Productivity parameters, antimicrobial consumption, and prevalence of enteric pathogens before and after intramuscular vaccination against Lawsonia intracellularis in naturally infected Danish weaner and finisher pig herds.

In Danish pig production, gastro-intestinal diseases account for most of the antimicrobials (AM) used in growing pigs. Diarrhoea is most frequently caused by Lawsonia intracellularis (LI), Brachyspira pilosicoli (BP), E coli fimbria type F4 (F4) and E. coli fimbria type F18 (F18). With a new LI vaccine available from 2019, it was relevant to investigate the effect of this vaccine in a Danish field study including both weaner and finisher sites. The aim was to evaluate the efficacy of Porcilis® Lawsonia Vet. in naturally LI-infected pig herds by comparing of productivity parameters, AM consumption and dynamics of enteric pathogens over two 6-months periods before and after LI vaccination. Further, faecal sock samples were collected from each site before and after vaccination and analysed by qPCR for excretion levels of LI, BP, F18 and F4. In total, 28 weaner and 41 finisher sites were included in the study. Vaccination reduced Feed Conversion Ratio by 0.12 Feed Unit/kg (p = 0.029) and 0.08 Feed Unit/kg (p = 0.005) in weaners and finishers, respectively. Increased Average Daily Weight Gain of 45.6 gr./day (p < 0.001) was found in the finishers. Mortality risk fell by 8.8% in weaners (RR = 0.912; p < 0.001). AM prescriptions for oral group treatments were reduced by 38.8% active compound/kg pig produced (p = 0.005) or 33.3% Weighted Animal Daily Doses per 100 animals per day in finishers (p = 0.004). LI prevalence was reduced in weaners and finishers (both p < 0.001) and BP prevalence was reduced in finishers (p = 0.043). Mean excretion levels of LI and BP decreased at weaner sites (-1.32 and -1.02 log(10) copies/gr faeces, respectively; both p < 0.001) and at finisher sites (-1.04 and -1.16 log(10) copies/gr faeces, respectively; both p < 0.001). Prevalence and excretion levels of F18 and F4 were unaffected by LI vaccination. In conclusion, vaccination against LI using Porcilis® Lawsonia Vet. improved productivity parameters, cut AM consumption, and reduced prevalence and excretion levels of LI and BP in naturally LI-infected pig herds.

Effect of intramuscular vaccination against Lawsonia intracellularis on production parameters, diarrhea occurrence, antimicrobial treatment, bacterial shedding, and lean meat percentage in two Danish naturally infected finisher pig herds.

Lawsonia intracellularis (LI) is an economically important enteric pathogen in pigs with a worldwide endemic prevalence. The objective of this study was to evaluate the effect of an intramuscularly administrated LI vaccine (Porcilis®Lawsonia Vet.) in Danish finisher pigs (30-115 kg) measured on key production figures, antimicrobial (AB) treatments, occurrence of diarrhea and LI shedding. The study was a group-randomized block-trial with parallel groups in two herds, Herd 1 and Herd 2, experiencing a natural subclinical-clinical LI infection in early finisher period. Vaccination occurred at weaning, but the study focused on the first eight weeks in the finisher period. Further, slaughterhouse data were included. In total, 52 and 50 finisher pens comprising 2184 and 2254 finisher pigs were included in each of two herds, respectively. LI vaccination significantly reduced feed conversion ratio (FCR) by 0.05 and 0.09 FU/kg (p = 0.007 and p < 0.001) alongside a significantly increased average daily weight gain (ADWG) by 31 and 43 gr/day (p = 0.001 and p < 0.001) in each of the herds, respectively. In the vaccinated group, less variation was found in ADWG compared to the control group (p < 0.001 in both herds) as an expression of a more uniform growth, which was further confirmed by less variation in lean meat percent in the vaccinated group in one herd (p = 0.007). No significant difference between groups were found in mortality and pigs excluded due to welfare reasons. AB flock treatment against diarrhea was significantly reduced in Herd 1 with all pens treated in the control group compared to 30.8 % in the vaccinated group (p < 0.001). In Herd 2, the difference was non-significant with 68.0 % in the control group compared to 50.0 % in the vaccination group (p = 0.252). Low levels of individual treatments against diarrhea were seen in both herds (≤ 5.0 %) but still significantly reduced in vaccinated pigs compared to control pigs (p < 0.050 in both herds). Mean diarrheic blot counts were significantly reduced in vaccinated pens compared to control pens (p < 0.001 in both herds). In vaccinated pigs, shedding of LI was reduced in both prevalence (p < 0.001 in both herds), excretion level in positive samples (p < 0.001 in both herds) and, in one herd, also in duration (p = 0.003) when compared to control pigs. In conclusion, pigs vaccinated with Porcilis®Lawsonia Vet against LI in both of two high-health and high-productive finisher herds had, compared to non-vaccinated pigs, significantly improved key production figures, and reduced AB treatment, occurrence of diarrhea, LI shedding, and growth variation.

Efficacy of neomycin dosing regimens for treating enterotoxigenic Escherichia coli-related post-weaning diarrhoea in a Danish nursery pig herd not using medicinal zinc oxide.

Neomycin is a concentration-dependant aminoglycoside antimicrobial used to treat enterotoxigenic Escherichia coli (ETEC)-related post-weaning diarrhoea (PWD) in pigs. The objective was to compare the efficacy of neomycin administered in a single high dose (50,000 IU/kg) and a standard dose and frequency (25,000 IU/kg daily for 3 consecutive days) in reducing the number of pigs with clinical PWD. We also aimed to evaluate the development of antimicrobial resistance in E. coli following neomycin treatment. The study was performed in a Danish herd not using medicinal zinc oxide and experiencing outbreaks of PWD caused by ETEC in the first week after weaning. Pigs from six batches with perianal faecal staining on days 4-6 after weaning and a faecal score of 3-4 were ear tagged and treated with neomycin. Pens were randomly assigned to a treatment group before inclusion. A total of 772 pigs (471 in the control group and 301 in the experimental group) were included and treated orally. The apparent prevalence of diarrhoea on the first day of inclusion across six batches (n = 1,875) was 27%. The efficacy of the neomycin treatment strategy was 86% for the control group and 91% for the single high-dose group (p = 0.043), and the mean percentage (standard deviation (sd)) of haemolytic E. coli-like colonies was 12% (26) and 26% (37) (p < 0.001), respectively. Neomycin resistance did not differ between groups. Before treatment, all analysed isolates were identified as ETEC (n = 142), while after treatment, 91% were identified as ETEC (n = 69) and 9% (n = 7) as non-ETEC E. coli (without fimbria or toxins). A higher cure rate in the single high-dose group suggests that ETEC-related PWD can be treated with a single high dose of 50,000 IU/kg of neomycin, thereby reducing antimicrobial use by 33% compared to the standard treatment of 25,000 IU/kg for 3 consecutive days. The study indicated a higher number of haemolytic E. coli in the single high-dose group after treatment, but no evidence of increased neomycin resistance in coliforms was observed compared to the standard treatment.

Prevalences and excretion levels of Lawsonia intracellularis, Brachyspira pilosicoli and Escherichia coli F4 and F18 in fecal sock samples from Danish weaner and finisher pig batches and the association with diarrhea.

BACKGROUND: Bacterial enteritis in growing pigs is a matter of concern in Danish pig production challenging herd health as well as production economy, and antimicrobial usage. The aim of this observational study using fecal sock samples was to determine the prevalence and excretion level of Lawsonia intracellularis (LI), Brachyspira pilosicoli (BP), Escherichia coli F4 (F4) and F18 (F18) and to investigate associations between prevalence or excretion levels of the bacteria and diarrhea. The study was performed in the late weaner and the early finisher period in herds with a history of diarrhea. Every weaner and finisher herd contributed with one sample each. RESULTS: In total, 47 weaner and 59 finisher herds were sampled. The overall prevalence and excretion levels (median and range in log(10) copies/gram of feces) were for LI 84.0% (median 6.2; range 3.0-7.7), for BP 45.2% (median 5.6; range 3.0-6.6), for F18 20.8% (median 5.7; range 4.7-7.7), and for F4 4.7% (median 5.5; range 5.2-6.0). In both diarrheic and non-diarrheic samples, the most prevalent bacteria were either LI alone or LI and BP in combination. In general, no association was found between increasing total bacterial excretion levels and diarrhea, but prevalence (p = 0.04) and excretion (p < 0.01) level of F18 was found to be significantly higher in diarrheic samples. Further, a significant association was found between low LI excretion level and lack of diarrhea in weaner herds (p = 0.03). A significant positive correlation was found between excretion levels of LI and BP in diarrheic weaner herd samples (p = 0.02). CONCLUSION: Enteric pathogens were prevalent in a wide range of bacterial excretion levels in both diarrheic and non-diarrheic samples. Especially LI and BP were frequently found and with a positive correlation between excretion levels. Even in the absence of diarrhea, high prevalence and excretion levels of LI and BP were detected, thus making the status of diarrhea an insufficient tool for assessing the severity of their infections.

Post-weaning diarrhea in pigs weaned without medicinal zinc: risk factors, pathogen dynamics, and association to growth rate.

BACKGROUND: Porcine post-weaning diarrhea (PWD) has reemerged as an important topic in pig production, as common control strategies based on prophylactic use of antimicrobials and zinc oxide have been deemed unsustainable. The objectives of this study were to estimate the cumulative incidence of porcine post-weaning diarrhea with different etiologies in production systems weaning without zinc oxide and prophylactic antimicrobials, to assess risk factors for post-weaning diarrhea, and to estimate the impact of post-weaning diarrhea on growth rate. A cohort study was conducted at two commercial indoor producers weaning without medicinal zinc oxide and prophylactic antimicrobials. RESULTS: Piglets were included at birth (n = 300) and 272 survived until weaning. After insertion to the nursery units, the piglets were clinically examined every day for 14 days, and rectal swabs were collected and analyzed for enterotoxigenic Escherichia coli (ETEC) and rotavirus A. The cumulative incidences of PWD the first 14 days after insertion to the nursery units were 41.8% (CI 33.6, 50.4) and 51.1% (CI 42.3, 60.0) at the two producers, respectively. We found a low incidence of cases associated to ETEC, and detected a substantial proportion of cases associated to rotavirus. We observed a biphasic pattern in the assumed etiology with rotavirus occurring first, and then a shift towards cases associated to ETEC/non-ETEC hemolytic E. coli. Being offspring of older sows was a protective factor for the development of PWD (Hazard ratio = 0.88 [CI 0.78, 0.99] per unit increase in parity of the dam). Low birth weight reduced the post-weaning growth rate (- 5.2 g/day [CI - 7.5, - 2.9] per 100 g decrease in birthweight) and increased the hazard of developing PWD (Hazard ratio for birthweight below 1100 g: 2.30 [CI 1.41-3.74]). The combined effect of having diarrhea for 2 days or more and receiving antimicrobial treatment was associated with an increased average daily weight gain. CONCLUSIONS: This study suggests novel insights regarding pathogen dynamics and risk factors for PWD in productions not using prophylactic antimicrobials and medicinal zinc. The findings may have important implications for both antimicrobial usage and prevention strategies.

A Very Low CP Level Reduced Diarrhoea and Productivity in Weaner Pigs, but No Differences between Post-Weaning Diets Including Soybean Meal or Soy Protein Concentrate Were Found.

Soy protein concentrate improves nutrient utilization and growth performance compared to soybean meal, and diets with a low crude protein (CP) level decreases diarrhoea. The objectives were to (1) test a low CP diet based on different soy products, and (2) to test a very-low CP diet (15.1%) with amino acids (AA) on diarrhoea and productivity. A total of 5,635 weaned pigs (~28 days), were assigned to five dietary treatments; PC (positive control): Standard CP levels (192, 189, 191 g/kg CP) with 2500 ppm ZnO; NC (negative control): Same as PC without ZnO; SP (Soy protein concentrate): Low CP levels (176, 174, 191 g/kg CP); SB (Soybean meal): Low CP levels (177, 176, 191 g/kg CP); and XLA (X-low CP + AA): Very low CP levels (154, 151, 191 g/kg CP) with AA. The PC and XLA diets reduced diarrhoea by 41 and 61%, respectively, compared to the NC group, while no difference between SB and SP were observed. The XLA diet reduced feed intake and daily gain compared with PC and NC, where SP, SB, and XLA had a poorer feed conversion compared with PC. Conclusively, the SP and SB low-protein diets did not reduce diarrhoea or growth performance, whereas the XLA diet decreased both diarrhoea and performance.

Risk factors for gastric ulceration in nursery pigs.

Pars oesophageal gastric ulceration is a prevalent condition in swine production and often observed in finisher pigs at slaughter, with reported prevalence ranging from 32 % to 65 %. Although feeding practices in nursery and finisher pigs are similar, little is known about the prevalence and risk factors associated with this condition in nursery pigs. This prospective cohort study aimed to identify risk factors for gastric ulceration in nursery pigs. The objectives were: (1) to estimate the association of inherent pig characteristics (birth weight, sex, parity of sow, litter size, and pen fouling behavior) and management characteristics (antibiotic treatment during suckling and post weaning, piglet rotation during lactation, weaning age and weight) with pars oesophageal lesions; (2) to assess differences in gastric dry matter content, pH, and percentage of solid particles in the sediment in relation to pars oesophageal lesions. A total of 58 piglets tracked from birth to 20 days of age and 210 nursery pigs from birth to 10 weeks of age were euthanized and assessed for pars oesophageal gastric lesions. All nursery pigs were fed finely ground pelleted feed ad libitum. Based on a macroscopic assessment, the pigs' stomachs were categorized as either LPO "lesions in the Pars oesophagea" or NLPO "no lesions in the Pars oesophagea". We observed an overall prevalence of 26.2 % (CI 95 % 0.67 - 0.79) for LPO in nursery pigs at 10 weeks of age, while no lesion were observed in piglets at 20 days of age. A mixed multivariable logistic regression model showed that piglets with a birthweight ≤ 1.11 kg had a 2.11 times higher odds of LPO (CI 95 % 1.02-4.37, p = 0.04) than their heavier counterparts. There was a weak association of LPO in nursery pigs from pens with pen fouling with a 2.14 times higher odds (CI 95 % 0.97-4.72 p = 0.05). Nursery pigs with LPO presented an increased gastric content fluidity with a reduced (p ≤ 0.01) gastric dry matter content (18.7 %), solid phase (74.7 %), and pH (3.22) compared to NLPO nursery pigs (20.5 %, 85.09 %, 3.56, respectively). We have demonstrated that LPO are present in nursery pigs as early as 10 weeks of age. We observed that low birth weight is associated with higher odds for LPO in nursery pigs. These observations might help explain differences between individual pigs in terms of LPO development when groups are exposed to similar environments, management, and feeding conditions.

Association between gastric content fluidity and pars oesophageal ulcers in nursery pigs: a cross-sectional study of high-risk Danish herds using commercial feed.

BACKGROUND: The objective of this cross-sectional study was to assess the within-herd prevalence of pars oesophageal ulcers (POU) in high-risk Danish herds using commercial diets. Furthermore, we aimed to estimate the association between gastric content fluidity and POU using a generalised additive model (GAM). The study included 200 clinically healthy nursery pigs randomly selected from ten farms (20 pigs from each farm). The 10 farms were selected based on a suspected high prevalence of gastric ulcers. Post-mortem gastric ulcer assessment was based on macroscopic lesions, and gastric content fluidity was assessed based on the solid particle sedimentation percentage (solid phase). RESULTS: We observed an overall prevalence of 35.5% for POU in nursery pigs. Within-herd prevalence varied considerably among farms, with values ranging from 0% in Farm 1 to 84% in Farm 4. Our model showed strong associations between POU and gastric content fluidity (P < 0.001), as well as between POU and farm of origin (P < 0.001). In addition, we observed that the risk of POU decreased non-linearly as the gastric content solid phase percentage increased, i.e. as the gastric content became more solid. CONCLUSION: We have demonstrated that pars oesophageal ulcers are present in Danish herds with nursery pigs fed commercial diets. Furthermore, we have established that gastric content fluidity is strongly associated with POU in nursery pigs. Even so, we cannot conclude that gastric content fluidity is solely responsible for POU. Future research should look into the association between pars oesophageal ulcers and both farm management activities and individual pig factors.

Color of Colon Content of Normal and Intrauterine Growth-Restricted Weaned Piglets is Associated with Specific Microbial Taxa and Physiological Parameters.

A well-balanced gut microbiome is associated with improved health outcomes, but to date, the GM of IUGR piglets have only been sparsely investigated. Here, we investigated GM composition, color of colon content, and blood parameters of 20 IUGR and 20 normal 24-day-old piglets. No significant differences were detected in colon microbiota composition between IUGR and the normal piglets with respect to alpha and beta diversity measures. The colon content of these piglets displayed three colors: brown, black, and yellow. Interestingly, the color of the colon content varied with microbial community composition, with significant differences in the relative abundance of taxa belonging to Fusobacteria and Treponema. Fusobacteria were most abundant in yellow fecal samples, with a mean relative abundance around 5.6%, whereas this was 0.51% within brown and 0.02% for the black fecal samples. Fusobacteria positively correlated with total blood protein, albumin, and triglycerides. Contrarily, Treponema was at 0.9% the most abundant in black fecal samples, while present at 0.1% of relative abundance in brown fecal samples and 0.01% in yellow samples, correlating positively with blood iron content. This study indicates that colon/fecal content color can be used as indicator for specific GM and metabolite signatures.

Comparison of antimicrobial resistance in E. coli isolated from rectal and floor samples in pens with diarrhoeic nursery pigs in Denmark.

INTRODUCTION: The prudent use of antibiotics in veterinary medicine necessitates the selection of antibiotic compounds with narrow-spectrums targeted against the specific pathogens involved. The same pathotype of enterotoxigenic E. coli (ETEC) was recently found both in diarrhoeic pigs and in samples from the pen floor where the pigs were housed. The first objective of this study was to compare resistance profiles from ETEC isolates and Non-ETEC isolates. The second objective was to evaluate the agreement between resistance profiles of ETEC isolated from pen floor samples and from individual rectal samples from pigs. Across three Danish pig herds, faecal samples were collected from the floors of 31 pens that had a within-pen diarrhoea prevalence of >25%, and from rectal samples of 93 diarrhoeic nursery pigs from the same pens. A total of 380 E. coli isolates were analysed by PCR and classified as ETEC when genes for adhesin factors and enterotoxins were detected. Minimum inhibitory concentrations of 13 antimicrobial agents were determined by the broth micro dilution method. Isolates were classified as resistant based on clinical breakpoints. RESULTS: Based on logistic regression models, the odds of Non-ETEC isolates (n=291) being pan-susceptible were significantly higher compared to ETEC isolates (n=89), (P<0.001, OR=20.22, CI95%=6.35-64.35). The odds of ETEC isolates having multidrug resistance were significantly higher compared to Non-ETEC isolates (p<0.001, OR: 7.21, CI95%: 2.87-18.10). The odds of an isolate being resistant were significantly higher in ETEC isolates compared to Non-ETEC isolates for ampicillin (p<0.001), apramycin (p=0.003), sulphamethoxazole (p<0.001) and trimethoprim (p<0.001). No overlap of resistance patterns between the three study herds was observed in the sampled ETEC isolates. In addition, there was generally good or excellent agreement when comparing resistance profiles from isolates from the same pen (pen floor and pig samples), and perfect agreement (Kappa=1.000, SE=0.316) was observed for ampicillin, apramycin, gentamycin, sulphamethoxazole, tetracycline and trimethoprim. CONCLUSIONS: We found that ETEC isolates were more resistant than Non-ETEC isolates. Furthermore, this study indicates that resistance testing of ETEC isolates from pen floor samples can be used as a convenient sampling method for resistance testing and in the selection of clinically relevant antimicrobial agents in the treatment of diarrhoeic pigs. The herd-level variation of resistance in ETEC isolates emphasises the importance of performing antimicrobial susceptibility testing at farm level when selecting antimicrobial agents for the treatment of E. coli-related diarrhoea.

Batch medication of intestinal infections in nursery pigs-A randomised clinical trial on the efficacy of treatment strategy, type of antibiotic and bacterial load on average daily weight gain.

INTRODUCTION: Previous research projects have demonstrated the need for better diagnostic tools to support decisions on medication strategies for infections caused by Escherichia coli F4 (F4) and F18 (F18), Lawsonia intracellularis (LI) and Brachyspira pilosicoli (PILO). This study was carried out as a randomised clinical trial in three Danish pig herds and included 1047 nursery pigs, distributed over 10 batches and 78 pens. The objectives of this study were: (1) to assess the effect of four 5-day treatment strategies (initiated at clinical outbreak of diarrhoea or at fixed time points 14, 21, or 28days after weaning) on average daily weight gain (ADG); (2) to compare the effect of treatment with doxycycline or tylosine on diarrhoea prevalence, pathogenic bacterial load, and ADG; (3) to evaluate PCR testing of faecal pen floor samples as a diagnostic tool for determining the optimal time of treatment. RESULTS: (1) The four treatment strategies had a significant overall effect on ADG (p=0.01). Pigs starting treatment 14days after weaning had a significantly higher ADG (42 g) compared to pigs treated on day 28 (p=0.01). (2) When measured 2days after treatment, doxycycline treatment resulted in fewer LI-positive pens (p=0.004), lower excretion levels of LI (p=0.013), and fewer pens with a high level of LI (p=0.031) compared to pens treated with tylosine. There was no significant difference in F4, F18 and PILO levels after treatment with the two antibiotic compounds. There was a significant difference (p=0.04) of mean diarrhoea prevalence on day 21 of the study between pens treated with tylosine (0.254, 95% CI: 0.184-0.324), and doxycycline (0.167, 95% CI: 0.124-0.210). The type of antibiotic compound was not found to have a significant effect on ADG (p=0.209). (3) Pigs starting treatment on day 14 in pens where F4, F18, LI or PILO were detected by qPCR on the pen floor had a statistically significant increase in ADG (66g) compared to pigs treated on day 14 in pens where no enteric pathogens were detected (p=0.04). CONCLUSIONS: The results of this study showed that the highest ADG was achieved when treatment was initiated 14days after weaning in pens where intestinal pathogens were detected. Doxycycline was more effective in reducing diarrhoea and LI excretion levels than treatment with tylosine.

Occurrence of diarrhoea and intestinal pathogens in non-medicated nursery pigs.

BACKGROUND: Intestinal disease in nursery pigs is the most common cause of antibiotic usage in pigs in Denmark. The decision to initiate batch medication of intestinal diseases in nursery pigs is typically made by the stock personnel based on clinical assessments of pigs and counting of diarrhoeic faecal pools on the pen floor. The target population of this study was batches of nursery pigs (10-66 days after weaning) where the stock personnel assessed the pigs to be without signs of intestinal disease and therefore did not needed treatment. The objective was to determine the within-herd prevalence of diarrhoea, and to determine the prevalence of Escherichia coli F4 and F18, Lawsonia intracellularis and Brachyspira pilosicoli by quantitative PCR in pigs with and without diarrhoea. RESULTS: The overall apparent prevalence of diarrhoeic pigs across sixteen herds was 32.6 % (CI 95 % 27.9-37.3). The prevalence of diarrhoea increased (p ≤ 0.001) with age of the pigs (days after weaning) with an odds ratio of 1.04 (CI 95 % 1.02-1.05) per extra day. Diarrhoeic pools were observed in 51 % of the pens. L. intracellularis, B. pilosicoli, E. coli F4 and F18 were detected in 20, 17, 13 and 11 % of the 256 faecal samples analysed by quantitative PCR respectively. There was no association between detection of pathogens and diarrhoea status of the individual pigs and between detection of pathogens in a pen and diarrhoea floor pools. In 51 % of the samples from diarrhoeic pigs, pathogens were not detected. Only 5 % of the 3060 pigs examined had clinical signs of diseases other than diarrhoea. CONCLUSIONS: One-third of non-medicated nursery pigs had diarrhoea when clinically examined even though they were assessed as healthy by stock personnel. Diarrhoeic status of the pigs and diarrhoeic pools in pen was a poor indicator of intestinal infections with E. coli F4 and F18, L. intracellularis and B. pilosicoli and subclinical infections were common. Therefore, clinical examination and counting of diarrhoea pools should be supported by microbiological testing as decision tools for initiation of batch treatments of intestinal infections in nursery pigs.

Fetal lipopolysaccharide exposure modulates diet-dependent gut maturation and sensitivity to necrotising enterocolitis in pre-term pigs.

Uterine infections during pregnancy predispose to pre-term birth and postnatal morbidity, but it is unknown how prenatal bacterial exposure affects maturation of the immature gut. We hypothesised that a prenatal exposure to gram-negative lipopolysaccharide (LPS) has immunomodulatory effects that improve resistance towards necrotising enterocolitis (NEC) in pre-term neonates. At approximately 85 % gestation, pig fetuses were injected intramuscularly with saline or LPS (0·014 mg/kg), or intra-amniotically with LPS (0·4 mg/kg). Pigs were delivered by caesarean section 3-5 d later and fed colostrum (C) or formula (F) for 48 h. Gut indices did not differ between pigs injected intramuscularly with saline or LPS, and these groups were therefore pooled into two control groups according to diet (control-F, n 32 and control-C, n 11). Control-F pigs showed reduced villus heights, mucosal structure, gut integrity, digestive enzymes, elevated NEC incidence (38 v. 0 %, P < 0·05) and several differentially expressed immune-related genes, relative to control-C pigs. Compared with the control-F and control-C groups, values in formula-fed pigs given intra-amniotic LPS formula (n 17) were intermediate for villus height, enzyme activities, intestinal permeability and NEC incidence (18 %, P = 0·2 relative to control-F), and numbers of differentially expressed immune genes. In conclusion, prenatal exposure of the fetal gut to Gram-negative bacteria may modulate the immediate postnatal response to an enteral diet and colonising bacteria.