Thought for food

Despite strong evidence that diet can impact colorectal cancer (CRC) risk, our understanding is limited by the lack of human studies and insight into the underlying mechanisms of these relationships.

Many long-term randomised controlled trials (RCTs) with cancer incidence as the outcome are needed to inform dietary guidance in a complex and evolving nutritional landscape – and that’s a real challenge, because conducting these studies is extremely difficult given the long adherence timeframe, high costs, and large number of participants required.

Because of this, it remains unclear how a low-carbohydrate (ketogenic or “keto” diet) affects CRC risk. This diet is commonly followed in the Western world – by 10-15% of the UK population – for weight loss and general health improvement, yet whether it increases or decreases CRC risk, and through which mechanisms, is still unknown.

Our team has taken on the challenge of exploring this question using a streamlined study design that brings together the strengths of clinical trials and genetic epidemiology. The approach allows us to examine whether short-term dietary changes influence biological pathways that may, over time, affect cancer risk – without requiring decades-long trials to detect disease outcomes. By focusing on molecular changes that occur early in response to dietary interventions, we can gain crucial understanding in a fraction of the time typically required by traditional approaches.

Weeks, not years…

Instead of waiting decades for cancer outcomes to emerge from dietary trials, we are leveraging molecular changes that occur early – weeks, not years – after a dietary intervention.

In a recently completed RCT, 42 participants followed two 4-week dietary interventions: a standard control diet, and a very low-carbohydrate diet designed to induce ketosis.

Blood and adipose tissue samples were collected at baseline and at 4 weeks to measure molecular changes, including circulating proteins and gene expression. These are the biological footprints that could link diet to disease. However, our analysis goes beyond biomarker measurement. Using Mendelian randomisation (MR) – an epidemiological method that leverages naturally occurring genetic variations as proxies for long-term exposure – we aim to test whether the molecular traits altered by the ketogenic diet have a causal effect on CRC risk. With MR, we can infer causality in a way that observational studies cannot.

Importantly, we’ll be looking at CRC by anatomical site, sex, and early-onset cases (diagnosed under 50 years old), where rates are rising at an alarming pace.

This project brings together expertise from across the Southwest of the UK. At Bath, researchers with deep experience in human RCTs, especially focused on understanding fat and carbohydrate metabolism in humans, led the clinical side of the study. At Bristol, experts in MR and cancer epidemiology, are driving the next phase. The complimentary skills shared between our research groups strengthens our study design, and brings together molecular, clinical and epidemiological expertise, creating a robust foundation for tackling the diet-cancer relationship.

Helping inform choice

Throughout the project, we plan to work closely with people with experience of cancer to enhance the transferability and relevance of our research.

Their feedback, which in part has emphasised the urgency of understanding the role of diet in rising CRC rates among young people, has helped to sharpen our focus and keep our work grounded in real-world impact.

We believe that combining the molecular precision of modern science with the lived experience of patients is the best way to ensure our research matters – not just to the scientific community, but to people everywhere trying to make informed choices about their health.

Our goal is to identify molecular changes triggered by a ketogenic diet that play a causal role in CRC risk. By uncovering whether these biological shifts are protective or harmful, we hope to inform future dietary recommendations and open the door to targeted interventions. Ultimately, our findings will clarify whether ketogenesis could be a viable strategy for reducing CRC risk – and help unravel the mechanisms behind it.

Perhaps more importantly, our study design itself – linking short-term RCTs to long-term outcomes using MR – could become a template for future research on diet and cancer prevention.

In a world where nutrition advice evolves more quickly than public health guidelines can keep up, resource-efficient studies are crucial to pushing the field forward.