Osteoarthritis (OA) is a complex disease characterised by involvement of multiple tissues in the synovial joint. It is a leading cause of pain and disability in older adults. It has long been hypothesised that subchondral bone plays an important role in the development and progression of the disease. This thesis aims to investigate how subchondral bone measures of the knee such as subchondral bone mineral density (sBMD), bone size, and bone marrow lesions (BMLs) are associated with important disease outcomes in OA.

A population-based sample of older adults aged 50–80 years (51% female;​ mean age 62 years) participated at baseline and approximately 3.0 years later. sBMD was assessed using dual-energy x-ray absorptiometry (DXA). Cartilage volume, cartilage defects, bone area, and BMLs were determined using magnetic resonance imaging (MRI). X-ray was used to assess radiographic osteoarthritis [joint space narrowing (JSN) and osteophytes]. Blood samples were collected to assess vitamin D and high-density lipoprotein (HDL) cholesterol. Multiple questionnaires were used to assess pain, function, dietary intake, physical activity, sun exposure, and total knee replacement surgery.

The first study examined the cross-sectional correlates of sBMD and found that many factors were associated with sBMD including age, sex, body mass index (BMI), vitamin D, sun exposure, physical activity, and knee structural measures. The most novel structural measure was cartilage defects and a longitudinal study was required to address causality.

In the second study, bone area at baseline predicted cartilage defect development and cartilage volume loss. Baseline sBMD predicted cartilage defect development, which confirmed the cross-sectional findings above. These associations were independent of each other, indicating there are multiple mechanisms by which subchondral bone may lead to cartilage damage.

In the third and fourth study, 43% of participants presented with a BML at baseline with 25% improving in size and 24% worsening in size over time. Baseline BMLs predicted cartilage defect development and cartilage volume loss, suggesting BMLs may have a local effect on cartilage homeostasis. Baseline cartilage defects predicted BML progression, which may represent increased bone loading adjacent to defects. These results suggest BMLs and cartilage defects are interconnected and play key roles in knee cartilage volume loss;​ thus, both should be considered targets for intervention. BMLs also predicted total knee replacement surgery. A change in BML size was associated with a change in pain, only in those participants without radiographic osteoarthritis. Importantly a decrease in BML size was associated with a decrease in pain.

In the final study, baseline energy, carbohydrate and sugar intake (but not fat) were positively associated with a change in BML size. Baseline HDL cholesterol was negatively associated with BML change.

In conclusion, this series of related studies indicate that subchondral bone plays a significant role in OA pathogenesis. Features of the subchondral bone contribute to knee pain and predict important disease outcomes such as cartilage loss and joint replacement surgery. This suggests that subchondral bone is an attractive target for therapeutic intervention in OA. Future work should consider subchondral bone treatments when developing disease-modifying OA drugs (DMOADs).