Antimicrobial resistant E. coli shedding and farm stress
Stress events before slaughter did not significantly change antimicrobial resistant E. coli shedding in pigs and poultry in a six-farm UK pilot study. Shedding stayed statistically flat across thinning, depopulation and transport compared with baseline. The more important finding is that shedding was already high at every time point, so the contamination risk window is constant rather than event triggered, which matters for antibiotic reduction programs.
Why shedding at the slaughter threshold matters
Antimicrobial resistance is a continuing threat to both public and animal health. In the United Kingdom, livestock antibiotic use has fallen substantially over the past decade, yet resistant bacteria still enter farms through feed, water, people and the environment. Common management events such as thinning, transport and depopulation are known to stress animals, and stress can raise bacterial shedding through immune changes. These stress events in livestock are routine, but their effect on shedding was exactly the question this pilot set out to test. That raises a practical question for anyone running an antibiotic reduction program: do the moments just before slaughter push resistant bacteria out in larger numbers, and is that where control should be focused.
The background numbers show why this matters. National veterinary surveillance data from 2024 put the pig sector at the highest antibiotic use among UK livestock at 50 mg per kg, while broilers used 4.7 mg per kg. Fully susceptible Escherichia coli (E. coli) isolated from broilers dropped from 45 percent in 2022 to 32 percent in 2024, and E. coli with limited treatment options appeared in 22 percent of pig isolates in 2024. Shedding of resistant organisms at the processing stage is a direct route into the food chain, so understanding when it spikes is the first step to controlling it.
What the study measured
Researchers ran a longitudinal observational pilot across six commercial farms: three broiler houses and three pig finisher units. Each farm was sampled at several points in its production cycle so that shedding could be compared before and after routine stress events.
| Time point | Broilers | Pigs |
|---|---|---|
| Baseline (at least one week before first event) | yes | yes |
| Post-thinning | yes | no |
| Post-depopulation (loaded and sent to abattoir) | yes | yes |
| Post-transport (the lorry that carried them) | yes | yes |
At baseline, broilers were 20 to 30 days old and pigs were 18 to 30 weeks old. Sampling ran from September 2022 to February 2024. On each visit, six pooled faecal samples were collected from the floor of the house or, for transport, from the lorry, giving a 95 percent confidence interval and a 5 percent detection limit for E. coli. One post-transport time point was excluded after a logistical delay left samples unrefrigerated for 48 hours.
In the laboratory, pooled faeces were suspended in phosphate-buffered saline and serially diluted tenfold. Each dilution was spread plated on CHROMagar ECC without antibiotic and on CHROMagar ECC containing selected antibiotics, then incubated at 37 degrees Celsius for 24 hours. Suspect colonies were confirmed on MacConkey agar and 5 percent sheep blood agar using indole and oxidase tests, and the result was expressed as log colony-forming units per gram of faeces. The antibiotic concentrations used to screen for resistance were chosen from baseline susceptibility and aligned with EUCAST clinical breakpoints for Enterobacterales.
| Antibiotic | Poultry screen (mg/L) | Pig screen (mg/L) | Class note |
|---|---|---|---|
| Ampicillin | 8 | 8 | veterinary importance |
| Nalidixic acid | 32 | not used | |
| Cefotaxime | 1 | 1 (farms 5 and 6) | critically important |
| Ciprofloxacin | 0.5 | 0.5 (farms 5 and 6) | critically important |
| Gentamicin | 2 | not used | |
| Tetracycline | not used | 8 | |
| Sulfamethoxazole | not used | 32 (farm 4) | |
| Apramycin | not used | 32 (farms 5 and 6) |
Stress events and antimicrobial resistant E. coli shedding: what the data showed
Shedding varied a great deal between farms and time points, but the statistical test found no significant difference in E. coli counts after stress events compared with baseline, for both pigs and poultry and for both total and resistant E. coli. The analysis used Poisson generalised linear mixed models with the farm treated as a random effect. For pig farms, the estimated incidence rate ratio was above 1 at every time point, but the 95 percent confidence interval always included values below 1, so the result is not significant.
The raw levels were high throughout. Poultry showed the widest spread, from 4.0 to 8.1 log cfu/g, while pigs ranged from 4.1 to 6.4 log cfu/g. The single highest broiler value was 8.1 log cfu/g for both total and ampicillin resistant E. coli at the Farm 2 baseline. The highest pig value was 6.4 log cfu/g for total E. coli at the Farm 6 post-depopulation sample.
| Metric | Broilers | Pigs |
|---|---|---|
| Log cfu/g range across study | 4.0 to 8.1 | 4.1 to 6.4 |
| Highest single value | 8.1, total and ampicillin resistant, Farm 2 baseline | 6.4, total E. coli, Farm 6 post-depopulation |
| High-density shedder threshold (over 4.00 log cfu/g) | all farms, all time points | all farms, all time points |
| Dominant resistance profile | ampicillin more than nalidixic acid | tetracycline more than ampicillin (farms 4 and 5); ampicillin dominant (farm 6) |
Among broilers, ampicillin resistant E. coli was more common than nalidixic acid resistant E. coli at every time point. One farm showed an instructive wobble: nalidixic acid resistant E. coli fell 24.3 percent from post-thinning to post-depopulation, then rose 17.5 percent by the post-transport sample taken the same day. That pattern points to intermittent shedding rather than a clean stress response.
The counterintuitive result: the risk window is constant
The headline that stress did not move shedding is interesting, but the more useful insight sits one level down. Every farm in the study sat above the 4.00 log cfu/g line that earlier work used to define a high-density shedder, at every single time point including baseline. In other words, the animals were already shedding resistant E. coli heavily before any stress event happened.
That reframes the whole control problem. If shedding is high and roughly stable, then the contamination risk at processing is not created by transport or depopulation. It is present from the start and merely carried into the abattoir. For an antibiotic reduction program, this means the defensive line cannot be drawn only around the few days before slaughter. The gut load of resistant E. coli has to be managed continuously, because the window of risk is the whole production cycle, not the stressful finish.
It also shifts where hygiene investment pays off. Abattoir cross-contamination control remains essential as the last barrier, but it is a backstop. Lowering what arrives at the gate, week after week, is what actually shrinks the problem upstream.
Temporal and farm factors that actually moved shedding
Although the stress events did not register, the models did find factors that changed shedding in broilers. In univariate analysis, sampling in May and in spring were both linked to lower expected colony counts, while larger flock size was linked to higher counts. All three effects were significant, but only for nalidixic acid resistant E. coli, and none applied to pigs.
| Variable (broiler, nalidixic resistant E. coli) | Incidence rate ratio | 95 percent CI | p value | Direction |
|---|---|---|---|---|
| Sampling in May | 0.78 | 0.59 to 0.97 | 0.013 | lower |
| Sampling in spring | 0.75 | 0.56 to 0.94 | 0.003 | lower |
| Flock size 45,500 birds | 1.27 | 1.08 to 1.46 | 0.012 | higher |
| Flock size 95,500 birds | 1.33 | 1.14 to 1.52 | 0.003 | higher |
The researchers caution that these signals are probably driven by a single farm. The one broiler unit sampled in May and spring also had the smallest flock, at 36,500 birds, and it showed the lowest mean nalidixic acid resistant shedding of the broiler group. With only three broiler farms, a single outlier can create an apparent seasonal or size effect. Pigs showed no significant explanatory variable at all.
What this means for antibiotic reduction programs
For producers and the companies that supply them, the practical reading is straightforward. An antibiotic reduction strategy that treats the pre-slaughter period as the main risk window is aiming at the wrong target. Shedding of antimicrobial resistant E. coli is already elevated well before any handling stress, so the lever is a continuous gut health and biosecurity program rather than a last-minute intervention.
Three implications stand out. First, interventions such as acidifiers, competitive exclusion and improved litter and air quality should be judged on whether they lower the standing gut load across the cycle, not on a single post-event measurement. Second, because resistance to critically important antibiotics such as ciprofloxacin and cefotaxime was detected on some farms, any alternative approach must be validated against resistance outcomes, not only against production metrics. Third, larger flocks carried more shedding in this dataset, so higher-density systems need proportionally stronger management rather than weaker.
None of this argues against reducing antibiotics. It argues for doing it with eyes open: the resistant organism is already there, so the goal of reduction is to keep the load low and the food chain clean without letting stress or gaps in management tip it upward.
Practical steps to keep shedding low
- Run a continuous gut health program instead of a pre-slaughter push, because baseline shedding is already high.
- Treat abattoir hygiene as the non-negotiable backstop, since contaminated lorries and surfaces spread the organism between batches.
- Monitor the resistant profile on your own farm, especially for critically important antibiotics, rather than assuming a clean status.
- Watch flock or herd size as a risk multiplier, and tighten management as density rises.
- Close the feed withdrawal knowledge gap: it was not measured here but is known to raise shedding in some species, so standardise and record it.
- Repeat sampling across the cycle, not just at one point, because shedding is intermittent and a single sample can mislead.
- Treat pre-slaughter stress as one risk among many, not the main source, because the load is already there before any handling begins.
- Watch for high-density shedders by sampling across the cycle, not only at the finish, so you catch the standing load rather than a single spike.
- Protect gut health in livestock continuously, because that is what keeps the standing load down and limits food chain contamination.
Limitations of the pilot
This was a small pilot with only six farms, so statistical power was limited and the results are preliminary. The broiler farms came from one integrated company, so how representative they are of the wider sector is unknown. The baseline was assumed to be a low-stress state, but other factors could have been stressing the animals all along. One post-transport sample was excluded after a handling delay. These constraints mean the “no significant effect of stress” result should be read as “not detected here”, not as proof that stress never matters.
Bottom line
A six-farm UK pilot found no significant change in antimicrobial resistant E. coli shedding after thinning, depopulation or transport compared with baseline. The real signal is that shedding was already high at every time point, so the risk window is constant and antibiotic reduction programs must manage gut load continuously rather than only around slaughter.
FAQ
Does stress increase E. coli shedding in pigs
A pilot study of three UK pig finisher farms found no significant increase in antimicrobial resistant E. coli shedding after depopulation or transport compared with baseline. The estimated incidence rate ratio was above 1 at each point, but every confidence interval included values below 1. Shedding levels were already high before any stress event, so the study indicates stress was not the main driver in this small sample.
Does stress increase E. coli shedding in poultry
The broiler part of this pilot also showed no significant change in antimicrobial resistant E. coli shedding after thinning, depopulation or transport versus baseline. Broiler shedding varied more than pig shedding, from 4.0 to 8.1 log cfu/g, but that variation was not tied to the stress events. As with pigs, the levels were high from the first baseline sample onward, which suggests the risk is present throughout the cycle.
What are high-density E. coli shedders
A high-density shedder is an animal or flock whose faecal E. coli load exceeds 4.00 log colony-forming units per gram of faeces. Earlier work used that threshold to sort farms into categories for contamination risk. In the pilot discussed here, every farm sat above that line at every sampling time point, including baseline, which is why abattoir hygiene matters so much for keeping the food chain clean.
How is E. coli shedding measured
Researchers collect pooled faecal samples from the house floor or transport lorry and suspend them in buffer with tenfold dilutions. Each dilution is spread plated on CHROMagar ECC with and without antibiotics, incubated at 37 degrees Celsius, and confirmed with MacConkey agar and indole and oxidase tests. Results are reported as log colony-forming units per gram, which lets farms compare loads across time points.
Why is E. coli shedding a food safety risk
E. coli shed in faeces can contaminate the processing environment, transport lorries and carcasses during slaughter. When the organism is resistant to antibiotics, especially critically important ones such as ciprofloxacin and cefotaxime, treatment options narrow for both animals and people. Because shedding was high at every time point in the pilot, the contamination risk exists before any stress event occurs.
Can antibiotic reduction increase shedding
Cutting antibiotic use does not by itself raise shedding, but it removes one pressure that can suppress bacterial load. The pilot did not test reduction programs directly and did not collect on-farm usage data, so it cannot link lower use to higher shedding. The safer conclusion is that reduction must be paired with continuous gut health and hygiene so resistant organisms are not given room to climb.
What reduces E. coli shedding on farm
The evidence here points to continuous management rather than a single intervention. Useful steps include a steady gut health program, strict abattoir and lorry hygiene, routine resistance profiling, and tighter control as flock or herd size rises. Feed withdrawal was not measured but is known to raise shedding in some species, so standardising it is worth adding to any plan.
Does flock size affect E. coli shedding
In the broiler analysis, larger flocks were linked to higher expected shedding of nalidixic acid resistant E. coli. Flocks of 45,500 and 95,500 birds showed incidence rate ratios of 1.27 and 1.33 respectively, both statistically significant. The effect appeared only for that resistance type and only in broilers, and the researchers note it may reflect a single small-flock farm rather than a true size law.
Is E. coli shedding higher in poultry or pigs
Poultry shed more E. coli than pigs in this study, with broiler loads ranging 4.0 to 8.1 log cfu/g against 4.1 to 6.4 log cfu/g for pigs. The highest single value overall was a broiler at 8.1 log cfu/g. Both species exceeded the high-density threshold at every time point, so the absolute risk was high in each, but the poultry range sat consistently higher.
What are stress events in pig and poultry production
Stress events are routine management moments that disturb the animal and can weaken defenses. In pig and poultry cycles they include thinning, where part of a flock is removed, depopulation, when animals are loaded and sent to the abattoir, and transport, when the lorry itself is sampled. These were the events the pilot used to test whether shedding of resistant E. coli changed afterward.
Does transport to abattoir increase shedding
Transport was one of the stress events tested, and it did not produce a statistically significant rise in E. coli shedding in either species. In broilers, one farm even showed nalidixic acid resistant E. coli rebound by 17.5 percent from depopulation to transport on the same day, but that reflects intermittent shedding more than a transport effect. The study as a whole found no clear transport-driven spike.
How common is resistance to critically important antibiotics
The pilot detected resistance to ciprofloxacin or cefotaxime, both classified as critically important, on some pig and broiler farms. Broader national surveillance adds context: fully susceptible E. coli from broilers fell from 45 percent in 2022 to 32 percent in 2024, and limited-treatment-option E. coli appeared in 22 percent of pig isolates in 2024. These figures show why any reduction plan must track resistance outcomes.
