The movement blind spot most people never notice

Think about the last time you walked to the kitchen, crossed a car park, or climbed a flight of stairs. You almost certainly weren't thinking about it at all — and that is precisely the problem.

Gait is one of the most automated behaviours the human body performs. Once learned, it drops below conscious attention and runs on habit. The result is that most people have no objective picture of how they actually move. They assume symmetry because nothing hurts. They assume efficiency because they arrive where they intended to go. But assumption is not measurement.

Compensation patterns — a fractionally shortened stride on one side, a subtle forward lean of the trunk, a hip that drops slightly with each step — can develop over months or years without ever producing a clear symptom. An old ankle sprain quietly changes how load travels through the knee. A desk-bound posture gradually flattens the lumbar curve and reduces hip extension. None of it announces itself; the body simply adapts, and the adapted pattern becomes the new normal.

Clinical observation helps, but only so far. What one practitioner notices in a patient's walk, another may miss entirely. The assessment is real, but it is not reproducible, not quantified, and not something a person can track week to week.

Professor Paul Lee's argument in Practical Regeneration (2026) is direct: movement should be something shaped deliberately, not something that simply happens to you. The gap between those two states — habit and intention — is a measurement gap. Close it, and you can begin to act on what you actually find.

Why gait speed is a clinical vital sign

Walking pace, it turns out, is one of the most information-dense readings a clinician can take.

A landmark 2011 meta-analysis by Studenski and colleagues — now cited nearly 6,000 times — pooled data from tens of thousands of older adults and found that gait speed predicted survival as reliably as assessments accounting for multiple chronic conditions. The thresholds are striking in their simplicity: a walking speed above 1.0 m/s is associated with above-average life expectancy, whilst a speed below 0.6 m/s signals meaningfully elevated risk. The gradient between those points is precise — every reduction of 0.1 m/s correlates with a 12% rise in premature mortality risk. In one of the study's most arresting findings, a 75-year-old's ten-year survival probability could be estimated from just three variables: gait speed, age, and sex. That estimate matched the accuracy of far more complex clinical workups.

The reason gait carries so much signal is that walking is a full-system output. It draws on cardiovascular reserve, neurological coordination, musculoskeletal integrity, and balance simultaneously. A decline in any of those systems tends to surface in how a person moves before it announces itself in other ways.

None of this requires specialist equipment to appreciate. But the implications are clear: if walking pace carries genuine biological signal about longevity and function, then measuring it accurately — rather than estimating it by eye — becomes a rational priority, not a clinical luxury.

How markerless capture turns standard video into biomechanical data

For most of the history of biomechanics research, capturing how a person moves required a dedicated laboratory: a room lined with infrared cameras, a subject dressed in a tight suit studded with reflective markers, and a technician to place each one precisely over bony landmarks. Accurate, yes — but impractical outside a handful of specialist centres.

Markerless motion capture takes a different approach. Rather than tracking physical objects attached to the body, computer-vision algorithms analyse the raw pixel patterns of standard video — frame by frame — to infer where joints are and how they are moving. Systems based on OpenPose-style pose estimation identify key anatomical landmarks from visual cues alone, then reconstruct their positions in three dimensions across hundreds of frames per second. No suit, no calibration session, no specialised environment.

The accuracy of these systems has now been assessed in peer-reviewed clinical settings. A 2023 study found intraclass correlation coefficients of 0.835 to 0.957 when comparing video-based markerless assessment against gold-standard 3D laboratory capture; Bland–Altman analysis showed no significant systematic errors. A 2024 validation study in MDPI Sensors extended this to depth-estimated 3D reconstructions with similar results.

Reliability is strongest for the metrics most relevant to functional health: walking speed, cadence, step length, and sagittal-plane joint angles. Some limitations persist for complex rotational data in three dimensions — an honest caveat worth noting, though not one that undermines the core clinical utility. For the movement qualities that map most directly to longevity and function, the technology is clinically comparable to systems that cost many times more and exist only in research laboratories.

What MAI Motion measures and how it scores movement

Consider what gait analysis actually revealed for Steve, a 49-year-old former runner described in Professor Paul Lee's Practical Regeneration. He arrived with recurring hamstring strains and lower back pain — plausible enough, all pointing at the injury sites. A video-based assessment told a different story: minimal hip extension, a shortened stride, and anterior pelvic tilt. The problem was not in his hamstrings. Six weeks of foot drills, hip stability work, and glute reactivation produced significant improvement — because the intervention addressed the actual mechanism, not the symptom.

That gap between apparent cause and measured cause is precisely what MAI Motion® is built to close. The system tracks 15 body keypoints at 120 frames per second, extracting step length, cadence, joint angles, and loading symmetry from standard video without sensors or specialist equipment.

Findings are structured through the C.R.A.F.T. framework — Control, Range of motion, Asymmetry, Functional loading, and Tolerance to load. The five dimensions are designed to surface not only what a person's movement pattern looks like, but where the body is compensating — which is often where the most actionable information lives.

All of this resolves into a single output: a Motion Age score, a functional biological age derived by comparing an individual's movement signature against population norms. It is a tracking tool rather than a diagnostic verdict — a number designed to shift as training and habits improve.

The baseline assessment runs in clinic at Harley Street (approximately 30 minutes), producing a full report encoded into the Regen OS AI dashboard. Re-scans run via the MAI Motion home app on the same pipeline, so monitoring becomes continuous rather than episodic. Movement data sits alongside the 32-marker blood panel in one longitudinal record, making changes over months visible rather than estimated. No independent peer-reviewed trial specific to MAI Motion's pipeline has yet been published; the system's accuracy draws on company validation and on a broader technology class with solid peer-reviewed support — the class whose clinical reliability was described in the preceding section.

Turning a scan into a deliberate movement practice

The difference between knowing you have a movement problem and actually changing how you move is, in practice, considerable. Measurement bridges that gap — but only when it is continuous rather than occasional.

Once the baseline scan has established a starting Motion Age and flagged the specific patterns most worth addressing, the home app runs the same pipeline between clinic visits. Cadence, loading symmetry, and stride length can be checked at regular intervals, with each result added to the longitudinal record. The week-on-week comparison is what shifts behaviour: someone who sees loading asymmetry narrowing over six weeks has evidence that a specific change worked. Someone who sees cadence plateau knows something still needs adjusting.

That precision matters. C.R.A.F.T. feedback produces targeted responses — adjusting stride length, redistributing load, or working on hip extension through a specific drill — rather than the generic instruction to move more. The intervention is proportionate to the finding, which is exactly what makes it actionable.

Regen PhD reports that most members see their Motion Age fall meaningfully below chronological age within 16 weeks; this comes from company data rather than an independent published trial, but it aligns with the timeframes in which targeted gait retraining typically produces measurable shifts in spatiotemporal parameters.

As Professor Paul Lee writes in Practical Regeneration, MAI Motion helps to 'intervene earlier, track improvement with confidence and design movement strategies that hold up over time.' That durability is the point. Movement assessed and adjusted deliberately tends not to drift back into the compensatory patterns that generated the problem — because the next scan will show whether it has.

Movement as a Physics pillar inside a larger system

Professor Paul Lee's Regeneration by Design organises the drivers of functional ageing into four interdependent pillars: Physics (load, posture, movement), Chemistry (nutrition, hormones, inflammation), Biology (sleep, gut, nervous system), and Time (repair windows, monitoring, early action). MAI Motion® sits within the Physics layer — but movement data only becomes fully legible when read against the others.

A gait asymmetry flagged in a Motion Age scan may reflect a Chemistry imbalance such as systemic inflammation or a hormonal shift affecting connective tissue. It may equally point to a Biology signal — neuromuscular control degraded by poor sleep. When movement trends sit alongside blood panel data in the Regen OS dashboard, patterns that look purely mechanical can be interrogated against biochemical context. That is the Time pillar made practical: months of overlapping, cross-referenced data rather than isolated readings.

onMRI™, Professor Lee's complementary AI imaging platform, adds a structural dimension — quantitative tissue biomarkers from MRI — so the picture extends inward alongside the functional and biochemical layers.

This integration is the core argument of both books: health is not a collection of separate metrics but a system, and intervening in one pillar without reading the others is working half-blind. The Studenski survival gradient holds precisely because gait speed already integrates the heart, lungs, nervous system, and musculoskeletal system in a single number. MAI Motion makes that cross-pillar signal visible, trackable, and — for the first time — something a person can deliberately act on rather than passively experience.