Bone Marrow Adiposity Alterations in Postmenopausal Women With Type 2 Diabetes Are Site-Specific

Sammy Badr, Anne Cotten, Daniela Lombardo, Stefan Ruschke, Dimitrios C. Karampinos, Nassima Ramdane, Michaël Génin, Julien PaccouJournal of the Endocrine Society 2024 · DOI

TL;DR

This study reveals that type 2 diabetes (T2DM) alters bone marrow fat composition in postmenopausal women, but these changes are not uniform across the skeleton. Specifically, T2DM is associated with less fat in the femoral head, while fat content in the lumbar spine shows no significant difference. This suggests that targeting specific bone sites may be crucial for understanding and potentially treating diabetes-related bone fragility.

What Problem Does This Paper Address?

The musculoskeletal system is increasingly recognized as a target of damage in diabetes, with individuals with type 2 diabetes (T2DM) experiencing increased bone fragility and fracture risk despite often having normal or high bone mineral density (BMD). While various mechanisms are implicated, the role of bone marrow adipose tissue (BMAT) – the fat within the bone marrow – is gaining attention. Previous research, primarily in animal models, has suggested that T2DM leads to increased BMAT. However, human studies using MRI have yielded mixed results, with some showing differences in BMAT content and composition (like lipid unsaturation) and others finding no significant changes, particularly in the lumbar spine. These discrepancies may stem from differences in study populations (age, sex), measurement sites, and imaging techniques. This paper aims to clarify these site-specific alterations in BMAT in a well-characterized cohort of postmenopausal women with and without T2DM, using advanced MRI techniques.

Methods

The researchers employed a cross-sectional study design using data from the ADIMOS cohort. The core imaging technique was chemical shift-based water-fat separation imaging (WFI), specifically the mDixon-Quant sequence, to quantify the proton density fat fraction (PDFF). This method allows for the simultaneous assessment of fat and water content in the bone marrow. To further characterize BMAT composition, proton magnetic resonance spectroscopy (¹H-MRS) was used in a subgroup of participants to determine the apparent lipid unsaturation level (aLUL).

Key aspects of their methodology include:

  • Site-Specific Assessment: BMAT was measured at multiple skeletal sites: the lumbar spine (vertebrae L1-L4), femoral head, femoral neck, and femoral diaphysis.
  • Advanced MRI: WFI was used for quantitative PDFF measurements, accounting for T1 bias and T2* decay for improved accuracy. ¹H-MRS provided insights into the chemical composition of the marrow fat.
  • Statistical Adjustments: To isolate the effect of T2DM, imaging-derived PDFF and aLUL values were adjusted for potential confounders such as age, body mass index (BMI), and bone mineral density (BMD) using linear mixed models.
  • Mediation Analysis: The study also explored whether recent fractures or BMD mediated the relationship between T2DM and BMAT.

Datasets

The study utilized data from the ADIMOS cohort, a case-control investigation focused on fragility fractures in postmenopausal women. The cohort included 199 postmenopausal women, with 29 diagnosed with T2DM and 170 without. The data collected included demographic information, clinical history, laboratory measurements (including HbA1c), BMD measurements via dual-energy x-ray absorptiometry (DXA), and the MRI-derived BMAT parameters (PDFF and aLUL). No new datasets or benchmarks were introduced.

Key Insights and Ideas

  • Site-Specificity is Key: The most striking finding is that the impact of T2DM on bone marrow fat is not uniform. While the femoral head shows a significant reduction in fat content in women with T2DM, the lumbar spine does not. This challenges the idea of a generalized increase in BMAT due to diabetes.
  • Femoral Head as a Potential Biomarker: The lower PDFF in the femoral head of diabetic women, even after adjusting for confounders, suggests this site might be particularly sensitive to the metabolic changes associated with T2DM and could serve as a potential imaging biomarker.
  • Lipid Unsaturation Matters: ¹H-MRS data revealed that T2DM was associated with lower lipid unsaturation at the L3 vertebral level. This indicates a change in the composition of the bone marrow fat, not just its quantity, which could have implications for bone health.
  • BMD and Fractures as Mediators: While the study explored mediation, it found no significant mediating effect of recent fractures on the T2DM-BMAT association. Lumbar spine BMD showed a weak mediating effect on the T2DM-femoral head PDFF relationship, but this was considered exploratory.

Key Results

  • Demographics: Women with T2DM were significantly heavier and had higher BMD at the total hip, femoral neck, and lumbar spine compared to non-diabetic women. They also had higher HbA1c and glycemia levels.
  • Imaging-Based PDFF: After adjusting for age, BMI, and BMD, women with T2DM had significantly lower PDFF in the femoral head (88.0% vs. 90.6%, P < .001). No significant differences were found for the femoral neck or diaphysis. Lumbar spine PDFF showed a trend towards being higher in T2DM patients (61.2% vs. 57.4%), but this did not reach statistical significance (P = .053).
  • ¹H-MRS Data: At the L3 vertebral level, while PDFF showed no significant difference, the apparent lipid unsaturation level (aLUL) was lower in patients with T2DM (3.9% vs. 4.3%, P = .02). No significant differences were found for PDFF or aLUL at the femoral neck.
  • Correlations with HbA1c: A small positive correlation was found between lumbar spine PDFF and HbA1c in the overall cohort and in non-diabetic patients, but no significant correlations were observed for other sites or in the diabetic group specifically.

Limitations and Future Work

The authors acknowledge several limitations:

  • Sample Size: The study may not have been adequately powered to detect smaller differences, particularly for comparisons between the T2DM and non-diabetic groups.
  • Cross-Sectional Design: This design prevents the establishment of causal relationships or the assessment of temporal associations between T2DM and BMAT changes.
  • Control Group: The use of women with osteoarthritis as controls might introduce some bias, although they were selected to be free of fragility fractures.
  • Well-Controlled Diabetes: The T2DM patients in this cohort had relatively well-controlled diabetes, which might have masked more pronounced differences in BMAT.
  • Generalizability: Findings are limited to postmenopausal women and cannot be generalized to men or younger age groups.

Future work suggested includes:

  • Larger, Longitudinal Studies: To confirm these site-specific findings and establish temporal relationships.
  • Sex-Stratified Analyses: To investigate potential sex differences in T2DM-related BMAT alterations.
  • Broader Assessment of BMAT: Reporting BMAT content and composition at both proximal femur and lumbar spine sites in future studies.
  • Therapeutic Target Exploration: Further research is needed to define the role of marrow adiposity in diabetic skeletal fragility and to determine if it can be a therapeutic target.

Summary updated

Abstract

The content and composition of BMAT are modified in postmenopausal women with T2DM and these changes occur at specific sites.
BibTeX
@article{badr2024bone,
  author = {Sammy Badr and Anne Cotten and Daniela Lombardo and Stefan Ruschke and Dimitrios C. Karampinos and Nassima Ramdane and Michaël Génin and Julien Paccou},
  title = {{Bone Marrow Adiposity Alterations in Postmenopausal Women With Type 2 Diabetes Are Site-Specific}},
  journal = {Journal of the Endocrine Society},
  year = {2024},
  doi = {10.1210/jendso/bvae161},
}
Figure