Abstract Context Growth in healthy children is associated with changes in bone density and microarchitecture. Trabecular morphology is an additional important determinant of bone strength, but little is currently known about trabecular morphology in healthy young people. Objective To investigate associations of trabecular morphology with increasing maturity and with body composition in healthy girls. Design Cross-sectional study. Setting Academic research center. Participants Eighty-six healthy girls aged 9 to 18 years. Main Outcome Measures High-resolution peripheral quantitative computed tomography and individual trabecula segmentation were used to assess volumetric bone density, microarchitecture, and trabecular morphology (plate-like vs rod-like) at the distal radius and tibia. Results Plate-like bone volume divided by total volume (pBV/TV) increased statistically significantly at the tibia (R = 0.41, P < 0.001), whereas rod-like BV/TV (rBV/TV) decreased statistically significantly at both the radius and tibia (R = −0.34, P = 0.003 and R = −0.28, P = 0.008, respectively) with increasing bone age. In multivariable models, lean mass positively correlated with pBV/TV and plate number at the radius and with plate thickness at both sites. In contrast, fat mass negatively correlated with plate thickness at the tibia and plate surface at both sites. In addition, fat mass positively correlated with rBV/TV and number at the tibia. pBV/TV at both the distal radius and tibia was positively correlated with spine bone mineral density. Conclusions Increasing maturity across late childhood and adolescence is associated with changes in trabecular morphology anticipated to contribute to bone strength. Body composition correlates with trabecular morphology, suggesting that muscle mass and adiposity in youth may contribute to long-term skeletal health. Childhood and adolescence are critical times for skeletal development, with up to 95% of peak bone mass acquired by age 18 years (1). Evidence suggests that the peak mass achieved is an important predictor of bone density and, by extension, fracture risk in later life (2). However, bone mass is not the only determinant of bone strength; the structure and microarchitecture of the skeleton also strongly influence the risk of fracture (3). Accumulating cross-sectional studies using high-resolution peripheral quantitative computed tomography (HR-pQCT) to obtain three-dimensional imaging of the distal radius and tibia in children have illuminated the substantial changes in bone size, cortical thickness, and cortical porosity associated with growth (4–7). However, the trabecular compartment, particularly in girls, appeared to be more static in these studies (4–7). More recently, a longitudinal study in a large cohort of 1071 children detected a ∼20% increase in trabecular bone density across pubertal growth at both the radius and tibia in boys, whereas in girls, there was a 10% increase at the tibia with no detectable increase at the radius (8). Analysis of mother-daughter pairs suggests that skeletal microarchitecture in the adult is largely determined in childhood (9). These data suggest that factors that influence microarchitectural changes during this developmental window may affect fracture risk throughout the life span. Among these factors, body composition plays a major role (10). Lean mass has consistently been shown to promote accrual of bone mass and cross-sectional area (11, 12) and has recently been demonstrated to correlate with trabecular bone number and density in children as measured by HR-pQCT (13). The relationship of fat mass to bone mass and microarchitecture in growing children is less clear, with several studies demonstrating conflicting results (10), and may depend on sex, pubertal status, and bone site. Trabecular morphology is an increasingly recognized contributor to bone strength, with plate-like trabeculae conferring more strength than rod-like trabeculae and axial alignment along the direction of primary loading similarly being associated with increased strength (14, 15). Individual trabecula segmentation (ITS) is a recently developed imaging technique that enables the in vivo distinction of plate-like and rod-like trabeculae as well as their associated number, thickness, surface, and length (16). This technique has been used to elucidate skeletal fragility in premenopausal women with idiopathic osteoporosis (17) and postmenopausal women taking glucocorticoids (18), as well as to explore differences among women of different racial backgrounds (19, 20). Our group has recently used ITS analysis in a cohort of adolescent and young adult female athletes and found that those with oligomenorrhea had decreased density and number of plate-like trabeculae at the distal radius compared with eumenorrheic athletes—a difference that was not evident when total trabecular bone density was evaluated (21). The potential clinical importance of trabecular morphology is highlighted by several case-control studies of postmenopausal women in which altered trabecular morphology has been associated with prevalent fracture (22–24). The objective of this study thus was to investigate the correlations of age and body composition with trabecular morphology in a cohort of healthy girls aged 9 to 18 years. We hypothesized that trabecular bone would become increasingly plate-like with increased maturity and that lean mass would positively correlate with plate-like morphology as well. Materials and Methods Participants This was a cross-sectional investigation of skeletal microarchitecture in healthy girls aged 9 to 18 years. Recruitment strategies and inclusion and exclusion criteria have been previously reported (25). Briefly, participants were recruited from the community using flyers and through a direct mail campaign. Exclusion criteria included disorders known or suspected to affect skeletal metabolism, including hypogonadism, thyroid disease, gastrointestinal disease, renal disease, and severe obesity (body mass index ≥99th percentile for age), as well as bone-active medications, including vitamin D >1000 IU/d and hormonal birth control. Participants were limited to girls due to previously reported differences in the pattern and timing of bone microarchitecture development between boys and girls (4). Ninety participants were enrolled in this study, of whom 86 underwent HR-pQCT imaging and are included in this analysis. This study was approved by the Partners Human Research Committee. Informed consent was obtained from participants who were 18 years old and from a parent or guardian for participants who were minors. Informed assent was obtained from minors. This trial was registered as NCT01180946 at ClinicalTrials.gov. Clinical and biochemical investigations Participants self-identified race and ethnicity. A standardized physical examination, including breast Tanner staging and anthropometric measurements, was performed. Pubertal status was categorized as prepubertal (Tanner 1, n = 11), midpubertal (Tanner 2 to 4, n = 34), and postpubertal (Tanner 5, n = 41). All blood samples were drawn in the morning in the fasting state. Insulin was measured by a chemiluminescent immunoassay (Beckman-Coulter, Brea, CA) with a sensitivity of 3 μIU/mL and interassay coefficient of variation (CV) of 3.1% to 5.6%. 25-hydroxyvitamin D was measured by liquid chromatography–tandem mass spectrometry with a lower limit of detection of 6 ng/mL and an interassay CV of 6% to 9%. 1,25-dihydroxyvitamin D [1,25(OH)2D] was measured by column chromatography followed by radioimmunoassay (LabCorp, Burlington, NC) with a lower limit of detection of 5 pg/mL and an interassay CV of 3.4% to 7.2%. Parathyroid hormone was measured by chemiluminescent immunoassay (Beckman-Coulter) with a sensitivity of 1 pg/mL and intra-assay and interassay CVs of 3% and 6%, respectively. Insulin-like growth factor 1 (IGF-1) was measured by enzyme-linked immunosorbent assay (Immunodiagnostic Systems, Tyne and Wear, UK) with a sensitivity of 3.1 μg/L and intra-assay and interassay CVs of 7%. Serum intact fibroblast growth factor 23 was measured by enzyme-linked immunosorbent assay (Kainos, Tokyo) with a sensitivity of 3 pg/mL and intra-assay and interassay CVs of ≤3% and ≤4%, respectively. Bone age was determined by radiographs of the left hand according to the method of Greulich and Pyle (26); all films were read by a single pediatric endocrinologist. Dual-energy x-ray absorptiometry imaging Scans of the lumbar spine were obtained (Hologic QDR Discovery A, Marlborough, MA) and analyzed using Hologic Discovery version 188.8.131.52, and z scores were generated using Hologic Apex 3.3 software. Spine dual-energy x-ray absorptiometry (DXA) data were excluded for two participants, one because of overlying jewelry and one because of scoliosis. Whole-body imaging was used to measure body composition (total lean and fat mass) in all participants and visceral and subcutaneous adiposity in a subset of 51 participants. HR-pQCT Total volumetric density, compartment-specific densities, and microarchitecture of the distal radius and tibia were assessed using HR-pQCT (XtremeCT; Scanco Medical AG, Brüttisellen, Switzerland) as previously described (27). Scans consisting of 100 computed tomography (CT) slices over a 9.02-mm region of interest (ROI) were acquired with an isotropic voxel size of 82 μm. The nondominant arm or leg was scanned unless there was a prior fracture at that region, in which case the contralateral side was scanned. As many of our participants (n = 39, 45%) had a bone age of <15 years and thus had remaining growth potential, to account for limb lengthening during growth, rather than a fixed site, we defined the ROI as the 7% site at the radius and the 8% site at the tibia as described (28). We used semiautomated software to segment cortical and trabecular regions with a threshold-based algorithm, followed by individual review and manual adjustment as necessary. Same-day reproducibility for repeated measurements is 0.2% to 1.4% for density values, 0.3% to 8.6% for trabecular microarchitecture parameters, and 0.6% to 2.4% for cortical microarchitecture parameters. We obtained high-quality tibia scans from all 86 participants who had HR-pQCT scanning. We excluded radius scans from 9 participants due to motion artifact, leaving 77 participants for analysis. ITS The trabecular bone compartment was extracted from each HR-pQCT image, and the entire compartment underwent ITS-based morphological analysis as described (16). Morphological parameters measured included plate- and rod-like bone volume divided by total volume (pBV/TV and rBV/TV); plate and rod number density; plate and rod thickness; plate surface area; rod length; plate tissue fraction; plate-plate, plate-rod, and rod-rod junction density; and trabecular orientation (axial bone volume fraction). This method has previously been shown to correlate well with ITS parameters as determined by micro-CT (29). Statistical analyses We used Stata 12.1 (StataCorp LP, College Station, TX) for all analyses. This was an exploratory analysis of a previously described cohort (25), and thus, we did not use power calculations to determine the sample size. Clinical and laboratory data are reported as mean ± standard deviation or median (interquartile range) for skewed data. We used linear regression to assess the associations of clinical and biochemical variables with measures of trabecular density and morphology. Associations of bone outcomes with body composition and biochemical parameters were adjusted for bone age. Multivariable regression was used to assess independent contribution of fat and lean mass to trabecular morphology. Associations with a P value of <0.05 are reported as statistically significant. For comparisons of trabecular morphology by pubertal status, we used analysis of variance, followed by pairwise comparisons adjusted for multiple comparison by the Bonferroni method. Results Participant characteristics Table 1 describes the clinical characteristics of the study participants. As reported previously, bone age was on average slightly higher than chronological age and had a left-skewed distribution. Both visceral adipose tissue and fasting insulin had right-skewed distributions. As expected, several parameters were statistically significantly associated with age in these growing children, including height, weight, body mass index, fat mass, lean mass, and spine bone mineral density (BMD) (positively correlated); phosphate and 1,25(OH)2D (negatively correlated); and fasting insulin and IGF-1 (inverted U-shaped distribution). Table 1. Clinical Characteristics of Participants (n = 86) Characteristic Value Age, (y) 14.2 ± 2.6 Bone age (y) 15.0 (12.0, 17.0) Race White 67 (78%) Black 5 (6%) Asian 2 (2%) Multiple 8 (9%) None of the above 4 (5%) Hispanic ethnicity 9 (10%) Height (cm) 157.9 ± 10.5 Height z score 0.2 ± 1.1 Weight (kg) 53.0 ± 13.1 Weight z score 0.3 ± 1.0 BMI (kg/m2) 21.0 ± 4.1 BMI z score 0.3 ± 1.0 Breast Tanner stage 1 (prepubertal) 11 (13%) 2–4 (pubertal) 34 (39%) 5 (postpubertal) 41 (48%) Lean body mass (kg) 35.8 ± 7.6 Fat mass (kg) 16.7 ± 6.8 Visceral adipose tissue (g) 181 ± 101 Subcutaneous adipose tissue (g) 1055 ± 434 Spine BMD (g/cm2) 0.852 ± 0.181 Spine BMD z score −0.2 ± 1.1 Calcium (mg/dL) 9.6 ± 0.3 Phosphate (mg/dL) 4.3 ± 0.6 25OHD (ng/mL) 22.0 ± 6.7 1,25(OH)2D (pg/mL) 50.3 ± 15.1 PTH (pg/mL) 34.7 ± 11.8 IGF-1 (ng/mL) 193.4 ± 66.1 Intact FGF23 (pg/mL) 42.6 ± 11.1 HbA1c (%) 5.4 ± 0.3 Insulin (μIU/mL) 7.8 (5.8, 10.4) Characteristic Value Age, (y) 14.2 ± 2.6 Bone age (y) 15.0 (12.0, 17.0) Race White 67 (78%) Black 5 (6%) Asian 2 (2%) Multiple 8 (9%) None of the above 4 (5%) Hispanic ethnicity 9 (10%) Height (cm) 157.9 ± 10.5 Height z score 0.2 ± 1.1 Weight (kg) 53.0 ± 13.1 Weight z score 0.3 ± 1.0 BMI (kg/m2) 21.0 ± 4.1 BMI z score 0.3 ± 1.0 Breast Tanner stage 1 (prepubertal) 11 (13%) 2–4 (pubertal) 34 (39%) 5 (postpubertal) 41 (48%) Lean body mass (kg) 35.8 ± 7.6 Fat mass (kg) 16.7 ± 6.8 Visceral adipose tissue (g) 181 ± 101 Subcutaneous adipose tissue (g) 1055 ± 434 Spine BMD (g/cm2) 0.852 ± 0.181 Spine BMD z score −0.2 ± 1.1 Calcium (mg/dL) 9.6 ± 0.3 Phosphate (mg/dL) 4.3 ± 0.6 25OHD (ng/mL) 22.0 ± 6.7 1,25(OH)2D (pg/mL) 50.3 ± 15.1 PTH (pg/mL) 34.7 ± 11.8 IGF-1 (ng/mL) 193.4 ± 66.1 Intact FGF23 (pg/mL) 42.6 ± 11.1 HbA1c (%) 5.4 ± 0.3 Insulin (μIU/mL) 7.8 (5.8, 10.4) Data expressed as mean ± standard deviation, median (25th percentile, 75th percentile) or number (percent), as appropriate. Participants who self-identified as multiracial identified as white and black (n = 3), black and Native American (n = 1), and Asian and white (n = 4). Abbreviations: BMI, body mass index; HbA1c, hemoglobin A1c; PTH, parathyroid hormone; 25OHD, 25-hydroxyvitamin D. View Large Correlations of trabecular density and morphology with bone age Overall, volumetric bone density and microarchitectural parameters as measured by HR-pQCT had numerically stronger correlations with bone age than with calendar age (data not shown), and thus all analyses in this report use bone age as a proxy for maturity. As shown in Fig. 1, we did not observe a correlation of overall trabecular bone volume divided by total volume (BV/TV) as determined by standard HR-pQCT analysis with bone age at either the radius or tibia, although there was a trend toward an increase with increasing bone age at the tibia (Fig. 1A and 1B). In contrast, we observed an increase of pBV/TV with bone age at the tibia (Fig. 1D) and a decrease of rBV/TV with bone age at both the radius and tibia (Fig. 1E and 1F). We observed no statistically significant association of pBV/TV with bone age at the radius (Fig. 1C). The correlations were substantially similar when using age rather than bone age as the independent variable (R = 0.33, P = 0.002 for pBV/TV at the tibia, R = −0.38, P = 0.001 for rBV/TV at the radius, and R = −0.29, P = 0.007 for rBV/TV at the tibia). These correlations were also substantially similar when, in sensitivity analyses, we restricted the cohort to white participants (n = 67) (R = 0.45, P < 0.001 for pBV/TV at the tibia, R = −0.43, P = 0.001 for rBV/TV at the radius, and R = −0.49, P < 0.001 for rBV/TV at the tibia). Figure 1. View largeDownload slide (A and B) Scatterplots of bone age vs total BV/TV, (C and D) pBV/TV, and (E and F) rBV/TV. Figure 1. View largeDownload slide (A and B) Scatterplots of bone age vs total BV/TV, (C and D) pBV/TV, and (E and F) rBV/TV. As shown in Table 2, the increases in pBV/TV at the tibia were consistent with bone age–associated increases in plate number as well as plate size as measured by both thickness and surface area. At the radius, despite no statistically significant correlation of bone age with pBV/TV, we did observe a positive correlation with plate surface area. These relationships are highlighted in the scatterplots in Fig. 2. Negative correlations of rBV/TV with bone age were consistent with negative correlations of bone age with rod number at both the radius and tibia, although we did observe a statistically significant positive correlation of rod length with bone age at the tibia. In addition, axial bone volume fraction was positively correlated with bone age at the tibia but not at the radius. As shown in the line plots in Fig. 2C, 2F, and 2I, we observed a generally linear increase in plate thickness and surface across puberty, reaching statistical significance at the tibia in postpubertal girls vs both prepubertal and midpubertal girls, with an additional statistically significant difference in radius plate thickness in midpubertal vs prepubertal girls. Table 2. Correlations of Skeletal Outcomes (Trabecular Density and Morphology) With Bone Age Trabecular Parameter Radius Tibia R P R P BV/TV −0.14 0.238 0.20 0.065 Axial BV/TV −0.06 0.585 0.33 0.002 Plate BV/TV 0.06 0.602 0.41 <0.001 Plate BV/BV 0.20 0.087 0.43 <0.001 Plate TbN −0.03 0.824 0.24 0.026 Plate TbTh 0.14 0.235 0.52 <0.001 Plate TbS 0.24 0.036 0.43 <0.001 Rod BV/TV −0.34 0.003 −0.28 0.008 Rod TbN −0.32 0.004 −0.27 0.012 Rod TbTh 0.01 0.928 −0.05 0.655 Rod Tbℓ 0.07 0.537 0.26 0.017 Rod-Rod JuncD −0.34 0.003 −0.32 0.002 Rod-Plate JuncD −0.12 0.287 −0.06 0.588 Plate-Plate JuncD −0.02 0.872 0.18 0.103 Trabecular Parameter Radius Tibia R P R P BV/TV −0.14 0.238 0.20 0.065 Axial BV/TV −0.06 0.585 0.33 0.002 Plate BV/TV 0.06 0.602 0.41 <0.001 Plate BV/BV 0.20 0.087 0.43 <0.001 Plate TbN −0.03 0.824 0.24 0.026 Plate TbTh 0.14 0.235 0.52 <0.001 Plate TbS 0.24 0.036 0.43 <0.001 Rod BV/TV −0.34 0.003 −0.28 0.008 Rod TbN −0.32 0.004 −0.27 0.012 Rod TbTh 0.01 0.928 −0.05 0.655 Rod Tbℓ 0.07 0.537 0.26 0.017 Rod-Rod JuncD −0.34 0.003 −0.32 0.002 Rod-Plate JuncD −0.12 0.287 −0.06 0.588 Plate-Plate JuncD −0.02 0.872 0.18 0.103 Correlations with P < 0.05 shown in bold. Abbreviations: JuncD, junction density; Tbℓ, trabecular length; TbN, trabecular number; TbS, trabecular surface; TbTh, trabecular thickness. View Large Figure 2. View largeDownload slide (A and B) Scatterplots of bone age vs plate TbN, (D and E) plate TbTh, and (G and H) plate TbS. Panels (C) TbN, (F) TbTh, and (I) TbS are line plots depicting mean ± standard error across stages of puberty defined by breast Tanner stage (Tanner 1: prepubertal; Tanner 2 to 4: midpubertal, Tanner 5: postpubertal). Radius: dashed line; tibia: solid line. aP < 0.05 vs prepubertal; bP < 0.05 vs midpubertal. TbN, trabecular number; TbTh, trabecular thickness; TbS, trabecular surface. Figure 2. View largeDownload slide (A and B) Scatterplots of bone age vs plate TbN, (D and E) plate TbTh, and (G and H) plate TbS. Panels (C) TbN, (F) TbTh, and (I) TbS are line plots depicting mean ± standard error across stages of puberty defined by breast Tanner stage (Tanner 1: prepubertal; Tanner 2 to 4: midpubertal, Tanner 5: postpubertal). Radius: dashed line; tibia: solid line. aP < 0.05 vs prepubertal; bP < 0.05 vs midpubertal. TbN, trabecular number; TbTh, trabecular thickness; TbS, trabecular surface. Correlations of ITS parameters with body composition Given well-described associations of body composition with bone density and microarchitecture in children (10), we examined the correlations of fat and lean mass as measured by DXA with trabecular morphology. In univariate analyses adjusted for bone age, lean mass was positively associated with pBV/TV at both the radius and tibia (P = 0.022 and 0.025, respectively), whereas fat mass was positively correlated with rBV/TV at the tibia (P = 0.005). Because fat and lean mass were themselves positively correlated (R = 0.59, P < 0.001), we generated models for each ITS parameter with lean mass, fat mass, bone age, and height as covariates. As shown in Table 3, in our adjusted models, lean mass remained positively correlated with pBV/TV at the radius but not at the tibia. In addition, lean mass was associated with more numerous plates at the radius, thicker plates at both the radius and tibia, and thicker rods at the radius. In contrast, fat mass correlated inversely with plate thickness at the tibia and plate surface area at both sites. Fat mass also correlated positively with rBV/TV at the tibia, corresponding to an increase in rod number but a decrease in rod length. Finally, fat mass was associated with both rod and plate connectivity at the tibia. In addition, we found no associations of visceral adipose fat mass or of the visceral/subcutaneous fat ratio with any trabecular parameters (data not shown). In a sensitivity analysis restricting the cohort to the white participants only, the inverse correlations of fat mass with plate thickness at the tibia and plate surface area at both sites as well as the positive correlation of fat mass with rBV/TV and rod number at the tibia persisted. Lean mass showed a statistically significant inverse correlation with rod thickness at the tibia, but the positive correlations seen in the whole cohort were not observed when limited to white participants only. Table 3. Multivariable Correlations of Body Composition Parameters (Fat and Lean Mass) With Trabecular Density and Morphology Trabecular Parameter Radius Tibia Fat Mass Lean Mass Fat Mass Lean Mass BV/TV −0.01 0.48 0.06 0.47 Axial BV/TV −0.11 0.57a −0.03 0.27 Plate BV/TV −0.11 0.55a −0.05 0.39 Plate BV/BV −0.16 0.52a −0.13 0.17 Plate TbN −0.05 0.56a 0.21 0.27 Plate TbTh −0.20 0.56a −0.26a 0.56a Plate TbS −0.29a 0.43 −0.26a 0.18 Rod BV/TV 0.11 0.27 0.31a 0.12 Rod TbN 0.18 0.19 0.33a 0.16 Rod TbTh −0.27 0.61a −0.16 −0.39 Rod Tbℓ −0.21 −0.43 −0.46b −0.09 Rod-rod JuncD 0.23 0.07 0.32a 0.10 Rod-plate JuncD 0.05 0.44 0.38c 0.22 Plate-plate JuncD −0.01 0.52 0.26 0.26 Trabecular Parameter Radius Tibia Fat Mass Lean Mass Fat Mass Lean Mass BV/TV −0.01 0.48 0.06 0.47 Axial BV/TV −0.11 0.57a −0.03 0.27 Plate BV/TV −0.11 0.55a −0.05 0.39 Plate BV/BV −0.16 0.52a −0.13 0.17 Plate TbN −0.05 0.56a 0.21 0.27 Plate TbTh −0.20 0.56a −0.26a 0.56a Plate TbS −0.29a 0.43 −0.26a 0.18 Rod BV/TV 0.11 0.27 0.31a 0.12 Rod TbN 0.18 0.19 0.33a 0.16 Rod TbTh −0.27 0.61a −0.16 −0.39 Rod Tbℓ −0.21 −0.43 −0.46b −0.09 Rod-rod JuncD 0.23 0.07 0.32a 0.10 Rod-plate JuncD 0.05 0.44 0.38c 0.22 Plate-plate JuncD −0.01 0.52 0.26 0.26 Data expressed as standardized regression coefficient. All correlations adjusted for bone age and height. Abbreviations: JuncD, junction density; Tbℓ, trabecular length; TbN, trabecular number; TbS, trabecular surface; TbTh, trabecular thickness. a P < 0.05. b P < 0.001. c P < 0.01. View Large Correlations of ITS parameters with biochemical and hormonal determinants of skeletal metabolism We next examined the associations of total BV/TV, pBV/TV, and rBV/TV with several important determinants of bone growth and mineralization. These correlations were all adjusted for bone age (data not shown). Notably, serum phosphate positively correlated with total BV/TV and pBV/TV at both the distal radius (P = 0.041 and P = 0.011, respectively) and the tibia (P = 0.008 and P = 0.008, respectively). Although lean mass and fat mass were both inversely correlated with serum phosphate, this correlation became nonsignificant after adjustment for bone age. We found no correlations of 25OHD, 1,25(OH)2D, serum calcium, or FGF23 with trabecular parameters. PTH positively correlated with total BV/TV at the radius (P = 0.028) but not at the tibia and not with rBV/TV or pBV/TV at any site. Given the inverted U-shaped correlation of age with IGF-1 and fasting insulin, we generated three splines with knots placed at bone age ≤12 years and ≤16 years by visual inspection. Within these subgroups, we observed no correlations of IGF-1 or fasting insulin with trabecular parameters. Correlations of trabecular morphology with spine BMD The spine, a clinically important site due to its susceptibility to major osteoporotic fracture, is composed primarily of trabecular bone. We thus investigated to what extent measures of trabecular density and morphology were correlated with spine BMD as measured by DXA. As shown in Table 4, both total and plate-like trabecular density at the tibia were statistically significantly correlated with spine BMD. Of note, the correlation with pBV/TV was numerically stronger. pBV/TV at the radius also correlated statistically significantly with spine BMD; we observed no statistically significant correlation with total BV/TV. There was no correlation of rBV/TV at either site with spine BMD. Table 4. Correlations of Trabecular Density and Morphology With Spine BMD as Measured by DXA Trabecular Parameter Radius Tibia R P R P BV/TV 0.14 0.227 0.43 <0.001 pBV/TV 0.29 0.013 0.56 <0.001 rBV/TV −0.05 0.681 −0.17 0.117 Trabecular Parameter Radius Tibia R P R P BV/TV 0.14 0.227 0.43 <0.001 pBV/TV 0.29 0.013 0.56 <0.001 rBV/TV −0.05 0.681 −0.17 0.117 Correlations with P < 0.05 shown in bold. View Large Discussion Using ITS analysis, we here report what is, to our knowledge, a novel assessment of in vivo trabecular morphology in healthy preadolescent and adolescent girls. We demonstrate that, with increasing maturity, the density of plate-like trabeculae increases at the distal tibia and that the plate surface area increases at both the distal radius and tibia, with additional increases in plate number and thickness observed at the tibia. Conversely, the density of rod-like trabeculae decreases with age at both sites, driven by a decrease in rod number. These data are consistent with several previous studies that demonstrate minimal changes in overall trabecular density with age across pubertal growth (4–8) but reveal important age-related morphological changes predicted to increase bone strength. These results are consistent with previous data derived from histomorphometry of iliac crest biopsy samples, which demonstrated that, in growing children, remodeling of trabecular bone proceeds with a small but measurable positive balance such that trabecular width increases with age (30). Our data suggest that this positive balance promotes a transition from a more rod-like to a more plate-like morphology. This morphology has been empirically correlated with fracture resistance. For example, Zhou et al. (15) found that plate and axial bone volume fraction were strongly and positively related to yield strength in human trabecular bone samples, whereas rod-like parameters were weakly and inversely correlated. Similarly, Arlot et al. (14) found that the density of microcracks and diffuse damage in lumbar vertebrae was higher in specimens with a more rod-like phenotype. Trabecular rods have the additional disadvantage that they may be more susceptible to remodeling-induced transection, after which they no longer contribute mechanically to bone strength (31, 32). Identifying modifiable influences on bone accrual during growth is critical to minimize the risk of later osteoporosis. To this end, the independent correlations of lean and fat mass to trabecular morphology in our cohort extend the growing body of literature regarding the impact of body composition on bone accrual and fracture risk in children (10). We demonstrate that, after multivariable adjustment, lean mass is positively correlated with plate-like trabecular morphology at both sites, but particularly at the radius. This is consistent with previous literature and with studies that link muscular force and physical activity to bone accrual in children (33, 34), suggesting that mechanical loading of the skeleton is an important determinant of bone accrual. However, it is clear that loading is not the only mediator of the association between lean mass and bone accrual given that previous studies report similar standardized effect sizes for the relationship between lean mass and/or physical activity and trabecular parameters at both the tibia, a weight-bearing site, and the distal radius, a non-weight-bearing site (13, 34). Indeed, in our data, the effect size for lean mass as a determinant of plate-like parameters was in fact higher at the radius. This suggests that other factors such as secreted myokines may also mediate this muscle-bone relationship (35). By contrast, we found that fat mass is correlated with rod parameters at the tibia and, interestingly, is inversely correlated with plate thickness at the tibia and with plate surface area at both the radius and tibia. These data suggest that fat mass may impair a physiologic rod-to-plate transition with age. These data are consistent with HR-pQCT data from Farr et al. (13) demonstrating a negative relationship between trabecular thickness and fat mass in boys (although not in girls). In contrast to that report, we did not detect an association of fat mass with overall trabecular density at either site. The relationship between fat mass, obesity, and bone health is not straightforward; fat mass provides mechanical loading to bone while secreting a number of factors with positive, negative, and pleiotropic effects on bone accrual (36). Notably, a rodent model of diet-induced obesity demonstrated an initial increase in bone accrual, followed by a decrease in bone formation coincident with altered adipocyte metabolism (37). These data suggest that fat mass may have differing effects depending on the age of the organism and on the degree of fat accumulation and metabolic disruption. Several skeletal parameters “track” across childhood growth such that children who start at a lower percentile will persist at a lower percentile throughout development. These parameters include areal bone density as measured by DXA (38), volumetric density as measured by CT (39, 40), and cortical and medullary cross-sectional areas (39, 40). It is thus likely by analogy that trabecular morphology similarly tracks and that children with a more plate-like architecture maintain this architecture as adults. Importantly, the skeletal deterioration that occurs with aging and hypogonadism is associated with a transition back to a more rod-like morphology (41, 42). These considerations suggest that optimal development of trabecular morphology during the period of bone accrual may confer bone strength advantages not only in youth but also throughout the life span and that exposure to excess adiposity during this critical window may have a negative and persistent impact on bone strength. Because our data are cross-sectional, effects of age and body composition on development of trabecular morphology are not directly tested here. Of note as well, the bone age of our participants as a proxy for maturity was on average higher than chronological age, consistent with previously described secular trends in American adolescents (43). In addition, our data are limited to girls; patterns and predictors of microarchitecture are different in boys and girls (4, 5, 7, 8), and trabecular morphology likely differs as well. Given the small numbers of participants in our cohort who did not identify as white, we do not have the ability to evaluate differences in trabecular morphology in children of differing racial backgrounds, a factor known to be important in adults (19, 20). The absence of correlation of lean mass with plate-like parameters at the radius when limiting the cohort to white participants may be due to the smaller sample size but raises the possibility of racial differences in determinants of trabecular morphology in childhood, which should be further explored. We also do not have data on physical activity in our participants, which is an important determinant of bone accrual (34). There are also technical considerations important to consider in the interpretation of these data. Although plate-like trabecular parameters measured by HR-pQCT are generally highly correlated with those measured by the gold-standard micro-CT, rod-like parameters have weaker correlations, likely due to the larger voxel size of HR-pQCT and more limited ability to resolve smaller structures (29). In addition, the choice of the ROI to measure in a growing bone to compare the same “relative” region among children of different ages and heights is challenging (44). Rather than using a fixed site, we chose, like other investigators, to perform measurements in a volume of interest located relative to limb length (5, 6). Given that bone elongation does not proceed symmetrically and that the relative contribution of the distal and proximal growth plates to elongation may vary with age and pubertal stage, apparent variation in trabecular morphology may be in part due to functional positioning differences (44). The development of imaging technologies with higher resolution and the ability to scan a larger ROI may help resolve this challenge. In addition, HR-pQCT images the distal radius and tibia, which are sites less subject to fragility fracture; the correlations of plate-like trabecular density with DXA measures of the spine suggest our findings at the radius and tibia are representative of trabecular changes at this clinically important site. In conclusion, our data suggest that, across pubertal growth in healthy girls, trabecular bone morphology changes substantially in a manner predicted to increase bone strength and to better retain architectural integrity in the face of aging and hypogonadism. The observed associations of lean and fat mass with trabecular morphology are intriguing. If these factors are confirmed in longitudinal and mechanistic studies to directly affect trabecular morphology, an important determinant of the biomechanical properties of bone, this would suggest that body composition during youth may have long-term implications for skeletal health and fracture risk. Abbreviations: 1,25(OH)2D 1,25-dihydroxyvitamin D BMD bone mineral density BV/TV bone volume divided by total volume CT computed tomography CV coefficient of variation DXA dual-energy x-ray absorptiometry HR-pQCT high-resolution peripheral quantitative computed tomography IGF-1 insulin-like growth factor 1 ITS individual trabecula segmentation pBV/TV plate-like bone volume divided by total volume rBV/TV rod-like bone volume divided by total volume ROI region of interest. Acknowledgments Financial Support: The work described in this article was supported by National Institutes of Health Grants T32DK007028, F32HD071759, K23DK073356, K23DK105350, K24HD071843, S10RR023405, and UL1RR025758; Harvard Clinical and Translational Science Center; and the National Center for Research Resources. The work was also supported by a Massachusetts General Hospital Physician-Scientist Development Award, a Harvard Medical School Office for Diversity Inclusion and Community Partnership Award, and a Claflin Distinguished Scholar Award. Disclosure Summary: The authors have nothing to disclose. References 1. Baxter-Jones AD, Faulkner RA, Forwood MR, Mirwald RL, Bailey DA. Bone mineral accrual from 8 to 30 years of age: an estimation of peak bone mass. J Bone Miner Res . 2011; 26( 8): 1729– 1739. Google Scholar CrossRef Search ADS PubMed 2. Seeman E, Hopper JL, Bach LA, Cooper ME, Parkinson E, McKay J, Jerums G. Reduced bone mass in daughters of women with osteoporosis. N Engl J Med . 1989; 320( 9): 554– 558. Google Scholar CrossRef Search ADS PubMed 3. Bouxsein ML, Seeman E. Quantifying the material and structural determinants of bone strength. Best Pract Res Clin Rheumatol . 2009; 23( 6): 741– 753. Google Scholar CrossRef Search ADS PubMed 4. Kirmani S, Christen D, van Lenthe GH, Fischer PR, Bouxsein ML, McCready LK, Melton LJ III, Riggs BL, Amin S, Müller R, Khosla S. Bone structure at the distal radius during adolescent growth. J Bone Miner Res . 2009; 24( 6): 1033– 1042. Google Scholar CrossRef Search ADS PubMed 5. Wang Q, Wang XF, Iuliano-Burns S, Ghasem-Zadeh A, Zebaze R, Seeman E. Rapid growth produces transient cortical weakness: a risk factor for metaphyseal fractures during puberty. J Bone Miner Res . 2010; 25( 7): 1521– 1526. Google Scholar CrossRef Search ADS PubMed 6. Burrows M, Liu D, Moore S, McKay H. Bone microstructure at the distal tibia provides a strength advantage to males in late puberty: an HR-pQCT study. J Bone Miner Res . 2010; 25( 6): 1423– 1432. Google Scholar PubMed 7. Nishiyama KK, Macdonald HM, Moore SA, Fung T, Boyd SK, McKay HA. Cortical porosity is higher in boys compared with girls at the distal radius and distal tibia during pubertal growth: an HR-pQCT study. J Bone Miner Res . 2012; 27( 2): 273– 282. Google Scholar CrossRef Search ADS PubMed 8. Gabel L, Macdonald HM, McKay HA. Sex differences and growth-related adaptations in bone microarchitecture, geometry, density, and strength from childhood to early adulthood: a mixed longitudinal HR-pQCT study. J Bone Miner Res . 2016; 32( 2): 250– 263. Google Scholar CrossRef Search ADS PubMed 9. Wang Q, Ghasem-Zadeh A, Wang XF, Iuliano-Burns S, Seeman E. Trabecular bone of growth plate origin influences both trabecular and cortical morphology in adulthood. J Bone Miner Res . 2011; 26( 7): 1577– 1583. Google Scholar CrossRef Search ADS PubMed 10. Farr JN, Dimitri P. The impact of fat and obesity on bone microarchitecture and strength in children. Calcif Tissue Int . 2016; 100( 5): 500– 513. Google Scholar CrossRef Search ADS PubMed 11. Goulding A, Taylor RW, Grant AM, Jones S, Taylor BJ, Williams SM. Relationships of appendicular LMI and total body LMI to bone mass and physical activity levels in a birth cohort of New Zealand five-year olds. Bone . 2009; 45( 3): 455– 459. Google Scholar CrossRef Search ADS PubMed 12. Wey HE, Binkley TL, Beare TM, Wey CL, Specker BL. Cross-sectional versus longitudinal associations of lean and fat mass with pQCT bone outcomes in children. J Clin Endocrinol Metab . 2011; 96( 1): 106– 114. Google Scholar CrossRef Search ADS PubMed 13. Farr JN, Amin S, LeBrasseur NK, Atkinson EJ, Achenbach SJ, McCready LK, Joseph Melton L III, Khosla S. Body composition during childhood and adolescence: relations to bone strength and microstructure. J Clin Endocrinol Metab . 2014; 99( 12): 4641– 4648. Google Scholar CrossRef Search ADS PubMed 14. Arlot ME, Burt-Pichat B, Roux JP, Vashishth D, Bouxsein ML, Delmas PD. Microarchitecture influences microdamage accumulation in human vertebral trabecular bone. J Bone Miner Res . 2008; 23( 10): 1613– 1618. Google Scholar CrossRef Search ADS PubMed 15. Zhou B, Liu XS, Wang J, Lu XL, Fields AJ, Guo XE. Dependence of mechanical properties of trabecular bone on plate-rod microstructure determined by individual trabecula segmentation (ITS). J Biomech . 2014; 47( 3): 702– 708. Google Scholar CrossRef Search ADS PubMed 16. Liu XS, Sajda P, Saha PK, Wehrli FW, Bevill G, Keaveny TM, Guo XE. Complete volumetric decomposition of individual trabecular plates and rods and its morphological correlations with anisotropic elastic moduli in human trabecular bone. J Bone Miner Res . 2007; 23( 2): 223– 235. Google Scholar CrossRef Search ADS 17. Liu XS, Cohen A, Shane E, Stein E, Rogers H, Kokolus SL, Yin PT, McMahon DJ, Lappe JM, Recker RR, Guo XE. Individual trabeculae segmentation (ITS)–based morphological analysis of high-resolution peripheral quantitative computed tomography images detects abnormal trabecular plate and rod microarchitecture in premenopausal women with idiopathic osteoporosis. J Bone Miner Res . 2010; 25( 7): 1496– 1505. Google Scholar CrossRef Search ADS PubMed 18. Sutter S, Nishiyama KK, Kepley A, Zhou B, Wang J, McMahon DJ, Guo XE, Stein EM. Abnormalities in cortical bone, trabecular plates, and stiffness in postmenopausal women treated with glucocorticoids. J Clin Endocrinol Metab . 2014; 99( 11): 4231– 4240. Google Scholar CrossRef Search ADS PubMed 19. Liu XS, Walker MD, McMahon DJ, Udesky J, Liu G, Bilezikian JP, Guo XE. Better skeletal microstructure confers greater mechanical advantages in Chinese-American women versus white women. J Bone Miner Res . 2011; 26( 8): 1783– 1792. Google Scholar CrossRef Search ADS PubMed 20. Putman MS, Yu EW, Lin D, Darakananda K, Finkelstein JS, Bouxsein ML. Differences in trabecular microstructure between black and white women assessed by individual trabecular segmentation analysis of HR-pQCT images. J Bone Miner Res . 2017; 32( 5): 1100– 1108. Google Scholar CrossRef Search ADS PubMed 21. Mitchell DM, Tuck P, Ackerman KE, Cano Sokoloff N, Woolley R, Slattery M, Lee H, Bouxsein ML, Misra M. Altered trabecular bone morphology in adolescent and young adult athletes with menstrual dysfunction. Bone . 2015; 81: 24– 30. Google Scholar CrossRef Search ADS PubMed 22. Liu XS, Stein EM, Zhou B, Zhang CA, Nickolas TL, Cohen A, Thomas V, McMahon DJ, Cosman F, Nieves J, Shane E, Guo XE. Individual trabecula segmentation (ITS)–based morphological analyses and microfinite element analysis of HR-pQCT images discriminate postmenopausal fragility fractures independent of DXA measurements. J Bone Miner Res . 2012; 27( 2): 263– 272. Google Scholar CrossRef Search ADS PubMed 23. Liu XS, Wang J, Zhou B, Stein E, Shi X, Adams M, Shane E, Guo XE. Fast trabecular bone strength predictions of HR-pQCT and individual trabeculae segmentation-based plate and rod finite element model discriminate postmenopausal vertebral fractures. J Bone Miner Res . 2013; 28( 7): 1666– 1678. Google Scholar CrossRef Search ADS PubMed 24. Wang J, Stein EM, Zhou B, Nishiyama KK, Yu YE, Shane E, Guo XE. Deterioration of trabecular plate-rod and cortical microarchitecture and reduced bone stiffness at distal radius and tibia in postmenopausal women with vertebral fractures. Bone . 2016; 88: 39– 46. Google Scholar CrossRef Search ADS PubMed 25. Mitchell DM, Jüppner H, Burnett-Bowie SM. FGF23 is not associated with age-related changes in phosphate, but enhances renal calcium reabsorption in girls. J Clin Endocrinol Metab . 2017; 102( 4): 1151– 1160. Google Scholar CrossRef Search ADS PubMed 26. Greulich WW, Pyle SI. Radiographic atlas of skeletal development of the hand and wrist . 2nd ed. Stanford, CA: Stanford University Press; 1959. 27. Boutroy S, Bouxsein ML, Munoz F, Delmas PD. In vivo assessment of trabecular bone microarchitecture by high-resolution peripheral quantitative computed tomography. J Clin Endocrinol Metab . 2005; 90( 12): 6508– 6515. Google Scholar CrossRef Search ADS PubMed 28. Liu D, Burrows M, Egeli D, McKay H. Site specificity of bone architecture between the distal radius and distal tibia in children and adolescents: an HR-pQCT study. Calcif Tissue Int . 2010; 87( 4): 314– 323. Google Scholar CrossRef Search ADS PubMed 29. Liu XS, Shane E, McMahon DJ, Guo XE. Individual trabecula segmentation (ITS)–based morphological analysis of microscale images of human tibial trabecular bone at limited spatial resolution. J Bone Miner Res . 2011; 26( 9): 2184– 2193. Google Scholar CrossRef Search ADS PubMed 30. Parfitt AM, Travers R, Rauch F, Glorieux FH. Structural and cellular changes during bone growth in healthy children. Bone . 2000; 27( 4): 487– 494. Google Scholar CrossRef Search ADS PubMed 31. Mosekilde L. Consequences of the remodelling process for vertebral trabecular bone structure: a scanning electron microscopy study (uncoupling of unloaded structures). Bone Miner . 1990; 10( 1): 13– 35. Google Scholar CrossRef Search ADS PubMed 32. Wehrli FW, Gomberg BR, Saha PK, Song HK, Hwang SN, Snyder PJ. Digital topological analysis of in vivo magnetic resonance microimages of trabecular bone reveals structural implications of osteoporosis. J Bone Miner Res . 2001; 16( 8): 1520– 1531. Google Scholar CrossRef Search ADS PubMed 33. Wetzsteon RJ, Zemel BS, Shults J, Howard KM, Kibe LW, Leonard MB. Mechanical loads and cortical bone geometry in healthy children and young adults. Bone . 2011; 48( 5): 1103– 1108. Google Scholar CrossRef Search ADS PubMed 34. Gabel L, Macdonald HM, Nettlefold L, McKay HA. Physical activity, sedentary time, and bone strength from childhood to early adulthood: a mixed longitudinal HR-pQCT study. J Bone Miner Res . 2017; 32( 7): 1525– 1536. Google Scholar CrossRef Search ADS PubMed 35. Brotto M, Bonewald L. Bone and muscle: interactions beyond mechanical. Bone . 2015; 80: 109– 114. Google Scholar CrossRef Search ADS PubMed 36. Lecka-Czernik B, Rosen CJ. Skeletal integration of energy homeostasis: translational implications. Bone . 2016; 82: 35– 41. Google Scholar CrossRef Search ADS PubMed 37. Lecka-Czernik B, Stechschulte LA, Czernik PJ, Dowling AR. High bone mass in adult mice with diet-induced obesity results from a combination of initial increase in bone mass followed by attenuation in bone formation: implications for high bone mass and decreased bone quality in obesity. Mol Cell Endocrinol . 2015; 410: 35– 41. Google Scholar CrossRef Search ADS PubMed 38. Kalkwarf HJ, Gilsanz V, Lappe JM, Oberfield S, Shepherd JA, Hangartner TN, Huang X, Frederick MM, Winer KK, Zemel BS. Tracking of bone mass and density during childhood and adolescence. J Clin Endocrinol Metab . 2010; 95( 4): 1690– 1698. Google Scholar CrossRef Search ADS PubMed 39. Loro ML, Sayre J, Roe TF, Goran MI, Kaufman FR, Gilsanz V. Early identification of children predisposed to low peak bone mass and osteoporosis later in life. J Clin Endocrinol Metab . 2000; 85( 10): 3908– 3918. Google Scholar PubMed 40. Wang Q, Cheng S, Alén M, Seeman E; Finnish Calex Study Group. Bone’s structural diversity in adult females is established before puberty. J Clin Endocrinol Metab . 2009; 94( 5): 1555– 1561. Google Scholar CrossRef Search ADS PubMed 41. Akhter MP, Lappe JM, Davies KM, Recker RR. Transmenopausal changes in the trabecular bone structure. Bone . 2007; 41( 1): 111– 116. Google Scholar CrossRef Search ADS PubMed 42. Vogel M, Hahn M, Delling G. Relation between 2- and 3-dimensional architecture of trabecular bone in the human spine. Bone . 1993; 14( 3): 199– 203. Google Scholar CrossRef Search ADS PubMed 43. Calfee RP, Sutter M, Steffen JA, Goldfarb CA. Skeletal and chronological ages in American adolescents: current findings in skeletal maturation. J Child Orthop . 2010; 4( 5): 467– 470. Google Scholar CrossRef Search ADS PubMed 44. Seeman E, Ghasem-Zadeh A. Challenges in the acquisition and analysis of bone microstructure during growth. J Bone Miner Res . 2016; 31( 12): 2239– 2241. Google Scholar CrossRef Search ADS PubMed Copyright © 2018 Endocrine Society
Journal of Clinical Endocrinology and Metabolism – Oxford University Press
Published: Jan 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera