Loss of BMD linked to knee OA progression

View on the News

OA, BMD relationship needs further study

Dr. M. Kassim Javaid and Dr. Nigel K. Arden commented:

Understanding the association between OA and osteoporosis has proven to

be challenging because of the difficulty in disentangling the effects

of BMD on OA, and, conversely, the effects of OA on BMD and fracture


The answer might be found in the effects of bone size and

the BMD properties of local subchondral bone on progression of OA.

Patients with OA have been found to have larger bone size but not higher

volumetric BMD (vBMD), as measured by peripheral quantitative CT.

Another study that used CT to measure vBMD of the subchondral bone found

higher vBMD among patients with knee OA, Dr. Javaid and Dr. Arden


These factors may be intertwined by characteristics of bone

phenotype that cannot be assessed by measuring BMD, including bone

marrow lesions, metabolic activity, and bone turnover markers.


novelty of Dr. Lee and colleagues getting around this problem by

accounting for changes in femoral neck BMD over time helps to shed light

on the true contribution of BMD to OA progression, but it should be

noted that the study "does not address whether changes in [areal] BMD

are a cause vs. effect of cartilage loss. However it is likely the

relationship is bi-directional with bone altering chondrocyte and matrix

properties and vice versa," Dr. Javaid and Dr. Arden said.


also noted that the study by Dr. Lee and colleagues is limited by the

fact that only 13% of patients lost enough BMD to be deemed significant

and by the question of whether the knee OA of patients with a

Kellgren-Lawrence grade of 4 could progress.

Further studies will

need to focus on symptomatic progression as a primary outcome rather

than change in cartilage volume. A key randomized controlled trial

appeared to indicate that bisphosphonate treatment had no effect on OA,

but a cartilage-sparing effect of strontium ranelate has been confirmed

in a phase III trial on both symptoms and a structural end point for

knee OA (Ann. Rheum. Dis. 2013;72:179-86), Dr. Javaid and Dr. Arden noted.


Javaid and Dr. Arden, both with the University of Oxford (England), did

not report having any conflicts of interest. Their remarks were taken

from an editorial accompanying Dr. Lee’s study (Arthritis Rheum. 2013 March 12 [doi:10.1002/art.37924]).



Longitudinal loss of bone mineral density from the femoral neck was associated with prevalent knee osteoarthritis in an observational cohort study, lending support to the notion that osteoporosis treatments are worth further investigation for the prevention of osteoarthritis progression.

The study, led by Dr. Ji Y. Lee of the division of rheumatology at Tufts Medical Center, Boston, builds on previous research that loss of bone mineral density (BMD) is associated with the progression of radiographic joint space narrowing. The current study is the first to examine the relationship between BMD and knee osteoarthritis (OA) progression as measured by cartilage volume and thickness on MRI, providing a more sensitive measure of changes in knee cartilage (Arthritis Rheum. 2013 March 12 [doi:10.1002/art.37926]).

The study contrasts with previous research, which has shown a positive relationship between BMD and incident or prevalent radiographically measured knee OA. Other past studies have found no relationship between BMD and radiographic progression of OA or a paradoxical opposite relationship in which low BMD predicted radiographic progression, according to Dr. Lee and associates.

Because previous studies included patients who already had OA, there may have been selection bias (collider confounding) in which the variables of interest – BMD and cartilage volume and thickness – were affected by the same factors. Dr. Lee and colleagues hoped to avoid this selection bias by studying the effect of risk factors that change after disease onset – in this case, change in BMD.

The investigators analyzed a cohort of 127 patients with prevalent knee OA, defined as a Kellgren-Lawrence grade of 2 or more. The patients had at least two MRI scans over a 2-year period, but most had three scans: at baseline and at 1 and 2 years. The patients (41% men) had a mean age of 63 years and mean body mass index (BMI) of 30 kg/m2. Baseline BMD averaged across two femoral neck measurements was 0.95 g/cm2.

In multivariate linear regression models – adjusted for baseline values of age, gender, BMI, alignment status, and vitamin D treatment – BMD loss of 0.1 g/cm2 was associated with a 1.25% per year loss of cartilage volume. For patients who lost BMD at a rate considered to be significant (calculated by the investigators to be a loss of at least 4.7% from baseline), cartilage volume loss was 1.02% per year greater than for patients without BMD loss.

The models also showed that a BMD loss of 0.1 g/cm2 was associated with a significant loss of cartilage thickness at the tibia (0.028 mm/year). Those who lost at least 4.7% of BMD lost a mean of 0.021 mm in tibial cartilage thickness per year, compared with those who did not lose BMD, the investigators reported.

Baseline BMD, however, was not significantly associated with any cartilage outcomes in the study.

The biological mechanisms conjectured to link systemic BMD to cartilage loss in knee OA include the possibility that "BMD health might provide an environment that supports optimal subchondral bone turnover and remodeling in response to OA stressors, thus favoring joint stabilization" and the beneficial effect that optimal bone health might have on cartilage health via "humoral mechanisms," Dr. Lee and colleagues wrote. "Systemic BMD could also be a marker for a range of covariates that mediate or confound the relationship, such as systemic inflammation, circulating growth factors or hormones, physical activity, or frailty."

The original trial from which the observational cohort was derived, the Randomized Controlled Trial of Vitamin D for Knee OA, was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases. None of the authors had financial disclosures to report.

Next Article: