Progestogens: Thrombosis and Osteoporosis


Observations on osteoporosis from the sixth, century onwards have shown that stress is the main causative factor. This has been confirmed by animal experiments, using rabbits, with stress simulated by the administration of cortisone. For normal health, the hormonal anticatabolic/catabolic ratio should remain on the anticatabolic side. Stress, emotional or otherwise, can tip the balance over to the catabolic side. In the presence of progestational compounds (in the pill or hormone replacement therapy), much lower levels of stress are needed. The main types of osteoporosis are disuse, with a lower blood flow through bone, steroid-induced and thrombus-induced osteoporosis. When the anticatabolic/catabolic ratio tips to the catabolic side, cell membranes become more rigid. Osteogenic precursor cells coalesce to form osteoclasts, which remove bone. Memory cells are affected, and also those responsible for the absorption of food from the gut. Increased catabolic levels also lead to the production of abnormal megakaryocytes. These produce sticky platelets that immediately coalesce re form thrombi. Many can block blood vessels in cortical bone, killing the osteocytes. Later the dead bone is removed by phagocytic cells. Other thrombi are deposited elsewhere in the body, including the coronary arteries. In these ways, hormone replacement therapy may cause osteoporosis, thrombi, heart attacks and senility, rather than preventing them.

Keywords: osteoporosis, heart attacks, senility, stress, anticatabolic/catabolic ratio, osteoblasts, osteoclasts, osteocytes, canaliculi, haemopoietic tissue, abnormal megakaryocytes, sticky platelets, thrombosis, oral contraceptives, synthetic hormones, post-menopausal hormone replacement therapy, menopause.


In order to understand how progestogens cause osteoporosis the basic biological mechanisms need to be ascertained. In the sixth to seventh centuries AD, Paulus Aegineta wrote a comprehensive medical textbook [ 1]. He dealt with osteoporosis as it affected the hip and the vertebrae separately. He wrote that the stimuli were sometimes physical while sometimes "sorrow, care, watchfulness and the other passions of the mind can excite an attack of the disorder". In 1634, Peter Lovve wrote the first medical textbook in the English language [ 2]. He also realized that the `passions' can be involved, and cites sadness and fear, anger, envy and hatred. He dealt with hip fractures separately, and described the different appearance of the bone in pathological fractures in the elderly, as compared with traumatic fractures in the young, but again drew attention to the correlation with the psychological aspect.

Bell in 1824 [ 3] and Ward in 1838 [ 4] recorded more clinical and anatomical observations, but there was no further substantial increase in our knowledge of osteoporosis until the work of Albright in the late 1930s and 1940s [ 5-7]. Albright and his co-workers realized that there is a hormonal background. They showed that hormonally induced osteoporosis frequently occurs in women about 5-10 years after the menopause, but that it could occur in individuals of any age who had taken steroids for therapeutic purposes, They also showed that when disuse is superimposed on senile osteoporosis the effects tend to be magnified, and that old women are more susceptible than men of comparable age to the harmful effects of immobilization.

By 1948, Albright and Reifenstein [ 8] had evidence that an imbalance between anabolic hormones and corticosteroids was involved, and a few years later Reifenstein advocated the use of anabolic steroids to counter the effects of excess corticosteroids [ 9]. At that time, there was a failure to distinguish between `anabolic' and `anticatabolic' properties.


The natural and synthetic steroids are characterized by each having several properties, in different proportions, and with no two steroids `being exactly alike. The properties involved are androgenic, contraceptive, thrombogenic, progestational, catabolic, and so on. There is one term that has caused considerable confusion, because it has been (and is) used for two quite different properties, that may be present in different proportions, and that is `anabolic'.

Within cells, there are the `rough' membranes, responsible for the formation of the intercellular matrices of connective tissues and bone and also of muscle fibres. A compound that stimulates their activity is known as `anabolic', the opposite being `catabolic'.

There are also the `smooth' membranes of the cell walls. In the case of bone and connective tissue cells, and also a number of other cell types, catabolic compounds make these membranes of the cell walls more rigid, while what I have called `anticatabolic' compounds make them more flexible. This flexibility extends to the cells of sinusoid vessel walls, so that with the stimulation of vascular activity by an anticatabolic compound the type of pain due to pressure from engorged blood vessels on nearby nerves is relieved.

The property of rigidity or flexibility of cell surfaces is important for a number of reasons. Flexibility is required for cell proliferation, so that for normal proliferation to occur, as in bone turnover, the anticatabolic/catabolic ratio needs to stay on the anticatabolic side. It also affects the transport of minerals, such as zinc or magnesium, and other chemicals across the cell surface membrane, so that with too rigid a cell surface there can be a considerable mineral imbalance, Cells in the gut are also affected as the balance goes over to the catabolic side, so that food is absorbed less efficiently. Again, there are a variety of biotransformations that take place at cell surfaces, and these are modified.

Anticatabolic compounds are also anabolic, but sometimes only weakly anabolic. Some anabolic compounds are not anticatabolic. It is these that are the most effective when taken for muscle-building purposes. The fact that anabolic and anticatabolic are listed together in reference books has helped to lead to confused thinking.

Because corticosteroids have a diversity of actions, the anticatabolic agents found apparently diverse clinical uses. For normal health the anti catabolic/catabolic ratio has to stay on the anticatabolic side. If it tips over to the catabolic side, there can be a whole range of undesirable consequences. However, there is no convenient, or even meaningful, quantitative method available for measuring anticatabolic activity. Qualitatively it may he assessed by a combination of a variety of histological and observational methods [ 10-13]. It is then possible to explain clinical observations in terms of the biological mechanisms involved.

Stanozolol [ 14-16] is the most effective of the anticatabolic agents developed in the 1940s to 1960s. It has the least androgenic activity. It was originally developed to counter post-menopausal problems, which it did very successfully, partly because it had no oestrogenic or progestational activity, and partly because the post-menopausal symptoms are due to an anti catabolic/catabolic imbalance. It was soon shown to reverse the osteoporotic process when this was of the type caused by the action of excess corticosteroids on bone cells,

There are changes in the anticatabolic/catabolic ratio during life. Thus, under normal conditions, there are swings in the ratio during adolescence, with the gap sometimes larger and sometimes smaller. This is needed for proper bone growth and development [ 10, 11, 17]. The progestational compounds in oral contraceptives can tip the balance to the catabolic side. The consequences of this include interference with bone development that can lead to arthritis and other complications later in life.

After adolescence, there is a fairly large gap between the anticatabolic and catabolic levels, while the progestational compounds in oral contraceptives reduce that gap. At the time of the menopause, there are changes, completely natural changes, that result in a lowering of the anticatabolic levels, so that the gap narrows. At that time, there are changes in the intervertebral discs that have the effect of strengthening them [ 10]. If these do not take place, as a result of extra oestrogen being given, the spine becomes more vulnerable to the consequences of later osteoporotic changes.

In old age the production of both catabolic and anticatabolic hormones decreases, with the gap narrowing still further. This again is a completely natural change, but as it happens people become more vulnerable to the effects of stress.

The main causes of stress are the unpleasant emotions, frustration, trauma, pain and severe illness. Mild stress activates the adrenal medulla, producing adrenaline, which can cause temporary changes in liver metabolism, leading to fat deposits in the arteries. There is always a certain amount of cortisol in the blood, produced by the adrenal cortex, but more severe stress liberates extra cortisol, which may be sufficient to tip the anticatabolic/catabolic balance to the catabolic side. This can cause osteoporosis, thrombosis and other stress diseases.

In the case of severe stress, with high corticosteroid levels, emotional symptoms may vary, according to the level of stress, from a nervous state and depression, through a chronic anxiety state, to severe psychotic episodes. In many ways, this is also a description of senile dementia, with the chronic state partly due to the effects of increasing numbers of microthrombi (see the section on thrombus formation). Some steroids (e.g. some oestrogens) cause euphoria, but a genuine anticatabolic agent causes an increased feeling of well-being in people who are ill, but not in other individuals. It also prevents memory loss.

With the administration of stanozolol, doctors in geriatric hospitals have reported a definite increase in the feeling of well-being in their patients. The lifting of the symptoms of depression is followed by a better appetite for food and drink, and so there is increased health and mobility from that cause also.

The effect of the progestational compounds present in oral contraceptives and hormone replacement therapy (HRT) is to reduce considerably the level of stress required to produce symptoms. Since another of the properties of these progestational compounds--and also some oestrogenic compounds--is to produce large numbers of microthrombi, the symptoms of senility and senile dementia can appear at a much earlier age than previously, This seems to be the main reason why the name was changed to Alzheimer's disease. Yet the prevention of Alzheimer's disease is now being added to the prevention of osteoporosis and heart attacks as a selling point for HRT.


The blood flow through bone is partly the result of the pumping action of the heart pushing it in, and partly the pumping action of the muscles pulling it out. In the case of the long bones of the legs, this is about 50:50. Within the bones, as well as arteries and veins, there are also the very thin-walled blood vessels known as sinusoid vessels. Unlike the arteries and veins that remain open and full of blood all the time, these open and shut, with only about 10% of them open at any one time. Changes occur rhythmically every 1 or 2 min [ 17, 18].

As well as the cortical bone in the shafts of the long bones, there is also trabecular or cancellous bone. Ward [ 4] showed that this is laid down along lines of force to maintain the mechanical strength of the bone. The sinusoid vessels remain open the most frequently in those directions.

Given a suitable stimulus the cells in the sinusoid vessel walls divide to form osteogenic precursor cells [ 19-21]. Cell processes link them all together into a network. Those that are about to lay down new bone enlarge, and are then known as osteoblasts (Fig. 1). These lay down the organic matrix of the bone tissue, which soon has first phosphates and then calcium from the tissue fluids deposited on it, which form the hydroxyapatite of bone mineral. Johnson [ 18] used the full range of histological stains to follow these changes in detail. Figure 2 shows a section of an osteoblast, with its cell processes entering the bone tissue. The matrix in the immediate vicinity of the cell is not yet calcified. (Unlike the other illustrations, this photograph is of a white object on a dark background.) Once the osteoblasts have become buried in the bone tissue that they have made they are known as osteocytes. These remain joined together by the cell processes that pass through narrow channels in the bone, the canaliculi. When active bone production ceases, a row of inactive osteoblasts remains on the bone surfaces, to maintain the connections between the osteocytes and sinusoid vessels.

Figure 3 is from a section of the cortical bone of an adult human, stained to show the interlinked cell processes passing through the canaliculi. By means of these the osteocytes maintain their health, receive oxygen and various nutrients from the blood vessels to which they are attached, and discharge carbon dioxide and other waste products. In the absence of viable cells the organic matrix of the bone tissue partially degrades.

In both the formation of bone and its removal the bone mineral plays a passive role. Its composition depends on the composition of the underlying organic matrix [ 22]. The same mineral phase, hydroxyapatite only, is found in both normal and osteoporotic bone [ 23]. The changes which occur as a preliminary to bone resorption are in the organic matrix. A comparison of decalcified bone matrix in normal recent bone, unfixed bone from circa 150 AD and fossil bones showed intact collagen fibrils in each. By contrast, in regions of osteoporotic bone where there are no viable osteocytes, all the original fibrillar structure had gone after decalcification [ 24].

The osteoclasts that remove bone tissue are formed from two different precursor cells. Unlike osteoblasts, with their cytoplasm present as a gel, both types of osteoblast have a fluid cytoplasm. As they progress to their active phase, many organelles form in their cytoplasm to produce the enzymes that degrade bone matrix. In Fig. 4, an active osteoclast is heavily stained, while the lightly stained one is in the cavity that it has just produced. The life of an osteoclast is about 2 days.

One type of osteoclast is formed from osteogenic precursor cells. In the presence of excess corticosteroids the cell walls become less flexible, while attempting to achieve their minimum area. This results in the coalescence of cells to form osteoclasts. Figure 5 shows such osteoclasts in the process of formation.

In the bone, as in other tissues, there are phagocytic cells that go under a variety of names, but which are usually called macrophages. Their expansion is inhibited by corticosteroids. When the anticatabolic/catabolic balance reverts to the anticatabolic side, they expand if dead bone is in the vicinity, coalesce, and act as osteoclasts to remove the dead bone. The presence of thyroid and parathyroid hormones is also required for their formation [ 25].

From the early 1940s, there were substantial increases in the knowledge and understanding of osteoporosis, which I reviewed in 1973 [ 10, 26]. The three main types of osteoporosis recognized clinically are disuse, post-menopausal and senile. These roughly correspond with the three distinctive biological mechanisms--disuse, steroid-induced and thrombus-induced osteoporosis.

Disuse Osteoporosis
A total of partial diminution of muscle activity results in a diminished blood flow through the bone. This causes a local increase in carbon dioxide levels, as the removal of waste products from cells is hindered, The result of this is that osteoclasts formed from macrophages become active [ 27].

With a sudden diminution of activity, as when a limb is put in plaster, there is initially an increase in pressure and a build-up of carbon dioxide levels. It takes approximately 2-3 weeks before the pressure and carbon dioxide levels revert to normal [ 28, 29]. During this process, there is a remodelling of the cortical bone, to a lesser total quantity [ 30].

Should muscle activity be restored, the increase in blood flow, and thus the increase in available oxygen, stimulates osteoblast activity, so that more matrix is laid down, and with the restoration of a normal blood flow the quantity of bone tissue is restored to normal. The rate of bone formation, however, is only about a tenth of the rate of bone removal [ 18], so that the time for recovery is correspondingly longer.

Disuse osteoporosis is frequently superimposed on corticosteroid-induced and thrombus-induced osteoporosis. In the former case, exercise can help to reverse the problem. In the latter case, many thrombi are deposited in the muscles surrounding the bone, so that the blood supply can never be fully restored.

Corticosteroid-induced Osteoporosis
The main corticosteroid. in the blood is cortisol, a certain amount being needed for the normal functioning of the body. Problems arise when more is being produced as a result of stress, so that the anticatabolic/catabolic balance is tipped to the catabolic side. When corticosteroids are administered therapeutically, a lower level of stress produces the undesirable side-effects [ 31]. Similarly, with the pill, and now with combined HRT (which contains the same compounds), a lower level of stress is sufficient to tip the balance to the catabolic side, and a very low level when the progestational components also have catabolic properties. This is the main cause of the escalation in the number of cases of osteoporosis since about 1960.

In rabbit experiments [ 10-13, 32, 33], cortisone was used to simulate stress, because about a third of it is transformed to cortisol on cell surfaces. The cortisol is the active compound. It rapidly stabilizes cell membranes, with their mobility greatly reduced, and leads to a retraction--or attempted retraction--of the cell processes and a rounding of the cells [ 34, 35]. (Rabbits were chosen as experimental animals because the cells in their bone marrow behave in a similar manner to the cells in the bone marrow of humans. The hormones in rats and mice are so different from the human that they should not be used in experiments on bone, or on hormone or steroid properties, because the results of such experiments would almost certainly be misleading in a human context.)

When cortisone was given to simulate stress, osteoclasts were formed on every bone surface. Osteoclast activity continued until all the osteogenic precursor cells were used up. In the presence of the progestational compounds used in oral contraceptives and HRT, a lower dose of cortisone (or level of stress) was needed to have the same effect.

Oestrogens have two relevant properties. They double the half-life of cortisol in the blood, and they are only weakly anticatabolic. When they were administered alternately with cortisone, they caused osteogenic precursor cells to proliferate, but not produce new bone. In that way, with unfortunate timing, the greater part of the cancellous bone in the body can be removed, along with a substantial amount of the cortical bone--something that has also been observed clinically.

As osteoclasts remove bone tissue, by partially degrading the organic matrix [ 36, 37], the degradation products spread into the surroundings, and provide nutrition for the proliferation of developing blood cells, so that red marrow is formed wherever there is bone turnover [ 38].

With women who have corticosteroid-induced osteoporosis, intra-capsular fractures of the femur are common. The femur is weight-bearing, and as the trabeculae are thinned some of them break. Dead bone is then removed, and there is proliferation of haemopoietic tissue. In Fig. 6, the areas of active proliferation of blood precursors (the grey areas) show where this usually occurs, and this is where one finds the intra-capsular fractures. Other bones where pathological fractures due to this type of osteoporosis occur are the vertebrae and distal forearm (Colles fracture) [ 39, 40].

Thrombus Formation and Thrombus-induced Osteoporosis
The second type of stress-induced osteoporosis results from abnormalities in the megakaryocytes in the bone marrow when excess corticosteroids are present. Megakaryocytes are the cells that produce platelets. With excess corticosteroids present in the blood the formation of megakaryocytes is increased [ 32].

When rabbits were given testosterone and progestational steroids, together with cortisone to simulate stress, there were different appearances of the megakaryocytes for each compound [ 33]. Those differences showed in histological sections, but not in marrow smears [ 41]. The compounds investigated were testosterone, lynoestranol, norethandrolone, norethynodrel and norethisterone. Three different doses of cortisone were administered, but with the same dose for each of the progestational compounds. Norethynodrel and norethisterone produced some very large megakaryocytes. At intervals, full megakaryocytes shed their platelets, and then more formed round the same nuclei [ 33].

An obvious question was how comparable were these appearances in the rabbits to what happens in humans. In the case of testosterone, sections of rabbit bone were compared with sections from the vertebrae of men who had died from long and stressful illness. The appearance of the megakaryocytes was the same in both (Fig. 7). For the progestational compounds in oral contraceptives, Professor Mason showed me sections from women who had died as a result of taking the pill. In each case I named the correct progestational compound (unpublished work). Figure 8 shows the appearance of the megakaryocytes produced in the presence of norethysterone. (Haemopoietic tissue is not normally sectioned at post-mortem, while the distinctive appearances do not show up on marrow smears.)

Sharp [ 42] had demonstrated that there are two separate stages in the clumping of normal platelets. First, reversible aggregates are formed, then later there is irreversible fusion. By contrast, many of the platelets produced by the abnormal megakaryocytes were very sticky, and clumped together almost immediately into what looked like lumps of almost featureless gel, that soon stuck to vessel walls. Figure 9 shows a small thrombus sticking to the wall of a synovial vessel taken at operation. It shows the beginning of fibrous organization. Figure 10 shows a vessel in the lumbar region of an elderly man. An organized deposit more than half blocks the vessel.

Many of these thrombi are deposited in or near the bone. Figure I I shows a recently formed thrombus blocking a vessel in cancellous bone removed at operation from a patient with rheumatoid arthritis. Figure 12 shows another recently formed thrombus blocking a vessel in the compact bone of an elderly person. Some osteocytes in the surrounding bone are still viable, but others have died. Soon all the osteocytes supplied by that vessel will have died, and the solid deposit will have calcified. Bone in such patients contains many vascular canals with calcified deposits, surrounded by dead bone.

The organic matrix in such dead bone begins to degrade, and when for any reason (e.g. bed rest) the anticatabolic/catabolic ratio is restored to the anticatabolic side, osteoclasts formed from macrophages remove the dead bone. This considerably weakens the cortex, particularly in regions near the ends of the bones where most of the abnormal platelets are produced. Extra-capsular fractures of the hip result from this cause.

What is clear from this account is that the changes in the bones in these two types of osteoporosis, corticosteroid-induced and thrombus-induced, are quite different, although one may be superimposed on the other and disuse superimposed on either. One is reversible, the other can never be completely reversed. Also, in thrombus-induced osteoporosis the extent of dead bone does not become apparent until conditions become favourable for it to be resorbed, when there can be a rapid diminution of total. bone mass. In this situation, bone density measurements can give misleading results.

In both cases it needs to be remembered that the condition is more generalized, affecting all parts of the body. Thus, in the rabbits given cortisone together with the progestational agents, small thrombi were scattered all through the organs and tissues, including the liver and brain, and in some cases large thrombi were seen. As with cortisone administered by itself the cells in the liver enlarged and took on a watery appearance, while its weight almost doubled.

When thrombi blocked vessels in the liver, an area of the surrounding tissues became necrotic [ 33, 38]. In the presence of norethandrolone or testosterone, there was a proliferation of granulation tissue around the affected patch. Cells coalesced to form multi-nucleated phagocytic cells which removed the necrotic tissue, and the gap was filled by fibrous tissue. With norethynodrel and norethisterone the necrotic tissue calcified. With lynoestranol, both a mild granulation tissue reaction and calcification have been seen in the same necrotic patch.

As long ago as 1874, Sir William Osler [ 43] had shown that platelets clump and fuse together into amorphous masses in the blood of some patients with febrile illnesses. Since then, thrombi caused by sticky platelets have been observed in a variety of stressful conditions, even though the fact that they were abnormal platelets does not seem to have been realized.

With thrombus formation, strokes and coronary heart disease are comparatively frequent consequences. Here, it needs to be remembered that combined HRT contains the same progestational compounds that have been responsible for thrombus formation, strokes and heart attacks in women taking oral contraceptives.

In the normal course of events the first effects of stress that can be observed are the atheromatous deposits on artery walls [ 44]. It is only with higher levels of stress that thrombus formation becomes important. Post-mortem examination of young women who had died from thrombi produced as a result of taking progestational steroids presented a different picture [ 45]. There were multiple thrombi, but no evidence of the atherosclerosis that is an accompaniment of the thrombi produced by stress in the presence of natural hormones alone.

Similarly, there was no record of vertebrobasilar occlusion in otherwise healthy young women except for those who had taken contraceptive steroids [ 46]. As with the rabbit experiments, these and related observations indicate that, in the presence of progestational steroids, a lower level of stress suffices to raise cortisol levels to the height required to affect megakaryocyte behaviour.


From the osteoporotic mechanisms described, it is obvious that the action of the progestational compounds in oral contraceptives and HRT would increase both types of stress-induced osteoporosis [e.g. 47, 48]. It has been found that the lifetime risk of any of the types of fracture involved (hip, spine or distal forearm) has been three times more for women than for men over the age of 50 years. For the distal forearm fractures (mainly from corticosteroid-induced osteoporosis) the risk was over six times greater for young urban women than for men [ 40] in countries where progestational compounds have been prescribed extensively, such as the US, Scandinavia and the UK. Figure 13 (Figure 1 in Ref. 40) gives the observations for a typical US city. By contrast, among rural populations in Africa, Asia and the southern Mediterranean region, there are lower fracture rates and sex ratios approaching unity. (In considering the large increase in older women, it must be remembered that the pill has been prescribed since the early 1960s, while more recently progestational compounds have been prescribed as components of HRT.)

Further, even apart from work showing the thrombotic mechanism, it has been known since the 1960s that oral contraceptives can cause heart attacks and strokes. Now, after years of acceptance of misleading studies, epidemiologists are at last showing that HRT can also cause heart attacks [ 49, 50].

It is becoming more urgent to question how it has come about that two of the main adverse effects of steroids used in HRT, namely osteoporosis and heart attacks, have become the main selling points of HRT, with claims that it prevents them.

For this type of problem, many statistical surveys are almost meaningless. Complications that cannot effectively be taken into account arise because the level of stress needed to tip the balance over to the catabolic side varies from one individual to another; the causes and degrees of stress vary, while each of the thrombogenic compounds has different properties. Therefore, it would be most unwise to place reliance on the results of surveys where there are relevant factors that have not been taken into account.

Yet another factor that should not be overlooked is that there tend to be withdrawal symptoms when administration of HRT ceases [ 51]. This can be very stressful, so that we can expect an upsurge of osteoporosis, heart attacks and Alzheimer's disease at that stage.

Because of the considerable increase in the number of cases of osteoporosis as a result of the pill, HRT and overuse of corticosteroid therapy, various suggestions for treatment are being acted on without any genuine attempt to assess their validity. Thus, because bone is a calcified tissue, one simple-minded suggestion has been to give excess calcium. This ignores the fact that the appearance of calcium is the last stage in bone formation.

Similarly, because osteoclasts remove bone, compounds have been developed, and put on the market, that are said to inhibit osteoclast activity. This seems to have been done without any valid animal investigations. One might question the fate of the osteocytes in neighbouring bone. Their death could lead to massive resorption of dead bone at a later stage. Also, there would seem to have been no attempt to assess the haematological consequences. Blood cell formation depends primarily on osteoclasts breaking down bone matrix during the course of normal bone tissue turnover.

If the progestational compounds had been introduced for therapeutic reasons, instead of population planning reasons, they would have been banned by the middle of the 1960s. The only real answer to the clinical problems is still to ban the use of progestational compounds.

FIG. 1. Metaphyseal vessel in long bone of puppy. On either side, there are direct intercellular connections from cells in the vessel walls, through intermediate cells to the osteoblasts on the bone surface. The region occupied by the intermediate osteogenic cells has a high fluid content, resulting in empty spaces (white) in the dried specimen (x 280).

FIG. 2. Osteoblast with cell processes entering neighbouring bone tissue. Some of the newly laid down bone matrix has not yet calcified. (This photograph is of a white object on a dark background (x 700).).

FIG. 3. Mature human bone tissue, with silver stain to demonstrate the presence of canaliculi, which link one osteocyte to another, The cells are elongated parallel to the surface at which they were formed, and most cell processes are found at their sides (x 260).

PHOTO (BLACK & WHITE): FIG. 4. Cortical bone in healing fracture in rabbit, with osteoclasts; in the process of removing dead bone. An active osteoclast is heavily stained, while a lightly stained osteoclast is in the cavity that it has already produced (x 94).

PHOTO (BLACK & WHITE): FIG. 5. Metaphyseal bone of rabbit. Groups of osteoblasts and osteogenic precursor cells are in the process of coalescing to form osteoclasts (x 128).

PHOTO (BLACK & WHITE): FIG. 6. Section through head of femur from a woman with post-menopausal osteoporosis. This was taken during an active phase, and there is an abundance of haemopoietic tissue. In the area where there has been little or no weight-bearing, many of the trabeculae have been resorbed. This weakens the structure, and could lead to an intra-capsular fracture. Because of the effect of raised cortisol levels on cartilage, arthritic changes that would normally be caused by inadequate use of the hip are minimal.

PHOTO (BLACK & WHITE): FIG. 7 Bone marrow from rabbit given cortisone for 2 weeks, followed by cortisone and testosterone for 2 weeks. Many megakaryocytes in this section were in the process of losing platelets, while newly deposited thrombi were seen in the liver and elsewhere in the body (x 300).

PHOTO (BLACK & WHITE): FIG. 8. Bone marrow from rabbit which had received cortisone for 2 weeks, followed by cortisone together with norethisterone for 2 weeks. Mid-shaft: The appearance of the megakaryocytes, and also of the surrounding red marrow, is distinctive for each progestational compound (x 300).

PHOTO (BLACK & WHITE): FIG. 9. Vessel in synovial tissue taken at operation. A small deposit attached to the vessel wall shows the beginning of fibrous organization (x 145).

PHOTO (BLACK & WHITE): FIG. 10. Vessel in muscle in lumbar region, taken at post-mortem, after a lengthy terminal illness. Stained to show the internal elastic lamina. An organized deposit more than half blocks the vessel. A few muscle fibres can be seen in the comer of the photograph (x 145).

PHOTO (BLACK & WHITE): FIG. 11. Bone removed at operation from a patient with rheumatoid arthritis. A recently formed thrombus blocks a vessel in cancellous bone (x 145).

PHOTO (BLACK & WHITE): FIG. 12. A recently formed thrombus in compact bone of an elderly person. The Haversian canal is blocked. Some osteocytes in the surrounding bone are still viable, but others have died (x 145).

GRAPH: FIG. 13. Figure from Ref. [ 40], Age-specific incidence rates for hip, vertebral and distal forearm fractures in (a) men and (b) women. Data derived from the population of Rochester, Minnesota, USA. Reprinted by permission of Elsevier Science and W. B. Saunders Company Ltd.

[1] Paulus Aegineta. The seven books of Paulus Aegineta (trans. from the Greek by Francis Adams, London Sydenham Society 1844-7), Book 3, Sect. LXXVIII, 6th-7th Century AD.

[2] Peter Lovve. A Discourse of the Whole Art of Chyrurgerie. Compiled by Peter Lovve, Scottishman. 3rd edn; corrected and much amended. Printed by Thomas Purfoot, London 1634.

[3] Bell B. Remarks on Interstitial Absorption of the Neck of the Thighbone. Edinburgh: Maclachlan and Stewart, 1824.

[4] Ward FO. Outlines of Human Osteology, London: Henry Renshaw, 1838.

[5] Albright F, Bloomberg E, Smith PH. Postmenopausal osteoporosis. Trans Assoc Am Physns 1940; 55: 298.

[6] Albright F, Smith PH, Richardson AM.. Postmenopausal osteoporosis: its clinical features. JAMA 1941; 116: 2465.

[7] Albright F. Osteoporosis. Ann Intern Med 1947; 27: 861.

[8] Albright F, Reifenstein EC. Parathyroid Gland and Metabolic Bone Disease. Baltimore: The Williams and Wilkins Company, 1948.

[9] Reifenstein EC. The rationale for the use of anabolic steroids in controlling the adverse effects of corticoid hormones upon protein and osseous tissues. South Med J 1956; 49: 933.

[10] Little K. Bone Behaviour. London: Academic Press, 1973.

[11] Little K. Interactions between catabolic and anabolic steroids. Curr Ther Res 1970; 12: 658-76.

[12] Little K, Munuera L. Some mechanisms of action of stanozolol (Stromba) and its interactions with cortisone. Curr Ther Res 1970; 12: 291-305.

[13] Little K. Steroid levels and infection. BMJ 1968; 17 Aug: 432.

[14] Samuel J, Lumbroso A. Le Stanozolol dans le traitment des osteopathes decalcifiantes diffuses. La Semaine des Hopitaux (Semaine Therapeutique) 1964; 40: 412-5.

[15] Hioco D, Miravet L, Lumbroso A. Traitment des Osteopathies decalcifiantes avec Stanozolol, Annales d'Endocrinologie, Paris 1964; No. 2; 25: 253-6.

[16] Louyet P, Gaucher A, Benoit P. Clinical Study of a new anabolic agent. Annales Medicales de Nancy 1965; Feb 4: 203-11.

[17] Johnson LC. The Kinetics of Skeletal Remodelling. (Originally presented as a paper at the symposium "Structural Organization of the Skeleton". Johns Hopkins Hospital, Baltimore, 1965.) Publ. by Birth Defects Original Article Series (The National Foundation), 1966, Vol. II, No. 1, pp. 66-142.

[18] Johnson LC. Morphological analysis in pathology. In: Frost HM, ed. Bone Biodynamics. Boston: Little, Brown & Co, 1964; 543-654.

[19] Trueta J. La vascularization des os et l'osteogenese. Revue Chir Orthop 1958; 44: 3.

[20] Trueta J. The housing problem of the osteoblast. J Traumat Mal Prof 1961; 1: 5.

[21] Trueta J. The role of the vessels in osteogenesis. J Bone Joint Surg 1963; 45B: 402.

[22] Wells HG. Calcification and ossification. Harvey Lecture 1910-11. Chicago: JB Lippincott & Co, 1910; 102.

[23] Hardaway EM, Buhr AJ, Little K. Mineral phase in osteoporic bone. Nature 1962; 196: 181-2.

[24] Little K, Kelly M, Courts A. Studies on bone matrix in normal and osteoporotic bone. J Bone Joint Surg 1962: 44B: 503-19.

[25] Burkhart JM, Jowsey J. Parathyroid and thyroid hormones in the development of immobilization osteoporosis. Endocrinology 1967; 81: 1053-1062.

[26] Little K. Osteoporotic Mechanisms. J Int Med Res 1873; 1: 509-29.

[27] Jee WSS, Nolan PD. Origin of osteoclasts from the fusion of phagocytes. Nature (London) 1963; 200: 325.

[28] Geizer M, Trueta J. Muscle action, bone rarefaction and bone formation. J Bone Joint Surg 1958; 40B: 282.

[29] Trueta J. The dynamics of bone circulation. In: Frost HM, ed. Bone Biodynamics. Boston: Little, Brown & Co, 1964; 245-58.

[30] Jee WSS. The influence of reduced local vascularity on the rate of internal reconstruction in adult long bone cortex. In: Frost HM, ed. Bone Biodynamics. Boston: Little, Brown & Co, 1964; 259-77.

[31] Burrows FGO. Avascular necrosis of bone complicating steroid therapy. Br J Radiol 1965; 38: 309.

[32] Little K, White AM. Stimulation of megakaryocyte formation in rabbits by anabolic and progestational steroids administered in conjunction with cortisons. Biorheology 1968; 5: 185.

[33] Little K, The production of platelet thrombi. Curr Ther Res 1970; 12: 677-94.

[34] Dougherty TF, Schneebeli GL. The use of steroids as anti-inflammatory agents. Ann NY Acad Sci 1955; 61: 328.

[35] Dougherty TF, Berliner DL, Berliner ML. Corti costeroid--tissue interactions. Metabolism 1961: 10: 966.

[36] Carrel A, Baker LE. The chemical nature of substances required for colt multiplication. J Exp Med 1926; 44: 503.

[37] Baker LE, Carrel A. The effect of digests of pure protein on cell proliferation. J Exp Med 1928; 47: 353.

[38] Little K. Bone marrow and ageing. Gerontologia 1969; 15: 155-70.

[39] Buhr AJ, Cooke AM. Fracture patterns. Lancet 1959; i: 531.

[40] Cooper C. Epidemiology and public health impact of osteoporosis. Bailliere's Clin Rheumatol 1993; 7: 459-77. Cooper C, Melton U. Epidemiology of osteoporosis. Trends Endocrinol Metabolism 1992; 3: 224-9.

[41] Sharp AA. Personal communication, 1965-6.

[42] Sharp AA. Platelet (viscous) metamorphosis. Henry Ford Hospital Symposium "Blood Platelets". Boston: Little, Brown & Co, 1961; 67.

[43] Oster W. Account of certain organisims occurring in the liquor sanguis. Proc R Soc 1874; 22: 391.

[44] Mason JK. Asymptomatic disease of coronary arteries in young men. BMJ 1963; ii: 1234-5.

[45] Irey NS, Manion WC, Taylor HB, Vascular lesions in women taking oral contraceptives. Arch Pathol 1970; 79: 1-9.

[46] Salmon ML, Winkleman JZ, Gay AJ. Neuro-opthalmic sequelae in users of oral contraceptives. JAMA 1968; 206: 85-91.

[47] Cundy T, Evans M, Roberts H, et al. Bone density in women receiving depot medroxyprogesterone acetate for contraception. BMJ 1993; 303: 13-16.

[48] Sathyamala C, Shah P, Jain Y, et al. Use of injectable depot medroxyprogesterone acetate in lactating Indian women. Lancet 1994; 344: 134-5.

[49] Van der Graaf V, De Kleijn NJJ, van der Schoune YT. Menopause and cardiovascular disease. J Psychosom Obstet Gynecol 1997; 18: 113-20.

[50] Grant ECG, Thrombosis and heart attacks with contraceptive and menopausal hormones. J Nutr Environ Med 1998; 8: 159-67.

[51] White M, Grant ECG. Addiction to oestrogen and progesterone. I Nutr Environ Med 1998; 8: 117-20.


By KITTY LITTLE MA BSc DPHIL 8 Olney Court, Marlborough Road, Oxford OX1 4LZ, UK

Share this with your friends