In women osteoporotic fractures are a major health problem. The causation of fractures is complex, but bone fragility is a primary contributor to fracture risk. Bone mineral density (BMD) is a good surrogate measure of bone strength, and strong relationships between several measures of BMD and the probability of fracture have been documented in large prospective studies. As a result, a general recognition of the clinical usefulness of BMD measures has emerged. The most appropriate use of the available techniques remains controversial, but BMD measures are applied widely in postmenopausal women. Recent National Osteoporosis Foundation recommendations and expanded insurance coverage are extending their use further.

In men the situation is similar but at the same time quite different. Although not as common as in women, fractures in men result in tremendous morbidity and cost,(1) and an emerging awareness of the problem has appropriately led to an increasing demand for BMD measures. Unfortunately, the data that underlie the application of BMD measurements in women are much less mature in men. Most importantly, in men there are few prospective studies of the relationship between BMD and fracture risk. Although this research is underway, it is difficult to assign clinical or economic value to densitometry measures. Nevertheless, there is a present need to devise interim solutions to the problem of fracture risk assessment, especially because successful osteoporosis preventive and treatment strategies have now been identified for men.(2–5) Two salient questions exist. Which men should have BMD measures, and how should results be interpreted?

There are some populations of men for whom BMD measures are probably reasonable. Those include men who have suffered low trauma fractures, are discovered to have prevalent vertebral deformity, have radiographic osteopenia, or have conditions recognized to impart risk for bone loss and fractures (e.g., hyperparathyroidism, overt hypogonadism, or glucocorticoid therapy). In these men, preventive and therapeutic recommendations can be made more confidently when bone density is known. At present there seems to be little reason to consider these indications to be different in men and women.

The question of more broad-based screening for low BMD is more difficult. In part because the risk of fracture increases quickly after the age of 65 years in women, some recent recommendations include screening BMD measures for all women after that time. Fracture risk also increases dramatically in men but at a somewhat later age than in women, and if a goal is to intercept and prevent that trend perhaps BMD screening in men ≥70 years or 75 years is appropriate. On the other hand, case finding with a combination of clinical characteristics and BMD testing may be more attractive. For example, a positive family history, smoking, and low weight/weight loss, (among others) have been linked to low bone mass and fractures and could represent effective prompts for BMD measures. Modeling analyses based on the prevalence of low bone mass in the general population, expense, and expected benefits are underway to explore the cost-benefit implications of various detection strategies. Obviously, the cost and availability of BMD measures and the effectiveness and cost of intervention strategies will be critically important. Much of this information is not yet available from prospective studies, and until it is, widespread BMD screening would seem to be premature in men.

The BMD criteria (thresholds) that should be used to identify men in need of diagnostic and therapeutic intervention are of considerable importance and debate. Population-based male reference ranges are available for many BMD measurement techniques, and clinicians frequently use risk assessment paradigms that are parallel to those used in women (i.e., classifications based on standard deviation scores, with young male reference means as the denominator). Recently, some have suggested that for diagnostic purposes it is more appropriate to use the same absolute level of BMD in both genders. Unfortunately, there is little information from prospective trials in men to validate either of these approaches. As the use of BMD assessments in men increases, along with the prescription of preventive and treatment regimens based on BMD results, it is more and more important that some consensus emerges to guide their interpretation.

A strong relationship between BMD and fracture risk, a foundation for the use of BMD measures in clinical situations, has been documented in prospective studies in both men and women.(6–8) Thus, the basic hypothesis that BMD is a determinant of fracture risk in men is well founded. At issue is whether there are gender differences in the BMD-fracture risk relationship that should render the clinical application of BMD measures in men different than that in women. Of special importance is the choice of BMD levels that might prompt clinical intervention (diagnostic or preventive/therapeutic). Should BMD measures in men be interpreted differently than in women?

It is attractive to postulate that BMD should have a constant relationship with biomechanical strength and hence fracture likelihood regardless of gender. In fact, the relationship between volumetric BMD and fracture resistance in cancellous bone is not influenced by gender. However, volumetric BMD is but one of the factors that influence fracture risk and several other variables that are influenced by gender would be predicted to have an effect. For example, men and women obviously differ in body size and composition, in turn influencing the biomechanics of falls. Men also fall differently than women,(9) and in general sustain a different constellation of fractures than do women. Moreover, bone size and geometry strongly contribute to biomechanical strength independent of bone density and on average are quite different in men and women. Even after adjustment for height and weight, bone size tends to be greater in men.(10) On another level, age-related bone loss has a different pattern in men and women, with more trabecular dropout and endocortical wasting in women.(11) These differences may not be reflected in measures of BMD but may have important effects on fracture likelihood and in turn on the application of BMD in the prediction of fracture. In light of these gender-dependent variables that should affect the BMD-fracture risk relationship it would be surprising if the same algorithms could be used to relate absolute BMD to fracture risk in male and female patients.

Another complexity may affect the application of BMD measures in men. Gender differences in bone size should fundamentally alter the accuracy with which areal measures of BMD (e.g., dual-energy X-ray absorptiometry [DXA]) reflect true volumetric BMD. A larger (i.e., male) bone may have the same volumetric BMD as a smaller bone but a higher apparent BMD when assessed with an areal measurement technique.(12) This size artifact might be expected to alter the relationship between areal BMD and fracture risk and to be the source of gender differences in osteoporosis diagnostic thresholds. On the other hand, the larger bone also should have a greater resistance to fracture (by dint of its biomechanical advantage) and perhaps the BMD-fracture risk relationship described in smaller (i.e., female) bones is thus maintained fortuitously. In reality, how apparent BMD and biomechanical fracture resistance relate to the clinical prediction of fractures from BMD measures in men is unknown. Therefore, the most commonly used BMD measurement technique, DXA, is plagued by a size-related artifact that has uncertain effects on the criteria previously developed in women. In view of the uncertain impact of bone size on the BMD-fracture risk relationship, it is imprudent to assume that BMD levels associated with a given fracture risk in women should have the same usefulness in men. Other techniques (i.e., volumetric measures) may be less affected by this problem, but even when using volumetric methods the fact that men have larger bones still may alter the BMD-fracture risk relationship.

Virtually all trials, both cross-sectional and longitudinal, document that the average areal hip and spine BMD in men who suffer fractures is higher than women who experience similar fractures.(13) For instance, the Dubbo Osteoporosis Epidemiology Study reported that men who experienced low traumatic fractures had higher spine and femur DXA BMD (approximately 20% higher at each site) than women who experienced fractures.(7) Similarly, the European Quantitation of Osteoporosis Study found that proximal femoral BMD was considerably higher in men with hip fracture than in women with fracture.(14) These relationships probably can be summarized in Fig. 1, in which the number of fractures is shown as a function of BMD. More fractures occur in women, and in fracture patients the average BMD in women is lower than in men.

thumbnail image

Figure Figure 1. A model of the distribution of BMD in older patients with fracture. More fractures occur in women than in men, and the average BMD is lower in women.

Download figure to PowerPoint

Unfortunately, the essential question of how BMD is related to fracture risk is not illuminated by these data. Three ways to interpret the gender difference in average BMD in fractured patients (Fig. 1) are illustrated in Fig. 2 where the relationship between BMD and fracture risk is considered. First, the data in Fig. 1 are compatible with Fig. 2 slope a, in which the relationship between fracture risk and BMD is the same in men and women. In this scenario, the mean BMD in women with fractures is lower than in men primarily because more women have BMD levels in the extremely low range where fracture risk is highest. However, the fact that the average BMD in male fracture patients is higher than in similarly fractured women also is compatible with Fig. 2 slope b, in which the increase in fracture risk with lower BMD levels in men is similar to that in women but is shifted to the right. In other words, at a similar absolute fracture risk, men have a higher BMD. A final possibility is that the curve in men may be shifted to the right by virtue of a different slope (Fig. 2, slope c). If either Fig. 2 slope b or c is true, there is a need to use gender-specific criteria to determine which patients are at sufficient risk to be offered increased levels of intervention.

thumbnail image

Figure Figure 2. A model of the relationships between BMD and fracture risk in men and women. The nature of the BMD-fracture risk relationship is well established in women, and the figure shows how it could compare to possible relationships in men.

Download figure to PowerPoint

Some recently published studies suggest that fracture risk in men and women is the same at any absolute level of BMD (Fig. 2, slope a). Two prospective studies of the usefulness of calcaneal BMD measures in the prediction of fracture, either spinal fracture(15) or all fragility fractures,(16) support that contention. It is important that these two trials used calcaneal BMD measures, a site that probably is less subject to the effects of bone size than are BMD measures of the spine or hip. In fact, Cheng et al.(16) expressed their BMD results as a volumetric measure. In analyses from the Rotterdam study, in which the assumption was made that the relationship between relative reduction in BMD and fracture risk was the same in men and women, De Laet et al. projected that there was no gender effect on the risk of hip fracture at any level of BMD.(17) Subsequently, the model (which includes age, gender, and BMD) appeared to be useful in estimating fracture risk when compared with prospectively obtained data.(8) However, these studies do not directly examine the hypothesis that the BMD-fracture risk relationship is uninfluenced by gender.

Other data are more compatible with Fig. 2 slope b or c. Analyses from Rochester suggest that men have a different BMD-fracture risk relationship than women.(18) Also, an examination of results from recent treatment trials supports gender differences in the BMD-fracture risk relationship. Shown in Fig. 3 are average spine and femoral neck BMD levels at baseline in the placebo-treated subjects in several prospective trials.(19–22) Some subjects (FIT 1, MORE 1, and MEN 1) had prevalent vertebral fractures, while others did not (FIT 2, MORE 2, and MEN 2). All trials used DXA machines of a single manufacturer. Although the groups of men had higher spine and hip BMD levels, their vertebral fracture rates were quite similar to those in the groups of women. This experience would suggest that the relationship between hip and spine BMD measures and subsequent fracture risk is different in men and women. At this point, the available information is obviously contradictory and probably insufficient to definitively settle the issue.

thumbnail image

Figure Figure 3. The relationship between baseline BMD (spine and femoral neck) and vertebral fracture incidence in three longitudinal trials (FIT, osteoporotic women; MORE, osteoporotic women; and MEN, osteoporotic men).(19–22) Subjects in FIT 1, MORE 1, and MEN1 had prevalent vertebral fractures at baseline, whereas subjects in FIT 2, MORE 2, and MEN 2 had no prevalent vertebral deformities. Only placebo-treated patients from each trial are included, and in all trials BMD was assessed using densitometers of the same manufacturer.

Download figure to PowerPoint

A real problem with assuming a single, gender-neutral diagnostic criterion based on absolute levels of BMD derived from studies in women is that most men may not be recognized to be at risk, a situation clearly at odds with the commonality of osteoporosis in men. Because men experience fractures at a relatively higher BMD, many if not most of those that experience hip fracture would not be considered at risk using conventional standards in women (≤2 or 2.5 SD below the young female mean). Melton et al. recently reported that 19% of Rochester, MN men over the age of 50 years were found to have osteoporosis at one of three measurement sites (hip, spine, or radius using a criteria of ≤2.5 SD) when young male reference range was used, but merely 3% were so classified if young female reference levels were applied.(18) In the National Health and Nutrition Examination Survey (NHANES) experience 6% of men were classified as having osteoporosis (≤2.5 SD below the young normal male mean at the femoral neck), a figure that was reduced by ⅓ when the female criteria were applied.(23) Obviously, the public health implications of the choice of diagnostic criteria are large. Given the fact that up to 25% of white men over the age of 60 years may experience low trauma fractures in their lifetime,(24,25) the number of men identified as at risk using the female criteria seems inappropriately low.

One solution recently proposed would assign clinical status (risk of future fracture) based on a method that links absolute levels of BMD (perhaps gender specific) to future rates of fracture. With that information, the level of future fracture risk, regardless of gender, would determine the level of intervention warranted. For instance, the BMD at which the average 5-year fracture risk was 5% might be used to assign a certain level of therapeutic intervention, whereas the finding of a BMD associated with a 20% 5-year fracture risk might invoke a more aggressive treatment regimen. Prospective data obviously are needed to determine the BMD-fracture risk relationships that underlie this approach. Although they probably are available in women, the necessary large-scale prospective information is just being accumulated in men.

In summary, osteoporotic fractures in men are common and of enormous public health importance. Clearly, paradigms are needed to guide the application of risk assessment tools—including measures of BMD. Consensus concerning which men would benefit from BMD measures should be developed. Specifically, high-risk groups that deserve BMD testing should be identified, and the wisdom of more widespread screening should be reconsidered when sufficient data are available to make a reasoned judgement about its merit. More data also are clearly needed to guide the interpretation of BMD measures in men. Unfortunately, the current status of using BMD measures to assign levels of clinical concern is at best unsettled, regardless of gender. Even so, BMD measures are powerful tools for clinicians faced with the need to identify men at risk of fracture, and choices must be made in the context of insufficient data. Uncertainties remain about the effect of gender on the BMD-fracture risk relationship and current thresholds based on absolute levels of BMD used for women may unreasonably minimize the number of men who would be identified as at risk for fracture. Thus, until additional data are available, it is sensible to use gender-specific criteria to select men for preventive and therapeutic intervention.


  1. Top of page
  • 1
    Ray NF, Chan JK, Thamer M, Melton LJ 1997 Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: Report from the National Osteoporosis Foundation. J Bone Miner Res 12:2435.
  • 2
    Dawson-Hughes B, Dallai GE, Krall EA, Sadowski L, Sahyoun N, Tannenbaum S 1990 A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N Engl J Med 323:878883.
  • 3
    Adachi JD, Bensen WG, Brown J, Hanley D, Hodsman A, Josse R, Kendler DL, Lentle B, Olszynski W, Ste.-Marie L-G, Tenenhouse A, Chines AA 1997 Intermittent etidronate therapy to prevent corticosteroid-induced osteoporosis. N Engl J Med 337:382387.
  • 4
    Saag KG, Emkey R, Schnitzer TJ, Brown JP, Hawkins F, Goemaere S, Thamsborg G, Liberman UA, Delmas PD, Malice M-P, Czachur M, Diafotis AG 1998 Alendronate for the prevention and treatment of glucocorticoid-induced osteoporosis. N Engl J Med 339:292299.
  • 5
    Cohen S, Levy RM, Keller M, Boling E, Emkey RD, Greenwald M, Zizic TM, Wallach S, Sewell KL, Lukert BP, Axelrod DW, Chines AA 1999 Risedronate therapy prevents corticosteroid-induced bone loss. Arthritis Rheum 42:23092318.
  • 6
    Mussolino ME, Looker AC, Madans JH, Langlois JA, Orwoll ES 1998 Risk factors for hip fracture in white men: The NHANES I epidemiologic follow-up study. J Bone Miner Res 13:918924.
  • 7
    Nguyen T, Sambrook P, Kelly P, Jones G, Lord S, Freund J 1993 Prediction of osteoporotic fractures by postural instability and bone density. BMJ 307:11111115.
  • 8
    De Laet CEDH, Van Hout BA, Burger H, Weel AEAM, Hofman A, Pols HAP 1998 Hip fracture prediction in elderly men and women: Validation of the Rotterdam Study. J Bone Miner Res 13:15871593.
  • 9
    O'Neill TW, Varlow J, Silman AJ, Reeve J, Reid DM, Todd C, Woolf AD 1994 Age and sex influences on fall characteristics. Ann Rheum Dis 53:773775.
  • 10
    Gilsanz V, Boechat MI, Gilsanz R, Loro ML, Roe TF, Goodman WG 1994 Gender differences in vertebral sizes in adults: Biomechanical implications. Radiology 190:678682.
  • 11
    Seeman E 1997 From density to structure: Growing up and growing old on the surfaces of bone. J Bone Miner Res 12:509521.
  • 12
    Seeman E 1999 Osteoporosis in men. Osteoporos Int 9 (Suppl 2):S97S110.
  • 13
    Greenspan SL, Myers ER, Maitland LA, Kido TH, Krasnow MB, Hayes WC 1994 Trochanteric bone mineral density is associated with type of hip fracture in the elderly. J Bone Miner Res 9:18891894.
  • 14
    Kroger H, Lunt M, Reeve J, Dequeker J, Adams JE, Birkenhager JC, Diaz Curiel M, Felsenberg D, Hyldstrup L, Kotzki P, Laval-Jeantet A-M, Lips P, Louis O, Perez Cano R, Reiners C, Ribot C, Ruegsegger P, Schneider P, Braillon P, Pearson J 1999 Bone density reduction in various measurement sites in men and women with osteoporotic fractures of spine and hip: The European quantitation of osteoporosis study. Calcif Tissue Int 64:191199.
  • 15
    Ross PD, Lombardi A, Freedholm D, 1999 The assessment of bone mass in men. In: OrwollES (ed.) Osteoporosis in Men: The Effects of Gender on Skeletal Health. Academic Press, San Diego, CA, U.S.A., pp. 505525.
  • 16
    Cheng S, Suominen H, Sakari-Rantala R, Laukkanen P, Avikainen V, Heikkinen E 1997 Calcaneal bone mineral density predicts fracture occurrence: A five-year follow-up study in elderly people. J Bone Miner Res 12:10751082.
  • 17
    De Lait CEDH, van Hout BA, Burger H, Hofman A, Pols HAP 1997 Bone density and risk of hip fracture in men and women: Cross sectional analysis. BMJ 315:221225.
  • 18
    Melton LJ, Atkinson EJ, O'Connor MK, O'Fallon WM, Riggs BL 1998 Bone density and fracture risk in men. J Bone Miner Res 13:19151923.
  • 19
    Orwoll E, Ettinger M, Weiss S, Miller P, Kendler D, Graham J, Adami S, Weber K, Lorenc R, Pietschmann P 1999 Alendronate treatment of osteoporosis in men. J Bone Miner Res 14:S184.
  • 20
    Ettinger B, Black DM, Mitlak BH, Knickerbocker RK, Nickelsen T, Genant HK, Christiansen C, Delmas PD, Zanchetta JR, Stakkestad J, Gluer CC, Krueger K, Cohen FJ, Eckert S, Ensrud KE, Avioli LV, Lips P, Cummings SR 1999 Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: Results from a 3-year randomized clinical trial. JAMA 282:637645.
  • 21
    Cummings SR, Black DM, Thompson DE, Applegate WB, Barrett-Connor E, Musliner TA, Palermo L, Prineas R, Rubin SM, Scott JC, Vogt T, Wallace R, Yates AJ, LaCroix AZ 1998 Effect of alendronate on risk of fracture in women with low bone density but without vertebral fractures. JAMA 280:20772082.
  • 22
    Black DM, Cummings SR, Karpf DB, Cauley JA, Thompson DE, Nevitt MC, Bauer DC, Genant HK, Haskell WL, Marcus R, Ott SM, Torner JC, Quandt SA, Reiss TF, Ensrud KE 1996 Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet 348:15351541.
  • 23
    Looker AC, Orwoll ES, Johnston CC, Lindsay RL, Heinz WW, Dunn WL, Calvo MS, Harris TB, Heyse SP 1997 Prevalence of low femoral bone density in older U.S. adults from NHANES III. J Bone Miner Res 12:17611768.
  • 24
    Melton LJ 1997 Editorial: The prevalence of osteoporosis. J Bone Miner Res 12:17691771.
  • 25
    Nguyen TV, Eisman JA, Kelly PJ, Sambrook PN 1996 Risk factors for osteoporotic fractures in elderly men. Am J Epidemiol 144:258261.