Antimicrobial resistance (AMR) poses a critical challenge to both human and veterinary medicine, with pig production recognized as one of the major contributor due to intensive antimicrobial usage (AMU). This study aimed to explore the relationship between AMU and AMR patterns of Escherichia coli isolated from commercial pig farms, using data-driven analytical methods. Farm-level records were harmonized with microbiological data from 203 isolates collected in December 2023 across four Hungarian farms. AMU was summarized over 3-, 6-, 9-, and 12-month retrospective windows and expressed in modified population-corrected units, while AMR was quantified as mean minimum inhibitory concentration (MIC) and AMR rate under epidemiological and clinical breakpoints. The results revealed substantial variation in AMU among farms, with amoxicillin predominating across timeframes. Farm-specific comparisons indicated that higher AMU may not always coincide with elevated resistance levels, and data analysis did not consistently identify a direct association between use and resistance at the individual farm level, which warrants further investigation in larger datasets. Correlation analyses identified strong intra-class relationships among β-lactams and fluoroquinolones, as well as a cross-class linking, suggesting concurrent selection pressures. Overall, the integration of AMU and AMR data demonstrated the feasibility of farm-level surveillance for AMR modelling and provides a foundation for future predictive systems to support antimicrobial stewardship in livestock production.

Weaning disrupts intestinal structure and function in piglets and frequently results in post-weaning diarrhea, highlighting the need for effective nutritional alternatives to medicinal feed additives. This study evaluated non-fermented field peas and diets containing different inclusion levels of a fermented pea–wheat ingredient as substitutes for soybean meal in restricted liquid diets for weaned piglets. Fifty-six piglets were assigned to four dietary treatments for 14 days after weaning: non-fermented soybean meal, non-fermented peas, partial inclusion of the fermented pea–wheat ingredient, or full inclusion of the fermented pea–wheat ingredient. Fecal consistency, plasma inflammatory cytokines, small-intestinal morphology, crypt cell proliferation, and intestinal microbiota composition were assessed. Plasma interleukin-1β, interleukin-6, and tumor necrosis factor-α remained low and did not differ among treatments, indicating no measurable systemic inflammatory response. Diets containing the fermented pea–wheat ingredient were associated with differences in early intestinal morphology, particularly villus height on day 4 after weaning. Treatment- and segment-specific variation was also observed in crypt cell proliferation. Microbiota analysis showed differences in community composition among treatments, with diets containing the fermented pea–wheat ingredient associated with a higher relative abundance of lactic acid-producing genera. At the pen level, full inclusion of the fermented pea–wheat ingredient was consistently associated with low fecal scores, whereas partial inclusion was associated with higher fecal scores throughout the study period. Overall, these findings suggest that, under the controlled conditions of this study, diets containing the full inclusion level of the fermented pea–wheat ingredient were associated with indicators of intestinal adaptation and consistently low pen-level fecal scores. Further studies with replicated pen designs, longer-term performance evaluation, and functional analyses are needed before practical recommendations can be made.