- Email: [email protected]
Pathogenesis of postmenopausal osteoporosis
5 Pathogenesis of postmenopausal osteoporosis ROBERT
LINDSAY
The term osteoporosis refers to a group of conditions that are associated with loss of bone tissue and an accompanying architectural abnormality that occurs in the cancellous bone space. These alterations in the skeleton lead to an increase in the risk of skeletal failure or fracture (Consensus Development Conference, 1991). When the condition develops in postmenopausal women it is referred to as postmenopausal osteoporosis. The only important feature of osteoporosis for the clinician is that its presence in the skeleton increases the risk of fracture. Fractures occur commonly in postmenopausal women and most frequently involve the hip, spine and distal radius, although, as bone loss occurs throughout the skeleton, fracture of any bone can occur (Melton, 1988). The problem of postmenopausal osteoporotic fracture is sufficiently great that many countries consider it to be a major public health problem. This chapter describes the mechanisms by which postmenopausal women become osteoporotic. PEAK BONE MASS
Among young adult women, bone mass is distributed in a normal fashion in the population (Lindsay et al, 1992). Mean bone mass is slightly higher for black women than for caucasian women, and remains so throughout life (Luckey et al, 1989). Bone mass in other races is somewhat lower than in caucasians. The variability of bone mass in the population has been described in a number of cross-sectional studies. The standard deviation is usually found to be about 10% of the mean value, irrespective of the site of measurement. Bone mass in young women is primarily under genetic control (Pocock et al, 1987), although there is some evidence that the prescribed genetic amount may be influenced by a variety of factors during growth (Johnston et al, 1992). The importance of that is not fully established but may be a significant factor in the subsequent development of osteoporosis. One public health approach to the disease might then be to stress alterations in life-style and nutrition for growing children to maximize peak bone mass. It has been assumed that at peak bone mass the internal structure of the bone will be normal (Figure 1). This implies that those with low bone mass (say 2sDs below the mean) would have normal internal architecture, even Bailli~re' s Clinical Rheumatology--
Vol. 7, No. 3, October 1993 ISBN 0-7020-1711-6
499 Copyright9 1993,by Bailli6reTindall All rights of reproductionin any form reserved
500
R. LINDSAY
though the density of the skeleton as assessed by non-invasive technology is low. This has never been rigorously tested. It might be implied that only those who were genetically predisposed to have low bone mass would be those with normal structure, while those whose peak bone mass was lowered by factors negatively influencing the skeleton during growth would have some internal derangement of architecture. This hypothesis would be important clinically since the latter might be expected to be more at risk of osteoporotic fracture, requiring only a modest amount of bone loss to exaggerate the architectural problem and thereby preferentially increase fracture risk. If such differences in the premenopausal population could be observed, the capability to predict risk would be improved. At present, fracture prevalence and incidence is known to be related to low bone mass (Wasnich et al, 1985; Melton, 1988; Cooper et al, 1991; Cummings et al, 1993). Low bone mass is also a predictor for future risk of fracture. Therefore, entering the menopause with a skeletal mass less than average increases the risk for osteoporosis in the future. This predictability may be further refinable if we could detect those who, in addition to low bone mass, also had impaired bone quality, a feature that can be equated, for the most part, with disordered internal architecture (Figure 1). To ensure adequate synthesis of bone during growth, and thereby the achievement of genetic potential, requires a healthy childhood, sufficient activity to ensure adequate skeletal development, and good nutrition to supply the ingredients. It also requires adequate development of ovarian or testicular function, and its maintenance. Disorders of childhood, especially those of the neuromuscular system, reduce growth and impair the capacity to reach peak skeletal mass. These are primarily disorders associated with long-term immobility, or paralysis, since in other situations the skeleton has a remarkable capacity for 'catchup' growth. The prejudiced skeleton of the child paralysed at birth is well known, but such a skeleton may be sufficient for the everyday needs of that individual. Within the normal confines of activity, however, there is more doubt about the impact of exercise on the skeleton during growth and in adult life. It seems intrinsically reasonable to imply that there is some form of 'mechanostat' that allows the skeleton to judge the level of stress to which it is continuously being subjected and modify its structure appropriately (Frost, 1986). That such a mechanism may have greater effects during growth than at other times also seems obvious. The data to support this view are scant. It also seems self-evident that adequate nutrition is required to .provide the materials for skeletal growth. Guidelines, however, are equally difficult to provide. Even for calcium, which has been studied in greater detail than other nutrients, it is not clear whether current levels of intake are inadequate. The young have a higher efficiency for absorption of calcium across the intestine than adults, offsetting the higher requirement for calcium to supply sufficient to support skeletal growth. Support for increased calcium intake comes from an elegant double-blind controlled study of calcium supplementation in monovular twins (Pocock et al, 1987). This study suggested that an intake of around 1500mg per day increases accretion of skeletal mass in prepubertal children. However, that study
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
(a)
(b) Figure 1. (a) Normal and (b) osteoporotic cancellous bone.
501
502
R. LINDSAY
terminated before these children had reached peak skeletal mass, and therefore it is not clear whether this increased mass would be maintained. Our cross-sectional data suggest that peak bone mass may be modified by as much as 5-10% by ensuring adequate calcium (Kanders et al, 1988). This would be a clinically relevant amount as it would slow the onset of osteoporosis by some 5 years as the population aged after menopause. S K E L E T A L R E M O D E L L I N G AND THE M E N O P A U S A L C H A N G E
In the adult, especially in cancellous bone, there is an ongoing process by which old bone is replaced by new tissue. This process, called remodelling, occurs in packets that are distinct in both time and space (Parfitt, 1984). Each remodelling unit follows a defined sequence of events that has been characterized in histomorphometric studies of the skeleton (Figure 2). At a bone surface to be remodelled, in a poorly understood process termed activation, the cells lining the surface of the bone contract, exposing it to the marrow space. Osteoclasts, multinucleated giant cells of the macrophage lineage, are chemotactically attracted to the site, and begin the process of bone resorption. During a very few weeks, these cells remove bone to the approximate depth of about 50p~m, a depth that appears to be preprogrammed. At that point the osteoclasts disappear from the site, their fate unknown, and are replaced by mononuclear cells whose origin and purpose are not clear. As they link the process of resorption with that of formation, this time in the remodelling cycle has been labelled reversal. In some fashion, these cells may prepare the surface of the resorbed bone for the action of osteoblasts, perhaps secreting the so-called cement line, seen on histological sections as a boundary to the new packet of bone. They may also be responsible for the recruitment to the site of active osteoblasts, which synthesize an organic matrix that includes type 1 collagen called osteoid. The process of filling the cavity takes several weeks to months. The osteoid is subsequently mineralized by an extracellular process forming mature bone. In the best of homoeostatic conditions, bone removed is replaced by an equal quantity of new bone. In the premenopausal woman, this is essentially the case, although there may be somewhat of a deficit at some skeletal sites. As women become oestrogen deficient (or for that matter men testosterone deficient), there are significant changes that occur in the remodelling process. First, there is a clear increase in the frequency at which new 'remodelling cycles are created (Parfitt, 1988). As there is no apparent change in the life span of these cycles, the net effect is that at any single time point more bone surface is in the process of being remodelled. Consequently, there is a remediable deficit in bone mass, that probably amounts to a small percentage of the total cancellous bone mass. In addition, there is a less well documented, but necessary, deficit in the amount of new bone formed within each remodelling unit (Parfitt, 1988). The mechanism, from biochemical data, is presumed to be increased activity of the osteoclast (and consequently a deeper cavity) without an accompanying increase in osteoblast activity, but this has yet to be adequately documented histologically.
503
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
LC 1 QUIESCENCE
POC 2 ACTIVATION
3 RESORPTION
4 REVERSAL
5 FORMATION
LC 6 QUIESCENCE
[~OLD
BONE
[~NEW
BONE
[]OSTEOID
Figure 2.
Consequently, and measurably by techniques that estimate bone mass, there is an irreversible decline in the amount of bone present. Because bone remodelling is a surface phenomenon, and loss of bone tissue removes surface, the decline in mass is self-limited, but will be greatest in cancellous bone that has the highest surface to volume ratio. The proposed increase in the depth of the cavity created by osteoclasts has one further consequence that has been documented histologically and using scanning electron microscopy of cancellous bone (Dempster et al, 1986). This type of bone is constituted by trabecular plates that are of relatively uniform thickness. With increased number of remodelling units present, and increased depth to each cavity, there is a greater chance of penetration of the trabecular plate by osteoclasts. This penetration exaggerates the process of bone loss by removing surfaces upon which new bone could be deposited. As
504
R. LINDSAY
trabeculae are increasingly frequently penetrated they become thinner and more rod-like, until they are disconnected. At this point they are completely removed. The consequence is loss of the entire trabecular structure within the bone (Figure 2). In the vertebrae there is preferential loss of horizontal trabeculae, and thus early weakening of a structure that must withstand vertical loading (imagine a ladder without horizontal rungs). The consequence is vertebral fracture. In long bones, where cancellous bone fills the metaphyseal ends, the consequence is an increase in the risk of fracture on trauma and a subsequent increase in risk of Colles' and hip fractures. The precise mechanism by which these events occur is not entirely understood. Non-invasive studies of bone mass have confirmed the relationship between oestrogen deficiency and reduced bone density, irrespective of the cause of oestrogen deficiency. Thus, low bone mass occurs after menopause or ovariectomy (Aitken et al, 1973; Aloia et al, 1983; Slemenda et al, 1987), as a consequence of hyperprolactinaemia (Klibanski et al, 1980), or hypothalamic causes of amenorrhoea including anorexia (Rigotti et al, 1984) and excessive exercise (Drinkwater et al, 1984; Crilly et al, 1988). Additionally, the use of gonadotrophin releasing hormone agonists as prolonged treatment for endometriosis causes bone loss (Cann et al, 1987). However, although in each case there seems to be an increase in the risk for all fractures, only in the postmenopausal state is the classical fracture distribution commonly seen. (Anorexia, also associated with classical osteoporotic fractures, is a special case, probably because of the calorie deprivation and other metabolic dysfunctions.) Therefore, it appears that other factors must be superimposed upon oestrogen deficiency, modifying the rate and/or extent of bone loss to increase the risk of specific fractures in this population. RISK FACTORS FOR OSTEOPOROSIS
Epidemiologists often make careers from collecting risk factors for disease. This is particularly common among disorders of ageing that mostly have multiple factors involved in their pathogenesis. Osteoporosis is no exception. There is a long list of factors that are thought to increase the risk of osteoporosis among ageing women. Some, for example the use of glucocorticoids in therapeutic doses, are sufficiently potent in themselves to be specific causes of secondary osteoporosis (Lukert and Raisz, 1990). These are considered elsewhere (see Chapter 10) and will not be discussed further here, ' except to say that each is likely to exert a more detrimental effect on the skeleton of oestrogen-deficient women, who already are in negative bone balance. The factors to which the physician must pay attention are those that are potentially remediable (Tables 1 and 2). Thus, factors such as a positive family history, which may be of some value when talking to patients about risk, or may bring the patient to the physician to discuss risk, are clearly not of value when discussing changes to be effected by the patient that might reduce risk. For the most part, for healthy perimenopausal patients, risk factors that must be considered involve nutrition and life-style.
PATHOGENESIS OF POSTMENOPAUSALOSTEOPOROSIS
505
Table 1. Risk factors for osteoporotic fracture. Low bone mass Previous atraumatic fracture Risk factors for falls
Disorders of confusion (e.g. Alzheimer's disease) Medications causing confusion (long-acting tranquillizers, alcohol etc.) Medications: diuretics, hypotensive agents Neuromuscular disease (e.g. Parkinson's disease) and frailty Extrinsic factors: poor lighting, loose carpets and wires, and other home hazards Table 2. Risk factors for low bone mass.
Genetic Endocrine
Nutritional
Life-style Drugs
Race, sex, positive family history Low body-weight Ovarian insufficiency Cushing's syndrome, including therapeutic use of glucocorticoids Excess thyroid hormone Type 1 diabetes mellitus Hyperparathyroidism (low cortical bone mass only) Calcium insufficiencyand malabsorption Alcohol excess Low body-weight (e.g. anorexia nervosa, cyclicaldieting) Caffeine, protein and phosphate excess (?) Sodium, hydrogen ion excess (??) Physical inactivity Cigarette consumption Physical activity to excess (pituitary-ovarian dysfunction) Thyroid hormone Prednisone and other steroids Loop diuretics (frusemide) Anticonvulsants (?) Gonadotrophin releasing hormone agonists and antagonists Methotrexate
Nutrition
There are abundant data to indicate that calcium deficiency can produce osteoporosis in both animals and humans (Fanstat, 1962; Foresta et al, 1983). Insufficient calcium in the diet, or calcium malabsorption, results in the utilization of skeletal calcium to help maintain serum calcium levels within the normal range. In its most obvious state calcium deficiency results in overt secondary hyperparathyroidism. However, in normal clinical practice this is seen only in situations such as adult coeliac disease or gastrointestinal resection, where severe calcium malabsorption is occasionally seen (excluding renal disease which has different consequences). Most adult w o m e n have parathyroid h o r m o n e (PTH) levels within the normal range. Therefore, if calcium deficiency is present, the P T H level must be abnormally high for that individual, although still within the normal range. The analogy is with thyroid replacement therapy. H e r e , the practice used to be to give sufficient thyroid h o r m o n e to keep the circulating levels of T4 and T3 within the normal range. With the development of the supersensitive thyroid stimulating h o r m o n e (TSH) assay it has b e c o m e clear that m a n y of these
506
R. LINDSAY
individuals are over-replaced with completely suppressed TSH levels and require a downward adjustment in dose even though thyroid hormone levels are normal. From the data produced over many years by Heaney and co-workers (1978a; 1978b), it has become clear that calcium absorption is extremely variable across the normal population. Consequently, while one individual may be replete in calcium at an intake of say 600 mg/day, many others will not. The average intake required to bring 90% of the population into the calcium replete state appears to be about 1000-1500 mg/day. Because there is no test for calcium deficiency that is easily applied clinically (like ferritin for iron) and because intakes at that level appear to be safe, there is a general recommendation that adult women increase their calcium consumption to around these values (National Osteoporosis Foundation, 1992a). For most women, at that level of intake calcium deficiency will not add to the skeletal insult imposed by oestrogen deficiency. Other dietary factors that have been suggested as risk factors for osteoporosis include salt, caffeine, protein (especially animal protein) and phosphate (Mazess et al, 1990). Since these appear to be minor risk factors, if indeed they are risk factors at all, the only recommendation required is to ensure that intake is moderate. Several are risk factors for other diseases, so their intake is often limited, at least in the North American population. Alcohol links nutrition and life-style. Excessive alcohol intake is clearly detrimental to the skeleton and should be discouraged (Bikle, 1988). The level of intake that is toxic to bone is not clear but may well be within the range of 'social' drinking among adult women. Thus, we ask for complete abstinence, and expect some reduction in intake. Alcohol, in addition to its effects on the skeleton, also increases the risk of falling and injury among the elderly and will predispose to fracture for that reason. Thus, the elderly in particular should be discouraged from alcohol use.
Life-style There are normally two factors considered under the heading of life-style. The first is cigarette consumption, which is often linked to high alcohol intake. There are data of varying standard to suggest that cigarette consumption is an independent risk factor for osteoporotic fracture (Daniell, 1976). Moreover, since there are major health consequences from cigarette use, this becomes yet another reason to encourage patients to stop, irrespective of the quality of the data. The second life-style factor is physical activity. While it is clear that disuse increases bone loss (Minare et al, 1974), there is less information on the ideal level of physical activity within the normal range that would be most beneficial for the skeleton (Cooper et al, 1988; Gutin and Kasper, 1992). There is equal doubt about the types of activity, although most would agree that weight-bearing activity is best (National Osteoporosis Foundation, 1992b). It seems likely that the effects of activity are mostly, if not completely, mediated directly upon the skeleton. Consequently, to achieve an effect on a particular bone, that bone must be stressed as part of the
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
507
exercise programme (Lanyon, 1984). In the absence of specific recommendations, many physicians adopt the guidelines promoted for cardiovascular health and recommend 30 min of aerobic activity three times per week to their patients. In fact, the effects of activity may be mediated on fracture risk by improving strength and skeletal muscle function rather than by their effect on the skeleton. Thus, even for the elderly, an exercise programme is important. For those with low bone mass, or who already have fractures, initiating a programme of progressive strengthening under the supervision of a qualified physical therapist is often beneficial. Other risk factors
Superimposed on these risk factors are those of ill-health that, by virtue of their direct effects on the skeleton, indirect effects through nutrition or life-style, or the effects of the drugs used in their treatment, will increase the risk of osteoporosis. The medications vary from glucocorticoids, to antimitotic agents, to diuretics, thyroid hormone and anticonvulsants. Obtaining an adequate history of medication use is therefore important in evaluating the patient. The main risk factors for future fracture are low bone mass and the presence of a previous atraumatic fracture (Ross et al, 1991). For younger individuals who do not have an increased risk of falling, these seem to explain the majority of risk, while for older individuals they probably add to the risk from falling (Melton and Riggs, 1985). Thus, disorders such as Alzheimer's disease, general frailty and neuromuscular disorders such as Parkinson's disease all increase the risk of falls. In addition, there may be a component of frailty linked to vitamin D metabolism, which is often impaired in the elderly. Recent studies have demonstrated a significant reduction in incidence of hip fracture with just modest supplements of vitamin D and calcium in an elderly population (Chapuy et al, 1992). Since vitamin D insufficiency may be fairly common among the elderly, and is known to cause proximal muscle weakness, the remediation of this may be a possible mechanism for reduced fracture frequency. While it has been commonly assumed by many that low bone mass leads to an increase in the risk of fracture, it is only comparatively recently that data have been published to confirm this assumption (Wasnich et al, 1985; Melton, 1988; Cooper et al, 1991; Cummings et al, 1993). In prospective studies it has now been shown clearly that low bone mass is a predictor of fracture. The measurement appears to be predictive of all fractures, irrespective of the skeletal site at which the measurement is made, with roughly equal predictive power. However, preliminary data do indicate that measurement at a specific site will be more predictive of fracture at that site (Cummings et al, 1993). For example, measurements of the hip appear to be most reliable at predicting fractures of the femoral neck. Thus, for the clinician, bone mass becomes a tool by which the patient's risk of osteoporotic fracture can be assessed. This would be similar to using blood pressure as an indicator of the risk of cerebrovascular disease or cholesterol as a marker of risk for coronary heart disease.
508
R. LINDSAY
MECHANISMS OF BONE LOSS
As noted above, there are alterations in the remodelling cycle with increasing age that result in the loss of bone mass. The mechanisms by which these occur are not entirely understood. Of major importance in the pathogenesis of the disease among postmenopausal women is the decline in ovarian function that occurs at climacteric, resulting in the external clinical event termed the menopause. The role of ovarian insufficiency
The fact that ovarian failure occurs in all women at the average age of 51 years, and is associated with the alterations in skeletal remodelling outlined above, results in the generalization that bone loss occurs among all postmenopausal populations (Lindsay, 1988). It is this ubiquitous finding that makes the relationship between ovarian failure and menopause one of major importance in public health. Since women enter their menopausal years with variable amounts of bone, bone mass at time of menopause is a useful marker for those most at risk of osteoporotic fracture. There is also considerable variability in the rate (and probably duration) of bone toss among postmenopausal women. Thus, in addition to the use of bone mass as an indicator of risk, there has been considerable interest in attempting to predict rates of loss to add to the efficiency with which the clinician can identify those at maximal risk (Christiansen et al, 1987). To date, one algorithm that uses a combination of biochemical markers of remodelling, together with clinical factors, has been published, suggesting that this together with one measurement of bone mass will predict bone mass up to 12 years later with considerable accuracy (Christiansen et al, 1987). However, this has yet to be verified by other laboratories, and would add additional expense to the evaluation of the perimenopausal woman. Thus, biochemical evaluation cannot be recommended for general clinical use at the present time. The importance of the menopause in the pathogenesis of osteoporosis is substantiated by the effects of oestrogen replacement therapy on bone mass and fracture occurrence in the postmenopausal population (Lindsay et al, 1978; 1984; Hutchinson et al, 1979; Christiansen et al, 1980; Weiss et al, 1980; Naessen et al, 1990; Christiansen and Lindsay, 1991). Many studies have demonstrated that oestrogen intervention reduces bone remodelling to premenopausal levels, and inhibits loss of bone. Some studies have continued for as long as 15 years without evidence of bone loss in cortical bone, and cross-sectional data in which vertebral and femoral neck bone mass have been measured suggest that there is minimal if any bone loss at these sites of important osteoporotic fracture, at least for the first 10 years of the studies (Christiansen et al, 1980; Christiansen and Lindsay, 1991). These controlled clinical trial data are confirmed by epidemiological studies indicating that oestrogen intervention is associated with a reduction in the risk of osteoporotic fracture (Hutchinson et al, 1979; Weiss et al, 1980; Naessen et al, 1990). The majority of published studies have evaluated the effects of
P A T H O G E N E S I S OF P O S T M E N O P A U S A L O S T E O P O R O S I S
509
oestrogen on femoral neck fracture. All suggest that there is a significant effect; on average the reduction in risk conferred by oestrogen use appears to be about 50%. Similar data are available for Colles' fracture. Vertebral fracture data are more sparse. One epidemiological type study indicates a significant protection against vertebral fracture in the order of 60%, while one controlled clinical trial suggests similar, although slightly greater, protection using radiographic abnormalities as the end-point (Ettinger et al, 1985; Lindsay, 1987). Thus, interventional data support the concept that the menopause plays an important role in skeletal loss among the ageing female population. The mechanism by which this occurs is not well understood. A few years ago, for the first time oestrogen receptors were detected in cells of the osteoblast lineage (Eriksen et al, 1988; Komm et al, 1988). It has proven considerably more difficult to determine the presence of oestrogen receptors on osteoclasts, although the expression of these specific receptor types may not be required for a skeletal effect. Certainly, in vitro, it is difficult to demonstrate oestrogen effects on osteoclasts (Arnett et al, 1986), even when the culture is contaminated with osteoblasts. There is some suggestion that the degree of contamination with mononuclear cells, and specifically osteoblast-like cells, increases the likelihood that oestrogen will cause inhibition of resorption of devitalized bone in vitro (McSheehy and Chambers, 1986). However, the issue is still far from proven. The fact that oestrogens do not affect the function of osteoclasts directly implies that, if they affect osteoclast function in vivo, this must be mediated indirectly. However, the major effect of oestrogen is to reduce the activation of new remodelling sites (Steiniche et al, 1989), and activation frequency is more likely to be controlled by lining cells, which are thought to be of osteoblast origin (Parfitt, 1988). A variety of responses has been detected in cells derived from osteoblasts (Ernst et al, 1988; Gray et al, 1988). However, one study of osteoblasts derived from human bone specimens found little support for a direct effect on a number of parameters characteristic of osteoblasts (Keeting et al, 1991). Thus, the osteoblast as a mediator of oestrogen action on bone is far from proven. Others have suggested that other mononuclear cells may be involved in the oestrogen effect, principally cells from the marrow. Production of interleukin 1 (I1-1) by human mononuclear cells in the circulation appears to increase after the menopause, and can be reduced by oestrogen administration (Fournier et al, 1990; Pacifici et al, 1991). It has been suggested that stimulated I1-1 production correlates with bone turnover in postmenopausal patients. In another experimental model, I1-6 has been proposed as the mediator of oestrogen action, and the development of osteoporotic changes in the skeleton have been inhibited by antibodies to I1-6 (Jilka et al, 1992). The production of other factors may play a role in the oestrogen effect on bone. These include transforming growth factor [3TGF-beta, insulin-like growth factors i and 2 (Canalis et al, 1991), and an assortment of other general growth factors that occur in abundance in bone. Whatever the local effects of oestrogen it is clear that loss of oestrogen at the menopause also affects calcium homoeostasis. Several years ago it was
510
R. LINDSAY
suggested that the net effect of oestrogen, at the whole body level, is to render the skeleton resistant to the resorbing effects of P T H (Heaney, 1965). Recent data, in which the response of the skeleton to P T H was tested in postmenopausal women, confirmed this effect (Cosman et al, 1993). In addition, oestrogen status also modified the renal hydroxylase response to PTH, increasing the synthesis of 1,25(OH)2D in response to P T H (Cosman et al, 1991). Other studies have suggested that oestrogen can improve the efficiency with which the intestine absorbs calcium from diet, a process whose active component is controlled by 1,25(OH)2D (Heaney et al, 1978). Overall the effects of oestrogen increase the efficiency with which the body utilizes calcium, since urinary calcium concentration also falls as oestrogen status improves. Whether these are the byproducts of a primary oestrogen effect on bone (with reduced mobilization of skeletal calcium) or represent a prime effect of oestrogen in controlling, first, calcium homoeostasis and, only secondarily, skeletal homoeostasis is not at all clear.
SUMMARY Osteoporosis is a disorder of ageing that shares with other disorders of ageing a multifactorial pathogenesis. The important factors for osteoporosis include the diet, life-style and intercurrent factors such as disease. However, it is clear that loss of ovarian function is an important determinant of bone loss, and oestrogen appears to be the key factor involved. Thus, not only does toss of ovarian function result in bone loss, it can be stopped by adequate oestrogen intervention. Numerous techniques are available to measure bone mass non-invasively and to estimate the risk of future fracture. Thus, for the postmenopausal woman who is concerned about osteoporosis, and who is willing to accept intervention to prevent the disease, bone mass measurement allows the clinician to determine the risk of future osteoporotic fracture and to provide intervention if required. Future studies may elucidate whether determination of skeletal remodelling using biochemistry adds significantly to risk determination. This may be required when considering agents other than oestrogen for intervention among asymptomatic women, as these agents primarily affect the skeleton, while the effects of oestrogen are wide ranging in the body.
REFERENCES Aitken JM, Anderson JB & Horton PW (1973) Seasonal variations in bone mineral content after the menopause. Nature 241"59-60. Aloia JF et al (1983) Determinants of bone mass in postmenopausal women. Archives of Internal Medicine 143: 1700-1704. Arnett TR, Lindsay R & Dempster DW (1986) Effect of estrogens and anti-estrogens on osteoclast activity in vitro. Journal of Bone and Mineral Research 1:99 (abstract). Bikle DD (1988) Effects of alcohol abuse on bone. Comprehensive Therapy 14: 16-20. Canalis E, McCarthy TL & Centrella M (1991) Growth factors and cytokines in bone cell metabolism. Annual Review of Medicine 42" 17-24.
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
511
Cann CE, Henzl MR, Burrk K et al (1987) Reversible bone loss is produced by the GnRH agonist nafarelin. In Cohn DV, Martin TJ & Meunier PJ (cds) Calcium Regulation and Bone Metabolism: Basic and Clinical Aspects, pp 123-127. Amsterdam: Elsevier. Chapuy MC, Arlot ME, Duboeuf F et al (1992) Vitamin D 3 and calcium to prevent hip fractures in elderly women. New England Journal of Medicine 27: 163%1642. Christiansen C & Lindsay R (1991) Estrogens, bone loss and preservation. Osteoporosis International 1: 7-13. Christiansen C, Christiansen MS & McNair P (1980) Prevention of early postmenopausal bone loss: conducted 2 years study in 315 normal females. European Journal of Clinical Investigation 10: 273-279. Christiansen C, Riis BJ & Rodbro P (1987) Prediction of rapid bone loss in postmenopausal women. Lancet i: 1105-1107. Consensus Development Conference (1991) Prophylaxis and treatment of osteoporosis. Osteoporosis International 1:114-126. Cooper C, Barker DJ & Wickham C (1988) Physical activity, muscle strength and calcium intake in fractures of the proximal femur in Britain. British Medical Journal 297: 1443-1446. Cooper C, Wickham C & Walsh K (1991) Appendicular skeletal status and hip fracture in the elderly: 14-year prospective data. Bone 12: 361-364. Cosman F, Shen V, Herrington B & Lindsay RA (1991) Response of the parathyroid gland to infusion of (1-34)hPTH: demonstration of suppression of endogenous secretion using IRMA intact (1-84)PTH assay. Journal of Clinical Endocrinology and Metabolism 73: 1345-1351. Cosman F, Shen V, Xie F, Seibel M, Ratcliffe A & Lindsay R (1993) Estrogen protection against bone resorbing effects of parathyroid hormone infusion. Annals of Internal Medicine 118: 337-343. Crilly RG, Anderson C, Hogan D & Delaquerriere-Richardson L (1988) Bone histomorphometry, bone mass, and related parameters in alcoholic males. Calcified Tissue International 43: 269-276. Cummings SR, Black DM, Nevitt MC et al (1993) Bone density at various sites for prediction of hip fractures. Lancet 341: 72-75. Daniell HW (1976) Osteoporosis of the slender smoker. Archives of Internal Medicine 136: 298-304. Dempster DW, Shane E, Horbert W & Lindsay R (1986) A simple method for correlative light and scanning electron microscopy of human iliac crest bone biopsies: qualitative observations in normal and osteoporotic subjects. Journal of Bone and Mineral Research 1: 15-21. Drinkwater BD, Nilson KL & Chesnut CHIII (1984) Bone mineral content of amenorrheic and eumenorrheic athletes. New England Journal of Medicine 311: 277-281. Eriksen EF, Colvard DS, Berg NJ et al (1988) Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241: 84-86. Ernst MC, Schmid C, Frankenfoldt ER & Froesch ER (1988) Estradiol stimulation of osteoblast proliferation in vitro: mediator roles of TGFB, PGE2, IGF1. Calcified Tissue International 42 (supplement 1): 117. Ettinger B, Genant HK & Cann CE (1985) Long-term estrogen therapy prevents bone loss and fracture. Annals of Internal Medicine 102: 319-324. Fanstat T (1962) Hormonal burrs for collagen bundle generation in uterine stoma: extracellular studies of uterus. Endocrinology 71: 878-887. Foresta C, Busnardo B, Ruzza G, Zanattia G & Mioni R (1983) Lower calcitonin levels in young hypogonadic men with osteoporosis. Hormone and Metabolic Research 15: 206207. Fournier B, Feralli J & Rordorf C (1990) Induction of interleukin-6 (IL-6) by interleukin-1 beta (IL-f beta) in MG63 cells. Modulation by oestradiol. Calcified Tissue International 46 (supplement 2): A36 (abstract). Frost HM (1986) The pathomechanics of osteoporosis. Clinical Orthopaedics and Related Research 200: 198-225. Gray TK, Mohan S, Linkhart TA, Williams ME & Baylink DJ (1988) Estrogen may mediate its effects on bone ceils by signalling the observation of growth factors. Journal of Bone and Mineral Research 3: A552. Gutin B & Kasper MJ (1992) Can vigorous exercise play a role in osteoporosis prevention? A review. Osteoporosis International 2: 55.
512
R. LINDSAY
Heaney RP (1965) A unified concept of osteoporosis. American Journal of Medicine 39: 377-380. Heaney RP, Recker RR & Saville PD (1978a) Menopausal changes in calcium balance performance. Journal of Laboratory and Clinical Medicine 92: 953-963. Heaney RP, Recker RR & Saville PD (1978b) Menopausal changes in bone remodelling. Journal of Laboratory and Clinical Medicine 92: 964-970. Hutchinson TA, Polansky JM & Feinstein AR (1979) Postmenopausal oestrogens protect against fracture of hip and distal radius. Lancet ii: 705-709. Jilka RL, Hangoc G, Girasole G et al (1992) Increased osteoclast development after estrogen loss: mediation by interleukin-6. Science 257: 88-91. Johnston CC Jr, Miller JZ, Slemenda CW et al (1992) Calcium supplementation and increases in bone mineral density in children. New England Journal of Medicine 327: 82-87. Kanders B, Dempster DW & Lindsay R (1988) Interaction of calcium nutrition and physical activity on bone mass in young women. Journal of Bone and Mineral Research 3: 145-149. Keeting PE, Scott RE, Colvard DS, Hart IK, Spelsberg TC & Riggs BL (1991) Lack of a direct effect of estrogen on proliferation and differentiation of normal human osteoblast-like cells. Journal of Bone and Mineral Research 6: 29%304. Klibanski A, Neer RM, Beitins IZ, Ridgway C, Zervas NT & MacArthur J (1980) Decreased bone density in hyperprolactinemic women. New England Journal of Medicine 303: 1511-1514. Komm BS, Terpening CM & Benz DJ (1988) Estrogen binding, receptor mRNA, and biologic response in osteoblast-like osteosarcoma cells. Science 241: 81-84. Lanyon LE (1984) Functional strain as a determinant for bone remodeling. Calcified Tissue International 36: 556-561. Lindsay R (1987) Estrogen therapy in the prevention and management of osteoporosis. Obstetrics and Gynecology 156: 1347-1356. Lindsay R (1988) Sex steroids in the pathogenesis and prevention of osteoporosis. In Riggs BL (ed.) Osteoporosis: Etiology, Diagnosis and Management, pp 333-358. New York: Raven Press. Lindsay R, Hart DM, MacLean A, Clark AC, Kraszewski A & Garwood J (1978) Bone response to termination of oestrogen treatment. Lancet i: 1325-1327. Lindsay R, Hart DM & Clark DM (1984) The minimum effective dose of estrogen for prevention of postmenopausal bone loss. Obstetrics and Gynecology 63: 759-763. Lindsay R, Cosman F, Herrington BS & Himmelstein S (1992) Bone mass and body composition in normal women. Journal of Bone and Mineral Research 7: 55-63. Luckey MM, Meier DE, Mandell JP, DaCosta MC, Hubbard ML & Goldsmith SJ (1989) Radial and vertebral bone density in white and black women: evidence for radial differences in premenopausal bone homeostasis. Journal of Clinical Endocrinology and Metabolism 69: 762-770. Lukert BP & Raisz LG (1990) Glucocorticoid-induced osteoporosis: pathogenesis and management. Annals of Internal Medicine 112: 352-364. McSheehy PHJ & Chambers TJ (1986) Osteoblastic cells mediate osteoclastic responsiveness to parathyroid hormone. Endocrinology 118: 824-828. Mazess RB, Barden HS, Bisek JP & Hanson J (1990) Dual-energy X-ray absorptiometry for total-body and regional bone-mineral and soft-tissue composition. American Journal of Clinical Nutrition 51: 1106-1112. Melton LJ III (1988) Epidemiology of fractures. In Riggs BL & Melton LJ III (eds) Osteoporosis: Etiology, Diagnosis, and Management, pp 133-154. New York: Raven Press. Melton LJ & Riggs BL (1985) Risk factors for injury after a fall. Clinics in Geriatric Medicine 525: 539. Minare P, Meunier P, Edouard C, Bernard J, Courpron P & Bournet J (1974) Quantitative histological data on disuse osteoporosis. Comparison with biological data. Calcified Tissue Research 17: 57-63. Naessen T, Persson I, Adami HO, Bergstrom R & Bergkvist L (1990) Hormone replacement therapy and the risk for first hip fracture. A prospective, population-based cohort study. Annals of lnternal Medicine 113: 95-103. National Osteoporosis Foundation (1992a) Calcium Intake Recommendations. Washington, DC. National Osteoporosis Foundation (1992b) Recommendations on Exercise. Washington, DC.
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
513
Pacifici R, Brown C, Puscheck E et al (1991) Effect of surgical menopause and estrogen replacement on cytokine release from human blood mononuclear cells. Proceedings of the National Academy of Sciences of the United States of America 88: 5134-5138. Parfitt AM (1984) The cellular basis of bone remodeling: the quantum concept reexamined in light of recent advances in the cell biology of bone. Calcified TissueInternational 36: 37-45. Parfitt AM (1988) Bone remodeling: relationship to the amount and structure of bone, and the pathogenesis and prevention of fractures. In Riggs BL & Melton LJ III (eds) Osteoporosis: Etiology, Diagnosis, and Management, pp 45-93. New York: Raven Press. Pocock NA, Eisman JA, Hopper JL, Yeates MG, Sambrook PM & Ebers S (1987) Genetic determinants of bone mass in adults. Journal of Clinical Investigation 80: 706-710. Rigotti NA, Nussbaum SR, Herzog DB et al (1984) Osteoporosis in women with anorexia nervosa. New England Journal of Medicine 311: 1601-1606. Ross PD, Davis JW, Epstein RS & Wasnich RD (1991) Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Annals of Internal Medicine 114: 919-923. Slemenda C, Hui SL, Longcope C & Johnston CC (1987) Sex steroids and bone mass; a study of changes about the time of the menopause. Journal of Clinical Investigation 80: 1261-1269. Steiniche T, Hasling C, Charles P e t al (1989) A randomized study of the effects of estrogen/ gestagen or high dose oral calcium on trabecnlar bone remodeling in postmenopausal osteoporosis. Bone 10: 313-320. Wasnich RD, Ross PD, Heilbrun LK & Vogel JM (1985) Prediction of postmenopausal fracture risk with use of bone mineral measurements. American Journal of Obstetrics and Gynecology 153: 745-751. Weiss NS, Ure CL & BaUard JH (1980) Decreased risk of fractures of the hip and lower forearm with postmenopausal use of oestrogen. New England Journal of Medicine 303: 1195-1198.
LINDSAY
The term osteoporosis refers to a group of conditions that are associated with loss of bone tissue and an accompanying architectural abnormality that occurs in the cancellous bone space. These alterations in the skeleton lead to an increase in the risk of skeletal failure or fracture (Consensus Development Conference, 1991). When the condition develops in postmenopausal women it is referred to as postmenopausal osteoporosis. The only important feature of osteoporosis for the clinician is that its presence in the skeleton increases the risk of fracture. Fractures occur commonly in postmenopausal women and most frequently involve the hip, spine and distal radius, although, as bone loss occurs throughout the skeleton, fracture of any bone can occur (Melton, 1988). The problem of postmenopausal osteoporotic fracture is sufficiently great that many countries consider it to be a major public health problem. This chapter describes the mechanisms by which postmenopausal women become osteoporotic. PEAK BONE MASS
Among young adult women, bone mass is distributed in a normal fashion in the population (Lindsay et al, 1992). Mean bone mass is slightly higher for black women than for caucasian women, and remains so throughout life (Luckey et al, 1989). Bone mass in other races is somewhat lower than in caucasians. The variability of bone mass in the population has been described in a number of cross-sectional studies. The standard deviation is usually found to be about 10% of the mean value, irrespective of the site of measurement. Bone mass in young women is primarily under genetic control (Pocock et al, 1987), although there is some evidence that the prescribed genetic amount may be influenced by a variety of factors during growth (Johnston et al, 1992). The importance of that is not fully established but may be a significant factor in the subsequent development of osteoporosis. One public health approach to the disease might then be to stress alterations in life-style and nutrition for growing children to maximize peak bone mass. It has been assumed that at peak bone mass the internal structure of the bone will be normal (Figure 1). This implies that those with low bone mass (say 2sDs below the mean) would have normal internal architecture, even Bailli~re' s Clinical Rheumatology--
Vol. 7, No. 3, October 1993 ISBN 0-7020-1711-6
499 Copyright9 1993,by Bailli6reTindall All rights of reproductionin any form reserved
500
R. LINDSAY
though the density of the skeleton as assessed by non-invasive technology is low. This has never been rigorously tested. It might be implied that only those who were genetically predisposed to have low bone mass would be those with normal structure, while those whose peak bone mass was lowered by factors negatively influencing the skeleton during growth would have some internal derangement of architecture. This hypothesis would be important clinically since the latter might be expected to be more at risk of osteoporotic fracture, requiring only a modest amount of bone loss to exaggerate the architectural problem and thereby preferentially increase fracture risk. If such differences in the premenopausal population could be observed, the capability to predict risk would be improved. At present, fracture prevalence and incidence is known to be related to low bone mass (Wasnich et al, 1985; Melton, 1988; Cooper et al, 1991; Cummings et al, 1993). Low bone mass is also a predictor for future risk of fracture. Therefore, entering the menopause with a skeletal mass less than average increases the risk for osteoporosis in the future. This predictability may be further refinable if we could detect those who, in addition to low bone mass, also had impaired bone quality, a feature that can be equated, for the most part, with disordered internal architecture (Figure 1). To ensure adequate synthesis of bone during growth, and thereby the achievement of genetic potential, requires a healthy childhood, sufficient activity to ensure adequate skeletal development, and good nutrition to supply the ingredients. It also requires adequate development of ovarian or testicular function, and its maintenance. Disorders of childhood, especially those of the neuromuscular system, reduce growth and impair the capacity to reach peak skeletal mass. These are primarily disorders associated with long-term immobility, or paralysis, since in other situations the skeleton has a remarkable capacity for 'catchup' growth. The prejudiced skeleton of the child paralysed at birth is well known, but such a skeleton may be sufficient for the everyday needs of that individual. Within the normal confines of activity, however, there is more doubt about the impact of exercise on the skeleton during growth and in adult life. It seems intrinsically reasonable to imply that there is some form of 'mechanostat' that allows the skeleton to judge the level of stress to which it is continuously being subjected and modify its structure appropriately (Frost, 1986). That such a mechanism may have greater effects during growth than at other times also seems obvious. The data to support this view are scant. It also seems self-evident that adequate nutrition is required to .provide the materials for skeletal growth. Guidelines, however, are equally difficult to provide. Even for calcium, which has been studied in greater detail than other nutrients, it is not clear whether current levels of intake are inadequate. The young have a higher efficiency for absorption of calcium across the intestine than adults, offsetting the higher requirement for calcium to supply sufficient to support skeletal growth. Support for increased calcium intake comes from an elegant double-blind controlled study of calcium supplementation in monovular twins (Pocock et al, 1987). This study suggested that an intake of around 1500mg per day increases accretion of skeletal mass in prepubertal children. However, that study
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
(a)
(b) Figure 1. (a) Normal and (b) osteoporotic cancellous bone.
501
502
R. LINDSAY
terminated before these children had reached peak skeletal mass, and therefore it is not clear whether this increased mass would be maintained. Our cross-sectional data suggest that peak bone mass may be modified by as much as 5-10% by ensuring adequate calcium (Kanders et al, 1988). This would be a clinically relevant amount as it would slow the onset of osteoporosis by some 5 years as the population aged after menopause. S K E L E T A L R E M O D E L L I N G AND THE M E N O P A U S A L C H A N G E
In the adult, especially in cancellous bone, there is an ongoing process by which old bone is replaced by new tissue. This process, called remodelling, occurs in packets that are distinct in both time and space (Parfitt, 1984). Each remodelling unit follows a defined sequence of events that has been characterized in histomorphometric studies of the skeleton (Figure 2). At a bone surface to be remodelled, in a poorly understood process termed activation, the cells lining the surface of the bone contract, exposing it to the marrow space. Osteoclasts, multinucleated giant cells of the macrophage lineage, are chemotactically attracted to the site, and begin the process of bone resorption. During a very few weeks, these cells remove bone to the approximate depth of about 50p~m, a depth that appears to be preprogrammed. At that point the osteoclasts disappear from the site, their fate unknown, and are replaced by mononuclear cells whose origin and purpose are not clear. As they link the process of resorption with that of formation, this time in the remodelling cycle has been labelled reversal. In some fashion, these cells may prepare the surface of the resorbed bone for the action of osteoblasts, perhaps secreting the so-called cement line, seen on histological sections as a boundary to the new packet of bone. They may also be responsible for the recruitment to the site of active osteoblasts, which synthesize an organic matrix that includes type 1 collagen called osteoid. The process of filling the cavity takes several weeks to months. The osteoid is subsequently mineralized by an extracellular process forming mature bone. In the best of homoeostatic conditions, bone removed is replaced by an equal quantity of new bone. In the premenopausal woman, this is essentially the case, although there may be somewhat of a deficit at some skeletal sites. As women become oestrogen deficient (or for that matter men testosterone deficient), there are significant changes that occur in the remodelling process. First, there is a clear increase in the frequency at which new 'remodelling cycles are created (Parfitt, 1988). As there is no apparent change in the life span of these cycles, the net effect is that at any single time point more bone surface is in the process of being remodelled. Consequently, there is a remediable deficit in bone mass, that probably amounts to a small percentage of the total cancellous bone mass. In addition, there is a less well documented, but necessary, deficit in the amount of new bone formed within each remodelling unit (Parfitt, 1988). The mechanism, from biochemical data, is presumed to be increased activity of the osteoclast (and consequently a deeper cavity) without an accompanying increase in osteoblast activity, but this has yet to be adequately documented histologically.
503
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
LC 1 QUIESCENCE
POC 2 ACTIVATION
3 RESORPTION
4 REVERSAL
5 FORMATION
LC 6 QUIESCENCE
[~OLD
BONE
[~NEW
BONE
[]OSTEOID
Figure 2.
Consequently, and measurably by techniques that estimate bone mass, there is an irreversible decline in the amount of bone present. Because bone remodelling is a surface phenomenon, and loss of bone tissue removes surface, the decline in mass is self-limited, but will be greatest in cancellous bone that has the highest surface to volume ratio. The proposed increase in the depth of the cavity created by osteoclasts has one further consequence that has been documented histologically and using scanning electron microscopy of cancellous bone (Dempster et al, 1986). This type of bone is constituted by trabecular plates that are of relatively uniform thickness. With increased number of remodelling units present, and increased depth to each cavity, there is a greater chance of penetration of the trabecular plate by osteoclasts. This penetration exaggerates the process of bone loss by removing surfaces upon which new bone could be deposited. As
504
R. LINDSAY
trabeculae are increasingly frequently penetrated they become thinner and more rod-like, until they are disconnected. At this point they are completely removed. The consequence is loss of the entire trabecular structure within the bone (Figure 2). In the vertebrae there is preferential loss of horizontal trabeculae, and thus early weakening of a structure that must withstand vertical loading (imagine a ladder without horizontal rungs). The consequence is vertebral fracture. In long bones, where cancellous bone fills the metaphyseal ends, the consequence is an increase in the risk of fracture on trauma and a subsequent increase in risk of Colles' and hip fractures. The precise mechanism by which these events occur is not entirely understood. Non-invasive studies of bone mass have confirmed the relationship between oestrogen deficiency and reduced bone density, irrespective of the cause of oestrogen deficiency. Thus, low bone mass occurs after menopause or ovariectomy (Aitken et al, 1973; Aloia et al, 1983; Slemenda et al, 1987), as a consequence of hyperprolactinaemia (Klibanski et al, 1980), or hypothalamic causes of amenorrhoea including anorexia (Rigotti et al, 1984) and excessive exercise (Drinkwater et al, 1984; Crilly et al, 1988). Additionally, the use of gonadotrophin releasing hormone agonists as prolonged treatment for endometriosis causes bone loss (Cann et al, 1987). However, although in each case there seems to be an increase in the risk for all fractures, only in the postmenopausal state is the classical fracture distribution commonly seen. (Anorexia, also associated with classical osteoporotic fractures, is a special case, probably because of the calorie deprivation and other metabolic dysfunctions.) Therefore, it appears that other factors must be superimposed upon oestrogen deficiency, modifying the rate and/or extent of bone loss to increase the risk of specific fractures in this population. RISK FACTORS FOR OSTEOPOROSIS
Epidemiologists often make careers from collecting risk factors for disease. This is particularly common among disorders of ageing that mostly have multiple factors involved in their pathogenesis. Osteoporosis is no exception. There is a long list of factors that are thought to increase the risk of osteoporosis among ageing women. Some, for example the use of glucocorticoids in therapeutic doses, are sufficiently potent in themselves to be specific causes of secondary osteoporosis (Lukert and Raisz, 1990). These are considered elsewhere (see Chapter 10) and will not be discussed further here, ' except to say that each is likely to exert a more detrimental effect on the skeleton of oestrogen-deficient women, who already are in negative bone balance. The factors to which the physician must pay attention are those that are potentially remediable (Tables 1 and 2). Thus, factors such as a positive family history, which may be of some value when talking to patients about risk, or may bring the patient to the physician to discuss risk, are clearly not of value when discussing changes to be effected by the patient that might reduce risk. For the most part, for healthy perimenopausal patients, risk factors that must be considered involve nutrition and life-style.
PATHOGENESIS OF POSTMENOPAUSALOSTEOPOROSIS
505
Table 1. Risk factors for osteoporotic fracture. Low bone mass Previous atraumatic fracture Risk factors for falls
Disorders of confusion (e.g. Alzheimer's disease) Medications causing confusion (long-acting tranquillizers, alcohol etc.) Medications: diuretics, hypotensive agents Neuromuscular disease (e.g. Parkinson's disease) and frailty Extrinsic factors: poor lighting, loose carpets and wires, and other home hazards Table 2. Risk factors for low bone mass.
Genetic Endocrine
Nutritional
Life-style Drugs
Race, sex, positive family history Low body-weight Ovarian insufficiency Cushing's syndrome, including therapeutic use of glucocorticoids Excess thyroid hormone Type 1 diabetes mellitus Hyperparathyroidism (low cortical bone mass only) Calcium insufficiencyand malabsorption Alcohol excess Low body-weight (e.g. anorexia nervosa, cyclicaldieting) Caffeine, protein and phosphate excess (?) Sodium, hydrogen ion excess (??) Physical inactivity Cigarette consumption Physical activity to excess (pituitary-ovarian dysfunction) Thyroid hormone Prednisone and other steroids Loop diuretics (frusemide) Anticonvulsants (?) Gonadotrophin releasing hormone agonists and antagonists Methotrexate
Nutrition
There are abundant data to indicate that calcium deficiency can produce osteoporosis in both animals and humans (Fanstat, 1962; Foresta et al, 1983). Insufficient calcium in the diet, or calcium malabsorption, results in the utilization of skeletal calcium to help maintain serum calcium levels within the normal range. In its most obvious state calcium deficiency results in overt secondary hyperparathyroidism. However, in normal clinical practice this is seen only in situations such as adult coeliac disease or gastrointestinal resection, where severe calcium malabsorption is occasionally seen (excluding renal disease which has different consequences). Most adult w o m e n have parathyroid h o r m o n e (PTH) levels within the normal range. Therefore, if calcium deficiency is present, the P T H level must be abnormally high for that individual, although still within the normal range. The analogy is with thyroid replacement therapy. H e r e , the practice used to be to give sufficient thyroid h o r m o n e to keep the circulating levels of T4 and T3 within the normal range. With the development of the supersensitive thyroid stimulating h o r m o n e (TSH) assay it has b e c o m e clear that m a n y of these
506
R. LINDSAY
individuals are over-replaced with completely suppressed TSH levels and require a downward adjustment in dose even though thyroid hormone levels are normal. From the data produced over many years by Heaney and co-workers (1978a; 1978b), it has become clear that calcium absorption is extremely variable across the normal population. Consequently, while one individual may be replete in calcium at an intake of say 600 mg/day, many others will not. The average intake required to bring 90% of the population into the calcium replete state appears to be about 1000-1500 mg/day. Because there is no test for calcium deficiency that is easily applied clinically (like ferritin for iron) and because intakes at that level appear to be safe, there is a general recommendation that adult women increase their calcium consumption to around these values (National Osteoporosis Foundation, 1992a). For most women, at that level of intake calcium deficiency will not add to the skeletal insult imposed by oestrogen deficiency. Other dietary factors that have been suggested as risk factors for osteoporosis include salt, caffeine, protein (especially animal protein) and phosphate (Mazess et al, 1990). Since these appear to be minor risk factors, if indeed they are risk factors at all, the only recommendation required is to ensure that intake is moderate. Several are risk factors for other diseases, so their intake is often limited, at least in the North American population. Alcohol links nutrition and life-style. Excessive alcohol intake is clearly detrimental to the skeleton and should be discouraged (Bikle, 1988). The level of intake that is toxic to bone is not clear but may well be within the range of 'social' drinking among adult women. Thus, we ask for complete abstinence, and expect some reduction in intake. Alcohol, in addition to its effects on the skeleton, also increases the risk of falling and injury among the elderly and will predispose to fracture for that reason. Thus, the elderly in particular should be discouraged from alcohol use.
Life-style There are normally two factors considered under the heading of life-style. The first is cigarette consumption, which is often linked to high alcohol intake. There are data of varying standard to suggest that cigarette consumption is an independent risk factor for osteoporotic fracture (Daniell, 1976). Moreover, since there are major health consequences from cigarette use, this becomes yet another reason to encourage patients to stop, irrespective of the quality of the data. The second life-style factor is physical activity. While it is clear that disuse increases bone loss (Minare et al, 1974), there is less information on the ideal level of physical activity within the normal range that would be most beneficial for the skeleton (Cooper et al, 1988; Gutin and Kasper, 1992). There is equal doubt about the types of activity, although most would agree that weight-bearing activity is best (National Osteoporosis Foundation, 1992b). It seems likely that the effects of activity are mostly, if not completely, mediated directly upon the skeleton. Consequently, to achieve an effect on a particular bone, that bone must be stressed as part of the
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
507
exercise programme (Lanyon, 1984). In the absence of specific recommendations, many physicians adopt the guidelines promoted for cardiovascular health and recommend 30 min of aerobic activity three times per week to their patients. In fact, the effects of activity may be mediated on fracture risk by improving strength and skeletal muscle function rather than by their effect on the skeleton. Thus, even for the elderly, an exercise programme is important. For those with low bone mass, or who already have fractures, initiating a programme of progressive strengthening under the supervision of a qualified physical therapist is often beneficial. Other risk factors
Superimposed on these risk factors are those of ill-health that, by virtue of their direct effects on the skeleton, indirect effects through nutrition or life-style, or the effects of the drugs used in their treatment, will increase the risk of osteoporosis. The medications vary from glucocorticoids, to antimitotic agents, to diuretics, thyroid hormone and anticonvulsants. Obtaining an adequate history of medication use is therefore important in evaluating the patient. The main risk factors for future fracture are low bone mass and the presence of a previous atraumatic fracture (Ross et al, 1991). For younger individuals who do not have an increased risk of falling, these seem to explain the majority of risk, while for older individuals they probably add to the risk from falling (Melton and Riggs, 1985). Thus, disorders such as Alzheimer's disease, general frailty and neuromuscular disorders such as Parkinson's disease all increase the risk of falls. In addition, there may be a component of frailty linked to vitamin D metabolism, which is often impaired in the elderly. Recent studies have demonstrated a significant reduction in incidence of hip fracture with just modest supplements of vitamin D and calcium in an elderly population (Chapuy et al, 1992). Since vitamin D insufficiency may be fairly common among the elderly, and is known to cause proximal muscle weakness, the remediation of this may be a possible mechanism for reduced fracture frequency. While it has been commonly assumed by many that low bone mass leads to an increase in the risk of fracture, it is only comparatively recently that data have been published to confirm this assumption (Wasnich et al, 1985; Melton, 1988; Cooper et al, 1991; Cummings et al, 1993). In prospective studies it has now been shown clearly that low bone mass is a predictor of fracture. The measurement appears to be predictive of all fractures, irrespective of the skeletal site at which the measurement is made, with roughly equal predictive power. However, preliminary data do indicate that measurement at a specific site will be more predictive of fracture at that site (Cummings et al, 1993). For example, measurements of the hip appear to be most reliable at predicting fractures of the femoral neck. Thus, for the clinician, bone mass becomes a tool by which the patient's risk of osteoporotic fracture can be assessed. This would be similar to using blood pressure as an indicator of the risk of cerebrovascular disease or cholesterol as a marker of risk for coronary heart disease.
508
R. LINDSAY
MECHANISMS OF BONE LOSS
As noted above, there are alterations in the remodelling cycle with increasing age that result in the loss of bone mass. The mechanisms by which these occur are not entirely understood. Of major importance in the pathogenesis of the disease among postmenopausal women is the decline in ovarian function that occurs at climacteric, resulting in the external clinical event termed the menopause. The role of ovarian insufficiency
The fact that ovarian failure occurs in all women at the average age of 51 years, and is associated with the alterations in skeletal remodelling outlined above, results in the generalization that bone loss occurs among all postmenopausal populations (Lindsay, 1988). It is this ubiquitous finding that makes the relationship between ovarian failure and menopause one of major importance in public health. Since women enter their menopausal years with variable amounts of bone, bone mass at time of menopause is a useful marker for those most at risk of osteoporotic fracture. There is also considerable variability in the rate (and probably duration) of bone toss among postmenopausal women. Thus, in addition to the use of bone mass as an indicator of risk, there has been considerable interest in attempting to predict rates of loss to add to the efficiency with which the clinician can identify those at maximal risk (Christiansen et al, 1987). To date, one algorithm that uses a combination of biochemical markers of remodelling, together with clinical factors, has been published, suggesting that this together with one measurement of bone mass will predict bone mass up to 12 years later with considerable accuracy (Christiansen et al, 1987). However, this has yet to be verified by other laboratories, and would add additional expense to the evaluation of the perimenopausal woman. Thus, biochemical evaluation cannot be recommended for general clinical use at the present time. The importance of the menopause in the pathogenesis of osteoporosis is substantiated by the effects of oestrogen replacement therapy on bone mass and fracture occurrence in the postmenopausal population (Lindsay et al, 1978; 1984; Hutchinson et al, 1979; Christiansen et al, 1980; Weiss et al, 1980; Naessen et al, 1990; Christiansen and Lindsay, 1991). Many studies have demonstrated that oestrogen intervention reduces bone remodelling to premenopausal levels, and inhibits loss of bone. Some studies have continued for as long as 15 years without evidence of bone loss in cortical bone, and cross-sectional data in which vertebral and femoral neck bone mass have been measured suggest that there is minimal if any bone loss at these sites of important osteoporotic fracture, at least for the first 10 years of the studies (Christiansen et al, 1980; Christiansen and Lindsay, 1991). These controlled clinical trial data are confirmed by epidemiological studies indicating that oestrogen intervention is associated with a reduction in the risk of osteoporotic fracture (Hutchinson et al, 1979; Weiss et al, 1980; Naessen et al, 1990). The majority of published studies have evaluated the effects of
P A T H O G E N E S I S OF P O S T M E N O P A U S A L O S T E O P O R O S I S
509
oestrogen on femoral neck fracture. All suggest that there is a significant effect; on average the reduction in risk conferred by oestrogen use appears to be about 50%. Similar data are available for Colles' fracture. Vertebral fracture data are more sparse. One epidemiological type study indicates a significant protection against vertebral fracture in the order of 60%, while one controlled clinical trial suggests similar, although slightly greater, protection using radiographic abnormalities as the end-point (Ettinger et al, 1985; Lindsay, 1987). Thus, interventional data support the concept that the menopause plays an important role in skeletal loss among the ageing female population. The mechanism by which this occurs is not well understood. A few years ago, for the first time oestrogen receptors were detected in cells of the osteoblast lineage (Eriksen et al, 1988; Komm et al, 1988). It has proven considerably more difficult to determine the presence of oestrogen receptors on osteoclasts, although the expression of these specific receptor types may not be required for a skeletal effect. Certainly, in vitro, it is difficult to demonstrate oestrogen effects on osteoclasts (Arnett et al, 1986), even when the culture is contaminated with osteoblasts. There is some suggestion that the degree of contamination with mononuclear cells, and specifically osteoblast-like cells, increases the likelihood that oestrogen will cause inhibition of resorption of devitalized bone in vitro (McSheehy and Chambers, 1986). However, the issue is still far from proven. The fact that oestrogens do not affect the function of osteoclasts directly implies that, if they affect osteoclast function in vivo, this must be mediated indirectly. However, the major effect of oestrogen is to reduce the activation of new remodelling sites (Steiniche et al, 1989), and activation frequency is more likely to be controlled by lining cells, which are thought to be of osteoblast origin (Parfitt, 1988). A variety of responses has been detected in cells derived from osteoblasts (Ernst et al, 1988; Gray et al, 1988). However, one study of osteoblasts derived from human bone specimens found little support for a direct effect on a number of parameters characteristic of osteoblasts (Keeting et al, 1991). Thus, the osteoblast as a mediator of oestrogen action on bone is far from proven. Others have suggested that other mononuclear cells may be involved in the oestrogen effect, principally cells from the marrow. Production of interleukin 1 (I1-1) by human mononuclear cells in the circulation appears to increase after the menopause, and can be reduced by oestrogen administration (Fournier et al, 1990; Pacifici et al, 1991). It has been suggested that stimulated I1-1 production correlates with bone turnover in postmenopausal patients. In another experimental model, I1-6 has been proposed as the mediator of oestrogen action, and the development of osteoporotic changes in the skeleton have been inhibited by antibodies to I1-6 (Jilka et al, 1992). The production of other factors may play a role in the oestrogen effect on bone. These include transforming growth factor [3TGF-beta, insulin-like growth factors i and 2 (Canalis et al, 1991), and an assortment of other general growth factors that occur in abundance in bone. Whatever the local effects of oestrogen it is clear that loss of oestrogen at the menopause also affects calcium homoeostasis. Several years ago it was
510
R. LINDSAY
suggested that the net effect of oestrogen, at the whole body level, is to render the skeleton resistant to the resorbing effects of P T H (Heaney, 1965). Recent data, in which the response of the skeleton to P T H was tested in postmenopausal women, confirmed this effect (Cosman et al, 1993). In addition, oestrogen status also modified the renal hydroxylase response to PTH, increasing the synthesis of 1,25(OH)2D in response to P T H (Cosman et al, 1991). Other studies have suggested that oestrogen can improve the efficiency with which the intestine absorbs calcium from diet, a process whose active component is controlled by 1,25(OH)2D (Heaney et al, 1978). Overall the effects of oestrogen increase the efficiency with which the body utilizes calcium, since urinary calcium concentration also falls as oestrogen status improves. Whether these are the byproducts of a primary oestrogen effect on bone (with reduced mobilization of skeletal calcium) or represent a prime effect of oestrogen in controlling, first, calcium homoeostasis and, only secondarily, skeletal homoeostasis is not at all clear.
SUMMARY Osteoporosis is a disorder of ageing that shares with other disorders of ageing a multifactorial pathogenesis. The important factors for osteoporosis include the diet, life-style and intercurrent factors such as disease. However, it is clear that loss of ovarian function is an important determinant of bone loss, and oestrogen appears to be the key factor involved. Thus, not only does toss of ovarian function result in bone loss, it can be stopped by adequate oestrogen intervention. Numerous techniques are available to measure bone mass non-invasively and to estimate the risk of future fracture. Thus, for the postmenopausal woman who is concerned about osteoporosis, and who is willing to accept intervention to prevent the disease, bone mass measurement allows the clinician to determine the risk of future osteoporotic fracture and to provide intervention if required. Future studies may elucidate whether determination of skeletal remodelling using biochemistry adds significantly to risk determination. This may be required when considering agents other than oestrogen for intervention among asymptomatic women, as these agents primarily affect the skeleton, while the effects of oestrogen are wide ranging in the body.
REFERENCES Aitken JM, Anderson JB & Horton PW (1973) Seasonal variations in bone mineral content after the menopause. Nature 241"59-60. Aloia JF et al (1983) Determinants of bone mass in postmenopausal women. Archives of Internal Medicine 143: 1700-1704. Arnett TR, Lindsay R & Dempster DW (1986) Effect of estrogens and anti-estrogens on osteoclast activity in vitro. Journal of Bone and Mineral Research 1:99 (abstract). Bikle DD (1988) Effects of alcohol abuse on bone. Comprehensive Therapy 14: 16-20. Canalis E, McCarthy TL & Centrella M (1991) Growth factors and cytokines in bone cell metabolism. Annual Review of Medicine 42" 17-24.
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
511
Cann CE, Henzl MR, Burrk K et al (1987) Reversible bone loss is produced by the GnRH agonist nafarelin. In Cohn DV, Martin TJ & Meunier PJ (cds) Calcium Regulation and Bone Metabolism: Basic and Clinical Aspects, pp 123-127. Amsterdam: Elsevier. Chapuy MC, Arlot ME, Duboeuf F et al (1992) Vitamin D 3 and calcium to prevent hip fractures in elderly women. New England Journal of Medicine 27: 163%1642. Christiansen C & Lindsay R (1991) Estrogens, bone loss and preservation. Osteoporosis International 1: 7-13. Christiansen C, Christiansen MS & McNair P (1980) Prevention of early postmenopausal bone loss: conducted 2 years study in 315 normal females. European Journal of Clinical Investigation 10: 273-279. Christiansen C, Riis BJ & Rodbro P (1987) Prediction of rapid bone loss in postmenopausal women. Lancet i: 1105-1107. Consensus Development Conference (1991) Prophylaxis and treatment of osteoporosis. Osteoporosis International 1:114-126. Cooper C, Barker DJ & Wickham C (1988) Physical activity, muscle strength and calcium intake in fractures of the proximal femur in Britain. British Medical Journal 297: 1443-1446. Cooper C, Wickham C & Walsh K (1991) Appendicular skeletal status and hip fracture in the elderly: 14-year prospective data. Bone 12: 361-364. Cosman F, Shen V, Herrington B & Lindsay RA (1991) Response of the parathyroid gland to infusion of (1-34)hPTH: demonstration of suppression of endogenous secretion using IRMA intact (1-84)PTH assay. Journal of Clinical Endocrinology and Metabolism 73: 1345-1351. Cosman F, Shen V, Xie F, Seibel M, Ratcliffe A & Lindsay R (1993) Estrogen protection against bone resorbing effects of parathyroid hormone infusion. Annals of Internal Medicine 118: 337-343. Crilly RG, Anderson C, Hogan D & Delaquerriere-Richardson L (1988) Bone histomorphometry, bone mass, and related parameters in alcoholic males. Calcified Tissue International 43: 269-276. Cummings SR, Black DM, Nevitt MC et al (1993) Bone density at various sites for prediction of hip fractures. Lancet 341: 72-75. Daniell HW (1976) Osteoporosis of the slender smoker. Archives of Internal Medicine 136: 298-304. Dempster DW, Shane E, Horbert W & Lindsay R (1986) A simple method for correlative light and scanning electron microscopy of human iliac crest bone biopsies: qualitative observations in normal and osteoporotic subjects. Journal of Bone and Mineral Research 1: 15-21. Drinkwater BD, Nilson KL & Chesnut CHIII (1984) Bone mineral content of amenorrheic and eumenorrheic athletes. New England Journal of Medicine 311: 277-281. Eriksen EF, Colvard DS, Berg NJ et al (1988) Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241: 84-86. Ernst MC, Schmid C, Frankenfoldt ER & Froesch ER (1988) Estradiol stimulation of osteoblast proliferation in vitro: mediator roles of TGFB, PGE2, IGF1. Calcified Tissue International 42 (supplement 1): 117. Ettinger B, Genant HK & Cann CE (1985) Long-term estrogen therapy prevents bone loss and fracture. Annals of Internal Medicine 102: 319-324. Fanstat T (1962) Hormonal burrs for collagen bundle generation in uterine stoma: extracellular studies of uterus. Endocrinology 71: 878-887. Foresta C, Busnardo B, Ruzza G, Zanattia G & Mioni R (1983) Lower calcitonin levels in young hypogonadic men with osteoporosis. Hormone and Metabolic Research 15: 206207. Fournier B, Feralli J & Rordorf C (1990) Induction of interleukin-6 (IL-6) by interleukin-1 beta (IL-f beta) in MG63 cells. Modulation by oestradiol. Calcified Tissue International 46 (supplement 2): A36 (abstract). Frost HM (1986) The pathomechanics of osteoporosis. Clinical Orthopaedics and Related Research 200: 198-225. Gray TK, Mohan S, Linkhart TA, Williams ME & Baylink DJ (1988) Estrogen may mediate its effects on bone ceils by signalling the observation of growth factors. Journal of Bone and Mineral Research 3: A552. Gutin B & Kasper MJ (1992) Can vigorous exercise play a role in osteoporosis prevention? A review. Osteoporosis International 2: 55.
512
R. LINDSAY
Heaney RP (1965) A unified concept of osteoporosis. American Journal of Medicine 39: 377-380. Heaney RP, Recker RR & Saville PD (1978a) Menopausal changes in calcium balance performance. Journal of Laboratory and Clinical Medicine 92: 953-963. Heaney RP, Recker RR & Saville PD (1978b) Menopausal changes in bone remodelling. Journal of Laboratory and Clinical Medicine 92: 964-970. Hutchinson TA, Polansky JM & Feinstein AR (1979) Postmenopausal oestrogens protect against fracture of hip and distal radius. Lancet ii: 705-709. Jilka RL, Hangoc G, Girasole G et al (1992) Increased osteoclast development after estrogen loss: mediation by interleukin-6. Science 257: 88-91. Johnston CC Jr, Miller JZ, Slemenda CW et al (1992) Calcium supplementation and increases in bone mineral density in children. New England Journal of Medicine 327: 82-87. Kanders B, Dempster DW & Lindsay R (1988) Interaction of calcium nutrition and physical activity on bone mass in young women. Journal of Bone and Mineral Research 3: 145-149. Keeting PE, Scott RE, Colvard DS, Hart IK, Spelsberg TC & Riggs BL (1991) Lack of a direct effect of estrogen on proliferation and differentiation of normal human osteoblast-like cells. Journal of Bone and Mineral Research 6: 29%304. Klibanski A, Neer RM, Beitins IZ, Ridgway C, Zervas NT & MacArthur J (1980) Decreased bone density in hyperprolactinemic women. New England Journal of Medicine 303: 1511-1514. Komm BS, Terpening CM & Benz DJ (1988) Estrogen binding, receptor mRNA, and biologic response in osteoblast-like osteosarcoma cells. Science 241: 81-84. Lanyon LE (1984) Functional strain as a determinant for bone remodeling. Calcified Tissue International 36: 556-561. Lindsay R (1987) Estrogen therapy in the prevention and management of osteoporosis. Obstetrics and Gynecology 156: 1347-1356. Lindsay R (1988) Sex steroids in the pathogenesis and prevention of osteoporosis. In Riggs BL (ed.) Osteoporosis: Etiology, Diagnosis and Management, pp 333-358. New York: Raven Press. Lindsay R, Hart DM, MacLean A, Clark AC, Kraszewski A & Garwood J (1978) Bone response to termination of oestrogen treatment. Lancet i: 1325-1327. Lindsay R, Hart DM & Clark DM (1984) The minimum effective dose of estrogen for prevention of postmenopausal bone loss. Obstetrics and Gynecology 63: 759-763. Lindsay R, Cosman F, Herrington BS & Himmelstein S (1992) Bone mass and body composition in normal women. Journal of Bone and Mineral Research 7: 55-63. Luckey MM, Meier DE, Mandell JP, DaCosta MC, Hubbard ML & Goldsmith SJ (1989) Radial and vertebral bone density in white and black women: evidence for radial differences in premenopausal bone homeostasis. Journal of Clinical Endocrinology and Metabolism 69: 762-770. Lukert BP & Raisz LG (1990) Glucocorticoid-induced osteoporosis: pathogenesis and management. Annals of Internal Medicine 112: 352-364. McSheehy PHJ & Chambers TJ (1986) Osteoblastic cells mediate osteoclastic responsiveness to parathyroid hormone. Endocrinology 118: 824-828. Mazess RB, Barden HS, Bisek JP & Hanson J (1990) Dual-energy X-ray absorptiometry for total-body and regional bone-mineral and soft-tissue composition. American Journal of Clinical Nutrition 51: 1106-1112. Melton LJ III (1988) Epidemiology of fractures. In Riggs BL & Melton LJ III (eds) Osteoporosis: Etiology, Diagnosis, and Management, pp 133-154. New York: Raven Press. Melton LJ & Riggs BL (1985) Risk factors for injury after a fall. Clinics in Geriatric Medicine 525: 539. Minare P, Meunier P, Edouard C, Bernard J, Courpron P & Bournet J (1974) Quantitative histological data on disuse osteoporosis. Comparison with biological data. Calcified Tissue Research 17: 57-63. Naessen T, Persson I, Adami HO, Bergstrom R & Bergkvist L (1990) Hormone replacement therapy and the risk for first hip fracture. A prospective, population-based cohort study. Annals of lnternal Medicine 113: 95-103. National Osteoporosis Foundation (1992a) Calcium Intake Recommendations. Washington, DC. National Osteoporosis Foundation (1992b) Recommendations on Exercise. Washington, DC.
PATHOGENESIS OF POSTMENOPAUSAL OSTEOPOROSIS
513
Pacifici R, Brown C, Puscheck E et al (1991) Effect of surgical menopause and estrogen replacement on cytokine release from human blood mononuclear cells. Proceedings of the National Academy of Sciences of the United States of America 88: 5134-5138. Parfitt AM (1984) The cellular basis of bone remodeling: the quantum concept reexamined in light of recent advances in the cell biology of bone. Calcified TissueInternational 36: 37-45. Parfitt AM (1988) Bone remodeling: relationship to the amount and structure of bone, and the pathogenesis and prevention of fractures. In Riggs BL & Melton LJ III (eds) Osteoporosis: Etiology, Diagnosis, and Management, pp 45-93. New York: Raven Press. Pocock NA, Eisman JA, Hopper JL, Yeates MG, Sambrook PM & Ebers S (1987) Genetic determinants of bone mass in adults. Journal of Clinical Investigation 80: 706-710. Rigotti NA, Nussbaum SR, Herzog DB et al (1984) Osteoporosis in women with anorexia nervosa. New England Journal of Medicine 311: 1601-1606. Ross PD, Davis JW, Epstein RS & Wasnich RD (1991) Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Annals of Internal Medicine 114: 919-923. Slemenda C, Hui SL, Longcope C & Johnston CC (1987) Sex steroids and bone mass; a study of changes about the time of the menopause. Journal of Clinical Investigation 80: 1261-1269. Steiniche T, Hasling C, Charles P e t al (1989) A randomized study of the effects of estrogen/ gestagen or high dose oral calcium on trabecnlar bone remodeling in postmenopausal osteoporosis. Bone 10: 313-320. Wasnich RD, Ross PD, Heilbrun LK & Vogel JM (1985) Prediction of postmenopausal fracture risk with use of bone mineral measurements. American Journal of Obstetrics and Gynecology 153: 745-751. Weiss NS, Ure CL & BaUard JH (1980) Decreased risk of fractures of the hip and lower forearm with postmenopausal use of oestrogen. New England Journal of Medicine 303: 1195-1198.