Intermuscular adipose tissue and lean muscle mass assessed with MRI in people with chronic back pain in Germany: a retrospective observational study

Abstract

Background: Chronic back pain (CBP) affects over 80 million people in Europe, contributing to substantial healthcare costs and disability. Understanding modifiable risk factors, such as muscle composition, may aid in prevention and treatment. This study investigates the association between lean muscle mass (LMM) and intermuscular adipose tissue (InterMAT) with CBP using noninvasive whole-body magnetic resonance imaging (MRI).

Methods: This cross-sectional analysis used whole-body MRI data from 30,868 participants in the German National Cohort (NAKO), collected between 1 May 2014 and 1 September 2019. CBP was defined as back pain persisting >3 months. LMM and InterMAT were quantified via MRI-based muscle segmentations using a validated deep learning model. Associations were analyzed using mixed logistic regression, adjusting for age, sex, diabetes, dyslipidemia, osteoporosis, osteoarthritis, physical activity, and study site.

Findings: Among 27,518 participants (n = 12,193/44.3% female, n = 14,605/55.7% male; median age 49 years IQR 41; 57), 21.8% (n = 6003; n = 2999/50.0% female, n = 3004/50% male; median age 53 years IQR 46; 60) reported CBP, compared to 78.2% (n = 21,515; n = 9194/42.7% female, n = 12,321/57.3% male; median age 48 years IQR 39; 56) who did not. CBP prevalence was highest in those with low (<500 MET min/week) or high (>5000 MET min/week) self-reported physical activity levels (24.6% (n = 10,892) and 22.0% (n = 3800), respectively) compared to moderate (500-5000 MET min/week) levels (19.4% (n = 12,826); p < 0.0001). Adjusted analyses revealed that a higher InterMAT (OR 1.22 per 2-unit Z-score; 95% CI 1.13-1.30; p < 0.0001) was associated with an increased likelihood of chronic back pain (CBP), whereas higher lean muscle mass (LMM) (OR 0.87 per 2-unit Z-score; 95% CI 0.79-0.95; p = 0.003) was associated with a reduced likelihood of CBP. Stratified analyses confirmed these associations persisted in individuals with osteoarthritis (OA-CBP LMM: 22.9 cm³/kg/m; InterMAT: 7.53% vs OA-No CBP LMM: 24.3 cm³/kg/m; InterMAT: 6.96% both p < 0.0001) and osteoporosis (OP-CBP LMM: 20.9 cm³/kg/m; InterMAT: 8.43% vs OP-No CBP LMM: 21.3 cm³/kg/m; InterMAT: 7.9% p = 0.16 and p = 0.0019). Higher pain intensity (Pain Intensity Numerical Rating Scale ≥4) correlated with lower LMM (2-unit Z-score deviation = OR, 0.63; 95% CI, 0.57-0.70; p < 0.0001) and higher InterMAT (2-unit Z-score deviation = OR, 1.22; 95% CI, 1.13-1.30; p < 0.0001), independent of physical activity, osteoporosis and osteoarthritis.

Interpretation: This large, population-based study highlights the associations of InterMAT and LMM with CBP. Given the limitations of the cross-sectional design, our findings can be seen as an impetus for further causal investigations within a broader, multidisciplinary framework to guide future research toward improved prevention and treatment.

Funding: The NAKO is funded by the Federal Ministry of Education and Research (BMBF) [project funding reference numbers: 01ER1301A/B/C, 01ER1511D, 01ER1801A/B/C/D and 01ER2301A/B/C], federal states of Germany and the Helmholtz Association, the participating universities and the institutes of the Leibniz Association.

Keywords: Body composition; Chronic back pain; German national cohort; Muscle composition; Physical activity.

Fig. 1

MRI-Based Analysis of Intermuscular Adipose Tissue (InterMAT) and Lean Muscle Mass (LMM) in Chronic Back Pain (CBP) using data from the German National Cohort (NAKO). (a) Representative MRI segmentations of InterMAT (yellow) and LMM (red) in participants with and without CBP, stratified by InterMAT percentiles (25th, 50th, and 75th). For comparability, MR slices corresponding to the L2-3 region were used and participants with similar LMM values were selected. (b) Association of muscle composition with CBP symptom burden. Multivariable logistic regression analysis demonstrated that variations in muscle composition were significantly related to CBP symptom burden, with higher InterMAT linked to increased symptom burden and higher LMM linked to reduced symptom burden, independent of physical activity and musculoskeletal comorbidities. (c) Combinatorial associations of InterMAT and LMM on CBP probability. A heatmap displaying CBP odds ratios for different combinations of InterMAT and LMM Z-scores. Higher InterMAT and lower LMM were associated with greater likelihood of CBP, with a potential synergistic association observed in individuals with both unfavorable parameters. Notably, high InterMAT values were associated with greater likelihood of CBP regardless of LMM levels, while the association between low LMM values and CBP appeared less pronounced when InterMAT values were lower.

Fig. 2

Nomogram-based probability estimation and subgroup analysis of muscle Composition in Chronic Back Pain (CBP). (a) Nomogram for Assessing CBP Probability: This nomogram integrates intermuscular adipose tissue (InterMAT), lean muscle mass (LMM), sex, age, physical activity level, and comorbidities (dyslipidemia, diabetes mellitus, osteoporosis, and osteoarthritis) to estimate an individual's probability of CBP. Each variable contributes a weighted score that is summed to yield an overall estimated probability of CBP. The nomogram is intended to illustrate the associations between these factors and CBP probability without implying direct causation or serving as a clinical decision-making aid. InterMAT and LMM are most strongly associated with CBP, with higher InterMAT and lower LMM being linked to greater probability of CBP. (b) Subgroup Analysis for Osteoarthritis (OA) and Osteoporosis (OP): Boxplots illustrate InterMAT (%) and LMM (cm³/kg/m²) across different subgroups stratified by CBP presence. Osteoarthritis (OA): Individuals with OA exhibit significantly higher InterMAT and lower LMM compared to those without OA. Among OA patients, those with CBP show further increased InterMAT and reduced LMM (∗∗∗∗p < 0.0001). Osteoporosis (OP): OP is similarly associated with increased InterMAT and reduced LMM, with significant differences between CBP and non-CBP groups (∗∗p < 0.0019 to ∗∗∗∗p < 0.0001). However, LMM differences between OP-CBP and OP-no CBP are not significant (ns), suggesting a potentially different pathophysiological relationship.