Bone mineral density is cross sectionally associated with cartilage volume in healthy, asymptomatic adult females: Geelong Osteoporosis Study

Abstract

Introduction

The association between osteoporosis and osteoarthritis (OA) is controversial. Although previous studies have shown total body, lower limb, spinal and knee BMD and knee cartilage volume to be positively associated, the relationship between other distant site-specific measures of BMD and other knee structures is unknown. The aim of this study was to determine the associations between BMD at eight skeletal sites, and knee structure in asymptomatic young to middle-aged females without any clinical signs of OA.

Methods

One hundred and sixty healthy, asymptomatic females (29–50 yr) underwent magnetic resonance imaging of the knee. BMD was measured at the spine, hip, total body and forearm by dual energy X-ray absorptiometry, and at the calcaneus by quantitative ultrasound. BMD was tested for an association with cartilage volume, defects, and bone marrow lesions (BMLs).

Results

Medial cartilage volume was positively associated with BMD at the spine, total body, femoral neck, and Ward's triangle (all p < 0.05), with non-significant associations in the same direction at the trochanter (p = 0.07). Findings in the lateral compartment were similar. The presence of medial cartilage knee defects were also associated with BMD at the spine; defects in the lateral compartment were associated with BMD at the forearm (both p = 0.05). BMD was not associated with the presence of BMLs. No associations were observed with calcaneus BMD.

Conclusions

Site-specific BMD is associated with cartilage volume at the knee in asymptomatic young to middle-aged adults, with the direction and effects trending in the same direction. The magnitude of changes correlates with clinically relevant changes. QUS defined calcaneus BMD, showed no associations with knee structure. Although systemic factors may underlie the association between knee cartilage volume and axial/lower limb BMD, these data suggest that common local, possibly biomechanical factors may also play a role.

Introduction

The association between osteoporosis and osteoarthritis (OA) is controversial. The relationship between these diseases is most often assessed by linking bone mineral density (BMD), as a surrogate marker for osteoporosis, and the radiological features of OA. Cross-sectional studies consistently suggest that BMD is associated with the presence of knee OA as measured by osteophytes [1], [2], joint space narrowing [3]. Similarly, longitudinal studies have shown higher BMD to be associated with incidence of knee OA in older women [4], and to predict greater joint space narrowing [5]. However, there are conflicting data regarding the association between BMD and OA, for instance a recent longitudinal study identified that BMD was positively associated with increased odds of developing joint space narrowing and Kellgren–Lawrence grade ≥ 2 in subjects without OA, but was not associated with progression in those with prevalent disease [6]. Furthermore, in those with OA, despite increased BMD, it is not clear that fracture risk is reduced, with studies providing conflicting results [7], [8]. These discrepancies may be related to subchondral bone attrition, which as a key feature of OA, has been associated with systemic BMD [9].

One of the defining characteristics of OA is articular cartilage loss. By the time the first signs of radiographic OA are present, even with grade 1 joint space narrowing, 10% of articular cartilage volume loss has already occurred [10]. The use of magnetic resonance imaging (MRI) enables the direct visualisation of cartilage, and allows the direct examination of knee structure, before radiographic changes are present, allowing the characterisation of the pre-radiographic structural changes of OA. Reduced cartilage volume, increased prevalence and severity of defects and increased tibial plateau area have been related to the severity of radiographic OA [11], [12]. Bone marrow lesions (BMLs) have been related to pain and increased disease progression, measured by cartilage loss [13]. Initial work has shown that cartilage volume of older subjects without OA is positively associated with total body BMD in both genders, and also with hip BMD for men [1]. It is possible that factors associated with reduced cartilage volume in asymptomatic individuals will increase the risk of OA onset.

Whilst previous studies have examined the relationship between BMD and cartilage volume [1], the relationships between BMD (at sites distant to the knee) and other knee structures have not been examined. Furthermore, the mechanisms behind the relationship between BMD and knee cartilage volume are not known; thus, examination of the association between site specific measures of BMD and components of knee structure in asymptomatic young to middle-aged adult females without clinical OA may help us to understand these mechanisms. Additionally, the use of peripheral measures of BMD to determine properties of the knee may be useful in community-based settings, due to quick, cheap and easy access.

The aim of this study was to determine the associations between BMD at eight skeletal sites, and knee structure, including cartilage volume, defects, and BMLs in healthy, asymptomatic young to middle-aged female subjects without any clinical signs of OA.