Age-Related mineralization heterogeneity changes in trabecular bone of the proximal femur†
Version of Record online: 6 AUG 2004
Copyright © 2004 Wiley-Liss, Inc.
The Anatomical Record Part A: Discoveries in Molecular, Cellular, and Evolutionary Biology
Volume 281A, Issue 2, pages 1296–1302, December 2004
How to Cite
Bloebaum, R. D., Lundeen, G. A., Shea, J. E. and Whitaker, E. L. (2004), Age-Related mineralization heterogeneity changes in trabecular bone of the proximal femur. Anat. Rec., 281A: 1296–1302. doi: 10.1002/ar.a.20079
This article is a US government work and, as such, is in the public domain in the United States of America.
- Issue online: 22 NOV 2004
- Version of Record online: 6 AUG 2004
- Manuscript Accepted: 1 APR 2004
- Manuscript Received: 19 DEC 2003
- Office of Research and Development, Medical Research Service, Department of Veterans Affairs, Salt Lake City Health Care System, Salt Lake City, Utah
- trabecular bone;
- proximal femur;
- backscattered electron imaging technology;
Although there is extensive documentation in the literature regarding the importance of trabecular bone for proximal femoral integrity and fracture resistance, there remain gaps in our understanding of the basic mineral changes that may occur in trabecular bone attributable to aging. It is unclear what age-related changes take place in the trabecular bone of the proximal femur, a common fracture site in the elderly. It has been suggested that some explanation for conflicting reports on cancellous bone may be found at a microscopic level. The goal of this study was to document age-related changes in micromineralization in the proximal femur of Caucasian females using backscattered electron imaging technology. Proximal femurs were obtained from 11 young and 11 elderly females. Sections of bone from the superior and inferior neck and superior and inferior trochanter were analyzed in a scanning electron microscope using the backscatter technique to determine ash percent. Mean ash percent did not change with age in any of the four regions (P > 0.05). However, while the mean ash percent did not change, there was a dramatic increase in variability elderly age group and loss of mineral heterogeneity. This indicates that there are subpopulations with higher or lower ash percents than the mean in the elderly study group in this investigation. While variance changed dramatically, variance within individuals did not change significantly with age (P > 0.05). The results of this study suggest that changes in micromineralization may occur within an individual, adding a possible new dimension to our understanding of fracture risk in the elderly. Future studies should examine a longer population base to confirm this observation. Published 2004 Wiley-Liss, Inc.
One of the most important determinants of the strength and fracture resistance of the proximal femur is the structure and quality of the trabecular bone. Trabecular bone has been repeatedly shown to contribute significantly to the strength and shock absorption properties of bone (Weaver and Chalmers, 1966; Leichter et al., 1982; Werner et al., 1988; McCalden et al., 1997). Lotz et al. (1995) showed trabecular bone to carry as much as 70% of the load during gait. Trabecular bone loss is a commonly recognized problem with age (Fig. 1) and a significant factor for the fractures that occur in the elderly (Kyle et al., 1994). In addition to bone loss, bone quality is important in understanding trabecular bone structural integrity. Mineralization (ash percent or ash fraction) has been shown to affect the mechanical properties of bone, including compressive strength (Weaver and Chalmers, 1966; McCalden et al., 1997) and elasticity or stiffness (Martens et al., 1983; Currey, 1988). Even a minor drop in ash percent can cause a large decrease in fracture resistance (Leichter et al., 1982; Mosekilde et al., 1987) and modulus of elasticity (Currey, 1979).
Although there is extensive documentation in the literature regarding the importance of trabecular bone for femoral integrity and fracture resistance, there remain gaps in our understanding of the basic mineral changes that occur in aging trabecular bone. There are conflicting reports in the literature as to whether ash percent of trabecular bone changes with age. Mosekilde et al. (1987) demonstrated that ash percent of vertebral trabecular bone decreases 48–50% between 20 and 80 years. However, despite documented changes in bulk ash, researchers have often been unable to identify the changes they expect or explain those they find. Parfitt (1993) anticipated finding highly mineralized trabecular bone in elderly patients due to decreased remodeling. Instead, the trabecular bone had lower levels of mineralization in the older population. Two other studies, however, found that ash did not change with age (Smith et al., 1975; Martens et al., 1983). One used a bone mineral analyzer on whole bone in the radius (Smith et al., 1975), and the other used bulk ash fraction measurements in the proximal femur (Marten et al., 1983).
There are several possible explanations for these conflicting reports. One explanation may lie in anatomical site-specific changes. The studies cited have examined trabecular bone from many different anatomical sites that may not be comparable. Mineralization can vary greatly even in a cross-section of the same bone due to differing mechanical loads, such as tension and compression regions (Skedros et al., 1994). It has also been suggested that further explanation for these discrepancies in the literature may be found at a microscopic level. Boyce and Bloebaum (1993) suggested that fracture toughness of cortical bone is directly related to mineral content at a microscopic level based on the age-related changes they found. There is also interest in the properties of trabecular bone at a microstructural level to help explain the decrease in strength with age (Goldstein, 1987; Mosekilde et al., 1987). Trabecular bone is very heterogeneous, and there is evidence that the heterogeneity of individual trabeculae can affect the properties of the whole bone (Keaveny and Hayes, 1993). In any sample of bone, there are regional variations in mineral content due to varied rates of calcified tissue turnover and remodeling (Grynpas and Holmyard, 1988; Grynpas, 1993). It is therefore the goal of this study to provide data that will help clarify microscopic mineral content variation in the aging process of trabecular bone in the Caucasian female proximal femur. Our objective is to analyze the trabecular micromineralization of the proximal femurs from young and elderly subjects using backscattered electron (BSE) microscopy to determine the age-related changes in mineralization. The hypothesis being examined is that mineral heterogeneity will be altered with aging.
MATERIALS AND METHODS
In order to determine age-related changes in micromineralization, BSE analysis will be used. BSE provides a black-and-white image that is brighter in areas with a higher average atomic number and darker in regions of lower average atomic number (Bloebaum et al., 1990; Skedros et al., 1993a). It is similar to the image obtained from an X-ray or microradiography, but as the backscattered electrons only penetrate a few microns into the surface of the specimen, projection effect is minimal (Bachus and Bloebaum, 1992). Clarity and resolution of the image, therefore, are much better than when other techniques are utilized. The BSE images were divided into 256 gray-level bins to generate a gray-level histogram. There are numerous ways to analyze a histogram statistically, but not all of these methods have been determined to have biological meaning. The shape of the curve can be affected by topography, probe current, scan rate, and other factors, particularly at the margins of the histogram (Howell and Boyde, 1994; Vajda and Skedros, 1999; Vajda et al., 1999). Despite these variations, the weighted mean gray-level (WMGL) of a histogram has shown a strong linear correlation to the percent ash of bone (Bloebaum et al., 1990, 1997; Roschger et al., 1998). Ash percent in turn correlates with mechanical properties of bone (Currey et al., 1979). Because of this linear correlation, WMGL from a BSE image can be converted to ash percent using BSE images of mineral standards of known ash percent (Skedros et al., 1993a, 1993b). It is also valuable to compare the width of the histograms within a single study using full width at half maximum (FWHM) (Vajda et al., 1995; Vajda and Skedros, 1999). Therefore, WMGL and FWHM will be measured from the gray-level histograms obtained in this study.
Specimen Selection and Preparation
Twenty-two human donor femora were obtained from Caucasian female cadavers in an institutional review board-approved study of two age groups: young (premenopausal; n = 11; range, 17–35 years; mean, 28 ± 6) and elderly (n = 11; range, 76–95 years; mean, 85 ± 6). Only Caucasian females were used to eliminate possible complications of interracial or gender variations, and because this represents the group with the highest fracture risk (Kannus et al., 1996; Dubey et al., 1999). Individuals with metabolic bone disease, on pharmaceutical regimens, or with other pathologies that might affect the skeletal system were excluded from this study in accordance with standard bone banking procedures (Bloebaum et al., 1993a).
The specimen selection to compare the superior and inferior regions of the femoral neck and greater trochanter was based on previous investigations that showed the inferior neck region had the smallest decline in volume fraction during aging compared to the superior region (Lundeen et al., 2000). Since the greater trochanter has also been shown to be a high-risk fracture site with aging in the elderly, the superior and inferior regions were chosen for investigation based on the previous report by Lundeen et al. (2000).
Each femur was fixed in 70% ethyl alcohol for a minimum of 2 weeks and then manually cleaned of soft tissue. Using previously described osteometric techniques, the femurs were measured and marked to identify three locations that were comparable between all specimens: the base of the head, the neck-trochanter junction (base of neck), and through the base of the lesser and greater trochanters (Ruff et al., 1983; Bloebaum et al., 1993b; Kuo et al., 1998). The specimens were then cut at the specified locations using a band saw, yielding two sections per femur: a neck and a trochanter section (Fig. 2). Each section was then embedded in methylmethacrylate according to previously published protocols (Sanderson et al., 1990; Sanderson and Kitabayashi, 1994). Specimens were ground and polished and scribed with a stylus, dividing the section into superior and inferior aspects following the principal anatomical directions. Because of the chamber limitations in the SEM, and the requirement to include pure metals and bone mineral calibration standards, the cancellous bone specimen size was restricted to approximately 0.5 cm3. Specimens from both the superior and inferior regions were analyzed during the same imaging session (Fig. 2). The bone segments were then randomly assigned to one of three composite blocks, which were diamond-milled to an optical finish and carbon-coated for quantitative analysis using BSE imaging.
Each block included two mineral standards of known ash percent, a mule deer antler (low ash percent) and whale tympanic bulla (high ash percent) (Bloebaum et al., 1997). For every block, a piece of each mineral standard was cut into three pieces. The middle piece was included in the composite block, and the two outer pieces were ashed to determine the exact mineral content of the standard. Ash measurements were obtained by first defatting the bone segments for 20 days in a large volume of reagent-grade chloroform (Omnisolv, EM Industries, Gibbstown, NJ) under vacuum with constant stirring (Folch et al., 1957). Residual chloroform was removed by placing the specimens in an 80°C oven for 5 days (Skedros et al., 1993b). Following defatting and drying, each bone segment was weighed using a precision analytical balance (Mettler H51, Mettler Instruments, Hightstown, NJ) and then placed in a furnace at 550°C for 24 hr to remove organic constituents (Skedros et al., 1993b). The bone specimens were weighed again after ashing. Ash percent was calculated as 100 times the ratio of the weight of ashed bone (WAB) to the weight of dry defatted bone (WDB), or (WAB/WDB) × 100. The value used is the mean value of the two ashed segments from each bone.
Each carbon-coated composite block was placed inside a JEOL 6100 Scanning Electron Microscope (JEOL USA, Peabody, MA) for analysis with a BSE detector (Tetra, Oxford Instruments, Buckinshire, U.K.). Microscope calibration, imaging, and analysis were performed according to the previously published protocol of Bloebaum et al. (1997). All operating conditions were maintained at constant levels to ensure consistent analysis, including voltage (20 kV), probe current (0.7 ± 0.02 nA), aperture (50 μm), working distance (15 mm), and magnification (50×). Probe current was checked before each image capture and any minor fluctuations were corrected manually by adjusting the condenser lens strength. Two random images were captured from each region (Fig. 2). Following imaging, three analysis areas (30 × 30 pixels) from each image were selected, avoiding edges and central canals. The WMGL was calculated for each analysis area, yielding a total of six WMGLs per region (inferior neck and trochanter, superior neck and trochanter). The following equation was used to calculate the WMGLs:
where Ai is the area of the ith gray-level, GLi is the ith gray-level, and At is the total area imaged (Vajda et al., 1995). To ensure instrument stability, BSE image WMGLs were calibrated at 20-min intervals using pure aluminum and carbon as calibration standards following previously published protocols (Vajda et al., 1995; Bloebaum et al., 1997; Roschger et al., 1998). A mechanized stage was used to ensure that the same area of the standards was imaged throughout the analysis.
In a BSE image, lower WMGL measurements correspond to darker areas, indicating lower mineral content. WMGLs are linearly correlated to ash percent (mineral content independent of porosity) (Bloebaum et al., 1990; Skedros et al., 1993a; Roschger et al., 1998). The two mineral standards of known ash percent were used to create a linear regression line correlating WMGL to ash percent. Using this linear regression, the average ash percent was obtained for each region. The mineral content could then be compared between regions and age groups. In addition to WMGL, a gray-level histogram was obtained from each measured area. Skewness, kurtosis, and FWHM were calculated for each histogram.
Mean ash percent was compared between all ages and regions using two-way ANOVAs to calculate the effect of region and age on ash percent, as well as the interaction between the two factors. Tukey's posthoc test was used to analyze individual differences. Skewness, kurtosis, and FWHM were calculated for each gray-level histogram (six per region) and averaged. Skewness, kurtosis, and FWHM were then analyzed using two-way ANOVAs to test the effect of age and region. Tukey's posthoc test was used to detect individual differences.
Population variance of ash percent was obtained for each region. An F-test for comparing variances between two populations was then used to compare regions and age groups. To analyze the individual variances, the variance was obtained for the six ash percents from each region in each individual. Then the individual variances were averaged and compared by region and age group using the same F-test used for population variance.
Mean ash percent did not significantly change with age in any of the four regions (P > 0.05; Fig. 3). In addition, no statistical difference was found between the anatomical regions analyzed in this investigation (P > 0.05). While there was no change in mean ash percent with age, the population variance was significantly different with age (Fig. 4). In the inferior neck, the variance was 1.1 in the young and 3.4 in the elderly (P < 0.05). The variance in the superior neck was 8.2 in the young and 27.0 in the elderly (P < 0.05). In the superior trochanter, the variance was 1.7 in the young and 16.7 in the elderly (P < 0.05). While the variance increased with age in the three regions listed above, it significantly decreased with age in the inferior trochanter, from 11.2 in the young to 3.0 in the elderly (P < 0.05). Although the population variance changed significantly with age, individual variances did not change (Fig. 5). Table 1 summarizes ash percent means and standard deviations.
|Inferior neck||Inferior trochanter||Superior neck||Superior trochanter|
|Elderly||65.6 ± 1.8||65.8 ± 1.7||66.6 ± 5.2||66.5 ± 4.1|
|Young||64.7 ± 1.0||66.3 ± 3.3||65.6 ± 2.9||64.0 ± 1.3|
Skewness describes an asymmetric tail extending in one direction from the curve. There was a trend toward less negative skewness with age; in the superior neck, it increased from −2.6 ± 0.4 in the young to −2.3 ± 0.4 in the elderly (P = 0.068), with a slight trend in the other three regions. Kurtosis describes the peakedness of the curve and the steepness of the slopes. There was a trend toward decreasing kurtosis with age in the superior and inferior trochanters (13.2 ± 2.9 to 12.0 ± 5.8 and 12.2 ± 3.4 to 10.6 ± 2.3, respectively), and a clearer trend in the inferior neck (P = 0.055), decreasing from 14.3 ± 2.8 to 12.1 ± 2.2. Kurtosis decreased significantly with age in the superior neck, from 15.2 ± 2.0 to 11.6 ± 2.6 (P < 0.01). FWHM increased significantly with age, from 10.9 ± 0.1 to 11.4 ± 0.1 (P = 0.008), but did not change significantly between regions.
Mean ash percent was not found to change significantly with age in this study, which agrees with previous reports for trabecular bone bulk ash and mineral content (Smith et al., 1975; Parfitt, 1993; Lundeen et al., 2000). It is of interest that the ash percents derived from WMGLs in this study are comparable to those found directly by Lundeen et al. (2000) using bulk ash, which confirms the utilization of the described BSE technique to determine mineralization of human bone.
Skewness and kurtosis also were unchanged between the young and the elderly groups, with the exception that kurtosis decreased significantly with age in the superior neck. However, it is unclear what the small changes occurring at the extremes of the gray-level histograms mean. What may appear to be extreme high and low mineralized bone at the edges of the histogram may be largely attributed to topographical artifact and edge effects (Vajda et al., 1999), even if differences in these regions were statistically significant. Therefore, it is more relevant to analyze the changes in WMGL mean and variance and FWHM since these measures reflect the bulk of the histogram rather than the outlying points and are known to correlate with ash percent in human bone.
The most striking observations were the changes in population variance with both age and location within the proximal femur comparing the two groups. While the mean ash percent did not change, there was a dramatic increase in variability between individuals within the elderly age group. Again, these microscopic results mirror the changes reported in whole trabecular bone (Lundeen et al., 2000). This indicates that although the mean ash percent remained the same between the two groups, in this study, individuals within the elderly group had ash percents that were much higher or lower than the younger group. These observations could have implications for hip fracture since it has been demonstrated that even small changes in mineral content can have a great effect on the mechanical properties of bone (Leichter et al., 1982; Mosekilde et al., 1987; Currey, 1988). Individuals with hypermineralized bone tissue would be at an increased risk for fracture because their high level of mineralization causes increased brittleness (Boyce and Bloebaum, 1993). Individuals with hypomineralized tissue have lower compressive strength (Leichter et al., 1982; Martens et al., 1983; McCalden et al., 1997). The possible existence of these subpopulations with excessively high or low mineralization is very important because it may explain the confusion in the literature regarding the existence and extent of aging changes in mineralization (Smith et al., 1975; Mosekilde et al., 1987; Parfitt, 1993; Lundeen et al., 2000). In order to define the subgroups with high or low trabecular mineralization compared to the population mean, the ash percent data from this study was compared with the donor data. Postmineral content analysis revealed that no correlations were found between mineralization measurements and weight, height, or cause of death in the donor groups. Another possible limitation to this investigation may be the limited sample size and regions analyzed to determine mineral content. The current study was limited since the sample size and regions imaged were small and the information regarding the donors was insufficient. It is believed that a more extensive study is necessary to define the subgroups that exist in the older population.
Past studies have suggested that variations in mineral content at a microscopic level or heterogeneity of individual trabeculae may affect fracture toughness of bone (Goldstein, 1987; Keaveny and Hayes, 1993). In the current study, variance within each individual was calculated, and these variances were not found to change significantly with age. This is further supported by a lack of statistical change in both the skewness and kurtosis of the gray-level histograms. This demonstrates that changes in mineralization between the two age groups in this study are not the result of increased variability within individuals. Instead, the data demonstrate that a subpopulation of individuals exist within the elderly group whose trabecular bone is hypermineralized or hypomineralized relative to the average.
The most important finding that is supported by the data in this study is the contrast between population variance and individual variances in the mineralization of trabecular bone in the elderly female proximal femurs investigated. As individuals age, they may gain or lose trabecular mineralization in the proximal femur as a whole, rather than developing regions of high and low mineralization. The current study does not support differences in mineralization within the proximal femur in the age group studied. The clinical significance for those elderly individuals that are at the extremes of mineralization is difficult to identify, as there is no technology that can noninvasively measure mineralization independent of porosity (Lundeen et al., 2001). Future research should focus on investigating the risk factors and indicators that cause subpopulations of elderly females to develop a significant overall increase or decrease in trabecular mineralization, such as medications or health problems, as well as techniques to measure these changes in vivo.
The authors thank Greg Stoddard for assistance with statistical analysis, Gwenevere Shaw for assistance in manuscript preparation, as well as Jennifer Holmes and Eric Vajda for assistance with quantitative BSE analysis.
- 1992. Projection effect errors in biomaterials and bone research. Cells Mater 2: 347–355. , .
- 1990. Backscattered electron imaging: the role in calcified tissue and implant analysis. J Biomater Appl 5: 56–85. , , .
- 1993. Cortical aging differences and fracture implications for the human femoral neck. Bone 14: 769–778. , .
- 1993a. Roentgenographic procedure for selecting proximal femur allograft for use in revision arthroplasty. J Arthroplasty 8: 347–360. , , , , , , .
- 1993b. Comparison of human and canine external femoral morphologies in the context of total hip replacement. J Biomed Mater Res 27: 1149–1159. , , , .
- 1997. Determining mineral content variations in bone using backscattered electron imaging. Bone 20: 485–490. , , , , .
- 1979. Mechanical properties of bone tissues with greatly differing functions. J Biomech 12: 313–319. .
- 1988. The effect of porosity and mineral content on the young's modulus of elasticity of compact bone. J Biomech 21: 131–139. .
- 1999. Hip fracture epidemiology: a review. Am J Orthop 28: 497–506. , , .
- 1957. A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 226: 497–509. , , .
- 1987. The mechanical properties of trabecular bone: dependence on anatomic location and function. J Biomech 20: 1055–1061. .
- 1988. Changes in quality of bone mineral on aging and in disease. Scanning Microsc 2: 1045–1054. , .
- 1993. Age and disease-related changes in the mineral of bone. Calcif Tissue Int 53(Suppl 1): S57–S64. .
- 1994. Monte Carlo simulations of electron scattering in bone. Bone 15: 285–291. , .
- 1996. Epidemiology of hip fractures. Bone 18: 57S–63S. , , , , , .
- 1993. A 20-year perspective on the mechanical properties of trabecular bone. Transact ASME 115: 534–542. , .
- 1998. Comparison of human, primate, and canine femora: implications for biomaterials testing in total hip replacement. J Biomed Mater Res 40: 475–489. , , .
- 1994. Fractures of the proximal part of the femur. J Bone Joint Surg 76A: 924–950. , , , , , , , .
- 1982. The relationship between bone density, mineral content and mechanical strength in the femoral neck. Clin Orthop 163: 272–281. , , , , , , , .
- 1995. Stress distributions within the proximal femur during gait and falls: implications for osteoporotic fracture. Osteoporos Int 5: 252–261. , , .
- 2000. Age-related cancellous bone loss in the proximal femur of Caucasian females. Osteoporos Int 11: 505–511. , , .
- 2001. The contribution of cortical and cancellous bone on dual energy X-ray absorptiometry measurements in the female proximal femur. Osteoporos Int 12: 192–198. , , , , .
- 1983. The mechanical characteristics of cancellous bone at the upper femoral region. J Biomech 16: 971–983. , , , , .
- 1997. Age-related changes in the compressive strength of cancellous bone. J Bone Joint Surg 79A: 421–427. , , .
- 1987. Biomechanical competence of vertebral trabecular bone in relation to ash density and age in normal individuals. Bone 8: 79–85. , , .
- 1993. Bone age, mineral density, and fatigue damage. Calcif Tissue Int 53(Suppl 1): S82–S86. .
- 1998. Validation of quantitative backscattered electron imaging for the measurement of mineral density distribution in human bone biopsies. Bone 23: 319–326. , , , .
- 1983. Cross-sectional geometry of Pecos Pueblo femora and tibiae: a biomechanical investigation: I, method and general patterns of variation. Am J Phys Anthropol 60: 359–381. , .
- 1990. Polypropylene containers for safe and predictable embedding of specimens in PMMA. J Histotechnol 13: 131–133. , , .
- 1994. Parallel experience of two different laboratories with the initiator perkadox 16 for polymerization of methylmethacrylates. J Histotechnol 17: 343–348. , .
- 1993a. The meaning of gray-levels in backscattered electron images of bone. J Biomed Mater Res 27: 47–56. , , , .
- 1993b. Influence of mineral content and composition on gray-levels in backscattered electron images of bone. J Biomed Mater Res 27: 57–64. , , , , .
- 1994. Analysis of a tension/compression skeletal system: possible strain-specific differences in the hierarchial organization of bone. Anat Rec 239: 396–404. , , , .
- 1975. The loss of bone mineral with aging and its relationship to risk of fracture. J Clin Invest 56: 311–318. , ,
- 1995. Consistency in calibrated backscattered electron images of calcified tissues and mineral analyzed in multiple imaging sessions. Scanning Microsc 9: 741–755. , , .
- 1999. Letter to the editor. Bone 24: 619–621. , .
- 1999. Influence of topography and specimen preparation on backscattered electron images of bone. Scanning 21: 379–386. , , , .
- 1966. Cancellous bone, its strength and changes with aging and an evaluation of some methods for measuring its mineral content: I, age changes in cancellous bone. J Bone Joint Surg 48A: 289–299. , .
- 1988. Contribution of the trabecular component to mechanical strength and bone mineral content of the femoral neck: an experimental study on cadaver bones. Scand J Clin Lab Invest 48: 457–460. , , .