Machine Learning Unveils Metabolic Biomarkers for Predicting Cancer Risk

By Sunil Sonkar
2 Min Read
Machine Learning Unveils Metabolic Biomarkers for Predicting Cancer Risk

A study conducted by researchers at the University of South Australia has leveraged the power of machine learning to uncover metabolic biomarkers that hold the potential to predict the risk of developing cancer. This innovative research explored data from an impressive group of 459,169 individuals participating in the UK Biobank.

The findings identified a remarkable total of 84 distinct features within this data that serve as indicators of an elevated risk of cancer. Beyond this critical discovery, the research also shed light on markers associated with chronic kidney or liver disease, hinting at intriguing connections between these ailments and the onset of cancer.

Dr. Amanda Lumsden, one of the lead researchers, highlights the significance of this study in clarifying the mechanisms that contribute to cancer risk. She highlights one particularly noteworthy finding: high levels of urinary microalbumin emerged as the most potent predictor of cancer risk, following age. Microalbumin, typically a serum protein required for tissue growth and healing, takes on a different role when found in urine. In this context, it serves not only as an indicator of kidney disease but also as a signal for heightened cancer risk.

The research further spotlighted indicators of compromised kidney function, such as elevated levels of cystatin C and urinary creatinine, as well as lower total serum protein. All these are commonly said to be linked to an increased risk of cancer. In addition to these kidney-related findings, the study simultaneously revealed that heightened levels of C-reactive protein, a marker of systemic inflammation, and the enzyme gamma glutamyl transferase (GGT), which reflects liver stress, were also associated with a higher likelihood of developing cancer.

Share This Article