AI-Detected INSULIN RESISTANCE Increases the Risk of 12 Types of Cancer
In this article
CANCER’S HIDDEN BIOLOGICAL ENGINE
AI-IR, a new digital score developed using artificial intelligence to estimate insulin resistance in the body, sheds light not only on a person’s weight but also on their metabolic risk. This feature offers a notable innovation for early cancer diagnosis and more accurate selection of individuals at risk.
We have long known that diabetes and obesity increase cancer risk. However, “insulin resistance”, the main bridge between these two conditions, was a parameter that was difficult to measure in routine clinical practice (because gold-standard methods are laborious) and often overlooked. Published in Nature Communications, this study shows that a “digital insulin resistance score” (AI-IR), generated by artificial intelligence from 9 basic clinical data points without complex laboratory tests, can illuminate not only an individual’s diabetes risk but also their susceptibility to 12 different types of cancer like a beacon.
A digital biomarker: how does the AI-IR model work?
Instead of the complex “clamp” technique used to detect insulin resistance, the researchers developed a machine learning algorithm that brings together parameters routinely measured in every hospital. AI-IR is the name of this model, which achieves a high accuracy of 88% (AUC: 0.88) by analyzing the following 9 parameters:
The 9 Clinical Data Points That Feed the Algorithm
This massive analysis of UK Biobank data from 372395 cancer-free individuals proved that the AI-IR score is far superior to traditional methods such as BMI and metabolic syndrome in predicting diabetes.
The relationship between 12 types of cancer and insulin resistance
The study revealed that cancer risk increased significantly in individuals with high AI-IR scores. The statistically strongest associations (after “Bonferroni” correction) were identified for the following types:
It stands out as the cancer type most strongly associated with insulin resistance.
In addition to these main cancers, colon, breast (in women), stomach, liver, gallbladder, leukemia, and lung cancer risks were also observed to increase in direct proportion to the insulin resistance score. In the composite cancer analysis, an age- and sex-adjusted increase in risk of %25 was calculated.
Weight is not everything: BMI-independent risks
One of the most striking findings for the oncology community is that the effect of insulin resistance cannot be explained simply by “being overweight”. Even when the researchers adjusted the data for Body Mass Index (BMI), they found that the effect of insulin resistance on cancer persisted.
In contrast, the relationship between insulin resistance and skin cancer was found to be inverse (HR: 0.85) in the study, leaving a question about how metabolic processes can produce different responses in different tissues.
Healing begins with metabolic honesty
In our story with cancer, we often focus on a tumor’s genetic code, but sometimes neglect the metabolic health of the “soil” (the body) that harbors it. This study whispers to us that insulin resistance is a silent storm that disturbs the calm of cells and pushes them toward becoming cancerous.
This digital biomarker (AI-IR) offered by artificial intelligence is part of our vision of precise guidance in oncology. Rather than simply saying “watch your weight”, we can now map an individual’s metabolic risk and seize the chance for early diagnosis through targeted screening. Healing means restoring harmony to the body’s biological orchestra (the balance of insulin and sugar).
Scientific sources and further reading
- 1. Lee CL, et al. (2026). Machine learning-predicted insulin resistance is a risk factor for 12 types of cancer. Nature Communications.
- 2. UK Biobank Prospective Cohort Study (2006-2022). Cardiovascular and Cancer Incidence Analysis.
- 3. HOMA-IR and Hyperinsulinemic-euglycemic clamp validation studies. Journal of Clinical Investigation.
Was this article helpful?
Your feedback helps us improve our content.
Reader comments
Share your comment
Your name and comment are published only after administrator approval. Your email address is never shown to visitors.
Be the first to comment.