Secondary Causes of Osteoporosis in Men
In men with osteoporosis, it is critical to exclude underlying pathological causes as these are more likely to be present than in women (Riggs and Melton, 1986). Previous studies have shown that between 30 and 60% of men evaluated for vertebral fractures have another illness contributing to the presence of bone disease (Ebeling, 1998a; Orwoll and Klein, 1995; Francis et al., 1989; Seeman and Melton, 1983; Resch etal., 1992; Seeman, 1993); however, there have been fewer studies of the bone disease present in men sustaining proximal femoral (Nyquist et al., 1998; Boonen et al., 1997a) or other peripheral fractures. This post will focus on the major conditions causing osteoporosis in men which are enumerated in Table I.
I. Glucocorticoid-Induced Osteoporosis
Glucocorticoid excess (predominantly exogenous) is the most commonly identified etiological factor, accounting for 16-18% of cases (Francis et al., 1989; Seeman and Melton, 1983). The pathophysiology of glucocorticoid- induced osteoporosis is similar in women and men, and the most important mechanism is a direct inhibition of the actions of osteoblasts, including the elaboration of growth factors promoting collagen proliferation (McCarthy et al., 1990; Raisz and Simmons, 1985), and decreased osteoblast recruitment (Chuyn etal., 1984). However, muscle weakness, immobility, impaired intestinal calcium absorption, hypercalciuria, and a reduction in serum total and free testosterone levels also may all contribute to glucocorticoid-associated bone loss (MacAdams et al., 1986; Fitzgerald et al., 1997). The cause of the reduction in serum testosterone levels of men receiving treatment with glucocorticoids is incompletely understood. The mechanisms may include central inhibition of GnRH release, suppression of pituitary sensitivity to GnRH, and a direct inhibition of testicular steroidogenesis (Veldhuis et al., 1992). Importantly, it is not uncommon for more than one risk factor to be operative in the same patient, such as long-standing tobacco and alcohol abuse in a man requiring oral glucocorticoid therapy for asthma.
II. Pulmonary Disease and Immunosuppressive Drugs
In men with asthma (Ebeling et al., 1998b), low spinal bone density is related to both the cumulative inhaled glucocorticoid dose and the cumulative exposure to oral prednisolone. However, the cumulative exposure to oral prednisolone is a more important determinant of bone density at both the spine and the proximal femur. Vertebral deformities are common in older men with chronic pulmonary disease and the likelihood of vertebral fracture is also greatest in those men using continuous systemic glucocorticoid therapy (McEvoy et al., 1998). Osteoporosis is also a common accompaniment in adult male survivors with cystic fibrosis. In men with cystic fibrosis aged 25-45 years, overall fracture rates were twofold greater than in the general population, whereas vertebral compression and rib fractures were 100- and 10-fold more common, respectively, than expected in the general population (Aris et al., 1998). Mean standardized bone mineral densities (BMDs) at the spine, femur, and total body were two standard deviations lower than expected. The cumulative prednisolone dose, body mass index, and age at puberty were the strongest predictors of bone density.
TABLE I Etiology of Osteoporosis in Men
Glucocorticoid excess (exogenous or endogenous)
Other immunosuppressive drugs (e.g., cyclosporine A)
Hypogonadism (including treatment for prostatic carcinoma, glucocorticoid- induced and renal insufficiency)
Ineffective skeletal estrogen action (estrogen receptor defects, aromatase deficiency)
Alcohol excess Smoking
Chronic obstructive pulmonary disease and asthma Cystic fibrosis
Gastrointestinal disease (coeliac, Crohn’s diseases, short bowel syndrome, postgastrectomy, total parenteral nutrition, lactase deficiency)
Pernicious anemia Hypercalciuria
Anticonvulsants (phenytoin, phenobarbitone)
Thyrotoxicosis Hyperparathyroidism Immobilization Osteogenesis imperfecta Homocystinuria
Neoplastic disease (multiple myeloma, lymphoma)
Ankylosing spondylitis and rheumatoid arthritis Systemic mastocytosis
In animal models, cyclosporine A therapy leads to high bone turnover and rapid bone loss (Movsowitz etal., 1988). In humans, immunosuppressive therapy with cyclosporine A or tacrolimus for the prevention and treatment of graft versus host disease also results in osteoporosis. The combination of glucocorticoids and cyclosporine A results in rapid bone loss following cardiac, renal, single lung and bone marrow transplantation, and vertebral fractures are not infrequent in this increasingly common clinical setting (Katz and Epstein, 1992). Following allogeneic bone marrow transplantation, bone loss at the femoral neck and lumbar spine is related to the cumulative prednisolone dose and the duration of cyclosporine A therapy used to treat graft versus host disease, as well as the pretransplantation levels of bone resorption (Ebeling et al., 1999).
Although there is no abrupt cessation of testicular function or “andro- pause” comparable with the menopause in women, both total and free testosterone concentrations decline irrevocably with age in men. The decrease in free testosterone concentrations is greater than that for total testosterone because sex hormone-binding globulin levels also increase with age. Finally, age-related decreases in adrenal androgens are even greater than for testosterone and may have more impact on age-related bone loss (Clarke et al., 1994). A limited correlation exists between free testosterone levels with bone density at some, but not all, skeletal sites, and these findings have been inconsistent between studies and skeletal sites (Kelly etal., 1990). In a study of 90 healthy men, after controlling for age, free but not total testosterone concentrations were related to BMD at the femoral neck and Ward’s triangle, but not at the lumbar spine, while an inverse relationship existed between total testosterone and fat mass (Ongphiphadhanakul et al., 1995). The distribution of adipose tissue is also altered in men with postpubertal hypogonadism. Although measurements of visceral fat were similar to eugonadal men, subcutaneous and muscle fat areas were higher (Katznelson et al., 1998).
In contrast to the studies of testosterone, a recent study has shown that BMD at all skeletal sites was significantly positively associated with serum estradiol concentrations in men aged over 65 years (Slemenda et al., 1997). Body weight, age and serum sex steroids accounted for 30% of the variability of BMD in men with consistent positive associations between BMD and
estradiol concentrations. Importantly, within the normal range, lower serum testosterone levels were not associated with low BMD. Khosla et al. (1998) also showed that bioavailable estradiol levels correlated positively with BMD and negatively with bone resorption in normal men aged 23-90 years. Currently, it is not therefore possible to identify a threshold serum testosterone concentration, below which bone loss will occur.
Hypogonadism is one of the most common secondary causes of osteoporosis in men, being present in up to 30% of men with osteoporotic vertebral fractures (Orwoll and Klein, 1995), and low free testosterone and 25(OH) vitamin D concentrations are common in men with hip fractures (Boonen et al., 1997a). Free testosterone and 25(OH) vitamin D concentrations are both independently negatively related to deoxpyridinoline excretion in men with hip fractures, suggesting that increased bone resorption rates in these men may be important in contributing to bone loss at the femoral neck. Fragility fractures and low bone density are also common in men with prostate cancer. The cumulative incidence of osteoporotic fractures in men 7 years after orchidectomy for prostate cancer was approximately 28% (Danieli, 1997). The incidence of fragility fractures in men with prostate cancer treated with luteinizing hormone-releasing hormone agonists for 22 months, on average, was somewhat lower at 5% (Townsend et al., 1997).
Androgens are important in both attainment of peak bone mass and the maintenance of bone mass in adult men. Reduced bone mass is present in men with Klinefelter’s and Kallman’s syndromes and also in men who have had constitutionally delayed puberty. In young men with constitutional delay of puberty (CDP), bone turnover was increased and BMD was decreased, whereas bone turnover was normal and BMD was decreased in men with idiopathic hyogonadotrophic hypogonadism (IHH). Interestingly, after testosterone therapy, BMD increased only in men with CDP and not in men with IHH, indicating that high baseline bone turnover could be a useful predictor of responses to testosterone in men with hypogonadism (Lubushutzky etal., 1998).
In hypogonadism beginning in the prepubertal years, the bone deficit is more marked in cortical bone than in hypogonadism commencing after puberty. It is likely, therefore, that androgens are important in bone modeling and subperiosteal apposition. This is supported by rat studies that show an increase in subperiosteal bone formation rates after oophorectomy compared with a decrease after orchidectomy (Orwoll and Klein, 1995). By contrast, in postpubertal hypogonadism, vertebral bone loss, with reduced trabecular numbers, is greater than appendicular bone loss (Finkelstein et al., 1989).
There is an interaction between testosterone concentrations and vitamin D metabolism. Francis and Peacock (1986) found that l,25-(OH)2D levels were decreased in European hypogonadal males with fragility fractures. Both bone formation parameters and plasma l,25-(OH)2D levels increased after testosterone treatment. However, in animal studies using vitamin D replete,
sexually immature male chicks, circulating l,25-(OH)2D levels decreased, whereas the tissue levels of l,25-(OH)2D and the biological responses of bone and intestine to l,25-(OH)2D increased (Otremski et al., 1997). A French histomorphometric study of men with chronic hypogonadism also demonstrated low bone formation rates (Delmas and Meunier, 1981). By contrast, North American studies of men with fragility fractures have not demonstrated reduced bone formation rates, but a slight increase in mean remodeling rates comparable with postmenopausal estrogen deficiency in women (Jackson and Kleerekoper, 1987). Thus, nutritional vitamin D deficiency may have contributed to the European study findings. It is notable, however, that in all studies trabecular loss was a consistent finding.
Osteoblasts in men contain low numbers of androgen and estrogen receptors (Colvard et al., 1989; Orwoll et al., 1991; Eriksen et al., 1988). Androgens and estrogens affect bone cells by indirect and direct mechanisms secondary to changes in concentrations of both systemic and local factors. Several of these effects including proliferation, growth factor and cytokine production, and bone matrix protein production (type I collagen, osteocalcin, osteopontin) are mediated by the androgen receptor (AR). The mechanisms whereby sex hormones interact with local growth factors to exert their effects on bone cells in men needs to be further elucidated. For example, it is possible that an interaction occurs between decreasing serum concentrations of testosterone and insulin-like growth factor-1 (IGF-I) with aging resulting in reduced bone formation rates and an age-related increase in bone fragility (Boonen etal., 1997b).
Acute testosterone deficiency following orchidectomy is associated with a phase of rapid bone loss and increases in biochemical markers of bone turnover as it is in postmenopausal women (Stepan and Lachman, 1989; Goldray et al., 1993). This is followed by a phase of slower bone loss associated with low bone turnover. Histomorphometric studies have also shown that bone formation rates are reduced in chronic hypogonadism (Otremski etal., 1997; Delmas and Meunier, 1981; Jackson and Kleerekoper, 1987). In addition, it is uncertain whether testosterone therapy of eugonadal men results in significant increases in bone density. One small, nonrandomized study of parenteral testosterone therapy of eugonadal older men detected increases in spinal, but not proximal femur, bone mass (Tenover, 1992); however, a double blind, placebo-controlled trial of transdermal testosterone showed no effect (Orwoll and Oviatt, 1992). A recent open, uncontrolled study of testosterone treatment of eugonadal men with osteoporosis for 6 months showed that a 5% increase in spinal BMD was associated with large decreases in urinary excretion of deoxypyridinoline and N-telopeptide. The increase in spinal BMD was correlated with changes in estradiol but not in testosterone (Anderson et al., 1997).
This is in keeping with increasing evidence that estrogens (Vanderscheueren et al., 1996), as well as androgens, play an important role in skeletal maintenance in men. In men peripheral aromatization of androgens to estrogens occurs, and aromatization also occurs within osteoblasts. Two human models currently exist for inefficient actions of estrogen on the male skeleton. A man with a stop mutation in the estrogen receptor gene and high circulating estradiol concentrations had failure of epiphyseal fusion, continued skeletal growth, and severe osteoporosis (Smith et al., 1994). Similarly, an aromatase-deficient man developed tall stature and osteoporosis, and low-dose estrogen therapy resulted in a large increase in bone density (Kenan etal., 1995; Morishima et al., 1998).
IV. Alcohol and Osteomalacia
Alcoholic men have a high risk of osteoporotic fracture. The link between alcohol abuse and bone diseases has been well established by epidemiological studies (Seeman and Melton, 1983; Resch et al., 1992). In 25-50% of men who present for medical assistance because of excessive drinking, osteopenia will be present (Orwoll and Klein, 1995). The habitual consumption of alcohol is also significantly negatively related to bone mass in men (Seeman and Melton, 1983; Resch etal., 1992; Slemenda etal., 1992), and in a longitudinal study, Slemenda et al. (1992) showed that high alcohol consumption was associated with increased rates of bone loss.
However, the dose dependence of alcohol-related bone loss is unclear. For example, recent studies suggest that modest alcohol intakes may be associated with increases in bone mass. Even those men in the Framingham heart study cohort who were the heaviest drinkers (> 414 mL/week) had BMD that was, on average, 3.9% higher at all sites after adjustment for age, weight, height and smoking (Felson et al., 1995). In addition, men aged 50 years and over from the European Vertebral Osteoporosis Study Group with vertebral deformities did not have a higher frequency of alcohol intake than men without vertebral deformities (Naves Diaz et al., 1997). There was also a small and nonsignificant protective effect of moderately frequent alcohol intake (1-4 days per week) in men aged 65 years and older (OR = 0.81; 95% Cl = 0.62, 1.08). Overall, further prospective data are required to determine at what level of alcohol ingestion the positive benefits of regular intake are overtaken by the increased risk of fragility fracture from alcohol excess.
In many cases of secondary osteoporosis in men, including alcohol- related bone disease, gastrointestinal disease and anticonvulsant use, it is important to exclude vitamin D deficiency as a contributing factor. In the individual man with metabolic bone disease, the degree of vitamin D deficiency may vary from severe (osteomalacia) to mild, but the latter is also important to detect because it may lead to secondary hyperparathyroidism if it is untreated. In men with alcoholic cirrhosis and alcoholic bone disease the free concentrations of both 25(OH)D and l,25-(OH)2D are normal despite low levels of the total hormones (Bikle et al., 1984, 1986). Trans-iliac bone biopsy is the definitive test to diagnose osteomalacia. Nevertheless, a measure of serum 25(OH)D concentration will provide useful information on the degree of vitamin D deficiency present (with the exception noted above).
Tobacco use is associated with decreased bone mass in women (Hopper and Seeman, 1994). In men, hip fracture rates are higher in current smokers (Paganini-Hill etal., 1991) and the relative risk of vertebral fracture in smokers is 2.3 based on a previous cohort study (Seeman and Melton, 1983). This risk is independent of alcohol consumption. Both risk factors had adverse effects on BMD in men who were aged greater than 60 years and whose average duration of smoking was 36 years, indicating that prolonged exposure may be required for effects on fracture incidence. Slemenda etal. (1992) demonstrated in twins discordant for smoking that radial bone loss was 40% greater in the smoking twin than the nonsmoking twin. The number of cigarettes smoked also correlated with the rapidity of bone loss.
A recent study (Vogel et al., 1997) examined effects of smoking on bone density, and of current smoking on rates of bone loss in men. Current and past smokers had significantly lower bone density, particularly at the calcaneus (approximately 4.5%) and distal radius (approximately 2.5%). The effect was linked both to the smoking duration and to the number of cigarettes smoked. The rates of bone loss measured over an average of 5 years in current smokers were 20.5, 27.2, and 9.7% greater at the calcaneus, distal, and proximal radius, respectively, than in the never smokers, but these changes did not reach significance. However, smokers of more than 20 cigarettes per day had a significantly higher rate of bone loss (by 77.6% at the distal radius), consistent with an increase in fracture risk of 10-30% per decade of smoking. This effect was similar to that demonstrated in male twins and slightly greater than the effect on spinal BMD in female twins discordant for smoking, whose BMD deficit was 9.3% for 20 pack-years (one pack-year is equivalent to 20 cigarettes per day for one year) (Hopper and Seeman, 1994).
The mechanism of the adverse effect of smoking on bone mass in men is unknown but it may be related to decreased body weight, decreased calcium absorption, and decreased estrogen levels as it is in women (Krall and Dawson-Hughes, 1991). Smoking may also have a direct toxic effect on bone metabolism. However, its effects on androgen concentrations or growth factor production in men are unknown.
Several nutrients including amino acids, calcium, magnesium, and phosphorus and the fat-soluble vitamins D and K are important for the maintenance of skeletal health in normal men. Gastrointestinal disease predisposes to bone disease as a result of intestinal malabsorption of these nutrients. In men with hip fractures, a low serum albumin concentration, a nonspecific marker of nutrition, was the strongest independent variable correlated with fracture (Thiebaud etal., 1997). Low femoral neck BMD was also correlated with hip fracture risk. Serum insulin-like growth factor-binding protein 3 (IGF-BP3) concentrations were lower in men with hip fractures and also correlated with BMD and albumin. Lactase deficiency can lead to low dietary calcium intakes, increased bone turnover, and osteoporotic fractures in men (Tamatani et al., 1998). In addition to vitamin D deficiency, decreased vitamin K1 and K2 levels are correlated with BMD in osteopenic elderly Japanese men, suggesting that both deficiencies may cooperatively play a role in the etiology of type II osteoporosis in men (Laroche etal., 1995).
In particular, gastrectomy is commonly associated with vertebral osteoporosis in men (Francis et al., 1989; Seeman, 1993). Between 10 and 40% of men have low BMD following gastrectomy (Garrick et al., 1971). It is not currently known whether prolonged use of powerful inhibitors of gastric acid secretion such as omeprazole may also predispose to bone loss in men, but by analogy with pernicious anemia this seems unlikely (see following discussion). Other gastrointestinal diseases, particularly small bowel disease (coeliac disease, Crohn’s disease, small intestinal resection), cause bone disease equally in men and women (Orwoll and Klein, 1995). However, large bowel diseases are uncommonly associated with osteopenia.
Vitamin D deficiency and osteomalacia are associated with malabsorption. However, lesser degrees of nutrient malabsorption more commonly result in other forms of bone loss. For example, Parfitt has described focal osteomalacia (increased osteoid thickness with normal osteoid surface) or atypical osteomalacia (increased osteoid surface with normal osteoid thickness) in patients with postgastrectomy bone disease and without obvious vitamin D deficiency, secondary hyperparathyroidism, or hypophosphatemia (Parfitt and Duncan, 1982). Similar forms of bone disease are also seen in small bowel disease. All are characterized by low mineralization rates. Low turnover osteoporosis may also be seen in men with bowel disease (Rao and Honasoge, 1996). Its etiology is unknown, but in some men treatment with glucocorticoids may be partially or completely responsible.
In cholestatic or alcoholic liver disease, there is an increased incidence of metabolic bone diseases, including osteomalacia and osteoporosis. In men with cirrhosis following necrotic hepatitis, spinal BMD, serum osteocalcin and 25-(OH)D concentrations were lower than in control men. The spinal BMD was correlated with serum 25-(OH)D concentrations or the clinical severity of cirrhosis (Chen et al., 1996). It is therefore important to exclude vitamin D deficiency in men with cirrhosis of any cause or severity.
Hypercalciuria is more than twice as common in men as in women (Smith, 1989). Currently there are data that suggest a link between hypercalciuria and bone loss in some men with osteoporosis. Hypercalciuria or nephrolithiasis in men is associated with osteopenia (Pietschmann et al., 1992). In men with renal hypercalciuria, this may be related to a negative calcium balance with secondary hyperparathyroidism, increased 1,25 (OH)2 vitamin D levels and increased bone turnover rates (Zerwekh et al., 1992) .
Anticonvulsants (phenobarbitone, phenytoin) cause a spectrum of disorders of bone and mineral metabolism associated with hyperosteoidosis in 10-60% of men and women using these drugs (Orwoll and Klein, 1995). This is particularly important in the elderly who are at risk of epileptic seizures as a result of previous stroke or tumor (Cohen et al., 1997) and in institutionalized epileptics. Nilsson et al. (1986) found that institutionalized epileptics had a fracture rate of 10%/year. Bone histological findings from the fracture patients revealed increased osteoid volume, increased osteoclastic resorptive activity, and reduced trabecular bone volume compared with age-matched controls. Thus, the bone disease can include a combination of osteoporosis, osteomalacia, and hyperparathyroidism. Anticonvulsants increase hepatic metabolism of vitamin D and 25-(OH)D to inactive metabolites via induction of cytochrome P4,(l enzyme activity. Circulating and tissue levels of biologically active vitamin D metabolites are decreased, and hypocalcemia, decreased intestinal calcium absorption and hypophosphatemia, secondary hyperparathyroidism, and alterations in bone remodeling result.
In epileptics receiving phenytoin, serum 25(OH)D concentrations correlate positively with serum calcium levels and BMD (Hahn et al., 1972). Total body bone mineral content was 10-30% lower than age-matched normal values. The effect of anticonvulsants on BMD was related to the number of drugs used, the total daily dose, and the duration of drug therapy (Wolin- sky Friedland, 1995). Phenytoin also may directly decrease intestinal calcium absorption by inhibition of cellular calcium fluxes. It inhibits PTH-induced bone resorption in a dose-dependent manner and inhibits collagen synthesis in vitro. In boys, sodium valproate (but not carbamazepine) therapy, resulted in significantly lower BMD by 14 and 10%, at axial and appendicular sites, respectively (Sheth et al., 1995).
IX. Pernicious Anemia
In men and women with pernicious anemia, there is an increased risk of fragility fractures (Goerss et al., 1992). In comparison with fracture rates from the general community, patients with pernicious anemia had a 1.9-fold increase in proximal femur fractures, a 1.8-fold increase in vertebral fractures, and a 2.9-fold increase in distal forearm fractures. The rate of proximal humerus or pelvic fractures was not increased. The mechanism of bone loss in this condition is unknown because women with pernicious anemia have normal true fractional calcium absorption and normal levels of calciotrophic hormones despite achlorhydria, indicating gastric acid is not required for absorption of dietary calcium (Eastell et al., 1992).
X. Thyrotoxicosis and Thyroidectomy
Thyrotoxicosis and a past history of thyrotoxicosis are associated with osteopenia in both men and women. There are few, if any, studies of the effects of clinical or subclinical hyperthyroidism on bone in men. Studies in women suggest that thyrotoxic bone disease and osteopenia are potentially reversible disorders (Diamond et al., 1994). In women with subclinical hyperthyroidism, cortical bone is adversely affected more than trabecular bone (Ross, 1994). No data exist on an increased fracture risk of subclinical hyperthyroidism in either men or women. In men with a history of thyroidectomy, the relative risk of fractures of any of the vertebrae, proximal humerus, distal radius, pelvis, or proximal femur was increased only 1.5-fold (95% confidence interval, 0.7-3.2) compared with age-matched controls (Nguyen et al., 1997). This difference was entirely due to a statistically significant increase in the rate of proximal femur fractures in men who had had a thyroidectomy. Other risk factors for hip fractures included being older at the time of surgery, a greater extent of surgery, and the presence of other risk factors for osteoporosis. Thyroidectomy, performed mainly for adenoma or goiter, had little influence on the risk of fracture at other sites.
In the original bone densitometric studies of patients with primary hy-perparathyroidism, osteopenia occurred predominantly in postmenopausal women and less commonly in premenopausal women or men (Pak et al., 1975) . Nevertheless, vertebral compression fractures are the mode of pres-entation in approximately 4% of patients with surgically proven primary hyperparathyroidism (Dauphine et al., 1975). Survival in men and women following surgical treatment for primary hyperparathyroidism is usually not impaired, but the presence of osteoporosis and muscle weakness and the absence of a history of renal calculi have been associated with reductions in survival (Soreide et al., 1997). In men with severe congestive heart failure, there is a high prevalence of osteoporosis and osteopenia, low vitamin D metabolite concentrations, and high bone turnover rates. Conversely, higher parathyroid hormone (PTH) concentrations were associated with better left ventricular function (Shane et al., 1997). In men with chronic renal failure treated with chronic ambulatory peritoneal dialysis and l-a-(OH)D, for 2 years, there was no change in bone mineral content of the distal radius, whereas in age-matched postmenopausal women, it decreased by 6% per year. The greater bone loss in women could therefore indicate an additive effect of hypogonadism on bone loss in secondary hyperparathyroidism in uremia, as also exists for primary hyperparathyroidism (Lyhne and Pedersen, 1995).
Prolonged bed rest is similar to weightlessness in that it can result in bone loss. Bone lost during immobilization can be regained after weight-bearing is recommenced, but a deficit in bone mass is likely to result (Donaldson etal., 1970). Osteopenia is also a rapid, constant, and permanent accompaniment of quadriplegia and paraplegia. In men with hemiplegia, the ipsilateral femoral neck BMD, measured by DXA, was lower than in the contralateral femur, but the difference was less than in women. The duration of immobilization comprised only 5% of the total variance of bone loss (del Puente et al., 1996) . In men with a history of a tibial fracture 9 years prior, spinal BMD measured by dual energy x-ray absorptiometry (DXA) was 12.3 or 9.5% lower in men with a history of primary nonunion and union, respectively, compared with age-, weight- and height-matched normal men (Kannus et al., 1994). Therefore, bone loss due to immobilization may be long-lived.
XIII. Osteogenesis Imperfecta
Osteogenesis imperfecta in its milder forms may present as osteoporosis in either men or women (Spotila et al., 1991). It is important to assess men with osteoporosis carefully for signs of osteogenesis imperfecta such as ligamentous laxity, increased elasticity of the skin, and blue sclerae. A family history of multiple and severe fragility fractures may also be present.
Osteoporosis occurs commonly in homocystinuria. By the age of 15 years, spinal osteoporosis is detected in 64% of patients with vitamin B6-unresponsive compared with 36% of vitamin B6-responsive patients (Mudd etal., 1985). Deficient collagen cross-linking has been proposed as a cause of osteoporosis in homocystinuria. In ten patients with homocystinuria, concentrations of amino- and carboxy-terminal propeptides of types III and I collagen, respectively, were similar to those in healthy age-matched controls. However, the urinary excretion of the carboxy-terminal telopeptide cross-link (ICTP) was decreased consistent with normal collagen synthesis and reduced pyridinium cross-link formation in homocystinuria (Lubec et al., 1996).
XV. Neoplastic Disease (Multiple Myeloma, Lymphoma)
Multiple myeloma and lymphoma are associated with bone destruction, low bone mass, and fragility fractures due to the production of cytokines. In patients with multiple myeloma, new bone formation is also inhibited. Possible mediators of this effect include lymphotoxin, interleukin-1(5, parathyroid hormone-related protein (PTHrP), and interleukin-6. They can be produced by the myeloma cells or by marrow stromal cells in response to myeloma cells, and all have been implicated as possible osteoclast-activating factors (OAF) (Roodman, 1997). The cytokines were originally labeled osteoclast- activating factors because of their ability to stimulate osteoclastic bone resorption. Production of these cytokines is inhibited by glucocorticoids, and intravenous bisphosphonate therapy with monthly infusions of pamidronate for nine months reduced the skeletal morbidity associated with multiple myeloma and improved quality of life (Berenson et al., 1996).
XVI. Ankylosing Spondylitis and Rheumatoid Arthritis
Cytokine production, particularly interleukin-6 (Papanicolaou et al., 1998) , as well as immobility and cumulative glucocorticoid therapy dose all play important roles in the etiology of osteoporosis in ankylosing spondylosis and rheumatoid arthritis. In men with ankylosing spondylitis of recent onset, spinal and femoral neck BMDs were lower than in age-matched control men, suggesting early loss of trabecular bone (Will etal., 1989). Later in the disease process, spinal BMD is increased due to syndesmophyte formation. Vertebral compression fractures are present in 10-41% of patients with ankylosing spondylitis (Sivri et al., 1996; Donnelly et al., 1994). Patients with fractures are more likely to be male, to be older, and to have longer disease duration and more advanced spinal limitation with less mobility. There were no consistent deficits in either spinal or femoral neck BMD in fracture patients, and spinal BMD does not reliably predict the risk of vertebral fracture in men with ankylosing spondylitis.
In men with rheumatoid arthritis, osteoporosis is also more evident at the hip and radius than the spine (Dequeker et al., 1995). Low-dose methotrexate therapy may increase the risk of fragility fractures and lower extremity pain syndromes in men and women with rheumatoid arthritis, but further prospective studies are required to prove this hypothesis. Few studies have addressed any problems specific to men with rheumatoid arthritis, but risk factors for osteoporosis in men are likely to be similar to those in women. In particular, consideration should be given to the treatment of underlying hypogonadism.
XVII. Systemic Mastocytosis
Mastocytosis is a spectrum of disorders in which aberrant mast cell proliferation may occur in a number of different organs (Table II). Mutations of the c-kit proto-oncogene, which encodes for a receptor tyrosine kinase and plays an important role in mast cell growth and differentiation during hematopoiesis, have been identified in some cases of human mastocytosis (Pignon, 1997). Mastocytosis may be indolent or aggressive, and in 75% of cases bone disease will be present (Travis et al., 1988). Diffuse osteoporosis is the most common finding (28%) followed by osteosclerosis (19%); mixed sclerosis and demineralization occurs in 10% of the patients studied. The skeletal abnormalities are dependent on mast cell mediator production. Both heparin and prostaglandin D, may cause increased bone resorption, whereas histamine promotes bone fibrosis.
TABLE II Classification of Mastocytosis
Cutaneous disease Ulcer disease Malabsorption
Bone marrow mast cell aggregates
Skeletal disease (osteopenia, osteosclerosis, mixed)
Aggressive Mastocytic leukemia
Lymphadenopathic mastocytosis with eosinophilia
Recently there has been renewed interest in the prevalence of systemic mastocytosis in men with no other obvious cause of osteoporosis. In a large retrospective study by Ritzel et al. (1997), the prevalence of mastocytosis in patients having biopsies for osteoporosis was 1.1%, and men were equally represented with women. Osteoporosis in mastocytosis was associated with an increase in osteoid volume and osteoid surface, and an increase in eroded surface compared with the control group. Trabecular bone volume and trabecular thickness were decreased. Because osteoclast numbers were similar in both groups, these observations suggest an effect of mast cell products on osteoclasts together with an uncoupling of osteoblast and osteoclast function. It is possible the increased osteoid is due in part to calcium and/or vitamin D malabsorption caused by mastocytosis-associated gastrointestinal dysfunction.
In a smaller retrospective study of 136 bone biopsies from men with idiopathic osteoporosis, de Gennes et al. (1992) detected only four cases of skeletal mastocytosis, a prevalence of 3%. In a large teaching hospital in the United Kingdom, a search of all bone biopsies submitted over 5 years revealed evidence of only six cases of skeletal mastocytosis, four of whom were men (Andrew and Freemont, 1993). On the other hand, mastocytosis may present with diffuse osteopenia in up to 50% of cases with or without vertebral compression fractures and in the absence of systemic symptoms or its classical cutaneous manifestation of urticaria pigmentosum (Chines et al., 1993) . In addition, mastocytosis may be more common as a cause of severe generalized osteopenia in younger men without clinical evidence of mast cell mediator disease (Lidor et al., 1990). The treatment of men with mastocytosis has included antihistamines (HI and H2 blockers), ketotifen, disodium chromoglycate, bisphosphonates (clodronate and pamidronate) and, most recently, interferon alpha-2b (Gruchalla, 1995). In one man with mastocytosis (Marshall et al., 1997), a single infusion of 105 mg of pamidronate resulted in a 16% increase in spinal bone density over one year as measured by dual-energy x-ray absorptiometry. In two small series of men treated with either daily or thrice weekly subcutaneous injections of interferon alpha-2b, trabecular BMD increased by 16%, on average, over 8 months (Lehmann et al., 1996; Weide etal., 1996).