Obscore AI Spots Disease Risk in Overweight Patients
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- Obscore, the new AI tool, uses interpretable machine learning on data from almost 200,000 UK Biobank participants with BMI 27 or above to predict 10-year risk of 18 obesity-related complications — from stroke to gout — using 20 features including age, sex, total cholesterol, and creatinine levels.
- Prof Nick Wareham of the University of Cambridge said the tool is about "more rational resource allocation," allowing therapy to be prescribed to those most likely to need and benefit, as NHS access to weight-loss jabs is currently limited to those with high BMI and existing obesity-related health problems.
- Kamil Demircan of Queen Mary University of London noted that for conditions including type 2 diabetes, the highest-risk category included "a considerable proportion of people who are overweight rather than obese" — a group that "may be overlooked if we only look at BMI and not other risk factors."
- The researchers validated Obscore using UK Biobank data and two independent health studies, and applied a version of the tool to data from a tirzepatide randomized control trial, confirming that highest-risk patients would experience similar weight loss to others on the drug.
- Naveed Sattar of the University of Glasgow, who was not involved in the study, cautioned that many obesity-related conditions are closely interrelated, that robust risk scores already exist for some, and that several Obscore metrics are not routinely available in the NHS — requiring "substantial further development and validation" before clinical adoption.
- The study, published in Nature Medicine, comes as recent data shows about two-thirds of adults in England are overweight or obese.
Why it matters: Current NHS access to weight-loss jabs is rationed by BMI and pre-existing conditions, but Obscore's data show that overweight (not obese) people can rank in the highest risk tier for type 2 diabetes. If adopted, the tool could redirect limited drug supplies toward patients more likely to benefit — though independent experts flag routine NHS data gaps.
