Fighting antimicrobial resistance requires looking beyond antibiotic use

Antibiotics are essential medicines for both people and animals. But their effectiveness is threatened by antimicrobial resistance (AMR), in which bacteria acquire or carry mechanisms that allow them to survive antibiotic treatment. Livestock production is an important part of the global AMR challenge, as it uses up to 80% of globally produced antibiotics. Since it is known that antibiotic use can select for resistant bacteria and resistance genes, one of the main strategies to tackle the probleblem of AMR is reducing unnecessary use. But what happens when antibiotic use is already very low?

Sweden provides an opportunity to ask precisely this question. Compared with many other countries, Swedish livestock production uses very small amounts of antibiotics and relies heavily on disease prevention and individual treatment of sick animals.

We studied ten Swedish pig farms by following one group of pigs on each farm throughout an entire production cycle to see how the resistome changed as the pigs grew and whether these changes could be explained by antibiotic use in this pig group. Environmental samples were collected from pig pens approximately every four weeks and analysed using metagenomic sequencing. This allowed us to investigate both the microbial communities and the complete collection of antibiotic resistance genes – the resistome – present in the farm environment.

Despite low antibiotic use in the studied pig groups, we found a diverse resistome. Resistance genes associated with 16 different antibiotic classes were detected, although antibiotics from only six classes were used to treat the studied pigs during the observation period. A particularly important result was that antibiotic use alone could not explain how the resistome developed.

Resistance patterns changed considerably as pigs grew, and these changes were more pronounced over time than the differences between individual farms. Moreover, the microbiome and resistome followed different trajectories, indicating that resistance dynamics depend on more than simply which bacterial species are present.

Some resistance genes remained abundant despite little or no recorded use of the corresponding antibiotics. Tetracycline resistance provides a striking example: genes associated with tetracycline resistance were abundant and increased with pig age, although tetracycline treatment was recorded in the studied pig group on only one of the ten farms.

This does not reduce the importance of responsible antibiotic use. Minimising unnecessary treatment remains essential because antibiotic exposure creates selective pressure favouring resistance. Instead, our results reveal another part of the AMR challenge: resistance can persist in livestock environments even when current antibiotic pressure is low. This raises the next question: What else is maintaining resistance on farms?

The environment factors and farm management may all play a role. Understanding the contribution of these factors is the next step in our research.

This work is important for farmers, veterinarians and policymakers alike. Sweden shows what can be achieved through decades of responsible antibiotic use, but our results suggest that further progress against AMR may require looking beyond antibiotic use alone. Identifying the factors that maintain resistance on farms could help develop more targeted prevention, biosecurity and management strategies.

Ultimately, tackling AMR requires a One Health perspective that considers animals, people and their shared environment. Sweden’s low-antibiotic livestock system provides a valuable setting for understanding what more can be done once antibiotic use has already been reduced.

Publication