SEARCH

SEARCH BY CITATION

Keywords:

  • AGING;
  • POPULATION STUDIES;
  • OSTEOPOROSIS

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Previous studies using dual-energy X-ray absorptiometry (DXA) have demonstrated that age is a major predictor of bone fragility and fracture risk independent of areal bone mineral density (aBMD). Although this aBMD-independent effect of age has been attributed to poor bone “quality,” the structural basis for this remains unclear. Because high-resolution peripheral quantitative computed tomography (HRpQCT) can assess bone microarchitecture, we matched younger and older subjects for aBMD at the ultradistal radius and assessed for possible differences in trabecular or cortical microstructure by HRpQCT. From an age-stratified, random sample of community adults, 44 women aged <50 years (mean age 41.0 years) were matched to 44 women aged ≥50 years (mean age 62.7 years) by ultradistal radius aBMD (mean ± SEM, younger and older aBMD 0.475 ± 0.011 and 0.472 ± 0.011 g/cm2, respectively), and 57 men aged <50 years (mean age 41.3 years) were matched to 57 men aged ≥50 years (mean age 68.1 years; younger and older aBMD both 0.571 ± 0.008 g/cm2). In these matched subjects, there were no sex-specific differences in trabecular microstructural parameters. However, significant differences were noted in cortical microstructure (all p < 0.05): Older women and men had increased cortical porosity (by 91% and 56%, respectively), total cortical pore volume (by 77% and 61%, respectively), and mean cortical pore diameter (by 9% and 8%, respectively) compared with younger subjects. These findings indicate that younger and older women and men matched for DXA aBMD have similar trabecular microarchitecture but clearly different cortical microstructure, at least at an appendicular site represented by the radius. Further studies are needed to define the extent to which this deterioration in cortical microstructure contributes to the aBMD-independent effect of age on bone fragility and fracture risk at the distal radius and other sites of osteoporotic fractures. © 2012 American Society for Bone and Mineral Research


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Although areal bone mineral density (aBMD) is an important predictor of subsequent fracture risk, age itself is a major factor determinant of fracture risk, independent of aBMD.1 Indeed, based on the known relationship between aBMD and fracture risk, hip fracture risk would be predicted to increase 4-fold between the ages of 50 and 80 years. In fact, the risk of hip fracture increases 30-fold over this age range, indicating that over the course of life, changes in age are approximately 7-fold more important than changes in aBMD.2 As such, age is a key component of the World Health Organization's fracture prediction algorithm, FRAX.2

Although the effects of age on fracture risk could be owing to a number of factors, including the increased risk of falling,3 it has been suggested that with aging there is a deterioration in bone “quality” that is not captured by aBMD.4 These age-related changes in bone include microstructural deterioration, such as trabecular perforation, thinning, and loss of connectivity, as well as cortical thinning and increased porosity,5 along with possible changes in bone material properties, such as the composition and degree of collagen cross-linking.4 Although techniques to measure bone material properties noninvasively may be validated and available in the future,6 the advent of high-resolution peripheral quantitative computed tomography (HRpQCT) now allows for the assessment of bone microstructure, at least at the appendicular sites of the radius and tibia. HRpQCT has a voxel size of 82 µm, can define trabecular and cortical microstructure separately, and shows excellent correlation with ex vivo µCT imaging (resolution of 20 µm or better).7, 8 We9 and others10, 11 have recently used this technique in adults to obtain an in vivo assessment of bone microarchitecture across life in women and men, making it possible to obtain “noninvasive bone biopsy” data in population studies.9

In the current investigation, we sought to more clearly define the effects of age on bone microarchitecture by matching younger (aged <50 years) and older (aged ≥50 years) women and men by dual-energy X-ray absorptiometry (DXA) aBMD at the ultradistal (UD) radius and then examining trabecular and cortical bone microstructural parameters by HRpQCT at the same site in these matched subjects. This study thus addressed the clinically important question of how, on average, bone microstructure might differ in a younger and older woman or man with identical aBMD values.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Study subjects

We recruited subjects from an age-stratified, random sample of Rochester, MN, USA, residents, as previously described.12 This population is highly characteristic of the US white population, but blacks and Asians are underrepresented.13 From this sample, we matched 44 women aged <50 years with 44 women aged ≥50 years as well as 57 men aged <50 years with 57 men aged ≥50 years by UD radius aBMD by DXA (see Statistical Analyses for details regarding the matching procedure). None of the subjects were on an osteoporosis treatment drug (bisphosphonates, teriparatide, estrogen) at the time of study. All studies were approved by the Mayo Institutional Review Board, and written, informed consent was obtained from all subjects before evaluation.

DXA measurements

Areal bone mineral density measurements of the UD radius were made with the Lunar Prodigy (GE Medical Systems, Waukesha, WI, USA). The site of measurement for the aBMD measurement at the UD radius was set according to the standard forearm scan protocol for the Prodigy system. For this, the reference line was autoplaced on the styloid process of the ulna. The radiation dose for the UD DXA scan was <10 mrad.

HRpQCT measurements

Details regarding the HRpQCT imaging used in this cohort have previously been reported9 and are summarized briefly here. Measurements were obtained from the nondominant wrist and tibia on all subjects using the Xtreme CT (Scanco Medical AG, Brüttisellen, Switzerland). At the radius, a scout view was used to set a reference line at the intersection of the joint space with the radio-ulnar junction. The measurement protocol then acquired a 3D stack of 110 slices (9.02 mm) starting 9.5 mm from the reference line, with an isotropic voxel size of 82 µm. Similar measurements are made at the distal tibia, although the scan started 22.5 mm from the reference line. The scans were graded as “good,” “acceptable,” or “not acceptable” using the standard operating procedures provided by the manufacturer. All of the scans used in the study were either “good” or “acceptable.” The radiation dose for the HRpQCT scans was 65 mrad. Figure 1 provides the sites at the forearm scanned by the Xtreme CT as well as the comparable region included in the UD radius DXA scans.

thumbnail image

Figure 1. Relative scanning sites for the Xtreme CT (red lines) and UD DXA scans (orange box).

Download figure to PowerPoint

Trabecular parameters

Bone volume/total volume (BV/TV, %) was derived from trabecular volumetric bone mineral density (vBMD), assuming a mineral density of fully mineralized bone of 1.2 g hydroxyapatite/cm3. Recognizing that individual trabeculae would not be resolved at their correct thickness (∼100 µm) because of partial volume effects, a thickness-independent structure extraction was employed to identify three-dimensional ridges (center points of the trabeculae);14 trabecular number (TbN, mm−1) was then taken as the inverse of the mean spacing of the ridges.15 Analogous with standard histomorphometry,16 trabecular thickness (TbTh, µm) was calculated using the formula TbTh = BV/TV ÷ TbN, and trabecular spacing (TbSp, µm) was calculated as TbSp = (1–BV/TV) ÷ TbN. Validation studies show excellent correlation (r ≥ 0.96) for these parameters compared with the gold-standard ex vivo µCT technique.7 Additional trabecular parameters, including the SD of the trabecular spacing (Tb 1/N SD, trabecular heterogeneity), connectivity density (Conn Dens), and structure model index (SMI), which indicates whether trabeculae are more platelike (lower values) or more rodlike (higher values), were derived from these parameters using the manufacturer's protocol.

Cortical parameters

The cortex was segmented from the grayscale image with a Gaussian filter and threshold.15 Details regarding delineation of the endocortical surface have previously been provided.17 Briefly, the approach uses morphological operators (dilation erosion) and component labeling to find the endocortical contour. We used the extended cortical analysis available from the manufacturer to obtain cortical vBMD, cortical tissue mineral density (Ct.TMD, which is the density of the bone material excluding the cortical pores), and cortical thickness (CtTh).

For the cortical microstructural parameters, we used the manufacturer's software, which uses an approach published recently by several groups10, 18 and described in detail by Burghardt and colleagues.17 This analysis provides the following data: cortical pore volume (CtPoV), a direct voxel-based measure of the volume of the intracortical pore space; cortical porosity (CtPo), a relative voxel-based measure of the volume of the intracortical pore space normalized by the sum of the pore and cortical bone volume; cortical pore diameter (CtPoDm), the mean 3D diameter of the intracortical pore space; and cortical pore diameter distribution (CtPoDmSD), the standard deviation of the 3D diameters of the intracortical pore space. The specific technical approach and validation of the methods for obtaining these cortical parameters have been described by Burghardt and colleagues.17 Briefly, the outer and inner contours of the cortex are first detected. Within the cortical compartment, the CtPoV is calculated and CtPo is then that volume normalized to the total cortical compartment volume. Using direct distance transformation methods,19 the pore diameter of each pore is calculated and their average (CtPoDm) and standard deviation (CtPoDmSD) determined. Similar to trabecular heterogeneity (Tb 1/N SD), where greater heterogeneity is associated with an increase in fracture risk, a larger CtPoDmSD indicates greater disruption of cortical microstructure.20, 21

Statistical analyses

The younger group was defined as subjects aged <50 years old. Subjects in the older group were aged ≥50 years and chosen by using an optimal matching algorithm.22, 23 Briefly, from the total cohort of 328 subjects, there were 44 younger women and 57 younger men, and we looked for an optimal match among the 111 older women and 116 older men, respectively. In doing so, we did not specify percentage limits for the matching but selected older subjects with the UD radius aBMD value closest to each younger subject. As such, we used all of the younger subjects available (because we had smaller numbers of these) and 44 of the available older women and 57 of the available older men. All 57 of the male pairs had matched UD radius aBMD within 0.016 mg/cm2 (approximately 3%) and 41 of the 44 female pairs matched within 0.02 mg/cm2 (approximately 4%). The other 3 female pairs had a UD radius aBMD difference of 0.023, 0.039, and 0.056 mg/cm2. Comparisons between groups were made using two-sample, unpaired t tests. Linear regression models were used to compare the groups after adjusting for height and weight. A p value <0.05 was considered significant, and all data are presented as mean ± SEM.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Table 1 shows the clinical characteristics and aBMD at the UD radius of the younger and older women and men matched for aBMD by DXA. As is evident, our matching procedure resulted in the subjects having similar aBMD values. Of note, the younger women and men had average radius aBMD Z-scores within 1 SD of 0; however, the older subjects were 1.2 SD (women) and 1.6 SD (men) above the mean. On average, the older women were 21.7 years and the older men 26.8 years older than the younger women and men, respectively. There was a trend for the older women to be shorter and have higher weight and body mass index (BMI) compared with the younger women. Only 7% of the younger women were postmenopausal compared with 82% of the older women. The younger men were significantly taller than the older men, but BMI and body weight were not significantly different in both groups.

Table 1. Clinical Characteristics and aBMD by DXA at the Radius of the Younger and Older Women and Men Matched for UD Radius aBMD
 YoungerOlderp Value
  1. aBMD = areal bone mineral density; DXA = dual-energy X-ray absorptiometry; UD = ultradistal; BMI = body mass index.

  2. Data are mean ± SEM (min, max).

Women
 No.4444
 Age (years)41.0 ± 0.9262.7 ± 1.60<0.001
 UD radius aBMD, g/cm20.475 ± 0.0110.472 ± 0.0110.833
 UD radius aBMD, Z-score0.1 ± 0.3 (–2.7, 3.8)1.2 ± 0.2 (–2.2, 4.0)0.002
 Height (m)1.65 ± 0.011.62 ± 0.010.051
 Weight (kg)75.5 ± 2.9182.7 ± 2.080.046
 BMI (kg/m2)27.9 ± 1.1031.5 ± 0.760.009
 Postmenopausal, No. (%)3 (7%)36 (82%)<0.001
Men
 No.5757
 Age (years)41.3 ± 0.7868.1 ± 1.51<0.001
 UD radius aBMD, g/cm20.571 ± 0.0080.571 ± 0.0080.978
 UD radius aBMD, Z-score0.9 ± 0.2 (–2.1, 3.0)1.6 ± 0.2 (–0.7, 6.2)<0.001
 Height (m)1.81 ± 0.011.76 ± 0.01<0.001
 Weight (kg)96.9 ± 2.3091.6 ± 2.260.104
 BMI (kg/m2)29.6 ± 0.6229.5 ± 0.650.927

As shown in Table 2, matching for UD radius aBMD resulted in the older women and men having virtually identical trabecular microstructural parameters (BV/TV, TbN, TbTh, TbSp, trabecular heterogeneity, and Conn Dens) compared with the younger women and men, respectively, as assessed by HRpQCT at the same site. The only exception was SMI, which was slightly but significantly higher (by 8%) in the older compared with the younger men, although this difference was no longer statistically significant after adjusting for height and weight.

Table 2. Trabecular Bone Microstructural Parameters at the Radius of the Younger and Older Women and Men Matched for UD Radius aBMD
 YoungerOlderp Value (unadjusted)p Value (adjusted)
  1. UD = ultradistal; aBMD = areal bone mineral density; BV/TV = bone volume/total volume; TbN = trabecular number; TbTh = trabecular thickness; TbSp = trabecular separation; Tb 1/N SD = SD of the trabecular spacing, trabecular heterogeneity; Conn Dens = connectivity density; SMI = structure model index.

  2. Data are mean ± SEM. The p values are unadjusted and adjusted for height and weight.

Women
 BV/TV0.129 ± 0.0050.129 ± 0.0050.9770.576
 TbN, 1/mm1.74 ± 0.051.78 ± 0.050.4970.800
 TbTh, mm0.074 ± 0.0020.073 ± 0.0020.7010.363
 TbSp, mm0.525 ± 0.0220.514 ± 0.0250.7520.984
 Tb 1/N SD, mm0.232 ± 0.0150.240 ± 0.0240.7750.714
 Conn Dens, 1/mm33.34 ± 0.173.46 ± 0.160.6140.934
 SMI2.15 ± 0.072.25 ± 0.060.3180.146
Men
 BV/TV0.164 ± 0.0040.159 ± 0.0030.3110.249
 TbN, 1/mm2.06 ± 0.032.05 ± 0.030.6870.676
 TbTh, mm0.080 ± 0.0010.078 ± 0.0010.3720.256
 TbSp, mm0.412 ± 0.0090.418 ± 0.0100.6320.595
 Tb 1/N SD, mm0.172 ± 0.0050.179 ± 0.0080.4810.441
 Conn Dens, 1/mm34.78 ± 0.134.67 ± 0.120.5380.542
 SMI1.70 ± 0.051.83 ± 0.040.0430.063

Table 3 shows the cortical parameters (cortical vBMD, cortical tissue mineral density, CtTh, periosteal and endosteal circumference) in the study subjects. Cortical vBMD and cortical tissue mineral density but none of the other cortical parameters were significantly lower (by 3% and 2%, respectively) in the older compared with the younger women matched for aBMD. In the men, cortical vBMD was similar in the younger and older men, whereas cortical tissue mineral density was slightly higher in the older men, although this did not remain significant after adjusting for effects of height and weight. The other cortical parameters were similar in the younger and older men, except for periosteal circumference, which was slightly lower in the older men after adjustment for height and weight.

Table 3. Cortical Bone Parameters at the Radius of the Younger and Older Women and Men Matched for UD Radius aBMD
 YoungerOlderp Value (unadjusted)p Value (adjusted)
  1. UD = ultradistal; aBMD = areal bone mineral density; vBMD = volumetric bone mineral density.

  2. Data are mean ± SEM. The p values are unadjusted and adjusted for height and weight.

Women
 Cortical vBMD (mg/cm3)951 ± 4.80919 ± 9.860.0060.003
 Cortical tissue mineral density (mg/cm3)1046 ± 41026 ± 70.0110.010
 Cortical thickness (mm)0.974 ± 0.0190.964 ± 0.0320.7880.352
 Periosteal circumference (mm)62.9 ± 0.84563.3 ± 0.7860.6990.216
 Endocortical circumference (mm)44.8 ± 0.72444.7 ± 0.7910.8760.473
Men
 Cortical vBMD (mg/cm3)881 ± 4.97880 ± 7.300.9220.876
 Cortical tissue mineral density (mg/cm3)980 ± 4993 ± 50.0400.106
 Cortical thickness (mm)0.992 ± 0.0211.017 ± 0.0280.4700.654
 Periosteal circumference (mm)80.7 ± 0.78780.6 ± 0.9100.9410.033
 Endocortical circumference (mm)57.6 ± 0.73756.5 ± 0.8410.3400.286

In contrast to these minimal differences in standard trabecular or cortical parameters, CtPo (Fig. 2A) was significantly higher in the older women and men matched for aBMD with the younger subjects (by 91% and 56% in the women and men, respectively). This was accompanied by significant increases in the older subjects in the CtPoV (of 77% and 61% in the women and men, respectively; Fig. 2B), in the CtPoDm (of 9% and 8% in the women and men, respectively; Fig. 2C), and in the CtPoDmSD (of 17% and 13% in the women and men, respectively; Fig. 2D).

thumbnail image

Figure 2. (A) Cortical porosity, (B) cortical pore volume, (C) cortical pore diameter, and (D) cortical pore diameter SD at the distal radius in younger and older women and men matched for UD radius aBMD by DXA. Open bars, younger subjects; hatched bars, older subjects. The p values are unadjusted (first value) and adjusted for height and weight (second value).

Download figure to PowerPoint

Figure 3 shows representative cross-sectional images of HRpQCT sections from younger versus older women and men matched for aBMD, visually demonstrating the increase in cortical porosity in the older subjects. Thus, the younger and older women in this example (top panels) had UD radius aBMD values within 3% of each other, yet cortical porosity was ∼3.5-fold increased in the older subject. Similarly, the younger and older men (bottom panels) had UD radius aBMD values within 3% of each other; again, cortical porosity was increased ∼2.7-fold in the older subject.

thumbnail image

Figure 3. Representative cross-sectional images from HRpQCT scans from younger and older women (top panels) and younger and older men (bottom panels) with similar UD radius aBMD values (within 3% of each other) but marked differences in cortical porosity.

Download figure to PowerPoint

Although we did not have analogous aBMD measures at the tibia, we did have bone structural variables at the tibia from the HRpQCT. As shown in Table 4, matching younger and older subjects for UD radius aBMD resulted in these subjects having virtually identical values for BV/TV not only at the radius but also at the tibia. Again, the trabecular parameters did not differ between the groups. Cortical vBMD was significantly lower in the older women and men compared with the younger subjects (by 8% and 2%, respectively). Similar to the findings at the radius, cortical tissue mineral density was significantly lower in the older compared with the younger women but not men. In addition, the younger women had significantly higher CtTh and lower periosteal circumference than the older women. However, as for the radius, the major differences between the younger and older subjects were in the cortical porosity parameters, with an increase in CtPo of 92% and 28% and in CtPoV of 78% and 30%, in the women and men, respectively (Table 4). CtPoDm and CtPoDmSD were similar in younger and older women as well as in the younger and older men.

Table 4. Trabecular and Cortical Bone Microstructural Parameters at the Tibia of the Younger and Older Women and Men Matched for UD Radius aBMD
 YoungOldp Value (unadjusted)p Value (adjusted)
  1. UD = ultradistal; aBMD = areal bone mineral density; BV/TV = bone volume/total volume; TbN = trabecular number; TbTh = trabecular thickness; TbSp = trabecular separation; Tb 1/N SD = SD of the trabecular spacing, trabecular heterogeneity; Conn Dens = connectivity density; SMI = structure model index; vBMD = volumetric bone mineral density; CtTh = cortical thickness; CtPo = cortical porosity; CtPoV = cortical pore volume; CtPoDm = cortical pore diameter; CtPoDmSD = cortical pore diameter distribution.

  2. Data are mean ± SEM. The p values are unadjusted and adjusted for height and weight.

Women
 Trabecular parameters
  BV/TV0.136 ± 0.0050.146 ± 0.0040.1040.327
  TbN (1/mm)1.75 ± 0.051.85 ± 0.040.1270.441
  TbTh (mm)0.077 ± 0.0020.079 ± 0.0020.4580.493
  TbSp (mm)0.524 ± 0.0280.470 ± 0.0110.0790.198
  Tb 1/N SD (mm)0.249 ± 0.0200.211 ± 0.0070.0800.143
  Conn Dens (1/mm3)3.76 ± 0.184.07 ± 0.150.1970.762
  SMI1.82 ± 0.061.74 ± 0.050.2990.612
 Cortical parameters
  Cortical vBMD (mg/cm3)905 ± 5.72828 ± 11.3<0.001<0.001
  Cortical tissue mineral density (mg/cm3)1009 ± 4963 ± 7<0.001<0.001
  CtTh (mm)1.28 ± 0.031.23 ± 0.040.3200.071
  Periosteal circumference (mm)101 ± 0.95104 ± 1.260.0400.003
  Endosteal circumference (mm)82 ± 0.8484 ± 1.260.1000.010
  CtPo (%)3.13 ± 0.156.00 ± 0.39<0.001<0.001
  CtPoV (mm3)32.8 ± 1.7658.5 ± 3.32<0.001<0.001
  CtPoDm (µm)193 ± 3.30200 ± 3.080.1560.196
  CtPoDmSD (µm)87.9 ± 2.5292.8 ± 2.200.1460.200
Men
 Trabecular parameters
  BV/TV0.171 ± 0.0040.168 ± 0.0040.5270.685
  TbN (1/mm)2.19 ± 0.042.11 ± 0.050.1960.497
  TbTh (mm)0.078 ± 0.0010.080 ± 0.0020.3870.676
  TbSp (mm)0.385 ± 0.0080.407 ± 0.0120.1280.304
  Tb 1/N SD (mm)0.171 ± 0.0050.187 ± 0.0090.1220.280
  Conn Dens (1/mm3)5.36 ± 0.165.13 ± 0.190.3480.650
  SMI1.49 ± 0.051.57 ± 0.050.2370.349
 Cortical parameters
  Cortical vBMD (mg/cm3)855 ± 5.34837 ± 6.290.0350.039
  Cortical tissue mineral density (mg/cm3)967 ± 4964 ± 40.6280.518
  CtTh (mm)1.37 ± 0.031.44 ± 0.040.1560.707
  Periosteal circumference (mm)120 ± 1.22120 ± 1.230.9740.002
  Endosteal circumference (mm)97 ± 1.1095 ± 1.210.3600.049
  CtPo (%)5.22 ± 0.246.68 ± 0.29<0.001<0.001
  CtPoV (mm3)72.5 ± 4.3194.1 ± 4.740.001<0.001
  CtPoDm (µm)187 ± 2.21193 ± 2.900.1010.179
  CtPoDmSD (µm)86.3 ± 1.6089.2 ± 2.140.2770.415

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

In this study of younger and older subjects matched for aBMD by DXA at the UD radius, we found that the older women and men had significantly worse measures of cortical microstructure (CtPo, CtPoV, CtPoDm, and CtPoDmSD) but generally similar measures of trabecular microstructure compared with younger subjects. These findings thus demonstrate that at least for appendicular sites represented by the radius and tibia, the major effect of age independent of aBMD is on cortical parameters.

It is important to note that these data do not imply that trabecular parameters do not deteriorate with age. Indeed, we9 and others10, 11 have previously demonstrated significant age-related decreases in trabecular BV/TV at the radius and tibia. Thus, in the larger cohort from which we selected the subjects used in this study, we found that cross-sectional decreases in BV/TV were 27% and 26% in women and men, respectively, between ages 20 and 90 years.9 However, these changes in BV/TV appear to be reflected in the age-related changes in DXA aBMD: When we match younger and older subjects by aBMD at the radius, they have similar values for BV/TV as well as for the other trabecular parameters (TbN, TbTh, and TbSp). Rather, our data indicate that the key parameters that are different between younger and older subjects with the same aBMD and that are not captured by DXA are those related to cortical porosity.

Although our data are novel in highlighting the importance of cortical porosity as perhaps the key feature related to bone “quality” that deteriorates with age independent of aBMD, these findings are consistent with recent work on changes in cortical porosity with aging. Thus, Burghardt and colleagues10 found that in 151 subjects (57 men, 94 women) between the ages of 20 to 78 years, CtPo increased significantly with age; in fact, porosity-related measures provided the only significant decade-wise discrimination in the radius for women in their fifties versus those in their sixties. Consistent with these data, Macdonald and colleagues11 observed marked increases in CtPo over life in women (of 176%) and men (of 84%) accompanied by decreases in trabecular BV/TV, as noted in our previous study.9 Similar findings have been reported by Zebaze and colleagues.5 Our observed age-related alterations in cortical bone at the radius and tibia are also consistent with previous work showing deficits in cortical structure with aging at the hip as an important basis for fracture pathogenesis.24–27 Collectively, these data indicate that cortical porosity is an important component of bone quality that deteriorates with age.

In a study design somewhat similar to ours, Kazakia and colleagues28 matched subjects aged 20 to 78 years for simulated aBMD calculated from HRpQCT data and found that CtPo was one of three parameters that reflected the greatest variation within aBMD subgroups. However, these investigators did not specifically address the question of age-related changes in bone microstructure independent of aBMD. Nonetheless, it is of interest that CtPo emerged as an important variable in that analysis as well as ours. Kazakia and colleagues28 also identified vBMD in the central portion of the trabecular compartment as well as trabecular heterogeneity (greater heterogeneity is associated with an increase in fracture risk20, 21) as showing significant variability within aBMD subgroups. We did not specifically assess trabecular vBMD in the central compartment of the radius or tibia scans, but we did find a modest increase in trabecular heterogeneity in the tibia in older men matched with younger men for radius aBMD.

In addition to the cortical porosity variables, cortical vBMD was reduced at the radius and tibia in the older compared with the younger women and at the tibia in the older men. Cortical tissue mineral density, which is the density of the bone material excluding the pores) was reduced at the radius and tibia in the older compared with the younger women but not in the men at either site. Although cortical thickness was not consistently different in the younger versus older subjects, this may, in part, be because of the limited resolution even of the HRpQCT device.

We recognize potential additional limitations of our study. First, we were only able to assess bone microstructure using HRpQCT at the radius and tibia and not at the spine or hip. Thus, although we found that younger and older subjects matched for radius aBMD by DXA had similar values for trabecular parameters, this may not be true for trabecular microarchitecture at the spine or hip in younger and older subjects with equivalent aBMD at those sites. As such, although our data at the appendicular sites highlight the importance of aBMD-independent changes in cortical microstructure with aging, our findings do not exclude the possibility of additional aBMD-independent changes in trabecular microstructure at these central sites, which we did not assess. Second, we recognize that although HRpQCT provides better resolution for imaging bone microstructure than previously available, assessment of cortical porosity using this technique is likely providing an approximation or “index” of cortical porosity rather than true cortical porosity. Third, by virtue of our matching procedure, we inevitably selected for older subjects who had average radius aBMD values somewhat above the mean (average Z-scores of 1.2 and 1.6 for the older women and men, respectively). These limitations notwithstanding, our data do provide evidence that of all of the available microstructural parameters by HRpQCT, matching for DXA aBMD leads to significant differences principally in indices of cortical porosity. Further studies are needed to define the extent to which this deterioration in cortical microstructure contributes to the aBMD-independent effect of age on bone fragility and fracture risk at the radius as well as other sites of osteoporotic fractures, such as the spine and hip.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

This work was supported by National Institutes of Health Grants AR027065 and UL1-RR24150 (Center for Translational Science Activities).

We thank Mr James Peterson for help with manuscript preparation.

Authors' roles: Study design: KMN, EJA, SK. Study conduct: KMN, SK. Data collection: KMN, EJA. Data analysis: KMN, EJA, SK. Data interpretation: KMN, SA, EJA, BLR, LJM, SK. Drafting manuscript: KMN, SK. Revising manuscript content: SK. Approving final version of manuscript: KMN, SA, EJA, BLR, LJM, SK. SK takes responsibility for the integrity of the data analysis.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References