Bowel cancer is surging in under-50s, but AI could help us catch it early

New Zealand has seen a striking rise in bowel cancer among people under 50, with researchers investigating everything from diet and obesity to nitrates, plastics, and gut bacteria as possible causes. What we do know though is that earlier and easier detection would save lives, and this will depend on better screening technologies.

It was only a generation ago that bowel cancer was considered a disease of retirement. Today, Kiwis in their thirties and forties are increasingly hearing the words no one wants to hear: “You have bowel cancer.” New Zealand has experienced one of the fastest increases in early onset bowel cancer in the world.

This surge began in the mid-1990s and has accelerated since. Between 2000 and 2020, incidence rates have risen by an average of 26 percent per decade.

Historically, high rates of cancer have been a byproduct of modern medicine: as we are less likely to die from preventable, infectious diseases, we live long enough to develop ageing-associated conditions such as cancer. But this early onset bowel cancer spike breaks that rule.

What is driving the increase?

We know it is highly unlikely to be genetic. Human DNA simply does not mutate rapidly enough across a population in just two or three decades to cause a spike of this magnitude. The cause is likely a complex web of environmental and lifestyle factors.

There are known risk factors that are likely involved. The rapid normalisation of ultra-processed foods and shifting global diets have fundamentally changed what we eat. Obesity rates have climbed alongside this. Environmental factors such as the presence of microplastics and nitrates in our water systems are also being investigated. Some researchers are looking backward to early development, investigating whether birthweight could serve as a biomarker for future risk.

The cause is likely to be complicated though, and consequently so is prevention. The good news is that we do have technical capacity to prevent people from dying from the disease by diagnosing those at risk earlier, and these technologies can only get better.

New Zealand’s bowel screening programme has already saved countless lives by detecting cancers before symptoms develop. The faecal immunochemical test is simple, inexpensive and remarkably effective at identifying people who need a colonoscopy. Arriving in a little green envelope in the mail, this at-home stool test has been responsible for thousands of early diagnoses by the National Bowel Screening Programme.

The test by itself doesn’t tell you if you have cancer; but it does detect microscopic blood in your stool, indicating that you should be referred for a colonoscopy.

Many people who receive the free kit in the mail do not return it, often because they feel uncomfortable collecting a stool sample or are worried about what the result might mean. Increasing participation in bowel screening remains one of the simplest ways to detect cancers sooner.

Bowel screening programmes are evolving around the world. Several countries, including the United States and Australia, have lowered the recommended age for routine bowel cancer screening, and large studies such as in Sweden and the Netherlands are investigating whether screening intervals and referral pathways can be tailored to an individual’s risk rather than relying solely on age. The common goal is clear: detect more cancers while avoiding unnecessary colonoscopies.

Outside the National Bowel Screening Programme, New Zealand introduced the FIT for Symptomatic pathway nationwide in July 2026. Unlike the age-based screening programme for people without symptoms, this pathway is available to people of any age who present to their GP with bowel symptoms. It uses the same faecal immunochemical test tubes to help estimate a patient’s risk of bowel cancer, allowing clinicians to prioritise urgent colonoscopies for those most likely to benefit while safely managing many lower-risk patients in primary care.

However, the future of bowel cancer screening is unlikely to rely on a single biomarker. The next generation of tests will probably combine multiple sources of information from blood, DNA, the microbiome and other molecular signals, to provide a more personalised assessment of risk from the same simple stool sample.

In my lab, we are hoping to investigate what can be revealed from tiny bits of poo left behind in the tubes that currently get thrown away. When a standard test is processed, residual sample remains in the tube. That leftover material is rich in information – a goldmine of bacterial DNA.

By extracting and analysing this DNA, we can map a patient’s microbiome; the trillions of bacteria living in our gut influence digestion, immunity and inflammation. A growing body of evidence suggests that the gut microbiome plays a role in bowel health and may provide useful biomarkers for colorectal cancer. Last year, for instance, research spotlighted a specific bacterial strain of E. coli capable of mutating human DNA and promoting tumour growth in early onset cases.

To make sense of this massive amount of biological data, we intend to use Artificial Intelligence. Not generative text bots like ChatGPT, but the statistical machine-learning models that scientists have used for decades. By training AI to recognise the microscopic bacterial signatures of early stage cancer, we could better predict who is at risk before a tumour ever bleeds. Like most game-changing research, achieving this will depend on funding.

Solving the mystery of why bowel cancer is striking younger New Zealanders may take years. Until we understand why, our best defence is to find it earlier.