Diagnostic accuracy
Do blood ageing markers predict death better as a bundle than one-by-one?
Open access · cc by · source: Europe PMC
In 777 Newcastle 85+ participants, a 40-biomarker frailty index predicted 7-year mortality (HR 1.05 per 1% FI-B; AUC 0.66) better than any single marker (AUC ≤0.61); combining with the clinical FI raised AUC to 0.75.
How this study connects
Role on claims
Each row is a claim on a concept or method page where this paper supports, challenges, or qualifies the statement. Roles are hand-checked — not a model guess.
The two indices correlated only weakly (r = 0.16), so the blood markers add information rather than restating the clinical picture.
A composite index outperformed each of its components. A frailty index built from blood biomarkers beat every individual biomarker and was robust to dropping items, with each 1% rise raising hazard by 5.4%; combining it with a clinical frailty index gave AUC 0.75 against 0.71 and 0.66 alone.
Evidence for the claim as stated.
How much a validation study can claim depends on its sample and its cut-offs, and these differ by an order of magnitude. The p-tau comparison rests on a single memory clinic with cohort-specific thresholds and only 36 CSF samples, while the frailty index draws on a large cohort but reports AUCs around 0.75 that its authors say are not bedside tests.
Evidence for the claim as stated.
Open questions
Tensions this paper is part of
From concept pages' “where studies disagree.” Disagreement means the same question; scope means different assays, populations, or outcomes.
How much a validation study can claim depends on its sample and its cut-offs, and these differ by an order of magnitude. The p-tau comparison rests on a single memory clinic with cohort-specific thresholds and only 36 CSF samples, while the frailty index draws on a large cohort but reports AUCs around 0.75 that its authors say are not bedside tests.
Study at a glance
- Design
- Cohort — Newcastle 85+ cohort: baseline biomarker/clinical frailty indices vs up to 7-year mortality.
- N
- N=845 · 845 enrolled (mean age 85.5); FI-B calculable in 777 (60.9% women).
- Population
- Very old adults in the Newcastle 85+ Study (UK), mean age 85.5 years.
- Outcome
- All-cause mortality predicted by a 40-item biomarker frailty index (FI-B) vs clinical FI-CD, Fried phenotype, and single biomarkers.
Structured fields used in claim comparison tables when every cited study has a complete layer.
Inspect source passages
Do blood ageing markers predict death better as a bundle than one-by-one?
These passages occur verbatim in the stored paper text. They support the points listed below; they do not verify every statement in the explanation.
Newcastle 85+ analytic frame was n=845, mean age 85.5; FI-B in 777.
We analyzed baseline data and up to 7-year mortality in the Newcastle 85+ Study (n = 845; mean age 85.5).
Mean FI-B was 0.35 vs FI-CD 0.22; each 1% FI-B raised HR by 5.4%.
The mean FI-B was 0.35 (SD, 0.08), higher than the mean FI-CD (0.22; SD, 0.12); no participant had an FI-B score <0.12.
Key findings
Mean FI-B 0.35 vs FI-CD 0.22; nobody had FI-B <0.12. Each 1% FI-B rise raised HR by 5.4%. FI-B beat every individual biomarker and was robust to dropping items. Combined FI-CD/FI-B AUC was 0.75 vs 0.71 and 0.66 alone.
Methodology
Built FI-B from 40 cellular-ageing, inflammation, haematology, and immunosenescence markers, compared it with a clinical deficit FI and the Fried phenotype, and tested 7-year mortality with Cox models, Kaplan–Meier strata, random biomarker subsamples, and ROC AUCs.
Limitations
Moderate AUCs are not bedside tests; FI-B and FI-CD correlated only weakly (r=0.16), so blood markers do not simply duplicate clinical frailty.
Related papers in this topic
Same topic cluster — not a recommendation engine.