Education and debate

Changing perceptions in osteoporosis

Commentary: Bone density can be used to assess fracture risk

Changing perceptions in osteoporosis

Terence J Wilkin, professor. 

Plymouth Postgraduate Medical School, University of Plymouth, Plymouth PL4 8AA

twilkin{at}plymouth.ac.uk

A recent meta-analysis of 11 separate study populations and over 2000 fractures concluded that bone mineral density "cannot identify individuals who will have a fracture."¹ So why do we measure it? The question is worth asking when the number of dual energy x ray absorptiometry machines installed in the United Kingdom exceeds 130 (National Osteoporosis Society, personal communication), and the number of Medline citations incorporating the term dual energy x ray absorptiometry has grown from 26 in 1988 to 464 in 1997. The concept of "fracture threshold" has led to the recommendation that preventive treatment be given to women once their bone density lies (arbitrarily) more than 1 SD below the mean for premenopausal womena state of osteopenia, according to the WHO definition. But should we be managing osteoporosis by numbers?

	Osteoporosis and osteomalacia

Bone comprises a matrix framework on which mineral is deposited. Osteoporosis is caused by the disintegration of the matrix, and osteomalacia by a failure to mineralise it. When mineralised matrix disintegrates, calcium is inevitably lost. The negative calcium balance observed with matrix loss gave rise to erroneous beliefs that the calcium requirements of postmenopausal women were higher than those of premenopausal women and that osteoporosis could be prevented by calcium supplementation.² Calcium is crucial during the development of bone, but it cannot replace disintegrating matrix or prevent its loss.³ As Heaney has pointed out, calcium is a nutrient and not a drug, and the only disorder it can be expected to alleviate is calcium deficiency.⁴ What is more, excess calcium supplementation will suppress the secretion of parathyroid hormone and slow the natural turnover of bone. "Fossilised" bone is at risk from microfractures. Calcium intakes of between 1 g and 1.5 g daily, commonly recommended in postmenopausal women, are associated with an increased rather than decreased risk of fracture.⁵ Preventing osteoporosis does not depend on calcium, but rather on preserving bone matrix, a living tissue whose structure, strength, and integrity depend, in turn, on fine control of its turnover.

Bone turnover is maintained by two groups of cellsosteoclasts, which dig pits in mineralised matrix, and osteoblasts, which refill the pits. Osteoclastic activity is constrained by the action of sex steroids, and coordination with the osteoblasts is normally maintained such that there is no net change in bone mass during early adult life. After the menopause, circulating oestrogen concentrations fall rapidly and osteoclastic activity accelerates, outstripping the attempts of osteoblasts to keep pace. The net result is bone loss which, over a period of years, may amount to 20% or more of the skeleton.

	Effect of antiresorptive drugs

The widespread view that treatment with antiresorptive drugs restores lost bone, and that the increase in bone density observed is responsible for the subsequent reduction in the fracture rate, needs careful rethinking.⁶ There is no evidence that treatment for osteoporosis restores bone, although (unlike calcium supplements) antiresorptive drugs undoubtedly prevent its further loss. The small increases in bone mineral density gained during the early years of treatment with hormone replacement therapy or bisphosphonate drugs soon level off,⁷ and probably reflect no more than the filling by osteoblasts of the myriad pits dug before treatment by uncontrolled osteoclasts. ^{7 8} The challenge to the view that there is a causal relation between bone mineral density and risk of fracture lies in the results of large, placebo controlled, double blind, multicentre trials of bisphosphonate drugs (even those conducted in ageing women with established osteoporosis). These show a reduction in the risk of fracture at the hip and spine of more than 50%, but an increase in bone mineral density at these sites of only 5-8%,⁹ and even less in earlier studies of spinal fracture.⁷ It is difficult to attribute such a spectacular clinical result to such a small increase in bone mass.

	Factors in fracture risk

Hui et al constructed a family of curves relating fracture risk to bone density in different age groups. For the same bone density, the risk of fracture rose eightfold to 10-fold from age <45 years to 80 years.¹⁰ In a sample of 5800 Dutch men and women over 55 years of age, the risk of hip fracture rose 13-fold with age, to which the decrease in bone density contributed only 1.9 in women and 1.6 in men.¹¹ These are important observations, because they suggest that something very important in the ageing process influences fracture risk independently of bone density.

Healthy, dense bone

Apart from further undermining the validity of bone density as a surrogate for individual fracture risk and its clinical use in identifying those to whom treatment should be given, the observation calls for an explanation. One obvious mechanism is that ageing people fall more frequently and, for a given bone density, sustain fracture more often as a result. Fractures seldom occur without some impact, however fragile the bone, but there may be an additional and more subtle explanation to account for the differences in risk, one which can reconcile the data from clinical trials on fracture risk, bone density, and bone resorption.

It is intuitively likely that bone depends for strength more on its architecture than on its mass. Bone in a younger person is structurally normal, whereas in ageing people its architecture is compromised in two ways. Firstly, the progressive erosion of trabeculas will leave them weakened and, in some cases, disconnected.⁷ Microscopy shows that trabecular erosions and disconnectivity cannot be reversed by treatment.¹² Secondly, and possibly more importantly, the rate of bone turnover in women who are deficient in oestrogen will inevitably be higher, mass for mass, than in women who are oestrogen replete. If a bolt or two at a time were removed from a cantilever bridge for replacement, architectural strength might not be affectedbut if a thousand were removed at one time (a high turnover state), architectural strength could be compromised critically, with little loss in mass. Fractures occur in situations in which high bone turnover is combined with frequent impact. Bone density does not contribute greatly to the individual's risk.

While measurements of bone density emphasise how little bone is regained during antiresorptive treatment, measures of matrix loss show a dramatic and immediate slowing of bone turnover. In postmenopausal women, N-telopeptide type I collagen (a marker for osteoclast activity) fell from 3 SD above the mean to the premenopausal mean within one month of starting treatment, an effect which remained throughout the 15 month study.¹³ The bone specific alkaline phosphatase value (a marker for osteoblastic activity) showed a similar pattern within six months (the delay probably contributed to the 5% increase in spinal bone mass observed over the study period). The fossa intervention trial (FOSIT) observed over 1000 postmenopausal women with low bone density in a placebo controlled trial of alendronate that lasted more than 12 months. It showed an 80% reduction in the bone resorption rate in relation to increases in bone density of 5% at the lumbar spine and just 2% at the neck of the femur.¹⁴

Critics have pointed out that markers of bone turnover are "poorly predictive of bone mineral density ... and cannot be used to diagnose osteoporosis or to select patients for subsequent densitometry," but this is to miss the point.¹⁵ If the modest gain in bone density seen with treatment is insufficient to account for the substantial reduction in fracture risk, a state of high bone turnover, rather than its prevailing mass, may be the responsive element in fracture prevention, no matter at what age it is encountered. The clinical implications are important, as there is no evidence that bone density identifies those people who will sustain a fracture, but abundant evidence that restoring bone turnover to normal values universally reduces the risk. A switch in emphasis from bone density, which declines irreversibly, to bone turnover, which rises, but is fully reversible, makes reducing the risk of fracture a viable consideration at any age after the menopause.

A fall in bone mass, erosion of trabeculas, and ultimately their loss of connectivity must, in themselves, render bone progressively more fragile. However, the evidence cited suggests that fracture risk can be reduced appreciably, even in older women who are already osteoporotic, and without a noticeable restoration of bone mass. The widely held notion that it is too late to treat women with established osteoporosis because bone cannot be restored once it is lost is flawed because bone density and fracture risk are poorly correlated and a high turnover of bone can be normalised at any age or bone density. Bone densitometry might arguably have only a limited role in strategies to prevent osteoporotic fracture.

If high bone turnover rather than low bone density is the major component of fracture risk, antiresorptive drugs should reduce the risk of fracture to the same extent at any age after the menopause, independently of bone density, and when they are stopped their protective effect should be lost before bone density decreases. A recent report by Michaelsson et al provides the first evidence that bears out both these predictions.¹⁶ In a large, population based, case-control study, hormone replacement therapy reduced the risk of hip fracture to the same extent, whether treatment began at the menopause or decades later. After hormone replacement therapy was stopped, protection against fracture fell rapidly in relation to the interval since it was last used, and became statistically insignificant after five years.

In 1995, it was claimed that "it is now possible to accurately determine individuals' risk of osteoporosis, and to monitor their response to treatment, by means of bone densitometry."⁷ In the light of current understanding, this is probably not the case. People who will sustain a fracture cannot be identified by bone densitometry, and if bone turnover is the key responsive element in treatment, measures of bone resorption rather than of its density should be the focus of technical developments in monitoring response to treatment.

	Programmes for preventing osteoporotic fracture

How much, then, can we be certain of in formulating programmes for the prevention of osteoporotic fracture? We know that antiresorptive drugs taken at the time of the menopause will maintain bone density for as long as they are continued. However, we also know that without them few women sustain fractures before the age of 65 years, probably because they tend not to fall. So has anything been gained in prescribing treatment during 15 years of low fracture risk? Certainly, the NHS can afford neither to screen the perimenopausal population nor to treat blindly. We also know that antiresorptive drugs reduce the risk of fracture equally at any age after the menopause, and that this protection is lost rapidly when they are stopped. Finally, bone density benefits little from antiresorptive agents, while bone turnover returns to its premenopausal state.

If bone density determination cannot distinguish between those who will and will not sustain osteoporotic fractures, it has no place in selecting people for prophylaxisnotwithstanding recent Department of Health guidelines.¹⁷ Current knowledge indicates that prophylaxis should be targeted not at maintaining bone density throughout menopausal life but at restoring normal bone turnover in those who, for whatever reason, become infirm and at greatest risk of impact. Such an approach to the prevention of osteoporotic fracture would be evidence based, could be budgeted for, and would rationalise management of the individual patientsomething that treatment by numbers does not.

	Acknowledgments

I thank Debbie Hobbs for preparing this manuscript.

	Footnotes

Competing interests: None declared.

	References

1. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996; 312: 1254-1259[Abstract/Free Full Text].
2. Nordin BE, Horsman A, Marshall DH, Simpson H, Waterhouse GH. Calcium requirement and calcium therapy. Clin Orthop 1979; 140: 216-239.
3. Kreiger N, Gross A, Hunter G. Dietary factors and fracture in post-menopausal women: a case-controlled study. Int J Epidemiol 1992; 21: 953-958[Abstract/Free Full Text].
4. Heaney RP. Calcium in the prevention and treatment of osteoporosis. J Intern Med 1992; 231: 169-180[Medline].
5. Cumming RG, Cummings SR, Nevitt MG, Scott J, Ensrud KE, Vogt TM, et al. Calcium intake and fracture risk; results from the Study of Osteoporotic Fractures. Am J Epidemiol 1997; 145: 926-934[Abstract/Free Full Text].
6. Liberman UA, Weiss SR, Broll J, Minne HW, Quan H, Bell NH, et al. Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. N Engl J Med 1995; 333: 1437-1443[Abstract/Free Full Text].
7. Peel N, Eastell R. Osteoporosis. BMJ 1995; 310: 989-992[Free Full Text].
8. Christiansen C. Skeletal osteoporosis. J Bone Miner Res 1993; 8(suppl 2): S475-S480.
9. Black DM, Cummings SR, Karpf DB, Cauley JA, Thompson DE, Nevitt MC, et al. Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. N Engl J Med 1996; 348: 1535-1541 [the correct citation for this article is Lancet 1996; 348:1535-1541] .
10. Hui SL, Slemenda CW, Johston Jr CC. Age and bone mass as predictors of fracture in a prospective study. J Clin Invest 1988; 81: 1804-1809.
11. DeLaet CEDH, Van Hoat BA, Banger H, Hotman A, Pols HA. Bone density and risk of hip fracture in men and women: cross sectional analysis. BMJ 1997; 315: 221-225[Abstract/Free Full Text].
12. Kanis J. Treatment of osteoporotic fracture. Lancet 1984; i: 27-33.
13. Garnero P, Shih WJ, Gineyts E, Karpf DB, Delmas PD. Comparison of new biochemical markers of bone turnover in late post-menopausal osteoporotic women in response to alendronate treatment. J Clin Invest 1994; 79: 1693-1700.
14. Wilkin TJ. Multi-national, placebo-controlled study of alendronate in post-menopausal osteoporosisresults of 1000 patients from the FOSIT study. Acta Obstet Gynecol Scand 1997; 76 (suppl 167): 61[Medline].
15. Eastell R, Blumsohn A. The value of biochemical markers of bone turnover in osteoporosis. J Rheumatol 1997; 24: 1215-1217[Medline].
16. Michaelsson K, Baron JA, Farahmand BY, Johnell O, Magnusson C, Persson P-G, et al. Hormone replacement therapy and risk of hip fracture: population based case-control study. BMJ 1998; 316: 1858-1863[Abstract/Free Full Text].
17. Department of Health. Clinical guidelines for strategies to prevent and treat osteoporosis. London: Department of Health , 1998)

Commentary: Bone density can be used to assess fracture risk

Richard Eastell, professor. 

Section of Medicine, University of Sheffield Division of Clinical Sciences, Northern General Hospital, Sheffield S5 7AU

r.eastell{at}sheffield.ac.uk.

Wilkin identifies two areas of weakness in the use of bone mineral density measurements. Firstly, he argues that the risk of fracture cannot be predicted from someone's bone mineral density as the relation between the two described in cohort studies is too weak.¹ Secondly, he argues that bone mineral density cannot be used as a surrogate for fracture risk in clinical trials as the relation between a change in density and a change in fracture risk is not as predicted from these cohort studies.¹

I believe that it is reasonable to use an individual's bone mineral density to assess the risk of osteoporotic fracture and to make decisions about treatment with antiresorptive drugs. The relation between bone mineral density and fracture is analogous to that between blood pressure and stroke, and it is just as strong. The problem with bone mineral density has arisen when it has been used to diagnose osteoporosis. This approach raises a risk factor for fracture to the status of a diagnostic criterion, and ignores the importance of other determinants of bone strength and of factors that increase the risk of falls.² Low bone mineral density is one of the strongest risk factors for osteoporotic fracture. Nonetheless, it is preferable to use this together with other predictive factors that are common and easy to ascertain, such as low body weight (<58 kg), current smoking, a first degree relative with low trauma fracture, and personal history of low trauma fracture.³

An increase in bone mineral density of 5-10% is commonly reported for antiresorptive treatments for osteoporosis, such as hormone replacement therapy and bisphosphonate drugs. The relation between bone mineral density and fracture risk provides a prediction that the reduction in fracture should be about 25%yet it is closer to 50%.² Wilkin proposes that since the change in bone mineral density only partly explains the change in fracture risk, these antiresorptive drugs protect against fracture by reducing the turnover rate of bone.

Osteoporotic bone

The importance of a high rate of bone turnover as a risk factor for fracture has been considered by others. Riggs et al, using data from their clinical trial with hormone replacement therapy, concluded that an increase in bone density and a decrease in bone turnover at the spine had approximately equal effects in reducing the risk of spinal fracture.⁴ Parfitt proposed that antiresorptive treatment reduces the rates of bone remodelling, and since these remodelling sites act as stress risers and promote microfractures, their reduction lowers the risk of fracture.⁵ Wilkin is therefore correct to say that the change in bone turnover resulting from antiresorptive treatment may reduce the risk of fracture. The practical consequence of this is that it is reasonable to use markers of bone turnover to monitor treatment. However, in using these markers in clinical practice, care needs to be taken to minimise their variability.⁶

This marker hypothesis cannot be the entire explanation for the unexpected relation between a change in bone mineral density and a reduced risk of fracture. Treatment with alendronate results in the hypermineralisation of bone tissue. This may change the biomechanical properties of bone⁷ and explain the increase in bone mineral density seen in the third year of alendronate treatment.⁸ Parathyroid hormone treatment increases bone density and turnover, yet it reduces the risk of spine fracture.⁹ We still have much to learn about the mechanism by which treatments for osteoporosis reduce the risk of fracture.

	References

1. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996; 32: 1254-1259.
2. Eastell R. Treatment of postmenopausal osteoporosis. N Engl J Med 1998; 338: 736-746[Free Full Text].
3. Eddy DM, Johnston CC, Cummings SR, Dawson-Hughes B, Lindsay R, Melton III LJ, et al. Osteoporosis: review of the evidence for prevention, diagnosis, and treatment and cost-effectiveness analysis. Osteoporosis Int 1998; (suppl 4): S7-80.
4. Riggs BL, Melton III LJ, O'Fallon WM. Drug therapy for vertebral fractures in osteoporosis: evidence that decreases in bone turnover and increases in bone mass both determine antifracture efficacy. Bone 1996; 18: 197-201S[Medline].
5. Parfitt AM. Pathophysiology of bone fragility. In: Christiansen C, Riis BJ, eds. Osteoporosis. Copenhagen: Osteopress, 1993:164-166.
6. Blumsohn A, Eastell R. The performance and utility of biochemical markers of bone turnover: do we know enough to use them in clinical practice? Ann Clin Biochem 1997; 34: 449-459.
7. Chavassieux PM, Arlot ME, Reda C, Wei L, Yates AJ, Meunier PJ. Histomorphometric assessment of the long-term effects of alendronate on bone quality and remodeling in patients with osteoporosis. J Clin Invest 1997; 100: 1475-1480[Medline].
8. Liberman UA, Weiss SR, Broll J, Minne HW, Quan H, Bell NH, et al. Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. N Engl J Med 1995; 333: 1437-1443.
9. Lindsay R, Nieves J, Formica C, Henneman E, Woelfert L, Shen V, et al. Randomised controlled study of effect of parathyroid hormone on vertebral-bone mass and fracture incidence among postmenopausal women on oestrogen with osteoporosis. Lancet 1997; 350: 550-555[Medline].