Oracle and the Buck Institute for Research on Aging have begun a collaboration that will apply advanced analytics to more than 122 million de‑identified longitudinal health records to search for early indicators of ageing and age‑related disease.
What the project will do
The Buck Institute will use Oracle Life Sciences Data Intelligence — a platform that combines a vast repository of real‑world clinical records with tools for cohort construction and analytics — to probe long time series of patient data. The stated aim is to identify subtle clinical changes that may foreshadow frailty or chronic disease years before clinical onset, and to build a statistically validated model that predicts who is most at risk and when.
The work is part of a wider research effort led by Stanford University called THRIVE (Transforming Health: Reclaiming Intrinsic Vitality for Everyone), which itself operates under the PROSPR programme funded by the Advanced Research Projects Agency for Health (ARPA‑H), an agency in the US Department of Health and Human Services. THRIVE seeks to develop a validated measure of physical and mental capacity — sometimes described as an "intrinsic capacity" or vitality score — that could predict long‑term health trajectories and support earlier interventions.
"Understanding how and why we age is one of the most complex challenges in modern medicine," said Dr David Furman, PhD, of the Buck Institute for Research on Aging.
Why this matters
A validated, FDA‑grade measure of intrinsic capacity would serve multiple purposes. If robust, it could:
- Help clinicians identify patients at elevated risk of decline well before symptoms become obvious;
- Provide a standard outcome for clinical trials of interventions aimed at prolonging healthy years of life;
- Offer population‑level insight into patterns and drivers of ageing that could inform health services planning and prevention strategies.
The Buck Institute will act as one of the analytics and bioinformatics engines within the THRIVE consortium. Using Oracle’s platform, investigators plan to construct cohorts from decades of records, then apply advanced statistical and computational methods to detect patterns of small changes that, in aggregate, might indicate future decline.
Data and governance
Oracle’s dataset is described as de‑identified and longitudinal, which means records follow individuals over time while removing personal identifiers. The collaboration explicitly aims to pursue qualification of the proposed vitality metric to an FDA‑grade standard, a process that requires rigorous validation against clinical outcomes and adherence to regulatory expectations.
| Item | Detail |
|---|---|
| Dataset | 122 million+ de‑identified longitudinal records (Oracle Life Sciences Real‑World Data) |
| Consortium | THRIVE (Stanford‑led), part of PROSPR under ARPA‑H |
| Primary goal | Develop and validate an "Intrinsic Capacity Score" for healthy ageing |
Limitations and next steps
Large administrative and clinical datasets offer scale, but they also present challenges. De‑identified records can reduce risk to patient privacy, yet they can limit linkage to certain clinical details or to social and environmental data that also influence ageing. Electronic health records may capture some conditions variably across providers and over time, which complicates efforts to find subtle preclinical signals.
The Buck Institute team plans to use the data to build and statistically validate predictive models. If successful, the next stage would be supporting clinical trials of interventions informed by early detection — a process that can take years. Pursuing an FDA‑grade qualification for an intrinsic capacity metric also requires explicit demonstration that the score is reproducible, clinically meaningful and robust across populations.
Broader implications
A validated vitality score could reshape how clinicians and researchers approach ageing, shifting more attention to early detection and prevention rather than late‑stage treatment. For countries such as South Africa, where the population is ageing alongside a persistent burden of communicable and non‑communicable diseases, standardised measures of functional capacity could improve targeting of scarce resources and the design of public health interventions — provided such tools are validated across diverse populations.
The collaboration highlights two trends in modern biomedical research: the growing use of real‑world data at scale, and the movement to translate complex computational outputs into regulatory‑grade measures that can be used in clinical settings.
