Dr Eric Topol has a good dissection of this over on his substack [1]

The Nature paper itself is here [2]

Briefly: "The model was as good or slightly better than the best current single disease predictive models for dementia or heart disease (see below), and most others, but worse for Type 2 diabetes using HbA1c as single marker. Importantly, it predicted not just one but the multiple diseases (←when that risk was present) for an individual and the timing of appearance for each disease."

[1] https://erictopol.substack.com/p/dawn-of-a-new-era-of-primar...

[2] https://www.nature.com/articles/s41586-025-09529-3