Is Bone Mineral Density Predictive of Fracture Risk Reduction?

Charles A. Cefalu


Curr Med Res Opin. 2004;20(3) 

In This Article

Other Mechanisms of Fracture Risk Reduction

The studies of osteoporosis therapies mentioned previously indicate that the loose correlation between fracture risk and BMD increase may be explained by other variables. Fracture risk is related to bone strength, which is derived from bone integrity. Bone density alone does not reflect bone integrity, and therefore, increasing BMD does not necessarily fully protect against fractures. Strength of bone is dependent on two main interdependent physical/structural factors: quantity and quality.[7] Bone quantity consists of density and size, which together constitute mass. The increased fracture risk associated with decreased BMD is well recognized, but there is also an association with both smaller bone diameter and decreased cortical thickness.[30,31] Bone quality consists of structural (microarchitecture and macroarchitecture) and material properties. The balance between bone production and resorption (turnover) also contributes to bone quality.

The interconnecting structure, or microarchitecture, of the bone contributes significantly to its strength.[7] The structure of trabecular bone, which includes prevalence, location, and connectivity of horizontal trabeculae, has a major impact on fracture risk independent of measures of bone quantity.[31,32] Historically, the necessity for bone biopsy to measure bone structure (microarchitecture) made routine study of bone strength difficult. Newer, noninvasive techniques, such as 3-dimensional magnetic resonance microimaging, microcomputed tomography, and direct image analysis, are now enabling some determinants of bone structure to be measured.[33,34] However, these techniques are still in development and are not yet accepted for clinical use. Recent data indicate that bisphosphonate therapy may improve bone structure. Bisphosphonates increase bone diameter and cortical thickness and decrease porosity, possibly by increasing periosteal bone formation.[35,36] There are emerging data both in minipigs[37] and in postmenopausal women [38]to suggest that risedronate preserves bone architecture, including bone volume; trabecular thickness, number, and connectivity; and bone surface/bone volume. A study in postmenopausal women suggested that alendronate increased wall thickness and tended to decrease erosion depth.[39] This effect is dissimilar to that observed with teriparatide. The increase in the remodeling space and cortical porosity seen within the first months of teriparatide administration does not seem to compromise bone strength.[40] Material properties of bone are also affected by the nature of bone matrix deposition and mineralization and by accumulation of unrepaired microdamage. The latter contributes to the loss of elastic properties, accumulates with age, and is theorized to contribute to fracture risk.[41] One factor that partially compensates for the loss of bone strength during the aging process is the periosteal apposition of new bone. A 19-year prospective analysis of women, all of whom were 48 years of age (premenopausal) at the time of enrollment, found that women lose bone density after menopause but have an increase in skeletal size as a result of periosteal apposition.[42]

In accelerated bone turnover, resorption by osteoclasts exceeds bone formation by osteoblasts, producing a general thinning of the trabeculae that can lead to irreversible loss of some trabeculae and to disconnections within the matrix, contributing to fracture risk.[43,44] It is not possible to routinely assess trabecular connectivity, but high levels of bone resorption markers can predict fracture risk in untreated individuals, and changes in markers can predict BMD response to antiresorptive therapy during treatment.[45,46,47,48,49] Recent studies of antiresorptive therapies for osteoporosis have focused on the relationship between bone turnover and fracture risk reduction.

Routinely measured biochemical markers of bone turnover found in blood and urine are proteins or other products of osteoblasts, which reflect bone formation, and collagen breakdown products, which reflect bone resorption.[45,46,50] Osteoblast products include osteocalcin (OC), alkaline phosphatase (ALP), and type I collagen propeptides (PINP or PICP). Collagen breakdown products include hydroxyproline (Hyp), hydroxylysines (Gal-Hyl, Glc-Gal-Hyl), pyridinolines (PYD, DPD), and urinary C-or N-terminal telopeptides cross-linked collagen type I (CTX, NTX), and the matrix metalloproteinase for CTX (CTX-MMP).

Pooled data from the VERT trials showed that as little as 6 months of therapy with risedronate 5 mg/day reduced the risk of clinical vertebral fracture compared with placebo (incidence: 0.1 vs 1%).[51] The early decrease in vertebral fracture risk with risedronate therapy may be explained by an early decrease in bone resorption. A subset of 693 women from the VERT trials had bone marker measurement at baseline and at least once during follow-up.[52] Urinary NTX decreased 51% and urinary CTX decreased 60% (p < 0.001 vs placebo for each) during the first 3-6 months of risedronate therapy (Figure 5). Vertebral fracture risk fell 75% at 1 year and 50% at 3 years in this subset, a magnitude similar to that reported for the whole group,[11,12] and was significantly associated with changes in NTX and CTX at 3-6 months (p < 0.001 for each measure). The direct relationship between change in bone resorption marker value and risk of vertebral fracture was not linear, because a threshold appeared to be reached after a 40% decrease in NTX and a 60% decrease in CTX, after which there was no further decrease in vertebral fracture risk. According to Cox regression analysis, approximately half of the treatmentrelated change in vertebral fracture risk reduction at 1 year and approximately two thirds of the risk reduction at 3 years was explained by changes in bone resorption markers at 3-6 months.[52] We are not aware of any alendronate studies monitoring changes in bone turnover markers and fracture risk. A subset of 2622 women from the MORE trial had bone marker measurement at baseline and after 6 and 12 months of participation.[53] Change in bone turnover (i.e., serum OC, serum bone-specific ALP, and urine CTX) during the first 6 and 12 months of raloxifene therapy was significantly related to risk of vertebral fracture during the 3-year trial duration, even after adjusting for baseline vertebral fracture status and BMD (p < 0.05). The change in BMD over 12 and 24 months was not correlated with fracture risk in any of the analyses.

Relationship between bone resorption marker changes and vertebral fracture incidence in a subset of patients from the VERT studies. (Reprinted from Eastell R, Barton I, Hannon RA, et al. J Bone Miner Res 2003[52])


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.