Correlates of knee bone marrow lesions in younger adults

Subchondral bone marrow lesions (BMLs) play a key role in the pathogenesis of osteoarthritis (OA) and are associated with pain and structural progression in knee OA. However, little is known about clinical significance and determinants of BMLs of the knee joint in younger adults. We aimed to describe the prevalence and environmental (physical activity), structural (cartilage defects, meniscal lesions) and clinical (pain, stiffness, physical dysfunction) correlates of BMLs in younger adults and to determine whether cholesterol levels measured 5 years prior were associated with current BMLs in young adults.

Subjects broadly representative of the Australian young adult population (n = 328, aged 31-41 years, female 48.7 %) underwent T1- and proton density-weighted fat-suppressed magnetic resonance imaging (MRI) in their dominant knee. BMLs, cartilage defects, meniscal lesions and cartilage volume were measured. Knee pain was assessed by Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and physical activity was measured by the International Physical Activity Questionnaire (IPAQ). Cholesterol levels including high-density lipoprotein (HDL) were assessed 5 years prior to MRI.

The overall prevalence of BML was 17 % (grade 1: 10.7 %, grade 2: 4.3 %, grade 3: 1.8 %). BML was positively associated with increasing age and previous knee injury but not body mass index. Moderate physical activity (prevalence ratio (PR):0.93, 95 % CI: 0.87, 0.99) and HDL cholesterol (PR:0.36, 95 % CI: 0.15, 0.87) were negatively associated with BML, while vigorous activity (PR:1.02, 95 % CI: 1.01, 1.03) was positively associated with medial tibiofemoral BMLs. BMLs were associated with more severe total WOMAC knee pain (>5 vs ≤5, PR:1.05, 95 % CI: 1.02, 1.09) and WOMAC dysfunction (PR:1.75, 95 % CI: 1.07, 2.89), total knee cartilage defects (PR:2.65, 95 % CI: 1.47, 4.80) and total meniscal lesion score (PR:1.92, 95 % CI: 1.13, 3.28).

BMLs in young adults are associated with knee symptoms and knee structural lesions. Moderate physical activity and HDL cholesterol are beneficially associated with BMLs; in contrast, vigorous physical activity is weakly but positively associated with medial tibiofemoral BMLs.