Osteoarthritis is one of the world’s fastest-growing causes of disability. With no disease-modifying treatments and global prevalence expected to reach one billion by 2050, the need for biologically informed therapies is urgent. A new study in Nature combines genetic data from almost two million individuals with functional analysis across joint tissues, creating the most detailed genomic map of osteoarthritis to date¹.

The authors identify more than 960 associated loci, including over 500 that are newly reported, and increase the number of candidate effector genes tenfold. They link these to early skeletal development, uncover shared pathways between pain, cartilage damage, and inflammation, and highlight existing drugs with repurposing potential. The findings shift how osteoarthritis is understood and where the next wave of therapies might begin.

Scale and Scope

This genome-wide association meta-analysis spans 1.96 million individuals across 87 cohorts, including almost 490,000 osteoarthritis cases. The authors identify 962 independent associations across multiple joint types, more than half of which are newly reported. Rare variant burdens show stronger effect sizes than typical common variants, particularly for genes such as ADAMTSL3, VIT, and IL11. While most data came from individuals of European ancestry, the authors report some improvement in diversity compared to earlier studies.

Despite the expanded scale, polygenic risk scores performed modestly. The best predictive model, for hip osteoarthritis with body mass index included, reached an area under the curve of 66%. However, the scale of the study allows for a new level of biological insight.

Pathways and Cell Types

Functional genomic analysis pinpoints specific skeletal cell types linked to osteoarthritis risk. Using single-cell data from early human development, the study finds signal enrichment in embryonic chondrocytes, osteoblasts, and tenocytes. These findings support the idea that adult disease may have its roots in developmental processes that shape joint form and function.

The 700 effector genes identified are connected to eight major biological pathways. These include extracellular matrix organisation, WNT, FGF, BMP, and TGF-β signalling, retinoic acid metabolism, circadian rhythm regulation, and glial cell processes. Many of these pathways influence joint structure, inflammation, and pain. 

From Genomics to Therapy

Around 10% of the effector genes encode proteins already targeted by approved drugs. Some, such as FGF18 and romosozumab, are already being tested in patients. Others point to new angles, such as timing drug administration to match circadian rhythms, or targeting glial cell processes to manage pain.

The inclusion of circadian pathways is particularly notable. Several risk loci influence genes that regulate time-of-day patterns in inflammation, cartilage repair, and pain signalling. Synchronising drug delivery with these biological cycles could improve treatment response in osteoarthritis, as seen in other conditions. It may also help explain why patients with similar levels of joint damage can experience very different symptoms, one of the biggest challenges in osteoarthritis care and clinical trial design.

Genetic risk profiles could guide patient selection in clinical trials. For example, alleles that influence COLGALT2 or TGFB1 expression may help identify patients more likely to benefit from collagenase or TGF-β modulators. The study also raises the possibility that long-term use of certain drugs could influence osteoarthritis risk, depending on dose, tissue, and direction of effect.

Caveats and Next Steps

This study confirms that self-reported osteoarthritis status is reliable for genetic discovery and that current imaging definitions offer limited additional power. However, clinical stratification remains a challenge, and deeper phenotyping will be needed to address the gap between structural damage and symptom severity. The authors also highlight the need for broader global representation to ensure findings translate across diverse populations.

Despite these limitations, the study expands the number of candidate effector genes by an order of magnitude and links developmental biology to lifelong joint disease, setting a foundation for both discovery and therapeutic development.

References

1. Hatzikotoulas K, Southam L, Stefansdottir L, et al. Translational genomics of osteoarthritis in 1,962,069 individuals. Nature. 2025;641(8065):1217-1224. doi:10.1038/s41586-025-08771-z