Overview, Vol 9, Issue 2

Calcium, vitamin D, both – yes – no – maybe

There is little convincing evidence that calcium reduces vertebral and nonvertebral fractures because there have been no replicated and credible trials (randomized, double-blind, placebo-controlled, large samples, prolonged follow up, few dropouts) demonstrating a reduction in vertebral or nonvertebral fractures in children, women or men. Nevertheless, the believers keep believing data if it suits their opinion and criticize the data if it does not. This is dangerous, particularly when it comes from a respected authority or prestigious journal.

Surrogate endpoints of fracture are used and inferences such as rise in BMD are made even though a change in BMD is a poor predictor of fracture risk reduction. Lambert et al report a randomized trial of calcium (792 mg/d) with follow up 2 years after supplement withdrawal in 96 girls with low calcium intakes (mean: 636 mg/d). Compared with the controls, the supplemented group showed greater gains in BMD. Resorption markers and parathyroid hormone were lower. After 42 months, gains in BMD and differences in bone resorption markers were no longer evident. Calcium effects are short-lived because remodelling is suppressed and then reversed on supplement withdrawal. The question is whether the structural effects produced by remodelling suppression might reduce fracture risk while treatment continues. Am J Clin Nutr 2008;87:455-62

Zhu et al report that among 120 community-dwelling women aged 70-80 years given calcium 1200 mg/d, calcium with 1000 IU/d vitamin D2 or double placebo in a randomised controlled double-blind trial, hip BMD was preserved in treated groups and not in controls during 5 years. When split by the median, benefits were seen in those with baseline 25OHD levels below the median (68 nmol/L). At year one, Ca and CaD groups had lower alkaline phosphatase and lower urinary DPD/Cr ratio. At 5 years, this suppression was maintained in the CaD group. CaD reduced PTH at 3 and 5 years in those with baseline PTH above the median. A lot of subgroup analyses were undertaken. The authors infer that addition of vitamin D to calcium has long term beneficial effects on BMD, but does that mean calcium alone does not? J Clin Endocrinol Metab 2008;93:743-9

Prince et al report that, in a 1-year double-blind, randomized trial of 302 community-dwelling ambulant women aged 70-90 years with a serum 25-hydroxyvitamin <24.0 ng/mL and a history of falling, ergocalciferol at 1000 IU/d reduced the risk of having at least 1 fall after adjustment for height (OR, 0.61; 95% CI, 0.37-0.99). Ergocalciferol reduced the risk of a first fall in winter and spring, but not in summer and autumn, and reduced the risk of having one but not multiple falls. There were no data concerning fractures reported (Arch Intern Med 2008;168:103-8). Richy et al report that in 14 trials including 21,268 subjects, vitamin D analogs provided a lower risk for falling than native vitamin D: RR = 0.79 (0.64-0.96) vs. 0.94 (0.87-1.01) (intergroup difference P= 0.049). Calcif Tissue Int 2008;82:102-07

Van Schoor et al studied 1311 community-dwelling older men and women. During 6 years, 115 persons (8.5%) had one or more osteoporotic fractures. serum 25(OH)D cut point of 12 ng/ml gave the best discrimination between persons with and without fractures. The lowest percentage of fractures (5.6%) was found above 30 ng/ml. After adjustment for age, serum 25(OH)D below or equal to 12 ng/ml was associated with an increased fracture risk in the youngest (HR = 3.1; 95% CI: 1.4-6.9), not the oldest age group (HR = 1.3; 95% CI: 0.7-2.2). Cut points of 25(OH)D (<10 ng/ml, 10-19.9 ng/ml, 20-29.9 ng/ml, ≥30 ng/ml) were not associated with fractures. Bone 2008;42:260-6

Looker and Mussolino studied 1917 men and women over 65 yr of age. There were 156 incident hip fractures. Serum 25(OH)D exceeding 60 nM was related to hip fracture risk. For example, the multivariate-adjusted RR was 0.64 (0.46-0.89) among individuals with serum 25(OH)D >62.5 nM compared with those with values below this level. The multivariate-adjusted RRs for the second, third, and fourth vs. the first quartile were 0.50 (0.25-1.00), 0.41 (0.24-0.70), and 0.50 (0.29-0.86), respectively. Serum 25(OH)D was related to a lower hip fracture risk. J Bone Miner Res 2008;23:143-50

It’s not calcium in milk anyway – angiogenin

Morita et al report that a factor responsible for inhibiting osteoclast-mediated bone resorption is present in the basic protein fraction of bovine milk (milk basic protein, MBP). The purified bovine angiogenin inhibited the pit forming activity of both unfractionated bone cells and purified osteoclasts in a dose-dependent manner, and the inhibitory activity was suppressed by anti-bovine angiogenin antibody. The inhibitory activity was confirmed in mice in vitro and in vivo. Treatment of osteoclasts with bovine angiogenin resulted in an impairment of the formation F-actin ring and a reduction in the mRNA levels of TRAP and cathepsin K. Bone 2008;42:380

Risedronate and reduced osteoclastogenesis

Bisphosphonates reduce bone resorption by inducing osteoclast apoptosis. D'Amelio et al examined the influence of risedronate on the formation of osteoclast precursors and cytokine production. In 25 patients treated with risedronate 5 mg/d, there was a reduction in the number and degree of differentiation of osteoclast precursors, osteoclast formation, vitality and activity, in the level of RANKL and TNF in cultures and of TNF and OPG in serum, and no changes in the group treated with calcium and vitamin D. J Bone Miner Res 2008;23:373-9

Sustained fracture risk reduction and antiresorptives

Few anti-fracture efficacy trials extend longer than 3 years with retention of at least the majority of the cohort was originally allocated to treatment or placebo. Lack of controls has obvious implications; nothing can be said about fracture rates without controls – have they decreased during treatment, remained unchanged or increased ……. relative to controls. We certainly don’t mean relative to pre-treatment fracture rates in the same cohort because no study has ever been done to measure fracture rates for say 1-3 years before treatment, then compared fracture rate during treatment in year 1, 2, and 3 with the pretreatment period.

As shown in the figure, when an antiresorptive is given, the resorptive cavities excavated and present before treatment go to completion with the bone formation phase of the remodelling cycle. As the bone tissue deposited mineralizes, BMD rises. This increase is unopposed by the birth of the same numbers of new remodelling units because the antiresorptive has reduced this birth rate. Fracture rates continue in control and treated subjects, but less frequently in treated subjects. The precise morphological basis for the reduced progression of bone fragility is uncertain because it has not been studied.

As time goes by, steady state is achieved and the rise in BMD continues slowly, but the morphological basis for the rise changes. Now the increase in BMD is the result of secondary mineralization (crystal enlargement) of the newly deposited bone that has undergone primary mineralization. Concurrently there is continued secondary mineralization of the existing bone tissue which has not been recently remodelled. The total bone tissue may be decreasing as remodelling with the negative BMU balance continues to remove bone from bone slowly (because remodelling is suppressed during treatment). So bone tissue density increases in a tissue that is decreasing. The net effect is a rise in BMD as the densitometer does not ‘see’ tissue mass decreasing.

What is in question is whether the fracture risk reduction seen in many studies within the first 12-36 months is sustained. In most clinical trials there is a dropout rate of about 10% per year, so after 5 years half the sample is lost. Has randomization been violated? It is very difficult to determine whether the Kaplan-Myer curves continue to diverge, which means the risk reduction is maintained. If they become parallel, this suggests fracture incidence is the same in the treated and control group; and if the lines converge, this suggests the fracture rates are higher under treatment than without treatment.

Loss of subjects in a trial potentially violates randomization. Randomization ensures that known and unknown covariates influencing fracture (independent of the drug) are equally present in treatment and placebo arm of the trial. Is this potential important? Of course it is. Say healthy people in the placebo arm drop out, the fracture rate will be higher and exaggerate the benefit of the drug. If sicker people drop out of the placebo arm, then fracture rates will be lower and the benefit of the drug will be obscured. A range of scenarios result when dropouts occur in the treatment arm. Is this corrected by bringing the last measurement point forwards or doing other analyses? If the design and execution of a trial are flawed, nothing can save it, nothing can give it credibility, not even statistics.

Watts et al report that patients who received risedronate 5 mg daily (N=398) or placebo (N=361) during the VERT-NA study stopped therapy after 3 years but continued taking vitamin D and calcium for one year. In the year off treatment, BMD decreased but remained higher than baseline and placebo at the spine and hip. Urinary NTX and BSAP decreased with treatment, were not different from placebo after 1 year off treatment. The incidence of new morphometric vertebral fractures was 46% lower in the former risedronate than former placebo (RR 0.54, 0.34, 0.86). Osteoporos Int 2008;19:365-72

Ensrud et al studied 10,101 postmenopausal women ≥55 yr of age with coronary heart disease. Women received 60 mg raloxifene or placebo daily for a median of 5.6 yr. No risk reduction for nonvertebral fractures was observed but women treated with raloxifene had a lower risk of clinical vertebral fractures (64 vs. 97 events; HR, 0.65, 0.47-0.89) (J Bone Miner Res 2008;23:112-20). This is probably one of the longest follow up studies reported and provides compelling data regarding the anti-vertebral fracture efficacy of this drug and the lack of it for nonvertebral fractures.

Strontium ranelate under the microscope

As illustrated in the figure, ideally a drug should increase periosteal apposition, reduce the volume of bone resorbed by each BMU and increase the volume of bone formed by each BMU so that the net effect is a positive bone balance. If the bone balance is positive, it is of interest to make bone remodelling rapid so that each remodelling event deposits bone on bone to help reconstruct the skeleton. If bone balance is not restored by the BMU, then the drug should also slow remodelling because each remodelling event robs bone of bone. If there is a positive balance produced by the BMU, then it is of interest for the drug to increase the rate of remodelling or at least maintain it.

Arlot et al report that in 141 transiliac bone biopsies from 133 postmenopausal osteoporotic women, 49 biopsies after 1-5 yr of 2 g/d strontium ranelate and 92 biopsies at baseline or after 1-5 yr of placebo, strontium ranelate was associated with 9% higher mineral apposition rate in cancellous bone. Osteoblast surfaces were 38% higher and 3D analysis of 3-yr biopsies suggested 18% higher cortical thickness and 14% higher trabecular number with a 22% lower structure model index and 16% lower trabecular separation with no change in cortical porosity. J Bone Miner Res 2008;23:215-22

These are encouraging results but they are cross-sectional, so ‘change’, ‘increased’ or ‘decreased’ should be ‘difference’, ‘higher’ or ‘lower’, respectively. Activation frequency, the measure of the birth rate of new remodelling units did not change in this study – it decreases with antiresorptive agents and increases with PTH. The importance of the remodelling rate depends on what the net effect of the volumes of bone resorbed and deposited by each BMU during the life cycle as described above.

Bonnelye et al studied primary mouse osteoblasts and osteoclasts derived from calvaria and spleen cells. Strontium ranelate continuously or during proliferation or differentiation phases stimulated osteoblast formation. After 22 days of continuous treatment, expression of the osteoblast markers ALP, BSP and OCN increased, and was combined with an increase in bone nodule numbers. The number of mature osteoclasts decreased after treatment. Osteoclast resorbing activity was also reduced, with strontium ranelate being associated with a disruption of the osteoclast actin-containing sealing zone. Bone 2008;42:129-38

Seeman et al report that data was pooled from the Spinal Osteoporosis Therapeutic Intervention study (SOTI; n=1649) and the TReatment Of Peripheral OSteoporosis (TROPOS; n=5091) to evaluate the antivertebral fracture efficacy of strontium ranelate in women with lumbar spine (LS) osteopenia with any BMD value at the femoral neck (FN; N=1166) and in 265 women with osteopenia at both sites (intention-to-treat analysis). The women were randomized to strontium ranelate 2 g/d orally or placebo for 3 yr. In women with LS osteopenia, treatment reduced the risk of vertebral fracture by 41%. In women with osteopenia at both sites, treatment reduced the risk of fracture by 52%. J Bone Miner Res 2008;23:433-8

Osteonecrosis and bisphosphonates

Etminan et al report that for each case of osteonecrosis, 10 controls were randomly selected. The initial cohort consisted of 87,837 subjects. In the primary analysis, the adjusted RR for AON among bisphosphonate users was 2.87 (1.71-5.05). The adjusted RR for alendronate, etidronate, and risedronate were 2.87 (1.46-5.67), 2.43 (1.05-5.62), and 3.34 (1.04-10.67), respectively. There were no differences in RR for AON among current users and past users of bisphosphonates. J Rheumatol 2008 [Epub ahead of print]

Anabolic therapy – are we there yet?

Mukherjee et al report that Bortezomib (Bzb), a proteasome inhibitor used in multiple myeloma, induces MSCs to undergo osteoblastic differentiation, in part, by modulating transcription factor Runx-2 in mice. Mice implanted with MSCs showed increased ectopic ossicle and bone formation when recipients received low doses of Bzb. Bone formation increased and bone loss was rescued in a mouse model of osteoporosis. Tissue-resident adult stem cells in vivo can be pharmacologically modified to promote a regenerative function in adult animals. J Clin Invest 2008;118:491-504

Boonen et al report that 245 women with osteoporosis had 2 years teriparatide and were stratified to alendronate (n=107), risedronate (n=59), etidronate (n=30), non-bisphosphonate (n=49). Increases in bone formation markers occurred in all groups after 1 month teriparatide. Spine BMD increased while a transient decrease in hip BMD reversed. BMD change was similar in all prior antiresorptives. Duration of prior antiresorptive and lag time between stopping prior therapy and starting teriparatide did not affect the BMD response. Teriparatide induces positive effects on BMD and markers of bone formation in postmenopausal women with osteoporosis, regardless of prior exposure to antiresorptives. Clin Endocrinol Metab 2008;93:852-60

Adami et al report that following a year teriparatide 20 µg/day, women with osteoporosis were assigned to raloxifene 60 mg/day or placebo for a year, then a year of raloxifene. The raloxifene and placebo groups showed a decrease in spine (LS) BMD in year 2 (-1.0%, P=0.004; and -4.0%, P<0.001, respectively); the decrease was less with raloxifene. Open-label raloxifene reversed the LS BMD decrease with a placebo, resulting in similar decreases at 2 years: -2.6% (raloxifene-raloxifene) and -2.7% (placebo-placebo). At study end, LS and femoral neck (FN) BMD were higher than pre-teriparatide, with no differences between the raloxifene-raloxifene and placebo-raloxifene groups, respectively (LS: 6.1% vs. 5.1%; FN: 3.4% vs. 3.0%). Sequential raloxifene prevented rapid bone loss at the LS and increased FN BMD, whether raloxifene was started immediately or after a one-year delay following teriparatide. Osteoporos Int 2008;19:87-94

Fluoride won’t go away – should it ?

Vestergaard et al report the results of a meta-analysis including 25 studies. Spine BMD increased 7.9% and hip BMD by 2.1%. Overall, there was no effect on the risk of vertebral or nonvertebral fracture. Daily dose of ≤20 mg fluoride (152 mg monofluorophosphate/44 mg sodium fluoride) were associated with a reduction in vertebral (OR=0.3, 0.1-0.9) and nonvertebral (OR=0.5, 0.3-0.8) fracture risk. With a daily dose >20 mg fluoride equivalents, there was no reduction in vertebral or nonvertebral fracture risk (Osteoporos Int 2008;19:257-68). It’s a shame to discard a drug like fluoride which has an anabolic effect but, at least in the way it is used, is producing poor quality bone. Low dose fluoride may well be worth studying alone or in combination with an antiresorptive but who will fund this sort of study ?


Kanis et al report alendronate 70 mg weekly for 5 years was cost-effective for the primary fracture prevention in women with osteoporosis irrespective of age as was treatment of women with a prior fracture irrespective of BMD. NICE guidelines are misguided. Bone 2008;42:4-15

Lekander et al reported the cost-effectiveness of 50 year old women. Hormone therapy (HT) compared to no treatment was cost-effective for most subgroups of hysterectomised women; whereas for women with an intact uterus without a previous fracture, HT was dominated by no treatment. Fracture risks were the single most important determinant of the cost-effectiveness. Bone 2008;42:294-306

Compliance with therapy

Sinsky et al assessed how patient and provider compliance with osteoporosis guidelines varies when efficacy is presented as relative risk (35% RRR) vs. absolute risk reduction (1% ARR). Compliance fell when the expression of treatment benefit was switched from RRR to ARR for both patients (86% vs. 57% compliance; P<0.001) and physicians (97% vs. 56% compliance; P<0.001). Increasing drug copayment from 0% to 10% of total drug cost decreased patient compliance with CPGs from 80% to 57% (P<0.001) but did not impact physician compliance. J Gen Intern Med 2008;23:164-8

Bone loss before menopause

When I asked the late Harold Frost when ageing began he replied, “at birth”. It is likely that bone loss begins shortly after completion of growth. There are several studies suggesting this and a recent study from Larry Riggs sheds some interesting insights into this process. The authors studied an age- and sex-stratified sample (n=553) by QCT for up to 3 years. Substantial cortical bone loss began in middle life in women but began mainly after age 75 in men. Trabecular bone loss began in young adult women and men at all skeletal sites and continued with acceleration during perimenopause. Women experienced 37% and men experienced 42% of their total lifetime trabecular bone loss before age 50 compared with 6% and 15%, respectively, for cortical bone. Median rates of change in trabecular bone (%/yr) were -0.40, -0.24, and -1.61 in young adult women and -0.38, -0.40, and -0.84 in young adult men at the DR, DT, and LS, respectively (all p<0.001). J Bone Miner Res 2008;23:205-14

The question is, what are the effects on bone strength? If remodelling is slow and the loss of bone is due more to reduced bone formation than increased rate of remodelling or increased volume of bone resorbed in each remodelling unit, then the same loss of bone produces less loss of strength when bone loss proceeds, producing thinning of trabeculae rather than loss of connectivity. In addition, as periosteal apposition continues, any loss of bone from the endocortical surface may be offset by the independent and continued gain of bone on the periosteal surface. The authors reported little change in cortical bone, but they did not report marrow cavity size, which would be of interest.

Ursus arctos horribilis and remodelling balance

McGee et al report cortical bone turnover during hibernation is balanced, bone formation and resorption replace the same volume of bone removed by each BMU in grizzly bear femurs, which avoids bone loss. Hibernating grizzly bear femurs were less porous and more mineralized and did not have changes in cortical bone geometry or mechanical properties. Activation frequency was 75% lower but mineral apposition rate was unchanged, so turnover decreases but osteons continue to refill at normal rates. Grizzly bears prevent bone loss during disuse by decreasing bone turnover and maintaining balanced formation and resorption, which preserves bone structure and strength. Bone 2008;42:396-404

Osteocytes – damage prevention and removal

This osteocytic-canalicular system functions in damage prevention by orchestrating adaptive remodelling, and damage removal by orchestrating reparative remodelling. Osteocytes detect strain and initiate modelling and remodelling to adapt bone’s material properties and structural design to offset the strain that will otherwise damage bone. Adaptation can be viewed as a damage-prevention mechanism. The change in bone size, shape and mass distribution during growth achieved by modelling and remodelling is successful adaptation; it is damage prevention by pre-emptive modification of structural strength in response to increasing stresses imposed by growth.

Microdamage, when it accumulates, compromises bone strength and must be removed. The second important function of the osteocyte is the detection of damage and initiation of focal remodelling to remove and replace damage with new bone. The osteocyte is pivotal in detecting damage and initiating remodelling and in adapting bone structure to its loading circumstances. Sclerostin, the protein product of the Sost gene, inhibits bone formation and is found nearly exclusively in osteocytes, the cell type that has is implicated in sensing and initiating mechanical signalling. Osteocytes control mechanotransduction by adjusting sclerostin (Wnt inhibitory) signal output to modulate Wnt signalling in the effector cells.

Robling et al report that Sost transcripts and sclerostin protein levels were reduced by ulnar loading. Portions of the ulnar cortex receiving a greater strain stimulus were associated with a greater reduction in Sost staining intensity and sclerostin-positive osteocytes than were lower-strain portions of the tissue. Hindlimb unloading yielded an increase in Sost expression in the tibia. Modulation of sclerostin levels allows osteocytes to coordinate regional and local osteogenesis in response to increased mechanical stimulation, perhaps via releasing the local inhibition of Wnt/Lrp5 signalling. J Biol Chem 2008;283:5866-75

The purpose of modelling and remodelling during adulthood is to maintain bone strength achieved during growth but in accordance with the prevailing loading circumstances. Part of the notion of maintaining strength is the detection and removal of damaged bone. The osteocyte plays a pivotal role in bone modelling and remodelling by sacrificing itself. Microcracks sever osteocyte processes in their canaliculi, producing osteocyte apoptosis. Osteocyte apoptosis is likely to be one of the first events signalling the need for remodelling.

In vivo, osteocyte apoptosis occurs within 3 days of immobilization and is followed within 2 weeks by osteoclastogenesis. In vitro, death of the osteocyte-like MLO-Y4 cells induced by scratching results in the formation of TRACP positive (osteoclast-like) cells along the scratching path.

You et al report that osteocytes cocultured with osteoclast precursors support osteoclast formation and activation. Mechanical stimulation of MLO-Y4 osteocyte-like cells decreases their ability to facilitate osteoclastogenesis, suggesting soluble factors are released by mechanically stimulated MLO-Y4 cells that inhibit osteoclastogenesis. Soluble RANKL and OPG were released by MLO-Y4 cells and the expressions of both are mechanically regulated. Mechanical loading decreases the osteocyte's potential to induce osteoclast formation by direct cell-cell contact. Bone 2008;42:172-79


Bone’s material composition and structural design determine its strength. These two components interact and adapt to ensure whole bone strength is appropriate to the loading requirements. Changes in one trait can result in adaptive changes in another so the whole bone strength is maintained. There are many examples of this, but one of the best known is the MOV13 mouse model. The mutant produces abnormal collagen and the material abnormality is compensated for by greater periosteal apposition, which offsets the loss of strength (Bonadio et al. J Clin Invest 1993; 92:1697-1705).

Tommasini et al measured slenderness (area/length) and tissue level mechanical properties from tibias from 14 female (22-46 yr old) and 17 male (17-46 yr old) donors. Ash content correlated negatively with slenderness and marrow area indicating that slender bones were constructed of tissue with higher mineralization. Slender tibias were compensated by higher mineralization and a greater area fraction of bone suggesting that bone adapts by varying the relative amount of cortical bone within the diaphysis and by varying matrix composition. J Bone Miner Res 2008;23:236-46

Bone loss determines structure determines bone loss

Bone remodelling is surface based so that higher remodelling on trabecular bone than cortical bone is partly the result of trabecular bone being fashioned with more surface than cortical bone – it has a higher surface to volume ratio. As remodelling events in adulthood usually remove bone from bone (because the volume of bone removed during the resorptive phase of a remodelling event is greater than the volume of bone deposited), there is a change in structure that accompanies this loss of bone.

Net loss of bone occurs on bone surfaces, so trabeculae thin; and when vigorous enough remodelling causes perforation, so the surface disappears and remodelling intensity on the trabecular surface decreases or stops. In cortical bone, remodelling produces intracortical porosity which increases the amount of surface within cortical bone, making it look like trabecular bone so that the intensity of remodelling in the intracortical compartment increases. So this is a self defeating process, this change in structure within cortical bone – trabecularizing it increases bone remodelling, increases bone loss and structural decay – hence the title.

Squire et al report that 21 days of unloading produces greater trabecular bone loss in the distal femur and proximal tibia in the metaphyses than in the epiphyses and 2-fold greater in females than in males. Disuse-induced changes were also greater in trabecular than in cortical bone. Bone loss was inversely related to baseline bone volume fraction (R2 = 0.51 for females and 0.43 for males) and directly related to baseline bone surface to volume ratio (R2 = 0.69 for females and 0.60 for males). Trabecular bone loss correlated osteoclast surface to bone surface ratios. Baseline bone morphology modulates bone loss; anatomical regions with high surface-to-volume ratios, and high levels of osteoblastic and osteoclastic activity are particularly susceptible to disuse. Bone 2008;42:341-9

Time to see architecture not shadows

Boutroy et al report that in 33 postmenopausal women with a prior wrist fracture, areal and volumetric densities, cortical thickness, trabecular number, and mechanical parameters were associated with wrist fracture. Five independent components explained 86.2% of the total variability of bone characteristics. The first component included FE-estimated failure load, areal and volumetric BMD, and cortical thickness, explaining 51% of the variance with an OR for wrist fracture = 2.49. The second component included trabecular architecture, explaining 12% of the variance, with an OR=1.82. The third component included the proportion of the load carried by cortical vs. trabecular bone, assessed by FEA, explaining 9% of the variance, and had an OR=1.61. Thus, the proportion of load carried by cortical vs. trabecular bone seems to be associated with wrist fracture independently of BMD and microarchitecture. J Bone Miner Res 2008;23:392-9

Cortical thickness is the net result of periosteal apposition and endocortical resorption, the absolute and relative movement of these surfaces during growth and ageing determine the total bone cross-sectional size, its external shape and the distance the cortical mass is placed from the neutral or long axis of the long bone. This 3D structural organisation determines bone strength. Bone strength cannot be understood using bone densitometry.

Lauretani et al report that 345 men and 464 women, 21 to 102 years of age, had tibial QCT measured during 6-yr. Periosteal apposition was higher in younger than in older men; whereas in women, the rate of apposition was homogenous across age groups. The age-related medullary expansion was higher in women than men. In women, not men, endocortical resorption was not matched by periosteal apposition caused loss of cortical bone. Endocortical resorption causes bone loss in older women despite periosteal apposition. J Bone Miner Res 2008;23:400-8

Marshall et al report dimensions and vBMD in the femoral neck and shaft obtained using QCT in 3,305 men >65 yr of age in the Osteoporotic Fractures in Men (MrOS) study. All groups had similar femoral neck integral (total) volume. Blacks and Asians had 6% greater mean cortical volume as a percent of integral volume, integral vBMD was 6-10% greater, and trabecular vBMD was 33-36% greater than Whites. Shaft cross-sectional area was similar in Blacks and Whites, but smaller among Asians than Whites. Mean shaft cortical area was greater among Blacks but similar among Asians and Whites, resulting in mean cortical thickness being 5% greater among Black and Asian men. Blacks also had greater mean cortical vBMD in both the femoral neck and shaft. Blacks and Asians have features in the proximal femur that may confer advantages for bone strength. J Bone Miner Res 2008;23:121-30

Collagen crosslinking – the good, bad and ugly

Material of bone is composed of collagen containing crystals of calcium hydroxyapatite-like mineral. The mineral confers stiffness, the collagen confers toughness or the ability to absorb energy by deforming without cracking. If the collagen molecules are crosslinked with advanced gylcation products (AGEs) they lose their ability to deform, and so they cannot absorb energy which is, therefore, dissipated in the worst of all possible ways by causing failure of the material – cracking.

Yamamoto et al report that increased bone pentosidine is associated with its plasma levels and bone fragility in type 2 diabetics with and without VFs. Although BMD did not differ, pentosidine was higher in women with VFs than in those without VFs (0.0440±0.0136 vs. 0.0321±0.0118 µg/ml, p<0.001) (J Clin Endocrinol Metab 2008;92:1013-9). Shiraki et al report that in 432 Japanese elderly women followed for 5.2 years, 97 incident vertebral fractures occurred in 72 subjects. Urinary pentosidine was a predictor of vertebral fracture (hazard ratio, 1.33; 95% CI, 1.01-1.76, P=0.04). J Bone Miner Metab 2008;26:93-100

Viguet-Carrin et al report an in vitro model of young bovine cortical bone specimens incubated in a sugar (ribose - an inducer of AGEs). Pentosidine concentration increased in specimens incubated with ribose, an effect inhibited by AMG. Bone 2008;42:139-49

Byrjalsen et al report that collagen maturation measured as the ratio between ααCTX and ββCTX showed that bisphosphonate induced a collagen profile consistent with an older matrix with a 52% (alendronate) and 38% (ibandronate) reduction in the ratio between the two CTX isoforms vs. 3% and 15% with HRT or raloxifene, respectively. Antiresorptives had different effects on the endogenous profile of bone collagen maturation. Osteoporos Int 2008;19:339-48

Allen et al report neither alendronate nor risedronate altered the strength-density relationship compared to control. The energy absorption-density relationship was altered by alendronate, resulting in lower energy absorption capacity at a given aBMD compared to both controls (-22%) and risedronate (-14%). After adjusting for increased aBMD, vertebrae from animals treated with bisphosphonates had similar strength as those from untreated animals. Conversely, when adjusted for increased aBMD, alendronate reduced the energy required for vertebral fracture. Osteoporos Int 2008;19:95-9

Allen et al treated female beagles for 1 year with vehicle, risedronate, alendronate or raloxifene. Vertebral trabecular bone collagen isomerization (α/βCTX), enzymatic (PYD and DPD), and non-enzymatic (pentosidine) crosslinks. Bisphosphonates increased pentosidine (+34 to 58%) and PYD/DPD (+14 to 26%), and decreased α/βCTX (-29 to 56%), raloxifene did not. Bone turnover correlated to pentosidine (R = -0.664), α/βCTX (R=0.59), and PYD/DPD (R = -0.47). Osteoporos Int 2008;19:329-37

Morbidity and mortality

Vestergaard et al report 169,145 hip fractures in Denmark between 1977 and 2001. Compared to 524,010 controls, the cases had twice the prevalence of comorbidity (HR=2.26, 95% CI: 2.24-2.27). Adjustments for confounders changed little the excess mortality risk. The mortality after hip fracture was divided into an excess mortality of 19% within the first year following the fracture (relative survival = 0.81 compared to controls), and an excess mortality of 1.8% per year (relative survival 0.982) for every additional year following the fracture. The major causes of the excess mortality were due to fracture event complications (70.8% within the first 30 days). Osteoporos Int 2007;18:1583-93


Note from the Editor

The purpose of Progress in Osteoporosis is to provide the reader with a summary of the most important literature published in the preceding three to four months in the field of osteoporosis. Most reviews and original research are cited. In addition, summaries and figures are provided for readers who may not have easy access to all the specialist literature. The summaries are based on the contents of abstracts, which have been abbreviated to concisely convey the main theme. The contents of the abstracts and figures should be used only as a means of directing the reader to the original literature and should not be quoted verbatim or cited as a reference. The opinions expressed in the Overview are my own and do not necessarily reflect those of the International Osteoporosis Foundation.