Sexual Dimorphism in Skeletal Size, Density, and Strength

Where is the wisdom we have lost in knowledge?Where is the knowledge we have lost in information? From The Rock by T. S. Eliot Techniques are not ends in themselves, they are only as good as the answers they provide. The answers are only as good as the questions, and the questions are only as good as the insights generated from observations made using the technique. Problems arise when the technique takes over, like HAL in Stanley Kubrick’s 2001: A Space Odyssey, driving inquiry because of the accessibility of its “ON” button, but misdirecting inquiry by producing flawed inferences due to the technique’s unrecognized limitations (1, 2). Although the bone densitometer has a central role in the diagnosis, prevention, and treatment of osteoporosis in clinical practice, it has taken over in clinical research (“I’m sorry Dave, … . . I can’t do that… . ”), and its printout, the bone “density” measurement, has become an end in itself, producing conceptual errors because of its technical limitations. The use of densitometry has led to several misconceptions: 1) volumetric bone mineral density (BMD) increases during growth. It does not. Growth builds a bigger, not more dense, skeleton; 2) peak volumetric BMD is higher in men than women. It is not. Bone size is greater; 3) women lose more trabecular bone from the spine than men. No so, there are qualitative, not quantitative, differences; 4) cortical bone loss is less in men than women during aging. This is not so; the amount of bone resorbed from the endocortical (inner) surface of the cortical shell is similar in men and women. The amount of bone formed on the periosteal (outer) surface during aging is greater in men so that “net” cortical bone loss is less in men; and 5) bone mass is reduced in patients with fractures. It certainly is, but the deficit is exaggerated in patients with smaller bones than controls and underestimated in patients with larger bones than controls. These erroneous notions are produced by uncritical use of densitometry. Bone consists of the mineralized matrix fashioned into a cortical shell and a trabecular network of plates within a marrow cavity. Only the mineralized mass of bone is seen by the photons so that the degree of attenuation of the photons during their transmission through the bone is a measure of the bone mineral content (BMC) of the whole bone, not all of which is mineralized tissue. The image produced during scanning is like a silhouette, just a two-dimension image of the three-dimensional structure, an “areal” projection of the bone in the coronal plane (Fig. 1A). The length and width of the scanned bone is known, but not its depth. Because the depth of the bone is not seen, a bone with greater depth will attenuate more photons and will be reported as being more dense. The bone is not more dense, it does not have a greater amount of bone mineral distributed (as thicker trabeculae, more trabeculae, a thicker cortex) within the periosteal envelope, the bone is just larger; there is more mineral, but this mineral is distributed in the larger volume of the whole bone. a, Photons are attenuated during transmission, producing an attenuation wave profile proportional to the mass of mineralized bone in the scanning path. The attenuation produced by each passage is summated to derive the BMC subtended by the projected area of the bone in the coronal plane. [Reproduced with permission from E. Seeman.]. b: A, Age-related bone loss reduces the amount of bone within the periosteal envelope so that the photon attenuation profile is reduced (more are transmitted through the porous bone); B, Age-related periosteal apposition offsets bone loss, the photons are more greatly attenuated even though the same bone loss has occurred inside the bone. [Reproduced with permission from E. Seeman.] The measurement obtained is the BMC per unit projected bone area of the bone in the coronal plane, or an areal BMC (g/cm2). “Areal” is deleted, and“ content” is replaced by density, so BMC per unit projected area is called “bone mineral density” or “BMD.” The problems that follow from this poor choice of language affect the interpretation of every aspect of the pathophysiology of skeletal growth, aging, and drug therapy in which densitometry is used. The areal BMD measurement is a function of the size of the bone and the amount of bone within its periosteal envelope; its volumetric BMD. The problems arise because bone size and the amount of bone within it change independently during growth and aging, and often in opposite directions. For example, during growth, external bone size increases with a proportional increase in the amount of bone within its periosteal envelope so that the volumetric BMD remains constant, but areal BMD increases because the bone is bigger. During aging, the amount of bone within the periosteal envelope decreases as bone resorption results in thinning and loss of trabeculae, thinning, and increased porosity of the cortex (Fig. 1B). However, aging is also associated with simultaneously occurring periosteal apposition, which enlarges the bone and offsets resorptive removal of the mineralized matrix (3–5). Consequently, the net loss of bone (the sum of bone deposit on the outside and removed on the inside) underestimates the absolute bone lost by resorptive removal inside the bone. Patients with spine fractures have smaller bones than controls (6). Because the areal BMD measurement only sees less bone mineral mass without distinguishing whether its due to a smaller bone, less bone in the bone or both, the lower areal BMD in patients with fractures overestimates the deficit in bone mass inside the smaller bone (produced by excessive bone loss or reduced accrual). The densitometric measurement does not see these opposing surface-specific changes but rather integrates their net effects. The purpose of this commentary is to describe the structural and biomechanical basis of: 1) the greater gain bone strength during growth in men than in women; 2) the lesser decline in bone strength in men than in women during aging; and 3) bone fragility in patients with fractures drawing attention to the conceptual errors produced by densitometry in the clinical context of skeletal growth and aging. Bone mass increases during growth because bone length and diameter increase. The mass of bone contained within the periosteal envelope of the bone also increases. The growing bone is fashioned into a structure adapted to carry out specific functions. The vertebral bodies consist of an interconnecting trabecular network of plates and sheets forming a marrow filled spongiosa, a mass of cancellous bone contained within a thin cortical shell. The design secrets of weight-bearing structures were known by the Ancients and are still standing despite thousands of years of gravity. Long bones like the radius and femur grow in length and diameter. The mineralized bone mass within their periosteal envelope is not formed a solid column that will easily tolerate loads but will remain immovable no matter how hungry the owner, or how delicious the meal scurrying by. The mineralized mass of bone within the periosteal envelope of the long bones is modeled into hollow cylinders with splayed ends containing trabecular bone. These structures are architectural masterpieces of minimalism, created using the least amount of material needed for optimum function—lightness for speed, strength for loading—ensuring the organism can find dinner, rather than be dinner. Growth fashions a bigger, not denser, skeleton. BMD increases during the pre- and peripubertal years because long bones increase in size (Fig. 2, top) (7). The densitometer sees a bigger bone and prints out that the bigger bone has a higher areal BMD, a higher density, giving the impression that as children grow their bones have somehow become more dense. Similarly, in late puberty, when boys have long bones of greater diameter than girls, the densitometer sees a bigger bone and prints out a higher areal BMD in boys, giving the impression that somehow boys have grown denser bones than girls. This is not so. Femoral shaft areal BMD increases with age, but volumetric BMD is independent of age in males and females. [Reproduced with permission from P. W. Lu et al.: J Clin Endocrinol Metab 81:1586–1590, 1996 (7 ),© The Endocrine Society.] As the bone grows in length and diameter, the mass of bone inside the periosteum increases in proportion to the enlarging volume of the whole bone including the marrow space. This mass of bone is fashioned into a cortex of a given thickness. Volumetric BMD of long bones, like the radius or femur, is independent of age and no different in boys and girls (Fig. 2, bottom) (7–9). In young adulthood, men and women have the same volumetric BMD, but men have a bigger bone than women. Thus, if the strength of the radius or femur is greater in older than younger children, or greater in males than females, it is due to differences in size, not density, not the amount of bone within the periosteal envelope of that bone. Volumetric BMD also remains independent of age for the vertebral body, at least until puberty (10). In this case, the mass of bone is fashioned as a trabecular network. The length and thickness of the intersecting trabecular sheets increases in proportion to the enlarging medullary space containing them. The number of trabeculae, established at the growth plates, do not increase with age (11). At puberty, trabecular BMD increases and does so to a similar degree in boys and girls (Fig. 3) (10), perhaps because trabeculae continue to thicken while growth in external size of the vertebral body ceases. Growth in external size ceases sooner than the accrual of bone within its periosteal envelope (12–14). However, longitudinal studies documenting the precise temporal of of growth in external size and accrual of mass within the vertebral body are trabecular volumetric BMD is similar in boys and girls and is independent of age until puberty increases by from et Thus, and growth, and in the grown young males and have the same trabecular BMD and thickness of within the vertebral is vertebral body have a and only vertebral Thus, growth does not a denser in males than females, it builds a bigger skeleton. the strength of the vertebral body is greater in young males than females, it is due to differences in size, not density, not the amount of bone within the periosteal envelope of that bone. The of volumetric BMD in is The of volumetric BMD puberty has if volumetric BMD is independent of age, the of an volumetric BMD be volumetric BMD of the spine in the same in the as age Volumetric BMD is a function of the growth in size of the whole bone and the mass within not the absolute growth in size or Thus, there be and of bone size and mineral accrual that that growth of size is with a given proportional growth in mass within Consequently, a volumetric BMD at or through growth the same in the at of age as at of This is in the for size, and for there to be in volumetric BMD, more bone per unit external volume of bone than to be at the less bone per unit external volume of bone to be at the have less cortical or trabeculae per unit external bone volume than at greater for fragility fractures during biomechanical and for the in volumetric the the growth of the in size and the mass within the in shaft cortical trabecular number for a given bone will not until we the and of these structural of areal BMD areal and the or that for of the in each of these structural areal The absolute and of the periosteal and bone size and its The in length is the of a of growth in the During this growth is more in the than spine The growth in boys, and growth is greater and more than in girls. At puberty, growth of the and ceases growth of the spine to that of the and The in at peak women and men is in not not The long bones and are also but the are only in men than women. The absolute and of the periosteal and the diameter of the long bone, the mass of cortical bone, cortical and the this cortical mass is from of the bone. periosteal and endocortical in cortical thickness of the enlarging bone will remain endocortical is excessive to the of periosteal the enlarging bone will have a which is not periosteal with no endocortical or less periosteal with endocortical the same cortical thickness is However, the bone will be and its cortical mass will be the long of the bone so that strength will be apposition increases bone width in boys and girls during growth. is associated with periosteal apposition with less endocortical in boys, in of the bone diameter, cortical and an increase in the medullary diameter. of the increase in cortical thickness is by periosteal apposition, a that is to be by the and In of and puberty, periosteal apposition is due to the of on periosteal bone endocortical bone is cortical thickness and the medullary This and surface-specific in studies of the of the bones (Fig. 4) also be occurring less at weight-bearing as the femur than or radius diameter of the bones does not puberty in boys and girls. During puberty the periosteal diameter in boys and ceases to in girls medullary diameter remains in boys growth but in girls. [Reproduced with permission from S. The in bone size and strength is established in The greater periosteal and endocortical puberty and endocortical during puberty in men the cortical bone mass from the of the long bone in men than women. cortical thickness is similar in males and females, the mass of cortical bone is greater because of the greater of the bigger Bone from the by periosteal apposition more strength in than bone by endocortical apposition the of the long bone In a of the et reported the results of a of the growth in mass and strength of the radius in and males Bone by as BMC of a of cortical bone at the differences at the age of BMC increased in men and women because the bone by periosteal in males and and endocortical in females. higher BMC due to of bone from the of the long bone. Bone strength increased in but the in strength the of the in bone not bone The same volumetric BMD of the radius in an so that the is in this of the of the amount of bone in the growing bone Thus, the absolute and growth of these by the of bone, and in degree at every the length of bone from to and so that these long bones in size, cortical and medullary diameter their whole At each these will be by the bone on the and the of endocortical bone resorption and to For the endocortical to resorption bone for endocortical to bone resorption and are there will be no net of the endocortical The and periosteal and endocortical of long bone are created by the and occurring in that the same of this surface-specific and is in of the studies of skeletal growth (11). growth of the of the is by periosteal apposition and endocortical resorption on the cortex and endocortical apposition and net periosteal resorption on the endocortical resorption of the cortex and periosteal apposition on the cortex when these the same of the of bone during growth. The differences in the of these at puberty that puberty structural boys will lose the of periosteal apposition, a smaller bone with a cortex but medullary diameter (Fig. BMD will be reduced because bone size is and volumetric BMD be if the in size and accrual are will lose the endocortical apposition, a or larger bone apposition in of the of with a cortex and larger medullary diameter (Fig. Thus, areal BMD be or reduced on bone size, but volumetric BMD will be reduced because the or larger bone will have less bone mass accrual within Because of reduced bone size in boys and bone size in girls, the biomechanical of puberty be in boys than girls, even though areal BMD be In boys, puberty periosteal apposition, a smaller bone with a cortex but medullary diameter In girls, puberty in reduced endocortical apposition, a or larger bone periosteal apposition in of the of with a cortex and larger medullary diameter [Reproduced with permission from E. Seeman.] areal BMD is reported in boys with puberty of the studies reported volumetric BMD the deficit in areal BMD only due to smaller bone This and but the is more than bone size for the areal BMD, but the amount of bone in the smaller bone is observations are in areal BMD is reduced because bone size is but volumetric BMD is will when the limitations of densitometry are The three-dimensional be seen using of or long bones (1, 2). During aging, periosteal apposition as it during growth but more Bone at or on the and of the (inner) surface of the skeleton. However, the within each is because the volume of bone removed by bone resorption is not by the same volume of bone formed within the The bone within each is the structural basis of bone The loss of bone within each cortical thinning and the bone trabecular thinning, loss of trabecular plates, and loss of The using bone densitometry is that men lose less bone than women at the spine (Fig. top) This is produced by the that densitometry the bone on the periosteal surface the loss occurring inside bone. of the changes that within the trabecular inside bone and on its periosteal surface that the amount of trabecular bone lost during aging in women and men is or only less in men than women (Fig. This has using at the at the and of The in areal BMD body using photon [Reproduced with permission from et al.: J Clin and The in vertebral body trabecular and cortical BMD of the vertebral body using from et The of trabecular bone loss by occurring by thinning in men and by loss of in women bone loss in in women due to the increased surface that at The of trabecular bone loss to bone loss decreases as trabecular plates and because there is less trabecular surface for In there is no of bone because men do not through but bone loss due to reduced bone and thinning of greater of results in of the trabecular for so that trabecular bone loss in men than in women. is trabecular bone volume is similar in men and women. The amount of trabecular surface for bone in age to be greater in men than women The periosteal bone occurring during aging offsets bone loss from the in men and but for that are not periosteal apposition is greater in men than women so that the loss of bone inside bone is more greatly in men than in women For example, and greater differences in periosteal bone than bone resorption at area absolute periosteal bone increased by in men and in women. area absolute bone increased in men and women Thus, the lesser in areal BMD at the spine in men than women is due to the lesser in cortical bone in men than in women (Fig. bottom) However, the lesser in cortical mass is the of greater periosteal bone in not less resorptive removal of bone from the The densitometer is to all this surface-specific bone and The lesser in areal BMD in men than women in studies is for the that men lose less bone than women. The is bone loss is less in men than women because men more bone during aging on the outside surface of the bone, not because the lose less bone on the inside of the bone. The in peak bone size is not men than women have fragility fractures. The larger during growth in men a bone than in women a bone that larger absolute this is not the for the lower of fragility fractures in men than in women. has a bigger and bone than a but has to with the generated by an bigger A on the loads the radius of a with the generated by the and not a and The absolute on the vertebral body is greater in young men than women because men are and However, the per unit area on the vertebral no different in young men and because the larger bone is to larger loads The of periosteal of the radius to area is the same in boys and girls is in bigger bone has a bigger that is to be by size, not the amount of fractures are in young men and women because loads by and are the of the bone to by area and volumetric during aging in men and women because of the the and the to tolerate that apposition has it increases the area of the bone more in so that per unit area decreases more than in women. greater periosteal apposition more bone to the of the bone in men than women the bone loss more in men than women so that the in volumetric density of the whole bone is less in men than in women. Thus, men and women to age, the per unit area of bone decreases and strength of the bone decreases in men than in women so that a lower proportion of men than women have bone size and volumetric BMD reduced a the loads on the bone are greater than the bones structural to tolerate less in men than and women because the and bone strength is in men than in women Patients with fragility fractures have reduced volumetric BMD, but they also have reduced bone size to controls As the bone size is not taken into in the BMC measurement, and only taken into in the areal BMD measurement, the deficit in bone mass is exaggerated if the bone size is reduced to controls. This deficit will be to reduced or (as the to excessive bone Thus, the is by the areal BMD deficit if the of the smaller bone size is not bone size in is increased to as reported in women with fractures and E. for the deficit in BMC or areal BMD will be underestimated and deficit in volumetric BMD due to reduced accrual or excessive bone to for differences in bone size in and controls will the to that there is no deficit in volumetric BMD no producing a deficit in volumetric BMD. A is that in bone size and volumetric BMD in women and men with spine fractures are the of reduced periosteal apposition during aging periosteal apposition during aging will 1) in a smaller bone that loads less and will 2) to bone loss so that volumetric BMD the whole will be lower than it be periosteal apposition Growth a bigger not denser in males and (Fig. periosteal growth during puberty a that is just the size to the larger of the The smaller of the is just the size to the smaller skeletal mass to its apposition of the cortical mass in apposition at puberty of cortical thickness at bones in the it less biomechanical during than a amount produced by periosteal strength is needed at and it that endocortical apposition less cortex at weight-bearing this endocortical accrual is the for of a without the bone strength is a in of During growth, areal BMD increases because bone size increases. BMD is higher in men because bone size is greater than in women; there is no in peak volumetric BMD. In long bones, volumetric BMD is independent of At the volumetric BMD increases at During aging, men and women lose similar of bone by have greater periosteal apposition than so that net bone loss is less in men than women. bone loss by loss of trabecular in in women and by trabecular thinning in men. [Reproduced with permission from E. Seeman.] have a that to aging by greater periosteal apposition that increases bone size and offsets bone loss more than in women so males than are at for in Patients with fractures have smaller and less bone in the smaller bone. The of each of these the of the of the during growth and aging. Bone densitometry is an for the clinical of It has made in of the and of bone a of and so the size of the of BMD is a of and so is in at of be for These are and clinical of the For clinical the excessive on bone densitometer problems in the structural basis of the of bone strength during growth, bone fragility during aging, as as the structural basis of the of bone fragility associated with drug therapy because densitometry is to bone size and of bone by the and it in the of a and the of adulthood, and to the of age, its like and fractures. The for and by the to the pathophysiology of bone the with “bone the of … bone the of … bone the and treatment of … bone this the and of the pathophysiology of growth and aging of the skeleton. the of the structural basis of of bone strength during growth, and its during aging, is and young the will for a of a there is no of The absolute and of the periosteal and produced by bone and bone resorption during growth and aging the size, and of the so, its The of the will remain while the of bone fragility remains in the of and structural areal these are the of bone fragility will in its Bone mineral bone mineral