The Cheapest Longevity Test in Medicine Takes 11 Seconds. We Can Now Measure It Every Day.

A new study in JAMA Network Open followed 5,472 women aged 63 to 99 for an average of 8.4 years. The researchers were not looking at expensive biomarkers, MRI scans, or genetic panels. They timed how long it took each woman to stand up from a chair five times in a row, without using her arms.

The women in the slowest group, taking 16.7 seconds or more, died at a rate of 60.2 per 1,000 person-years. The women in the fastest group, finishing in 11.1 seconds or less, died at less than half that rate – 26.5 per 1,000 person-years. After controlling for age, race, education, body weight, smoking, blood pressure, comorbidities, walking aid use, self-rated health, accelerometer-measured aerobic activity, sedentary time, and walking speed, the fastest group still had a 34% lower mortality risk than the slowest.

The chair stand test. Five repetitions. No equipment. No clinical setting required.

This is the cheapest, most predictive longevity test in medicine, and it has been sitting in geriatric assessment toolkits for decades. The question we have been thinking about at VRsteps is: why is it still administered once a year in a doctor’s office, when the act it measures – getting up from a chair – happens dozens of times a day in every home?

Why the sit-to-stand transition carries this much signal

The sit-to-stand transition is a uniquely informative biomechanical event. In a few seconds it integrates lower-limb power, hip and knee extensor strength, ankle dorsiflexion range, postural control, vestibular function, balance reactions, motor planning, and the willingness of the central nervous system to commit to a destabilizing movement. There is a reason it predicts mortality more strongly than walking speed in older adults: it is harder than walking, it is performed more frequently than most people realize, and the moment a person starts avoiding it – using their arms, sliding forward, rocking, hesitating – that avoidance signals a cascade of declines that often go unreported at clinic visits.

The JAMA authors note something important. The correlation between grip strength and chair stand time in their cohort was only -0.13. They are largely measuring different things. Grip strength looks like a clean biomarker of skeletal muscle output. Chair stand performance looks more like a general aging biomarker – fatigability, motor control, fear of falling, joint pain, and strength all rolled into one number. That is exactly the kind of composite signal that benefits from continuous measurement rather than annual snapshots.

What a single annual measurement misses

A patient performs the chair stand test in February. They score 14 seconds. They pass. The clinician moves on.

What the timer did not catch:

• That on most mornings they now need both arms on the armrests
• That they avoided the test during their last two doctor’s visits because they were embarrassed
• That their time has drifted from 11 seconds in 2024 to 14 in 2025 to 15.5 by November
• That the asymmetry between left and right loading has doubled

None of this is captured. A single number, taken in a single setting, with a single attempt that may have benefited from adrenaline and the desire to perform.

This is the gap that Pedisteps and VRsteps Family Wellness were built to close, though we had not articulated it specifically around sit-to-stand until this study landed.

What we can already measure today, and what we are building next

Pedisteps’ 40 plantar pressure sensors per insole, sampling at 10 Hz, capture the full center-of-pressure trajectory through a sit-to-stand. This is everything an instrumented force plate would measure, except the patient is wearing the sensors at home, every day, in their kitchen, in their living room, getting up to answer the door.

From that signal stream we can derive – and in current product roadmap order – the following:

Already in the platform:

• Plantar pressure distribution and per-foot loading asymmetry during stance phase, validated in the ALYN Hospital stance-phase study
• Real-time weight-bearing alerts when load on a healing limb crosses a clinical threshold
• Event log of weight-bearing transitions, including chair-to-stand and stand-to-chair

Within reach of the existing sensor stream – these are the new analyses we should be running on data we already collect:

• Sit-to-stand transition detection from pressure signatures (the unloading-to-loading pattern is highly distinctive)
• Time-to-stand measurement for each individual transition, accumulated across the day
• Five-stand bursts when the patient performs them naturally, with automatic timing
• Asymmetry index during the rise phase, which is a known early marker of weakness or pain on one side
• Time-of-day patterns, since morning stiffness vs evening fatigue tell different stories
• Trend detection – is the patient’s median sit-to-stand time drifting upward week over week?

On the AI gait model roadmap for Summer 2026:

• Automated classification of compensation strategies: arm push, forward lean, momentum swing, multi-attempt rises
• Estimated chair-stand-test equivalent score from passively collected daily transitions, so we can give therapists a clinical-grade SPPB-comparable number without the patient performing a formal test
• Predicted decline curves with confidence intervals, the same way Mobileye predicts a collision risk envelope

The training side: VRsteps Family Wellness and Body Control mode

Measurement alone is not the product. The reason the JAMA finding matters is that chair stand performance is trainable. The American Heart Association’s 2024 Scientific Statement on resistance training, cited in the same paper, recommends muscle-strengthening activity at least twice a week. In the OPACH cohort, women not meeting aerobic guidelines still had lower mortality if their muscular strength was higher. This decouples strength training from cardio in a way that is operationally important for our elderly population, many of whom cannot reach 150 minutes of MVPA per week but can absolutely do sit-to-stand sets at home.

The VRsteps Family Wellness app, in Body Control mode, is positioned to be the home training engine for exactly this. The patented 3D avatar mirrors the patient’s actual body movement via Pedisteps and Tzora device inputs. We can build, and should prioritize building, a sit-to-stand training program with the following properties:

• The avatar performs each rep alongside the patient, providing the motivational and pacing cue that bare repetition lacks
• The patient’s actual sit-to-stand cadence drives the avatar – if they push off with their left leg dominantly, the avatar shows that asymmetry, externalizing a problem the patient cannot see themselves
• Reps are counted, timed, and scored from the insole signal, not from accelerometer estimates
• Progressive overload is automatic: as 5 reps become easy, the program asks for 7, then 10, then introduces tempo manipulation (slow eccentrics) and pauses
• The therapist dashboard shows the trainable progression alongside the passive monitoring data – the same signal stream serves both surveillance and intervention

This is the closed loop that distinguishes us from fitness trackers and from app-only competitors. The sensor knows what the patient did. The avatar shows the patient what they should do next. The therapist sees both.

A specific product proposal: the “Daily Stand Score”

If I were prioritizing a single new feature for the next release cycle that builds directly on this research, it would be this:

Daily Stand Score – a passively computed, 0-to-100 score derived from every sit-to-stand transition the patient performed that day. The score combines median rise time, asymmetry, attempt-count (did they try and fail and try again?), and trend slope against the patient’s own 30-day baseline. It surfaces in the patient app as a single tile next to the Pedisteps gait score, and in the therapist dashboard as a longitudinal chart with an SPPB-equivalent overlay.

The clinical pitch writes itself. A therapist gets the equivalent of a chair stand test administered hundreds of times per month per patient, without the patient ever walking into the clinic. The patient gets a measurable, trainable target that takes 30 seconds a day to improve. The family member gets an interpretable number. And we, as a platform, get a new piece of evidence in conversations with clinics that we are not just another remote monitoring tool – we are running a continuous version of one of the most validated mortality predictors in geriatric medicine.

What this study changes in how we talk about the product

For the clinic outreach work, the JAMA paper gives us a new line that I think lands harder than the FRAT framing alone:

The single best non-imaging predictor of all-cause mortality in older women is how fast they can stand up from a chair five times. We measure that, automatically, every day, in their living room.

For the families-with-elderly-parents segment, the framing is simpler and more emotional:

You probably do not know how long it takes your mother to stand up from her chair. She knows, and she is hiding it. We can show you the number, every day, without asking her to perform.

For the post-surgery segment, the connection is mechanical:

Recovery is not just about how far you walk. It is about whether you can stand up on your own. We measure both, and we alert you the moment one starts going backwards.

A few honest caveats

The JAMA study was in women only, postmenopausal, ambulatory at baseline, drawn from the Women’s Health Initiative. The associations are robust but observational – they show that chair stand performance predicts mortality, not that improving chair stand performance causes lower mortality. The interventional evidence for that causal link comes from a separate body of resistance-training literature, which is strong but is not what this paper measured.

There is also a real measurement question we will need to address: a chair stand test performed in a clinic with a stopwatch is standardized. A sit-to-stand transition detected from insole pressure during normal daily activity is not. The chair height, the patient’s clothing, whether they used the armrests, whether the dog was in the way – all of that varies. Our derived score will need its own validation against the formal SPPB chair stand test in a small clinical study. ALYN is the obvious partner. This is probably a 30-to-50-patient comparison study with a paired protocol: formal SPPB chair stand in the clinic, plus two weeks of Pedisteps daily wear, with regression of the derived passive score against the gold-standard timed test.

That is a study worth doing in 2026.

To explore how Pedisteps smart insoles and the VRsteps Family Wellness platform bring the predictive power of the chair stand test into the home — measured automatically, every day — visit vrsteps.io.

Frequently Asked Questions

Closing

For most of the history of geriatric medicine, the chair stand test has been a five-second clinical curiosity – useful but episodic. The JAMA paper reminds us that it carries enormous predictive weight. Continuous sensors and a patient-facing avatar app finally make it possible to operationalize what the test has been telling us all along: how someone stands up from a chair, every single day, is a signal worth listening to.

That signal is in our data already. We just need to read it.

---

David Drabkin is the Founder and CEO of VR STEPS Ltd., developing Pedisteps smart insoles and the VRsteps Family Wellness rehabilitation platform.

Reference: LaMonte MJ, Hyde ET, Nguyen S, et al. Muscular Strength and Mortality in Women Aged 63 to 99 Years. JAMA Netw Open. 2026;9(2):e2559367. doi:10.1001/jamanetworkopen.2025.59367