PJ, Leontsini E, Torres Gayoso C, López Romero S, Gilman RH, et al. Domestic poultry-raising practices in a Peruvian shantytown: implications for control of Campylobacter jejuni-associated diarrhea. Acta Trop. 2003;86(1):41–54. pmid:12711102
- 83. Oberhelman RA, Gilman RH, Sheen P, Cordova J, Zimic M, Cabrera L, et al. An intervention-control study of corralling of free-ranging chickens to control Campylobacter infections among children in a Peruvian periurban shantytown. Am J Trop Med Hyg. 2006;74(6):1054–9. pmid:16760519
- 84. Berglund B. Environmental dissemination of antibiotic resistance genes and correlation to anthropogenic contamination with antibiotics. Infect Ecol Epidemiol. 2015;5:28564. pmid:26356096
- 85. Sclar GD, Garn JV, Penakalapati G, Alexander KT, Krauss J, Freeman MC, et al. Effects of sanitation on cognitive development and school absence: a systematic review. Int J Hyg Environ Health. 2017;220(6):917–27. pmid:28697975
- 86. Contreras JD, Islam M, Mertens A, Pickering AJ, Kwong LH, Arnold BF, et al. Influence of community-level sanitation coverage and population density on environmental fecal contamination and child health in a longitudinal cohort in rural Bangladesh. Int J Hyg Environ Health. 2022;245:114031. pmid:36058111
- 87. Hitchings GH. Mechanism of action of trimethoprim-sulfamethoxazole. J Infect Dis. 1973;128:S433–6.
- 88.
Akhavan BJ, Khanna NR, Vijhani P. Amoxicillin. In: StatPearls. Treasure Island (FL): StatPearls Publishing; 2025. Available from: http://www.ncbi.nlm.nih.gov/books/NBK482250/
- 89. Barathe P, Kaur K, Reddy S, Shriram V, Kumar V. Antibiotic pollution and associated antimicrobial resistance in the environment. J Hazard Mater Lett. 2024;5:100105.
- 90. Khuluza F. In-vitro evaluation of the quality of paracetamol and co-trimoxazole tablets used in Malawi based on pharmacopoeial standards. Malawi Med J. 2014;26(2):38–41. pmid:25157315
- 91. Khuluza F, Kigera S, Heide L. Low prevalence of substandard and falsified antimalarial and antibiotic medicines in public and faith-based health facilities of Southern Malawi. Am J Trop Med Hyg. 2017;96(5):1124–35. pmid:28219993
- 92. Rogawski ET, Platts-Mills JA, Seidman JC, John S, Mahfuz M, Ulak M, et al. Use of antibiotics in children younger than two years in eight countries: a prospective cohort study. Bull World Health Organ. 2017;95(1):49–61. pmid:28053364
- 93. Ercumen A, Mertens AN, Butzin-Dozier Z, Jung DK, Ali S, Achando BS, et al. Water, sanitation, handwashing, and nutritional interventions can reduce child antibiotic use: evidence from Bangladesh and Kenya. Nat Commun. 2025;16(1):556. pmid:39788996
- 94. Ishii S, Yan T, Vu H, Hansen DL, Hicks RE, Sadowsky MJ. Factors controlling long-term survival and growth of naturalized Escherichia coli populations in temperate field soils. Microbes Environ. 2010;25(1):8–14. pmid:21576846
- 95. Fahimipour AK, Hartmann EM, Siemens A, Kline J, Levin DA, Wilson H, et al. Daylight exposure modulates bacterial communities associated with household dust. Microbiome. 2018;6(1):175. pmid:30333051
- 96. Chopra I, Roberts M. Tetracycline antibiotics: mode of action, applications, molecular biology, and epidemiology of bacterial resistance. Microbiol Mol Biol Rev. 2001;65(2):232-60; second page, table of contents. pmid:11381101
- 97. Khan SJ, Osborn AM, Eswara PJ. Effect of sunlight on the efficacy of commercial antibiotics used in agriculture. Front Microbiol. 2021;12:645175. pmid:34140934
- 98. Kou R, Li X, Wang H, Wu Y, Zou D, Hu J, et al. Panel data analysis of the effect of ambient temperature on antimicrobial resistance in China. Sci Rep. 2025;15(1):18391. pmid:40419687
- 99. Lin D, Du S, Zhao Z, Zhang T, Wang L, Zhang Q, et al. Climate warming fuels the global antibiotic resistome by altering soil bacterial traits. Nat Ecol Evol. 2025;9(8):1512–26. pmid:40468041
- 100.
Padda IS, Nagalli S. Cefotaxime. In: StatPearls. Treasure Island (FL): StatPearls Publishing; 2025. Available from: http://www.ncbi.nlm.nih.gov/books/NBK560653/
- 101. Shaikh S, Fatima J, Shakil S, Rizvi SMD, Kamal MA. Antibiotic resistance and extended spectrum beta-lactamases: types, epidemiology and treatment. Saudi J Biol Sci. 2015;22(1):90–101. pmid:25561890