MRI Muscle Markers and Metabolic Risk

Paraspinal muscle composition on MRI can provide measurable information about cardiometabolic risk in adults who do not report established disease. A deep learning workflow quantified intermuscular adipose tissue, known as IMAT, and lean muscle mass, known as LMM, within the erector spinae and multifidus muscles. In a 2026 Radiology publication using data from the German National Cohort, more than 11,000 participants underwent whole-body MRI, blood testing and clinical examination across German imaging sites. IMAT reflected fat within the muscle compartment, while LMM reflected muscle quantity after adjustment for body size. The risk factors assessed were hypertension, dysglycaemia and atherogenic dyslipidaemia. The findings connect information already visible on MRI with early metabolic warning signs in people without known cardiometabolic disease.

Muscle Quality and Quantity on MRI

The MRI workflow focused on the erector spinae and multifidus muscles on both sides of the spine, covering the thoracic to sacral region. Whole-body MRI used a standardised 3-T protocol, and an automated segmentation tool separated the paraspinal muscles from surrounding structures. A statistical image-processing method then separated fat from non-fat muscle tissue. IMAT represented the share of intermuscular fat in the total muscle volume, while LMM represented fat-free muscle volume adjusted for body mass index. This approach turned routine anatomical information into measurable body-composition markers.

Clinical and lifestyle data came from the baseline German National Cohort examinations. Standardised procedures captured blood pressure, laboratory results, age, sex, self-reported medical history and physical activity. Activity levels came from a questionnaire and were grouped as low, moderate or high. The included participants had complete imaging and laboratory data and no self-reported cardiometabolic or major musculoskeletal conditions. This selection allowed muscle composition to be evaluated as an early marker in adults who appeared free of established disease at enrolment. The automated workflow also addressed a practical barrier, as manual IMAT measurement has limited earlier work to smaller samples. It also used consistent spine coverage across sites to support comparable measurement within the included cohort.

Must Read: AI Maps Abdominal Fat on MRI

Risk Factors Emerge Despite No Known Disease

Although the cohort excluded people with known pre-existing cardiometabolic conditions, clinical examination and laboratory testing revealed previously unrecognised risk factors. Atherogenic dyslipidaemia appeared most often, while hypertension and dysglycaemia also emerged in a smaller but notable share of participants. Men had a higher body mass index and a higher frequency of newly detected hypertension and atherogenic dyslipidaemia than women. Education level and relative income position did not differ meaningfully between groups with and without newly detected risk factors, supporting a low likelihood that these social factors drove the observed group differences. The cohort contained very few participants of non-European ethnicity, so ethnicity-based subgroup assessment did not follow.

Muscle composition followed clear demographic and activity patterns. IMAT increased with age in both sexes, while LMM declined with age. Men generally had lower IMAT and higher LMM than women. In women, LMM remained relatively stable into midlife before declining later. Physical activity also aligned with muscle composition. Low activity corresponded to higher IMAT than moderate activity and lower LMM than both moderate and high activity. Moderate and high activity did not show a clear difference for either muscle measure, suggesting that the strongest contrast occurred between low activity and higher activity levels.

Combined Muscle Profiles Show Higher Risk Burden

Higher IMAT aligned with higher odds of hypertension, dysglycaemia and atherogenic dyslipidaemia in both sexes. This pattern remained after accounting for age, sex, physical activity and imaging site. LMM showed a different pattern. Higher LMM aligned with lower odds of all three risk factors in men, while the corresponding associations in women were weaker and did not reach statistical significance. These sex-specific differences were clearest for hypertension and atherogenic dyslipidaemia. IMAT and LMM did not strongly interact for hypertension or dysglycaemia, indicating that each measure contributed distinct information.

The combined profile sharpened risk stratification. Participants with higher IMAT and lower LMM had the highest prevalence of cardiometabolic risk factors. A broader burden pattern also appeared across age and sex groups: as the number of risk factors increased, IMAT tended to rise and LMM tended to fall. Evaluating both measures together may therefore capture a wider range of metabolic impairment than either marker alone. These associations remain descriptive rather than causal because the design captured a single time point. The population was comparatively young, mostly European and asymptomatic, and physical activity relied on self-reporting. The MRI method also focused on paraspinal muscles and T2-weighted imaging, so other muscle regions and fat compartments may need separate evaluation.

MRI-derived paraspinal muscle composition adds a measurable view of cardiometabolic risk in adults without known pre-existing disease. Higher IMAT consistently aligns with newly identified risk factors, while higher LMM aligns with lower risk in men. The combination of greater fat infiltration and lower lean muscle mass marks the highest risk-factor burden. These results support further work on muscle quality and quantity as complementary markers, while the cross-sectional design, population profile and imaging focus limit immediate clinical translation and require prospective confirmation before clinical use.

Source: Radiology

Image Credit: iStock