A latest artificial intelligence model developed by researchers at the University of Tokyo and collaborating institutions can predict an individual’s risk of developing insulin resistance – and, crucially, link that resistance to a significantly increased chance of developing 12 different types of cancer, according to research published Monday in Nature Communications.
The model, termed AI-IR, utilizes nine routinely collected clinical parameters to estimate insulin resistance, outperforming traditional measures like body mass index (BMI) in predicting both diabetes and cancer risk. The findings represent a significant step toward identifying individuals at higher risk and enabling earlier, more targeted screening.
Insulin resistance, a condition where cells become less responsive to insulin, often precedes Type 2 diabetes and is as well associated with cardiovascular, kidney, and liver diseases. While a connection between insulin resistance and cancer has long been suspected, establishing large-scale evidence has been hampered by the difficulty of accurately assessing insulin resistance in clinical settings.
“While a possible link between insulin resistance and cancer has been suggested, large-scale evidence has been limited due to the difficulty of evaluating insulin resistance in the clinic,” said Yuta Hiraike, a researcher from the University of Tokyo Hospital and a senior author of the study. “But with AI-IR, we have provided the first population-scale evidence that insulin resistance is a risk factor for cancer.”
The AI-IR model was trained and validated using data from over half a million participants in the UK Biobank, as well as independent cohorts from the United States, and Taiwan. Researchers found that AI-IR not only accurately predicted insulin resistance when compared to direct measurements – typically only available in specialized diabetes clinics – but also identified individuals at risk who might be missed by relying solely on BMI.
BMI, a common measure of body fat, can produce both false positives – identifying metabolically healthy obese individuals as high-risk – and false negatives – failing to identify insulin resistance in individuals with a normal BMI. AI-IR, by incorporating a broader range of clinical parameters, offers a more nuanced and accurate assessment.
“By combining nine clinical parameters into a single metric, AI-IR can detect insulin resistance that BMI alone cannot explain,” Hiraike explained. “And since the nine input parameters for AI-IR are obtained through standard health checkups, AI-IR could be easily implemented to identify high-risk individuals and enable focused screening of diabetes, cardiovascular disease and cancer.”
The research team is now focused on understanding the genetic factors that influence insulin resistance and cancer risk, and on integrating large-scale human data with molecular biology studies to develop more effective strategies for prevention and treatment. The University of Tokyo team intends to further refine the model and explore its potential for personalized medicine approaches.
