Introduction
Intervertebral disc degeneration (IVDD) is a common condition, characterized by the degradation of extracellular matrix and the loss of hydrophilic matrix molecules in the nucleus pulposus [
1,
2]. Structural and biomechanical alterations due to the loss of disc height lead to changes in load distributions across the spine, eventually causing degeneration, segmental instability, and facet joint arthropathy [
3,
4].
Fatty infiltration of skeletal muscle, i.e., myosteatosis, is associated with muscle quality and skeletal muscle deterioration [
5]. Muscle deterioration may have a significant impact on the onset, progression, and severity of osteoarthritis (OA) [
6,
7]. Moreover, previous studies reported an association of paraspinal myosteatosis and IVDD, raising the possibility of a “whole-organ pathology” [
8‐
10].
Vertebral bone marrow adipose tissue (BMAT) plays an important role in bone health. Changes in its composition have been associated with various metabolic diseases, such as type 2 diabetes mellitus (T2DM) or osteoporosis [
11,
12]. The avascular intervertebral disc is supplied with nutrients via microvessels of the vertebral bodies. Since the conversion of hematopoietic bone marrow into bone marrow fat is associated with a decrease in blood flow and thus nutrient supply to the intervertebral discs, the amount of BMAT may play a role in the development of IVDD [
13‐
15].
Chemical shift encoding-based water-fat magnetic resonance imaging (MRI) is a non-invasive method that provides quantitative information about the biochemical water-fat composition of bone marrow and skeletal muscle in vivo [
16,
17]. It allows measurements of the vertebral BMAT fat fraction (BMAT-FF) and the paraspinal muscle proton density fat fraction (PDFF) [
18,
19].
Despite spatial and possible functional relationships of paraspinal PDFF and vertebral BMAT-FF, their interactions and potential role in the development of IVDD remain uncertain. Therefore, we systematically analyzed the correlation between paraspinal PDFF and vertebral BMAT-FF and investigated their associations with IVDD in individuals from the general population. We hypothesized that paraspinal PDFF is correlated with vertebral BMAT-FF and that both high paraspinal PDFF and high vertebral BMAT-FF are associated with IVDD.
Discussion
In this population-based study, quantitative MRI techniques were used to assess the correlation between paraspinal myosteatosis and vertebral bone marrow composition and their associations with IVDD. We found a significant correlation between paraspinal myosteatosis and vertebral bone marrow fatty infiltration. However, both paraspinal PDFF and vertebral BMAT-FF were positively associated with the presence of IVDD and our results suggest that increased fatty infiltration of paraspinal muscle and lumbar bone marrow may be both independent as well as additive risk factors of IVDD. Although we were not able to prove causality, these findings underline the complex biomechanics of muscle, bone, and intervertebral disc that play a role in spinal degeneration.
The paraspinal muscles are spatially and functionally connected to the vertebral column and are of major importance for spine stability [
29]. Despite this close relationship, only two studies assessed the association between vertebral bone marrow and paraspinal muscle composition using quantitative MRI [
30,
31]. While Burian et al found no significant association between bone marrow and muscle fat composition at the lumbosacral junction, Sollmann et al reported a significant correlation of paraspinal muscle PDFF and vertebral body PDFF in postmenopausal women in a small female-only cohort [
30,
31]. In contrast, no significant correlation between muscle and bone marrow PDFF was observed in premenopausal women [
30]. Here, we found a significant correlation between paraspinal fatty infiltration and bone marrow adipose tissue, measured by chemical shift encoding-based water-fat MRI. Of note, the participants of this study were overweight on average (BMI 25.0–29.9), similar to those investigated by Sollmann et al. A further parameter of interest is menopausal status. Our sample was too small to support analyses stratified for this variable since we would not be able to disentangle effects of age and menopause. In line with Sollmann et al, we therefore report results for the whole sample [
30]. However, future studies with larger sample sizes should further investigate this parameter.
Current concepts suggest that fat accumulation in muscle and bone follow similar mechanisms [
5]. Considering bone, adipocyte accumulation occurs in the marrow cavities. Fatty infiltration of the paraspinal muscle has two components: intra- and extramyocellular lipid, both contributing to the fat component in chemical encoding-based water-fat MRI in our study [
32,
33]. In the last years, several studies assessed the association of paraspinal muscle fat composition in patients with spinal disorders, including lower back pain, osteoporosis, spinal stenosis, and radiculopathy [
10,
34‐
37]. Significant associations of paraspinal muscle fatty degeneration with decreased bone mineral density (low bone mass/osteoporosis), measured by dual-energy X-ray absorptiometry (DXA) and quantitative computed tomography (qCT), have recently been described [
36,
37]. Also, several studies found associations between osteopenia/osteoporosis and higher MRI-based fat fraction measurements [
38‐
43]. In summary, these correlations of paraspinal myosteatosis and BMAT-FF highlight a potential (inter-)relationship between bone and muscle in the development of spinal pathologies such as osteoporosis.
Intramyocellular lipids display metabolic activity by producing proinflammatory mediators. Since it is known that pro-inflammatory mediators initiate and accelerate the process of IVDD [
44], there may be a link between myosteatosis and IVDD. Numerous previous studies reported an association between paraspinal myosteatosis and IVDD [
8,
9,
45‐
48]. In line with these findings, we also report a significant positive association between paraspinal PDFF and IVDD. However, it has to be mentioned that these studies used qualitative or semi-quantitative measurements such as the cross-sectional area (CSA) or the visual Goutallier-score for determining paraspinal fatty infiltration. It is known that semi-quantitative visual grading systems such as the Goutallier-score are prone to significant inter- and intraobserver variability [
49]. Thus, a strength of our study is the use of DIXON-based quantitative MRI for myosteatosis measurements, a reliable method with good concordance to histology and validated inter- and intraobserver reliability [
19,
33,
50].
In a retrospective study of 72 community-based subjects, Teichtahl et al reported a positive association between lumbar IVDD, Modic changes of the vertebral endplates, and a high paraspinal muscle fat content, suggesting that IVDD may be considered a “whole-organ” pathology affecting intervertebral discs, muscle, and the bone [
8]. In line, we found a positive association of vertebral BMAT-FF with the presence of IVDD. Also, there was an increase in paraspinal PDFF and BMAT-FF according to the maximum lumbar spine Pfirrmann grade and the presence of severe disc degeneration at two or more lumbar motion segments. Moreover, our findings suggest that the associations of vertebral BMAT-FF and paraspinal PDFF with IVDD are likely independent of each other, which further supports the hypothesis of a “whole-organ” pathology. Prior to the present analysis, only two studies performed quantitative MRI to assess the association between vertebral bone marrow fat and IVDD [
13,
51]. In agreement with our results, Krug et al and Ji et al found significant associations between disc degeneration adjacent vertebral body bone marrow fat [
13,
51].
Our study has several limitations. First, our results are limited by relatively small sample size and the cross-sectional study design, therefore requiring confirmation in larger, longitudinal cohort studies. In addition, MR-based results of myosteatosis measurements were not compared to histopathology which is still considered the current gold standard for quantification of fat content. Sample PDFF measurements on a single axial slice at the level of the L3 vertebra, as performed in this study, may not reflect the exact distribution of lipid storage within the whole. Yet, previous studies have demonstrated the validity and reproducibility of a standardized, anatomic landmark-based quantification of skeletal muscle fat via measurement of PDFF and good concordance to histology [
19,
33,
50].
In this study, we used the L1 and L2 vertebral bodies for BMAT-FF measurements. These levels may not be representative of the entire lumbar spine, as previous studies found differences in vertebral bone marrow fat fractions with an increase of BMAT-FF in the craniocaudal direction [
13,
52]. Yet, considering that the lower lumbar spine (L3-L5) usually is the level of most severe degenerations and the described craniocaudal increase of BMAT-FF in lumbar vertebrae, it may be speculated that the effects reported in our study are rather conservative estimates of the association between lumbar spine BMAT-FF and IVDD. However, further studies are needed to confirm this hypothesis.
Our MRI sequence for BMAT-FF measurements did not account for T2* effects, therefore likely overestimating fat fraction when compared to other methods like MR spectroscopy [
53]. Also, it did not allow sub-stratification of different lipids [
54]. Despite this, our analysis focused on relative differences between groups and was not focused on providing reference values.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.