AI Movement Model Connects Gait Patterns to Multisystem Health
A study in Nature Medicine analyzed movement sequences from 3,328 adults to detect markers across 18 body systems.
Analysis of Three-Dimensional Movement across Multiple Body Systems
Artificial intelligence (AI) models analyzing three-dimensional movement can detect physiological markers across cardiovascular, metabolic, and liver systems.
The peer-reviewed study, published in Nature Medicine and led by Adam Gabet, evaluated 3,328 adult participants from the Human Phenotype Project research cohort. Researchers collected three-dimensional movement data using camera-based Movement Lab technology developed by Newton Tech during five standardized motor assessments.
The research team processed 17,589 movement sequences through an AI gait foundation model designed to evaluate complex mechanics beyond basic metrics like step count.
Predictive Signals Beyond Body Mass Index and Age
Movement patterns provided predictive health data that remained statistically significant after adjusting for age, height, body mass index (BMI), and visceral fat.
The model identified significant associations with 1,887 of the 3,210 evaluated health measures spanning 18 distinct body systems. These measures included blood tests, liver imaging, bone density, sleep metrics, mental health markers, and physical frailty.
Accounting for established physical covariates improved predictions across all 18 body systems evaluated in men and 17 of 18 systems in women.
Shifting from Single-Metric Walk Tests to Multidimensional Tracking
Standard functional motor evaluations often compress physical capacity into a single number such as walking speed or the time required to complete a set distance.
The foundation model instead tracks joint-level coordination, mechanical effort, balance strategy, and compensatory movement patterns. Newton Tech is deploying this multi-angle analysis in clinical drug trials to give biotechnology sponsors a sensitive method to monitor disease progression and treatment responses.
Movement carries biological signals across organ systems that traditional clinic walk tests cannot capture.
Study Scope and Observational Constraints
The reported findings demonstrate observational correlations between movement characteristics and biological markers rather than direct causal mechanisms.
Newton Tech co-authored the paper and provided the proprietary Movement Lab technology used for data collection and processing. In addition, the system currently operates in clinical research cohorts and pharmaceutical trial environments rather than routine primary care settings.
Individuals tracking metabolic health cannot yet access this specific foundation model during standard clinical checkups.
Related Articles
Study Maps Hypothalamus Shifts in Alzheimer's Disease
Researchers mapped hypothalamic subregions in Alzheimer's disease, finding distinct structural...
Read article RecoveryCluster Headache and Burnout Symptoms Linked in New Study
A Karolinska Institutet study finds two-thirds of patients in an active cluster headache phase...
Read article RecoveryBrain Waste Clearance Waves During Sleep Decline in Older Adults
MIT researchers found cerebrospinal fluid waves that clear brain waste during deep sleep are...
Read article RecoveryOral GLP-1 Orforglipron Reduced Sleep Apnea Events in Trials
Daily oral orforglipron reduced sleep apnea episodes by 18 events per hour in adults with excess...
Read articleWant help fixing your sleep for real?
Opt Health treats sleep as a measurable system — cortisol rhythm, thyroid, testosterone, inflammation — and works the biology behind bad sleep with a physician, not a sleep-hygiene checklist.
See Opt Health's sleep program →Topic updates
Get the weekly recovery roundup
Sleep, HRV, fatigue, burnout, cortisol rhythm, sleep apnea, overtraining, and stress physiology.
Understand Your Health Better
Our free assessment maps your symptoms to the most relevant guides, tools, and product recommendations.
Take the Free Assessment →Free · Takes 5 minutes · Instant results