Researchers report that insulin resistance (IR) is linked to a higher risk of multiple cancers, with a new AI tool identifying metabolic risk even in people who appear healthy by BMI. Published February 16 in Nature Communications, the study found people flagged as insulin-resistant had a 25% higher risk of developing 12 different cancers, and a 134% increased risk of uterine (endometrial) cancer.
The AI model, called AI-IR, was developed by teams at the University of Tokyo and Taichung Veterans General Hospital. It combines nine routinely collected parameters — age, sex, race, body mass index (BMI) and five blood tests — and uses artificial intelligence to predict insulin resistance. The developers chose these inputs because primary care clinicians commonly record them, making real-world implementation feasible.
AI-IR picked up elevated risks for diabetes, heart disease and cancer. The six cancers most strongly associated with predicted insulin resistance were uterine, kidney, esophagus, pancreas, colon and breast. Six additional cancers showed weaker but still significant associations: renal pelvis, small intestine, stomach, liver and gallbladder, leukemia, and bronchial and lung cancers.
Uterine cancer showed the strongest link. Experts say that association is not surprising — endometrial cancer has long been connected to excess weight and metabolic dysfunction — but AI-IR added predictive power beyond BMI alone. Even after adjusting for body weight, the model identified some patients at increased uterine cancer risk, suggesting metabolic dysfunction itself can drive risk independent of body size. Mechanistically, insulin resistance may promote cancer growth through hormonal and growth-factor signaling.
A key advantage of the AI-IR approach is its ability to detect metabolic dysfunction in people with normal BMI who would be cleared by standard screening. In particular, the model identified elevated lung and bronchial cancer risk independently of BMI. Outside experts noted the importance of recognizing cancer risk in normal-weight patients who have high insulin levels, insulin resistance, or high body fat (a phenotype sometimes called normal-weight obesity). Prior research has shown that normal-weight individuals with high body fat can have substantially increased cancer risk.
The model was trained using datasets from U.S. and Taiwanese populations and validated on nearly 400,000 participants in the United Kingdom. Because the UK validation cohort was predominantly ethnically European, the authors acknowledge limitations in generalizing findings across all ethnic groups.
AI-IR is not yet available for clinical use, but the markers it relies on — including hemoglobin A1C and body fat percentage — are measurable in routine care. Experts suggest practical steps people and clinicians can take now to assess and reduce risk:
– Measure body composition. Ask about body fat percentage. A DEXA scan is most accurate; bioelectrical impedance (BIA) scales are a reasonable, accessible option. Watch trends rather than a single reading; many experts suggest aiming for total body fat below about 30% (individual targets vary).
– Rethink bloodwork. Routine hemoglobin A1C testing for diabetes screening can also reveal metabolic risk. An A1C in the 5.5–5.7% range or higher may indicate elevated cardiometabolic — and possibly cancer — risk worth discussing with a clinician.
– Improve lifestyle factors. Increased physical activity, healthier diet, and weight loss when needed reduce insulin resistance and lower risk for diabetes, heart disease and many cancers.
Researchers and outside oncologists view this work as a step toward more personalized cancer risk prediction that moves beyond BMI to metabolic drivers of disease. If validated and implemented clinically, tools like AI-IR could help identify at-risk patients earlier and guide interventions to prevent diabetes, heart disease and cancer.
