How can mathematical modelling help us to understand the threat posed by antimicrobial resistance?

What will the world look like in 2050—flying cars and a 3-day workweek, or ecological catastrophe and untreatable infections? While it might be more fun to think about all the wonderful things the future may have in store for us, it’s now clear that without drastic measures, the future may also bring major threats to our way of life.

By 2050, researchers have projected that climate change may cause 600,000 deaths a year and slash world GDP by 10%. Meanwhile, a report commissioned by the UK government predicts that, by the same year, antimicrobial-resistant infections could kill 10 million people yearly and suck 100 trillion USD out of the global economy. These estimates are massively uncertain, but they underline the scale of the problems we face in the 21st century.

Global warming and antimicrobial resistance (AMR) are similar beyond their potential for devastation. Each represents an unforeseen and potentially disastrous consequence stemming from the widespread use of a valuable technology. While using up the energy in fossil fuels has led to a warming of the earth’s climate, curing infections with antibiotics has promoted the spread of superbugs.

Moreover, both problems demand worldwide action. While the response to climate change has been slow, global accords such as the Paris Agreement may be a step towards mitigating its impact. Governments have also started to take notice of AMR: the UK just announced its ambition to reduce antibiotic prescriptions by 15% by 2024.

Yet while estimates of the impact of global warming are based on a scientific understanding of the Earth’s climate and underpinned by detailed mathematical models run on supercomputers, we lack similarly robust models for how antimicrobial resistance spreads. This means that we have very little certainty about the magnitude of this threat.

In a new study, I and my LSHTM colleagues Katherine E. Atkins, Stefan Flasche, and Mark Jit, have developed a dynamic mathematical model of resistance evolution that can explain current patterns of antibiotic resistance across 30 European countries.

While projecting this model into the future will be an important next step, it was a challenge just to be able to explain the trends we see today. Our work suggests that competition between resistant and non-resistant bacteria may be an important piece of the puzzle.

Why does this competition matter? The bacteria that live in our bodies are not isolated, but grow in vast aggregations of hundreds or thousands of species in the gut, nose, skin, and other sites. Living in such close proximity to one another puts these different strains in intense competition for the resources that our bodies provide in the form of space to grow and food to eat.

While we are used to thinking of E. coli and other bacteria as dangerous, for the most part we live in harmony with these and other “commensal” bacteria: they are a normal part of our microflora, only occasionally causing disease, and in some cases they may even help to protect us from nastier bugs.

Our work shows how competition between different strains of bacteria may impact upon the spread of antibiotic resistance. When people take antibiotics, non-resistant bacteria are killed and so the resistant bacteria that live inside them do comparatively better. Conversely, if a person isn’t taking antibiotics, resistant bacteria do worse because these bacteria are wasting energy on costly traits—like enzymes that actively pump antibiotics out of the cell—that are of no use when antibiotics aren’t present. Even a small amount of wasted energy can spell doom for bacteria—it’s a microbe-eat-microbe world out there. In our work, we found that explicitly accounting for this antibiotic-mediated competition within individual hosts helped us to explain patterns of resistance we observe in the wider population.

To our knowledge, our paper contains the first pan-European dynamic mathematical model of resistance evolution. Now, we aim to extend our predictions to a global scale and into the future. One day soon, dynamic mathematical models of resistance evolution may give us a better sense of the scale of this problem—and help us to plan how to manage it.


Nicholas G. Davies, Stefan Flasche, Mark Jit, Katherine E. Atkins. (2019) Within-host dynamics shape antibiotic resistance in commensal bacteria. Nature Ecology & Evolution. doi: 10.1038/s41559-018-0786-x

Short courses

LSHTM's short and specifically designed courses provide the opportunity for intensive study in specialised topics.

These courses enable participants to refresh their skills and keep up to date with the latest research and knowledge in public and global health.