Vertebral trabecular microarchitecture changes in the normally ageing population

Introduction

Osteoporosis is a skeletal disorder characterized by bone mass loss and trabecular microarchitecture deterioration, causing skeletal fragility. It usually occurs in postmenopausal women and older men, leading to increased fracture risk of the vertebral body, proximal femur, and pelvic bone (1). Typically, vertebral fractures are asymptomatic and overlooked in radiologic reports, resulting in body height loss, reduction in quality of life, and decreased mobility (2-5). Hip and pelvic fractures cause the most severe outcomes, and they result in high morbidity and mortality (6). According to an epidemiologic survey, approximately 50% of postmenopausal women and 20% of men older than 50 years will have at least one fragility fracture (7). Therefore, the methods of screening and monitoring pharmaceutical treatment for osteoporosis are imperative.

The gold standard for diagnosing osteoporosis is bone mineral density (BMD) as measured by dual-energy X-ray absorptiometry (DXA) (8,9). For men aged at least 50 years and post-menopausal women, the normal BMD T-score is −1.0 or higher, while osteopenia is −1.0 to −2.5 and osteoporosis is −2.5 or less (9). This BMD quantitatively measures the bone mineral in the lumbar spine, femur, and forearm. Additional advantages of DXA include short examination time and low radiation dose; therefore, it is widely used in clinical practice as a tool for diagnosing osteoporosis and monitoring patient prognosis (10). However, considering the major clinical concern of osteoporosis outcome—bone fracture—the ability of BMD from DXA to provide bone fracture risk prediction and explanation power is limited (11). Fractures are highly affected by bone strength, which is a combination of bone material ratios and bone structure (cortical and trabecular microarchitectures) (12,13). The trabecular microarchitecture can be quantified indirectly by the trabecular bone score from DXA images. However, it is difficult to calculate the three-dimensional (3D) trabecular microarchitecture from two-dimensional (2D) images, because there is insufficient information on the microarchitecture in the z direction.

Computed tomography (CT) has benefited from technological advances and image processing improvements, prompted studies using multi-detector CT (MDCT) to produce high-resolution images of the trabecular microarchitecture (14,15). To determine the 3D morphology of trabecular bone, several quantitative parameters were developed: bone volume to total volume (BV/TV; as a percentage), trabecular thickness (Tb.Th), trabecular number (Tb.N), trabecular separation (Tb.Sp), and fractal dimensions [2D fractal dimension (D2D) and 3D fractal dimension (D3D)]. These are measured using micro-computed tomography (µCT) imaging or peripheral quantitative computed tomography (pQCT). Studies have shown that these parameters along with BMD can predict fracture risk and explain vertebral fracture assessment (14). However, changes in the trabecular microarchitecture associated with normal aging have not yet been assessed using routine chest or abdominal CT images.

An extensive CT scanning range may cause the risk of radiation doses and create significant obstacles to routine monitoring of treatment in clinical practice. To address the scan range and dose used in these cases, Palepu et al. (16) reported that, throughout all the lumbar vertebral levels, the L1 vertebral body is the largest, and the cortical thickness is the thinnest. Importantly, L1 has been recognized as a sentinel vertebra for osteoporosis-related fracture risk: it is among the most common sites of osteoporotic vertebral compression fractures, often occurring early in the disease process. Moreover, morphometric assessment of L1 is incorporated into vertebral fracture assessment and correlates strongly with systemic skeletal fragility. Additionally, MDCT emits a radiation dose of less than 1 mSv on L1 (17), which could be appropriate for a routine or short-term follow-up exam protocol. In this study, we analyzed routine clinical CT images to quantitatively characterize trends and differences in vertebral trabecular microarchitectures based on age and sex in normal older adults. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1453/rc).

Methods

Patients and study design

Taipei Medical University-Joint Institutional Review Board approved this study (TMU-JIRB approval No. N202204098) on April 29, 2022, and informed consent was waived because of the retrospective nature of the study. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. We retrospectively searched the medical records of patients who underwent routine MDCT scans with the lumbar spine protocol between January 1, 2014, and December 31, 2021. Inclusion criteria were as follows: (I) age between 50 and 80 years; (II) availability of a complete medical history; and (III) availability of lumbar spine CT images. Exclusion criteria included: (I) compression fracture at the L1 vertebral level; (II) history of anti-osteoporosis pharmaceutical treatment; and (III) diagnosis of diabetes mellitus, chronic kidney disease, or any malignancy. The final cohort of 360 eligible participants (180 women) was randomly selected and classified into six groups according to age intervals of 5 years (Figure 1).

Figure 1 Flowchart shows the selection of patients. OPD, outpatient department; PE, physical health examinations.

CT imaging acquisition protocol

All CT images without contrast agent administration were obtained by a dual-source MDCT scanner equipped with 256-slice and 78 cm aperture (Somatom Definition Flash; Siemens Healthineers, Erlangen, Germany). All imaging protocols were performed by identical and high resolution CT imaging acquisition parameters: scans with collimation of 256 mm × 0.6 mm, tube voltage of 120 kVp, smart tube current of 50–250 mA at least with automatic exposure control (AEC) technology, a gantry rotation speed of 0.5 sec/rotation, ultra-high bone kernel of 170 µm resolution (30lp/cm@0%MTF), z-sharp interpolation technology of 256 slice in the Z-axis and sinogram affirmed iterative reconstruction (SAFIRE) applied to improve and investigate the resolution of MDCT in a single breath hold so that we can capture the morphological structure of trabecular bones. Finally, we will use the Otsu Binarization algorithm to improve and rebuild the 3D images of trabecular bone using Quibim Software. Scan coverage was taken from the lower half of T12 to the upper half of L2. The volume of interest (VOI) (Figure 2) for trabecular microarchitecture analysis is an elliptical cylinder containing 50% of the vertebral body. The position was chosen to reduce regional variations.

Figure 2 Example trabecular images of L1 vertebra. (A) FOV in a high-resolution CT image at L1. (B) 3D reconstruction of the elliptic cylinder used to analyze trabecular microarchitecture. 3D, three-dimensional; CT, computed tomography; FOV, field of view.

Trabecular microarchitecture analysis

Trabecular microarchitecture analysis was performed by a senior technologist using promising software (QUIBIM Precision 2.8, QUIBIM SL, Valencia, Spain). The morphological trabecular microarchitecture was characterized using the following quantitative parameters: BV/TV percentage, Tb.Th, Tb.Sp, Tb.N, D2D, and D3D.

BV/TV percentage

Given that the reconstructed trabecular bone is a composite of several tiny voxels, BV/TV is the ratio of the total number of voxels containing trabecular bone (n_voxel of BV) to the total number of voxels (n_voxel of TV) within the VOI:

BV/TV=n_voxelofBVn_voxelofTV[1]

Tb.Th

Each trabecular bone is divided into several contours of a slice in the axial view; additionally, the contour of the trabecular bone is a tree-like structure with numerous skeletons. Thus, Tb.Th is estimated as the minimum 2D width of each skeleton (2d_width). The average width among all skeletons and contours is defined as Tb.Th:

Tb.Th=∑j∈contour∑i ∈Skmin(2d_width)n_skeletonn_contour[2]

Tb.N

Because Tb.N cannot be counted directly from a CT image, it is computed as the ratio between BV/TV and Tb.Th:

Tb.N=BV/TVTb.Th[3]

Tb.Sp

The Tb.Sp is defined as Tb.Th subtracted from Tb.N divided into 1,000:

Tb.Sp=1000Tb.N−Tb.Th[4]

D2D and D3D

The counting box calculates the D2D and D3D by using the pixel and voxel methods, respectively. The program employed several resolution boxes to fit the following equations:

log(N)=−D2Dlog(λ)+k[5]

log(N)=−D3Dlog(λ)+k[6]

Inter-operator reproducibility

One experienced operator performed two repeated measures across 44 randomly selected individuals to determine reproducibility in trabecular microarchitecture analyses. The two measures were separated by at least 1 month apart.

Statistical analysis

Given the relatively large sample size in each subgroup (n=30 per sex per 5-year age group), approximate normality was assumed based on the central limit theorem. Descriptive statistics are presented as mean ± standard deviation (SD) with minimum and maximum values. An independent t-test was adopted to compare the sexes, whereas analysis of variance (ANOVA) with Scheffe post-hoc multiple comparisons was used to compare age groups. Correlation analysis and regression analysis were conducted to show the trends of trabecular microarchitecture during aging. A statistical threshold of P<0.05 was used to find significant differences. All statistical analyses and illustrations were performed using Python (version 3.7.4 with numpy version 1.20.1, matplotlib version 3.1.3, scipy version 1.6.1, and scikit-learn version 0.24.1) and SPSS (version 22; IBM, Armonk, NY, USA).

Results

A total of 360 participants were recruited in the study. Each sex was divided into six 5-year age groups, with 60 participants per group. There is a significant and strong correlation of the trabecular microarchitecture for all parameters in the inter-operator repeatability test (Pearson correlation coefficient, 0.933 to 0.994; P<0.001 for all) (Figure 3 and Figure S1). These results confirm that measurement precision and reproducibility are sufficient for subsequent analyses.

Figure 3 Reproducibility results. Forty-four patients were randomly selected for repeated testing. (A) BV/TV percentage. (B) Tb.Th mean. (C) Tb.Sp mean. (D) Tb.N. (E) D2D. (F) D3D. The ρ values indicate the Pearson correlation coefficients. Asterisks (***) indicate statistical significance at P<0.001. BV/TV, bone volume to total volume; D2D, 2D fractal dimension; D3D, 3D fractal dimension; Tb.N, trabecular number; Tb.Sp, trabecular separation; Tb.Th, trabecular thickness.

Results of the 360 participants (180 women) show that Tb.N, D2D, and D3D gradually declined while Tb.Th mean and Tb.Sp mean gradually increased during normal ageing (Table 1, Table S1, Figure 4, and Figure S2). In men, BV/TV percentage did not change with age. In men, the age of 65 years was the crucial threshold for Tb.Sp mean and Tb.N. In women, however, the age of 65 years was significant for Tb.Th mean, Tb.N, and D2D. All other parameters showed a less clear change, and at the age of 65 to 69 years.

Figure 4 Age-based trends in the parameters of trabecular microarchitecture. (A) BV/TV percentage. (B) Tb.Th mean. (C) Tb.Sp mean. (D) Tb.N. (E) D2D. (F) D3D. Asterisks (*, **, and ***) indicate statistical significance at P<0.01, P<0.05, and P<0.001, respectively. BV/TV, bone volume to total volume; D2D, 2D fractal dimension; D3D, 3D fractal dimension; Tb.N, trabecular number; Tb.Sp, trabecular separation; Tb.Th, trabecular thickness.

Men and women did not differ in BV/TV percentage, but in two age groups (55–59 and 75–80 years), they did differ in Tb.Th mean, Tb.Sp mean, and Tb.N. Tb.N demonstrated a consistent decline with advancing age in both sexes, reaching the lowest values in the 75–80 years group. A significant sex difference was observed in 55–64 group, with women exhibiting higher Tb.N than men. D2D and D3D progressively decreased with age, indicating a reduction in the structural complexity of the trabecular network.

From age 60 to 69 years, BV/TV percentage, Tb.N, D2D, and D3D gradually decreased, whereas Tb.Th mean and Tb.Sp mean increased. Remarkably, the rates of decline in Tb.N and D2D were greater compared to the others, and the rates of increase in Tb.Sp mean were greater compared to the others (Figure 5 and Figure S3).

Figure 5 The ageing trends of quantitative trabecular microarchitecture parameters fitted to third-order polynomials. Triangles symbolize men; squares symbolize women. The three dotted lines are the fitted curves, where blue, red, and black lines represent men, women, and all sexes. (A) BV/TV percentage. (B) Tb.Th mean. (C) Tb.Sp mean. (D) Tb.N. (E) D2D. (F) D3D. BV/TV, bone volume to total volume; D2D, 2D fractal dimension; D3D, 3D fractal dimension; Tb.N, trabecular number; Tb.Sp, trabecular separation; Tb.Th, trabecular thickness.

Age was significantly associated with all trabecular microarchitectural parameters (P<0.001), with varying correlation strengths (Table 2 and Table S2). The strongest negative correlations were found in Tb.N (r=−0.542, R²=0.294), indicating lower values in men. Moderate negative correlations were observed for BV/TV percentage and D2D, while Tb.Th mean and Tb.Sp mean showed positive correlations, suggesting thicker trabecular but wider Tb.Sp in men. D3D showed the weakest correlation (r=−0.252, R²=0.064), indicating minimal sex-related differences in 3D complexity.

Discussion

We quantitatively determined the age-specific and sex-specific differences of vertebral trabecular microarchitecture in a normal older population by applying reconstructed high-resolution CT images from routine abdominal imaging. Results indicate that D3D was the most sensitive to sex in most age groups; other parameters showed significant differences only in two age groups (50–64 and 65–80 years). They also indicate that most of the quantitative trabecular microarchitecture parameters showed two-stage changes during the ageing process, implying that routine abdominal CT scans with high-resolution reconstructions can characterize changes in vertebral trabecular microarchitecture.

Normal ageing is marked by correlations between age and Tb.Th mean (ρ=0.376), and Tb.Sp mean (ρ=0.397) being moderately positive, D2D (ρ=−0.462) and Tb.N (ρ=−0.542) being moderately negative. Notably, Tb.N was the most sensitive parameter of trabecular microarchitecture: it was the most strongly negatively correlated variable in both men (ρ=−0.574; P≤0.001) and women (ρ=−0.512; P≤0.001). Others have also shown that Tb.Sp mean and that Tb.N and BV/TV decrease during normal ageing (18-21); however, the results for Tb.Th mean was more controversial. For instance, Yeni et al. (18) reported that Tb.Th mean increased with normal ageing, whereas Hulme et al. (19) and Chen et al. (20) showed no significant trend. Using a weight-bearing skeletal site, Greenwood et al. (22) showed that Tb.N and Tb.Sp of non-fractured femoral heads were significantly but weakly negatively correlated to age in both sexes.

Our findings regarding the age-related decline in Tb.N and relative preservation of Tb.Th are consistent with prior histomorphometric and high-resolution-pQCT (HR-pQCT) studies (2,14). However, unlike these studies, which relied on dedicated research-grade imaging or invasive biopsies, we demonstrate that such age- and sex-related trabecular modifications can be reliably quantified using routine abdominal MDCT scans enhanced with high-resolution reconstruction. This approach highlights the clinical feasibility of anti-osteoporosis drugs, especially anabolics, which can shorten the follow-up period to achieve treatment results, rather than waiting for 1–2 years by DXA scans.

Others have characterized sex-based differences in trabecular microarchitecture using pQCT at the radius and tibia or µCT at other skeletal sites (20,22-24). Whittier et al. (23) enrolled 1,236 adult participants (768 women) aged 18 to 90 years and found that, at the radius and tibia, the parameters of trabecular microarchitecture significantly declined among postmenopausal women and men aged more than 50 years. Specifically, the trabecular BMD declined at the radius at 50 years of age and at the tibia at 70 years of age. Because certain skeletal sites, such as the femur or vertebrae, function to support the body weight, the trabecular microarchitecture can be expected to show different degradation patterns in men and women. Greenwood et al. (22) used µCT to compare femoral heads with traumatic fractures from 37 osteoporotic patients with non-fractured femoral heads from 83 cadavers. They reported that both men and women with fractures showed smaller values for Tb.Th and larger values for Tb.Sp and that women also showed smaller values for Tb.N. Studies that used vertebrae found that BV/TV and Tb.N declined faster in men (19% and 24%, respectively) than in women (16% and 22%, respectively) in those aged 62 to 92 years (18,20,21). Our results were similar in that Tb.N, Tb.Th, and Tb.Sp were more heterogeneous in men than in women, particularly in those aged 55–59 and 75–80 years. Although the sexes did not differ in terms of BV/TV percentage, our correlation analyses showed that BV/TV and Tb.N decreased and Tb.Th mean and Tb.Sp mean increased faster in men than women; the reverse was true for D2D and D3D. We therefore conclude that the trabecular microarchitecture remodels more quickly in men than in women. This gives an indirect explanation for the greater fracture risk in women compared to men.

Others used µCT on cadavers because there was no radiation dose concern. Similarly, at the tibia and radius, pQCT does not require a high radiation dose. In our study, the resolution of in-plane raw clinical CT was insufficient for calculating trabecular microarchitecture parameters. Therefore, we leveraged a high-resolution reconstruction method and medical imaging processing to improve the spatial resolution, thus meeting the minimal requirements of trabecular microarchitecture analysis. Guha et al. (25) showed that µCT was highly correlated with four clinical CTs (HR-pQCT, dental cone beam CT, MDCT, and extremity cone beam CT) when determining each trabecular microarchitecture parameter. While a lower resolution might lead to a rough trabecular microarchitecture, it would be more difficult to observe small or regional changes. We therefore performed our protocol with only specific age groups and parameters.

This study has some limitations. First, we omitted areal BMD (aBMD) and volumetric BMD (vBMD) in a cross-comparison because prior DXA examinations were lacking; we substituted BV/TV percentage because it is measured by the bone volume ratio and is therefore similar to BMD. Second, we enrolled no younger participants for comparison; however, our focus was to characterize the vertebral trabecular microarchitecture during normal ageing in an older population. Third, all our data were collected from a single MDCT machine at a single institution; the reproducibility and sensitivity of vertebral trabecular microarchitecture parameters must be further investigated across other institutions. Fourth, although MDCT with high-resolution reconstruction improves spatial resolution, its intrinsic in-plane resolution is still lower than that of µCT or HR-pQCT. This limitation may reduce the sensitivity of certain morphometric parameters to detect subtle sex-related differences within 5-year age subgroups.

Conclusions

Vertebral trabecular microarchitecture undergoes distinct age- and sex-related alterations, with Tb.N as the most sensitive marker of structural decline. Men exhibit earlier and greater declines in Tb.N and D3D, while women maintain higher D3D values, reflecting greater architectural complexity. These findings indicate accelerated remodeling in men and support D3D as a robust, sex-sensitive marker of microarchitectural integrity. Incorporating high-resolution reconstruction into routine CT may enhance BMD assessments and improve individualized fracture risk evaluation in older adults.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1453/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1453/dss

Funding: This work was supported in part by a grant from the Ministry of Science and Technology (MOST) (No. 110-2314-B-038-083-MY3).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1453/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Taipei Medical University-Joint Institutional Review Board approved this study (TMU-JIRB approval No. N202204098) on April 29, 2022, and informed consent was waived because of the retrospective nature of the study. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Cummings SR, Melton LJ. Epidemiology and outcomes of osteoporotic fractures. Lancet 2002;359:1761-7. [Crossref] [PubMed]
2. Parfitt AM, Mathews CH, Villanueva AR, Kleerekoper M, Frame B, Rao DS. Relationships between surface, volume, and thickness of iliac trabecular bone in aging and in osteoporosis. Implications for the microanatomic and cellular mechanisms of bone loss. J Clin Invest 1983;72:1396-409. [Crossref] [PubMed]
3. Wainwright SA, Marshall LM, Ensrud KE, Cauley JA, Black DM, Hillier TA, Hochberg MC, Vogt MT, Orwoll ES. Hip fracture in women without osteoporosis. J Clin Endocrinol Metab 2005;90:2787-93. [Crossref] [PubMed]
4. Molino G, Dalpozzi A, Ciapetti G, Lorusso M, Novara C, Cavallo M, Baldini N, Giorgis F, Fiorilli S, Vitale-Brovarone C. Osteoporosis-related variations of trabecular bone properties of proximal human humeral heads at different scale lengths. J Mech Behav Biomed Mater 2019;100:103373. [Crossref] [PubMed]
5. Tamimi I, Cortes ARG, Sánchez-Siles JM, Ackerman JL, González-Quevedo D, García Á, Yaghoubi F, Abdallah MN, Eimar H, Alsheghri A, Laurenti M, Al-Subaei A, Guerado E, García-de-Quevedo D, Tamimi F. Composition and characteristics of trabecular bone in osteoporosis and osteoarthritis. Bone 2020;140:115558. [Crossref] [PubMed]
6. Melton LJ, Cooper C. Chapter 21 - Magnitude and Impact of Osteoporosis and Fractures. In: Marcus R, Feldman D, Kelsey J. editors. Osteoporosis. 2nd ed. San Diego: Academic Press; 2001:557-67.
7. Sambrook P, Cooper C. Osteoporosis. Lancet 2006;367:2010-8. [Crossref] [PubMed]
8. The International Society For Clinical Desitometry. ISCD Official Positions for Adults. 2019. Accessed Aug 22, 2021. Available online: https://iscd.org/learn/official-positions/adult-positions/
9. Kanis JA. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: synopsis of a WHO report. WHO Study Group. Osteoporos Int 1994;4:368-81. [Crossref] [PubMed]
10. Liang CL, Wang HK, Syu FK, Wang KW, Lu K, Liliang PC. Repeated vertebral augmentation for new vertebral compression fractures of postvertebral augmentation patients: a nationwide cohort study. Clin Interv Aging 2015;10:635-42. [Crossref] [PubMed]
11. Compston JE, McClung MR, Leslie WD. Osteoporosis. Lancet 2019;393:364-76. [Crossref] [PubMed]
12. Brandi ML. Microarchitecture, the key to bone quality. Rheumatology (Oxford) 2009;48:iv3-8. [Crossref] [PubMed]
13. Seeman E, Delmas PD. Bone quality--the material and structural basis of bone strength and fragility. N Engl J Med 2006;354:2250-61. [Crossref] [PubMed]
14. Ito M, Ikeda K, Nishiguchi M, Shindo H, Uetani M, Hosoi T, Orimo H. Multi-detector row CT imaging of vertebral microstructure for evaluation of fracture risk. J Bone Miner Res 2005;20:1828-36. [Crossref] [PubMed]
15. Lespessailles E, Chappard C, Bonnet N, Benhamou CL. Imaging techniques for evaluating bone microarchitecture. Joint Bone Spine 2006;73:254-61. [Crossref] [PubMed]
16. Palepu V, Rayaprolu SD, Nagaraja S. Differences in Trabecular Bone, Cortical Shell, and Endplate Microstructure Across the Lumbar Spine. Int J Spine Surg 2019;13:361-70. [Crossref] [PubMed]
17. Alshamari M, Geijer M, Norrman E, Lidén M, Krauss W, Wilamowski F, Geijer H. Low dose CT of the lumbar spine compared with radiography: a study on image quality with implications for clinical practice. Acta Radiol 2016;57:602-11. [Crossref] [PubMed]
18. Yeni YN, Zinno MJ, Yerramshetty JS, Zauel R, Fyhrie DP. Variability of trabecular microstructure is age-, gender-, race- and anatomic site-dependent and affects stiffness and stress distribution properties of human vertebral cancellous bone. Bone 2011;49:886-94. [Crossref] [PubMed]
19. Hulme PA, Boyd SK, Ferguson SJ. Regional variation in vertebral bone morphology and its contribution to vertebral fracture strength. Bone 2007;41:946-57. [Crossref] [PubMed]
20. Chen H, Shoumura S, Emura S, Bunai Y. Regional variations of vertebral trabecular bone microstructure with age and gender. Osteoporos Int 2008;19:1473-83. [Crossref] [PubMed]
21. Lochmüller EM, Matsuura M, Bauer J, Hitzl W, Link TM, Müller R, Eckstein F. Site-specific deterioration of trabecular bone architecture in men and women with advancing age. J Bone Miner Res 2008;23:1964-73. [Crossref] [PubMed]
22. Greenwood C, Clement J, Dicken A, Evans P, Lyburn I, Martin RM, Stone N, Zioupos P, Rogers K. Age-Related Changes in Femoral Head Trabecular Microarchitecture. Aging Dis 2018;9:976-87. [Crossref] [PubMed]
23. Whittier DE, Burt LA, Hanley DA, Boyd SK. Sex- and Site-Specific Reference Data for Bone Microarchitecture in Adults Measured Using Second-Generation HR-pQCT. J Bone Miner Res 2020;35:2151-8. [Crossref] [PubMed]
24. Chen H, Zhou X, Fujita H, Onozuka M, Kubo KY. Age-related changes in trabecular and cortical bone microstructure. Int J Endocrinol 2013;2013:213234. [Crossref] [PubMed]
25. Guha I, Klintström B, Klintström E, Zhang X, Smedby Ö, Moreno R, Saha PK. A comparative study of trabecular bone micro-structural measurements using different CT modalities. Phys Med Biol 2020; [Crossref]