New blood test could help spot children at risk of serious illnesses, study finds | Lipid analysis using existing machines could give warning of type 2 diabetes, liver and heart disease, researchers say
https://www.theguardian.com/science/2024/sep/20/new-blood-test-could-help-identify-childrens-health-risks-study-shows
3 Comments
From the article: Scientists say a new blood test that analyses lipids could make it easier to identify children at risk of serious health conditions including type 2 diabetes, liver and heart disease.
Researchers at King’s College London said the test capitalised on a link between lipids and diseases affecting metabolism in children, and could serve as an early warning system for potentially life-threatening illnesses.
Using machines that test blood plasma in babies and that already exist in hospitals, doctors could spot early signs of disease in children quicker and help them access treatment, the researchers said.
The study’s principal author, Dr Cristina Legido-Quigley, said: “For decades, scientists have relied on a classification system for lipids that have split them into good and bad cholesterol, but now with a simple blood test we can assess a much broader range of lipid molecules that could serve as vital early warning signs for illness.”
The development had significant implications, said Legido-Quigley, a group leader in systems medicine at King’s and head of systems medicine at the Steno Diabetes Center Copenhagen.
“In the future, this has the potential to be an entirely new way to evaluate someone’s personal risk of disease and by studying how to change lipid molecules in the body, we could even prevent metabolic diseases like diabetes altogether.”
The team’s findings were published in the journal [Nature Medicine](https://www.nature.com/articles/s41591-024-03279-x).
Haha, oh dear. I think (respect to the author’s courage in calling it a “simple blood test”) that they are certainly overestimating the number of NHS hospitals that have access to a LC Q-TOF mass spectrometer, the expertise necessary to interpret a lipidomics panel, and the willingness to go through verification-validation for the sake of a test that shows a 4% improvement over a Liver Function Test (which costs about 38p in reagents and is done on a routine analyser in about 20 minutes) for predicting hepatic steatosis in these kids. The hospital I work at has one, but it’s not actually used for regular patient testing.
Reading the paper, it’s just a wash of vague allusions toward ‘these 47 lipids are associated with that biomarker’ ‘These go up when the kids lose weight’ ‘This appears to mediate 23% of that’.
If this was something like “Kids with this particular pattern of lipids at baseline clearly did not improve clinical outcome X even after significant weight loss AND our test is 10x better than any conventional test at predicting which kids don’t improve” and then maybe later they intervene on those particular lipids with drugs and show an improvement, then sure I could get behind the idea that lipidomics is clinically useful.
Having read the source article, I think the title is a bit generous.
There was a lot of statistical magic. Math did a lot of heavy lifting and there were a lot of variable corrected analyses which makes me weary of accepting results as anything more than a starting point for further analysis.
Granted I was reading it with my morning coffee, it was not an easy thing to follow overall. The big takeaways for me were a small, homogenous sample (~1000 European, caucasian children) with a lot of measured biomarkers.
It is good quality research, but likely just the start of a line of inquiry before any broader assumptions can be made.