New review from global geroscience experts maps the current state of digital biomarkers – and where they must go next.
From sleep scoring to lens elasticity, respiration to step count, the growing menu of digital biomarkers is promising to transform how we monitor aging – and eventually, how we intervene. A new review published in The Lancet Healthy Longevity by researchers from institutions including Harvard, Stanford, NUS Singapore and the Buck Institute takes stock of this diverse field, offering both a classification of available digital biomarkers and a candid look at their readiness for clinical use.
In focusing on technologies that are non-invasive, wearable or portable, and suitable for use in community-dwelling adults, the authors have mapped the evidence for digital biomarkers across eight physiological systems – nervous, musculoskeletal, respiratory, cardiovascular, digestive, endocrine, reproductive and integumentary – organizing them according to validity, responsiveness, cost-effectiveness and generalizability [1].
Longevity.Technology: Digital biomarkers are finally catching up to their promise – but this comprehensive review underlines how fragmented the field still is. Step count and sleep remain the usual suspects, yet aging touches every system in the body. Why, then, are so few trials integrating metrics for cognitive decline, endocrine shifts or skin glycation? Take the LensAge index, for example – a deep learning tool that analyzes smartphone photos of the eye to assess lens elasticity and transparency [2]. It’s non-invasive, scalable and associated with age, functional decline and mortality risk – and yet it still isn’t used to track intervention outcomes. The technology is outpacing the trial design.
For that to change, digital biomarkers must graduate from nice-to-have secondary endpoints to essential indicators of therapeutic effect; but validation won’t be enough if access remains an afterthought. As longevity tools expand into mainstream healthcare and consumer tech, we must design with equity in mind – ensuring digital literacy and affordability aren’t barriers to participation. The authors rightly call out the underutilization of these tools in aging trials. For a sector built on precision and personalization, we’re overdue a shift from sporadic inclusion to systematic deployment.
Function over form
The study offers a pragmatic framework for evaluating digital biomarkers – favoring those linked to age, function or mortality, and those responsive to intervention. While heart rate, physical activity and sleep are among the most widely studied and accessible, several less familiar indicators are emerging with promise. These include circadian rhythm derived from wearable accelerometry, lactate captured via sweat sensors, and advanced glycation end-products detected through smartwatch-based skin autofluorescence.
That said, enthusiasm is tempered with realism, as author Dr Jesse Poganik, Founding Co-Director of the Biomarkers of Aging Consortium told us.
“Some of the challenges are similar to those we’ve talked about before for omics biomarkers: they need to be standardized, validated across populations, and shown to be robust for prediction of future outcomes,” he said, noting a specific complication for wearables – the sheer number of devices and overlapping functionalities. “Some of them measure the same or similar features, which introduces another standardization challenge.”
What sets digital biomarkers apart, he added, is their ability to capture function where it matters – in normal, everyday life. “Of course, the great advantages of many digital biomarkers are their noninvasive nature and the opportunity for continuous monitoring, and the direct measurement of function, the loss of which is very impactful to the lives of aging individuals,” he said, arguing that digital biomarkers are more immediately interpretable and potentially more clinically useful than omics biomarkers.
“But as with omics biomarkers, more validation is needed.”
Still on the sidelines in most studies
Although digital biomarkers are gaining traction in trials, they are often consigned to secondary outcomes. The authors found only a handful of studies – most focusing on physical activity – that used digital markers to track the effects of interventions [1]. Yet, as they note, “digital biomarkers enable the monitoring of aging trajectories in non-clinical environments, in which most aging-related changes occur and ongoing clinical oversight might be low” [1].
This observation strengthens the case for their inclusion in trials of lifestyle, pharmaceutical or nutraceutical interventions aiming to slow biological aging – particularly as digital technologies enable continuous, low-cost data capture at scale.
Equity, ethics and implementation
For digital biomarkers to shift from promising adjunct to core tool, technical accuracy must be matched by thoughtful implementation. “Digital biomarkers of aging can revolutionize healthcare by enabling personalized, real-time monitoring, but to integrate them responsibly, we need rigorous validation, clear regulatory pathways, and strong ethical safeguards to ensure equity, privacy, and scientific integrity,” author Dr Mahdi Moqri, also a Founding Co-Director of the Biomarkers of Aging Consortium, told us.
The review briefly touches on the limitations of current studies, many of which are focused on older White or Asian populations, with limited data on variability across age, sex or ethnic background. It also points to future directions – notably AI integration, multi-system models and the move beyond wrists and fingers to include sweat, breath and saliva-based sensors [1].
Toward a richer biomarker palette
As epigenetic clocks, molecular signatures and imaging-based tools continue to evolve, digital biomarkers offer a unique bridge between daily life and clinical insight. They may not yet have the precision of omics, nor the regulatory weight of lab-based diagnostics – but their potential for longitudinal, personalized, functionally relevant measurement remains strong.
Whether that promise translates into clinical or consumer-grade impact will depend on more than technology. Validation, standardization and the will to move from proof-of-concept to practice will all play a part – as will the commitment to ensure these tools are built for everyone, not just those already tracking their sleep.
[1] https://www.thelancet.com/journals/lanhl/article/PIIS2666-7568%2825%2900044-3/fulltext
[2] https://www.nature.com/articles/s41467-023-42934-8
