THE ARISING · Longevity & Healthspan
Blood Protein “Organ Age” Scores Added Little to Alzheimer’s Prediction
A large UK Biobank preprint found a modest gain over a basic prediction model, not enough to establish a useful clinical test.
The idea behind a blood test for “organ age” is appealing: blood proteins might reveal whether the brain or another organ is aging faster than a person’s birthday suggests. The harder question is whether the resulting scores tell a clinician anything useful that simpler information cannot.
An analysis of 52,667 UK Biobank volunteers took a direct look at that question. The researchers used blood protein data to compare three methods of building scores across 15 organ systems and two measures spanning the whole body. One measured the gap between a person’s actual age and an age estimated by a model. Another summarized patterns of proteins without first training the model to predict age. A third linked proteins to traits associated with how organs work. The study also examined brain MRI data in 5,078 of those participants.
The scores captured related but distinct signals. Adding all the organ age-gap scores produced the largest gain for predicting later recorded diagnoses of Alzheimer’s disease. The area under the receiver operating characteristic curve, an estimate of how well a model ranks people who do and do not develop disease, rose by 0.037 over a model using basic demographic information. That number is not a percentage-point change in someone’s risk, and it does not show whether using the test would improve care.
The work is observational and the report is a preprint, not peer-reviewed. A useful next step would be to test the approach in a separate, more representative population and against a stronger clinical baseline. Even then, the question would remain: does an “organ age” result change whom to evaluate or how to follow up? This study did not establish that.