Heterogeneity in the Growth of the Axial and Appendicular Skeleton in Boys: Implications for the Pathogenesis of Bone Fragility in Men

Authors

  • Michelle Bradney,

    1. Endocrine Unit and Department of Medicine, Austin and Repatriation Medical Center, University of Melbourne, Melbourne, Australia
    Search for more papers by this author
  • Magnus K. Karlsson,

    1. Endocrine Unit and Department of Medicine, Austin and Repatriation Medical Center, University of Melbourne, Melbourne, Australia
    Search for more papers by this author
  • Yunbo Duan,

    1. Endocrine Unit and Department of Medicine, Austin and Repatriation Medical Center, University of Melbourne, Melbourne, Australia
    Search for more papers by this author
  • Stephen Stuckey,

    1. Department of Radiology, Alfred Hospital, Melbourne, Australia
    Search for more papers by this author
  • Shona Bass,

    1. Endocrine Unit and Department of Medicine, Austin and Repatriation Medical Center, University of Melbourne, Melbourne, Australia
    Search for more papers by this author
  • Ego Seeman M.D.

    Corresponding author
    1. Endocrine Unit and Department of Medicine, Austin and Repatriation Medical Center, University of Melbourne, Melbourne, Australia
    • Endocrine Department, Austin and Repatriation Medical Center, Heidelberg, Melbourne, 3084, Australia
    Search for more papers by this author

Abstract

Men with spine fractures have reduced vertebral body (VB) volume and volumetric bone mineral density (vBMD). Men with hip fractures have reduced femoral neck (FN) volume and vBMD, site-specific deficits that may have their origins in growth. To describe the tempo of growth in regional bone size, bone mineral content (BMC), and vBMD, we measured bone length, periosteal and endocortical diameters, BMC, and vBMD using dual-energy X-ray absorptiometry in 184 boys aged between 7 and 17 years. Before puberty, growth was more rapid in the legs than in the trunk. During puberty, leg growth slowed while trunk length accelerated. Bone size was more advanced than BMC in all regions, being ∼70% and ∼35% of their predicted peaks at 7 years of age, respectively. At 16 years of age, bone size had reached its adult peak while BMC was still 10% below its predicted peak. The legs accounted for 48%, whereas the spine accounted for 10%, of the 1878 g BMC accrued between 7 and 17 years. Peripubertal growth contributed (i) 55% of the increase in leg length but 78% of the mineral accrued and (ii) 69% of the increase in spine length but 87% of the mineral accrued. Increased metacarpal and midfemoral cortical thickness was caused by respective periosteal expansion with minimal change in the endocortical diameter. Total femur and VB vBMD increased by 30–40% while size and BMC increased by 200–300%. Thus, growth builds a bigger but only slightly denser skeleton. We speculate that effect of disease or a risk factor during growth depends on the regions maturational stage at the time of exposure. The earlier growth of a regions size than mass, and the differing growth patterns from region to region, predispose to site-specific deficits in bone size, vBMD, or both. Regions further from their peak may be more severely affected by illness than those nearer completion of growth. Bone fragility in old age is likely to have its foundations partly established during growth.

INTRODUCTION

FRAGILITY FRACTURES are a public health problem in men. For example, the prevalence of vertebral body (VB) fractures in men is similar to or only slightly less than that observed in women, whereas one-third of all hip fractures occur in men.(1–6) Moreover, the mortality from fractures is higher in men than in women.(7)

Men with VB fractures have reduced VB volume and reduced VB volumetric bone mineral density (vBMD) with more modest deficits at the femoral neck (FN), whereas men with FN fractures have reduced FN volume and reduced FN vBMD with modest trait deficits at the VB.(8) Deficits in bone size may be caused by reduced bone growth, failed periosteal expansion during aging, or both; deficits in vBMD may be caused by reduced accrual, excessive bone loss, or both. The contribution of growth-related factors is suggested by the observation that the offspring of men with fractures have reduced areal bone mineral density (aBMD) at the site of fracture in their fathers.(9)

Although data are lacking in males, in females, growth of the appendicular and axial skeleton have a different temporal sequence.(10–12) For example, appendicular growth is more rapid than axial growth before puberty. During puberty, appendicular growth decelerates whereas axial growth accelerates.(11,12) We have proposed that this normal heterogeneity in the temporal pattern of growth in bone size relative to its mineral accrual and the differing pattern of growth of one region relative to another may facilitate the occurrence of region-specific deficits in bone size and mass in adulthood should illness intervene.(12,13) Thus, (i) exposure to diseases or risk factors in the early prepubertal years may affect both the axial and the appendicular skeleton, (ii) exposure shortly before puberty may selectively affect the appendicular skeleton, and (iii) exposure late in puberty may selectively affect the axial skeleton.(14) The purpose of this study was to describe the age- and pubertal stage-specific tempo of growth in bone size and bone mineral content (BMC) of the axial and appendicular skeleton in males.

MATERIALS AND METHODS

Subjects

One hundred and eighty-four volunteers from Ivanhoe Boys Grammar School were recruited after written permission was obtained from the boys, their parents, and the school principal. This is a private school situated in a suburb in Melbourne, Australia. All but eight boys were white Australians. Six boys were Asian, one was South African, and one was Sri Lankian. Data were collected annually for 2 years. Fourteen adult white Australian males aged from 18 to 48 years were recruited to calculate peak adult values. Three boys were omitted from the study because of illnesses noted during recruitment (intellectual disability and cardiac abnormality). All the subjects were healthy with no diseases or exposure to drugs known to affect the skeleton. The study was approved by the Austin and Repatriation Medical Center Ethics Committee.

Measurements

Measurements were made at baseline, year 1 and year 2. One hundred and eighty-four boys had a single (baseline) measurement, 107 had a second measurement, and 70 had a third measurement. Standing and trunk length were measured using a Holtain stadiometer. A Harpenden anthropometer was used to measure the length of the tibia, femur, humerus, and radius, and femoral intercondylar, biacromial, bi-iliac, and bitrochanteric widths. Femur length was the distance from the inferior border of the lateral epicondyle to the superior border of the greater trochanter. The CV ranged from 0.4% to 1.5%. Pubertal status was assessed using Tanner staging (1, prepubertal; 2–4, peripubertal; 5, postpubertal) and testicular size was measured using a Prader orchidometer. Skeletal age was determined from radiographs of the hand using the Greulich and Pyle method.

BMC was measured using dual-energy X-ray absorptiometry (DPX-L; Lunar Corp., Madison WI, U.S.A.; version 1.3z).(15) Results were expressed as BMC (g) and as a vBMD (g/cm3). Both pediatric and adult software versions were used. Any child under 30 kg was scanned in the pediatric software. All subjects remained in the same software for the duration of the study. Regional BMC was determined by using the “region of interest” option from the total body scan. Spine BMC refers to the spine including third cervical to fifth lumbar vertebrae. The CV was 2–4%. VB estimated vBMD (BMC/volume) of the third lumbar vertebra was estimated using the method of Carter et al.(16) The CV was 0.9-2.9%. The vBMD calculation is likely to be an overestimate because rectilinear posteroanterior (PA) scanning includes the mass of the posterior processes; the growth in size of these structures is not taken into account by the Carter method.(17) Estimated total femur vBMD was derived as BMC total femur/estimated femur volume (π × r2 × femur length), where r = midshaft diameter/2, assuming the femur to be cylindrical. The pediatric PA spine program and the ruler function were used to measure periosteal and endocortical widths and cortical thickness of the midshaft of the femur.(12) The CV was 1.5%. Radiographs of the hand were used to determine morphometry at the third metacarpal midshaft using a Vernier caliper. The CV was 0.6-2% for periosteal and endocortical diameters. The CV (interobserver) for reanalysis of the midfemoral shaft scan was 1.5%. The CVs were derived using three measurements in five individuals measured at different times. The machine is calibrated daily using spine phantoms and a calibration block. Precision for metacarpal morphometry was based on three repeat measurements of five individual radiographs at different times.

Statistical analyses

The data are presented in absolute terms, percentage terms, and as a rate of change (mm/year, cm/year, and g/year). To define the relative patterns of growth in bone size and BMC, each trait was expressed as a percentage of the predicted adult peak, derived from data in 14 men, aged 31.0 ± 2.8 years, height, 175.6 ± 2.3 cm; trunk length, 93.2 ± 0.8 cm; leg length, 82.3 ± 2.0 cm; total body BMC, 3160 ± 103 g; spine BMC, 254 ± 12 g; and leg BMC, 1283 ± 46 g. Third lumbar VB values were BMC, 20.2 ± 0.9 g; height, 3.7 ± 0.05 cm; width, 4.5 ± 0.1 cm; and vBMD, 0.30 ± 0.01 g/cm3. No significant differences were found in height or leg BMC comparing these controls with a larger group of 56 men aged 20–40 years (height, 178 ± 2 cm; leg BMC, 1280 ± 35 g). Small differences in the spine BMC (291 ± 10 g; p = 0.06) suggesting some spine accrual may still occur. We have not included the 56 control subjects because segment length data were not available.

A mixed cross-sectional longitudinal design was used because this gives an accurate age-specific mean for the trait being studied.(18) There was continuity in the cross-sectional and longitudinal data; the absolute values for anthropometric and BMC measurements did not differ in the subjects tested for the first time at a given age and subjects at the same age having their first measurement 2 years earlier. For the mixed cross-sectional and longitudinal analyses, analysis of variance (ANOVA) followed by the Fisher method of multiple comparisons was used to compare mean values at each age interval. A significance level of p < 0.05 was used for the ANOVA and 95% confidence level was used for the Fisher comparisons. A polynomial regression was used to plot femur and third lumbar vertebral traits versus age.

RESULTS

At the onset of the study, the boys had a mean chronological age and bone age of 11 years (chronological age range, 7–17 years; bone age range, 6–19 years). Chronological age and bone age did not differ (mean difference = 0.12 ± 0.08 years). Of the 184 boys, 104 were prepubertal with a chronological age of 9.9 years (range, 7–13 years) and a bone age of 9.5 years (range, 6–13 years); 67 were peripubertal with a chronological age of 13.6 years (range, 10–16 years) and a bone age of 13.7 years (range, 10–19 years); 3 were postpubertal with a chronological age of 15.9 years (range, 15–16 years) and bone age 16.5 years (range, 15–19 years); and 10 refused pubertal assessment.

Growth of appendicular and axial skeletal size and mass

As shown in Table 1, across 7–17 years of age, of the 55 ± 2 cm gain in height, 39% (22 ± 2 cm) was gained before puberty (7-11 years) and 61% (34 ± 1 cm) was gained peripubertally (11-17 years; p < 0. 001). Of the 22 ± 2 cm prepubertal gain in height, 59% (13 ± 1 cm) occurred in the legs and 41% (9 ± 1 cm) was truncal (p < 0.01). Of the 34 ± 1 cm peripubertal gain in height, the legs contributed 16 ± 1 cm (47%) and the trunk contributed 18 ± 1 cm (53%; p < 0.01).

Table Table 1.. Cross-Sectional Data: Anthropometry (Mean ± SEM)
original image

As shown in Table 2, of the 1878 ± 109 g mineral accrued between 7 and 17 years, 21% (392 ± 109 g) was accrued before puberty and 79% (1486 ± 46 g) was accrued peripubertally (p < 0.001). Of the 392 ± 109 g mineral accrued before puberty, 51% (198 ± 55 g) occurred in the legs and 6% (25 ± 10 g) occurred in the spine (p < 0.01; the remaining 163 g was accrued at other sites). Of the 1486 ± 46 g accrued peripubertally, 47% (704 ± 21 g) occurred in the legs and 11% (166 ± 4 g) occurred at the spine (p < 0.001; the remaining 824 g was accrued at other sites).

Table Table 2.. Cross-Sectional Data: Total Body and Regional BMC (Mean ± SEM)
original image

As shown in Fig. 1, before puberty, longitudinal growth velocity was constant and more rapid in the legs than in the trunk. During puberty, leg length continued at a relatively constant velocity while trunk length accelerated. Both regions decelerated in the postpubertal period but growth velocity of the trunk at 16–17 years of age was ∼1.5 cm/year whereas growth velocity of the legs was ∼0.4 cm/year. Total body BMC accrual doubled between 11 and 13 years and then decelerated. Mineral accrual was 56 g/year at the legs, increasing almost 3-fold to 150 g/year and then decelerating to 36 g/year, while mineral accrual at the spine was 14 g/year, reaching a peak of 34 g/year and then decelerating to 10 g/year at 17 years.

Figure Figure 1.

Growth velocity for total, leg, and spine length (cm/year) and total body, leg, and spine BMC; BMC (g/year) as a function of chronological age (mean ± SEM). The shaded region depicts the peripubertal growth period (Tanner stage 2–4) mean ± 1 SD.

Relative growth of appendicular and axial skeletal size and mass

As shown in Fig. 2, the maturational development of bone size was more advanced than mineral accrual at all sites; even at a chronological age of 7 years, trunk and leg length and vertebral dimensions were about 70% of their respective peaks, while BMC of the legs and spine was ∼35% of its respective predicted adult peaks.

Figure Figure 2.

Leg and trunk length; leg and spine BMC, third lumbar vertebra (L3), width, height, and BMC, expressed as a percentage of the predicted young adult value (mean ± SEM). The shaded region depicts the peripubertal growth period (Tanner stage 2–4) mean ± 1 SD.

The peripubertal period (between 11 and 17 years) accounted for 61% of the increase in height between 7 and 17 years but 79% of the increase in BMC. The pre- and peripubertal period contributed 13 ± 0.1 cm and 16 ± 1 cm, respectively, to leg growth (p < 0.05), but almost four times more mineral was accrued peripubertally (704 ± 21 g) than prepubertally (198 ± 55 g; i.e., 78% of the total BMC gain in the legs was peripubertal; p < 0.001). The peripubertal period accounted for double the increment in trunk length (18 ± 1 cm) than the prepubertal period (9 ± 1 cm; 69% of 26 ± 1 cm trunk growth; p < 0.001) but a 5-fold higher mineral accrual (166 ± 4 g vs. 25 ± 10.0 g; i.e., 87% of 191 ± 10 g trunk accrual; p < 0.001).

Volumetric BMD

Fig. 3 shows BMC, bone volume, and vBMD of the total femur and L3 VB. BMC increased by 300% and bone volume increased by 200% (i.e., a doubling or more) while vBMD increased by 30–40%.

Figure Figure 3.

Total femoral shaft and third lumbar vertebra (L3) BMC, bone volume, and vBMD as a function of chronological age. The shaded region depicts the peripubertal growth period (Tanner stage 2–4) mean ± 1 SD.

Other regions

The femur and tibia had similar growth patterns with minimal acceleratory change at puberty (in contrast to the trunk; Fig 4). Longitudinal growth velocity of the humerus increased (like the trunk) while radius growth remained constant (like the legs). As growth velocity of the proximal and distal segments decelerated, BMC accelerated. Before puberty, biacromial and bi-iliac widths grew at a similar velocity. During puberty (11-17 years), growth velocity of biacromial width increased by ∼70%, whereas bi-iliac velocity increased by ∼20%. Both regions decelerated in the postpubertal period. Biacromial width increased by 13 ± 1 cm, while bi-iliac width increased by 8 ± 0.4 cm across at 7–17 years of age (Table 1). Of the 13 ± 1 cm increase in biacromial width, 34% (4 ± 0.3 cm) was gained during the prepubertal period and 66% (8 ± 0.3 cm) was gained during the peripubertal period (p < 0.001). By contrast, the prepubertal/peripubertal contributions to the 8 ± 0.4 cm increase in bi-iliac width was 41% (4 ± 0.4 cm) and 59% (5 ± 0.3 cm; p = 0.06).

Figure Figure 4.

Growth velocity for regional bone lengths (mm/year) and regional BMC; BMC (g/year) as a function of chronological age (mean ± SEM). The shaded region depicts the peripubertal growth period (Tanner stage 2–4) mean ± 1 SD.

As shown in Fig. 5, the periosteal diameter of the third metacarpal increased while endocortical diameter increased more modestly and then contracted. Thus, 25% of final metacarpal cortical thickness was caused by endocortical contraction. At the femoral midshaft, periosteal diameter increased with minimal expansion of the endocortical diameter. However, there was no detectable endocortical contraction within the age range studied at this site.

Figure Figure 5.

Third metacarpal and femoral midshaft dimensions and cortical thickness, as a function of chronological age (mean ± SEM). The shaded region depicts the peripubertal growth period (Tanner stage 2–4) mean ± 1 SD.

DISCUSSION

In this study of males, we observed a differing tempo of growth in size of a region and the mineral accruing within the periosteal envelope of the growing bone, and a differing tempo of growth from one region to another, as described in females.(12) Before puberty, growth of the legs was more rapid and contributed more to growth in height than the trunk. During puberty, growth of the trunk accelerated while growth of the legs decelerated. Both regions contributed relatively equally to growth in height but a greater proportion of trunk length was gained during puberty.

A greater proportion of mineral accrual (∼78%) occurred during the peripubertal period than during the prepubertal period at all sites, but the respective mineral accrual was 6-fold greater at the spine (166 g vs. 25 g) and 3-fold greater at the legs (704 g vs. 198 g). The increase in bone size preceded the increase in BMC; peak height velocity occurred at 13 years, and peak mineral accrual occurred at 14 years. Even at 7 years of age, bone size was ∼70% of its predicted adult peak value, while BMC was 35% of its peak. At 16 years of age, bone size was ∼100% of its predicted adult peak value, while BMC was still 10% below the predicted adult peak value.

We have suggested that the differing temporal patterns of growth in bone size and the mass within the periosteal envelope of a region and the differing tempo of growth of one region relative to another is likely to predispose to the development of deficits in bone size and vBMD found in adult females should illness intervene in growth.(12) Illness or exposure to a risk factor may produce a deficit in bone size, mineral accrual, or both at one site but not another, depending on the maturational level of the trait at the time of exposure to illness. A region or trait (e.g., BMC) further from its peak when illness occurs will be more severely affected than a region or trait nearer its adult peak at the time of exposure to illness.

For instance, exposure to an illness in early puberty is likely to affect both legs and spine. However, the spine will be affected selectively when exposure occurs in late puberty when the legs have completed their growth. We have reported this observation in a study of 210 women with anorexia nervosa diagnosed at different ages.(13) Because cessation of leg growth and acceleration of trunk growth are sex hormone dependent, delayed puberty in gymnasts may have greater effects on trunk than leg length while cessation of gymnastics is associated with acceleration of truncal, not leg growth.(14) In Klinefelter's syndrome, leg length is increased because of continued chondrocyte proliferation and epiphyseal patency.(19) In this way, depending on the age of exposure, diseases in childhood may produce a range of deficits in bone size of the legs, spine, or both in adulthood.

The increase in BMC and aBMD during growth is caused by the increase in bone size and mineral accrual within the growing bone. Both expressions of bone “mass” or “density” are size dependent.(20,21) Thus, if size is reduced the deficit in BMC (caused by reduced accrual within the smaller bone) will be exaggerated; if size is increased the deficit (caused by reduced accrual within the larger bone) will be obscured.(8) For example, in men with estrogen receptor defects or aromatase deficiency, the deficits in BMC or aBMD are greater at the spine than at the femur because VB size is reduced (relative to controls) and proximal femur size is increased (relative to controls).(22,23)

vBMD depends on the relative growth in size and mass of the region, not the absolute growth of either trait. If growth of both size and mineral accrual within the growing bone are similarly reduced, vBMD will be normal, as occurs in Turner's syndrome.(24) If size increases proportionally more than the mineral accrued within it, vBMD of the larger bone will be reduced. For example, in female rats, gonadectomy results in increased growth in femur length with disproportionately lower mineral accrual producing reduced vBMD in a bigger bone.(25) If mineral accrual is reduced proportionately more than the reduction in size, vBMD of the smaller bone will be reduced. For example, gonadectomy in male rats reduces accrual more than it reduces bone size, producing a smaller bone with reduced vBMD. Similarly, growth hormone administration in growing rats produces a greater increase in bone size than the increase in mineral accrual, resulting in reduced vBMD in the larger bone.(26)

During normal growth in females, vBMD is constant before puberty because of the proportional increases in size and mass.(12,27) Similar observations were made in males. vBMD of the femur was constant or increased slightly at puberty because of an increase in cortical thickness, itself caused by the lesser expansion of the endocortical than periosteal diameter. The increase in cortical thickness of the third metacarpal was caused by periosteal expansion with little change in the endocortical diameter, because of endocortical contraction as observed in women.(12,28) By contrast, VB vBMD increased throughout growth because of the increase in trabecular thickness, not numbers.(29,30) (The increase in vBMD at the spine reported here is likely to be an overestimate because of inclusion of the posterior processes in the measurement; see the Materials and Methods section.)

In summary, growth of bone is site specific, varying according to whether a region is axial or appendicular, cortical or trabecular, or whether the surface is periosteal or endosteal (endocortical, intracortical, or trabecular). Based on the observations made in this study, in the study of growing girls,(12) retired gymnasts(31) and women with anorexia nervosa,(13) we propose that the differing tempo of growth of size and mass of a region and the differing tempo of growth of one region relative to another in males and females predispose to region-specific deficits in bone size, BMC and vBMD should an illness or risk factor intervene. These deficits established in growth may in turn contribute to the differing patterns of fractures in later life. The pathogenesis of bone fragility at the spine or hip is likely to have its origins partly established during growth.

Acknowledgements

The authors thank the students and staff of the Ivanhoe Boys Grammar School, Ms. Vanessa De Luca and Sisters Jane Wilmont and Jan Edmonds for their contribution to this work. This study was supported by the Dairy Research and Development Corp. of Australia.

Ancillary