Continuous multi-domain phenotyping of knee osteoarthritis identifies independent structural and symptom axes and clinically invisible subgroups at elevated arthroplasty risk
Authors' conclusion
Knee OA showed no stable discrete subtypes and instead varies continuously along largely independent structural and symptom axes.
Study context
- Population
- Knees with osteoarthritis (OA) from the FNIH OA Biomarkers Consortium (discovery) and the Osteoarthritis Initiative (validation)8,697 validation knees
- Participants
- 600 participants
- Outcome
- Total knee replacement (TKR)
- Follow-up
- 12 years
Methodological detailsMulti-Omics Factor Analysis across six domains (886 variables), clustering tested with four algorithms and bootstrap stability assessment, Clinical Shadow model projected into validation cohort
- Multi-Omics Factor Analysis across six domains (886 variables)
- Clustering tested with four algorithms and bootstrap stability assessment
- Clinical Shadow model projected into validation cohort
Main finding
As reported in the abstract
- Discrete clustering (GMM, Ward, HDBSCAN, k-means): no stable discrete subtypes; best silhouette score
- 0.178
- Structural and symptom axes continuous and largely independent, correlation r (p = 0.004)
- −0.117
- Multi-domain phenotypic variance accounted for by routine clinical assessment
- 19%
- Silent Progressors (high structure, low symptoms), discovery knees
- 23.5%
- Silent Progressors, validation knees
- 9.7%
- Total knee replacement (TKR) in Silent Progressors (discovery) across Multi-Omics Factor Analysis (MOFA) risk-score tertiles (p=0.0002)
- 10.6% to 44.7%
- Structure-symptom interaction, RERI on additive (risk-difference) scale, super-additive
- 0.139 (95% CI 0.013-0.247, p=0.02)
- RERI on odds-ratio scale (95% CI −1.97 to 12.08), inflated, shown for comparison
- 5.05
- AUC of factor model for TKR prediction, external validation
- 0.836
Observational data: associations, not effects.
Original publication
Continuous multi-domain phenotyping of knee osteoarthritis identifies independent structural and symptom axes and clinically invisible subgroups at elevated arthroplasty risk. Osteoarthritis Cartilage. Published 30 July 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →